Impact of Insecticide on Pollinator Communities in a Forested System:
A Model System Using Eastern Hemlock, Tsuga canadensis,
Rosebay Rhododendron, Rhododendron maximum, and Imidacloprid

Mortality of eastern hemlock, Tsuga canadensis (L.) Carrière (Pinales: Pinaceae), has occurred at a high rate since the arrival of the invasive hemlock wooly adelgid (HWA), Adelges tsugae Annand (Hemiptera: Adelgidae). The systemic neonicotinoid pesticide imidacloprid is soil-applied to hemlocks for effective control of HWA. However, is this pesticide translocated by incidental non-target plants under hemlock trees and, if so, does it impact non-target insects, such as pollinators? One commonly encountered flowering understory associate is Rhododendron maximum L. (Ericales: Ericaceae). Research has demonstrated that imidacloprid is translocated to leaves, nectar and pollen of R. maximum. The goal of this research was to assess if quantifiable differences in pollinators are associated with imidacloprid-treated HWA-infested hemlocks in Great Smoky Mountains National Park (GRSM). Research objectives were to: 1) determine the influence of imidacloprid treatment of hemlocks on pollinators of non-target R. maximum, 2) assess the influence of imidacloprid treatment of hemlock on seed production and germination of seed collected from R. maximum growing near treated hemlocks, and 3) determine species incidence, species composition, and seasonality of pollinators of R. maximum in GRSM.Insect visitors to R. maximum flowers were observed and some specimens were collected. High variability in pollinator abundance and diversity was documented at the four sites and no significant difference was identified between pollinators in treated and untreated areas. Seedpods and seedfall collected from rhododendrons had high seed number variability, and no significant differences in seed characteristics were identified between treated and untreated sites. A moderate positive correlation between pollinator visitation rate and number of seeds collected in buckets was observed. Through 64.5 hours of cumulative observations of R. maximum in GRSM, a total of 711 insects were observed visiting flowers. Hymenoptera comprised 89% of visitors with Bombus spp. the most common taxa observed. Bombus spp. were more active in the evening than first morning period (8am – 10am). Halictidae were more active in the two midday periods than morning. More sites and better classification of the sites by assessing landscape factors that impact pollinator populations would help obtain more definitive results to discern possible pesticide induced responses.