Environmental niche differences may enable the establishment and maintenance of polyploids by facilitating geographic separation from their diploid progenitors. Most studies of diploids and their polyploid derivatives detect environmental differentiation, although the nature of those differences is not consistent across studies. Whether the degree and nature of differentiation, and changes in niche breadth and optimum, are predictable remains controversial. If they are predictable, then they should be apparent in polyploid derivatives of closely related species in similar environments. We characterized environmental niche differences between diploids and polyploids in three closely related species of Phlox that grow in mountainous regions of southwestern North America: P. amabilis, P. nana, and P. woodhousei. Each species has three cytotypes, diploids, tetraploids, and hexaploids, although in P. woodhousei there are too few hexaploid populations to model. Ploidy was determined by flow cytometry linked to chromosome counts, and environmental differentiation was studied using a PCA approach to model niches in environmental space (PCA-env). We tested cytotypes for differences in niche equivalency, niche optimum, and niche breadth in environmental space.
Results/Conclusions
Tetraploid niches were distinct from diploid niches for P. nana and P. woodhousei, but not for P. amabilis. Hexaploid niches were distinct from diploid niches for P. amabilis but not for P. nana. Niche breadth and optimum were different in every comparison made. In P. amabilis and P. nana, most polyploids had narrower niches on both principal component (PC) axes but tetraploids of P. nana had a broader niche than diploids on PC1. Likewise, P. woodhousei tetraploids had a broader niche than diploids. The analyses of niche optimum on PC1 suggest that polyploids of P. amabilis and P. woodhousei occur in wetter environments but polyploids of P. nana occur in drier habitats. Analyses of niche optima on PC2 suggest that polyploids typically occur in less continental climates than diploids—only the hexaploid cytotype of P. amabilis is an exception. In summary, we found measurable differences in environmental niches between diploids and polyploids in all three species, suggesting that environmental differentiation plays a role in polyploid establishment. However, we found little evidence for a predictable effect of polyploidy, even in closely related species occurring in similar habitats, limiting our ability to generalize about the effects of polyploid differentiation.