COS 14-1 - Quantifying long-term changes in rocky intertidal community structure of the Gulf of Maine using network-based metrics

Abstract: Global ecosystems are facing increasingly intense and frequent disturbances, including climate change, overexploitation, and habitat loss. Because species interact and depend on each other, the negative impacts of disturbances can spread from one species to another, creating a “domino effect” that alters species abundances and the composition of the entire community. Thus, recognizing the patterns of interactions among species (i.e., network structure) is critical for understanding how communities transform over time and for predicting community-level responses to disturbances. However, characterizations of empirical changes in network structure over time have been rare due to the scarcity of long-term repeated community surveys paired with associated data on trophic interactions.

We quantify how network structure has changed between 1972-2014 in the Gulf of Maine (GOM) rocky intertidal ecosystem. In addition to being one of the most rapidly warming ecosystems in the world, it also experiences frequent disturbances such as marine heat waves. This ecosystem is particularly well-studied, with repeated observational data collected at sites ranging from Massachusetts to Northern Maine, allowing us to examine long-term changes in network structure. To do so, we standardized and combined community survey data from multiple sources to create a long-term dataset recording species abundances at study sites across the GOM. We also conducted an extensive literature review to compile data on over 1,400 trophic interactions among species in the GOM rocky intertidal ecosystem. Combining these two datasets, we constructed a time series of food webs for each study site based on the observed species and their known interactions. We then analyzed these food webs to identify trends in network structure along multiple dimensions.

We find that, in general, communities observed between 1972-76 differed significantly from those observed between 2013-14 in multiple aspects of network structure, including: the number of species present (p < < 0.001), connectance (p < < 0.001), and transitivity (p < < 0.001). After accounting for site-level fixed effects, we find that connectance has decreased over time (p < 0.05). We also find that the number of species present has increased over time while transitivity decreased, though these effects were not statistically significant. Our analysis reveals how communities in the GOM have changed over the last 50 years. Previous work has shown that connectance and transitivity are both positively correlated with community stability, suggesting that rocky intertidal ecosystems are more vulnerable than ever to disturbance.