COS 88-2 - Evidence for mate selection in a hummingbird pollinated plant, Heliconia tortuosa

Background/Question/Methods

A plant’s mating system has major consequences for the ecology and evolution of plants, influencing genetic structure and diversity, effective population size, and offspring fitness, as well as metapopulation dynamics and possibly the rate and trajectory of evolution. The mating success of an individual generates a mating portfolio that tells a compelling story. At our research site in Costa Rica, pollen dye experiments of Heliconia tortuosa show that there is frequent pollen transfer to nearby plants, and that pollen is widely transferred within a small forest patch. However, paternity analysis from three small forest patches in the same study system indicates that very few seeds are sired by mates from within the same patch. To investigate this discrepancy, we evaluated if, for a given pair of near-neighbour plants that successfully mated, whether the father is, on average, less related to the maternal plant, than a random sample of the available local near-neighbours. In each of the 18 forest patches in Coto Brus, Costa Rica, we sampled leaf tissue and seeds from a group of 4-5 near-neighbour plants, for a total of 71 H. tortuosa maternal plants and 814 seeds; we then genotyped all mothers and seeds at 11 microsatellite loci.

Results/Conclusions

Paternity analysis identified 10 local mating events, where an outcrossed seed was sired by one of the 3 - 4 genotyped near-neighbour plants. Paired t-tests for three measures of relatedness, the proportion of shared alleles and two kinship coefficients: the Loiselle Estimator and the Ritland Estimator, showed that these successful near-neighbour mates, on average, were significantly less related to the maternal plant than the average of the (available) genotyped near-neighbour plants. This result suggests the presence of a selection mechanism against inbred pollen. Field experiments will be needed to establish this putative mate selection process and to test whether its strength varies with forest fragmentation, and thus, whether this plant's mating system is adapting to changes in pollination related to landscape change.