Pollinating insects are important for food security and ecosystem function, providing over $200 billion annually in pollination services. Recent honeybee declines have underlined the importance of native pollinators and their ability to provide effective pollination services. However, native bees are also affected by multiple pressures including pathogens, pesticide use and poor nutrition. RNA viruses, once considered to be specific to European honey bees, are among the suspected threats to native bumble bees. However, very little is known about how these viruses are transmitted, their effects on bumble bees, and whether other stressors promote conditions for virus replication. Filling these knowledge gaps is critical for making management recommendations that will lessen the risk of virus infection to important crop pollinators. We a series of controlled laboratory experiments to first examine the effect of deformed wing virus (DWV) on bumble bees and whether exposure of imidacloprid, a neonicotinoid pesticide, influences viral infection, food intake, and bumble bee mortality. Food intake was significantly lower for bumble bees exposed to 10 and 20 parts per billion (ppb) of imidacloprid but virus loads and mortality were not affected by imidacloprid exposure. Results indicate that exposure to imidacloprid has non-lethal adverse affects for bumble bees. In a second set of experiments, we examined the role of plants in virus transmission between honey bees and bumble bees. We demonstrated that honey bees leave behind viruses while foraging on flowers and that deposition may vary across plant species. Bumble bees that foraged on infected flowers did not develop virus infections. In light of these results and past field observations, we do not discredit the possibility of virus transmission between bee species through flowers and suggest further experimentation to examine the full transmission route.
Bumble bees are important pollinators of both food crops and wild plants. In 2015, Vermont awarded legal protection to three native species of bumble bees; listing them as either threatened (Yellow-banded Bumble Bee, Ashton’s Cuckoo Bumble Bee) or endangered (Rusty-patched Bumble Bee) (Vermont Fish and Wildlife 2015). Many crops vital to Vermont’s food system rely on these pollinators; among them, blueberries, raspberries, apples, squash and tomatoes. A decline in bumble bee populations may have serious environmental and economic consequences.
RNA viruses, once thought to be specific to honey bees, have been detected in wild bumble bees (Singh et al. 2010; Fürst et al. 2014; McMahon et al. 2015). While relatively well-studied in honey bees, few studies have examined the effect or transmission of these pathogens in bumble bees. In honey bees, viruses may persist in the host as subclinical infections but are capable of replicating quickly under certain conditions. In honey bees, pesticide exposure may result in higher virus titers and symptomatic bees (Di Prisco et al. 2013). However, the conditions necessary for virus replication in bumble bees is completely unknown. Additional studies are needed to understand how the effects of RNA viruses on bumble bees and whether pesticide exposure increases virus loads.
In honey bees, RNA viruses may be transmitted horizontally between nest mates and through Varroa mite (Varroa destructor) vectors (Chen et al. 2006; Di Prisco et al. 2011). Since mites are not known hosts of bumble bees, it is unclear how RNA viruses are transmitted to bumble bees. One hypothesis is that viruses are spilling over from managed honey bees to wild bumble bees through the shared use of flowers (Singh et al. 2010; Mcart et al. 2014). In previous surveys throughout Vermont, we have found that the prevalence of RNA viruses is higher both in bumble bees and on floral surfaces collected from managed honey bee apiaries, indicating that honey bees are significant contributors of RNA viruses to both bumble bees and the landscape (Alger et al. unpublished). While these results provide evidence for transmission through flowers, a controlled study is needed to directly test this transmission route. If spillover from managed bees is occurring, the spread of disease may be reduced by better controlling disease loads in managed honey bee colonies.
Examining the effects of viruses and pesticides exposure as well as transmission routes is critical to making recommendation to safeguard these important crop pollinators.
This research has two main objectives: 1) To examine the effect of deformed wing virus (DWV) and how pesticide exposure influences bumble bee mortality, food intake, and virus infection in bumble bees and 2) To examine the role of flowering plant species in viral transmission between bee species.
To accomplish objective 1, purified deformed wing virus isolate was prepared at the University of Maryland by collaborator Humberto Bonchristiani. We obtained five commercial bumble bee colonies and tested each for three RNA viruses (black queen cell virus, Israeli acute paralysis virus, and deformed wing virus) upon arrival. Of the five colonies, all were infected with black queen cell virus (BQCV) and four were infected with deformed wing virus (DWV). These results presented many challenges, as we needed virus-free colonies for upcoming experiments. Despite these challenges, we conducted two experiments to determine a. the effectiveness of the virus inoculum and b. effect of imidacloprid on bumble bee survivorship and DWV virus load that was already present in the colonies.
To test the effect of DWV inoculum on bumble bees, we conducted an experiment in which we transferred one hundred bumble bee (Bombus impatiens) workers to individual containers and assigned each to one of 5 treatments: 4 different concentrations of DWV and a control. After a 5-hour period without food, we fed each bee 10 ul of an inoculum containing DWV and 50% sucrose. The control bees only received 10 ul of 50% sucrose. We provided all bees with pollen and 30% sucrose ad libitum for 14 days and recored mortality for all groups. After 14 days, we transferred all surviving bees to -80°C. Finally, we extracted total RNA following Qiagen RNeasy mini kit protocols and used real time quantitative polymerase chain reaction (RT-qPCR) to analyze DWV loads in two of the groups (the control and highest DWV concentration) to examine whether the inoculum resulted in higher virus load than the control group.
In a second experiment, we tested the effect of different concentrations of imidacloprid (a commonly used neonicotinoid pesticide) on bumble bee survivorship and sucrose intake. Since the commercial bumble bee colonies arrived already infected, we also tested whether exposure of imidacloprid influenced the viral loads already present, rather than administering the DWV inoculum. We assigned twenty bees to each of 4 treatments and a control. To each treatment group, we provided pollen and 30% sucrose ad libitum inoculated with different concentrations of imidacloprid: 0.1, 1, 10, and 20 parts per billion (ppb) for 8 days. The control received 30% sucrose only. We measured sucrose consumption and mortality for five days after which, all surviving bees were transferred to -80°C. Lastly, we extracted RNA and used RT-qPCR to test bees to see if virus levels were affected by the pesticide exposure.
To address objective 2: the role of flowers in acting as bridges for viral deposition and transmission, we exposed flowers to infected honey bees and then allowed micro-colonies of bumble bees to forage on these same flowers in a screened enclosure (figures 1 and 2). We then tested flowers and bees for the presence of virus after each experiment. We examined if the transmission between bees was influenced by: 1) plant species, 2) plant diversity, 3) multiple exposures to infected plant and, 4) if direct contact or co-mingling is necessary for viral transmission by allowing bumble bees and honey bees to forage at the same time on the same plants.
Setup and Pre-screening
We grew plants of red clover (Trifolium pratense), white clover (Trifolium repens), and birdsfoot trefoil (Lotus corniculatus) from seed in the greenhouse in 8 in. diameter, 6.5 in. deep plastic pots. We planted approximate 100 seeds per pot and then allowed natural thinning to occur. These species grow to flowering quickly, are used as cover crops in agricultural systems, and are highly attractive to bees. In addition, we had detected virus on these species collected n the field. To verify that our plants were free of virus prior to the experiments, we haphazardly collected composite samples of each plant species and tested them for DWV and BQCV using RT-qPCR protocols.
We collected virus-positive honey bees from colonies known to be infected by testing with RT-qPCR. We obtained 7 bumble bee (Bombus impatiens) colonies from a commercial supplier and tested them for DWV and BQCV. All were negative at the start of the experiment. We created microcolonies of bumble bees of 12 adult bees, provided them with 30% sucrose solution ad libitum, and allowed them to acclimate for 5 days before using them in experiments.
We conducted all experiments in three 10 x 10 x 10 foot screened enclosures with tarp bottoms. Each screened enclosure was assigned a single treatment: honey bee tent, bumble bee control tent, or bumble bee treatment. To conduct foraging experiments with bumble bees, we released bumble bees into smaller hoop houses in each of the two bumble bee control and treatment tents. Hoop houses (3 x 4 x 2.5 ft) were constructed of fabric stretched and stapled over two pieces of arching PVC tubes secured to a wooden frame.
At the start of each trial, we placed plants within the screened enclosure with two honey bee colonies and allowed bees to forage for 9 hours. We then transferred plants to a holding tent and allowed nectar to be replenished for 15 hours. After this, we transferred the plants previously visited by honey bees to a bumble bee tent and evenly distributed them among the three hoop houses.
For the control bumble bee tent, we transferred clean flowering plants from the greenhouse directly into each of three hoop houses within the tent. For the bumble bee treatment tent, we transferred plants that had been previously visited by honey bees. We allowed each micro colony of bumble bees to forage within their hoop house enclosures for 6 hours and then collected all inflorescences and bumble bees. Inflorescences were then stored at -80°C for later analysis. We kept the bumble bee micro colonies in new containers and fed them 30% sucrose ad libitum for one week as an incubation period for the virus infection to take hold. Bees were fed only sugar water and not pollen to reduce the chance of detecting inactive virus particles from consumed pollen still in the bees’ guts. After one week, we collected all bees and stored them in -80°C until RNA extraction and virus assays.
We conducted the experiment three times using white clover, red clover, and birdsfoot trefoil and used these replicates to test if plant species affects virus transmission. To test if plant diversity affects virus transmission, we allowed bees to forage on floral arrays containing all 3 plant species. To test if multiple exposures to infected plants is necessary for virus transmission, we used only white clover and repeatedly exposed bumble bees by allowing them to forage on infected white clover over the course of three days.
We also tested if co-mingling on flowers is necessary for transmission of DWV between bee species. To do so, we used bumble bee colonies containing 75-100 workers and white clover arrays. We placed two honey bee colonies, a single bumble bee colony, and 6 pots of white clover plants in a 10 x 10 x 10 ft tent. We allowed bees to forage and co-mingle on plants for 7 hours. After this time, we collected all bees and returned them to their colonies. We repeated this procedure over three days using the same, infected honey bee colonies but different bumble bee colonies. We fed the bumble bee colonies pollen and 30% sucrose ad libitum for three weeks to allow time for DWV to spread throughout the colony. We then starved bees of pollen, in microcolonies of 12 bees each, and collected these bees after one week, After three weeks, we created micro colonies consisting of 12 bees each which we then pollen-starved for one week to clear their guts of virus particles on pollen. We collected these bees and stored them at -80°C. We extracted total RNA from honey bees, bumble bees, and plants, using Qiagen RNeasy mini kit protocols and RT- qPCR to test all tissues for RNA viruses.
In our pilot experiment, we tested the effect of DWV inoculum on bumble bees both to ensure the efficacy of the inoculum and examine mortality. Mortality rates did not differ between groups. Using RT-qPCR, bees fed the DWV inoculum had significantly higher DWV levels than the control group, indicating that the DWV inoculum was effective.
In our second experiment, we tested the effect of imidacloprid on bumble bee mortality, sucrose consumption, and DWV virus load. Mortality and DWV virus load did not differ between treatments. However, bees in the 10 ppb and 20 ppb group consumed significantly less sucrose (figure 3), indicating that high concentrations of imidacloprid negatively afffects bumble bee food consumption rates and/or foraging behavior. Field realistic doses of imidacloprid from a variety of crops and studies are 0.7-10 ug/L (Cresswell, 2011), indicating that the effects we observed have important implications for wild bees foraging on treated crops. Due to the food aversion affect of imidacloprid, bees in this group did not receive the full treatment and thus effects of the the higher imidacloprid doses on virus load remain unknown. While we did not see differences in virus load between the other treatment groups, it is possible that results were masked since bumble bees arrived already infected with the virus. Obtaining virus-free colonies is crucial for future experiments.
In this experiment, we tested whether honey bee could leave viruses behind on flowers while foraging and whether bumble bees could become infected after visiting these flowers. We detected both BQCV and DWV viruses on plants visited by honey bees, thus confirming that honey bees can deposit viruses on plants while foraging (figure 4). The two viruses were not equally detected across the three plant species (figure 5). This suggests that different viruses may be shed in different ways and plant morphology may mediate this process. Bumble bees did not transfer viruses to flowers, as all control flowers were negative for RNA viruses. Under our experimental conditions, bumble bees did not develop an infection after direct contact with honey bees through co-mingling or indirect contact through shared flowers. These results could indicate that transmission of viruses between bee species is a less likely occurrence or that shared flowers are not the primary transmission route. While we did not demonstrate virus transmission to bumble bees in our experiment, we believe it is possible since honey bees leave viruses behind while foraging and bumble bees can be orally innoculated with RNA viruses (Fürst et al. 2014). Virus transmission is dependent on many factors including the probability a bumble bee will pick up a virus particle, the bumble bee’s immune system, and virus virulence. Since our experiment was successful in demonstrating virus deposition to flowers, future controlled experiments should focus on the second half of the transmission cycle and examine whether bumble bees can ‘pick up’ virus particles from flowers that have been inoculated by hand.
In our first pilot experiment, the bumble bee colonies arrived infected with RNA viruses. Of the five colonies, all were infected with BQCV and four were infected with DWV. These results presented challenges for our experiment but also provide evidence that commercial bumble bee colonies may be contributing to RNA virus spread. Captive bumble bee colonies may become infected through the pollen feed that is collected from honey bees colonies. In our previous work, we have tested multiple sources of honey bee collected pollen and detected RNA viruses on small amount of pollen grains. Fortunately, gamma irradiation can inactivate RNA virus particles on pollen grains (Meeus et al. 2014). We suggest that all commercially available pollen feed for both honey bees and bumble bees undergo gamma irradiation treatment to reduce the risk of disease spread to bees.
Our study shows that sucrose intake is reduced when bumble bees are exposed to imidacloprid levels at and above field realistic doses (10 ppb or 10ug/L) (Cresswell 2011). Presently, neonicotinoid pesticide use and regulations are under scrutiny by legislative committees. This is especially true for Vermont where neonicotinoid sales and use were the topic of three legislative bills over the past seven years (H.34, H.236, H.688). Our results provide much needed empirical evidence that imidaclorprid, a commonly used neonicotinoid, has adverse effects for bumble bees. Future studies should investigate its effects on full bumble bee colonies, with a focus on foraging behavior.
Our study also confirms a widely accepted, but largely untested hypothesis: honey bees deposit viruses on flowers while foraging. While we did not experimentally demonstrate that bumble bees can become infected after visiting infected flowers, our results indicate that honey bees are significant contributors to RNA viruses to the landscape. In a previous field survey we conducted, we also found that bumble bees were more likely to be infected when collected from honey bee apiaries (Alger et al. unpublished). Given these results, it is crucial that beekeepers take precaution to ensure healthy honey bee colonies to reduce the risk of spillover to wild bees. For example, since high mite infestations are linked to high virus loads, beekeepers should monitor and treat mite infestations. We suggest improving education opportunities for beekeepers. Currently, the Vermont Beekeepers Association is working to improve their mentoring program to help provide education to new beekeepers. In addition, we suggest improving and growing state apiary inspection programs. While these programs help to monitor colony health and provide beekeeper education, they are often understaffed and underfunded. We are currently working with the Vermont Apiary Inspection Program to improve the apiary database to better prioritize inspections.
Chen, Y., Pettis, J.S., Collins, A. & Feldlaufer, M.F. (2006). Prevalence and transmission of honeybee viruses. Appl. Environ. Microbiol., 72, 606–611.
Cresswell, J.E. (2011). A meta-analysis of experiments testing the effects of a neonicotinoid insecticide (imidacloprid) on honey bees. Ecotoxicology, 20, 149–157.
Fürst, M. a, McMahon, D.P., Osborne, J.L., Paxton, R.J. & Brown, M.J.F. (2014). Disease associations between honeybees and bumblebees as a threat to wild pollinators. Nature, 506, 364–6.
Mcart, S.H., Koch, H., Irwin, R.E. & Adler, L.S. (2014). Arranging the bouquet of disease: Floral traits and the transmission of plant and animal pathogens. Ecol. Lett., 17, 624–636.
McMahon, D.P., Fürst, M. a., Caspar, J., Theodorou, P., Brown, M.J.F. & Paxton, R.J. (2015). A sting in the spit: widespread cross-infection of multiple RNA viruses across wild and managed bees. J. Anim. Ecol., 84, 615–624.
Meeus, I., Mosallanejad, H., Niu, J., de Graaf, D.C., Wäckers, F. & Smagghe, G. (2014). Gamma irradiation of pollen and eradication of Israeli acute paralysis virus. J. Invertebr. Pathol., 121, 10–13.
Di Prisco, G., Cavaliere, V., Annoscia, D., Varricchio, P., Caprio, E. & Nazzi, F. (2013). Neonicotinoid clothianidin adversely affects insect immunity and promotes replication of a viral pathogen in honey bees, 110, 18466–18471.
Di Prisco, G., Pennacchio, F., Caprio, E., Boncristiani, H.F., Evans, J.D. & Chen, Y. (2011). Varroa destructor is an effective vector of Israeli acute paralysis virus in the honeybee, Apis mellifera. J. Gen. Virol., 92, 151–155.
Singh, R., Levitt, A.L., Rajotte, E.G., Holmes, E.C., Ostiguy, N., Vanengelsdorp, D., et al. (2010). RNA viruses in hymenopteran pollinators: Evidence of inter-taxa virus transmission via pollen and potential impact on non-Apis hymenopteran species. PLoS One, 5, e14357.
Vermont Fish and Wildlife. (2015). Endangered and Threatened Animals of Vermont.
Education & Outreach Activities and Participation Summary
We aim to bring awareness that the so-called ‘honey bee viruses’ can infect wild bees. Therefore, we work to inform beekeepers that sound beekeeping practices that reduce pathogens in managed honey bees can benefit wild bees living near apiaries by reducing the risk of pathogen spillover. I have presented my research to a number of beekeeping and farming groups including the Bennington beekeeping club, the Vermont Beekeeping Association, the Addison County Beekeepers Association, the Southern Adirondack Beekeepers Association. I worked with the Bennington beekeeping club to hold a ‘bee pathogen workshop’ for small-scale beekeepers to gain hands-on experience identifying and monitoring honey bee pathogens. I have also presented my research at several other venues including the Bumble Bee Working Group Meeting in at the University of Sussex in the UK, the International Pollinator Conference at Penn State, the Garden Club of America in Lenox Massachusetts, and the Vermont Grazing and Livestock Conference. I provided testimony at the VT Statehouse hearing for two separate bills: H.236: Bill proposes to ban the use, sale, or application of neonicotinoid pesticides; H.539: Bill proposes to establish a Pollinator Protection Committee (PPC). After the PPC was established, I presented my research at one of their meetings at the state house. An overview of my research was published in March 2017 in Bee Culture. These studies will be submitted to peer reviewed journals for publication in 2018. I am also currently working with the Vermont Pesticide Certification and Training coordinator to include an article about pollinators in the Spring newsletter for pesticide applicators. Lastly, I am working with the Vermont Apiary Inspection Program to improve honey bee monitoring efforts and beekeeper education.
Our project joins a body of research that provides empirical evidence for the adverse affects of neonicotinoids on bees. Since bees are crucial for the pollination of flowering plants, it is critical that direct and indirect effects of agricultural chemicals on pollinators be examined to ensure agricultural sustainability. Our research also highlights the need to improve honey bee disease management to reduce the risk of disease spillover to wild bees through flowering plants. Helping beekeepers to reduce disease loads in honey bee colonies is important to reduce honey bee losses, reduce disease impacts to wild pollinations, and ultimately improve pollination services to crops.
Through this research, it became clear that honey bee monitoring programs need to be improved to reduce the risk of disease spillover to wild bees. As a result, we have taken on several projects aimed at examining and reducing the role of honey bees in disease spread. In a crowd funded project, we examined the role of migratory beekeeping in the spread of disease. The results from the experiment were submitted earlier this year to a peer reviewed journal for publication. To help improve the Vermont Apiary Inspection Program, we worked with the state inspector to develop a beekeeper census to begin collecting data on beekeeper management practices and hive losses in Vermont. Results will help to prioritize inspections this coming season as well as highlight educational opportunities. Lastly, we a grant from the North American Pollinator Protection Campaign (NAPPC) for $9,980 to further study the bee-flower transmission route