This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Agrochemical formulations are composed of two broad groups of chemicals: active ingredients, which confer pest control action, and ‘inert’ ingredients, which facilitate the action of the active ingredient. Most research into the effects of agrochemicals focusses on the effects of active ingredients. This reflects the assumption that ‘inert’ ingredients are non-toxic. A review of relevant research shows that for bees, this assumption is without empirical foundation. After conducting a systematic literature search, we found just 19 studies that tested the effects of ‘inert’ ingredients on bee health. In these studies, ‘inert’ ingredients were found to cause mortality in bees through multiple exposure routes, act synergistically with other stressors and cause colony level effects. This lack of research is compounded by a lack of diversity in study organism used. We argue that ‘inert’ ingredients have distinct, and poorly understood, ecological persistency profiles and toxicities, making research into their individual effects necessary. We highlight the lack of mitigation in place to protect bees from ‘inert’ ingredients and argue that research efforts should be redistributed to address the knowledge gap identified here. If so-called ‘inert’ ingredients are, in fact, detrimental to bee health, their potential role in widespread bee declines needs urgent assessment.

Pollinators, particularly wild bees, are suffering declines across the globe, and pesticides are thought to be drivers of these declines. Research into, and regulation of pesticides has focused on the active ingredients, and their impact on bee health. In contrast, the additional components in pesticide formulations have been overlooked as potential threats. By testing an acute oral dose of the fungicide product Amistar, and equivalent doses of each individual co-formulant, we were able to measure the toxicity of the formulation and identify the ingredient responsible. We found that a co-formulant, alcohol ethoxylates, caused a range of damage to bumble bee health. Exposure to alcohol ethoxylates caused 30% mortality and a range of sublethal effects. Alcohol ethoxylates treated bees consumed half as much sucrose as negative control bees over the course of the experiment and lost weight. Alcohol ethoxylates treated bees had significant melanisation of their midguts, evidence of gut damage. We suggest that this gut damage explains the reduction in appetite, weight loss and mortality, with bees dying from energy depletion. Our results demonstrate that sublethal impacts of pesticide formulations need to be considered during regulatory consideration, and that co-formulants can be more toxic than active ingredients.

The impact of human-mediated environmental change on the evolutionary trajectories of wild organisms is poorly understood. In particular, species’ capacities to adapt rapidly (in hundreds of generations or less), reproducibly and predictably to extreme environmental change is unclear. Silene uniflora is predominantly a coastal species, but it has also colonised isolated, disused mines with phytotoxic, zinc-contaminated soils. To test whether rapid, parallel adaptation to anthropogenic pollution has taken place, we used reduced representation sequencing (ddRAD) to reconstruct the evolutionary history of geographically proximate mine and coastal population pairs and found largely independent colonisation of mines from different coastal sites. Furthermore, our results show that parallel evolution of zinc tolerance has occurred without gene flow spreading adaptive alleles between mine populations. In genomic regions where signatures of selection were detected across multiple mine-coast pairs, we identified genes with functions linked to physiological differences between the putative ecotypes, although genetic differentiation at specific loci is only partially shared between mine populations. Our results are consistent with a complex, polygenic genetic architecture underpinning rapid adaptation. This shows that even under a scenario of strong selection and rapid adaptation, evolutionary responses to human activities (and other environmental challenges) may be idiosyncratic at the genetic level and, therefore, difficult to predict from genomic data.