Traits and functional diversity of a bee assemblage are linked to aridity

Authors:
Publisher website
Google Scholar
Abstract
Climate change in the American Southwest is reshaping species assemblages. However, the resulting patterns in mean trait values and functional diversity are poorly understood. Bee assemblages in Southwestern drylands are exceptionally diverse, with species varying greatly in their morphological traits. In this study, we addressed two questions: Do climate variables associated with aridity predict community-weighted mean trait values? Has bee assemblage functional diversity declined under increasing aridity? To answer these questions, we utilized 16 years of abundance data for 33 focal bee species within a hyperdiverse bee assemblage at the Sevilleta National Wildlife Refuge (NM, USA), combined with measurements of a suite of five morphological traits hypothesized to mediate climate change responses: body size, wing size, lightness, face hairiness, and thorax hairiness. Our models reveal that community-level responses to climate are trait-specific and vary seasonally: aridity is more predictive in the spring, while temperature and precipitation are more predictive of trait shifts during the monsoon season. With less precipitation, higher temperatures, and increased aridity, bee species that have large bodies, large wings, are lightly colored, and are hairy appear favored. The functional richness of the Sevilleta bee assemblage increased with aridity in the spring and decreased with precipitation in the monsoon season. Functional evenness and functional divergence did not co-vary with any climate variable. We conclude that characteristics of bee assemblages such as community average trait values and functional diversity may change predictably with the increased aridity forecast for this region in the future.
Funding Information
  • Sevilleta Long Term Ecological Research and the University of New Mexico Department of Biology Program (NSF DEB 1655499)

This publication has 71 references indexed in Scilit: