Influence of the agricultural landscape surrounding Apis mellifera colonies on the presence of glyphosate and its metabolite ampa in honey

Abstract

During foraging flights, honey bees settled in agricultural ecosystems may encounterglyphosate residues on flowers of both cultivated and native plants growing in semi-naturalhabitats within the landscape. This product has been considered non-toxic to bees based onacute contact and oral toxicity tests (> 100 μg / per bee) performed mostly under lab conditions;however, several authors have described various deleterious effects of this herbicide onhoneybees behavior and survival. The configuration and heterogeneity of the landscape andthe agricultural management in which the bee colonies are located, have direct and indirectimplications on the presence and survival of bees. In this work, we were interested to analyzethe relationship between landscape configuration and the presence of pesticides in honey. Inparticular, we tested the relationship between the presence of glyphosate and AMPA residuesin honey and some metrics of habitat configuration. We expect a higher presence of glyphosateand AMPA as the proportion of crop lands in the landscape is increased and the connectivityof semi-natural habitats is decreased. A total of 30 honey samples belonging to the 2019 and2020 harvests were analyzed in various areas of the southeast of the province of Buenos Airesand a frequency of 50% of glyphosate positivity was found in the analyzed simples meanwhileAMPA was found in 30% of the samples using a method of analysis of residues of Glyphosateand its main metabolite of degradation to Aminophosphoric Acid (AMPA) in honey developedusing UPLC MSMS. We identified a total of 24 different sites to test the relationships betweenhabitat configuration and residues of pesticides in honey because six of the samples wereobtained in the same apiary (during the same of different season; see below). The range ofvalues found in the samples was between 2 ? 27,5 μg/kg for Glyphosate and 1,9 ? 18,1 μg/kgfor AMPA. The metrics for the near landscape (1000m ratio) surrounding the apiaries showednegative higher correlations than those for the larger one (2225 m ratio), particularly with AMPAresidues. Nevertheless, these correlations were weak and thus the amount of crop lands withindustrial agriculture is not enough to clear explain the values of glyphosate and AMPAresidues detected in the honey bee colonies. Based on the pesticide prevalence in theenvironment, as well as our findings of its presence in honey samples, we propose that detailedstudies at the apiary scale should be conducted in Argentina to determine if glyphosate is infact a contributing driver to the colony losses. Geospatial analysis like the performed in thisstudy can help honey producers to estimate risks to pesticide exposure in landscapes withintensive agriculture.