Modeling the correlation between potato disease spread and climate variables to guide fungicide applications in Cundinamarca, Colombia

Abstract

Potato production relies on frequent fungicide spraying, damaging the environment, increasing production costs, and eventually leading to crop abandonment. To better guide fungicide spraying in the field, the influence of weather on pathogen characteristics should be understood. To achieve this, we sampled a potato pathogen, Phytophthora infestans, in twelve commercial potato fields in the department of Cundinamarca, Colombia. Disease samples were collected to estimate the number of sporangia. Fields were planted with different potato cultivars of varying disease resistance/susceptibility, and chemical management was performed according to standard agricultural practices. The relationship between climate conditions and sporangia production not only depended on climate conditions on the day of collection, but also on weather variables one or two days prior. Based on climate data, we also evaluated the potential non-linear effects of climate variables and used this analysis to perform a forecast simulation. We used the GeoSimCast model to estimate the number of fungicide applications required to control late blight in this region. The number of applications suggested by the model run for 60% relative humidity was very different from the observed number of applications. The model results for 90% relative humidity estimated a number closer to the observed value, suggesting that potato growers always assume optimal conditions for disease development and spray fungicides accordingly. Our results indicate that this model has the potential to guide seeding plans and the chemical management of late blight and other plant diseases in Colombia.