Plant-herbivore interactions on the Yukon-Kuskokwim Delta: The effect of goose herbivory on arrowgrass

Abstract

I examined effects of herbivory by black brant geese (Branta bernicla nigricans) on the small herbaceous perennial Triglochin palustris (arrowgrass) in a subarctic saltmarsh in SW Alaska. I investigated effects of biomass removal, and indirect effects of geese (changes in resource availability and competition) to compare the role of selective herbivory in this mixed-species environment with that of herbivory in monospecific saltmarsh communities. I manipulated nutrient availability, light availability, and salinity in a transplant experiment, and manipulated size of arrowgrass, and neighbor size and feces deposition in exclosure experiments. Additional experiments examined relationships between size, biomass allocation, survival and reproduction, and explanations for low rates of sexual reproduction in arrowgrass. A cellular automata model was used to investigate potential long-term effects of changes in grazing intensity. Direct effects of geese were smaller than indirect effects: biomass removal had little effect on rates of population growth or plant size, and resulting changes in biomass allocation did not affect survival or reproduction. For unclipped arrowgrass, feces deposition resulted in increased competition for light, which was ameliorated by consumption of neighboring plants, but some species may provide protection from grazing. Expansion into neighboring communities is limited by physical factors on the sea-side end of the distribution, and by competition for light and high selectivity on the inland end. Overall effects of changes in grazing pressure will depend on changes in goose foraging behavior and selectivity. Trade-offs exist between sexual reproduction and all other functions, and sexual reproduction may increase risk of herbivory. Goose effects occur at several spatial and temporal scales: immediately (through biomass removal), within a growing season (through changes in competition and resource availability), over several growing seasons (through feedbacks to foraging behavior), and over long periods (through changes in reproduction). Model results suggest increased grazing intensity may not decrease arrowgrass populations under some conditions, and that spatial distribution of geese affects population dynamics of arrowgrass. There is no evidence that feces deposition results in greater productivity of preferred species. More detailed knowledge of goose foraging behavior at several spatial and temporal scales is needed in order to understand the dynamics of this system.