John J Schenk1. 1. Department of Biological Science, Florida State University, Tallahassee, Florida 32306-4295, USA.
Abstract
PREMISE OF THE STUDY: Dispersal is a major feature of plant evolution that has many advantages but is not always favored. Wide dispersal, for example, leads to greater seed loss in oceanic-island endemics, and evolution has favored morphologies that limit dispersal. I tested the hypothesis that selection favored limited dispersal on gypsum islands in western North America, where edaphic communities are sparsely vegetated except for a specialized flora that competes poorly with the surrounding flora. • METHODS: I applied a series of comparative phylogenetic approaches to gypsophilic species of Mentzelia section Bartonia (Loasaceae) to investigate the evolution of limited dispersal function in seed wings, which increase primary dispersal by wind. Through these tests, I determined whether narrowed wings were selected for in gypsophilic species. • KEY RESULTS: Gypsophily was derived four to seven times. Seed area was not significantly correlated with gypsophily or wing area. Wing area was significantly smaller in the derived gypsum endemics, supporting the hypothesis in favor of limited dispersal function. A model-fitting approach identified two trait optima in wing area, with gypsum endemics having a lower optimum. • CONCLUSIONS: Evolution into novel ecologies influences morphological evolution. Morphological characters have been selected for limited dispersal following evolution onto gypsum islands. Selection for limited dispersal ability has occurred across animals and plants, both in oceanic and terrestrial systems, which suggests that reduced dispersal ability may be a general process: selection favors limited dispersal if the difference in survival between the habitat of the parent and the surrounding area is great enough.
PREMISE OF THE STUDY: Dispersal is a major feature of plant evolution that has many advantages but is not always favored. Wide dispersal, for example, leads to greater seed loss in oceanic-island endemics, and evolution has favored morphologies that limit dispersal. I tested the hypothesis that selection favored limited dispersal on gypsum islands in western North America, where edaphic communities are sparsely vegetated except for a specialized flora that competes poorly with the surrounding flora. • METHODS: I applied a series of comparative phylogenetic approaches to gypsophilic species of Mentzelia section Bartonia (Loasaceae) to investigate the evolution of limited dispersal function in seed wings, which increase primary dispersal by wind. Through these tests, I determined whether narrowed wings were selected for in gypsophilic species. • KEY RESULTS: Gypsophily was derived four to seven times. Seed area was not significantly correlated with gypsophily or wing area. Wing area was significantly smaller in the derived gypsum endemics, supporting the hypothesis in favor of limited dispersal function. A model-fitting approach identified two trait optima in wing area, with gypsum endemics having a lower optimum. • CONCLUSIONS: Evolution into novel ecologies influences morphological evolution. Morphological characters have been selected for limited dispersal following evolution onto gypsum islands. Selection for limited dispersal ability has occurred across animals and plants, both in oceanic and terrestrial systems, which suggests that reduced dispersal ability may be a general process: selection favors limited dispersal if the difference in survival between the habitat of the parent and the surrounding area is great enough.