Evaluation of Nannochloropsis

gaditana raw and hydrolysed biomass

at low inclusion level as dietary
functional additive for gilthead
seabream ( Sparus aurata ) juveniles -
eBook PDF
Visit to download the full and correct content document:
https://ebooksecure.com/download/evaluation-of-nannochloropsis-gaditana-raw-and-
hydrolysed-biomass-at-low-inclusion-level-as-dietary-functional-additive-for-gilthead-s
eabream-sparus-aurata-juveniles-ebook-pdf/
 Aquaculture 556 (2022) 738288

Contents lists available at ScienceDirect

Aquaculture
journal homepage: www.elsevier.com/locate/aquaculture

Evaluation of Nannochloropsis gaditana raw and hydrolysed biomass at low

inclusion level as dietary functional additive for gilthead seabream (Sparus
aurata) juveniles
María Isabel Sáez a, Alba Galafat a, Antonio Jesús Vizcaíno a, Elena Chaves-Pozo b,
María Dolores Ayala c, Marta Arizcun b, Francisco Javier Alarcón a, María Dolores Suárez a,
Tomás Francisco Martínez a, *
a
Departamento de Biología y Geología, Escuela Superior de Ingeniería, CEIMAR, Universidad de Almería, 04120 Almería, Spain
b
Centro Oceanográfico de Murcia, Instituto Español de Oceanografía (IEO – CSIC), Carretera de la Azohía s/n, Puerto de Mazarrón 30860, Murcia, Spain
c
Departamento de Anatomía y Anatomía Patológica Comparada, Facultad de Veterinaria, Universidad de Murcia, 30100 Murcia. Spain

A R T I C L E I N F O A B S T R A C T

Keywords: Abundant research is being carried out in the last years aimed at exploring microalgal biomass as nutrient source
Bioactive compounds for different species of aquacultured fish. Some microalgae species, such as Nannochloropsis gaditana, have thick
Cellulase cell walls rich in cellulose, which might well reduce the bioavailability of intracellular active compounds. Among
Functional additive
the alternatives aimed at overcoming this limitation, cellulase enzyme hydrolysis is proposed as a convenient and
Lipid oxidation
Microalgae hydrolysis
practical solution. In this regard, an in vitro assay was carried out, in which N. gaditana biomass was treated with
cellulase (5% w/w basis) and the release of soluble compounds (reducing sugars, free amino acids and total
phenolics) into the reaction medium was measured and compared to untreated raw biomass. The results
confirmed increased yields of those compounds as a result of the enzyme pre-treatment. A 90-d feeding trial was
also carried out in order to assess in vivo the influence of the inclusion of N. gaditana in feeds on juvenile gilthead
seabream (Sparus aurata) growth, digestive physiology and body composition. Microalgal biomass was added at
two inclusion levels (25 and 50 g kg− 1 dry weight) in four experimental feeds, either crude or enzymatically pre-
treated. Animals (15.1 g initial body weight) were randomly assigned to five dietary treatments (two inclusion
levels, 2.5 and 5%, and two microalgae formats, raw and enzymatically hydrolysed, plus a microalgae-free
control), and distributed triplicate tanks per dietary treatment. Fish were withdrawn after 45 and 90 days,
and proximate composition, muscle fatty acid and amino acid profiles, muscle and liver lipid oxidation,
instrumental skin colour, digestive enzyme activities, as well as structural and ultrastructural changes in the
intestinal mucosa were determined. No differences attributable to the dietary treatments were found with regard
to fish growth or proximate composition at the end of the feeding trial. On the contrary, the inclusion of
microalgal biomass, irrespectively of the cellulase pre-treatment, caused beneficial effects on some physiological
parameters (namely digestive mucosa structure and functionality, oxidative status of muscle lipids, and instru­
mental colour). The only clear improvement found in fish attributable to the cellulase pre-treatment of the
microalgal biomass was related to the prevention of muscle lipid oxidation. Overall, the results suggest that
N. gaditana used as additive (at inclusion level below 5%) in feeds might represent a valuable nutritional strategy
for S. aurata juveniles, even if growth was not affected.

1. Introduction aquaculture feed manufacturing (Yarnold et al., 2019).

However, they are also drawing the attention of nutritionists as a
Interest in microalgae has emerged strongly in recent years, as they valuable source of bioactive compounds, many of which remain to be
present a valuable and still relatively unexploited potential to reduce identified. Different studies have pointed out that these substances can
dependence on unsustainable ingredients, namely fishmeal, in exert positive effects on several aspects of fish physiology, even if added

* Corresponding author.
E-mail address: tomas@ual.es (T.F. Martínez).

https://doi.org/10.1016/j.aquaculture.2022.738288
Received 23 December 2021; Received in revised form 30 March 2022; Accepted 22 April 2022
Available online 28 April 2022
0044-8486/© 2022 Elsevier B.V. All rights reserved.
M.I. Sáez et al.
at low inclusion level (e.g., less than 10%) in feeds (Becker, 2003; Kiron,
2012). Given their current production costs, the interest in microalgae is
turning from large-scale use as main ingredient towards their use as
functional additives at low inclusion levels in feeds. The search for
bioactive compounds aimed at improving not only fish growth, but also
the general condition of the animals is nowadays a thriving field in
aquaculture research. This concept is based upon numerous reports
indicating that many microalgae species are valuable sources of essen­
tial n-3 long chain polyunsaturated fatty acids (n3-PUFAs), vitamins,
minerals, pigments, and polyphenols, among others (Sansone et al.,
2020; Teimouri et al., 2013; Tibaldi et al., 2015; Shah et al., 2018). In
this context, Nannochloropsis gaditana, by virtue of its richness in eico­
sapentanoic acid (EPA, C20:5n-3), pigments and other natural antioxi­
dants and other bioactive compounds (Kilian et al., 2011; Tibbetts et al.,
2017; Cerón-García et al., 2018), together with their availability at in­
dustrial or semi-industrial scale (Heredia et al., 2021; Kavitha et al.,
2021), is a promising candidate as a commercial additive in aquafeeds.
However, it has been reported in the literature that the theoretical
nutritional potential of microalgae might not always be reflected
straight on the animals, neither in terms of fish growth nor on physio­
logical condition (Cerezuela et al., 2012; Cardinaletti et al., 2018). This
could be explained by the existence of a cell wall that limits the
bioavailability of inner microalgal compounds (Wu et al., 2017; Yong
et al., 2020). Specifically, the existence of a cellulose-rich cell wall in
certain genus, such as Nannochloropsis, together with the lack of diges­
tive cellulase activity in fish, might well limit their further practical
utilization. In fact, cellulose accounts for 75% of cell walls dry matter in
N. gaditana (Scholz et al., 2014).
When it comes to improving bioavailability prior to their inclusion in
aquafeeds, several strategies have been proposed in the literature aimed
at weakening or disrupting microalgae cell walls, all of them with ad­
vantages and disadvantages (Gomes et al., 2020; Rojo et al., 2021;
Pagels et al., 2021). Studies have reported that both enzymatic (Agboola
et al., 2019; Batista et al., 2020; Galafat et al., 2020) and physical
(Timira et al., 2021; Rojo et al., 2021) disruption strategies (such as bead
milling, ultrasonication and high-pressure homogenization, Lee et al.,
2010; Günerken et al., 2015) are able to increase the yield of antioxidant
compounds (Almendinger et al., 2021), as well as the nutrient avail­
ability and digestibility of algae by fish (Teuling et al., 2019). However,
species-specific factors must be taken into account when selecting the
most appropriate method to disrupt microalgal cell walls (Batista et al.,
2020).
Even if successful at laboratory scale, however, methods of physical
disruption may not end up in scalable, and economically feasible pro­
cedures applicable to the feed processing industry, taking into account
that additional costs should be added to the production prices of
microalgal biomass (Batista et al., 2020). Consequently, there is still
considerable scope for developing simple, economical, and cost-
effective cell wall disruption protocols. The use of hydrolytic enzymes
capable of weakening cell walls prior to its incorporation into feeds
could overcome many of the existing limitations. Fibrolytic enzymes,
not least cellulases, have a wide range of industrial applications, and
hence they are available at affordable prices, and generally speaking,
any bioprocess including enzymes could be certainly scalable at indus­
trial level.
In this context, we hypothesize that the use of cellulases capable of
weakening N. gaditana cell walls may represent a valuable strategy to
improve the bioavailability of nutrients and bioactive compounds of the
biomass when added into gilthead seabream (Sparus aurata) experi­
mental diets. To this end, either crude or enzymatically hydrolysed
N. gaditana was added at low inclusion levels (2.5 and 5% w/w) to diets
for gilthead seabream juvenile. This is, the microalgal biomass was
assessed as a potential functional additive rather than as a main ingre­
dient. A 90-d feeding trial was carried out, and the occurrence of po­
tential effects of the microalgae on fish growth, muscle composition,
lipid oxidative status, skin pigmentation, and digestive structure and
2
Aquaculture 556 (2022) 738288
functionality were assessed.
2. Materials and methods
2.1. Microalgae biomass and enzyme hydrolysis
Nannochloropsis gaditana was cultured in tubular photobioreactors at
the pilot plant (EU-H2020 SABANA facilities) of the Universidad de
Almería (Spain) as reported by Menegol et al. (2019). The culture pH
was maintained at 8 by the on-demand addition of CO2. The culture was
harvested daily by centrifugation (at a dilution rate of 0.3 d− 1), then the
concentrated biomass was freeze-dried. Raw microalgae biomass
(approx. 15% dry matter) was freeze-dried and stored at − 20 ◦ C until
further use. The proximal composition of N. gaditana meal yielded
44.5% crude protein, 33.3% carbohydrates, 4.5% ash, and 17.7% crude
lipid on dry matter basis.
Enzymatic hydrolysis was carried out by mixing the microalgal
biomass, at a final concentration of 150 g dry weight L− 1 in 50 mM
sodium citrate buffer solution (pH 5.5), and incubated at 45 ◦ C under
continuous agitation for 5 h. Commercial cellulase (from Aspergillus
oryzae, Sigma-Aldrich, Madrid, Spain) was added at an enzyme-to-
microalgae ratio of 0.05 (50 g cellulase kg− 1 dry microalgae). The
estimation of the microalgae hydrolysis was carried out by monitoring
the amount of reducing sugars (expressed as g of glucose equivalents
100 g− 1 biomass, according to Miller, 1959) and total amino acids
(expressed as g L-leucine 100 g− 1 protein, according to Church et al.,
1983) released into the reaction vessel at different sampling times (0, 15,
30, 60, 90, 120, 180, 240, and 300 min). Total polyphenols (expressed as
mg gallic acid equivalents 100 g− 1 biomass, according to Singh et al.,
2002) were also measured in reaction vessels at the beginning and at the
end of the in vitro hydrolysis. A control assay was also carried out under
the same experimental conditions, without the addition of cellulase
enzyme.
Following the hydrolysis, the mixture was immediately used for
manufacturing aquafeeds.
2.2. Experimental diets
Four isonitrogenous and isolipidic experimental diets containing
Nannochloropsis gaditana biomass were elaborated at the CEIA3-Uni­
versidad de Almería facilities (Servicio de Piensos Experimentales, htt
p://www.ual.es/stecnicos_spe). Two inclusion levels (25 and 50 g
kg− 1 w/w), and two microalgae formats (raw and enzymatically
hydrolysed) were considered. Therefore, diets were designed as R25 and
R50 for raw microalgae lots, and H25 and H50 for enzymatically-
hydrolysed biomass. A microalgae-free diet was used as control (CT).
The formulation and proximal composition of the experimental diets is
shown in Table 1. In addition, fatty acid and amino acid profiles of each
diet are presented in Table 2 and Fig. 1, respectively. Feed ingredients
were finely ground and mixed in a vertical helix mixer (Sammic 13 M-
11, 5-L capacity, Sammic SA, Azpeitia, Spain) for 20 min. Then the
microalgae (crude or hydrolysed) were added at the specified inclusion
level, and water content was adjusted to provide 400 mL per kg of the
ingredient mixture, in order to obtain homogenous dough. The dough
was passed through a single screw laboratory extruder (Miltenz 51SP, JS
Conwell Ltd., New Zealand), provided with matrixes so as to obtain 2
and 3 mm pellets, according to the size of fish. The feeds were dried with
forced-air circulation (Airfrio, Almería, Spain) at 30 ◦ C for 24 h, and
then stored at − 20 ◦ C until use.
2.3. Fish maintenance and experimental design
The feeding trial was carried out at the aquaculture facilities (REGA:
ES300261040017) of Centro Oceanográfico de Murcia (Mazarrón,
south-eastern Spain), Instituto Español de Oceanografía (IEO-CSIC).
Gilthead seabream juveniles (15.06 ± 1.40 g average initial body
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Table 1 keeping seawater renewal rate (37‰ salinity) at 500 L h− 1 and ammonia
Ingredients and composition of the experimental diets. and nitrite values (<0.1 mg L− 1) suitable for gilthead seabream culture.
Diets Animals were kept under natural photoperiod (ranging from 10 to 13
daylight hours per day) and the water temperature was kept at 18 ±
Ingredients CT R25 R50 H25 H50
(% dry matter) 0.5 ◦ C. Light intensity ranged from 100 to 150 lx. Tanks were equipped
with aerators to maintain an adequate level of oxygenation (above 6 mg
Fish meal LT941 15.0 15.0 15.0 15.0 15.0
Raw N. gaditana2 – 2.5 5.0 – –
L− 1).
Hydrolysed N. gaditana – – – 2.5 5.0 After 45 and 90 days of the feeding trial, twenty fish per tank (60
Squid meal3 2.0 2.0 2.0 2.0 2.0 animals per dietary treatment) were withdrawn at each sampling time
CPSP904 1.0 1.0 1.0 1.0 1.0 and killed by anaesthetic overdose (200 mg L− 1 clove oil; isoeugenol)
Krill meal5 2.0 2.0 2.0 2.0 2.0
followed by spine severing. Body weight was recorded, and growing
Gluten meal6 15.0 15.0 15.0 15.0 15.0
Soybean protein parameters were calculated as follows: feed conversion rate (FCR): (total
40.0 38.8 37.3 38.8 37.3
concentrate7 feed being consumed/weight gain) and specific growth rate (SGR) (% d
Fish oil8 11.4 11.0 10.5 11.0 10.5 − 1): 100 × {(ln final weight – ln initial weight)/days}. Immediately
Soybean lecithin9 1.0 1.0 1.0 1.0 1.0 after slaughtering, instrumental colour parameters were determined on
Wheat meal10 5.4 4.5 4.0 4.5 4.0
Choline chloride11 0.5 0.5 0.5 0.5 0.5
the right side of the anterior dorsal skin of fish. Then, sampled fish were
Betain12 0.5 0.5 0.5 0.5 0.5 dissected, and the digestive tract and dorsal muscle were removed. The
Lysine13 1.5 1.5 1.5 1.5 1.5 liver and portions of muscle (5 g per fish) were stored at − 80 ◦ C for lipid
Methionine14 0.6 0.6 0.6 0.6 0.6 oxidation determinations (TBARS). The rest of individual muscle sam­
Vitamin and mineral
2.0 2.0 2.0 2.0 2.0 ples were freeze-dried and stored at − 20 ◦ C for further analysis of
premix15
Vitamin C16 0.1 0.1 0.1 0.1 0.1 proximate composition and fatty acids. Small segments (around 5 mm
Guar gum17 2.0 2.0 2.0 2.0 2.0 long) of the anterior intestine from five fish per tank were collected for
Crude protein
45.2 ± 46.1 ± 46.4 ± 45.4 ± 45.9 ± examination by optical, transmission (TEM) and scanning (SEM) elec­
0.2 1.1 0.1 0.5 1.0 tron microscopy. For digestive enzyme activity determinations, in­
15.2 ± 15.9 ± 15.5 ± 15.8 ± 15.8 ±
Crude lipid
0.1 0.3 0.1 0.7 0.1
testines from 15 fish per tank were randomly pooled (5 fish per pool, 3
11.8 ± 11.9 ± 12.2 ± 12.0 ± 11.7 ± pools per tank and sampling time) to obtain three enzymatic extracts
Ash
0.3 0.1 0.0 0.1 0.3 from each experimental tank (a total of 9 pooled extracts per dietary
Moisture
5.4 ± 5.6 ± 5.7 ± 5.3 ± 5.6 ± treatment and sampling time).
0.2 0.4 0.1 0.2 0.1

CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal 2.4. Proximate composition, fatty acid and amino acid analysis
biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydro­
lysed microalgae, respectively. Moisture (drying at 105 ◦ C in an oven, J.P. Selecta S.A., Barcelona,
1
69.4% crude protein, 12.3% crude lipid (Norsildemel, Bergen, Norway); Spain), crude protein (Kjeldahl Nx6.25, Foss KT200 Kjeltecdistiller,
2
Nannochloropsis gaditana (44.5% crude protein, 33.3% carbohydrates, 4.5%
FOSS, Denmark), and ash (mufla oven 1100 ◦ C, J.P. Selecta S.A., Bar­
ash, and 17.7% crude lipid);
3, 4,5 celona, Spain) were determined in feeds and muscle samples according
purchased from Bacarel (UK). CPSP90 is enzymatically pre-digested fish­
meal.
to AOAC (2000) procedures.
6
78% crude protein (Lorca Nutrición Animal SA, Murcia, Spain). Lipids were extracted from samples following Folch (1957) meth­
7
Soybean protein hydrolysate, 65% crude protein, 8% crude lipid (DSM, odology using chloroform/methanol (2:1 v/v) as solvent, and crude lipid
France). content was calculated gravimetrically. Fatty acid (FA) profiles of
8
AF117DHA (Afamsa, Spain). N. gaditana, feeds and muscle samples were determined by gas chro­
9
P700IP (Lecico, DE). matography (Hewlett Packard, 4890 Series II, Hewlett Packard Com­
10
Local provider (Almería, Spain). pany, Avondale, PA) following the method described in Rodríguez-Ruiz
11,12, 13,14
Lorca Nutrición Animal SA (Murcia, Spain). et al. (1998), using a modification of the direct transesterification
15
Lifebioencapsulation SL (Almería, Spain). Vitamins (mg kg− 1): vitamin A method described by Lepage and Roy (1984) that requires no prior
(retinyl acetate), 2000,000 UI; vitamin D3 (DL-cholecalciferol), 200,000 UI;
separation of the lipid fraction. Briefly, 150 mg of sample were mixed
vitamin E (Lutavit E50), 10,000 mg; vitamin K3 (menadione sodium bisulphite),
with 1 mL of n-hexane, and an internal standard solution (pentadecanoic
2500 mg; vitamin B1(thiamine hydrochloride), 3000 mg; vitamin B2 (ribo­
flavin), 3000 mg; calcium pantothenate, 10,000 mg; nicotinic acid, 20,000 mg; acid, C15:0) was added to each tube. Then, 1 mL of methylation mixture
vitamin B6 (pyridoxine hydrochloride), 2000 mg; vitamin B9 (folic acid), 1500 (methanol and acetyl chloride 20:1 v/v) was added to each tube and
mg; vitamin B12 (cyanocobalamin), 10 mg vitamin H (biotin), 300 mg; inositol, placed in a thermoblock (100 ◦ C, 30 min) with the purpose of obtaining
50,000 mg; betaine (Betafin S1), 50,000 mg. Minerals (mg kg− 1): Co (cobalt the corresponding fatty acid methyl esters (FAMEs). Once methylated, 1
carbonate), 65 mg; Cu (cupric sulphate), 900 mg; Fe (iron sulphate), 600 mg; I mL distilled water was added to each tube in order to remove the hexane
(potassium iodide), 50 mg; Mn (manganese oxide), 960 mg; Se (sodium sele­ phase, and tubes were centrifuged (3500 g, 5 min). The hexane phase
nite), 1 mg; Zn (zinc sulphate) 750 mg; Ca (calcium carbonate), 18.6%; that accumulated the FAMEs was removed from the tubes, and then
(186,000 mg); KCl, 2.41%; (24,100 mg); NaCl, 4.0% (40,000 mg). inserted in vials for fatty acid identification and quantification by gas
16
TECNOVIT, Spain.
17
chromatography.
EPSA, Spain.
Amino acid profiles of N. gaditana and feeds were determined by ion
exchange chromatography with post column derivatization with
weight) were selected and randomly distributed in 15 tanks (triplicate ninhydrin (Biochrom 30+ amino acid analyser, Biochrom LTD Cam­
tanks per dietary treatment) of 500 L capacity to reach an initial average bridge, UK) after hydrolysis of the samples (20 mg mL− 1 HCl 6 M,
biomass of 1200 g m− 3 (40 fish tank− 1). 110 ◦ C, 24 h, under N2 atmosphere), using norleucine as standard.
Fish were fed with CT diet (microalgae-free) during a 15-d acclima­
tion period prior to the beginning of the feeding trial. Afterwards, the 2.5. Digestive enzyme activities
experimental diets were offered thrice per day (9:00, 14:00 and 18:00)
at 2% of the biomass, until triplicating initial body weight. The amount For intestinal extracts, samples were homogenized in distilled water
of feed ingested was recorded daily in each tank. (0.5 g mL− 1) at 4 ◦ C. Supernatants were obtained after centrifugation
The 90-d feeding trial was carried out in an open flow circuit, (16,000 g, 12 min, 4 ◦ C) and stored in aliquots at − 20 ◦ C until further

3
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Table 2
Fatty acid profile of N. gaditana meal and experimental diets (% of total fatty acids).
Diets

Fatty acids N. gaditana CT R25 R50 H25 H50 P1

14:0 5.60 ± 0.01 3.11 ± 0.00 3.09 ± 0.03 3.13 ± 0.02 3.15 ± 0.04 3.17 ± 0.037 n.s.
16:0 22.4 ± 0.02 21.60 ± 0.10 21.51 ± 0.04 21.49 ± 0.18 22.03 ± 0.35 21.70 ± 0.07 n.s.
18:0 21.30 ± 0.02 5.95 ± 0.05 b 5.77 ± 0.01 ab 5.59 ± 0.09 a 5.89 ± 0.07 b 5.67 ± 0.03 a 0.007
16:1n7 4.21 ± 0.02 a 4.59 ± 0.03 b 5.06 ± 0.04 b 4.63 ± 0.03 c 5.06 ± 0.06 c < 0.001
18:1n7 2.45 ± 0.01 b 2.42 ± 0.00 ab 2.35 ± 0.02a 2.45 ± 0.02 b 2.36 ± 0.03 a 0.008
18:1n9 15.17 ± 0.12 b 14.91 ± 0.04 b 14.42 ± 0.14 a 14.95 ± 0.11 b 14.54 ± 0.09 a 0.004
20:1n9 4.0 ± 0.01 1.40 ± 0.02 1.42 ± 0.01 1.62 ± 0.36 1.42 ± 0.01 1.38 ± 0.06 n.s.
18:2n6 11.64 ± 0.11 b 11.54 ± 0.10 b 11.34 ± 0.04 ab 11.51 ± 0.13 ab 11.12 ± 0.10 a 0.020
18:3n3 3.7 ± 0.01 1.61 ± 0.07 1.50 ± 0.02 1.51 ± 0.11 1.44 ± 0.08 1.41 ± 0.06 n.s.
16:2n4 0.74 ± 0.02 0.56 ± 0.22 0.69 ± 0.00 0.69 ± 0.00 0.56 ± 0.19 n.s.
16:3n4 0.94 ± 0.01 0.83 ± 0.17 0.93 ± 0.01 0.97 ± 0.00 0.82 ± 0.18 n.s.
18:4n3 0.46 ± 0.02 a 0.41 ± 0.02 a 0.66 ± 0.06 b 0.71 ± 0.01 b 0.67 ± 0.00 b < 0.001
20:4n6 0.31 ± 0.01 0.27 ± 0.00 0.27 ± 0.00 0.31 ± 0.01 0.28 ± 0.04 n.s.
20:4n3 9.5 ± 0.02 1.41 ± 0.01 a 1.57 ± 0.00 b 1.68 ± 0.04 c 1.50 ± 0.01 b 1.80 ± 0.01 c < 0.001
20:5n3 (EPA) 33.4 ± 0.05 6.10 ± 0.02 a 6.52 ± 0.05 b 7.10 ± 0.08 c 6.47 ± 0.06 b 7.15 ± 0.02 c < 0.001
22:5n3 1.23 ± 0.04 1.26 ± 0.03 1.28 ± 0.06 1.26 ± 0.05 1.25 ± 0.08 n.s.
22:6n3 (DHA) 17.14 ± 0.27 b 16.31 ± 0.21 a 15.60 ± 0.18 a 16.10 ± 0.12 a 15.77 ± 0.11 a 0.003
Others 4.54 ± 0.56 5.45 ± 0.78 5.23 ± 0.50 4.53 ± 0.94 5.30 ± 0.82 n.s.
∑
SFA 30.65 ± 0.15 30.38 ± 0.05 30.21 ± 0.29 31.07 ± 0.47 30.53 ± 0.07 n.s.
∑
MUFA 23.23 ± 0.13 23.34 ± 0.02 23.45 ± 0.16 23.45 ± 0.15 23.34 ± 0.12 n.s.
∑
PUFA 38,14 ± 0.22 39.91 ± 0.29 39.44 ± 0.37 39.29 ± 0.13 39,45 ± 0.25 n.s.
∑
n-3 27.95 ± 0.15 27.56 ± 0.33 27.83 ± 0.31 27.48 ± 0.20 28.05 ± 0.27 n.s.
∑
n-6 11.95 ± 0.12 b 11.83 ± 0.10 ab 11.60 ± 0.04 ab 11.818 ± 0.14 ab 11.40 ± 0.14 a 0.025
n3/n6 2.34 ± 0.01 a 2.33 ± 0.01 a 2.40 ± 0.02 b 2.33 ± 0.01 a 2.46 ± 0.01 c < 0.001
EPA/DHA 0.36 ± 0.01 a 0.40 ± 0.00 b 0.46 ± 0.00 c 0.40 ± 0.00 b 0.45 ± 0.00 c < 0.001
1
The statistical comparison was carried out among the experimental diets, excluding N. gaditana meal. Therefore, P-values illustrate the statistical significance of
differences among CT, R25, R50, H25 and H50 diets. CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and
H50: diets including 25 and 50 g kg− 1 hydrolysed microalgae, respectively Different lower-case superscripts indicate significant differences among diets within each
row (P < 0.05). Values (n = 3) are mean ± standard deviation. n.s.: not significant.

Fig. 1. Amino acid profile of N. gaditana meal and the experimental diets. A: essential amino acids; B: non-essential amino acids. Results (n = 3) are expressed as % of
total amino acids.

use. Total soluble protein was determined according to Bradford (1976)
using bovine serum albumin as standard.
Total alkaline protease activity in intestinal extracts was measured
spectrophotometrically following the procedure described by Alarcón
et al. (1998), using 5 g L− 1 casein in 50 mM Tris-HCl (pH 9.0) as sub­
strate. One unit of total protease activity was defined as the amount of
enzyme that released 1 μg of tyrosine per min in the reaction mixture,
considering an extinction coefficient of 0.008 μg− 1 mL− 1 cm− 1 for
tyrosine, measured at 280 nm wavelength. Trypsin and chymotrypsin
activities were assayed using 0.5 mM BAPNA (N-α-benzoyl-DL-arginine-
4-nitroanilide) as substrate according to Erlanger et al. (1961), and 0.2
mM SAPNA (N-succinyl-(Ala)2-Pro-Phe-p-nitroanilide) according to Del
Mar et al. (1979), respectively, in 50 mM Tris-HCl, 10 mM CaCl2 buffer,
pH 8.5. Leucine aminopeptidase activity was determined by using 2 mM

4
M.I. Sáez et al.
L-leucine-p-nitroanilide in 100 mM Tris-HCl buffer, pH 8.8, as substrate
(Pfleiderer, 1970), and alkaline phosphatase was assayed using 450 mM
p-nitrophenyl phosphate in 1 M diethanolamine, 1 mM MgCl2 buffer, pH
9.5 (Bergmeyer, 1974) as substrate. For trypsin, chymotrypsin, and
leucine aminopeptidase activities, one enzyme activity unit (U) wasde­
fined as the amount of enzyme releasing 1 μmol of p-nitroanilide (pNA)
per minute, considering as extinction coefficient 8800 M cm− 1,
measured spectrophotometrically at 405 nm. For alkaline phosphatase,
one activity unit was defined as the amount of enzyme that released 1 μg
of nitrophenyl per min considering an extinction coefficient of 17,800 M
cm− 1 for p-nitrophenol, measured also at 405 nm. All assays were per­
formed in triplicate, and specific enzymatic activity was expressed as
units (U) g tissue− 1.
2.6. Histology of the intestinal mucosa
Intestine samples were fixed in phosphate-buffered formalin (4% v/
v, pH 7.2) for 24 h, then dehydrated and embedded in paraffin according
to standard histological techniques, as described in Vizcaíno et al.
(2018). Samples were cut in transversal sections (5 μm), and the slides
were stained with haematoxylin-eosin (H&E). The preparations were
examined under light microscope (Olympus ix51, Olympus, Barcelona,
Spain) equipped with a digital camera (CC12, Olympus Soft Imaging
Solutions GmbH, Muenster, Germany). Images were analysed with
specific software (Image J, National Institutes of Health, USA). The
length of mucosal folds and total enterocyte height were determined in
intestinal samples (10 independent measurements performed at 4
different optical areas of each section from 5 fish per tank; 2 sections per
fish).
2.7. Ultrastructure of the intestinal mucosa
Intestine samples for scanning electron microscopy (SEM) were
washed with 1% S-carboxymethyl-L-cysteine (Sigma Chem.) for 20 s,
with the aim of removing the epithelial mucus, prior to fixation. Then,
specimens were fixed in phosphate-buffered formaldehyde (4% v/v, pH
7.2) for 24 h; next excess glutaraldehyde was removed by washing
samples in 0.1 mol L− 1cacodylate buffer, pH 7.2, and then dehydrated
with a series of increasing concentrations of ethanol (50% to 100% v/v).
Samples were critical point dried in absolute ethanol as intermediate
fluid, and CO2 as transition fluid (CDP 030 Critical point dryer, Leica
Microsystems, Madrid, Spain). After drying, specimens were mounted
on aluminium stubs, immobilized with graphite (PELCO® Colloidal
Graphite, Ted Pella INC., Ca, USA), and then gold sputter coated (SCD
005 Sputter Coater, Leica Microsystems). Finally, all samples were
screened with a scanning electron microscope (HITACHI S-3500, Hitachi
High-Technologies Corporation, Japan).
Samples for transmission electron microscopy (TEM) were fixed (4 h,
4 ◦ C) in 25 g L− 1 glutaraldehyde, 40 g L− 1 formaldehyde in phosphate
buffer saline (PBS), pH 7.5. Next, intestine sections were washed with
PBS for 20 min and then, a post-fixation step with 20 g L− 1 osmium
tetroxide was carried out. Samples were dehydrated by consecutive
immersion in increasing concentrations of ethanol, embedded for two
hours, in 1:1 mixture of Epon resin and 100% (v/v) ethanol under
continuous shaking, and then, included in pure Epon resin, and let
polymerize at 60 ◦ C. Finally, ultrathin cuts were obtained from resin
blocks, and placed on a 700 Å copper mesh and stained with uranyl
acetate and lead citrate. The mesh observation was performed with a
Zeiss 10C TEM at 100 Kv (Carl Zeiss, Barcelona, Spain).
SEM and TEM visualization fields were recorded and digital images
were analysed using UTHSCSA ImageTool software (University of Texas
Health Science Center, San Antonio, TX). Microvilli length (ML) and
microvilli diameter (MD) were determined in TEM micrographs ac­
cording to (Vizcaíno et al., 2014). SEM images were used to obtain
measurements of enterocyte apical area (EAA). Finally, data obtained
from TEM and SEM images were used to estimate the total absorption
5
Aquaculture 556 (2022) 738288
surface per enterocyte (TAS) according to Vizcaíno et al. (2014).
2.8. Lipid oxidation
Lipid oxidation was estimated by thiobarbituric acid-reactive sub­
stances (TBARS) analysis in muscle and liver according to the method of
Buege and Aust (1978). Briefly, samples (2 g each) were homogenized in
4 mL 50 mM NaH2PO4, 0.1% (v/v) Triton X-100 solution. The mixture
was centrifuged (10,000 g, 20 min, 4 ◦ C) and supernatants were mixed in
a ratio 1:5 (v/v) with 2-thiobarbituric acid (TBA) reagent (0.375% w/v
TBA, 15% w/v TCA, 0.01% w/v 2,6-dibutyl hydroxytoluene (BHT) and
0.25 N HCl). The mixture was heated for 15 min and then centrifuged
(3600 g, 10 min, 4 ◦ C), and the absorbance of supernatants was
measured at 535 nm. The amount of TBARS was expressed as mg of
malonyl dialdehyde (MDA) per kg of muscle after comparing with a
MDA standard.
2.9. Instrumental colour determination
For all fish samples colour was measured on skin dorsal portion by
L*, a*, and b* system (CIE, 1986), using a Minolta Chroma meter CR400
device (Minolta, Osaka, Japan). The parameter lightness (L*, on a 100-
point scale from black to white), redness (a*, assesses the position be­
tween red, positive values, and green, negative values), and yellowness
(b*, assesses the position between yellow, positive values, and blue,
negative values) were recorded.
2.10. Statistics
The effect of the categorical variables “pre-treatment” and “doses”,
as well as their interactions, were determined for each numeric
parameter studied by fitting a generalized lineal multifactorial statistical
model (GLM analysis) that relates measured parameters to predictive
factors, using specific software (SPSS 25, IBM Corporation Inc.). Least
square means were tested for differences using Fisher’s least significant
difference (LSD) procedure. Unless otherwise is specified, a significance
level of 95% was considered to indicate statistical differences (P < 0.05).
When measurements were expressed as a percentage (e.g., fatty acids
profile), arcsine transformation of their square root was carried out in
order to normalize data prior to the statistical analysis.
3. Results
3.1. Microalgae hydrolysis
The concentration of reducing sugars in the reaction vessels
increased throughout the in vitro assay owing to the addition of the
commercial cellulase enzyme (Fig. 2A), being differences significant (P
< 0.001) at each sampling time throughout the hydrolysis. Results
indicate that enzyme-treated (5% cellulase) microalgal biomass yielded
final values in the region of 8 g glucose equivalents (GE) per 100 g
microalgae dry mass. This value was about 4-fold the amount of
reducing sugars released from untreated raw algae (control), which
accounted for stable values about 2–2.5 g GE throughout the complete
assay (300 min).
Analogously, the total amount of amino acids released (Fig. 2B)
during the assay indicated that cellulase hydrolysis increased signifi­
cantly (P < 0.001) their concentration in the reaction vessels, compared
to raw biomass. Under the assay conditions, final concentration of free
amino acids in enzyme-treated batches reached 12 g 100 g protein− 1,
compared to 6–7 g free amino acids 100 g protein− 1 measured in con­
trols (Fig. 3).
The quantification of total polyphenols at the beginning and at the
end of the enzymatic assay revealed a 70% increase of these substances
when the algal biomass was enzymatically treated (Fig. 3). The final
concentration of total phenolics in supernatants was significantly higher
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Fig. 2. Time-course of the concentration in reducing sugars (A, expressed as D-glucose equivalents, GE, 100 g dry biomass-1) and total free amino acids (B, expressed
as g L-leucine, L-Leu, 100 g protein− 1) measured from raw and cellulose-hydrolysed biomass of N. gaditana during the in vitro assay. Within each sampling time,
asterisks indicate significant differences between values (P < 0.05). Values (n = 6) are mean ± standard deviation.

Table 3
Fish biometric parameters and muscle proximate composition at the end of the
feeding trial (90 d).
Diets

CT R25 R50 H25 H50 P

Final BW 49.10 ± 51.30 ± 49.50 ± 49.90 ± 50.2 ± n.

(g) 5.69 4.75 5.69 5.69 5.69 s.
1.07 ± 1.04 ± 1.01 ± 1.01 ± 1.01 ± n.
FCR 0.09 0.09 0.19 0.09 0.19 s
SGR (% 1.47 ± 1.49 ± 1.54 ± 1.54 ± 1.52 ± n.
d− 1) 0.09 0.09 0.09 0.09 0.09 s
Crude
protein 17.68 ± 18.36 ± 18.29 ± 17.99 ± 17.86 ± n.
(% DM) 0.28 0.34 0.33 0.48 0.27 s.
Crude lipid 7.33 ± 7.05 ± 7.19 ± 7.21 ± 7.18 ± n.
(% DM) 0.87 0.77 0.75 0.81 0.74 s.
Ash (% 1.81 ± 1.76 ± 1.75 ± 1.75 ± 1.77 ± n.
DM) 0.06 0.04 0.05 0.06 0.11 s.
Moisture 72.59 ± 72.61 ± 71.91 ± 71.9 ± 72.18 ± n.
(%) 0.64 0.66 0.79 0.73 0.49 s.

Fig. 3. Total phenolics released from raw and cellulase-hydrolysed N. gaditana
biomass at the beginning and at the end of the in vitro hydrolysis. Results are
represented as mg gallic acid equivalents (GAE) 100 g microalgal dry bio­
mass− 1. Within each sampling time, asterisks indicate significant differences
between values (P < 0.05). Data (n = 6) are mean ± standard deviation.
(P < 0.01) in cellulase-treated N. gaditana (reaching 70 mg gallic acid
equivalents (GAE) 100 g dry microalgae− 1), than that measured in
controls (40 mg GAE 100 g− 1). In absence of cellulase, no significant
differences were observed in total phenolics measured in the reaction
vessel after the 5-h (300 min) incubation period.
3.2. Fish biometric parameters, muscle proximate composition and fatty
acid profile
Experimental diets were well accepted by the fish, and feed intake
was similar in all groups. Mortality was below 1%. During the experi­
mental period, no differences were observed regarding growth param­
eters (final BW, FCR and SGR) among experimental lots at the end of the
feeding trial (Table 3). Throughout this period, final body weight
(approx. 50 g) at least triplicated initial values (approx. 15 g). No
BW: final body weight. FCR: feed conversion rate. SGR: specific growth rate. CT:
control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal
biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydro­
lysed microalgae, respectively. Different lower-case superscripts indicate sig­
nificant differences among diets within each row (P < 0.05). Values are mean ±
standard deviation. For proximate composition n = 15. For biometric parame­
ters n = 90. n.s.: not significant.
significant differences (P > 0.05) were observed for any of the param­
eters of muscle proximate composition. Although not significantly,
microalgae-enriched diets tended to decrease slightly total muscle lipid
content compared to CT diet, no matter the microalgae concentration or
treatment considered.
Overall results on muscle fatty acid profile indicated that the inclu­
sion of raw or hydrolysed microalgae yielded significant changes in FA
profile (Table 4), especially with regard to MUFAs and PUFAs, which
displayed opposing tendencies. Thus, microalgae-enriched diets reduced
total MUFAs compared to CT, being this decrease more evident in diets
including raw biomass (R25 vs. H25, and R50 vs. H50; P < 0.05).
Regarding individual MUFAs, oleic acid (18:1n9) was the predominant
FA, and its tendency paralleled that of total MUFA values. On the other

6
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Table 4 compared to controls (Table 5), with the exception of leucine amino­
Effects of the dietary inclusion of N. gaditana on fatty acid profile of gilthead peptidase (LAP) activity. Also considering all the activities as a whole,
seabream muscle after a 90-d feeding trial (% of total fatty acids). significant differences were attributable to sampling time, as values
Diets measured at the end of the assay (90 d) where significantly higher (P <
Fatty CT R25 R50 H25 H50 P
0.05) than those measured at day 45, irrespectively of the dietary
acids treatment, with some exceptions for LAP activity again.
After 45 days of feeding, significant differences (P < 0.05) were
2.00 ± 1.51 ± 1.47 ± 1.78 ± 1.86 ±
14:0
0.01d 0.00a 0.02a 0.04b 0.02c
0.020 observed in trypsin activity attributable to both biomass pre-treatment
16.54 ± 16.57 ± 16.83 ± 16.68 ± 16.96 ± and inclusion level, whereas only microalgae hydrolysis influenced
16:0 0.015
0.14a 0.15a 0.15ab 0.12a 0.26b total alkaline protease and alkaline phosphatase activities (P < 0.05). No
4.70 ± 5.51 ± 5.52 ± 5.10 ± 5.10 ± changes due to these factors were observed for leucine aminopeptidase.
18:0 0.022
0.04a 0.00c 0.04c 0.05b 0.08b
4.69 ± 3.58 ± 3.82 ± 4.63 ± 4.40 ±
At the end of the feeding trial, the influence of both variables was sig­
16:1n7 0.033 nificant on total alkaline protease, trypsin and chymotrypsin and ac­
0.02c 0.02a 0.05b 0.05c 0.02c
2.69 ± 2.53 ± 2.53 ± 2.69 ± 2.44 ± tivities, but only a hydrolysis-dependent effect was observed for and
18:1n7 0.036
0.02b 0.12ab 0.01b 0.11b 0.09a leucine aminopeptidase activity.
19.64 ± 14.76 ± 15.76 ± 17.68 ± 18.35 ±
18:1n9 0.024
0.18e 0.31a 0.14b 0.07c 0.09d
1.49 ± 1.32 ± 1.28 ± 1.36 ± 1.48 ± 3.4. Intestinal mucosa histology
20:1n9 0.006
0.05b 0.14ab 0.09a 0.09ab 0.05b

18:2n6
8.37 ± 8.06 ± 8.14 ± 8.50 ± 8.52 ±
0.014 The histological characteristics of intestinal sections obtained from
0.08b 0.03a 0.08a 0.16b 0.07b fish receiving the experimental dietary treatments at the end of the
0.97 ± 0.86 ± 0.86 ± 1.03 ± 1.05 ±
18:3n3
0.01b 0.02a 0.02a 0.02b 0.08b
0,014 feeding trial are shown in Fig. 4, and results of the measurements carried
0.59 ± 0.49 ± 0.51 ± 0.55 ± 0.56 ± out on haematoxylin-eosin stained sections are summarized in Table 6.
16:4n3 0.019
0.01c 0.02a 0.06a 0.01b 0.01b Neither evidence of lipid droplet accumulation in the intestinal epithe­
0.60 ± 0.43 ± 0.44 ± 0.55 ± 0.54 ± lium nor inflammatory changes in the lamina propria were observed.
18:4n3 0.019
0.01c 0.01a 0.02a 0.05bc 0.01b
Consequently, no apparent damage attributable to any of the dietary
1.72 ± 2.20 ± 2.37 ± 1.78 ± 1.88 ±
20:4n6
0.02a 0.01c 0.03d 0.06ab 0.11b
0.033 treatments was found. Enterocytes presented aligned nucleus, homog­
0.52 ± 0.53 ± 0.50 ± 0.51 ± 0.53 ± enous supranuclear vacuolization and adequate cell shape (columnar
20:4n3 n.s.
0.00 0.01 0.01 0.02 0.01 and high). Intercellular spaces were not visible between enterocytes, and
20:5n3, 4.76 ± 5.63 ± 5.72 ± 5.23 ± 5.25 ± goblet cells were evenly dispersed throughout the epithelium.
0,014
EPA 0.03a 0.04c 0.09c 0.11b 0.04b
2.22 ± 2.12 ± 2.12 ± 2.14 ± 2.19 ±
Image analysis of the preparations indicated that no significant dif­
22:5n3 0.031 ferences in fold length or enterocyte height were found among the di­
0.02c 0.01a 0.09a 0.05ab 0.03ab
22:6n3, 22.84 ± 23.75 ± 23.96 ± 23.42 ± 23.56 ± etary treatments. However, differences attributable to the inclusion
0.001
DHA 0.08a 0.07c 0.16c 0.10b 0.26bc level were observed, as the animals receiving R50 and H50 diets
5.17 ± 9.52 ± 7.76 ± 5.61 ± 4.70 ± <
Other FA showing a significantly thinner lamina propria than the rest of the
0.38a 0.51d 0.36c 0.25 ab 0.40b 0.001
∑ 23.24 ± 23.58 ± 23.67 ± 23.70 ± 23.92 ± experimental batches, irrespectively of the enzyme pre-treatment.
SFA n.s
0.18 0.16 0.16 0.23 0.37
∑
MUFA
28.50 ± 22.19 ± 23.39 ± 26.35 ± 26.67 ±
0.032 3.5. Ultrastructure of the intestinal mucosa
0.26d 0.31a 0.06b 0.17c 0.16c
∑ 41.82 ± 43.62 ± 44.10 ± 43.16 ± 43.53 ± <
PUFA
0.17a 0.02c 0.12d 0.16b 0.17bc 0.001
TEM and SEM observations indicated that none of the experimental
∑ 31.73 ± 33.33 ± 33.59 ± 32.87 ± 33.13 ± < diets caused any perceptible damage on the enterocyte brush border
n-3
0.19a 0.03b 0.12bc 0.03c 0.20b 0.001 ultrastructure (Fig. 5), since specimens from animals from all the
∑ 10.09 ± 10.28 ± 10.50 ± 10.29 ± 10.40 ± experimental groups presented a well-defined and organized intestinal
n-6 0.026
0.08a 0.09b 0.10b 0.13b 0.09b
brush border membrane. Moreover, no intercellular spaces were visible
3.15 ± 3.24 ± 3.20 ± 3.19 ± 3.19 ± <
n3/n6
0.02a 0.01b 0.01b 0.02b 0.03ab 0.001 in the apical zone of the epithelium. Image analysis (Table 7) showed
EPA/ 0.21 ± 0.24 ± 0.24 ± 0.22 ± 0.22 ± < that only 5% inclusion level caused changes in all the parameters
DHA 0.00a 0.00c 0.01c 0.01 b 0.00b 0.001 studied. R50 treatment yielded higher microvilli diameter (MD) than the
CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal rest of treatments, whereas only H50 caused significant increase in
biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydro­ microvilli length (ML). Both R50 and H50 treatments increased signifi­
lysed microalgae, respectively. Values with different lowercase superscript cantly enterocyte apical area (EAA). The factor that accounted for most
within each row indicate significant differences in muscle lipids attributed to of the changes observed in total enterocyte absorption surface (TAS) was
dietary treatments (p < 0.05). SFA: saturated fatty acids; MUFA: mono­ the biomass pre-treatment, given that both H25 and H50 batches
unsaturated fatty acids; PUFA: polyunsaturated fatty acids; EPA: eicosapentae­ showed significantly higher values for this parameter compared to CT
noic acid; DHA: docosahexaenoic acid. Values (n = 15) are expressed as average and R25.
± standard deviation. n.s.: not significant.
3.6. Muscle and liver lipid oxidation (TBARS)
hand, increased total PUFAs in muscle was observed in fish fed on
microalgae-containing diets compared to CT lot. As indicated for total The overall tendency for TBARS values (Table 8) indicates that CT
MUFAs, although with opposite trend, the hydrolyzed biomass was batch yielded the highest values for this parameter, irrespectively of the
responsible for significantly higher values (P < 0.05) of muscle PUFAs tissue, sampling time, and dietary treatment considered (P < 0.001).
than raw biomass within each inclusion level. It is worth mentioning Nevertheless, not all factors were responsible for significant differences
that both EPA and DHA paralleled such increase. in all cases.
With regard to muscle, R50, H25 and H50 treatments decreased
3.3. Digestive enzyme activities significantly lipid oxidation compared to CT batch at both sampling
times (45 and 90 d). Within each dietary treatment, only H50 showed
In general, the supplementation with N. gaditana increased signifi­ differences attributable to sampling time.
cantly (P < 0.05) the enzyme activities measured in intestinal extracts The effects of the microalgae-enriched diets were even more evident

7
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Table 5
Enzyme activities (U g tissue− 1) measured in intestinal extracts of Sparus aurata juveniles fed with the experimental diets.
Diets

Sampling time Enzyme activity CT R25 R50 H25 H50 P

Total alkaline protease 491.55 ± 50.47 a,I 567.54 ± 23.68 b,I 589.56 ± 24.92 b,I 553.14 ± 37.69 b,I 560.90 ± 28.50 b,I <0.001
45 d Trypsin (x10− 3) 16.60 ± 2.30 a,I 33.02 ± 1.76 c,I 35.77 ± 2.22 d,I 28.71 ± 1.58 b,I 28.90 ± 1.26 b,I <0.001
Chymotrypsin 2.49 ± 0.17 a,I 2.45 ± 0.13 a,I 3.05 ± 0.32 b,I 2.34 ± 0.20 a,I 2.32 ± 0.13 a,I <0.001
LAP (x10− 3) 0.43 ± 0.04 I 0.41 ± 0.04 I 0.41 ± 0.04 0.44 ± 0.04 0.42 ± 0.04 n.s.
Alkaline phosphatase 11.0 ± 0.36 a,II 10.44 ± 0.51 a,I 10.78 ± 0.40 a,I 13.25 ± 0.40 b,I 13.38 ± 0.27 b,I <0.001
Total alkaline protease 676.46 ± 38.27 a,II 909.67 ± 57.50 b,II 1015.58 ± 22.37 c,II 849.97 ± 49.89 b,II 862.73 ± 29.72 b,II <0.001
90 d Trypsin (x10− 3) 22.85 ± 2.03 a,II 46.57 ± 1.75 c,II 53.31 ± 2.57 d,II 36.78 ± 2.44 b,II 34.46 ± 2.30 b,II <0.001
Chymotrypsin 2.10 ± 0.10 a,II 3.21 ± 0.15 c,II 3.79 ± 0.11 d,II 2.81 ± 0.27 b,II 2.80 ± 0.11 b,II <0.001
LAP (x10− 3) 0.37 ± 0.04 a,I 0.52 ± 0.05 c,II 0.41 ± 0.05 ab 0.51 ± 0.04 c 0.43 ± 0.07 b <0.001
Alkaline phosphatase 14.60 ± 0.81 a,II 15.22 ± 0.97 a,II 16.59 ± 0.49 ab,II 17.25 ± 0.92 b,II 15.09 ± 0.78 a,II <0.001

CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydrolysed
microalgae, respectively. LAP: Leucine aminopeptidase. Values (n = 9) are mean ± standard deviation. Values in the same row with different lowercase superscript
indicate significant differences owing to dietary treatments (P < 0.05). Values of each enzyme activity with different superscript in Roman numerals indicate sig­
nificant differences due to sampling time (P < 0.05). n.s.: not significant.

Fig. 4. Light microscopy details of intestine sections of S. aurata juveniles fed on the experimental diets for 90 days. H&E stain, magnification x100 (upper images)
and ×400 (lower images). CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and H50: diets including 25 and
50 g kg− 1hydrolysed microalgae, respectively.

Table 6
Measurements in histological preparations of the intestinal mucosa of S. aurata juveniles fed with the experimental diets during 90 days.
Parameters Diets

CT R25 R50 H25 H50 P

Fold length (μm) 940.82 ± 216.24 1173.12 ± 414.47 1193.48 ± 269.95 1053.49 ± 205.41 1013.00 ± 239.95 n.s.
Enterocyte height (μm) 46.94 ± 7.47 57.21 ± 8.26 46.95 ± 3.95 45.61 ± 4.94 50.01 ± 8.33 n.s.
Lamina propria (μm) 13.76 ± 3.62b 13.06 ± 2.50b 6.35 ± 1.40a 11.23 ± 1.95b 7.63 ± 2.39a < 0.001

Values in the same row with different lowercase indicate significant differences (P < 0.05) owing to dietary treatments. CT: control diet. R25 and R50: diets including
25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydrolysed microalgae, respectively. Values are expressed as
mean ± standard deviation. n.s.: not significant.

in liver, given that all treatments, irrespectively of biomass hydrolysis or
inclusion level, yielded lower TBARS values than CT group. In this tis­
sue, significant differences attributable to biomass hydrolysis (lower
values for H25 and H50 in comparison with R25 and R50, respectively)
and inclusion level (lower values for R25 and H25 compared to R50 and
H50, respectively) were also observed. Same as found in muscle, also in
liver only H50 showed differences attributable to sampling time.
N. gaditana were significantly lower thanthose of CT group, indicating a
more greenish colorationat both 45 and 90 d. Nevertheless, no differ­
ences attributable to biomass pre-treatment (R25 vs. H25; R50 vs. H50)
or inclusion level (2.5% vs. 5.0%) were observed (P > 0.05). Similarly,
all treatments including algae biomass tended to increase b* parameter
(more yellowish pigmentation of the skin) compared to CT batch,
although differences were significant only for R50 treatment at day 45,
and for R25, R50 and H25 specimens at day 90.

3.7. Instrumental colour determinations

4. Discussion

After 45 days of the feeding trial, skin L* values were similar in all
Given that cellulose accounts for the most abundant structural car­
lots (Table 9). The same lack of significant differences was found at the
bohydrate in N. gaditana, the breakage of cell walls by hydrolysis with
end of the experimental period for this parameter, although L* values
cellulase enzymes emerges as a promising alternative aimed at
were influenced for the factor sampling time (P < 0.05). Skin a*
improving nutrient bioavailability and digestibility. The following may
parameter presented negative values in all specimens, irrespectively of
be cited as advantages of this procedure: i) owing to the wide variety of
the sampling time. Figures for fish fed with any diet supplemented with

8
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Fig. 5. Transmission (upper images, A) and scanning (lower images, B) electron microscopy micrographs from the anterior intestinal region of juvenile gilthead
seabream at the end of the feeding trial (90 days). CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and
H50: diets including 25 and 50 g kg− 1 hydrolysed microalgae, respectively.

Table 7 Table 8
Microvilli morphological parameters obtained from transmission electron mi­ Estimation of lipid oxidation (TBARS) in muscle and liver of juvenile fish fed on
croscopy ultramicrographs of the anterior intestine of S. aurata juveniles fed the different experimental diets.
with the experimental diets during 90 days. Diets
Parameters Diets
Time CT R25 R50 H25 H50 P
CT R25 R50 H25 H50 P
2.17 1.42
1.68 ± 1.31 ± 1.04 ± <
2.07 ± 2.16 ± 2.59 ± 2.28 ± 2.73 ± 45 d ± ±
ML (μm) 0.008 0.14bc 0.14ab 0.13a,I 0.001
0.59a 0.15a 0.20ab 0.13ab 0.14b 0.37c 0.12b
Muscle
0.11 ± 0.10 ± 0.13 ± 0.10 ± 0.11 ± 2.01 1.15
MD (μm) 0.026 1.71 ± 1.26 ± 0.92 ± <
0.01a 0.01a 0.01b 0.01a 0.01a 90 d ± ±
0.05b 0.20a 0.09a,II 0.001
19.36 20.75 26.14 20.45 25.92 0.39b 0.23a
EAA (μm2) <0.001
± 3.57a ± 4.20a ± 4.03b ± 4.02a ± 4.03b 4.62 3.19
3.86 ± 2.77 ± 2.52 ± <
587.63 620.47 692.94 772.71 917.96 45 d ± ±
0.39c 0.19a 0.18a,I 0.001
TAS (μm ) 2
± ± ± ± ± 0.006 0.43d 0.17b
Liver
88.60a 55.46a 81.93ab 38.01b 56.62b 4.64 3.03
3.73 ± 2.67 ± 2.15 ± <
90 d ± ±
ML: microvilli length; MD: microvilli diameter; EAA: enterocyte apical area; 0.26d 0.20b 0.07a,II 0.001
0.19e 0.09c
TAS: total enterocyte absorption surface. CT: control diet. R25 and R50: diets
including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25 and H50: CT: control diet. R25 and R50: diets including 25 and 50 g kg− 1 raw microalgal
diets including 25 and 50 g kg− 1 hydrolysed microalgae, respectively. Values in biomass, respectively. H25 and H50: diets including 25 and 50 g kg− 1 hydro­
the same row with different lowercase superscripts indicate significant differ­ lysed microalgae, respectively. TBARS stands for thiobarbituric acid reactive
ences (P < 0.05) owing to the dietary treatment. Values (n = 15) are expressed as substances, expressed as mg malonyldialdehyde (MDA) kg tissue− 1. Values in
mean ± standard deviation. the same row with different lowercase superscripts indicate significant differ­
ences (P < 0.05) owing to dietary treatments. Within each tissue, values with
different superscript in Roman numerals indicate significant differences due to
industrial applications, cellulases are reasonably inexpensive; ii) no sampling time (P < 0.05). Values (n = 15) are mean ± standard deviation.
complex equipment is needed for hydrolysis, as incubators widely uti­
lized in the food and feed industry can be used; iii) negative impacts on
line with previous studies on this microalgae genus carried out on
thermolabile compounds are not expected, as mild temperatures are
several species of commercial fish (Qiao et al., 2019; Sørensen et al.,
involved in the hydrolysis process; and iv) given the specificity of the
2017; Walker and Berlinsky, 2011). Nevertheless, other reports pointed
catalytic action on cellulose, the remaining released compounds would
to improved fish growth owing to microalgae inclusion in diets, but
not be hydrolysed by the enzymatic pre-treatment.
higher inclusion levels were evaluated (up to 10% inclusion in diets for
Under this perspective, this study evaluatedin a 90-d feeding trial the
Nile tilapia, Abdel-Tawwab and Ahmad, 2009; from 10 to 39% for
effects of a cellulase pre-treatment on Nannochloropsis gaditana biomass
S. aurata, in Vizcaíno et al., 2014, 2016, 2018). It is likely that the low
prior to its incorporation into feeds for gilthead seabream juveniles. It
inclusion levels considered have accounted for this lack of effect on
was expected that nutrient bioavailability would increase as a result of
growth.
the enzyme pre-treatment, and according to the results of the enzyme
With regard to fish muscle proximal composition, overall, no dif­
hydrolysis (Figs. 2 and 3), the increased release of reducing sugars, free
ferences attributable to the experimental diets were observed, in
amino acids and polyphenols taking place in the reaction vessels seem to
agreement with previous studies on N. gaditana enriched diets (Qiao
confirm this hypothesis.
et al., 2019; Vizcaíno et al., 2018; Sales et al., 2021). Only slight, but not
Even if the enzyme treatment increased in vitro bioavailability,
significant differences in lipid and protein contents were measured
however, no impact on fish growth was observed for any of the exper­
among the experimental groups (Table 3). The study by Galafat et al.
imental batches throughout the 90-d feeding trial. These results are in

9
M.I. Sáez et al. Aquaculture 556 (2022) 738288

Table 9 the contrary, control diet showed the highest DHA figures, but yielded
Instrumental colour determinations on the skin surface of juvenile fish fed with the lowest content in fish muscle. In this regard, previous studies have
the different experimental diets. also reported certain selective retention of this structural FA owing to
Diets the addition of both macro and microalgae (Hussein et al., 2013; Viz­
Time Colour CT R25 R50 H25 H50 P
caíno et al., 2014, 2016; Kousoulaki et al., 2016; Sáez et al., 2020). In
parameters short, the results suggested that N. gaditana, even at the low inclusion
levels studied, was responsible for some selective retention of n-3-PUFA
73.74 74.98 73.36 72.40 72.13
L* ± 4.22 ± 3.24 ± 3.44 ± 5.19 ± 5.33 n.s. in muscle (mostly owing to DHA, the most abundant n3-PUFA), whilst
I I I I I the opposite effect was observed with regard to MUFAs (Table 4). Such
- 1.00 - 1.88 - 1.85 - 1.89 - 1.80 decrease in total MUFAs is in agreement with the evidenced lower oleic
45 d a* ± ± ± ± ± 0.047 acid content, which is the main MUFA in muscle of gilthead seabream.
0.39a 0.44b 0.40b 0.39b 0.32b
5.25 7.79 8.54 7.09 7.88
Nevertheless, disparate results have been reported in the literature
b* ± ± ± ± ± 0.033 regarding the effects of dietary microalgae on S. aurata lipid metabolism
1.40a 1.81ab 1.04b 1.09ab 1.36ab (Perera et al., 2020), and further studies aimed at fully understanding
89.75 87.91 87.91 88.07 87.37 the intrinsic mechanisms underlying the results observed are needed.
L* ± 1.41 ± 3.24 ± 3.25 ± 2.06 ± 1.45 n.s.
II II II II II On the other hand, microalgae are acknowledged as valuable source
- 1.48 - 2.01 - 2.16 - 1.89 - 1.94 of pigments and phenolic compounds with antioxidant capacity
90 d a* ± ± ± ± ± 0.021 (Koyande et al., 2019; Almendinger et al., 2021; Sáez et al., 2021), many
0.16a 0.38b 0.11b 0.12b 0.08b of which remain unidentified (Sansone et al., 2020).
5.53 8.16 8.28 8.02 7.25 Due to the interest of the pharmaceutical industry in pigments, these
<
b*
substances have received more attention than phenolics, but some au­
± ± ± ± ±
0.001
1.13a 0.46b 0.46b 0.92b 1.07ab
thors suggested that both groups of substances might contribute simi­
Values in the same row with different lowercase superscripts indicate significant larly to the antioxidant activity of microalgae (Almendinger et al.,
differences (P < 0.05) owing to dietary treatments. CT: control diet. R25 and
2021). Nevertheless, the relative contribution of phenolics and pigments
R50: diets including 25 and 50 g kg− 1 raw microalgal biomass, respectively. H25
to the antioxidant capacity of most microalgae species remains to be
and H50: diets including 25 and 50 g kg− 1 hydrolysed microalgae, respectively.
Values (n = 30) are mean ± standard deviation. Parameters L*, a* and b* as
ascertained (Goiris et al., 2012). N. gaditana contains chlorophylls,
defined in M&M section. n.s.: not significant. β-carotene, violaxanthin y vaucheriaxanthin, as well as trace amounts of
astaxanthin (Cerón-García et al., 2018), which might explain our results
pointing to higher antioxidant response in muscle and liver of fish
(2020) reported a significant increase inmuscle protein in juvenile
supplemented with the microalgal biomass. Teimouri et al. (2016, 2019)
gilthead seabream fed enzymatically hydrolysed Arthrospira platensis
also described this effectas a result of the inclusion of microalgae in
added at 2% inclusion level, as well as a significant decrease in total
feeds, and more specifically, Qiao et al. (2019) reported lower TBARS
lipids when added at 4% inclusion level. Reduced muscle lipid storage
values both in liver and serum in Scophtalmus maximus juveniles fed on
has also been reported not only for microalgae species (Hussein et al.,
5% N. gaditana diets. A recent study (Sales et al., 2021) has shown that
2013; El-Sheekh et al., 2014; Vizcaíno et al., 2014, 2016), but also for
purified extracts of the unsaponifiable fraction of N. gaditana, rich in
macroalgae (Ortiz et al., 2006; Yildirin et al., 2009; Sáez et al., 2020).
carotenoids, included in feeds to partially replace fish oil yielded potent
These findings suggest the existence of bioactive compounds in algae
antioxidant effects in muscle of S. aurata juveniles.
capable of influencing protein and lipid metabolism, although the na­
The cellulase pre-treatment of the microalgal biomass was respon­
ture of such substances or the underlying mechanisms involved in such
sible for a trend towards increased in vivo antioxidant effects on muscle
effects have not yet been fully ascertained. Recent evidence in this re­
lipids (Table 8), compared to untreated N. gaditana. Such increase
gard was provided by Perera et al. (2020), although microalgae-
attributable to enzyme hydrolysis reached statistical significance in the
containing commercial products rather than pure microalgae biomass
case of liver lipids. These results suggest that increased release and
were considered in the study.
further bioavailability of some inner bioactive compounds contained in
Whilst no quantitative differences in muscle lipid content were
the cells might have occur, as was the case for total phenolics (Fig. 3).
observed, however, qualitative differences were found in this analytic
Galafat et al. (2020) also found lower TBARS values in muscle of
component. It is known that fish muscle lipids reflect dietary FA profiles
S. aurata juveniles fed with Arthrospira sp. protease hydrolysates
(Grigorakis et al., 2002; Grigorakis, 2007; Yildiz, 2007), and this might
included at low inclusion level (2 and 4%) in diets. In agreement, and
explain the significant increase of EPA muscle content of fish fed with all
with regard to phenolics, N. gaditana contains certain amount of these
the algae-containing diets (Table 4), compared to control batch.
substances in raw biomass, in line with previous studies (Kherraf et al.,
N. gaditana is rich in EPA (33% of total FA), and consequently, all
2017; Haoujar et al., 2019), which might explain the potent antioxidant
microalgae-supplemented diets, either raw or hydrolysed (Table 4),
effects found on fish lipids in our study. Noticeably, as mentioned, total
were enriched in this specific FA in a dose-dependent manner, but not
phenolics measured in the reaction vessels increased as a result of the
influenced by the enzymatic pre-treatment of the biomass (no significant
cellulase treatment (Fig. 3).
differences, between R25 and H25, or between R50 and H50, Table 2).
Physical treatments, even if valuable when it comes to increasing the
Although all the microalgae-enriched diets yielded muscle EPA
yield of microalgae main compounds (i.e. protein and lipid fractions),
contents higher than those found in CT batch (Table 4), this effect wasn’t
might jeopardize the chemical integrity of thermolabile minor com­
dose-dependent. Interestingly, and contrary to what was expected, the
pounds (Schafberg et al., 2020), and consequently, impair their func­
enzymatic treatment of the biomass yielded lower EPA in muscle
tional activity.
compared to the raw microalgae. A possible explanation would be that
Given the susceptibility of pigments, especially carotenoids, to
EPA released from cells could be more susceptible to structural damage
different factors (temperature, oxygen, light, acidic pH, etc., Schieber
than that remaining within the microalgae cells. In other words, intact
and Weber, 2016), and even if the extraction procedures increase the
cell walls might have acted as a sort of “natural microcapsule” for EPA.
releasing of inner compounds, it should also be born in mind that
All the experimental batches fed with the supplemented diets yielded
microalgal biomass, as part of the ingredient mixture, will be extruded
significantly higher DHA muscle content than control fish. This fact
during the elaboration of the experimental diets, a process involving
can’t be explained by differences in this FA in the experimental diets
high pressure and temperature. Consequently, doubts could arise related
(Table 2), given that no DHA was measured in N. gaditana biomass. On
to the integrity and the subsequent in vivo bioavailability of some of the

10
M.I. Sáez et al.
compounds released in vitro. Previous research suggests that the
resulting balance of disrupting strategies is favourable to the enrichment
of aquafeeds (Schafberg et al., 2020), and our results coincide with that
idea. But given the diversity and complex nature of the antioxidant
substances involved, this specific issue deserves further research. Espe­
cially the balance between phenolics and carotenoids in a given
microalgae species is likely that could determine the persistence of the
antioxidant effects in feeds after processing procedures.
Although instrumental colour measurements at early stages of the
productive cycle have no interest in practical terms of fish quality
assessment, they can still provide valuable information about pigment
deposition and antioxidant effects in growing fish tissues. The favour­
able influence of microalgae on fish colour parameters found in our
study (increased a* and b*, Table 9) has been documented previously
(Teimouri et al., 2013; Cardinaletti et al., 2018; Galafat et al., 2020;
Kousoulaki et al., 2020; Sales et al., 2021). Although tendencies
observed for skin pigmentation suggest that raw microalgae intensified
the effects compared to hydrolysed biomass (Table 9), however, no
significant differences attributable to the enzymatic treatment were
found. This might well be explained again by the fact that some pig­
ments contained in the hydrolysed biomass might have been damaged to
a higher extent than those from raw biomass due to feed processing.
The activity of digestive enzymes is acknowledged not only as a
marker of their digestive and absorptive capacity (Alarcón et al., 1998),
but also as a reliable indicator of the nutritional status of aquacultured
fish. More specifically, the activity of some brush border membrane
enzymes, such as leucine aminopeptidase (LAP) and alkaline phospha­
tase, reveals the integrity and the absorptive capability of the intestinal
mucosa (Silva et al., 2010). Disparate results have been reported on the
effects of raw microalgae on digestive enzyme activities, and thus, Qiao
et al. (2019) found increased trypsin activity in juvenile turbot supple­
mented with N. gaditana biomass at 7.5% inclusion level after a 10-week
feeding trial. On the contrary, Jorge et al. (2019) observed no effects on
total alkaline protease, trypsin, α-amylase and lipase activities in
response to dietary N. gaditana supplementation, although the low in­
clusion levels considered (0.5, 1, and 1.5%) together with the short
duration of the feeding trial (37 d) might well have accounted for such
lack of effects. Few studies are available assessing the physiological
consequences of microalgae enzyme hydrolysates on such activities.
Galafat et al. (2020) described higher trypsin and LAP activities as a
result of protease hydrolysates of Arthrospira sp. at low inclusion level (2
and 4%) in S. aurata juveniles. More recently, Galafat et al. (2022) re­
ported increased trypsin, chymotrypsin and leucine aminopeptidase
activities in gilthead seabream fry as a result of supplementing starting
diets with Arthrospira platensis at 5 and 10% compared to control fish. In
addition, within each inclusion level, animals fed with diets that
included the hydrolysed biomass yielded consistently higher digestive
enzyme activities than those receiving the crude biomass. The results
obtained in this study indicate that N. gaditana supplementation, even at
the low inclusion levels tested, overall increased the enzyme activities
assayed compared to control fish, irrespectively of sampling time (45 or
90 d), or biomass pre-treatment, with the exception of leucine amino­
peptidase activity at day 45 (Table 5). It is also worth mentioning that
the favourable effects of the experimental diets on intestinal digestive
activities observed in this work concur, roughly speaking, with the ul­
trastructural determinations (Table 7) carried out on the intestinal
mucosa at the end of the feeding trial, especially at the highest con­
centration assayed (5%).
Overall, concerning the enzyme pre-treatment considered in this
work, no decisive effects were observed in terms of fish growth, muscle
composition, or digestive functionality, but the remarkable influence of
this treatment on the oxidative status of fishlipids could result in
beneficial effects on other parameters linked to the health status of
aquacultured fish, a fact that deserves further investigation as well. It
should also be born in mind that any feeding trial under controlled
conditions and short duration has evident limitations in terms of further
11
Aquaculture 556 (2022) 738288
applicability on-farm, bearing in mind the numerous additional factors
involved in the operation of long-term production cycles in commercial
fish farms.
5. Conclusions
Although N. gaditana biomass at low inclusion level in feeds had no
impact on growth and muscle proximal composition, however, it is
worth mentioning that the lack of detrimental effects, together with
some beneficial effects on other physiological parameters (digestive
structure and functionality, oxidative status of muscle and liver lipids,
and skin colour), overall indicate that might represent a valuable ad­
ditive in long-term S. aurata production cycles.
The results obtained evidenced the effectiveness of the cellulase pre-
treatment when it comes to in vitro releasing of intracellular compounds
from N. gaditana cells, which could improve not only extraction yields in
industrial applications, but also increase the bioavailability of certain
metabolites with potentially bioactive and functional effects. However,
no conclusive evidence was found regarding the impact of this strategy
on most of the physiological parameters tested, with the exception of the
enhanced effects on lipid oxidation.
It would be interesting to carry out further research aimed at
assessing the possible influence of N. gaditana hydrolysates on other
valuable physiological aspects, such as their influence on the intestinal
microbiome, the intermediary metabolism (not least lipid metabolism)
and the immune response of gilthead seabream, especially at early
stages of the production cycle.
Authors’ contributions
Sáez, M.I., Alarcón, F.J. and Martínez T.F. conceived and designed
the experiments. Alarcón, F.J. and Galafat, A. prepared the aquafeeds.
Galafat, A., Sáez, M.I., Vizcaíno,A.J. and Martínez, T.F. performed fish
sampling. Arizcun, M., Chaves-Pozo, E.and Ayala, M.D. participated in
sampling, fish care and maintenance, and in biometric and proximal
analysis. Suárez, M.D. performed and interpreted fatty acid analysis.
Galafat, A., Sáez, M.I., Martínez, T.F., Suárez, M.D., Arizcun M., and
Chaves-Pozo E. performed analytical analysis and discussed the data.
Sáez, M.I., Alarcón, F.J. and Martínez, T.F. drafted the manuscript. T.F.
Martinez and M. Arizcun obtained the necessary funds for conducting
the research. All authors critically revised and approved the manuscript.
Funding information
This research was funded by Spanish MCIU-FEDER (grant #
RTI2018-096625-B-C33 and grant # RTI2018–096625-B-C31), SABANA
project (the European Union’s Horizon 2020 Research and Innovation
program, grant # 727874), AquaTech4Feed (grant # PCI2020-112204)
granted by MCIN/AEI/10.13039/501100011033, the EU “NextGener­
ationEU”/PRTR within the ERA-NET BioBlue COFUND). Servicio de
piensos experimentales was granted by EQC2018-004984-P and
EQC2019-006380-P.
Statement of informed consent, human/animal rights
The authors state that no conflicts, informed consent, human or
animal rights are applicable. All studies involving fish were conducted in
accordance with the requirements of the Directive 2010/63/EU, and the
Spanish legislation (Real Decreto 53/2013, as amended by RD218/
2021), regarding the ethical rules applicable in research involving lab­
oratory animals. Thereby, all the procedures were authorized by the
Bioethics and Animal Welfare Committee of the Instituto Español de
Oceanografía (REGA code ES300261040017) with the approval of the
Ministry of Water, Agriculture and Environment of the Autonomous
Community Region of Murcia (Spain; A13200101).
M.I. Sáez et al.
Declaration of Competing Interest
The authors declare that they have no conflict ofinterest.
References
Abdel-Tawwab, M., Ahmad, M.H., 2009. Live Spirulina (Arthrospira platensis) as a growth
and immunity promoter for Nile tilapia, Oreochromis niloticus (L.), challenged with
pathogenic Aeromonas hydrophila. Aquac. Res. 40, 1037–1046. https://doi.org/
10.1111/j.1365-2109.2009.02195.x.
Agboola, J.O., Teuling, E., Wierenga, P.A., Gruppen, H., Schrama, J.W., 2019. Cell wall
disruption: an effective strategy to improve the nutritive quality of microalgae in
African catfish (Clariasgariepinus). Aquac. Nutr. 25, 783–797. https://doi.org/
10.1111/anu.12896.
Alarcón, F.J., Díaz, M., Moyano, F.J., Abellán, E., 1998. Characterization and functional
properties of digestive proteases in two sparid; gilthead sea bream (Sparus aurata)
and common dentex (dentex dentex). Fish Physiol. Biochem. 19, 257–267. https://
doi.org/10.1023/A:1007717708491.
Almendinger, M., Saalfranka, F., Rohnab, S., Kurthc, E., Springerd, M., Pleissner, D.,
2021. Characterization of selected microalgae and cyanobacteria as sources of
compounds with antioxidant capacity. Algal Res. 53, 102168 https://doi.org/
10.1016/j.algal.2020.102168.
AOAC, 2000. Official methods of analysis. In: 17th Edition. The Association of Official
Analytical Chemists. Gaithersburg, MD, USA. Methods 925.10, 65.17, 974.24,
992.16.
Batista, S., Pintado, M., Marques, A., Abreu, H., Silva, J.L., Jessen, F., Valente, L.M.P.,
2020. Use of technological processing of seaweed and microalgae as strategy to
improve their apparent digestibility coefficients in European seabass (Dicentrarchus
labrax) juveniles. J. Appl. Phycol. 32, 3429–3446. https://doi.org/10.1007/s10811-
020-02185-2.
Becker, W., 2003. Microalgae in human and animal nutrition. In: Handbook of
Microalgal Culture. Blackwell Publishing Ltd., Oxford, UK, pp. 312–351. https://doi.
org/10.1002/9780470995280.ch18.
Bergmeyer, H.V., 1974. Methods of enzymatic analysis. In: Phosphatase, vol. 2.
Academic Press, New York, pp. 856–860.
Bradford, M., 1976. A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dyebinding. Anal.Biochem. 72,
248–254.
Buege, J.A., Aust, S.D., 1978. Microsomal lipid peroxidation. In: Methods in enzymology,
Vol. 52. Academic Press, pp. 302–310.
Cardinaletti, G., Messina, M., Bruno, M., Tulli, F., Poli, B.M., Giorgi, G., Tibaldi, E., 2018.
Effects of graded levels of a blend of Tisochrysis lutea and Tetraselmissuecica dried
biomass on growth and muscle tissue composition of European sea bass
(Dicentrarchuslabrax) fed diets low in fish meal and oil. Aquaculture 485, 173–182.
https://doi.org/10.1016/j.aquaculture.2017.11.049.
Cerezuela, R., Guardiola, F.A., Gonzalez, P., Meseguer, J., Esteban, M.A., 2012. Effects of
dietary Bacillus subtilis, Tetraselmischuii, and Phaeodactylumtricornutum, singularly or
in combination, on the immune response and disease resistance of sea bream (Sparus
aurata L.). Fish Shell. Immunol. 33, 342–349. https://doi.org/10.1016/j.
fsi.2012.05.004.
Cerón-García, M.C., González-López, C., Camacho-Rodríguez, J., López-Rosales, L.,
García-Camacho, F., Molina-Grima, E., 2018. Maximizing carotenoid extraction from
microalgae used as food additives and determined by liquid chromatography
(HPLC). Food Chem. 257, 316–324. https://doi.org/10.1016/j.
foodchem.2018.02.154.
Church, F.C., Swaisgood, H.E., Porter, D.H., Catignani, G.L., 1983. Spectrophotometric
assay using o-phthaldehyde for determination of proteolysis in milk proteins.
J. Dairy Sci. 66, 1219–1227.
CIE, 1986. Recommendations on uniform color spaces, color difference equations,
psychometric color terms, supplement 2 to CIE publication, vol. 15. Central Bureau
of the Commission Internationale de l’Éclairage Vienna, Austria (E1.3.1) 1971/
(TC1.3).
Del Mar, E.G., Largman, C., Broderick, J.W., Geokas, M.C., 1979. A sensitive new
substrate for chymotrypsin. Anal. Biochem. 99, 316–320. https://doi.org/10.1016/
s0003-2697(79)80013-5.
El-Sheekh, M., El-Shourbagy, I., Shalaby, S., Hosny, S., 2014. Effect offeeding Arthrospira
platensis (Spirulina) on growth and carcass composition of hybrid red tilapia
(Oreochromis niloticus x Oreochromis mossambicus). Turk. J. Fish. Aquat. Sci. 14,
471–478. https://doi.org/10.4194/1303-2712-v14_2_18.
Erlanger, B., Kokowsky, N., Cohen, W., 1961. The preparation and properties of two new
chromogenic substrates of trypsin. Arch. Biochem. Biophys. 95, 271–278. https://
doi.org/10.1016/0003-9861(61)90145-X.
Folch, J., 1957. A simple method for the isolation and purification of total lipids form
animal tissues. J. Biol. Chem. 226, 497–509. https://doi.org/10.1016/S0021-9258
(18)64849-5.
Galafat, A., Vizcaíno, A.J., Sáez, M.I., Martínez, T.F., Jerez-Cepa, I., Mancera, J.M.,
Alarcón, F.J., 2020. Evaluation of Arthrospira sp. enzyme hydrolysate as dietary
additive in gilthead seabream (Sparus aurata) juveniles. J. Appl. Phycol. 32,
3089–3100. https://doi.org/10.1007/s10811-020-02141-0.
Galafat, A., Vizcaíno, A.J., Sáez, M.I., Martínez, T.F., Arizcun, M., Chaves-Pozo, E.,
Alarcón, F.J., 2022. Assessment of dietary inclusion of crude or hydrolysed
Arthrospira platensis biomass in starter diets for gilthead seabream (Sparus aurata).
Aquaculture 548, 737680. https://doi.org/10.1016/j.aquaculture.2021.737680.
12
Aquaculture 556 (2022) 738288
Goiris, K., Muylaert, K., Fraeye, I., Foubert, I., De Brabanter, J., De Cooman, L., 2012.
Antioxidant potential of microalgae in relation to their phenolic and carotenoid
content. J.Appl.Phycol. 24, 1477–1486. https://doi.org/10.1007/s10811-012-9804-
6:1-10.
Gomes, T.A., Zanette, C.M., Spier, M.R., 2020. An overview of cell disruption methods for
intracellular biomolecules recovery. Prep.Biochem.Biotech. 50, 635–654. https://
doi.org/10.1080/10826068.2020.1728696.
Grigorakis, K., 2007. Compositional and organoleptic quality of farmed and wild gilthead
sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax) and factors affecting it:
a review. Aquaculture 272, 55–75. https://doi.org/10.1016/j.
aquaculture.2007.04.062.
Grigorakis, K., Alexis, M.N., Taylor, K.D.A., Hole, M., 2002. Comparison of wild and
cultured gilthead sea bream (Sparus aurata); composition, appearance and seasonal
variations. Int. J. Food SciTechn37(5), 477–484. https://doi.org/10.1046/j.1365-
2621.2002.00604.x.
Günerken, E., D’Hondt, E., Eppink, M.H.M., Garcia-Gonzalez, L., Elst, K., Wijffels, R.,
2015. Cell disruption for microalgae biorefineries. Biotechnol. Adv. 33, 243–260.
https://doi.org/10.1016/j.biotechadv.2015.01.008.
Haoujar, I., Cacciola, F., Abrini, J., Mangraviti, D., Giuffrida, D., Oulad El Majdoub, Y.,
SkaliSenhaji, N., 2019. The contribution of carotenoids, phenolic compounds, and
flavonoids to the antioxidative properties of marine microalgae isolated from
Mediterranean Morocco. Molecules 24, 4037. https://doi.org/10.3390/
molecules24224037.
Heredia, H., Pruvost, J., Gonçalves, O., Drouin, L., Marchal, L., 2021. Lipid recovery from
Nannochloropsis gaditana using the wet pathway: investigation of the operating
parameters of bead milling and centrifugal extraction. Algal Res. 56, 102318 https://
doi.org/10.1016/j.algal.2021.102318.
Hussein, E.E.S., Dabrowski, K., El-Saidy, D.M., Lee, B.J., 2013. Enhancing the growth of
Nile tilapia larvae/juveniles by replacing plant (gluten) protein with algae protein.
Aquac. Res. 44, 937–949. https://doi.org/10.1111/j.1365-2109.2012.03100.x.
Jorge, S.S., Enes, P., Serra, C.R., Castro, C., Iglesias, P., Oliva Teles, A., Couto, A., 2019.
Short-term supplementation of gilthead seabream (Sparus aurata) diets with
Nannochloropsisgaditana modulates intestinal microbiota without afecting intestinal
morphology and function. Aquac. Nutr. 25, 1388–1398. https://doi.org/10.1111/
anu.12959.
Kavitha, S., Gajendranb, T., Saranyaa, K., Selvakumarc, P., Manivasagana, V., 2021.
Study on consolidated bioprocessing of pre-treated Nannochloropsisgaditana biomass
into ethanol under optimal strategy. Ren.Energy 172, 440–452. https://doi.org/
10.1016/j.renene.2021.03.015.
Kherraf, A., Tehami, W., Boufeldja, W., Yahla, I., Dra, G.A., Mansour, I.F.Z., Benali, M.,
2017. Determination of the nutritional and functional metabolites of
Nannochloropsisgaditana produced in Algeria and evaluation of its antioxidant
activity. Der. Pharma. Chem. 9 (14), 8–13. https://doi.org/10.38150/sajeb.7(1).
Kilian, O., Benemann, C.S.E., Niyogi, K.K., Vick, B., 2011. High-efficiency homologous
recombination in the oil-producing alga Nannochloropsis sp. Proc. Natl. Acad. Sci. U.
S. A. 108, 21265–21269. https://doi.org/10.1073/pnas.1105861108.
Kiron, V., 2012. Fish immune system and its nutritional modulation for preventive health
care. Anim. Feed Sci. Technol. 173, 111–133. https://doi.org/10.1016/j.
anifeedsci.2011.12.015.
Kousoulaki, K., Mørkøre, T., Nengas, I., Berge, R.K., Sweetman, J., 2016. Microalgae and
organic minerals enhance lipid retention efficiency and fillet quality in Atlantic
salmon (Salmo salar L.). Aquaculture 451, 47–57. https://doi.org/10.1016/j.
aquaculture.2015.08.027.
Kousoulaki, K., Gerd Marit, B., Mørkøre, T., Krasnov, A., Baeverfjord, G., Ytrestøyl, T.,
Ruyter, B., 2020. Microalgal Schizochytriumlimacinum biomass improves growth and
filet quality when used long-term as a replacement for fish oil, in modern salmon
diets. Front. Mar. Sci. 7 https://doi.org/10.3389/fmars.2020.00057.
Koyande, A.K., Chew, K.W., Rambabu, K., Tao, Y., Chu, D.T., Show, P.L., 2019.
Microalgae: a potential alternative to health supplementation for humans, food Sci.
Human Wellness 8, 16–24. https://doi.org/10.1016/j.fshw.2019.03.001.
Lee, J.Y., Yoo, C., Jun, S.Y., Ahn, C.Y., Oh, H.M., 2010. Comparison of several methods
for effective lipid extraction from microalgae. Bioresour. Technol. 101, S75–S77.
https://doi.org/10.1016/j. biortech.2009.03.058.
Lepage, G., Roy, C.C., 1984. Improved recovery of fatty acid through direct
transesterification without prior extraction or purification. J. Lipid Res. 25,
1391–1396.
Menegol, T., Romero-Villegas, G.I., Lopez-Rodríguez, M., Navarro-Lopez, E., Lopez-
Rosales, L., Chisti, Y., Molina-Grima, E., 2019. Mixotrophic production of
polyunsaturated fatty acids and carotenoids by the microalga
Nannochloropsisgaditana. J. Appl. Phycol. 31, 2823–2832. https://doi.org/10.1007/
s10811- 019-01828-3.
Miller, G.L., 1959. Use of dinitrosalicylic acid reagent fordetermination of reducing
sugar. Anal. Chem. 31, 26–428.
Ortiz, J., Romero, N., Robert, P., 2006. Dietary fiber, amino acid, fatty acid and
tocopherol contents of the edible seaweeds Ulva lactuca and Durvillaea Antarctica.
Food Chem. 99, 98–104. https://doi.org/10.1016/j.foodchem.2005.07.027.
Pagels, F., Pereira, R.N., Vicente, A.A., Guedes, A.C., 2021. Extraction of pigments from
microalgae and cyanobacteria - a review on current methodologies. Appl. Sci. 11,
5187. https://doi.org/10.3390/app11115187.
Perera, E., Sánchez-Ruiz, D., Sáez, M.I., Galafat, A., Barany, A., Fernández-Castro, M.,
Martos-Sitcha, J.A., 2020. Low dietary inclusion of nutraceuticals from microalgae
improves feed efficiency and modifies intermediary metabolisms in gilthead sea
bream (Sparus aurata). Sci. Rep. 10, 18676. https://doi.org/10.1038/s41598-020-
75693-3.
Pfleiderer, G., 1970. Particle-bound aminopeptidase from pig kidney. Met.Enzymol. 19,
514–521. https://doi.org/10.1016/0076-6879(70)19038-0.
M.I. Sáez et al.
Qiao, H., Hu, D., Ma, J., Wang, X., Wu, H., Wang, J., 2019. Feeding effects of the
microalga Nannochloropsis sp. on juvenile turbot (Scophthalmus maximus L.). Algal
Res. 41, 101540 https://doi.org/10.1016/j.algal.2019.101540.
Rodríguez-Ruiz, J., Belarbi, E.H., Sanchez, J.L.G., Alonso, D.L., 1998. Rapid
simultaneous lipid extraction and transesterification for fatty acid analyses. Biotech.
Techn. 12 (9), 689–691. https://doi.org/10.1023/A:1008812904017.
Rojo, E.M., Piedra, I., González, A.M., Vega, M., Bolado, S., 2021. Effect of process
parameters on the valorization of components from microalgal and microalgal-
bacteria biomass by enzymatic hydrolysis. Bioresour. Technol. 335, 125256 https://
doi.org/10.1016/j.biortech.2021.125256.
Sáez, M.I., Vizcaíno, A.J., Galafat, A., Anguís, V., Fernández-Díaz, C., Balebona, M.C.,
Martínez, T.F., 2020. Assessment of long-term effects of the macroalgae Ulva ohnoi
included in diets on Senegalese sole (Solea senegalensis) fillet quality. Algal Res. 47,
101885 https://doi.org/10.1016/j.algal.2020.101885.
Sáez, M.I., Suárez, M.D., Alarcón, F.J., Martínez, T.F., 2021. Assessing the potential of
algae extracts for extending the shelf life of rainbow trout (Oncorhynchus mykiss)
fillets. Foods. 10, 910. https://doi.org/10.3390/foods10050910.
Sales, R., Galafat, A., Vizcaíno, A.J., Sáez, M.I., Martínez, T.F., Cerón-García, M.C.,
Alarcón, F.J., 2021. Effects of dietary use of two lipid extracts from the microalgae
Nannochloropsisgaditana (Lubián, 1982) alone and in combination on growth and
muscle composition in juvenile gilthead seabream, Sparus aurata. Algal Res. 10,
2270. https://doi.org/10.1016/j.algal.2020.102162.
Sansone, C., Brunet, C., Noonan, D.M., Albini, A., 2020. Marine algal antioxidants as
potential vectors for controlling viral diseases. Antioxidants 9, 392. https://doi.org/
10.3390/antiox9050392.
Schafberg, M., Loest, K., Müller-Belecke, A., Rohn, S., 2020. Impact of processing on the
antioxidant activity of a microorganism-enriched fish feed and subsequent quality
effects on fillets of rainbow trout (Oncorhynchus mykiss). Aquaculture 518, 734633.
https://doi.org/10.1016/j.aquaculture.2019.734633.
Schieber, A., Weber, F., 2016. Carotenoids. In: Handbook on natural pigments in food
and beverages.Industrial applications for improving food color. Woodhead
Publishing Series in Food Science, Technology and Nutrition, pp. 101–123. https://
doi.org/10.1016/B978-0-08-100371-8.00005-1.
Scholz, M.J., Weis, T.L., Jinkerson, R.E., Jing, J., Roth, R., Goodenough, U., 2014.
Ultrastructure and composition of the Nannochloropsisgaditana cell wall. Eukaryot.
Cell 13, 1450–1464. https://doi.org/10.1128/EC.00183-14.
Shah, M.R., Lutzu, G.A., Alam, A., Sarker, P., Chowdhury, M.K., Parsaeimehr, A.,
Daroch, M., 2018. Microalgae in aquafeeds for a sustainable aquaculture industry.
J. Appl. Phycol. 30 (1), 197–213. https://doi.org/10.1007/s10811-017-1234-z.
Silva, F.C.P., Nicoli, J.R., Zambonino-Infante, J.L., Le Gall, M., Kaushik, S., Gatesoupe, F.
J., 2010. Influence of partial substitution of dietary fishmeal on the activity of
digestive enzymes in the intestinal brush border membrane of gilthead sea bream,
Sparus aurata and goldfish, Carassius auratus. Aquaculture 306, 233–237. https://
doi.org/10.1016/j.aquaculture.2010.05.018.
Singh, R.P., Murthy, K.N.C., Jayaprakasha, G.K., 2002. Studies on the antioxidant
activity of pomegranate (Punicagranatum) peel and seed extracts using in vitro
models. J. Agric. Food Chem. 50, 81–86. https://doi.org/10.1021/jf010865b.
Sørensen, M., Gong, Y., Bjarnason, F., Vasanth, G.K., Dahle, D., Huntley, M., Kiron, V.,
2017. Nannochloropsis oceania-derived defatted meal as an alternative to fishmeal in
Atlantic salmon feeds. PLoS One 12 (7), e0179907. https://doi.org/10.1371/journal.
pone.0179907.
Teimouri, M., Amirkolaie, A.K., Yeganeh, S., 2013. The effects of Spirulina platensis meal
as a feed supplement on growth performance and pigmentation of rainbow trout
(Oncorhynchus mykiss). Aquaculture 396, 14–19. https://doi.org/10.1016/j.
aquaculture.2013.02.009.
Teimouri, M., Yeganeh, S., Amirkolaie, A.K., 2016. The effects of Spirulina platensis meal
on proximate composition, fatty acid profile and lipid peroxidation of rainbow trout
13
Aquaculture 556 (2022) 738288
(Oncorhynchus mykiss) muscle. Aquac.Nutr. 22, 559–566. https://doi.org/10.1111/
anu.12281.
Teimouri, M., Yeganeh, S., Mianji, G.R., Najafi, M., Mahjoub, S., 2019. The effect of
Spirulina platensis meal on antioxidant gene expression, total antioxidant capacity,
and lipid peroxidation of rainbow trout (Oncorhynchus mykiss). Fish Physiol.
Biochem. 45 (3), 977–986. https://doi.org/10.1007/s10695-019-0608-3.
Teuling, E., Wierenga, P.A., Agboola, J.O., Gruppen, H., Schrama, J.W., 2019. Cell wall
disruption increases bioavailability of Nannochloropsisgaditana nutrients for juvenile
Nile tilapia (Oreochromis niloticus). Aquaculture 499, 269–282. https://doi.org/
10.1016/j.aquaculture.2018.09.047.
Tibaldi, E., ChiniZittelli, G., Parisi, G., Bruno, M., Giorgi, G., Tulli, F., Poli, B.M., 2015.
Growth performance and quality traits of European seabass (D. labrax) fed diets
including increasing levels of freeze-dried Isochrysis sp. (T-ISO) biomass as a source
of protein and n-3 long chain PUFA in partial substitution of fish derivatives.
Aquaculture 440, 60–68. https://doi.org/10.1016/j.aquaculture.2015.02.002.
Tibbetts, S.M., Mann, J., Dumas, A., 2017. Apparent digestibility of nutrients, energy,
essential amino acids and fatty acids of juvenile Atlantic salmon (Salmo salar L.) diets
containing whole-cell or cell-ruptured Chlorella vulgaris meals at five dietary
inclusion levels. Aquaculture 48, 25–39. https://doi.org/10.1016/j.
aquaculture.2017.08.018.
Timira, V., Meki, K., Li, Z., Lin, H., Xu, M., Pramod, S.N., 2021. A comprehensive review
on the application of novel disruption techniques for proteins release from
microalgae. Crit. Rev. Food Sci. Nutr. 1–17, Web. https://doi.org/10.1080/
10408398.2021.1873734.
Vizcaíno, A.J., López, G., Sáez, M.I., Jiménez, J.A., Barros, A., Hidalgo, L., Alarcón, F.J.,
2014. Effects of the microalga Scenedesmus almeriensis as fishmeal alternative in diets
for gilthead sea bream, Sparus aurata, juveniles. Aquaculture 431, 34–43. https://
doi.org/10.1016/j.aquaculture.2014.05.010.
Vizcaíno, A.J., Saéz, M.I., López, G., Arizcun, M., Abellán, E., Martínez, T.F., Alarcón, F.
J., 2016. Tetraselmissuecia and Tisochrysis lutea meal as dietary ingredients for
gilthead seabream (Sparus aurata L.) fry. J. Appl. Phycol. 28 (5), 2843–2855. https://
doi.org/10.1007/s10811-016-0845-0.
Vizcaíno, A.J., Rodiles, A., López, G., Sáez, M.I., Herrera, M., Hachero, I., Alarcón, F.J.,
2018. Growth performance, body composition and digestive functionality of
Senegalese sole (Solea senegalensis Kaup, 1858) juveniles fed diets including
microalgae freeze-dried biomass. Fish Physiol.Biochem. 44, 1–17. https://doi.org/
10.1007/s10695-018-0462-8.
Walker, A.B., Berlinsky, D.L., 2011. Effects of partial replacement of fish meal protein by
microalgae on growth, feed intake, and body composition of Atlantic cod. N. Am. J.
Aquac. 73, 76–83. https://doi.org/10.1080/15222055.2010.549030.
Wu, C., Xiao, Y., Lin, W., Li, J., Zhang, S., Zhu, J., Rong, J., 2017. Aqueousenzymatic
process for cell wall degradation and lipid extraction fromNannochloropsissp.
Bioresour. Technol. 223, 312–316. https://doi.org/10.1016/j.biortech.2016.10.063.
Yarnold, J., Karan, H., Oey, M., Hankamer, B., 2019. Microalgal aquafeeds as part of a
circular bioeconomy. Trends Plant Sci. 24, 959–970. https://doi.org/10.1016/j.
tplants.2019.06.005.
Yildirin, Ö., Ergün, S., Yaman, S., Türker, A., 2009. Effect of two seaweeds (Ulva lactuca
and Enteromorpha linza) as a feed additive in diets on growth performance, feed
utilization, and body composition of rainbow trout (Oncorhynchus mykiss).
KafkasÜniv. Vet. FakültesiDergisi 15, 455–460.
Yildiz, M., Şener, E., Timur, M., 2007. Effects of differences in diet and seasonal changes
on the fatty acid composition in fillets from farmed and wild sea bream (Sparus
aurata, L) and sea bass (Dicentrarchus labrax, L). Int. J. Food Sci. Technol. 43,
853–858.
Yong, T.C., Chiu, P.-H., Chen, C.H., Hung, C.H., Che, C.N.N., 2020. Disruption of thin-
and thick-wall microalgae using high-pressure gases: effects of gas species, pressure
and treatment duration on the extraction of proteins and carotenoids. J. Biosci.
Bioeng. 129, 502–507. https://doi.org/10.1016/j.jbiosc.2019.10.008.
Another random document with
no related content on Scribd:
 — Mies, kuka oletkin, astu syrjään, sillä minulla on asiaa
jumalanaiselle.

— Minulla myös, oli kylmä vastaus, — sillä herramme Sandi on

pannut minut tänne, ja minä olen kuin hän; täällä olen seisonut joka
yö paitsi yhtä.

Mackineylla oli revolveri kädessään, mutta hän ei uskaltanut

ampua peläten hälyttävänsä majan asukkaat.

— Anna minun mennä, sanoi hän. Hän tunsi, pikemmin kuin näki,
pitkän keihään, joka oli ojennettu hänen rintaansa vastaan
pimeässä. — Anna minun olla, niin minä annan sinulle monta säkkiä
suolaa ja putkia enemmän kuin metsässä on puita.

Hän kuuli naurahduksen pimeästä.

— Annat liian paljon vähästä, sanoi ääni. — Oi Mlaka!

Mackiney kuuli jalkojen tömisevän; hän oli satimessa, sillä jossakin

hänen edessään aseistetut miehet sulkivat polun.

Hän kohotti revolverin ja ampui kaksi laukausta olentoon.

Keihäs vihelsi mennessään hänen ohitseen, hän syöksähti

eteenpäin ja kävi käsiksi polulla olevaan mieheen.

Hän oli vahva kuin nuori leijona, mutta hänen kurkkuunsa tarttuva
käsi ei myöskään ollut heikko. Hetkisen he kamppailivat, ja sitten he
kaatuivat kierien toistensa yli polulla.

Mackiney tavoitti toista revolveriaan. Hän sai kiinni perästä, kun

hän samassa tunsi väristyksen — jokin sattui hänen pehmeästi,
vasempaan kylkeen — jokin, joka pani hänen jokaisen hermonsa
tuskasta värisemään.

— Oh, taivas, sanoi Mackiney englanniksi.

Sitten hän ei puhunut enää.

*****

— Arabialainen vai valkoihoinen, sitä en tiedä, sanoi Monrovian

Bosambo, — eikä ole ketään sitä sanomassa, sillä mieheni olivat
nopeat tappamaan, ja vain yksi hänen joukostaan on elossa, eikä
hän tiedä mitään.

— Mitä olet tehnyt tälle arabialaiselle? kysyi Sanders.

He pitivät palaveria lähetysasemalla hämärän ensi hetkenä, ja

tyttö, kalpeana ja peloissaan, istui pöydän ääressä silmäillen vuoroin
toista, vuoroin toista, sillä hän ymmärsi kieltä heikosti.

— Herra, sanoi Bosambo, — hautasin hänet mieleni mukaan niin,

ettei kukaan tietäisi tästä hyökkäyksestä, koska se panisi pahoja
ajatuksia heidän päähänsä.

— Teit viisaasti, sanoi Sanders.

Hän palasi päämajaan hieman ymmällä, sillä hän ei tiennyt

tapauksen yksityiskohdista mitään.

Ja kun kuukautta myöhemmin hänelle saapui tärkeä tiedustelu,

koskeva muuatta Burney Mackineyta, hän vastasi
totuudenmukaisesti, ettei hän voinut antaa mitään tietoja.
KAUNOPUHEINEN NAINEN

Ngombilaisten joukossa oli nainen, jolla oli suloinen kieli. Kun hän
puhui, miehet kuuntelivat innokkaasti, sillä sensukuinen hän oli,
syntymästään asti voimakaspuheinen.

Hän villitsi oman kylänsä asukkaat niin, että he tekivät

hyökkäyksen Ranskan alueelle tuottaen suuren häpeän hänen
isälleen, sillä Sanders saapui kiiruusti pohjoiseen, ja nyt seurasi
voimakkaat pieksäjäiset, melkeinpä hautajaiset. Siitä isä havaitsi
viisaaksi naittaa tämän naisen miehelle, joka voisi pitää hänen
kielensä aisoissa.

Niin hän naitti hänet eräälle päällikölle, joka kuului ngombilaisiin, ja

tämä päällikkö rakasti häntä niin paljon, että teki hänet
päävaimokseen ja rakensi hänelle majan aivan omansa viereen.
Kaulassaan hän kantoi suurta messinkirengasta, suunnilleen
kahdenkymmenen neljän naulan painoista — suuri huomionosoitus,
jota miehen toiset vaimot kadehtivat.

Tämä päävaimo oli noin viidentoista vuoden ikäinen — joka on

Joella melkein keski-ikä — ja oli kaikin puolin viisas miesten tapaan.
Liiankin viisas, jotkut ajattelivat, ja varmaankin hänen herrallaan oli
valittamisen syytä, kun hän palatessaan metsästysretkeltään päivää
tai paria aikaisemmin kuin oli otaksuttu tapasi vaimonsa
onnellisempana kuin hän olisi toivonut, eikä ensinkään yksinäisenä.

— Mfasimbi, sanoi hän, kun nainen polvistui hänen eteensä

käsivarret ristissä paljaalla, ruskealla povellaan, — isäni aikoina
olisin taivuttanut kimmoisen puun ja sitonut sinut kaulastasi siihen, ja
kun pääsi olisi irronnut ruumiistasi, olisin polttanut sinut ja hänet,
joka häväisi minut. Mutta se ei ole vaikeitten miesten laki, ja minä
pidän sinua arvottomana naisena saattaakseni sinun tähtesi niskani
vaaraan.

— Herra, minusta ei ole paljon hyvää, sanoi nainen.

Koko päivän hän makasi maassa ympärillään koko kylän väki, jolle
hän puhui työmiesten sahatessa poikki messinkirengasta hänen
kaulastaan. Päivän mentyä rengas saatiin pois ja päällikkö lähetti
hänet takaisin vanhempansa luo, jolta hän oli tytön ostanut suurella
summalla. Hänen toimenpiteensä kohtasi suurta vastustusta, sillä
nainen oli käyttänyt aikansa hyödyllisesti, ja koko kylä oli niin
kuohuksissaan, että se oli valmis kapinaan.

Sillä ei yksikään nainen eroa miehestään — olipa hänellä sitten

Pariisin silkit ja hepenet tai Ngombin kamferttipuuta ja öljyä — ilman
vihan ja koston tunteita, ja tuskin oli Mfasimbi melottu miehensä
kylästä, kun hän suunnitteli kostoa.

Häntä seurasi maanpakoon se mies, jonka kanssa ja jonka tähden

hän oli pannut peliin ja hävinnyt niin paljon. Hänen nimensä oli
Otapo, ja hän oli hyvin penseä.

Heidän meloessaan nainen nousi polvilleen miehen taa ja sanoi:

 — Otapo, mieheni on tehnyt minulle suurta vääryyttä ja pannut
multaa pääni päälle, etkä sinä sano mitään.

— Miksi minä puhuisin, kun sinä olet puhunut liikaa? kysyi Otapo
kylmästi. — Kiroan sen päivän, jolloin sinut näin, Mfasimbi, sillä
erehdykseni on maksanut minulle kalaverkon, joka oli kylän paras, ja
palan uutta kangasta, jonka ostin kauppamieheltä; ne on herramme
päällikkö ottanut.

— Jos sinulla olisi ollut miehen mieli, niin Namani, mieheni, olisi
nyt kuollut, ivasi nainen.

— Olen tappanut itseni ja menettänyt verkkoni, sanoi Otapo, — ja

vielä kangaskappaleeni.

— Olet kuin akka.

— Toivon, että äitini olisi synnyttänyt tytön, kun hän synnytti minut,
sanoi Otapo, — silloin en olisi joutunut häpeään.

Nainen meloi vaiteliaana hetkisen ja sanoi sitten äkkiä:

— Tämä nainen on joko hullu tai hänelle on tehty suuri vääryys.

Hän sai pian siitä tiedon, sillä nainen juoksi rinnettä hänen
luokseen ja polvistui syleillen hänen jalkojaan.

— Iwa! Kuolema miehelleni Namanille, joka on valehdellut minusta

ja lyönyt minua, oi isien isä! huusi hän.

— Nainen, sanoi isä, — mitä tämä on?

Nainen kertoi hänelle tarinan — julman tarinan. Myöskin hän

kertoi, mikä tärkeintä, Otapon surman.
 — Tämä mies, puolustaakseen minua, vei minut mieheni luota,
joka pieksi minua, nyyhkytti hän, — ja mieheni seurasi, ja kun me
istuimme aterioimassa joen rannalla, mieheni keihästi hänet
takaapäin. Oi voi! — Ja hän kieriskeli tomussa isänsä jaloissa.

Päällikkö oli vihainen, sillä hän oli Namania korkeampi herra ja sitä
paitsi valvoi seudun rauhaa komissaarin puolesta.

— Tämä on tappopalaver, ja se on liian raskas minun

ratkaistavakseni, sanoi hän, — ja kun sinä sitä paitsi olet minun
tyttäreni, voitaisiin luulla, etten jaa oikeutta rehellisesti miehen ja
miehen välillä.

Niin hän astui kanoottiinsa ja matkasi Isauun, jossa Sanders oleili.

Komissaari oli selviämässä kuumepuuskasta eikä ollut mielissään

päällikön tulosta. Vielä vähemmän hän oli hyvillään kuultuaan
»kaunopuheisen naisen» tarinan.

— Menen surmapaikalle ja katson, mitä voin nähdä. — Hän meni

»Zairelle», ja hyvää vauhtia pieni siipialus kiiti paikalle, jonka nainen
osoitti. Sanders nousi maalle siinä, missä kanootin jälki vielä näkyi
hiekalla, sillä Joki ei nouse eikä laske huomattavasti kuukausiin.

Hän seurasi Mfasimbia metsään ja näki siellä Otapon maalliset

jäännökset, ja hän näki keihään.

Mfasimbi katseli häntä läheltä.

— Herra, sanoi hän uikuttaen, — tällä paikalla Namani tappoi

nuorukaisen Otapon, kun istuimme aterioimassa.

Sandersin tarkka silmä tähysti paikkaa.

 — En näe tulen jälkiä, sanoi Sanders äkkiä.

— Tulen, herra? änkytti nainen.

— Mihin ihmiset käyvät aterialle, siihen he tekevät tulen, sanoi

Sanders lyhyesti, — eikä tässä ole ollut tulta sitten maailman alun.

Hän vei naisen takaisin laivaan ja nousi jokea Namanin kylään.

— Mene, sanoi hän salaa hausakersantille, — ja ellei päällikkö

tule luokseni, niin vangitse hänet, ja jos hän tulee, niin ota huostaasi
hänen majansa ja naisensa.

Namani odotti tervehtiäkseen häntä, ja Sanders komensi hänet

laivalle.

— Namani, sanoi Sanders, — tunnen sinut kunnon mieheksi, eikä

sanaakaan ole sinua vastaan sanottu. Nyt tämä nainen, sinun
vaimosi, sanoo, että sinä olet murhaaja ja olet murhannut Otapon.

— Hän on valehtelija! sanoi Namani tyynesti. — En tiedä mitään

Otaposta.

Tarkka kuulustelu, jota kesti kaksi päivää, ei voinut osoittaa

päällikköä syylliseksi. Se pikemminkin loi epäilyttävän varjon
Mfasimbin luonteeseen; mutta maassa, jossa naisella on joukoittain
rakastajia, hän kärsi siitä vähän.

Kahden päivän kuluttua Sanders langetti tuomion.

— Olen tyytyväinen siihen, että Otapo on kuollut, sanoi hän, —

monesta syystä en ole varma siitä, että Namani tappoi hänet. Uskon,
että Mfasimbi on pahojen töiden nainen ja suuri lörpöttelijä, ja sen
vuoksi karkoitan hänet kaukaiseen maahan vieraiden ihmisten pariin.

Hän otti naisen laivalle, ja »Zaire» lähti.

Kahdenkymmenenneljän tunnin kuluttua hän tuli »metsän

kaupunkiin», Ochoriin, ja hänen laivansa vihellyksen kuultuaan kylän
väki tuli juosten rantaan.

Bosambo, Ochorin päällikkö, saapui viimeisenä, sillä hän saapui

juhlasaatossa ison taivaansinisen sateenvarjon suojassa, yllään
kiiltokullalla koristeltu vaippa, ja hänen edellään kulki kymmenen
vanhinta kantaen kullattuja keppejä.

Sanders katseli päällikön tuloa laivansa sillalta, eivätkä hänen

kasvonsa ilmaisseet minkäänlaisia tunteita. Kun Bosambo oli
päässyt laivalle, kysyi komissaari häneltä:

— Mitä lastenkujeita tämä on, Bosambo?

— Herra, sanoi Bosambo,— tällä tavoin saapuvat suuret kuninkaat

suurempien kuninkaiden luo, sillä olen nähnyt jumalanaisen luona
eräitä kuvia kirjoissa, ja niistä olen saanut oppia.

— Sillä tavoin ihmiset myöskin pukeutuvat, kun he menevät

ilveilemään, sanoi Sanders epämiellyttävästi. — Nyt olen tuonut
sinulle naisen, joka puhuu liian paljon ja jonka eräs mies karkoitti
luotaan ja joka on luullakseni murhannut toisen miehen, ja toivon,
että hän asuu sinun kylässäsi.

— Herra, niin kuin käsket, sanoi kuuliainen Bosambo ja katseli

tyttöä arvostelevasti.
 — Naita hänet, kun hän toivoo, sanoi Sanders, — mutta hänen
tulee kuulua sinun sukuusi ja sinä olet vastuussa hänestä siihen
saakka. — Herra, hän menee naimisiin tänä iltana, sanoi Bosambo
vakavasti.

Kun Sanders oli mennyt ja hänen katoavan laivansa savu häipynyt

puiden taa, kutsui Bosambo päämiehensä ja vanhimpansa
neuvotteluun.

— Miehet, sanoi hän, — herra Sandi, joka rakastaa minua

suuresti, on saapunut tuoden lahjoja — tämän naisen.

Hän viittasi kädellään sievään tyttöön, joka seisoi hänen

vieressään pikku kummulla, jolla neuvottelumaja oli.

— Hän on Ngombin ihanin nainen, sanoi Bosambo, — ja hänen

nimensä on Nlaminsafo, joka merkitsee Helmi, ja Sandi on maksanut
hänestä suuren summan, sillä hän tanssii kuin leopardi leikissään ja
hänellä on monta rakastettavaa ominaisuutta.

Tyttö ymmärsi Ochorin outoa murretta kylliksi, että tiesi ansioitaan

lueteltavan, ja kohotti jalkaansa kömpelösti.

— Hän on vaimojen vaimo, sanoi Bosambo painokkaasti, —

lempeä luonteeltaan ja suloinen, mainio maniokin keittäjä ja
tarinoiden kertoja — mutta minä en kuitenkaan voi mennä naimisiin
hänen kanssaan, sillä minulla on monta vaimoa ja minä olen kuin
vaha heidän käsissään. Niin että teistä voi ottaa hänet se, joka
maksaa reilusti ja pelottomasti, sillä ostattehan sitä, mikä on vuohia
ja suolaa arvokkaampi.
 Kymmenestä vuohesta ja tuhannesta putkesta Sandin »lahja»
siirtyi erään päämiehen omaisuudeksi.

Puhuessaan päävaimonsa kanssa asiasta Bosambo sanoi:

— Näin on Sandia toteltu, näin olen minäkin tyytyväinen ja kaikki

on tapahtunut Jumalan tahdon mukaan.

— Jos sinä olisit ottanut hänet, Muhamed, sanoi vaimo, joka oli
kanonainen ja harras uskovainen, — olisit tullut surulliseksi.

— Kirkkaan valon helmi, sanoi Bosambo vaatimattomasti, — sinä

olet elämäni ensimmäinen, niin kuin Jumala tietää; sinun tähtesi olen
luopunut muista jumalista ja uskon yhteen, voimakkaaseen ja
laupiaaseen; sinun tähtesi olen myöskin ostanut suuren
sateenvarjon kanokuninkaiden maihin.

Seuraavana päivänä Bosambo meni metsään eikä palannut,

ennen kuin viikko oli vierähtänyt.

Ochorilaisten tapana on, kuten muidenkin heimojen, mennä

päällikköä vastaan tämän palatessa metsästämästä, ja outoa oli,
ettei kukaan tullut häntä vastaan laulaen norsun laulua.

Kahdenkymmenen miehensä kera hän saapui melkein

huomaamatta majalleen.

Puolitiessä kylän kadulla hän tapasi vanhahkon miehen, joka

juoksi hänen luokseen.

— Herra, sanoi tämä, — älä mene Fabadinon, ylimmän

päämiehesi majalle.
 — Onko hänessä tauti? kysyi Bosambo.

— Pahempaa, herra, sanoi vanha kyynikko. — Hänellä on vaimo,

ja kuusi päivää ja suurimman osan kuudesta yöstä koko kylä on
istunut kuunnellen häntä.

— Mitä loruja hän puhuu? kysyi Bosambo.

— Hän puhuu kaikki asiat selviksi, sanoi vanhus, — ja kaikilla

hänen sanoillaan on tarkoitus, ja hän heittää valoa niin kuin itse
aurinko pimeihin aivoihin, ja kaikki näkevät hänen kanssaan.

Bosambolla oli mukanaan kaksikymmentä miestään, joihin hän

saattoi luottaa. Pimeys oli tulossa, ja kylän kauimmaisessa kolkassa
hän näki suuren nuotion, jonka luona »puhelias nainen» puhui ja
puhumistaan puhui.

Hän meni ensin majalleen. Hän tapasi kanovaimonsa yksikseen,

muut hänen majansa vaimot olivat kaikonneet.

Herra, en odottanut näkeväni sinua elävänä, sanoi nainen, niin

että varroin kuolemaa, kun sen aika tulee.

— Siihen menee monta vuotta, virkkoi Bosambo.

Hän lähetti vaimon kahden miehen saattamana metsään; loput

hänen miehistään lähtivät kaksittain ja kolmittain päämiehen majan
luona olevan joukon jatkoksi.

Ison nuotion loimuavat liekit valaisivat majan edessä levitetyin

käsin seisovaa hentoa olentoa.
 —… Kuka teki teidät orjan orjiksi — Bosambon orjiksi? Kuka antoi
hänelle vallan sanoa: »Menkää» tai: »Jääkää»? Ei kukaan. Sillä eikö
hän ole ihminen kuin tekin, ei erilainen rakenteeltaan, ei
tarkempisilmäinen, ja jos pistätte häntä keihäällä, eikö hän kuole
kuin tekin kuolette?

Ja Sandi, eikö hänkin ole ihminen, vaikka valkea? Onko hän

voimakkaampi kuin Efambi tai Elaki tai Jako? Minä sanon, että te
ette ole vapaita ihmisiä niin kauan kuin Bosambo elää tai Sandi elää.

Bosambo oli mies, jolla oli eläimen vaistot. Hän tunsi ilmassa
levottomuutta. Jokainen väräjävä hermosäie toi hänelle viestin:
hänen miehensä olivat luisumassa pois hänen vallastaan. Hän ei
epäröinyt.

Suuri joukko kansaa seisoi hänen ja naisen välillä. Hän ei päässyt

tähän käsiksi.

Irroittamatta katsettaan naisesta hän pani kätensä kilven alle ja otti

pitkän heittokeihään. Hänellä oli tilaa heittoon; tasapainossa ja
notkuvana ase oli hänen aivan suoraksi ojennetussa kädessään.

Hiu-ii!

Kiiltävä kärki suhahti ilman läpi nopeammin kuin katse voi seurata.

Mutta nainen oli nähnyt ojennetun käden ja tunsi heittäjän, ja hän

juoksi syrjään.

Keihäs sattui hänen takanaan seisovaan mieheen — Fabadimoon,

ylimmäiseen päämieheen, joka kuoli ääntä päästämättä.
 — Bosambo! kirkui tyttö ja viittasi. — Bosambo, tappakaa —
tappakaa!

Bosambo kuuli keihäiden kalinan ja pakeni pimeyteen.

*****

Sanders makasi riippumatossa, jonka toinen pää oli kiinnitetty

kuistinkaiteeseen ja toinen hänen bungaloonsa seinään iskettyyn
koukkuun. Hän luki, tai ainakin koetti lukea pitkää ja virallista kirjettä,
joka oli saapunut päämajasta. Se koski erästä Sandelsin lähettämää
veromäärää, ja arvatenkin tämä lähetys oli ollut eräissä suhteissa
liian pieni.

— Herra, tällä on kirja, sanoi mies.

Sanders oli heti hereillä ja pystyssä riippumatostaan.

Linnun punaisen jalan ympärille oli pitkällä kuminauhalla kiinnitetty

kahden savukepaperin suuruinen silkkipaperi. Hän avasi sen.

Kopiokynällä oli kirjoitettu arabiaksi muutamia sanoja:

Abibulta, Jumalan palvelijalta, Sandille, kansansa ainavalvovalle

isälle.

Rauha sinulle ja talollesi. Julistaen, että on vain yksi Jumala,

totinen ja jakamaton, ilmoitan sinulle, että sinun Bosambolle
antamasi nainen saa aikaan suurta hämminkiä. Tämän ilmoittavat
sanansaattajani, Bosambon paettua kahdenkymmenen miehen
kerällä Isisin rajalle.
 Kirjoitettu Isisi-joella paikassa, jossa krokotiilisärkät kohtaavat
nuolen muotoisina.

Abibu oli jätetty kylään, josta Mfasimbi oli karkoitettu. Hänet oli
jätetty selvittämään Otapon kuoleman salaisuutta, eikä häntä
helposti säikytetty.

Vedettyään jalkaansa hyttyssaappaat Sanders meni

hausaupseerin taloon.

Hän tapasi tämän herran yksinään teetä juomassa.

— Tarvitsen sinua, sanoi Sanders. — Ochorissa on meteli.

Upseeri kohotti kulmiaan. Hän oli kahdenkymmenenviiden vuoden

kyynillisellä puolella oleva nuori mies.

— Eihän vain lempeä Bosambo, vastusteli hän, — eihän vain tuo

ritarillisuuden perikuva?

— Älä ole koomillinen, Hamilton, sanoi Sanders. — Ochorilaiset

ovat kuohuksissa, ja jossakin siellä lähellä on naislähetyssaarnaaja.

Kapteeni ponnahti jaloilleen.

— Siunatkoon, unohdin naisen tykkänään! huudahti hän. Hän otti

rintataskustaan pillin ja puhalsi, ja paljasjalkainen mies juoksi
vahtikojusta pienen kentän poikki.

— Ta-ta-ta! sanoi kapteeni, ja joukon kokoontumiskutsu soi.

— Mikä tämä jupakka oikeastaan on? tivasi hän, ja lyhyesti

Sanders selitti tilanteen.
 Täyttä höyryä »Zaire» puski jokea ylös. Päivän ja yön se höyrysi,
kunnes tuli »paikkaan, missä krokotiilisärkät kohtaavat toisensa
nuolen muotoisina», ja Sanders pysähtyi ottamaan Abibun ja
kourallisen hausoja.

Sanders kuuli helpotuksekseen, ettei taistelu ollut ulottunut

lähetysasemalle saakka.

— Entä Bosambo? kysyi hän.

— Elävä vai kuollut, en tiedä, sanoi Abibu filosofisesti, — ja jos

hän on kuollut, hän kuoli uskovaisena, sillä kanonainen, jonka hän
otti vaimokseen, uskoo yhteen Allahiin ja Muhamediin, hänen
profeettaansa.

— Kaikki tämä voi olla totta, sanoi Sanders kärsivällisesti, — mutta

minua kuitenkin vähemmän huvittaa hänen
kuolemattomuussuunnitelmansa kuin hänen ruumiinsa nykyinen tila.

Siitä ei Abibu osannut kertoa mitään, paitsi että Bosambo oli

vetäytynyt eräälle kymmentä mailia kauempana olevalle saarelle ja
oli ollut siellä vielä kaksi päivää sitten.

Näillä paikoin joki mutkittelee ja polveilee, eikä keskellä olevalla

saarella aavistettu mitään, ennen kuin »Zaire» kääntyi äkkimutkan
takaa esiin.

— Asettukaa konekivääreille! sanoi Sanders terävästi.

Hausakapteeni istahti toisen messinkivaippaisen konekiväärin

satulaan, ja Abibu otti toisen.

Vesi kiehui kanootteja.

 Ochorilaiset hyökkäsivät saarta vastaan; »Zairen» rattaan jyske
hukutti kaiken muun äänen.

— Tappelevat mainiosti, virkahti kapteeni. — Mitä sanot, Sanders?

Sanders seisoi käsi ruorirattaassa, odottavana, katse kiinteästi

eteen tähystäen.

Nyt hän näki selvästi. Yksi joukko oli astunut maihin, ja siellä kävi
kiivas käsirysy.

— Antaa mennä, sanoi hän, ja kaksi paukkuvaa tulisuihkua

syöksyi konekivääreistä.

Vastaukseksi kanootit kuin taikavoimasta muodostivat ensin kaksi,

sitten kolme, sitten neljä taistelulinjaa, ja hurjaa vauhtia ne tulivat
myötävirtaa.

Sitten toinen konekivääri lakkasi toimimasta; tuli keihäskuuro, ja

yksi hipaisi Sandersia.

Hetkessä pieni alus oli piiritetty — konekiväärin taika oli pettänyt

ensi kertaa Joella. Se oli niin odottamatonta, niin hämmästyttävää,
että miehelle voitaisiin antaa anteeksi hämmentyminen sellaisessa
tilanteessa; mutta Sandersin käsi ei vavissut, kun hän pyöräytti
peräsinrattaan ympäri ja höyrylaiva teki täyskäännöksen.

Hausat ampuivat karbiineillaan täysilaidallisia; päähän

haavoittunut kapteeni korjasi järkkymättömänä konekiväärin lukkoa.

»Zaire» kiiti täyttä höyryä myötävirtaa alas. Kanootit eivät

pysyneet sen matkassa, lukuunottamatta yhtä, joka oli kiinnitetty sen
kylkeen; hausat tappoivat pistimillä siinä olijat pyytämättä ja
kuulematta selityksiä.

— Olen saanut konekiväärin korjatuksi, sanoi kapteeni ja pani

syöttäjään uuden panosvyön.

Sanders nyökkäsi. Sana peränpitäjälle, ja »Zaire» kääntyi. Se

palasi kanoottilinjoja kohti konekivääri säännöllisesti paukkuen.

Toinen linja joutui sekasortoon ja murtui, kolmas ei kehittynyt

lainkaan. Pakenevien kanoottien keskellä oli muuan toisia suurempi.
Sen perässä seisoi nainen viittoillen ja puhuen.

— Abibu, sanoi Sanders, ja hausa antoi konekiväärinsä toisen

huostaan ja tuli herransa luo. — Näetkö tuon naisen tuolla
kanootissa?

— Herra, näen hänet, sanoi Abibu.

— Minusta tuntuu, sanoi Sanders vakavasti, — että hänen olisi

parasta kuolla.

Hän soitti konehuoneeseen pysähtymismerkin, ja koneiden häly

lakkasi. »Zaire» kulki eteenpäin tärisemättä, ja Abibu asettui
pitkäkseen kannelle, pani kiväärin poskelleen ja tähtäsi huolellisesti.

*****

He tapasivat Bosambon tunnottomana suuren ruumiskasan alta.

Hän makasi poikittain kanonaisen päällä, joka myöskin oli elossa,
sillä Bosambo oli saanut ne iskut, jotka olivat häneen tähdätyt.

Hänessä oli, kuten Sanders laski, kaksikymmentäviisi haavaa.

 — Herra, hän kuiskasi, kun Sanders seisoi hänen vieressään, —
enkö sanonut, että ochorilaiset osaavat tapella?

— He ovat tapelleet erään asian vuoksi, lapseni, sanoi Sanders

irvistäen.

Bosambo irvisti hiukan.

— Herra, sanoi hän pehmeästi, — kun menen takaisin heidän

luokseen, niin he tulevat surullisiksi.

Ja he tulivat myöskin.
RUKOILEVA MAURILAINEN

Abibu kertoi Sandersille, että muuan arabialainen oli tullut

tapaamaan häntä, ja arabialaiset ovat Rannikolla harvinaisia, vaikka
muutamilla tummaihoisilla, seemiläisistä polveutuvilla miehillä onkin
tämä kunnioitettava titteli.

Sanders tuli kuistille odottaen tapaavansa kanolaisen ja

hämmästyi nähdessään siellä kyykkimässä oikeaa maurilaista
tyyppiä olevan miehen. Tämä istui kädet polvillaan, kietoutuneena
tahrattomaan, valkeaan dzellabiin.

— Oletko Marokosta, kysyi Sanders arabiaksi, — vai Dakarista?

Mies nyökkäsi.

— Dakarilaiset ovat koiria, sanoi hän tarinoitsija-ammattilaisen

laulavalla äänellä. — Eräs mies, joka oli äitini serkku, varasti
talostani kaksikymmentä douroa ja palasi Dakariin rannikkolaivalla,
ennen kuin ehdin tavoittaa ja piestä hänet. Toivon, että hänet on
tapettu ja hänen perheensä myöskin. Bismallah. Jumala on hyvä.

Sanders kuunteli, sillä hän tiesi tangerilaisten olevan puheliaita.

 Mies jatkoi. — En välitä, olipa mies Alin tai Sulin sukua.
Molemmissa on varkaita.

— Miksi tulit tänne? kysyi Sanders.

— Tunsin kerran miehen, joka istui suuressa sokissa. (Sanders

antoi hänen kertoa tarinansa omalla tavallaan.) Ja kaikki maan
asukkaat, jotka toivat kasviksia ja sysiä torille, suutelivat hänen
dzellabinsa lievettä ja antoivat hänelle lantin. Hän oli vanha,
valkopartainen mies ja istui rukousnauha sylissään lukien Koraanin
suria. Tangerissa ei ollut ketään, joka ei olisi suudellut hänen
dzellabiaan ja antanut hänelle viittä senttimoa, paitsi minä. Kun
kaukaisten kylien ihmiset tulivat, oli tapanani mennä erääseen
paikkaan lähelle hänen pienen valkean talonsa ovea ja katsella
hänelle tulevaa rahavirtaa. Päivänä muutamana, kun aurinko oli
hyvin kuuma ja minä olin viipynyt kauan sen jälkeen, kun viimeinen
vieras jo oli mennyt, hadsi kutsui minut luokseen ja minä istuuduin
maahan hänen eteensä. Hän katsoi minuun mitään puhumatta, siveli
vain hitaasti valkeata partaansa. Pitkän aikaa hän istui siten silmät
tutkien minun sieluani. »Poikani», sanoi hän viimein, »mikä on sinun
nimesi?» »Abdul as Israel», minä vastasin. »Abdul», hän sanoi,
»monet tuovat minulle lahjoja, mutta sinä et milloinkaan». »Jumalan
ja Hänen Profeettansa kautta», vannoin minä, »olen köyhä mies,
joka usein kärsin; minulla ei ole ystäviä». »Mitä minulle puhut, on
valhetta», sanoi pyhä mies, sitten hän taas oli vaiti. Viimein hän
puhui. »Rukoiletko, Abdul?» kysyi hän. »Neljästi päivässä», minä
vastasin. »Rukoilet neljästi päivässä, mutta joka päivä uudessa
paikassa», ja hän viittasi kädellään näin.

Abdul Asrael vei kätensä hitaasti silmiensä ohi.