Geostatistical modelling improves prediction of Macrophomina phaseolina abundance and distribution in soybean fields

Phytopathology. 2024 Nov 19. doi: 10.1094/PHYTO-04-24-0139-R. Online ahead of print.

Abstract

Charcoal rot, caused by the soilborne fungus Macrophomina phaseolina (Mp) poses a serious threat to soybean health and harvests at a global scale. Mp exhibits varying distribution patterns across fields, which complicates our ability to predict disease occurrences and outbreaks. Therefore, determining the spatial distribution of Mp abundance and its relation with soil physicochemical properties would help to inform precision management decisions for mitigating charcoal rot. To achieve this, Mp colony forming units (CFU) and edaphic properties were evaluated in 297 soybean fields located in the main soybean growing regions across 7 Departments of Paraguay. A pattern of decreasing CFU density was observed from the south-eastern to the western part of the country. While several edaphic factors are positively correlated with Mp CFU, pH showed a significant negative correlation with CFU. Both spatial and non-spatial model suggest that cation exchange capacity, percentage of clay, and pH could be potential predictors of Mp CFU abundance. Including spatial dependence of edaphic factors improved the prediction of Mp CFU more effectively than classical statistical models. We demonstrated that the occurrence of Mp shows a significant spatial clustering pattern as indicated by Moran's I. Our findings will help growers and policy-makers make informed decisions for managing Mp by improving our ability to predict which agricultural fields and soils are at greatest risk for charcoal rot.

Keywords: Epidemiology; Fungal Pathogens; Modelling; Pathogen Detection.