Summary: | 1.Warmer summer temperatures and enhanced soil fertility increase shrub growth in tundra ecosystems, and these factors have likely contributed to shrub expansion at the circumpolar scale over the last decades. Conversely, large herbivores have the potential to counteract the positive impacts of climate change on shrub growth. Indeed, by stripping the leaves, herbivores have the potential to control the growth of shrub species and, consequently, limit their expansion. 2.To disentangle the impacts of climate change and herbivory on Betula glandulosa Michx., we conducted a 5-yr factorial experiment near Deception Bay, Nunavik, Canada, in which we simulated warmer temperatures, increased nitrogen availability and three caribou browsing intensities during the growing season. At the end of the experiment, we harvested the aboveground biomass of B. glandulosa and conducted dendrochronological analyses on stems. 3.Fertilized plots under ambient temperature had 34% greater shrub biomass than plots assigned to the combined treatment of nitrogen addition and warmer temperatures. Browsing intensity had no effect on final biomass. Nitrogen addition increased radial growth (18-33%; 3 years out of 5). Overall, browsing had a cumulative negative impact on B. glandulosa radial growth during the 5-yr experiment. While browsing had no effect in the first year of the experiment, moderate browsing (leaves stripped on 25% of available shoots) decreased radial growth by 27% at year 2, 32% at year 4 and 27% at year 5. Heavy browsing (leaves stripped on 75% of available shoots annually) decreased radial growth by approximately 27% at year 2, 37% at year 3, 50% at year 4 and 48% at year 5. We did not observe significant interactions between browsing, temperature and nitrogen availability. 4.Synthesis. Our results clearly showed that caribou browsing may limit the growth of B. glandulosa, and thus can potentially limit its expansion. Herbivory should thus be considered when predicting tundra vegetation changes in the Arctic, at least in ...
|