THE GENETIC BASIS OF THE ECOLOGICAL AMPLITUDE OF SPARTINA PATENS. II. VARIANCE AND CORRELATION ANALYSIS

Reciprocal transplantations of Spartina patens genotypes from adjacent salt marsh, swale, and dune habitats provided evidence for genetic differentiation among subpopulations, due at least in part to contrasting selection regimes. Genet survival in the different habitats was related to the amount of genetic divergence. In the dune habitat, marsh ramets showed the lowest survival, swale ramets showed intermediate survival, and dune ramets showed the highest survival. This relationship was not reciprocal, however. The marsh habitat afforded an environment where survival was maximal for all genotypes. Thus, by comparison, the dune environment appeared to impose a more intense selection pressure, and the swale an intermediate selection pressure on Spartina patens. In each site resident genotypes tended to show greater relative fitness than aliens. This evidence for genetic divergence corroborates that previously reported on morphometric (Silander and Antonovics, 1979) and allozymic traits (Silander, 1984). High levels of phenotypic plasticity may permit greater adaptation to the spatially and temporally heterogeneous environment occupied by S. patens than would genetic variation alone. Dune and swale genets were more phenotypically plastic across traits examined than were marsh genotypes. The higher plasticity in these peripheral subpopulations may confer increased fitness among residents and compensate for observed declines in genetic variation. A slight decrease in genetic variability was evident from marsh to dune subpopulations. However, since the differences in genetic variation among subpopulations were small, and disparities did occur, it is unlikely that evolutionary divergence is retarded primarily by a lack of genetic variability in the characters considered. Evidence is presented to indicate that evolutionary divergence among subpopulations may be retarded by negative or unfavorable correlations among characters being selected simultaneously. These negative correlations may increase extinction probabilities in small peripheral populations, such as those represented by the dune or swale, and are likely to lower fitness. Based on these observations, I hypothesize that further microevolution may be retarded in peripheral dune and swale subpopulations, primarily by unfavorable genetic correlation structures among fitness components or characters under simultaneous selection. Contributing factors may include lowered genetic variance and higher levels of phenotypic plasticity.