Insect Pests on the Move: Climate, Soil, Land Use, and the Search for Contingent Generality

Climate change is reshaping the geographic distributions of insect pests, with major consequences for agriculture, forestry, and ecosystem stability. Species distribution models (SDMs) are widely used to project these changes, yet most rely primarily on climatic predictors and implicitly assume a degree of generality in species responses that may not hold across diverse taxa. Here, we evaluate how 62 globally important insect pest species respond to climate change and to non-climatic predictor variables, asking whether their responses exhibit global generality, idiosyncrasy, or intermediate forms of contingent generality. We constructed correlative SDMs using multiple algorithms and predictor sets (climate only; climate + land use; climate + soil; climate + land + soil) and projected distributions under two emissions scenarios (SSP1–2.6 and SSP5–8.5) and two future time horizons (2020s and 2080s). Climate change produced strong aggregate trends, including poleward shifts and declining similarity between future and historical distributions, but species-level responses were highly heterogeneous in both magnitude and direction. Trait-based analyses revealed patterns of contingent generality structured by taxonomic order, feeding guild, and habitat association. Adding land-use variables consistently improved model performance more than adding soil variables, while combining both predictors was typically subadditive. Further analyses showed that land-use effects reorganize climate-response predictions for specific subsets of species, whereas soil effects were weaker and more context-dependent. Together, these results demonstrate that insect pest responses to climate change and model complexity are neither uniform nor random, but structured by ecologically meaningful contingencies. Recognizing and explicitly incorporating contingent generality can improve the interpretation, performance, and policy relevance of SDMs under global change.