Conditional probability and integrated pest management: Using a nonlinear Kriging technique to predict infectious levels of Verticillium dahliae in Michigan potato fields

Abstract

A recent survey of potato (Solanum tuberosum) growers in the state of Michigan identified that soilborne pathogens were causing concerns as to whether growers would be able to continue to meet the high demands for marketable potatoes. Of these soilborne pathogens, Verticillium dahliae is one of the most concerning due to its direct correlation with yield decline and its persistence in the soil. Following the survey a statewide soil study was conducted to study soilborne pathogens and their interactions with multiple abiotic and biotic factors. The use of geostatistics and geographical information systems (GIS) were incorporated into this study to assess the spatially distribution of colonies of V. dahliae across a field and to use geostatistical methods to determine V. dahliae inoculum levels throughout the entire field from 20 soil samples. Furthermore, the research team incorporated the use of a nonlinear indicator Kriging method to create conditional probability maps of soilborne pathogen inoculum levels and predict where inoculum levels would be high enough to result in infection. The methods presented in this paper evaluated conditional probability mapping of soilborne plant pathogens for the potential to become a practical crop management tool for commercial potato growers.