Abstract:
Maize production is an important component of food security in Ethiopia and ranks first in
productivity and second in area coverage after tef. Fusarium is one of the major fungal genera
associated with maize in the world. This genus comprises several toxigenic species including F.
verticillioides and F. proliferatum, which are the main fumonisins producers. Fumonisin
occurrence in maize grains is a main concern for human and animal health. Hence the study was
conducted with the objectives to: 1) determine occurrence of Fusarium spp. and fumonisin
contamination in fresh harvested and three-month stored maize samples in eastern Ethiopia; 2)
determine the effects of farmers’ practices on levels of fumonisin contamination in fresh
harvested and three-month stored maize samples; 3) determine the effects of season, location and
maize hybrid and their interaction on kernel infection by F. verticillioides and fumonisin
contamination; 4) analyze environmental factors associated with F. verticillioides kernel infection
and fumonisin contamination; 5) develop preliminary model that would assist in the prediction of
fumonisin concentration in maize grain at harvest using weather variables around silking date and
insect damage at harvest; and 6) assess the antagonistic effects of indigenous Trichoderma spp.
against F. verticillioides under in vitro conditions. Two-round surveys were conducted to collect
127 fresh harvested and 127 three-month stored maize samples from five districts (Girawa,
Goromuti, Haramaya, Meta and Tullo) of Eastern Ethiopia during 2017/18 main cropping
seasons. All samples were analyzed for occurrence of Fusarium species and fumonisin
contamination in maize grain in the study areas. During the surveys, farmers’ agronomic and
post-harvest practices were recorded through questionnaires with selected farmers. In addition, six
maize varieties (Alemaya Composite, BH 140, BH 545, BH 661, Gibe-2 and Raare-1) were
planted in a randomized complete block design (RCBD) in three replications at Girawa,
Haramaya and Hirna fields during 2017 and 2018 cropping seasons. At harvest, 65 plants were
randomly taken from each plot, evaluated for insect damage severity, hand-shelled and analyzed
for F. verticillioides kernel infection and fumonisin contamination. Fusarium species were
isolated and identified using direct plating technique by plating 15 kernel on Malachite green agar
(MGA2.5) and Carnation leaf agar (CLA), while fumonisin concentration was analyzed by
xvii
ELISA protocols. Isolation of antagonistic Trichoderma species from rhizosphere soil was made
using serial dilution technique at 10−2
, 10−3
and 10−4
. Antagonistic potential of Trichoderma
isolates against F. verticillioides was evaluated in dual culture assay on PDA. In this result, higher
numbers of Fusarium isolates were recorded in fresh harvested (2376 isolates) than in three month stored maize samples (1897 isolates). They were recorded in 97 and 90% of fresh
harvested and three-month stored maize samples along with average kernel infection of 62 and
50%, respectively. A total of seven Fusarium species (F. andiyazi, F. graminearum, F.
oxysporum, F. proliferatum, F. solani, F. subgulitnans and F. verticillioides) were identified
based on morphological characteristics. Fusarium verticillioides was dominant in both fresh
harvested and three-month stored maize samples, followed by F. proliferatum and F.
subgulitnans in fresh harvested and three-month stored samples, respectively. The frequencies of
F. proliferatum and F. verticillioides were higher in fresh harvested maize samples than in three month stored maize samples. Fumonisin was detected in all fresh harvested and 92% of three month stored maize samples (concentration ranging from 105 to 5,460 µg kg−1
). The average
(2,509 µg kg−1
) fumonisin concentration detected in fresh harvested maize samples was higher
than that of three-month stored maize samples (1,668 µg kg−1
). Similarly, percentage of samples
with fumonisin level above the regulatory limit was the highest in fresh harvested maize samples
than in three-month stored maize samples. Thus, 84% of the fresh harvested and 94% of the three month stored maize samples met the standard set by US Food and Drug Administration (FDA) in
raw maize for human consumption (≤4000 µg kg−1
), while 63% of fresh harvested and 83% of
three-month stored maize samples met the standard (<2000 µg kg−1
) set by East African
Community (EAC). Farmers’ agronomic practices, such as intercropping, late planting, late
harvesting and insect pest attack in the field, tended to increase the likelihood of fumonisin
contamination at levels above 2000 µg kg−1 in both fresh harvested and three-month stored maize,
while application of fertilizer decreased the likelihood of fumonisin contamination at levels above
2000 µg kg−1 in fresh harvested maize samples. Post-harvest practices, such as drying maize in the
field and sorting, showed decreasing trend on likelihood of fumonisin contamination to the level
above 2000 µg kg−1
in three-month stored maize. Since some field environmental conditions may
not be controlled, farmers should adopt pre- and post-harvest practices that reduce fumonisins
contamination to the level below the permissible regulatory limits. High Fusarium kernel
infection and fumonisin contamination recorded in the study area could indicate risk of exposures
to fumonisin as almost all maize produced in the study areas were used for human consumption.
The observed prevalence of the other Fusarium species, indicates the possibility of contamination
of maize kernels by several other mycotoxins. Thus, variety of harmful mycotoxins such as
xviii
deoxynivalenol, nivalenol and zearalenone should also be considered in the future assessments.
Higher Fusarium species kernel infection and fumonisin contamination recorded in fresh
harvested maize indicates Fusarium infection and fumonisin production was accorded in the field
prior to harvest. Therefore, any future study regarding Fusarium species and fumonisin
contamination should focus on pre-harvest environmental condition. In the two years of field
experiments, the average F. verticillioides kernel infection (36.6%) and fumonisin concentration
(2.24 µg g−1) detected in 2018 season was higher than kernel infection (34.6%) and fumonisin
concentration (1.9 µg g−1) detected in 2017 season. Among the fields, maize harvested from Hirna
field was highly (47.87%) infected by F. verticillioides and fumonisin concentration (3.45 µg
g
−1
). On average, Alemaya Composite variety had the lowest (28.2%) kernel infection by F.
verticillioides and the lowest fumonisin concentration (1.31 µg g−1
). Using 36 data points (year x
location x varieties), stepwise multiple linear regression detected several weather variables in four
7-day critical periods around silking date and insect damage as predictor of fumonisin
contamination. Fumonisin was well predicted when data from both insect damage and weather
variables were included in the model and the model explained 92% of the variation in fumonisin
concentration. The current model did not take into account all maize growth stage and weather
variables. Thus, other variables need to be identified and included in future cycles of model
improvement and verification. A total of ten Trichoderma isolates (Tr1, Tr2, Tr3, Tr4, Tr5, Tr6,
TrNS2, TrN3, TrNS4 and TrNS5) were identified and grouped into two Trichoderma spp. (T.
harizanum and T. viride). All Trichoderma isolates were analyzed for antagonistic potential
against F. verticillioides in in vitro condition in a dual culture assay after 6 day of incubation. Six
Trichoderma isolates (Tr1, Tr3, Tr5, TrNS2, TrN3 and TrNS4) exhibited inhibition percentage
˃80% after six day of dual culture incubation. The tested Trichoderma isolates (T. harizanum and
T. viridea) showed varying degrees of antagonism against F. verticillioides. However, further
studies on the effectiveness of Trichoderma spp. to control F. verticillioides should be conducted
under in vivo conditions. The overall results of this present study are important in creating new
management strategies to prevent maize kernel infection by Fusarium spp. and grain
contamination by fumonisins. Management decisions to use maize genotypes resistant for F.
verticillioides infection and fumonisin accumulation are more likely to produce maize grain that
meet tolerable fumonisin concentrations. Since there no regulatory limit for any mycotoxin
contamination in food in Ethiopia at the moment, efforts are needed to intervene the situation in
order to have safe food for consumer.