Differences between neighbouring tree species in phenology could isolate populations of host-plant generalists that depend on matching the phenology of their host. We studied the relationship between the budburst phenology of two co-occurring tree species with early (hornbeams) and late (oaks) budburst, and the egg-hatching date of associated winter moths (Operophtera brumata) during two seasons (autumns starting in 2020 and 2021). A previous study in spring 2019 had found no winter moth larvae on the focal oaks, while we found them mainly on oaks with hornbeam neighbours in 2022. Congruently, adult winter moths were rarely encountered during the autumns of 2018 and 2019 and sparsely in 2020 and 2021, and then mainly near hornbeams.The vast majority of winter moths had early-hatching eggs when both hornbeams and oaks were present, matching the hornbeams rather than the oaks. Where hornbeam was absent in the neighbourhood, the few winter moths had late-hatching eggs.These results suggest that winter moth populations on hornbeam and oak are to some degree isolated from each other despite spatial proximity, so that recolonization of oaks by populations synchronized with budburst of hornbeam is slow. Therefore, trees would benefit from differing from their neighbours in budburst phenology with respect to herbivore damage.

Study system: We selected 35 individual mature sessile oaks and recorded the species identity of neighboring trees (trees with which the crown is in touch) in the Puszcza Zielonka forest in Western Poland [52°33′12″N 17°06′48″E; 17]. This forest consists mainly of sessile oaks and pines (Pinus sylvestris L., not suitable for winter-moth larvae), with a sub-canopy of common hornbeam in some areas [18]. The selected oaks were located in two clusters near the villages Kamińsko and Zielonka. The neighbourhoods of focal oak trees near Kamińsko often included hornbeams (4 out of 7 sites), but hornbeams were absent in sites near Zielonka village (6 sites). In this region, hornbeams typically are among the first trees to burst buds, and oaks the last.

Winter-moth larvae are generalist feeders, attacking leaves and flowers of most woody angiosperms in temperate regions, including oaks and hornbeam. They can cause extensive defoliation in forest stands during cyclic outbreaks, leading to reduction in radial growth of trees. The wingless females eclose from the leaf litter during evenings of autumn of early winter, orient towards tree trunks where they mate with winged males, and then climb up to deposit eggs in the canopy. The egg-hatching phenology is highly heritable. When larvae hatch from eggs in spring, they can wait for several days for buds to burst, and they may also disperse by ballooning to avoid starvation. In case eggs hatch when leaves have already developed, their food quality is low and this results in lower fitness. Overall, host-plant choice is limited to larvae making the risky decision to leave a tree. Larvae finally pupate in the leaflitter below the tree.

Abundance of larvae: We used a slingshot and ropes to obtain canopy branches from ten focal oaks [sampled before in 2019; 17] on May 11^th and 16^th 2022. From each tree, we collected two to three branches on a sheet and counted the number of winter-moth larvae. Three of these trees were in neighbourhoods with hornbeams, while the others were far from hornbeams.

 Budburst phenology: We determined the budburst date of focal oak trees, their neighbouring oaks (in total 126 individual oaks) and hornbeams (24 individuals) in 2022. To estimate the day of 50% budburst for each tree, we visited trees every three to four days during the period of budburst. On each observation day, one observer scored the per cent budburst of each tree crown on branches from four cardinal directions using binoculars. We then interpolated or extrapolated the day of 50 per cent budburst.

Egg-hatching phenology: Our attempts to collect adult winter moths in 2018 and 2019 were largely unsuccessful. We collected mating pairs of winter moths in October and November 2020 and 2021 (between 5 pm and 8 pm) from the trunks of focal oak trees and their neighbours into plastic vials with foam stoppers. That same evening, we transferred mated females to vials with rolled tissue paper for egg-laying and kept the vials with eggs outdoors in the shade. We then determined the date-of-first-egg-hatching during the following spring for the eggs of each female.

These are Excel files with raw data, plus the R code for the eg-hatching data.