Many factors can directly affect species abundances in aquatic ecosystems. However, the effects of key factors such as temperature or salinity on the abundance of a particular species can sometimes be unclear if its abundance is also affected by other species in the ecosystem. This complexity can make it difficult to predict how an ecosystem might respond to environmental change, so it is important to understand how species in an ecosystem interact with one another before conclusions can be drawn about the effects of other factors on species abundance.
The existence and strengths of species interactions in an aquatic ecosystem can be determined by examining how the abundance of each species changes over time in comparison to the abundances of all the other species in the system. A statistical analysis called a multivariate autoregressive (MAR) model has been designed to do just that. This analysis works well with detailed datasets composed of frequent abundance measurements for many species over a long period of time. In aquatic ecosystems, datasets that focus on planktonic species are particularly ideal for analysis because of plankton’s importance in aquatic food webs, their relatively rapid response to change, and the lack of direct human harvesting effects on their numbers.
The MAR model has been applied successfully to long-term freshwater plankton datasets to reveal species interactions in those ecosystems. However, it is more difficult to use this analysis with datasets from marine ecosystems. Plankton sampling in marine systems is often done less frequently and involves more error than in freshwater systems, so the resulting datasets are not ideal for MAR model application. The goals of this project are to adjust the MAR model so it can be used with lower-quality marine datasets and to test the new model on existing marine datasets. The successful characterization of marine species interactions through the use of this adapted model will help us better understand how various marine ecosystems function and differ from one another. This understanding will in turn help us develop better strategies for the management of marine ecosystems.