Intraspecific trait variation improves the detection of deterministic community assembly processes in early successional forests, but not in late successional forests

Aims: Intraspecific trait variation (ITV) has been increasingly recognized to play an important role in understanding the underlying processes influencing community assembly. However, gaps remain in our understanding of how incorporating ITV will influence the relative importance of deterministic (e.g. habitat filtering, limiting similarity) and stochastic processes in driving community assembly at different successional stages. Methods: We used data for eight functional traits from 55 woody species in early (24 ha) and late (25 ha) successional temperate forest plot in northeast China. We employed an approximate Bayesian computation approach to assess the relative contribution of stochastic processes, habitat filtering and limiting similarity in driving community structure. We then compared the results with and without intraspecific trait variation to investigate how ITV influences the inferred importance of each process. Important Findings: We found that when analyzing interspecific trait variation only (i.e. without ITV), stochastic processes were observed most frequently in driving community composition, followed by habitat filtering and limiting similarity in both forests. However, ITV analyses showed that the relative importance of both deterministic processes (habitat filtering and limiting similarity) increased in early successional forest, but remained virtually unchanged in late successional forest. Our study reveals the distinctive influence of ITV on the inference of underlying processes in a context of succession and reinforces the need to estimate ITV for making correct inferences about underlying ecological processes.