Understanding internal and external effects on the composition of marine ecosystems

Andrea Tabi, Institute for Cross-Disciplinary Physics and Complex Systems (IFISC), Palma, Spain, Luis J. Gilarranz, Swiss Federal Institute of Aquatic Science and Technology (Eawag), Dübendorf, Switzerland and Serguei Saavedra, Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, MA

Background/Question/Methods
Understanding the factors modulating the composition of marine ecosystems is paramount for improving restoration and sustainability strategies. In this direction, size-based ecosystem modeling has had a long tradition in marine research. Body size is considered as a 'master trait' scaling with many physiological and behavioral traits, which has been useful in approximating the composition of marine ecosystems. However, the extent to which internal (such as the particular collection of species in a community) and external factors (such as protected areas or temperature variability) affect species composition has not been properly measured (i.e., what body size cannot capture). Yet, this is critical in the face of global environmental change, which continues to cause biodiversity loss and climatic changes. To address this gap, we estimate internal effects by approximating species interactions using a niche overlap framework that estimates the intersection between species’s body size density functions. We integrate these interactions under a Lotka-Volterra dynamics in order to analytically estimate the community composition given a collection of species. We then quantify the extent to which this methodology explains what body size alone cannot. Then, we identify external factors (e.g. climatic variables, human impact, site-specific factors) linked to the capacity of species interactions (removing the effect of body size) to explain community composition.
Results/Conclusions
We applied our methodology to the FishBase database containing 3464 sampling sites around the planet, where population data of 3086 non-benthic marine species are available with individual body size information. As a validation of our methodology, we found a high association of the analytically estimated composition and the observed composition for the majority of sites. Additionally, we found that once we control for body size, the association between the estimated and observed compositions displayed a large variability. This variability was not driven by community size or sampling effort. Importantly, we found that poor associations were linked to sites subject to harsh environmental conditions, revealing the conditions under which external factors may overrule the effect of internal effects, i.e. species interactions. Our findings open new possibilities to differentiate the role of internal and external factors affecting the composition of marine ecosystems.