Aggregating taxa and the influence of scale: Potential concerns for analysis of stability using functional measures

Abstract A functional group approach to analyzing ecological processes calls for aggregating species populations' data (e.g., biomass, abundance) into larger functional units. This method offers a more direct insight into the coarse community and ecosystem dynamics than a species level. Since aggregation changes the scale of analysis, it is reasonable to expect that the results and inferences will also change. To gauge the nature and size of this effect, we examined two sets of communities using the same methodology at taxonomic and functional levels. We asked how asynchrony among constituent populations stabilizes the regional biomass or abundance in two systems: zooplankton in North American temperate lakes and rock pool invertebrates in Jamaica. We aggregated species into pairs by organism size for the functional level analysis because it correlates well with life history traits. We hypothesized that taxonomical and functional level analyses would yield different pictures and inferences. We analyzed asynchrony among local populations, metapopulations, and populations of different species at different locations. We found that in lakes, different asynchrony classes contributed to the stabilization of the regional metric than in the rock pools. We further found that aggregation of taxa into functional units changed the pattern of differences between lakes and rock pools revealed by taxonomical analysis and that each set of communities responded differently to species lumping. While aggregating taxa leads to different results, the taxonomic and functional lenses add new insights to the biological interpretation of mechanisms stabilizing communities when interpreted jointly. Specifically, population asynchrony—a known stabilizing factor—can provide complementary stabilizing mechanisms when seen through taxonomic and functional lenses. Without exposing these modes, inferences about abundance/biomass/richness stabilizing contributions arising from asynchrony will remain tenuous. Exposing these modes offers a path for significant breakthroughs in linking local, spatial, and functional structures to multispecies system dynamics.