Inflorescence architecture influences how pollinators interact with flowers, and thus affects plant fertilization and reproduction. Although pollinator species are known to differ in the amount of pollen they transfer per flower visit, few studies have investigated the behaviors underlying these differences. To explore how pollinator behavior affects the pollination of different plant species, we made detailed observations of bees as they visited flowers of Phacelia tanacetifolia and Monarda fistulosa, and then collected stigmas and counted pollen grains to link behavior to pollen deposition. We identified each visiting bee to one of eight morphogroups during its visit. With our visit and pollen deposition data, we asked: (1) How do bee behavior and bee interaction with the floral architecture of each plant vary across bee morphogroups; and (2) are any bee behavioral patterns associated with the level of pollination delivered? Results were analyzed with likelihood ratio X2 tests (bee behavior variation across morphogroups) or t-tests (behavioral patterns and pollen deposition), corrected for multiple comparisons with a False Discovery Rate of 0.05.
Results/Conclusions
We observed 195 bees visit P. tanacetifolia and 133 bees visit M. fistulosa. Because P. tanacetifolia and M. fistulosa have many flowers per inflorescence, each observed bee visited many stigmas, resulting in more stigmas than bees (n=1533 stigmas for P. tanacetifolia and n=1244 for M. fistulosa). For both plant species, we found that bee morphogroups interacted with floral architecture in different ways. For example, when visiting multiple flowers within a P. tanacetifolia inflorescence, bees in the genera Hylaeus and Bombus tended to land on and walk across petals, while bees in other genera did not (p=0.002). Similarly, when visiting multiple flowers within a M. fistulosa inflorescence, bees in the genus Bombus walked between flowers while bees in other genera flew (p=0.010). For P. tanacetifolia, but not M. fistulosa, behavioral patterns were linked with the level of pollination delivered. Specifically, bees that walked between stigmas of P. tanacetifolia (i.e. Hylaeus and Bombus) deposited 6.5 times as many pollen grains per visit compared with bees that flew between stigmas (p=0.006). Our results show that bee morphogroups have different behaviors when visiting flowers, and that these behaviors can translate into differences in pollen deposition.