COS 61-2 - Morphological trait matching between bees and flowers drives pairwise interactions

Wednesday, August 10, 2016: 1:50 PM
Grand Floridian Blrm B, Ft Lauderdale Convention Center

ABSTRACT WITHDRAWN

Joseph Zientek, Rutgers University; Michael E. Roswell, Rutgers University; Daniel P. Cariveau, University of Minnesota; Rachael Winfree, Rutgers University

Background/Question/Methods

Pairwise interactions between species influence the stability and function of ecological networks. Morphological trait matching between pollinators and flowers, such as matching between bee tongue length and floral corolla length, partly determines pairwise interactions and thus network structure. However, because bee tongue length is difficult to measure, studies have not measured corolla-tongue matching on the community level with large data sets or high-resolution measurements. We ask if corolla depth imposes a lower limit on tongue length of visiting bees. First, to measure pairwise interactions, we net-collected all bees visiting all species of flowers in 16 old fields throughout New Jersey 4 times per summer over 4 years. Second, we developed standardized, repeatable specimen handling and microscopy methods for measuring tongue length and corolla depth. We used quantile regression with mixed effects to analyze the rate of change of different quantiles of tongue length attributable to corolla depth. We included sampling year and round as fixed effects and site as a random effect in our model to account for the difference in floral and bee community composition at each sampling event. We expected the 10thpercentile (smallest) tongues would increase with corolla depth, while the larger tongues would show no response.

Results/Conclusions

We collected a total of 8,165 bees (129 species) from 66 flower species. Separately, we measured 786 tongues from 100 species and estimated tongue length using intertegular distance for an additional 189 bee species. We also measured 1622 corollas from 66 flower species. The 10th percentile tongue length increased in response to increasing corolla depth with a slope of 0.49 (p<0.001). The 75th percentile of tongue length had a slope indistinguishable from zero, suggesting that bees with long tongues are not restricted by corolla depth. The lower limit of tongue length imposed by corolla depth is defined as y = 0.49x + 4.31, with y representing tongue length and x representing corolla depth (mm). Therefore, tongue length is longer than corolla depth until corolla depth surpasses 8.45mm. Our empirical community-level findings support the existence of trait matching on the network scale while using higher quality measurements and larger data sets than those used in previous studies. Our work provides strong quantitative evidence for trait-driven ecological complementarity in bee foraging preferences. This study demonstrates a mechanism for bee and flower coevolution through the morphological conditioning of interactions and underlines the importance of floral diversity for supporting a diverse fauna of pollinators.