Interspecific pollen loss by hummingbirds visiting flower mixtures: effects of floral architecture

Nevertheless, species-specific patterns of pollen placement utterly failed to explain the effect of interspecific visits on loads of Palicourea pollen already in place. A single visit to a flower of any other species drastically altered the number and distribution of Palicourea grains carried by the hummingbird [ILLUSTRATION FOR FIGURE 2 OMITTED]. In all cases, the great majority of those Palicourea grains that had been carried dorsally disappeared. Grains carried ventrally were affected less. There was no evidence that Palicourea grains were simply displaced to novel sites on the bird's bill or head. Examining sexual architecture of the second flower alone, we could not have predicted these impacts. For example, Dicliptera, the only species whose sexual parts mainly contacted hummingbirds from below [ILLUSTRATION FOR FIGURE 1 OMITTED], had the least impact, of any species we examined, on the ventral portion of the Palicourea pollen load [ILLUSTRATION FOR FIGURE 2 OMITTED]. Hansteinia, despite the dorsal contact its exserted sexual parts made with hummingbirds [ILLUSTRATION FOR FIGURE 1 OMITTED], removed most extant Palicourea grains from both dorsal and ventral surfaces of the bird's bill [ILLUSTRATION FOR FIGURE 2 OMITTED]. Effects of female-phase Besleria flowers [ILLUSTRATION FOR FIGURE 2 OMITTED] on Palicourea pollen loads closely resembled effects of male-phase Besleria flowers [ILLUSTRATION FOR FIGURE 2 OMITTED].

A pre-existing load of Palicourea pollen also influenced, if less strongly, subsequent pollen placement by a second species [ILLUSTRATION FOR FIGURE 2 OMITTED]. Palicourea-laden birds picked up fewer grains from the second flower than did clean birds, especially in Besleria and Dicliptera trials [ILLUSTRATION FOR FIGURE 2 OMITTED].

Interspecific pollen transfer to sexual structures

Thus far, the evidence suggests that many Palicourea pollen grains were scraped off when birds probed flowers of a second species, not only during the experiments reported in Fig. 2 but also in the pollen-transfer experiments reported in Table 3. Were the sexual structures of the other flowers directly responsible? Examinations of anthers and stigmas from intervening flowers in the pollen-transfer experiments provided little evidence. Those structures did carry some Palicourea grains (Table 5), but the numbers involved were generally far too low to account for the effects documented in Table 3A. Likewise, the pollen-load data [ILLUSTRATION FOR FIGURE 2 OMITTED] suggest that the grains lost to heterospecific anthers and stigmas alone (Table 5) constituted only small subsets of the Palicourea loads carried, and lost, by hummingbirds. Among the intervening species, those that reduced Palicourea pollen transfer the most (Table 3) had not necessarily accumulated the most Palicourea grains on their own anthers and stigmas (Table 5). Stigmas and anthers of Besleria accumulated more Palicourea grains than did those of other intervening species; stigmas and anthers of Hansteinia accumulated the fewest (Table 5), in line with the Sexual Architecture Hypothesis. However, Palicourea pollen transfer suffered more from Hansteinia than from Besleria (Table 3C).