Cyclically outbreaking geometrid moths in sub-arctic mountain birch forest: the organization and impacts of their interactions with animal communities

In sub-arctic mountain birch forest in northern Fennoscandia, the 2 geometrid moth species Epirrita autumnata (autumnal moth) and Operophtera brumata (winter moth) show high-amplitude population cycles with regular 10-year periodicity. During some population peaks, moth populations attain outbreak densities and cause region-wide defoliation and mortality of mountain birch. The severity and duration of moth outbreaks presently appears to be increasing, owing to climate-driven range-expansions of both native and novel (see below) moth species in the system. The causal mechanisms of moth population cycles have been widely studied, with research focusing on the role of parasitoids during the last decade. This research has focused on total parasitism rates and has paid little attention to parasitoid community organization and its consequences for the functionality of parasitoid communities. Study I – III of this PhD project addressed this knowledge gap for larval parasitoids, which have received more attention than other parasitoid guilds in the research on parasitism the birch-moth system. Study I explored the possibility of stochastic extinction-recolonization dynamics – induced by fluctuations in moth host populations – as a driver of the spatial distribution patterns of different larval parasitoid species. The study documented large-scale spatial segregation in the prevalence of different parasitoid species in O. brumata, which may have resulted from stochastic extinction-recolonization processes. However, the alternative explanation that the observed patterns were caused by spatial gradients in habitat characteristics could not be rejected. Further, the study found that the magnitude of total larval parasitism rates at a given location was independent of which parasitoid species was locally dominant. Study II mapped out the phenology of attack of the larval parasitoid species of O. brumata. The study showed that the attacks of different parasitoid species followed each other in a successional manner throughout the larval season, so that all larval instars were attacked by at least 1 parasitoid species. The study argued that this phenological diversity within the larval parasitoid guild would reduce the probability of climate-induced phenological mismatches between larvae and many parasitoid species within a single season, hence buffering total larval parasitism rates against stochastic climatic variation. Study III compared larval parasitoid species richness and prevalence rates among E. autumnata, O. brumata and Agriopis aurantiaria (scarce umber moth). E. autumnata is native species to the mountain birch forest, while O. brumata and A. aurantiaria invaded this system by range-expansion approximately a century and 15 years ago, respectively. The study found that E. autumnata and O. brumata hosted similar numbers of larval parasitoid species in the mountain birch system, while the larval parasitoid guild of A. aurantiaria was strongly species-impoverished compared to the 2 other moth species. Based on this, the study argued that invasive moth species take at least a century to acquire a larval parasitoid guild with native levels of species richness in the mountain birch forest. Total larval parasitism rates were similar among all 3 moth species, suggesting that invasive geometrid moths do not enjoy release from larval parasitism in the mountain birch forest, despite having species-impoverished larval parasitoid guilds. Taken together, study I – III pointed towards high levels of functional redundancy among larval parasitoid species in the birch-moth system. This could act to stabilize total larval parasitism rates in space and time. The studies also highlighted that it is necessary to resolve numerous uncertainties surrounding parasitoid taxonomy in order to make further progress in parasitoid community ecology in this system. While the causes of moth population cycles and outbreaks have been intensively studied, comparatively little attention has been paid to the ecological consequences of moth outbreaks. Some of the most serious knowledge gaps relate to the consequences of outbreak-induced forest damage for animal communities in the mountain birch ecosystem. Study IV and V addressed this issue by investigating the short-term responses of saproxylic (i.e. associated with dead wood) beetles and passerine birds, respectively, to an outbreak that had caused widespread mortality of birch forest 3 – 5 years before the outset of the studies. Study IV showed that the proportion of obligate saproxylic species in the beetle community was only about 10 % higher in damaged than undamaged birch forest. The study thereby indicated that saproxylic beetles have limited ability to respond numerically to the enormous amounts of dead wood that are generated by moth outbreaks. Climatic constraints on beetle activity and diversity in my sub-arctic study region, and species-specific preferences for dead wood in certain stages of decay, were suggested as explanations for the weak response of the saproxylic beetle community. The study raised the possibility that saproxylic beetles, owing to weak numerical responses, may play a minor role in wood decomposition in the immediate aftermath of moth outbreaks. This highlighted that there is need to learn more about the role of microbial wood-decomposer communities after outbreaks. The results of study V mirrored those of study IV, by indicating a weak response of bird communities to outbreak-induced forest damage. In 1 of my 2 main study areas (Kirkenes), the total abundance of birds was roughly 25 % lower in damaged than undamaged forest. Bird species-richness showed an even smaller reduction in damaged forest. Meanwhile, in the other study area (Tana), there were no consistent differences in bird abundance or richness between damaged and undamaged forest. The observed reduction in bird abundance in damaged forest in Kirkenes was mainly driven by the Willow warbler (Phylloscopus trochilus); a foliage gleaning species which may have suffered loss of foraging habitat due to outbreak-induced mortality of trees. By documenting a weak response to forest damage in the studied bird community, study V suggested that this community has a high degree of resistance to the habitat disturbance caused by outbreaks. This may be explained by the fact that many of the studied bird species are habitat generalists. It was also suggested that bird populations in the outbreak area might have been maintained by surviving trees and by standing birch trunks, which could serve to maintain the vertical structure of the forest habitat. Study IV and V were limited in their conclusions by their short-term time perspectives. Thus, both studies highlighted the need for more long-term research on the responses of animal communities to outbreak-induced forest damage in the mountain birch ecosystem