Differential Importance of Beetle Species Pollinating Dieffenbachia Longispatha (Araceae)

Dieffenbachia longispatha (Araceae) is pollinated by scarab beetles (Cyclocephala spp. and Erioscelis sp.) at the La Selva Biological Station, Costa Rica. The relative contributions of each beetle species to female reproductive success of D. longispatha were determined by observing inflorescences during the flowering seasons of 1982 and 1983 and recording the number of visiting beetles of each species. Fruit production was determined for each inflorescence as fruits ripened, 9 mo after flowering. The three most abundant beetle species differed in abundance, recapture rates, pollen loads carried, and behavior at inflorescences. Despite these differences, the effect of the single visit on fruit—set of D. longispatha (the "pollinator effectiveness" of each species) did not vary among the beetle species. The "importance" of each pollinator species, estimated in previous studies from the abundance and effectiveness of each species and determined in this study from examining fruit—sets of inflorescences visited by varying numbers of beetles, was not a positive, linear function of beetle abundance. In 1982, when visitation rate of all beetle species was low, fruit—set was positively correlated with the number of beetles in an inflorescence but "per visit" effectiveness decreased with increasing visitation, resulting in a saturation of fruit—set at 42%. Beetle identity (to species) did not explain a significant proportion of the variance in fruit—wet in 1982. In 1983, beetle abundance quadrupled relative to 1982 and fruit—set was positively correlated with beetle numbers up to four, but visitation by more than four beetles resulted in a decreased in fruit—set. A high proportion of Cyclocephala gravis produced higher than predicted fruit—set, while a high proportion of Erioscelis yielded fruit—sets lower than was predicted by total number of beetles. The presence of large numbers of Eriosceliis within inflorescences in 1983 and the low fruit—sets of those inflorescences explained the significantly negative slope of the regression for inflorescences fisted by more than four beetles. Previous models have predicted saturation of fruit—set with increasing pollinator visitation (as seen in 1982) but have not predicted a significant decline in fruit—set with pollinator abundance. The proportion of the variance in fruit—set explained by beetle abundance was low for both years (2—8%), suggesting that other factors contribute to female reproductive success in Dieffenbachia (i.e., number of potential mates and the distances to potential mates in the population). Use of common indices of pollinator contribution to seed set, such as pollinator abundance and the number of seeds produced as a result of a single visit, would lead to erroneous conclusions if applied to Dieffenbachia and its beetle pollinators, or to other pollination systems where there is cumulative floral damage with increasing visitation frequencies.