Accounting for temporal change in multiple biodiversity patterns improves the inference of metacommunity processes

Area of Science:

• Metacommunity ecology
• Community dynamics
• Ecological assembly processes

Background:

• Metacommunity research often overlooks temporal dynamics, focusing on spatial patterns.
• Key assembly processes include dispersal limitation and environmental filtering.
• Understanding these processes is vital for predicting community structure.

Purpose of the Study:

• To develop a "virtual ecologist" approach for simulating and evaluating metacommunity assembly processes.
• To assess the utility of spatial and temporal community structure statistics for inferring assembly processes.
• To determine the necessity of temporal data for disentangling ecological processes.

Main Methods:

• Simulated metacommunities with varying density-independent, density-dependent, and dispersal processes.
• Calculated spatial and temporal community structure summary statistics.
• Employed random forests to evaluate the explanatory power of statistics for inferring assembly processes.

Main Results:

• Both spatial and temporal data are essential for disentangling metacommunity processes.
• Temporal data increased the explanatory power of random forests by up to 59%.
• Different statistics effectively distinguish between density-independent, density-dependent, and dispersal processes.

Conclusions:

• Temporal data, especially repeated observations, is critical for inferring metacommunity assembly, particularly dispersal.
• Statistics like coefficient of variation in species abundances and evenness metrics are highly informative.
• A combination of methods and statistics is necessary for a comprehensive understanding of metacommunity assembly through space and time.