Host use diversification during range shifts shapes global variation in Lepidopteran dietary breadth

Analysing a global dataset of almost 7,000 dietary records from >4,400 species of butterflies and moths, the author shows that latitudinal variation in diet breadth can be better explained by the relative position of a population within its geographic range, than by latitude per se. Niche breadths tend to be greater at higher latitudes. This pattern is frequently assumed to emerge from the cumulative effects of multiple, independent local adaptation events along latitudinal environmental gradients, although evidence that generalization is more beneficial at higher-latitude locations remains equivocal. Here I propose an alternative hypothesis: that latitudinal variation in niche breadths emerges as a non-adaptive consequence of range shift dynamics. Based on analysis of a global dataset comprising more than 6,934 globally distributed dietary records from 4,410 Lepidopteran species, this hypothesis receives robust support. Population-level dietary niche breadths are better explained by the relative position of the population within its geographic range and the species' poleward range extent than by the latitude of diet observation. Broader diets are observed closer to poleward range limits and in species that have attained higher latitudes. Moreover, latitudinal variation in diet breadth is more prominent within and among species undergoing rapid, contemporary range shifts than for species with more stable ranges. Together these results suggest that latitudinal patterns in niche breadth represent a transient and emergent property of recent geographic range dynamics and need not require underlying gradients in selective agents or fitness trade-offs. The results have wide-ranging implications for global ecology and for anticipating changes in host use during ongoing distributional shifts of pests and disease vectors.