Stream Salamander and Benthic Macroinvertebrate Community Responses to Imidalcoprid Exposure

Crayton, Sara

ABSTRACT: In this study, I investigated whether the usage of a common insecticide for forest conservation has non-target effects on stream organisms. Hemlock woolly adelgid (HWA; Adelges tsugae) is an invasive insect that has caused wide-spread loss of hemlock trees in the eastern United States. Imidacloprid is a neonicotinoid insecticide and is the most widely used treatment to mitigate hemlock mortality from HWA infestations, but application of imidacloprid can result in lethal and sublethal effects on non-target taxa. I assessed non-target effects of HWA treatments using imidacloprid on benthic macroinvertebrates and stream salamanders in centralWest Virginia. In Chapter 1, I provide a literature review of the impact of HWA on forest systems and known non-target effects of HWA treatments on benthic macroinvertebrates and amphibians. I also review studies that investigated bioaccumulation of imidacloprid in amphibians and other taxa. In Chapter 2, I assess whether stream salamanders or benthic macroinvertebrates bioaccumulate imidacloprid or imidacloprid metabolites through exposure to HWA treatments. I also evaluate relationships between imidacloprid bioaccumulation and exposure and sublethal effects in salamanders, including corticosterone hormone levels and body condition indices. Of 107 stream salamanders, 29 individuals had detectable levels of imidacloprid (n =14) and/or imidacloprid-olefin (n = 19) in their tissues. Of 15 stream invertebrate samples, 15 had detectable levels of imidacloprid and 13 had detectable levels of imidacloprid-urea in their tissues. Corticosterone levels increased with increasing numbers of imidacloprid applications and body condition decreased as water imidacloprid concentration increased. Our study demonstrates that stream salamanders and invertebrates can bioaccumulate imidacloprid from HWA treatments and imidacloprid exposure likely has sublethal effects on stream salamanders. In Chapter 3, I evaluate whether abundances of five stream salamander species are affected by imidacloprid presence. We surveyed stream salamanders 5–7 times at 48 sites, 27 of which were directly adjacent to HWA treatments. We used N-mixture models and a model selection approach to identify important predictors of abundance for Desmognathus fuscus, D. monticola, D. ochrophaeus, Eurycea spp., and Gyrinophilus porphyriticus. We identified influential detection and habitat variables for each species and then tested models containing imidacloprid predictors. For all species, there was support for models containing predictors of imidacloprid exposure. Number of imidacloprid treatments and presence of environmental imidacloprid had a negative relationship with Desmognathus spp. abundances, but we did not detect negative effects of imidacloprid exposure on Eurycea spp. or G. porphyriticus. In Chapter 4, I evaluate whether imidacloprid exposure affects benthic macroinvertebrate communities in the Monongahela National Forest and two units of the National Park Service in West Virginia. We sampled benthic macroinvertebrates from 47 sites, 26 of which were directly adjacent to HWA treatments. We calculated seven benthic macroinvertebrate community metrics and then compared those indices to four imidacloprid predictors with linear or beta regressions. There was support for effects of imidacloprid exposure on all community metrics, and the relationship between community metrics and imidacloprid differed by locality with GLIMPSS and WVSCI being lower in the Monongahela National Forest in sites with imidacloprid exposure. Additionally, we compared functional traits of benthic macroinvertebrate communities to metrics of imidacloprid exposure. Functional traits analysis showed no significant differences in community functional traits composition between sites with or without imidacloprid presence, and with or without treated trees. Variation in functional traits was higher in sites in MNF with environmental imidacloprid exposure...