Abstract
Aim Determining why community composition changes from place‐to‐place (spatial β‐diversity) is central to advancing biodiversity theory and conservation of threatened species. Spatial β‐diversity can occur most simply due to random sampling effects from the larger species‐pool, such as the random placement and distribution of individuals and species. However, ecological processes can also generate intraspecific aggregation changes that also produce changes in spatial β‐diversity. We decomposed changes in β‐diversity among birds in geographically isolated wetlands to understand the relative contributions of intraspecific aggregation and random sampling processes. Specifically, we compared two analytical approaches from the Measurement of Biodiversity framework: multi‐metric versus multi‐scale. We also examined how estimates of β‐diversity are sensitive to changes in spatial extent and environmental context. Location Continental United States. Time Period 2010–2024. Major Taxa Studied Birds. Methods We integrated a dataset of n = 207 geographically isolated wetlands with eBird data from 2010 to 2024 from across the continental United States to investigate how β‐diversity changes in these wetland systems at two different spatial extents: (1) within ecoregions and (2) between wetlands continentally, regardless of ecoregion. We used both the multi‐metric and the multi‐scale approaches of the Measurement of Biodiversity framework, which account for changes in abundance and the species‐abundance distribution, to test if β‐diversity in these systems reflects intraspecific aggregation across ecoregions. Results The multi‐metric β‐diversity approach was largely consistent with random sampling effects due to variation in community size and the species‐abundance distribution at both ecoregion and continental extents evidenced by low β C . However, our multi‐scale β‐diversity analysis using spatially explicit rarefaction curves revealed a strong divergence from the random sampling effects indicating that β‐diversity is also due to intraspecific aggregation generated by ecological processes both at the ecoregion and continental extents. Main Conclusions Our two analytical frameworks uncovered different explanations for spatial β‐diversity in wetlands. The β‐diversity metrics did not differ from random sampling presumably due to the homogenous distribution of a few hyperabundant species. However, the spatial accumulation of species was slower than expected under random sampling which indicates that ecological processes, such as environmental filtering and/or dispersal limitation, are driving nearby wetlands to have more similar composition than expected. Both of these signals were present within and across ecoregions, highlighting the need for a pluralistic approach to analysing β‐diversity and disentangling these non‐mutually exclusive processes.