Structural equation modeling approach to the diversity–productivity relationship of Wadden Sea phytoplankton

Biodiversity–ecosystem functioning (BEF) research has been a major topic in ecology for over 2 decades, and recent meta-analyses have confirmed biodiversity to be a driver of ecosystem processes and services. To date, the vast majority of BEF studies have been conducted experimentally, and it is unclear whether their outcomes can be transferred to natural communities and ecosystems. The major challenge faced in the analysis of observational data is to incorporate direct and indirect processes which influence the response variable of interest. Consequently, the statistical methods used to analyze such relationships must accommodate the multivariate nature of these data. One multivariate approach, viz. structural equation modeling, has already been applied to BEF research in terrestrial and freshwater ecosystems. In this study, we applied a structural equation model to monitoring data on marine phytoplankton communities, including data on environmental parameters, community structure, and measures of productivity. Our aim was to ascertain whether similar patterns and processes driving BEF relationships as described for other ecosystem types are evident in marine phytoplankton communities. We found that different aspects of biodiversity (richness, evenness) are significantly linked to ecosystem functions (productivity, resource use efficiency). These relationships are embedded in a multitude of direct and indirect links between environmental factors, community diversity, and productivity. Overall, our analysis confirms patterns observed in terrestrial and freshwater ecosystems and highlights the importance of incorporating multivariate methods for a better understanding of BEF processes in natural ecosystems.