COS 33-2 - Pollinator services in fragmented landscapes: Variation in wild bee community composition between farms and prairies

Tuesday, August 8, 2017: 8:20 AM
B110-111, Oregon Convention Center
Stephen Hendrix, University of Iowa and Andrew A. Forbes, Department of Biology, University of Iowa, Iowa City, IA
Background/Question/Methods

Increasing alarm over loss of pollinator services has increased pressure to better understand the distribution of wild bee pollinators across different landscape elements and how pollinator services can be enhanced. In this study we use a phylogenetic approach to examine for potential differences in bee communities at six produce farms, eight large prairie preserves, and five naturally small prairies in Iowa. These bee communities were sampled using identical methods consisting of pan-trapping and timed netting from flowering plants in one-hectare plots. We sampled replicate plots at seven of the large prairies. We constructed a phylogenetic tree of all bee species at these sites (150 species, 38 genera) using previously published phylogenies of wild bees and then constructed null models to compare against the communities from the 26 sample plots examined. We examined MPD (Mean Phylogenetic Distance) and MNTD (Mean Nearest Taxon Distance) for both binary and abundance weighted data.

Results/Conclusions

Our analysis reveals that five of the six farms and one small prairie were significantly more clustered than expected using binary MPD data, largely because these sites lacked Andrena species and Andrenid bees in general except for one species (Calliopsis andreniformis). Samples from large prairie preserves and most small prairies typically had 2-8 Andrenidae species. These differences between farms and prairies largely disappeared with abundance weighted MPD data because Andrena are not abundant at farm sites (6% of all individuals) and their loss had little impact on abundance-weighted values. Binary MNTD analysis showed only one site in each category as significantly clustered; these differences were weakened using abundance weighted MNTD values. The absence of Andrena spp. at farm is not because these species are absent in the local environment. Ongoing studies reveal up to ten Andrena spp. in the local pool with flight periods coinciding with sampling. More likely, Iowa produce farms do not attract Andrena species because they lack the floral important resources, particularly species in the Asteraceae in summer and autumn. Such deficiencies could be easily corrected. In addition, the number of species of Lasioglossum (Dialectus) at the farms was typically double that of prairies. These species may benefit because of their small size and favorable nesting conditions of exposed soil at farms.