Kirsty H. Macphie, Jelmer Menno Samplonius, Jarrod Hadfield, James Pearce-Higgins, Albert Phillimore 

Phenological shifts are well documented biological responses to warming. While many studies have focused on the mean timing of an event, there is growing appreciation that the height and width of the phenological distribution will also impact on species interactions. A temperate deciduous forest food chain of oak trees – arboreal caterpillars – insectivorous passerines has become paradigmatic in research on phenological mismatch. This focus on oak-dominated woodlands means that we have limited insight into whether (i) caterpillar phenological distributions vary among tree taxa and habitats and (ii) oak is an exceptional host, which has implications for the potential for buffering of interactions on a local and landscape scale. Here, we survey caterpillar abundance and mass throughout spring on 10 tree taxa for 10 years across 44 Scottish woodland sites. We found substantial variation in caterpillar abundance among host taxa, with oak, birch and willow yielding similarly high numbers of caterpillars, and evidence that caterpillar abundance increases with the density of oak foliage within a woodland stand, but not with the density of other taxa. Considering variation in the phenological distribution of caterpillars on different host taxa, we found the main axis of variation to be the maximum abundance/total biomass reached, which was highest on oak. We found significant variation in the mean timing of abundance and duration of abundance and total biomass among hosts, though effect sizes were quite small, and little evidence for among host variation in the phenological distribution of individual caterpillar mass. In woodlands where oak is abundant, our findings are consistent with the presence of other tree taxa providing little local buffering of phenological mismatch. Whereas, in the absence of oak, birch and willow have the potential to support similarly substantial caterpillar abundances. These findings have implications for conservation, resilient forestry planting and management decisions.