REGIONAL AND LOCAL DRIVERS OF MASON BEE (GENUS OSMIA) DECLINE ACROSS THE EASTERN SEABOARD
Centrella, Mary Louise
Though there has been much focus on honey bee (Apis mellifera L.) decline, wild bees, which are essential to both ecosystem functioning and crop pollination, are also facing declines across Europe and North America. Potential drivers of theses declines include competition with non-native species, landscape simplification, increased pesticide risk, and reduced diet diversity. Here, I ask 1. how an introduced wild bee impacts a closely related, native congener at a regional scale, 2. how landscape simplification, diet diversity, and pesticide risk interact to impact wild bee populations in apple agroecosystems, and 3. how pesticide risk levels compare between wild bees and honey bees in the same apple orchards during bloom. To assess the impact of non-native Osmia cornifrons on the decline of native O. lignaria across the Eastern Seaboard, Ie used historical specimen records from 36 insect collections over 120 years. I found no evidence that O. cornifrons influenced O. lignaria decline; instead, their abundance (relative to other bees) has been decreasing since 1890, long before the 1977 introduction of O. cornifrons. Next, I was interested in exploring the drivers of Osmia performance in agroecosystems. Due to limited availability of O. lignaria, I assessed the response of nesting female O. cornifrons to landscape simplification, pesticide risk, and floral diet diversity in 17 NY apple orchards in 2015. In simplified landscapes, O. cornifrons produced fewer female offspring that weighed less, via reduced diet diversity and increased fungicide risk levels from Rosaceae (likely apple) pollen. Reductions in female offspring number and weight could lead to O. cornifrons population decline over time, as smaller-bodied bees produce fewer offspring and have shorter life-spans, suggesting that further studies of wild bee declines should focus on landscape simplification, pesticide risk, and floral diet diversity as potential drivers. To assess whether the historic use of honey bees as models for wild bee decline is adequate, I directly compared one driver of bee decline, pesticide risk levels, in O. cornifrons and A. mellifera pollen in 14 apple orchards during bloom in 2015. For O. cornifrons, increasing apple land cover resulted in increased pesticide risk levels in their pollen provisions, via increased Malus (crop) pollen collected. However, these relationships were not significant for honey bees, suggesting that their use as a model for all bee species may lead to inaccurate assessments of pesticide risk to some wild bee populations in agroecosystems. My results show that Osmia lignaria decline is not necessarily exacerbated by Osmia cornifrons at the regional scale. At the local scale, I show that landscape simplification, increased pesticide risk, and reduced diet diversity could potentially lead to O. cornifrons population decline, via reduced offspring number and size. Finally, I show that it is essential to continue studying the drivers of wild bee decline, as honey bees do not provide an adequate model with which to assess health of all bee species. By continuing to research the underlying causes of wild bee decline at both the regional and local scales, we can better preserve these important pollinators.
bee health; introduced species; landscape simplification; mason bees; pesticide risk; pollen diet; Ecology; Entomology
Danforth, Bryan N.
Mcart, Scott Harold; Scott, Jeffrey Graham; Poveda, Katja Andrea
Doctor of Philosophy
dissertation or thesis