The integrity of wetland ecosystems is largely determined by hydrological functionality, degree of connectivity to like ecosystems, and permeability to external influence. Land use changes in upland areas adjacent to wetland ecosystems may influence hydrology and connectivity while introducing novel biotic and abiotic materials. There is an increasing trend toward the use of remote assessment techniques to determine the degree of impact of external influences on adjacent wetlands. Remote assessment and predictive capabilities are provided by indices such as the Landscape Development Intensity Index (LDI) (Brown and Vivas 2005) which may be beneficial in determining site condition, and which have the added benefit of providing a quantitative gradient of human impact. This study assessed the predictive ability of the LDI in cypress ecosystems, by testing its correlations with plant community metrics including an index of floral quality calculated using coefficients of conservatism (CC)(Cohen et al. 2004), plant species diversity, and fluctuation in community composition assessed by changes in the wetland status and native status of component plant species. LDI was also compared against an independent measure of disturbance which was used to construct an a priori disturbance gradient. Overall, diversity measures showed little correlation with any of the disturbance indices, while CC scores were significantly correlated. Models were constructed in an attempt to explain each of the variables of plant community response to development in the surrounding landscape. The length of time since the development of the land adjacent to the cypress domes was a predictor of plant community response only when included in models with other variables. LDI was the strongest predictor in all models except where increases in land use associated with hydrological changes helped predict or better predicted proportions of exotic and upland species.

In this thesis, I combined field, common-garden and greenhouse experiments to examine the ecological and evolutionary consequences of plant-herbivore interactions in the perennial herb Lythrum salicaria. More specifically I examined (1) whether resistance and tolerance to damage from herbivores vary with latitude and are positively related to the intensity of herbivory in natural populations, (2) whether effects of herbivory on plant fitness vary with latitude, (3) whether populations are locally adapted and whether herbivory influences the relative fitness of populations, and (4) whether the intensity and effects of insect herbivory on reproductive output vary locally along a disturbance gradient and are associated with differences in plant resistance. A common-garden and a greenhouse experiment demonstrated that plant resistance decreased whereas plant tolerance increased with latitude of origin among populations sampled along a latitudinal gradient in Sweden. Oviposition and feeding preference in the greenhouse and leaf damage in the common-garden experiment were negatively related to natural damage in the source populations. Experimental removal of insect herbivores in three populations sampled along the latitudinal gradient demonstrated that intensity of herbivory and its effects on plant fitness decreased towards the north. A reciprocal transplant experiment among the same three populations showed that herbivory affected the relative fitness of the three populations, but did not detect any evidence of local adaptation. Instead the southernmost population had the highest relative fitness at all three sites. A herbivore-removal experiment conducted in nine populations in an archipelago in northern Sweden demonstrated that insect herbivory strongly influenced among-population variation in reproductive output. However, variation in resistance was not related to differences in intensity of herbivory at this spatial scale. Taken together, the results demonstrate that resistance and tolerance to herbivory vary with latitude but in opposite directions, that intensity of herbivory is a major determinant of flowering and seed output, and that the strength of herbivore-mediated selection varies among populations in Lythrum salicaria. They further indicate that both physical disturbance regime and latitudinal variation in abiotic conditions may strongly influence the performance and abundance of perennial herbs because of their effects on interactions with specialized herbivores.

No description available.