Biophysical factors and agronomic practices associated with Fusarium ear rot and fumonisin contamination of maize in multiple agroecosystems in Ethiopia

In Ethiopia, Fusarium ear rot (FER) reduces yield, quality, and safety of maize and negatively impacts the economics of maize growers. However, there is almost no published information on geographic distribution, severity, or causal agents of FER in Ethiopia or levels of preharvest fumonisin contamination, biophysical factors, and agronomic practices associated with FER. Therefore, we surveyed 480 agricultural fields located in 10 Ethiopian zones to assess the incidence and severity of FER as well as agricultural practices and biophysical factors associated with each field. We also determined the concentration of fumonisin in samples taken from the fields. Results revealed that FER was prevalent with varied disease intensity and yield loss levels among the assessed zones. The highest percent severity index (PSI) was recorded in Mirab Wellega (33.2%), with a correspondingly maximum estimated mean grain yield loss of 13.6%, while the minimum disease severity 11.3% and yield loss 5.4% were recorded from the Ilubabor zone. Fumonisins (FB1, FB2, FB3, and FB4) were detected in 98% of the samples, and the levels ranged from 5 to 7517 mu g kg(-1) (mean, 540 mu g kg(-1) and median 82.4 mu g kg(-1)). The mean fumonisin contamination of 1524.4, 1316.8, 744.5, and 742.6 mu g kg(-1) was recorded for Mirab Welega, Misraq Wellega, Assosa, and Hawassa zones, respectively. In total, about 11.3% of the maize samples exceeded the maximum tolerable total fumonisin limit set by the European Union in maize flour (>1000 mu g kg(-1)) and are not suitable for human consumption. Disease severity highly significantly (p < 0.0001) correlated with biophysical factors such as insect infestation (r = 0.60, p < 0.0001) and weed infestation (r = 0.64, p < 0.0001) and agronomic practices such as nitrogen phosphorus sulfur fertilizer application rate (r = -0.39, p < 0.0001) and insecticide spray frequency (r = -0.55, p < 0.0001). Other agronomic practices such as earlier or later sowing date and inter-cropping systems significantly reduced (p < 0.001) FER incidence, PSI, and the associated grain yield loss. There was significant (p < 0.0001) variation among maize cultivars currently under production in reaction against FER disease intensity. Moreover, the result of the present study attests a strong positive association (p < 0.001) between disease intensity and maize fumonisin concentrations. Generally, the study revealed a wider distribution of FER and fumonisin contamination of maize in the southwest major maize-growing belts of Ethiopia. Thus, appropriate integrated pre-harvest ear rot management strategies including the use of resistant cultivars and suitable agronomic practices are recommended for the management of FER and the associated mycotoxins in Ethiopia.