SYMP 13-4 - Indirect risks to ecosystem functioning through coextinctions in ecological networks

Changes in biodiversity are degrading ecosystem functions and services. Most studies on how biodiversity loss alters functioning, typically primary productivity, have focused on a single trophic level, though increasing attention is being paid to functions provided at multiple trophic levels. However, many such studies fail to recognize the knock-on effects of the loss of species and functioning for the broader food web in which these species are embedded. For example, the loss of a basal resource species may lead to coextinctions of consumers and higher trophic levels, with unknown consequences for the overall functioning of the food web. In other words, we seek to understand: if a food web collapses, will its ecosystem function be maintained?

In this study, we develop a new framework for predicting the responses of multiple ecosystem functions to coextinctions and implement this framework using a case study from the Gulf of Maine (GOM) rocky intertidal food web. We first develop a conceptual framework for direct and indirect functional redundancy in ecological networks. While 'direct functional redundancy' refers to the number of species contributing to the ecosystem function directly, we propose a new concept of 'indirect functional redundancy' where a function is indirectly redundant if the species directly performing that function relies on different resource species. These resource species are thus indirectly important for supporting ecosystem function, and we will discuss ways to computationally identify these supporting species in a network.

We then implement this approach using a newly assembled food web for the GOM intertidal with 276 species and simulated food web dynamics. We measured three multitrophic ecosystem functions (annual primary production, annual secondary production, water filtration) in this system using field studies on Allen Island, Maine. The functions were then embedded as nodes in the food web, linked to the species providing them. Species losses were then simulated on the food web, from which we predicted a community shift that reduced water filtration potential, but increased secondary production of herbivores. Our approach can help identify species that both directly and indirectly perform important functions and services, to increase awareness of their role in the system.