Ecological genetics of co-occurring understory grasses in ponderosa pine forests of northern Arizona

To understand how selective and non-selective forces affect genetic variation of plant species, I studied genetic variation of two understory grass species, Koeleria macrantha (Ledeb.) Schult. (Junegrass) and Elymus elymoides (Raf.) Swezey ssp. elymoides (squirreltail) in ponderosa pine (Pinus ponderosa var. scopulorum Dougl. ex Laws.) forests of northern Arizona at three levels: (1) allozyme variation; (2) ecophysiological variation; and (3) quantitative genetics. Chapter Two presented results of the allozyme variation in the two grass species and ponderosa pine, revealing higher genetic differentiation and higher gene-environment correlation in the inbreeding grass, squirreltail. This grass species also showed genetic association with ponderosa pine genotypes. As most allozyme variation was neutral in the three species, some marker variation may be adaptive. Higher linkage among genes in inbreeding species may cause higher correlation between genotypes and environment Chapter Three investigated genetic variation in ecophysiological and survival responses to drought in the two grass species. Higher net photosynthetic rate was found in low elevation populations in both grass species, and higher stomatal conductance was found in low elevation populations of squirreltail. This Chapter also showed longer survival time in low elevation Junegrass but not in squirreltail. The differences in gas exchange traits and survival time between elevations in the two species can be explained by their different leaf traits such as leaf area, suggesting that divergent selection along an elevation and water availability gradient appears to have caused genetic differentiation. The Fourth chapter introduced a genetic model to compare population structure in quantitative traits and allozyme markers. Using this model, I found different evolutionary forces were shaping population divergence of the two grass species. However, the interpretation of this model needs to consider other factors such as mating system of species. Taken together, the results suggest that morphological and physiological traits are more likely to be affected by selection and more informative for restoration than genetic markers.