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

Ecology, March, 1996 by Carolina Murcia

RESULTS

Floral traits and bird feeding

Although flowers of all six species had enclosed corollas roughly similar in length, other pollination-related traits varied widely (Table 2). Sexual architecture of Palicourea most resembled that of its close relative Cephaelis, and least that of the two Acanthaceae. Anthers and the tiny stigmas of the Acanthaceae contacted the forehead and crown (Hansteinia) or chin and upper throat (Dicliptera) of hummingbird visitors, whereas sexual parts of the other four species contacted the bill [ILLUSTRATION FOR FIGURE 1 OMITTED]. Anthers of Besleria, containing numerous tiny [TABULAR DATA FOR TABLE 3 OMITTED] grains (Table 2), primarily contacted the dorsal surface. The short, poricidally deshiscent anthers of Satyria [ILLUSTRATION FOR FIGURE 1 OMITTED] contacted the bill tip but dusted pollen much more widely over the bill surface.

Effect of intervening flowers on Palicourea pollen transfer

Regardless of sexual architecture, intervening flowers of other species strongly influenced pollen transfer from pins (donors) to thrums (recipients) of Palicourea (Table 3A). All treatments involving a second species led to significant reductions, relative to Palicourea-only trials, in numbers of pollen tubes found in recipients' styles (Table 3A, B, D). Reductions averaged 76%. Contrary to the prediction (see Introduction), the effect was strongest when the intervening flowers were Hansteinia or Dicliptera (Table 3A, c), whose exserted sexual parts differed most from those of Palicourea (Table 2, [ILLUSTRATION FOR FIGURE 1 OMITTED]).

[TABULAR DATA FOR TABLE 4 OMITTED]

Pollen transfer should be viewed not only from the perspective of recipients (i.e., receipt of compatible grains, hence potential for producing seeds) but also from that of pollen donors (i.e., dispersal of grains to compatible recipients, hence potential for siring seeds). Carry-over distance is one measure of a donor's potential for siring seeds (Feinsinger and Busby 1987, Feinsinger et al. 1988). Here, intervening flowers had no significant effect on maximum (Table 4) or median carry-over distance, nor did effects of the two Acanthaceae differ from the other species (Table 4).

Pollen loads on birds

As expected based on their sexual architecture [ILLUSTRATION FOR FIGURE 1 OMITTED], pin flowers of Palicourea deposited most pollen close to the bill tip, virtually all within the distal 10 mm [ILLUSTRATION FOR FIGURE 2 OMITTED]. Despite the flowers' radial symmetry, much more pollen was deposited on the dorsum than on the renter of hummingbird bills, presumably because [TABULAR DATA FOR TABLE 5 OMITTED] cause hummingbirds feeding at Palicourea typically approached the flower from above the plane of flower orientation and pressed most closely against the dorsal anthers, as also occurs in the field. Besleria, with its exclusively dorsal anthers [ILLUSTRATION FOR FIGURE 1 OMITTED], deposited almost all pollen on the bill's dorsum, again near the tip [ILLUSTRATION FOR FIGURE 2 OMITTED]. Most Satyria pollen remained within 14 mm of the bill tip, although some reached the bird's throat. Pollen placement by the two Acanthaceae matched the architecture of their sexual parts, Hansteinia depositing pollen on top of the bird's head and Dicliptera exclusively below [ILLUSTRATION FOR FIGURE 2 OMITTED]. Indeed, each species-specific pattern of pollen placement [ILLUSTRATION FOR FIGURE 2 OMITTED] matched stamen architecture quite closely (Table 2, [ILLUSTRATION FOR FIGURE 1 OMITTED]) and, taken alone, lent credence to the Sexual Architecture Hypothesis. For example, pollen placement by Hansteinia and Dicliptera clearly differed from that by Palicourea.