An ecological network approach to predict ecosystem service vulnerability to species losses

Aislyn Keyes, Ecology and Evolutionary Biology, University of Colorado, Boulder, CO, John P. McLaughlin, Ecology, Evolution and Marine Biology, University of California at Santa Barbara, Santa Barbara, CA, Allison K. Barner, Department of Environmental Science, Policy, & Management, University of California, Berkeley, Berkeley, CA and Laura Dee, University of Colorado-Boulder, Boulder, CO

Background/Question/Methods
Human-driven threats are changing biodiversity, impacting ecosystem services. The loss of one species can trigger secondary extinctions of additional species, because species interact -- yet the consequences of these secondary extinctions for ecosystem services remain underexplored.
Here, we examine how species losses lead to direct and indirect threats to ecosystem services by extending a network property— robustness, which measures the tolerance of a food web to species loss— to ecosystem services. We measure ecosystem service robustness as the tolerance of ecosystem services to species loss, allowing us to compare food web robustness to ecosystem service robustness. We ask: (1) Is food web robustness correlated with ecosystem service robustness across different patterns of species extinctions? (2) Does the robustness to species losses vary across ecosystem services? And (3) Are the species that contribute to ecosystem services critical to food web persistence? We apply robustness analysis to ecological networks augmented to include 7 ecosystem services (e.g., fishery, birdwatching) under 12 patterns of species loss. We utilize data from three salt marsh food webs in California, USA and Baja, Mexico. We hypothesized that ecosystem services robustness is related to food web robustness, and that this relationship varies based on the pattern of species lost. Further, we hypothesized that ecosystem service providers- critical to ecosystem services - are not critical for food web robustness, but that supporting species are important for both food web and ecosystem service robustness.
Results/Conclusions
We found that food web and service robustness are highly correlated, but that the robustness of both the food webs and ecosystem services vary depending on the pattern of species losses. We found that ecosystem services with high numbers of species providing them and low trophic levels were more robust to species losses than those with few species providing them at higher trophic levels. Ecosystem service providers (species directly providing services) were not critical to food web robustness, but supporting species (species indirectly supporting services through interactions) were critical to both food web and ecosystem service robustness. Together, these results reveal indirect risk to ecosystem services through secondary species losses and predictable differences in vulnerability across ecosystem services.