Providing suitable floral resources is a key element of bee conservation. Many pollinator friendly plant lists are available, but few have been tested for their effects on the attracted pollinator community composition. Moreover, several pollinator plant lists include non-native plants that are considered to be high quality food resources due to e.g. large quantities of nectar. It remains unknown how non-native plant resources affect native bee communities.
In an experimental field study on farmland of the University of Maryland, we compared native and non-native pollinator friendly seed mixes in their attractiveness to bees and their effect on bee visitation networks to answer the following research questions: Do non-native flowers attract more individuals and species? Is the level of network specialization lower for non-native than native pollinator friendly seed mixes?
Experimental plots were established at three different sites each comprising one native and one non-native plant plot. The plant mixes included each 20 species of wildflowers as well as two grasses and were seeded in April 2016. After the onset of flowering in June, we observed plant bee interactions twice a month until September 2016 and additionally captured bees with pan traps over the flowering season.
The number of bee individuals and species was higher for non-native plant mixes, which can partly be explained by faster growing and earlier flowering plant species. Ten non-native and eight native plant species successfully flowered. The start of flowering of native plants was delayed by 1.5 months, decreasing the number of sampling events on native plant plots. In total, 214 bee individuals of 27 species were sampled from non-native plants and 149 individuals of 18 species from native plants, representing 46.5 and 34.9 % of the species caught in pant traps, respectively. Thirteen bee species were unique to non-native plots, while three bee species occurred only on native plots.
Non-native plant visitor networks showed more variable, but on average lower levels of specialization than native plant networks (H2’ non-natives: 0.22 ± 0.25; H2’ natives: 0.43 ± 0.10). Accordingly, niche overlap of bee species was higher in networks of non-native plants (0.79 ± 0.24) compared to native plant mixes (0.44 ± 0.20).
Observations will be repeated in 2017, but these first results suggest that non-native plant mixes provide suitable resources for a broader spectrum of wild bees (including specialists). However, they also indicate a possible shift towards more generalized networks.