Environmental rather than spatial factors determine trait-based community structure: a case study with zooplankton
Ecological community structure, which has traditionally been described in terms of taxonomic units, is driven by a combination of dispersal and environmental filters. Traits have recently been recognized as alternative units for quantifying community parameters, but they may have important differences with taxonomic units. For example, as taxon-based community structures are determined by the identities of individual species, they may be more limited by species dispersal constraints, whereas trait-based community structures may be more regulated by environmental conditions because functional traits should have stronger linkages to habitat types rather than their specific locations. This suggests that the relative importance of dispersal and environmental filters may vary depending on the units employed to define community structure, yet how the contributions of these filters compare remains unclear. Zooplankton in lakes and reservoirs are an ideal system for examining the relative importance of these filters as they have moderate dispersal ability and inhabit communities in isolated habitats. In this study, we examined zooplankton assemblages in 87 artificial reservoirs throughout the Japanese archipelago to test the hypothesis that environmental filters play a more dominant role in determining trait-based community structures than taxon-based structures. To describe trait-based communities, we categorized zooplankton taxa into 12 functional groups using ecologically meaningful traits, including body size, feeding habits and mode, and population growth rate. We then examined the effects of both spatial configuration (reflecting dispersal filters) and environmental variables such as reservoir size and depth, trophic conditions, and fish assemblages for taxon- and trait-based community structures. Although variation in the taxon-based structure was explained equally well by spatial and environmental variables, variation in the trait-based structure explained by environmental variables was nearly twice that of spatial variables. These results support the idea that environmental filters play a more central role in determining trait-based community structures, and show that the relative importance of spatial and environmental filters changes with the way we define community structure. Open research statementAll the data used in this study was obtained from disclosed sources including the Database of Dams in Japan (http://mudam.nilim.go.jp/home) for the location and watershed areas; the Japan Dam Foundation (http://damnet.or.jp/) for the dam specification data; and the River Environmental Database (Red: http://www.nilim.go.jp/lab/fbg/ksnkankyo/index.html) for water chemistry, fish, and zooplankton. Data obtained from local reports of the national reservoir management office, Japan, can be provided by the authors upon request. All analyses were conducted in the R programming language and all functions used in the present study originate from existing packages.