Gene flow, pollen competition, and introgression among cultivated and wild populations of Lobelia cardinalis

To understand the potential loss of local adaptation in native populations due to the influx of genes from non-local populations, I investigated the consequences of introducing non-local plants to a wild population of Lobelia cardinalis. Non-local plants included both horticultural wild types and a cultivated variety. I evaluated the potential for gene dispersal from small plantings into wild, populations, the probability of hybridization, and the consequences of introgression and found that the likelihood of gene movement from ornamental plants to conspecific wild populations is high, and the chances of establishment of hybrids in natural populations are good.;Gene flow distances for Lobelia cardinalis were investigated in both artificial and natural populations. Gene movement declined with distance from the pollen source but pollinators did not discriminate between local and non-local pollen sources. At distances of 1km from a pollen source, half of available flowers were fertilized. Gene dispersal was a common occurrence within a 150m stretch of a continuous population.;I conducted pollen competition experiments to discern whether local pollen had any advantage over non-local on local stigmas. While all populations were able to sire seeds, I found that the identity of the sire population determined the outcome of competition. A neighboring population sired more offspring than expected, while the cultivated variety was not very successful in competition for ovules.;We must understand the performance of hybrids in the field to anticipate the evolutionary consequences of gene flow from ornamental plantings. A large field experiment including hybrids between local and four non-local populations revealed a high likelihood of hybrid establishment. While local and neighboring parental plants survived better than parentals from other populations, F 1 hybrids survived better than local parental types. Hybrid generations of all types reproduced in the field, laying the groundwork for introgression of non-local traits.;A similar study under greenhouse conditions was designed to investigate the genetics of hybrid performance. In controlled conditions, parentals showed little genetic differentiation, as did the resultant hybrids. Either the greenhouse conditions do not aid in detecting epistatic interactions, or the field performance of L. cardinalis is strongly influenced by environmental factors.