Does autopolyploidy contribute to range size in the giant goldenrod Solidago gigantea?

Abstract

Autopolyploidy is an under-recognized evolutionary phenomenon in angiosperms, creating cryptic patterns of reproductive isolation and phenotypic divergence within a single species. Abundant evidence shows that autopolyploid cytotypes can exhibit different phenotypes, which could lead to subtle niche differentiation. In this way a species comprising an autopolyploid series could exhibit a broad geographic distribution that encompasses vast abiotic/biotic differences due to the collective ranges of its cytotypes. We aim to test this hypothesis in the giant goldenrod Solidago gigantea (Ait), an abundant plant species found throughout much of eastern North America. Diploid (2n=18), tetraploid (2n=36), and hexaploid (2n=54) cytotypes are known, and previous studies suggest that they are non-randomly distributed across the species' range. Previous work also suggests that these cytotypes are morphologically distinguishable based on leaf vestiture and dimensions. We evaluate these claims by combining two datasets: genotype-based estimates of the cytotype of herbarium specimens and a large set of previous chromosome counts. Together these provided a larger sample size, and one spanning the extensive range of S. gigantea. Our results show that S. gigantea cytotypes are non-randomly distributed and that their abiotic niches are not equivalent. This suggests that genome doubling and/or subsequent natural selection confer greater ecological amplitude on S. gigantea polyploid cytotypes. This pattern was more evident in the tetraploid to hexaploid transition. Finally, while morphological trends are apparent, leaf vestiture and dimensions are not sufficient to consistently discriminate cytotypes.