Temporal Modelling of the Spread of Late Blight Infestation on Potato at Pundibari (a Part of Coochbehar District)

Abstract

The paper is devoted to study the dynamics of the infestation of potatoes owing to the occurrence of late blight disease over two successive years (2014 and 2015) in the Terai region of West Bengal. Nonlinear models have been fitted on the potato late blight data (i.e. percent disease index data). The goodness-of-fit tests on different models have been performed by the application of the following criteria, namely coefficient of determination, mean absolute error, mean absolute percentage error, average relative predictive error, Akaike’s information criterion and Bayesian information criterion, respectively. The validation of the models has been carried out by the Shapiro-Wilk test and running test for conforming to the assumptions of normality and independence of the errors, respectively. Based on the application of the goodness-of-fit tests on the models, it was found that for 2014 and 2015, cubic and Gompertz models provided the best-fitted models based on the above sets of data, respectively. Using the best-fitted models referred to above, the values of two important parameters, namely (1) maximum rate of growth of the disease and (2) maximum disease severity, were determined. For the year 2014, it was observed that the maximum rate of growth of the disease occurred at 31 days after planting (DAP) and the maximum disease severity occurred at 48 DAP; however, even after the attainment of the maximum rate of the growth of the disease, the severity of the disease may increase, so it can be concluded that an additional application of fungicidal spray is necessary (when the disease is located in the field). On the other hand, for 2015, we observed that the maximum rate of growth of the disease occurred at 43 DAP. It indicates that any protection measure should be adopted at this stage (here, 43 DAP). Importantly, in an unprotected field, the late blight disease grows indefinitely as time advances infinitely.