The evolution of species' geographical range limits: an empirical test using two coastal dune plants, Camissonia cheiranthifolia (Onagraceae) and Abronia umbellata (Nyctaginaceae)

Abstract

Every species has a limit to its geographic distribution. The problem is that we don’t really know why. Classical explanations propose that range limits occur where the ecological conditions overwhelm adaptation and populations are no longer self-sustaining. It is generally expected that population abundance and fitness decline towards range margins across a gradient of declining habitat quality. Current evolutionary and theoretical explanations of range limits predict that this geographic pattern of demography will result in genetic constraints in marginal populations, such that range expansion is thwarted by reduced evolutionary potential. In this thesis, I tested the key assumptions and predictions of range limit theory through an empirical evaluation of two coastal dune, endemic plants; Camissonia cheiranthifolia (Spreng.) Raim. (Onagraceae) and Abronia umbellata L. (Nyctaginaceae). In geographic wide surveys of a large proportion of populations across both species’ ranges, neither species exhibited declining abundance or performance towards its range limits. Central populations of C. cheiranthifolia tended to have a higher production of seeds per unit area than marginal populations. Although this pattern demonstrated the potential for gene flow from central sites to swamp selection in marginal sites, results from a transplant experiment suggested that this was unlikely. Experimental populations of C. cheiranthifolia originating from ≥ 675 km south of the northern range limit exhibited similar levels of fitness when planted at the range margin. Along a 200 km transect across the limit, and in contrast to expectations, fitness of all populations increased towards the limit and generally remained high beyond the limit. Individuals from all populations reproduced and matured fruit beyond the limit, suggesting that if individuals dispersed beyond the limit that populations would establish. The species’ abrupt distributional limit also did not correspond to an abrupt shift in ecological conditions, despite the association of fitness with plant community and microhabitat variables. Overall, ecological tolerances to fitness or niche-based explanations to range limits do not adequately describe the distributions of either species. Constraints on dispersal rate, the influence of anthropogenic factors on habitat dispersion and limited genetic variability for fitness related traits are addressed.