Friendly neighbours? Investigating ecological facilitation between Thunbergia atriplicifolia (Acanthaceae) and Exochaenium grande (Gentianaceae).

Abstract

Wildflowers that co-exist, share similar floral signals and flower simultaneously, may interact with the same pool of pollinators, therefore sharing pollinator species and individuals. Floral similarity could be a result of a common ancestor, convergent floral syndrome or mimicry. Pollinator sharing may have consequences for the pollination success and fecundity of one or more of the interacting wildflowers. This study was conducted in grassland vegetation of southern KwaZulu-Natal to investigate an observed floral similarity between Thunbergia atriplicifolia and Exochaenium grande, document pollinator sharing between the two species and investigate the consequences of pollinator sharing on (1) interspecific pollen transfer between the two species and (2) seed set of individuals of E. grande that grow within sparse and dense populations of T. atriplicifolia. It was hypothesized that E. grande is a mimic of T. atriplicifolia. Investigated aspects of floral similarity include: flowering phenology, flower colour, flower size and nectar production. Weekly surveys indicated that both species flower simultaneously but E. grande occurs at much lower density and peaks in flowering later than does T. atrplicifolia. Flowers of the two species have similar spectral reflectance and in terms of a bee vision model can be considered indistinguishable to bees. There is also overlap in flower size of the two species. They provide similar amounts of nectar but nectar of E. grande has a higher sugar concentration. Both species are dependent on pollinators for seed production and are pollinated by generalist solitary bees belonging to the Apideae and Hactilidae family, with some species shared by both plant species. Stigmas of T. atriplicifolia and E. grande carried heterospecific pollen in varying degrees. Estimates of pollen production indicated that E. grande produces more pollen than T. atriplicifolia. Although interspecific pollen transfer was confirmed, it is not yet clear whether this improper pollen transfer affects seedset. Plant density of E. grande and T. atriplicifolia was not a significant predictor of seed set in E. grande. However, there were very high levels of seed predation in E. grande, and that hinders our ability to rule out the any effects of pollinator sharing in seed set. These findings reveal a striking similarity and pollinator sharing between the two species, but despite this, there appears to be no negative or positive effects of this on seed set. Seed set only accounts for half of fitness, however, and the effects of pollinator sharing may therefore be evident on other measures of fitness. The role of trait similarity for pollinator attraction still requires further experimental investigation before the hypothesis of ecological facilitation (or mimicry) can be accepted.