Journal of Stored Products Research 88 (2020) 101664

Contents lists available at ScienceDirect

Journal of Stored Products Research

journal homepage: www.elsevier.com/locate/jspr

Hermetic storage technologies reduce maize pest damage in

smallholder farming systems in Mexico
~ o , Ariel Rivers 1, Nele Verhulst *
Sylvanus Odjo , Juan Burguen
n, Texcoco, Mexico
International Maize and Wheat Improvement Center (CIMMYT), Carretera Mexico-Veracruz Km 45, El Bata

a r t i c l e i n f o a b s t r a c t

Article history: In Mexico, smallholder farmers use a variety of technologies to store their maize grain for several
Received 9 May 2020 months, which may result in high losses in quantity and quality of grain. This work compared the
Received in revised form effectiveness of different storage technologies for minimizing losses in smallholder conditions in 109
11 June 2020
different locations from 21 to 2816 m above sea level (asl) across different agroecological zones of
Accepted 11 June 2020
Mexico, under “controlled” (i.e. managed by researchers), and “non-controlled” conditions (i.e. on-farm
Available online 20 July 2020
managed by extension agents). Depending on the common practice at each site, conventional storage
technologies (polypropylene bag with and/or without insecticide) were compared to alternative storage
Keywords:
Corn
technologies (selected from hermetic metal silos, hermetic bags, recycled plastic containers, silage plastic
Postharvest losses bags, and inert dusts-micronized and standard lime) during one to 12 months. Data on grain damages
Storage were collected at the beginning and end of the storage period. Climatic variables and initial grain
Polypropylene bag infestation with pests influenced the ability of a technology to minimize losses, particularly under
Hermetic technologies tropical conditions. After six months of storage, percentages of insect-damaged grain with polypropylene
bags, the most common farmers’ practice, were 39.4% and 4.1%, respectively, in lowlands (<500 m asl)
and highlands (>2000 m asl). With hermetic metal silos, percentages of insect-damaged grain after six
months of storage were 3.8% on average in the highlands and similar in lowlands, with 2.9%. Hermetic
technologies, which prevent the introduction of oxygen, were effective in reducing losses under farmers’
conditions across agroecological areas, regardless of storage time. Recycled hermetic containers had
similar results and were a viable low-cost alternative to more expensive options like hermetic metal
silos. With adequate technical support for their appropriate use, hermetic technologies have the po-
tential to reduce grain losses during storage and strengthen food security in Mexico and Latin American
countries with similar conditions.
© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction smallholders face many challenges, including postharvest losses,

Agriculture in Mexico is characterized by the coexistence of an
intensive and large-scale sector located mainly in the country’s
northwestern regions and a “traditional” small-scale sector gener-
ally encountered in the central and southern regions that uses
fewer inputs and mainly produces for family consumption
(Sweeney et al., 2013). Maize (Zea mays L.) is the main staple crop in
Mexico, produced by millions of smallholders mostly under rain-
fed conditions (Appendini and Quijada, 2016). In this context,
* Corresponding author.
E-mail addresses: arielrivers@yahoo.com (A. Rivers), n.verhulst@cgiar.org
(N. Verhulst).
1
Present address: Fin del Río Consulting PO Box 2571 Livermore, CA 94551 USA.
which can reach 40% of harvests in Mexico’s tropical regions
(García-Lara and Bergvinson, 2007). These postharvest losses are
directly associated with food insecurity and poverty (Bellon et al.,
2005; Stathers et al., 2013).
From on-farm primary production to home consumption, grains
go through several stages, including drying, harvesting, transport,
shelling, and storage, during which improper handling in associa-
tion with abiotic (ambient temperature, relative humidity) and
biotic factors (insects, fungi, rodents) may lead to losses at any
point of the process (Bradford et al., 2018; Kumar and Kalita, 2017).
Generally, in a smallholder farming system, farmers let their grain
dry in the field, which may result in insect infestation (Tigar et al.,
1994). In such a system, the timing of harvest affects losses
(Borgemeister et al., 1998) due to the possible development of fungi
and the production of mycotoxins such as aflatoxin (a potent liver

https://doi.org/10.1016/j.jspr.2020.101664
0022-474X/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
2 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
Abbreviations
AGMI initial number of live S. cerealella
AGMF final number of live S. cerealella
Asl above sea level
FunI initial percentage of fungi-damaged grain
FunF final percentage of fungi-damaged grain
GMCI initial grain moisture content
GMCF final grain moisture content
GTI initial grain temperature
GTF final grain temperature
HBT GrainPro Hermetic SuperGrainbag® Farm
HBZ GrainPro hermetic SuperGrainbag® Premium RZ
with zip
HMS hermetic metal silo
Hrmax monthly average maximum relative humidity
Hrmin monthly average minimum relative humidity
InsI initial percentage of insect-damaged grain
InsF final percentage of insect-damaged grain
LLGBI initial number of live P. truncatus
LLGBF final number of live P. truncatus
carcinogen) if harvest occurs at a moment when the grain is too
moist. When insect and fungi infestations start in the field, these
pests may develop and increase postharvest losses without an
appropriate threshing-cleaning process and adequate storage
conditions. In Mexico, farmers store their grain during several
months using a variety of storage technologies, including tradi-
tional wooden structures or polypropylene bags, sometimes in
combination with insecticides (García-Lara and Bergvinson, 2007;
Manuel Rosas et al., 2007).
Several studies, mainly in Sub-Saharan Africa, have demon-
strated the effectiveness of hermetic storage technologies in con-
trolling postharvest pests (Sheahan and Barrett, 2017). These
technologies work as a barrier that stops exchanges of oxygen and
moisture between the stored grain and its environment. Biological
activity of the living organisms inside the grain consumes oxygen
and produces carbon dioxide. As a result, most insects in hermet-
ically sealed containers become inactive (feeding activity, insect
growth) due to the low oxygen level and eventually die by desic-
cation (Murdock et al., 2012). Hermetically sealed containers such
as metal silos or improved plastic bags are suitable for smallholders
by maintaining grain at safe storage moisture content without
pesticides (Bern et al., 2013). These technologies are increasingly
used worldwide, especially in Sub-Saharan Africa, Asia, Central, and
Latin America (De Groote et al., 2013).
There is a need for studies under smallholders’ conditions and
farmers’ management practices across different agroecological re-
gions, considering interactions with environmental parameters and
initial grain conditions, particularly in Mexico and Central America.
Through the POSTCOSECHA program, the use of metal silos (in
combination with aluminum phosphide tablets) in Central America
has been promoted from the early 1990s until the beginning of the
2000s (Hellin and Kanampiu, 2008). In Mexico, a more advanced
version of metal silos with a superior conical cup and a bronze plug
that works solely on the principle of hermeticity has been devel-
oped and summarized in a Mexican norm (NMX-FF-123-SCFI-2015
Norma Mexicana, 2015). However, data on the effectiveness of this
hermetic metal silo, in comparison with other storage technologies,
is still lacking.
Although the economic effectiveness of hermetic technologies
has been demonstrated (Ndegwa et al., 2016), the initial acquisition
LMWI initial number of live S. zeamais
LMWF final number of live S. zeamais
NA not applicable
PP polypropylene bag
PP_AP polypropylene bag with aluminum phosphide
PP_DM polypropylene bag with deodorized malathion
PP_ML polypropylene bag with micronized lime
PP_SL polypropylene bag with standard lime
PBA plastic barrel
PBO plastic bottle
Q1 storage time for one to three months in “non-
controlled” experiments
Q2 storage time for four to six months in “non-
controlled” experiments
Q3 storage time for seven to ten months in “non-
controlled” experiments
SPB silage plastic bag
Tmax monthly average maximum temperature
Tmin monthly average minimum temperature
WL weight loss
costs are too high for many smallholders. An alternative is to use
recycled and hermetic containers such as sealed metal drums to
control pests (Murdock et al., 2003). Another way to minimize
losses during storage could be to use inert dusts that are non-toxic
and unreactive in nature but effective in controlling postharvest
pests when mixed with grains. These particles act as irritating,
abrasive, and desiccating agents that damage insect cuticular
waxes, which leads to their death through desiccation. Inert dusts
include clay, sand, rock phosphate, ashes, diatomaceous earth, and
_
synthetic silica (Fields and Korunic, 2002; Krzyzowski et al., 2019).
The Aztecs of ancient Mexico used lime to control postharvest pests
(Golob, 1997).
This study was conducted for two years in different agroeco-
logical zones of Mexico to address questions regarding the effec-
tiveness of storage technologies for reducing postharvest losses in
maize: (1) Which storage technologies are effective in avoiding
postharvest losses regardless of the environmental conditions? (2)
Do the initial grain conditions have an impact on the effectiveness
of storage technologies? (3) Are improved storage technologies
effective under smallholder farmer conditions across agroecologi-
cal zones and using farmers’ practices?
2. Materials and methods
2.1. Study sites and description of experiments
CIMMYT and its network of local collaborators conducted the
experiments in 2017 and 2018 in different agroecological areas of
Mexico from elevation of 21 to 2816 m asl. On the chosen experi-
mental sites, most farmers produced and stored their grain through
family farming practices and identified postharvest losses as a
significant concern. Most experimental sites were in temperate and
tropical savanna agroecological areas (Fig. 1).
Controlled and non-controlled experiments evaluated storage
technologies. Under “controlled conditions”, researchers closely
followed-up on the experiment throughout the storage period and
included three repetitions of each type of storage technology.
CIMMYT and its local collaborators conducted experiments under
controlled conditions in six different sites in 2017 and again in
2018 at the same sites (short municipal name of the site mentioned
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664 3

Fig. 1. Map of study sites and agroecological zones.

in parenthesis will be used throughout the rest of the paper): experiments under “non-controlled conditions” in 47 sites in 2017
Texcoco de Mora (Texcoco), at 2282 m asl in the state of Mexico; (Fig. 1) with ten sites in semi-arid conditions, four sites in sub-
Zacualtipa 
n de Angeles n) at 2070 m asl in the state of
(Zacualtipa tropical conditions, 16 sites in temperate conditions, three sites in
Hidalgo; San Juan Cotzoco n (Cotzoco
n) at 113 m asl, Santo Domingo tropical humid conditions and 14 in tropical savanna conditions. In
n (Yanhuitla
Yanhuitla n) at 2138 m asl, and San Pedro Pochutla 2018, CIMMYT collaborators conducted experiments in "non-
(Pochutla) at 202 m asl in the state of Oaxaca; and Peto (Peto) at controlled conditions” in 56 sites with four sites in semi-arid con-
36 m asl in the state of Yucat
an. The average monthly temperature ditions, four sites in subtropical conditions, 23 sites in temperate
during storage time varied from 7.7 to 34.4, while the average conditions, seven sites in tropical humid conditions and 18 sites in
monthly relative humidity varied between 39.9 and 98.1% (Table 1). tropical savanna conditions (Table A1). Each site included one
Extension workers trained by CIMMYT conducted on-farm storage container of each type.

Table 1
Climatic characteristics and grain varieties evaluated in “controlled” experimental sites in 2017 and 2018.

Site Elevation (m Type of Average climatic data during storage Varieties evaluated
asla) climate
Year Minimum Maximum Minimum relative Minimum relative Type Name (company,
temperature (
C) temperature (
C) humidity (%) humidity (%) country)

Texcoco 2282 Temperate 2017 7.7 ± 2.87 21.2 ± 2.96 39.3 ± 19.84 87.5 ± 7.85 Hybrid HC-8 (Aspros,
Mexico)
2018 9.1 ± 1.62 21.3 ± 2.90 44.9 ± 20.47 88.3 ± 7.73 Hybrid Albatros (Asgrow,
Mexico)
n 2138
Yanhuitla Temperate 2017 9.4 ± 2.72 21.3 ± 2.96 49.5 ± 14.18 96.8 ± 2.17 Hybrid Ares (Unisem,
Mexico)
2018 11.4 ± 0.89 20.7 ± .1.94 60.4 ± 9.59 98.1 ± 0.44 Native Jaltepec
n 2070
Zacualtipa Temperate 2017 14.4 ± 1.61 24.8 ± 2.90 59.4 ± 11.39 97.4 ± 0.97 Native e
2018 14.5 ± 0.87 25.2 ± 1.32 58.6 ± 7.84 97.5 ± 0.89 Native e
Pochutla 202 Tropical 2017 21.3 ± 0.99 26.6 ± 1.33 63.7 ± 13.39 87.2 ± 3.79 Hybrid H-565 (Inifap,
savanna Mexico)
2018 21.6 ± 0.50 27.0 ± 1.04 61.5 ± 6.62 87.3 ± 2.77 Hybrid VS-536 (Inifap,
Mexico)
n
Cotzoco 113 Tropical 2017 19.4 ± 2.81 28.5 ± 3.81 61.1 ± 6.32 95.5 ± 1.17 Hybrid 7575 (Asgrow,
humid Mexico)
2018 20.2 ± 2.00 30.4 ± 3.19 56.1 ± 9.21 94.5 ± 1.97 Hybrid Nv 15 (Novasem,
Mexico)
Peto 36 Tropical 2017 21.7 ± 1.89 34.4 ± 1.58 40.9 ± 12.25 96.1 ± 3.11 Hybrid Dk-390 (Dekalb,
savanna Mexico)
2018 21.7 ± 0.60 33.9 ± 1.38 44.5 ± 7.64 98.1 ± 1.24 Native Tsit bacal
a
asl: above sea level.
4 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
2.2. Description of evaluated storage technologies
A variety of storage methods are available in Mexico; this study
evaluated four categories of storage technologies: conventional
storage technologies, alternative conditioning agents, hermetic
technologies, and alternative hermetic technologies.
Conventional storage technologies included all storage tech-
niques commonly used by farmers, mainly polypropylene bags
(PP), which are usually used to store around 50 kg of grain. In areas
where postharvest pests pressure is high, aluminum phosphide
tablets are used inside polypropylene bags (PP_AP). The manufac-
turers generally suggest using 4e6 tablets per ton of grain. One
tablet weighs approximately 3 g and contains 56% of aluminum
phosphide. However, since farmers in Mexico commonly apply one
tablet per 50 kg of grain (Gonza lez Regalado et al., 2017; Odjo et al.,
2018), this dose was used in all experiments. In some areas, mainly
in southern Mexico, aluminum phosphide was replaced by deo-
dorized malathion (PP_DM) in an insecticide powder formulation
at 4e5% of the active ingredient and at a dose of approximately 50 g
per 50 kg of grain. The experiments used these conventional
methods as controls.
Alternative conditioning agents grouped all methods that use
grain-conditioning agents, other than chemical insecticides, that
potentially have an insecticidal effect and are used in combination
with polypropylene bags. This includes inert dusts: standard
(PP_SL) and micronized lime (PP_ML) (calcium hydroxide), both
used at a dose of 4 kg/ton of grain (Maredia and Mihm, 1995; Odjo
et al., 2018). Micronized lime is reportedly more efficient because of
its particle size, as it adheres well to grain and insect cuticles and is
more likely to enter insect spiracles. All the doses of grain treat-
ments used in these experiments were locally recommended.
Hermetic technologies are storage structures that are sealed to
create a modified atmosphere inside the containers, which results
in the control of pests. This study evaluated three hermetic tech-
nologies: hermetic metal silos (HMS) and two types of hermetic
bags marketed by GrainPro Inc. The HMS is a cylindrical tank,
topped with a cone made of galvanized sheet, and with capacities
ranging from 25 to 2000 kg; common sizes in this study were 25,
100, 500 and 1000 kg. The hermeticity of the silos was verified
before usage, mainly using water or compressed air. Burning can-
dles inside the HMS after filling with grain depleted the oxygen, as
described by De Groote et al. (2013). The present study also eval-
uated GrainPro Inc. hermetic plastic bags, the only hermetic bags
commercially available in Mexico. The bags are made of poly-
ethylene multilayer plastic that prevents gas exchange and there-
fore limits the introduction of oxygen once sealed. Both hermetic
bags can store between 10 and 69 kg of grain, but the two bags
differ in how they are sealed: (1) the GrainPro Hermetic Super-
Grainbag® Farm (HBT) is sealed by twisting the bag closed, after
which a wire and piece of tire (to prevent damage to the bag) are
tied around the closure; (2) the GrainPro Hermetic SuperGrainbag®
Premium RZ (HBZ) is sealed by a zip system. The bags’ hermeticity
was verified before usage by inflating them and searching for any
perforation. Both types of hermetic bags were placed inside a
polypropylene bag for additional protection.
Alternative hermetic technologies are other plastic containers,
such as recycled plastic bottles and plastic barrels, and work under
the same principle of hermeticity. Plastic bottles (PBO) usually used
for water are one example and are sometimes used to store seeds.
The PBO used is this study had a capacity ranging from 10 to 20 L
(8e17 kg of maize grain). Alternatively, larger storage options exist
in the form of plastic barrels (PBA) with a capacity of around 200 L
(around 180 kg of grain) that are adequately sealed. Additionally,
silage plastic bags (SPB) with a capacity of approximately 50 kg
offer a similar method for storing grain as GrainPro Inc. bags but are
made of lower quality plastic and thus more affordable.
Table 2 provides a summary of the storage technologies evalu-
ated in this study. Polypropylene bags, plastic bottles, barrels and
silage bags, aluminum phosphide tablets, insecticide powder of
deodorized malathion, standard (74% of calcium hydroxide), and
micronized (95% of calcium hydroxide lime) were obtained in local
markets. Hermetic metal silos were purchased from trained local
blacksmiths who followed the Mexican norm for manufacturing
hermetic silos (NMXeFFe123-SCFI-2015) while local providers of
GrainPro Inc. products provided the hermetic bags.
2.3. Details of experiments and data collection
Each experiment compared the conventional storage technol-
ogy prevalent in the area to other storage technologies. Climatic
data during storage time were obtained using the business.
weather.com algorithm. Monthly average data (minimum temper-
ature, maximum temperature, minimum relative humidity,
maximum relative humidity) were obtained for the months during
which grain was stored at each site and averaged to obtain average
climatic data during storage. Grain was cleaned and screened
before storage, using sieves with a 2.5e5 mm mesh size, depending
on the size of the grain, to remove impurities, crop residues and
most live insects and fragments of broken grain.
Under controlled conditions, depending on the area’s post-
harvest practices, maize grains of native or hybrid varieties were
evaluated during 1, 3, 4, 6, 9, and/or 12 months. The selection of
grain variety used in each experiment depended on the grain
available locally and common farmer practice in the area. Local
collaborators provided grains used in controlled experiments while
participating farmers provided grains used in non-controlled ex-
periments for the time they were willing to store their grain. Ex-
periments under non-controlled conditions evaluated storage
technologies between 36 and 308 days, with 173 storage containers
evaluated during one to three months (Q1), 193 during four to six
months (Q2), and 72 assessed during seven to ten months (Q3).
The moisture content and temperature of grain were taken in
triplicate at the beginning of each experiment using a hand-held
grain moisture tester (John Deere Moisture Check Plus Grain
Moisture, Illinois, USA). Before filling the storage containers, three
representative maize samples (500e1000 g) were obtained to
quantify damages. As described by Boxall (2002), “grain damage”
refers to physical spoilage and apparent evidence of deterioration,
such as holes, cracks, crevices, and discoloration. Insects, fungi, and
rodents are the main agents causing this deterioration (Kumar and
Kalita, 2017). Trained collaborators visually checked samples to
identify different types of damages. (1) Insect damaged grain was
grain that presented perforations of galleries caused by insects.
Internal grain damage caused by larvae feeding is also considered
within insect-damaged grain and was evaluated using visible light.
(2) Fungi damaged grain was grain that, on the surface, the germ or
embryo, presented partial or total development of fungi mycelium,
generally characterized by discoloration (bluish, blackish, green,
orange or yellowish). (3) Grain with other types of damage pre-
sented any physical or chemical alteration caused by other agents,
e.g. rodent-damaged grain (grain with rodent bites), heat-damaged
grain, broken grain, or grain other than maize grain.
Before identifying the damage, samples were weighed, and the
percentage of impurities (including all organic and inorganic ma-
terials other than maize, or part of grains that sieved through a
4.5 mm mesh) was determined. The grain was separated for each
kind of damage and the number and weight of damaged grain were
recorded. As some grains had more than one type of damage (for
example, insect and fungi damage), these grains were weighed in
each damage category. These data were used to estimate the
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664 5

Table 2
Postharvest storage technologies evaluated in of “controlled” and “non-controlled” experimental sites.

Type of storage technologies Storage technologies Abbreviation Grain conditioning Number of evaluation
sites

Controlled Non-
controlled

Conventional method Polypropylene bag PP NA 4 45

Polypropylene bag with aluminum phosphide tablet PP_AP Aluminum phosphide (20 tablets/ton of 3 19
grain)
Polypropylene bag with deodorized malathion PP_DM Malathion (1 kg/ton of grain) 0 4
Alternative conditioning Polypropylene bag with standard lime (calcium PP_SL Calcium hydroxide (4 kg/ton of grain) 1 5
agents hydroxide)
Polypropylene bag with micronized lime (calcium PP_ML Calcium hydroxide (4 kg/ton of grain) 3 15
hydroxide)
Hermetic technology GrainPro hermetic SuperGrainbag® Farm HBT NA 5 33
GrainPro hermetic SuperGrainbag® Premium RZ with HBZ NA 3 46
zip
Hermetic metal silo HMS NA 6 69
Alternative hermetic Silage plastic bag SPB NA 1 10
technology Plastic bottle PBO NA 4 0
Plastic barrel PBA NA 0 5

NA: not applicable.

percentage of each type of damage by dividing the weight of the

damaged kernels by the initial weight of the sample. The per-
centage of total damaged grain was obtained by weighing all
damaged grain together and dividing by the initial weight of the
sample.
The numbers of live maize weevils (Sitophilus zeamais Mot-
schulsky) (Coleoptera: Curculionidae), large grain borers (Proste-
phanus truncatus Horn) (Coleoptera: Bostrichidae), and Angoumois
grain moths (Sitotroga cerealella Olivier) (Lepidoptera: Gelechiidae)
were recorded by counting them in the collected representative
sample and expressed as number of insects per 500 g. The present
study focused on these three insects because they are the most
important postharvest pests in Mexico (García-Lara and
Bergvinson, 2007).
The average moisture content at the beginning of the experi-
ments was, 11.4% and 11.3%, respectively, under controlled and non-
controlled conditions (Table 3). Total damages varied between 5.1%
and 8.7%. Live insects were found, mainly S. zeamais and sometimes
P. truncatus. The presence of the Angoumois grain moth was
recorded under non-controlled conditions.
At the end of storage time, a representative sample of around
500e1000 g was obtained from each container, and the same set of
analyses done on the initial samples was completed (temperature,
moisture content, grain damage, and insect counts). The data ob-
tained were used to estimate weight loss (WL) as suggested by
Boxall (1986) and using the following equation (1):
ðWu Nd  Wd Nu Þ*100
WL ¼ (1)
ðWu ÞðNu þ Nd Þ
where Wu represents the weight of undamaged grains, Wd, the
weight of damaged grains, Nu the estimated number of undamaged
grains within the sample, and Nd the number of damaged grains.
2.4. Statistical analyses
Descriptive statistics such as mean and standard deviations
were used to summarize the data. Only data collected before and
after six months of storage were presented for experiments under
controlled conditions, while data from non-controlled experiments
were grouped in three categories depending on storage time: 1e3
months (Q1) for short-term storage, 4e6 months (Q2) for mid-term
storage, and more than 6 months (Q3) for long-term storage.
Partial least squares, also known as projection to latent struc-
tures, is a robust, multivariate linear regression technique suitable
for analyzing noisy and highly correlated data. To study the effect of
different factors, including storage technologies on pest damage
under controlled conditions, a response matrix and a predictive
matrix were built. Both matrices have as rows the different com-
binations of site by year for each storage technology. The columns
of the response matrix correspond to the percentages of insect- and
fungi-damaged grain and the weight loss at the end of storage time.
The square root of the number of live S. zeamais, P. truncatus, and
S. cerealella was also included. The predictive variables were storage
technology, variety (native or hybrid), storage time, monthly
average minimum and average maximum temperature, monthly

Table 3
Average (±standard error) grain quality parameters at the beginning of the experiments.

Grain quality parameters Experiments under controlled conditions (N ¼ 12) Experiments under non-controlled conditions (N ¼ 103)


Temperature ( C) 26.8 ± 1.58 24.2 ± 0.51
Moisture content (%, wet basis) 11.4 ± 0.48 11.3 ± 0.22
Grain impurities (%) 0.3 ± 0.10 0.7 ± 0.14
Insect damage (%) 1.7 ± 0.60 1.0 ± 0.18
Fungi damage (%) 3.6 ± 0.87 2.2 ± 0.24
Total grain damage (%) 8.7 ± 1.38 5.1 ± 0.45
Average live S. zeamais (number per 500 g) 2.0 ± 0.75 2.3 ± 0.46
Average live P. truncatus (number per 500 g) 0.2 ± 0.13 0.3 ± 0.14
Average live S. cerealella (number per 500 g) 0 0.1 ± 0.04
6 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
average minimum and average maximum relative humidity, initial
grain moisture content, initial grain temperature, initial count of
live S. zeamais, initial count of P. truncatus (initial count of live
S. cerealella was not considered because all values were zero), and
initial percentages of insect-damaged and fungi-damaged grains.
In the case of controlled experiments, previous to partial least
squares analysis, each individual variable was analyzed by a
generalized linear mixed model to obtain the adjusted mean in
each treatment and environment (variety, site, and storage time).
Percentages were analyzed assuming a Beta distribution and using
the Napierian logarithm (ln) as the link function. Variables with
values of zero were transformed, adding 1E-5. The minimum least
square means for each treatment in each environment were used in
partial least squares analysis.
The same analysis was done of the non-controlled data with the
same variables except weight loss, which was not estimated in non-
controlled experiments. Not all the variables were included in the
results; some of them were discarded because their contribution to
the models was negligible. The final number of factors was decided
using a cross-validation approach one-at-a-time and selecting the
number of factors with the lowest minimum root mean of the
predicted residual error sum of squares statistic. The results of
partial least squares were shown in correlation loading plots, which
summarize the characteristics of the two factors models. Variable
importance plots based on the variable importance for projection
statistic of Wold (1995) were also presented. As suggested by Wold,
a variable importance for projection value of less than 0.8 is
considered “small”. All analyses were performed with SAS software
version 9.4, using SAS/GLIMMIX, SAS/PLS, and SAS/SGPLOT soft-
ware capabilities, version 2015.
3. Results
3.1. Effectiveness of storage technologies under controlled
conditions
The percentages of insect damage, total damage, and weight loss
followed the same trend for each storage technology (Table 4).
Grain stored in polypropylene bags had more damage than grain
stored in other methods, especially in tropical conditions, with
39.4%, 48%, and 8%, respectively, of insect-damaged grain, total
damaged grain, and weight loss in Peto, averaged over 2017 and
2018. Large numbers of live S. zeamais and P. truncatus were also
found in general with this technology. However, in some experi-
ments, particularly in the highlands, the percentage of insect-
damaged grain with polypropylene bag was low. In Texcoco, in
temperate conditions, damages over the two years’ experiments
were, on average, 2.2%, 8.3%, and 0.3% respectively, of insect-
damaged grain, total damaged grain, and weight loss. Overall, in-
sect damages and weight loss with polypropylene bags on sites
>2000 m asl (Texcoco, 2017; Texcoco, 2018; Yanhuitl an, 2017;
Zacualtipan, 2018) averaged 4.1% and 0.4%, respectively, while the
same damages on sites <500 m asl (Peto, 2017, Peto, 2018) were
39.4% and 8%, respectively.
Damages with aluminum phosphide tablets in a polypropylene
bag were variable. In Yanhuitlan in 2017, for example, damages to
grain stored by this method were low, with 0.7% insect-damaged
grain, 8.1% total damaged grain, and 0.1% weight loss, similar to
damages without insecticide at this temperate site. However, in
tropical conditions in Cotzoco n in 2017, for example, grains
exhibited 46.3%, 59.8%, and 4.8% insect-damage, total damage, and
weight loss, respectively. Overall, the use of aluminum phosphide is
less effective in limiting damages in experimental sites below
500 m asl (Cotzoco  n, 2017, Cotzoco n, 2018; Pochutla, 2017) with
21.3% of insect-damaged grain and 3.5% of weight loss in
comparison with experimental sites >2000 m asl (Yanhuitla n, 2017,
Yanhuitlan, 2018) where percentages of insect-damaged grain and
weight loss were respectively 4.3% and 0.2%.
The use of standard lime in combination with a polypropylene
bag had variable results, with 0.7%, 2.8%, and 0.2% of insect-
damaged, total damaged, and weight loss, respectively, in
Pochutla in 2017. However, in the same area in 2018, damages were
higher, with 36.7%, 41.3%, and 8.2% of insect-damaged grain, total
damaged grain and weight loss of grain, respectively, and large
quantities of live S. zeamais and P. truncatus. Results with micron-
ized lime were variable as well. For example, in Peto in 2017
damages were low (2.9%, 7.3% and 0.6% of insect-damaged grain,
total damaged grain, and weight loss respectively), while damages
were higher in 2018 with 27.1% of insect-damaged grain, 30.7% of
total damaged grain and 6.9% of weight loss.
Silage plastic bags and plastic bottles were very effective in
controlling pests, with the maximum percentage of insect-
damaged grain and weight loss being 1.2% and 0.2% for silage
plastic bags and 1.5% and 0.2% for plastic bottles. Upon opening,
very few live insects were found in both technologies. Hermetic
technologies (GrainPro hermetic bags and hermetic metal silos)
showed similar results regardless of the areas where experiments
were performed. Average damages on sites over 2000 m asl (Tex-
coco, 2017, Texcoco, 2018; Yanhuitla n, 2017, Yanhuitlan, 2018;
Zacualtipan, 2017, Zacualtipa n, 2018) for hermetic metal silo were
low with 3.8% of insect-damaged grain and 0.4% of weight loss,
similar to results on sites below 500 m asl (Cotzoco  n, 2017,
Cotzoco n, 2018; Pochutla, 2017, Pochutla, 2018; Peto, 2017, Peto,
2018) with 2.9% and 1.3% of insect-damaged grain and weight loss,
respectively. There were, however, some notable exceptions. The
percentage of insect-damaged grain for hermetic metal silos was
13.5% in Zacualtipa n (2017), likely attributable to the high number
of live S. zeamais inside the silos. A high percentage of insect-
damaged grain (8.1%) was also found under the same conditions
for GrainPro Hermetic SuperGrainbag® Farm with however a low
percentage of weight loss (0.7% of weight loss for GrainPro Her-
metic SuperGrainbag® Farm and 1.3% of weight loss for hermetic
metal silo in 2017).
Percentages of fungi-damaged grain ranged between 0.0 and
13.4% across experimental sites (Table 4). The effect of storage
technologies on the percentage of fungi-damaged grain varied
across sites and years. In the case of Texcoco, the experimental site
at the highest elevation, for example, percentages of fungi-
damaged grain did not significantly vary between storage tech-
nologies in 2017 and 2018. In the case of Peto, the experimental site
at the lowest elevation, percentages of fungi-damaged grain were
3.1%, 3.8%, 3.6%, and 2.6%, respectively, for polypropylene bag,
polypropylene bag with micronized lime, GrainPro hermetic
SuperGrainbag® Farm and hermetic metal silos in 2017, and 12.8%,
1.1%, 0.0%, 0.5% and 0.9% for polypropylene bag, polypropylene bag
with micronized lime, plastic bottle, GrainPro hermetic Super-
Grainbag® Farm and hermetic metal silos, respectively, in 2018. In
Peto and Cotzoco  n in 2017, under tropical conditions, conditioning
grain with micronized lime increased the percentage of fungi-
damaged grain. A slight increase in the percentage of fungi-
damaged grain was sometimes observed with hermetic metal
silos, for example, in Zacualtip an (2018), Yanhuitla n 2017, and
Yanhuitlan 2018.
3.2. Effectiveness of storage technologies under non-controlled
conditions
The results were highly variable in the non-controlled condi-
tions (Table 5), mainly due to the diverse conditions of technology
evaluation. Similar to the controlled experiments, the
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664 7

Table 4
Average grain damage (±standard deviation) during six months of storage under controlled conditions.

Storage technology Insect damaged grain (%) Fungi damaged grain (%) Total damaged grain (%) Weight loss (%) Live insects per 500 g
and experiment
S. zeamais P. truncatus S. cerealella

Initial average value 1.7 ± 2.07 3.6 ± 3.02 8.7 ± 4.79 e 2.0 ± 2.59 0.2 ± 0.45 0
(N ¼ 12)
Polypropylene bag (PP), N ¼ 3
Texcoco, 2017 0.2 ± 0.19 5.1 ± 0.41 7.8 ± 0.54 0.1 ± 0.05 0 0.7 ± 1.15 0
Texcoco, 2018 4.2 ± 1.44 2.0 ± 0.41 8.9 ± 1.92 0.5 ± 0.04 0 4.7 ± 2.52 0
Yanhuitlan, 2017 0.8 ± 0.13 2.1 ± 0.34 5.9 ± 0.68 0.2 ± 0.11 4.7 ± 0.58 0 0
Zacualtipan, 2018 11.4 ± 7.04 9.1 ± 5.24 37.2 ± 10.35 1.0 ± 0.65 98.7 ± 40.08 1.0 ± 1.00 0
Peto, 2017 20.0 ± 5.69 3.1 ± 0.10 23.6 ± 5.60 4.6 ± 1.31 1.3 ± 2.31 2.0 ± 2.65 0
Peto, 2018 58.8 ± 7.50 12.8 ± 12.08 72.4 ± 5.33 11.4 ± 1.46 205.7 ± 48.58 12.7 ± 7.64 0
Polypropylene bag with aluminum phosphide tablet (PP_AP), N ¼ 3
Yanhuitlan, 2017 0.7 ± 0.26 3.3 ± 1.27 8.1 ± 1.44 0.1 ± 0.05 1.0 ± 1.00 0 0
Yanhuitlan, 2018 8.0 ± 2.22 7.9 ± 0.36 19.1 ± 1.47 0.4 ± 0.26 4.0 ± 2.00 0 0
Pochutla, 2017 15.5 ± 4.06 1.3 ± 1.54 17.7 ± 3.36 5.4 ± 1.42 139.3 ± 94.35 2.3 ± 2.52 0
Cotzoco n, 2017 46.3 ± 20.03 10.8 ± 7.48 59.8 ± 14.43 4.8 ± 1.14 233.3 ± 186.87 7.0 ± 12.12 0.3 ± 0.58
Cotzoco n, 2018 2.2 ± 0.34 4.4 ± 0.20 8.8 ± 0.59 0.4 ± 0.06 10.0 ± 12.73 0 0.5 ± 0.71
Polypropylene bag with standard lime (calcium hydroxide) (PP_SL), N ¼ 3
Pochutla, 2017 0.7 ± 0.39 0.3 ± 0.10 2.8 ± 0.57 0.2 ± 0.09 0 0.3 ± 0.58 0
Pochutla, 2018 36.7 ± 11.22 2.7 ± 2.73 41.3 ± 13.68 8.2 ± 2.50 126.7 ± 28.71 33.7 ± 10.21 0
Polypropylene bag with micronized lime (calcium hydroxide) (PP_ML), N ¼ 3
Texcoco, 2017 0.0 ± 0.02 6.3 ± 1.01 8.4 ± 1.27 0 0 0 0
Cotzoco n, 2017 5.6 ± 0.85 13.2 ± 7.63 25.1 ± 8.02 3.4 ± 1.45 40.0 ± 33.04 19.3 ± 22.37 1.0 ± 1.73
Peto, 2017 2.9 ± 1.49 3.8 ± 0.23 7.3 ± 1.67 0.6 ± 0.39 0 1.0 ± 1.00 0
Peto, 2018 27.1 ± 13.79 1.1 ± 0.64 30.7 ± 14.10 6.9 ± 4.25 195.3 ± 81.44 37.7 ± 19.86 0
Silage plastic bag (SPB), N ¼ 3
Cotzoco n, 2018 1.2 ± 0.32 4.6 ± 0.30 7.4 ± 0.17 0.2 ± 0.03 1.3 ± 1.53 0.3 ± 0.58 0
Plastic bottle (PBO), N ¼ 3
Texcoco, 2018 0.5 ± 0.26 1.8 ± 0.49 4.6 ± 0.76 0.1 ± 0.11 0 0.3 ± 0.58 0
Pochutla, 2018 1.2 ± 0.24 6.7 ± 1.64 12.9 ± 2.52 0.2 ± 0.03 0 0.3 ± 0.58 0
Cotzoco n, 2018 1.5 ± 0.35 4.4 ± 0.39 8.2 ± 0.67 0.2 ± 0.02 0 0 0
Peto, 2018 0.4 ± 0.35 0.0 ± 0.00 2.0 ± 1.78 0.1 ± 0.10 0.3 ± 0.58 0 0
GrainPro hermetic SuperGrainbag® Farm (HBT), N ¼ 3
Yanhuitlan, 2017 0.7 ± 0.11 4.3 ± 0.53 8.4 ± 1.04 0.1 ± 0.06 0 0 0
Yanhuitlan, 2018 3.5 ± 1.93 7.2 ± 1.00 14.8 ± 2.78 0.9 ± 0.52 0 0 0
Zacualtipan, 2017 8.1 ± 0.62 11.1 ± 2.93 24.6 ± 2.62 0.7 ± 0.62 3.7 ± 3.79 0 0
Pochutla, 2017 0.5 ± 0.00 0.2 ± 0.02 2.1 ± 0.13 0.3 ± 0.10 0 0 0
Pochutla, 2018 5.2 ± 1.89 2.8 ± 0.51 10.6 ± 1.96 1.6 ± 1.57 1.3 ± 1.15 1.3 ± 2.31 0
Cotzoco n, 2017 0.5 ± 0.36 2.5 ± 0.48 9.6 ± 0.57 0.0 ± 0.01 0 0 0
Cotzoco n, 2018 0.8 ± 0.33 3.0 ± 0.42 5.2 ± 0.78 0.2 ± 0.08 0 0 0
Peto, 2018 0.7 ± 0.12 0.5 ± 0.23 2.3 ± 0.64 0.4 ± 0.31 0 0 0
GrainPro hermetic SuperGrainbag® Premium RZ with zip (HBZ), N ¼ 3
Texcoco, 2017 0.0 ± 0.03 5.9 ± 0.62 8.2 ± 0.73 0.0 ± 0.04 0 0 0
Texcoco, 2018 0.8 ± 0.47 1.5 ± 0.47 4.5 ± 0.22 0.2 ± 0.11 0 0.7 ± 0.58 0
Zacualtipan, 2018 3.5 ± 3.74 7.2 ± 2.23 32.9 ± 4.89 0.3 ± 0.37 0.3 ± 0.58 0 0
Peto, 2017 0.5 ± 0.29 3.6 ± 0.53 4.9 ± 0.55 0.3 ± 0.23 0 0 0
Hermetic metal silo (HMS), N ¼ 3
Texcoco, 2017 0.1 ± 0.14 4.6 ± 0.23 6.8 ± 0.08 0.0 ± 0.02 0 0 0
Texcoco, 2018 1.1 ± 0.81 2.0 ± 0.13 5.2 ± 0.81 0.3 ± 0.19 0 0 0
Yanhuitlan, 2017 0.8 ± 0.29 4.1 ± 1.12 9.3 ± 1.30 0.1 ± 0.02 1.3 ± 0.58 0 0
Yanhuitlan, 2018 4.5 ± 0.74 6.7 ± 0.76 14.1 ± 0.42 0.1 ± 0.01 0 0 0
Zacualtipan, 2017 13.5 ± 1.65 10.7 ± 3.96 27.7 ± 6.60 1.3 ± 0.33 67.3 ± 3.21 0.7 ± 0.58 0
Zacualtipan, 2018 2.7 ± 0.78 13.4 ± 1.50 34.9 ± 12.66 0.7 ± 0.74 4.3 ± 2.52 0 0
Pochutla, 2017 0.5 ± 0.22 0.2 ± 0.03 1.9 ± 0.10 0.2 ± 0.10 1.3 ± 0.58 0 0
Pochutla, 2018 7.3 ± 0.53 2.3 ± 0.68 12.1 ± 2.14 1.9 ± 1.49 3.3 ± 2.08 3.7 ± 1.53 0
Cotzoco n, 2017 4.5 ± 1.69 2.3 ± 0.81 14.0 ± 2.28 4.5 ± 1.73 0 0 0
Cotzoco n, 2018 2.1 ± 0.41 6.0 ± 0.57 10.4 ± 0.68 0.6 ± 0.27 0 0.3 ± 0.58 0
Peto, 2017 0.9 ± 0.13 2.6 ± 0.89 4.2 ± 1.12 0.2 ± 0.02 0.3 ± 0.58 0 0
Peto, 2018 2.1 ± 0.12 0.9 ± 0.12 6.8 ± 1.25 0.5 ± 0.32 0 0 0

polypropylene bag exhibited the highest percentage of insect-
damaged grain with, on average, 12.5%, 31.7, and 32.4% for short-
term, mid-term, and long-term storage, respectively. The average
percentage of insect-damaged grain during mid-term storage with
polypropylene bags on sites <500 m asl was 51.8 ± 29.68% (N ¼ 7)
while on sites >2000 m asl it was 3.2 ± 4.86% (N ¼ 12; data not
shown). Likewise, the polypropylene bags also contained a high
number of live insects in some sites. The percentage of insect-
damaged grain was not related to storage time. Polypropylene
bag with aluminum phosphide tablets succeeded in controlling
pests for short-term storage with 1.9% of insect-damaged grain,
while an increase was recorded for mid-term and long-term stor-
age (8.7% and 7.2%, respectively). Conditioning with deodorized
malathion (PP_DM) produced consistent results for different stor-
age times, with 1.0%, 0.4%, and 0.2% for short-term, mid-term, and
long-term storage, respectively. Regarding inert dusts, standard
lime revealed higher losses than micronized lime, but both tech-
nologies resulted in high percentages of insect-damaged and fungi-
damaged grain in certain places. Overall, hermetic technologies and
alternative hermetic technologies reduced the number of live in-
sect pests upon opening of the technologies, and grain losses were
lower regardless of storage time, compared to conventional
8 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664

Table 5
Average grain damage (±standard deviation) under non-controlled conditions.

Storage Insect damaged grain (%) Fungi damaged grain (%) Total grain damaged (%) Live insects per 500 g
technologies
S. zeamais P. truncatus S. cerealella

Initial average 1.0 ± 1.86 2.2 ± 2.40 5.1 ± 4.57 2.3 ± 4.67 0.3 ± 1.40 0.1 ± 0.42
value (N ¼ 103)
Polypropylene bag (PP)
Q1 (N ¼ 30) 12.5 ± 15.31 4.7 ± 10.57 18.7 ± 21.39 23.9 ± 24.58 1.5 ± 3.69 1.7 ± 4.18
Q2 (N ¼ 28) 31.7 ± 35.58 2.9 ± 4.66 38.3 ± 36.60 114.7 ± 198.63 11.1 ± 24.15 17.8 ± 73.24
Q3 (N ¼ 18) 32.4 ± 36.61 1.0 ± 1.00 42.7 ± 38.72 10.8 ± 39.86 3.5 ± 14.12 1.0 ± 2.00
Polypropylene bag with aluminum phosphide tablet (PP_AP)
Q1 (N ¼ 11) 1.9 ± 1.81 5.6 ± 5.96 9.9 ± 7.41 4.6 ± 8.45 0.1 ± 0.30 0.3 ± 0.90
Q2 (N ¼ 21) 8.7 ± 15.36 3.3 ± 4.33 14.7 ± 15.57 31.5 ± 79.34 3.0 ± 6.16 0.0 ± 0.22
Q3 (N ¼ 3) 7.2 ± 11.48 0.4 ± 0.47 14.8 ± 24.10 0 13.0 ± 22.52 0
Polypropylene bag with deodorized malathion (PP_DM)
Q1 (N ¼ 4) 1.0 ± 1.64 1.4 ± 1.10 3.4 ± 1.66 1.0 ± 2.00 0.5 ± 1.00 0
Q2 (N ¼ 4) 0.4 ± 0.42 2.2 ± 1.36 4.0 ± 2.02 0 0.5 ± 1.00 0.5 ± 1.00
Q3 (N ¼ 3) 0.2 ± 0.12 16.0 ± 4.54 20.4 ± 4.25 0 0 0
Polypropylene bag with standard lime (calcium hydroxide) (PP_SL)
Q1 (N ¼ 4) 3.8 ± 1.57 8.7 ± 4.63 15.2 ± 2.34 0.5 ± 0.58 1.0 ± 1.41 0
Q2 (N ¼ 5) 10.3 ± 11.15 8.4 ± 7.48 21.6 ± 7.63 1.6 ± 2.51 25.2 ± 47.19 0.4 ± 0.55
Q3 (N ¼ 1) 9.4 2.6 15.1 25 0 0
Polypropylene bag with micronized lime (calcium hydroxide) (PP_ML)
Q1 (N ¼ 13) 3.1 ± 4.39 3.9 ± 6.73 7.9 ± 8.02 4.4 ± 12.91 2.5 ± 5.03 0.2 ± 0.55
Q2 (N ¼ 14) 6.9 ± 8.05 1.4 ± 1.11 12.0 ± 8.08 2.3 ± 3.29 8.7 ± 21.34 0.9 ± 1.75
Q3 (N ¼ 3) 5.9 ± 7.99 8.1 ± 3.54 18.0 ± 8.84 0 0 0
Silage plastic bag (SPB)
Q1 (N ¼ 6) 2.8 ± 2.32 1.4 ± 2.41 5.1 ± 4.56 5.0 ± 7.13 0 1.8 ± 2.14
Q2 (N ¼ 10) 6.1 ± 8.33 2.6 ± 2.80 13.9 ± 5.91 2.2 ± 3.58 7.7 ± 17.79 0
Plastic barrel (PBA)
Q1 (N ¼ 2) 5.0 ± 7.07 3.2 ± 1.13 12.6 ± 2.83 0 0 0
Q2 (N ¼ 4) 0.4 ± 0.73 2.9 ± 1.54 8.7 ± 2.79 1.5 ± 2.38 0.7 ± 0.96 0
Q3 (N ¼ 3) 0.5 ± 0.58 1.3 ± 0.61 2.9 ± 0.83 0 0 0
GrainPro hermetic SuperGrainbag® Farm (HBT)
Q1 (N ¼ 31) 1.3 ± 2.16 3.7 ± 4.62 7.9 ± 6.61 0.3 ± 1.44 0.0 ± 0.18 0
Q2 (N ¼ 29) 0.9 ± 0.93 3.6 ± 5.02 8.0 ± 6.52 0.2 ± 0.47 0 0
Q3 (N ¼ 3) 0.9 ± 0.81 0.3 ± 0.23 2.5 ± 1.92 0 0 0
GrainPro hermetic SuperGrainbag® Premium RZ with zip (HBZ)
Q1 (N ¼ 24) 0.9 ± 1.74 1.0 ± 1.41 3.1 ± 3.55 1.3 ± 3.64 0.1 ± 0.28 0.2 ± 1.03
Q2 (N ¼ 30) 1.0 ± 1.43 2.9 ± 3.14 7.0 ± 4.43 0.2 ± 0.75 0 0.0 ± 0.18
Q3 (N ¼ 19) 1.5 ± 3.83 3.9 ± 5.03 8.2 ± 9.57 0 1.6 ± 6.88 0.3 ± 1.15
Hermetic metal silo (HMS)
Q1 (N ¼ 48) 1.0 ± 2.03 2.9 ± 3.67 5.8 ± 5.87 0.7 ± 2.34 0.0 ± 0.20 0
Q2 (N ¼ 48) 1.6 ± 2.56 2.8 ± 4.29 7.2 ± 6.74 0.5 ± 1.81 0.1 ± 0.87 0.1 ± 0.58
Q3 (N ¼ 19) 0.8 ± 1.12 3.3 ± 4.34 7.8 ± 9.62 0.2 ± 0.69 0 0

Q1: storage time from one to three months; Q2: storage time from four to six months; Q3: storage time from seven to ten months.

technologies.
3.3. The effect of different parameters on postharvest losses under
controlled and non-controlled conditions
The first two partial least squares factors described 26.5% and
27.8% of the variability of the X and the Y data in the correlation
loading plot for controlled conditions (Fig. 2). This may be
explained by the high variability of the data collected in this study.
The loading plot, however, showed a strong and positive relation-
ship between the monthly average maximum relative humidity of
the experiment site (Hrmax), the initial percentage of fungi-
damaged grain (FunI), and the final percentage of fungi-damaged
grain (FunF). The loading plot revealed that weight loss (WL), the
final percentage of insect-damaged grain (InsF), and the final
number of live S. zeamais (LMWF) were strongly correlated with the
final number of live P. truncatus (LLGBF), the final number of live
S. cerealella (AGMF) and technologies like polypropylene bag,
polypropylene bag with aluminum phosphide and polypropylene
bag with micronized lime. The performance of these non-hermetic
storage technologies was also correlated to the monthly average
relative humidity of the experiment site, hence the percentage of
fungi-damaged grain. Grain variety (distinguishing between native
and hybrid maize) was positively correlated with the initial number
of live P. truncatus and negatively correlated with the initial per-
centage of fungi-damaged grains. The variable importance plot al-
lows identifying variables that contributed more than others to the
variance. Regarding the initial grain conditions, the main parame-
ters were the initial grain temperature (GTI), the initial number of
live P. truncatus (LLGBI), the initial percentage of fungi-damaged
grains (FunI), and the grain variety (Fig. 2). Regarding climatic
factors, the most important parameters were the monthly average
relative humidity (Hrmax), the monthly minimum and maximum
temperature (Tmin and Tmax). While the storage time seemed to
be an important factor, the storage technologies that contributed
the most to the explanation of predicted variables were poly-
propylene bag, GrainPro hermetic SuperGrainbag® Farm, and
polypropylene bag with aluminum phosphide (Fig. 2).
Under non-controlled conditions, the first two partial least
squares factors described 10.2% and 25.3% of the variability of the X
and the Y data in the correlation loading plot (Fig. 3). Trends similar
to those recorded under controlled conditions were also found, as
the final percentage of insect-damaged grain (InsF) was strongly
correlated with the final number of live S. zeamais (LMWF) and the
use of polypropylene bag (PP). As noted under controlled condi-
tions, the final percentage of fungi-damaged grain (FunF) was also
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664 9

Fig. 2. Partial least squares correlation loading plot (on the left) and variable importance plot (on the right) under controlled conditions.
Legend: Time: storage time; Variety: grain variety; Tmax: monthly average maximum temperature; Tmin: monthly average minimum temperature; Hrmax: monthly average
maximum relative humidity; Hrmin: monthly average minimum relative humidity; GTI: initial grain temperature; GTF: final grain temperature; GMCI: initial grain moisture
content; GMCF: final grain moisture content; LLGBI: initial number of live large grain borers; LLGBF: final number of live large grain borers; LMWI: initial number of live maize
weevils; LMWF: final number of live maize weevils; AGMI: initial number of live Angoumois grain moths; AGMF: final number of live Angoumois grain moths; FunI: initial
percentage of fungi-damaged grain; FunF: final percentage of fungi-damaged grain; InsI: initial percentage of insect-damaged grain; InsF: final percentage of insect-damaged grain;
PP: polypropylene bag; PP_AP: polypropylene bag with aluminum phosphide; PP_DM: polypropylene bag with deodorized malathion; PP_SL: polypropylene bag with standard
lime; PP_ML: polypropylene bag with micronized lime; PBO: plastic bottle; SPB: silage plastic bag; HMS: hermetic metal silo; HBT: GrainPro hermetic SuperGrainbag® Farm; HBZ:
GrainPro hermetic SuperGrainbag® Premium RZ with zip.

Fig. 3. Partial least square loading plot (on the left) and variable importance plot (on the right) under non-controlled conditions.
Legend: Time: storage time; Variety: grain variety; Tmax: monthly average maximum temperature; Tmin: monthly average minimum temperature; Hrmax: monthly average
maximum relative humidity; Hrmin: monthly average minimum relative humidity; GTI: initial grain temperature; GTF: final grain temperature; GMCI: initial grain moisture
content; GMCF: final grain moisture content; LLGBI: initial number of live large grain borers; LLGBF: final number of live large grain borers; LMWI: initial number of live maize
weevil; LMWF: final number of live maize weevils; AGMI: initial number of live Angoumois grain moths; AGMF: final number of live Angoumois grain moths; FunI: initial per-
centage of fungi-damaged grain; FunF: final percentage of fungi-damaged grain; InsI: initial percentage of insect-damaged grain; InsF: final percentage of insect-damaged grain; PP:
polypropylene bag; PP_AP: polypropylene bag with aluminum phosphide; PP_DM: polypropylene bag with deodorized malathion; PP_SL: polypropylene bag with standard lime;
PP_ML: polypropylene bag with micronized lime; PBO: plastic bottle; SPB: silage plastic bag; HMS: hermetic metal silo; HBT: GrainPro hermetic SuperGrainbag® Farm; HBZ:
GrainPro hermetic SuperGrainbag® Premium RZ with zip.
10 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
related to the initial percentage of fungi-damaged grain and the
monthly average maximum relative humidity (Hrmax). The main
parameters shown by the variance importance plots were the
initial percentage of fungi-damaged grain and the initial number of
live S. zeamais regarding initial grain conditions, the monthly
average maximum temperature, the monthly average maximum
relative humidity, and the elevation regarding parameters related
to environmental conditions (Fig. 3). The storage technologies that
were most important in predicting dependent variables under non-
controlled conditions were polypropylene bag and hermetic
technologies.
4. Discussion
Data were highly variable, due to the variability of the condi-
tions of evaluation and the high number of people involved in the
collection of data, but general trends emerge. The most common
grain storage technology for Mexico’s smallholders, polypropylene
bag, was ineffective in preventing losses caused by insects and
fungi, particularly under tropical conditions. This result is similar to
those of other studies implemented elsewhere. Polypropylene bags
are, for example, commonly used by farmers to store their grain in
sub-Saharan Africa, where these bags also generally fail to protect
grain from postharvest losses (Abass et al., 2018, 2014; Chigoverah
and Mvumi, 2016; De Groote et al., 2013).
Conditioning grain with fumigants may be a solution to avoid
pest development. In this study, the use of aluminum phosphide in
combination with polypropylene bag was somewhat effective
during short-term storage, but became less effective with pro-
longed storage time, even at a high dosage. Some insect develop-
ment stages, including eggs and pupae, may tolerate the action of
phosphine (the gas generated during the use of aluminum phos-
phide) due to the lower uptake of the gas, duration the grains are
treated with the fumigant, and distribution of the gas within grain
(Howe, 1973). Hole et al. (1976) also found that phosphine was less
effective at lower temperatures, highlighting the importance of
improving the exposure time in those cases. This may explain why
aluminum phosphide was ineffective in minimizing losses in grain
stored by smallholders: different development stages of insects
may be present in stored grain, resulting in posterior infestations of
grain after eggs and pupae hatch. The inefficiency of fumigation
with aluminum phosphine can also be explained by the develop-
ment of resistance of some strains of stored product pests. Strains of
postharvest pests resistant to phosphine have been reported in
Mexico in the past (Hole, 1981) and recently elsewhere (Afful et al.,
2018; Pimentel et al., 2009); however, the current status of popu-
lation resistant strains of common postharvest pests in Mexico is
unknown. In comparison with fumigants, the control effect of
contact insecticides seems to be permanent. The deodorized mal-
athion used in these experiments, which is still used in some areas,
consistently avoided losses the entire storage time. However, this
molecule has been classified as carcinogenic (Reuber, 1985), and its
use by smallholders should be discouraged. The use of insecticides
by smallholders in Mexico is a public health concern as intoxication
cases are frequently reported (Bernardino Herna ndez et al., 2019;
Villa-Manzano et al., 2019). This study identified safe alternatives
for storage that need to be promoted.
Inert dusts may be a low-cost and non-toxic solution that works
as contact insecticides. The results of this study are highly variable,
with inert dusts sometimes being effective against insects and
sometimes less effective, particularly in the lowlands. As inert dusts
are used with non-hermetic polypropylene bags, grain stored in
this way is still exposed to the environment. Grains may, in these
conditions, absorb moisture allowing for further fungal and insects
growth and damages. The efficacy of inert dusts, therefore, depends
on environmental conditions. Data on the use of lime to preserve
grain are scarce, as this practice seems to be limited to Mexico and
Central America (Golob, 1997). However, there are numerous re-
sults on other inert dusts, mainly diatomaceous earth, showing that
their insecticidal potential can be influenced by factors such as the
type of grain, the type of insect pest and its stage of development,
grain moisture content and storage conditions including tempera-
ture and relative humidity (Arthur, 2002; Athanassiou et al., 2007;
Fields and Korunic, 2000; Nielsen, 1998; Plumier et al., 2019). For
example, Stathers et al. (2004) found that diatomaceous earth was
less effective in preventing losses by insects when relative humidity
increased from 50 to 60%, and when storage time increased as the
dust may absorb moisture from the environment. The same result
was found with nanostructured alumina (Stadler et al., 2012), ze-
olites formulations (Eroglu et al., 2019) and wood ashes from
different tree species (Bohinc et al., 2018), highlighting the impor-
tance of temperature and relative humidity on the effectivity of
inert dusts (Korunic et al., 2016). In this study, micronized lime
failed to minimize losses under tropical humid and savanna con-
ditions (Cotzoco n and Peto), i.e., conditions of high relative hu-
midity. Other factors, including the type of grain and its oil content,
may also impact the efficacy of inert dusts (Athanassiou et al.,
2008). The use of lime and micronized lime in Mexico and Cen-
tral America is important as it is used in the nixtamalization pro-
cess. Further research is required to understand the impact of lime
and micronized lime on pests, and to determine the adequate
doses, exposure time and formulations and the impact on grain
quality parameters.
Overall, hermetic technologies and alternative hermetic tech-
nologies effectively reduced damage due to insects under
controlled and non-controlled conditions. The effectiveness of
hermetic technologies in reducing postharvest losses has been
widely demonstrated, as oxygen depletion due to biological activ-
ities inside the containers alone is sufficient to control pests (Abass
et al., 2018, 2014; De Groote et al., 2013; Mutambuki et al., 2014;
Ndegwa et al., 2016; Quellhorst et al., 2018; Tefera et al., 2011).
Reaching hermeticity is essential for hermetic technologies to be
effective and can be challenging to achieve, either because the
technologies are not manufactured correctly or because the con-
tainers are not correctly used and sealed. Results of Zacualtipa n in
the present study (i.e., relatively high losses in the hermetic tech-
nologies) showed the importance of the initial grain conditions and
hermeticity. In the case of hermetic metal silos, it is important to
train local blacksmiths on how to manufacture them, as welding is,
for example, a critical phase. Regarding all hermetic technologies, it
is important to train extension workers and farmers on the correct
sealing as this impacts effectiveness (Kharel et al., 2018). Farmers’
access to good quality postharvest technologies is important for
scaling up postharvest storage technologies while giving them the
adequate technical support on the use of these technologies.
Hermetic technologies were also effective in suppressing the
spread of fungi. Previous research in Sub Saharan Africa on her-
metic bags had similar results, as the low-oxygen environment
created by hermeticity can limit the proliferation of fungi and the
production of mycotoxins such as aflatoxins and fumonisins
(Ng’ang’a et al., 2016; Tubbs et al., 2016; Williams et al., 2014).
Additionally, relative humidity remains fairly constant in hermetic
technologies while variation is larger in non-hermetic containers
(Baoua et al., 2018). Since the initial grain moisture content was less
than 12% in this study, keeping it constant in hermetic containers
could have limited fungal growth. In the context of Mexico and
Central America with a high amount of maize consumed per capita
(Ranum et al., 2014), hermetic bags can be used to decrease the
dietary exposure of smallholders to aflatoxins and the associated
health risks (Mendoza et al., 2017). A slight increase in fungi
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
infestation in hermetic metallic silos was sometimes observed, e.g.,
in Zacualtipa n (2018), Yanhuitla n 2017, and Yanhuitla n 2018. Pre-
vious results have shown that the proliferation of fungi inside
hermetic containers is related to the grain moisture content,
especially when it is more than 14% (Ng’ang’a et al., 2016; Quezada
et al., 2006; Tubbs et al., 2016). In the present study, grain moisture
content was less than 12%, suggesting only the presence/growth of
xerophilic fungi. However, the effect of hermetic technologies on
the percentage of fungi-damaged grain with low moisture content
has been less investigated. Further research should focus on iden-
tifying the species of fungi present in grain with low moisture
content and their effect on mycotoxin production.
Partial least square analyses showed that grain moisture content
did not affect storage losses, as all of the grain evaluated had a
moisture content of less than 12%, suggesting the grain underwent
prolonged periods of field drying. As much as possible, it is
important to reduce the prior infestation of grain with insects and
fungi by choosing the right moment for harvesting and storing.
Compared to hermetic technologies, grain damage in non-hermetic
technologies was more affected by environmental parameters
(elevation, temperature, and relative humidity, three correlated
parameters as highlighted by the variable importance plots) and
initial grain conditions (percentage of damage and presence of live
P. truncatus and S. zeamais). These results demonstrated the
importance of postharvest practices such as winnowing, sorting,
and cleaning. The results of the partial least square analyses also
suggest that the percentage of fungi-damaged grain strongly de-
pends on the initial percentage of fungi and maximum relative
humidity. Cotty and Jaime-Garcia (2007) already mentioned the
relationship between climate conditions, fungi infestation, and
aflatoxin production. It is then important to implement pre-harvest
solutions to avoid on-field fungi infestation. Grain variety is also an
important parameter since some maize varieties are more sus-
ceptible to insect and fungi infestation. Resistance of stored grain to
insect attack has been attributed to factors such as grain hardness
or phenolic compounds (Arnason et al., 1993). Breeding efforts have
increased the resistance of grain to postharvest pests (Abebe et al.,
2009). However, most smallholder farmers in Mexico chose the
varieties they plant according to their home consumption strate-
gies (Appendini and Quijada, 2016). The temperature of grain at the
beginning of storage also seems to be an important factor, as shown
in variable importance plots. Decreasing grain temperature may be
challenging in a warmer climate, and reducing the equilibrium
relative humidity through airtight technologies is a low-cost solu-
tion in tropical areas (Bradford et al., 2018).
In conclusion, agroecological conditions impacted the perfor-
mance of non-hermetic storage technologies, particularly under
tropical conditions where they failed to minimize postharvest
losses. With polypropylene bags, the most common conventional
storage technology, the percentage of insect-damaged grain was, on
average, almost ten times higher on sites <500 m asl than on sites
>2000 m asl. Hermetic technologies, on the contrary, were a good
option for minimizing losses across agroecologies and under
farmers’ conditions with less than 4% of insect-damaged grain.
Alternative hermetic technologies, including recycled containers,
were also a good low-cost option. Compared to technologies that
involve the use of insecticides, inert dusts were a better option for
smallholder farmers when hermetic technologies were not avail-
able, although they were less effective under tropical conditions
than hermetic technologies. More research is needed on the correct
use of these adjuvants and their impact on grain quality.
Funding
CIMMYT implemented this study as part of the projects (1)
11
MasAgro Productor, (2) Milpa Sustentable en la Península de
Yucatan, (3) MasAgro Guanajuato and (4) MasAgro Quere taro made
possible by the generous support of SADER, Mexico; Fundacio n
Haciendas del Mundo Maya and Fomento CitiBanamex; SDAYR,
Guanajuato, Mexico and SEDEA, Quere taro, Mexico. This work is
also part of the CGIAR Research Program on Maize (MAIZE) with
support from W1&W2 donors, including national governments,
foundations, development banks, and other public and private
agencies. Any opinions, findings, conclusions, or recommendations
expressed in this publication are those of the authors and do not
necessarily reflect the view of the donors mentioned previously.
CRediT authorship contribution statement
Sylvanus Odjo: Methodology, Investigation, Data curation,
Formal analysis, Visualization, Writing - original draft, Writing -
review & editing. Juan Burguen ~ o: Methodology, Formal analysis,
Writing - review & editing. Ariel Rivers: Methodology, Investiga-
tion, Writing - review & editing. Nele Verhulst: Conceptualization,
Methodology, Investigation, Project administration, Supervision,
Writing - review & editing.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgments
We thank Jessica Gonz alez Regalado, Rodolfo Vilchis Ramos,
Abel Saldivia Tejeda, Luis Castillo Villasen~ or, Blanca Aide  Albarra
n
Contreras, Fabian Enyanche Vel azquez, Francisco Javier Vargas Lara,
Paullette Galaviz Soto and Juan Arista Corte s for their help in data
collection. We are grateful to all of our collaborators in Mexico and
to the farmers who participated in the non-controlled experiments.
We would like to especially thank the institutions hosting
controlled experiments: Universidad Tecnolo  gica de la Sierra
Hidalguense; Universidad Auto  noma de Yucata n; Instituto Nacional
de Investigaciones Forestales, Agrícolas y Pecuarias (sitio experi-
mental Mixteca Oaxaquen ~ a); Unio  n de Productores Agrícolas y
Pecuarios de Cotzoco  n; Sociedad Agroecolo gica de la Costa de
Oaxaca.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jspr.2020.101664.
References
Abass, A.B., Fischler, M., Schneider, K., Daudi, S., Gaspar, A., Rüst, J., Kabula, E.,
Ndunguru, G., Madulu, D., Msola, D., 2018. On-farm comparison of different
postharvest storage technologies in a maize farming system of Tanzania Central
Corridor. J. Stored Prod. Res. 77, 55e65. https://doi.org/10.1016/
j.jspr.2018.03.002.
Abass, A.B., Ndunguru, G., Mamiro, P., Alenkhe, B., Mlingi, N., Bekunda, M., 2014.
Post-harvest food losses in a maize-based farming system of semi-arid
savannah area of Tanzania. J. Stored Prod. Res. 57, 49e57. https://doi.org/
10.1016/j.jspr.2013.12.004.
Abebe, F., Tefera, T., Mugo, S., Beyene, Y., Stephen, V., 2009. Resistance of maize
landraces to the maize weevil Sitophilus zeamais Motsch. (Coleoptera: Curcu-
lionidae). Afr. J. Biotechnol. 8, 5937e5943. https://doi.org/10.1007/s13744-012-
0054-8.
Afful, E., Elliott, B., Nayak, M.K., Phillips, T.W., 2018. Phosphine resistance in north
American field populations of the lesser grain borer, Rhyzopertha dominica
(Coleoptera: Bostrichidae). J. Econ. Entomol. 111, 463e469. https://doi.org/
10.1093/jee/tox284.
Appendini, K., Quijada, M.G., 2016. Consumption strategies in Mexican rural
12 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
households: pursuing food security with quality. Agric. Hum. Val. 33, 439e454.
https://doi.org/10.1007/s10460-015-9614-y.
Arnason, J.T., Baum, B., Gale, J., Lambert, J.D.H., Bergvinson, D., Philogene, B.J.R.,
Serratos, J.A., Mihm, J., Jewell, D.C., 1993. Variation in resistance of Mexican
landraces of maize to maize weevil Sitophilus zeamais, in relation to taxonomic
and biochemical parameters. Euphytica 74, 227e236. https://doi.org/10.1007/
BF00040405.
Arthur, F.H., 2002. Survival of Sitophilus oryzae (L.) on wheat treated with diato-
maceous earth: impact of biological and environmental parameters on product
efficacy. J. Stored Prod. Res. 38, 305e313. https://doi.org/10.1016/S0022-
474X(01)00041-8.
Athanassiou, C.G., Kavallieratos, N.G., Peteinatos, G.G., Petrou, S.E., Boukouvala, M.C.,
Tomanovi, E., 2007. Influence of temperature and humidity on insecticidal effect
of three diatomaceous earth formulations against larger grain borer (Coleop-
tera: bostrychidae). J. Econ. Entomol. 100, 599e603. https://doi.org/10.1093/jee/
100.2.599.
Athanassiou, C.G., Kavallieratos, N.G., Vayias, B.J., Panoussakis, E.C., 2008. Influence
of grain type on the susceptibility of different Sitophilus oryzae (L.) populations,
obtained from different rearing media, to three diatomaceous earth formula-
tions. J. Stored Prod. Res. 44, 279e284. https://doi.org/10.1016/
j.jspr.2008.02.006.
Baoua, I.B., Bakoye, O., Amadou, L., Murdock, L.L., Baributsa, D., 2018. Performance of
PICS bags under extreme conditions in the sahel zone of Niger. J. Stored Prod.
Res. 76, 96e101. https://doi.org/10.1016/j.jspr.2018.01.007.
Bellon, M.R., Hodson, D., Bergvinson, D., Beck, D., Martinez-Romero, E., Montoya, Y.,
2005. Targeting agricultural research to benefit poor farmers: relating poverty
mapping to maize environments in Mexico. Food Pol. 30, 476e492. https://
doi.org/10.1016/j.foodpol.2005.09.003.
Bern, C.J., Yakubu, A., Brumm, T.J., Rosentrater, K.A., 2013. Hermetic storage systems
for maize stored on subsistence farms. Am. Soc. Agric. Biol. Eng. Annu. Int. Meet
1, 802e809, 2013, ASABE 2013.
Bernardino Hern andez, U.H., Torres Aguilar, H., Sanchez Cruz, G., Reyes Velasco, L.,
Zapie n Martínez, A., 2019. Uso de plaguicidas en el cultivo de maíz en zonas
rurales del Estado de Oaxaca. Me xico. Rev. Salud Ambient. 19, 23e31.
Bohinc, T., Horvat, A., Andri c, G., Pra
zi
c Goli
c, M., Kljaji
c, P., Trdan, S., 2018. Com-
parison of three different wood ashes and diatomaceous earth in controlling
the maize weevil under laboratory conditions. J. Stored Prod. Res. 79, 1e8.
https://doi.org/10.1016/j.jspr.2018.06.007.
Borgemeister, C., Adda, C., Se tamou, M., Hell, K., Djomamou, B., Markham, R.H.,
Cardwell, K.F., 1998. Timing of harvest in maize: effects on post harvest losses
due to insects and fungi in central Benin, with particular reference to Proste-
phanus truncatus (Horn) (Coleoptera: Bostrichidae). Agric. Ecosyst. Environ. 69,
233e242. https://doi.org/10.1016/S0167-8809(98)00109-1.
Boxall, R.A., 2002. Damage and loss caused by the larger grain borer. Integrated Pest
Manag. Rev. 7, 105e121. https://doi.org/10.1023/A:1026397115946.
Boxall, R.A., 1986. A Critical Review of the Methodology for Assessing Farm-Level
Grain Losses after Harvest. Trop. Dev. Res. Institute, U.K., p. 139
Bradford, K.J., Dahal, P., Van Asbrouck, J., Kunusoth, K., Bello, P., Thompson, J., Wu, F.,
2018. The dry chain: reducing postharvest losses and improving food safety in
humid climates. Trends Food Sci. Technol. 71, 84e93. https://doi.org/10.1016/
j.tifs.2017.11.002.
Chigoverah, A.A., Mvumi, B.M., 2016. Efficacy of metal silos and hermetic bags
against stored-maize insect pests under simulated smallholder farmer condi-
tions. J. Stored Prod. Res. 69, 179e189. https://doi.org/10.1016/
j.jspr.2016.08.004.
Cotty, P.J., Jaime-Garcia, R., 2007. Influences of climate on aflatoxin producing fungi
and aflatoxin contamination. Int. J. Food Microbiol. 119, 109e115. https://
doi.org/10.1016/j.ijfoodmicro.2007.07.060.
De Groote, H., Kimenju, S.C., Likhayo, P., Kanampiu, F., Tefera, T., Hellin, J., 2013.
Effectiveness of hermetic systems in controlling maize storage pests in Kenya.
J. Stored Prod. Res. 53, 27e36. https://doi.org/10.1016/j.jspr.2013.01.001.
Eroglu, N., Sakka, M.K., Emekci, M., Athanassiou, C.G., 2019. Effects of zeolite for-
mulations on the mortality and progeny production of Sitophilus oryzae and
Oryzaephilus surinamensis at different temperature and relative humidity levels.
J. Stored Prod. Res. 81, 40e45. https://doi.org/10.1016/j.jspr.2018.11.004.
Fields, P., Korunic, Z., 2000. The effect of grain moisture content and temperature on
the efficacy of diatomaceous earths from different geographical locations
against stored-product beetles. J. Stored Prod. Res. 36, 1e13. https://doi.org/
10.1016/S0022-474X(99)00021-1.
Fields, P.G., Korunic, Z., 2002. Post-harvest Insect Control with Inert Dusts. In:
Encycl. Pest Manag.
García-Lara, S., Bergvinson, D.J., 2007. Programa integral para reducir pe rdidas
poscosecha en maiz. Agric. Te cnica Mexico 33, 181e189.
Golob, P., 1997. Current status and future perspectives for inert dusts for control of
stored product insects. J. Stored Prod. Res. 33, 69e79. https://doi.org/10.1016/
S0022-474X(96)00031-8.
Gonza lez Regalado, J., Rivers, A., Verhulst, N., 2017. El sistema poscosecha en
Me xico: diagno  stico mediante encuestas a pequen ~ os productores. EnlACe la
Rev. la Agric. Conserv. An ~ o IX 42e48. Número 40 Cctubre - Noviembre 2017.
Hellin, J., Kanampiu, F., 2008. Metal silos bring food security to El Salvador. Appropr.
Technol. 35, 69e70.
Hole, B.D., 1981. Variation in tolerance of seven species of stored product Coleoptera
to methyl bromide and phosphine in strains from twenty-nine countries. Bull.
Entomol. Res. 71, 299e306. https://doi.org/10.1017/S0007485300008324.
Hole, B.D., Bell, C.H., Mills, K.A., Goodship, G., 1976. The toxicity of phosphine to all
developmental stages of thirteen species of stored product beetles. J. Stored
Prod. Res. 12, 235e244. https://doi.org/10.1016/0022-474X(76)90039-4.
Howe, R.W., 1973. The susceptibility of the immature and adult stages of Sitophilus
granarius to phosphine. J. Stored Prod. Res. 8, 241e262. https://doi.org/10.1016/
0022-474X(73)90041-6.
Kharel, K., Mason, L.J., Williams, S.B., Murdock, L.L., Baoua, I.B., Baributsa, D., 2018.
A time-saving method for sealing Purdue Improved Crop Storage (PICS) bags.
J. Stored Prod. Res. 77, 106e111. https://doi.org/10.1016/j.jspr.2018.04.002.
Koruni c, Z.K., Rozman, V., Liska, A., Lucic, P., 2016. A review of natural insecticides
based on diatomaceous earths. Poljoprivreda 22, 10e18. https://doi.org/
10.18047/poljo.22.1.2.
_
Krzyzowski, M., Francikowski, J., Baran, B., Babczyn  ska, A., 2019. Physiological and
behavioral effects of different concentrations of diatomaceous earth on com-
mon stored product pest Callosobruchus maculatus. J. Stored Prod. Res. 82,
110e115. https://doi.org/10.1016/j.jspr.2019.05.004.
Kumar, D., Kalita, P., 2017. Reducing postharvest losses during storage of grain crops
to strengthen food security in developing countries. Foods 6, 8. https://doi.org/
10.3390/foods6010008.
Manuel Rosas, I., Gil Mun ~ oz, A., Ramírez Valverde, B., Hern andez Salgado, J.H.,
Bellon, M., 2007. Calidad física y fisiolo gica de semilla de maíz criollo almace-
nada en silo met alico y con me todos tradicionales en Oaxaca. Me xico. Rev.
Fitotec. Mex. 30, 69e78.
Maredia, K.M., Mihm, J.A., 1995. Integrated management of stored grain insect pests
in developing countries. In: Jewell, D.C., Waddington, S.R., Ransom, J.K.,
Pixley, K.V. (Eds.), Maize Research for Stress Environments: Proceedings of the
Fourth Eastern and Southern Africa Regional Maize Conference. Fourth Eastern
and Southern Africa Regional Maize Conference; Harare, Zimbabwe; 28 March -
1 April, 1994. CIMMYT, Mexico D.F., Mexico, pp. 205e2010.
Mendoza, J.R., Sabillo  n, L., Martinez, W., Campabadal, C., Hallen-Adams, H.E.,
Bianchini, A., 2017. Traditional maize post-harvest management practices
amongst smallholder farmers in Guatemala. J. Stored Prod. Res. 71, 14e21.
https://doi.org/10.1016/j.jspr.2016.12.007.
Murdock, L.L., Margam, V., Baoua, I., Balfe, S., Shade, R.E., 2012. Death by desicca-
tion: effects of hermetic storage on cowpea bruchids. J. Stored Prod. Res. 49,
166e170. https://doi.org/10.1016/j.jspr.2012.01.002.
Murdock, L.L., Seck, D., Ntoukam, G., Kitch, L., Shade, R.E., 2003. Preservation of
cowpea grain in sub-Saharan Africa - bean/Cowpea CRSP contributions. Field
Crop. Res. 82, 169e178. https://doi.org/10.1016/S0378-4290(03)00036-4.
Mutambuki, K., Affognon, H., Baributsa, D., 2014. Evaluation of triple layer hermetic
storage bag (PICS) against Prostephanus truncatus and Sitophilus zeamais. 11th
Int. Work. Conf. Stored Prod. Prot. 355e362. https://doi.org/10.14455/
DOA.res.2014.59.
Ndegwa, M.K., De Groote, H., Gitonga, Z.M., Bruce, A.Y., 2016. Effectiveness and
economics of hermetic bags for maize storage: results of a randomized
controlled trial in Kenya. Crop Protect. 90, 17e26. https://doi.org/10.1016/
j.cropro.2016.08.007.
Ng’ang’a, J., Mutungi, C., Imathiu, S., Affognon, H., 2016. Effect of triple-layer her-
metic bagging on mould infection and aflatoxin contamination of maize during
multi-month on-farm storage in Kenya. J. Stored Prod. Res. 69, 119e128. https://
doi.org/10.1016/j.jspr.2016.07.005.
Nielsen, P.S., 1998. The effect of a diatomaceous earth formulation on the larvae of
Ephestia kuehniella Zeller. J. Stored Prod. Res. 34, 113e121. https://doi.org/
10.1016/S0022-474X(97)00060-X.
Nmx-Ff-123-Scfi-2015 Norma Mexicana, 2015. Silo meta lico herme tico-proceso de
fabricacio  n-tecnica de referencia basica.
Odjo, S., Gonza lez Regalado, J., Vilchis Ramos, R., Castillo Villasen ~ or, L., Saldivia
Tejeda, A., Galaviz Soto, P., Albarra n Contreras, B., Verhulst, N., 2018. Poscosecha.
Resultados 2017. In: Fonteyne, S., Verhulst, N. (Eds.), Red de Plataformas de
Investigacio n Cimmyt. Resultados PV 2017 y OI 2017-18. Agricultura de Con-
servacio n En La Agricultura Familiar. CIMMYT, Texcoco, Mexico, pp. 100e121.
Pimentel, M.A.G., Faroni, L.R.D.A., Guedes, R.N.C., Sousa, A.H., To tola, M.R., 2009.
Phosphine resistance in Brazilian populations of Sitophilus zeamais Motschulsky
(Coleoptera: Curculionidae). J. Stored Prod. Res. 45, 71e74. https://doi.org/
10.1016/j.jspr.2008.09.001.
Plumier, B., Zhao, Y., Cook, S., Ambrose, R.P.K., 2019. Adhesion of diatomaceous earth
dusts on wheat and corn kernels. J. Stored Prod. Res. 83, 347e352. https://
doi.org/10.1016/j.jspr.2019.08.001.
Quellhorst, H.E., Williams, S.B., Murdock, L.L., Baributsa, D., 2018. Cumulative oxy-
gen consumption during development of two postharvest insect pests: Callos-
obruchus maculatus Fabricius and Plodia interpunctella Hübner. J. Stored Prod.
Res. 77, 92e95. https://doi.org/10.1016/j.jspr.2018.03.006.
Quezada, M.Y., Moreno, J., V azquez, M.E., Mendoza, M., Me ndez-Albores, A., Mor-
eno-Martínez, E., 2006. Hermetic storage system preventing the proliferation of
Prostephanus truncatus Horn and storage fungi in maize with different moisture
contents. Postharvest Biol. Technol. 39, 321e326. https://doi.org/10.1016/
j.postharvbio.2005.10.004.
Ranum, P., Pen ~ a-Rosas, J.P., Garcia-Casal, M.N., 2014. Global maize production, uti-
lization, and consumption. Ann. N. Y. Acad. Sci. 1312, 105e112. https://doi.org/
10.1111/nyas.12396.
Reuber, M.D., 1985. Carcinogenicity and toxicity of malathion and malaoxon. En-
viron. Res. 37, 119e153. https://doi.org/10.1016/0013-9351(85)90054-4.
Sheahan, M., Barrett, C.B., 2017. Food loss and waste in Sub-Saharan Africa: a critical
review. Food Pol. 70, 1e12. https://doi.org/10.1016/j.foodpol.2017.03.012.
Stadler, T., Buteler, M., Weaver, D.K., Sofie, S., 2012. Comparative toxicity of nano-
structured alumina and a commercial inert dust for Sitophilus oryzae (L.) and
 S. Odjo et al. / Journal of Stored Products Research 88 (2020) 101664
Rhyzopertha dominica (F.) at varying ambient humidity levels. J. Stored Prod.
Res. 48, 81e90. https://doi.org/10.1016/j.jspr.2011.09.004.
Stathers, T., Lamboll, R., Mvumi, B.M., 2013. Postharvest agriculture in changing
climates: its importance to African smallholder farmers. Food Secur 5, 361e392.
https://doi.org/10.1007/s12571-013-0262-z.
Stathers, T.E., Denniff, M., Golob, P., 2004. The efficacy and persistence of diato-
maceous earths admixed with commodity against four tropical stored product
beetle pests. J. Stored Prod. Res. 40, 113e123. https://doi.org/10.1016/S0022-
474X(02)00083-8.
Sweeney, S., Steigerwald, D.G., Davenport, F., Eakin, H., 2013. Mexican maize pro-
duction: evolving organizational and spatial structures since 1980. Appl. Geogr.
39, 78e92. https://doi.org/10.1016/j.apgeog.2012.12.005.
Tefera, T., Kanampiu, F., De Groote, H., Hellin, J., Mugo, S., Kimenju, S.C., Beyene, Y.,
Boddupalli, P.M., Shiferaw, B., Banziger, M., 2011. The metal silo: an effective
grain storage technology for reducing post-harvest insect and pathogen losses
in maize while improving smallholder farmers’ food security in developing
countries. Crop Protect. 30, 240e245.
13
Tigar, B.J., Key, G.E., Flores-s, M.E., Vazquez-a, M., 1994. Field and post-maturation
infestation of maize by stored product pests in Mexico. J. Stored Prod. Res.
30, 1e8.
Tubbs, T., Baributsa, D., Woloshuk, C., 2016. Impact of opening hermetic storage
bags on grain quality, fungal growth and aflatoxin accumulation. J. Stored Prod.
Res. 69, 276e281. https://doi.org/10.1016/j.jspr.2016.10.003.
Villa-Manzano, A.I., Zamora-Lo pez, X.X., Huerta-Viera, M., Va zquez-Solís, M.G.,
Flores-Robles, G.D., 2019. Mortalidad y factores asociados con intoxicacio n por
fosfuros: estudio de cohorte prospectiva. Rev. Med. Inst. Mex. Seguro Soc. 57,
156e161.
Williams, S.B., Baributsa, D., Woloshuk, C., 2014. Assessing Purdue Improved Crop
Storage (PICS) bags to mitigate fungal growth and aflatoxin contamination.
J. Stored Prod. Res. 59, 190e196. https://doi.org/10.1016/j.jspr.2014.08.003.
Wold, S., 1995. PLS for multivariate linear modeling. In: van de Waterbeemd, H.
(Ed.), Chemometric Methods in Molecular Design, Methods and Principles in
Medicinal Chemistry. Verlag Chemie, Weinheim, Germany, pp. 195e218.