ABSTRACT
Many species of animal-pollinated flowers are known to vary widely in the nectar content of flowers. Some proportion of flowers in many species is apparently nectarless,and such flowers are believed to be 'cheaters'. Cheating may explain a part of the variability in nectar content.If cheating exists as a qualitatively different strategy then we expect bimodality in the distribution of nectar content of flowers. It has been shown in a multispecies study that gregarious species have a higher proportion of cheater flowers. We studied the frequency distribution of total nectar sugar in two gregariously flowering species Lantana camara and Utricularia purpurascens, which differed in other floral and ecological characters. At the population level, both the species showed significant bimodality in the total sugar content of flowers. The obvious sources of heterogeneity in the data did not explain bimodality. In Lantana camara, bimodality was observed within flowers of some of the individual plants sampled. In Utricularia purpurascens the proportion of nectarless flowers was more in high-density patches, suggesting that the gregariousness hypothesis may work within a species as well. The results support the hypothesis of cheating as a distinct strategy since two distinct types of flowers were observed in both the species. The effect of density in Utricularia purpurascens also supports the gregariousness hypothesis.
