Population ecology of the floodplain herb Macbridea caroliniana (Lamiaceae) with investigations on the species\u27 habitat, breeding system and genetic diversity

The perennial herbaceous mint Macbridea caroliniana is known from 36 locations in discrete watersheds of the Carolinas and Georgia. It is one of relatively few conspicuously flowering herbs that occupy bottomland hardwood forests. The general project goal was to gain knowledge that is applicable to the species\u27 conservation both at the Congaree National Park (CNP) where the largest known population of this species occurs and range-wide. Specific objectives were to (1) quantify the population size and describe the distribution of M. caroliniana within CNP, and determine the extent of co-occurrence with wild hogs and the non-native Murdannia keisak; (2) identify habitat characteristics associated with the presence of M. caroliniana; (3) describe the breeding system; and (4) describe the genetic diversity and structure of the species across the range and within the CNP. Data relevant to the first three objectives were collected at CNP from field plots and experiments conducted in two large of seepage forest herein referred to as ECC and EDB. For the fourth objective, leaves were collected across the species range and allozyme systems were characterized using starch gel electrophoresis. Important findings included the following: CNP has the largest known population of M. caroliniana, with the greatest concentration in the seepage forest ECC, and new CNP locations of M. caroliniana were found; hog rooting activity negatively affects M. caroliniana patches in the short-term, but the long-term threat is unknown; the invasive plant, Murdannia keisak, is a frequent co-occurring plant, but is not a clear threat to the study species; in a patch-scale habitat study, the best model tested to predict the presence of M. caroliniana included the variables (+) herb richness, and two soil nutrients, (+) phosphorus and (-) potassium; in a forest-scale study, the variables above were not statistically different between the two forested areas, but they were in the direction predicted, that is, ECC with a larger M. caroliniana population has greater herb richness, more phosphorus and less potassium and this may explain in part why there is less M. caroliniana at EDB; M. caroliniana is not autogamous, but it is self-compatible, dependant on pollinators to set fruit, and likely pollen-limited; floral visitation was infrequent, but the most common floral visitors were Poanes zabulon and Bombus impatiens; in the species-wide genetic study, the genetic structure of the species is greatly influenced by river basin; the ECC and EDB populations ranked highest for conservation priority based on genetic measures, further emphasizing their importance to the species; at ECC, gap patches are more like each other genetically than are the patches from closed canopies and this suggests more gene flow between the gap patches likely from floral visitors. Conservation implications based on this research may apply to other perennial, herbaceous, insect-pollinated species that occupy naturally fragmented or disjunct wetland habitats. Additional information needed to improve conservation efforts include an understanding of the relationship of M. caroliniana to canopy gap dynamics, of its ability to compete for scarce resources, and of demographic patterns and processes