To understand the extraordinary diversity and complexity of organisms requires comprehensive studies of the mechanisms by which populations diverge to become different species. The importance of ecology in speciation has long been recognized; however, there is little direct evidence that ecological factors are the principal isolating barriers between closely related species. To examine how adaptation to ecological factors may contribute to reproductive isolation and lead to speciation, I studied the isolating mechanisms between two closely related Neotropical herbs, Costus allenii and C. villosissimus. These perennial species occur in the same geographic region, but occupy distinct habitats. Costus allenii is located along shady ravines in mature forests, while C. villosissimus is found in drier, open sites along forest edges. Both species are pollinated by euglossine bees and can be crossed to produce fully fertile hybrids. I conducted field studies in central Panama, where the two species co-occur, to quantify the strength of multiple isolating barriers and total reproductive isolation. I also conducted reciprocal transplant experiments in the field that were coupled with greenhouse studies to determine whether the two species are locally adapted to ecological factors in their respective habitats, and to identify putative adaptive traits that may contribute to local adaptation and speciation.There are four chapters in my dissertation. Chapter 1 discusses the mechanisms contributing to sexual isolation in these species. I found that pollinator-mediated barriers and gametic isolation restrict heterospecific gene flow asymmetrically between C. allenii and C. villosissimus. Chapter 2 describes the parapatric distribution of the two species along a water availability gradient. The results of reciprocal transplant experiments at two life stages, seeds and juveniles, suggest that local adaptation contributes to strong microhabitat isolation and asymmetrical extrinsic postzygotic isolation between C. allenii and C. villosissimus. As habitat isolation has been found to be strong in this system, the environmental factors contributing to this form of isolation are of great interest. Chapter 3 summarizes comparisons of the two parental habitats and shows differences in putative adaptations to these habitats among C. allenii, C. villosissimus, and their F1 hybrids in the greenhouse. Higher leaf mass per area was found in C. allenii, which occupies habitats with lower light availability, while higher drought tolerance was found in C. villosissimus, which occupies habitats with lower soil moisture. The F1 hybrids had leaf mass per area similar to that of C. villosissimus, although hybrid fitness was not reduced in C. allenii habitats compared to pure C. allenii transplants. The F1 hybrids had intermediate drought tolerance, which is consistent with their lower seedling survival in C. villosissimus habitats. Chapter 4 presents an examination of multiple isolating barriers and comparisons of their relative contribution to speciation in C. allenii and C. villosissimus. Total reproductive isolation was found to be high between the two species, and the major isolating barriers are ecogeographic and microhabitat isolation. Because ecogeographic isolation represents spatial isolation based on genetic differences between species, presumably due to local adaptation, and microhabitat isolation is found to be a consequence of local adaptation (Chapter 2), I conclude that local adaptation is the primary mechanism of speciation in C. allenii and C. villosissimus.In this research, I used a comprehensive approach to the study of speciation by linking adaptation to the origin of reproductive isolation. My dissertation provides empirical evidence for how local adaptation to different environmental conditions contributes to reproductive isolation and lead to speciation.
Read
