Competitive context drives pollinator behavior: linking foraging plasticity, natural pollen deposition, and plant reproduction
With ongoing global pollinator declines it is important to understand the functional impact of pollinator species losses. While network-based simulation models of pollinator declines predict that plant communities will be robust to losses of pollinator species, these predictions have never been tested empirically. In four chapters, my dissertation uses both empirical and modeling approaches to explore the impacts of losing pollinator species in alpine plant communities.
First, I test the hypothesis that interspecific interactions among pollinators (rather than fixed species roles) dynamically alter the functional contribution of species in a community and that these dynamic roles can sometimes reduce plant reproductive function. Experimental removal of the most abundant bumble bee pollinator species from an alpine bumblebee community led to a reduction in floral fidelity in the remaining pollinators. Importantly, this behavioral response in the remaining pollinators reduced plant reproduction by increasing the transfer of ineffective heterospecific pollen.
Second, I evaluate how patterns of naturally deposited heterospecific pollen relate to the reproductive output of Delphinium barbeyi, a common subalpine perennial herb in the Rocky Mountains. I found a significant negative interaction between conspecific pollen and the amount of heterospecific pollen whereby the relationship between conspecific pollen and viable seed production became weaker with increasing heterospecific pollen deposition on stigmas.
Third, I explore how traits—specifically pollinator tongue length, which dictates pollinator resource partitioning—influences how pollinators responded to reduced interspecific competition. I found that bees vary in their baseline floral fidelity and that their tongue length explained a large part of this variation. Bees with shorter tongues moved between plant species (floral infidelity) more often than bees with longer tongues. I did not find significant variation in how bee species responded to reduced interspecific competition, but rather saw a guild-wide reduction in floral fidelity in response to the removal of the dominant bee species. Interestingly, I found that competitive context determines the floral fidelity of the pollinators in a community. In this case, as the tongue length of the most abundant bee increases, the site level foraging fidelity decreases.
Network model simulations predict plant populations will be resilient when faced with pollinator extinctions. Importantly, these models are built on the assumption that all interactions in network are positive, potentially overestimating the resilience of plant-pollinator networks. A wide range of studies (including my field work) have shown that many pollinators are actually ineffective at pollinating plant species they visit, and can therefore have negative consequences for plant reproduction. Fourth, I used interaction network simulations to understand how the addition of negative interactions impacts the predictions of resilience when plant communities are faced with pollinator species losses. I found that the addition of negative interactions makes networks less robust to pollinator extinctions, but not always in the same way. This work suggests metrics specific to interaction networks may be important in determining how robust plant-pollinator interactions are to extinctions.
Finally I provide overview on some of the causes and consequences of pollinator decline in the US and offer insight in to how my basic research into pollinator ecology can help pinpoint vulnerable species and communities that should be targeted for conservation efforts.