Datos de Densidad de Microalga

Intensive Microalgae and Copepod Culture
Workshop 16-18 December 2013

Course Compendium
IMProvement of AQuaculture high quality fish fry production (IMPAQ)
COpepod egg Mass production in Aquaculture (COMA)
Roskilde University (RUC)
Graduate School of Environmental Stress Studies (GESS)

16. December: Intensive Microalgae Production
1. Microalgae as biodiesel & biomass feedstocks: Review & analysis of the biochemistry,
energetic & economics (Williams & Laurens, 2010) (Obligatory reading).
2. Biodiesel from microalgae (Chisti, 2007) (Obligatory reading).
3. Microalgal Reactors: A review of enclosed system designs and performances
(Carvalho et al. 2006) (Recommended reading).
17. December: Intensive Copepod Production

4. The elusive copepods: their production and suitability in marine aquaculture (Støttrup, 2006)
(Obligatory reading).
5. A first step towards improving copepod cultivation using modelling: the effects of density,
crowding, cannibalism, tank design and strain selection on copepod egg production yields
(Drillet & Lombard, 2013) (Obligatory reading).
6. Intensive cultivation of the calanoid copepod Gladioferens imparipes (Payne & Rippingale,
2001) (Obligatory reading).
7. Effects of different monoalgal diets on egg production, hatching success and apoptosis
induction in a Mediterranean population of the calanoid copepod Acartia tonsa (Dana)
(Zhang et al. 2013) (Obligatory reading).

18. December: Cost Initiative

8. COST (obligatory reading).
9. COST SESA guidelines (recommended reading).

Monday 16/12-2013
Intensive Microalgae Production

10.00: We meet at Søminestationen for a cup of coffee/Thee and a piece of bread
10.30: Welcome, Practical info, house rules and workshop programme (PMJ)
11.00: Parameters for large-scale microalgal cultivation (SLN)
12.00: Lunch
13.00: Practical experience from existing intensive algae tube reactor, does and don’ts (BZI)
13.30: Process monitoring and control – for algae systems (PMJ)
14.45: Coffee break
15.15: Group exercises: Which algae strain? Algae system design? (Facilitator PMJ)
17.30: Discussion/presentation about group exercise (Facilitator PMJ)
19.00→: Dinner
Tuesday 17/12-2013
Intensive Copepod Production

9.00: Existing Intensive Calanoid copepod systems (PMJ)
9.30: Calanoid copepods resting eggs – for aquaculture (BWH)
10.00: Automatic copepod estimation (MINH)
10.30: Coffee break
11.00: Densities in copepod cultures, a model approach (GD)
12.00: Lunch
13.00: Intensive RAS farm – with emphasis on live feed production (BH)
14.00: Process monitoring and control – for live feed systems (MBM)
15.00: Coffee break
15.15: Group exercise: Which calanoids strain? Copepod system design? (Facilitator PMJ)
17.30: Discussion/presentation about group exercise (Facilitator PMJ)
19.00→: Gala dinner
Wednesday 18/12-2013
Network, Innovation and concluding remarks

9.00: Presentation of different consortiums, groups, industries etc.
9.30: Best practices for applied research and industrial applications grant founding (BWH)
10.30: Coffee break
11.00: Identification Knowledge exchange – COST networks (Facilitator PMJ)
12.00: Lunch
13.00: Building a broad consortium for a COST network on intensive algae and copepod production
16.00: Closing discussion with coffee (Facilitator PMJ). End of workshop – please clean up and
leave – drive safe
List of participants:
Name – Institution – Nationality ‐ contact info
1. Aliona Novac – University of Copenhagen – Moldova ‐ aliona_novac@yahoo.com
2. Ana Cecília Gomes Silva ‐ Universidade Federal do Rio Grande (FURG) – Brazil ‐
ceciliaoceano@hotmail.com
3. Andreas Thymann – Roskilde University – Denmark ‐ thymann@ruc.dk
4. Benni W. Hansen (BWH) – Roskilde University – Denmark – bhansen@ruc.dk
5. Bent Højgaard (BH) – AkvaGroup – Denmark – bhojgaard@akvagroup.com
6. Bibi Ziersen (BZI) – AgroTech ‐ bzi@agrotech.dk
7. Christina Kjellerup – Roskilde University – Denmark – chkjje@ruc.dk
8. Christina Thoisen – Aarhus university – Denmark ‐ cht@dmu.dk
9. Elenora Bruno – DTU Aqua – Italy ‐ elebr@aqua.dtu.dk
10. Evandro Malanski – DTU Aqua – Brazil ‐ evma@aqua.dtu.dk
11. Guillaume Drillet (GD) – DHI Singapore – France – gdr@dhigroup.com
12. Jakob Linnemann – Aquamind – Denmark – jlr@aquamind.dk
13. Karen Riisgaard – DTU Aqua – Denmark ‐ krii@aqua.dtu.dk
14. Katarína Boriová – Roskilde University – Slovakia ‐ boriova@ruc.dk
15. Mark W. Holm – Roskilde University – Denmark ‐ mwholm@ruc.dk
16. Mie Hylstofte Sichlau – DTU Aqua – Denmark ‐ mhsi@aqua.dtu.dk
17. Minh Vu Thi Thuy (MINH) – Roskilde University – Vietnam – minhvu@ruc.dk
18. Morten Bjørn‐Mortensen (MBM) – AkvaGroup – Denmark – mbjorn@akvagroup.com
19. Niels O. G. Jørgensen ‐ University of Copenhagen – Denmark ‐ nogj@plen.ku.dk
20. Per M. Jepsen (PMJ) – Roskilde University – Denmark – pmjepsen@ruc.dk
21. Sara Castanho ‐ Portuguese Institute for the Ocean and Atmosphere/Aquaculture Research Station –
Portugal ‐ scastanho@ipma.pt
22. Rasmus W. Sørensen – AkvaGroup – Denmark ‐ rwsorensen@akvagroup.com
23. Søren L. Nielsen (SLN) – Roskilde University – Denmark – Nielsen@ruc.dk
24. Thomas A. Rayner – Roskilde University – Denmark – tar@ruc.dk

16. December Intensive

Microalgae production
2013
GESS WORKSHOP AND PHD COURSE
View Online
REVIEW www.rsc.org/ees | Energy & Environmental Science
Microalgae as biodiesel & biomass feedstocks: Review & analysis of the

biochemistry, energetics & economics†
Peter J. le B. Williams* and Lieve M. L. Laurens‡
Received 26th November 2009, Accepted 18th February 2010
First published as an Advance Article on the web 22nd March 2010
DOI: 10.1039/b924978h
Following scrutiny of present biofuels, algae are seriously considered as feedstocks for next-generation
biofuels production. Their high productivity and the associated high lipid yields make them attractive
options. In this review, we analyse a number aspects of large-scale lipid and overall algal biomass
Published on 22 March 2010 on http://pubs.rsc.org | doi:10.1039/B924978H
production from a biochemical and energetic standpoint. We illustrate that the maximum conversion
efficiency of total solar energy into primary photosynthetic organic products falls in the region of 10%.
Biomass biochemical composition further conditions this yield: 30 and 50% of the primary product
mass is lost on producing cell protein and lipid. Obtained yields are one third to one tenth of the
Downloaded by Roskilde University on 23 September 2011
theoretical ones. Wasted energy from captured photons is a major loss term and a major challenge in
maximising mass algal production. Using irradiance data and kinetic parameters derived from reported
field studies, we produce a simple model of algal biomass production and its variation with latitude and
lipid content. An economic analysis of algal biomass production considers a number of scenarios and
the effect of changing individual parameters. Our main conclusions are that: (i) the biochemical
composition of the biomass influences the economics, in particular, increased lipid content reduces
other valuable compounds in the biomass; (ii) the ‘‘biofuel only’’ option is unlikely to be economically
viable; and (iii) among the hardest problems in assessing the economics are the cost of the CO2 supply
and uncertain nature of downstream processing. We conclude by considering the pressing research and
development needs.
1. Introduction and historical background reduce the emissions of greenhouse gases. Although alternatives
exist for land-based transport – notably electrical power – there
1.1 History of biofuels development is an understandable reluctance to abandon liquid hydrocar-
Without question our society will need to move away from its bons and internal combustion engines for many transport
strong dependence upon fossil fuels as sources of energy and to purposes. Further, in the case of aviation transport and ship-
ping, there is no practical alternative in the foreseeable future.
These are amongst a number of reasons for the drive for the
School of Ocean Sciences, University of Bangor, Menai Bridge, Anglesey, industrial development of liquid biofuels. Systems to produce
UK LL595PP biodiesel and bioethanol from crop plants (so-called first-
† Electronic supplementary information (ESI) available: Appendices
I–IV. See DOI: 10.1039/b924978h generation biofuels) have been developed and optimised over
‡ Current address: National Bioenergy Center, National Renewable the past several decades and are currently run as profitable
Energy Laboratory, 1617 Cole Blvd, Golden, CO 80401, USA businesses.
Broader context
There is a rapidly growing interest in the potential of microalgae as feedstocks for the next generation of biofuels. Working from
fundamental biochemical principles, we consider the potential yields of organic production by photosynthesis. The maximum
theoretical energy conversion of full spectrum sunlight to organic material lies in the region of 10%. The yields are constrained by
thermodynamics and stoichiometry. The yields obtained with outdoor cultures are characteristically one third to one tenth of this
theoretical yield; the losses are primarily due to the inability of the photosynthetic system to process the captured photons at the rate
they are absorbed and so energy is lost. Overcoming this problem is a major challenge. The mass and, to a lesser extent, the energy
yields are further reduced following the conversion of the primary photosynthetic products to the spectrum of biochemicals required
by the cell. With a simple light-driven model, we explore the potential economics of a number of production scenarios of algae of
varying lipid content (15–50% lipid of the cell’s dry weight) at low (0–30 ), intermediate (35–45 ) and high (45–55 ) latitudes. The
main conclusions were that (i) the ‘‘fuel only’’ option is not viable, markets need to be found for the other major and minor
components of the cell; (ii) high lipid containing algae may not necessarily be the most favourable candidate organisms. We conclude
that although the potential does appear to exist for economic production of algal biofuels, a major R & D programme would be
called for to convert the concept to a reality.
554 | Energy Environ. Sci., 2010, 3, 554–590 This journal is ª The Royal Society of Chemistry 2010
View Online
Although the production and development of biodiesel and freshwater. Thus, their negative ecological impact is much
bioethanol has increased rapidly since the start of biofuels reduced as compared with higher plants.1 The potential to
development and production in the 1970s, at 30–40% per annum, produce the feedstock on waste or desert land and the reduced
the total energy content in both biofuels is still less than 1% of the freshwater requirements addresses some of the concerns in the
world’s energy production. The 2008 production figure for bio- Gallagher report.
ethanol was 2.5 1018 joules and that for biodiesel an order The mass production of microalgae for lipid production has
of magnitude less by comparison to a total global energy use of a long history. Prior to the present interest in producing biofuels
5 1020 joules. to mitigate CO2 release, there had been an extended period of
Despite their relatively small contribution (1%) to the overall research interest motivated by security of oil supply. This was
production of liquid fuels, these first-generation biofuels have initially prompted by the 1973 oil crisis. In response to the
come under considerable international scrutiny and criticism. recognition of the potential vulnerability of oil supplies, Exxon
The main criticisms are the encroachment of the biofuels feed- Research and Engineering Company, for example, funded work
stock production on valuable crop and virgin land, and the effect of lipid production by algae.2 At about the same time (1978 to be
biofuels have on food commodity prices. These issues have raised precise), the U. S. Department of Energy set up the Aquatic
major question marks surrounding their social benefits, as Species Program (ASP). This program reached peak funding in
summed up in the 2008 Gallagher report to UK House of the mid-1980s; after which the funding dwindled, until the
Commons.129 The executive summary of that report noted that program was eventually closed down in 1996. All told, just over
‘‘feedstock production must avoid agricultural land that would $25 million was invested in the program. The findings, which
otherwise be used for food production’’ and ‘‘The introduction of were extensive and valuable, are collected in a major report.3
biofuels should be significantly slowed until adequate controls to There was a concurrent Japanese programme, which purportedly
address displacement effects are implemented and are demon- cost in excess of $100 million, but which produced very little
strated to be effective’’ (the embolding is ours). This mood, which accessible science. The main conclusion of the ASP report was
although not universal, is widespread and has given rise to that algal biofuels are a potential valuable alternative to ‘tradi-
searches for alternative, so-called ‘‘second-generation’’ biofuels tional’ biofuels, however, considering the oil prices of the early
and has re-stimulated interest in mass algal biomass production. 1990s, the calculated economics turned out to not be profitable.
Nevertheless, considerable headway was made with regards to
1.2. Algal biofuels potential basic biology, strain selection, metabolic engineering and large-
scale growth cultures of selected strains.
We have restricted the review to microalgae. Macroalgae Algae have long been known to produce a great variety of
(seaweeds) have long been grown commercially and may have lipids, hydrocarbons and other complex oils (reviewed recently in
a role as biofuels and although a number of the aspects of their ref. 4,5). Cultured algae have been used as feeds for aquaculture
physiology are shared with the microalgae, their mass production applications because of their high content of nutritionally
involves a substantially different set of considerations. essential polyunsaturated fatty acids. A further motivation for
The advantages of microalgae as a feedstock for biodiesel algal culture has been the production of high value by-products
production, over terrestrial plants, are that there is no require- such as the pigments astaxanthin (a food colorant and antioxi-
ment for soil fertility and, if marine algae are used, there is no dant from Haematococcus pluvialis) and b-carotene (a food
need to draw upon valuable and often scare supplies of additive produced from Dunaliella species). These sorts of
Peter Williams is professor Dr Lieve Laurens is a researcher

emeritus at University of Ban- at the National Renewable
gor, UK. Originally trained as Energy Laboratory (NREL) in
an industrial biochemist, he Golden, Colorado. Prior to
came to the view that oceanog- joining NREL, she was
raphy offered more interesting a Research Associate at the
challenges. He has held posts at Centre for Applied Marine
the Woods Hole Oceanographic Science in the School of ocean
Institute, Southampton Univer- Sciences, at the University of
sity, Bigelow Lab in Maine and Bangor, UK, where she started
the University of Gothenburg, research on biodiesel from
returning to the UK to become microalgae and the work leading
Professor of Marine Biogeo- to this review was initiated. Her
Peter J: le B: Williams chemistry at Bangor Uni- Lieve M: L: Laurens recent work focuses on the
versity’s School of Ocean quantification and analysis of
Sciences. His long term research interest has been the dynamics of lipid yield and composition in microalgae using high-throughput
organic material in the oceans, but an opportunity provided by spectroscopic methods. She holds a PhD from the Department of
Shell, opened up a new interest in the science behind algal biomass Metabolic Biology at the John Innes Centre in Norwich, UK.
and biofuel production.
This journal is ª The Royal Society of Chemistry 2010 Energy Environ. Sci., 2010, 3, 554–590 | 555
View Online
production processes have been in place since the early 1950s, so the eukaryotes and bacteria. They are able to photosynthesise,
many aspects of the technology of mass algal growth may be and are found in a range of different habitats, from fresh to
considered mature.6 Much historical research has focused on the marine and hyper-saline environments.11 The large number of
lipid composition from either a taxonomic or nutritional stand- species are generally subdivided into 10 taxonomic groups which
point. There is, however, a sharp difference between the growth include the green algae (Chlorophyceae), diatoms (Bacillar-
of algae for nutritional and for fuel use. These nutritional iophyceae), yellow-green (Xanthophyceae), golden algae
products are high value: astaxanthin for example commands (Chrysophyceae), red algae (Rhodophyceae), brown algae
a price in the region of $3 million tonne1,7 compared with less (Phaeophyceae), dinoflagellates (Dinophyceae), Prasinophyceae
than a $1000 tonne1 for crude oil. Thus the economics are and Eustigmatophyceae.12 The blue-green algae (Cyanophyceae)
profoundly different. As a result, the production of biofuels will were originally grouped with the eukaryotic algae; however it
require a fundamental change in the approaches to production was subsequently realised that they belong to the bacterial
compared with nutraceutical and aquaculture feed-grade prod- domain, hence their present common name – cyanobacteriax.
ucts and without question fresh, major challenges. A major and significant difference between the bacteria and
Algal biomass can serve as a feedstock for the production of the eukaryotes is that the former lack discrete internal, subcel-
a variety of different biofuels, e.g. biodiesel, hydrogen, methane lular structures, organelles (chloroplasts, mitochondria, nuclei).
and bioethanol. Furthermore, its production is also being seri- Organelles are surrounded by lipid membranes whose bilayer
ously considered for the removal of carbon dioxide from the structure requires strongly polar molecules. The triglycerides, the
flue gases of fossil fuel power stations. In order to maintain predominant lipid class in higher plant oils, lack the required
focus, we limit our discussion to the use of algae as feedstock polarity to form a sufficiently stable bilayer. The more polar
for biodiesel and biomass production. There have been claims in phospholipids and the glycolipids are the major components of
the literature over recent years hailing algae as the solution to cell membranes and as a result, in actively growing cells, these
the global energy crisis (e.g. ref. 8). In this review we aim to two groups are major components of algal lipids, with important
present the principles behind large-scale production of algal consequences.
biomass and biodiesel production along with a critical and The metabolism of organisms, particularly microorganisms, is
objective review of the literature to date. As others (e.g. ref. 9), strongly influenced by their surface-to-volume ratio. Simple,
we build our review on the biochemical fundamentals of single celled organisms can be approximated to spheres, thus
photosynthesis and biomass production, and accordingly begin their metabolism, and therefore growth rate, is inversely
(Section 2) with a brief discussion of the biology and proportional to their cell diameters (see insert in Fig. 1). The
biochemical composition of microalgae, with particular regard growth rates of microorganisms can be very high; whilst some
to the major biochemical categories: proteins, carbohydrates algae are able to divide once every 3–4 h, most divide every 1–2
and lipids. In Section 3, starting from first principles, we days under favourable conditions (see Fig. 1). Accordingly their
establish the overall potential efficiency of photosynthetic scope for growth is colossal and this, in major part, is the basis
production of algae. A major challenge in algal biofuels for the interest in their potential as biomass producers. In prin-
production is maximising this stage, as economic models for ciple algal biomass crops may be harvested either daily or every
biodiesel production from seed oils have acknowledged that the few days (e.g. ref. 15, 16).
feedstock cost comprises a substantial portion of the overall
biodiesel cost.10 Accordingly, in Section 4 we address the growth
of algae, the controls on growth and the reported yields. These 2.1. Major biochemical groups, their presence and function
are then combined with the theoretical photosynthetic yields to
It is conventional to consider four principal biochemical classes
give predictions of biomass yields for season and latitude.
of molecules: carbohydrates, proteins, nucleic acids and lipids.
Section 5 contains a review of the harvest and processing of the
Table 1 gives the broad cell content of these major fractions, their
biomass relevant to published processes. In Section 6, we bring
elemental composition and energetic properties.
together the findings of the two preceding sections in the form
of an economic analysis of various scenarios for low, mid and
(i) Carbohydrates. Microorganisms contain a wide variety of
high latitudes. Finally, in Section 7, we explore the answers to
carbohydrates, both monomers and polymers. Carbohydrates
two questions: (i) can the production of biofuels from algae be
serve both structural and metabolic functions and, as the early
economically viable; and (ii) what R & D is needed to achieve
products of photosynthesis, they serve as the starting point for the
a profitable outcome.
synthesis of the other biochemicals. Different classes of algae
produce specific types of polysaccharides. For example, green
2. Biology and biochemical composition of algae produce starch as an energy store, consisting of both
microalgae amylose and amylopectin, similar to higher plants. The green alga
In this section we introduce the aspects of algal biology and

x Strictly the term ‘‘alga’’ should be limited to the eukaryotic
biochemistry that are relevant in the economical considerations phototrophs, since cyanobacteria (formerly known as blue-green algae)
(Section 6). The composition of the algal biomass with regards to are a form of bacteria, with very different genetics and evolutionary
lipids, carbohydrates and proteins will greatly determine its history. It is however very cumbersome to qualify every reference to
these two groups of single-celled phototrophs as ‘‘algae and
overall value.
cyanobacteria’’, or as ‘‘micro-photoautotrophs’’ so, unless there is need
Microalgae are single cell organisms, found in either colonies to be specific, for convenience the two are referred in the text simply as
or individual cells and are comprised of representatives of both ‘‘algae’’.
View Online
Proteins have important commodity value as animal feed.

Critical to this is their amino acid composition, as a number of
amino acids are dietary essentials for mammals yet they are unable
to synthesise them, for example. Fig. 2 shows the mean amino acid
spectrum for a number of algae. It can be seen that the spectrum
compares favourably with proteins of high nutritional quality.
(iii) Nucleic acids. Nucleic acids (RNA and DNA), in

conjunction with proteins and their monomers, provide the basis
for algal division and growth. The nucleic acids comprise a small
fraction of cellular biomass but the major part of the cell’s phos-
phate and the second most important site of nitrogen (see Table 2).
(iv) Lipids. As with carbohydrates, lipids serve both as

energy reserves and structural components (membranes) of the

cell. The simple fatty acid triglycerides are important energy
Fig. 1 Frequency analysis of algal growth rates (main pane), derived reserves. Membranes are mainly constructed from phospholipids
from data in ref. 13 and algal growth-size relationships from ref. 13, 14. and glycolipids, where the hydrophilic polar phosphate or sugar
moieties and the level of saturation of the fatty acyl chains
Tetraselmis suecica, accumulates 11 and 47% of its dry weight as determine the fluidity of the membranes. The microalgae have
starch in nutrient replete and deplete conditions respectively.19 the facility for rapid adaptation to new environments
Red algae synthesise a carbohydrate polymer known as floridean (e.g. changes in temperature), through the de novo synthesis and
starch, consisting mostly of amylose.20 A commonly found poly- recycling of fatty acids to maintain the membrane characteristics.
saccharide in a large number of algal species is chrysolaminarin, A large fraction of the cell’s phosphate may be present in
a linear polymer of b(1 / 3) and b(1 / 6) linked glucose units.21 phospholipids. The membrane lipids associated with the thyla-
Chrysolaminarin often accumulates in pyrenoids, which are koids (the internal chloroplast membranes, also the sites for
centres of high activity of carbon assimilation in the chloroplast. photosynthetic activity in eukaryotes, see Section 3.1(i)) contain
a sulfur lipid (sulfoquinovosyldiacyl glycerol), which is the major
(ii) Proteins. As do carbohydrates, proteins also have both site of the sulfur in algal lipids.11 Table 3 gives a summary of lipid
structural and metabolic functions. As enzymes, they are the class distribution derived from published analyses. A number of
prime catalysts for cell metabolism and so facilitate growth. factors give rise to the variations seen in the table. Storage lipids,
Second, they serve a structural role. For example, they provide which are predominantly triglycerides, may gain a greater
the scaffold upon which the chlorophyll molecules are assembled proportion of the overall lipid fraction as the metabolic rate
in the light harvesting complexes of the chloroplast (see Section slows down. As a consequence, shifts in lipid composition occur
3.1(i)) and also are embedded in the lipid membranes, where they through the various phases of growth.
serve a similar structural role, as well as a metabolic role. Allowing for the scatter due to differences in analytical tech-
Furthermore, it is known that the structural cell wall of Chla- niques, the frequency distribution suggests a minimum cell lipid
mydomonas reinhardtii consists primarily of cross-linked content in the region of 15% (see Fig. 3). There appears to be
hydroxy-proline-rich glycoproteins.22 a marked difference between the total lipid content of the
Table 1 Elemental composition of algal biochemical components. Data for the proteins and fatty acids were derived from means from the analyses
shown in Fig. 3 and 5. The values for the lipid classes are derived from the mean fatty acid elemental composition and the composition of the other
molecular structures in the overall molecule. That for the overall algal lipids is derived from the values for the individual lipid classes and the mean lipid
class composition given in Table 3. The values for the nucleic acids and carbohydrates are taken from ref. 17. The calorific values (as kJ g1) of lipids and
their derivatives have been calculated according to an equation for higher heating values (equivalent to the bomb calorific value) ¼ 35.17C + 116.25H
11.1O + 6.28N + 10.47S, where C, H, O, N, P and S are the mass fractions of the constituent elements.18 This equation gives a good approximation for
these compounds. The hydrogen and oxygen atoms associated with phosphate and sulfate were not included in the calculation. The calorific values for
proteins and nucleic acids are calculated from the calorific values of their component molecules, using 18 kJ per mol of water as the energy lost on
polymerisation. The range given in the right hand column derives from the data compilations used to produce Fig. 5 (only the middle 90% of the data is
used)
Characteristic elemental
Biochemical component composition Calculated calorific value/kJ g1 Range of typical cell content (%)
Algal lipids C1H1.83O0.17N0.0031P0.006S0.0014 36.3 15–60

Acylglycerides C1H1.83O0.096 40.2 —
Glycolipids C1H1.79O0.24S0.0035 33.4 —
Phospholipids C1H1.88O0.173N0.012P0.024 35.3 —
Algal fatty acid C1H1.91O0.12 39.6 —
Methyl esters C1H1.92O0.05 43.0 —
Protein C1H1.56O0.3N0.26S0.006 23.9 20–60
Nucleic acid C1H1.23O0.74N0.40P0.11 14.8 3–5
Polysaccharide C1H1.67O0.83 17.3 10–50
View Online
Fig. 2 Mean amino acid composition of algal protein derived from 28 species (9 classes) of eukaryotic algae.23 The upper bars are the standard
deviations, the lower the standard errors. The low value for tryptophan will derive in part from loss of the molecule during acid hydrolysis of the protein.
Characteristic values for high quality proteins are shown for comparison purposes.
eukaryotic and the prokaryotic algae (the cyanobacteria) – the proteins preferentially run down or do the carbohydrates and
latter containing less than the former (see Fig. 3), likely to be due to proteins decrease in relative proportions. These three trajectories
the absence of internal membranes in the prokaryotes (see Section are shown in the triangular plot in Fig. 5. In the case of first of the
2.1 (i)). So, although the cyanobacteria are easier organisms to three scenarios listed above, we might expect maintenance of
manipulate genetically (as a result of their simpler DNA), their growth rate with increases in lipid content up to 40–50%, whereas
potential as lipid producers would not appear to be promising. in the others we would anticipate a reduction in growth rate as
The fatty acid composition of algal lipids (Fig. 4) is well docu- the lipid content increased (see discussion below in Section 2.3).
mented, with a high occurrence of unsaturated (and poly- The data we have been able to assemble in Fig. 5 does not give
unsaturated) fatty acids, with half of the fatty acids having a clear sign of any systematic compositional trend and probably
a carbon number less than C18. The majority of the unsaturated a dedicated study would be needed to elicit one, if it indeed exists.
fatty acids occur in the membrane lipids, where their main func-
tion is to maintain membrane fluidity under different conditions. 2.3. Relationship between lipid content and growth and nutrient
The high concentrations of unsaturated fatty acids in the extracted status
lipids, and ultimately in the resulting fuel, will be an important fuel
quality determinant. The preponderance of the shorter chain fatty Fig. 6 contains data from two papers2,42 and an analysis of the
acids has significance for their potential as diesel fuels (defined as relationship between lipid content and growth rate. There is
alkyl chains of between 12 and 18 carbon atoms long). The level of scatter in the data and the absolute rates differ, however there is
unsaturation affects biodiesel properties.36 For example, fuels a significant inverse relationship between growth rate and lipid
with a higher level of unsaturation of the acyl chains have a higher content (p value for ref. 2 data set is 0.0006 and that for ref. 42
cloud point, which is desirable, but are also much more susceptible data set is 0.0014). A similar inverse relationship is also reported
to oxidation. The highly unsaturated fatty acids found in algae in ref. 131. The product of the lipid content and growth rate will
may need to be hydrogenated to improve their potential fuel give a lipid-normalised production rate (in units of days1) and
properties. Furthermore, a high level of unsaturated fatty acids in the relationship between this latter property and lipid content will
a fuel increases the danger of polymerisation in the engine oil and take on a quadratic form, with an intermediate maximum. The
can cause problems with oxidative stability of the fuel. first derivative of the quadratic equation will give the lipid content
for the maximum lipid production rate (mass of lipid produced
per mass of lipid present in the culture per day). To demonstrate
2.2. Shifts in the biochemical composition with increasing lipid
the outcome, and potential, of this procedure, we have processed
As the lipid content increases, the percentage of the sum of the the parameters from the two equations given in Fig. 6. The
other components must go down. Fig. 5 gives a summary of the analysis gives a maximum lipid production rate at a lipid content
reported relative proportions of the three major biochemicals – of ca. 15% (from the analysis in ref. 42’s data set) and ca. 30% (for
protein, lipid and carbohydrate. For two reasons the shift in ref. 2). From these values, an optimum growth rate for lipid
protein content is critical. First these molecules set the level of the metabolism may be calculated from the initial equation. Esti-
cell’s metabolism and in conjunction with the nucleic acids mated maximum lipid production and growth rates are valuable
determine the growth rate potential. Second, from an economic for process design and management, e.g. to allow culture dilution
point of view proteins are valuable bulk components of the cell rates to be optimised without deleterious effect on overall lipid
(see Section 6). Thus, in designing a strategy to search and select yield. However, bespoke data sets are needed to establish to what
for the economically most suitable algae, it is of prime impor- extent the relationship seen in Fig. 6 is general. Overall produc-
tance to establish whether or not we can generalise over the tion of lipid will be more complex than this simple analysis, as
change in protein content associated with changes in total lipid there are many other considerations, but the calculation brings
content. As the lipid content increases, the question arises, are home that high lipid content alone cannot regarded as the sole,
the carbohydrates preferentially run down and the protein perhaps not even the major, consideration when searching for
content held constant as long as possible. Alternatively, are the suitable strains or optimum growth conditions.
View Online
Table 3 Mean ( standard error) lipid class content as a percent of total
Table 2 Calculated distribution of major elements in the four principal biochemical fractions for model low (15%), medium (25%) and high (50%) lipid-containing algae. The calculation is made using
High
the elemental compositions and calorific values given in Table 1 and assumes that, as the lipid content varies, the nucleic acid content remains constant at 5% and protein : carbohydrate ratio remains
0.4
lipid of individual algal species, plus a composite from all reports for
36
64
0
0
microalgae.24–35
Medium
Simple Glyco- Phospho-
lipids lipids lipids
15
85
0
0
0
Low
Diatoms
0.5
91
37 16 36 8 25 8
S
Chaetoceros species
0
0
Phaeodactylum tricornutum 54 6 34 5 11 1
High
Range from Borowitzka (1988)130 14–60 13–44 10–47
1.1
52
48
0
0
Green algae
Chlamydomonas species 48 10 44 13 63
Medium
Dunaliella tertiolecta 71 67 1 25 0
0.81
Dunalliella viridis 13 1 44 3 42 2
35
65
0
0
Low
0.7
Blue green algae
25
75
P
0
0
High
5.4
82
15
Others
2
Nannochloropsis oculata 22 1 39 0 38 1
Medium
Isochrysis species 36 3 35 1 27 3
7.8
Composite from all reports (n ¼ 46) 35 ± 3 40 ± 2 25 ± 2

88
11
1
0
Low
8.7
90
10
N
It has been known since the 1940s that cell lipid content
High
increases during nutrient limitation (see Fig. 7). As nutrient

33
24
36
24
limitation also affects growth rate, this provides an explanation

7
for the apparent inverse relationship between growth rate and

Medium
lipid content (see Fig. 6). Ref. 43 drew the broad conclusion that
14
32
29
48
when the nitrogen ran out, the organisms were forced to termi-
Percentage major element in various biochemical fractions
nate the production of nitrogen containing material (proteins

Low
constant at 3 : 2 (see legend to Fig. 5). These simplifications mean that the data are mainly illustrative
and nucleic acids) but continued to synthesise lipids and carbo-

30
34
52
O
hydrates. Whereas it may explain the lipid–nutrient relationship,

High
it cannot explain observations where growth was controlled by

8.6
63
22
13
2
light rather than inorganic nitrogen and where again a negative

Medium
relationship was seen.42 This, and other work, suggests that the
matter may be more complex than ref. 43 suggested. Regardless
7.6
36
39
23
2
of the biochemical mechanism, the evidence for of a negative

Low
relationship between growth rate and lipid content (see Fig. 6)

7.2
23
47
28
H
has an important bearing on the strategy adopted for biomass

High
production, especially when products other than the lipid

60
59
24
13
component are valuable (see e.g. Fig. 24).

3
Medium
Lipid level
55
33
41
23
2.4. Algal lipids as fuels

3
Low
From their molecular composition, we can anticipate a number

53
20
49
27
C
of properties relevant to the potential of algal lipids to serve as

High
fuels. A summary of the fuel properties of biodiesel from

28.5
50
27
18
5
a microalga (heterotrophically grown Chlorella protothecoides)44

composition (as %
Medium
is given in Table 4. Algal oils differ from higher plant oils in their
Biochemical
high phospholipid and glycolipid concentrations. These lipid

Calorific value/kJ g1 23.2 24.7
total mass)
Lipid level
25
42
28
classes contain nitrogen, phosphorous and sulfur that may be

Element total (as % dry weight)
5
problematic with regards to engine performance, if present in

Low
15
48
32
fuels. However, it is likely that the sulfur, phosphorus and

5
nitrogen-containing compounds would end up in the water-

soluble fraction following transesterification, so one would
Polysaccharides
expect these elements to be low to non-existent in algal biodiesel.

Nucleic acids
Since about 30% of the original lipid mass can be lost to the polar
phase during esterification, the lipid class composition will also
Proteins
Lipids
greatly affect the potential fuel yield by transesterification. Thus

triglycerides have a > 99% biodiesel yield compared to a < 70%
View Online
Fig. 3 Frequency distribution of the lipid content (as % dry weight) of eukaryotic algae and Cyanobacteria. The values in parentheses are the number of
data sets used for the analysis. Primary data and references in Appendix I (provided in the ESI†).
Fig. 4 Upper histogram: mean fatty acid composition of various eukaryotic algal groups and Cyanobacteria. The figures in parenthesis are the numbers
of analyses from which the means are derived. The upper error bars are standard deviations, the lower standard errors. Full data set and references in
Appendix III (provided in the ESI†). Lower histogram: data for comparison purposes for higher plant fatty acids (C. Price, Shell Global Solutions,
personal communication).
yield for phospholipids.45 If the relative distribution of the lipid the figure of 41 kJ g1 found in ref. 44. This derives from a slightly
classes given in Table 3 is used, then the overall mass yield on higher value in our case for the C : H ratio (see Table 4). The
producing biodiesel is calculated to be about 80%. This figure difference is small and, considering the two approaches used are
does not take into consideration the consumption of methanol so different, we do not place any great significance on the vari-
during esterification; if the methanol used is considered as part of ance between these two sets of values.
the original organic supply, as it could well be, then the overall The fuel characteristics will be greatly determined by the fatty
organic yield is reduced to ca. 72%. acid composition and will thus vary with species and growth and
The overall heating value of crude algal lipid is somewhat nutrient conditions of the algal culture. As shown in Fig. 4, the
depressed to a value in the region of 36 kJ g1 (Table 1), due to chain length distribution of fatty acids occurring in microalgal
the lower calorific value of the glyco- and phospholipids. species is more diverse compared with that of higher plants,
However, as these fractions would be almost certainly separated making it possible that certain species will be cultured for
off in the transesterification process, their low calorific value selected fuel properties. For example, strains that primarily
would have no effect on the properties of the final biodiesel. accumulate shorter (<C16) chain fatty acids may be more
From their calculated elemental composition (based on their amenable for use in the production of jet fuels, compared to the
reported fatty acid composition), we determine a heating value application of very long chain fatty acids (>C20) in the lubricant
for the fatty acid methyl esters of 43 kJ g1, slightly higher than market.
View Online
Fig. 5 A: triangular plot of the proportions of lipid, protein and

carbohydrate content of algae, data taken from ref. 2, 27 and 37–42. The Fig. 7 The effect of nitrogen limitation on the lipid content of eukary-
red circle is the mean of the whole dataset for active growth (lipid 24.2%, otic algae. The values in parenthesis are number of observations from
protein 48.3%, carbohydrate 27.5%). The solid circles show the shift in which the mean was derived. The upper error bar is the standard devi-
composition from unlimited logarithmic growth (red) to N-limited ation, the lower the standard error. Full data set and references in
growth (brown). The green line shows the trajectory on increasing lipid Appendix II (provided in the ESI†).
composition assuming the protein : carbohydrate ratio remains constant,
the red line that if the carbohydrate content diminishes as the lipid
content increases and the magenta line that if the protein content electrons. These electrons are driving the dissociation of water,
diminishes. Throughout this review we have used a rounded off value of generating protons and oxygen. The protons and the associated
3 : 2 for the protein carbohydrate ratio.
electrons enable the reduction of carbon dioxide to organic
material, whereby oxygen is essentially a waste product.
3. Principles and efficiency of photosynthesis
Since the basis of algal biomass production is directly propor-
tional to the efficiency with which the algal cells assimilate carbon 3.1. Mechanism of photosynthesis
from the atmosphere through photosynthesis, we discuss the basic The simple stoichiometry of photosynthesis may be written as:
principles and efficiency of photosynthesis in this section. The aim
of this section is to provide a background to the theoretical and H2O + CO2 / [CH2O]{ + O2
practical yield calculations in the subsequent sections so that the
limitations to mass algal production can be made clear. However, the above simple equation implies that at least 50%
Photosynthesis is the predominant process maintaining of the atoms of the oxygen produced must come from the carbon
a whole host of elements (notably carbon, nitrogen and sulfur) dioxide, whereas both derive from water, thus the more appro-
out of thermodynamic equilibrium, and thereby driving their priate equation is:
global geochemical cycles. It is the basis of the food supply for
most life on Earth, maintaining the biosphere; it is also the 2H2O + CO2 / [CH2O] + O2 + H2O
ultimate source of all fossil fuels. Photosynthesis, in turn, is
driven by photons, which, when absorbed by chlorophyll mole-
This overall reaction can be separated into two phases: (i) a set
cules, give rise to a charge separation and the ejection of
of photochemical and redox reactions (conventionally called the
‘‘light reaction’’) and ii) a sequence of enzymatic reactions, often
referred to as ‘‘dark reactions’’ but better as ‘‘light-independent
reactions’’, as they occur both in the light and the dark.
Light reaction: 2H2O / 4[H] + O2 + energy
Light-independent reactions: 4[H] + CO2 / [CH2O] + H2O
(The notation [H] refers to the combination of the reduced

coenzyme nicotinamide adenine dinucleotide phosphate
(NADPH) and an electron.)
These reactions + O2 + energy are intimately connected within
Fig. 6 The variation of growth rate with lipid content. The data set from the cell. The light reaction operates on very short time
ref. 2 derives from 15 strains of freshwater and 11 strains of marine scales (from femtoseconds to milliseconds) whereas the
eukaryotic algae; the data used were taken during active logarithmic
growth. The data from ref. 42 comes from 8 eukaryotic marine algae { The notation [CH2O] is commonly used as shorthand for organic
commonly used in aquaculture. In this case growth rate was controlled by material in biology – in the present context its use is restricted to
varying the irradiance and the cultures were sampled during active growth. organic material with the same elemental ratio as monosaccharides.
View Online
Table 4 Comparison of properties of biodiesel from microalgae and diesel fuel and the ASTM biodiesel standard. The data from ref. 44 comes from
a heterotrophically grown Chlorella protothecoides culture. There are small differences between their observations and the values we calculate (given in
square parentheses) from elemental composition and calorific values given in Table 1
Properties Micro-algal Higher plant Diesel fuel ASTM biodiesel standard
Density/kg dm3 0.864 0.877–0.887 0.838 0.86–0.90

Viscosity/mm2 s1@ 40 C 5.2 3.3–5.2 1.9–4.1 3.5–5.0
Flash point/ C 115 — 75 Min. 100
Solidifying point/ C 12 — –50–10 —
Filter plugging point/ C 11 — 3.0 (max. 6.7) Summer max. 0
Winter max. < 15
Acid value/mg KOH g1 0.374 0.16–0.43 Max. 0.5 Max. 0.5
Heating value/kJ g1 41 [43] 39.5–40.3 40–45 —
H/C ratio 1.8 [1.9] 1.8 —
light-independent reaction operates over timespans of seconds to

hours. This profound mismatch of timescales gives rise to inef-
ficiencies in fluctuating environmental circumstances, notably
irradiance and temperature variations and is a major problem in
maximising yields of mass algal culture.
(i) Light, its absorption and the formation of ‘‘reducing

capacity’’. Not all incoming radiation is available for photosyn-
thesis. The solar spectrum is described by Planck’s radiation
distribution equation; the spectrum of light arriving at the
surface of the planet has been attenuated by some 30% by the
gases in the atmosphere and losses due to light scattering and
absorption by clouds. The irradiance spectrum is shown in Fig. 8.
The primary pigment involved in photosynthesis is chlorophyll a,
which has strong absorption bands in the regions 400–450 and
650–700 nm (see Fig. 9). This delimits the useful range of
incoming radiation to 400–700 nm – so called photosynthetically
active radiation (PAR). PAR amounts to 45–50% of the total
incoming radiation, the exact value mainly being determined by
the moisture content of the atmosphere attenuating the infrared
part of the spectrum.47 Since we use clear sky radiation as the
Fig. 9 The spectrum of incoming radiation (black line) and the
absorption spectrum of chlorophyll and accessory photosynthetic
pigments. Redrawn with permission from ref. 46.
basis for further discussion here, the 45% end of the spread is the
more appropriate value to adopt.
Because of its low absorption in the range 450–650 nm, chlo-
rophyll a itself only is able to capture some 30–40% of PAR.
Plants have overcome this by introducing additional light
capturing pigments (e.g. b-carotene) that fill in much of the
chlorophyll a window of the spectrum (see Fig. 9) so increasing
the portion of the spectrum that can be used for photosynthesis.
Algae can change the quantity of these accessory pigments
(subject to phylogenetic constraints) to optimise light capture,
but the adaptation process is slow compared with the time scales
of the photochemical reactions. The photosynthetically active
parts of the chlorophyll spectrum lie at 680 and 700 nm (see
Fig. 9). The energy of photons, captured at shorter wavelengths,
can be transferred to the 680–700 nm region very efficiently on
a quantum basis. However, there is an inevitable loss of energy
resulting from the transfer from high-energy, shorter wavelength
Fig. 8 The distribution of energy (blue line) and photons flux (red line) to lower energy, longer wavelength, photons. Thus, although the
in incoming radiation. The lower green line shows the residual energy resultant quantum efficiency may be more or less constant
after it has been transferred to a frequency equivalent to that of chlo- throughout the PAR spectrum, there is a net loss of some 21% of
rophyll a at 680–700 nm absorption bands. the original energy; this is illustrated in Fig. 8.
View Online
Fig. 10 Illustration of the light reactions of photosynthesis (the so-called Z-scheme). The major functional units are represented as oval shapes;
photosystem II (PSII), plastoquinone (PQ), plastocyanin (PC), cytochrome b6f complex (Cyt b6f), photosystem I (PSI), ferredoxin (Fd), ferredoxin-
NADP reductase (FNR) (in order of electron transport chain) and ATP synthase. P680 and P700, refer to the reaction centres of photosystem II (PSII)
and I (PSI) respectively, the asterisk (*) indicates the excited state. The inset shows a schematic close-up of the light harvesting complex (LHC).
The broad mechanism of the conversion of the photochemical The electrons pumped by the two reaction centres eventually
energy into metabolic energy and reducing capability is complex, give rise to the production of the reducing agent (NADPH) used
although comparatively well understood. Briefly, there are two in the process of carbon assimilation. Two molecules of NADPH
functionally separate sites of photon absorption, coupled in are being produced per four electrons transported with a free
tandem by a chain of redox carrier molecules. The photon energy gain of 220 kJ mol1 NADPH. At the same time protons
absorption elicits a charge separation of at two reaction sites, at are pumped across the membrane into the inner cavity of the
photosystem II (PSII; l ¼ 680 nm) there is a charge separation of thylakoid (the lumen). This sets up a charge gradient. On their
1.7 V and 1.6 V at photosystem I (PSI; l ¼ 700 nm) requiring return, the protons spin a molecular rotor,** which gives rise to
respectively 164 and 154 kJ per einstein.k This is depicted in the synthesis of adenosine triphosphate (ATP), the biological
Fig. 10 as the so-called ‘‘Z-scheme’’ of photosynthesis. The energy currency. In total, three ATP molecules are formed per
electron flow away from the chlorophyll molecules draws elec- 12 protons transported, with a free energy gain of ca. 50 kJ mol1
trons from water. Whereas the formation of electrons operates ATP;†† along with the two NADPH molecules; the total
on a one-photon-per-electron basis, the dissociation of water potential energy yield is 590 kJ per 4 moles electrons transported.
requires an accumulation of four electrons to effect the produc- Although in theory this could be supplied by four photons (two
tion of one molecule of oxygen. A tetra-manganese complex acts einsteins at 680 nm and two at 700 nm would yield 692 kJ),
as an accumulator, being oxidised stepwise at four stages, the experimental observations suggest 8–10 photons‡‡ are needed
fully oxidised form then draws the four electrons from two per four electrons transported.
molecules of water in one step, producing a molecule of oxygen Some of the intermediates in the system have a limited lifespan,
plus four protons. The functioning of the water splitting system is and the statistical probability of arrival of 8 photons at a single
described in detail in ref. 48. This whole complex of photon
capturing mechanisms, charge separation, generation of meta-
** It has been held that Nature never evolved a wheel – it however did
bolic energy and reducing capability, and the water splitting
a billion or more years ago and on the nanoscale! Only recently have
system is embedded in the lipid membrane of flattened sac-like we developed the skills to observe these molecular mechanisms.
structures present in the chloroplast, known as thylakoids. †† Depending upon the circumstances the DG for the hydrolysis of ATP
varies from 45 to 55 kJ mol1. A mean figure of 50 kJ mol1 is used
in the present account.
‡‡ There have been two schools of thought over the number of photons
k An einstein is defined as a mole of photons, of unprescribed required per molecule of O2 split: either 3–6 photons or 8–10 photons (see
wavelength. ref. 56). Present consensus favours the latter.
View Online
capturing site within the required time frame is low. Nature has been a puzzle why an enzyme, of such crucial importance to
overcome this problem by linking together approximately 2000 life on Earth, has seemingly remained so inefficient since there
chlorophyll molecules in a light harvesting complex (LHC) with have been a billion or more years for improvements to have
a much smaller number of reaction centre chlorophyll molecules. evolved. It appears, however, that the enzyme is nearly
On average there are about 300 chlorophyll molecules per active perfectly optimised49 and that the perceived inefficiency is
reaction centre. This is termed the photosynthetic unit (see inset a fallacy.
in Fig. 10). The captured energy of photons very rapidly passes The first stable products of the reaction are 3-carbon organic
through this network of chlorophyll molecules, reaching the acids. It is from these compounds that all major biochemicals
reaction centre within a timescale of ca. 1010 s. (fats, fatty acids, sugars, proteins etc.) are eventually formed
(see Fig. 11). Carboxylation of Ru5BP in the Calvin cycle leads
(ii) The ‘‘light-independent’’ reactions. The ‘‘reducing power’’ to the production of two molecules of 3-phosphoglyceric acid
(NADPH + H+) and energy (ATP) produced by the light reaction (3-PGA), which is subsequently phosphorylated to
is used in the enzymatic ‘‘light-independent’’ part of photosyn- 1,3-bisphosphoglycerate (1,3-BPGA) and reduced to glyceral-
thesis to enable the incorporation of CO2 into organic material dehyde-3-phosphate (G3P). This reduction step is where the
and its subsequent reduction (see Fig. 11). The calorific value of reducing power generated during the photochemical reactions is
the product is around 469 kJ mol1 C. The first committed step is used. These set of three carbon compounds are the building
the carboxylation of the sugar ribulose 1 : 5 bisphosphate blocks for the synthesis of the basic biochemical fractions (see
(Ru5BP) by the enzyme ribulose bisphosphate carboxylase insert in Fig. 11). The biochemistry is complex50 and incurs
oxygenase (RuBisCO). RuBisCO exists as large hetero- further demands on energy and, where acetyl-CoA in an inter-
multimeric protein complex (536 kDa) and is an outstandingly mediate (lipids and some amino acids), loss of carbon and
sluggish catalyst, fixing 2–10 molecules of CO2 per active site per oxygen, and therefore biomass, as carbon dioxide. A simplified
second. In addition, it has the curious property of catalysing both analysis is made in Box 1 for triglycerides and proteins based on
the carboxylation and oxidation of its substrate. Which reaction hexoses as the starting point. Fig. 12 shows the biomass loss
predominates depends upon the ratio of the partial pressures of with increase in lipid production. Two matters are important to
the two gases; CO2 and O2. The oxidation reaction is wasteful of note here: first high lipid content is achieved at a cost of
energy. Most, but not all, aquatic microalgae overcome the biomass loss; second, whereas ‘‘theoretical’’ calculations of
predominance of oxygen over carbon dioxide in oceans photosynthetic efficiency are based on hexose production, the
(ca. 30-fold) by actively pumping in CO2 and so increasing its cells ultimately produce biomass which comprises lipids as well
concentration around RuBisCO. as proteins, this results in something in the region of a 30–50%
RuBisCO plays a central role in all plant photosynthesis, it loss of the original photosynthetically produced organic mate-
accounts for a major fraction of all living protein. It has long rial (see Fig. 12).
Fig. 11 Schematic simplified representation of the Calvin-Benson cycle in three parts, (i) CO2 fixation, (ii) reduction and (iii) regeneration. The average
cycling time of one ‘round’ of CO2 assimilation is 100 to 500 ms. The necessary energy (ATP) and reductant (NADPH+ + H+) (not shown stoichio-
metrically) are originating from the photosynthetic light reactions. RuBisCO: ribulose bisphosphate carboxylase/oxygenase; RuBP: ribulose-1,5-
bisphosphate; 3-PGA: 3-phosphoglycerate; 1,3-BPGA: 1,3 bisphosphoglycerate; G3P: glyceraldehyde-3-phosphate; F6P: fructose-6-phosphate; G6P:
glucose-6-phosphate; PEP: phosphoenolpyruvate.
View Online
Box 1: Biomass and energy losses incurred during the biosynthesis of lipids and proteins
Lipid synthesis: The biosynthesis of high calorific value compounds such as lipids from the primary products of photosynthesis –
hexoses – must incur a metabolic penalty in the form of loss of mass. A simple minimum calculation may be made from the change
in calorific value from 15.7 kJ g1 (a hexose) to 36 kJ g1 (a generic algal lipid). Thus, to conserve energy there must be a mass loss
of about 2.5. This, however, is a minimum calculation, as the 2nd law of thermodynamics calls for some energy, and therefore
additional mass, loss. This can be estimated from the biochemical stoichiometry of the various reactions during the trans-
formation of hexoses into lipids. If there is a negative energy balance, then as well as the mass losses during the biochemical
transformations the production of CO2 (see equation below), additional mass will be lost to provide the extra energy, which must
come from respiration. The overall synthesis of the primary lipid product – palmitic acid – from glucose may be divided into two
phases: (i) the formation of acetyl CoA and (ii) and the conversion of acetyl CoA to palmitic acid.
4C6H12O6 / 8 acetyl CoA + 8CO2 + 8ATP + 16 NAD(P)H (1)

8 acetyl CoA + 7ATP + 14 NAD(P)H / CH3(CH2)14COOH (2)
Overall:
4C6H12O6 / CH3(CH2)14COOH + 8CO2 + ATP + 2NAD(P)H (3)
Thus, there is a small energy gain in the form of an ATP and two NAD(P)H; however there will be the loss of 3ATP molecules
for the formation of the triglyceride and additional losses when palmitic acid is desaturated to form other fatty acids. These energy
gains and losses are small (100–200 kJ per palmitic acid molecule) as compared with the overall energy present in the molecule (ca.
10 000 kJ mol1) and can be largely ignored. There will be a small energy gain if the triose phosphates, rather than the hexose
sugars are considered to be the starting point for lipid biosynthesis.
Overall, the conversion of 4 molecules of a hexose to 1 mole of palmitic acid results in a drop in calorific value from 11 446 to 10 117
kJ. Thus, these calculations imply that, whereas there is a small loss of energy (ca. 10%), there is a substantial (2.9-fold) loss of mass.
Algal lipid contains a substantial amount of phospholipid and glycolipid. The mass loss in making phospholipid may be estimated to
be about 2.7 and making glycolipid 2.3 – these values are not as certain as the triglyceride figure. If we take the mean figures for the
proportions of these components of algal lipid given in Table 3, a weighted mean value of 2.6-fold loss of mass is obtained for algal
lipid formation from hexose; this has been used for the calculation of the mass yield of cells with varying lipid content.
Protein synthesis: The parallel calculation for the conversion of the photosynthetic products to proteins is more complex. The
elemental composition of algal protein may be calculated from the amino acid analysis as C1H1.56O0.30N0.26S0.006 (see Table 1); the
loss of mass on production protein from a hexose, assuming no respiration of carbon, is 1.36-fold, much less than the formation of
lipid. The energy gain on a mass basis is much the same as the mass loss (1.33) on a carbon basis. However, there will also be
metabolic work at two stages: (i) during the synthesis of the individual amino acids; and (ii) on their conversion to proteins (ref.
125 and personal communication) makes a very comprehensive analysis of the mass yield of synthesising the amino acids for
protein synthesis and obtains a value of 1.65 g g1 protein. The energy for the second reaction, which incurs a considerable
reduction in entropy, is estimated to be 120 kJ per peptide bond.46 The energy required per gram of protein synthesised can be
estimated from the nitrogen content of the protein and is small fraction (1.4 kJ g1, i.e. 6%) of the overall calorific value of the
protein, thus we may take a round figure of 1.7 g for the mass of hexose requires to synthesise 1 g of protein.
Nucleic acid synthesis: The calculation for the nucleic acids is extremely complex and, in view of their relatively small contri-
bution to the overall mass of the cell, the calculation has not been made.
Calculating the mass yields: given the values calculated above, the general equation for the biomass yield is
Y¼ 1/(1.11C + 1.7P + 2.6L) (4)
where C is the carbohydrate content (assuming it to be a hexose-based polysaccharide), P is the protein content and L the lipid
content, Y is the biomass yield in mass per mass hexose synthesised. Note: C + P + L ¼ 1. If it is assumed that the protein to
carbohydrate ratio is 3 : 2 and that this ratio remains constant with increase in lipid content, then the equation simplifies to:
Y ¼ 1/(1.46 + 1.14L) (5)
Estimating mass loss from calculated the photosynthetic quotient (PQ): If there is little metabolic work involved in the conversion
of photosynthetically produced hexose to another biochemical category, then the hexose mass required for its synthesis may be
simply calculated as 30 PQ/FW (g g1), where FW is the formula weight (with carbon set as 1) and 30 is the formula weight of
CH2O. For lipid, it underestimates the demand 9%, in the case of protein by about 15%, reflecting the scales of the metabolic work
needed.
Note: NAD(P)H is used to refer to both NADH and NADPH.
View Online
compensation irradiance is seen as an offset on the photosyn-

thesis–irradiance curve (see Fig. 13A); reported values show
a wide range from 03 to 12 mE m2 s1 PAR (equivalent to 0.01
to 0.46 MJ m2 d1 of incoming solar radiation). There are
thought to be differences between taxonomic groupings of algae
(see Fig. 13B) and ref. 55. The analysis in ref. 53 suggests that the
major part of the variation can be attributed to individual cell
biomass: Ec is approximately proportional to the square root of
cell mass. The compensation irradiance takes on importance in
dense algal cultures (see Section 4.4(ii) and Box 2).
3.3 Overall efficiency of photosynthesis

The energetics of the photochemical reactions in photosynthesis
and the linked enzymatic reactions associated with carbon

Fig. 12 Graph of the calculated yields of biomass (green line) from
hexose for biomass with varying lipid contents. The pro-
assimilation have been covered above. We now consider the
tein : polysaccharide ratio is taken to be constant at 3 : 2, this assumption relationship between the photon flux (the irradiance) and the
resultant photosynthetic rate. Characteristically the relationship
has a minor effect on the outcome of the calculations. The calculation is

based on eqn (5), Box 1. Shown also is the hexose energy required to is nonlinear (see Fig. 14). In the near-linear initial part of the
synthesise biomass of varying lipid content. The hexose is assumed to curve, there is excess enzymatic activity, with light limiting the
have a calorific value of 15.7 kJ g1. photosynthetic rate, and maximum quantum yields are obtained.
The flattening off of the curve at higher irradiances is a conse-
3.2 Respiration quence of the limitation of the rate of photosynthesis by
At first it may seem surprising that plants, that have the ability to
create metabolic energy de novo, should need a supplementary
energy generating process. There are a number of reasons, the
two main ones being: (i) to sustain their metabolism over periods
of darkness, so plants need some energy generating process
independent of light; and (ii) the mismatch of time scales between
the light-dependent and independent reactions in photosynthesis
can give rise to ‘‘traffic jams’’ along the electron transport chain
when the former reaction gets ahead of the latter. Respiration-
like electron disposing reactions have evolved to alleviate this
congestion. However, the penalty is wastage of the energy from
the captured photons. This is a major limitation to achieving
maximum photosynthetic efficiency in other than ideal labora-
tory situations. Photosynthesis can in most part be treated as one
of two alternative pathways. By contrast respiration is more
complex. In addition to the widespread so-called ‘‘mitochondrial
respiration’’, ref. 51 list four other forms of respiration in plants.
In the case of microalgae, overall respiration can be treated
operationally as a composite of two components: (i) a biomass-
associated component, which probably represents some basic
maintenance energy; and (ii) a photosynthetically-associated
component, which may result from the above decongestion
process but also from repair or replacement of the components of
the light gathering system which suffer photo-damage. A careful
analysis of these two forms of respiration for the major algae
groupings has been made.53 Normally, other than at very low
rates of photosynthesis, the dominant loss term is the photo-
synthetically associated respiration and the fraction of photo-
synthesis lost to concurrent respiration lies in the range 8–12%.
As respiration continues in the dark, these numbers are roughly
twice as great over a 24 h diel cycle. By comparison, respiration-
Fig. 13 A: Schematic of the various components of respiration and their
associated losses in terrestrial crop plants fall in the range of 30– effect on net metabolism. B: Analysis of data obtained from cultures of
60%.54 The point at which the respiration and photosynthesis algae in ref. 53, which illustrates the offset due to the compensation
curves intersect is the ‘‘compensation point’’ and the irradiance at irradiance. Gonyaulax tamarensis, a large dinoflagellate, would not be
this level is known as the compensation irradiance (Ec). Net a candidate organism for mass algal culture but is included as the
growth only occurs at light levels above this value. The compensation irradiance is clearly shown with this organism.
View Online
Box 2: Estimating the light threshold for growth of dense algal cultures
In dense, well-mixed algal cultures, growth will only occur when the daily average irradiance (E) through the culture is greater
than the compensation irradiance (Ec) i.e.
(E)/z $ Ec (1)
where z is the depth of the culture. The minimum incoming irradiance level that will allow growth (Emin 0 ) may be calculated
according to Sverdrup’s critical depth theory.127 Assuming the distribution of light with depth complies with Beer’s law, then the
above integral with depth has the following solution
Pz
0E ¼ (E0/k)(1ekz) (2)
where k is the beam attenuation coefficient (extinction coefficient). In dense cultures ekz1, thus the equation simplifies to
Pz
0E ¼ E0/k (3)
Combining eqn (1) and (3) gives the minimum irradiance for growth
Emin
0 ¼ Eckz (4)
For cultures with a biomass of 0.25 kg dry weight m3, the light extinction coefficient will be in the region of 50 m1. This
assumes chlorophyll, the main light absorbing pigment, to be 2% of the cell dry weight and to have a specific light attenuation of
10 m2 g1 (derived from data in ref. 128). Thus, for a raceway 15 cm deep, using the above numbers, eqn (4) reduces to:
Emin
0 ¼ 7.5Ec (5)
For candidate algae, the 24 h compensation irradiance, based on total incoming radiation, has a wide scatter, generally falling in
the region 0.03 to 0.3 MJ m2 d1 (see ref. 52, 53 and 55), with a geometric mean of 0.12 MJ m2 d1. This would give a range for
2
Emin
0 from 0.2 to 2 MJ m d1, with a median value in the vicinity of 1 MJ m2 d1.
enzymatic reactions. In this zone, the energy from the excess achieved by reducing the size of the light-collecting antenna, this
captured photons has to be disposed of in some manner, either by lowers the slope of the initial part of the curve, such that the
fluorescence or by one of the respiratory decongesting mecha- photosynthetic system saturates at higher irradiances. Thus, the
nisms. It is in the context of this effect that the time scale algae can exist as two physiological types – low light adapted
mismatch of the photochemical and enzymatic parts of photo- (high chlorophyll content) and high light adapted (low chloro-
synthesis becomes significant. If the high irradiances are sus- phyll content). It would appear that the low light adapted form is
tained for an hour or so, then the algae will adapt to them; by the default state. The phenomenon of light adaptation has
either increasing their enzymatic capacity but more commonly by a major influence on the yields of optically dense cultures as will
reducing the capturing efficiency of photons. The latter is be discussed in a later section.
It is important to stress that the following calculation of the
efficiency of photosynthesis restricts its attention to the ideal
circumstance. Even under these circumstances it is not
a straightforward calculation and all too often the complications
are overlooked or bypassed. A rigorous analysis of the thermo-
dynamics was made in ref. 56, and the energetics are concisely
outlined in ref. 54. The present calculation is made for the
production of one mol of organic carbon (of a formula CH2O),
assumed to have a heat of combustion of 469 kJ mo11;
a quantum yield of 8 moles of photons (8 einsteins) captured per
mol of oxygen produced (and CO2 assimilated) is assumed. A
quantum yield of 10 would give rise to a proportionately lower
yield. A quantum yield of 8 (4 quanta at 680 nm and 4 at 700 nm)
Fig. 14 The photosynthesis versus irradiance (PvE) relationship. The would have a total energy content of 1384 kJ per 8 einsteins. The
dashed black line is a projection of the initial rate, and the hatched area consequential voltage jumps of 1.7 V (PSII) and 1.6 V (PSI)
between this and the blue curved line for photosynthesis is an indication (equivalent to 1267 kJ in total) imply very efficient energy
of the photon wastage. The zone between the blue dashed line and the transfer at this stage. The reactions following these charge
solid blue line for photosynthesis is that of unexpressed enzyme activity. separations give rise to the formation of 3 moles of ATP and
View Online
Fig. 15 Energy losses during photosynthesis. A. The stepwise loss of energy during the production of 1 mole of organic carbon as glucose (calorific
value taken to be 469 kJ mol1 C). B. The fractional energy losses at the major steps during photosynthesis. The hatched lines are stages at which the
losses are variable.
2 moles of NADPH a combined yield of 590 kJ. These products composition of the cell will affect both biomass and energy
of the light reaction in turn give rise to the formation of 1 mole of yields. This is discussed in Box 1 and the effect of varying cell
organic carbon with an elemental composition of CH2O. Taking lipid content is illustrated in Fig. 12. A significant change in
this to be a hexose, with a calorific value of 469 kJ mol1 C, the yield can result from the state of oxidation of the nitrogen
overall efficiency of the formation of 1 mole of organic carbon source – whether the cells are using ammonia (or urea) or
from the 8 absorbed moles of photons themselves would be 34%, nitrate. If nitrate is used as a nitrogen source as opposed to
with the major part of the loss occurring during the light reac- reduced sources such as ammonia, then the reduction of nitrate
tions (see Fig. 15A and 15B) calls for extra energy, characteristically increasing the energy
There are also inevitable energy losses preceding and following demand by 20–30%. A summary of photosynthetic yields
these two sets of reactions. Photons are gathered from across the calculated from the present analysis, along with the various
whole 400–700 nm range – the balance will depend upon the yield terms are used in the study of the energetics of plant
photosynthetic pigments present in the cell. First, whereas culture, is given in Table 5.
the quantum yield can be remarkably constant across the PAR The above efficiencies may be regarded essentially as fixed
range, energy losses will be incurred transferring the energy to the ceilings for photosynthetic efficiency under natural conditions, as
lower frequencies of the longer wavelength photons (see Fig. 9). the losses are determined by the stoichiometry (metabolic losses)
This loss is about 20%. This reduces the photosynthetic efficiency and the thermodynamics of the processes – for which there is
to 27% (see Table 5). Second, the radiation available for limited scope for variation.
photosynthesis is only some 45% of the incoming radiation. There are two further forms of loss, which are variable, where
Thus, incorporating these losses, 3908 kJ of incoming radiation is there may be possibilities for manipulation of the photosynthetic
required to produce one mol of organic carbon, reducing the apparatus and/or cell metabolism: (i) concurrent respiration
overall yield to 12%. (Fig. 13); and (ii) photon wastage (Fig. 14). Taking a median
There are also energy losses associated with the metabolic estimate of the 24 h loss to respiration as 20% of photosynthesis,
conversion of the primary photosynthetic products (e.g. the overall yield of the incoming radiation is reduced to around
hexoses) to biomass i.e. lipids, proteins, polysaccharides, etc. 10%. A quantum requirement of 10, rather than 8, would reduce
This loss will be dependent upon two factors. The biochemical the overall maximum yields to 8% after respiration is considered.
Table 5 Calculation of maximum photosynthetic efficiencies. Yield terminology and symbols adopted from ref. 57. CHO refers to organic material with
a hexose-type elemental ratio, and a calorific value of 469 kJ mol1 C (15.6 kJ g1). CHON refers to the calculated carbon-normalised major element
composition (C1H1.64O0.44N0.13) for a cell containing 42% protein, 28% carbohydrate, 25% lipid and 5% nucleic acid, with a calorific value of 541 kJ
mol1 C (24.7 kJ g 1, data taken from Table 1). This gives photosynthetic quotients (DO2/DCO2) of 1.1 on ammonia as a N-source and 1.3 on nitrate.
Values are calculated on PAR except those within brackets, which are rounded off values based on total solar radiation
Type of yield Units CHO CHON (NH3) CHON (NO3)
Bioenergetic yield (J%) % (kJ kJ1) 26.7 [12] 17.7 [8] 14.3 [6.4]
Energetic yield (JkJ) g kJ1 17 103 [7.7 103] 7.2 103 [3.2 103] 5.8 103 [2.6 103]
Quantum yield (JE) g einstein1 3.8 — 2.1 — 1.7
Photons/mol C fixed einstein mol C1 8 — 10.2 — 12.6
View Online
Table 6 Observed and projected yields for crop plants and microalgae.
Figures for annual insolation are calculated as for Fig. 18A; theoretical
values are given for three latitude zones, the same as in Fig. 24. * allowing
for respiration, ** assuming a calorific value of 20 kJ g1 dry weight for
the crop as a whole, *** assuming a biomass calorific value of 24.7 kJ g1
dry weight (see Table 5) and a respiratory loss of 20%
Maximum
biomass
yield/tonnes dry
Crop weight ha1 a1 Source
Higher plants
Theoretical* (C3 plants) Low 210, mid 170, 54
c% ¼ 4.6% high 140**
Theoretical* (C4 plants) Low 270, mid 220, 54
c% ¼ 6.0% high 190**
Sugar cane 74–95 67–68

Switchgrass 8–20 19,67
Corn (grain) 8–34 66
Poplar wood chips 11 69
Soya 4.6–5.5 66 Fig. 16 Open ponds (image courtesy of Nature Beta Technologies Ltd,
Rape seed 4.5–6 66 Eilat, Israel, subsidiary of Nikken Sohonsha Co. Gifu, Japan).
Oil palm 8.7 66
Microalgae 4.1. Algal growth and production
Theoretical c% ¼ 12% (Table 5) Low 410, mid 330, Present
high 280*** account Initial work on mass culture of algae was carried out in ponds
Projected raceway, algae 110–220 16 and tanks of various forms, frequently with some form of
unspecified
Bioreactor raceway, algae 175 4 agitation. From these beginnings, two types of growth system
unspecified have evolved: (i) raceways; (ii) photobioreactors (PBR). The
Projected raceway, algae 127 4 former (see Fig. 16) are simple engineering developments of the
unspecified
pond/tank type system. At their simplest, they are oval in shape
Best case raceway, algae 120–153 9
unspecified with depths 100–300 mm and the water is kept in circulation with
Achieved bioreactor 182 70 paddle wheels – the most energy efficient system. Bioreactors (see
(Phaeodactylum) Fig. 17) are enclosed systems of various geometries, character-
Achieved raceway (Pleurochrysis) 60 71
over 10 months istically, although not invariably, with a light capturing depth of
less than 100 mm, often having internal dimensions as little as
30 mm. They may take the form of narrow tubes, horizontal or
By comparison the estimated equivalent yield for higher plants54 inclined tubes, vertical coils or columnar structures as well as flat
(see Table 6) lies in the range 4.6 to 6%, the difference being due plate structures. Many of the financial disasters in mass culture
to the greater respiration losses in higher plants. of algae have been associated with bioreactor-based systems (see
If we take the irradiance information used to create Fig. 18A ref. 58). The main problems with algal culture in bioreactors are
and the more cautious quantum requirement of 10, then, allowing the maintenance of the necessary turbulent flow in long lengths
for respiration, the maximum ‘‘theoretical yields’’ of a carbohy- of narrow tubing and the possibility of inhibition of
drate type molecule would be 410 tonnes dry weight ha1 a1
(112 g m2 d1) at low latitudes less than 30 , falling to 280 tonnes
dry weight ha1 a1 (75 g m2 d1) at latitudes 45–55 . Actual
yields (see Table 6) fall short for a number of reasons; photon
wastage (which can be up to 80%) being the major problem, when
attempting to maximise the output from mass cultures of algae.
Photon wastage is the principal reason why the actual yields fall
short of the theoretical. This phenomenon continues to tax the
ingenuity of researchers and engineers and is far from resolved.
We return to this problem in the following section.
4. Production rates: background and observations

Building on the background information on photosynthesis, we
consider the theoretical limitations to biomass production,
based on the incoming radiation and efficiency of energy
transfer from photons to biomass. We discuss two types of algal
biomass production systems, closed bioreactors and open Fig. 17 Photobioreactors (Greenfuels image courtesey of http://
ponds. www.flickr.com/photos/jurvetson/58591531/).
View Online
photosynthesis due to the accumulation of oxygen in the closed activation energy of ca. 50 000 kJ mol1. Biologists customarily
system. Active pumping with air-lift pumps induces turbulence use a simple logarithmic alternative to the Arrhenius equation –
and also helps to drive off the accumulating oxygen. On the the Q10 – defined as the change in rate over 10 C – the value is
positive side, bioreactors, being closed, are much less prone to characteristically close to 2. Temperature is different to nutrients
contamination than open raceways. A coupling of the above and light, in that it is treated as a determinant of mmax, rather
systems has been used and found to be very reliable and than a moderator. Relevant to outdoor mass algae culture (and
successful. In these coupled systems, large diameter (ca. 300 mm) in common with light), as well as the immediate response, algae
bioreactors act as a ‘‘nursery’’ stage, where a pure culture is also have some capability to adapt to the ambient temperature.
maintained, after which the cultures are periodically flooded into There are also physiological interactions between the responses
large area raceways, which serve as the ‘‘grow-on’’ phase. Such to temperature and light, as the photochemical system is essen-
systems have been successfully used for growth of the astax- tially temperature insensitive, whereas the enzymatic system is
anthin-producing algae Haematococcus pluvialis.7,16 temperature sensitive. Photoinhibition, and associated losses, is
Microbial production (or yield) may be regarded as a product high at low temperatures and high irradiances.60
of biomass (crop) and net growth rate i.e.:
4.3. Yields
Production (M L2 T1 or M L3 T1) ¼ biomass (M L2 or M (i) ‘‘Theoretical’’ yields. From the information derived in
L3) growth rate (T1) Section 3 (summarised in Table 5) and the calculated incoming
radiation, it is possible to calculate maximum ‘‘theoretical’’ yields

Biomass may be expressed as moles or mass of organic carbon of biomass per unit area for various latitudes and times of the
or, more commonly in algal mass culture, as ash-free dry weight year. One must stress that, for a number of reasons, this partic-
(per unit area or volume). Growth rate is subject to a number of ular calculation is intended to provide nothing more than a high
modifying physical, chemical and biological factors. Maximising ceiling value.
the output of these systems, with the number of interacting The calculation of incoming radiation is as follows: with a clear
factors is complex, as without exception, the relationships are sky, the incident solar radiation with the sun vertically overhead
non-linear. Further, with probably only a single exception (the may be taken as 1000 W m2, of which some 50% (500 J m2 s1or
apparent Arrhenius constant in the temperature response curve), 2300 mE m2 s1) lies within the PAR region. Given this, the irra-
the values for the parameters cannot be anticipated from theory diance may be calculated for each hour angle of the day and for the
with any degree of certainty. cycle of the solar declination angle through the year. Finally
a correction is made for reflective losses from the surface of the
water using Fresnel’s equation – although this correction is very
4.2. Controls on algal growth small. Details and the overall calculation may be found in ref. 47
and 61. The calculated daily irradiances for various latitudes are
Growth of microorganisms is usually modelled as a species- given in Fig. 18A. For comparison purposes, field data of total
specific maximum growth rate (mmax), reduced by one or incoming radiation, derived from satellite observations, are given
a combination of growth limiting factors – the two key ones in in Fig. 19. The point made by this latter figure is that local atmo-
the case of the algae are light and inorganic nutrients spheric conditions – notably cloud cover – give rise to a much more
(e.g. nitrogen, carbon or phosphorus). Characteristically the complex distribution in space and time of incident radiation than
effect is non-linear, with the initially controlling factor reaching the simple calculation from which Fig. 18A is derived.
a point where it no longer affects the growth rate. This is Once the irradiances have been obtained, then a simple linear
modelled by the biological equivalent of the Langmuir isotherm calculation of organic production may be made from the product
(Michaelis–Menten equation), giving broad nutrient-limited and of the irradiance and the energetic yield (see Table 5). The curves
nutrient-unlimited zones. In the case of light, there is a further given in Fig. 18B are for a ceiling value of bioenergetic yield of
zone of light inhibition (see Fig. 14). A number of equations have 10% this is obtained by correcting the 12% yield for a hexose
been devised to model this latter curve (see ref. 57 and 59). It is (Table 5) for a 20% loss due to respiration. To transform the
debated whether growth is controlled by the most severe limiting energy yields to mass yields (i.e. kJ to g dry weight) an energy/
factor or a product of all factors at one instance in time. mass conversion factor is required. Commonly (see ref. 56) this is
Two additional factors affect the growth rate of algae; both are calculated theoretically from elemental composition or estab-
of considerable importance in modelling and designing growth lished empirically by bomb calorimetry. These approaches do
conditions. The first is respiration (R), which essentially subtracts not take account of the metabolic work on transformation from
from growth (m), i.e.: hexoses (taken as the primary photosynthetic product) to other
biochemicals.xx A more correct value may be derived from the
mnet ¼ m R
xx Bomb calorimetry involves the conversion the nitrogen in the molecule
to dinitrogen, whereas the starting and end nitrogen compound in
The second factor is temperature. The overall temperature microalgal growth is commonly ammonia. Strictly a correction should
response has three cardinal points – a temperature minimum, be made to the reported bomb calorimitry values for the heat of
formation of ammonia from dinitrogen (46 kJ mol1) otherwise the DG
a temperature maximum and a temperature optimum. The zone
of the reaction will be overestimated. The error is small, about 0.4 kJ g1
from the temperature minimum to approaching the temperature in the case of protein and 0.2 or less in the case of algal biomass, and is
optimum closely follows the Arrhenius equation, giving effective customarily ignored.
View Online
Fig. 18 A. Monthly average daily clear sky total isolation for various latitudes. The calculation takes into account the loss from the surface due to
reflection. Calculation made following ref. 47 and 61. B. Estimated monthly averaged ‘‘theoretical’’ daily photosynthesis (as biomass dry weight,
composition as in Table 5, assumed calorific value 24.7 kJ g1) through the year at various latitudes; the calculation (see text) is based on a 10%
bioenergetic yield (see Table 5); no biological losses, respiration or organic excretion are taken into account. C. Variation of calculated productivity with
latitude, assuming 3 and 10% bioenergetic yields (again no biological losses are taken into account) and also using the tanh relationship shown in Fig. 22.
D. Monthly averaged ‘‘theoretical’’ daily productivity (as dry weight) through the year at various latitudes again using the same tanh relationship as
insert C.
to control by other factors such as temperature or inorganic

nutrients – so in this respect the rates obtained perhaps are best
regarded as maximal ones, unless major improvements can be
made to the light capturing system.
(ii) Observed yields. Fig. 20A summarises an extensive

collation of reports of area yields of algal biomass derived from
various culture systems. The reported rates have been plotted
against time, as there have been improvements over the years,
and separated into ponds, bioreactor systems and raceways.
They have been further separated into average (or sustained
values) and maximum values. Two exceptionally high rates
(130 and 174 g m2 d1) have been omitted, as they appear close
Fig. 19 Global distribution of total incoming radiation through the first
to or beyond accepted ‘‘theoretical’’ yields. Generally values
half of the year.62
above 40 g m2 d1 come from narrow path length (<50 mm)
bioreactors. The clusters of low rates (>20 g m2 d1) for race-
mass of hexose required to produce a given mass of biochemical ways in the post 2000 period come from observations during the
and the calorific value of hexose itself (taken as 15.7 kJ g1, ref. solar winter at latitudes greater than 30 . This is not anticipated
63, 64). The former can be estimated from either the protein/ from the radiation calculations and may be due to temperature
polysaccharide/lipid content (see Box 1, eqn (4)) or from the lipid limitation arising from evaporative cooling effect on the race-
content, assuming a fixed protein/carbohydrate ratio of 3 : 2 ways (see the following section). Furthermore, some of the data
(Box 1, eqn (5)). The 3 : 2 ratio for the protein carbohydrate come from high lipid algae, which will result in a depression of
proportions comes from the analysis of data in Fig. 5. the yield (see Box 1) and Fig. 12.
Fig. 18D shows the results of a calculation using a non-linear Ignoring the low cluster (see above) there would appear
relationship between photosynthesis and radiation – the tanh a general upward drift due to improvements with time to a figure
relationship between photosynthesis and irradiance (see discus- in the region of 35–40 g m2d1 – very much the prediction in ref.
sion in Section 4.4 (i) and Fig. 21). In Fig. 18C the data has been 65. The means of the maximum and average/sustained rates are
integrated through the year giving the distribution of calculated given in Fig. 20B (a geometric mean has been used to avoid biases
annual yields with latitude, the calculations are made for a 10% due to the wide scatter). Where the appropriate information is
bioenergetic yield, a 3% yield (an average obtained value, see available, the bioenergetic yield (j%) has been calculated (see
Fig. 20) and the above tanh relationship. The calculation Fig. 20C). The data set however is very small; some 19 sets of
assumes that only light limits growth – no consideration is given observations in total and predominantly from high performance
View Online
due to the fact that, in the case of the higher plants, the reported
crop (e.g. grain) may only be part of the above-ground plant
yield, and the below-ground yield (roots) is almost inevitably
omitted. There are major uncertainties with the estimations for
algal biomass yields in that there are very few reports of long time
series (i.e. data for a full year). When better sets of data become
available, we may find some downward revision of the numbers.
However, this may be offset by improvements in strains and
culturing techniques. We simply have to wait and see.
4.4 Controls on yields
Broadly, the yields commonly achieved in outdoor mass culture

systems lie between one tenth and one third of the ‘‘theoretical’’
value of 10% of the energy in the total incoming radiation

(cf. Table 5 and Fig. 20C). A number of factors give rise to this.
In Section 4.1, light, temperature and oxygen were identified as
factors that would affect photosynthesis; further controls may be

set by carbon dioxide concentrations, turbulence, inhibitors and
contamination.
(i) Light. Given the above efficiencies, a large fraction (60–

90%) of the incoming radiation is not used for photosynthesis.
Some 50% of the total incoming radiation is outside the range
useful for photosynthesis and so cannot be used. But this is not
the full story. Losses may result from light not being collected by
the culture or that the photons collected by the algae are in large
part wasted and their energy is lost as heat or fluorescence. The
former loss can be essentially disregarded with cultures with light
paths of 100 mm and biomasses of 1 kg m3, as the fraction of the
light emerging from the culture is a very small proportion (<104)
Fig. 20 A. Analysis of published production rates of outdoor cultures,
of that entering. The main loss is associated with the wastage of
grouped according to type of growth system. Colours designate the type
photons after absorption by the algae, which results in the
of system, crosses maximum rates, solid circles sustained rates, PBR:
photobioreactors, RW: raceways. B. Average (geometric means)
culture as a whole using light very inefficiently: the algae in the
production rates of ponds, bioreactors and raceways. Upper line is the first 5–20 mm of the light path being light-saturated or light
standard deviation, lower the standard error. C. Reported or calculated inhibited, whereas those in remaining 90–95% of the culture are
bioenergetic yield (j%, see Section 3.3 and Table 5), based on photo- severely light limited. The resolution of this problem has been,
synthetically available radiation (PAR; n ¼ 15) and total solar radiation and continues to be, the major challenge to maximising the
(TSR; n ¼ 4). ‘‘TSR’’ is the TSR efficiencies, plus the PAR values cor- productivity of mass cultures (see Section 4.5(i)).
rected to TSR values by dividing by 2 (see Section 3.1(i)). The ‘‘theoret- There is a paucity of reported data on the seasonality of
ical’’ maximum is taken from Table 5 allowing for 20% respiration. Data production rates for PBRs and raceways, which allows the
and sources given in Appendix IV (provided in the ESI†). overall relationship between daily irradiance and production
rates to be explored. Fig. 21 and 22 show the sets of data we have
photobioreactors, where the mean bioenergetic yield (based on been able to locate.70,72,73 The relationship between light and
PAR) is close to 10%; the mean for raceway observations is much areal production rates was analysed using the tanh model for the
lower at 2%, with these values being around 30% and 10% PvE relationship (P ¼ Pmax*tanh(a*E/Pmax), where Pmax is the
respectively of the ‘‘theoretical’’ yield (see Table 5). The generally maximum photosynthetic rate, a is the initial slope, E the irra-
lower values for raceways, compared with bioreactors, may in diance; see ref. 47. The fitting to the data was undertaken using
some part be due to design limitations, but more likely that the a least squares analysis. We found it necessary to allow for an
bioreactors are far less susceptible to contamination and the offset on the irradiance axis. The tanh model was chosen as it
populations in raceways may be growing sub-optimally. provides a direct parameterisation of the initial slope (a), from
Table 6 sets out yield data for crop plants along with those which the photosynthetic efficiency (j%, see Table 5) may be
available for microalgal cultures. In round figures, the higher calculated. In order to calculate the photosynthetic efficiency,
plant annual yields fall generally in the range of 5–30 tonnes crop a calorific value has to be prescribed for the organism, this was
dry weight per hectare (sugar cane being the striking exception), calculated from the reported lipid content for the organism and
the algal crops 50–150 tonnes. The basis for the difference, the calorific values reported in Table 1. The data have been
although striking, is not wholly clear. In part, it may be that the separated into PBRs and raceways. At first sight there would
longer growing period in the case of higher plants results in appear to be a significant difference between the rates in race-
respiration taking a higher toll on the yields. It is also likely to be ways and PBRs (as seen also in Fig. 20B and C) but in ref. 72 the
View Online
Table 7 Parameters derived from the analysis of data shown in Fig. 21

and 22. Sources: ref. 70, organism cultured: Phaeodactylum tricornutum,
ref. 72, organism cultured: Tetraselmis suecica, ref. 71, organism cultured:
Pleurochrysis carterae
Growth system PBR PBR PBR RW RW
Dimensions/mm diameter or depth 30 60 35 150 150

Reference 70 70 72 72 71
Pmax/g m2 d1 1400 86 20 20 41
a/g MJ1 1.7 3.5 2.9 1.9 2.3
Offset/MJ m2 d1 4.2 0.5 4.6 3.3 9.7
Lipid content (% dry weight) 16 16 26.5 26.5 22.5
Calculated calorific value/kJ g1 23.3 23.3 25.0 25.0 24.3
Observed photosynthetic efficiency 3.9 8.2 7.1 4.6 5.5
(J%)
Fig. 21 Analysis of three data sets70,72 of the seasonal changes in biomass

yield in outdoor photobioreactors. The study in ref. 70 was undertaken in seen in the data of ref. 72 and 73. This offset is not as evident, or
southern Spain at a latitude 36 480 N; the culture organism was perhaps absent, in the data from ref. 70, but that data set lacks
Phaeodactylum tricornutum. The study in ref. 72 study was made in observations at the low irradiances needed to establish whether
central Italy at 43 80 N; the culture organism was Tetraselmis suecica. or not there is an intercept, furthermore, the plot for the 30 mm
Using a least squares analysis, the data were fitted to the tanh expression PBR certainly looks anomalous. Thresholds, equivalent to 5–9
P ¼ Pmax*tanh(a*(E Emin)/Pmax), where P is the observed rate of MJ m2 d1, were reported for mass cultures in ref. 74 (see
photosynthesis, Pmax is the maximum photosynthetic rate, a is the initial Fig. 14.4 and Table 14.5 in ref. 74). The simple explanation for
slope, E the irradiance.47 The lines were not forced through zero, but an the threshold would be that it represents the compensation
offset on the irradiance axis (Emin) allowed. The derived parameters are
irradiance (see Section 3.2) but an analysis for dense cultures (see
given in Table 7.
Box 2) suggests this would give rise to a maximum value of 2 MJ
same organism was grown in parallel in PBRs and raceways and m2 d1, distinctly below the observed range of 3–9 MJ m2 d1;
the rates are essentially identical, so the apparent differences may so it cannot be the sole explanation. Water temperature also
be more due to the organism or the environmental or operating varies with irradiance and will have an effect upon photosyn-
circumstances. thetic, and respiration rates, thus the observed offset may be
The derived parameters from the analyses and the photosyn- a combined effect of temperature and irradiance. A temperature-
thetic efficiency and its calculation are summarised in Table 7 parameterised model would be needed to resolve this. The data in
and a number of features are worth noting. Most important is Table 7 have been used as the basis of parameterisation for the
low spread of values for the initial slope (a); the mean value is production model used elsewhere is the review.
2.4 0.75 g MJ1 gives a calculated photosynthetic efficiency of
5.9 1.8%; allowing for a 20% loss incurred as respiration, (ii) Temperature. The incoming radiation has a marked
a value for the primary photosynthetic efficiency of 7.4% is diurnal effect upon the temperature of the systems. ref. 71, 75, 76
obtained, very close to the theoretical value of 8% for the observed 10–25 C increases in temperature during the day due to
production of biomass (see Table 5). The second important solar heating. In the case of shallow open systems there is
outcome is the apparent irradiance offsets (3.3 to 9.7 MJ m2 d1) a counter effect on temperature due to evaporative cooling.
Evaporative losses at middle to low latitudes fall in the region of
2 m per annum. On a daily basis this results in a daily cooling
effect equivalent to some 12 MJ m2 (ca. 300 calories cm2). With
150 mm deep raceways this generates a potential cooling of 20 C
per day. Given a Q10 of 2, this would give rise close to a 4-fold
reduction in metabolic rate and possibly a shift in the photo-
synthesis/respiration ratio. The matter is clearly more complex
than this, as there will be some moderating feedback between
cooling and evaporation – but as evaporation rates are propor-
tional to degrees Kelvin, the feedback will be weak.
Temperature is very likely a major factor determining the
productivity of outdoor growth systems open to the atmosphere
and may limit the latitudinal extent to which they can be
successfully used. Clearly supplementary heating is impractical,
as the energy required would far exceed the energy gain from
Fig. 22 Analysis of two data sets71,72 of the seasonal changes in biomass photosynthesis. It does not seem practical to use solar panels as
yield in outdoor raceways. The study in ref. 71 was undertaken in Perth in heat collectors to offset the heat loss due to cooling, as the
Western Australia (latitude 31 570 S), the organism cultured: was Pleu- incoming radiation (10–25 MJ m2 d1, see Fig. 18A) is, as it
rochrysis carterae. The irradiance data were obtained from the Australian should be, not too dissimilar from that associated with the
Bureau of Metrology database. Other details as Fig. 21. evaporative associated losses (12 MJ m2 d1), so the heat
View Online
collectors would need to be broadly of the same area as the showed that, to have a significant effect, the dark period needed
growth systems and would almost certainly not be economic. to be 10 times greater than the light, consistent with the above
generalisation on relative processing rates of the photochemical
(iii) Oxygen. Oxygen is a by-product of photosynthesis and it and enzymatic reactions. The gains in efficiency are lost at flash
has been long known that high oxygen concentrations inhibit periods shorter than 10 ms. In dense cultures, where perhaps 90%
photosynthesis (the so called ‘‘Warburg’’ effect). Part of this of the culture is in effective darkness, turbulence cycles the cells
derives from the competing oxygenase activity of RuBisCO (see through a dark/light cycle. There has been extensive theoretical
section 3.1(ii)). Oxygen concentrations in the region of 400% and experimental work on this effect. Ref. 80 undertook a very
saturation (ca. 2 mM) essentially switch off photosynthesis. detailed study with dense algal cultures (3–6 kg m3) and found
Commonly encountered photosynthetic rates in mass cultures the flashing light effect to cut out at about 10 Hz. Ref. 81
(ca. 700 g dry weight m3 d1) result in net oxygen production concluded that the response in outdoor tubular bioreactors
rates of 30 mol m3 d1 or a rise in oxygen concentrations of saturated at frequencies greater than 1 Hz. Ref. 74 concluded
30 mM over the daylight period; well in excess of the 2 mM that power consumption considerations would limit flows to
inhibitory level. The exchange of gases across the air/water 0.15–0.2 ms1. At these flow rates, in raceways 150 to 200 mm
interface results in some 20 cm depth of water being re-aerated deep, the cross transfer frequency due to turbulent diffusion
per hour. Thus with open raceway systems, provided the system would be of the order of 0.1 Hz,{{ thus there is potential for
is turbulent (flow rates in excess of 0.05 m s1 ensure this, see the improvement of production rates by increasing the turbulence in
following section), accumulation of oxygen should not occur. In raceways. Ref. 82 used simple hydrofoil spoilers to induce
the case of enclosed bioreactors, oxygen may be calculated to turbulence and with a raceway system was able to achieve
accumulate at rate of about 4 mol m3 h1 (ref. 60 gives a figure of a mixing frequency of 0.5 to 1 Hz, which gave rise to a two-fold
6 mol m3 h1 for Spirulina), thus reaching inhibitory concen- or more gain in areal production. In a subsequent paper, using
trations in 20–30 min. Bioreactor systems characteristically use the same arrangement, ref. 83 report sustained production rates
airlift pumps to circulate the water, which should strip off the of 40 g m2 d1, with short term, maximum rates in the region of
accumulating oxygen, however, these rapid rates of oxygen 60–70 g m2 d1. With the exception of a couple of questionably
production impose constraints on the maximum length of run values, these rates are the highest achieved with raceway cultures,
and the minimum flow rates within the bioreactors. There is and are equivalent to bioenergetic efficiencies (based on total
apparently a combined effect of temperature and oxygen. Ref. 60 incoming radiation) of 5.5%, with a high value of 9.5%. Whether
reported that a fall in temperature from 35 to 25 C, accompa- such systems can be made economic remains an open question.
nied with a rise in oxygen concentrations from 0.69 mM to 1.9
mM, gave rise to a substantial (60%) fall in photosynthetic rates.
4.5 Maximising yields
(iv) Carbon dioxide. Many of the details for oxygen are (i) Maximising biomass yields. A full discussion of max-
shared with carbon dioxide, although the implications are imising the overall output and economy of algae biomass/biofuel
profoundly different. Consider the example of seawater, where production, as implied before, is complex and beyond the scope
there is a broad chemical predictability. Seawater total inorganic of the present review. We restrict our focus on maximising the
carbon dioxide concentrations are approximately 2 moles m3. utilization of photons, i.e. maximising the photosynthetic effi-
At daily production rates of 30 moles m3 (see above), the carbon ciency. This has long been seen as crucial to the success of algal
dioxide species in seawater will be exhausted within an hour of biomass production.15,78 The problem is peculiar to mass algal
so. The rate of CO2 replacement is about one tenth that of culture. Near ‘‘theoretical’’ yields can be achieved in short light
oxygen, as the driving force of the reaction is associated with the path lengths of dilute cultures at low irradiances. Mass algal
dissolved CO2, which is only a minor component of the seawater production involves optically dense cultures (biomasses 1 kg
carbonate system. The consequence is that re-carbonation of the m3, giving light attenuation coefficients of 200 m1) and supra-
water by simple exchange with the atmosphere will not sustain optimal irradiances. The consequence is that 90% of the photons
the rates of removal, thus the CO2 needs to be replaced contin- are absorbed in the first 10 mm by algae suffering severe light
uously by the addition of gaseous or liquid CO2. The supply of inhibition and so use photons very inefficiently – wasting perhaps
CO2 could well be a major constraint on the siting and conse- some 80–90%. Thus the remaining 90% or more of the culture is
quential scale of algal biomass production facilities. using photons very efficiently, but exists in virtual darkness.
Mixing by inducing turbulence alleviates the problem to some
(v) Turbulence. It had been demonstrated early on in the extent but would not appear to solve it fully. In essence, the root
study of the mechanism of photosynthesis that the yield of of the problem is that, in mass cultures, the algae are presented
oxygen per photon was greater in intermittent light.77 The basis with a very unnatural situation and they are not equipped
for this is that a single 400 chlorophyll photon collector can physiologically to deal with it.
process some 2000 excitations per second, whereas downstream Photosynthesis occurs as three events with badly matching
enzymatic carbon assimilation can process at best 100–200 time responses:84 photon capture rates have timescales of 106 s
electrons per second.78 The dark period, between the flashes, or less, whereas enzymatic CO2 fixation and reduction have
allows the slower enzymatic reactions to catch up with the
photochemical reaction. The optimum dark period is tempera-
{{ This is calculated as the transverse timescale ¼ h3/(0.06kz) ¼
ture-dependent but generally falls in the region of 50 ms. Ref. 79 h3/(0.06u*xh) ¼ h2/(0.06 0.05 u) z 30h2/u; where h is the depth
extended the early studies in ref. 77 for growing cultures and and u is the velocity, all in SI units.
View Online
timescales 1 s and the adaptation to changes to elevated light photons cannot be expected to be complementary. It is not
levels even longer timescales 103 s. Thus, the adaptive response possible without a sophisticated model to determine the
is far too slow to resolve the problem of super-optimal irradi- combined effect of these two approaches. The cautious conclu-
ances. It is possible to mix the culture on time scales of fractions sion is that given these developments one could reasonably
of a second, but not on the required microsecond time scale, so expect to approach 50% of the ‘‘theoretical’’ efficiency, giving
although mixing will alleviate the congestion in the electron a maximum potential bioenergetic yield in the region of 5%.
chain to some extent with a consequential increase yield, it does
not solve the fundamental problem. The solution lies in either (ii) Maximising lipid yields. At first sight a logical strategy to
speeding up the enzymatic reactions or slowing down the improve the economy of algal biodiesel production would seem
photochemical ones. The slow working, apparently inefficient to be to obtain and grow algae of high lipid content. This was
enzyme RuBisCO has been an obvious target based on the a major research thrust of the US Aquatic Species Program.
supposition being that one could improve on evolution. Broadly, two lines of research were followed, the first was to
However, the analysis made in ref. 49 would appear to show this isolate and screen algae for high lipid producers. The second
assumption to be false. Even if one were able to engineer approach aimed to increase the extent of lipid production by
improvements in the reaction rate of RuBisCO, or modify the cell identifying the rate limiting enzyme and modify the cell’s genetics
to produce more, the probability is that another enzyme would to increase its expression and so increase yield. The enzyme
then become limiting and the problem would end up processing acetyl-CoA carboxylase, which catalyses the carboxylation of
through the complex cascade of enzymes of the dark reaction. acetyl-CoA to malonyl-CoA, is the first committed step in lipid
Solving the other end of the problem, by reducing the photon synthesis and was regarded as the limiting reaction. It had been
capturing capacity, has long been seen as the more likely solution possible to increase the level of expression of the enzyme two- to
(e.g. ref. 15, 78). When exposed to sustained high irradiances, three-fold; however, no associated increase in lipid levels was
algae adapt by reducing the size of the light harvesting antennae found (see ref. 3, p. 138). Unfortunately this work was in its
and so reduce the rate of collection of photons. The problem is infancy when the ASP was terminated, so the feasibility of raising
that the time-scale of the adaptation is long as compared with the lipid yields by such means has not been fully explored. An
other timescales in photosynthesis and relies upon the algae alternative approach to producing lipid-rich cells is to drive the
being held for periods of tens of minutes to an hour or so at high cells into nutrient deficiency, when lipid accumulates (see Section
irradiances. In mass culture, the algae exist in a rapidly fluctu- 2.3, Fig. 7).
ating light regime making it impossible for them to adapt to the There are, however, penalties in producing cultures with high
high irradiances. Furthermore, unless the high irradiances are lipid content. First, as the fraction of lipid is increased some
sustained, the algae appear to revert back to the low light/high other cell component (e.g. protein, nucleic acid or carbohydrate)
chlorophyll state.85 The solution is to manipulate the physio- must decrease. The proteins and nucleic acids are major deter-
logical control mechanism and to switch the organism perma- minants of maximum growth rates, and it is far from clear how
nently into the high-light adapted state. This problem has these biochemical groups change as lipid content changes (see
attracted research interest and headway has been made.85–91 The Fig. 5). The coefficient of relationships shown in Fig. 6 implies
chlorophyll antennae are assembled on a scaffold of so called that an increase in lipid content from 15 to 30% would give rise to
‘‘light-harvesting’’ proteins and it is these proteins that set the size a 50% or greater reduction in growth rate. The reduced growth
of the chlorophyll antennae. It has been possible to down regu- rate is accompanied with a proportionate increase in the period
late these proteins and so reduce the photon capturing ability of the organisms would need to be held in the growth system and
the algal photosynthetic system (see e.g. ref. 86). The photo- a consequential increase in costs. Increasing their lipid content,
synthetic unit associated with photosynthetic system II can be by driving the algae into nutrient deficiency, ultimately gives rise
reduced some 3-fold from some 300 chlorophyll molecules down to the same problem; it extends culturing time and increases the
to 95, but this is not achieved in the case of photosynthetic system space and capital requirements of the production plant. At the
I.88 The reduction in the size of the light collecting antenna gives present it would appear that there is limited potential to increase
rise to 2-fold or more enhancement in photosynthesis and growth the lipid content of the algae and for a number of reasons 35%
at high irradiances,86,88 importantly without apparently lipid content appears to be a realistic maximum target.
compromising enzymatic capacity and there appears to be no An apparently unrecognised penalty of growing algae of high
penalty due to rise in respiration rate. lipid content is the significant overall biomass loss that neces-
The results from a simple spreadsheet model show that if the sarily occurs when lipids are synthesised from the primary
chlorophyll content is reduced, a greater number of photons are products (hexoses) of photosynthesis. This is analysed in Box 1.
delivered to the deeper parts of the culture, resulting in a more There is a 50% loss in mass resulting from removal of oxygen
efficient photon usage and greater areal productivity. The effect during triglyceride synthesis; a further third is lost as CO2. In
is most pronounced at intermediate irradiances and as a conse- combination, and allowing for the different component lipids
quence higher latitudes. This simple model suggests that 2-fold and their proportions in algae, we estimate there to be around
gains in photosynthesis can be expected, consistent with the a 2.6-fold reduction in biomass during the synthesis of microalgal
supposition in ref. 15. However, it is unlikely that one will be able cell lipid from a hexose – this effect can be seen in the yield
wholly eliminate the problem. calculations for cells of varying lipid contents in Fig. 23. If the
Improving productivity of the culture in this way probably economic analysis focuses exclusively on biodiesel and lipid
negates some of the gains that can be achieved by mixing, so production, then this may not be a problem. However, if the
a priori the two approaches to improving the utilization of components of the cell that are diminished at the expense of lipid
View Online
Fig. 23 Estimates of algal areal biomass yields (both daily and annual) for a range of latitudes. The calculations are based on calculated clear sky total
insolation (see Fig. 18A) coupled to a tanh model of the relationship between photosynthesis and irradiance, described in Fig. 21. The parameters used
were derived from the analysis of the data in Fig. 21 and 22, summarised in Table 7. The value for the initial slope (a) was the average of the full data set,
the irradiance offset (Emin) was derived from ref. 70, 72; the dataset from70 lacked the observations at the low irradiances necessary to adequately
constrain the intercept. Pmax was taken from ref. 71 – the maximum rates in ref. 72 are low when compared with the data shown in Fig. 20A and
accordingly were not thought to be appropriate. The initial calculation is of primary photosynthetic production of a hexose, so the individual values for
a and Pmax have to be scaled up to take account of the fact that the observations were of biomass, which has incurred a loss of mass on converting hexose
to proteins, polysaccharides and lipid (see Box 1). The final values used were a ¼ 3.55 (dimensionless), Emin ¼ 5.86 MJ m2 d1, Pmax ¼ 76.9 g m2 d1; the
calculation also requires the calorific value for the primary photosynthetic product, which was taken as 15.7 kJ g1 (ref. 63, 64). The yields for various
lipid concentrations are derived from primary hexose production and the lipid content using the eqn (5), Box 1. This assumes a protein to carbohydrate
ratio of 3 : 2, the mean value found in Fig. 5. The field observation of yields (vertical bars) placed at latitude 20 come from ref. 16; the values at placed at
latitude 35 come from ref. 71, 72.
increase are an important part of the economic equation, then observed integrated rates and the predicted ones for algae with
this loss on conversion could well be a problem, especially if they realistic biochemical compositions. The calculation is made from
were to command a price comparable to or greater than biolipid clear sky radiation data, which will overestimate the incoming
(see Section 6). Fig. 12, which derives from eqn (5), Box 1, shows radiation (cf. Fig. 18A and 19) so overestimate growth rates.
the effect of the cell lipid content on the total biomass and lipid There is little to be gained at this stage by making a detailed
yields from the primary hexose photosynthetic products. calculation based on observed insolation rates. As a counter to
this, the calculation makes no attempt to anticipate improve-
(iii) Predicting maximum yields. We have attempted to ments of the light capturing apparatus of the algae which we may
incorporate the findings of this section into a model of algal expect to occur (see Section 4.5(i)).
biomass production (see Fig. 23). In essence we have used the There is some steepening of the poleward gradient between the
calculated clear sky irradiance values (Section 4.3(i) and latitudes 35–55 . This may limit year-round outdoor culture
Fig. 18A) and parameters derived in major part from an analysis systems to latitudes less than 35 (see also Fig. 19). This appears
of the data reported in ref. 71 to produce biomass production to be the case with commercial production facilities. The long-
rates. We chose the Moheimani data set as it comes from race- standing, and successful, Earthrise Farms in southern California
ways, which it is consistent with our economic analysis (see (latitude 33 080 N) only operates its raceways for 9 months of the
Section 6). year, whereas the systems at Kanahoe Point on the Big Island of
The initial calculation we make is the rate production of Hawaii (latitude 19 470 N) are operated successfully on a year-
hexose. In order to do this we need to back calculate the rates of round basis, with no evident seasonal reduction of produc-
hexose formation from the biomass-derived rates. This is done by tivity.16,92 The experimental facility at Roswell (33 N), which
reversing the calculation of the mass loss during the biomass. was part of the ASP programme, ran into problems of snow and
Ref. 71 observed a mean lipid content of 36.5% of dry weight. ice during the winter period. In the model, the trend with latitude
This (from eqn (5), Box 1) gives a figure of 1.88 for the biomass to derives mainly from the reduction in annually integrated inso-
conversion hexose and scales up the Pmax value from 42 g m2 d1 lation; the effect will very likely be much greater in reality as one
(Table 7) to 77 g m2 d1 and the initial slope (a) from 2.43 to may expect additional strong temperature effects on growth.
3.55 g MJ1. The data sets of ref. 71, 72 show a mean offset on the This affirms the need for temperature-parameterised models of
irradiance axis of 5.86 MJ m2 d1, this mean has been used for photosynthesis and respiration.
the calculation. The calculated photosynthetic hexose produc- Fig. 23 shows the very marked negative effect of the lipid
tion rates are then converted to production for biomasses of content on the yields, and this questions the search for high lipid
varying lipid contents (15, 35 and 50% of cell dry weight) using species for, unless they possess uniquely high intrinsic photo-
eqn (5) in Box 1, assuming a protein : carbohydrate ratio of 3 : 2. synthetic rates, they will give low biomass yields. There may be
Given the above parameterisations, we can calculate the yields an additional effect due to the reduction in growth rate at high
per unit area, and their variation with latitude from the insola- lipid content, resulting in the algae failing to achieve the rates
tion data derived in Fig. 18A and the results are shown in Fig. 23. that the irradiance would allow. Thus, there would appear to be
Shown for comparison purposes are results from three reports of diminishing returns with increased lipid content (see also
seasonal productivity abstracted from published sources. It can Fig. 12); this is seen when the economics are considered (see
be seen in Fig. 23 that there is broad agreement between the Fig. 24).
View Online
5. Harvesting and processing of microalgal biomass manually. An alternative design is the solid bowl centrifuge in
which the bowl contains a disc stack, causing the biomass to
Before considering the economics of the process (Section 6), we collect at the outer part of the bowl. However this set-up also
want to highlight the challenges associated with the harvest and requires manual intervention in removing the biomass from the
processing of microalgal biomass. The methods that are most bowl. A third type of centrifuge is the nozzle-type centrifuge,
used for harvesting of algal biomass are discussed in this section, similarly containing internally stacked discs, and in which the
with an emphasis on the use of flocculation and centrifugation. shape of the bowl is modified so that the conical section of the
With regards to biomass processing, we briefly discuss the prin- bowl has evenly spaced orifices (nozzles). The biomass is
ciples behind the methods relevant to the economical process we collected in a conical space on the side of the bowl where it is
discuss in section 6. automatically discharged. This last type is a true continuous
Out of all steps considered in an algal biomass production centrifugation system with minimal manual intervention98 and
process, harvesting and processing of the biomass contains the could be considered the most appropriate type to harvest the
largest uncertainty surrounding cost and effectiveness. The large algae cultures. These three types of centrifuges vary in the
separation step in producing algal biomass as feedstock for volume processed per hour, with a maximum of 50 to 100 m3 h1
biodiesel is difficult and inherently more expensive by compar- (Table 8; ref. 15 and centrifuge industry representatives, personal
ison with oilseed crops. For algal biomass, there are three main communication).
phases in the preparation process prior to biodiesel production: An alternative to settling is flotation of biomass, where
(i) harvesting of the algae; (ii) disrupting the cells; and (ii)
dispersed air bubbles in the culture adhere to flocs formed during

extraction of the lipids from the biomass. The efficiency and flocculation and cause the cell aggregates to float. This process
methodology used in these three steps can have a major impact can operate more efficiently and rapidly than sedimentation or
on the economics of commercial algal biofuel production. These settling and can also achieve a higher solids fraction.15,97 This
processes have been largely avoided in past reviews, despite being process is potentially scalable to large-scale algal ponds.15
identified as a high priority for the future R&D.93–95 However, it is probable that an extra concentration step (possibly
centrifugation) may be necessary because of the relatively low
final product concentration.
5.1 Biomass harvest Flocculation, in combination with centrifugation or dissolved
Cell harvest we take as the initial 50–200-fold concentration step air flotation, is also widely method used for harvesting micro-
of the algal cultures yielding a biomass product that can be used algal cells. Flocculation refers to the aggregation of particles in
for further processing. The choice of harvesting methodology, suspension to form clumps that can overcome the Stokes’ drag
likely to be a combination of individual methods, depends highly force and increase the settling rate. This reduces or avoids the
on the biomass type and requirements of the downstream pro- need of energy intensive separation mechanisms like centrifuga-
cessing (Sections 5.2 through 5.4). tion. It has been suggested that this process could be adapted to
Settling rate (vs) of microalgae is a functon of two forces (drag, be the lowest cost harvesting process.15 However, flocculation
Fd and gravity, Fg) and the diameter (rc) and shape of the cells. rate and efficiency are highly dependent on cellular characteris-
The relationship at low Reynolds’ numbers for a sphere is tics, such as cell volume and cell wall properties, or culture
described by Stokes’ law: characteristics, such as cell number, culture age and growth
phase.99,100
Two mechanisms that give rise to flocculation are: (i) charge
vs ¼ Fg/Fd ¼ 2/9grc2 (rcell rfluid)/h
neutralisation, i.e. the attraction between the negatively charged
algal cell surface and the positively charged flocculentl; and (ii)
The settling rate of cells in suspension follows a square rela- bridging, when larger flocculent molecules span a distance large
tionship with the radius of the cells and a linear relationship with enough to adsorb to multiple particles and hence ‘bridge’ the
the difference in density (r) between the cell and the fluid and the particles.101,102 Several different types of flocculants can be used:
kinematic viscosity of the fluid (h). polyvalent metal salts, synthetic cationic polymers, high pH and
Centrifugation is the most common method for harvesting bioflocculants. The low cost of metal salt flocculants make these
algae from large volume cultures and is widely applicable to the more popular flocculants. However, the potential toxicity of
a variety of algal species.15,96,97 Although this process is effective, the wastewater fraction and the undesirable presence of metals
it is also expensive and energy intensive. The operational in the biomass to be processed are concerns that arise and may
variables, such as centrifugal force, flow rate, biomass settling inhibit their widespread applicability in the microalgal biomass
rate and settling distance will determine the efficiency of production process. This would be particularly the case if the
centrifugation. spent medium were to be recycled.
There are three different types of in-line industrial centrifuges: The main advantage of flocculation as a harvesting method is
multi-chamber, nozzle and solid bowl. All three types are the relatively cheap nature of the process and high volume pro-
continuous in that there is a continuous flow of culture through cessing. However, disadvantages include the species and growth
the centrifuges. A multi-chamber centrifuge consists of a number stage-dependent nature of flocculation efficiency. Until now, no
of tubular bowls arranged coaxially, causing each chamber to clear correlation has been demonstrated between flocculation
collect particles of a specific size due to the distinct centrifugal efficiency, dose and algal taxonomic group. Another disadvan-
forces exerted in each chamber. A solids content of 20% can be tage is the low solids content (i.e. < 10%) of the biomass har-
achieved, however the solids need to be removed from the bowl vested by flocculation, which requires further concentration
View Online
using other methods. As a result, flocculation is often combined

Table 8 Summary of comparisons between relative harvesting throughput, applications, capital and operating costs. Settling refers to autoflocculation, changes in pH in a stagnant culture will induce
h1 basis, are at best approximate since these were obtained from earlier reports and simply adjusted for inflation, not taking improved engineering into account. c Values given in brackets indicate the
Centrifugation is subdivided into multi-chamber, nozzle and solid bowl centrifugation. The relative harvesting costs include reference to operating, maintenance and capital costs and are based on
Values presented are averaged costs of biomass from cultures grown at unregulated pH, pH 9.6 or pH 9. e The cost calculations have been adjusted for inflation and converted into today’s US dollars
h1 by flotation was 20% of the cost of centrifugation. These figures did not include the cost of flocculant. Assuming a relative cost of 0.31 for flocculation, we could derive a relative harvesting figure of
settling of algal biomass, flotation refers to the separation process based on the adhesion of air bubbles to flocs of suspended algae after which these float to the surface and separation can occur.
Moheimani
Final product concentration refers to the algal biomass dry weight in the harvested product (e.g. 10% ¼ 10 g dry biomass from 100 g of centrifuged paste). b Capital costs presented, expressed on a 1 m3
harvesting costs, for the sake of comparison the calculation uses a 1 kg m3 cell concentration upon harvesting, assuming a 0.4 and 0.079 g dw L1 cell concentration for ref. 37 and 103 respectively.
(conversion factor 1 AUS $ ¼ 0.81 USD). f The cost of harvesting using flotation was reconsidered in ref. 15, where according to equipment and information available in 1996 the cost of harvesting 1 m3
(2004)103 d,e
1.1 [0.08]
5.2 [0.41]
4.6 [0.36]
3.8 [0.29]
4.1 [0.32]
in a process with centrifugation to increase the solids content of
the harvested biomass.
—
Cost per kg/$ c Another method for harvesting the biomass is filtration. The
vacuum drum filters and chamber filter press units are most
0.61 [0.24]
1.39 [0.56]
1.71 [0.68]
commonly employed for harvesting fairly large microalgae.
(1994)37
Becker
However, because filtration efficiency is highly dependent on the
—
—
size range of the species to be harvested, certain types of filtration
are ineffective for species in the few micrometre size range, such
(2004)94
Molina
as Scenedesmus, Dunaliella and Chlorella due to the rapid clog-

Grima
ging of the filters.97 This species-dependent nature of the filtra-

—
—
—
—
—
—
1
tion process makes this less attractive compared to centrifugation
and flocculation and will not be considered further.
Benemann
(1996)15
An exact economic calculation and comparison of methods for

0.6–0.8
0.4–0.6
Relative costs/m3
0.05–1
harvesting is complex and practically impossible without infor-

—
—
—
1
mation on the species chosen, production system and growth

stage for harvesting. We have summarised the costs published in
(1988)97
Mohn
the literature to date in Table 8. This table contains the best-

0.91f
0.31
0.82
0.26
—
characterised processes and we have compared these on the basis

1
of processing volume, species application, capital and operating

Grima (2004)94
costs. The purpose of the overview table is to provide relative

costs estimates, not absolute comparisons. The cited literature
figures from ref. 97, 15 and 94. The capital costs presented are adjusted for inflation and converted into US dollars (USD, $)
Molina
lacked detailed information on, e.g. the type and energy

requirements of the centrifuges used. Furthermore, species and
—
—
—
—
—
11
0.51. g Current centrifugation engineering allows processing of up to 100 m3 of culture per hour; na – not applicable.
9
developmental information is largely lacking in the cited reports,

Capital cost/1000 $ m3 b
and together with large uncertainties on capital an operating

Benemann
costs it makes exact cost comparisons difficult (if not impossible).

(1996)15
19–105
The harvesting costs are taken into consideration further in

—
—
—
55
—
—
Section 6, where the economic analysis of the complete algal

production process is considered.
(1988)97
Mohn
In summary, most of the literature on the economic consid-

erations of harvesting processes has focused on the need for
—
—
33
—
15
12
4
highly purified products for the food and feed industry and waste
Final product
concentration
water purification. The application to harvesting biomass for

(% solids)a
biofuels will not require the stringent product purity that is the
case for the nutraceutrical industry, which will make the process
12–27
<15
>20
1–3
cheaper. It has been mentioned before97 that there is not

20
7
8
a universally valid (one-size-fits-all) optimum separation process.

However, the harvesting process is a major consideration in the
processed/m3 h1
set-up of the production process. The biomass concentration and

form will affect (and most likely be dictated by) the downstream
159–1900
Volume
15–20g
treatment processes.
2.5–5
0.5–3
na
na
20
5.2 Biomass pre-treatment and extraction of the lipid fraction

Continuous
The harvested biomass (consisting of about 5–20% solids,

process?
depending on the harvesting method used) has to be processed

N
N
N
Y
further to produce the feedstocks (i.e. lipid fraction) for biofuels

Y
production. In most economic analyses the exact harvest,

Flotation (dissolved air flotation)
extraction and process details are not discussed. This is largely

due to the uncertainty surrounding the effectiveness, extraction
yield and the costs associated with the catalytic upgrading of the
Settling/autoflocculation
solid bowl or decanter
lipid fraction.
3-chamber centrifuge
Most of the economic analyses published extrapolate the oil

nozzle centrifuge
extraction procedures from a seed oil extraction process to algal

Centrifugation
Flocculation
biomass. Seeds require very little pretreatment or processing and

the oils (mainly triglycerides) are traditionally extracted either by
pressing the oil out of the biomass or using an apolar solvent to
selectively extract the triglycerides from the biomass. Because of
a
View Online
large differences between algal and seed biomass properties and more, valuable than the biodiesel itself. The process that is
oil composition, specific procedures will have to be developed, utilized in the cost model discussed in section 6 includes anaer-
applied and costed for an algal production process. obic digestion as an option for residual biomass utilization.
Pretreatment, such as drying or cell disruption, of algal Anaerobic digestion is a long established technology in
biomass has been shown to aid lipid extraction.94,104,105 One wastewater management for the disposal of organic waste and
advantage of processing dried biomass is better percolation of the recouping of energy. In essence, in the absence of oxygen,
the solvents in the biomass, leading to an increase in lipid bacteria turn to a variety of alternative proton acceptors to
extraction.104,106,107 However drying is not considered an dispose of the protons the organisms abstract from their organic
economical option for biomass pretreatment for a biofuels substrates. If available, bacteria will use nitrate, sulfate and
production process because of the high-energy requirements of carbon dioxide in that order of preference. If nitrate and sulfate
the drying process. are absent, or at low concentration, carbon dioxide is the main
Alternative pretreatment methods include cell rupture of the proton acceptor and the products are methane and CO2. The
harvested biomass, which could aid in the lipid extraction yield of methane can be high and the energy recovery likewise
process by avoiding or reducing the use of solvents. A range of high, a figure of 70% may be calculated from the data in ref. 15. If
cell rupture techniques, mechanical, chemical and enzymatic, has decomposition can be driven to near completion then the process
been described in the literature as applied to oilseeds and has the potential to recycle the inorganic nitrogen (as ammonia)
algae.104,108–112 However, no such method has been adopted for and phosphate, thus saving costs and the carbon footprint
algae on a large-scale. This is, most likely, because the effec- associated with the Haber–Bosch production of ammonia. The
tiveness of the cell rupture methodology depends heavily on the recycling of nitrogen, of course, could only occur if the protein
physical properties of the algal cell wall and thus will change or and nucleic acid constituents of algal biomass were put to
have to be adapted with particular species used in the culture. digestion. If the protein were sold for animal feed, then there
The composition and strength of the cell wall varies considerably would be very limited recycling of nitrogen, as 80–90% or so of
across species and throughout the algal growth cycle. For the nitrogen resides with these molecules (see Table 2). The
example, Chlorella sp. and Scenedesmus sp. both have robust cell remainder of the nitrogen is principally associated with nucleic
walls with high cellulose content, rendering the cells difficult to acids. This class of compounds have limited nutritional value and
digest or rupture.113 Alternatively, species like Spirulina could usefully be put to anaerobic digestion, were it economic – it
(a cyanobacterium) lack a rigid cell wall and are thus easier to would also result in recycling a major part of the cell’s phos-
disrupt,114 however, cyanobacteria are characteristically low in phorus (see Table 2).
lipid content. Despite its obvious benefits, the economics of anaerobic
digestion are not promising; ref. 15 (p. 141) ‘‘Thus, power
generation (from anaerobic digestion) is not a major profit centre’’.
5.3. Production of biofuels
Our own analysis of recycling the protein and carbohydrate
The lipids obtained by the above processes can then be catalyt- fractions via anaerobic digestion likewise suggests that after
ically converted to biodiesel (methyl esters of the fatty acid allowing for the offsetting of the cost of electricity from the
portion of lipids) or green diesel (deoxygenated fuel derived from energy derived from the methane, there is no net revenue from
fatty acids). Conventional biodiesel production occurs via the anaerobic digestion (see Fig. 24).
transesterification of pre-extracted lipids in the presence of an
alcohol and a catalyst, which can be a strong acid or base to
5.5 Summary
produce fatty acid methyl esters. Green diesel refers to the
production of alkanes through for example catalytic decarbox- From the foregoing discussion it can be seen that there are
ylation of fatty acids. A detailed discussion of the methodology a number of routes available for the separation and processing of
and efficiency of these processes falls outside of the scope of this the algal culture. In many cases their effectiveness is dependent
review and can be found elsewhere.115 It is worth mentioning that upon specific features of the algae, which may vary between
the factors determining the yield and conversion efficiency, such species and with their physiological and growth status. Thus, no
as temperature, time of incubation and catalyst concentration, general strategy can be specified, although some may be elimi-
are highly dependent on the nature and composition of the nated based on the current state of technology (e.g. filtration).
feedstocks. No systematic study has been undertaken to opti- Uncertainties surrounding the optimum strategy for harvest
mise, develop and price the cost of a biodiesel production process and biomass processing would not be a major concern if the
specifically for algal lipids or biomass. A detailed costing scheme separation stage made an insignificant contribution to the overall
was developed for biodiesel production from soybean oil.10 budget. However, this does not appear to be the case – in the
However, there are too many unknowns in the algal biodiesel economic analysis undertaken in the present study (Section 6),
process to be able to provide an estimate for the cost of this harvest and extraction accounted for 15% or so of the operating
process. Furthermore, it is anticipated that this process will be budget (50% of the labour costs has been attributed to these
species, growth- and environmental conditions-dependent. processes). Ref. 15 gives a lower figure for operating costs but
attribute a higher contribution (12%) to capital costs. Pres-
ently, the estimates for the cost of harvesting and separating
5.4. Processing of the residual biomass and co-products
should be seen as best guesses, and without doubt they are major
Critical to the economic success of algal biofuel production is the sources of uncertainty in any economic analysis. Until target
use to which the cell co-products are put, as these can be as, if not organisms have been identified this may have to remain to be the
View Online
case. Furthermore, advances in engineering and improvements in discussion of the analysis of scenarios. A hybrid system, such as
centrifuge technology could also alter the costing. that described in ref. 16, comprising a front end PBR ‘‘nursery
Similar uncertainties surround the fate of the residual biomass. stage’’ and ‘‘grow-on’’ stage in raceways is used as the physical
The high protein content of the residual biomass could provide arrangement as it appears to be the most cost effective option. An
an excellent source of feed. However, if the lipid extraction outline of the costings in the model is given in Table 9. Without
process is solvent-based, prior to considering animal feed as an doubt it can be added to and improved upon.
option, toxicology tests will have to be carried out. The carbo- Recently a comprehensive and thorough review of the
hydrates in the residual biomass could be used for ethanol economics of algal biomass production report has been prepared
fermentation, however, not enough information is present in the for the British Columbian Innovation Council.119 These authors
literature to date regarding the composition of individual sugars worked largely, although not entirely, from fresh estimates. Their
and concentration in the residual biomass as well as large analysis was not available when our spreadsheet was generated.
unknowns about the recalcitrance of microalgal structural We are encouraged to find that their capital costs ($209 500 ha1,
carbohydrates. The potential of the different scenarios on the use converted from Canadian dollars using a Can/US exchange rate
of the residual biomass will be discussed in the following section. of 0.8) are almost identical to our estimate ($209 000 ha1). Their
estimate for operating costs, with the capital repayment costs

removed, was $20 425 per year; our estimate was $26 375;
6. Economics of biomass and biofuel production
however, the similarity in these values is almost certainly fortu-
Widely varying estimates for the cost of algal biodiesel may be itous, since there is considerable variance in the estimates for
found in the recent literature (see Table 10). This, in large part, several major individual items between the two budgets. Their
reflects the immaturity of the subject. The aim of this section is energy costs are about 15% of ours; (deriving in part from the low
not to produce a definitive costing – presently there are too many cost of hydropower in British Columbia) whilst their labour costs
uncertainties to allow this – but give a basis to examine the are around four times greater (indicating that ours may need
pattern of the economics for various management scenarios. To revisiting). There is, however, no significant change to the
this end a set of objective costings is needed. There is only economic pattern seen in Fig. 24 if their labour costs are
a limited amount of information to draw upon to construct substituted into our economic analysis.
a costing algorithm. The DOE Aquatic Species Program In the economic analysis we consider three lipid levels (15, 35
commissioned a report (ref. 116) to produce costings; ref. 15 and 50% lipid as dry weight) and three latitudes: low (0–30 ), mid
incorporated the essence of this in the 1996 Report to the Pitts- (35–45 ) and high (45–55 ) latitude bands using the production
burgh Energy Technology Centre. A detailed costing was rates shown in Fig. 23. We have explored three scenarios
prepared for a pond/raceway type system.117 Although the surrounding the exploitation of the non-lipid fraction: (i) the
capital costs were specified, it is difficult to extract the opera- carbohydrate and protein part of the cell debris is sold as animal
tional costs with any certainty. Ref. 94 gives a costing for feed (60% protein, taken from the 3 : 2 protein to carbohydrate
a tubular bioreactor and ref. 103 costed a raceway system. Some ratio used throughout this work, see legend to Fig. 5); (ii) all of
of the numbers in the report in ref. 103 are recycled from ref. 15 the non-lipid cell debris (plus the glycerol and other water soluble
and 116, so it cannot be taken as a wholly independent analysis – products from the transesterification process) is put to anaerobic
this is also the case for a number of other reports and to some digestion, with the production of methane (used to generate
small extent the present one. energy – in the process the nitrogen and phosphate are recycled);
(iii) the carbohydrate from the algae and the glycerol and other
water soluble products are digested anaerobically when all the
6.1 Parameterising the economic model
phosphate will be recycled, but only the small fraction of
We have based our analysis upon a recent set of costings the nitrogen in nucleic acids and phospholipids, about 10–15% of
prepared for the US Department of Energy’s Office of Energy the total nitrogen (see Table 2). The protein is sold as high grade
Efficiency and Renewable Energy118 for a combined photo- protein-rich material, which could have a market as supplements
bioreactor (PBR)/raceway system; all costings were made de for low protein animal feeds such as corn meal (15% protein).
novo. This has been updated with published costings and from Assigning cash values to these various products (algal lipid,
discussions with colleagues in the field. From this a spreadsheet protein and carbohydrate plus protein mix) has to be a matter of
model has been generated. All the items we have been able to cost speculation as no market exits for them as yet. Animal feeding
are built into the spreadsheet. However, there are a number of trials on algal preparations in general have been underway for
processes we have been unable to cost with any certainty: the some time.120 However, they will need to be extended to look at,
processing of the cell debris to separate the carbohydrate and amongst other things, the possible effects of materials used
protein fractions and the subsequent isolation and processing of during harvesting and processing and it will be some time before
the protein fraction to animal feed. There are additional uncer- a product value will be established. We have approached the
tainties surrounding the harvesting step. Estimating the cost of problem by setting a starting figure from the expected chemical
the necessary CO2 supply is problematic as it will be very much property of the product. The Food and Agriculture Organization
dependent upon the nature of the overall facility, i.e. whether the (FAO) statistics132 for the prices of animal feed show a broad
algal growth system is an adjunct to a CO2 source (i.e. a power correlation with protein content, thus given a protein content we
station) or an isolated facility. Our solution to this problem is, can derive an objective estimate of price. For feed with 60%
rather than to put in a costing that is circumstance-specific, to protein content a value of $750 tonne1 is obtained. An alter-
leave it out of the costing algorithm but to return to it in our native calculation may be made pro rata from the price of soya
View Online
meal (protein content 45%; median commodity value ca. $400

tonne1); this gives a lower figure of $533 tonne1. We have Table 9 Basic structure of the spreadsheet costing model, shown for one
example: a low latitude situation, with a 28 g m2 d1 (103 tonnes ha1 a1)
conservatively used a rounded down value of the lower figure – productivity and a lipid content of 35% of dry weight. In this example, the
$500 tonne1. For the protein-only material the protein/price protein is separated from the carbohydrate and sold at $900 tonne1 as
analysis is not regarded to be reliable as it exceeds the available high-grade animal feed, the carbohydrate and other water soluble
data, a figure of $890 may be calculated pro rata from soya, this products are processed by anaerobic digestion. The biodiesel, as methyl
esters, is sold at $900 tonne1. The costings are intended to be for a scale
has been rounded off to $900 for the purpose of calculation. It’s where they are no longer dependent of the size of the plant (i.e. there is no
important to note that these values are only used as starting further economy of scale), but determined, where appropriate, by the size
values for scenario analysis, so their exact values are not critical. of the culture area and the scale of processing. To convert biodiesel from
mass to volume units, it is assumed to have a specific gravity of 0.88 and
Somewhat more critical is the price we assign to algal biodiesel.
a barrel is taken as 159 litres. All large numbers have been rounded off to
We have drawn a mean value of $916 73 tonne1 from recent 3 significant digits; dw ¼ dry weight
analyses by market specialistskk and have rounded this down to
$900 tonne1 – this approximates to $125/barrel. Whether algal ASPECT OF PRODUCTION FACILITY UNITS
biodiesel will be more valuable than ‘traditional’ biodiesel
Operating Parameters & Assumptions

remains to be determined and will depend on the market when its Areal production 28 g dw m2 d1
fuel properties are better established, as well as social attitudes Annual tonnage produced 103 tonnes dw ha1 a1
towards higher plant biodiesel (see highlighted comment in the Lipid content 35% dw
Protein : carbohydrate ratio 3 : 2 mass
Gallagher report, Section 1). As a starting point we have used the Lipid production 36 tonnes dw ha1 a1
above value, plus two additional values: $550 tonne1 and $700 Bidiesel production (at 70% biolipid 25.2 tonnes dw ha1 a1
tonne1, these are rounded off values for $75/barrel and $100/ production)
barrel. Carbohydrate used for anaerobic 100%
digestion
The required nutrients, nitrogen and phosphorus, are assumed Protein used for anaerobic digestion 0%
to be supplied as a mixture of di-ammonium phosphate and Annual tonnage used for anaerobic 35 tonnes dw ha1 a1
ammonia (costed at $325 tonne1 and $300 tonne1, respectively) digestion
Electricity yield from anaerobic digestion 139 000 MJ ha1 a1
matched to the calculated nitrogen and phosphorous content for
Growth area, as % site area 88%
cells of low (15%), medium (35%) and high (50%) lipid content DAP costs per tonne biomass 0 $ tonne1
(see Table 2). When all the non-lipid fraction is put to anaerobic NH3 costs per tonne biomass 21 $ tonne1
digestion, the nutrient cost is set to zero (this assumes 100% Capital costs
Site prepa 41 800 $ ha1 a1
recovery). In the scenario where the protein fraction is not Culture systema 148 000 $ ha1 a1
digested but sold, the phosphate addition is set to zero and Engineeringa 22 000 $ ha1 a1
ammonia alone is added to satisfy the nitrogen requirement. Harvestb 3570 $ ha1 a1
The model assumes that the cell’s protein : carbohydrate ratio Extractionc 4560 $ ha1 a1
Anaerobic digester @ $225 tonne1 annum1 8730 $ ha1 a1
is 3 : 2 and does not vary with lipid content. This assumption throughputd
needs verification as it has a marked effect on some aspects of the TOTAL CAPITAL COSTS 228 000 $ ha1 a1
costing. Furthermore, the separation of the protein and carbo- Running costs
Laboura 4430 $ ha1 a1
hydrate fraction will have influence on the costing, however, no Growth – power requirement – electricity 7020 $ ha1 a1
data on this separation step was found in the literature. There are @ $0.07/kWha
three losses that need to be taken into consideration: (i) the Harvest – power requirement – electricity 2660 $ ha1 a1
production of extracellular dissolved organic material during @ $0.07/kWha
Dewatering – power requirement – electricity 105 $ ha1 a1
growth; (ii) an allowance for operational downtime; and (iii) @ $0.07/kWha
losses during harvesting. There is a substantial literature Diammonium phosphate (DAP)b 0 $ ha1 a1
surrounding the losses due to the release of organic material (a Ammoniab 2210 $ ha1 a1
Transesterification – methanolc 1650 $ ha1 a1
figure of 5% is used) and downtime is assumed to be 1 week per
Transesterification – power requirement – 166 $ ha1 a1
annum (a 2% loss). Losses on harvesting are presently not natural gasc
established. A round figure of 10% is taken for the combined loss Waste – energy value @ $0.07 kWh1 d 1120 $ ha1 a1
from these three sources, although we suspect that this may be an Anaerobic digester operating costsd 2720 $ ha1 a1
Captial costs @10% per annum 22 800 $ ha1 a1
underestimate. TOTAL COSTS 42 700 $ ha1 a1
The spreadsheet (see Table 9) scales the costings, where Product costs & revenues
appropriate, to the tonnage cell material produced, the tonnage Biomass cost 0.41 $ kg1
of lipid processed and the tonnage put to anaerobic digestion. Biodiesel cost 1.69 $ kg1
Biodiesel cost 237 $ bbl1
The capital and operating costs of anaerobic digestion were Revenue from biodiesel (@ $900 tonne1 22 700 $ ha1 a1
supplied by a UK manufacturer. The figures are not too Revenue from pure protein @ $900 36 200 $ ha1 a1
dissimilar from costings (inflation adjusted) used by ref. 15 and tonne1
Total revenue 58 900 $ ha1 a1
67. Anaerobic digestion is assumed to lose 25% of the calorific Total expenditure 42 700 $ ha1 a1
value of the feedstock during the production of methane, the NET REVENUE 16 200 $ ha1 a1
latter is then converted to electricity assuming 35% efficiency, a
Scaling of parameters: scaled to culture area. b Scaled to biomass
production. c Scaled to lipid processed. d Scaled to tonnage digested.
kk Values for Europe, Asia and USA from ICIS pricing and for Europe
from Kingsman (http://www.kingsman.com).
View Online
Fig. 24 Results of a spreadsheet model of costings and economics for the production of biodiesel and associated product streams. The calculation
assumes biodiesel to have a commodity value of $900 tonne1 (ca. $125 bbl1). The details of the model are given in Table 9 and the text. A. Net revenue.
Lipid levels: blue 15%, green 35%, red 50%. B. The value the separate products, biodiesel at $900 tonne1, animal feed at either $500 tonne1 (as low grade
protein – dark blue bars) or at $900 tonne1 (as high grade protein – pale blue bars). The net revenue calculation for anaerobic digestion involves the
revenue gain from electricity production and the recycling of nitrogen (where it occurs) and phosphate minus the operating and capital costs of the
anaerobic digestion plant. C. The cost of biomass ($ tonne1) and biodiesel ($ bbl1) assuming them to be the sole marketed product.
View Online
Fig. 25 A matrix of the outcome of analyses of economic Scenarios 1 and 3 for the net revenue resulting from various commodity prices for the algal
products.
waste heat from the generators is not put to any specific use in commodity. Whether the alternative of fermenting the carbo-
our analysis, although it certainly would be in the plant design. hydrate and glycerol to ethanol offers better economic rewards
The digester is assumed to enable 100% recycling of the nitrogen would be worth exploring.
and phosphate in the feedstock for use in the culture media. All The analysis makes the point that the co-products, notably
of these assumptions need confirmation. protein, have the potential to be very valuable. Perhaps the most
important finding is that the simple aim to increase lipid yields
using high lipid containing algae may not necessarily be a sound
6.2 Analysis of scenarios
economic strategy if the biochemical and other losses associated
An illustration of the three end-product management scenarios is with lipid and biodiesel production are taken into consideration,
given in Fig. 24A. A general trend of reducing revenue with gains from high lipid production only appear at biodiesel prices
increasing lipid content is apparent. Furthermore, only at a lipid above $125 bbl1.
content of 50% of the cell’s dry weight does the revenue of bio- It is appropriate at this point to recognise that no costing has
diesel exceed that from selling the algal proteins. However at 50% been made for CO2 supply. At liquid CO2 prices of $10 tonne1
lipid levels, the assumption in the productivity model that suffi- the pattern in Fig. 24 remains much the same. At $50 tonne1, the
ciently high growth rates can be sustained and only light limits price determined from a recent study in ref. 121 for carbon
growth, becomes uncertain (see Fig. 6). capture and sequestration, profitable scenarios were only found
A summary of a matrix analysis for Scenarios 1 and 3 is shown at the higher price levels, at $100 tonne1 profitable scenarios
in Fig. 25. Given a value of $125 bbl1 for biodiesel, positive appear only with biodiesel prices of $125 and above.
economic solutions can be found if the value of the protein + There is no doubt that our analysis can be improved upon and
carbohydrate product (Scenario 1) is above $350 tonne1, and for that there are uncertainties associated with many of the numbers
the protein-only scenario (Scenario 3) above values of $600 used in the calculations. Large uncertainties surround the costing
tonne1. The analyses are not particularly sensitive to lowering of of the plant – notably the harvesting and subsequent treatment of
oil prices – an analysis for $350 tonne1 (nominally $50 bbl1) is the biomass and the eventual market values of the various
much the same as the analysis for $550 tonne1. However, a rise biomass components. Uncertainties over the estimated areal
of the same scale to from $100 to $125 bbl1 would make all the production rates will condition the reliability of our economic
scenarios in Fig. 25 profitable on the costings used. analysis. There is active ongoing research and this is one area
Anaerobic digestion (Scenario 2), even allowing for savings where one can anticipate substantial advances in the not too
arising from the recycling of energy and nutrients, is not a net distant future. The present model assumes that light alone limits
source of revenue (Fig. 24B). This was also the earlier conclusion production, the algae are presumed not to enter nutrient limi-
in ref. 15. The costs of anaerobic digestion come from temperate tation, it this respect it may be providing optimistic estimates. On
latitudes, where a considerable proportion of the energy yield has the other side of the coin, the sense exists that there is potential to
to be used to maintain the temperature of the digester; one may improve upon present yields. Potential gains in productivity
expect these circumstances to be very different at low latitudes. could arise from modifications of the cell’s light capturing
This may make the anaerobic digestion of protein, along with apparatus (see Section 4.5). We have chosen not to make
carbohydrate, a more attractive proposition. This would be allowance for these potential improvements. By far the majority
essential if the intention were for biodiesel to be the sole of algal ‘‘farms’’ to date have focused on high value products, so
Table 10 A comparison of published costings for algal biomass and biodiesel production. We have only taken the median value from ref. 122. To convert biodiesel from mass to volume units, it is
assumed to have a specific gravity of 0.88 and a barrel is taken as 159 litres. The estimates have been adjusted to 2007 prices, the currency conversions used are: $ Aus/$ US ¼ 0.81, $ Can/$ US ¼ 0.8 and
V/$ US ¼ 1.28
Biomass costs Biodiesel costs

Assumed productivity Assumed lipid content
$ tonne1 $ tonne1 [$ bbl1]
Source Organism g dw/m2 d1 % dw PBR RW Hybrid PBR RW Hybrid
584 | Energy Environ. Sci., 2010, 3, 554–590

Borowitzka (1992)117 Not specified 20 — — 1 6800 — — — —
Various 3.2–30 — — 350–70 000 — — — —
Mean (+StDev) (14.4 8.2) — — (1 4600 8200) — — — —
Benemann & Oswald (1996)15 Not specified 30 50 — 241 — — 490 [69] —

Not specified 60 50 — 148 — — 300 [42] —
Molina Grima et al. (2004)94 Phaeodactylum Not given Not given 30 400 — — — — —
Moheimani (2005)103 Dunaliella 20 Not given — 6100 — — — —

Pleurochrysis 16–20 Not given — 4700–6900 — — — —
van Harmelen & Oonk (2006)122 Not specified 27 30 — 370 — — 1200 [168] —
Chisti (2007)4 Not specified 72 30 470 — — 1600 [224] — —

Not specified 35 30 — 600 — — 2050 [287] —
Huntley & Redalje (2007)16 Not specified 19–60 40 — — 110–360 — — 280–910 [39–127]
Carlsson et al. (2007)123 Not specified 2.7–14 Not given — 2000–15 000 — — — —
Alabi et al. (2009)119 Not specified 15.3 25 5900 — — — — —

Not specified 9.4 15 — 2140 — — — —
Pienkos & Darzins (2009)95 Not specified 10 15 — — — — 7500 [1050] —

Not specified 25 25 — — — — 2250 [315] —
Not specified 50 50 — — — — 750 [105] —
Present account Not specified 18–37 15–50 — — 360–650 — — 900–3500 [126–490]
This journal is ª The Royal Society of Chemistry 2010

View Online
View Online
the incentive to undertake expensive R & D to increase yields has digestion. Our finding is in agreement with ref. 122, in which it
been low. This investment will need to be made to increase sus- was concluded that the ‘‘fuel only’’ option was not economically
tained algal biomass production rates if we are to use algae for plausible – other revenue sources were needed to bring the system
relatively low value products such as biodiesel and animal feed. into profit. If the first option is adopted, i.e. to market other
biomass products, notably protein as animal feed, in addition to
6.3 Comparison with other economic analyses biolipid, the market for protein will control the overall scale of
production. This has implications, potentially setting a limit to
We have compared the costings published in the literature. Table the scale of production.
10 gives a collation of published costings. At first sight they seem There are broadly four areas of uncertainty in the economic
disturbingly disparate: estimates for biodiesel costs range for analysis. First, we have not been able to obtain costings for
$280–$7500 tonne1 ($39–$1050 bbl1) and biomass from $110– anaerobic digestion at low latitudes. This could extend the
$30 400 tonne1. Ref. 117 contains his own estimate and economic viability. A second major uncertainty relates to
summarises earlier work – the average of these estimates in the the cost of harvesting. This is probably the most complex of the
region $15 000 tonne1. Part of the reason for the high costs in problems, as the properties of the alga used, its nutritional and
the latter study are the conservative assumed yield rates (15 to physiological status, will determine the strategy that can be used
20 g dry weight m2 d1) used in the analysis, but also in Bor- for harvesting. Harvesting costs will have a major, if not over-
owitzka’s study an apparently high operational cost. The esti- riding, effect on strain selection. Estimates of harvesting costs
mate in ref. 103, which again is high, likewise contains high
(capital and operating costs), as a percentage of total costs, vary

values for manpower and harvesting, which seems out of line considerably in different economic analyses. The analysis in ref.
with other costings. Further, the numbers must also be consid- 103 gave a range of 40–50% for the harvest and processing costs
ered in context – the high estimate reported in ref. 94 derives for the calcium carbonate-containing coccolithophorid Pleuro-
from an advanced PBR system, intended for a high value chrysis carterae and 63% for Dunaliella salina. The analysis in ref.
product. The costings in ref. 123 are based on an extremely 15 gave range of 20–25%, and the analysis in ref. 117 analysis
cautious assumption of growth yields (2.7–14 g dry weight m2 implies a similar figure: 15–25% of the capital costs and 20% of
d1). We find it difficult to see how systems can be run in profit the operating costs were associated with harvesting. We derive
with biomass yields below 20 g dry weight m2 d1. The careful a figure in the region of 10% in the present analysis, similar to the
analysis in ref. 119 also stands out as high-cost estimates; estimate of 12% in ref. 119. Much of this variability derives not
however, they made their analysis for the latitude of British from the uncertainty in the costings themselves, over which there
Columbia ($50 N), where biomass yields would around half is reasonable consistency, but in the uncertainty over what
those from latitudes of 0–30 . If we consider the remaining procedures will be used. In particular, whether or not and to
analyses and disregard the high cost, non-viable, versions of the what extent centrifugation, which has high capital and operating
present report and the worst case scenario of the analysis in ref. costs, will be used; i.e. whether it has to be used for primary
95, the spread is still large but reduced to $110–$600 tonne1 for separation (which seems unlikely) or final concentration.
biomass and $280–$2050 tonne1 ($39–$315 bbl1) for biodiesel. Doubling the harvesting costs in our analysis gives rise to
Despite the uncertainties, the broad patterns in the economics a 10–15% fall in revenue of the more profitable scenarios at low
are probably correct, although it would be unrealistic to expect latitudes, but a greater (15–30%) loss at high latitudes.
the exact values to be. A third area of uncertainty deals with the future market prices
Thus, given the most favourable form of culture – a photo- for the various commodities – notably the value and acceptability
bioreactor/raceway hybrid – we can envisage a number of of algal protein as an animal feed and, of course, the market price
scenarios that would appear to have the potential to yield net of biodiesel. If legislation for minimum additions of biodiesel to
profit. fuels persists and the ecological concerns over higher plant fuels
continues or rises, then this may inflate the value and so price of
7 Prospects and research needs algal biodiesel.
In summing up the review we explore the answers to two ques- The fourth major uncertainty lies in the prediction of biomass
tions that arose through the course of the review: yields. As discussed in Section 6, yields have a pronounced effect
(1) Can the production of algal biomass (or biolipid) be made upon the economics. If, for example, the yields are reduced to
profitable? two-thirds from about 30–40 g dry weight m2 d1 (100–140
(2) What R & D is needed to achieve these aims? tonnes ha1 a1) to 20–25 g dry weight m2 d1 (67–93 tonnes ha1
a1) then we find it difficult to produce economic scenarios.
Although predicting yields is a complex matter, we would see in
7.1 Can the production of algal biomass (or biolipid) be made
many respects that producing models of algal production is
profitable?
a more tractable problem than many of the others above and the
The evidence from our analysis is a conditioned yes. First, in our potential exists to produce robust models in the not too distant
analysis in Section 6 we found that, with prices for biodiesel up to future.
$1400 tonne1 ($200/barrel), we are only able to find profitable
outcomes if all the major biochemical products are processed and
7.2 What R & D is needed to achieve these aims?
marketed. We are unable to find economic solutions for the
marketing of biodiesel alone and the associated recycling of the A number of reports have considered the R & D required to
nutrients and energy in the non-lipid fractions by anaerobic produce algal biomass economically, and on an industrial
View Online
scale.3,5,15 The list of R & D needs is very long and there is no be solely attributable to the compensation irradiance, but there
need to detail these at length. The aim here is to give emphasis to may also be a secondary effect of light mediated through
the R & D requirements that this review has indicated to be of temperature. It may set an upper latitude boundary for year-
greatest importance and urgency: optimising yields, advancing round production. In ref. 122 a latitude boundary of 35 was
our understanding of harvesting and processing strategies and suggested, they used a mean annual temperature of 15 C as
resolving the problems of scaling up. a criterion. Such latitudinal boundaries would exclude year-
round production in the EU, Russia and Canada, and
(i) Optimising yields. This may be subdivided into two broad geographically limit it in the US and China. Most of Africa, the
problems – increasing production rates and maximising the Middle East, Central and South America, India, South-East Asia
economics of product yields. and Australia would however be well placed to develop such
In Section 3 we developed the broad basis for the theoretical technologies. There is a pressing need to establish whether or not
photosynthetic yields – a figure in the region of 10% of the energy this apparent irradiance threshold is a general phenomenon and,
in incident irradiation incorporated into organic material is if it is we need to gain a better understanding of its physiological
a widely accepted value. This maximum yield, we point out, is basis.
constrained by stoichiometry and thermodynamics and, for all Many of the previous analyses of R & D needs have pointed to
practical purposes, may be considered to be immovable. Actual the requirement to isolate or develop high lipid producing algae.
yields fall short of this, often being in the region of one third to In relation to this, a pressing matter is to determine whether the
one tenth of the theoretical maximum (see Section 4.3(ii)). As far apparent negative relationship between lipid content and growth
as we can determine, there is no fundamental barrier to achieving rate exists (see Fig. 7 and the associated discussion in Section
the theoretical yields, however finding solutions that are 2.3). This brings to the fore the need to distinguish between high
economically viable present severe practical challenges. The lipid containing algae and high lipid yielding algae. Lipid content
losses are principally due to wastage of the energy of the collected by itself would seem of limited value, if it has to be achieved at
photons at high photon flux rates. This derives mainly from the expense of growth rate. This is particularly the case when
a mismatch between the processing rates of the photochemical products other than lipids are part of the overall economic
and enzymatic reactions. The problem was discussed at length in equation (see Section 6 and Fig. 24).
Section 4.5(i) and its probable resolution lies not in increasing the Following on from the above, as soon as the economics are
rates of the dark reaction, there seems limited capability to do extended beyond the production of lipids and biodiesel, the value
this, but by slowing down the photochemical reaction by and the dynamics of the quantitative distribution of the other
reducing the photon capturing capability of the cell. There has major biochemicals within the cell become a major issue. A
been work done on this and progress made. This will require collation of the reported proportions of the major biochemical
modifying the control of the cell’s light capturing apparatus, by fractions (lipids, carbohydrates and proteins) was analysed in
some intrusive procedure – essentially locking the cell into the Fig. 3 and no pattern was evident. This may well be the case, but
high light adapted state. It is very unlikely to come from the compilation came from very disparate studies and in many
screening of natural species but rather through genetic engi- cases required data conversions and any pattern may have been
neering. In ref. 15 it was noted that the type of strain needed lost or concealed within the resultant noise. Network models for
would be as ‘‘as common as hen’s teeth’’ in Nature, as they would the flows of material during metabolism have been developed for
have little to no competitive capability. A major issue is whether higher plants (see ref. 50 and 125). These could be adapted to
this could be achieved by classical plant breeding techniques – microalgae. However, these models will need dedicated studies
induced mutations (so called ‘‘accelerated evolution’’ techniques) on microalgae in order to calibrate them. Once they are available,
or by direct manipulation of the genome. Some major players in one could use these models as powerful predictive tools.
the field, sensitive to the negative public reaction towards
genetically modified organisms (GMOs), have set their face (ii) Advancing our understanding of strategies for harvesting
against going down the latter route. This, as other issues that and processing of algal biomass. Reviews and reports invariably
surround the use of GMOs, will need to be debated. point to processing, and particularly harvesting, as a major area
In a growing microalgal culture there is a complex of inter- of uncertainty that has major economic implications (as dis-
actions between the organism and the physical and chemical cussed in section 5.1). The present review reaffirms this. The
environments, involving a number of positive and negative point has been made on a number of occasions that centrifuga-
feedbacks (see Section 4.4). Commonly used, simple first order tion, although conceptually and operationally simple, may not be
empirical models based on energetics, such as that employed in an option as a primary form of separation on cost grounds.
Section 4.5(iii), cannot incorporate these feedbacks in any However, advances in centrifugation engineering may improve
explicit way and thus can only be considered as first order the economics. Similarly, the current status of filtration and
solutions. Fundamental physiological models of algal growth micro-straining technology does not allow for a viable option for
exist,124 but do not appear to be extensively used, as yet, in the harvesting algae on a large scale. Again, technological advances
field of mass production of algae. They will be needed to optimise in membrane and systems engineering could change this suppo-
production rates – further they have the potential to serve as an sition. Furthermore, the strong species-dependent nature of the
exploratory tool – to determine the environmental criteria filtration process makes it difficult to draw conclusions about the
necessary for economic production. One matter that needs general applicability of this process. One is left with some form of
urgent consideration is the apparent irradiance threshold (see flocculation, agglutination or co-precipitation in combination
Fig. 22, and text in Section 4.4(ii)). We conclude that this cannot with a centrifugation step to harvest cells. To provide some
View Online
strategic basis we need a much more detailed understanding of a number of special considerations. To make a significant
the surface properties of the cell and how this varies with species, contribution to global fuel production one needs to consider the
physiology and culture conditions. As we start from a position of question of sustainable production approaching the near giga-
very little existing knowledge in the case of microalgae, it will call tonne scale. This, not surprisingly, brings along a host of prob-
for a major research effort, however without some basic under- lems: political, social and economic as well as scientific. There are
standing we will be committed to continue to fumble our way a number of potential ‘‘show stoppers’’ to sustained production
through the problem by trial and error. on this scale, three of them merit early consideration and
Once a concentrated slurry of cells has been produced we have research. Algal mass culture, in addition to land and sunlight,
a number of processing options. Three scenarios, which involved requires, at the minimum, significant quantities of water and
overall various ways of handling the major biochemicals, were nutrients. A first possible ‘‘show stopper’’ is the availability of the
considered in Section 6 (see Fig. 24). These raise major R&D nutrients, nitrogen and phosphorus in particular, required for
questions; the cost, practicality and utility of separating the algal mass culture. Of the three major nutrients, nitrogen might
protein, lipid and carbohydrate fractions from one another, and be argued not to be a problem as, given energy, it can be
the further separation of valuable, more specialised biochemical, generated from atmospheric nitrogen. However even if the
nutraceutical and potential pharmaceutical products. In the energy were derived from within the algal production facility,
analysis in Section 6, carbohydrate was relegated to a low value and thereby carbon neutral, nitrous oxide, a powerful green-
commodity; research is needed to determine whether the bulk house gas, is a by-product of the Haber–Bosch process (used for
carbohydrate fraction or a sub-fraction can be put to better use ammonia synthesis) and its release into the atmosphere will need
than simply recouping a fraction of its energy by anaerobic to be incorporated into any life cycle analysis. In contrast to
digestion. Compositional analysis of the algal biomass will nitrogen, supplies of phosphate would need to come from very
determine whether the carbohydrate fraction can be used to finite reserves. At present rates of exploitation the global reserves
generate a second biofuels stream; e.g. bioethanol after fermen- are estimated to have a lifespan no more than 50–100 years.126
tation of the structural sugars. However, current published One-off use of phosphate on a gigatonne scale of algal lipid
carbohydrate information does not allow predictions of their production would add significantly to the present demand and
potential yield. shorten the lifespan of the global phosphorus reserves. This
Major questions surround the application of anaerobic returns us again to the problem of the recycling of phosphate
digestion to the processing of ‘‘waste’’ algal debris. Whereas we within the overall scheme of production and the successful, or
have extensive knowledge of the application of anaerobic otherwise, use of anaerobic digestion to achieve this.
digestion to human and animal waste and a growing knowledge A second possible ‘‘show stopper’’ is the availability and
of the treatment of domestic waste, by comparison we have delivery of significant quantities of CO2. As discussed in Section
a very limited knowledge of the digestion of microalgal debris. 4.4(iv), in order to achieve rapidly growing dense cultures,
We are, however, not without experience of applying anaerobic carbon dioxide has to be added to the culture in some way, as the
digestion to algal material, as this was undertaken as part of the natural exchange across the atmosphere is too slow. It may be
high oxidation rate ponds used in some of the US west coast possible to achieve this by active aeration relying upon the
sewage treatment plants. How easy it would be to transfer this ambient concentration in air, although this would be costly or it
knowledge to the residual cell material from mass algal culture may be necessary to actively separate the CO2 from the atmo-
needs to be explored. We need insight on a number of matters sphere by liquefaction and add pure CO2 to the culture. Neither
surrounding anaerobic digestion: (i) the speed and completeness is likely to be a low cost solution.
with which nitrogen (presumably as ammonia) and particularly The third potential ‘‘show stopper’’, not surprisingly, is water.
phosphate are released and therefore recycled; (ii) the energetic As soon as one anticipates production plants of considerable
budget and whether or not the net energy yield improves with the scales, local water supplies are unlikely to be adequate and so
increasing ambient temperature at lower latitudes; and (iii) the water (for a number of reasons most probably seawater) will
fraction that remains undigested over some economic time scale need to be pumped onto the culturing site and after use disposed
and whether it has value as a fertiliser or soil conditioner. of in some manner. The engineering and energy costs for
There remains the question whether, rather than separating pumping could be considerable and almost certainly one would
and processing the individual biochemical fractions, there is need to resort to recycling water to some extent. Recycling of the
value in subjecting algal biomass to some form of thermal spent culture medium is embedded in the conceptual scheme in
upgrading for producing biofuels. Work on this was recently ref. 95. Recycling of used water brings along a number of
reviewed in ref. 115; however, a great deal more will be required problems, the spent culture medium will contain soluble organic
before we can establish the practicalities and their relative products released during algal growth; these will accumulate and
economics. may have a positive or negative effect upon the growth rate of
subsequent cultures. Similarly any residual additives, for
(iii) Resolving the problems of scaling up. There are two levels example material added to induce flocculation will be present or
of scaling up: (i) from the laboratory to the pilot plant/modest will have to be removed from the spent medium. Biological
production plant scale, (ii) from the pilot plant scale to a scale contaminants, e.g. competitors and predators, and, perhaps
that would make a significant and sustainable contribution to more problematic, pathogens such as viruses if not removed will
global liquid hydrocarbon fuel production. The first is essentially be recycled. This may call for further treatment e.g. sterilisation
scaling up from gram to tonnage production and is associated of the recycled water with inevitable cost implications. There is
with the R&D needs outlined above. The second brings along experience of these matters from the existing production plants,
View Online
whether this is sufficient to plan large-scale production facilities Biodiesel from Algae, National Renewable Energy Laboratory,
or whether further systematic R & D is needed will require Golden, 1998.
4 Y. Chisti, Biotechnol. Adv., 2007, 25, 294–306.
careful consideration. 5 Q. Hu, M. Sommerfeld, E. Jarvis, M. Ghirardi, M. Posewitz,
In conclusion, the production of biodiesel and other algal M. Seibert and A. Darzins, Plant J., 2008, 54, 621–639.
biomass components on an industrial scale does appear to have 6 A. Richmond, in Handbook of Microalgal Culture, Blackwell, 2004,
the potential to be a sound economic proposition. It would, p. 566.
7 G. R. Cysewski and R. T. Lorenz, in Hangbook of Microalgal
however, be unwise to underestimate the scale of R & D required Culture, Blackwells, 2004, pp. 281–288.
to make this a reality. The diversity of knowledge required is such 8 Y. Chisti, Trends Biotechnol., 2008, 26, 126–131.
that it is unlikely that a single institution or corporation would 9 K. M. Weyer, D. R. Bush, A. Darzins and B. D. Willson, Bioenerg.
Res., 2010, DOI: 10.1007/s12155–009–9046-x.
have the resources to undertake the work exclusively ‘‘in house’’. 10 M. J. Haas, A. J. McAloon, W. C. Yee and T. A. Foglia, Bioresour.
The review National Algal Biofuels Technology Roadmap Technol., 2006, 97, 671–678.
prepared as part of the US Department of Energy Biomass 11 P. G. Falkowski and J. A. Raven, Aquatic Photosynthesis, Princeton
Program makes a strong play for openness in the R & D and the University Press, 2007.
12 C. Van den Hoek, D. G. Mann, and H. M. Jahns., Algae. An
value of collaborative work and exchange of information Introduction to Phycology, Cambridge University Press,
between industry and academic centres. This steer should be Cambridge, 1995.
noted. 13 M. J. Furnas, J. Plankton Res., 1990, 12, 1117–1151.
14 G. Sarthou, K. R. Timmermans, S. Blain and P. Treguer, J. Sea Res.,
It is widely accepted that it will need considerable political will
2005, 53, 25–42.
and social change if we are to find solutions to the problem of 15 J. R. Benemann, W. J. Oswald, Systems and Economic Analysis of
greenhouse gas emissions. The same will and changes in attitude Microalgae Ponds for Conversion of CO2 to Biomass, Final Report,
and approach will be needed if we are to contemplate putting Pittsburgh Energy Technology Center, 1996.
16 M. E. Huntley and D. G. Redalje, Mitigation Adapt. Strategies
algal biomass and biofuel production on a scale where it is Global Change, 2007, 12, 573–608.
a significant part of the overall equation. James Lovelock once 17 P. J. l. B. Williams and J. E. Robertson, J. Plankton Res., 1991, 13,
observed that it characteristically takes at least 25 years for new 153–169.
major technologies to come on line, this broadly matches the 18 S. Ringen, J. Lanum and F. P. Miknis, Fuel, 1979, 58, 69–71.
19 G. C. Dismukes, D. Carrieri, N. Bennette, G. M. Ananyev and
timescale of many of the major political aspirations and promises M. C. Posewitz, Biotechnology, 2008, 19, 235–240.
surrounding the tackling the problem of CO2 emissions. Thus, 20 D. A. McCracken and J. R. Cain, New Phytol., 1981, 88, 67–71.
that the task may seem daunting should not deter us. 21 P. Roessler, J. Phycol., 1987, 23, 494–498.
22 K. Roberts, M. Gurney-Smith and G. J. Hills, J. Ultrastruct. Res.,
1972, 40, 599–613.
23 Y. K. Chau, C. L. Chuecas and J. P. Riley, J. Mar. Biol. Assoc. U. K.,
Acknowledgements 1967, 47, 543–554.
24 D. L opez Alonso, E. H. Belarbi, J. M. Fernandez Sevilla,
We have valued the discussions with many of our colleagues, in J. Rodriguez Ruiz and E. M. Molina Grima, Phytochemistry,
particular P. W. has appreciated inputs and guidance from Ian 2000, 54, 461–471.
25 D. L opez Alonso and F. Garcia Maroto, Biotechnol. Adv., 2000, 18,
Archibald (Shell Global Solutions), John Beardall (Monash
481–497.
University), Ian Joint (Plymouth Marine Laboratory), Phillip 26 S. M. Budge and C. C. Parrish, Phytochemistry, 1999, 52, 561–
Williamson (University of East Anglia), Dylan Evans (Anglesey 566.
Sea Zoo) and David Bowers (School of Ocean Sciences, 27 J. P. Fidalgo, A. Cid, E. Torres, A. Sukenik and C. Herrero,
Aquaculture, 1998, 166, 105–116.
University of Bangor) and L. L. has appreciated input from 28 Z. Gombos and N. Murata, Plant CellPhysiol., 1991, 32, 73–77.
Marti~ na Ferreira-Novio (Universidade de Santiago de Com- 29 F. J. L. Gordillo, M. Goutx, F. L. Figueroa and F. X. Niell, J. Appl.
postela, Spain) in Section 5.1. Al Darzins and Andy Aden Phycol., 1998, 10, 135–144.
(NREL) are gratefully acknowledged for their comments on the 30 P. A. Hodgson, R. J. Henderson, J. R. Sargent and J. W. Leftley,
J. Appl. Phycol., 1991, 3, 169–181.
manuscript. We thank Jasvir Atwal, Account Manager of ICIS- 31 M. Kates and B. E. Volcani, Biochim. Biophys. Acta, 1966, 116, 264–
ICIS Pricing & ICIS News for providing a copy of their publi- 278.
cation Market. P. W. also thanks J€ org Schwender of the Biology 32 M. G. Kleinschmidt and V. A. McMahon, Plant Physiol., 1970, 46,
286–289.
Department of the Brookhaven National Laboratory for help 33 A. T. Lombardi and P. J. Wangersky, Hydrobiologia, 1995, 306, 1–6.
with the calculations of the conversion efficiency of protein given 34 F. Pernet, R. Tremblay, E. Demers and M. Roussy, Aquaculture,
in Box 1. We thank Carolyn Walburton for help with preparing 2003, 221, 393–406.
the images and Alice Bowers for digitising the Pedroni et al. data 35 H. Tatsuzawa, E. Takizawa, M. Wada and Y. Yamamoto,
J. Phycol., 1996, 32, 598–601.
used in Fig. 20 and 21. Finally, much of the data present in the 36 T. W. Ryan, L. G. Dodge and T. J. Callahan, J. Am. Oil Chem. Soc.,
Appendices and P. W.’s understanding of the subject are an 1984, 61, 1610–1619.
outcome of contracts from Shell and the company is thanked for 37 E. W. Becker, Microalgae – Biotechnology and Microbiology,
Cambridge University Press, Cambridge, 1994.
their support and for their willingness to allow this material into
38 M. R. Brown, J. Exp. Mar. Biol. Ecol., 1991, 145, 79–99.
the public domain. 39 M. R. Brown, G. A. Dunstan, S. J. Norwood and K. A. Miller,
J. Phycol., 1996, 32, 64–73.
40 T. R. Larson and T. A. V. Rees, J. Phycol., 1996, 32, 388–393.
References 41 S. O. Lourenço, E. Barbarino, J. Mancini-Filho, K. P. Schinke and
E. Aidar, Phycologia, 2002, 41, 158–168.
1 P. J. le B. Williams, Nature, 2007, 450, 478. 42 P. A. Thompson, P. J. Harrison and J. N. C. Whyte, J. Phycol., 1990,
2 N. S. Shifrin and S. W. Chisholm, J. Phycol., 1981, 17, 374–384. 26, 278–288.
3 J. Sheehan, T. Dunahay, J. R. Benemann and P. Roessler, A Look 43 P. G. Roessler, J. Phycol., 1990, 26, 393–399.
Back at the U.S Department of Energy’s Aquatic Species Program- 44 H. Xu, X. Miao and Q. Wu, J. Technol., 2006, 126, 499–507.
View Online
45 N. Nagle and P. R. Lemke, Appl. Biochem. Biotechnol., 1990, 24–25, 81 E. Molina Grima, F. G. Acien Fernandez, F. Garcia Camacho,
355–361. F. Camacho Rubio and Y. Chisti, J. Appl. Phycol., 2000, 12, 365–
46 C. K. Mathews and K. E. Van Holde, Biochemistry, Benjamin/ 367.
Cummings Publishing Company, Redwood City, 1990. 82 E. A. Laws, K. L. Terry, J. Wickman and M. S. Chalup, Biotechnol.
47 J. T. O. Kirk, Light and Photosynthesis in Aquatic Ecosystems, Bioeng., 1983, 25, 2319–2335.
Cambridge University Press, 1994. 83 E. A. Laws, S. Taguchi, J. Hirata and L. Pang, Biotechnol. Bioeng.,
48 W. Lubitz, E. J. Reijerse and J. Messinger, Energy Environ. Sci., 1986, 28, 191–197.
2008, 1, 15–31. 84 J. C. Merchuk and X. Wu, J. Appl. Phycol., 2003, 15, 163–169.
49 G. G. B. Tcherkez, G. D. Farquhar and T. J. Andrews, Proc. Natl. 85 A. Melis, J. Neidhardt and J. R. Benemann, J. Appl. Phycol., 1999,
Acad. Sci. U. S. A., 2006, 103, 7246–7251. 10, 515–529.
50 D. K. Allen, I. G. L. Libourel and Y. Shachar-Hill, Plant, Cell 86 J. H. Mussgnug, S. T. Hall, J. Rupprecht, A. Foo, V. Klassen,
Environ., 2009, 32, 1241–1257. A. McDowall, P. Schenk, O. Kruse and B. Hankamer, J. Plant
51 J. A. Raven and J. Beardall, in Respiration in Aquatic Ecosystems, ed. Technol., 2007, 5, 802–814.
P. A. d. Giorgio and P. J. le B. Williams, Oxford University Press, 87 Y. Nakajima and R. Ueda, J. Appl. Phycol., 1997, 9, 503–510.
2005. 88 J. E. W. Polle, J. R. Benemann, A. Tankaka and A. Melis, Planta,
52 C. Langdon, J. Plankton Res., 1987, 9, 459–482. 2000, 211, 335–344.
53 C. Langdon, Mar. Sci. Symp., 1993, 197, 69–78. 89 J. E. W. Polle, S. Kanakagiri, E. Jin, T. Masuda and A. Melis, Int.
54 X. G. Zhu, S. P. Long and D. R. Ort, Curr. Opin. Biotechnol., 2008, J. Hydrogen Energy, 2002, 27, 1257–1264.
19, 153–159. 90 J. E. W. Polle, S. Kanakagiri and A. Melis, Planta, 2003, 217, 49–59.
55 K. Richardson, J. Beardall and J. A. Raven, New Phytol., 1983, 93, 91 S. D. Tetal, M. Mitra and A. Melis, Planta, 2007, 225, 813–829.
157–191. 92 M. Olaizola, J. Appl. Phycol., 2000, 12, 499–506.
56 J. Pirt, New Phytol., 1986, 102, 3–37. 93 J. R. Benemann, Biofixation of CO2 and Greenhouse Gas Abatement
57 E. Molina Grima, J. M. Fernandez Sevilla, J. A. Sanchez Perez and with Microalgae – Technology Roadmap, National Energy
P. Garcia Camacho, J. Biotechnol., 1996, 45, 59–69. Technology Laboratory, 2003.
58 M. Tredici, in Handbook of Microalgal Culture, ed. A. Richmond, 94 E. Molina Grima, F. G. Acien Fernandez and A. Rogbies Medina, in
Blackwell, 2004, pp. 178–214. Handbook of Microalgal Culture, ed. A. Richmond, Blackwells,
59 J. Huisman and F. J. Weissing, Ecology, 1994, 75, 507–520. 2004, pp. 215–251.
60 G. Torzillo, T. Goksan, C. Faraloni, J. Kopecky and J. Masojidek, 95 P. T. Pienkos and A. Darzins, Biofuels, Bioprod. Biorefin., 2009, 3,
J. Appl. Phycol., 2003, 15, 127–136. 431–440.
61 NASA, http://edmall.gsfc.nasa.gov/inv99Project.Site/Pages/science- 96 F. H. Mohn, Arch. Hydrobiol., 1978, 1, 228–253.
briefs/ed-stickler/ed-irradiance.html, 2009. 97 F. H. Mohn, in Algal Biomass, ed. G. Shelef, Soeder, C. J., Elsevier,
62 NASA, http://eosweb.larc.nasa.gov/sse/, Atmospheric Science Data Amsterdam, 1980, pp. 547–471.
Center, NASA Surface meteorology and Solar Energy: Global/ 98 L. Svarovsky, in Solid-Liquid Separation, ed. L. Svarovsky,
Regional Data, 2009. Butterworth-Heinemann, Oxford, 2001, pp. 1–29.
63 E. S. Domalski, J. Phys. Ref. Data, 1972, 1, 221–277. 99 L. M. Lubian, Aquacult. Eng., 1989, 8, 257–265.
64 M. S. Kharasch, Bur. Stand. J. Res., 1929, 2, 359. 100 J. Morales, J. Delanoue and G. Picard, Aquacult. Eng., 1985, 4, 257–
65 J. C. Goldman, Water Res., 1979, 13, 119–136. 270.
66 G. Fischer, H. van Velthuizen, M. Shah, F. Nachtergaele, Global 101 N. Levy, S. Magdassi and Y. Baror, Water Res., 1992, 26, 249–254.
Agro-Ecological Assessment for Agriculture in the 21st Century: 102 H. Salehizadeh and S. A. Shojaosadati, Biotechnol. Adv., 2001, 19,
Methodology and Results, International Institute for Applied 371–385.
Systems Analysis (IIASA), Vienna, 2002. 103 N. R. Moheimani, Ph D Thesis, Murdoch University, Australia,
67 D. L. Klass, Biomass for Renewable Energy, Fuels, and Chemicals, 2005.
Elsvier Incorporation, 1998. 104 S. J. Lee, B. D. Yoon and H. M. Oh, Biotechnol. Tech., 1998, 12,
68 I. C. Macedo, J. E. A. Seabra and J. E. A. R. Silvac, Biomass 553–556.
Bioenergy, 2008, 32, 582–595. 105 M. M. Mendes-Pinto, M. F. J. Raposo, J. Bowen, A. J. Young and
69 H. S. Kheshgi, R. C. Prince and G. Marland, Annu. Rev. Energy R. Morais, J. Appl. Phycol., 2001, 13, 19–24.
Environ., 2000, 25, 199–244. 106 G. Ahlgren and L. Merino, Arch. Hydrobiol., 1991, 121, 295–306.
70 F. G. Acien Fern andez, F. Garcia Camacho, J. A. Sanchez Perez, 107 M. L. Eltgroth, R. L. Watwood and G. V. Wolfe, J. Phycol., 2005,
J. M. Fernandez Sevilla and E. Molina-Grima, Biotechnol. Bioeng., 41, 1000–1009.
1998, 58, 605–616. 108 J. Darbyshire, Top. Enzyme Ferment. Biotechnol., 1981, 5, 147–186.
71 N. R. Moheimani and M. A. Borowitzka, J. Appl. Phycol., 2006, 18, 109 J. Fleurence, J. Appl. Phycol., 1999, 11, 313–314.
703–712. 110 G. E. Napolitano, J. Phycol., 1994, 30, 943–950.
72 P. M. Pedroni, G. Lamenti, G. Prosperi, L. Ritorto, G. Scolla, 111 A. Rosenthal, D. L. Pyle and K. Niranjan, Enzyme Microb. Technol.,
F. Capuano, M. Valdiserri, in Greenhouse Gas Control 1996, 19, 402–420.
Technologies 7. Proceedings of the 7th International Conference on 112 K. H. Wiltshire, M. Boersma, A. Moeller and H. Buhtz, Aquat.
Greenhouse Gas Control Technologies 5– September 2004, Ecol., 2000, 34, 119–126.
Vancouver, Canada, ed. E. S. Rubin, Keith, D. W., Gilboy, C. F., 113 D. S. Domozych, K. D. Stewart and K. R. Mattox, J. Mol. Evol.,
Wilson, T., Morris, T., Gale, J. Thambimuthu, K., Elsevier 1980, 15, 1–12.
Science Ltd, Oxford, 2005, pp. 1037–1042. 114 T. Ogawa and G. Terui, J. Ferment. Technol., 1970, 48, 361–367.
73 N. R. Moheimani and M. A. Borowitzka, Biotechnol. Bioeng., 2007, 115 B. Smith, H. C. Greenwell and A. Whiting, Energy Environ. Sci.,
96, 27–36. 2009, 2, 262–271.
74 W. J. Oswald, in Micro-Algal Biotechnology, ed. M. A. Borowitzka 116 J. C. Weissman and R. P. Goebel, Design and Analysis of Microalgal
and L. J. Borowitzka, Cambridge University Press, Cambridge, Open Pond Systems for the Purpose of Producing Fuels, Solar Energy
1988, pp. 357–394. Research Institution, Golden, 1987.
75 M. R. Tredici and G. C. Zittelli, Biotechnol. Bioeng., 1998, 57, 187– 117 M. A. Borowitzka, J. Appl. Phycol., 1992, 4, 267–279.
197. 118 C. B. Raleigh, M. E. Huntley, Subtask 3B-2/3: Feedstock Processing
76 G. C. Zittelli, V. Tomasello, E. Pinzani and M. R. Tredici, J. Appl. and Microbial Biomass Assessment. DE-FC36–04GO14248, Hawaii
Phycol., 1996, 8, 293–301. Natural Energy Institute, 2008.
77 R. Emerson and W. Arnold, J. Gen. Physiol., 1932, 15, 391–420. 119 A. O. Alabi, M. Tampier, E. Bibeau, Microalgae Technologies and
78 R. Radmer and B. Kok, BioScience, 1977, 27, 599–605. Processes for Biofuels/Bioenergy Production in British Columbia:
79 B. Kok, in Algal Culture from Laboratory to Pilot Plant, ed. J. S. Current Technology, Suitability and Barriers to Implementation:
Burlew, Carnegie Institution of Washington Publication No. 600, Final Report, British Columbia Innovation Council, Victoria,
Washington D.C, 1953, pp. 63–84. British Columbia, 2009.
80 J. U. Grobbelaar, L. Nedbal and V. Tichy, J. Appl. Phycol., 1996, 8, 120 A. P. Singh, C. A. Avramis, J. K. G. Kramerand and
335–343. A. G. Marangoni, J. Dairy Res., 2004, 71, 66–73.
View Online
121 S. Paltsev, J. M. Reilly, H. D. Jacoby, A. C. Gurgel, G. E. Metcalf, 127 H. U. Sverdrup, J. Conseil Int. Exploration Mer, 1953, 18, 287–295.
A. P. Sokolov and J. F. Hola, Clim. Policy, 2008, 8, 395–420. 128 T. R. Parsons and M. Takahashi, Limnol. Oceanogr., 1973, 18, 511–
122 T. van Harmelen, H. Oonk, Microalgae Biofixation Processes: 515.
Application and Potential Contributions to Greenhouse Gas 129 E. Gallagher, The Gallagher Review of the Indirect Effects of Biofuels
Mitigation, TNO, Appeldoorn, 2006. Production, 2008, Renewable Fuels Agency.
123 A. S. Carlsson, J. B. Van Beilen, R. Moeller, D. Clayton, Micro-and 130 M. A. Borowitzka in Micro-Algal Biotechnology, ed. M. A.
Macro Algae: Utility for Industrial Applications, CPL Press, Borowitzka and L. J. Borowitzka, Cambridge University Press,
Newbury, 2007. Cambridge, 1988, pp. 257–287.
124 K. J. Flynn, J. Plankton Res., 2001, 23, 977–997. 131 L. Rodolfi, G. C. Graziella, N. Bassi, G. Padovani, N. Bionini,
125 J. Schwender, Curr. Opin. Biotechnol., 2008, 19, 131–137. G. Bionini and M. R. Tredici, Biotechnol. Bioeng., 2009, 102, 100–
126 D. Cordell, J.-O. Dangert and S. White, Global Environ. Change, 112.
2009, 19, 292–305. 132 www.fao.org/es/esc/prices.
Biotechnology Advances 25 (2007) 294 – 306
www.elsevier.com/locate/biotechadv
Research review paper

Biodiesel from microalgae
Yusuf Chisti ⁎
Institute of Technology and Engineering, Massey University, Private Bag 11 222, Palmerston North, New Zealand
Available online 13 February 2007
Abstract
Continued use of petroleum sourced fuels is now widely recognized as unsustainable because of depleting supplies and the
contribution of these fuels to the accumulation of carbon dioxide in the environment. Renewable, carbon neutral, transport fuels are
necessary for environmental and economic sustainability. Biodiesel derived from oil crops is a potential renewable and carbon
neutral alternative to petroleum fuels. Unfortunately, biodiesel from oil crops, waste cooking oil and animal fat cannot realistically
satisfy even a small fraction of the existing demand for transport fuels. As demonstrated here, microalgae appear to be the only
source of renewable biodiesel that is capable of meeting the global demand for transport fuels. Like plants, microalgae use sunlight
to produce oils but they do so more efficiently than crop plants. Oil productivity of many microalgae greatly exceeds the oil
productivity of the best producing oil crops. Approaches for making microalgal biodiesel economically competitive with
petrodiesel are discussed.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Biofuels; Biodiesel; Microalgae; Photobioreactors; Raceway ponds
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
2. Potential of microalgal biodiesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
3. Microalgal biomass production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
3.1. Raceway ponds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
3.2. Photobioreactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
4. Comparison of raceways and tubular photobioreactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
5. Acceptability of microalgal biodiesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
6. Economics of biodiesel production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
7. Improving economics of microalgal biodiesel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
7.1. Biorefinery based production strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
7.2. Enhancing algal biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
7.3. Photobioreactor engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
⁎ Tel.: +64 6 350 5934; fax: +64 6 350 5604.

E-mail address: Y.Chisti@massey.ac.nz.
0734-9750/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.biotechadv.2007.02.001
Y. Chisti / Biotechnology Advances 25 (2007) 294–306 295
1. Introduction Box 1
Microalgae are sunlight-driven cell factories that Biodiesel production

convert carbon dioxide to potential biofuels, foods, Parent oil used in making biodiesel consists of
feeds and high-value bioactives (Metting and Pyne, triglycerides (Fig. B1) in which three fatty acid
1986; Schwartz, 1990; Kay, 1991; Shimizu, 1996, molecules are esterified with a molecule of glycerol.
2003; Borowitzka, 1999; Ghirardi et al., 2000; Akker- In making biodiesel, triglycerides are reacted with
man et al., 2002; Banerjee et al., 2002; Melis, 2002; methanol in a reaction known as transesterification or
alcoholysis. Transestrification produces methyl esters
Lorenz and Cysewski, 2003; Metzger and Largeau,
of fatty acids, that are biodiesel, and glycerol (Fig. B1).
2005; Singh et al., 2005; Spolaore et al., 2006; Walter The reaction occurs stepwise: triglycerides are first
et al., 2005). In addition, these photosynthetic micro- converted to diglycerides, then to monoglycerides and
organisms are useful in bioremediation applications finally to glycerol.
(Mallick, 2002; Suresh and Ravishankar, 2004; Kalin
et al., 2005; Munoz and Guieysse, 2006) and as
nitrogen fixing biofertilizers Vaishampayan et al.,
2001). This article focuses on microalgae as a potential
source of biodiesel.
Microalgae can provide several different types of
renewable biofuels. These include methane produced by
anaerobic digestion of the algal biomass (Spolaore et al., Fig. B1. Transesterification of oil to biodiesel. R1–3 are
hydrocarbon groups.
2006); biodiesel derived from microalgal oil (Roessler
et al., 1994; Sawayama et al., 1995; Dunahay et al., 1996;
Sheehan et al., 1998; Banerjee et al., 2002; Gavrilescu Transesterification requires 3 mol of alcohol for each
mole of triglyceride to produce 1 mol of glycerol and
and Chisti, 2005); and photobiologically produced
3 mol of methyl esters (Fig. B1). The reaction is an
biohydrogen (Ghirardi et al., 2000; Akkerman et al.,
equilibrium. Industrial processes use 6 mol of methanol
2002; Melis, 2002; Fedorov et al., 2005; Kapdan and for each mole of triglyceride (Fukuda et al., 2001). This
Kargi, 2006). The idea of using microalgae as a source of large excess of methanol ensures that the reaction is
fuel is not new (Chisti, 1980–81; Nagle and Lemke, driven in the direction of methyl esters, i.e. towards
1990; Sawayama et al., 1995), but it is now being taken biodiesel. Yield of methyl esters exceeds 98% on a
seriously because of the escalating price of petroleum weight basis (Fukuda et al., 2001).
and, more significantly, the emerging concern about Transesterification is catalyzed by acids, alkalis
global warming that is associated with burning fossil (Fukuda et al., 2001; Meher et al., 2006) and lipase
fuels (Gavrilescu and Chisti, 2005). enzymes (Sharma et al., 2001). Alkali-catalyzed
transesterification is about 4000 times faster than
Biodiesel is produced currently from plant and
the acid catalyzed reaction (Fukuda et al., 2001).
animal oils, but not from microalgae. This is likely to
Consequently, alkalis such as sodium and potassium
change as several companies are attempting to com- hydroxide are commonly used as commercial catalysts
mercialize microalgal biodiesel. Biodiesel is a proven at a concentration of about 1% by weight of oil.
fuel. Technology for producing and using biodiesel has Alkoxides such as sodium methoxide are even better
been known for more than 50 years (Knothe et al., 1997; catalysts than sodium hydroxide and are being increas-
Fukuda et al., 2001; Barnwal and Sharma, 2005; ingly used. Use of lipases offers important advantages,
Demirbas, 2005; Van Gerpen, 2005; Felizardo et al., but is not currently feasible because of the relatively
2006; Kulkarni and Dalai, 2006; Meher et al., 2006). In high cost of the catalyst (Fukuda et al., 2001). Alkali-
the United States, biodiesel is produced mainly from catalyzed transesterification is carried out at approxi-
soybeans. Other sources of commercial biodiesel mately 60 °C under atmospheric pressure, as methanol
boils off at 65 °C at atmospheric pressure. Under these
include canola oil, animal fat, palm oil, corn oil, waste
conditions, reaction takes about 90 min to complete. A
cooking oil (Felizardo et al., 2006; Kulkarni and Dalai,
higher temperature can be used in combination with
2006), and jatropha oil (Barnwal and Sharma, 2005). higher pressure, but this is expensive. Methanol and oil
The typically used process for commercial production of do not mix, hence the reaction mixture contains two
biodiesel is explained in Box 1. Any future production liquid phases. Other alcohols can be used, but
of biodiesel from microalgae is expected to use the same methanol is the least expensive. To prevent yield loss
process. Production of methyl esters, or biodiesel, from
microalgal oil has been demonstrated (Belarbi et al., (continued on next page)
296 Y. Chisti / Biotechnology Advances 25 (2007) 294–306
Box 1 (continued) Table 2

Oil content of some microalgae
due to saponification reactions (i.e. soap formation),
Microalga Oil content (% dry wt)
the oil and alcohol must be dry and the oil should have a
minimum of free fatty acids. Biodiesel is recovered by Botryococcus braunii 25–75
repeated washing with water to remove glycerol and Chlorella sp. 28–32
Crypthecodinium cohnii 20
methanol.
Cylindrotheca sp. 16–37
Dunaliella primolecta 23
Isochrysis sp. 25–33
2000) although the product was intended for pharma- Monallanthus salina N20
ceutical use. Nannochloris sp. 20–35
Nannochloropsis sp. 31–68
Neochloris oleoabundans 35–54
2. Potential of microalgal biodiesel
Nitzschia sp. 45–47
Phaeodactylum tricornutum 20–30
Replacing all the transport fuel consumed in the Schizochytrium sp. 50–77
United States with biodiesel will require 0.53 billion m3 Tetraselmis sueica 15–23
of biodiesel annually at the current rate of consumption.
Oil crops, waste cooking oil and animal fat cannot
realistically satisfy this demand. For example, meeting demonstrated biomass productivity in photobioreactors,
only half the existing U.S. transport fuel needs by as discussed later in this article. Actual biodiesel yield
biodiesel, would require unsustainably large cultivation per hectare is about 80% of the yield of the parent crop
areas for major oil crops. This is demonstrated in oil given in Table 1.
Table 1. Using the average oil yield per hectare from In view of Table 1, microalgae appear to be the only
various crops, the cropping area needed to meet 50% of source of biodiesel that has the potential to completely
the U.S. transport fuel needs is calculated in column 3 displace fossil diesel. Unlike other oil crops, microalgae
(Table 1). In column 4 (Table 1) this area is expressed as grow extremely rapidly and many are exceedingly rich in
a percentage of the total cropping area of the United oil. Microalgae commonly double their biomass within
States. If oil palm, a high-yielding oil crop can be 24 h. Biomass doubling times during exponential growth
grown, 24% of the total cropland will need to be devoted are commonly as short as 3.5 h. Oil content in microalgae
to its cultivation to meet only 50% of the transport fuel can exceed 80% by weight of dry biomass (Metting,
needs. Clearly, oil crops cannot significantly contribute 1996; Spolaore et al., 2006). Oil levels of 20–50% are
to replacing petroleum derived liquid fuels in the quite common (Table 2). Oil productivity, that is the
foreseeable future. This scenario changes dramatically, mass of oil produced per unit volume of the microalgal
if microalgae are used to produce biodiesel. Between 1 broth per day, depends on the algal growth rate and the
and 3% of the total U.S. cropping area would be oil content of the biomass. Microalgae with high oil
sufficient for producing algal biomass that satisfies 50% productivities are desired for producing biodiesel.
of the transport fuel needs (Table 1). The microalgal oil Depending on species, microalgae produce many
yields given in Table 1 are based on experimentally different kinds of lipids, hydrocarbons and other
complex oils (Banerjee et al., 2002; Metzger and
Table 1 Largeau, 2005; Guschina and Harwood, 2006). Not all
Comparison of some sources of biodiesel algal oils are satisfactory for making biodiesel, but
Crop Oil yield Land area Percent of existing suitable oils occur commonly. Using microalgae to
(L/ha) needed (M ha) a US cropping area a produce biodiesel will not compromise production of
Corn 172 1540 846 food, fodder and other products derived from crops.
Soybean 446 594 326 Potentially, instead of microalgae, oil producing
Canola 1190 223 122 heterotrophic microorganisms (Ratledge, 1993; Ratledge
Jatropha 1892 140 77 and Wynn, 2002) grown on a natural organic carbon
Coconut 2689 99 54
source such as sugar, can be used to make biodiesel;
Oil palm 5950 45 24
Microalgae b 136,900 2 1.1 however, heterotrophic production is not as efficient as
Microalgae c 58,700 4.5 2.5 using photosynthetic microalgae. This is because the
a
For meeting 50% of all transport fuel needs of the United States. renewable organic carbon sources required for growing
b
70% oil (by wt) in biomass. heterotrophic microorganisms are produced ultimately by
c
30% oil (by wt) in biomass. photosynthesis, usually in crop plants.
Production of algal oils requires an ability to

inexpensively produce large quantities of oil-rich
microalgal biomass.
3. Microalgal biomass production
Producing microalgal biomass is generally more

expensive than growing crops. Photosynthetic growth
requires light, carbon dioxide, water and inorganic salts.
Temperature must remain generally within 20 to 30 °C.
To minimize expense, biodiesel production must rely on
freely available sunlight, despite daily and seasonal
variations in light levels.
Growth medium must provide the inorganic elements
that constitute the algal cell. Essential elements include
Fig. 1. Arial view of a raceway pond.
nitrogen (N), phosphorus (P), iron and in some cases
silicon. Minimal nutritional requirements can be
estimated using the approximate molecular formula of a constant rate and the same quantity of microalgal
the microalgal biomass, that is CO0.48H1.83N0.11P0.01. broth is withdrawn continuously (Molina Grima et al.,
This formula is based on data presented by Grobbelaar 1999). Feeding ceases during the night, but the mixing
(2004). Nutrients such as phosphorus must be supplied of broth must continue to prevent settling of the bio-
in significant excess because the phosphates added mass (Molina Grima et al., 1999). As much as 25% of
complex with metal ions, therefore, not all the added P is the biomass produced during daylight, may be lost
bioavailable. Sea water supplemented with commercial during the night because of respiration. The extent of
nitrate and phosphate fertilizers and a few other this loss depends on the light level under which the
micronutrients is commonly used for growing marine biomass was grown, the growth temperature, and the
microalgae (Molina Grima et al., 1999). Growth media temperature at night.
are generally inexpensive. The only practicable methods of large-scale produc-
Microalgal biomass contains approximately 50% tion of microalgae are raceway ponds (Terry and
carbon by dry weight (Sánchez Mirón et al., 2003). Raymond, 1985; Molina Grima, 1999) and tubular
All of this carbon is typically derived from carbon photobioreactors (Molina Grima et al., 1999; Tredici,
dioxide. Producing 100 t of algal biomass fixes roughly 1999; Sánchez Mirón et al., 1999), as discussed next.
183 t of carbon dioxide. Carbon dioxide must be fed
continually during daylight hours. Feeding controlled in 3.1. Raceway ponds
response to signals from pH sensors minimizes loss of
carbon dioxide and pH variations. Biodiesel production A raceway pond is made of a closed loop
can potentially use some of the carbon dioxide that recirculation channel that is typically about 0.3 m deep
is released in power plants by burning fossil fuels (Fig. 1). Mixing and circulation are produced by a
(Sawayama et al., 1995; Yun et al., 1997). This carbon paddlewheel (Fig. 1). Flow is guided around bends by
dioxide is often available at little or no cost. baffles placed in the flow channel. Raceway channels
Ideally, microalgal biodiesel would be carbon neutral, are built in concrete, or compacted earth, and may be
as all the power needed for producing and processing the lined with white plastic. During daylight, the culture is
algae would come from biodiesel itself and from fed continuously in front of the paddlewheel where
methane produced by anaerobic digestion of biomass the flow begins (Fig. 1). Broth is harvested behind
residue left behind after the oils has been extracted. the paddlewheel, on completion of the circulation loop.
Although microalgal biodiesel can be carbon neutral, it The paddlewheel operates all the time to prevent
will not result in any net reduction in carbon dioxide that sedimentation.
is accumulating as a consequence of burning of fossil Raceway ponds for mass culture of microalgae have
fuels. been used since the 1950s. Extensive experience exists
Large-scale production of microalgal biomass on operation and engineering of raceways. The largest
generally uses continuous culture during daylight. In raceway-based biomass production facility occupies an
this method of operation, fresh culture medium is fed at area of 440,000 m2 (Spolaore et al., 2006). This facility,
owned by Earthrise Nutritionals (www.earthrise.com),

is used to produce cyanobacterial biomass for food.
In raceways, any cooling is achieved only by
evaporation. Temperature fluctuates within a diurnal
cycle and seasonally. Evaporative water loss can be
significant. Because of significant losses to atmosphere,
raceways use carbon dioxide much less efficiently than
photobioreactors. Productivity is affected by contami-
nation with unwanted algae and microorganisms that
feed on algae. The biomass concentration remains low
because raceways are poorly mixed and cannot sustain
an optically dark zone. Raceway ponds and other open
culture systems for producing microalgae are further
discussed by Terry and Raymond (1985).
Production of microalgal biomass for making biodie-
sel has been extensively evaluated in raceway ponds in
studies sponsored by the United States Department of Fig. 3. A fence-like solar collector.
Energy (Sheehan et al., 1998). Raceways are perceived
to be less expensive than photobioreactors, because they and back to the reservoir. Continuous culture operation is
cost less to build and operate. Although raceways are used, as explained above.
low-cost, they have a low biomass productivity com- The solar collector is oriented to maximize sunlight
pared with photobioreactors. capture (Molina Grima et al., 1999; Sánchez Mirón et al.,
1999). In a typical arrangement, the solar tubes are
3.2. Photobioreactors placed parallel to each other and flat above the ground
(Fig. 2). Horizontal, parallel straight tubes are sometimes
Unlike open raceways, photobioreactors permit arranged like a fence (Fig. 3), in attempts to increase the
essentially single-species culture of microalgae for number of tubes that can be accommodated in a given
prolonged durations. Photobioreactors have been suc- area. The tubes are always oriented North–South
cessfully used for producing large quantities of micro- (Fig. 3). The ground beneath the solar collector is often
algal biomass (Molina Grima et al., 1999; Tredici, 1999; painted white, or covered with white sheets of plastic
Pulz, 2001; Carvalho et al., 2006).
A tubular photobioreactor consists of an array of
straight transparent tubes that are usually made of plas-
tic or glass. This tubular array, or the solar collector, is
where the sunlight is captured (Fig. 2). The solar col-
lector tubes are generally 0.1 m or less in diameter. Tube
diameter is limited because light does not penetrate too
deeply in the dense culture broth that is necessary for
ensuring a high biomass productivity of the photobior-
eactor. Microalgal broth is circulated from a reservoir
(i.e. the degassing column in Fig. 2) to the solar collector
Fig. 4. A 1000 L helical tubular photobioreactor at Murdoch

University, Australia. Courtesy of Professor Michael Borowitzka,
Fig. 2. A tubular photobioreactor with parallel run horizontal tubes. Murdoch University.
(Tredici, 1999), to increase reflectance, or albedo. A high cally return to a degassing zone (Fig. 2) that is bubbled
albedo increases the total light received by the tubes. with air to strip out the accumulated oxygen. Typically, a
Instead of being laid horizontally on the ground, continuous tube run should not exceed 80 m (Molina
the tubes may be made of flexible plastic and coiled Grima et al., 2001), but the exact length depends on
around a supporting frame to form a helical coil tu- several factors including the concentration of the bio-
bular photobioreactors (Fig. 4). Photobioreactors such mass, the light intensity, the flow rate, and the con-
as the one shown in Fig. 4 are potentially useful for centration of oxygen at the entrance of tube.
growing a small volume of microalgal broth, for ex- In addition to removing the accumulated dissolved
ample, for inoculating the larger tubular photobior- oxygen, the degassing zone (Fig. 2) must disengage all
eactors (Fig. 2) that are needed for producing the gas bubbles from the broth so that essentially bubble-
biodiesel. Other variants of tubular photobioreactors free broth returns to the solar collector tubes. Gas–liquid
exist (Molina Grima et al., 1999; Tredici, 1999; Pulz, separator design for achieving complete disengagement
2001; Carvalho et al., 2006), but are not widely used. of bubbles, has been discussed (Chisti and Moo-Young,
Artificial illumination of tubular photobioreactors is 1993; Chisti, 1998). Because a degassing zone is gen-
technically feasible (Pulz, 2001), but expensive com- erally optically deep compared with the solar collector
pared with natural illumination. Nonetheless, artificial tubes, it is poorly illuminated and, therefore, its volume
illumination has been used in large-scale biomass needs to be kept small relative to the volume of the solar
production (Pulz, 2001) particularly for high-value collector.
products. As the broth moves along a photobioreactor tube, pH
Biomass sedimentation in tubes is prevented by increases because of consumption of carbon dioxide
maintaining highly turbulent flow. Flow is produced (Camacho Rubio et al., 1999). Carbon dioxide is fed in the
using either a mechanical pump (Fig. 2), or a gentler degassing zone in response to a pH controller. Additional
airlift pump. Mechanical pumps can damage the biomass carbon dioxide injection points may be necessary at
(Chisti, 1999a; García Camacho et al., 2001, 2007; intervals along the tubes, to prevent carbon limitation and
Sánchez Mirón et al., 2003; Mazzuca Sobczuk et al., an excessive rise in pH (Molina Grima et al., 1999).
2006), but are easy to design, install and operate. Airlift
pumps have been used quite successfully (Molina Grima
et al., 1999, 2000, 2001; Acién Fernández et al., 2001).
Airlift pumps for use in tubular photobioreactors are Table 3
designed using the same methods that were originally Comparison of photobioreactor and raceway production methods
developed for designing conventional airlift reactors Variable Photobioreactor Raceway ponds
(Chisti et al., 1988; Chisti and Moo-Young, 1988, 1993; facility
Chisti, 1989). Airlift pumps are less flexible than me- Annual biomass 100,000 100,000
chanical pumps and require a supply of air to operate. production (kg)
Periodically, photobioreactors must be cleaned and sani- Volumetric productivity 1.535 0.117
tized. This is easily achieved using automated clean-in- (kg m− 3 d− 1)
Areal productivity 0.048 a 0.035 b
place operations (Chisti and Moo-Young, 1994; Chisti,
(kg m− 2 d− 1) 0.072 c
1999b). Biomass concentration 4.00 0.14
Photosynthesis generates oxygen. Under high irradi- in broth (kg m− 3)
ance, the maximum rate of oxygen generation in a typical Dilution rate (d− 1) 0.384 0.250
tubular photobioreactor may be as high as 10 g O2 m− 3 Area needed (m2) 5681 7828
Oil yield (m3 ha− 1) 136.9 d 99.4 d
min− 1. Dissolved oxygen levels much greater than the
58.7 e 42.6 e
air saturation values inhibit photosynthesis (Molina Annual CO2 183,333 183,333
Grima et al., 2001). Furthermore, a high concentration consumption (kg)
of dissolved oxygen in combination with intense sun- System geometry 132 parallel tubes/unit; 978 m2/pond; 12 m
light produces photooxidative damage to algal cells. To 80 m long tubes; wide, 82 m long,
0.06 m tube diameter 0.30 m deep
prevent inhibition and damage, the maximum tolerable
Number of units 6 8
dissolved oxygen level should not generally exceed a
Based on facility area.
about 400% of air saturation value. Oxygen cannot be b
Based on actual pond area.
removed within a photobioreactor tube. This limits the c
Based on projected area of photobioreactor tubes.
maximum length of a continuous run tube before oxygen d
Based on 70% by wt oil in biomass.
e
removal becomes necessary. The culture must periodi- Based on 30% by wt oil in biomass.
Photobioreactors require cooling during daylight with raceway ponds (Table 3). Both raceway and
hours. Furthermore, temperature control at night is also photobioreactor production methods are technically
useful. For example, the nightly loss of biomass due to feasible. Production facilities using photobioreactors
respiration can be reduced by lowering the temperature and raceway units of dimensions similar to those in
at night. Outdoor tubular photobioreactors are effective- Table 3 have indeed been used extensively in com-
ly and inexpensively cooled using heat exchangers. A mercial operations (Terry and Raymond, 1985; Molina
heat exchange coil may be located in the degassing Grima, 1999; Molina Grima et al., 1999; Tredici, 1999;
column (Fig. 2). Alternatively, heat exchangers may be Pulz, 2001; Lorenz and Cysewski, 2003; Spolaore
placed in the tubular loop. Evaporative cooling by water et al., 2006).
sprayed on tubes (Tredici, 1999), can also be used and Recovery of microalgal biomass from the broth is
has proven successful in dry climates. Large tubular necessary for extracting the oil. Biomass is easily
photobioreactors have been placed within temperature recovered from the broth by filtration (Fig. 5), cen-
controlled greenhouses (Pulz, 2001), but doing so is trifugation, and other means (Molina Grima et al.,
prohibitively expensive for producing biodiesel. 2003). Cost of biomass recovery can be significant.
Selecting a suitable microalgal biomass production Biomass recovery from photobioreactor cultured broth
method for making biodiesel requires a comparison of costs only a fraction of the recovery cost for broth
capabilities of raceways and tubular photobioreactors. produced in raceways. This is because the typical
biomass concentration that is produced in photobior-
4. Comparison of raceways and tubular eactors is nearly 30 times the biomass concentration
photobioreactors that is generally obtained in raceways (Table 3). Thus,
in comparison with raceway broth, much smaller
Table 3 compares photobioreactor and raceway volume of the photobioreactor broth needs to be pro-
methods of producing microalgal biomass. This com- cessed to obtain a given quantity of biomass.
parison is for an annual production level of 100 t
of biomass in both cases. Both production methods 5. Acceptability of microalgal biodiesel
consume an identical amount of carbon dioxide
(Table 3), if losses to atmosphere are disregarded. For user acceptance, microalgal biodiesel will need
The production methods in Table 3 are compared for to comply with existing standards. In the United States
optimal combinations of biomass productivity and the relevant standard is the ASTM Biodiesel Standard D
concentration that have been actually achieved in 6751 (Knothe, 2006). In European Union, separate
large-scale photobioreactors and raceways. Photobior- standards exist for biodiesel intended for vehicle
eactors provide much greater oil yield per hectare use (Standard EN 14214) and for use as heating oil
compared with raceway ponds (Table 3). This is be- (Standard EN 14213) (Knothe, 2006).
cause the volumetric biomass productivity of photo- Microalgal oils differ from most vegetable oils in
bioreactors is more than 13-fold greater in comparison being quite rich in polyunsaturated fatty acids with
four or more double bonds (Belarbi et al., 2000). For
example, eicosapentaenoic acid (EPA, C20:5n-3;
five double bonds) and docosahexaenoic acid (DHA,
C22:6n-3; six double bonds) occur commonly in algal
oils. Fatty acids and fatty acid methyl esters (FAME)
with 4 and more double bonds are susceptible to
oxidation during storage and this reduces their ac-
ceptability for use in biodiesel. Some vegetable oils
also face this problem. For example, vegetable oils
such as high oleic canola oil contain large quantities of
linoleic acid (C18:2n-6; 2-double bonds) and linolenic
acid (C18:3n-3; 3-double bonds). Although these fatty
acids have much higher oxidative stability compared
with DHA and EPA, the European Standard EN 14214
Fig. 5. Microalgal biomass recovered from the culture broth by
filtration moves along a conveyor belt at Cyanotech Corporation
limits linolenic acid methyl ester content in biodiesel
(www.cyanotech.com), Hawaii, USA. Photograph by Terry Luke. for vehicle use to 12% (mol). No such limitation exists
Courtesy of Honolulu Star-Bulletin. for biodiesel intended for use as heating oil, but
acceptable biodiesel must meet other criteria relating refining expenses (19%), distribution and marketing
to the extent of total unsaturation of the oil. Total (9%). If taxes and distribution are excluded, the average
unsaturation of an oil is indicated by its iodine value. price of petrodiesel in 2006 was $0.49/L with a 73%
Standards EN 14214 and EN 14213 require the iodine contribution from crude oil and 27% contribution from
value of biodiesel to not exceed 120 and 130 g iodine/ refining.
100 g biodiesel, respectively. Furthermore, both the Biodiesel from palm oil costs roughly $0.66/L, or
European biodiesel standards limit the contents of 35% more than petrodiesel. This suggests that the
FAME with four and more double bonds, to a maxi- process of converting palm oil to biodiesel adds about
mum of 1 % mol. $0.14/L to the price of oil. For palm oil sourced
In view of the composition of many microalgal oils, biodiesel to be competitive with petrodiesel, the price
most of them are unlikely to comply with the European of palm oil should not exceed $0.48/L, assuming an
biodiesel standards, but this need not be a significant absence of tax on biodiesel. Using the same analogy, a
limitation. The extent of unsaturation of microalgal oil reasonable target price for microalgal oil is $0.48/L for
and its content of fatty acids with more than 4 double algal diesel to be cost competitive with petrodiesel.
bonds can be reduced easily by partial catalytic Elimination of dependence on petroleum diesel and
hydrogenation of the oil (Jang et al., 2005; Dijkstra, environmental sustainability require reducing the cost
2006), the same technology that is commonly used in of production of algal oil from about $2.80/L to $0.48/
making margarine from vegetable oils. L. This is a strategic objective. The cost reduction
necessary declines to $0.72, if the algal biomass is
6. Economics of biodiesel production produced in photobioreactors and contains 70% oil by
weight. These desired levels of cost reduction are
Recovery of oil from microalgal biomass and substantial, but attainable.
conversion of oil to biodiesel are not affected by whether Microalgal oils can potentially completely replace
the biomass is produced in raceways or photobioreac- petroleum as a source of hydrocarbon feedstock for the
tors. Hence, the cost of producing the biomass is the only petrochemical industry. For this to happen, microalgal
relevant factor for a comparative assessment of photo- oil will need to be sourced at a price that is roughly
bioreactors and raceways for producing microalgal related to the price of crude oil, as follows:
biodiesel.
For the facilities detailed in Table 3, the estimated cost Calgal oil ¼ 6:9 10−3 Cpetroleum ð1Þ
of producing a kilogram of microalgal biomass is
$2.95 and $3.80 for photobioreactors and raceways, where Calgal oil ($ per liter) is the price of microalgal oil
respectively. These estimates assume that carbon dioxide and Cpetroleum is the price of crude oil in $ per barrel. For
is available at no cost. The estimation methods used have example, if the prevailing price of crude oil is $60/barrel,
been described previously (Humphreys, 1991; Molina then microalgal oil should not cost more than about
Grima et al., 2003). If the annual biomass production $0.41/L, if it is to substitute for crude oil. If the price of
capacity is increased to 10,000 t, the cost of production crude oil rises to $80/barrel as sometimes predicted, then
per kilogram reduces to roughly $0.47 and $0.60 for microalgal oil costing $0.55/L is likely to economically
photobioreactors and raceways, respectively, because of substitute for crude petroleum. Eq. (1) assumes that algal
economy of scale. Assuming that the biomass contains oil has roughly 80% of the energy content of crude
30% oil by weight, the cost of biomass for providing a petroleum.
liter of oil would be something like $1.40 and $1.81 for
photobioreactors and raceways, respectively. Oil recov-
ered from the lower-cost biomass produced in photo-
bioreactors is estimated to cost $2.80/L. This assumes
that the recovery process contributes 50% to the cost of
the final recovered oil. In comparison with this, during
2006, crude palm oil, that is probably the cheapest
vegetable oil, sold for an average price of $465/t, or
about $0.52/L.
In the United States during 2006, the on-highway
petrodiesel price ranged between $0.66 and $0.79/L. Fig. 6. Microalgal biodiesel refinery: producing multiple products
This price included taxes (20%), cost of crude oil (52%), from algal biomass.
7. Improving economics of microalgal biodiesel Box 2

Light saturation and photoinhibition
Cost of producing microalgal biodiesel can be
reduced substantially by using a biorefinery based pro- Light saturation is characterized by a light satura-
tion constant (Fig. B2), that is the intensity of light at
duction strategy, improving capabilities of microalgae
which the specific biomass growth rate is half its
through genetic engineering and advances in engineer-
maximum value, μmax. Light saturation constants for
ing of photobioreactors. microalgae tend to be much lower than the maximum
sunlight level that occurs at midday. For example, the
7.1. Biorefinery based production strategy light saturation constants for microalgae Phaeodac-
tylum tricornutum and Porphyridium cruentum are
Like a petroleum refinery, a biorefinery uses every 185 μE m− 2 s− 1 (Mann and Myers, 1968) and
−2
component of the biomass raw material to produce use- ˜ 200 μE m s− 1 (Molina Grima et al., 2000),
able products. Because all components of the biomass respectively. In comparison with these values, the
are used, the overall cost of producing any given product typical midday outdoor light intensity in equatorial
is lowered. Integrated biorefineries are already being regions is about 2000 μE m− 2 s− 1. Because of light
saturation, the biomass growth rate is much lower
operated in Canada, the United States, and Germany for
than would be possible if light saturation value could
producing biofuels and other products from crops such
be increased substantially.
as corn and soybean. This approach can be used to Above a certain value of light intensity, a further
reduce the cost of making microalgal biodiesel. increase in light level actually reduces the biomass
In addition to oils, microalgal biomass contains growth rate (Fig. B2). This phenomenon is known as
significant quantities of proteins, carbohydrates and photoinhibition. Microalgae become photoinhibited at
other nutrients (Sánchez Mirón et al., 2003). Therefore, light intensities only slightly greater than the light level
the residual biomass from biodiesel production process- at which the specific growth rate peaks. Photoinhibi-
es can be used potentially as animal feed (Fig. 6). Some tion results from generally reversible damage to the
photosynthetic apparatus, as a consequence of
of the residual biomass may be used to produce methane
excessive light (Camacho Rubio et al., 2003). Elim-
by anaerobic digestion, for generating the electrical
ination of photoinhibition or its postponement to
power necessary for running the microalgal biomass higher light intensities can greatly increase the
production facility. Excess power could be sold to average daily growth rate of algal biomass.
defray the cost of producing biodiesel.
Although the use of microalgal biomass directly to
produce methane by anaerobic digestion (Mata-Alvarez
et al., 2000; Raven and Gregersen, 2007) is technically
feasible, it cannot compete with the many other low-cost
organic substrates that are available for anaerobic digestion.
Nevertheless, algal biomass residue remaining after the
extraction of oil can be used potentially to make methane. A
microalgal biorefinery can simultaneously produce biodie-
sel, animal feed, biogas and electrical power (Fig. 6).
Extraction of other high-value products may be feasible,
depending on the specific microalgae used.
7.2. Enhancing algal biology

Fig. B2. Effect of light intensity on specific growth rate of
Genetic and metabolic engineering are likely to microalgae.
have the greatest impact on improving the economics of
production of microalgal diesel (Roessler et al., 1994;
Dunahay et al., 1996). Genetic modification of microalgae
has received little attention (León-Bañares et al., 2004).
Molecular level engineering can be used to potentially: 2. enhance biomass growth rate;
3. increase oil content in biomass;
1. increase photosynthetic efficiency to enable in- 4. improve temperature tolerance to reduce the expense
creased biomass yield on light; of cooling;
5. eliminate the light saturation phenomenon (Box 2) so Various attempts have been made to estimate the
that growth continues to increase in response to frequency of light–dark cycling (Molina Grima et al.,
increasing light level; 1999, 2000, 2001; Sánchez Mirón et al., 1999; Janssen
6. reduce photoinhibition (Box 2) that actually reduces et al., 2003; Richmond, 2004), but this problem remains
growth rate at midday light intensities that occur in unresolved. Distinct from the productivity enhancing
temperate and tropical zones; and effect of light–dark cycling, turbulence in a dense
7. reduce susceptibility to photooxidation that damages culture reduces photoinhibition and photolimitation by
cells. ensuring that the algal cells do not reside continuously in
either the well lit zone or the dark zone for long periods.
In addition, there is a need to identify possible In principle, motionless mixers installed inside
biochemical triggers and environmental factors that photobioreactor tubes can be used to substantially
might favor accumulation of oil. Stability of engineered enhance the mixing between the peripheral lit zone
strains and methods for achieving stable production in and the interior dark zone (Molina Grima et al., 1999,
industrial microbial processes are known to be impor- 2001; Sánchez Mirón et al., 1999). Such mixers have
tant issues (Zhang et al., 1996), but have been barely proved useful in other tubular reactors (Chisti et al.,
examined for microalgae. 1990; Chisti, 1998; Thakur et al., 2003). Unfortunately,
existing designs of motionless mixers are not satisfac-
7.3. Photobioreactor engineering tory for photobioreactors because they substantially
reduce penetration of light in the tubes. New designs of
Although a capability for reliable engineering and motionless mixers are needed.
operation of tubular photobioreactors has emerged Like cells of higher plants (Moo-Young and Chisti,
(Acién Fernández et al., 1997, 1998, 2001; Camacho 1988) and animals (Zhang et al., 1995; Chisti, 2000,
Rubio et al., 1999; Molina Grima et al., 1999, 2000, 2001; García Camacho et al., 2005), microalgae are
2001; Sánchez Mirón et al., 1999, 2000; Janssen et al., damaged by intense hydrodynamic shear fields that
2003; Carvalho et al., 2006), problems remain. occur in high-velocity flow in pipes, pumps and mixing
Photobioreactor tubes operated with high-density tanks (Chisti, 1999a; García Camacho et al., 2001, 2007;
culture for attaining high productivity, inevitably con- Sánchez Mirón et al., 2003; Mazzuca Sobczuk et al.,
tain a photolimited central dark zone and a relatively 2006). Some algae are more sensitive to shear damage
better lit peripheral zone (Molina Grima et al., 1999, than others. Shear sensitivity can pose a significant
2001). Light intensity in the photolimited zone is lower problem as the intensity of turbulence needed in
than the saturation light level (Box 2). Turbulence in the photobioreactors to generate optimal light–dark cycling
tube causes rapid cycling of the fluid between the light (Grobbelaar et al., 1996; Camacho Rubio et al., 2003) is
and dark zones. The frequency of light–dark cycling difficult to achieve (Molina Grima et al., 2000, 2001;
depends on several factors, including the intensity of Camacho Rubio et al., 2004) without damaging algal
turbulence, concentration of cells, optical properties of cells. Methods have been developed to reduce the
the culture, the diameter of the tube, and the external damage associated with turbulence of limited intensity
irradiance level (Molina Grima et al., 2000, 2001). (García Camacho et al., 2001; Mazzuca Sobczuk et al.,
Under conditions of sufficient and excess external irra- 2006). Intensities of shear stress are not easily
diance, light–dark cycling of above a certain frequency determined in bioreactors (Chisti and Moo-Young,
can increase biomass productivity relative to the case 1989; Chisti, 1989, 1999a), but improved methods for
when the same quantity of light is supplied continuously doing so are emerging (Sánchez Pérez et al., 2006).
over the same total exposure time (Philliphs and Myers, Some algae will preferentially grow attached to the
1953; Terry, 1986; Grobbelaar, 1994; Nedbal et al., internal wall of the photobioreactor tube, thus preventing
1996; Grobbelaar et al., 1996; Camacho Rubio et al., light penetration into the tube and reducing bioreactor
2003). Light–dark cycling times of 10 ms, for example, productivity. Robust methods for controlling wall growth
are known to improve growth compared with continu- are needed. Wall growth is controlled by some of the
ous illumination of equal cumulative quantity. Benefi- following methods: 1. use of large slugs of air to
cial effects of rapid light–dark cycling under light intermittently scour the internal surface of the tube;
saturation conditions are associated with the short dark 2. circulation of close fitting balls in continuous run tubes
period allowing the photosynthetic apparatus of the cells to clean the internal surface; 3. highly turbulent flow; and
to fully recover from the excited state of the previous 4. suspended sand or grit particles to abrade any biomass
illumination event. adhering to the internal surface. Potentially, enzymes that
digest the polymer glue that binds algal cells to the tube Borowitzka MA. Pharmaceuticals and agrochemicals from microalgae.
walls, may be used for controlling wall growth. In: Cohen Z, editor. Chemicals from microalgae. Taylor & Francis;
1999. p. 313–52.
Bioprocess intensification approaches (Chisti and Camacho Rubio F, Acién Fernández FG, García Camacho F,
Moo-Young, 1996; Chisti, 2003) that have proved so Sánchez Pérez JA, Molina Grima E. Prediction of dissolved
successful in improving the economics of various bio- oxygen and carbon dioxide concentration profiles in tubular photo-
technology based processes have been barely assessed for bioreactors for microalgal culture. Biotechnol Bioeng 1999;62:
71–86.
use with photobioreactors.
Camacho Rubio F, García Camacho F, Fernández Sevilla JM, Chisti Y,
Molina Grima E. A mechanistic model of photosynthesis in
8. Conclusion microalgae. Biotechnol Bioeng 2003;81:459–73.
Camacho Rubio F, Sánchez Mirón A, Cerón García MC, García
As demonstrated here, microalgal biodiesel is techni- Camacho F, Molina Grima E, Chisti Y. Mixing in bubble columns:
cally feasible. It is the only renewable biodiesel that can a new approach for characterizing dispersion coefficients. Chem
Eng Sci 2004;59:4369–76.
potentially completely displace liquid fuels derived from Carvalho AP, Meireles LA, Malcata FX. Microalgal reactors: a review
petroleum. Economics of producing microalgal biodiesel of enclosed system designs and performances. Biotechnol Prog
need to improve substantially to make it competitive with 2006;22:1490–506.
petrodiesel, but the level of improvement necessary Chisti Y. An unusual hydrocarbon. J Ramsay Soc 1980–81;27–28: 24–6.
appears to be attainable. Producing low-cost microalgal Chisti Y. Airlift bioreactors. Elsevier; 1989. p. 355.
Chisti Y. Pneumatically agitated bioreactors in industrial and
biodiesel requires primarily improvements to algal environmental bioprocessing: hydrodynamics, hydraulics and
biology through genetic and metabolic engineering. Use transport phenomena. Appl Mech Rev 1998;51:33–112.
of the biorefinery concept and advances in photobior- Chisti Y. Shear sensitivity. In: Flickinger MC, Drew SW, editors.
eactor engineering will further lower the cost of Encyclopedia of bioprocess technology: fermentation, biocatalysis,
production. In view of their much greater productivity and bioseparation, vol. 5. Wiley; 1999a. p. 2379–406.
Chisti Y. Modern systems of plant cleaning. In: Robinson R, Batt C, Patel
than raceways, tubular photobioreactors are likely to be P, editors. Encyclopedia of food microbiology. Academic Press;
used in producing much of the microalgal biomass 1999b. p. 1806–15.
required for making biodiesel. Photobioreactors provide a Chisti Y. Animal-cell damage in sparged bioreactors. Trends Biotechnol
controlled environment that can be tailored to the specific 2000;18:420–32.
demands of highly productive microalgae to attain a Chisti Y. Hydrodynamic damage to animal cells. Crit Rev Biotechnol
2001;21:67–110.
consistently good annual yield of oil. Chisti Y. Sonobioreactors: using ultrasound for enhanced microbial
productivity. Trends Biotechnol 2003;21:89–93.
References Chisti Y, Moo-Young M. Prediction of liquid circulation velocity in
airlift reactors with biological media. J Chem Technol Biotechnol
Acién Fernández FG, García Camacho F, Sánchez Pérez JA, 1988;42:211–9.
Fernández Sevilla JM, Molina Grima E. A model for light Chisti Y, Moo-Young M. On the calculation of shear rate and apparent
distribution and average solar irradiance inside outdoor tubular viscosity in airlift and bubble column bioreactors. Biotechnol
photobioreactors for the microalgal mass culture. Biotechnol Bioeng 1989;34:1391–2.
Bioeng 1997;55:701–14. Chisti Y, Moo-Young M. Improve the performance of airlift reactors.
Acién Fernández FG, García Camacho F, Sánchez Pérez JA, Chem Eng Prog 1993;89(6):38–45.
Fernández Sevilla J, Molina Grima E. Modelling of biomass Chisti Y, Moo-Young M. Clean-in-place systems for industrial
productivity in tubular photobioreactors for microalgal cultures. bioreactors: design, validation and operation. J Ind Microbiol
Effects of dilution rate, tube diameter and solar irradiance. 1994;13:201–7.
Biotechnol Bioeng 1998;58:605–11. Chisti Y, Moo-Young M. Bioprocess intensification through bioreactor
Acién Fernández FG, Fernández Sevilla JM, Sánchez Pérez JA, engineering. Trans I Chem E 1996;74A:575–83.
Molina Grima E, Chisti Y. Airlift-driven external-loop tubular Chisti Y, Halard B, Moo-Young M. Liquid circulation in airlift
photobioreactors for outdoor production of microalgae: assessment reactors. Chem Eng Sci 1988;43:451–7.
of design and performance. Chem Eng Sci 2001;56:2721–32. Chisti Y, Kasper M, Moo-Young M. Mass transfer in external-loop
Akkerman I, Janssen M, Rocha J, Wijffels RH. Photobiological airlift bioreactors using static mixers. Can J Chem Eng
hydrogen production: photochemical efficiency and bioreactor 1990;68:45–50.
design. Int J Hydrogen Energy 2002;27:1195–208. Demirbas A. Biodiesel production from vegetable oils via catalytic and
Banerjee A, Sharma R, Chisti Y, Banerjee UC. Botryococcus braunii: non-catalytic supercritical methanol transesterification methods.
a renewable source of hydrocarbons and other chemicals. Crit Rev Pror Energy Combust Sci 2005;31(5–6):466–87.
Biotechnol 2002;22:245–79. Dijkstra AJ. Revisiting the formation of trans isomers during partial
Barnwal BK, Sharma MP. Prospects of biodiesel production from hydrogenation of triacylglycerol oils. Eur J Lipid Sci Technol
vegetables oils in India. Renew Sustain Energy Rev 2005;9:363–78. 2006;108(3):249–64.
Belarbi E-H, Molina Grima E, Chisti Y. A process for high yield and Dunahay TG, Jarvis EE, Dais SS, Roessler PG. Manipulation of
scaleable recovery of high purity eicosapentaenoic acid esters from microalgal lipid production using genetic engineering. Appl
microalgae and fish oil. Enzyme Microb Technol 2000;26: 516–29. Biochem Biotechnol 1996;57–58:223–31.
Fedorov AS, Kosourov S, Ghirardi ML, Seibert M. Continuous H2 Mallick N. Biotechnological potential of immobilized algae for
photoproduction by Chlamydomonas reinhardtii using a novel wastewater N, P and metal removal: a review. Biometals 2002;15:
two-stage, sulfate-limited chemostat system. Appl Biochem 377–90.
Biotechnol 2005;121124:403–12. Mann JE, Myers J. On pigments, growth and photosynthesis of
Felizardo P, Correia MJN, Raposo I, Mendes JF, Berkemeier R, Phaeodactylum tricornutum. J Phycol 1968;4:349–55.
Bordado JM. Production of biodiesel from waste frying oil. Waste Mata-Alvarez J, Mace S, Llabres P. Anaerobic digestion of organic
Manag 2006;26(5):487–94. solid wastes. An overview of research achievements and
Fukuda H, Kondo A, Noda H. Biodiesel fuel production by perspectives. Bioresour Technol 2000;74:3–16.
transesterification of oils. J Biosci Bioeng 2001;92:405–16. Mazzuca Sobczuk T, García Camacho F, Molina Grima E, Chisti Y.
García Camacho F, Molina Grima E, Sánchez Mirón A, González Pascual Effects of agitation on the microalgae Phaeodactylum tricornutum
V, Chisti Y. Carboxymethyl cellulose protects algal cells against and Porphyridium cruentum. Bioprocess Biosyst Eng 2006;28:
hydrodynamic stress. Enzyme Microb Technol 2001;29: 602–10. 243–50.
García Camacho F, Belarbi EH, Cerón García MC, Sánchez Mirón A, Meher LC, Vidya Sagar D, Naik SN. Technical aspects of biodiesel
Chile T, Chisti Y, et al. Shear effects on suspended marine sponge production by transesterification — a review. Renew Sustain
cells. Biochem Eng J 2005;26:115–21. Energy Rev 2006;10:248–68.
García Camacho F, Gallardo Rodríguez J, Sánchez Mirón A, Cerón Melis A. Green alga hydrogen production: progress, challenges and
García MC, Belarbi EH, Chisti Y, et al. Biotechnological prospects. Int J Hydrogen Energy 2002;27:1217–28.
significance of toxic marine dinoflagellates. Biotechnol Adv Metting FB. Biodiversity and application of microalgae. J Ind
2007;25:176–94. Microbiol 1996;17:477–89.
Gavrilescu M, Chisti Y. Biotechnology — a sustainable alternative for Metting B, Pyne JW. Biologically-active compounds from microalgae.
chemical industry. Biotechnol Adv 2005;23:471–99. Enzyme Microb Technol 1986;8:386–94.
Ghirardi ML, Zhang JP, Lee JW, Flynn T, Seibert M, Greenbaum E, Metzger P, Largeau C. Botryococcus braunii: a rich source for
et al. Microalgae: a green source of renewable H2. Trends hydrocarbons and related ether lipids. Appl Microbiol Biotechnol
Biotechnol 2000;18:506–11. 2005;66:486–96.
Grobbelaar JU. Turbulence in algal mass cultures and the role of light/dark Molina Grima E. Microalgae, mass culture methods. In: Flickinger MC,
fluctuations. J Appl Phycol 1994;6:331–5. Drew SW, editors. Encyclopedia of bioprocess technology: fermen-
Grobbelaar JU. Algal nutrition. In: Richmond A, editor. Handbook of tation, biocatalysis and bioseparation, vol. 3. Wiley; 1999. p. 1753–69.
microalgal culture: biotechnology and applied phycology. Blackwell; Molina Grima E, Acién Fernández FG, García Camacho F, Chisti
2004. p. 97–115. Y. Photobioreactors: light regime, mass transfer, and scaleup.
Grobbelaar J, Nedbal L, Tichy V. Influence of high frequency light/dark J Biotechnol 1999;70:231–47.
fluctuations on photosynthetic characteristics of microalgae photo Molina Grima E, Acién Fernández FG, García Camacho F, Camacho
acclimated to different light intensities and implications for mass Rubio F, Chisti Y. Scale-up of tubular photobioreactors. J Appl
algal cultivation. J Appl Phycol 1996;8:335–43. Phycol 2000;12:355–68.
Guschina IA, Harwood JL. Lipids and lipid metabolism in eukaryotic Molina Grima E, Fernández J, Acién Fernández FG, Chisti Y. Tubular
algae. Prog Lipid Res 2006;45:160–86. photobioreactor design for algal cultures. J Biotechnol 2001;92:
Humphreys K. Jelen's cost and optimization engineering. 3rd ed. 113–31.
McGraw-Hill; 1991. Molina Grima E, Belarbi E-H, Acién Fernández FG, Robles Medina A,
Jang ES, Jung MY, Min DB. Hydrogenation for low trans and high Chisti Y. Recovery of microalgal biomass and metabolites: process
conjugated fatty acids. Comp Rev Food Sci Saf 2005;4: 22–30. options and economics. Biotechnol Adv 2003;20:491–515.
Janssen M, Tramper J, Mur LR, Wijffels RH. Enclosed outdoor Moo-Young M, Chisti Y. Considerations for designing bioreactors for
photobioreactors: light regime, photosynthetic efficiency, scale-up, shear-sensitive culture. Biotechnology 1988;6:1291–6.
and future prospects. Biotechnol Bioeng 2003;81:193–210. Munoz R, Guieysse B. Algal–bacterial processes for the treatment of
Kalin M, Wheeler WN, Meinrath G. The removal of uranium from hazardous contaminants: a review. Water Res 2006;40:2799–815.
mining waste water using algal/microbial biomass. J Environ Nagle N, Lemke P. Production of methyl-ester fuel from microalgae.
Radioact 2005;78:151–77. Appl Biochem Biotechnol 1990;24–5:355–61.
Kapdan IK, Kargi F. Bio-hydrogen production from waste materials. Nedbal L, Tichý V, Grobbelaar JU, Xiong VF, Neori A. Microscopic
Enzyme Microb Technol 2006;38:569–82. green algae and cyanobacteria in high-frequency intermittent light.
Kay RA. Microalgae as food and supplement. Crit Rev Food Sci Nutr J Appl Phycol 1996;8:325–33.
1991;30:555–73. Philliphs JN, Myers J. Growth rate of Chlorella in flashing light. Plant
Knothe G, Dunn RO, Bagby MO. Biodiesel: the use of vegetable oils Physiol 1953;29:152–61.
and their derivatives as alternative diesel fuels. ACS Symp Ser Pulz O. Photobioreactors: production systems for phototrophic
1997;666:172–208. microorganisms. Appl Microbiol Biotechnol 2001;57:287–93.
Knothe G. Analyzing biodiesel: standards and other methods. J Am Ratledge C. Single cell oils — have they a biotechnological future?
Oil Chem Soc 2006;83:823–33. Trends Biotechnol 1993;11:278–84.
Kulkarni MG, Dalai AK. Waste cooking oil — an economical source Ratledge C, Wynn JP. The biochemistry and molecular biology of lipid
for biodiesel: A review. Ind Eng Chem Res 2006;45:2901–13. accumulation in oleaginous microorganisms. Adv Appl Microbiol
León-Bañares R, González-Ballester D, Galváan A, Fernández E. 2002;51:1–51.
Transgenic microalgae as green cell-factories. Trends Biotechnol Raven RPJM, Gregersen KH. Biogas plants in Denmark: successes
2004;22:45–52. and setbacks. Renew Sustain Energy Rev 2007;11:116–32.
Lorenz RT, Cysewski GR. Commercial potential for Haematococcus Richmond A. Biological principles of mass cultivation. In: Richmond
microalga as a natural source of astaxanthin. Trends Biotechnol A, editor. Handbook of microalgal culture: biotechnology and
2003;18:160–7. applied phycology. Blackwell; 2004. p. 125–77.
Roessler PG, Brown LM, Dunahay TG, Heacox DA, Jarvis EE, Singh S, Kate BN, Banerjee UC. Bioactive compounds from
Schneider JC, et al. Genetic-engineering approaches for enhanced cyanobacteria and microalgae: an overview. Crit Rev Biotechnol
production of biodiesel fuel from microalgae. ACS Symp Ser 2005;25:73–95.
1994;566:255–70. Spolaore P, Joannis-Cassan C, Duran E, Isambert A. Commercial
Sánchez Mirón A, Contreras Gómez A, García Camacho F, Molina applications of microalgae. J Biosci Bioeng 2006;101:87–96.
Grima E, Chisti Y. Comparative evaluation of compact photo- Suresh B, Ravishankar GA. Phytoremediation — a novel and
bioreactors for large-scale monoculture of microalgae. J Biotechnol promising approach for environmental clean-up. Crit Rev
1999;70:249–70. Biotechnol 2004;24:97–124.
Sánchez Mirón A, García Camacho F, Contreras Gómez A, Molina Terry KL. Photosynthesis in modulated light: quantitative dependence
Grima E, Chisti Y. Bubble column and airlift photobioreactors for of photosynthesis enhancement on flashing rate. Biotechnol
algal culture. AIChE J 2000;46:1872–87. Bioeng 1986;28:988–95.
Sánchez Mirón A, Cerón García M-C, Contreras Gómez A, García Terry KL, Raymond LP. System design for the autotrophic production
Camacho F, Molina Grima E, Chisti Y. Shear stress tolerance and of microalgae. Enzyme Microb Technol 1985;7:474–87.
biochemical characterization of Phaeodactylum tricornutum in Thakur RK, Vial C, Nigam KDP, Nauman EB, Djelveh G. Static
quasi steady-state continuous culture in outdoor photobioreactors. mixers in the process industries — a review. Chem Eng Res Des
Biochem Eng J 2003;16:287–97. 2003;81:787–826.
Sánchez Pérez JA, Rodríguez Porcel EM, Casas López JL, Fernández Tredici MR. Bioreactors, photo. In: Flickinger MC, Drew SW, editors.
Sevilla JM, Chisti Y. Shear rate in stirred tank and bubble column Encyclopedia of bioprocess technology: fermentation, biocatalysis
bioreactors. Chem Eng J 2006;124:1–5. and bioseparationWiley; 1999. p. 395–419.
Sawayama S, Inoue S, Dote Y, Yokoyama S-Y. CO2 fixation and oil Vaishampayan A, Sinha RP, Hader DP, Dey T, Gupta AK, Bhan U, et al.
production through microalga. Energy Convers Manag 1995;36: Cyanobacterial biofertilizers in rice agriculture. Bot Rev 2001;67:
729–31. 453–516.
Schwartz RE. Pharmaceuticals from cultured algae. J Ind Microbiol Van Gerpen J. Biodiesel processing and production. Fuel Process
1990;5:113–23. Technol 2005;86:1097–107.
Sharma R, Chisti Y, Banerjee UC. Production, purification, charac- Walter TL, Purton S, Becker DK, Collet C. Microalgae as bioreactor.
terization, and applications of lipases. Biotechnol Adv 2001;19: Plant Cell Rep 2005;24:629–41.
627–62. Yun YS, Lee SB, Park JM, Lee CI, Yang JW. Carbon dioxide fixation
Sheehan J, Dunahay T, Benemann J, Roessler P. A look back at the U.S. by algal cultivation using wastewater nutrients. J Chem Technol
Department of Energy's Aquatic Species Program — biodiesel from Biotechnol 1997;69:451–5.
algae. National Renewable Energy Laboratory, Golden, CO; 1998. Zhang Z, Chisti Y, Moo-Young M. Effects of the hydrodynamic
Report NREL/TP-580–24190. environment and shear protectants on survival of erythrocytes in
Shimizu Y. Microalgal metabolites: a new perspective. Annu Rev suspension. J Biotechnol 1995;43:33–40.
Microbiol 1996;50:431–65. Zhang Z, Moo-Young M, Chisti Y. Plasmid stability in recombinant
Shimizu Y. Microalgal metabolites. Curr Opin Microbiol 2003;6: 236–43. Saccharomyces cerevisiae. Biotechnol Adv 1996;14:401–35.
1490 Biotechnol. Prog. 2006, 22, 1490−1506
Microalgal Reactors: A Review of Enclosed System Designs and Performances
Ana P. Carvalho, Luı́s A. Meireles, and F. Xavier Malcata*

Escola Superior de Biotecnologia, Universidade Católica Portuguesa, Rua Dr. António Bernardino de Almeida, P-4200-072
Porto, Portugal
One major challenge to industrial microalgal culturing is to devise and develop technical apparata,
cultivation procedures and algal strains susceptible of undergoing substantial increases in
efficiency of use of solar energy and carbon dioxide. Despite several research efforts developed
to date, there is no such thing as “the best reactor system”- defined, in an absolute fashion, as
the one able to achieve maximum productivity with minimum operation costs, irrespective of
the biological and chemical system at stake. In fact, choice of the most suitable system is situation-
dependent, as both the species of alga available and the final purpose intended will play a role.
The need of accurate control impairs use of open-system configurations, so current investigation
has focused mostly on closed systems. In this review, several types of closed bioreactors described
in the technical literature as able to support production of microalgae are comprehensively
presented and duly discussed, using transport phenomenon and process engineering methodologi-
cal approaches. The text is subdivided into subsections on: reactor design, which includes tubular
reactors, flat plate reactors and fermenter-type reactors; and processing parameters, which include
gaseous transfer, medium mixing and light requirements.
Contents to replace those from animal sources, which were difficult to

obtain during World War II (1). In the U.S., research on mass
1. Introduction 1490
culture of microalgae began as a collateral development of
2. Reactor design 1492
fundamental studies on photosynthesis; in attempts to translate
2.1. Tubular reactors 1492 the biological requirements of microalgal culture into engineer-
2.1.1. Vertical tubular reactors 1492 ing specifications, a large-scale culture plant was made at
2.1.2. Horizontal tubular reactors 1493 Stanford Research Institute, back in 1948. During 1951, Arthur
2.1.3. Helical tubular reactors 1494 D. Little, Inc. (Cambridge, MA) made further advances through
2.1.4. r-Shaped reactors 1494 construction and operation of a Chlorella pilot plant for the
2.2. Flat plate reactors 1495 Carnegie Institution (2). Other studies followed in Japan, under
2.3. Fermenter-type reactors 1496 the guidance of Tamiya (2). Although experimental results
3. Processing parameters 1496 showed that continuous culture was possible, many subjects still
needed improvement, as microalgal proteins could not compete
3.1. Gaseous transfer 1496
with such inexpensive plant sources as soybean meal.
3.1.1. Carbon source 1496
Other developments were achieved by Oswald, who started
3.1.2. Transport process 1497
studying the role of microalgal photosynthesis in ponds and
3.1.3. CO2 transfer systems 1498 accordingly developed a high-rate algal pond for photosyn-
3.1.4. O2 removal 1500 thetic wastewater treatment. In the 1960s, Nichoporovich and
3.2. Mixing 1501 Semenenko intensively studied closed culture systems, for
3.3. Light requirement 1502 extraterrestrial life support during prolonged missions in outer
4. Conclusions 1503 space; this subject also received considerable attention by NASA
(1). The advent of the oil crisis in the 1970s led researchers to
investigate microalgae as potential sources of biomass, aiming
1. Introduction at methane production (3); more recently, advances have focused
on the production of fine chemicals and secondary metabolites,
From an economic point of view, microalgae may be
described as microorganisms with the ability to “harvest the which may reach high prices in the world market.
sun” and hence transform its radiant energy into valuable The evolution in goals throughout time has therefore been
products, at the expense of (theoretically) inexpensive natural driven by two major streamlines: (i) requirement for alternative
resources (viz., CO2 and H2O). The idea of producing microal- sources of several products, which were scarce as a result of
gae at the technical scale first occurred to German scientists, in political or economic reasons; and (ii) perception that microal-
concerted attempts to devise inexpensive sources of protein able gal-mediated processes (as happens with most biotechnological
ones) are usually characterized by noncompetitive production
* To whom correspondence should be addressed. Tel: 351 225 580 004. costs, the economic feasibility of which relies heavily on the
Fax: 351 225 090 351. Email: fxmalcata@esb.ucp.pt. market value of the resulting compounds. As expected, the
10.1021/bp060065r CCC: $33.50 © 2006 American Chemical Society and American Institute of Chemical Engineers
Published on Web 11/15/2006
Biotechnol. Prog., 2006, Vol. 22, No. 6 1491
Table 1. Generic Description of Commercial Microalgal Culture Systems Currently in Use

microorganism metabolite commercial use culturing system location
Chlorella spp. astaxanthin pigmenting agent circular ponds with Taiwan, Japan,
rotating arm U.S., Thailand
ferredoxin laboratory use raceway ponds
fermenters (heterotrophic)
Dunaliella salina β-carotene pigmenting agent extensive open ponds Australia, China, India,
raceway ponds Chile, U.S., Israel
Euglena gracilis arachidonic acid laboratory use
Haematococcus pluVialis astaxanthin pigmenting agent U.S.
Isochrysis galbana docosahexaenoic acid functional food additive
Phaeodactylum tricornutum eicosapentaenoic acid medical use
Spirulina platensis phycocyanin food coloring
Crypthecodinium cohnii docosahexaenoic acid functional food additive fermenters (heterotrophic) U.S.
nature of these compounds has evolved with time, in response ently reached their upper limit, with little room for further
to changing market demands. technological improvement.
A broad list of applications of microalga cultures has been In view of these difficulties, another approach was envisaged,
described and discussed in the literature. Those that have attained which is based on the alternative use of closed systems. These
commercial expression encompass the areas of healthy foods, are more appropriate for sensitive strains (which grow in non-
food additives, pigments, diets for aquaculture, growth-regulat- extreme environments) or when the final product is highly
ing agents, secondary metabolites and wastewater treatments susceptible to microbial degradation (e.g., bacterial metaboliza-
(4, 5) (see Table 1). The production of several bioactive tion of amino acids and polysaccharides). The closed config-
compounds such as hydrocarbons, isotopes, polysaccharides, and uration makes the control of contaminants easier, hence allowing
antifungal, antitumor, antibacterial and antiviral substances is growth in photo-autotrophic, heterotrophic or mixotrophic
currently under study; uses of microalgae for CO2 fixation, modes; because of the higher cell mass productivities attained
removal of nitric oxide from flue gas, fuel production, recovery (up to 3-fold those obtained in open systems) (3), harvesting
of heavy metals from effluents and in outer space technologies costs per unit mass can also be significantly reduced. Neverthe-
are also in order (3, 5-11). Nevertheless, despite such enormous less, the costs of closed systems are higher than their open-
potential, the number of applications that has reached the system counterparts, in addition to several other disadvantages
industrial scale is comparatively rather limited. (see Table 2). In fact, despite their higher volumetric productiv-
From the pioneer commercial large-scale cultures of microal- ity, closed systems were not a consensual industrial choice until
gae in the 1960s in Japan using Chlorella (12), only a few more only recently; intensive capital investments and high production
species have been employed industrially ever since, which costs account for this realization. In order to minimize produc-
include Spirulina sp. and Scenedesmus sp. for healthy food and tion costs, the major factors that play a role in the process ought
phycocyanin synthesis (a blue colorant in food and cosmetics), to be identified and their specific contributions comprehensively
Haematococcus pluVialis for production of astaxanthin (a food studied, in order to maximize advantages and minimize disad-
colorant), and Dunaliella salina for the manufacture of β-car- vantages. The choice of which configuration is preferable
otene (a vitamin A substitute and food colorant). In addition, depends obviously on the objective function considered; e.g.,
Crypthecodinium cohnii and Schizochytrium sp. are also com- wastewater treatment would preclude closed systems, owing to
mercially used for production of polyunsaturated fatty acids the unacceptably high costs that arise from the large volumes
(e.g., docosahexaenoic acid), although following a fermentative to be processed and the low added value of the feedstock
process (13). handled.
Open ponds were the ancient configurations proposed for When improvement in efficacy of a closed photo-bioreactor
microalga production and are still the most widely applied in is the goal, light and CO2 supply are key processing parameters,
industrial processes. They usually consist of either circular ponds especially owing to the difficulties associated with their control
with a rotating arm to mix the culture, or long channels in a (viz., assurance of stability throughout time and uniformity
single or multiple loop configuration stirred by paddle wheels throughout space); most of the so-called novel bioreactors do
(3), although simpler configurations also exist. The main in fact attempt to overcome the constraints related to control of
constraints related to operation of these open systems are the said parameters (14). However, the technology that supports
impossibility to control contamination, the difficulty to keep supply of adequate amounts of CO2 to microalgal cells is still
the culture environment constant and the cost of the harvesting poorly developed, which contrasts with the substantial research
stage. In order to avoid microbial contamination, highly selective efforts currently underway on the genetic improvement of native
conditions are necessary, so as to guarantee dominance by the freshwater and marine species for specific applications. On the
selected strain (e.g., D. salina dominance requires highly salted other hand, although the problem of light supply has sometimes
media, whereas Spirulina platensis demands high pH values); been circumvented by growing the microalgae heterotrophically,
unfortunately, both of these conditions are not suitable for most not all microalgae (or microalgal products, for that matter) can
microalgal species. The direct effect of weather conditions on be produced in this way.
the characteristics of the open-pond cultivation media also makes Therefore, issues such as contamination control, gaseous
it very difficult to keep preset values for the environmental exchanges, mixing patterns, suitability of light supply (which
parameters. Regarding the harvesting phase, the huge volume comprises light quality and quantity), geometrical configuration
of culture to be harvested (because of the low cell densities and building material are considered relevant, and will accord-
attained) magnifies the cost of processing, thus substantially ingly be discussed to some extent in this review. The aspects
increasing the final cost of the product. Because of the of nutrient supply, as well as pH and temperature control, will
aforementioned serious constraints, open systems have appar- not be considered here, because no major improvements are
1492 Biotechnol. Prog., 2006, Vol. 22, No. 6
Table 2. Main Design Features of Open and Closed Photobioreactors

feature open systems closed systems
area-to-volume ratio large (4-10 times higher than closed counterpart) small
algal species restricted flexible
main criteria for species selection growth competition shear-resistance
population density low high
harvesting efficiency low high
cultivation period limited extended
contamination possible unlikely
water loss through evaporation possible prevented
light utilization efficiency poor/fair fair/excellenta
gas transfer poor fair/high
temperature control none excellent
most costly parameters mixing oxygen control, temperature control
capital investment small high
a Dependent on transparency of construction material.
expected therefrom, as their control is rather simple. There has 2.1.1. Vertical Tubular Reactors. The airlift and bubble
been indeed substantial research on those topics, and the column reactors are examples of vertical tubular reactors (VTR),
solutions available to date already address conveniently the regularly composed of polyethylene or glass tubes (Figure 1),
intended purposes (15). Despite its relative importance, the which are sufficiently transparent to allow good light penetration
harvesting phase is also out of the scope of this review. but are manufactured with sufficiently common materials so as
to be nonexpensive. Air is bubbled at the bottom-a strategy
2. Reactor Design that provides good overall mixing, sufficient supply of CO2,
The main parameter that affects reactor design is provision and efficient removal of O2. Polyethylene bags have frequently
for light penetration, which implies a high surface-to-volume been used, with advantage taken from their particularly low cost,
ratio; such penetration is crucial if one wants to improve the high transparency and good sterility at startup-due to the high
photosynthetic efficiency, which is in turn a sine qua non temperatures used during film extrusion (29, 30); 32 cm × 250
condition to reach high product and biomass productivities. In cm (ca. 25 L) bag reactors were employed by Cohen (31) for
order to achieve said high surface-to-volume ratio, several cultivation of Porphyridium sp., which were able to reach cell
shapes have been developed that met with success. These shapes concentrations 3-fold those typically attained in open ponds.
can be grouped in three basic types, viz., tubular, flat plate and Trotta (30) described a reactor composed of several 30 cm ×
fermenter-type; the former two are specifically designed for 180 cm (ca. 50 L) polyethylene bag reactors, with various
efficient harvest of sunlight, whereas the latter requires artificial closing devices, and complemented with air and medium
illumination. supplies. Martinéz-Jerónimo (32) also reported cultivation in
Tubular and flat plate reactors are undoubtedly the most 16.8 cm × 224 cm (ca. 40 L) bags to be practical, and to exhibit
popular choices (12), considering that the light source required an improved area-to-volume ratio. Tredici and Rodolfi (33)
is free and readily available. Those reactor types are based on improved this idea by using a culture chamber made of flexible
the same principle, viz., to guarantee the highest possible area- transparent plastic film contained in a rigid metal framework,
to-volume ratio while ensuring reasonable working volume, so as to form a vertical panel of reduced width. More recently,
mixing pattern and carbon dioxide level. Both reactor configura- Chae (34) reported a pilot-scale photo-bioreactor that uses
tions may work with a separate unit for gas transfer, and several sunlight and flue gas, and consists of a vertical tubular part (kept
layouts have been already tested with success (16-20). Such in the dark) and a horizontal tubular part (subject to sunlight).
systems comprise: a light-harvesting unit, which employs small
Although cultivation of microalgae in the above systems is
diameter tubing so as to provide a high area-to-volume ratio
simple and hence widely employed (including in hatcheries),
that favors high photosynthetic activity, and a gas exchange unit,
the corresponding technology is somehow primitive, with
in which CO2 is supplied and biomass harvesting is processed.
obvious constraints derived from the high fragility and the low
The culture is circulated between those two units by a pump,
versatility of the material in stake (21). Furthermore, scale-up
which needs to be carefully designed and operated in order to
of these systems was initially thought to be easy, but ac-
prevent shear forces from disrupting cell integrity (17, 21-27).
cumulated experience (32) has indicated that increases in culture
Several reactor designs and corresponding productivities are
volume decrease bag productivity.
tabulated in Table 3.
2.1. Tubular Reactors. Most configurations of tubular Rigid VTR have also been frequently used. A 33.7 cm ×
reactors (TR) are one of the following three types: (i) simple 250 cm (ca. 40 L) polyethylene reactor was discussed by Laing
airlift and bubble column, which is composed of vertical tubing (29), in which temperature was controlled by a refrigeration
(in the form of a vertical tubular reactor) that is transparent so fluid flown through its double wall and in which artificial light
as to allow for light penetration and where CO2 is supplied via was provided from the inside. Myamoto (35), James (36) and
bubbling; (ii) horizontal tubular reactor, which is composed of Fukami (37) presented similar reactor configurations, but using
horizontal transparent tubing, usually bearing gas transfer direct sunlight; their main advantages were low cost and ease
systems attached to the connections; and (iii) helical tubular of operation. Unfortunately, scale-up is not straightforward;
reactor, which is composed of a flexible plastic tube coiled in furthermore, in order to provide enough culture volume, as well
a circular framework. Another such reactor that deserves as efficient gas transfer rate, the reactor diameter should be
particular attention is the R-shape tubular reactor, initially relatively high when compared to flat plate or tubular loop
conceived by Lee (28), because of its unique engineering design reactors, a requirement that in turn decreases the area-to-volume
that is characterized by a unidirectional, high liquid flow rate, ratio and consequently constrains photosynthetic efficiency.
concomitant with a low air flow and an excellent angle relative Another major drawback is the large angle relative to the
to sunlight. direction of sunlight, which causes a high fraction of incident
Table 3. Main Design Features of Closed Photobioreactors

light harvesting degree of land area
reactor type efficiency control required scale-up productivity (g L-1 d-1); species ref
vertical tubular medium medium medium possible 0.5; P. cruentum 27
horizontal tubular good medium poor possible 0.25; S. platensis 26; 30
0.7; Nannochloropsis sp.
helical medium good excellent easy 0.4; S. platensis 14
R-shaped excellent good poor very difficult 18
flat-plate excellent medium good possible 0.85; Nannochloropsis sp. 46; 51
2.15; S. platensis
fermenter type poor excellent excellent difficult 0.03-0.05; several 48
energy to be reflected back and thus lost in terms of biomass sp. (able to grow at 46 °C) was then used, and the maximum
growth purposes (38). temperature could be kept at 44 °C. A few companies have
2.1.2. Horizontal Tubular Reactors. Horizontal tubular adopted similar procedures for their industrial operation. For
reactors (HTR) have been on the rise; gas transfer takes place example, the Aquasearch (currently MERA Pharmaceuticals)
in the tube connection or via a dedicated gas-exchange unit, growth module (42), which is essentially a long HTR for cultiv-
and the angle toward sunlight is particularly adequate for ation of H. pluVialis (a species tolerant to temperature changes),
efficient light harvesting. Such systems can handle large working was cooled down by water spraying whenever necessary.
volumes, because they are not susceptible to contamination. On Gudin and Chaumont (43) have been developing tubular
the other hand, they may generate considerable amounts of heat, reactor configurations since 1984. They reported a reactor
which may attain temperature amplitudes of 20 °C within a system bearing a capacity of 7000 L in all, for which a
single day if (costly) temperature control systems are not productivity of 36 g m-2 d-1 was claimed when using
provided; thus, it will likely pose a problem for regular operation Phorphyridium cruentum (44). This apparatus was made of
(39-41). several 70-L glass VTR, all connected to a gas exchange unit,
A long tubular reactor, 14 cm in diameter, placed horizontally and the culture was mechanically pumped between the light
and made of Plexiglas, was reported by Torzillo (40); said device harvesting and the gas exchange units-whereas the temperature
included a diaphragm pump designed to drive the culture to a was controlled by submerging the culture in a water pool as
feeding tank, and gas transfer was allowed in the tube connec- deemed necessary. A major limitation of this reactor was its
tions. The maximum working volume was 8000 L in a land relatively high cost, coupled with the intrinsic fragility of its
area of 80 m2. However, when the area-to-volume ratio was constitutive glass.
set to that prevailing in open ponds, the photosynthetic efficiency Richmond (17) and Grima (45) reported similar layouts of
did not improve consistently: the maximum productivity HTR, which led to productivities as high as 1.5 g L-1 d-1 with
attained was ca. 0.25 g L-1 d-1 when using S. platensis. The S. platensis and 0.32 g L-1 d-1 with Isochrysis galbana. These
major problem encountered was indeed the control of temper- systems were based on an external unit designed for light
ature; three different methods were therefore tested, viz., shading harvesting, connected to a gas exchange tower (placed on top)
the tubes with dark plastic, overlapping the tubes and water- via an airlift pump (see Figure 2A and B, respectively); the
spraying the surface of the tubes. The latter was shown to be working pressure was set so as to provide a low shear stress
efficient, although assurance of low temperatures required a and a good rate of homogenization, with typical linear velocities
large consumption of water. A thermo-tolerant strain of Spirulina of ca. 50 and 30 cm s-1, respectively. The main difference
between those two configurations pertains to the light harvesting
unit: Grima (45) presented a 2.6 cm × 80.8 m Plexiglas loop
submerged in a thermostatic pool, whereas Richmond (17)
described parallel sets of 3.0 cm × 20 m polycarbonate
transparent tubes connected by a manifold, with temperature
controlled by water spraying. Both of these systems were found
rather efficient and cost-effective. Richmond (17) claimed that
the best way to increase the working volume is to add more
tube sets, in order to maintain an adequate linear velocity and
avoid oxygen buildup. Therefore, scale-up in this case poses in
principle no problems, because the same tower exchange unit
can be duly connected to several extra sets of tubes. The major
drawback is the land area occupied; in fact, the tubes in the
light-harvesting unit are very narrow, so industrial volumes (i.e.,
5 000-10 000 L) can be reached only at the expense of a
paramount number of tubes, which are supposed to be placed
horizontally. Therefore, economic feasibility should be assessed
in advance; otherwise this type of apparatus may not be cost-
effective at all, as discussed elsewhere (46).
Another advance in microalgal cultivation systems is the near-
horizontal tubular reactor (NHTR) designed by Tredici (47) (see
Figure 3), which consists of sets of parallel tubes made of
flexible plastic (typically 6.4 m in length, 43 mm diameter and
Figure 1. Schematic representation of airlift (A) and bubble column 0.15 mm thick), connected by PVC manifolds. The upper
(B) reactors. manifolds are used as degassers, and a perforated pipe inserted
Figure 2. Schematic representation of horizontal tubular reactor with a degassing unit and a light harvesting unit, composed of parallel sets of
tubes (A) or a loop tube (B).
ratio on the one hand, and the requirement of only a small land
area for relatively large volumes on the other. Placing a light
source inside the coil and then providing a reasonable control
of light intensity may compensate for the large angle toward
sunlight. Scale-up of this type of system is easy; one has simply
to increase the number of parallel layers of tubes in the coil,
thus maintaining the hydraulic head essentially unchanged (21).
However, use of a centrifugal pump to drive the culture to the
upper side of the tubing might increase shear stress, which then
would become a limiting step for biomass productivity. There-
fore, not all algal species are suitable for culture in this system;
the recirculation pump may damage some, whereas others may
be engaged in fouling on the inside of the reactor tubing (12).
The system described by Travesio (20) operates with an airlift
column instead of a centrifugal pump, and may be useful toward
reduction of shear stress and consequent minimization of cell
damage.
Morita (51) proposed a conical helical reactor that encom-
Figure 3. Schematic representation of near horizontal tubular reactor. passes a light-harvesting unit composed of PVC tubing (0.16
m of internal diameter) coiled in a conical framework (see Figure
in the lower manifolds is used to inject air in each individual 4B), a degassing unit placed above and a heat exchanger. An
tube. The tubes are placed at an angle of 5-7° relative to the air pump is used to force the culture from the heat exchanger,
horizontal plan. Temperature is controlled by activating water in an ascending trajectory, to the light-harvesting unit, which
spraying onto the reactor when the culture temperature exceeds is in turn connected to the degasser, placed at the same level;
a preset value. The maximum volume tested, 4000 L, obtained the culture is then returned to the heat exchanger unit in a
with a set of 8 parallel tubes 44 m in length, was associated descending trajectory. The same degasser and heat exchanger
with a mean productivity of 0.7 g L-1 d-1 in the case of can be connected to several light-harvesting units, as also tested
Nannochloropsis sp. The main advantage of this system is the by the same author. The main advantage of this system its the
high area-to-volume ratio and its easy scale-up, coupled with high efficiency in light harvesting; the conical shape distributes
the possibility for a higher degree of control (41). Surprisingly, the radiant energy input to a larger photo-receiving area, i.e., it
the temperature control was not efficient, and a day-night improves spatial distribution of light. The heat exchanger has
temperature gap of ca. 14 °C during summer and of ca. 22 °C proven efficient in maintaining temperature within a narrow
during spring was actually recorded. Such a poor control led to interval (28 ( 3 °C). One major disadvantage is that, unlike
different biochemical compositions of the microalga throughout the Biocoil, this arrangement cannot be easily scaled-up. In fact,
the year. However, the major drawback of this system was although the conical light receptor is very efficient, the angle
probably its low gas transfer rate, arising from a large length and height are strictly defined, so the only way to maintain high
and a small diameter (48). photosynthetic rates is to increase the number of light harvesting
2.1.3. Helical Tubular Reactors. Helical tubular reactors units, which also leads to larger energy losses in the complicated
(HeTR) are a suitable alternative to straight TR. The most branches of the flow networks (52); hence the land area
frequently used layout is the Biocoil, initially proposed by productivity will be significantly reduced.
Robinson (49) and currently traded by Biotechna (Melbourne, 2.1.4. R-Shaped Reactors. Lee (28) described a 300-L
Australia). This reactor is composed of a set of polyethylene R-shaped tubular reactor, with sets of 2.5 cm × 25 m transparent
tubes (3.0 cm of inner diameter) coiled in an open circular tubes made of PVC; it used an airlift pump to promote an
framework, coupled with a gas exchange tower and a heat ascending/descending trajectory, with several CO2 injection
exchange system; a centrifugal pump drives the culture broth points along its path (see Figure 5). This configuration presents
through the long tube to the gas exchange tower (see Figure several advantages from the engineering point of view. For
4A). A few authors (12, 18, 20, 21, 50) experimented with and example, the fluid is pumped in a single direction (except in
eventually improved such a design, so it ranks at present among the airlift tubes), so a high flow rate is possible at the expense
the most effective ones as a result of its high area-to-volume of relatively low air supply rates in the rising tubes. The
Figure 4. Schematic representation of helical tubular reactors: Biocoil (A) and conical framework (B).
Figure 5. Schematic representation of R-shaped reactor.
ascending and descending trajectories are placed at a 45° angle sp.; the maximum volumetric productivity, 0.85 g L-1 d-1, was
toward sunlight, so light harvesting is rather efficient. attained with the minimum light path, i.e., 1.3 cm.
2.2. Flat Plate Reactors. Flat plate reactors (FPR) are A different FPR was reported by Iqbal (55), who described
conceptually designed to make efficient use of sunlight; hence, a V-shaped apparatus characterized by unusually interesting
narrow panels are usually built so as to attain high area-to- engineering features, viz., very high mixing rate and very low
volume ratios (see Figure 6A). In the early 1980s, FPR were shear stress (see Figure 6B); scale-up of its reduced capacity (2
considered expensive and were even claimed to exhibit deficien- L) is, however, still to be done.
cies in culture flow control (53). A 500-L FPR was developed Introduction of alveolar panels (see Figure 6C), made of PVC,
by Pulz (16), in which the culture was circulated from an open polycarbonate or polymethyl methacrylate, for microalga cul-
gas exchange unit through several parallel panels placed hori- tivation has meanwhile emerged as a successful concept, because
zontally. The culture flew at a high linear speed (viz., 1.2 m of their high versatility and commercial availability. Several
s-1), but hydrodynamic parameters usually lay in a safe oper- systems using that type of panels have been built and duly tested.
ating range for the sake of cell integrity. The greatest advantage Tredici (56-59) used double-row sets of alveolar plates placed
of this system is its provision of an open gas transfer unit, which horizontally, where culture was circulated in the upper row and
has proven efficient in overcoming the problem of oxygen thermostated water was circulated in the lower row. Tredici (59)
buildup; however, such an open zone restricts effectiveness of also described a bubble column FPR, in which alveolar plates
contamination control, as compared with completely closed were mounted vertically, and the culture was mixed and
reactors. Richmond (54) presented a similar system, composed degassed simply by air bubbling at the bottom of each channel.
of several 200-L units potted together, each unit being composed Its productivity was very high using S. platensis (2.15 g L-1
of 200 cm × 100 cm × 10 cm glass plates. The main difference d-1), when compared with that obtained in open ponds under
was the absence of a gas transfer unit and instead bubbling of similar conditions (0.15 g L-1 d-1).
compressed air at the bottom, through a perforated plastic tube. In general, the main advantages of FPR are their high
A closed system of water spraying was employed to control productivity and uniform distribution of light and, in the specific
temperature; the sprayed water was then collected in troughs case of bubbled column FPR, the absence of a driving pump.
and recirculated through a ventilated water column for refrigera- Furthermore, these reactors can be oriented toward the sun,
tion. One such reactor, with an overall volume of 1000 L, was hence permitting a better efficiency in terms of energy absorbed
tested with various light paths for cultivation of Nannochloropsis from incident sunlight. Pusparaja (60) discussed a reactor
Figure 7. Schematic representation of fermenter-type bioreactor.
Several laboratory-scale FTR have been developed (63-68),

but there is scarce information available on their large-scale
counterparts. The main advantage of these systems is of course
the accurate control of processing parameters, including light,
coupled with the vast experience accumulated in food and
pharmaceutical industries in terms of the scale-up thereof.
Figure 6. Schematic representation of flat panel reactors: flat panel
bubbled in the bottom (A), V-shaped panel (B) and alveolar panel (C). Therefore, should productivity be enhanced, FTR would cer-
tainly become a competitive alternative for industrial manufac-
encompassing an alveolar panel system oriented toward the sun, ture of biochemical products brought about by microalgae.
coupled with an open raceway for gas transfer. The use of such
3. Processing Parameters
alveolar panels as solar receptors increased volumetric produc-
tivity from 0.18 g L-1 d-1 in open ponds to 0.31 g L-1 d-1. Effective supply of light and carbon dioxide to the whole
Although the volumetric productivity attained inside the panels cell culture, as well as other physical and nutritional require-
is higher, open raceways are the most often used cultivation ments, demands uniform dispersion of microalgae in a non-
systems for microalgae, so said combination may be of great limiting nutrient culture medium. Since adequate mixing levels
practical significance. The main disadvantage of alveolar panels in many reactor configurations are often obtained via injection
is oxygen buildup, which arises from the high photosynthetic of gas into the system, mixing and gas transfer are thus closely
activity reached, coupled with the small diameter of the reactors related to each other.
used (57, 59).
2.3. Fermenter-Type Reactors. The least expanded systems 3.1. Gaseous Transfer
for microalga cultivation are conventional fermenter-type reac- 3.1.1. Carbon Source. Because nearly 50% of the whole
tors (FTR) (see Figure 7). These apparata present indeed an microalgal biomass is made up of carbon (69), this element is
intrinsic disadvantage: the area-to-volume ratio is quite low, a major nutrient for cell growth. When grown photo-litotro-
so sunlight harvesting efficiency is poor. To overcome this phycally, all microalgae use inorganic carbon sources to
nuclear drawback, sophisticated systems of internal illumination synthesize organic compounds (70). In aqueous environments,
were developed, which are able to provide a more homogeneous inorganic carbon may exist in several alternative chemical forms,
distribution of light. When possible, microalgae may be het- CO2 (aq), H2CO3, HCO3- and CO32- (71), which are intercon-
erotrophically cultivated in FTR, using appropriate organic vertible via reactions controlled by temperature and pH (69).
carbon sources. A 250-L FTR was built by Pohl (61) with The possibility that microalgae use carbon in CO2 form only
stainless steel and illuminated internally by fluorescent lamps or that they also take up the HCO3- and CO32- forms is not a
placed inside narrow glass (or Plexiglas) tubes; CO2-enriched critical issue, because reactions that interconvert CO2, H2CO3,
air was bubbled at the bottom, through a V-shaped (i.e., low HCO3- and CO32- in soluble form are sufficiently fast not to
shear stress) stirrer. Such a system was operated batch-, be limiting steps in CO2 demand by cells (14, 71). Besides,
semicontinuous- and continuous-wise. The operation parameters e.g., for Chlorella Vulgaris, inorganic carbon affinity at the cell
could be fully controlled, so axenic cultures were maintained surface was found to be very low (not above 1 µg carbon/L)
for long periods, as considered crucial for production of certain (71), and synthesis of carbonic anhydrase (which catalyzes the
high-value metabolites. Although biomass productivity was quite aforementioned inorganic carbon interconversion) is enhanced
low (typically 30-50 mg L-1 d-1), there is still plenty of room when cells are exposed to a low-CO2 environment.
for enhancement of growth parameters with this reactor con- Detailed studies on the influence of the carbon source upon
figuration. microalga productivity (71) have indicated that, although HCO3-
Ogobona (62) has also developed an FTR using both sun and is easily absorbed by cells, it is a poor source of carbon when
artificial light, which may be a leap forward in terms of compared with CO2. In fact, it is possible to achieve a linear
reduction of operating costs. response in microalgal carbon biomass with mass input of
Table 4. Values for KL and KLa Using Several Gas Transfer Methods
transfer methoda KLa (min-1) A (m2) a (m-1) KL (m/s) ref
HIFO 1.48 × 10-2 1.7 1700 1.45 ×10-7 50
HIFI 1.33 × 10-2 1.4 × 10-1 140 1.59 × 10-6 50
HIFO (3.6-7.5) × 10-3 2.38 × 10-3 4.76 (1.26-2.64) × 10-5 62
bubbling 7.00 × 10-3 1.1 × 10-3 1.05 1.11 × 10-4 50
bubbling (7.59-21.7) × 10-2 nab 2.15-6.21 (5.83-5.88) × 10-4 60
bubbling (9-94) × 10-2 nab 42.8-147.2 (3.51-10.6) × 10-5 61
a HIFO, hollow hydrophobic fiber membrane; HIFI, hollow hydrophilic fiber membrane. b Not available.
carbon (which corresponds to an efficiency of virtually 100%) transfer resistance dominates overall resistance, the rate of mass
only if limited inputs of inorganic carbon and narrow pH ranges transfer of CO2 (NCO2) is approximately given by
are permitted; e.g., Phaeodactylum tricornutum produced up to
25 mg d-1 of algal carbon when pH of the culture was below NCO2 ) kLa(CCO2L* - CCO2L) (1)
9.0 (71). Beyond a given threshold, pH control becomes difficult,
so chemical precipitation of salts containing CO32-, OH- and where kL is the liquid-phase mass transfer coefficient, a is the
PO43- will likely occur, hence leading to chemical deterioration specific area available for mass transfer, CCO2L* is the concen-
of the medium and possible cell injury. The upper threshold of tration of CO2 in the culture medium that would equilibrate its
productivity is therefore considerably lower than that resulting actual partial pressure on the gas side, and CCO2L is the
from light limitation, and about 1/10 that obtained by replacing concentration of CO2 in the bulk of culture medium.
HCO3- by CO2. A general consensus exists about preference The lumped parameter kLa characterizes the CO2 mass
of microalgae for CO2 (as inorganic carbon source), as it is easily transfer capability of the reactor and thus determines whether a
controlled and produces minor pH changes. Note that CO2 in certain reactor will be able to sustain a given rate of cell growth.
the open air accounts for only ca. 0.03% (v/v) (69), so fluxes Regarding transfer of CO2, said parameter is of the utmost
of carbon transfer to the culture are small, even in the presence importance in design, scale-up and operation steps for a biomass
of extended interface areas or enhanced mixing. Consequently, culture system (72). Some properties of the bubbles affect
CO2-enriched air is the most commonly employed nutrient gas mainly kL, whereas others affect mainly a. Furthermore, the
mixture, so as to force light to become the sole limiting factor. amount of metabolites present in the medium depends on cell
3.1.2. Transport Process. When CO2 is injected at a given concentration, which will also affect kLa since, in general,
point in a culture, a concentration gradient builds up as it is metabolites modify the surface tension of the medium and thus
consumed by cells and/or lost to the atmosphere. According to act as an extra barrier to mass transfer (73).
the two-film theory, mass transfer of CO2 from the gas phase Studies (73, 74) have been undertaken to compare kLa values
to the cell phase occurs through sequential stages: transport obtained under bubbling (at different gas and liquid flow rates
from the bulk of the gas to the thin gaseous film at the immediate and at distinct bubble sizes) and under diffusion through hollow-
vicinity of the interface; diffusion through this gas film; transport fiber membranes (of different areas and distinct construction
across the gas/liquid interface; diffusion through the adjacent materials); kLa typically increases with increasing liquid tan-
liquid film; transport from the thin liquid film to the bulk of gential flow rate, because of thinning of the liquid boundary
the liquid; transport from the bulk of the liquid to the thin liquid layer. When comparing the rate of mass transfer obtained under
film at the immediate vicinity of the cell wall; diffusion through bubbling or by diffusion (with all other operating conditions
the outer cell liquid film; and finally, metabolic uptake by the remaining similar) (75), it was observed that kLa values were
cell. The overall resistance over the entire path distance can be higher in the latter case: an overall volumetric coefficient for
calculated by adding up the aforementioned single resistances, CO2 transfer of 1.48 × 10-2 min-1 was found for a hydrophobic
as they occur in series; however, most of the resistance actually membrane, 1.33 × 10-2 min-1 for a hydrophilic membrane and
lies on the liquid film. In an efficiently stirred bioreactor, 7.0 × 10-3 min-1 for plain bubbling (see Table 4). Such
concentration gradients within the bulk liquid are marginal, so increasingly lower values are a consequence of the substantially
(even in the presence of dense microalgal cultures) resistance shorter interfacial area available; in fact, the individual volu-
within the gas bubbles is larger than at the boundary layer of metric mass transfer coefficient values (kL) were higher for
cells (14). Consequently, resistance raised by the liquid film bubbling (1.11 × 10-2 m s-1) than for diffusion (1.45 × 10-7
adjacent to the interface essentially limits the rate of CO2 and 1.59 × 10-6 m s-1 for hydrophobic and hydrophilic
transfer. In fact, comparative studies (14) encompassing overall membranes, respectively), because of the turbulent conditions
mass-transfer resistances and gas/liquid mass-transfer resistance associated with the former. Therefore, use of microporous
revealed their similarity in order of magnitude, thus confirming hollow fibers instead of plain bubbling offers technological
that CO2 transport is mainly controlled by resistance offered enhancements in effectiveness of mass transfer at the expense
by the liquid film. of larger a. In addition, this type of system offers opportunity
The rate of CO2 uptake by cells eventually determines the to recirculate gas and hence permits lower gas pressures to be
rate at which it will be transferred to the medium, if steady- used, which in turn reduce operating costs. However, Carvalho
state conditions prevail. The rate of mass transfer is, in general, and Malcata (75) have demonstrated that overall metabolic
proportional to the driving force for said transfer (expressed in enhancement in microalgal growth is essentially nil, probably
terms of a concentration difference involving one actual bulk because CO2- and light-rich periods did not coincide with each
concentration of one phase and an equivalent bulk concentration other (both in space and time).
of the other phase) and the area available for transfer. The Other studies (71) pertaining to continuous cultures limited
proportionally coefficient is the sum of the reciprocals of all in carbon afforded interesting relationships between bubble size,
resistances to transfer and is usually denoted as an overall mass- gas flow rate and partial pressure of CO2. Optimum biomass
transfer coefficient. In the present case, since liquid-phase mass productivity was obtained by using either high bubbling rate
Table 5. Generic Description of Methods for Gas Transfer in Photobioreactors

feature
gas hydrody-
process type of reactor alga ref transfer mixing namic stress scale-up
Passive Mode
surface unstirred shallow D. salina 5 poor very poor very low difficult
aeration ponds
membrane hollow-fiber Nannochloropsis sp. 50 excellent uniforma medium medium
transfer membrane
Active Mode
gas injection at bottom of air-lift P. tricornutum 79 high uniform low medium
into culture reactor bubble columnb P. tricornutum 33 fair fair low medium
stirred-tank filamentous 48 low/high uniform high difficult
at given point flat-plate Spirulina 45 high uniform low/high difficult
of reactor tubular Spirulina 11 fair/high uniform low/high difficult
gas tubular solar P. cruentum 27 good good (by medium difficult
exchanger receptor pumping)
a Depends on mixing device. b Large bags.
(with small sized bubbles) with low inlet pressure of CO2 or mixed ponds of ca. 1 ha; or (ii) via membranes, through which
low bubbling rate with high inlet pressure of CO2 (irrespective gas diffuses into the culture. No examples have been reported
of bubble size). Although more efficient (47% vs 14%, in terms pertaining to surface aeration in closed systems, probably owing
of assimilation efficiency), the former option could bring about to the huge areas that would be involved. Conversely, membrane-
problems of cell flotation and consequent washout. At 1% (v/ mediated transfer has already been tested in closed systems
v) CO2, productivity was essentially independent of bubble size, (74-76).
probably because partial pressure in the culture was already high When membrane aeration is considered, gas diffuses through
enough; it is indeed possible to achieve light limitation under a permeable membrane, which can be either microporous or
these conditions. In conclusion, total carbon input flux is made of a material possessing high gas permeability (e.g.,
determined by both gas bubbling rate and partial pressure of silicone). The membrane is often arranged as a coil or bank of
CO2. Selecting the proper combination of these two variables tubes, placed inside the reactor vessel (77). Alternatively, the
is the key to avoid carbon limitation in intensive culturing. bank of tubes may be contained in an appropriate housing, which
In terms of biomass produced, the differences in value will work as a gas exchanger, connected to the reactor vessel
between the several parameters affecting it make an overall by plastic tubing (75). One configuration tested with success
comparison difficult to establish. Concerning production costs, encompasses diffusion of pure CO2 through permeable silicone
the literature is scarce; hence, reliable economic data are to be tubing, coiled up in order to maximize transfer area in a reduced
generated in the future, before a consistent comparison of space (76). This kind of system theoretically offers several
performances can be accurately done. advantages when compared with bubbling, viz., exclusion of
3.1.3. CO2 Transfer Systems. In algal mass culture systems, CO2 losses to atmosphere, possibility to accurately control CO2
it is important to obtain a reliable prediction of CO2 transfer transfer rates, and no requirement for air/CO2 mixing chamber
rates for accurate design, scale-up and operation. The occurrence or even highly pure CO2 (as it will not be in close contact with
of chemical reactions between CO2 and OH-, H2O and NH3 in the culture). Better overall efficiencies are expected for this
the liquid phase may lead to enhanced rates of CO2 absorption system when compared with bubbling (13-20%) or gas
by the culture medium (72); therefore, the CO2 exchange flux exchange (25-65%) (76). However, such a system suffers from
from the gaseous to liquid phase is governed not only by severe drawbacks: (i) since the transfer rate is proportional to
diffusion kinetics, but also (depending on the relative magnitude the membrane area, long tubing membranes are normally
of the reaction rates observed) by kinetics of the reactions in required, which raise the investment costs (21); (ii) especially
the liquid film at the vicinity of the interface. for highly salted media, high inner pressures are necessary in
The critical concentration of CO2 necessary for optimal order to balance transfer rates produced by bubbling, which force
growth of a particular microalga cannot be stated in general, as the use of thick membrane walls (the only ones that can handle
it strongly depends on the delivery system implemented in the such high pressures); and (iii) pressurization of the membrane
culture vessel. Since transfer of CO2 occurs through the interface often promotes expansion of the material, thus originating
between the gaseous mixture and the liquid medium culture, microspacing within the polymer network of the membrane, to
two main processes to increase such an interface area can be which bacteria may adhere and eventually grow, hence reducing
devised: (i) passive mode, where extensive gas/culture interface contact area between membrane and culture, and decreasing
areas are used and gas diffuses into the culture; and (ii) active transfer rate throughout time.
mode, where use of an extra apparatus for aeration either by An alternative possibility relies on the potential of mi-
injecting the gas into the medium or spraying the medium into croporous hollow-fiber membranes, i.e., bundles of polymeric
the gas forces expansion of the contact area between gas and porous fibers, potted to inlet/outlet ports in their ends and
culture (69) (see Table 5). contained in adequate housings (74, 75) (see Figure 8B), which
Regarding passive mode, large interface areas between the may also function as gas exchangers in the setup or as the whole
gas and the culture medium can be obtained by: either (i) using reactor itself. Commercial fiber modules usually contain tiny
large open-air ponds, where gas exchanges occur by surface hollow fibers, with typical inner diameters of ca. 250 µm; their
driven aeration (see Figure 8A), e.g., cultivation of D. salina permeability obviously depends on the construction material.
in Australia, in natural lakes of 50-300 ha, or similar cultivation As a result of the usually large number of fibers inside each
of Chlorella and Spirulina spp. in paddle-wheel (or circular) module, the ratio between membrane exposed area and external
Figure 8. Schematic representation of major methods of gas exchange in microalgal reactors: surface driven aeration (A), microporous hollow-
fiber membranes (B), airlift loop (C), bubble column (D), stirrer blade bubbling (E), and gas exchanger system (F).
volume of the module is typically high. Although such an favorably, pressures are upperly constrained by mechanical
indirect way of supplying CO2 by diffusion has similarities with limits of the equipment (typically 40 psi). Hence, flux enhance-
the process described above involving silicone tubing (76), this ment usually takes advantage of increases in recirculation flow
system configuration makes it possible to use lower gas pres- rate, whereas inlet pressure is maintained close to the maximum
sures, as no need to counterbalance hydrostatic heads exists permitted. Use of hollow fiber devices has been reported for
(75). culture of mammalian cells (78), in culture of the microalga
According to Fick’s law, mass transfer flux increases with Chlamydomonas reinhardtii (79), in continuous culturing of the
transmembrane pressure and recirculation flow rate. Although microalga P. tricornutum in seawater (80), and in culture of
rises in either of these operating parameters would contribute the cyanobacteria Anabaena Variabilis (81).
In regard to active mode, two main transfer processes are adequate sampling time. The use of a model to predict future
usually considered: (i) gas mixture is circulated together with behavior based on on-off signals helps to anticipate (and account
culture, either injected at a certain point of the system (riser for) the delay associated with a cycle time, taking also into
tube) (16, 17, 28, 41, 45, 53, 54, 82-90) or bubbled at the account the on-off nature of the control signal (95). Other types
bottom of the reactor vessel (29-32, 34-36, 46, 51, 52, 55, of mathematical models can predict the photosynthesis rate
56, 61-63, 91, 92); or (ii) gas is inserted into the culture when (based on O2 generation rate and CO2 consumption rate) as a
it passes through a gas exchanger (44, 50) (see Table 5). function of solar irradiance (89). Such models also permit
Bubbling of CO2-enriched air at the bottom of the reactor is accurate determination of the CO2 requirements of the system,
the most frequent mode of operation; it is usually performed thus allowing optimized use of this resource as well.
via sintered porous stones or pipes where tiny holes were 3.1.4. O2 RemoWal. As stressed above, light intensity is
previously drilled, although gas can also be bubbled from the regarded as an essential issue to be addressed in microalgal
perforated blades of a stirrer (see Figure 8C-E). For larger systems; however, it is not the only potentially limiting factor
reactor vessels, holes are usually drilled in the upper part of a in photosynthesis, which is chemically described by
long, sealed pipe, which is placed longitudinally at the bottom
energy + CO2 + H2O T sugar + O2 (2)
of the reactor. Such parameters as bubble size, gas flow rate
and CO2 pressure can then be adjusted to meet the requirements
In fact, if oxygen build-up occurs (i.e., when the concentration
of each specific culture (72). In order to avoid microalgal
of dissolved oxygen in the culture is above its counterpart in
accumulation in certain (stagnant) areas of the bottom region,
equilibrium with its partial pressure in the atmosphere), the
the gas stream can be injected from the sides rather than on the
aforementioned reversible reaction is shifted to the left, thus
axis at the bottom, coupled with continuous stirring of the
decreasing photosynthetic efficiency; in general, oxygen con-
bottom with a magnetic stirrer (69). Unfortunately, two major
centrations above air saturation inhibit photosynthesis in mi-
drawbacks are often observed: (i) obstruction of the gas transfer
croalgae (46). Besides, accumulation of O2 in the liquid culture
devices by biofouling, which requires frequent stops for
medium is one of the most difficult problems to overcome (58),
cleaning; and (ii) losses of CO2 to the atmosphere, owing to
because it may become toxic above a certain threshold; dissolved
the low residence times of the gas in the culture (which increase
oxygen concentrations above 35 mg/L, which are toxic to most
considerably operation costs, as CO2 is an expensive utility)
microalgae, were reached in outdoor cultures of Spirulina (70).
(21). Losses to the atmosphere can be reduced via installation
This problem has not been reported in the literature pertaining
of a covered sparging-diffusion system, consisting of a transpar-
to most types of small-scale culture devices, probably because
ent material tightened to a frame mounted below the waterline
the gas transfer systems employed were sufficiently powerful
(69); nevertheless, it is usually difficult to reach efficiencies
to provide enough carbon dioxide and promptly remove the
above 10% (69). By introducing the gas directly into the circuit
oxygen concomitantly formed. However, when larger systems
or via interconnected gas exchangers (see Figure 8F), it is
are implemented, the magnitude of said problem increases
technically possible to saturate the culture suspension. In the
exponentially; it has even been claimed (46) to be responsible
former situation, the gas is introduced into the circulation path for failures in a commercial device constituted by several-
and is distributed over one or more parts of the reactor (usually kilometer-long piping.
of tubular shape) through tiny holes. When using gas exchang- Accumulation of photosynthetically generated oxygen be-
ers, CO2 utilization is, in principle, rather efficient because comes a particularly serious problem in high area-to-volume
inflowing algal suspension is dispersed with a baffle plate and ratio closed photobioreactors operated outdoors; among these,
flows down the walls of the cylinder as a thin film, through HTR are described as particularly problematic. Most solutions
which gas exchange can rapidly occur (69). However, this proposed to date rely on use of a degasser (or gas exchange
configuration is effective, in terms of adequate gas supply, only unit), where dissolved oxygen can be released (17, 20, 51, 53,
in small reactor systems. In fact, although levels of ca. 30 ppm 83, 87, 89, 90); however, to attain an effective separation
of CO2 on the cell surface are considered sufficient to maintain between gas and liquid, the distance between entrance and exit
unlimited photosynthesis (69), experimental evidence has shown of the degasser should be such that the smallest bubbles will
that, after a 100-m path, microalgae have already used all gas have a sufficient time to disengage from the liquid before it
available, so further gas needs to be supplied (93). leaves the unit. In closed tubular photobioreactors, in which
Increasing awareness of the importance of CO2 led to axial gradients in concentration of gases can develop, connec-
development of control systems, able to regulate the pH of the tions between the several tubes can also incorporate a narrow
culture and thus, indirectly, control the amount of CO2 supplied. tube for oxygen degassing (40). Other authors (41) tested the
The most common system employed for pH control is the on- performance of a NHTR, essentially consisting of a layer of
off type, in which CO2 is injected into the culture when pH is tubes arranged in parallel and connected by two manifolds: the
above a desired setpoint. However, the dynamics of the system lower was used to inject air into the culture, whereas the higher
(e.g., mixing times) do not allow a fast response to disturbances. acted as a degasser; however, productivities were lower than
An alternative is the use of model-based predictive control expected, probably due to build-up of oxygen tension during
(MPC), which relies on a model able to predict the process periods of high light intensity. Especially when the exhausting
output at future time points (horizon)-by calculating the control gas is recirculated, accumulation of oxygen may be avoided by
sequence that minimizes a certain objective function and bubbling the exhaust gas through a sodium sulfite solution prior
implementing a receding strategy so that at each sampling instant to its return to the reactor (11).
the horizon is shifted toward the future. When classical on-off It should be emphasized that the aforementioned devices for
control was replaced by MPC (taking solar irradiance effect into oxygen degassing are only necessary when the reactor config-
account), carbon losses in tubular reactors with P. tricornutum uration does not provide an interface between culture and
were reduced from 19.7% to 5.5% (94). The main advantage surrounding atmosphere, e.g., reactors with tubular or flat panel
of using on-off MPC instead of a classical on-off controller is configuration. In stirred tanks and in the various bubbling-type
the predictive capability of the former and the selection of an reactors available, oxygen leaves the culture when it reaches
Table 6. Generic Description of Methods for Mixing in Photo-Bioreactors

feature
mixing gas hydrody-
process type of reactor microalga ref efficiency transfer namic stress scale-up
Pumping Mode
centrifugal system with
more than
one unit
rotary positive displacement fair low medium easy
peristaltic C. reinhardtii 66
diaphragm S. platensis 29
lobe
Mechanical Stirring Mode
stirring with two or more blades cylindrical tank filamentous 48 uniform fair/high high medium
Gas Mixing Mode
injection of gas air-lift I. galbana 54 uniform high low medium
into culture bubbling I. galbana 20 fair fair low medium
the surface, hence allowing simultaneous degassing of the whole is random, i.e., not all algal cells are submitted to a uniform
culture. To overcome this incompatibility between a high area- pattern of movement into and out of the illuminated area of the
to-volume ratio (which characterizes tubular and flat panel reactor. Furthermore, mixing should be supplied to proper
reactor configurations) and efficient gas transfer (of both oxygen extents, because very low values promote settling and hence
and carbon dioxide), a vertical alveolar panel was devised (56), emergence of dead zones (especially toward the bottom of the
in which air with extra carbon dioxide is sparged at the bottom reactor) where anaerobic conditions prevail, thus leading to cell
plates and oxygen bubbles leave the liquid when they reach deterioration and consequent decrease of productivity; in the
the top. limit, toxic compounds will form that might compromise
Since the efficiencies of all techniques used to date to provide viability of the whole culture (69, 70). Inadequate mixing also
oxygen removal from microalgal cultures are still not at a permits clumping of cells into aggregates of varying sizes, hence
satisfactory level, a competitive mode of operation (especially leading to development of a three-phase system (gas/liquid/solid)
in large culture systems) consists of using bubbling devices, inside the reactor, which is prone to decreasing mass transfer
because they eliminate the need for (the still ineffective) rates (6). On the other hand, high mixing rates may lead to shear-
degassing devices. Another alternative relies on use of hollow- induced injury of cells (22, 23), which hamper their viability.
fiber membrane apparata, as they seem to lead to lower dissolved The most important mechanism of cell damage in sparged
oxygen concentrations than bubbling systems do (74). The use cultures agitated at low intensities is bubble break-up at the
of several small-sized reactors instead of one larger unit may liquid-gas interface (i.e., microalgal cells are captured by the
also help alleviate this problem. Nevertheless, further research rising gas bubbles and carried to the surface, where they die as
should be conducted in order to accurately match the amount bubbles collapse). At higher agitation intensities, cell damage
of CO2 supplied to the actual requirement of the growing caused by fluid-mechanical stress acquires a greater significance
microalga, and the amount of O2 removed to the actual amount as the specific aeration rate decreases (97).
of O2 produced. The major mixing/recirculation systems commercially used
In view of the above, one concludes that there is no are described in Table 6. They can be tentatively divided in
universally optimum gas transfer device; the choice of the most pumping (used when more than one vessel is present), mechan-
suitable one depends on the microalga considered and the ical stirring (usually present when only one vessel is used) and
objective function preselected. gas mixing (which takes advantage of a gas, usually CO2-
3.2. Mixing. After fulfillment of nutritional growth require- enriched air, injected in the culture to promote turbulent mixing
ments and assurance that environmental factors are non-limiting, and recirculation through the various vessels). Combinations
mixing becomes a critical parameter. In reactor configurations of these systems are also possible. Former recirculation systems
that consist of more than one vessel (e.g., a carbonation tower used exclusively pumps, either centrifugal, positive displace-
for gas exchange, coupled to one or several tubes for light ment, peristaltic or diaphragm ones. As research progressed, it
supply), adequate mixing is particularly required in order to was soon realized that cultures recirculated with distinct
recirculate the culture between vessels. Even in the case of single pumping apparata attained different maximum specific growth
vessel reactors, efficient mixing should be provided in order to rates, and that the adverse effects caused by centrifugal and
produce a uniform dispersion of microalgae within the culture rotary positive displacement pumps were essentially proportional
medium, thus eliminating gradients of light, nutrient concentra- to their rotation speeds (38). The extent of damage observed in
tion (including CO2) and temperature; note that thermal various pumps was accordingly measured as a function of the
amplitudes of up to 8 °C have been recorded between the top number of passes of algal suspension through the pumps in a
and bottom of unmixed cultures, leading to irreversible damage closed loop, and the concept of hydrodynamic stress was thus
(96). Finally, the effect of “mutual shading”, i.e., continuous introduced, which provided a framework to assess parameters
cell movement from and to dark/light zones, has been claimed that determine magnitude of shear stress phenomena: geometry
(69) to be essential to guarantee high biomass productivity. In of the reactor, type of pump involved, and morphological and
a dense culture, the region where microalgae receive enough physiological conditions of the microalga (22, 24, 25).
light for photosynthesis can be quite shallow (2-5 cm) (69), Gas mixing systems, i.e., bubble column systems, cause less
so vigorous mixing is necessary to provide cells with a uniform extensive damage to fragile microalgal species than mechanical
average exposure to light. pumping does. This is especially the case of air-lift units, in
Current mixing techniques are often insufficient to attain the which mixing is achieved by fluid flow obtained from sparging
aforementioned requirements, because the induced turbulence air into a central draught tube (riser), where it decreases bulk
liquid density hence causing the liquid to rise. The liquid then where r and ro are the inner and outer radius of the tubes,
flows downward through the outer tube, thus creating a natural respectively, provided that the tubes hold no gaps between them.
circulation. Although these systems appear to cause the least Applying the well-studied microbial-growth energy equation,
extensive degree of cell damage (21), they are not completely after the reasoning detailed by Pirt (83), one is led to
devoid of shear stress: a cell-damaging hydrodynamic effect
has been reported (46) in bubble columns and airlift reactors, µ
q) +m (4)
which was associated with so intense turbulence patterns that YG
the length scale of the fluid microeddies approached cellular
dimension. Barbosa (26) reported bubble formation at the where q is the specific rate of light uptake, µ is the specific
sparger as the main event leading to cell death. Furthermore, rate of growth, YG is the maximum growth yield on light, and
increasing cell density leads to extra difficulties in producing m is the maintenance coefficient. The value of q may, in turn,
homogeneous cultures, so the need to provide mechanical be calculated via
agitation arises (associated with lower speeds, so as to keep
shear stress to a minimum) (98). Subsequent studies on the φ‚I0‚AV
influence of increasing power supply on decreasing of mi- q) (5)
X
croeddy length scale demonstrated that this scale was always
much higher than cell size, and therefore shear stress rather than where φ is the fraction of photosynthetically available light, Io
turbulence was responsible for the existence of stress phenomena is the total incident irradiance, and X is the biomass concentra-
(27). Extra stirrers may provide such a mechanical agitation, tion.
with one or more blades. Another alternative lies in the Combining eqs 4 and 5 and assuming a light-limited culture
introduction of baffles along the culture path, so as to create a with a maintenance coefficient virtually nil, one obtains the
controlled turbulence pattern. Degen (99) used horizontal baffles biomass output rate (µ‚X) as
to subdivide the riser on a flat panel airlift photo-bioreactor,
thus creating a defined circulation path and exposing the cells µ‚X ) φ‚Io‚YG‚AV (6)
to intermittent light, hence creating a flashing-light effect.
When the reactor configuration corresponds to only one In a reactor operated under a constant AV and a given Io, it is
vessel, the microalgal suspension can be agitated by means of possible to adjust µ and X (eq 6) so as to achieve the maximum
a stirrer, which usually consists of a central tube with two or biomass yield (YG), which then also leads to the maximum PE
more blades. In order to enhance mixing without damaging cells, (83).
a hybrid configuration has also been described (69) in which a If a proper reactor design is sought, it is crucial to obtain the
stream of CO2-enriched air is passed via the central tube into maximum biomass yield at the expense of the light actually
the blades, from where it penetrates the medium through available. The latter parameter depends on the reactor config-
capillary holes. This mixing device is the most usual in uration and location. If one considers reactors illuminated by
cylindrical bioreactors. direct sunlight, the maximum Io depends on the latitude of the
As distinct microalgae have different sensitivities to shear, facility, so it can be controlled only by changing the reactor
there is no single optimum mixing system for all microalgal inclination toward sunlight (28). Several reactors have indeed
species; the microorganism in use will again be the determinant been developed or tuned accordingly (59). On the other hand,
issue when evaluating mechanical and hydrodynamic shear in artificially illuminated reactors, the maximum Io depends on
effects that are admissible during regular operation of the photo- the nature, intensity and relative position of the light source,
bioreactor. all of which can be accurately manipulated and controlled.
3.3. Light Requirement. Light is the basic energy source Most reactors are normally designed so as to exhibit a high
for photo-autotrophic organisms, as microalgae are; hence, their AV; it is thus possible to work at higher biomass concentration
light harvesting efficiency is of crucial importance for bioreactor for a given yield (YG). Good examples of reactors designed to
engineering encompassing those microrganisms. The photosyn- harvest the maximum light are the alveolar panel reactor by
thetically active radiance (PAR) is normally assumed to be ca. Tredici (59), and the tubular reactors by Grima (45), Richmond
43-45% in the wavelength range 400-700 nm (100, 101). The (17) and Zittelli (41). Reactors that combine natural and arti-
photosynthetic efficiency (PE) is defined as the fraction of ficial lights have also been described (3, 62). In terms of
available light that is eventually stored as chemical energy in artificially illuminated reactors, the need of small diameters to
biomass; however, there is no general consensus on how to increase AV can be circumvented via provision of internal
calculate its actual value. Note that PE is a theoretical upper illumination. The higher biomass yield (YG) can be even more
limit. In practice, especially in long-term cultures, PE is normally important in the case of artificially illuminated reactors, as the
below 6.5%, and under optimal conditions, only 40% of the light provided represents a cost itself that adds to the overall
theoretical maximum yields can typically be reached (86). running costs of the process. Nevertheless, such cost may be
Hence, increase in PE is likely the best approach in attempts to kept limited, as the latest few years have witnessed the
enhance microalga productivity, which is a sine qua non development of methods for in situ growth monitoring-from
condition to guarantee commercial competitiveness of the flow injection analysis systems based on turbidimetric measure-
associated bioprocess. ments (68) to techniques based on real-time monitoring and
In order to estimate PE for a selected species in a given control of biomass by using light transmittance sensors (102).
reactor, one departs from calculation of the illuminated surface The light energy available for each microalga cell (Iav)
area per unit volume of the culture, AV; in the case of a VTR depends on several factors, e.g., Io and biomass concentration.
(probably the most popular reactor configuration), one has (53) Several authors (103) have demonstrated that increasing light
intensity increases growth rate, but only until a saturation point
ro is reached. Above this threshold, increasing light produces no
AV ) 2 2 (3) positive effect on growth rate, hence generating a dramatic
π‚r decrease in YG, as a major fraction of light cannot be used by
Table 7. Generic Description of Methods for Light Supply in Photo-Bioreactors

feature
mode cost intensity control land area required reactor design productivity
sunlight small problematic high must be designed for high depends on weather
surface/volume ratio and latitude
artificial light high excellent small allows great flexibility constant
the culture. Furthermore, the light will influence not only overcome the extra investment cost required, and thus become
biomass productivity, but also its biochemical profile (15, 19, profitable and fully competitive.
45, 64-67). To attain maximum productivity, several parameters are to
An interesting method to assess Iav on tubular reactors, based be controlled, viz., sufficient nutrient level and optimum
on Io and X, was developed by Evers (104) and later applied by temperature and pH (topics not covered here, owing to limitation
Grima (64). This model assumes that light attenuation caused of space), as well as adequate mixing and turbulence, coupled
by mutual shading is dependent on Beer-Lambert law, which with provision for oxygen degassing and light control (which
calculates the light available at each single point at a given were the focus of this specific review). From the many reactor
distance from the surface, followed by integration on both the configurations built to date (viz., HTR, VTR, HeTR, FPR and
cross sectional and the outer surface areas. Use of this model FTR), none is able to effectively control all of those parameters
allows one to fit Io in order to work below the saturation point, simultaneously. Some are indeed optimized for better sun-
hence avoiding a decrease in YG and/or an adjustment of Iav for harvesting capacity, but scale-up is not suitable; conversely,
a more consistent biochemical profile. others can be better controlled, but sun-harvesting efficiency
Sunlight and artificial illumination data are compared in Table becomes poor.
7. The source of artificial illumination is of crucial importance, On the other hand, since continuous supply of CO2 to
because the wavelength spectrum should be specific in order microalgal cultures is expensive, it is necessary to provide
to take full advantage of the total Io made available. As seen devices that supply it in a discontinuous fashion. In fact, during
before, PAR lies within the wavelength range 400-700 nm, so the first few hours (or days, depending on the duration of the
a lamp with a vast spectrum and with its maximum radiation in lag phase) following inoculation of a batch culture, there is a
that range is normally chosen. However, different lamps will large amount of CO2 dispersed in the medium; during the
generate distinct spectra; on the other hand, each microalga exponential growth phase, essentially all CO2 supplied is used
species possesses its own absorption optimum, so each indi- up; and finally, in the stationary phase, the rate of utilization of
vidual case must be studied per se. You and Barnett (105) related CO2 declines again. The supply of CO2 to the medium is in
a dependence of the exponential growth rates on the energy of general controlled via monitoring pH; nevertheless, such
radiation and spectrum of light, and concluded that blue light underlying relationship may lead to erroneous values due to
(400-500 nm) increased cell growth and polysaccharide the influence of other dissolved ions on pH. Consequently, it
production in P. cruentum. would be more correct to measure CO2 directly with a probe or
The method to evaluate the efficiency of conversion, de- to choose another parameter closely related therewith. Maxi-
scribed by Simmer (106), takes into account the photon flux mization of CO2 transfer may proceed via hollow-fiber mem-
(P) emitted by the lamp in the PAR wavelength, the specific branes, which enhances mass transfer and decreases costs, by
absorption coefficient of each species and the quantum capacity, recirculating unused gas and using lower gas pressures.
hence allowing one to estimate the efficiency of conversion In scale-up of culture devices, some of the conditions that
according to were easily controlled at a small scale may become difficult
problems. One such situation concerns degassing of oxygen,
ERF ) ∫ P(λ)dλ (7) which is believed to be responsible for failure of commercial
devices. There is not a single, straightforward solution, and a
feasible approach consists of using several small units for growth
where ERF is the emitted radiant flux, which is a function of
rather than one larger unit. In this way, the various units may
the wavelength, λ.
function separately (depending on the production needs), and
the gaseous transfers will be more efficient, while minimizing
4. Conclusions
O2 build-up. However, economic studies should be performed,
It is our belief that long-term trends of research in the field in order to adequately evaluate the feasibility of this solution.
of microalgae should encompass design and development of When using hollow-fiber membrane devices to bring about
reactor systems, as well as microalgal strains able to increase exchange of gases within the culture, the content of dissolved
specific metabolite productivity. oxygen was found to be lower than with the bubbling system,
Whereas open raceway ponds are presently the primary so it may constitute a suitable alternative in terms of gaseous
systems used in commercial production of microalgal biomass changes. Another approach relies on use of bubbling devices,
outdoors, there is not a single recommended design in what as they eliminate the need for (the still ineffective) degassing
concerns closed systems. Closed systems are indeed more devices.
complex than their open counterparts, thus requiring improved Concerning light intensity, operational issues are also not clear
technological skills to operate them and higher investment costs cut. On the one hand, use of sunlight is cheaper but the light
to house them; otherwise, they possess the advantage of more cycle cannot be controlled, which often precludes higher biomass
accurate control of various parameters during regular operation, productivities. On the other hand, artificially illuminated reactors
hence enhancing productivities. are typically expensive. The use of optical fibers (contaminated
The main challenge to the microalga research community now with impurities that will lead to gradual dispersion of light) that
appears to lie on the design of a closed system so effective in can introduce light inside reactors with large volumes and
biomass and metabolite productivity that it will eventually effective monitoring of the light available for photosynthesis
will probably constitute a breakthrough-that will eventually (11) Acién-Fernández, F. G.; Férnandez-Sevilla, J. M.; Egorova-
turn closed reactors into fully competitive options. Zachernyuk, T. A.; Molina-Grima, E. Cost-effective production of
13
In the past few years, studies on photo-bioreactor sensing C, 15N stable isotope-labelled biomass from phototrophic microal-
gae for various biotechnological applications. Biomol. Eng. 2005,
have also appeared, which presented innovative approaches to 22, 193-200.
assess photobioreactor states, hence allowing estimation of such (12) Borowitzka, M. A. Commercial production of microalgae: ponds,
parameters as biomass density, specific growth rate, dissolved tanks, tubes and fermenters. J. Biotechnol. 1999, 70, 313-321.
oxygen concentration, photosynthetic efficiency and average (13) Becker, C. C.; Kyle, D. J. Developing functional foods containing
light intensity (107). Such type of control is expected to help algal docosahexaenoic acid. Food Technol. 1998, 52, 68-71.
in decreasing operation costs. In addition, the continuous reuse (14) Markl, H. CO2 transport and photosynthetic productivity of a
of biomass for harvesting of high-value compounds from continuous culture of algae. Biotechnol. Bioeng. 1977, 19, 1851-
microalgae (“milking of microalgae”) has also been addressed 1862.
as a potential technology of production (108). Although such a (15) Grima, E. M.; Fernández, F. G. A.; Camacho, F. G.; Chisti, Y.
Photobioreactors: light regime, mass transfer, and scaleup. J.
method has been employed for the production of β-carotene Biotechnol. 1999, 70, 231-247.
from D. salina, it is not applicable to all strains or products. (16) Pulz, O. Open-air and semi-closed cultivation systems for the mass
Studies on the physiological behavior of cells, particularly cell cultivation of microalgae. In First Asia-Pacific Conference on Algal
membranes (among others), are still necessary before developing Biotechnology, Kuala Lumpur, Malaysia, 1992.
a generic “milking” process. (17) Richmond, A.; Boussiba, S.; Vonshak, A.; Kopel, R. A new
Finally, the intrinsic characteristics of the microalgae selected tubular reactor for mass production of microalgae outdoors. J. Appl.
for use should also guarantee technological suitability. Selection Phycol. 1993, 5, 327-332.
of strains should thus emphasize such features as specific growth (18) Chrismada, T.; Borowitzka, M. A. Growth and lipid production
of Phaeodactylum tricornutum in a tubular photobioreactor. In Algal
rate, tendency to clump and settle, ability to foam, adherence Biotechnology in the Asia-Pacific Region, Kuala Lumpur, Malaysia,
to container walls, and shear and temperature sensitivities. In 1994.
addition, existing species should be carefully screened in order (19) Grima, E. M.; Perez, J. A. S.; Camacho, F. G.; Fernandez, F. G.
to select the most performant strains, which hold the capacity A.; Sevilla, J. M. F.; Sanz, F. V. Effect of dilution rate on
to overproduce specific end products with a market value. The eicosapentaenoic acid productivity of Phaeodactylum tricornutum
performance of those species could be even reinforced via Utex-640 in outdoor chemostat culture. Biotechnol. Lett. 1994, 16,
1035-1040.
genetic engineering, using tools already available.
(20) Travesio, L.; Hall, D. O.; Rao, K. K.; Benı́tez, F.; Sánchez, E.;
Borja, R. A helical tubular reactor producing Spirulina in a
Acknowledgment semicontinuous mode. Int. Biodeterg. Biodegrad. 2001, 47, 151-
Mrs. Gisela Oliveira is hereby gratefully acknowledged for 155.
her kind help in bibliographic search. Financial support for (21) Borowitzka, M. A. Closed algal photobioreactors: design con-
A.P.C. (grant BD/2838/93-IF) and L.A.M. (grant BD/15890/ siderations for large-scale systems. J. Mar. Biotechnol. 1996, 4, 185-
98-FCT) by PRAXIS XXI (Portugal) is also to be noted. Partial 191.
funding was obtained from a project grant (MICROPESCA/ (22) Gudin, C.; Chaumont, D. Cell fragility-the key problem of
0072/04), issued by program MARE (Portugal). microalgae mass production on closed photobioreactors. Bioresour.
Technol. 1991, 38, 145-151.
References and Notes (23) Thomas, W. H.; Gibson, C. H. Effects of small-scale turbulence
on microalgae. J. Appl. Phycol. 1990, 2, 71-77.
(1) Soeder, C. J. A historical outline of applied algology. In Handbook (24) Jaouen, P.; Vandanjon, L.; Quéméneur, F. The shear stress of
of Microalgal Mass Culture; Richmond, A., Ed.; CRC Press: Boca microalgal cell suspensions (Tetraselmis suecica) in tangential flow
Raton, FL, 1986; pp 25-41. filtration systems: the role of pumps. Bioresour. Technol. 1999, 68,
(2) Burlew, J. Current status of the large-scale culture of algae. In Algal 149-154.
Culture: From Laboratory to Pilot Plant; Burlew, J. S., Ed.; (25) Vandanjon, L.; Rossignol, N.; Jaouen, P.; Robert, J. M.; Quémé-
Carnegie Institution of Washington Publications: Washington, DC, neur, F. Effects of shear on two microalgae species. Contribution
1976; pp 3-23. of pumps and valves in tangential flow filtration systems. Biotechnol.
(3) Chaumont, D. Biotechnology of algal biomass production: a review Bioeng. 1999, 63, 1-9.
of systems for outdoor mass culture. J. Appl. Phycol. 1993, 5, 593- (26) Barbosa, M. J.; Hadiyanto; Wijffels, R. H. Overcoming shear stress
604. of microalgae cultures in sparged photobioreactors. Biotechnol.
(4) Borowitzka, M. A. Microalgae for aquaculture: Opportunities and Bioeng. 2004, 85, 78-85.
constraints. J. Appl. Phycol. 1997, 9, 393-401.
(27) Alı́as, C. B.; López, M. C. G. M.; Fernández, F. G. A.; Sevilla,
(5) Yamaguchi, K. Recent advances in microalgal bioscience in Japan,
J. M. F.; Sánchez, J. L. G.; Grima, E. M. Influence of power supply
with special reference to utilization of biomass and metabolites: A
in the feasibility of Phaeodactylum tricornutum cultures. Biotechnol.
review. J. Appl. Phycol. 1996, 8, 487-502.
Bioeng. 2004, 87, 723-733.
(6) Panda, A. K.; Mishra, S.; Bisaria, V. S.; Bhojwani, S. S. Plant cell
reactors-a perspective. Enzyme Microb. Technol. 1989, 11, (28) Lee, Y. K.; Ding, S. Y.; Low, C. S.; Chang, Y. C. Design and
386-397. performance of an R-type tubular photobioreactor for mass cultiva-
(7) Wang, G.; Chen, H.; Li, G.; Chen, L.; Li, D.; Hu, C.; Chen, K.; tion of microalgae. J. Appl. Phycol. 1995, 7, 47-51.
Liu, Y. Population growth and physiological characteristics of (29) Laing, I.; Jones, E. A turbidostat vessel for the continuous culture
microalgae in a miniaturized bioreactor during space flight. Acta of marine microalgae. Aquacult. Eng. 1988, 7, 89-96.
Astronautica 2006, 58, 264-269. (30) Trotta, P. A simple and inexpensive system for continuous
(8) Travieso, L.; Pellón, A.; Benı́tez, F.; Sánchez, E.; Borja, R.; O’ monoxenic mass culture of marine microalgae. Aquaculture 1981,
Farrill, N.; Weiland, P. BIOALGA reactor: preliminary studies for 22, 383-387.
heavy metal removal. Biochem. Eng. J. 2002, 12, 87-91. (31) Cohen, E.; Arad, S. M. A closed system for outdoor cultivation
(9) Scragg, A. H.; Illman, A. M.; Carden, A.; Shales, S. W. Growth of Porphyridium. Biomass 1989, 18, 59-67.
of microalgae with increased calorific values in a tubular bioreactor. (32) Martı́nez-Jerónimo, F.; Espinosa-Chavez, F. A laboratory-scale
Biomass Bioenergy 2002, 23, 67-73. system for mass culture of freshwater microalgae in polyethylene
(10) Nagase, H.; Yoshihara, K.; Eguchi, K.; Okamoto, Y.; Murasaki, bags. J. Appl. Phycol. 1994, 6, 423-425.
S.; Yamashita, R.; Hirata, K.; Miyamoto, K. Uptake pathway and (33) Tredici, M. R.; Rodolfi, L. University of Florence, Italy. Reactor
continuous removal of nitric oxide from flue gas using microalgae. for industrial culture of photosynthetic micro-organisms. PCT WO
Biochem. Eng. J. 2001, 7, 241-246. 2004/074423 A2, 2004.
(34) Chae, S. R.; Hwang, E. J.; Shin, H. S. Single cell protein (55) Iqbal, M.; Grey, D.; Stepan-Sarkissian, F.; Fowler, M. W. A flat-
production of Euglena gracilis and carbon dioxide fixation in an sided photobioreactor for continuous culturing microalgae. Aquac-
innovative photo-bioreactor. Bioresour. Technol. 2006, 97, 322- ulture Eng. 1993, 12, 183-190.
329. (56) Tredici, M. R.; Carlozzi, P.; Zittelli, G. C.; Materassi, R. A vertical
(35) Miyamoto, K.; Wable, O.; Benemann, J. R. Vertical tubular reactor alveolar panel (VAP) for outdoor mass cultivation of microalgae
for microalgae cultivation. Biotechnol. Lett. 1988, 10, 703-708. and cyanobacteria. Bioresour. Technol. 1991, 38, 153-159.
(36) James, C. M.; al-Khars, A. M. An intensive continuous culture (57) Tredici, M. R.; Materassi, R. From open ponds to vertical alveolar
system using tubular photobioreactors for producing microalgae. panels: the Italian experience in the development of reactors for
Aquaculture 1990, 87, 381-393. the mass cultivation of phototrophic microorganisms. J. Appl. Phycol.
(37) Fukami, K.; Nishimura, S.; Ogusa, M.; Asada, M.; Nishijima, T. 1992, 4, 221-231.
Continuous culture with deep seawater of a benthic food diatom (58) Tredici, M. R.; Zittelli, G. C.; Biagiolini, S.; Materassi, R. Novel
Nitzchia sp. Hydrobiology 1997, 358, 245-249. photobioreactor for the mass cultivation of Spirulina spp. Bull. Inst.
(38) Lee, Y. K. Enclosed bioreactors for the mass cultivation of Oceanogr. 1993, 89-96.
photosynthetic microorganisms: the future trend. TIBTECH 1986, (59) Tredici M. R. Bioreactors, photo. In Encyclopedia of Bioprocess
July, 186-189. Technology: fermentation, biocatalysis and bioseparation; Flick-
(39) Richmond, A. The challenge confronting industrial microagricul- inger, M.C., Drew, S.W., Eds.; Wiley: New York, 1999; Vol 1, pp
ture: high photosynthetic efficiency in large-scale reactors. Hydro- 395-419.
biology 1987, 151/152, 117-121. (60) Puspararaj, B.; Pelosi, E.; Tredici, M. R.; Pinzani, E.; Materassi,
(40) Torzillo, G.; Puspararaj, B.; Bocci, F.; Balloni, W.; Materassi, R. An integrated culture system for outdoor production of microalgae
R.; Florenzano, G. Production of Spirulina biomass in closed and cyanobacteria. J. Appl. Phycol. 1997, 9, 113-119.
photobioreactors. Biomass 1986, 11, 61-74. (61) Pohl, P.; Kohlhase, M.; Martin, M. Photobioreactors for the axenic
(41) Zittelli, G. C.; Lavista, F.; Bastianini, A.; Rodolfi, L.; Vincenzini, mass cultivation of microalgae. In Algal Biotechnology; Stadler, T.,
M.; Tredici, M. R. Production of eicosapentaenoic acid by Nan- Mollion, J., Verdus, M.-C., Karamanos, Y., Morvan, H., Christiaen,
nochloropsis sp. cultures in outdoor tubular photobioreactors. J. D., Eds.; Elsevier: New York, 1988; pp 209-218.
Biotechnol. 1999, 70, 299-312. (62) Ogbonna, J. C.; Soejima, T.; Tanaka, H. An integrated solar and
artificial light system for internal illumination of photobioreactors.
(42) Olaizola, M. Commercial production of astaxanthin from Heama-
J. Biotechnol. 1999, 70, 289-297.
tococcus pluVialis using 25,000 liter outdoor photobioreactors. J.
Appl. Phycol. 2000, 12, 499-506. (63) Eriksen, N. T.; Geest, T.; Iversen, J. J. L. Phototrophic growth in
the lumostat: a photo-bioreactor with on-line optimization of light
(43) Gudin, C.; Chaumont, D. Solar biotechnology study and develop-
intensity. J. Appl. Phycol. 1996, 8, 345-352.
ment of tubular solar receptors for controlled production of
(64) Grima, E. M.; Perez, J. A. S.; Camacho, F. G.; Sanchez, J. L. G.;
photosynthetic cellular biomass for methane production and specific
Alonso, D. L. n-3 PUFA productivity in chemostat cultures of
exocellular biomass. Sol. Energy R&D Eur. Community, Ser. E 1984,
microalgae. Appl. Microbiol. Biotechnol. 1993, 38, 599-605.
5, 184-193.
(65) Grima, E. M.; Camacho, F. G.; Perez, J. A. S.; Sanchez, J. L. G.
(44) Chaumont, D.; Thepenier, C.; Gudin, C.; Junjas, C. Scaling up a
Biochemical productivity and fatty acid profiles of Isochrysis galbana
tubular photobioreactor for continuous culture of Porphyridium
Parke and Tetraselmis sp. as a function of incident light intensity.
cruentum from laboratory to pilot plant (1981-1987). In Algal
Process Biochem. 1994, 29, 119-126.
Biotechnology; Stadler, T., Mollion, J., Verdus, M.-C., Karamanos,
(66) Grima, E. M.; Perez, J. A. S.; Camacho, F. G.; Sevilla, J. M. F.;
Y., Morvan, H., Christiaen, D., Eds.; Elsevier: New York, 1988;
Fernandez, F. G. A. Effect of growth-rate on the eicosapentaenoic
pp 199-208.
acid and docosahexaenoic acid content of Isochrysis galbana in
(45) Grima, E. M.; Pérez, J. A. S.; Camacho, F. G.; Sánchez, J. L. G.; chemostat culture. Appl. Microbiol. Biotechnol. 1994, 41, 23-27.
Fernández, F. G. A.; Alonso, D. L. Outdoor cultivation of Isochrysis
(67) Grima, E. M.; Camacho, E. G.; Perez, J. A. S.; Fernandez, E. G.
galbana ALII-4 in a closed tubular photobioreactor. J. Biotechnol.
A.; Sevilla, J. M. F. Growth yield determination in a chemostat
1994, 37, 159-166.
culture of the marine microalga Isochrysis galbana. J. Appl. Phycol.
(46) Mirón, A. S.; Gómez, A. C.; Camacho, F. G.; Grima, E. M.; Chisti, 1996, 8, 529-534.
Y. Comparative evaluation of compact photobioreactors for large- (68) Meireles, L. A.; Azevedo, J. L.; Cunha, J. P.; Malcata, F. X. On-
scale monoculture of microalgae. J. Biotechnol. 1999, 70, 249- line determination of biomass in a microalga bioreactor using a novel
270. computerized flow injection analysis system. Biotechnol. Prog. 2002,
(47) Tredici, M. R.; Zittelli, G. C. Efficiency of sunlight utilization: 18, 1387-1391.
tubular versus flat photobioreactors. Biotechnol. Bioeng. 1998, 57, (69) Becker, E. W. Large-scale cultivation. In Microalgae: Biotech-
187-197. nology and Microbiology; Becker, E. W., Ed.; Cambridge University
(48) Tredici, M. R. Closed photobioreactors: basic and applied aspects. Press: New York, 1994; pp 63-171.
In Proceedings of Marine Biotechnology: Basics and Applications, (70) Richmond, A. Technological aspects of mass cultivation-a general
Matalascañas, Spain, 2003; p 1. outline. In CRC Handbook of Microalgal Mass Culture; Richmond,
(49) Robinson L. F.; Morrison A. W.; Bamforth M. R. Improvements A., Ed.; CRC Press: Boca Raton, FL, 1986; pp 245-264.
relating to biosynthesis. European Patent 261,872, 1988. (71) Goldman, J. C.; Dennett, M. R.; Riley, C. B. Inorganic carbon
(50) Chrismada, T.; Borowitzka, M. A. Effect of cell density and sources and biomass regulation in intensive microalgal cultures.
irradiance on growth, proximate composition and eicosapentaenoic Biotechnol. Bioeng. 1981, 23, 995-1014.
acid production of Phaeodactylum tricornutm grown in a tubular (72) Talbot, P.; Gortares, M. P.; Lencki, R. W.; de la Noue, J.
photobioreactor. J. Appl. Phycol. 1994, 6, 67-74. Absorption of CO2 in algal mass culture systems: a different
(51) Morita, M.; Watanable, Y.; Saiki, H. Investigation of photobiore- characterization approach. Biotechnol. Bioeng. 1991, 37, 834-842.
actor design for enhancing the photosynthetic productivity of (73) Grima, E. M.; Sanchez-Perez, J. A.; Garcia-Camacho, F.; Robles-
microalgae. Biotechnol. Bioeng. 2000, 69, 693-698. Medina, A. Gas-liquid transfer of atmospheric CO2 in microalgal
(52) Morita, M.; Watanable, Y.; Okawa, T.; Saiki, H. Photosynthetic cultures. Chem. Technol. Biotechnol. 1993, 56, 329-337.
productivity of conical helical tubular photobioreactors incorporating (74) Ferreira, B. S.; Fernandes, H. L.; Reis, A.; Mateus, M. Mi-
Chlorella sp. under various culture medium flow conditions. croporous hollow fibres for carbon dioxide absorption: mass transfer
Biotechnol. Bioeng. 2001, 74, 135-144. model fitting and the supplying of carbon dioxide to microalgal
(53) Pirt, S. J.; Lee, Y. K.; Walach, M. R.; Pirt, M. W.; Balyuzi, H. cultures. Chem. Technol. Biotechnol. 1998, 71, 61-70.
H. M.; Bazin, M. J. A tubular bioreactor for photosynthetic (75) Carvalho, A. P.; Malcata, F. X. Transfer of carbon dioxide within
production of biomass from carbon-dioxide-design and performance. cultures of microalgae: plain bubbling versus hollow-fiber modules.
J. Chem. Technol. Biotechnol. B 1983, 33, 35-58. Biotechnol. Prog. 2001, 17, 265-272.
(54) Richmond, A.; Cheng-Wu, Z. Optimization of a flat plate glass (76) Lee, Y.-K.; Hing, H.-K. Supplying CO2 to photosynthetic algal
reactor for mass production of Nannochloropsis sp. outdoors. J. cultures by diffusion through gas-permeable membranes. Appl.
Biotechnol. 2001, 85, 259-269. Microbiol. Biotechnol. 1989, 31, 298-301.
(77) Aunins, J. G.; Henzler, H. Aeration in cell culture bioreactors. In production of microalgae. Arch. Hydrobiol. Beih. Ergebn. Limnol.
Biotechnology; Stephanopoulos, G., Ed.; Wiley-VCH: Weinheim, 1978, 11, 254-258.
1993; Vol. 3 (Bioprocessing), p 223. (94) Sánchez, J. L. G.; Berenguel, M.; Rodrı́guez, F.; Sevilla, J. M.
(78) Gallagher, S. L.; Tharakan, J. T.; Chau, P. C. An intercalated- F.; Alias, C. B.; Fernández, F. G. A. Minimization of carbon losses
spiral wound hollow fiber bioreactor for the culture of mammalian in pilot-scale outdoor photobioreactors by model-based predictive
cells. Biotechnol. Tech. 1987, 1, 91-96. control. Biotechnol. Bioeng. 2003, 84, 533-543.
(79) Chen, F.; Johns, M. R. A strategy for high cell density culture of (95) Berenguel, M.; Rodrı́guez, F.; Acién, F. G.; Garcı́a, J. L. Model
heterotrophic microalgae with inhibitory substrates. J. Appl. Phycol. predictive control of pH in tubular photobioreactors. J. Process
1995, 7, 43-46. Control 2004, 14, 377-387.
(80) Marsot, P.; Cembella, A. D.; Mouhri, K. Croissance de la biomasse
azotée du Phaeodactylum tricornutum (Bacillariophyceae) en culture (96) Oswald, W. J. Large-scale algal culture systems (engineering
discontinue dialysante et non-dialysante. Can. J. Microbiol. 1992, aspects). In Micro-Algal Biotechnology; Borowitzka, M. A., Borow-
38, 945-952. itzka, L. J., Eds.; Cambridge University Press: New York, 1988;
(81) Markov, S. A.; Bazin, M. J.; Hall, D. O. Hydrogen photoproduc- pp 357-394.
tion and carbon dioxide uptake by immobilized Anabaena Variabilis (97) Camacho, F. G.; Gómez, A. C.; Sobczuk, T. M.; Grima, E. M.
in a hollow-fiber photobioreactor. Enzyme Microb. Technol. 1995, Effects of mechanical and hydrodynamic stress in agitated, sparged
17, 306-310. cultures of Porphyridium cruentum. Process Biochem. 2000, 35,
(82) Pirt, S. J.; Panikov, N.; Lee, Y.-K. The miniloop: a small-scale 1045-1050.
air-lift microbial culture vessel and photobiological reactor. J. Chem. (98) Singh, G. Reactor design for plant cell culture of food ingredients
Technol. Biotechnol. 1979, 29, 437-441. and additives. Food Technol. 1997, 51, 62-66.
(83) Pirt, S. J. Microbial photosynthesis in the harnessing of solar- (99) Degen, J.; Uebele, A.; Retze, A.; Schmid-Staiger, U.; Trösch, W.
energy. J. Chem. Technol. Biotechnol. 1982, 32, 198-202. A novel airlift photobioreactor with baffles for improved light
(84) Laws, E. A.; Terry, K. L.; Wickman, J.; Chalup, M. S. A simple utilization through the flashing light effect. J. Biotechnol. 2001, 92,
algal production system designed to utilize the flashing light effect. 89-94.
Biotechnol. Bioeng. 1983, 25, 2319-2335. (100) Laws, E. A.; Taguchi, S.; Hirata, J.; Pang, L. Continued studies
(85) Laws, E. A.; Taguchi, S.; Harata, J.; Pang, L. High algal of high algal productivities in a shallow flume. Biomass 1987, 11,
production rates achieved in a shallow outdoor flume. Biotechnol. 39-50.
Bioeng. 1986, 28, 191-197.
(86) Laws, E. A.; Taguchi, S.; Harata, J.; Pang, L. Optimization of (101) Pirt, S. J.; Lee, Y. K.; Richmond, A.; Watts-Pirt, M. The
microalgal production in a shallow outdoor flume. Biotechnol. photosynthetic efficiency of Chlorella biomass growth with reference
Bioeng. 1988, 32, 140-147. to solar energy utilization. J. Chem. Technol. Biotechnol. 1980, 30,
(87) Lee, C.-G.; Palsson, B. Ø. High-density algal photobioreactors 25-34.
using light-emitting diodes. Biotechnol. Bioeng. 1994, 44, 1161- (102) Sandnes, J. M.; Ringstad, T.; Wenner, D.; Heyerdahl, P. H.;
1167. Källqvist, T.; Gisler¢d, H. R. Real-time monitoring and automatic
(88) Muller-Feuga, A.; Guédes, R. L.; Hervé, A.; Durand, P. Com- density control of large-scale microalgal cultures using near infrared
parison of artificial light photobioreactors and other production (NIR) optical density sensors. J. Biotechnol. 2006, 122, 209-215.
systems using Porphyridium cruentum. J. Appl. Phycol. 1998, 10, (103) Dubinsky, Z.; Matsukawa, R.; Karube, I. Photobiological aspects
83-90. of algal mass culture. J. Mar. Biotechnol. 1985, 2, 61-65.
(89) Fuentes, M. M. R.; Sánchez, J. L. G.; Sevilla, J. M. F.; Fernández, (104) Evers, E. G. A model for light-limited continuous cultures-
F. G. A.; Pérez, J. A. S.; Grima, E. M. Outdoor continuous culture growth, shading, and maintenance. Biotechnol. Bioeng. 1991, 38,
of Phorphyridium cruentum in a tubular photobioreactor: quantitative 254-259.
analysis of the daily cyclic variation of culture parameters. J. (105) You, T.; Barnett, S. M. Effect of light quality on production of
Biotechnol. 1999, 70, 271-288. extracellular polysaccharides and growth rate of Porphyridium
(90) Fernández, A. F. G.; Sevilla, J. M. F.; Pérez, J. A. S.; Molina- cruentum. Biochem. Eng. J. 2004, 19, 251-258.
Grima, E.; Chisti, Y. Airlift-driven external-loop tubular photobiore-
actors for outdoor production of microalgae: assessment of design (106) Simmer, J.; Tichy, V.; Doucha, J. What kind of lamp for the
and performance. Chem. Eng. Sci. 2001, 56, 2721-2732. cultivation of algae? J. Appl. Phycol. 1994, 6, 309-313.
(91) Marsot, P.; Fournier, R.; Blais, C. Culture à dialyse: emploi de (107) Li, J.; Xu, N. S.; Su, W. W. Online estimation of stirred-tank
fibres creuses dialysantes pour la culture massive de phytoplankton. microalgal photobioreactor cultures based on dissolved oxygen
Can. J. Fish. Aquat. Sci. 1981, 38, 905-911. measurement. Biochem. Eng. J. 2003, 14, 51-65.
(92) Camacho, F. G.; Gómez, A. C.; Fernández, F. G. A.; Sevilla, J. (108) Hejazi, M. A.; Wijffels, R. H. Milking of microalgae. TIBTECH
F.; Grima, E. M. Use of concentric-tube airlift photobioreactors for 2004, 22, 189-194.
microalgal outdoor mass cultures. Enzyme Microb. Technol. 1999,
24, 164-172. Received March 7, 2006. Accepted August 3, 2006.
(93) Heussler, P.; Castillo, S. J.; Merino, M. F.; Vasquez, V. V.
Improvements in pond construction and CO2 supply for the mass BP060065R
17. December Intensive

Copepod production
2013
Aquaculture Research, 2000, 31, 703±711
The elusive copepods: their production and

suitability in marine aquaculture
J G Stùttrup
Danish Institute for Fisheries Research, Department for Fish Biology, North Sea Centre, DK-9850 Hirtshals, Denmark
Correspondence: J G Stùttrup, Danish Institute for Fisheries Research, Department of Fish Biology, North Sea Centre, Box 101,
DK-9850 Hirtshals, Denmark
production systems for copepods, their nutritional

Abstract
value as live prey for ®sh and their present and
Despite the fact that the suitability of copepods as potential use in aquaculture.
live prey for marine ®sh larvae is now well
established, their use in aquaculture remains
Introduction
sporadic. Although of lower nutritional value,
the relative ease of production of rotifers Rearing larvae of most marine ®sh species, ranging
(Brachionus spp.) and Artemia nauplii continues from the coldwater species, such as cod Gadus
to ensure their predominance. Studies in the morhua (L.) and halibut Hippoglossus hippoglossus
literature have highlighted differences in the (L.), to more temperate species, such as turbot
levels and ratios of fatty acids, lipid classes and Scophthalmus maximus (L.), and warmwater species,
pigments between copepods and traditional live such as sea bass Dicentrarchus labrax (L.) and
prey. Such differences may have important seabream Sparus auratus (L.), requires the provision
consequences for ®sh larval nutrition. The con- of live prey for a variable period from the onset of
sequences of poor nutrition during ®sh larval exogenous feeding. A common feature of these
development may be obvious, for example defor- species is the production of small pelagic eggs. The
mities or malpigmentation, but in many cases larvae generally hatch at an early stage in their
may be obscure, as in affects on temperature development; many lack functional organs at the
tolerance or growth during later life stages. In time of hatching, and they have relatively small
some aquaculture systems, copepods are cultured yolk reserves. The onset of exogenous feeding is
in large quantities in outdoor, extensive or semi- concurrent with the development of the digestive
intensive units. Intensive-rearing systems for system as well as the development of organs critical
copepods require further development. However, for successful feeding, such as vision and motor
it is now established that intensive rearing on development.
monoalgal diets does not result in severe dete- In intensive-rearing systems, ®rst-feeding marine
rioration of the nutritional value of the copepods, ®sh larvae are fed on rotifers Brachionus spp., with a
at least in terms of their highly unsaturated fatty range in lorica size from 100 mm to 340 mm
acid (HUFA) content. In some cases, the provision (Fukusho 1989), succeeded during the later larval
of copepods for a short period of time during the stages by the larger sized nauplii of the brine shrimp
larval stage is suf®cient to ensure normal Artemia spp. However, some ®sh larvae, such as
development. However, it remains to be demon- halibut and plaice, are large enough at ®rst feeding
strated that the extra costs involved result in to attack and ingest Artemia. The brine shrimp,
increased pro®ts as a result of higher percentages which can be purchased as eggs, come from a large
of normally pigmented juvenile ®sh, improved variety of strains, varying in size and in energetic
growth and survival and reduced incidence of and nutritional content. Newly hatched Artemia
disease. This paper reviews the advances in nauplii measure 422 mm to 517 mm in length
# 2000 Blackwell Science Ltd, 703

Copepods in marine aquaculture J G Stùttrup Aquaculture Research, 2000, 31, 703±711
(LeÂger, Bengtson, Simpson & Sorgeloos 1986). lagoons. The bags are initially ®lled with ®ltered
Rotifers and Artemia do not ful®l all the known (80 mm) sea water. Copepods are collected from the
nutritional requirements of marine ®sh larvae and lagoon and concentrated within a speci®c size range
must be ènriched' before feeding. Emulsions, before being transferred to the bags for feeding to
powders or microcapsules are commercially avail- the ®sh larvae. To provide a speci®c prey size range,
able for enrichment and, together with improved double ®lters such as those described by van der
culture techniques, the rearing of many marine ®sh Meeren & Naas (1997) are used. For ®rst-feeding
larvae using rotifers and Artemia has been success- larvae, mesh sizes of 80 and 200 mm are typically
ful. Despite these improvements, external signs of used. Copepods often dominate in this size fraction,
nutritional de®ciencies in reared marine ®sh are but other organisms such as rotifers also occur. The
common, e.g. abnormal pigmentation in turbot and Norwegian lagoon enclosure system therefore re-
in marbled sole Limanda yokohamae (Kanazawa presents a semi-intensive system. Van der Meeren &
1991) and halibut (McEvoy, Naess, Bell & Lie Naas (1997) provided a review of extensive and
1998). Rearing ®sh larvae on natural zooplankton semi-intensive rearing techniques including a
produced higher rates of normal pigmentation description of Maximus A/S, a turbot hatchery in
(Nñss, Germain-Henry & Naas 1995; McEvoy et al. Denmark, which produces around half a million
1998) and improved resistance to stress (Kraul, turbot juveniles annually. This system is described
Brittain, Cantress, Nagao, Ako, Ogasawara & in more detail by Stùttrup, Shields, Gillespie, Gara,
Kitagawa 1993). Sargent, Bell, Henderson, Tocher, Sutherland,
Nutritional de®ciency at the larval stage can lead Naess, Mangor Jensen, Naas, van der Meeren,
to less obvious effects. In experiments with sole Solea Harboe, Sanchez, Sorgeloos, Dhert & Fitzgerald
solea, tolerance to low temperatures during the (1998). It comprises six outdoor ponds 12 m in
juvenile stage was shown to be related to the larval diameter and 6 m high. These are kept empty
diet (enriched vs. non-enriched Artemia) (Howell during the winter and ®lled with ®ltered saltwater
1994). However, these de®ciencies were not detect- before the introduction of ®rst-feeding ®sh larvae.
able during or at the end of the larval stage, using During this period, natural phyto- and zooplankton
parameters such as growth and survival, commonly are monitored daily and, if necessary, agricultural
used to measure the success of different diets. fertilizers may be applied to enhance production.
Several studies have shown that rearing marine The objective is to time the introduction of ®rst-
®sh on natural zooplankton can ameliorate these feeding larvae with an increasing production of
nutritional de®ciencies. However, the provision of zooplankton nauplii and an ample supply of algae.
copepods is not as convenient as producing rotifers One pond is reserved exclusively for copepod
or Artemia. Extensive reviews on the production of production, which can be used as a supplement to
harpacticoid and calanoid copepods are provided by the ®sh-rearing tanks if necessary. The ®rst-feeding
Chandler (1986) and Ikeda (1973) and on the use ®sh larvae feed primarily on nauplii within a limited
of different live food for ®sh larval rearing by Nellen size range but, as they grow older, they feed on a
(1981). This paper reviews developments in ex- broader size range (van der Meeren 1991; Cunha &
tensive rearing systems based on natural zooplank- Planas 1995). It is important when monitoring the
ton and in larger scale production techniques for system to consider the number of food organisms
copepods for intensive ®sh-rearing systems. The within the appropriate size range in relation to the
advantages and disadvantages of using copepods for size of the larvae. This is particularly important
®sh larval rearing are discussed, and conclusions on during ®rst feeding when the prey size range is
the state of the art and prospects for future limited by the mouth gape of the larvae. As the ®sh
development in this area presented. larvae grow, their energetic demand may exceed the
number of available food organisms within such
systems, and Artemia nauplii may be added to
Copepods produced in extensive
maintain high growth rates (J Larsen, pers. com-
systems
mun.).
The use of copepods is widespread in extensive ®sh Planktonic copepods commonly found in these
larval rearing systems such as Norwegian lagoon systems include calanoid genera such as Acartia,
enclosures, in which the larvae are reared on Centropages and Temora, but harpacticoid copepods
natural zooplankton or in plastic bags placed in also proliferate and are frequently found in the
704 # 2000 Blackwell Science Ltd,, Aquaculture Research, 31, 703±711

Aquaculture Research, 2000, 31, 703±711 Copepods in marine aquaculture J G Stùttrup
stomachs of ®rst-feeding ®sh in extensive systems limited duration (weeks to months). The different
(van der Meeren 1991). Because of their benthic culture apparatus, culture medium and food used
habit, harpacticoids are more dif®cult to monitor. re¯ect the diverse life and feeding forms of copepods.
Examples of extensive copepod cultures are listed in Consequently, development of this ®eld has been
Table 1. In addition, Fukusho (1980) described a fragmentary, and progress towards mass-culture
semi-intensive system, in which Tigriopus japonicus techniques has been slow.
were reared in co-culture with rotifers in large A basic knowledge of the physiological processes
outdoor tanks on monoalgal cultures and enriched and population dynamics of a species is a prerequi-
yeast. site for the development of rearing techniques.
In extensive or semi-intensive systems, the Extensive cultivation of copepods has been shown
practice of using copepods harvested directly from to be biologically and economically feasible and has
the natural environment or allowed to bloom in been adopted in commercial hatcheries. However,
natural ponds may lead to a risk of parasitic scaling-up small-scale intensive techniques is com-
contamination. Some copepods are known to be plicated, and the development of economically
intermediate hosts for several teleost parasites, viable, large-scale culture systems for calanoid or
such as trematodes, which have a compulsory harpacticoid copepods remains to be demonstrated.
mollusc host and a de®nitive teleost host Ikeda (1973) gave up his quest to ®nd the
(Marcogliese 1995). Cestodes may also be found perfect aquaculture live food that could satisfy
in copepods. The cestode Bothriocephalus gregarius, several simultaneous requirements, such as prey
which infects turbot Psetta maxima has been size, optimal temperature, growth rate and
identi®ed in three species of copepods (Robert & fecundity. Certainly, the ®rst attribute to consider
Gabrion 1991). Four out of ®ve identi®ed is the body size of the adult and nauplii copepods
parasites claimed as the cause for 90% of in relation to the mouth size of the ®sh species to
mortality in halibut larvae were parasites with be reared. Species such as the calanoid Acartia
copepods as intermediate hosts (Bristow 1990). tonsa and the harpacticoid Tisbe holothuriae have
Being intermediate hosts between compulsory relatively small ®rst-stage nauplii, measuring
hosts, the risk of infection diminishes if the adult around 145 mm and 100 mm respectively (J G
copepods are collected and allowed to breed in a Stùttrup, unpubl. data). Reared within their
holding tank. The nauplii are then collected and optimal temperature range (A. tonsa 16±18 °C:
used as food for ®sh larvae. T. holothuriae 18 °C), these species have genera-
tion times lasting around 20 and 12 days
respectively. Larger sized species, such as
Copepods produced in intensive
Calanus ®nmarchicus, have also been considered
systems
for cultivation (J G Stùttrup, unpubl. data). This
In intensive hatchery systems, manipulated brood- species has ®rst-stage nauplii measuring 200 mm
stocks provide ®sh eggs all year round. Because of in length. However, because C. ®nmarchicus is a
the seasonality in production of zooplankton in higher latitude species with lower temperature
extensive systems in temperate areas, a regular optima, the culture was maintained at 15 °C,
supply of copepods cannot be guaranteed. Sporadic resulting in slower growth compared with A.
attempts have been made to culture zooplankton tonsa and T. holothuriae. Marcus & Alatalo (1989)
intensively to avoid such problems, but the state of measured a generation time of 4 weeks at 14 °C.
the art is in its infancy. Most research on copepods This is unpractical for culture purposes, as it is
has focused on their nutrition, reproduction or necessary to supplement the adult cultures
physiological aspects, and culture techniques have regularly. It is therefore important to consider
primarily been developed at the laboratory scale to how the temperature tolerance of the copepod
provide small numbers of organisms needed for species interacts with their growth rate and
experimental work. This is re¯ected in the overviews fecundity.
by Chandler (1986) and Ikeda (1973) on methods Although salinity may in¯uence copepod pro-
used to culture or maintain harpacticoid and ductivity, many species may be reared at
calanoid copepods. A common feature of the suboptimal salinity. For example, the estuarine
copepod culture trials is their relatively small scale, species Acartia tonsa DFH.AT1 (Stùttrup,
usually in small volumes of one to a few litres and of Richardsen, Kirkegaard & Pihl 1986) was shown
# 2000 Blackwell Science Ltd,, Aquaculture Research, 31, 703±711 705

706
Copepods in marine aquaculture J G Stùttrup

Table 1 Different methods for the larger scale mass culture of calanoid or harpacticoid copepods, which were harvested at regular intervals and/or used as live prey for marine ®sh
larvae
Culture species Culture size Densities Productivity Food/conditions Reference
Acartia tsuensis 24 m3 20% ER = 150 mg DW L±1 20% ER = 30 mg DW L±1 day±1 Natural phytoplankton, extensive Ohno, Takahashi & Taki
(1990)
Acartia tsuensis 25 L 30% ER = 100 mg DW L±1 Max 30% exploitation rate Nannochloropsis oculata, intensive Ohno et al. (1990)
Acartia tonsa 1890 L 232 L±1 2±75 nauplii/adult Natural phytoplankton, extensive Ogle (1979)
Acartia tonsa 200±450 L 50±100 L±1 200±220 eggs L±1 culture Rhodomonas baltica, Isochrysis galbana, intensive Stùttrup et al. (1986)
Eurytemora af®nis 30 m3 300±500 L±1 copepodites 7±10% Algae and detritus, extensive Nellen, Quantz, Witt,
and adults > 1000nauplii L±1 Kuhlmann & Koske (1981)
Calanoid species c. 2800 m3 Natural phytoplankton; extensive B. Urup (pers. commun.)
Tisbe spp. 1.5-L ¯oating baskets 92±115 mL±1 Not estimated Mytilus powder lettuce pieces Kahan, Uhlig, Schwenzer
# 2000 Blackwell Science Ltd,, Aquaculture Research, 31, 703±711
in 200-L ®sh tanks & Horowitz (1982)

Schizopera elatensis 29/mL
Tigriopus japonicus 210 m3 10±22 mL±1 4±5 kg at regular intervals Chlorella minutissima, w-yeast, baker's yeast Fukusho (1980)
co-culture with rotifers; outdoor tanks,
semi-extensive
Aquaculture Research, 2000, 31, 703±711

Tisbe spp. 32 L 1 mL±1 adults 1.4 nauplii/ind day±1 Microfeast L-10 larval diet or Isochrysis galbana Nanton & Castell (1997)
31 mL < 200 mm
Tisbe holothuriae 5-L trays 8 mL±1 300 000 (20 mg) nauplii/tray day±1 Rhodomonas baltica, batch intensive Stùttrup & Norsker (1997)
Tisbe holothuriae 150-L closed tank Not registered 500 000 ind day±1 mixed nauplii Rhodomonas baltica, continuous intensive Stùttrup & Norsker (1997)
and copepodites in food-limited cultures
ER, exploitation rate; DW, dry weight.

to have a faster development at 17 psu compared Starting up cultures with sediments containing
with 25 psu (Toudal & RiisgaÊrd 1987). For resting eggs is also another viable option, as
production purposes, this species was adapted to described by Nñss (1996). These cultures, which
laboratory conditions and high salinity (27 psu were run in outdoor tanks, were dominated by
and 35psu) corresponding to that of the sea Eurytemora af®nis, but other calanoids, such as
water available at the two laboratories where this Centropages hamatus and Acartia spp., were also
species was cultured (Stùttrup et al. 1986). This present.
organism was maintained in culture for several A major reason for the lack of progress in the
hundred generations since it was isolated from a development of reliable copepod production systems
plankton sample taken from the éresund, may, of course, be attributed to the lack of economic
Denmark, in 1981. One main disadvantage of incentives for the industry. It remains to be proven
this culture was the requirement for low-density that the added costs of maintaining copepod
(< 100 adults L±1), leading to low-volume output cultures are offset by increased revenue from
(» 530 eggs L±1 culture volume; Stùttrup et al. increased ®sh growth and survival and better
1986). This species is a free spawner, and the production ef®ciency in terms of lower percentages
eggs can be removed from the adult culture by of individuals with morphological or developmental
siphoning the bottom of the tank once or twice defects.
daily. The eggs are incubated for about 2 days,
with hatching percentages averaging 45%. In
contrast, the benthic, egg-sac-bearing harpacticoid Copepods as live prey for marine ®sh
Tisbe holothuriae (Stùttrup & Norsker 1997) could larvae
be maintained at much higher densities (> 13000
In a recent review of current production techniques
adults L±1), resulting in a more than 200-fold
for rearing cod, turbot and halibut (Stùttrup et al.
increase in output (average of 100 000 nauplii L±1
1998), the initial food source for cod and halibut
culture volume). Thus, despite the lower fecundity
was primarily extensive, copepod-based systems,
(eggs/female day±1) in egg-sac-bearing species
whereas turbot rearing used both extensive, cope-
compared with free spawners (Kiùrboe & Sabatini
pod-based systems and intensive rotifer- and
1995), the overall productivity of the system was
Artemia-based systems. As larval-rearing technology
higher because of the much higher densities
and knowledge of the species' developmental
obtainable. This benthic species can be fed a
requirements improves, a trend towards more
variety of food sources and has a high tolerance intensive rearing for all three species is occurring.
to a wide range of environmental conditions. Accompanying this trend are increasing problems
Calanoids and harpacticoids are the main cope- with abnormally pigmented juveniles, and other
pod species targeted for production and use as live de®ciencies have been noted. In response, there is
prey. Copepods with planktonic nauplii are preferred an increasing interest in intensive cultivation of
as most ®sh larvae hunt and feed in the water copepods, which can provide larval ®sh food all year
column. These species are more demanding in round. Such a trend has occurred in halibut rearing.
culture, requiring phytoplankton and low adult This species was initially reared in ¯oating plastic
densities. On the other hand, benthic copepods, bags (see review by van der Meeren & Naas 1997),
which tolerate higher densities, do not require after the rather prolonged (30±40 days) yolk-sac
planktonic microalgae as food and can be fed (prefeeding) stage, in large indoor units (silos) ®rst
various inert diets (Norsker & Stùttrup 1994). described by Rabben, Jelmert & Huse (1987). More
Only recently, Nanton & Castell (1997) described recently, halibut larvae have been reared in
a mass-culture system of Tisbe spp. with an average intensive systems using rotifers and Artemia but,
production of 1.4 nauplii per adult over the ®rst because of problems with abnormal pigmentation,
6 days of culture. Using an inert diet as food copepod supplements are still required (Nñss & Lie
(Microfeast), high adult densities approaching 1998). By switching to copepods from Artemia for
1000 L±1 were obtained over a period of 16 days. 7 days during the early larval stage, halibut
This experimental culture was terminated after juveniles develop normal pigmentation compared
16 days because of contamination by other harpac- with those fed exclusively on Artemia (Nñss & Lie
ticoid species. 1998).

The documented advantages of using copepods Measurable economic advantages during the on-
for larval rearing are listed below as well as growing stage may rekindle interest in improving
potential bene®ts, which have been claimed but larval nutrition.
are yet to be proven. Although research has focused on the nutritional
aspects of wild zooplankton compared with Artemia,
other aspects of the extensive system may have
Advantages
advantages, e.g. low stocking densities or the
Bene®ts stated in the literature include: presence of phytoplankton in the system. The
1. superior nutritional value resulting in normal green-water technique re¯ects an attempt to imitate
pigmentation and development and eliminating the natural environment and to mitigate `deleter-
the need for enrichment (Nellen 1981; ious' effects of the intensive system. The fact that the
Watanabe, Oowa, Kitajima & Fujita 1983; majority of commercial hatcheries use the green-
Witt, Quantz, Kuhlmann & Kattner 1984; water technique demonstrates the existence of basic
Norsker & Stùttrup 1994; Heath & Moore water-quality problems in intensive systems.
1997; Nñss & Lie 1998);
2. optimal lipid to protein content for fast-growing
Do copepods retain their nutritional
®sh larvae preventing dietary lipid overloading
value under culture?
(ConceicaÄo 1997; natural zooplankton);
3. supplement of T. holothuriae imparted improved In culture, copepods are usually fed monospeci®c
growth rate to juvenile sole (Heath & Moore algal diets to reduce costs, but mixed diets may be
1997); more appropriate, in terms of both copepod pro-
4. improved stress resistance in larval mahimahi ductivity and effects on their nutritional value for
(Kraul et al. 1993; Euterpina acutifrons); ®sh larvae. Fatty acid composition in adult cala-
5. no alternative live food because of size for very noids has been shown largely to re¯ect that of their
small ®rst-feeding ®sh larvae (Morais, de Tito & diet in Paracalanus parvus, C. ®nmarchicus, C.
Bodiou 1984); helgolandicus and Pseudocalanus elangatus (Moreno,
6. may serve as `tank cleaners', thus helping to De Moreno & Brenner 1979; Kattner & Krause
maintain tank hygiene (Uhlig 1965; Stùttrup 1989), in T. holothuriae (Norsker & Stùttrup 1994)
1993; Tisbe holothuriae); and in Acartia tonsa (Stùttrup, bell & Sargent 1999).
7. help to initiate learning of hunting behaviour When A. tonsa females were solely fed Dunaliella
through visual and/or chemical stimuli (Stùttrup tertiolecta, they ceased feeding and producing eggs,
& Norsker 1997; Tisbe holothuriae); although the reason was unclear (Stùttrup & Jensen
8. source of exogenous digestive enzymes that 1990). One reason may be that D. tertiolecta
improve digestion of prey in early-stage ®sh contains low levels of highly unsaturated fatty acids
larvae in which the gut is not fully functional (HUFAs). These are considered essential for cope-
(Munilla-Moran, Stark & Barbour 1990; pods (Fraser et al. 1989), as well as for marine ®sh
Eurytemora hirundoides); larvae. Another reason could be that this alga is
9. help maintain algal cultures in suspension in unpalatable for this copepod species. The ability of
green-water systems (Stùttrup, Gravningen & copepods to discriminate between nutritious and
Norsker 1995; Tisbe holothuriae). non-nutritious particles has been demonstrated by
Potential bene®ts of using copepods for larval Huntley, Sykes, Rhoan & Marin (1986). This ability
rearing include and improvement in weaning is important as food quality affects the number of
success and an improvement in larval condition, eggs produced, as shown for A. tonsa (Donaghay
resulting in higher resistance to disease. 1985). The number of eggs produced by Calanus
The claim of improved weaning success and ®nmarchicus was shown to be in¯uenced by previous
higher resistance to disease in the juveniles deriving dietary history and, speci®cally, the total lipid
from the copepod-based systems has not been content of the phytoplankton (Gatten, Sargent,
substantiated to date. The veri®cation of potential Forsberg, O'Hara & Corner 1980). Egg production
bene®ts is important for the development of this is related to food availability and quality (Stùttrup &
®eld, as increased post-weaning survival and Jensen 1990; Gatten et al. 1980). GueÂrin & Gaudy
decreased labour costs associated with regular (1977) suggested a relationship between lipid
disease outbreaks or with slow growth would result. content and fecundity in T. holothuriae. It may

therefore be expected that nutritionally suboptimal in wild zooplankton. It is therefore proposed that
diets would diminish the productive performance of copepod eggs may be stored for shorter periods
the copepods. If dietary HUFA levels are insuf®cient without impairment of their dietary value in terms
and the copepods are unable to synthesize these of dietary lipid.
fatty acids at the required rates, egg production will In the above discussion, only lipid nutrition was
decline. It can therefore be proposed that algal considered, although other dietary characteristics
quality may affect the output of the copepod may contribute to the nutritional value of copepods.
production system, as well as the subsequent As such, any changes in these characteristics
nutritional quality of the copepods when used as should be monitored in monocultures over several
food for ®sh larvae. This may have been the case for generations. The ®rst step, however, remains to
A. tonsa feeding on D. tertiolecta (Stùttrup & Jensen identify the key components that are essential for
1990), but egg production in T. holothuriae has been the normal development of marine ®sh larvae.
shown to be independent of the HUFA content of The challenge of devising reliable, intensive
their diet (Norsker & Stùttrup 1994). T. holothuriae production systems for copepods remains open for
were fed D. tertiolecta for several generations with- the next century. It would seem sensible to
out any dramatic impact and, more recently, concentrate copepod-rearing efforts on planktonic,
Nanton & Castell (1998) showed another species estuarine species, which are naturally adapted to
of Tisbe to be equally independent of dietary variable conditions and tolerant to rearing at high
provision of HUFAs. Possibly, D. tertiolecta could densities. Until then, the convenience of using
be used by T. holothuriae because of an ability to Artemia and rotifers will ensure their continued
chain elongate and desaturate shorter chain fatty use over the next few years.
acids such as 18:3n-3 (Norsker & Stùttrup 1994).
HUFA levels, especially docosahexanoic acid
(DHA), in A. tonsa and T. holothuriae were higher Acknowledgments
than in enriched Artemia even when the copepods Part of the copepod review was carried out in
were reared on algae with a low HUFA content connection with a Concerted Action funded by the
(Stùttrup et al. 1999; Norsker & Stùttrup 1994). The EU, AIR3-CT94-2094. I am indebted to Dr Clive Fox
inability to boost Artemia DHA levels to high levels for correcting and improving my English.
can be attributed to retroconversion of DHA to
eicosapentaenoic acid (EPA), as demonstrated using
radiolabelled fatty acids (Navarro, McEvoy, Bell, References
Amat, Hontoria & Sargent 1997). Further, in the
Bristow G.A. (1990) Dùdelighet hos kveitelarver og yngel i
attempt to boost HUFA levels in Artemia, the total startforingsfasen. Norsk Fiskeoppdrett 15, 40±43.
lipid content in this organism may be raised to levels Chandler G.T. (1986) High-density culture of meiobenthic
that impair lipid metabolism and subsequently harpacticoid copepods within a muddy sediment sub-
growth, as demonstrated for turbot larvae strate. Canadian Journal of Fisheries and Aquatic Sciences
(ConceicaÄo 1997). 43, 53±59.
A great advantage of Artemia cysts is that they ConceicaÄo L.E.C. (1997) Growth in Early Life Stages of
can be stored for up to several years. Similar Fishes: An Explanatory Model. PhD Thesis, Dept of Fish
techniques for copepods would be a major advance. Culture and Fisheries, Wageningen Agricultural
The use of resting eggs harvested from sediment has University, Netherlands.
been attempted (Nñss 1996). Cultured copepod Cunha I. & Planas M. (1995) Ingestion rates of turbot
eggs may be stored in a refrigerator for a period of larvae (Scophthalmus maximus) using different-sized live
prey. ICES Marine Science Symposia 201, 16±20.
weeks to months, although the hatching rate
Donaghay P.L. (1985) An experimental test of the relative
decreases with time (J G Stùttrup unpubl. data).
signi®cance of food quality and past feeding history to
DHA levels in A. tonsa eggs have been shown to
limitation of egg production of the estuarine copepod
decrease over a period of 12 weeks of storage in a Acartia tonsa. Archiv fuÈr Hydrobiologie. Ergebnisse der
refrigerator, whereas EPA levels were only slightly Limnologie 21, 235±245.
affected, and arachidonic acid (20:4n-6) increased Fraser A.J., Sargent J.R. & Gamble J.C. (1989) Lipid class
almost twofold (Stùttrup et al. 1999). During the and fatty acid composition of Calanus ®nmarchicus
storage period, the DHA/EPA ratio fell from 2.3 to (Gunnerus), Pseudocalanus sp. and Temora lonicornia
1.7, which is still, however, within the range found (MuÈller) from a nutrient enriched sea water enclosure.

Journal of Experimental Marine Biology and Ecology 130, LeÂger P., Bengtson D.A., Simpson K.L. & Sorgeloos P.
81±92. (1986). The use and nutritional value of Artemia as a
Fukusho K. (1980) Mass production of a copepod, Tigriopus food source. Oceanography and Marine Biology Annual
japonicus in combination culture with a rotifer Review 24, 521±623.
Brachionus plicatilis fed w±yeast as a food source. McEvoy L.E., Naess T., Bell J.G. & Lie é. (1998) Lipid and
Bulletin of the Japanese Society of Fisheries Science 46, fatty acid composition of normal and malpigmented
625±629. Atlantic halibut (Hippoglossus hippoglossus) fed enriched
Fukusho K. (1989) Biology and mass production of the Artemia; a comparison with fry fed wild copepods.
rotifer, Brachionus plicatilis. International Journal of Aquaculture 163, 237±250.
Aquaculture and Fisheries Technology. 1, 232±240. Marcogliese D.J. (1995) The role of zooplankton in the
Gatten R.R., Sargent J.R., Forsberg T.E., O'Hara S.C.M. & transmission of helminth parasites to ®sh. Reviews in
Corner E.D.S. (1980) On the nutrition and metabolism Fish Biology and Fisheries 5, 336±371.
of zooplankton. XIV. Utilisation of lipid by Calanus Marcus N.H. & Alatalo P. (1989) Conditions for rearing
helgolandicus during maturation and reproduction. Calanus ®nmarchicus (Gunnerus, 1770) (Copepoda,
Journal of the Marine Biological Association UK 60, Calanoida) through multiple generations in the labora-
391±399. tory. Crustaceana 57, 101±103.
GueÂrin J.-P. & Gaudy R. (1977) Etude des variations du van der Meeren T. (1991) Selective feeding and prediction
poids sec et de la constitution chemique eÂleÂmentaire de of food consumption in turbot larvae (Scophthalmus
Tisbe holothuriae (Copepoda: HarpacticoõÈda) eÂleveÂ sur maximus L.) reared on the rotifer Brachionus plicatilis
diffeÂrents reÂgimes arti®ciels. Marine Biology 44, 65±70. and natural zooplankton. Aquaculture 93, 35±55.
Heath L.P. & Moore C.G. (1997) Rearing Dover sole larvae van der Meeren T. & Naas K.E. (1997) Development of
on Tisbe and Artemia diets. Aquaculture International 5, rearing techniques using large enclosed ecosystems in
29±39. the mass production of marine ®sh fry. Reviews in
Howell B.R. (1994) Fitness of hatchery-reared ®sh for Fisheries Science 5, 367±390.
survival in the sea. Aquaculture and Fisheries Moreno V.J., De Moreno J.E.A. & Brenner R.R. (1979)
Management 25, 3±17. Fatty acid metabolism of the calanoid copepod
Huntley M., Sykes P., Rhoan S. & Marin V. (1986) Paracalanus parvus: 2. palmitate, stearate, oleate and
Chemically-mediated rejection of dino¯agellate prey by acetate. Lipids 14, 318±322.
the copepods Calanus paci®cus and Paracalanus parvus: Munilla-Moran R., Stark J.R. & Barbour A. (1990) The role
mechanisms, occurrence and signi®cance. Marine of exogenous enzymes in digestion in cultured turbot
Ecological Progress Series 28, 105±120. larvae (Scophthalmus maximus L.). Aquaculture 88,
Ikeda T. (1973) On the criteria to select copepod species for 337±350.
mass culture. Bulletin of the Plankton Society of Japan 20, Nñss T. (1996) Benthic resting eggs of calanoid copepods
41±48. in Norwegian enclosures used in mariculture: abun-
Kahan D., Uhlig G., Schwenzer D. & Horowitz L. (1982) A dance, species composition and hatching. Hydrobiologia
simple method for cultivating harpacticoid copepods 320, 161±168.
and offering them to ®sh larvae. Aquaculture 26, Nñss T. & Lie é. (1998) A sensitive period during ®rst
303±310. feeding for the determination of pigmentation pattern in
Kanazawa A. (1991) Nutritional mechanisms causing Atlantic halibut, Hippoglossus hippoglossus L., juveniles:
abnormal pigmentation in cultured marbled sole larvae, the role of diet. Aquaculture Research 29, 925±934.
Limanda yokohamae (Heterosomata). In: Larvi ¢91 (ed. by Nñss T., Germain-Henry M. & Naas K.E. (1995) First
P. Lavens, P. Sorgeloos, E. Jaspers & E. Ollevier), pp. feeding of Atlantic halibut (Hippoglossus hippoglossus)
20±22. Special Publication no. 15. European using different combinations of Artemia and wild
Aquaculture Society, Belgium. zooplankton. Aquaculture 130, 235±250.
Kattner G. & Krause M. (1989) Seasonal variations of Nanton D.A. & Castell J.D. (1997) Mass Culture of the
lipids (wax esters, fatty acids and alcohols) in calanoid Harpacticoid Copepod, Tisbe sp. ICES CM Marine Fish
copepods from the North Sea. Marine Chemistry 26, Culture Committee, 10 pp.
261±275. Nanton D.A. & Castell J.D. (1998) The effects of dietary
Kiùrboe T. & Sabatini M. (1995) Scaling of fecundity, fatty acids on the fatty acid composition of the
growth and development in marine planktonic cope- harpacticoid copepod, Tisbe sp., for use as a live food
pods. Marine Ecology Progress Series 120, 285±298. for marine ®sh larvae. Aquaculture 163, 251±161.
Kraul S., Brittain K., Cantress R., Nagao T., Ako H., Navarro J.C., McEvoy L.A., Bell M.V., Amat F., Hontoria F.
Ogasawara A. & Kitagawa H. (1993) Nutritional factors & Sargent J.R. (1997) Effect of different dietary levels of
affecting stress resistance in the larval mahimahi docosahexaenoic acid (DHA, 22:6w3) on the DHA
Coryphaena hippurus. Journal of the World Aquaculture composition of lipid classes in sea bass larvae eyes.
Society 24, 186±193. Aquaculture International 5, 509±516.

Nellen W. (1981) Live animal food for larval rearing in Stùttrup J.G., Richardsen K., Kirkegaard E. & Pihl N.J.
aquaculture. ± non-Artemia organisms. In: Realism in (1986) The cultivation of Acartia tonsa Dana for use as
Aquaculture: Achievements, Constraints, Perspectives (ed. live food source for marine ®sh larvae. Aquaculture 52,
by M. Bilio, H. Rosenthal & C.J. Sindermann), pp. 87±96.
215±260. European Aquaculture Society, Bredene, Stùttrup J.G., Gravningen K. & Norsker N.H. (1995) The
Belgium. role of different microalgae on the growth and survival
Nellen W., Quantz G., Witt U., Kuhlmann D. & Koske P.H. of turbot larvae in intensive cultivation systems. ICES
(1981) Marine ®sh rearing on the base of an arti®cial Marine Science Symposia Series 201, 173±186.
food chain. European Mariculture Society Special Stùttrup J.G., Shields R., Gillespie M., Gara M.B., Sargent
Publication 6, 133±147. J.R., Bell J.G., Henderson R.J., Tocher D.R., Sutherland
Norsker N.H. & Stùttrup J.G. (1994) The importance of R., Nñss T., Mangor Jensen A., Naas K., van der Meeren
dietary HUFAs for fecundity and HUFA content in the T., Harboe T., Sanchez F.J., Sorgeloos P., Dhert P. &
harpacticoid, Tisbe holothuriae Humes. Aquaculture 125, Fitzgerald R. (1998) The production and use of
155±166. copepods in larval rearing of halibut, turbot and cod.
Ogle J. (1979) Adaptation of a brown water culture Bulletin of the Aquaculture Association of Canada 4,
technique to the mass culture of the copepod Acartia 41±46.
tonsa. Gulf Research Reports 6, 291±292. Stùttrup J.G., Bell J.G. & Sargent J.R. (1999) The fate of
Ohno A., Takahashi T. & Taki Y. (1990) Dynamics of lipids during development and cold-storage of eggs in
exploited populations of the calanoid copepod, Acartia the laboratory-reared calanoid copepod Acartia tonsa
tsuensis. Aquaculture 84, 27±39. Dana, and in response to different algal diets.
Rabben H., Jelmert A. & Huse I. (1987) Production Aquaculture 176, 257±269.
Experiment of Halibut Fry (Hippoglossus hippoglossus Tito de Morais L. & Bodiou J.Y. (1984) Predation on
L.) as a Function of Temperature Regime. ICES CM 1987/ meiofauna by juvenile ®sh in a Western Mediterranean
F:42. ¯at®sh nursery ground. Marine Biology 82, 209±215.
Robert F. & Gabrion C. (1991) Experimental approaches to Toudal K. & RiisgaÊrd H.U. (1987) Acute and sublethal
the speci®city in ®rst intermediate hosts of bothrioce- effects of cadmium on ingestion, egg production and life-
phalids (Cestoda, Pseudophyllidea) from marine ®sh. cycle development in the copepod Acartia tonsa. Marine
Acta Oecologia 12, 617±632. Ecology Progress Series 37, 141±146.
Stùttrup J.G. (1993) First feeding in marine ®sh larvae ± Uhlig G. (1965) Die mehrgliedrige Kultur litoraler
nutritional and environmental aspects. Sarsia. In: Folliculiniden. HelgolaÈnder Wissenschaftliche
Physiology and Biochemistry of Fish Larval Development Meeresuntesuchungen 12, 52±60.
(ed. by B.T. Walther & H.J. Fyhn), pp. 123±131. Watanabe T., Oowa C., Kitajima C. & Fujita S. (1983)
University of Bergen, Norway. Nutritional values of live organisms used in Japan for
Stùttrup J.G. & Jensen J. (1990) In¯uence of algal diet on mass propagation of ®sh: a review. Aquaculture 34,
feeding and egg production of the calanoid copepod 115±143.
Acartia tonsa Dana. Journal of Experimental Marine Witt U., Quantz G., Kuhlmann D. & Kattner G. (1984)
Biology and Ecology 141, 87±105. Survival and growth of turbot larvae Scophthalmus
Stùttrup J.G. & Norsker N.H. (1997) Production and use of maximus L. reared on different food organisms with
copepods in marine ®sh larviculture. Aquaculture 155, special regard to long chain polyunsaturated fatty acids.
231±247. Aquacultural Engineering 3, 177±190.

Aquaculture Research, 2013, 1–10 doi:10.1111/are.12317
A first step towards improving copepod cultivation

using modelling: the effects of density, crowding,
cannibalism, tank design and strain selection on
copepod egg production yields
Guillaume Drillet1 & Fabien Lombard2,3

1
DHI-NTU Research Centre and Education Hub, Singapore, Singapore
2
Observatoire Oceanographique de Villefranche sur Mer, Villefranche sur Mer, France
3
Universite Pierre et Marie Curie, Paris, France
Correspondence: G Drillet, DHI-NTU Research Centre and Education Hub, 1 CleanTech Loop, #03-05 CleanTech One, Singapore
637141, Singapore. E-mail: gdr@dhigroup.com
Abstract Introduction
Copepods are the optimal live feed for hatcheries and Copepods are an ideal live feed item that could
improvement of cultivation techniques, to provide a benefit the entire aquaculture and aquarium trade
constant food source, is crucial for the expansion of industry if raised in high quantity at a low cost
the industry. However, studies based on experimen- (Drillet, Frou€el, Sichlau, Jepsen, Højgaard, Joarder
tal work and real observations can be labour inten- & Hansen 2011b). The use of these natural live
sive and expensive. A simple model was developed feed have been shown to increase survival,
based on the well-known life history traits of Acartia decrease mal-pigmentation and improve growth of
tonsa to describe batch cultures and their productiv- fish larvae (Næss, Germain-Henry & Naas 1995;
ity. Model results were compared to observations Rajkumar & Vasagam 2006; Wilcox, Tracy & Mar-
from real cultures. For maximizing egg production cus 2006; Busch, Falk-Petersen, Peruzzi, Rist &
yields, the optimal stocking density of copepods Hamre 2010). However, problems with providing
should be adapted to the design (depth) of the culture large and sustainable productions of copepods
tanks. At high densities, stress due to encountering persist, requiring fine-tuning of production methods.
conspecifics, as well as cannibalism of eggs by adults, Optimization of industrial productions of all type
limits egg production yields. Using this model, the can benefit from using modelling as a tool to investi-
potential selection efficacy of copepod strains was gate possible bottlenecks in production (e.g. Xie & Li
also evaluated in order to increase production yields. 2012). Within aquaculture, modelling approaches
Selecting larger copepods increases the egg produc- have been used as decision support systems (Gut-
tion per litre of culture, but decreases the optimal ierrez-Estrada, Pulido-Calvo, de la Rosa & Marchini
stocking density and the range of densities at which 2012), however, the optimization of copepod pro-
egg production yield is high, and vice-versa. Select- ductions never benefited from such tools. Because of
ing copepods that are less affected by stress due to its relatively low cost, modelling can help providing
conspecifics only affect production yields at very high theoretical zoo-technical and managerial solutions
adult densities. However, selecting copepods with a for the production of well described species without
high Specific Growth Rate (SGR), or improving their making great investments in terms of experimental
SGR, was found to be an alternative which did not work. Also, in a global context where selection of
affect the optimal cultivation densities but improved aquatic organisms is a key to the development of
egg production yields. aquaculture (Duarte, Marb a & Holmer 2007; Drillet
et al. 2011b), the use of models can define the most
Keywords: live feed, copepod, crowding, sensitive life history traits to select populations with
modelling, cannibalism, production, Acartia tonsa high productivity potential.
© 2013 John Wiley & Sons Ltd 1

Modelling copepod cultures G Drillet and F Lombard Aquaculture Research, 2013, 1–10
Eggs of some copepods species can be produced, (Kiørboe, Mohlenberg & Riisg
ard 1985 and refer-
processed, cold-stored and easily transported across ence therein).
the world for use as live feed in aquacultures.
Such copepod products are, in their utilization,
Ac ¼ ð1:11 105 PL2:92 Þ103 ð3Þ
similar to products like Artemia cysts (Marcus &
Murray 2001; Drillet, Iversen, Sørensen, Ramløv,
4 ED 3
Lund & Hansen 2006; Holmstrup, Overgaard, Ec ¼ p ECD ð4Þ
3 2
Sørensen, Drillet, Hansen, Ramløv & Engell-Søren-
sen 2006; Drillet, Lindley, Michels, Wilcox & Mar-
The specific growth (SGR; day1) which repre-
cus 2007). Companies producing copepods eggs
sent the proportion of carbon an individual could
could therefore centralize their production and
allocate to growth and reproduction compared to
market eggs as a separated product and/or farms
its own weight was set at 0.45 day1 at 17°C
could produce and store eggs on site (Drillet et al.
(Bergreen et al. 1988) and was influenced by tem-
2011b).
perature following a Q10 value of 2.7 according to
The development of a simple model was based
Hirst & Bunker (2003). Thus, SGR can be
on data from the literature to understand the effect
expressed using the following equation:
of adult stocking density on egg production yields
in batch cultures of the copepod Acartia tonsa. This SGR ¼ 0:45 Q10 ð
T17
10 Þ ð5Þ
model was tested and validated using laboratory
observations. Subsequently, copepod size, SGR and
Adults swim continuously at an average veloc-
stress related to stocking density were modelled to
ity (Sv in mm s1) which was dependent on water
assess their potential effect on the egg production
density and therefore temperature following
yield of A. tonsa.
Larsen, Madsen & Riisg
ard (2008) as described:
Sv ¼ 1:37eð0:0525T Þ ð6Þ
Material and method
The model is based on experimental data from the Copepod jumps to avoid their conspecifics when
literature and compared to independent experi- reaching Mean Nearest Neighbour Distance (MNND
mental data. This model was simplified as much as in mm) as defined by Leising & Yen (1997).
possible to keep only the most important variables
that are likely to influence cultures. The salinity
was theoretically kept at 30 and food items were MNND ¼ 0:59PL103 þ 1:6 ð7Þ
available ad libitum to mimic an ideal aquaculture
production. A. tonsa is considered to behave only The volume of water scanned by an adult (VSA
as a filter feeder that feeds continuously, although in L day1) in which an adult could meet another
A. tonsa occasionally behave as a raptorial feeder one was therefore defined from the MNND and its
(Jonsson & Tiselius 1990). Adult prosome length swimming speed (with 86400 s in one day)
(APL in lm) and egg diameter (ED in lm) were lin-
VSA ¼ 86400pMNND2 Sv 106 ð8Þ
early dependent of the temperature (T) (Hansen,
Drillet, Kozmer, Madsen, Pedersen & Sørensen
The number of adult encounter L1 day1 (AE)
2010) and using their data the following relation-
was also dependent on the density of adults (DA
ships were obtained:
adult L1) and follows:
APL ¼ 900:55 þ ð8:3695 TÞ ð1Þ
AE ¼ VSA DA ð9Þ
ED ¼ 86:229 þ ð0:2624 TÞ ð2Þ
Adults feed continuously except when encoun-
The carbon contents of eggs (EC in pgC)) and tering a conspecific and recovering from it. This
adults (AC in pgC) were dependent on their size. succession of events, i.e., encounter, jump and
The model used the weight-length relationship for recovery was defined as the encountering reaction
A. tonsa adults from Bergreen, Hansen & Kiørboe time (ERT) and was set as a 3 s event in the
(1988) and considered eggs as spherical using model. The foraging time ratio (FR) was defined as
an egg carbon density (ECD) of 0.14 pgC lm3 the proportion of time an adult had to feed:
2 © 2013 John Wiley & Sons Ltd, Aquaculture Research, 1–10

Aquaculture Research, 2013, 1–10 Modelling copepod cultures G Drillet and F Lombard
86400 ðAE ERT Þ volume scanned for eggs was defined as (VSE in
FR ¼ ð10Þ
86400 litres adult1 day1):

APW 2
Correction for the SGR of adults (CSGR) with the VSE ¼ 86400p Sv 106 ð16Þ
foraging time ratio: 2
CSGR ¼ SGRFR ð11Þ Taking into account the foraging time ratio, we
defined the maximum number of egg encountered
A sex ratio of 1:1 was assumed because natural per adult per day (MaxE) as:
selection should result in this proportion of each
gender in a population reproducing sexually MaxE ¼ VSE FR EWC ð17Þ
(Fisher 1930). In the model, only females produce
eggs (i.e. half of the adults) and because their The egg handling time (HT) was removed from
growth is arrested at adult stage, all the carbon the proportion of time adults had to cannibalize
available for growth is allocated to egg production. other eggs, but since eggs are considered as optimal
Therefore, the egg production (EP in food (Drillet et al. 2006), cannibalism did not affect
egg L1 day1) was defined as: the CSGR. The HT was set to 150.4 sec, based on
maximum observation from (G. Drillet, R. Maguet,
DA AC CSGR M.-S. Mahjoub, F. Roullier, M.J. Fielding, submitted)
EP ¼ ð12Þ
2 EC and therefore we defined the proportion of time lost
to encounter new eggs (TLC in ind1 day1):
The egg density was set to 1.082 in the model
MaxEHT
(Miller & Marcus 1994, average at 31 salinity) TLC ¼ ð18Þ
86400FR
and sediment in the water column at rates follow-
ing a fitting of results from Stoke’s equation for
The total number of eggs cannibalized per day
eggs at a salinity of 30 and at different tempera-
(TOTCANN in egg L1day1) was:
tures (egg size therefore changes; fitted
R2 = 0.9971). The egg sedimentation rate (ESR in TOT CANN ¼ ðMaxE ðMaxETLC ÞÞDA ð19Þ
m day1) is given as:
ESR ¼ 0:3577T þ 10:566 ð13Þ And the total egg production per litre per day
(TOT EPROD in egg L1day1):
Because eggs can be produced in the entire TOT EPROD ¼ EP TOT CANN ð20Þ
water column we considered that they were pro-
duced in the middle of the water column, of tanks The present model therefore only evaluates
with a total water depth D (m). We consider that instantaneous rates (feeding, eggs production,
once an egg has reached the bottom of the tank it cannibalism) of simulated copepods cultures using a
was safe because it could be harvested. We there- variety of environmental conditions (temperature,
fore defined the egg exposure to cannibalism (EExp) copepods density, tank depth) as forcing variables.
as follows: Because some parameters used in the model are
1 D subject to some variations, notably depending on
EExp ¼ ð14Þ the food provided and different in behaviour that
2 ESR
copepods may have, the adequacy of those param-
eters was assessed by doing a sensitivity analysis.
The number of eggs L1 in the water column
(EWC) at one time was: Parameters such as copepod sizes Specific Growth
Rates (SGR) Mean Nearest Neighbour Distance
EWC ¼ EP EExp ð15Þ (MNND) or the encountering reaction time (ERT)
were alternatively varied from 15% of their stan-
Egg cannibalism occurs when encountered dard values and eggs productions results evaluated
within the swimming path of an adult which was against a standard simulation. This exercise has
defined as the disk represented by the adult pro- two advantages: it allows to visualise which
some width (APW) (defined as APW = 0.3459.APL parameters have the higher influence on final
from measurements on 52 females, pers. obs.). The model results, and consequently should be
© 2013 John Wiley & Sons Ltd, Aquaculture Research, 1–10 3

intensively measured to increase model precision, The expected proportion of eggs which were
but it also indicates which copepods life history cannibalized increased with the depth of the tanks
trait should be selected for in order to achieve as well as with the stocking density of copepods
higher egg productions. (Fig. 1c). Including the cannibalistic effects
Results from the model were compared to data revealed that the highest egg production per litre
from Florida State University Marine Laboratory of culture was reached under low tank depths and
(FSUML), where a culture was held at 17°C and at densities of copepods between ~3000 ind L1
fed ad libitum with Rhodomonas salina and R. lens and ~5000 ind L1 (Fig. 1d).
(as described by Drillet et al. 2007). In this cul- Optimizing production in space (per m2) is impor-
ture, the number of adults and the sex ratio was tant from a zoo-technical point of view. At adult
counted weekly. Copepodites and nauplii numbers densities above ~1200 ind L1, deep tanks clearly
were ignored in the comparison because they did have a negative effect due to increased cannibalism
not produce eggs and may not cannibalize eggs. incidents. Culture tanks deeper than ~1.5 m seem
The egg harvest was counted daily (5 days a to be productive only for very low copepod densities
week) and averaged as a daily egg production so (Fig. 1e). As expected, lower densities of copepods
that one data per week was produced to compare coupled with shallower tanks will reduce the loss of
to the model. For these comparisons, the model potential production due to interactions between
was changed accordingly to the culture conditions adults and cannibalism (Fig. 1f). At very high densi-
in terms of sex ratio and depth of tanks. ties >5000 ind L1, copepod egg production per
female is low (Fig. 1f), but adults spent most of their
time encountering/disturbing each other and there-
Results
fore have little chances to cannibalize the few eggs
The effect of temperature on the egg production per produced (Fig 1c,d,e).
litre of culture was first tested assuming a tank of The extent by which egg production was tested
1.3 meters depth (Fig. 1a). Two optimum stocking was per tank with various factors influencing
densities were observed; one at a stable low copepod copepod size, specific growth rates, sensitivity to
density (~1000 ind L1) and a second one at higher conspecifics (MNND, jumping or recovering time
densities and which varied with temperature from from encounter). By testing the sensitivity of these
~3500 ind L1 at 30°C to >7000 ind L1 at 15°C. parameters in the model, the life history traits on
For both optimal densities, the egg production per which to force selection could be tested to achieve
litre increased with temperature. The higher density higher eggs production yields. As earlier a sensitiv-
for egg production was directly due to the influence ity analysis in comparison to a standard simula-
of temperature on copepod size, swimming speed tion at 20°C and in a 1.3 m deep tank was used.
and egg sedimentation speed (equations 3, 6 and We first tested these effects in the model for the
13) which all impacted the rate of cannibalism. selection of copepods 15% bigger or smaller than
However, because of the lack of data on productivity average A. tonsa size (Fig. 2a). Bigger copepods
and behaviour of A. tonsa under extreme densities increased the egg production yields per litre but
and temperatures (see discussion), the study focused decreased the range of densities at which they can
on the overall model properties at a fixed 20°C be stocked to achieve these productions. Inversely,
which is realistic for the production of A. tonsa (Hol- reducing copepod size led to a lower egg production
ste & Peck 2006; Drillet et al. 2006, 2007). but eggs were produced over a larger range of densi-
When ignoring cannibalism in the model, only ties, and without intermediate densities where egg
the copepod stocking density and the loss of production was depleted due to cannibalism
production due to the decrease in foraging time (Fig. 2a). Alternatively selecting copepods that nat-
(interaction adults to adults resulting in a decreased urally exhibited higher or lower SGR (15%)
SGR, equation 11) affected the egg production per affected the egg production yields accordingly, with-
litre (Fig. 1b). Under these idealistic conditions, an out affecting the optimal stocking density ranges
egg production of ~15000 to ~20000 egg L1 can (Fig. 2b). Also tested was if selecting copepods that
be expected at densities of ~3000–4000 ind L1 were more or less sensitive to the presence of con-
but the egg production would be reduced at higher specifics (parameter MNND and recovery time from
stocking densities because the copepods produce encounter) would affect the productivity. None of
less eggs per individual. these parameters affected the maximum egg

(a) (b)
(c) (d)
(e) (f)
Figure 1 (a) Egg production per litre in function of temperature and copepod density for a tank depth of 1.3 m. (b)
optimal production of copepod per litre at 20°C as a function of adult density and tank depth, without cannibalism.
(c) proportion of egg cannibalized by the copepods at 20°C as a function of adult density and tank depth. (d) Egg
production per litre at 20°C as a function of adult density and tank depth, taking into account cannibalism. (e)
maximum egg production yields per m2 at 20°C as a function of adult density and tank depth. (f) Maximum egg
production per female per day at 20°C as a function of copepod stocking density and tank depth.
production reached, only the range of stocking den- increasing MNND affected the range of densities
sities that result in these yields. Changing MNND where cannibalism occurred at high rates (~2000
(15%) affected the stocking density of the second to ~4000 ind L1) because encountering rates were
peak of production (Fig. 2c), but not the initial one increased (and vice versa). Changing the recovery
which occurred at lower stocking densities. The time after an encounter affected the general trend

(a) (b)
(c) (d)
Figure 2 Results from selection tests for 1.3 m deep tanks and cultures held at 20°C. Results are presented as egg
production per litre as a function of copepod stocking density in ind L1. (a) selection of copepod sizes (15%). (b)
selection of copepod Specific Growth Rates (SGR) (15%). (c) selection of the Mean Nearest Neighbour Distance
(MNND) (15%). (d) selection of the recovery from the jump carried to escape from conspecifics (15%).
further, but increasing recovery time is still driving the first time, modelling was used as a tool to
the model, as with increasing the MNND (see ERT in assist the improvement of copepod cultures that
material and methods). The sensitivity of adults to will eventually offer access to the breeding of new
each other clearly affected the range at which can- commercial finfish species for the benefit of aqua-
nibalism occurred (Fig. 2c,d). Finally, handling time culture.
of eggs during cannibalism was tested, but did not
change the model results significantly (data not
Results from the model
shown).
Comparing the results from the model against The present model offered the possibility to evalu-
real observations from the FSUML (Fig. 3a,b) ate/visualize the optimum stocking density of
showed that the models mimic the observations adults depending on the design of culture tanks.
nicely (Fig. 3a) and the 1:1 relationship between Cannibalism is an important factor that affects
predicted and observed eggs productions is hardly productivity of cultures at high densities (Camus &
differentiable, validating the accuracy of the exper- Zeng 2009; Drillet et al. unpublished; Liang, Uye &
imental model (Fig. 3b). There is however a large Onbe 1994). By taking into account the sedimen-
variability in the data set (Fig. 3a,b). tation of eggs in the water column, we showed
that the depth of the tanks drastically affected the
cannibalistic pressure and confirmed that eggs
Discussion
should be separated from swimming stages as
Improving copepods cultures is crucial to support reviewed by Drillet et al. (2011b and reference
the development of aquaculture worldwide. For therein). In order to achieve the goal of producing

6 6
(a) x 10 (b) x 10
3.5 4
Observed
Observed eggs production (nb. tank−1)

Predicted 3.5
3
Eggs production (nb tank–1) 3
2.5
2.5
2
2
1.5
1.5
1
1
0.5 0.5
0 0
0 200 400 600 800 1000 0 1 2 3 4
Adult density (ind. L–1) Predicted eggs production (nb tank–1) x 106
Figure 3 Comparison between observed and modelled egg production. The plain line represents the best fit between
observed and modelled egg production. Dashed lines are the 95% confidence intervals and the dotted line represents
the 1:1 relationship which is hardly distinguishable from the relationship between predicted and observed eggs
productions.
large number of eggs it was documented that the reach high egg production levels, however, careful
adult stocking density and the design of the tanks attention to the stocking density would be
should be adjusted carefully; notably when using required, because the range of densities at which
deep tanks (over ~1 m), a low stocking density is ideal production is reached is narrower than with
optimal for obtaining high egg productions. smaller copepods. Other traits such as MNND and
A high-density optimum was also predicted, but post encounter recovering period were also tested
this probably corresponds to a low profit scenario and showed that improvement/changes in optimal
because high densities would need large amount densities would be possible, but would not result
of algae to feed the copepod population and will in greater production yields. Improvements would
have a correspondingly low egg production per only affect the densities at which the maximum
individual (Fig 1f). Under optimal conditions, Acar- egg production is reached. Long-term monocul-
tia tonsa must be raised for 10–15 days from eggs tures of copepods, where predators are absent and
to adults and females will produce eggs for approx- there is no foraging pressure, likely naturally
imately 3–4 weeks. It is therefore important to result in selection at high densities, as proposed by
optimize the production of each copepod raised Tiselius, Hansen, Jonsson, Kiørboe, Nielsen, Piont-
until the reproductive stage. kovski & Saiz (1995).
This model tests the effects of potential trait Improvement of SGR was a good solution to
selection on copepods. Selection is one potential improve production yields in cultures, but did not
way to improve culture productivity (Drillet et al. necessarily result in selection. SGR could for
2011b), but traits to select on should be defined, instance be improved by improving the food acces-
as all traits may not be transmittable and the sible to the copepods (Koski, Breteler, Schogt,
effects of such selection should be evaluated. In Gonzalez & Jakobsen 2006) or by using additives
crustacea, Briski, Van Stappen, Bossier & Sorgeloos with pro/pre-biotic effects in the water (Drillet,
(2008) selected fast hatching Artemia but work on Rabarimanantsoa, Frou€el, Lamson, Christensen,
copepods is rare (but see Lee, Kang, Park & Dahms Kim-Tiam & Hansen 2011a).
2012). The sensitivity of the model, which is
somehow equivalent to selection (though in the
Validity of the model
environment this sensitivity could as well be
attributed to several other parameters ignored here Comparing data extracted from the model to real
that could induce different responses in the observation was necessary to validate the model.
model), was also tested. Theoretically, when select- The relationship between observed and predicted
ing larger copepods, fewer adults are required to eggs production (Fig. 3b; plain line) was hardly

distinguishable from the identity line (dash line) Future improvements:

which validates the accuracy of the model. How-
Water mixing
ever, the variability in the production estimations
from the observed data was quite high and proba- To improve this model, the mixing properties of
bly due to the changing conditions in the culture the tanks based on its optimal design, natural con-
held at FSUML, where food quantity and quality vection or an air supply could be taken into
may have changed over time and was not account and could also be used to develop egg
controlled. harvesting mechanisms. This would necessitate
coupling a physical model to the biological one.
Limitations of the model Egg quality
By definition, a model presents a simplified version The effect of water quality and the potential effects
of what happens in a system and therefore one of chemicals on the egg quality produced were
must keep in mind the possible limitations. In the ignored in the system. Copepods have been shown
present model, the SGR increased with the temper- to produce diapause or delayed hatching eggs
ature following a Q10 value. By using a Q10, set- under high density (Ban & Minoda 1994; Camus
up did not require any maximum values while & Zeng 2009). Understanding these effects prop-
every species present an optimum range of temper- erly could help to improve the present model by
ature at which it productivity is the highest (e.g. calculating the optimal water exchange rate neces-
Lombard, Labeyrie, Lea, Spero & Michel 2009). sary in cultures to avoid the production of
Acartia tonsa has for instance an optimal produc- diapause eggs (in case A. tonsa also produces dia-
tion at 24.7°C (Holste & Peck 2006) and this may pause eggs under high density conditions as it was
vary as for every life history parameter from one shown for other species of this genus). Neverthe-
population to another as reported by Drillet, Jep- less, other water quality parameters may also be
sen, Højgaard, Jørgensen & Hansen (2008). improved by an adequate recirculation of the
Because of the limitation inherent to the use of a water.
Q10, and confirmed by previous experimental
Food requirement and financial efficiency
observations, it is likely that the model would be
accurate at the extremes of the temperature range In the present model, copepods are fed optimally
(<10°C and >25°C). and therefore it ignores the potential effects of food
Also, it is a possibility that an egg is preyed quality changes/fluctuations that may affect cope-
upon and released due to another encounter pods productivity and behaviour. It also does not
resulting in sloppy feeding and mortality of eggs. integrate the commercial efficiency of using differ-
This would decrease production or at least the ent food sources though these inter-linked parame-
apparent hatching rate of the eggs. This is not ters could offer a very good insight on the
taken into account in the model because of the commercialisation capacity of copepod culture
complexity and the absence of data to paramete- systems (Drillet et al. 2011b).
rise these processes. However, this is very likely in
culture tanks and the second peak of production
Conclusion
under high adult density may be an artefact of the
model. A simple model was produced to understand the
Finally, in the present model, A. tonsa is consid- limitation of production yields in batch cultures of
ered as an obligatory filter feeder because the food the calanoid copepod A. tonsa. As proven there is
conditions used for the culture of this copepod are an optimal density of copepods which changes
considered optimal and based on small flagellates. with the design of the culture tanks to reach the
However, under contamination conditions such as maximal egg harvest. Cannibalism impacts the
when the ciliates Euplotes sp. are present in cul- productivity of the system with increasing density
tures, A. tonsa behaviour may switch to a more of adults, but at very high density the disturbance
raptorial mode and this explains the proposal of of adults with each other becomes the main con-
decreasing the algal feeding to regulate the pres- cern, as the foraging time is decreased. At these
ence of ciliates in cultures (Jonsson & Tiselius very high densities cannibalism may still occur
1990; Drillet & Dutz 2013). but is likely underestimated by the model because

it does not account for sloppy feeding/partial Busch K.E.T., Falk-Petersen I.B., Peruzzi S., Rist N.A. &
cannibalism. Hamre K. (2010) Natural zooplankton as larval feed in
Using the present model to evaluate the efficacy intensive rearing systems for juvenile production of
of selection of copepod strain was also shown to be Atlantic cod (Gadus morhua L.). Aquaculture Research
41, 1727–1740.
a promising tool. Larger copepods produce more
Camus T. & Zeng C. (2009) The effects of stocking den-
eggs but must be stocked in lower densities and the
sity on egg production and hatching success, cannibal-
range of densities that results in high egg harvest is ism rate, sex ratio and population growth of the
limited (and vice versa). Selecting copepods that can tropical calanoid copepod Acartia sinjiensis. Aquaculture
cope well with high encountering rate of conspecif- 287, 145–151.
ics mainly affect the density at which a second peak Drillet G. & Dutz J. (2013) Dealing with the presence of
is appearing in the model, and because of the ques- the ciliate Euplotes sp. in cultures of the copepod Acar-
tions remaining at this densities for which no data tia tonsa. Aquaculture International. doi:10.1007/
are available, it is proposed to focus on improving s10499-013-9647-4.
SGR as well as egg harvest systems. Drillet G., Iversen M.H., Sørensen T.F., Ramløv H., Lund
This is the first model of this kind and it shows T. & Hansen B.W. (2006) Effect of cold storage upon
eggs of a calanoid copepod, Acartia tonsa (Dana) and
that using such a tool for improving copepod pro-
their offspring. Aquaculture 254, 714–729.
ductions is very useful, but requires having access
Drillet G., Lindley L.C., Michels A., Wilcox J. & Marcus
to literature describing basic life history traits of N.H. (2007) Improving cold storage of subitaneous
the species of interest. eggs of the copepod Acartia tonsa Dana from the Gulf
of Mexico (USA-Florida). Aquaculture Research 38,
457–466.
Acknowledgements
Drillet G., Jepsen P.M., Højgaard J.K., Jørgensen N.O.G. &
We are grateful to PM Jepsen and Linda Holste as Hansen B.W. (2008) Strain specific vital rates in four
well as the other reviewers for their comments on Acartia tonsa cultures II: life history traits and bio-
a previous version of the manuscript and to NH chemical contents of eggs and adults. Aquaculture 279,
Marcus for allowing us using data produced at the 47–54.
Drillet G., Rabarimanantsoa T., Frou€el S., Lamson J.S.,
Florida State University Marine Laboratory where
Christensen A.M., Kim-Tiam S. & Hansen B.W.
research was supported by Florida Sea Grant to
(2011a) Do microbial preparations improve life history
NH Marcus. The present work was supported by traits of the copepod Acartia tonsa? Marine Biotechnol-
the Danish Ministry for Independent Research, ogy 13, 831–836.
Ung Elite Forsk grants 10-093759 and Drillet G., Frou€el S., Sichlau M.H., Jepsen P.M., Højgaard
10-094773 to G Drillet and the Agence Nationale J.K., Joarder A.K. & Hansen B.W. (2011b) Status and
de la Recherche project ANR-10-PDOC-005-01 recommendations on marine copepod cultivation for
‘Ecogely’ grant to F Lombard. M Lilley and Nichole use as live feed. Aquaculture 315, 155–166.
Chan corrected the English. Duarte C.M., Marb a N. & Holmer M. (2007) Rapid
domestication of marine species. Science 316, 382–
383.
References Fisher R.A. (1930) The Genetical Theory of Natural Selec-
tion. Oxford University Press, New York.
Ban S. & Minoda T. (1994) Induction of diapause egg
Gutierrez-Estrada J.C., Pulido-Calvo I., de la Rosa I. &
production in Eurytemora affinis by their own metabo-
Marchini B. (2012) Modeling inflow rates for the water
lites. Hydrobiologia 292, 185–189.
exchange management in semi-intensive aquaculture
Bergreen U., Hansen B. & Kiørboe T. (1988) Food size
ponds. Aquaculture Engineering 48, 19–30.
spectra, ingestion and growth of the copepod Acartia
Hansen B.W., Drillet G., Kozmer A., Madsen K.V., Peder-
tonsa during development: implications for determina-
sen M.F. & Sørensen T.F. (2010) Temperature effects
tion of copepod production. Marine Biology 99, 341–
on copepod egg hatching: does acclimatization matter?
352.
Journal of Plankton Research 32, 305–315.
boe T., Mohlenberg F. & Riisg ard H. U. (1985) In situ
Hirst A.G. & Bunker A.J. (2003) Growth of marine plank-
feeding rates of planktonic copepods: a comparison of
tonic copepods: global rates and patterns in relation to
four methods. Journal of Experimental Marine Biology
chlorophyll a, temperature, and body weight. Limnolo-
and Ecology 88, 67–81.
gyand Oceanography 48, 1988–2010.
Briski E., Van Stappen G., Bossier P. & Sorgeloos P.
Holmstrup M., Overgaard J., Sørensen T.F., Drillet G.,
(2008) Laboratory production of early hatching Art-
Hansen B.W., Ramløv H. & Engell-Sørensen K. (2006)
emia sp cysts by selection. Aquaculture 282, 19–25.

Influence of storage conditions on viability of quiescent rates of planktic foraminifera. Marine Micropaleontology
copepod eggs (Acartia tonsa Dana): effects of tempera- 70, 1–8.
ture, salinity and anoxia. Aquaculture Research 37, Marcus N.H. & Murray M. (2001) Copepod diapause
625–631. eggs: a potential source of nauplii for aquaculture.
Holste L. & Peck M. (2006) The effects of temperature Aquaculture 201, 107–115.
and salinity on egg production and hatching success of Miller D.D. & Marcus N.H. (1994) The effects of salinity and
Baltic Acartia tonsa (Copepoda: Calanoida): a laboratory temperature on the density and sinking velocity of eggs of
investigation. Marine Biology 148, 1–10. the calanoid copepod Acartia tonsa Dana. Journal of Experi-
Jonsson P.R. & Tiselius P. (1990) Feeding behavior, prey mental Marine Biology and Ecology 179, 235–252.
detection and capture efficiency of the copepod Acartia Næss T., Germain-Henry M. & Naas K.E. (1995) First
tonsa feeding on planktonic ciliates. Marine Ecology feeding of Atlantic halibut (Hippoglossus hippoglossus)
Progress Series 60, 35–44. using different combinations of Artemia and wild zoo-
Koski M., Breteler W.K., Schogt N., Gonzalez S. & Jakob- plankton. Aquaculture 130, 235–250.
sen H.H. (2006) Life-stage specific differences in exploi- Rajkumar M. & Vasagam K.P.K. (2006) Suitability of the
tation of food mixtures: diet mixing enhances copepod copepod, Acartia clausi as a live feed for seabass larvae
egg production but not juvenile development. Journal (Lates calcarifer Bloch): compared to traditional live-
of Plankton Reseach 28, 919–936. food organisms with special emphasis on the nutri-
Larsen P.S., Madsen C.V. & Riisg ard H.U. (2008) Effect of tional value. Aquaculture 261, 649–658.
temperature and viscosity on swimming velocity of the Tiselius P., Hansen B., Jonsson P., Kiørboe T., Nielsen
copepod Acartia tonsa, brine shrimp Artemia salina and T.G., Piontkovski S. & Saiz E. (1995) Can we use labo-
rotifer Brachionus plicatilis. Aquatic Biology 4, 47–54. ratory reared copepods for experiments: a comparison
Lee K.W., Kang J.H., Park H.G. & Dahms H.-U. (2012) of feeding behavior and reproduction between a field
Effect of strain and selection line on the fecundity of and a laboratory population of Acartia tonsa. ICES Jour-
the cyclopoid copepod Paracyclopina nana Smirnov, nal of Marine Sciences 52, 369–376.
1935. Aquaculture Research 44, 50–58. Wilcox J.A., Tracy P.L. & Marcus N.H. (2006) Improving
Leising A. & Yen J. (1997) Spacing mechanisms within live feeds: effect of a mixed diet of copepod nauplii
light-induced copepod swarms. Marine Ecology Progress (Acartia tonsa) and rotifers on the survival and growth
Series 155, 127–135. of first feeding larvae of the Southern Flounder, Para-
Liang T., Uye S. & Onbe T. (1994) Production and loss of lichthys lethostigma. Journal of the World Aquaculture
eggs in the calanoid copepod Centropages abdominalis Society 37, 113–120.
Sato in Fukuyama Harbor, the Inland Sea Japan. Bulle- Xie X. & Li J. (2012) Modeling, analysis and continuous
tin of the Plankton Society of Japan 41, 77–119. improvement of food production systems: a case study
Lombard F., Labeyrie L., Lea D., Spero H. & Michel E. at a meat shaving and packaging line. Journal of Food
(2009) Modelling the temperature dependent growth Engineering 113, 344–350.

Aquaculture 201 Ž2001. 329–342
www.elsevier.comrlocateraqua-online
Intensive cultivation of the calanoid copepod

Gladioferens imparipes
M.F. Payne, R.J. Rippingale )
Department of EnÕironmental Biology, School of Resource Science and Technology, Curtin UniÕersity of
Technology, GPO Box U1987, Perth, 6845, Australia
Received 20 September 2000; accepted 5 March 2001
Abstract
In aquaculture, inclusion of copepod nauplii in the diet of larval fish may increase the number
of fish species that can be successfully reared. Pelagic nauplii of calanoid copepods are more
available as food items for most marine fish larvae than the predominantly epibenthic nauplii of
harpacticoid copepods. However, sustained and substantial production of calanoid copepod nauplii
is difficult. Intensive copepod cultures allow reliable and sustained production while utilising less
space than extensive culture systems. Three intensive culture techniques for the temperate water
estuarine calanoid copepod Gladioferens imparipes are described. These techniques comprise 60-
and 500-l batch cultures and 1000-l semi-continuous cultures. Five hundred-litre cultures are part
of a recirculating system that features automated nauplius collection and water treatment.
Standardised nauplius production, expressed as number of nauplii produced per litre of culture
vessel per day, is given for each technique. 878 " 46 naupliirl culture vesselrday was achieved
over 420 days for G. imparipes cultured in 500-l vessels. Standardised nauplius production is
compared with published data from other copepod culture systems. q 2001 Elsevier Science B.V.
All rights reserved.
Keywords: Calanoid; Copepod; Culture
1. Introduction
With the increasing world-wide interest in aquaculture, finfish of various species are
under regular review as candidates for intensive cultivation. Many of these fish, such as
those in the family Sparidae, have larvae that are readily reared using rotifers as food.
)
Corresponding author. Tel.: q61-8-9266-7922; fax: q61-8-9266-2495.
E-mail address: R.Rippingale@info.curtin.edu.au ŽR.J. Rippingale..
0044-8486r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 4 - 8 4 8 6 Ž 0 1 . 0 0 6 0 8 - 1
330 M.F. Payne, R.J. Rippingaler Aquaculture 201 (2001) 329–342
However, there is an expanding list of species for which rotifers alone are not suitable
for larviculture. Some of these include fishes of the Lutjanidae ŽDoi et al., 1994; Schipp
et al., 1999. and Serranidae ŽToledo et al., 1997, 1999. families. When copepod nauplii
have been provided as food, larval growth and survival has improved for many of these
fish.
Use of cultured copepods in fish larviculture has mainly involved calanoids: e.g.
Acartia spp. ŽToledo et al., 1997; Schipp et al., 1999.; Eurytemora spp. ŽWitt et al.,
1984; Shields et al., 1999.; Pseudodiaptomus spp. ŽDoi et al., 1997; Toledo et al., 1999.
and harpacticoids: e.g. Euterpina acutifrons ŽKraul et al., 1992.; Tisbe spp. ŽStøttrup
and Norsker, 1997; Nanton and Castell, 1998.; Trigriopus japonicus ŽFukusho et al.,
1980. with the highest density copepod cultivation being achieved with harpacticoids.
However, calanoid nauplii are entirely pelagic and, as prey items, are more available to
pelagic fish larvae than are the nauplii of most harpacticoids, which may spend less time
in the water column and therefore are less readily predated. Also, the nauplii of many
calanoids are small enough to be taken as food by larvae of fish such as the lutjanids and
serranids, which have small gapes at the time of first feeding. Production of harpacticoid
copepods is clearly of value for larviculture, but the particular benefits that can be
provided with calanoids warrant continued efforts towards achieving the effective
cultivation of these copepods.
Commercial marine larviculture requires a reliable and substantial supply of live
food. Calanoid copepods, unlike rotifers Žand to a lesser extent harpacticoids., have
proved difficult to produce in consistently large quantities. One approach has been to
culture copepods in outdoor tanks ŽOgle, 1979; Ohno and Okamura, 1988.. However,
aside from space requirements, lack of precise control over food and water quality in
these tanks results in variable copepod production ŽOhno and Okamura, 1988.. More
closely controlled conditions may allow intensive copepod cultures to have consistent
production while minimising the requirement for space. These are important features of
intensive copepod systems developed for both harpacticoids ŽSun and Fleeger, 1995;
Støttrup and Norsker, 1997. and calanoids ŽStøttrup et al., 1986; Schipp et al., 1999..
Intensive calanoid copepod culture systems can be designed such that eggs or young
nauplii are selectively removed from broodstock cultures on a regular basis. This is an
efficient technique as it provides quantities of uniform-sized nauplii for feeding to larval
fish and ensures that most of the food added to cultures containing reproductive adults is
utilised for nauplius production rather than somatic growth of juvenile stages. Also,
many omnivorous calanoids, e.g. Acartia spp, will consume their own nauplii ŽLonsdale
et al., 1979.; hence, it is prudent to separate them from adults as soon as possible. Daily
separation of eggs from adult Acartia tonsa is featured in the culture system described
by Støttrup et al. Ž1986.. Egg or nauplius separation is important but it is time
consuming. Hence, the development of automated procedures for copepod culture
ŽStøttrup and Norsker, 1997..
The temperate water estuarine calanoid Gladioferens imparipes has shown potential
for intensive culture ŽRippingale and MacShane, 1991.. Much is known about this
species, including its biology and ecology ŽRippingale and Hodgkin, 1974a,b., perfor-
mance on different diets ŽTakano, 1971; Payne and Rippingale, 2000a. and efficacy as a
food for fish ŽPayne and Rippingale, 2000b.. This paper describes techniques developed
M.F. Payne, R.J. Rippingaler Aquaculture 201 (2001) 329–342 331
for intensive culture of G. imparipes in water volumes from 60 to 1000 l, with

automation to reduce labour and improve efficiency.
2. Materials and methods
G. imparipes used in this study were collected in 1995 from the upper reaches of the
Swan River estuary ŽPerth, Western Australia. using a plankton net Ž150 mm. towed
near the surface at dusk. Copepodid and adult stages of G. imparipes dominated the
sample, thus it was not necessary to isolate the species. Animals were maintained in
continuous laboratory culture for 5 years during and prior to the present work.
2.1. Water quality
In all cultures, copepods were reared in clean, filtered Ž1 mm. seawater diluted to
approximately 17‰ with either deionised water Ž60- and 500-l cultures. or tap water
filtered through activated carbon Ž1000-l cultures.. Aeration and water circulation was
achieved via one Ž60-l cultures. or two Ž500- and 1000-l cultures. air lift pipes attached
to the sides of vessels. Ammonia, nitrite, pH, and dissolved oxygen were measured
regularly in most copepod cultures.
2.2. Stocking density and feed rate
Copepod nauplii were stocked at 1rml in all culture vessels and provided with either
Isochrysis galbana ŽT-Iso; CSIRO culture code CS-177. or a combination of I. galbana
and Rhodomonas baltica ŽCSIRO culture code CS-201. as food. Feed rates in batch
cultures Ž60- and 500-l cultures. generally followed a standard protocol; young nauplii
were fed at the rate of 2 = 10 4 cellsrmlrday increasing gradually to 1 = 10 5
cellsrmlrday when copepods reached maturity. This higher feed rate was maintained
during the adult life of the copepods while nauplii were being produced. This higher
feed rate was varied daily in response to a subjective assessment of water clarity in
copepod culture vessels prior to the addition of food. If the water appeared turbid, the
feed rate was decreased to 6–8 = 10 4 cellsrmlrday and if it appeared very clear, the
rate was increased to 1.2–1.4 = 10 5 cellsrmlrday. Increased water turbidity was caused
by suspended particulate matter rather than the presence of uneaten algal cells. Adult
copepod densities in semi-continuous 1000-l cultures increased above those in the batch
cultures; hence, these cultures were fed at the higher rate of 2 = 10 5 cellsrmlrday. This
feed rate was also adjusted as described above.
2.3. Copepod culture techniques
2.3.1. 60-l stock cultures

Stock cultures were maintained as batch cultures at Curtin University ŽPerth, Western
Australia.. Copepods were maintained in two 60-l black plastic drums at 19 " 28C. The
microalga I. galbana, grown non-axenically in batch cultures in 5-l flasks ŽPyrex w .
using fr2 media ŽGuillard, 1972., was provided as food. Nauplius collection com-
menced when nauplii were observed in the culture water and was effected daily by
siphoning half the volume of water from the drums through a submerged 150-mm
screen. Drum volume was made up using clean, diluted seawater. The volume of
removed water containing nauplii was reduced to approximately 2 l by siphoning
through a submerged 53-mm screen, giving a relatively high density of copepod nauplii.
Collection of nauplii from copepod cultures continued until nauplius numbers began to
decline. At this point, the drums were drained, cleaned and restocked using recently
collected nauplii. Throughout the nauplius collection stage, detritus was siphoned from
the bottom of the drums regularly. Those adults removed during cleaning were held until
the detritus had settled and then poured back into the stock culture. Maintaining two
copepod stock cultures at different stages of development ensured that cultures did not
senesce at the same time.
2.3.2. Automated 500-l batch cultures

This batch culture system was established at Curtin University ŽPerth, Western
Australia.. It comprised two 500-l cylindro-conical fibreglass culture vessels, a 200-l
plastic harvest vessel and a 500-l fibreglass sump ŽFig. 1.. Water quality in the system
was maintained by intermittent recirculation. The system featured an automated nauplius
Fig. 1. Schematic diagram of the automated 500-l copepod culture system. Nauplius collection and water
recycling procedures are described as follows: Ž1. Harvest sequence initiated by pressing the ‘start’ button on
the PLC device. Ž2. Air supply to the culture vessels turned off for 5 min to stop water movement. Ž3. Light
turned on for 10 min to illuminate 150-mm screen submerged in culture vessel. Ž4. Solenoid valve opened so
that water Žcontaining nauplii. siphons through 150- mm screen into harvest vessel. Ž5. Float switch ŽFS. a1
activated which closes solenoid value, turns light off and turns pump a1 on. Ž6. FS a2 activated which turns
pump a1 off and pump a2 on Ždrawing through 53-mm screen.. Ž7. FS a3 deactivated which turns pump a2
off and air supply on. Ž8. Concentrated nauplii drained from harvest vessel.
collection and water treatment procedure. Water was maintained at 23 " 18C and
deionised water was used to replace water lost via evaporation.
Copepods were fed with I. galbana from batch algal cultures grown non-axenically
in 20-l carboys ŽNalgene w . on fr2 media. Copepod culture tanks were maintained as
static for approximately the first 21 days as copepods reached maturity. When ovigerous
females were abundant, nauplius collection commenced and continued daily until
nauplius numbers decreased Žapproximately 21 days.. During this time, detritus was
siphoned regularly from the bottom of vessels and those adults removed during cleaning
were returned to the culture vessel. Two copepod cultures were maintained concurrently
but not in synchrony; while one static culture was growing to maturity, the other mature
culture was subject to daily nauplius collection. Vessels containing senescent copepods
were drained, cleaned, refilled with clean, diluted seawater and restocked with nauplii.
Senescent adult copepods were enumerated at the termination of six separate 500-l
cultures.
Nauplius collection and water recycling procedures were fully automated, controlled
by a programmable logic control ŽPLC. device ŽToshibaw EX20.. These procedures are
illustrated and described in Fig. 1. Procedures incorporated light to attract nauplii
Fig. 2. Schematic diagram of a 1000-l copepod culture system. Nauplius collection and water exchange
procedures are described as follows: Ž1. Valve opened allowing half the volume of 1000-l culture vessel to
drain through 150-mm screen into submerged 53-mm screen. Pre-set valve ensures slow drainage Ž ;10
lrmin.. Ž2. When draining complete, valve closed and nauplii rinsed from 53-mm screen. Ž3. Culture vessel
refilled with diluted seawater from reservoir. Ž4. Copepods fed with algae.
towards a removal point, 150-mm screen to selectively retain adult copepods in culture
vessels whilst allowing the passage of nauplii and 53-mm screen to retain nauplii in the
collection vessel. Lids on culture vessels completely shaded copepods from ambient
light and a cartridge filter Ž5 mm. removed particulates from culture water as it was
transferred from collection vessel to sump ŽFig. 1.. Four open-ended cylinders of plastic
mesh Ž20 Žd. = 30 Žh. cm; 3 mm mesh. were floated in each culture vessel to increase
the internal surface area available for copepod attachment. Submerged foam filters,
operated by gentle air-lift, were located near the bottom of each culture vessel and were
very effective in collecting copepod faecal pellets while not trapping copepod nauplii.
These filters were cleaned daily. Water in the sump was continually air-lifted through a
biofilter and pumped through a foam fractionator. Algae was added to the culture vessels
to feed the adults after the nauplii were collected.
2.3.3. 1000-l semi-continuous cultures

This semi-continuous culture system was established at the Aquaculture Development
Unit ŽPerth, Western Australia.. The system comprised two 1000-l cylindro-conical
fibreglass vessels and a 200-l plastic barrel ŽFig. 2.. Water quality was maintained by
intermittent water flow-through. Water temperature was maintained at 21 " 28C.
Copepods were fed with I. galbana and R. baltica grown non-axenically as batch
cultures in 1000-l plastic tanks ŽNylex w . on fr2 media. Each day copepods were fed a
combination of I. galbana and R. baltica Ž3:1 by cell numbers.. Every 21 days,
copepodids and adult copepods were collected from culture vessels by slow draining
into a collection bucket with sides of 150-mm screen. These copepods were rinsed
gently for 5 min with diluted seawater and returned to the vessel after it was cleaned and
refilled.
Nauplius collection and water exchange procedures were performed manually approx-
imately every second day. These procedures are illustrated and described in Fig. 2.
Culture vessels contained submerged air-lift foam filters that were cleaned daily.
3. Results
3.1. Water quality
On average, pH and dissolved oxygen were within acceptable limits in all three
culture techniques ŽTable 1.. Ammonia and nitrite concentrations were generally high.
Table 1
Ammonia, nitrite, pH and dissolved oxygen ŽDO. recorded in three copepod culture systems: 60-l batch, 500-l
batch and 1000-l semi-continuous
Culture method Volume Ammonia Žppm., Nitrite Žppm., pH, DO Ž%.,
Žl. mean Žrange. mean Žrange. mean Žrange. mean Žrange.
Batch 60 not measured 0.2 Ž0–0.8. 8.1 Ž7.6–8.3. 93 Ž83–104.
Batch 500 0.2 Ž0–0.7. 0.1 Ž0–0.8. 8.0 Ž7.8–8.3. 95 Ž80–100.
Semi-continuous 1000 0.1 Ž0–0.3. not measured 7.8 Ž7.5–8.1. 97 Ž90–100.
Table 2
Summary of nauplius production from 60-l copepod batch cultures maintained at Curtin University
Total mean daily production expressed as mean"s.e.
Batch Duration of Duration of nauplius Daily nauplius collection, Total
no. culture Ždays. collection Ždays. mean Žrange. =10 3 nauplii =10 6
1 42 21 49 Ž3–133. 1.0
2 44 20 76 Ž5–142. 1.5
3 39 17 76 Ž15–156. 1.3
Total days: 58. Mean daily collection: 67"4.
3.2. Nauplius production
Nauplius production was standardised to the number of nauplii Žmean " s.e.. pro-
duced per litre of culture vessel volume per day Žnaupliirl culture vesselrday. to allow
comparison between culture systems of different volume.
3.2.1. 60-l stock cultures

Batch copepod cultures grown in 60-l drums produced an average of 67,000 " 4000
naupliirday for 58 days ŽTable 2.. This equates to 1117 " 59 naupliirl culture
vesselrday.
Table 3
Summary of nauplius production from 500-l copepod batch cultures maintained at Curtin University
The culture system incorporated automatic nauplius collection and intermittent water recycling. Mean daily
nauplius collection Žtotal expressed as mean"s.e.. was averaged over the entire nauplius collection period,
which included some days on which no collection was undertaken.
Batch Duration of Duration of nauplius Daily nauplius collection, Total
1 62 30 411 Ž112–1106. 12.3
2 61 40 540 Ž65–892. 21.6
3 59 34 538 Ž105–1140. 18.3
4 57 35 399 Ž71–767. 14.0
6 25 12 304 Ž74–636. 3.6
7 41 25 147 Ž70–233. 3.7
8 46 29 466 Ž79–1018. 13.5
9 42 24 551 Ž156–1113. 13.2
10 41 21 605 Ž280–966. 12.7
12 46 27 423 Ž144–713. 11.4
13 42 22 475 Ž290–648. 10.4
14 38 22 477 Ž90–925. 10.5
15 39 22 312 Ž20–546. 6.9
16 45 25 431 Ž174–640. 10.8
17 52 30 565 Ž173–912. 16.9
18 43 22 375 Ž164–1001. 8.3

Table 4
Summary of nauplius production from 1000-l copepod semi-continuous cultures maintained at the Aquaculture
Development Unit. Mean daily nauplius collection Žtotal expressed as mean"s.e.. was averaged over the
entire nauplius collection period, which included many days on which no collection was undertaken
Culture Duration of Duration of nauplius Daily nauplius collection, Total
1 156 121 406 Ž0–2800. 49.5
2 148 121 634 Ž0–4200. 83.6
3.2.2. Automated 500-l batch cultures

Automated batch copepod cultures grown in 500-l vessels produced an average of
439,000 " 23,000 naupliirday for 420 days ŽTable 3.. This equates to 878 " 46
naupliirl culture vesselrday. Survival averaged 29 " 9% at the termination of six
separate cultures.
3.2.3. 1000-l semi-continuous cultures

Copepods grown in 1000-l semi-continuous systems produced an average of 520,000
" 15,000 naupliirday for 184 days ŽTable 4.. This equates to 520 " 15 naupliirl
culture vesselrday.
3.3. Culture contamination
In total, six copepod cultures failed to reach nauplius production stage. In each case,
culture water was cloudy and had an unpleasant odour prior to copepod mortality. Also,
algal cultures used to feed these copepod cultures had an unpleasant odour and
contained large amounts of sediment. Chlorinationrdechlorination of copepod culture
water prior to stocking and improved algal hygiene eliminated this problem.
Poor nauplius production occurred when copepod cultures were contaminated with
harpacticoid copepods, mostly Tisbe sp. and Amphiascoides sp. When left untreated,
harpacticoid copepods completely replaced G. imparipes. Harpacticoids were readily
removed from G. imparipes cultures by rapidly reducing salinity from 17‰ to 5‰ and
maintaining this low salinity for 24 h.
Rotifers Ž Brachionus sp.. were observed in at least four copepod cultures. However,
in each case all rotifers were removed along with nauplii when the collection period
commenced.
4. Discussion
In the present work, standardised nauplius collection rates from intensive cultures of
G. imparipes appear greater than published data for calanoid culture systems using
Acartia spp. ŽTable 5.. However, definitive comparisons cannot be made as measures of
Table 5
Comparison of standardised copepod production Žcopepodsrl of culture vesselrday. from intensive culture
systems. The duration of the nauplius collection period is included. Copepod life cycle stages that were
collected are indicated in parentheses after production figures. Culture vessel volumes include all concurrent
vessels that contributed to total production
Copepod Culture vessel Duration of Standardised copepod production Reference
species volume Žl. collection Ždays. Žcopepodsrl culture vesselrday.
Calanoids
Acartia tonsa 4=1890 16 91 Žall stages. Ogle Ž1979.
A. tonsa 2=200 ; 28 213 Žnauplii. Støttrup et al.
Ž1986.
Gladioferens 2=1000 242 520"15 Žnauplii. present study
imparipes
G. imparipes 500 420 878"46 Žnauplii. present study
Harpacticoids
Amphiascoides 4=2000, 49 1144 Žall stages. Sun and Fleeger
atopus 2=430 Ž1995.
Tisbe holothuriae 150 43 3333 Žnauplii and copepodids. Støttrup and
Norsker Ž1997.
T. holothuriae 3 53 100,000 Žnauplii. Støttrup and
Norsker Ž1997.
variation for published production figures either are not provided or cannot be calculated
from the published data. Copepod stocking density and feed rate were considerably
higher in the present study than those used by Støttrup et al. Ž1986. to culture Acartia
tonsa. Recovery of A. tonsa nauplii was undoubtedly increased by daily separation of
eggs from cannibalistic broodstock. The technique used by Ogle Ž1979. cannot be
regarded as intensive as copepods were grown in water pumped from the ocean and
without food supplementation. However, it still provides an interesting comparison.
Schipp et al. Ž1999. describe a 1000-l culture system for Acartia spp. that did not
include daily removal of nauplii. Standardised nauplius production in this system Žbased
on increase of nauplius density over time. is estimated at a maximum of 440 naupliirl
culture vesselrday.
Collection of G. imparipes nauplii from 500- and 1000-l vessels was conducted over
a considerably longer period than in other studies ŽTable 5. and included sustained
periods of both below and above average nauplius production. Thus, the average
production rate is a true representation of nauplius production that can be expected from
G. imparipes cultures established in commercial facilities.
Standardised nauplius yields from intensive copepod cultures are generally higher for
harpacticoid copepods than for calanoid copepods ŽTable 5.. This results from the
generally lower adult densities achieved in calanoid cultures than for harpacticoid
cultures of similar volume. This may be explained by the differing ecology of the two
copepod classes; harpacticoid copepods are mostly epibenthic detritivores while calanoid
copepods are generally planktonic suspension feeders. Under culture conditions,
harpacticoid population density is limited by available internal surface area whereas
calanoid population density is generally limited by water volume. Internal surface area
in small vessels is readily increased by the addition of substrate such as plastic balls
ŽStøttrup and Norsker, 1997. to allow for higher populations of harpacticoids. In
contrast, increased calanoid populations generally require vessels of larger volume,
which places significant demands on space and water supply.
The success of intensive G. imparipes cultures may be partially explained by a
behaviour pattern which is unusual for calanoids. Like harpacticoids, late copepodid and
adult G. imparipes hold on to internal surfaces ŽRippingale, 1994.. However, unlike
most harpacticoids, G. imparipes continue to feed in the water column whilst remaining
attached to the substrate by the dorsal surface. This may allow energy investments in
growth and reproduction rather than free-swimming movement. Hence, high density
populations of this calanoid species are likely to benefit from increased internal surface
area. While G. imparipes were observed holding on to plastic mesh substrate in 500-l
vessels, quantified internal surface area was not correlated to copepod productivity. This
requires further examination.
In this study, G. imparipes were cultured in vessels of three different volumes. Initial
nauplius stocking densities were the same for each vessel and yet the number of nauplii
produced per litre of culture vessel per day decreased as vessel volume increased. Also,
while stocking density and feed rate remained constant in 60- and 500-l vessels,
increased copepod density and feed rate in 1000-l vessels still resulted in these vessels
recording the lowest standardised nauplius production rate. This trend also appears very
distinctly in the two harpacticoid culture systems reported by Støttrup and Norsker
Ž1997. ŽTable 5.. In copepod species that utilise internal surface area, such as G.
imparipes and harpacticoids, increased productivity in smaller vessels is probably a
function of increased surface area to volume ratio in these vessels. The highest
production rate of 100,000 naupliirl culture vesselrday was achieved by Støttrup and
Norsker Ž1997. in 3-l flat trays that would have a very high surface area to volume ratio.
There is insufficient published data of nauplius production from calanoid culture
systems to determine if this trend extends to fully pelagic copepod species. For G.
imparipes, this further indicates the need for continued work on the effect of increased
surface area on survival from nauplius to adulthood and nauplius production. It is
possible that increased yields of copepod nauplii is best achieved by greater numbers of
smaller culture vessels rather than fewer vessels of large volume.
Because calanoid copepods are generally regarded as having low tolerance to poor
water quality ŽHoff and Snell, 1987., efforts were focussed on reducing nitrogenous
wastes in G. imparipes cultures. Buttino Ž1994. determined that ammonia concentrations
of 0.12 ppm reduced egg production and viability in the calanoid Acartia clausi. In the
current study, ammonia concentrations exceeded this for much of the time, hence, are
likely to have affected copepod productivity, particularly survival from nauplius to
adulthood and fecundity. Determining the effect of ammonia and nitrate concentration
on G. imparipes survival and fecundity would allow much more precise culture
management. Submerged air-lift foam filters in 500- and 1000-l G. imparipes cultures
improved water quality. With gentle air-lift, these filters were very effective in collect-
ing copepod faecal pellets thus allowing their removal before degradation, and associ-
ated release of dissolved nitrogenous compounds was complete. Also, high concentra-
tions of non-nutritious suspended particulate matter may have decreased fecundity in G.

imparipes, as has been reported for the estuarine calanoid Eurytemora affinis ŽGasparini
et al., 1999..
In the present study, water quality was primarily maintained in the stock Ž60 l. and
1000-l cultures by intermittent water flow-through and in the 500-l culture vessels by
intermittent water recirculation. The latter incorporated a biofilter and a foam fractiona-
tor. Intermitted and continuous water flow-through was used by Støttrup and Norsker
Ž1997. to maintain water quality in culture systems, whilst Sun and Fleeger Ž1995. used
a continuous recirculating system. Intermittent water flow in copepod culture vessels, as
was used in the current work, has the benefit of allowing copepods to consume all added
food without wastage as some is lost from the vessel in a continuous flow. While
dedicated recirculating systems are more expensive to build and maintain than flow-
through systems, they are of considerable benefit to facilities with limited access to
clean water. Also, strict control of the culture environment prevents the negative effects
of unavoidable fluctuations in flow-through water quality.
Automation of the 500-l batch cultures proved very efficient. The automated nauplius
collection sequence was programmed to start at the same time each day. At the end of
the sequence, approximately 30 min was required to feed cultures, clean screens and
filters and count nauplii. Further advantages would be gained by incorporating an
automated feeding procedure, as used by Støttrup et al. Ž1986. to feed their Abasis
culturesB and Støttrup and Norsker Ž1997.. Despite the lack of automation, 1000-l
semi-continuous cultures also required little attention. Nauplii were collected from two
tanks concurrently by manually opening both valves and leaving tanks to drain.
Subsequent refilling, feeding, filter cleaning and nauplius counting for both tanks
required approximately 45 min.
In 500-l cultures, light was used to concentrate G. imparipes nauplii prior to
collection. When nauplius collection was conducted twice in succession, no nauplii were
collected on the second cycle, thus demonstrating the efficiency of this technique. With
complete removal, no nauplii remained in the culture vessel to grow to maturity, hence,
all nauplii collected from a single batch culture are offspring from the initial cohort.
With the senescence of this cohort, nauplius production declined rapidly and the batch
culture had to be discarded and restocked. In contrast, nauplius collection in 1000-l
vessels did not incorporate light to attract nauplii and presumably some nauplii remained
after collection. In consequence, nauplii production increased Žafter the initial cohort had
expired. when the next generation came to maturity. Thus, these cultures could be
maintained semi-continuously. Støttrup and Norsker Ž1997. also made use of light to
attract harpacticoid nauplii prior to collection. This culture system was continuous
suggesting that not all nauplii were removed during each collection.
I. galbana was chosen as the primary diet for G. imparipes cultures as this alga
maximises growth and fecundity for this copepod ŽPayne and Rippingale, 2000a.. While
R. baltica has not been properly assessed as a diet, it has previously supported good
productivity in G. imparipes cultures when used as a short-term replacement diet for I.
galbana. Also, Rhodomonas spp. have been used in combination with I. galbana in
other calanoid culture systems ŽStøttrup et al., 1986; Schipp et al., 1999.. Recently,
culture of some calanoid copepod species has been improved by provision of dinoflagel-
late diets ŽKleppel et al., 1998; Klein Breteler et al., 1999.. Use of these may improve
intensive culture of G. imparipes.
In the current study, mature stock and 500-l G. imparipes cultures were fed at a rate
that maximised fecundity. In some 500-l batch cultures, water clarity was maintained at
higher feed rates than others, probably reflecting varying survival of nauplii to adulthood
in different batches. Continued feeding at high rates when culture water was already
turbid prior to feeding resulted in increased ammonia concentrations and decreased
nauplius production. Thus, daily assessment of water clarity and adjustment of feed rate
was a simple but important technique for management of intensive G. imparipes
cultures. Quantified measurements of water clarity would allow standardisation of this
technique.
Survival of copepods from nauplius to adulthood was not recorded for G. imparipes.
Relatively low recorded survival at the termination of 500-l cultures was not surprising
considering that these cultures were senescing. Late copepodid and adult densities are
difficult to determine as very vigorous mixing is required to dislodge all animals from
internal surfaces. This treatment damages copepods which effectively reduces culture
production and longevity. Copepod cultures maintained in the present work were
required to produce nauplii for concurrent fish rearing trials, hence, culture disturbance
was minimised. In these cultures, subjective assessment of the number of adults attached
to internal surfaces was a useful indicator of copepod survival to adulthood. However, it
is clear that culture management would be improved if survival from nauplius to
adulthood is quantified in further studies.
Contamination of copepod cultures was not a significant problem in the present
study. Failure of some copepod cultures to reach nauplius production stage was probably
a result of the proliferation of bacteria in algal cultures and subsequent contamination of
copepod cultures. Chlorinationrdechlorination of copepod culture water prior to stock-
ing, used in the present study and by Schipp et al. Ž1999., and improved algal culture
hygiene was effective in preventing bacterial blooms. When in high numbers, harpacti-
coid copepods prevented G. imparipes adults from settling on internal surfaces and
eventually replaced them. Rapid reduction of salinity kills harpacticoids and not G.
imparipes as the latter is tolerant of low salinities ŽRippingale and Hodgkin, 1974b.,
presumably more so than the former. Rotifers are perhaps the most likely source of
contamination for copepod cultures in commercial facilities. Once established in a
copepod culture it is likely that they would rapidly out-compete copepods. Hence, it is a
significant benefit of the current G. imparipes culture systems that they do not allow
rotifer populations to become established.
No single copepod species will be suitable for rearing all fish species; a range of
copepods species is required to provide nauplii of different sizes which are tolerant to
different temperature and salinity regimes. Also, regional and international restrictions
on animal importrexports will limit availability of particular copepod species. Whilst
current culture techniques have been developed to suit particular copepod species, there
is little doubt that these systems, with or without minor modification, can be used
successfully to culture other copepod species with similar ecological requirements.
This study describes techniques required to reliably maintain intensive cultures of G.
imparipes of varying volumes and provides an accurate estimate of sustained nauplius
production from these cultures. When combined with quantitative growth and survival
data for fish larvae reared on a diet inclusive of G. imparipes nauplii, hatcheries can
assess the viability of incorporating G. imparipes cultures into their operations. Alterna-
tively, culture techniques described in this work may be adapted to suit the culture
requirements of other copepod species. As one example, this system will likely require
little, if any, modification to suit the culture of other Gladioferens species that occur in
eastern Australia and New Zealand.
Acknowledgements
Financial support was received from the Fisheries Research and Development
Corporation ŽFRDC; Project No. 96r398.. Thanks to Mr. Greg Jenkins, Mr. Ken
Frankish, Mr. Craig Poller and all staff at the Aquaculture Development Unit for their
assistance.
References
Buttino, I., 1994. The effect of low concentrations of phenol and ammonia of egg production rates, fecal pellet
production and egg viability of the calanoid copepod Acartia clausi. Mar. Biol. 119, 629–634.
Doi, M., Singhagraiwan, T., Singhagraiwan, S., Ohno, A., 1994. An investigation of copepods being applied
as initial food organisms for red snapper larvae. Thai Mar. Fish. Res. Bull. 5, 21–26.
Doi, M., Toledo, J.D., Golez, M.S.N., Santos, M.d.l., Ohno, A., 1997. Preliminary investigation of feeding
performance of larvae of early red-spotted grouper, Epinephelus coioides, reared with mixed zooplankton.
Hydrobiology 358, 259–263.
Fukusho, K., Arakawa, T., Watanabe, T., 1980. Food value of a copepod Tigropus japonicus, cultured with
v-yeast for larvae and juveniles of Mud Dab Limanda yokohamae. Bull. Jpn. Soc. Sci. Fish. 46, 499–503.
Gasparini, S., Castel, J., Irigoien, X., 1999. Impact of suspended particulate matter on egg production of the
estuarine copepod, Eurytemora affinis. J. Mar. Syst. 22, 195–205.
Guillard, R.L., 1972. Culture of phytoplankton for feeding marine invertebrates. In: Smith, W.L., Chanley,
M.H. ŽEds.., Culture of Marine Invertebrate Animals. Plenum, New York, pp. 29–60.
Hoff, F.H., Snell, T.W., 1987. Plankton Culture Manual. 3rd edn. Florida Aqua Farms, Dade City, 155 pp.
Klein Breteler, W.C.M., Schogt, N., Bass, M., Schouten, S., Kraay, G.W., 1999. Trophic upgrading of food
quality by protozoans enhancing copepod growth: role of essential lipids. Mar. Biol. 135, 191–198.
Kleppel, G.S., Burkart, C.A., Houchin, L., 1998. Nutrition and the regulation of egg production in the calanoid
copepod Acartia tonsa. Limnol. Oceanogr. 43, 1000–1007.
Kraul, S., Nelson, A., Brittain, K., Ako, H., Ogasawara, A., 1992. Evaluation of live feeds for larval and
postlarval Mahimahi Coryphaena hippurus. J. World Aquacult. Soc. 23, 299–307.
Lonsdale, D.J., Heinle, D.R., Siegfried, C., 1979. Carnivorous feeding behavior of the adult calanoid copepod
Acartia tonsa Dana. J. Exp. Mar. Biol. Ecol. 36, 235–248.
Nanton, D.A., Castell, J.D., 1998. The effects of dietary fatty acids on the fatty acid composition of the
harpacticoid copepod, Tisbe sp., for use as a live food for marine fish larvae. Aquaculture 163, 251–261.
Ogle, J., 1979. Adaption of a brown water culture technique to the mass culture of the copepod Acartia tonsa.
Gulf Res. Rep. 6, 291–292.
Ohno, A., Okamura, Y., 1988. Propagation of the calanoid copepod, Acartia tsuensis, in outdoor tanks.
Aquaculture 70, 39–51.
Payne, M.F., Rippingale, R.J., 2000a. Evaluation of diets for culture of the calanoid copepod Gladioferens
imparipes. Aquaculture 187, 85–96.
Payne, M.F., Rippingale, R.J., 2000b. Rearing West Australian seahorse, Hippocampus subelongatus,
juveniles on copepod nauplii and enriched Artemia. Aquaculture 188, 353–361.
Rippingale, R.J., 1994. A calanoid copepod Gladioferens imparipes, holding to surfaces. Hydrobiology
292r293, 351–360.
Rippingale, R.J., Hodgkin, E.P., 1974a. Population growth of a copepod Gladioferens imparipes Thomson.
Aust. J. Mar. Freshwater Res. 25, 351–360.
Rippingale, R.J., Hodgkin, E.P., 1974b. Predation effects on the distribution of a copepod. Aust. J. Mar.
Freshwater Res. 25, 81–91.
Rippingale, R.J., MacShane, M.G., 1991. A calanoid copepod for intensive cultivation. Mem. Queensl. Mus.
31, 457.
Schipp, G.R., Bosmans, J.M.P., Marshall, A.J., 1999. A method for hatchery culture of tropical calanoid
copepods, Acartia spp. Aquaculture 174, 81–88.
Shields, R.J., Bell, J.G., Luizi, F.S., Gara, B., Bromage, N.R., Sargent, J.R., 1999. Natural copepods are
superior to enriched Artemia nauplii as feed for halibut larvae Ž Hippoglossus hippoglossus. in terms of
survival, pigmentation and retinal morphology: relation to dietary essential fatty acids. J. Nutr. 129,
1186–1194.
Støttrup, J.G., Norsker, N.H., 1997. Production and use of copepods in marine fish larviculture. Aquaculture
155, 231–248.
Støttrup, J.G., Richardson, K., Kirkegaard, E., Pihl, N.J., 1986. The cultivation of Acartia tonsa Dana for use
as a live food source for marine fish larvae. Aquaculture 52, 87–96.
Sun, B., Fleeger, J.W., 1995. Sustained mass culture of Amphiascoides atopus a marine harpacticoid copepod
in a recirculating system. Aquaculture 136, 313–321.
Takano, H., 1971. Notes on the raising of an estuarine copepod Gladioferens imparipes Thomson. Bull. Tokai
Reg. Fish. Res. Lab. 64, 81–87.
Toledo, J.D., Golez, S.N., Doi, M., Ohno, A., 1997. Food selection of early grouper, Epinephelus coioides,
larvae reared by the semi-intensive method. Suisanzoshoku 45, 327–337.
Toledo, J.D., Golez, M.S., Doi, M., Ohno, A., 1999. Use of copepod nauplii during early feeding stage of
grouper Epinephelus coioides. Fish. Sci. 65, 390–397.
Witt, U., Quantz, G., Kuhlmann, D., Kattner, G., 1984. Survival and growth of turbot larvae Scophthalmus
maximus L. reared on different food organisms with special regard to long-chain polyunsaturated fatty
acids. Aquacult. Eng. 3, 177–190.
Aquaculture 400–401 (2013) 65–72
Contents lists available at SciVerse ScienceDirect
Aquaculture
journal homepage: www.elsevier.com/locate/aqua-online
Effects of different monoalgal diets on egg production, hatching success

and apoptosis induction in a Mediterranean population of the calanoid
copepod Acartia tonsa (Dana)
Jianshe Zhang a, b, Changwen Wu a, David Pellegrini b, Giovanna Romano c, Francesco Esposito c,
Adrianna Ianora c, Isabella Buttino b, c, d,⁎
a
College of Marine Science and Technology, Zhejiang Ocean University, Zhoushan, People's Republic of China
b
ISPRA Istituto Superiore per la Ricerca e Protezione Ambientale STS-Livorno,Piazzale dei Marmi Terminal Crociere, 57128 Livorno, Italy
c
Stazione Zoologica Anton Dohrn, Napoli, Villa Comunale 80121 Napoli, Italy
d
CRIAcq-Interdepartmental Research Center for Hydrobiological Resources Management and for Aquaculture, University of Naples “Federico II”, Italy
a r t i c l e i n f o a b s t r a c t
Article history: The influence of six monoalgal diets was tested on the reproductive success of the copepod Acartia tonsa over a
Received 5 September 2012 15-day period in order to define the most favorable diet for the optimization of this copepod species to be used in
Received in revised form 20 December 2012 aquaculture and in ecotoxicology applications. The cryptophytes Rhinomonas reticulata and Rhodomonas baltica
Accepted 26 February 2013
induced highest egg production rates (mean = 24.4 eggs female−1 day−1and 21.9 eggs female−1 day−1) and
Available online 7 March 2013
hatching success (mean = 76% and 86.1%) over the 15-day period, respectively. Lowest egg production rates
Keywords:
were recorded with both diatoms Phaeodactylum tricornutum (mean = 8.8 eggs female−1 day−1) and
Copepods Skeletonema marinoi (mean = 8.4 eggs female−1 day−1). These two diatoms also had detrimental effects on
Phytoplankton hatching success (mean = 44.1% and 46.5%, respectively) and adult survival. No adults survived for longer
Egg production than 13 days with a diet of P. tricornutum. Moreover, nauplii produced by females that had fed on diatoms for
Hatching success >10 d, were positively stained for the apoptotic fluorescent marker TUNEL, indicating imminent death. The
Fecal pellet production prasinophyta Tetraselmis suecica induced low fecundity (mean = 13.4 eggs female−1 day−1) and hatching suc-
cess (mean = 62%), and after 15 days only 37.5% of the adults survived. Isochrysis galbana induced low egg pro-
duction per female (b10 eggs female−1 day−1) after 15 days.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction Rearing techniques have focused on how to improve copepod fit-

ness and increase their productivity, especially in indoor systems
In recent years there has been a growing interest in the massive cul- (Buttino et al., 2012; Støttrup, 2000) since copepod production is
tivation of marine copepods to be used for different purposes, for exam- mainly dependent on the quality and quantity of food supplied
ple, in aquaculture, as live feed for fish larvae (Buttino et al., 2012; (Buttino et al., 2011; Ceballos and Ianora, 2003; Kleppel et al., 1998;
Drillet et al., 2006a; Olivotto et al., 2008, 2009), or in ecotoxicology stud- Zheng et al., 2011). Recently Drillet et al. (2011a) showed that a mi-
ies, as animal models to predict the impact of chemicals on marine zoo- crobial preparation in addition to an algal food, can enhance the per-
plankton physiology (Buttino et al., 2011; Carotenuto et al., 2012; Fang formance of Acartia tonsa in terms of egg production and hatching
et al., 2006; Gorbi et al., 2012). success.
Cultivation of copepods at high densities is difficult, especially for Copepods are at the base of the marine food web, linking primary
temperate species with longer developmental times compared to producers with higher trophic levels and, hence, the nutritional qual-
tropical and subtropical copepods (Payne and Rippingale, 2001; ity of phytoplankton will reflect the quality of fish larvae growing on a
Støttrup, 2000). Therefore, massive copepod production remains a food-based copepod diet (McKinnon et al., 2003; Olivotto et al.,
bottleneck for their extensive use in aquaculture or for other laborato- 2008). A. tonsa is one of the most investigated calanoid species, occur-
ry applications. ring in a wide range of geographic areas from temperate to subtropi-
cal waters (Paffenhöfer and Stearns, 1988). In the Mediterranean Sea,
this eurythermic and euryhaline species was introduced in the 1980s
(Farabegoli et al., 1989; Gaudy and Viñas, 1985), becoming the dom-
⁎ Corresponding author at: ISPRA Istituto Superiore per la Ricerca e Protezione
Ambientale STS-Livorno, Piazzale dei Marmi Terminal Crociere, 57128 Livorno, Italy.
inant species in Northern Adriatic lagoons (Comaschi et al., 2000; Sei
Tel.: + 39 0650074035. et al., 1996) and recently found in confined waters in Southern Italy
E-mail address: Isabella.buttino@isprambiente.it (I. Buttino). (Belmonte and Potenza, 2001).
0044-8486/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.aquaculture.2013.02.032
66 J. Zhang et al. / Aquaculture 400–401 (2013) 65–72
A. tonsa species may have a future in aquaculture because it is a cultured using F/2 medium (Guillard, 1975), with silicate for the diatom
cosmopolitan species and “easy to grow” compared to other calanoid cultures and without silicate for the other algae. The algae were
copepods, with a full cycle in the water column (see Drillet et al., cultured on a 14 h light:10 h dark cycle.
2011b for reviews; Støttrup, 2006) and with eggs that can be stored Algal cultures were supplied to copepods during their exponential
longer in temperature-controlled conditions (Drillet et al., 2006b, growth phases at the following concentrations corresponding to
2008). Moreover, this species has widely been proposed as an animal 500 μg C L −1: ISO 3.8 × 10 4 cells mL −1 (~ 65 μm 3 volume); RHINO
model in ecotoxicology standardized assays (Gorbi et al., 2012; ISO, 0.90 × 10 4 cells mL −1 (~ 321 μm 3); RHO 0.4 × 10 4 cells mL −1
1999; Widdows, 1993). (~ 942 μm 3); TETRA 1.1 × 10 4 cells mL −1 (~ 298 μm 3); SKE 2.5 ×
Many authors have tested a variety of phytoplankton cultures, in 10 4 cells mL −1 (~ 196 μm 3 volume); PHAE: 7.35 × 10 4 cells m −L −1
terms of cell density or strain quality, on A. tonsa production in different (~ 11 μm 3 volume). Carbon contents were converted from cell vol-
experimental rearing conditions: from massive cultivation in large umes according to the formula reported by Strathmann (1967).
volumes, to small aquariums in laboratory-based cultures (Amin et al.,
2011; Colin and Dam, 2002; Leandro et al., 2006; Marcus and Wilcox,
2007). Recently, Drillet et al. (2008, 2011c) found that populations of 2.1.1. Copepod rearing experiments
A. tonsa coming from different geographic areas, showed distinct mito- A. tonsa used for this study were reared, for more than 70 genera-
chondrial clades and life history traits, such as generation times and tions, at the ISPRA laboratory in Livorno (Italy). Adults, originally
productivity, which differed for each clade even if the same food condi- obtained from cultures coming from the University of Parma (courtesy
tions were supplied. Therefore, the relationship between feeding and of Gorbi G. and Sei S.), were reared in a 50 L aquarium containing 20 L of
productivity of A. tonsa must be re-considered in the light of these find- 0.22 μm filtered seawater (Millipore 90 mm holder YY3009000) at
ings and the origin of different populations must be taken into account 30‰ salinity and fed with a mixed algal diet of ISO, RHINO and RHO
before any generalizations can be made on the productivity of this cope- at a final concentration > 300 μg C L−1 per day. The aquarium was
pod species. maintained at 20 °C and a 14 h L: 10 h D photoperiod.
In our laboratory, the copepod A. tonsa collected in the lagoon area
in the Northern Adriatic Sea (Comacchio Valley) has been reared as
an animal model for ecotoxicological studies and for future applica- 2.2. Egg production, fecal pellets and hatching success
tions in aquaculture as live feed for fish larvae, for 4 years. In the pres-
ent study we tested which of 6 monoalgal diets was the best food, in After the in vitro population reached the adult stage, healthy mature
terms of egg production and egg viability, for the Mediterranean females and males, with intact appendages and that were actively
strain of A. tonsa for 15 days, and compared our results with those swimming, were selected from the tank and 8 couples were incubated
obtained by other authors for different populations of this copepod pair wise in 100 mL crystallizing dishes containing 50 mL 0.22 μm for
species. The aim of the present study was to define the most favorable each of the 6 test algal diets. Containers were maintained on a very
monoalgal diet for the optimization of A. tonsa to be used both in eco- slow agitator to favor algal suspension. After 24 h, copepods were trans-
toxicology applications, as an animal model to test the toxicity of dif- ferred to new crystallizing dishes containing fresh media and eggs and
ferent pollutants, and to be proposed as first-feed for fish larvae in fecal pellets were counted under an inverted microscope (x10, 20 mag-
aquaculture farming, as a supplement to classical Artemia and rotifer nification). Egg hatching success was determined 48 h after egg laying,
diets. by adding 0.2 mL Lugol solution and counting the number of nauplii
Of the algae tested, it is well documented that two diatoms that had settled on the bottom of the containers. Nauplii were consid-
(Skeletonema marinoi and Phaeodactylum tricornutum) produce toxic ered abnormal when the body or appendages were asymmetrical such
oxylipins which induce apoptosis in other copepod species (reviewed as previously described for Calanus helgolandicus (Poulet et al., 1995).
by Ianora and Miralto, 2010). Because there is no information about Egg production experiments were run for 15 days to test if the diet sig-
the effect of these diatom diets on A. tonsa reproduction, the aim of nificantly modified copepod reproductive responses. Experiments were
this study was also to test these algae on this copepod species. The run in duplicates for each algal diet (N = 96).
other phytoplankton species tested (Isochrysis galbana, Rhinomonas
reticulata, Rhodomonas baltica, Tetraselmis suecica) are not known to
produce toxic metabolites and are generally used as mixed diets for 2.3. Fluorescence labeling and image acquisition
copepod feed (Buttino et al., 2009, 2012; Carotenuto et al., 2012;
Mauchline, 1998). Apoptosis in nauplii was verified using TdT-mediated dUTP nick end
labeling (TUNEL) (Roche Diagnostics). A. tonsa nauplii, obtained from
2. Materials and methods females fed for 10–15 days on each of the monoalgal diets, were fixed
overnight in 4% paraformaldehyde and 0.2 M NaCl in PBS, pH 7.4, rinsed
2.1. Algal cultures in PBS, and frozen in liquid nitrogen to fracture the carapace. Samples,
incubated for 20 h in 0.3 unit mL−1 chitinase enzyme (Sigma-Aldrich)
Two diatom species P. tricornutum (FE1 corresponding to RCC 69) at 25 °C to permeabilize the chitin, were then treated according to the
(PHAE) and S. marinoi (Sarno et al., 2005) (CCMP 2092, SKE), two TUNEL manufacturer's instructions.
cryptophytes R. baltica (Zimmermann) (FE 202, RHO) and R. reticulata To verify if SKE and PHAE induced apoptosis in newly hatched
(FE 208 corresponding to CCAP 995/2, RHINO), the prymnesiophyte nauplii, we incubated nauplii with TUNEL and with Hoechst to visualize
I. galbana (FE 207 corresponding to CCMP 1323, ISO) and the prasinophyte all nuclei. A. tonsa nauplii were incubated with 10 μg mL−1 of the vital
T. suecica (Butcher) (FE 205 corresponding to CCMP 906, TETRA) were fluorescent dye Hoechst 33342 (Sigma-Aldrich) for 20 minutes at room
used as food for adult A. tonsa (Dana) copepods. These algae were cultured temperature to stain nucleic acids. Nauplii were then observed with the
in a temperature-controlled room, using 500 mL flasks filled with 0.22 μm confocal laser-scanning microscope Zeiss LSM-510, using a 488 nm
filtered seawater (FSW) with a salinity of 35‰ for the two diatoms PHAE wavelength laser, to visualize TUNEL-positive areas (green), with a
and SKE and 30‰ for the others. Filtered seawater was previously treated 543 nm wavelength laser, to visualize chlorophyll autofluorescence
for 24 h with HClO (0.04% v:v), and then with sodium thiosulphate (red), and with 405 nm wavelength laser to observe nuclei stained
12.5% (v:v) for another 24 h. Sea water was aerated for 24 h (Lavens with Hoechst (violet) (Buttino et al., 2004, 2011).
and Sorgeloos, 1996) to remove chloride residues. All algae, coming A. tonsa adults fed TETRA for 15 days were observed with a Zeiss
from cultures at the Stazione Zoologica Anton Dohrn in Naples, were Primovert Monitor inverted.
J. Zhang et al. / Aquaculture 400–401 (2013) 65–72 67
2.4. Data analysis 2. Each diet induced variable daily egg production rates during the 15-d
incubation period with lowest mean values of 8.4 eggs f−1 obtained
One-way ANOVA analysis of variance and Bonferroni's multiple with the diatom SKE, to a maximum of 24.4 eggs f −1 recorded with
comparison tests were used to analyze significant differences among the cryptophte RHINO.
treatments. All statistical analyses were conducted using GraphPad With a diet of ISO, egg production increased from 11.7 to a maxi-
Prisma Program. Daily data in the graphs are presented as means ± mum of 19.4 eggs f −1 after 4 days and remained at these levels until
standard errors (s.e.), means calculated over the whole experimental day 9. After 10 days egg production declined to reach a minimum of
period are reported in each graph ± standard deviation (s.d.). 5.0 eggs f −1 on day 13 (Fig. 1A). Mean egg production per female dur-
ing the whole experimental period was 12.7 ± 5 (s.d.). Fecal pellet pro-
duction followed a similar trend with a mean value of 95.2 ± 24.4 (s.d.)
3. Results fp couple −1 day−1 recorded during the whole experimental period.
Egg hatching success was high during the first week of feeding but
The number of eggs, fecal pellets (fp) and % egg hatching success for dropped to less than 50% after 12 days, with a mean value of 78.2%
A. tonsa copepods fed six monoalgal cultures are reported in Figs. 1 and (Fig. 1B).
ISO
eggs m= 12.78 5.0
m= 78.2 13.4
A fecal pellets m= 95.2 24.4 B
40 200 100
fecal pellets couple-1

80
% egg hatched
30 150
eggs female-1
60
20 100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
RHINO
eggs m= 24.4 7.0
m= 76.0 9.4
fecal pellets m= 138.0 15.8
C D
40 200 100
80
% egg hatched
30
eggs female-1
150
60
20 100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
RHO
eggs m= 21.9 5.9
m= 86.1 4.3
fecal pellets m= 108.6 15.8
E F
40 200 100
80
% egg hatched
30
eggs female-1
150
60
20 100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
Fig. 1. Acartia tonsa daily egg production per female, fecal pellet production per female and male couples and % of egg hatching success with monoalgal diets: Isochrysis galbana (A, B),
Rhinomonas reticulata (C, D) and Rhodomonas baltica (E, F).
TETRA
eggs m= 13.4 5.4
A fecal pellets m= 142.3 31 B m= 62.2 21.7
40
200 100

30
eggs female-1
80
% egg hatched
150
60
20
100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
PHAEO
eggs m= 8.8 4.6 m= 44.1 25.6
C fecal pellets m= 121.4 35.7 D
40 200
100
30 fecal pellets couple-1

eggs female-1
150 80
% egg hatched
60
20 100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
SKE
eggs m= 8.4 3.2 m= 46.5 16
E fecal pellets m= 136.8 31 F
40 200
100
30
eggs female-1
150 80
% egg hatched
60
20 100
40
10 50
20
0 0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
days days
Fig. 2. Acartia tonsa daily egg production per female, fecal pellet production per female and male couples and % of egg hatching success with monoalgal diets: Tetraselmis suecica (A, B),
Phaeodactylum tricornutum (C, D) and Skeletonema marinoi (E, F).
A RHINO diet induced a rapid increase in egg production after one When couples were fed with TETRA fecundity remained low
week, with a maximum of > 30 eggs f −1, followed by a slight decline (b 20 eggs f −1 day−1) and was strongly reduced after the first week.
after 11 days with a production of almost 20 eggs f −1 day−1 A minimum egg production rate (less than 5 eggs f −1) was recorded
(Fig. 1C). Mean egg production 24.21 ± 8.57 eggs f −1 day−1 (s.d.), at the end of the experiment and, on average, egg production was
calculated over the whole period, was the highest recorded for all of only 13.4 eggs per female during the whole experimental period
the tested diets. (Fig. 2A). In contrast, fp production was the highest recorded of the
Fecal pellet production was very high with a mean of 138 fp six diets, with a mean of 142 fp couple −1 day−1. Egg hatching success
couple −1 day −1. The % egg hatching success slowly declined during declined very quickly, from almost 100% on the first day to less than
the experimental period from greater than 85% to 60% after 15 days. 60% after 10 days, and to only 20% after 14 days (Fig. 2B). On average
On average, egg hatching success calculated over the whole period egg hatching success calculated for the whole experimental period,
was 76%. was 62.2%.
RHO diet induced a similar egg production rate as for RHINO with With the diatom diets PHAE and SKE, egg production was the lowest
almost 21.9 ± 5.9 eggs f −1 day −1 (s.d.) and the highest egg hatch- recorded (almost 8 eggs f −1d −1). In particular with PHAE egg produc-
ing success (86.1%) recorded for the entire period (Fig. 1E, F). tion was strongly reduced after 5 days from almost 15 eggs f−1d−1 to
less than 4 eggs after 8 days (Fig. 2C). Fecal pellet production declined
sharply during the experimental period, from an initial value of almost
200 fp couple−1 recorded the first day to about 50 fp on day 11,
suggesting that couples did not feed as at the beginning of the experi-
ment. Also egg hatching success declined very rapidly, and on day 9
only 20% of the eggs hatched (Fig. 2D). No apparent abnormal nauplii
were observed during the experiment, but those that survived were
sluggish and showed limited swimming capabilities. With this diet
only one couple survived after 12 days and the experiment was there-
fore terminated on day 12.
With the other diatom SKE, egg production remained low through-
out the experimental period, with a minimum of 3.8 eggs f−1 recorded
after 14 days (Fig. 2E). By contrast, fp production was very high at the
beginning, with more than 150 fp couple−1 on the first 3 days. It then
became very unstable: dropping to about 100 fp couple −1 for the fol-
lowing 4 days but increasing again on day 9 to 200 fp couple−1. Produc-
tion declined again during the second week. The % egg hatching success
declined in the first week, from 80% to less than 40% on day 7. During
the second week egg hatching success was more than 50% but then de-
clined again to less than 40% after three days (Fig. 2F).
Fig. 3 shows the % of female survival with the different diets. PHAE
and TETRA diets induced high adult mortality in females after 10 and
13 days, respectively.
In particular, after 13 days no adults that had fed PHAEO had sur-
vived while only 37.5% of adults fed TETRA survived after 15 days. By
day 13, A. tonsa adults fed TETRA showed filamentous structures near
the head and the mouth apparatus (Fig. 4A,B, arrows); their move-
ments were slow and sluggish which probably explained the increased
mortality of adults after 12 days of feeding on this diet (Fig. 3).
Nauplii, produced by females fed ISO were used as controls, due to
the high % egg viability recorded during the entire experimental period
(Fig. 1B). Fig. 5 shows fluorescent nauplii stained with both dyes and
observed with fluorescence microscopy. Fig. 5A is a normal nauplius
produced by a female fed ISO for 10 days; the only visible fluorescence
is the violet, due to Hoechst, and red autofluorescence due to chloro-
phyll inside the gut. A similar fluorescent staining was observed for
nauplii produced by A. tonsa females fed TETRA for 10 days (data not
reported), indicating that such nauplii were not undergoing any apo-
ptotic processes. Fig. 5B and C show nauplii produced by females fed
SKE for 10 and 12 days, respectively. Green fluorescence signals are
well visible in the legs and along the body; red fluorescence due to chlo-
rophyll is not visible. After 12 days many hatched nauplii appeared
strongly deformed with a relevant portion of tissues that were positive-
ly stained with TUNEL (Fig. 5C). These nauplii were characterized by Fig. 4. Acartia tonsa females fed Tetraselmis suecica for 13 days and observed with a
slow movements and did not survive long after observation. Zeiss Primovert Monitor microscope. Arrows indicate the filamentous structures at-
tached to the mouth appendages. (A) Dorsal view; (B) lateral view. (400×).
Nauplii produced by females fed PHAE appeared normal but showed

RHINO strong positivity to TUNEL in the legs and inner body (Fig. 5D) indicat-
RHO ing that apoptotic processes were under way.
100
ISO
4. Discussion
TETRA
% adult survival
75 SKE The copepod species A. tonsa has been proposed in aquaculture

PHAEO and as animal model in bioassay protocols to text marine pollutants
50 (Drillet et al., 2006a, 2011b; Gorbi et al., 2012). Therefore, it is impor-
tant to find better conditions to rear this copepod with reduced
efforts in terms of costs and time. Moreover, it has recently been
25
reported that different copepod strains, living in different geographic
areas, could have different physiology and life history traits (Drillet et
0 al., 2008, 2011c; Lauritano et al., 2012).
0 2 4 6 8 10 12 14
Our results on the productivity of A. tonsa copepods, collected in a
days
Mediterranean area and reared in laboratory for 4 years, indicated
Fig. 3. Acartia tonsa daily % adult survival with monoalgal diets: Isochrysis galbana,
that diatoms are deleterious for their fecundity and naupliar viability,
Rhinomonas reticulata, Rhodomonas baltica, Tetraselmis suecica, Phaeodactylum tricornutum while the best monoalgal diets are the cryptophytes RHINO and RHO.
and Skeletonema marinoi. Statistical comparison among A. tonsa fed six different monoalgal diets
Fig. 5. Acartia tonsa nauplii stained with Hoechst (violet) and Tunel (green), produced by adult females fed different monoalgal diets, and observed with the confocal laser scanning
microscope Zeiss LSM 510. Each image is a three-dimensional reconstruction of different z-stacks acquired along the whole body. (A) Nauplii produced by females Isochrysis galbana
fed 10 days, hoechst marks all body tissues; the red spot inside the gut is chlorophyll. (B, C) Nauplii produced by females fed Skeletonema marinoi 10 and 12 days, respectively.
Green fluorescence of the legs indicates apopotitc cells. B is slightly deformed whereas C is strongly abnormal due to longer maternal feeding on S. marinoi. (D) Nauplii produced
by females fed Phaeodactylum tricornutum for 9 days; green fluorescence in the legs and buccal region indicates apoptotic cells.
suggests three different groups with no significant differences in terms of 150 μg C L −1, but that this diet did not support their development
of productivity: the first one included two cryptophytes, with the until reproduction. However, Carotenuto et al. (2002) found that the co-
highest egg production and egg hatching success, the second group of pepod Temora stylifera was able to complete development from hatch-
ISO and TETRA with reduced egg production and relatively high egg ing to adulthood when reared with this flagellate. Our results indicate
hatching, and the diatoms PHAE and SKE, as a third group, with very that nauplii produced by females fed ISO developed normally and that
low fecundity and hatching success and with also a reduction in the % no more than 10% were positively stained for apoptotic fluorescent
of adult survival (Table 1). However, significant differences occurred marker TUNEL, however we did not follow development until adult-
among these three groups. With the two cryptophytes RHO and hood. The reduction in egg hatching success after 15 days of feeding
RHINO, egg production was the highest recorded, increasing 5-fold on ISO may be due to the poor quality of this alga which is known to
after the first 10 days with RHINO and 4-fold at the end of the experi- contain low levels of eicosapentaenoic acid (EPA) (Muller-Fuega et al.,
ment. RHO also induced the highest egg hatching success during the en- 2003; Tremblay et al., 2007).
tire experimental period, followed by ISO and RHINO. The excellence of Interestingly, also TETRA, commonly used to rear A. tonsa for eco-
a RHO diet has also been reported by other authors for different cope- toxicological studies in combination with ISO and RHINO (APHA,
pod species (Buttino et al., 2009; Dahl et al., 2009; Ismar et al., 2008) 1989; Gorbi et al., 2012) did not support high egg production rates
providing high survival rates and fast naupliar development at concen- in the Mediterranean population of A. tonsa, for a period longer than
trations ranging from 150 to 300 μg C L −1. Egg production per female 2 weeks. Colin and Dam (2002) considered Tetraselmis spp. as a con-
per day recorded for our Mediterranean A. tonsa population, did not trol monoculture in their 3-day feeding experiments; the authors
reach the production rates recorded by Holste and Peck (2006) for the reported that this alga stimulated ingestion and egg production
Baltic population (almost 50 eggs per female per day). However, this after three days. Our results agree in part with their findings: we
production was recorded supplying 5 times the concentration used in observed significantly higher fp production rates with this diet com-
our experiments suggesting that these two diets alone can support ex- pared to PHAE, RHO and ISO. If we consider only three days of feed-
tremely high production rates also for Mediterranean population of ing, TETRA seems to enhance egg production with values doubling
A. tonsa, when given at saturating food concentrations. By contrast, after two days of feeding. However, if we consider longer incubation
ISO alone is not a good diet as also reported by other authors for other periods, mean egg production rates decline with time with a strong
copepod species (Buttino et al., 2011; Carotenuto et al., 2002; Ismar et reduction in A. tonsa productivity. In consideration of our results,
al., 2008; Knuckey et al., 2005). Our results indicate that the % of egg this algae should be avoided with high advantages in terms of time
hatching success declined during the second week suggesting that and resources employed. In fact, reduction of the mixed diet will in-
this diet cannot support the cultivation of A. tonsa for periods longer fluence economic costs.
than one week. We also recorded a reduction in the % of adult survival Moreover, in our study adult survivorship was strongly reduced
after 12 days; to our knowledge, a similar adult mortality has never after 13 days. Even if TETRA did not significantly reduce adult survival
been reported before for A. tonsa fed ISO. Ismar et al. (2008) reported when the whole experimental period was considered, % mortality in-
that naupliar survivorship was high with a diet of ISO at concentrations creased from 25 to more than 60% during the last 12–15 days (Fig. 3)
confirming that this diet cannot sustain the rearing of A. tonsa for pe-
riod longer than 10 days. Females and males appeared covered by fil-
Table 1 amentous structures in the mouth regions that slowed down their
Comparison between mean fecundity, fecal pellet and hatching success for A. tonsa fed swimming movements. Probably, their appendages and filtration ap-
six different diets.
Different capital letters indicate differences statistically significant (df = 5, p > 0.001).
paratus were clogged by such structures and animals were unable to
feed. This is the first time that a similar effect has been described in
Fecundity Fecal pellets % hatching % adult survival copepods after ingestion of TETRA. Induction of apoptosis was not ev-
(eggs per (per couple success F = 8.305
ident in nauplii produced by females fed with this culture (b 10%).
female per day) per day) F = 16. 05
F = 22.57 F = 7.597 Hence we suggest that reduced egg production and high adult mor-
tality after 12 days could be due to a mechanical impediment in feed-
RHINO 24.4 ± 7.0A 138.0 ± 15.8AB 76.0 ± 9.4A 98.33 ± 4.39 A
RHO 21.9 ± 5.9A 108.6 ± 15.80C 86.11 ± 4.36A 95.00 ± 6.33 A ing due to filamentous structures which we were unable to identify.
PHAEO 8.8 ± 4.6C 121.4 ± 35.7B 44.1 ± 25.6C 60.83 ± 39.77 B On the other hand, we cannot explain why there was also a reduction
SKE 8.40 ± 3.2C 136.8 ± 31.0AB 46.5 ± 16.0C 98.33 ± 4.39 A in egg hatching success. It is also known that less food availability can
ISO 12.78 ± 5.0B 95.2 ± 24.40C 78.2 ± 13.4A 92.5 ± 12.32 A trigger delayed hatching eggs (resting eggs) (Drillet et al., 2011c) and
TETRA 13.4 ± 5.4B 142.3 ± 31.0A 62.2 ± 21.7B 88.30 ± 19.75 A
we cannot exclude the possibility that unhatched embryos could be
Table 2
Comparison of proteins and fatty acids content in the six algal species (SFA = saturated fatty acids; MUFA = monounsaturated fatty acids; PUFA = polyunsaturated fatty acids).
Nutritional content I. galbana Cryptophiceae T. suecica S. costatum P. tricornutum
Protein 23–29 20–31 25 30

(% dry weght)
SFA 30.3 14.4–22.6 21.7–24.7 32.0–33.0 35.2
(% total fatty acid)
MUFA 11.3 6–11 12.9–16.5 27.7–29.4 34.9
PUFA 33–56.4 75.2–65.8 27.2–52.5 36.8–34–8 17–27
EPA 20:5(n − 3) 0.9 2–4.4 4.3–4.8 28.4
(mg/g dry weight)
DHA 22:6(n − 3) 15.8 12–18 0.1–0.2 0.2–0.74
(mg/g dry weight)
References Albentosa et al., 1996; FAO, 1996; Li et al., 2006; Patil et al., 2007; Renaud et al., 1999
could be dormant even if dormant eggs have never been reported development of N. spinipes was inhibited further confirming that poor
for the Mediterranean A. tonsa copepod. It is well known that survival and development was due to the presence of toxins rather
subitaneous eggs of A tonsa can enter into quiescence when chal- than to poor food quality. On the contrary, SKE did not affect adult sur-
lenged with temperature decline, or when salinity and oxygen change vival but strongly reduced egg production and egg hatching success
(Drillet et al., 2006b; Hansen et al., 2010; Holmstrup et al., 2006), or after three days of feeding and until the end of the experiment. This is
due to food limitation (Drillet et al., 2011c). However, non-viable in agreement with our previous findings obtained for other copepod
eggs are clearly distinguished from resting eggs due to their shape species (e.g. Ianora et al., 2004). Fecal pellet production remained high
and opacity, and generally degenerate a few hours after spawning with mean values similar to TETRA and RHINO, suggesting that this
(also our observations). Knuckey et al. (2005) reported that Acartia alga was actively fed upon. It is interesting to note that nauplii produced
sinjiensis copepodites feeding on Tetraselmis spp. did not develop by females fed SKE and PHAE were strongly positively marked for apo-
beyond copepodite stage C1 and appeared strongly deformed show- ptosis. However, only those produced by females fed SKE were strongly
ing the absence of an eyespot and/or reduction in size. In our experi- deformed (Fig. 5C). Such deformities increased the longer the females
ments completely developed nauplii were visible within many of the ingested this food, suggesting that possible toxins present in both dia-
unhatched eggs indicating that development had proceeded normally toms could have different targets on copepod development. Apoptosis
until an advanced stage but was then blocked due to unknown in nauplii was accounted for more than 60% in nauplii produced by fe-
reasons. We conclude that this alga is a poor food item for A. tonsa re- males fed PHAE and SKE, after day 4 and 8, respectively.
production and development but it remains unclear as to why TETRA In conclusion, this study reveals the complexity in cultivating plank-
is such a poor diet. In a study by Dahl et al. (2009), who compared tonic copepods such as Acartia tonsa in controlled systems using
fatty acid content in different algae used to feed the harpacticoid monoalgal diets, and confirms the toxicity of two diatoms on A. tonsa re-
copepod Nitocra spinipes, TETRA was reported as having very low % production. Choosing the right feed item will benefit the ambition of
of HUFA (1.7) with respect to ISO and Rhodomonas salina, this latter mass cultivation of calanoids for aquaculture and bioassay purposes,
species having a 20 times higher highly unsaturated fatty acids also for temperate species.
(HUFA) concentration compared to TETRA. The nutritional contents
of the six algae used in our experiments are well known in literature
Acknowledgments
and the classical nutritional parameters considered essential for the
correct development of copepods are reported in Table 2. Diatoms
This work was partially supported by the Danish National Strate-
have very high proportions of EPA and DHA and very low linolenic
gic Research Council — IMPAQ — grant no. 10-093522 to IB, by
and linoleic acid content, while cryptophytes have very high propor-
Zhejiang Provincial Natural Science Foudation of PR-CHINA grant no.
tions of EPA and DHA as well as linolenic acid. Patil et al. (2007)
LQ12C19003 and by ISPRA-STS Ecotoxicology and Plankton Biology
reported that the highest amount of EPA was found in P. tricornutum
Lab. -Livorno-Italy. The authors thank the Inter-University Consortium
(28.4 mg g −1), followed by T. suecica (4.8 mg g −1) and R. baltica
for Marine Biology—CIBM in Livorno for logistic accommodation and
(4.4 mg g−1), while DHA was abundant only in I. galbana (15.8 mg g−1).
Prof. Barone Maria Carmela, director of the Inter-departmental Center
Finally, diatoms, as expected, showed strong effects on the produc-
for Hydrobiological Resource Management and Aquaculture — CRIAcq
tivity of A. tonsa (Table 1), as also described for other marine organisms
of University of Naples “Federico II” for the support in hosting Prof.
(reviewed by Caldwell, 2009; Ianora and Miralto, 2010). Two different
Zhang in Italy.
effects were observed with PHAE and SKE; the first diatom induced
a slight increase in egg production rates after the first week with a
simultaneous reduction in fp production and egg hatching success. A References
dramatic, significant decline in egg production, fp production, hatching A.P.H.A., A.W.W.A., W.P.C.F., 1989. Standard Methods for the Examination of Water and
success and adult survivorship were observed the second week. In par- Wastewater, 17th ed. American Public Health Association, Washington, DC, USA
ticular, after 13 days no adults that had fed PHAE had survived. Even 8.1–8.23.
Adolph, S., Poulet, S.A., Pohnert, G., 2003. Synthesis and biological activity of α, β, γ, δ-
though PHAE lacks the volatile aldehydes decadienal and decatrienal,
unsaturated aldehydes from diatoms. Tetrahedron 59, 3003–3008.
Pohnert (2000), Pohnert et al. (2002) reported the presence of the Albentosa, M., Labarta, U., Fernandéz-Reiriz, M.J., Pérez_Camacho, A., 1996. Fatty acid
aldheydic acids 9-oxo-(5Z,7E)-nonadienoic acid (9-ONDE) and 12- composition of Ruditapes decussates spat fed on different microalgae diets. Compara-
tive Biochemistry and Physiology 113A (2), 113–119.
oxo-(5Z,8Z,10E)-dodecatrienoic acid (12-ODTE) which affected sea
Amin, R.Md., Koski, M., Båmstedt, U., Vidoudez, C., 2011. Strain-related physiological
urchin development (Adolph et al., 2003). Dahl et al., 2009 reported and behavioural effects of Skeletonema marinoi on three common planktonic cope-
that the development of the harpacticoid copepod N. spinipes was af- pods. Marine Biology 158, 1965–1980.
fected when animals were fed P. tricornutum, with only 6.9% juvenile Belmonte, G., Potenza, D., 2001. Biogeography of the family Acartiidae (Calanoida) in
the Ponto-Mediterranean Province. Hydrobiologia 453 (454), 171–176.
survival after 7 days, and a total arrest in development from NI to CI Buttino, I., do Espirito Santo, M., Ianora, A., Miralto, A., 2004. Rapid assessment of cope-
stage. Even in combination with another algae, Dunaniella tertiolecta, pod embryo viability using fluorescent probes. Marine Biology 145, 393–399.
Buttino, I., Ianora, A., Buono, S., Vitiello, V., Sansone, G., Miralto, A., 2009. Are monoalgal ISO-International Organization for Standardization, 1999. Water Quality-Determination
diets inferior to plurialgal diets to maximize cultivation of the calanoid copepod of Acute Lethal Toxicity to Marine Copepods (Copepoda, Crustacea) ISO 14669.
Temora stylifera? Marine Biology 156, 1171–1182. Kleppel, G.S., Burkart, C.A., Houchin, L., 1998. Nutrition and the regulation of egg pro-
Buttino, I., Pellegrini, D., Romano, G., Hwang, J.-S., Liu, T.-M., Sartori, D., Sun, C.-K., duction in the calanoid copepod Acartia tonsa. Limnology and Oceanography 43,
Macchia, S., Ianora, A., 2011. Study of apoptosis induction using fluorescent and 1000–1007.
higher harmonic generation microscopy techniques in Acartia tonsa nauplii ex- Knuckey, R.M., Semmens, G.L., Mayer, R.J., Rimmer, M.A., 2005. Development of an op-
posed to chronic concentrations of nickel. Hydrobiologia 27 (S2), 97–104. timal micro diet for the culture of the calanoid copepod Acartia sinjiensis: effect of
Buttino, I., Ianora, A., Buono, S., Vitiello, V., Malzone, M.G., Rico, C., Langellotti, A., Sansone, algal species and feed concentration on copepod development. Aquaculture 249,
G., Gennari, L., Miralto, A., 2012. Experimental cultivation of the Mediterranean 339–351.
calanoid copepod Temora stylifera and Centropages typicus in a pilot re-circulating Lauritano, C., Carotenuto, Y., Miralto, A., Procaccini, G., Ianora, A., 2012. Copepod
system. Aquaculture Research 43, 247–259. population-specific response to a toxic diatom diet. PLoS One 7 (10), e47262.
Caldwell, G.S., 2009. The influence of bioactive oxylipins from marine diatoms on in- http://dx.doi.org/10.1371/journal.pone.0047262.
vertebrate reproduction and development. Marine Drugs 7, 367–400. Lavens, P., Sorgeloos, P., 1996. Manual on the production and use of live food for aqua-
Carotenuto, Y., Ianora, A., Buttino, I., Romano, G., Miralto, A., 2002. Is postembryonic de- culture. FAO Fisheries Technical Paper 361, Rome, Italy.
velopment in the copepod Temora stylifera negatively affected by diatom diets? Leandro, S.M., Tiselius, P., Queiroga, H., 2006. Growth and development of nauplii and
Journal of Experimental Marine Biology and Ecology 276, 49–66. copepodites of the estuarine copepod Acartia tonsa from southern Europe (Ria
Carotenuto, Y., Esposito, F., Pisano, F., Lauritano, C., Miralto, A., Perna, M., Ianora, A., 2012. deAveiro, Portugal) under saturating food conditions. Marine Biology 150, 121–129.
Multi-generation cultivation of the copepod Calanus helgolandicus in a re-circulating Li, J., Sun, S., Li, C.-l., Shang, Z., Tao, Z.-C., 2006. Effects of single and mixed diatom diets on
system. Journal of Experimental Marine Biology and Ecology 418–419, 46–58. the re production of copepod Calanus sinicus. Acta Hydrochimica et Hydrobiologica
Ceballos, S., Ianora, A., 2003. Different diatoms induce contrasting effects on the repro- 34, 117–125.
ductive success of the copepod Temora stylifera. Journal of Experimental Marine Marcus, N.H., Wilcox, J.A., 2007. A Guide to the Meso-scale Production of the Copepod
Biology and Ecology 294, 189–202. Acartia tonsa. Florida State University-Department of Oceanography Biological
Colin, S.P., Dam, H.G., 2002. Testing for toxic effects of prey on zooplankton using sole Oceanography 26.
versus mixed diets. Limnology and Oceanography 47, 1430–1437. Mauchline, J., 1998. The Biology of Calanoid Copepods. In: Blaxter, J.H.S., Southward,
Comaschi, A., Acri, F., Bianchi, F., Bressan, M., Camatti, E., 2000. Temporal changes of A.J., Tyler, P.A. (Eds.), Advances in Marine Biology. Academic Press, London, UK,
species belonging to Acartia genus (Copepoda, Calanoida) in the Northern Basin p. 710.
of the Venice Lagoon. Bollettino Museo Civico di Storia Naturale di Venezia 50, McKinnon, D., Duggan, S., Nichol, P.D., Rimmer, M.A., Semmens, G., Robin, B., 2003. The
189–193. potential of tropical paracalanoid copepods as live feeds in aquaculture. Aquaculture
Dahl, U., Lind, C.R., Gorokhova, E., Eklund, B., Breitholts, M., 2009. Food quality effects 223, 89–106.
on copepod growth and development, implications for bioassays in ecotoxicological Muller-Fuega, A., Moal, J., Kaas, R., 2003. The Microalgae of Aquaculture. In: Støttrup,
testing. Ecotoxicology and Environmental Safety 72, 351–357. J.G., McEvoy, L.A. (Eds.), Live Feeds in Marine Aquaculture. Blackwell Science, Oxford,
Drillet, G., Jorgensen, N.O.G., Sorensen, T.F., Ramlov, H., Hansen, B.W., 2006a. Biochemical pp. 206–243.
and technical observations supporting the use of copepods as live feed organisms in Olivotto, I., Buttino, I., Borroni, M., Malzone, M.G., Carnevali, O., 2008. The use of the
marine larviculture. Aquaculture Research 37, 756–772. Mediterranean calanoid copepod Centropages typicus in Yellowtail clownfish
Drillet, G., Iversen, M.H., Sørensen, T.F., Ramløv, H., Lund, T., Hansen, B.W., 2006b. Effect (Amphiprion clarkii) larviculture. Aquaculture 284, 211–216.
of cold storage upon eggs of a calanoid copepod Acartia tonsa (Dana) and their off- Olivotto, I., Avella, A.M., Buttino, I., Borroni, M., Cutignano, A., Carnevali, O., 2009. Calanoid
spring. Aquaculture 254, 714–729. copepod administration improve yellow tail clownfish (Amphiprion clarkii)
Drillet, G., Goetze, E., Jepsen, P.M., Højgaard, J.K., Hansen, B.W., 2008. Strain-specific vital larviculture: biochemical and molecular implications. AACL Bioflux 2, 355–367.
rates in four Acartia tonsa cultures, I: strain origin, genetic differentiation and egg sur- Paffenhöfer, G.-A., Stearns, D.E., 1988. Why is Acartia tonsa (Copepoda: Calanoida) re-
vivorship. Aquaculture 280, 109–116. stricted to nearshore environments? Marine Ecology Progress Series 42, 33–38.
Drillet, G., Rabarimanantsoa, T., Frouël, S., Lamson, J.S., Christensen, A.M., Kim-Tiam, S., Patil, V., Källqvist, T., Olsen, E., Vogt, G., Gislerød, H.R., 2007. Fatty acid composition of
Hansen, B.W., 2011a. Do inactivated microbial preparations improve life history 12 microalgae for possible use in aquaculture feed. Aquaculture International 15,
traits of the copepod Acartia tonsa? Marine Biotechnology 13, 831–836. 1–9.
Drillet, G., Frouël, S., Sichlau, M.H., Jepsen, P.M., Højgaard, J.K., Joarder, A.K., Hansen, Payne, M.F., Rippingale, R.J., 2001. Intensive cultivation of the copepod calanoid
B.W., 2011b. Status and recommendations on marine copepod cultivation for use Gladioferens imparipes. Aquaculture 201, 329–342.
as live feed. Aquaculture 315, 155–166. Pohnert, G., 2000. Wound-activated chemical defense in unicellular planktonic algae.
Drillet, G., Hansen, B.W., Kiørboe, T., 2011c. Resting egg production induced by food Angewandte Chemie International Edition in English 39, 4352–4354.
limitation in the calanoid copepod Acartia tonsa. Limnology and Oceanography Pohnert, G., Lumineau, O., Cueff, A., Adolph, S., Cordevant, C., Lange, M., Poluet, S.A.,
56 (6), 2064–2070. 2002. Are volatile unsaturated aldehydes from diatoms the main line of chemical
Fang, T.H., Hwang, J.-S., Hsiao, S.H., Chen, H.Y., 2006. Trace metals in seawater and cope- defence against copepods. Marine Ecology Progress Series 245, 33–45.
pods in the ocean outfall area off the northern Taiwan coast. Marine Environmental Poulet, S.A., Laabir, M., Ianora, A., Miralto, A., 1995. Reproductive response of Calanus
Research 61, 224–243. helgolandicus.1. Abnormal embryonic and naupliar development. Marine Ecology
FAO fisheries technical paper, 1996. Manual on the Production and Use of Live Food for Progress Series 129, 85–95.
Aquaculture, 361 31–35. Renaud, S.M., Thinh, L.-V., Parry, D.L., 1999. The grosso chemical composition and fatty
Farabegoli, A., Ferrari, I., Manzoni, C., Pugnetti, A., 1989. Prima segnalazione nel Mare acid composition of 18 species of tropical Australian microalgae for possible use in
Adriatico del copepode calanoide Acartia tonsa Dana. Nova Thalassia 10, 207–208. mariculture. Aquaculture 170, 147–159.
Gaudy, R., Viñas, M.D., 1985. Première signalisation en Mediterranée du copépode Sarno, D., Kooistra, H.C.F., Medlin, L.K., Percopo, I., Zingone, A., 2005. Diversity of the
pélagique Acartia tonsa. Rapport Commission International pour l'exploration genus Skeletonema (Bacillariophyceae). II. An assessment of the taxonomy of S.
scientifique de la Mer Mediterranee 29, 227–229. costatum-like species with the description of four new species. Journal of Phycology
Gorbi, G., Invidia, M., Savorelli, F., Faraponova, O., Giacco, E., Cigar, M., Buttino, I., Leoni, 41, 151–176.
T., Prato, E., Lacchetti, I., Sei, S., 2012. Standardized methods for acute and semi- Sei, S., Rossetti, G., Villa, F., Ferrari, I., 1996. Zooplankton variability related to environ-
chronic toxicity tests with the copepod Acartia tonsa. Environmental Toxicology mental changes in a eutrophic coastal lagoon in the Po Delta. Hydrobiologia 329,
and Chemistry 31, 2023–2028. 45–55.
Guillard, R.R., 1975. Culture of phytoplankton for feeding marine invertebrate animals. In: Støttrup, J.G., 2000. The elusive copepods: their production and suitability in marine
Smith, W.L., Chanley, M.H. (Eds.), Culture of Marine Invertebrate Animals. Plenum, aquaculture. Aquaculture Research 31, 703–711.
New York, pp. 29–60. Støttrup, J.G., 2006. Review on the status and progress in rearing copepods for marine
Hansen, B.W., Drillet, G., Kozmér, A., Madsen, K.V., Pedersen, M.F., Sørensen, T.F., 2010. larviculture. Advantages and disadvantages among calanoids, harpacticoids and
Temperature effects on copepod egg hatching: does acclimatization matter? Journal cyclopoids copepods. In: Cruz Suárez, L.E., et al. (Ed.), Avances en Nutriciόn
of Plankton Research 32, 305–315. Acuίcola VIII. VIII Simposium Internacional de Nutriciόn Acuίcola. 17–17, Nuevo
Holmstrup, M., Overgaard, J., Sørensen, T.F., Drillet, G., Hansen, B.W., Ramløv, H., Engel- Leόn, México (ISBN 970-694-333-5).
Sørensen, K., 2006. Influence of storage on viability of quiescent copepod eggs (Acartia Strathmann, R.R., 1967. Estimating the organic carbon content of phytoplankton from
tonsa Dana): effects of temperature, salinity and anoxia. Aquaculture Research 37, cell volume or plasma volume. Limnology and Oceanography 12, 411–418.
625–631. Tremblay, R., Cartier, S., Miner, P., Pernet, F., Quere, C., Moal, J., Muzellec, M.L., Mazuret,
Holste, L., Peck, M.A., 2006. The effects of temperature and salinity on egg production M., Samain, J.F., 2007. Effect of Rhodomans salina addition to a standard hatchery
and hatching success of Baltic Acartia tonsa (Copepoda: Calanoida): a laboratory diet during the early ontogeny of the scallop Pecten maximus. Aquaculture 262,
investigation. Marine Biology 148, 1061–1070. 410–418.
Ianora, A., Miralto, A., 2010. Toxigenic effects of diatoms on grazers, phytoplankton and Widdows, J., 1993. Marine and estuarine invertebrate toxicity tests. In: Calow, P. (Ed.),
other microbes: a review. Ecotoxicology 19, 493–511. Handbook of Ecotoxicology, vol. 1. Blackwell Science, Oxford, UK, pp. 145–164.
Ianora, A., Miralto, A., Poulet, S.A., Carotenuto, Y., Buttino, I., Romano, G., Casotti, R., Zheng, Y., Dam, H.G., Avery, D.E., 2011. Differential responses of populations of the co-
Pohnert, G., Wichard, T., Colucci-D'Amato, L., Terrazzano, G., Smetacek, V., 2004. Al- pepod Acartia hudsonica to toxic and nutritionally insufficient food algae. Harmful
dehyde suppression of copepod recruitment in blooms of a ubiquitous planktonic Algae 10, 723–731.
diatom. Nature 429, 403–407.
Ismar, S.M.H., Hansen, T., Sommer, U., 2008. Effect of food concentration and type of
diet on Acartia survival and naupliar development. Marine Biology 154, 335–343.
18. December European Cooperation in
Science and Technology
2013
European Cooperation Brussels, 29 May 2013
in Science and Technology
- COST -
——————————
Secretariat
-------
COST 4113/13
NOTE
To : COST Committee of Senior Officials (CSO)
Subject : COST Action Proposal Submission, Evaluation, Selection and Approval
Delegations will find attached the rules regarding the “COST Action Proposal Submission,
Evaluation, Selection and Approval” approved by the CSO on 15-16 May 2013.
________________
COST 4113/13 1
DG G III EN
COST Action Proposal Submission, Evaluation, Selection and Approval
The COST Committee of Senior Officials (hereafter referred to as “CSO”) having regard to
COST 4112/13 “Rules for Participation in and Implementation of COST Activities”
in accordance with its role as main decision-making body of COST,
Whereas:
(1) COST 4111/11, “Guidelines for Assessment of applications for COST Actions” refers to the
existing rules and process for the evaluation of COST Action Proposals.
(2) COST 4170/10, “Rules of Procedure for a COST Domain Committee” describes the general
framework for the tasks of the COST Domain Committees (hereafter referred to as the
“DC(s)”).
(3) COST 4157/11 “COST Strategy” has provided a mandate for the streamlining and
clarification of the COST procedures and documents so as to achieve the goals set by the CSO
therein.
has adopted the present rules for “COST Action Proposal Submission, Evaluation, Selection and
Approval” on the 15-16 May 2013 This Decision will enter into force on 1 June 2013.
Subject Matter and Scope
COST strives to implement transparent, efficient and simple evaluation and selection process.
This decision lays down the rules for the submission, evaluation, selection and approval of COST
Action proposals. The implementation of these rules is further detailed on the COST Website and in
the guidelines.1
Definitions
For the purpose of this Decision, the following definitions shall apply as follows:
1. COST Member Country: means any country that was a founding member of the COST
Intergovernmental Framework or joined COST as a full member following the approval of the
COST Committee of Senior Officials (CSO).
1
Available on the COST Website (www.cost.eu/opencall)
COST 4113/13 2
DG G III EN
2. COST Cooperating State: means a country that has been approved by the CSO to participate
without voting rights in the CSO.
3. COST National Coordinator (CNC): means the individual appointed by the COST Member
Countries and Cooperating State in charge of accepting the Action Memorandum of
Understanding, of the nomination process for the Domain Committee and Management
Committee members as well as providing information and support to national research
communities.
4. Domain Committee (DC): means the committee composed of scientific and technological
experts responsible for a particular research domain appointed by the COST Member Country
and Cooperating State.
5. Open Call for proposals: means the official announcement/publication with the description of
the objectives and criteria required for COST Action proposals to be evaluated and selected.
The Open Call allows submitting proposals on a continuous basis; the publication indicates
the Collection Dates.
6. Collection Date: means the date when the proposals for new COST Actions submitted during
a certain period are gathered and sent for evaluation.
7. Proposer: means the coordinator of the group of researchers who submit a proposal for a
COST Action in response to the Open Call.
8. COST Action: means the COST pan-European networking instrument allowing researchers
from COST Member Countries and Cooperating State to develop jointly their ideas and new
initiatives in a given field or topic of common interest.
9. Action’s Memorandum of Understanding (MoU): means the agreement accepted by a

minimum of five different COST Member Countries and/or Cooperating State describing the
Action objectives and the added value of networking. This document has to be endorsed by
any additional COST Member Country or Cooperating State joining the Action.
10. Action Participant: means any researcher who participates actively in a COST Action; this can
include researchers from Near Neighbour and International Partner Countries.
11. Management Committee (MC): means the group of researchers, nominated by the CNC, in
charge of the coordination, implementation, and management of an Action's activities as well
as supervising the appropriate allocation and use of the COST funding in view of achieving
the Action's scientific and technological objectives.
12. Researcher: means anyone engaged in the conception or creation of new knowledge, products,
processes, methods and systems, and in the management of the projects concerned. For the
purpose of participating in COST Actions, any individual, independently of his/her affiliation
or professional quality shall be considered as researcher.
COST 4113/13 3
DG G III EN
1. Basic principles
The rules reflect the principles of trans-nationality, fairness, transparency, openness, and
inclusiveness.
COST strives to avoid any conflict of interest2 in the evaluation of proposals.
All those involved in the evaluation of proposals must commit to confidentiality.
2. COST Open Call
The Open Call allows submitting proposals on a continuous basis; the call publication indicates the
Collection Dates; two cycles of evaluation are organised annually.
Collection Dates are communicated in the Official Journal of the European Union. An
announcement is also available on the COST website, together with the complete Open Call
schedule.
The objective of the COST Open Call is to enhance the scientific excellence and transparency of
COST through an accessible bottom-up opportunity with rigorous peer review. Proposals will be
evaluated on a competitive basis per domain, taking into account the available funds for the
particular Collection Date.
The Proposal is prepared by a Proposer acting on behalf of a network of researchers who see an
opportunity for advancing scientific, technological or social knowledge through the international
coordination support offered by COST.
The Proposer must be affiliated to a legal entity based in a COST Member Country or Cooperating
State. The network must be composed by researchers affiliated to legal entities from at least five
different COST Member Countries/Cooperating State.
The Open Call follows a two-stage submission process:
1. Preliminary Proposal: the eligible proposals are gathered at the Collection Date and
subsequently evaluated by the relevant Domain Committee (DC).
2. Full Proposal: Proposers of selected Preliminary Proposals will be invited to submit a Full
Proposal. Full Proposals will be peer reviewed by an External Expert Panel (EEP).
Proposers of selected Full Proposals will be invited to a Hearing by the relevant Domain
Committee.
2
COST 4160/10 “COST Code of Conduct”
COST 4113/13 4
DG G III EN
Preliminary Proposals may be submitted at any time using a dedicated secured online tool. They
will be considered for the upcoming collection. Login and proposal data are only valid during one
collection and are not transferable to the following collection.
The proposals must be written in English, as no translation is provided and peer reviewers come
from different countries.
3. The Preliminary Proposal
3.1. Submission
The Preliminary Proposal should provide an overview of the potential Full Proposal including
background information on the benefits, objectives, deliverables, scientific programme, and
organisational structure of the proposed new Action.
Preliminary Proposals may be submitted at any time during the year via a secured on-line tool
designated for this purpose; after registering a proposal, the Proposer will be able to access, edit
and submit the proposal until the Collection Date. Only the last submitted version of a proposal
is considered for evaluation.
3.2. Evaluation
A pre-check is performed by the COST Office in order to eliminate Preliminary Proposals that
do not meet the eligibility criteria as defined in the guidelines for assessment of applications for
COST Actions3.
Following the pre-check, the eligible Preliminary Proposals are allocated to the relevant COST
Domain Committees. The allocation is based on the Proposer’s stated preference at submission
and COST Office allocation procedures.
The Preliminary Proposals are evaluated online by the respective Domain Committee. Each
preliminary proposal will be evaluated by at least three experts.
The Proposer of the Preliminary Proposal can obtain feedback on the outcome of the evaluation
on his/her proposal webpage4.
The COST Office invites the submission of Full Proposals from selected Preliminary Proposals,
approximately three times the number of possible Actions for funding.
3
4
COST 4113/13 5
DG G III EN
4. The Full Proposal
4.1. Submission
Proposers of successful Preliminary Proposals can submit a Full Proposal within a period of
approximately two months via his/her proposal webpage.
Modifications to the submitted Full Proposals are allowed until the closing date. Only the last
submitted version of a proposal is considered for evaluation.
4.2. Evaluation
The evaluation of the Full Proposals is performed in two steps:
4.2.1. External Expert Panel evaluation (EEP)
The EEP is organised by the COST Office, in cooperation with the Domain Committee
Chair (or Vice-Chair). It is composed of external experts aiming to cover adequately the
required expertise and to ensure gender balance within the panels.
Within each Domain, the EEP will evaluate remotely the Full Proposals via a secured on-
line tool. Following the individual remote evaluation, the EEP convenes in a consensus
meeting. At this meeting, the EEP will agree upon a consensus score and comments for each
proposal. The EEP decides on the proposals to be invited to the subsequent evaluation step,
the Domain Committee Hearing (here after referred to as “DC Hearing”).
Feedback on the outcome of the EEP is made available to the Proposer on his/her proposal
webpage.
The European Commission liaison persons have access to all proposals and can submit
comments on the Full Proposals. These comments are made available both to the EEP and
the DC for consideration.
COST 4113/13 6
DG G III EN
4.2.2. DC Hearing
The Proposers retained by the EEP are invited to present orally their proposal to the Domain
Committee in a “DC Hearing”.
The DC Hearing will allow ranking the proposals, commenting on the oral presentation, and
highlighting any changes in the ranking proposed by the EEP.
Feedback on the outcome of the DC Hearing is made available to the Proposer via his/her
proposal webpage.
5. Approval
The COST Office prepares a final list of Action proposals based on the ranking and
recommendations of the Domain Committees.
The final decision for approval and funding of the Full Proposals for new Actions is taken by the
Committee of Senior Officials (CSO) on the basis of the final ranking of proposals further to the
EEP and DC Hearing.
The text of a successful full proposal approved by the CSO will form the basis of the Action’s
Memorandum of Understanding.
Final Provisions
The present rules for the submission, evaluation, selection and approval of COST Action proposals
shall be binding in their entirety.
Any change or derogations to the current rules are subject to the approval of the CSO.
___________________
COST 4113/13 7
DG G III EN
COST Action Proposal
Submission, Evaluation,
Selection and Approval
(SESA)
Guidelines
1
Table of Contents
Background ................................................................................................................................................... 3
Subject Matter and Scope ............................................................................................................................... 3
Introduction ................................................................................................................................................... 4
Preliminary Proposals for new COST Actions .................................................................................................. 5
In short ...................................................................................................................................................... 5
Submission steps ........................................................................................................................................ 5
Evaluation of Preliminary Proposals .......................................................................................................... 5
Eligibility check and criteria .................................................................................................................... 5
Evaluation by the Domain Committee (DC) .............................................................................................. 6
Full Proposals for new COST Actions ............................................................................................................. 6
Basic principles .......................................................................................................................................... 6
Submission steps ........................................................................................................................................ 6
Editorial instructions for Full Proposals .................................................................................................... 7
Evaluation of Full Proposals ...................................................................................................................... 7
Evaluation by the External Expert Panel (EEP) ....................................................................................... 7
Evaluation by the Domain Committee ...................................................................................................... 8
List of Annexes .............................................................................................................................................. 9
Annex I – Graphic representation of a COST Action evaluation process .................................................... 9
Annex II – Definitions .............................................................................................................................. 10
Annex III - Eligibility criteria for Preliminary Proposals ......................................................................... 12
Annex IV – Evaluation criteria for Preliminary Proposals ....................................................................... 13
Annex V - Preliminary Proposal template ................................................................................................ 14
Annex VI - Evaluation criteria for Full Proposals .................................................................................... 17
Annex VII - Full Proposal template.......................................................................................................... 20
FAQs ....................................................................................................................................................... 26
2
Background
In 2012, a revision exercise of the main COST documents led to the drafting and approval by the
CSO, of the revised COST Implementation Rules (COST doc. 4112/13) followed by the revised
rules for “COST Action Proposal Submission, Evaluation, Selection and Approval” (COST doc.
4113/13), the rules regarding the “COST Action Management” (COST doc. 4114/13) and the
“COST International Cooperation Rules” (COST doc. 4115/13).
A step further, is the introduction of a more detailed set of guidelines, such as the current ones,
addressing specific issues each time and linked to the revised COST rules, in order the assist in
submitting proposals under the revised rules.
Subject Matter and Scope

The current guidelines provide all the information necessary for COST Action proposers regarding
the submission, evaluation, selection and approval of such proposals.
These instructions and explanations depend COST doc. 4113/13 “COST Action Proposal
Submission, Evaluation, Selection and Approval”, which set the general framework as well as to
the information available on the Open Call webpage1.
The current Guidelines enter into force as of 18/09/2013.
1
www.cost.eu/opencall
3
Introduction
The submission of proposals for new COST Actions is a two-stage process. It is split between a
“Preliminary Proposal” and a “Full Proposal”.
More precisely the eligible “Preliminary Proposals” are gathered at the Collection Dates and
subsequently evaluated by the relevant Domain Committee (DC). For specific Collection Dates,
please refer to the Open Call schedule on the Open Call webpage2.
Proposers of selected Preliminary Proposals will be invited to submit a Full Proposal. Full
Proposals will be peer reviewed by an External Expert Panel (EEP). Proposers of selected Full
Proposals will be invited to a Hearing by the relevant DC.
Overall, a successful proposal for a new COST Action should entail the following
features/elements:
 create new networks rather than simply seek funding for existing networks;
 reach out for high scientific/technological quality in an innovative way;
 contribute substantially to the coordination and defragmentation of research efforts across
Europe and to the strengthening of Europe's scientific networking capacity (in the context
of the European Research Area), namely with the participation of relevant stakeholders;
 contribute strongly and visibly to European society, economic growth and welfare by
producing results of potential interest to important sectors such as public authorities, policy
institutions, standards bodies and/or private companies and industry;
 be based upon:
o careful consideration of the level of interest and relevant research resources in the
countries likely to participate in the Action;
o assessment of the added value expected from the coordination of national research
efforts by the Action;
 be flexible enough to permit at the implementation stage the inclusion of disciplinary
perspectives and activities not foreseen during the preparation of the proposal;
 identify and take into account R&D efforts supported by other national and international
funding schemes;
 encourage capacity building and the mobility of early-career European researchers.
2
www.cost.eu/opencall
4
Preliminary Proposals for new COST Actions
In short
The Preliminary Proposal should provide an overview of the potential Full Proposal (within 4 to
5 pages maximum), including background information on the benefits, objectives, deliverables,
scientific programme and organisational structure of the proposed new Action.
Submission steps
1. Create an account by registering via the applicant’s login page (see
http://www.cost.eu/opencall)
 You will be asked to enter your details (Name, Institution, Address and Email);
 An automatic email is sent to you containing your personal login information.
Note:
 The account is only active during the running collection and gives access to your personal
proposal webpage where you can edit your Preliminary/Full Proposal until the respective
deadline, retrieve any feedback from the evaluations and follow the status of your proposal.
 Attention: Proposals will not be transferred to any subsequent collection. If you want to
resubmit a proposal, you will need to create a new account linked to that collection.
2. Access the preliminary proposal online form by logging in with your credentials on the
applicants login page (see http://www.cost.eu/opencall).
3. Fill in the mandatory information on gender, Early Stage Researcher and re-submission.
4. Fill in the online form - Preliminary Proposal template (see 2.3). The template of the
preliminary proposal can be found in Annex III.
Evaluation of Preliminary Proposals
Eligibility check and criteria
During the three first weeks following the collection date, the COST Office performs an eligibility
check of all proposals against the “Eligibility criteria” (see Annex III). Proposers may be suggested
to choose a more appropriate Domain.
The results are published on the proposer’s personal page (for date see schedule on the Open Call
webpage).
Proposers whose proposal has been rejected based on eligibility criteria will also be informed by
e-mail.
5
Evaluation by the Domain Committee (DC)
The preliminary proposal evaluation process is chaired by the Domain Committee Chair assisted
by the COST Office.
During approximately four weeks, the evaluators, i.e. DC members or experts (who may be drawn
from the pool of "nominated DC experts” or from other sources), will evaluate the proposals within
their Domain. Each proposal will be evaluated by three evaluators at least. The DC Chairs do not
evaluate proposals in their Domain.
The evaluators mark their allocated Preliminary Proposals on the basis of the “Evaluation criteria”
(see Annex IV) by assigning marks between 1 and 6 to each of the six criterion and by adding an
appropriate and constructive comment.
The final marking (average of all the evaluator’s marks), the average mark for each criterion and
the individual comments are made available to the proposer on the proposer’s personal page (for
date see schedule on the Open Call webpage).
The COST Office normally invites, in total, around three times as many top ranked Preliminary
Proposals to submit Full Proposals as the total number of new Actions that can be supported by
the available funds.
Full Proposals for new COST Actions

Basic principles
Proposers of successful Preliminary Proposals are invited to submit a Full Proposal within a period
of approximately two months via his/her proposal webpage with the login provided for the
Preliminary Proposal submission.
The Full Proposal should provide on approximately 15 to 20 pages extensive information on the
background, objectives, benefits, scientific programme, organisation, deliverables, milestones and
dissemination plan of the proposed new Action. The text of a successful Full Proposal will
constitute the formal Technical Annex of the Memorandum of Understanding (MoU) of the COST
Action, and must therefore conform in all material respects, including formatting, to the template
provided by the COST Office (see below and examples of MoU on the Action’s web page).
Submission steps
a) Log in to your proposal webpage and access the Full Proposal form;
b) Fill in the online form - Full Proposal template (see Annex VII).
6
Editorial instructions for Full Proposals
 Follow the on-line template and make sure all the relevant parts are filled in following the
given heading structure. Please note that any modifications to the Full Proposals are
allowed until the closing date. The last saved version of the proposal is considered for
evaluation. No confirmation email is sent (the status is updated on the proposers webpage).
 Keep the format simple: use only a single font type and size (the default), left alignment,
simple bulleted and numbered lists, default line/paragraph spacing, no
footnotes/endnotes/headers/footers.
 Prefer to work directly in the on-line tool (conversion from MS Word for example can lead
to unexpected results, or text losses, if for example headings or automatic lists have been
used – in such case remove all automatic formatting before pasting to the on-line tool).
 The text limits are recommendations and can be slightly exceeded.
 The Proposal must be written in English. The Applicant is strongly advised to have the text
checked for correctness and clarity. COST does not provide translation or correction
services.
 No mentioning of individual scientists, institutes or organisations in the text of the
proposal.
 Check table layouts (ex: timetable).
 Use of capital letters for COST-specific and Action-related expressions; non-exhaustive
list: Action Proposal, Management Committee, Working Group, STSM (Short-Term
Scientific Mission), etc.
 Explain all acronyms (including those commonly used in the Framework Programme
context).
 Use of "Europe" or "COST countries" when referring to the overall geographical scope of
COST. "European Union" or "EU Member States" should only be used to refer to the EU
as a player ("EU legislation", "EU programmes", "EU policies" etc.) or when only EU
Member State(s) need to be explicitly mentioned, excluding COST countries not members
of EU.
 Avoid pronouns such as "I", "we"; rather use "the Action Proposal".
 Avoid any references to information contained in Part II - additional information (such as
"see attached list of experts").
Evaluation of Full Proposals
The evaluation of the Full Proposals is performed in two steps:
Evaluation by the External Expert Panel (EEP)
Within each Domain, during a period of approximately one month, the external experts will
evaluate remotely the Full Proposals against the evaluation criteria below.
At the end of this period, the EEP convenes in a consensus meeting. At this meeting, they agree
7
for each Full Proposal on consensus marks and comments, and recommend any number of
proposals to be invited to the DC Hearing among those that have reached the threshold mark (55).
Proposals marked below the threshold or not recommended to the DC Hearing are excluded from
further evaluation.
The consensus marks and comments, and the decision to be invited to the DC hearing or not, are
made available to the proposer via his/her webpage.
The Domain Committee Chair (or Vice-Chair) presents the EEP's outcome to the Domain
Committee.
Evaluation by the Domain Committee
The Proposers retained by the EEP are invited to present orally their proposal to the Domain
Committee (or its delegated Executive Group) in a “DC Hearing”, also addressing the comments
from the External Expert Panel.
The Domain Committee (or Executive Group) will rank and comment each proposal. The DC
comment will be made available to the Proposer via his/her proposal webpage.
The COST Office prepares the final list to be approved by the CSO based on the following
elements:
 The Domain’s share, which is generated by a filtering distribution that takes into account
the submission distribution per Domain. TDP is treated as a separate Domain;
 The number of Action Credits per Domain, which is generated from the past Collection
Dates, from the accumulated differences between the number of proposals actually
approved and the Domain’s share for those Collection Dates;
 The available funding.
The technical details of the methodology must be consensual between the DC Chairs and the
COST Office and any possible conflicts are to be resolved by JAF/CSO.
On the basis of the above procedure the JAF group will propose a definitive list for the Committee
of Senior Officials (CSO) who takes the final decision for approval and funding of the Full
Proposals for new Actions.
The COST Office will inform all the Applicants of the Full Proposals of the result of the selection
process.
8
List of Annexes
Annex I – Graphic representation of a COST Action evaluation process

Annex II – Definitions
Annex III - Eligibility criteria for Preliminary Proposals
Annex IV – Evaluation criteria for Preliminary Proposals
Annex V - Preliminary Proposal template
Annex VI - Evaluation criteria for Full Proposals
Annex VII - Full Proposal template
Annex I – Graphic representation of a COST Action evaluation process
9
Annex II – Definitions
For the purpose of the current guidelines, the following “Definitions” shall apply:
1. COST Member Country: means any country that was a founding member of the COST
Intergovernmental Framework or joined COST as a full member following the approval of
the COST Committee of Senior Officials (CSO).
2. COST Cooperating State: means a country that has been approved by the CSO to participate
without voting rights in the CSO.
3. COST National Coordinator (CNC): means the individual appointed by the COST Member
Countries and Cooperating State in charge of accepting the Action Memorandum of
Understanding, of the nomination process for the Domain Committee and Management
Committee members as well as providing information and support to national research
communities.
4. Domain Committee (DC): means the committee composed of scientific and technological
experts responsible for a particular research domain appointed by the COST Member
Country and Cooperating State.
5. Open Call for proposals: means the official announcement/publication with the description
of the objectives and criteria required for COST Action proposals to be evaluated and
selected. The Open Call allows submitting proposals on a continuous basis; the publication
indicates the Collection Dates.
6. Collection Date: means the date when the proposals for new COST Actions submitted during
a certain period are gathered and sent for evaluation.
7. Proposer: means the coordinator of the group of researchers who submit a proposal for a
COST Action in response to the Open Call.
8. COST Action: means the COST pan-European networking instrument allowing researchers
from COST Member Countries and Cooperating State to develop jointly their ideas and new
initiatives in a given field or topic of common interest.
9. Action Memorandum of Understanding (MoU): means the agreement accepted by a

minimum of five different COST Member Countries and/or Cooperating State describing
the Action objectives and the added value of networking. This document has to be endorsed
by any additional COST Member Country or Cooperating State joining the Action.
10. Action Participant: means any researcher who participates actively in a COST Action; this
can include researchers from Near Neighbour and International Partner Countries.
10
11. Management Committee (MC): means the group of researchers, nominated by the CNC, in
charge of the coordination, implementation, and management of an Action's activities as
well as supervising the appropriate allocation and use of the COST funding in view of
achieving the Action's scientific and technological objectives.
12. Researcher: means anyone engaged in the conception or creation of new knowledge,
products, processes, methods and systems, and in the management of the projects concerned.
For the purpose of participating in COST Actions, any individual, independently of his/her
affiliation or professional quality shall be considered as researcher.
11
Annex III - Eligibility criteria for Preliminary Proposals
 The text of the proposal follows the structure of the online template;
 The proposal fulfils the basic characteristics of a COST Action ;
 No research funding requested;
 Peaceful purposes;
 No duplication of existing research within COST, no overlap with an existing Action (but
the continuation of an existing Action is eligible);
 No duplication with another proposal within the same collection;
 No focus on a single network activity, such as a the set of up of a conference;
 Resubmissions should contain modifications, in particular with regards to evaluator
remarks;
 The text of the proposal must be anonymous (no disclosure of participants name or
institution);
 Are 5 or more COST Member States involved in the Proposal?
12
Annex IV – Evaluation criteria for Preliminary Proposals
1. Right for COST?
Is COST the best mechanism for achieving the Action's objectives?
High marks are given to proposals for which COST is the best adapted mechanism.
2. Public utility/science
Does the proposed Action address real current problems/ scientific issues?
High marks are given to highly exciting and interesting proposals on a very important and/or
timely topic.
3. Innovation
Is the proposed Action innovative?
High marks are given to highly innovative proposals.
4. Impact
Would the proposed network make a significant difference in terms of knowledge, capacity
building, social impacts, etc.?
High marks are given to proposals with high potential impact.
5. Networking
Are networking aspects well motivated and developed in the proposal?
High marks for proposals that both motivate the need for networking in the field and show how
the proposed networking will add value to the current state-of-the-art.
6. Presentation
Is the proposed Action presented in a clear, rational and understandable way?
High marks for proposals that are presented in a clear, rational and understandable way.
13
Annex V - Preliminary Proposal template
Remarks
- Timeout: to avoid losing data, do not forget to save regularly via the many “save changes”
buttons;
- You can access and edit your proposal any time before the collection deadline;
- Do not forget to press the submission button to submit your proposal (you will receive a
confirmation mail). Please note that even after pressing the submit button you can still edit your
proposal; the last submitted version (before the collection deadline) will be taken into account;
- Unsubmitted data will not be saved, nor transferred to the following collection;
- Anonymity: the proposers’ names are removed before the preliminary proposals are made
available to evaluators. The text of the proposal should not reveal the participants of the proposal,
whether by name, by institution name, or even by country (with the exception of International
Participants, whose country - and only country name - can be cited in the text. Be careful not to
disclose any participant if you add any references or short bibliography. A breach in the anonymity
rule can lead to the ineligibility of the proposal (see eligibility criteria) or the editing of the
proposal by the COST office so as to anonymize the proposal.
a) Proposal Title
- No character limit but it is strongly advised to avoid too long titles;
- Do not capitalize all the characters;
- Add a capitalized acronym at the end of the title and between brackets.
b) Proposal Abstract
Maximum 1000 characters, including blanks.
c) Keywords
Maximum 400 characters, including blanks. They should be separated by a comma.
d) Indicative COST Domain

Select from the list of Domains the one which fits best to your proposal (see also
http://www.cost.eu/domains_actions). Please note that during the eligibility check performed by
the COST Office after the collection date, you may be suggested to choose a more appropriate
Domain.
d) Text of proposal
Maximum 10000 characters, including blanks (a counter shows the number of characters left,
exceeding text will be lost).
Please do not overwrite the titles and respect their order in the text.
d.1) Background, Problems

This part should be an introduction to describe, in general terms, why it is desirable to launch the
COST Action in question. It should summarise the previous research and the current state of
knowledge in the field of the proposal. It could include an analysis of relevant research in the EU
14
Framework Programmes and other European fora. It may be useful also to compare the European
research with that in, for example, the USA, Canada, Japan or other parts of the World.
In addition it should explain the reasons for the proposed cooperation with a distinction between
the objectives, the expected results and the means to achieve them. As far as possible, this should
be done with emphasis on immediate or future applications envisaged, so that even a reader who
is not a specialist in the field obtains a clear picture of the expected benefits of the Action.
You may briefly describe also possible complementarity with and value-added to ongoing or
planned research in the EU Framework Programme and other European organisations as one of
the goals of COST is to avoid duplication of efforts in Europe.
Indicate the background of the proposal, the specific problems the network wants to solve and the
goal the network would like to achieve. This part should demonstrate that the proposal addresses
real current scientific and or technical issues with a high relevance for European society.
d.2) Benefits
This part should explain the expected benefits of the proposal itself without the networking
aspects. These benefits could be societal, scientific and/or in the field of technology. There may
be also other benefits for other areas which should be elaborated here.
d.3) Objectives, Deliverables and Expected Scientific Impact

This part should clearly indicate what one expects to achieve through the Action, in particular what
will be the expected impact of this Action. It is very important to explicitly state all the objectives,
whenever possible in quantitative terms making it easier to evaluate, how well the Action may
achieve its goals. As far as possible, the likely end-users of expected results should be clearly
indicated. In formulating objectives one has to distinguish between the aims (something toward
which effort is directed) and the means to achieve them (methods or ways for accomplishing
something). Carefully avoid all specifications of means - e.g. scientific problems to be solved as
well as research tasks - as they belong to part d) Scientific programme.
A clear list of deliverables accompanied by a list of milestones should be provided.
d.4) Scientific Programme and Innovation

Here the most important research tasks to be carried out should be described (the structure of the
work plan), with necessary explanation of how they will lead to achieving the objectives. In
particular the innovative elements of the proposals and its originality should be presented.
You should remember that scientists that have not participated in the preparation are also entitled
to join the network at a later stage if their countries sign the MoU. For that reason, the proposal
must provide an open and flexible framework making it possible for any interested country to join
the Action.
It will greatly enhance the clarity of the proposal if this Section is wholly focused on outlining the
scientific content of the Action, while all organisational matters such as setting up Working Groups
are dealt with in the following Section “Organisation”.
15
d.5) Organisation
The main purpose of this part is to give a clear picture of the arrangement of the Action. When
you have clarified the reasons for the proposed co-operation, you should explain why COST seems
to offer the best framework for it, for example as compared with the EU Framework Programme
and other European organisations. This can be explained by describing the advantages or benefits,
which should be gained from carrying out your project within the COST framework.
This part should clearly reflect the fact that a COST Action is implemented through a concerted
action. This means that the research is carried out in the participating countries and financed by
other sources than COST. COST provides via the COST Action the means for the co-ordination
of the research.
d.6) Participants interested in network (name, institution and country)

List here the participants of your network including yourself.
In order to guarantee the anonymity of the proposers and avoid any conflict of interest, only the
country list will be shown to the assessors.
The list must contain at least 5 participants from 5 different COST member countries and include
yourself.
It is not necessary to list all participants, especially from a same institution or country (you will
have this opportunity in the Full Proposal), prefer to highlight the geographical variety.
If your proposal involves more than 10 different participants, it is recommended to just include 10
of them; the full list will be requested if you are invited to submit a full proposal.
16
Annex VI - Evaluation criteria for Full Proposals
Section A. Science and Networking (marks from 1 to 4 - weight 2)

A.1 Does the proposed Action address real current problems/scientific issues?
4. The topic is very important and /or timely and proposal presents the correct approaches.
3. The topic is very important and /or timely, but proposal fails to present the correct approaches.
2. The topic is not important nor timely, although proposal presents the correct approaches.
1. Serious lack of substance and/or relevance.
A.2 Does the proposed Action show awareness of the state-of-the-art of the relevant scientific/
technical/socio-economic fields?
4. Excellent and up to date awareness of relevant scientific/technical fields
3. Good awareness of relevant fields.
2. Defective awareness of relevant fields.
1. Serious lack of awareness of relevant fields.
A.3 Is the proposed Action innovative?

4. Highly innovative: identifies a significant new problem and/or a significant new approach.
3. Innovative in some notable aspects.
2. Not very innovative: the topic is already well-studied and/or the proposal largely follows a well-
trodden approach.
1. Not at all innovative.
A.4 Does the proposed Action answer a need for the networking of experts in the field?
4. Networking in this field ranks amongst the best mechanisms to progress the state-of-the-art and
the proposal uses such a mechanism in a sound manner.
3. Networking in this field ranks amongst the best mechanisms to progress the state-of-the-art, but
the proposal fails to use such a mechanism in a sound manner.
2. Networking in this field is not amongst the best mechanisms to progress the state-of-the-art,
although the proposal uses such a mechanism in a sound manner.
1. Networking in this field is not amongst the best mechanisms to progress the state-of-the-art and
the proposal fails to use such a mechanism in a sound manner.
Section A
- Comments:
- Total Mark (Max 16 X 2 = 32)
Section B. Impact (marks from 1 to 4 - weight 2)

B.1 A COST Action may have impacts in various valuable directions. This Action mainly aims at
(choose between a, b, or c.): a. meeting European economic or societal needs / b. developing the
scientific or technological field / c. both a and b.
4. Important impacts very likely in several respects.
3. Some notable impacts likely.
2. May be some minor impacts.
17
1. Unlikely to make useful impacts.
B.2 Are there clear plans for stimulating the production of high quality outputs?
4. Plans for outputs are clear, wide-ranging and ambitious.
3. Plans for outputs are reasonable.
2. Plans for outputs are unambitious or defective.
1. Plans for outputs are minimal or absent.
B.3 Is attention given to the involvement of stakeholders in order to increase the potential
application of results (including, where appropriate, fostering their commercial exploitation)?
4. Stakeholders are already part of experts who took part in the preparation of the proposal.
3. Plans for implication of stakeholders are clear, wide-ranging and feasible.
2. Plans for implication of stakeholders are reasonable.
1. Plans for implication of stakeholders are unambitious or defective.
Section B
- Comments:
- Total Mark (Max 12 X 2 = 24)
Section C. Structure and organisation (marks from 1 to 4 - weight 1)

C.1 Is the proposal presented in a clear, convincing, and appropriate way?
4. Very clearly written with compelling argument; fully appropriate format.
3. Well written; argument is easy to follow; appropriate format but may need minor changes;
2. Poorly written, but argument can be followed with effort; and/or defective format.
1. Poorly written; argument is unclear; and/or inappropriate format.
C.2. Are the workplan and organisation appropriate?

4. Workplan and organisation make full, productive and cost-effective use of COST opportunities.
3. Workplan and organisation are reasonable, any defects are minor.
2. Workplan and/or organisation show significant defects.
1. Workplan and/or organisation are lacking or inappropriate or unclear.
C.3 Are the time schedule and the setting of milestones appropriate?
4. Schedule and milestones are well-defined and practical.
3. Schedule and milestones are reasonable.
2. Schedule and/or milestones show some defects.
1. Schedule and/or milestones are lacking or inappropriate or unclear.
C.4. Are appropriate plans made for monitoring and evaluating the achievement of objectives?
4. Monitoring and evaluation plans are well-defined and practical.
3. Monitoring and evaluation plans are reasonable.
2. Monitoring and evaluation plans show some defects.
1. Monitoring and evaluation plans are lacking or inappropriate or unclear.
18
Section C
- Comments:
- Total Mark (Max 16)
Section D. Contribution to wider COST goals (marks from 0 to 1 - weight 1)

D.1 How well does the proposed Action aim to involve early stage researchers?
1. An innovative plan is presented in addition to the standard template in Section E.4 of Full
Proposal
0. Otherwise.
D.2 How well does the proposed Action aim at gender balance?
1. An innovative plan is presented in addition to the standard template in Section E.4 of Full
Proposal
0. Otherwise.
D.3 Does the proposed Action have the potential to contribute to the solution of global challenges
in a global dimension?
1. Proposal will certainly attract interest from a wide range of non-COST Countries if approved
0. Otherwise.
Section D
- Comments:
- Total Mark (Max 3)
Total Mark for the Full Proposal (Threshold: 55 points out of 75)
E. Overall recommendation of the EEP
- Comments:
- Strength of proposal:
- Weakness of proposal:
- Invitation (or not) to the DC hearing
19
Annex VII - Full Proposal template
Part I - Draft Technical Annex
I.A. Abstract (maximum 200 words)

Be clear and precise as this section will form the basis for COST information – web site and
booklets – and reporting. The Abstract should include the broader scientific context of the Action
as well as the expected deliverables and benefits. It should also indicate the European added value
of the Action and the reasons for undertaking it in the COST framework.
Keywords: maximum 5 keywords or very short phrases.
I.B. Background (maximum 2-3 pages)
I.B.1 General background

- Define the research topic in such a way that it is clear that the network will address real current
problems or scientific issues.
- Inform about the wider relevance of the Action (why is it desirable to launch it as COST
Action).
- Explain why COST, which funds only networking and capacity-building activities and not
research, is the best mechanism for support. State reasons why COST seems to offer the
appropriate framework for the Action, compared to other research frameworks such as the EU
Framework Programme and other European organisations.
- Describe the advantages or benefits which should arise from carrying out your project within the
COST framework.
I.B.2 Current state of knowledge

- Summarise the previous research in the field of the proposal.
- Describe the current state of the art, including relevant research within the EU Framework
Programmes and other EU fora, comparison of EU research with that in other parts of the world.
- Explain how the Action will be innovative in addressing either a new problem or a new approach
to an existing problem.
I.B.3 Reasons for the Action

- Reasons for launching the Action, indicating the need for an experts network in the area and the
added value of the Action networking. Emphasise immediate and future benefits and envisaged
applications (understandable for non-specialists readers).
- Indicate whether the Action is mainly aimed at European economic/societal needs, or at
scientific/technological advance, or both.
- Clearly distinguish between objectives, expected results and the means that are needed to achieve
them. The impact of COST comes from concrete outcomes, not just activity; so indicate how the
Action will aim for maximally productive outcomes.
20
I.B.4 Complementarity with other research programmes (if appropriate)
Mention Relevant links to and complementarity with any current and/or planned European
research projects, such as the EU Framework Programme and other European organisations (bear
in mind that avoiding duplication is one of the goals of COST).
I.C. Objectives and Benefits (maximum 2 pages)
I.C.1 Aim
Standard text as first item of this section (as this sentence will be quoted word for word in point 2
of the Memorandum proper, it should be extremely concise – max 3 to 4 lines):
“The main objective of the Action is …”
The impact of COST comes from concrete outcomes, not just activity. Therefore indicate clearly
what should be achieved through the Action. Given that all COST Actions are networks of
scientists, the objectives should therefore clearly state the purpose of such networking, indicating
- where possible - clear expected deliverables, not only research activities to be undertaken.
However, if the proposed Action is of specially novel or "high risk" nature so that concrete
deliverables are difficult to envisage, this should be explained clearly in the proposal.
I.C.2 Objectives
List and explain secondary objectives (whenever possible in quantitative terms, which will make
it easier to evaluate how well the Action may achieve its goals).
I.C.3 How networking within the Action will yield the objectives?
Distinguish between objectives (aims of the Action) and means needed (manpower, equipment,
etc.) to achieve these objectives (avoid any reference to method and means – e.g. scientific
problems to be solved as well as research tasks – as they belong to section D (Scientific
programme) detailed below).
I.C.4 Potential impact of the Action

Describe expected benefits that will stem from the Action (with reference to section B).
I.C.5 Target groups/end users

Reflect on the likely stakeholders and end users that will exploit the expected results. Indicate
whether they were involved in the preparation of the proposal.
I.D. Scientific Programme (maximum 3-4 pages)

I.D.1 Scientific focus
- Describe the most important research tasks to be coordinated by the Action.
- Provide a structured, but not too detailed work plan flexible enough to permit the inclusion, at
the implementation stage, of disciplinary perspectives and activities not foreseen during the
preparation of the proposal. Keep the framework of the Action open and flexible.
21
- Explain the human and technical means to achieve the objectives described in section C.
- Remember that this section must be clear to non-specialists (even if the description may be more
“technical”).
I.D.2 Scientific work plan – methods and means

- Do not mention explicitly the names of individual scientists, specific research institutions or other
bodies (only exceptionally, if the Action cannot be implemented without the participation of a
specific Institution, you should clearly mention this with the relevant explanation).
Always remember that scientists who have not participated in the preparation are also entitled to
join if their countries accept the MoU.
- Focus on work plan and methods of the Action and not on its organisation.
- Mention for your Working Groups (see E.2 Working Groups), their objectives and what they will
achieve.
I.E. Organisation (maximum 2 pages)

General remark: You need not reiterate organisational features common for all COST Actions,
described in the "Rules and Procedures for Implementing COST Actions" (doc. COST 4112/13)
and in COST doc. 4114/13 “COST Action Management”. As a rule, organisational matters should
be mentioned only if you intend to apply them in some specific way. In order to avoid unnecessary
repetitions or contradictions, please refer to Rules and Procedures when drafting this section.
I.E.1 Coordination and organisation

- Give a clear picture of the management and organisation of the Action.
- Reflect the fact that a COST Action is implemented through a concerted action, which means
that the research is carried out in and financed by other sources than COST, while COST provides
the means for the necessary co-ordination.
- Use organisational features common to all COST Actions, but also allow for limited Action-
specific variations (e.g. you may want to introduce a Steering Group, an Editorial Board, STSM
manager, etc.). Consult “Rules and Procedures for implementing COST Actions”.
- Mention milestones – major achievements that are crucial to the future direction of the Action.
- Explain how the coordination of national research will be implemented (including the creation
of possible common research teams, conferences and workshops, short-term scientific missions or
other exchanges between laboratories, training schools, websites, etc.).
- Be aware of the obligation to set up an Action specific website that will not duplicate general
information already available from the COST website (e.g. signatory list, MC list, etc.) and to keep
it updated: Include a plan to keep this website up to date, both to serve the needs of the participants
and with the specific aim of ensuring the dissemination or exploitation of the results of the Action.
- As a rule, do not list names of interested research establishments and scientists. (This will be part
of the Additional Information – Part II.)
22
I.E.2 Working Groups
- Working Groups are a useful way of extending the Action beyond the membership of the
Management Committee and of sharing workloads.
- An Action has normally 3-6 but not more than 6 Working Groups.
- If you plan Working Groups, explain their organisation (without repeating unnecessarily the
“Scientific Programme” given under Section D).
I.E.3 Liaison and interaction with other research programmes

- Address possible liaisons and interaction with other COST Actions and other European and
international research programmes, such as the EU Framework Programme and other European
organisations.
- Indicate how these interactions will be organised: by exchange of information, meetings, by joint
seminars or any other means.
I.E.4 Gender balance and involvement of early-stage researchers

The following paragraph is compulsory:
“This COST Action will respect an appropriate gender balance in all its activities and the
Management Committee will place this as a standard item on all its MC agendas. The Action will
also be committed to considerably involve early-stage researchers. This item will also be placed
as a standard item on all MC agendas.”
- Please add any additional support the Action plans concerning gender balance and the
involvement of early-stage researchers, in particular with respect to the organisation of training
schools, STSMs etc.
- Explain how you intend to achieve a capacity building.
I.F. Timetable Maximum one page

- Give a clear picture of the timescale of the Action.
- The duration of a COST Action is four years, unless there are specific cases to be approved by
the CSO, on the basis of a justification provided in the proposal.
- Use relative time scales (Year 1, Year 2, etc.), not specific years.
- Please use a table.
I.G. Economic Dimension

- The Economic Dimension gives a rough estimate of the total research activity leveraged by the
Action based on a formula “Number of COST countries that have participated in the Action” X
EUR 4 million. No data entry is needed here, the system will calculate the amount automatically
based on the number of COST participants entered in part II – A.
I.H. Dissemination Plan Maximum 2 pages

I.H.1 Who?
Identify the target audiences for the dissemination of the results of the Action (in particular
findings and recommendations), e.g. other researchers working in the field; other research
frameworks; research Institutes and Academia; Standards Bodies; industry (represented by
23
manufacturers and service providers); European level policy makers; Government policy makers,
regional planners and policy makers; general public.
I.H.2 What?
- Describe the dissemination methods you intend to use.
- For each of your audiences you may choose several of the existing possibilities, e.g.
- Posting of general information on a public website;
- Posting of working documents on a password protected website;
- Set up of an electronic communication network (internet discussion forum, e-mail network,
etc.);
- Publications: state of the art reports, interim reports, case study reports, proceedings,
guidelines, manuals, final reports;
- Events: workshops, seminars and conferences organised by the MC, contributions to other
national and international conferences and symposia;
- Articles in peer-reviewed scientific and technical Journals;
- Non-technical publications.
I.H.3 How?
- Describe how these dissemination methods will be used.
- Note that dissemination goes beyond publication of results.
- Take into consideration the progress of the Action as well the results of its evaluation in updating
the dissemination plan during the course of the Action.
For additional information on dissemination activities, see the Guidelines for Monitoring,
Evaluation and Dissemination of Results of COST Actions.
Part II Additional Information (maximum 10 pages)

This part is not included in the Memorandum of Understanding (MoU).
The main purpose of the second part of the proposal is to facilitate the assessment of the proposal
and the nomination of National Representatives to the Management Committee (MC). This part
will not be part of the MoU. Note that only part A. List of Experts is mandatory as the information
given here is important for the later nominations to the Management Committee.
II.A. List of Experts

For COST country participants, list the Participants to the proposal (PtP), potential Management
Committee (MC) / Working Group (WG) members that expressed their interest (Participant detail
information and scientific expertise).
For participations from Near Neighbour Countries (NNC) , International Partner Countries (IPC)
or specific Organisations (European Commission, EU Agencies, European RTD Organisations
and International Organisations), a description of the mutual benefit and of the targeted scientific
activities, including the included Working Group(s) is requested.
24
II.B. History of the Proposal
The purpose of this section is to give the historical background of the proposal: how the idea of
the COST Action was born and how the subsequent definition of the objectives and the pre-
proposal planning was carried out.
II.C. Preliminary Work Programme

Especially if the proposal is very complex and based on participation of research teams from
different fields of research interacting in a specific way, you may wish to explain how this has
been envisaged, at a more concrete level than that indicated in the draft Technical Annex.
II.D. Recent Publications

In order to make it easier to assess the scientific merits of the proposal, you may wish to compile
a short list of recent scientific publications relating to the topic of the Action. If desired, you could
group all the publications authored or co-authored by you as a kind of scientific self-portrait. This
should be a maximum of 2 pages.
II.E. Further Remarks

In this subheading you may add any information or remarks but also comment on the following
assessment criteria.
- To what extent does the proposed network aim at involving early-stage researchers?
- To what extent does the proposed network aim at being gender balanced?
- Does the number of countries the Applicants come from reflect a wide European dimension?
- To what extent have provisions been made for monitoring and evaluating the achievement of
objectives?
- To what extent have provisions been made for assessing potential application, and fostering
exploitation, of results?
25
FAQs
1. Can I make changes to my proposal? Yes, any modifications to the proposals, whether it
is preliminary or full is allowed until the closing date. Preliminary proposal can be
submitted many times before the deadline (the last submission is considered for
evaluation). For the Full Proposal, the last saved version of the proposal is considered for
evaluation (there is no submission button, nor confirmation mail).
2. I have submitted to a previous collection, can I resubmit? Yes, but you need to create a
new account and re-enter all data. Do not forget to amend the proposal taking into account
the evaluators comments.
3. Where can I see sample of successful proposals: the Action Memorandum of
Understanding available in PDF version on the Action page contains the main information
included in a Full
4. Can I add pictures? No, only text or tables at both the Preliminary as the Full Proposal
stage.
5. Pasting to the online form from a word processor produces strange results. This happens if
you use features like automatic headings, bulleted/numbered lists, indents, justified text,
different font types and sizes. Please use a basic formatting following the editorial
instructions above.
6. Are amendments to the title possible? Yes, at the Full Proposal stage send an email to
opencall@cost.eu with the new title.
7. I have exceeded the text limits, are there any consequences? The chapter limits are
recommendations, slight excesses are not penalized.
8. Is there a redress procedure? No
9. Can proposals be transferred from one collection to another? No, a proposal is linked to
one collection and is cannot be transferred to any subsequent collection (however re-
submissions of the same proposal are allowed).
26

Datos de Densidad de Microalga

Uploaded by

Copyright:

Available Formats

You might also like

Datos de Densidad de Microalga

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Datos de Densidad de Microalga

Uploaded by

Copyright:

Available Formats

Intensive Microalgae and Copepod Culture

Workshop 16-18 December 2013

IMProvement of AQuaculture high quality fish fry production (IMPAQ)

COpepod egg Mass production in Aquaculture (COMA)

Roskilde University (RUC)

Graduate School of Environmental Stress Studies (GESS)

17. December: Intensive Copepod Production

18. December: Cost Initiative

Intensive Microalgae Production

Intensive Copepod Production

Network, Innovation and concluding remarks

REVIEW www.rsc.org/ees | Energy & Environmental Science

Microalgae as biodiesel & biomass feedstocks: Review & analysis of the

Peter Williams is professor Dr Lieve Laurens is a researcher

In this section we introduce the aspects of algal biology and

Proteins have important commodity value as animal feed.

(iii) Nucleic acids. Nucleic acids (RNA and DNA), in

(iv) Lipids. As with carbohydrates, lipids serve both as

energy reserves and structural components (membranes) of the

Algal lipids C1H1.83O0.17N0.0031P0.006S0.0014 36.3 15–60

Table 3 Mean ( standard error) lipid class content as a percent of total

Composite from all reports (n ¼ 46) 35 ± 3 40 ± 2 25 ± 2

increases during nutrient limitation (see Fig. 7). As nutrient

limitation also affects growth rate, this provides an explanation

for the apparent inverse relationship between growth rate and

nate the production of nitrogen containing material (proteins

and nucleic acids) but continued to synthesise lipids and carbo-

hydrates. Whereas it may explain the lipid–nutrient relationship,

it cannot explain observations where growth was controlled by

light rather than inorganic nitrogen and where again a negative

of the biochemical mechanism, the evidence for of a negative

relationship between growth rate and lipid content (see Fig. 6)

has an important bearing on the strategy adopted for biomass

production, especially when products other than the lipid

component are valuable (see e.g. Fig. 24).

2.4. Algal lipids as fuels

From their molecular composition, we can anticipate a number

of properties relevant to the potential of algal lipids to serve as

fuels. A summary of the fuel properties of biodiesel from

a microalga (heterotrophically grown Chlorella protothecoides)44

high phospholipid and glycolipid concentrations. These lipid

classes contain nitrogen, phosphorous and sulfur that may be

problematic with regards to engine performance, if present in

fuels. However, it is likely that the sulfur, phosphorus and

nitrogen-containing compounds would end up in the water-

expect these elements to be low to non-existent in algal biodiesel.

greatly affect the potential fuel yield by transesterification. Thus

Fig. 5 A: triangular plot of the proportions of lipid, protein and

Light reaction: 2H2O / 4[H] + O2 + energy

Light-independent reactions: 4[H] + CO2 / [CH2O] + H2O

(The notation [H] refers to the combination of the reduced

Properties Micro-algal Higher plant Diesel fuel ASTM biodiesel standard

Density/kg dm3 0.864 0.877–0.887 0.838 0.86–0.90

light-independent reaction operates over timespans of seconds to

maximising yields of mass algal culture.

(i) Light, its absorption and the formation of ‘‘reducing

4C6H12O6 / 8 acetyl CoA + 8CO2 + 8ATP + 16 NAD(P)H (1)

8 acetyl CoA + 7ATP + 14 NAD(P)H / CH3(CH2)14COOH (2)

4C6H12O6 / CH3(CH2)14COOH + 8CO2 + ATP + 2NAD(P)H (3)

Table 3 Mean ( standard error) lipid class content as a percent of total

Density/kg dm3 0.864 0.877–0.887 0.838 0.86–0.90

Source Organism g dw/m2 d1 % dw PBR RW Hybrid PBR RW Hybrid