LAUSR

The composition and richness of herbivore and plant assemblages change along climatic gradients, but knowledge about associated shifts in specialisation is scarce and lacks controlling for the abundance and phylogeny of interaction partners. Thus, we aimed to test whether the specialisation of phytophagous insects in insect - plant - interaction networks decreases towards cold habitats as predicted by the 'altitude niche-breadth hypothesis' to forecast possible consequences of interaction rewiring under climate change. We used a non-invasive, standardized metabarcoding approach to reconstruct dietary relationships of Orthoptera species as a major insect herbivore taxon along a broad temperature gradient (~12°C) in southern Germany. Based on orthopteran surveys, feeding observations, collection of faecal pellets from > 3,000 individuals of 54 species, and parallel vegetation surveys on 41 grassland sites, we quantified plant resource availability and its use by herbivores. Herbivore assemblages were richer in species and individuals at sites with high summer temperatures, while plant richness peaked at intermediate temperatures. Corresponding interaction networks were most specialised in warm habitats. Considering phylogenetic relationships of plant resources, however, the specialisation pattern was not linear but peaked at intermediate temperatures, mediated by herbivores feeding on a narrow range of phylogenetically related resources. Our study provides empirical evidence of resource specialisation of insect herbivores along a climatic gradient, demonstrating that resource phylogeny, availability, and temperature interactively shape the specialisation of herbivore assemblages. Instead of low specialisation levels only in cold, harsh habitats, our results suggest increased generalist feeding due to intraspecific changes and compositional differences at both ends of the microclimatic gradient. We conclude that this non-linear change of phylogeny-based resource specialisation questions predictions derived from the 'altitude-niche breadth hypothesis' and highlights the currently limited understanding of how plant-herbivore interactions will change under future climatic conditions.