Linking multidimensional functional diversity to quantitative methods: A graphical hypothesis-evaluation framework

Functional trait analysis is an appealing approach to study differences among biological communities because traits determine species' responses to the environment and their impacts on ecosystem functioning. Despite a rapidly expanding quantitative literature, it remains challenging to conceptualize concurrent changes in multiple trait dimensions ("trait space") and select quantitative functional diversity methods to test hypotheses prior to analysis. To address this need, we present a widely applicable framework for visualizing ecological phenomena in trait space to guide the selection, application, and interpretation of quantitative functional diversity methods. We describe five hypotheses that represent general patterns of responses to disturbance in functional community ecology and then apply a formal decision process to determine appropriate quantitative methods to test ecological hypotheses. As a part of this process, we devise a new statistical approach to test for functional turnover among communities. Our combination of hypotheses and metrics can be applied broadly to address ecological questions across a range of systems and study designs. We illustrate the framework with a case study of disturbance in freshwater communities. This hypothesis-driven approach will increase the rigor and transparency of applied functional trait studies.