INTRODUCTION
Recent declines in native and managed bee populations threaten the stability of pollination services that are vital for maintaining natural and agricultural ecosystems (Beismeijer et al. 2006, Potts et al. 2010). Several factors contribute to these declines, including the spread of multi-host pathogens, habitat loss, and climate change (Ricketts et al. 2008, Burkle et al. 2013, Furst et al. 2014). Losses in pollinator community biodiversity and abundance lead to changes in flower visitation patterns (Beismeijer et al. 2006, Albrecht et al. 2012, Burkle et al. 2013), as well as changes in the risk of infectious disease within reduced pollinator communities (Figueroa et al. 2020, Graystock et al. 2020, Fearon and Tibbetts 2021). Yet, it remains unclear how differences in floral visitation behaviors within pollinator communities affected by these declines may in turn affect the spread of pathogens.
Many pollinator pathogens and parasites (hereafter, ‘parasites’) are transmitted within and among species by visitation to flowers that were previously visited by infected bees (Durrer and Schmid-Hempel 1994, Graystock et al. 2015, Müller et al. 2019, Purkiss and Lach 2019). The likelihood of parasite deposition and subsequent transmission on flowers depends on multiple factors, including flower traits, flower morphology, pollinator behavior, and the environment (Durrer and Schmid-Hempel 1994, Alger et al. 2019, Figueroa et al. 2019, Russell et al. 2019). Depending on the parasite, different plant components, including the floral tissue, pollen, and nectar, are implicated in transmission among pollinators (reviewed by McArt et al. 2014). In particular, differences in the rates of parasite deposition and acquisition of microorganisms on various flower parts may depend on how bees interact with the flowers during foraging visits. For example, bees foraging for pollen had greater rates of microbe deposition and acquisition on flowers than did bees foraging for nectar (Russell et al. 2019). However, pollinator visitation behaviors have been shown to have a complex relationship with the prevalence of bee parasites on flowers. In a study on pollinator viruses, flowers receiving longer visits were more likely to host viruses, but those with high visitation rates were less likely to host viruses (Alger et al. 2019). In a different study, Crithidia bombii survived longer when deposited inside the corolla rather than on the bract, but infection occurring from an encounter with the bract resulted in more intense infection (Figueroa et al. 2019). Therefore, the ways in which infected bees interact with specific flower features and the duration and frequency of their visits will alter the likelihood of parasite deposition on floral surfaces and influence the probability of infection for later visitors. However, most studies on this topic have been conducted in the laboratory and have not fully considered the potential for parasite transmission via shared floral resources in natural settings.
Agricultural fields and the surrounding hedgerows may represent potential ‘hot spots’ for parasite transmission within and among bee species on shared floral resources. Managed honeybees (Apis mellifera) are frequently brought to agricultural fields to provide pollination services, where they have ample opportunity to interact with wild pollinators that are also attracted to plentiful crop flowers or nearby hedgerows with wildflowers (Goulson and Hughes 2015). The worldwide dispersal of A. mellifera (hereafter honeybees) and its many parasites has consequently led to spillover (i.e., parasite transmission from reservoir populations to sympatric wildlife) to many naïve wild pollinators (Daszak et al. 2000, Keesing et al. 2006, Goulson and Hughes 2015, Purkiss and Lach 2019). Since honeybee colonies tend to send generalist foragers to a few flower patches at a time (Visscher and Seeley 1982), it is possible that an infected colony may create localized floral hot-spots where wild bees may acquire parasites. Increasingly, parasites previously thought to only infect honeybees are found in diverse populations of wild pollinators and seem to be contributing to their decline (Furst et al. 2014, Arbulo et al. 2015, Goulson and Hughes 2015, Porrini et al. 2017, Müller et al. 2019, Purkiss and Lach 2019).
One parasite of particular concern is the widely-dispersed microsporidian parasite Vairimorpha (= Nosema )ceranae (Tokarev et al. 2020), which has been rapidly infecting honeybees and spilling over into wild bee populations over the past three decades (Paxton et al. 2007, Chen et al. 2008, Fries 2010). Although V. ceranae is transmitted within honeybee hives through contaminated feces and pollen stores, transmission may also occur when bees encounter spores on contaminated flowers (Higes et al. 2008b, 2010). Graystock et al. (2015) demonstrated that multiple pollinator parasites, including V. ceranae , can be effectively dispersed onto flowers by competent hosts and then vectored from flowers back to colonies by other pollinator species. Additionally, V. ceranae spores have been detected on the flowers of at least 14 plant genera in the field (Graystock et al. 2020). Therefore, contamination of shared flower resources is a likely mode of transmission for V. ceranae between different pollinator species, with dispersal potentially occurring through defecation on floral surfaces or through the rubbing off of spores that were attached to the bee cuticle (Graystock et al. 2015, Bodden et al. 2019, Piot et al. 2020). Furthermore, Graystock et al. (2015) found that V. ceranaetransmission was very rapid in small experimental flight cages, but they recognized that whether parasite dispersal is similar in nature will depend on the characteristics of pollinator communities and environmental conditions. Despite clear experimental evidence forV. ceranae transmission on flowers, the relationship between specific pollinator visitation patterns and V. ceranae prevalence across managed and wild pollinator species in the field has remained understudied.
Here, we examine whether the prevalence of V. ceranae in managed and wild bee populations is influenced by the floral visitation behaviors of bees in the natural environment. We conducted an observational study of V. ceranae in honeybee (A. mellifera ) and bumblebee (Bombus spp.) populations among different pollinator communities to understand how floral visitation patterns differ among pollinator species and whether the visitation patterns are linked with V. ceranae prevalence in both host species. Specifically, we investigated how V. ceranae prevalence is linked with the number of honeybee, bumblebee, and other pollinator visits to flowers and the time each bee species spent interacting with different parts of the flowers during each visit. We hypothesized that higher numbers of visits and longer visits by potentially infected bees would increase the likelihood of V. ceranae transmission and correlate with higher V. ceranae prevalence. These findings will be important for determining the pollinator visitation behaviors that contribute the most to V. ceranae exposure and subsequent infection in honeybees and bumblebees as well as helping to establish whether V. ceranae transmission on flowers occurs under field conditions.