Patterns of insect damage types reflect complex environmental signal in Miocene forest biomes of Central Europe and the Mediterranean.

Ecosystems are defined by the community of living organisms and how they interact together and with their environment. Insects and plants are key taxa in terrestrial ecosystems and their network determines the trophic structure of the environment. However, what drives the interactions between plants and insects in modern and fossil ecosystems is not well understood. In this study, we analyzed insect damage richness and frequency in 5000 fossil leaves deposited during the early Miocene at 20–17 Ma along a latitudinal gradient from Europe (two localities in Czech Republic) to Turkey (one locality) in a temperate climate setting. Damage frequency was mainly linked with abiotic factors (temperature, precipitation seasonality) whereas damage richness was mainly linked with biotic factors (plant richness, biome). Univariate analysis of insect damage types consistently suggested closer trophic similarity between the Mediterranean and either the one or the other Central European plant assemblage. In contrast, multivariate analysis of all insect damage types indicated closer similarity between the two Central European sites highlighting the importance of biogeographic legacy and geographic closeness to the plant-insect interaction patterns. Our results underscore the high complexity of the herbivory network and call for careful interpretations of plant-insect interaction patterns in palaeoecological studies. Finally, comparing the trophic similarity between different localities using total evidence plots as done in this work might be a promising complementary method in comparative studies of plant-insect interactions. • We compared insect damage types in Miocene leaf floras of Turkey and Czech Republic • DT richness was higher in subhumid laurel (TR) than mixed-mesophytic forest (CZ) • Damage frequency was linked to abiotic factors (t°, precipitation seasonality) • Damage richness was linked to biotic factors (species richness, biome) • A multivariate analysis of all DT richness recorded a strong geographic signal [ABSTRACT FROM AUTHOR]