EFFECTS OF HERBIVORY ON ARROWGRASS: INTERACTIONS BETWEEN GEESE, NEIGHBORING PLANTS, AND ABIOTIC FACTORS

Abstract

Herbivores may affect plants by removing biomass, altering competitive interactions, and altering the abiotic environment. Changes in the size and quality of forage species and in species composition as a result of herbivory, in turn, affect future herbivory. We investigated the direct and indirect effects of herbivory by Brant Geese (Branta bernicla nigricans) on Triglochin palustris (arrowgrass) in a subarctic salt marsh in southwestern Alaska. In the first experiment, we compared arrowgrass in exclosed plots, unexclosed plots with feces removed, and control plots. In the second experiment, we used a full-factorial design to examine the effects of clipping arrowgrass, clipping neighboring plants, depositing goose feces, and their interactions on arrowgrass size and biomass allocation. In the third experiment, we placed hand-reared goslings on premanipulated plots from the second experiment to examine the effects of arrowgrass size, density, and species composition on the probability of an individual arrowgrass plant being grazed. For unclipped plants, feces deposition resulted in reduced bulb mass, reduced percentage of biomass in bulb and roots, and increased percentage of biomass in leaves, whereas feces deposition had no effect on clipped plants. Clipping neighbors resulted in increased arrowgrass root and stolon mass only for unclipped plants. Feces deposition resulted in increased vegetative reproduction when neighbors were clipped, but had no effect on vegetative reproduction when neighbors were not clipped. Plants in exclosed plots were larger, had greater allocation to leaves, had higher concentrations of C and N, and were more likely to flower than were plants in unexclosed plots. These results indicate an increase in competition for light with neighbor plants under feces deposition, which may be ameliorated by biomass removal of neighbors. Our results predict that an increase in grazing pressure is not necessarily detrimental to arrowgrass, provided that it is accompanied by increases in consumption of neighbor plants. The number of arrowgrass plants completely removed was not related to arrowgrass density, but the number of plants partially grazed increased with arrowgrass density. The probability that an individual arrowgrass plant would be grazed was negatively related to biomass or percent cover of several other species (Potentilla egedei, Chrysanthemum arcticum, Carex spp., Salix spp.). These results suggest that some neighbor species may provide a measure of protection (associational refuge) from herbivory, and that an increase in grazing intensity may have a strong negative effect on arrowgrass populations by reducing this protection and through an increased likelihood of consumption of the arrowgrass remaining after neighbors are removed. We conclude that the way in which we view neighboring plants (as competitors or potential protectors) affects our predictions regarding the effects of changing herbivore populations. Feedbacks from the plant community to herbivores following grazing should be included in studies that aim to extrapolate to the population level the effects of herbivory on individuals.