Author contributions. LSA, REI and PCS conceived of and designed the study. PCS and IWF conducted chemical analysis of pollen and synthesized spermidines. AEF, RLM, PRA, LMC, PMD, and SL carried out bioassay experiments. AEF analyzed data and prepared figures. LSA wrote the manuscript with substantial contributions from PCS. All co-authors read and provided feedback on the manuscript.

1. Social insect castes and sexes differ in many ways, including morphology, behavior, and sometimes ploidy level. Recent studies have found that consuming sunflower pollen reduces the gut pathogen Crithidia bombi in workers of the common eastern bumble bee (Bombus impatiens). Here, this work is extended to the reproductive individuals that represent colony fitness-males and queens-to assess if the medicinal effects of sunflower pollen vary with bee caste and sex. 2. This study examined the effect of sunflower pollen compared to a diverse wildflower pollen mix on infection in worker, male, and daughter queen commercial B. impatiens. Bees were infected, fed either sunflower pollen or wildflower pollen for 7 days, and then infection levels were assessed. 3. Compared to wildflower pollen, sunflower pollen dramatically reduced Crithidia infection in workers and daughter queens, but not males. Infection levels were very low for both diets in males; this could be due to low pollen consumption or other mechanisms. 4. Reducing Crithidia infection in young queens before they undergo hibernation is important for population dynamics since infected queens are less likely to survive hibernation, and those that do are less likely to successfully establish a nest the following spring. Because sunflowers bloom in late summer when new queens are emerging, sunflowers could provide an important dietary component for queens during this critical life stage. Deepening our understanding of how diet impacts pathogens in reproductive bees, as well as workers, is crucial to maintain healthy pollinator populations.

Because carnivorous plants rely on arthropods as pollinators and prey, they risk consuming would-be mutualists. We examined this potential conflict in the Venus flytrap (Dionaea muscipula), whose pollinators were previously unknown. Diverse arthropods from two classes and nine orders visited flowers; 56% of visitors carried D. muscipula pollen, often mixed with pollen of coflowering species. Within this diverse, generalized community, certain bee and beetle species appear to be the most important pollinators, on the basis of their abundance, pollen load size, and pollen fidelity. Dionaea muscipula prey spanned four invertebrate classes and 11 orders; spiders, beetles, and ants were most common. At the family and species levels, few taxa were shared between traps and flowers, yielding a near-zero value of niche overlap for these potentially competing structures. Spatial separation of traps and flowers may contribute to partitioning the invertebrate community between nutritional and reproductive functions in D. muscipula.

Commercial bumblebees have become popular models to understand stressors and solutions for pollinator health, but few studies test whether results translate to other pollinators. Consuming sunflower pollen dramatically reduces infection by the gut parasite Crithidia bombi in commercially reared Bombus impatiens . We assessed the effect of sunflower pollen on infection in wild B. impatiens , Bombus griseocollis, Bombus bimaculatus and Bombus vagans . We also asked how pollen diet (50% sunflower pollen versus wildflower pollen) and infection (yes/no) affected performance in wild B. impatiens microcolonies. Compared to controls, sunflower pollen dramatically reduced Crithidia infection in commercial and wild B. impatiens, had similar but less dramatic effects in B. bimaculatus and B. vagans , and no effect in B. griseocollis . Bombus impatiens, B. bimaculatus and B. vagans are in the same subgenus, suggesting that responses to sunflower pollen may be phylogenetically conserved. In microcolonies, 50% sunflower pollen reduced infection compared to wildflower pollen, but also reduced reproduction. Sunflower pollen could control Crithidia infections in B. impatiens and potentially close relatives, but may hinder reproduction if other resources are scarce. We caution that research using managed bee species, such as B. impatiens , be interpreted carefully as findings may not relate to all bee species.

Parasites are linked to the decline of some bee populations; thus, understanding defense mechanisms has important implications for bee health. Recent advances have improved our understanding of factors mediating bee health ranging from molecular to landscape scales, but often as disparate literatures. Here, we bring together these fields and summarize our current understanding of bee defense mechanisms including immunity, immunization, and transgenerational immune priming in social and solitary species. Additionally, the characterization of microbial diversity and function in some bee taxa has shed light on the importance of microbes for bee health, but we lack information that links microbial communities to parasite infection in most bee species. Studies are beginning to identify how bee defense mechanisms are affected by stressors such as poor-quality diets and pesticides, but further research on this topic is needed. We discuss how integrating research on host traits, microbial partners, and nutrition, as well as improving our knowledge base on wild and semi-social bees, will help inform future research, conservation efforts, and management.