Tongue length appears to be important for structuring foraging preferences but we are lacking experimental work that evaluates how traits such as tongue length influence pollinators’ response to competition. We systematically manipulated interspecific competition in bumble bee communities through targeted single species removals and examined patterns of species-specific plasticity in resource use in the remaining pollinators. Specifically we examined to what extent tongue length explains species-specific difference in the pollinators’ foraging behavior in response to release from competition. Our sites are self assembled communities with natural diversity varying in plant and bee community composition. Utilizing removals in these natural communities allowed us to examine if the identity of the most abundant bee species, influences bee foraging behavior. We focused on the foraging response of the remaining bees in the community with respect to floral fidelity, or, within plant species movements within a single foraging bout. Floral fidelity is critical for many plants species’ reproductive success because transfer of conspecific pollen must occur in order for fertilization to take place.Study Sites: We worked in 28 subalpine meadow sites in the landscape surrounding the Rocky Mountain Biological Laboratory , plastic potting pots in the Gunnison National Forest, western Colorado, United States.
Eachsite consisted of a 20 × 20-m plot, all with the same dominant plant species to minimize, as much as possible, plant-community-driven differences in foraging behavior. A minimum distance of 1 km separated any two sites. We collected data over three summer growing seasons , in 2010, 2011 and 2013. Manipulations. We assessed each plot in a control state, waited one day, and then assessed in a manipulated state. We kept the interval between control and manipulated states short because of the rapid turnover in flowering phenology in our high-altitude system, allowing us to keep the plant community constant in our control–manipulation comparisons . Manipulations reduce interspecific competition through the temporary, nondestructive removal of the most abundant bumblebee species in each plot. We determined the most abundant bee via inventory of Bombus species richness and abundance on the control day using nondestructive aerial netting, with two field team members netting for a 20-min period, not including handling time . To avoid double-counting, we kept each bee in an individual glass vial, identified to species, and kept in a cool, dark cooler until the inventory time period was over, at which point bees were released. On the manipulation day we removed the most abundant bee species . The removals were accomplished through targeted hand-netting, and we minimized disturbance of other bees and vegetation by carefully placing the insect net over entire inflorescence and allowing bee individuals to fly up into the net . Captured bees were transferred to vials and placed in a cooler during the manipulation and released unharmed afterward.
We used as much time as necessary to remove essentially all individuals of the target species from the sample plot and immediately adjacent area . We left a period of at least 30 minutes between manipulative bee removals and subsequent sampling to minimize the impact of the disturbance on the foraging activities of other bees. We recorded both the abundance of removed individuals, as well as the number of un-captured “escapees” that were observed during bee sampling. Each site was only used once in a manipulated and controlled state per year . Foraging observations: We directly followed the foraging sequences of Bombus individuals in both the control and manipulated states. We recorded the identity of each plant species visited in a foraging sequence. We discontinued an observation when the bee was lost from sight, when it ventured more than 5 m outside of the plot, when it had been observed for 10 full minutes, or when we had tallied 100 individual plants visited. We discarded observations of bees that visited fewer than five plants. The number of individuals observed per/state/site varied due to bee abundance .Tongue Length measurements: The data on proboscis lengths of workers was taken from published measurements of bumblebees collected on the Front Range of the Colorado Rocky Mountains and overlap with the sites and species that we used for this study . Measurements indicate the sum of the individual lengths of the prementum and glossa. The mean tongue length of each bumble bee species was assigned to each individual bee and used in the trait analysis .
Floral fidelity was measured as the binomial counts of individual bee foraging movements that were conspecific vs. heterospecific . We used GLMMs with binomial errors using the logit link in the lme4 package for R to model the floral fidelity response variable . Data from individual bees foraging within a site cannot be considered independent and therefore we used site as a random effect . Relative to a binomial distribution, our data were overdispersed, which we corrected by including an individual-level random effect . We used the R statistical programming language for all models.Our results demonstrate that bees vary in their floral fidelity and that tongue length explains a large part of this variation. Bees with shorter tongues move between plant species more often than bees with longer tongues. We did not find significant variation in the response of bee species to a reduction in interspecific competition, but rather saw a guild-wide reduction in floral fidelity in response to the removal of the dominant bee species . Finally, our results suggest that tongue length of the most abundant bee species, a site-level attribute, explains much of the site-to-site variation in pollinator foraging behavior. In particular, we found that as the tongue length of the most abundant bee increases, the site level foraging fidelity decreases. We found that bumble bee species vary in the degree to which they move between different plant species within a single foraging bout, and tongue length explains much of the variation. Some suggest that long tongued bees should exhibit broader resource usage patterns because their traits permit them access to a wider range of flower types . In contrast, short-tongued bees should act as specialists, with a more restricted range of resource use options, rarely able to access the nectar at the base of the flowers with long corollas . Our results suggest the opposite pattern, that shorter tongue bees are more labile with their foraging patterns and on average move between plant species within a single foraging bout more often than longer tongue bees. We suggest the following interpretation, because long tongues enable bees to access flowers with better rewards and maintain a monopoly on those rewards, they may have less incentive than short-tongued species to move between plant species while foraging. While longer tongue bumble bees are capable of foraging on flowers with short corollas it would provide less energetic gain, potentially making behavioral plasticity less profitable . In contrast, the shorter-tongued bees in our system tend to have smaller bodies and are more likely to depend on resources within a more restricted foraging range . These limitations could favor a labile foraging habit, with shorter tongue bees constantly assessing the resource availability and competitive context in their community. As such, shorter-tongued bees more readily switch between plant species. While we saw little variation in species-specific foraging responses to our manipulations, we found an overall reduction in floral fidelity across sites after the removal of the most abundant bee species. Our results build on the findings of Brosiand Briggs , reaffirming a guild-wide reduction in floral fidelity in response to a reduction in interspecific competition. This study adds an additional two years, 8 sites and 165 bee individuals to our previous study, confirming that the guild-wide results found in Brosi and Briggs are robust. Variation in bumble bee floral fidelity is largely explained by the tongue length of the most abundant bee species in each site. This means that pollinator foraging behavior is context dependent and is determined by the most abundant bee species. In general, raspberry container growing short tongued bees exhibit lower floral fidelity than long tongued bees, and when they are in a site that has a long tongued bee removal, their reduction in floral fidelity is magnified. Bumble bees are large bodied insects that require many floral resources to keep their hive growing throughout the growing season.
As such, we might expect strong competition between these species, and tongue length, arguably the trait most relevant for resource acquisition, could dictate how resources are partitioned within a community, ultimately drive the assembly of bumble bees within communities . Pyke proposed that bumble-bee species with similar tongue lengths could not exist in his altitudinal alpine transects presumably because the bees compete for floral resources. But later studies did not support this pattern , and, as in our study, found that bumble bee species with similar tongue lengths co-ocurred within a community. This has left researchers to wonder if the coexistence of many bee species with substantial overlap in their life history requirements is possible because bumble bee species compete for something other than flower resources allowing so many similar species to co-occur . Our study suggests that in our system, bumble bees do in fact compete for floral resources and that longer tongue bees seem to elicit competition that is experienced across the range of trait values seen in our sites . The willingness of short tongue bees to exhibit behavioral plasticity may allow for such a large number of seemingly similar bee species to coexist in a community. Future work should examine the extent to which this plasticity is adaptive and assess the fitness costs that may result from the willingness to switch floral resources in response to a reduction in interspecific competition. Most pollinators are generalist foragers that can switch between plant species within a single foraging bout . When pollinators move between plant species, they can transfer heterospecific pollen to plant stigmas which in turn can reduce plant reproduction . We found an effect in which competition from a long tongued bee changes the foraging behavior of the rest of the bees in a way that could be detrimental to plant reproduction. From a plant’s perspective, not only do short tongue bees exhibit behavior that likely results in the transfer of heterospecific pollen, but when short tongue bees are in communities in which a longer tongue bee is most abundant, they exhibit even greater floral infidelity, making the likelihood of heterospecific pollen deposition even greater .Pollinator species are on the decline globally . Bumble bees in particular are experiencing population range contractions due to climate change as well reductions in abundance due to disease , agricultural intensification and pesticide use . In a changing world where we are likely to experience an emergence of new interactions via range shifts, introduced species and climate change, exploring how the competitive landscape shapes foraging plasticity will help us generalize to other plant pollinator systems and begin to better predict the functional implications of competitive interactions.Pollinator losses are increasing across the globe which could have potentially strong negative effects on the plants that rely on them for pollination . Network-based simulations suggest that plant communities will be very robust to pollinator extinctions . This robustness is likely driven by two features of network structure. First, pollination networks are dominated by generalist interactions; most plants and pollinators each interact with several species from the other group over the course of their lives . Second, pollination networks have a nested structure in which specialist plants and pollinators tend to interact with a subset of the species in the other group that generalists interact with . Nestedness leads to an asymmetric interaction structure where specialists from one group tend to act with generalists, not specialists, from the other, which could reduce linked extinctions if specialists are vulnerable to stochastic extinctions . One feature of binary-graph plant-pollinator simulation models that may overestimate network robustness is that all interactions in a binary-graph network are positive. A typical binary-graph simulation modeling approach is that if at least one link remains between a plant and a pollinator, the plant will continue to persist . While the assumption of “all interactions positive” is a reasonable starting point in a mutualistic network model, empirical evidence suggests at least two ways in which pollinators can have negative consequences for the reproduction of plants they interact with . First, the benefit that pollinators have on plant reproduction is sensitive to how “faithful” pollinators are to particular plant species in a single foraging bout.