Variation in the phylogenetic diversity of wild bees at produce farms and prairies

• Published in: Agriculture, ecosystems & environment Vol. 259; pp. 168 - 173
• Main Authors: Hendrix, Stephen D., Forbes, Andrew A., MacDougall, Caitlin E.D.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.05.2018; Elsevier BV
• Subjects: Abundance; Andrena; Andrenidae; Apoidea; Bees; Clustering; Communities; Composition; Data processing; Disturbance; Farms; Genetic diversity; Landscape; Lasioglossum; Nesting; Permutations; Phylogenetic diversity; Phylogenetics; Phylogeny; Pollinators; Prairie; Prairies; Produce farm; Species; Lasioglossum; Produce farm; Phylogenetic diversity; Andrena; Apoidea; Prairie
• ISSN: 0167-8809; 1873-2305
• DOI: 10.1016/j.agee.2018.03.005

•We used a phylogenetic approach to examine bee communities at farms and prairies.•Mean Phylogenetic Distance was significantly clustered at five of six farms.•Clustering at farms was due to a lack of Andrenidae and abundance of Halictidae bees.•Mean Nearest Taxon Distance did not show consistent differences between site types.•Phylogenetic approaches appear useful in measuring differences among bee communities. Declines in pollinators, particularly wild bees, along with rising demands for their services has intensified efforts to examine bee communities in different types of habitats. In this study we use a phylogenetic approach to compare and contrast bee communities associated with six small produce farms, eight large prairies, and five naturally small hill prairies in Iowa. We compare the mean phylogenetic distance (MPD) and the mean nearest taxon distance (MNTD) of bees in each community to expected mean MPD and MNTD values generated from 1000 random permutations of a tree composed of 144 species found across all sites. Standardized effect size scores for MPD using presence-absence data showed significant clustering of bee communities at five of six farms and two of the five hill prairies. Clustering at the farms and hill prairies was due primarily to the significantly lower number of species in the Andrenidae (0–5 species per site), especially species of Andrena, as well as significantly greater number of species in the Halictidae (9–22 species/site), particularly Lasioglossum (Dialictus). Lack of Andrenidae spp. may be related to a lack of appropriate floral resources, indicating that enriched prairie plantings for pollinators at farms could enhance the abundance of Andrena species and hence pollinator services at these sites. The higher richness of the ground-nesting Lasioglossum likely results from soil disturbance regimes at produce farms and the naturally shallow, rocky soil with exposed surface at hill prairies, respectively. Analyses of MNTD using either abundance weighted or presence-absence data and analyses of MPD using abundance data did not indicate consistent differences between the three site types, but do point to important differences between sites in phylogenetic composition of bee communities. Our results show that phylogenetic analyses of wild bee community diversity may be a useful tool for measuring how bee communities differ in composition as a result of natural variation and human-related changes in landscapes.