Transferability of a random forest model for resistance prediction between different regions in Europe

Abstract

Herbicides are an important technology in the Integrated Weed Management (IWM) tool box aiming to control weeds in modern agriculture. Prediction tools to evaluate the risk of resistance evolution will greatly help to choose the best IWM strategy adapted to the local field situation. In a previous work (HERRMANN et al., 2016) a random forest risk assessment model based on a data set comprising field history, management, and resistance status of Alopecurus myosuroides populations in Southern Germany was created. In this study transferability of the model with respect to regions and comparable weeds was analysed based on a similar dataset from a region in Northern France. The data from France also contained information on Lolium spp. The data related to Germany and France were subjected to a cross-validation procedure by interchanging test and training data. Results showed that acceptable predictions can be obtained for training data from Germany applied to France and vice versa. Resistance status in LOLSS samples from France can be predicted with a good accuracy based on a combined training set of A. myosuroides samples from Germany and France.

Keywords

• artificial intelligence
• geographical variation
• herbicide resistance prediction
• resistance management