|Capo-Chichi, L - AUBURN|
|Williamson, A - AUBURN|
|Patterson, M - AUBURN|
|Miller, J - USDA FOREST SERVICE|
|Van Santen, Edzard - AUBURN|
Submitted to: Southeastern Exotic Pest Plant Council
Publication Type: Abstract Only
Publication Acceptance Date: May 3, 2005
Publication Date: June 3, 2005
Citation: - Interpretive Summary: none required.
Technical Abstract: The genetic composition of an introduced, invasive species is influenced by its introduction as well as its life history characteristics (4). The initial genetic structure of a successful invasive population depends on several factors, including the effective population size of the introduction event(s), the genetic diversity of the source population(s) and the number of founding sources (6). Cogongrass (Imperata cylindrica (L.) Beauv.), introduced into the United States shortly after the turn of the twentieth century, is an aggressive, rhizomatous, invasive species that is generally considered a pernicious pest plant due to its ability to successfully disperse, colonize, spread, and subsequently compete with and displace desirable vegetation and disrupt ecosystems over a wide range of environmental conditions (5, 1, 2, 3). This species propagates itself sexually by seed and asexually by rhizomes. To understand the potential for colonizing and establishment of I. cylindrica, we used amplified fragment length polymorphic (AFLP) markers to evaluate genotypic diversity around the point of initial introduction in the southeastern United States. Among 9 populations, two AFLP primer pairs generated a total of 137 amplification products, of which 102 (74.4%) were polymorphic. Canonical discriminant analysis divided the nine populations into three major groups with distinct subgroups. The population collected from the original site of introduction formed a group in itself that was most distinct from the other eight populations. Cluster analysis separated all individuals into three mains clusters. It has to be remembered that the objectives of cluster analysis and canonical discriminant analysis are different. The former assesses the relationship based on the banding profile of each individual ramet and assigns these individuals to artificial (arbitrary) clusters, whereas the latter maximizes differences among existing groups (classes) based on the common profile of members of a group. Analysis of molecular variance (AMOVA) revealed that most of the genetic variation resided within populations (56%) while among populations component was 44% of the total genetic variation. Total gene diversity (HT) across all populations was estimated to be 0.18 and within population gene diversity ranged from 0.07 (population sampled closest to the point of introduction) to 0.16. Intra-population genetic diversity, number of observed and effective alleles in the initial population was observed to be lower than those of the population of adjacent areas. Gene flow (Nm), inferred from '-statistics, describing the genetic differentiation between pairs of populations ranged from 0.26 to 5.04. The genetic component of populations and the geographical pattern of AFLPs markers for I. cylindrica in its introduced range support conclusions drawn from historical evidence that the invasion of I. cylindrica into the southern United States arose from multiple introductions of genetically different materials. We found no significant relationship between gene flow and geographic distance or any correlation between genetic and geographic distances. This almost certainly will change as we expand our sampling range to include locations from central Florida to western Louisiana. We suggest that anthropogenic dispersal is one of the most powerful agents for local dispersal of I. cylindrica in the southern United States.