2017 ESA Annual Meeting (August 6 -- 11)

COS 145-4 - Seaweed subsidies influence island mouse ecology

Thursday, August 10, 2017: 2:30 PM
E142, Oregon Convention Center
Katie Davidson1,2, Brian M. Starzomski3,4, John D. Reynolds4,5, Rana W. El-Sabaawi6, Morgan Hocking3,7 and Chris T. Darimont1,4,8, (1)Geography, University of Victoria, Victoria, BC, Canada, (2)Hakai Institute, Heriot Bay, BC, Canada, (3)School of Environmental Studies, University of Victoria, Victoria, BC, Canada, (4)Hakai Institute, Quadra Island, BC, Canada, (5)Biological Sciences, Simon Fraser University, Burnaby, BC, Canada, (6)Biology, University of Victoria, Victoria, BC, Canada, (7)Ecofish Research Ltd., Victoria, BC, Canada, (8)Raincoast Conservation Foundation, Sidney, BC, Canada
Background/Question/Methods

Marine subsidies can shape the diets of terrestrial consumers, but the pathways of subsidy can be complex. In many areas, seaweed ("wrack") deposition provides a year-round subsidy to shorelines, fueling a food web of grazers (such as Talitrid amphipods) and potentially higher trophic level consumers. The responses of these consumers, particularly omnivorous terrestrial mammals, is still unclear. To examine this potentially complex pathway across ecosystem boundaries, we asked the following: i) do coastal mice consume marine-derived prey; ii) do diets differ among sites and islands that offer different amounts of terrestrial foods and marine subsidy and, if so; iii) what factors drive dietary differences? To address these questions, we sampled hair from Keen’s mice (Peromyscus keeni) from both a museum collection (c. 1930-1950) and live-trapped (2016) animals, with a focus on two islands on either end of the subsidy gradient in the Great Bear Rainforest of British Columbia, Canada. We used δ13C and δ15N signatures of mouse hair and their potential foods in a Bayesian isotope mixing model to determine the proportion of marine-derived resources in mouse diets, followed by a Generalized Linear Mixed-effects Model to determine ecological drivers of marine food consumption.

Results/Conclusions

Historical mouse specimens from islands with more wrack (n = 11) had significantly higher δ15N signatures than those with less wrack (n = 63; Wilcoxon Test, W = 16, p < 0.01), but exhibited no significant differences in δ13C signatures (Wilcoxon Test, W = 294, p = 0.47). Bayesian isotope modelling indicated that “subsidized” historical specimens had consumed a higher proportion of marine-derived prey (93% of diet) compared to those from less subsidized islands (41% of diet). In contemporary mice, individuals from islands with more wrack (n = 25) also had significantly higher δ15N signatures than those from an island with less wrack (n = 30; Wilcoxon Test, W = 220, p = 0.009), but exhibited no significant differences in δ13C signatures (Wilcoxon Test, W = 308, p = 0.26). Our preliminary analyses suggest that the degree of marine subsidy to an island will influence omnivorous mouse diets. We will show how variation in prey biomass, gender, reproductive status, and proximity to the shoreline predicts consumption of marine foods.