Causes of rarity in glade-endemic plants: Implications for responses to climate change

This dissertation documents the relationship between stress-adaptation and reproductive specialization in three endemic plant species (Delphinium treleasei, Echinacea paradoxa, and Scutellaria bushii) that are locally-abundant but restricted to glade habitats and their closely-related congeners (D. carolinianum, E. pallida, S. ovata, and S. parvula) that have broader habitat use, including glades.;In Chapter Two, I test two hypotheses concerning reproductive specialization in endemic plants and their common congeners (CC). I quantify morphological traits that are associated with stress-adaptation and reproductive specialization, pollinator behavior, plant specialization, and rates of autogamous self-fertilization. The locally-abundant, regionally-rare (LARR) species were significantly different from their CCs in vegetative traits associated with stress-adaptation (i.e., stem length and leaf area), but the direction of the differences was not consistent among genera or with predictions of stress-adaptation. On the other hand, two of the three LARR taxa had larger flowers and fewer dominant pollinators than their CCs, though there were no differences in autogamous selfing rates. This study supports a more nuanced theory of the relationship between reproductive specialization and rarity that addresses additional factors influencing rare taxa, such as stress-adaptation.;In Chapter 3, I examine the responses of the LARR and CC plants to drought and high heat conditions in order to explicitly test the prediction that the LARR species are more resistant to high-stress environments and allocate resources to fewer, higher-quality offspring in comparison to their widespread congeners.I exposed plants of both species in a congeneric pair (D. treleasei, D. carolinianum, E. paradoxa, E. pallida, S. bushii, and S. ovata) to experimental manipulations of water availability and temperature regimes that were consistent with those experienced in the glade habitat. The LARR species were more resistant to stress is some morphological traits that are associated with adaptation to stressful environments, and the CC species were not. Moreover, the LARR species had fewer, higher-quality offspring, whereas the CC species have more seeds of lower quality. This study indicates that plants that specialize on stressful environments differ from their CCs in morphological traits associated with stress-adaptation and in the allocation of resources to reproductive output, with implications for their interactions with pollinators.;I then test the alternative hypotheses that two LARR species are (1) poorer competitors for pollinators, as predicted by traditional theory of reproduction in rare species, (2) are better competitors for pollinators in stressful environments in comparison to a common congener, or (3) do not compete with their common congeners for pollination services (Chapter 4). I conducted pollinator competition trials at multiple localities, in order to control for spatial variation in pollinator assemblage. Naive plants were exposed to pollinators in two treatment arrays, either with conspecifics or heterospecifics, and constant density. I compared the pollination and reproductive success of the CC and LARR species in mixed groups and in monoculture. The congeneric pairs did not compete for pollinators but varied in their morphological traits and reproductive success across sites. The results of this study indicates that spatial variation in plant-pollinator interactions dominates pollination success in years with low pollinator abundance.;Finally, this dissertation explores the response of two LARR species and their CCs to regional climate change in order to test for the relative importance of intrinsic and extrinsic factors in determining species phenological patterns. Species responses to climate change have been well documented, but there is significant variation in responses both across and within taxa. I test the prediction that LARR species that specialize on stressful habitats are less responsive to climate change in comparison to their CC, since they are resistant to changes in abiotic conditions (intrinsic factors). This study is the first to document phenological responses to climate change by describing the development stage of individuals in time, which is a more biologically realistic estimate of phenology. (Abstract shortened by UMI.)...