Maureen Page collected and analyzed the data and led the writing of the manuscript. Both authors read and approved the final manuscript. Andrew Buderi, Staci Cibotti, Deanna Deterding, Lindsey Hack, Heather Spaulding, and Annie Zell assisted with field work. Skyler Burrows of the USDA Bee Lab in Logan, Utah, and Joel Gardner of the University of Manitoba, Canada helped us identify bees to species. We are also grateful to Randy and Eric Oliver who helped supplement sites in the Sierra with honey bees as well as landowners who allowed us to place hives on their land. Research support was provided by a U.S. Department of Defense Graduate Fellowship, Davis Botanical Society Grants, and a UC Davis Vansell Scholarship awarded to Maureen Page. Additional support was provided by NSF grant DEB1556885, the UC Davis H. Laidlaw Endowment, the California Almond Board, and multiple USDA grants awarded to Neal Williams.Introduced and invasive species are fundamentally altering the structure of ecological communities , leading to increased species extinctions and biodiversity loss . In addition to shifting community composition, exotic species can impact ecosystem functioning by altering the growth and fitness of primary producers and become costly agricultural pests . However, hydroponic vertical garden the impacts of exotic species may be more nuanced when they engage in keystone mutualisms like pollination.

In these cases, there is potential for direct negative impacts through competition with native species for shared resources but also potential to benefit other species through interactions that increase primary productivity and plant reproduction . Indeed, recent meta-analyses of the invasive species literature largely ignore exotic mutualists and we are only beginning to understand the impacts of invasive species on mutualistic interactions . The impacts of exotic species are even more contentious when the introduced species is actively managed for recognized benefits to humanity. For example, growing demand for agricultural pollination has led to steady increases in managed populations of the European honey bee , which has become a dominant floral visitor in many plant communities worldwide . Despite mounting evidence that honey bees compete with wild bees for floral resources with potential consequences for plant pollinator interactions and wild bee reproduction , the importance of honey bees as pollinators has led beekeeping to be promoted and even subsidized in some natural habitats . However, the importance of honey bees does not automatically imply that honey bee introductions will benefit plant populations . Indeed, we currently lack robust studies investigating how honey bee introductions impact pollination and this knowledge gap limits our ability to inform conservation policies that safeguard plant and pollinator populations. Assessing the overall impact of honey bee introductions on pollination is complicated because abundant honey bees can influence pollination directly, through their flower visits, but also indirectly, by competitively influencing visits from other pollinators.

Pollination is expected to increase with increased floral visitation and honey bees visit flowers frequently . However, a handful of studies have documented direct negative effects of high visitation rates by introduced pollinators, whereby increased visits increase pollen deposition but also damage stigmas or lead to clogging of styles with growing pollen tubes , ultimately reducing successful reproduction. Honey bees can also damage flowers while nectar robbing, increasing floral abortion . In addition to visit numbers, the relative quality of visits also influences pollination , and honey bees can be ineffective at depositing the pollen they extract . Though seemingly less dramatic than direct floral damage, ineffective pollinators can indirectly decrease pollination by reducing pollen available for deposition by more effective visitors . Furthermore, regardless of their relative pollination effectiveness, honey bees can deplete floral rewards that attract pollinators , thus diminishing other pollinator visits . These direct and indirect effects can add to one another, or they can cancel each other out if effects are of opposite sign but similar magnitude . Quantifying both direct and indirect impacts is needed to understand overall fitness consequences for plants, but few studies of honey bee introductions carefully partition direct and indirect effects. Indeed, across 29 studies of honey bee effects on pollination identified by Mallinger et al. , all but four were purely correlative studies and none investigated both direct and indirect effects of honey bee abundance simultaneously.

Studies that investigate both direct and indirect effects could shed light on how impacts vary across systems. For example, the generally positive direct effect of honey bee visits may be of greater importance in the absence of competition; in cases where native pollinators have become rare or locally extinct, honey bees often increase pollination and can even “rescue” plant populations from reproductive failure in isolated habitat fragments . However, negative indirect effects may occur and even outweigh direct effects when honey bees competitively displace native pollinators, especially when honey bees are ineffective substitutes . In this study, we investigated whether honey bee introductions in montane meadows competitively displace native bees and impact pollination of Camassia quamash , anherbaceous perennial plant which is an important floral resource for native bees and culturally important within indigenous communities in North America .We lacked sufficient power to directly test the relationship between honey bee abundance and native bee abundance and instead used focal plant visits to understand potential shifts in the community of C. quamash visitors. We evaluated how the number of honey bee visits responded to honey bee introductions by fitting a model which included honey bee abundance as a fixed effect and evaluated how native bee visits responded to honey bee introductions by fitting a model which included honey bee visits to focal plants and native bee abundance as fixed effects. Both models also included site and sample round as separate random effects. We fit this model using the lmer function in the lme4 package and tested for significance of fixed effects using likelihood ratio tests. All analyses were conducted in R . We determined the association between native bee and honey bee C. quamash visitation and three measures of pollination: pollen deposition, pollen tubes, and seed set. Because these measures were taken from the same plants, but not necessarily the same flowers, we performed separate analyses using generalized linear mixed effects models . Each model included as fixed effects the abundance of honey bees visiting C. quamash and and the abundance of native bees visiting C. quamash. We also included random intercepts for site and sample round. Pollen deposition and pollen tube data were over-dispersed, so we modeled responses using negative binomial distributions. We modeled seed set as a binary response where fertilized ovules were successes and unfertilized ovules were failures and included plant as a random effect to account for non-independence of flowers on the same plant. For all models, vertical vegetable tower we used the glmmTMB package , and calculated p-values using likelihood ratio tests. Using data from the controlled honey bee visit experiments described above, we assessed the direct relationship between increasing honey bee visits and C. quamash pollination by fitting a GLMM which included the number of honey bee visits as a fixed effect as well as date and plant ID as separate random effects to account for non-independence of flowers observed on the same plant and/or day. We modeled C. quamash pollination as a binomial response: successes were flowers that produced fertilized ovules and failures were flowers with no fertilized ovules. We tested for significance using likelihood ratio tests. We evaluated how pollen and nectar availability responded to honey bee introductions by fitting two separate GLMMs which included as fixed effects the abundance of honey bees in meadows, the abundance of native bees in meadows, and , to control for baseline pollen and nectar resources, either the mean pollen availability or the mean nectar availability in unvisited bagged flowers.

Both models included site and sample round as separate random effects. Data collectors varied in their ability to extract nectar from flowers, so we also included data collector as a random effect in both models. Nectar and pollen data were zero-inflated, so we modeled nectar and pollen availability as presence/absence binary responses. We calculated p-values using likelihood ratio tests. To assess whether native bees were more effective than honey bees as pollinators of C. quamash we first confirmed that pollinator taxon was an important predictor of effectiveness using generalized linear models. We modeled seed set as a binomial response where successes were flowers that produced fertilized ovules and failures were flowers that produced no fertilized ovules. Flies and large-bodied Andrena spp. were infrequent visitors , so we removed their visits from the analysis. Our maximal model used three predictors: the pollinator taxon observed, whether the stigma was contacted, and the day of the observation. We tested for significance of predictors by stepwise model simplification and performed Chi-square tests to compare individual taxa.Honey bees are ineffective pollinators of C. quamash – Although honey bees visit C. quamash frequently, they are ineffective pollinators compared to native bees and extract pollen and nectar without pollinating C. quamash flowers. Both visit frequency and visit quality determine the relative importance of different floral visitors as pollinators . In some other systems, frequent honey bee visits increase pollination, even when honey bees are less effective than other visitors on a per-visit basis . However, in our system, increased visit quantity by honey bees does not compensate for poor visit quality. As such, the direct contribution of honey bees to pollination in this system is negligible, and, if anything, negative. We suspect that honey bees are ineffective pollinators because of their behavior at flowers. Native bees contacted stigmas nearly six times more often than honey bees, who frequently removed nectar from behind petals without contacting reproductive structures. Such “robbing” is common for honey bees and results in low stigma contact compared to other pollinators . Indirect negative effects of honey bee visits may be severe when this behavior is frequent. Possible direct effects of honey bee abundance on pollination – If there is a direct negative effect of honey bee visits on C. quamash pollination, the mechanism is not obvious. We did not observe signs of stigma damage and, although excessive pollen receipt can lead to pollen tube competition , increased honey bee visits were not associated with changes in pollen deposition. Visitors that remove nectar without pollinating can directly reduce fitness by forcing plants to allocate resources to refilling nectar instead of fertilizing ovules . However, C. quamash does not refill nectar in the populations we studied and artificial nectar removal did not affect seed set for unvisited plants . Other possible direct negative effects of visitation include fungal infections , ovary damage by nectar-foragers , and floral abortion induced by nectar robbing , but these mechanisms are rarely documented. Thus, there might be direct negative effects of honey bee visitation, but indirect effects mediated by changes in the visitor community are more convincing. Clear indirect effects of honey bee abundance on pollination – Honey bee abundance indirectly decreased C. quamash pollination by reducing visits from more effective native bee pollinators. These reductions are likely the result of exploitative competition because both pollen and nectar availability declined with increased honey bee abundance, as has been shown in other systems , and resource competition can lead native bees to shift visits to different meadows or plant species . Although resource competition seems a more likely explanation, other competition avoidance behaviors, including scent-cues , could also reduce native bee visits in response to increased honey bee visits. Past studies have demonstrated that honey bees compete with wild bees for floral resources, but our study is unique in that we clearly document mechanistic evidence of floral resource depletion. Furthermore, this study is among the first to partition direct and indirect pathways through which introduced honey bees influence pollination. By isolating the minimally negative direct effect of honey bee visits, we can confidently conclude that indirect effects drive the magnitude of the negative association between honey bee abundance and pollination. Generalizability of findings – Honey bees were absent in most meadows before we experimentally introduced hives and the native bee community was abundant and diverse. In systems where honey bees are a natural element of bee communities or when other pollinator populations are diminished , the negative effects we observed might be lessened or even reversed.