The effects of climate change (CC) will impact species ranges, and crops are no exception. Anticipating these changes through forecasting the environmentally suitable area of crops would help to reduce or mitigate the impact and adapt ecological and economic strategies. To forecast the CC effects on crops, we describe here a model-assisted framework (hereafter SDM-CropProj) that combines two modelling steps to be implemented in sequence: i) a multi-technique calibration process and ensemble-forecasting approach to predict the current and future environmental suitability of target crops; ii) a parsimonious univariate log-log linear model to relate the average total annual production to the current SDM-based suitable area. Different metrics for assessing the model's predictive performance showed that:•Crop production is related to model-predicted suitable area, thus allowing to obtain future projections of total fruit production based on climate scenarios.•The SDM-CropProj framework can assess potential pathways and trends in annual production due to changes in the environmental suitability and the distribution of multiple crop varieties/types as a consequence of CC, offering insights to other areas and crop types.
Keywords: Agroecosystems; Climate change; Crop production; Long-term projections; Machine learning; Species distribution models.
© 2021 The Authors.