Thursday, November 16, 2006 - 9:15 AM
317-4

Molecular markers in the management of the invasive species cogongrass (Imperata cylindrica ).

Ludovic Capo-Chichi1, Edzard van Santen2, and Paul Raymer1. (1) Univ of Georgia, Crop and Soil Science Department, 1109 Experiment St, Griffin, GA 30223-1797, (2) Auburn University, 202 Funchess Hall, Auburn, AL 36849

Molecular markers contribute to our understanding of the entry, spread, and evolution of invasive species as a foundation upon which appropriate management strategies are developed.  Cogongrass (Imperata cylindrica L. Beauv.), introduced into the United States shortly after the turn of the twentieth century, is an aggressive, invasive weed species that colonizes a diverse range of habitats.  Genotypic diversity and spread dynamics at the point of initial introduction and its adjacent areas in the southern United States were estimated.  Clones evaluated from nine populations with two primer combinations produced a total of 137 AFLP loci of which 102 (74.4%) were polymorphic.  Among population variance accounted for 44% of the total variance.  The species exhibits no genetic structure inferred from 24 populations along putative routes of dispersal.  Only 6.9% (P > 0.05) of the total variance was partitioned between eastern and western areas of the State of Alabama.  Within population variance component was estimated to be 35.8 % (P < 0.05) of the total variation while among populations gathered  57.2% (P < 0.05), suggesting that asexual reproduction through rhizomes and sexual reproduction by seed significantly impact the genetic structure of I. cylindrica populations.  Gene flow (Nm) values ranged from 0.6 to 5.7.  The absence of significant relationship between geographic and genetic distances suggests that the spread dynamics of I. cylindrica into the southern United States is primary through human activities and to a lesser extent by natural dispersal mechanisms.  High levels of genetic differentiation were found between populations regardless of the geographic distance that separate them.  When assessed by Mantel's tests, there is no correlation between gene flow and geographic distance.