Biogeochemistry in Subarctic birch forests : Perspectives on insect herbivory

Herbivory can influence ecosystem processes, partly through long-term changes of the plant community compositions, but also more rapidly through the herbivores’ digestive alteration of the organic matter that is cycled through the soil and back to the primary producers. In the Subarctic mountain birch (Betula pubescens ssp. czerepanovii) forest (SMBF) in Northern Fennoscandia, outbreaks by the geometrid moths (Epirrita autumnata and Operophtera brumata) are well-described, widespread, and increasing with global warming. In contrast, the ecosystem effects of background insect herbivory (BIH) in this ecosystem lacks quantification, although belowground responses to aboveground perturbations in high-latitude systems may accelerate global warming due to their storage of large terrestrial organic carbon (C) pools. We quantified the ecosystem impact of BIH in the SMBF of Northern Sweden. An initial literature review showed that the clear increase in organic matter turnover rates under insect infestations was primarily driven by outbreak conditions. In line with this, our conversion of an average BIH-rate of ~1.6% of the leaf area to annual canopy-tosoil fluxes of nitrogen (~3.5% N) and phosphorus (~2.0% P) showed that the background rates were relatively small compared to internal recycling through litter, and inputs from external sources, such as atmospheric deposition, biological fixation and weathering.In addition, we showed that the insects themselves efficiently conserve N, as 70-80 % of the ingested N was converted to insect biomass, while respiring 30-50% of the ingested C. When insect excreta (frass) was added to the soil, we showed that another ~30 % of the C was respired by soil organisms. Hence, a total of ~60 % of the C ingested by insect herbivores would be respired during the first growing season, compared to ~10 % of the C added as senesced litter, suggesting a decreased litter C-sink in soils during outbreaks. In microcosm incubations, frass addition stimulated fungal growth more than bacterial growth while litter addition showed the opposite relationship. In contrast, under non-outbreak conditions along natural environmental gradients in the SMBF, decomposer bacterial growth was strongly correlated with BIH and other indicators of labile organic substrates, while fungal growth showed very little correlation with the potential driver variables. Yet, BIH did not explain a significant portion of the variation in the fraction of microbially assimilated C that was incorporated into soil microbial biomass, i.e. the soil carbon use efficiency (CUE). CUE was strongly controlled by respiration, but when this was controlled for, it increased with both bacterial and fungal growth rates. Further, CUE decreased with increasing soil temperature and the size of the soil microbial biomass pool. This suggests decreased soil C-sequestration with global warming, although an associated decrease in microbial biomass, which is often observed in warming experiments, may moderate this effect. Finally, gross N-mineralisation was also substantially higher after addition of insect frass (~17 % of added N) compared to litter (lower than control), so the availability of mineral N is higher under insect outbreaks increasing the risk of leaching losses.Finally, we challenged the assumption underlying space-for-time substitution studies, i.e. that variation along natural elevational gradients is scale invariant and universal, by showing that e.g. BIH exhibit contrasting trends with local and regional elevation. Although explorative, these findings merit further considerations of when spacefor-time-substitution is a feasible tool for inferring ecosystem responses to environmental change