PESTICIDE EXPOSURE EFFECTS ON BEES AND POLLINATION SERVICES AT THE LOCAL AND LANDSCAPE LEVEL

Abstract

Currently, about 11% of the Earth’s habitable land area is devoted to crops and 30% more is being used for grazing. These numbers are rapidly increasing in the tropics and subtropics where most of the ongoing deforestation is happening to allow agricultural activities. These changes in land use involve the loss of natural habitat for wildlife and the increased use of synthetic pesticides, which are two of the main drivers of insect decline. Bees are vitally important pollinators of wild and cultivated plants and while there is growing evidence that bee populations are decreasing, the area of pollinator-dependent crops is expanding. To maintain and potentially increase pollination services, it is important to understand how pesticide use and local and landscape land management impact bee populations. In this dissertation, at the local level, I ask how different prophylactic application methods of the insecticide thiamethoxam impact pest control and pesticide risk for bees in squash crops. At the landscape level, I explore whether natural habitat areas mitigate negative pesticide effects on tropical pollinator communities in Solanum quitoense fields. Also, how livestock-dominated landscapes mediate pesticide exposure to veterinary pesticides in the stingless bee Tetragonisca angustula. In squash crops, I found that in-furrow applications of thiamethoxam best prevented defoliation and resulted in the highest yield. However, it also produced the most frequent and highest concentrations of insecticide residues in flowers, reaching lethal levels of exposure for bees when calculated for the solitary specialist Eucera pruinosa, making this practice unsustainable. In S. quitoense crops, natural habitat areas surrounding fields can partially mitigate negative effects of pesticides on bee communities, property that can be lost if there are very high levels of pesticides. Finally, in landscapes with high proportions of pasture for cattle ranching, I found an increased exposure but also increased tolerance to avermectin pesticides in stingless bee colonies, with these residues likely originated from veterinary applications. My results show that pesticide exposure to bees is mediated by in-farm practices but also by how the land is used beyond farm borders. By integrating landscape characterization and chemical analysis is possible to better understand the risk of pesticides for bees and potential strategies for its mitigation.