We examined the genetic composition, habitat use, and morphological variation of a Phoxinus eos-neogaeus unisexual hybrid complex and its sexually reproducing progenitor species inhabiting beaver-modified drainages of Voyageurs National Park, Minnesota. In addition to the single diploid P. eos-neogaeus gynogenetic clone, triploid and diploid-triploid mosaic biotypes were present at our study sites. Both P. eos and P. neogaeus, and all three hybrid biotypes were ubiquitous throughout one intensively surveyed drainage, but abundances and relative frequencies of the parental species and hybrids varied considerably within and among successional environments. Data from a large number of additional sites indicated that the proportion of polyploid hybrids within an environment was negatively related to hybrid relative frequency, implying that the genomic constitution of hybrids is an important determinant of clonal fitness among successional environments. Statistical comparisons of variation along size-free multivariate body shape axes indicated that despite its genetic uniformity, the P. eos-neogaeus clone is no less variable than its sexual progenitors, suggesting that a single genotype may actually respond to environmental variation with as much phenotypic variation as a genetically variable sexual population. The incorporation and expression of a third genome in triploid and diploid-triploid mosaic biotypes derived from the gynogenetic clone significantly expanded phenotypic variation of the clone. This additional variation results in greater similarities in habitat use and morphological overlap with the parental species, primarily P. eos, the predominant sperm donor for gynogenetic hybrid females in this complex. Polyploid augmentation of a diploid gynogenetic clone appears to be typical in the P. eos-neogaeus complex, and the additional genetic and phenotypic variation that it generates has potentially significant ecological and evolutionary consequences for the success and persistence of a single genotype in highly variable environments.