Model Selection in Describing Disease Progress Curve of Fusarium Wilt (Fusarium oxysporum F.SP. Sesami) Disease in Sesame Varieties

Abstract

An epidemic of disease is the progress of the disease in time and space. The objectives of the present study are to understand and compare the four nonlinear models for disease progress curves of five sesame varieties. The regression parameters estimation, standard error, R-square, Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were estimated. The lowest values of standard error and the highest values of R-square were calculated from the monomolecular model. Also, the result showed that; the disease progress curve better fitted within the monomolecular model for each varieties with the smallest AIC and BIC values. This model is appropriate for modelling epidemics where there is a monocycle within a growing season. The Monomolecular model allows the estimation of the disease progression rate and an area under the disease progress curve was carried out to know the level of reaction to the disease. The lowest rate of fusarium wilt disease was recorded from Hirhir followed by Setit-2. However, the highest value was recorded from Setit-3 followed by Setit-1. A highest value of area under disease progress curve (AUDPC) was calculated from Setit-3. However, the lowest was calculated from Hirhir. Varieties with low disease incidence could be useful in breeding programs aimed at developing varieties with higher resistance to Fusarium wilt disease.