Honey bees are key agricultural pollinators, adding billions of dollars in increased crop yields (1), yet American beekeepers regularly lose over half their colonies annually (2). There are myriad biotic and abiotic stressors that contribute to these high annual losses, including exposure to pesticides, transmission of viruses and other pathogens, and poor nutrition (3–7). What is more, multiple stressors can occur simultaneously and have additive or synergistic effects on honey bee health and productivity (4, 8). In agricultural landscapes, honey bees are exposed to many classes of pesticides, some of which can exert direct effects on mortality, and others which can elicit sublethal effects that impair important physiological pathways (reviewed in (9)). Perhaps the best-studied example is the neonicotinoids. Since their introduction 30 years ago, neonicotinoid insecticides have received greater scrutiny for their sublethal effects on pollinators than other classes of insecticides, likely due to their rapid and widespread adoption by growers, and their ability to be taken up by plants systemically. Neonicotinoids can persist in plant tissues and the surrounding environment (10, 11), exposing off-target organisms to chronic sublethal doses. In an effort to protect pollinators, several neonicotinoids and other pesticides have been banned or restricted in some regions or countries (12–14).

There has been a push in recent years to develop new chemistries that are safer to honey bees and other pollinators. One such pesticide is flupyradifurone (the active ingredient in Sivanto^®), a novel type of butenolide insecticide publicized as being safe for bees (15). Flupyradifurone is a systemic insecticide that, like neonicotinoids, targets nicotinic acetylcholine receptors (nAChRs) in the nervous system (16). Neonicotinoid insecticides have been shown to produce many sublethal effects on honey bees, including impairment of immune response. In addition to other pathogens, honey bees are beset by dozens of viruses, which can be vectored by parasitic Varroa destructor mites or transmitted horizontally and vertically between nestmates (17, 18). The majority of surveyed honey bee colonies harbor viral infections at subclinical levels (19, 20), but these infections can become clinical if stressors impede the bees’ abilities to suppress viral replication (21, 22). Honey bees use several innate immune system pathways to fight viral infection, but pesticides can disrupt these pathways (4). For example, studies show that exposure to the neonicotinoid clothianidin can suppress expression of a key transcription factor used in antiviral immunity and lead to higher replication of Deformed wing virus (DWV) in honey bees (23). Neonicotinoid exposure can also reduce the phagocytotic function of hemocytes (insect immune cells) (24, 25), thereby compromising an insect’s ability to eliminate virus-infected cells (26). It is important to note that all of these effects were only observed in studies occurring after these insecticides were registered and in use in many countries.

Studies from regulatory bodies and industry scientists show that flupyradifurone has low acute toxicity to honey bees, with minimal impact on mortality and productivity when used under recommended guidelines (27–29). Given flupyradifurone’s recent introduction to the US market, however, there are limited data available on more subtle effects to honey bee biology, and concerns have been raised regarding its safety for bees. High doses of flupyradifurone can increase oxidative stress (30) and affect honey bee taste, cognition, and motor skills (31, 32), while field-realistic doses can impair behavior and affect survival and immunity (33–35). Flupyradifurone exposure has been shown to increase expression of several honey bee immune genes (36), and lead to higher pathogen loads of the microsporidian Nosema ceranae in some colonies (37). Thus, as with neonicotinoids, it is imperative to consider not only acute lethal effects of flupyradifurone but also sublethal effects that may be more difficult to track. One clear current gap is our understanding of how flupyradifurone affects response to virus infection.

Proper evaluation of pesticide safety also requires consideration of the season in which the trials take place (38–40). Outside of winter, individual honey bee workers usually live for less than two months, but the colony as a whole is actively foraging from early spring to fall (or varying periods, depending on local environments). Over the course of the year, needs of the colony shift, from high population growth and colony proliferation in the spring, to the cessation of foraging and preparation for overwintering in the fall. This is accompanied by corresponding changes in worker physiology (41), meaning bees sampled in the spring differ from those sampled in the fall. Moreover, there is a seasonal shift in flowering species available for forage (42, 43). The nectar and pollen consumed by workers and fed to developing larvae contain plant phytochemicals that can stimulate detoxification pathways, which can affect how tolerant bees are to pesticides. For example, phytochemical consumption can increase bees’ resistance to some pesticides but not others (44, 45). Moreover, some phytochemicals can stimulate immune pathways and make bees more resistant to viral infection (46, 47). Flupyradifurone has already been shown to have seasonally-dependent toxicity to honey bees, where bees in southern California are more sensitive to flupyradifurone in the summer than in the early spring (33).

Here, we examined how exposure to flupyradifurone or the neonicotinoid clothianidin affected response to virus infection in honey bee workers. We hypothesized that, like in other studies of neonicotinoids and DWV (23), sublethal exposure to dietary insecticides reduces the ability of honey bees to respond to a virus infection. Further, we also hypothesized that responses to our treatments would differ across seasons. To test these hypotheses, we used Israeli acute paralysis virus (IAPV) as a model virus (48). IAPV can cause a debilitating disease that can be lethal to all honey bee castes and development stages (49–52), and has been implicated in colony failure (53). Further, it can be used for repeatable assays on survivorship that have been shown to be affected by different dietary treatments (7, 47, 54). Thus, we used IAPV treatments to assess the interaction of virus infection with flupyradifurone and clothianidin, and measured treatment-dependent mortality in both the spring and the fall to look for any seasonal effects. In addition to tracking survivorship, we also measured IAPV titers to determine whether either of the pesticides affects the rate at which the virus replicates or is cleared by the host.

Evaluating the efficacy and safety of novel pest suppression chemistries requires careful consideration of the context in which the pesticides are administered. Flupyradifurone and the neonicotinoids that preceded it both act systemically and target acetylcholine receptors, but despite these similarities, honey bee workers respond differently to flupyradifurone and clothianidin. Our results indicate that there are substantial seasonal effects of pesticide-induced mortality. In the spring, bees exposed to flupyradifurone survived as well as control-treated bees, which is consistent with industry studies demonstrating low toxicity of this new chemistry (27). However, bees exposed to the same dose of flupyradifurone in the fall displayed significantly lower survival than control-treated bees. This corroborates studies from other groups which found seasonal effects of flupyradifurone toxicity on honey bees (33). While there are many biotic and abiotic factors that differ in the spring and fall, perhaps the most relevant is the availability of different flowering plants for forage. Pollen, the main source of dietary protein and lipids, has different nutrient profiles across different seasons, and this can affect worker development and ability to combat infections (57), as well as their overwintering success (58). Likewise, bees display seasonal variation in their sensitivity to pesticides (40, 59), and diet has been directly linked to pesticide tolerance (60). In fact, consuming pollen increases expression of detoxification enzymes and can increase bee survival when exposed to pesticides (61, 62). Our results here provide further evidence that honey bee sensitivity to some pesticides may be seasonally dependent and illustrate that restricting toxicity studies to a single season can create gaps in our knowledge. This may be particularly important for assessing the risk of pesticides used across different crops in different regions, where applications and thus bee exposure may differ. Because the effects on bees may be seasonally variable, a harmless exposure in one context may have larger ramifications in another. Further, because systemic pesticides can remain in plant tissue and soils for months (63, 64), plant materials foraged and stored by a colony (including nectar, pollen, and propolis) could result in low-level exposure throughout the year (65).

Another important finding was that clothianidin exposure appears to confer some benefit to bees in the springtime, at least at the dosage used here. Bees given clothianidin had ~7% higher survival than bees given the sucrose-only control, while bees given clothianidin+virus had ~26% higher survival than bees given the virus-only treatment, suggesting that clothianidin improves response to IAPV infection in bees during this time. Like other pesticides studied, clothianidin exposure shows a mix of both negative and neutral effects, and sometimes even positive outcomes for honey bees. On the one hand, clothianidin can impair worker memory (66), alter foraging and communication (67), diminish immune responses via lower hemocyte counts and decreased antimicrobial activity (68), and increase viral replication of DWV (23). On the other hand, clothianidin has shown to have no adverse effects on mortality, longevity, or overwintering success of honey bees (69). The higher survival of clothianidin-treated bees that we observed is likely due to hormesis, where consumption of small amounts of xenobiotics such as pesticides can activate detoxification and immunological pathways and increase bees’ resistance to chemical and biological stressors (47, 70–72). This corroborates other studies showing hormetic effects of clothianidin, including increased enzymatic activity of cytochrome P450s in honey bees (73), increased colony growth in bumble bees (74), and increased longevity in mason bees (75). This illustrates that sublethal doses of pesticides can elicit varied responses in off-target organisms. While much focus is aimed at understanding acute toxicity and deleterious effects on pollinators, it is important to highlight instances where hormetic effects of pesticide exposure may actually aid important physiological processes like detoxification and immunity. This provides a more complete picture of pesticide effects when weighing the costs and benefits of applying certain pest suppression strategies.

This study also shows that pesticides can have varied, often subtle, effects on honey bees. The significance of these subtle effects is hard to discern, but it illustrates that there are more factors to consider than just acute toxicity and LD50 measurements. Here, we observed that exposure to flupyradifurone may impair bees’ ability to clear a viral infection. When inoculated with IAPV, bees sampled at 24 hrs show high viral titers, but those sampled at 48 and 72 hrs post-inoculation have greatly reduced viral titers. However, when viral inoculation is paired with exposure to flupyradifurone, a greater proportion of bees sampled at 48 hours post inoculation maintain high viral titers. This is somewhat inconsistent with other recent studies examining flupyradifurone’s effect on 3 other honey bee viruses (76). In that study, flupyradifurone exposure did not affect titers of BQCV or DWV-A and -B at 7- and 14-days post inoculation. However, IAPV may have different host-pathogen interactions, as our data show bees not exposed to flupyradifurone are mostly free of infection within 2 days. Even though flupyradifurone appears to delay virus clearing, this did not lead to an increase in mortality among the relevant treatment groups. Viral titers were unaffected by clothianidin, as both clothianidin-treated and control-treated groups had similar proportions of individuals with high-level infections over the 3 days of sampling. This is somewhat contradictory to previous studies, in which a comparable dose of clothianidin led to increased replication of DWV (23). Thus, pesticide effects on antiviral immunity may be both pathogen-specific and pesticide-specific and could have ramifications on transmission dynamics. It should be noted there are challenges with measuring viral clearance. It is possible that lower titers observed in individuals sampled at days 2 and 3 are due to some degree of survivor bias, but given the prevalence of highly infected individuals sampled at random on day 1 this effect is likely minimal. An alternative strategy would be to re-measure the same individuals each day by extracting hemolymph samples, but the stress of repeatedly subjecting individuals to this type of treatment would likely affect their viral titers and overall health.

Taken together, our results highlight that pesticide effects on honey bees and other off-target organisms can be subtle and context-dependent. As the next generation of “pollinator-friendly” pesticides come to market, increased scientific scrutiny of these novel chemistries will inevitably reveal hitherto unknown sublethal effects. In the big picture, it will be important to assess how impactful these off-target effects are, and what alternative options are available for pest management. If approval of these novel chemistries is revoked or increased regulatory oversight prevents their registration altogether, growers may turn to older products, many of which are also demonstrated as unsafe for bees or are poorly studied. In the meantime, it is critical that we continue to investigate the broad array of sublethal effects caused by pesticides so that we can better inform policy and decision-making.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

GH: conceptualization, formal analysis, investigation, methodology, writing (original draft, review, editing). VP: investigation, methodology, formal analysis. AD: project administration, funding acquisition, conceptualization, methodology, writing (review and editing). All authors contributed to the articled and approved the submitted version.

This work was funded by the US Department of Agriculture grant 2019-67013-29300.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.