UNLABELLED: • PREMISE OF THE STUDY: An association between polyploidy and gender dimorphism has been noted in several plant lineages. Whereas the majority of Lycium species are diploid and have hermaphroditic flowers in cosexual populations, gender dimorphism (gynodioecy, dioecy) has been shown to be uniformly associated with polyploidy in previous studies. Preliminary field observations suggested that some populations of Lycium carolinianum were dimorphic, providing a test of this association.• METHODS: We assessed sexual systems and cytotype variation (to infer ploidy) across 17 populations of L. carolinianum. Comparison of flowers in cosexual and dimorphic populations were used to infer changes in reproductive morphology associated with the evolution of gynodioecy.• KEY RESULTS: The majority of populations were cosexual in gender expression, but dimorphism was present in the Yucatán and in some populations in Hawaii. Populations varied in ploidy and were either diploid or tetraploid. Floral sexual dimorphism was present in all gynodioecious populations, though the magnitude differed and was cryptic in some cases. Our results are consistent with the hypothesis that following the evolution of gynodioecy, flowers on hermaphrodites increased in size.• CONCLUSIONS: Dimorphic sexual systems have likely evolved convergently in L. carolinianum. In contrast to previous studies, dimorphism is not perfectly associated with polyploidy. Although our sample from the Yucatán was both tetraploid and dimorphic, all populations in Hawaii were diploid regardless of sexual system. Ongoing phylogeographic and mating system studies will contribute to our understanding of reproductive evolution in this widespread, polymorphic species.
UNLABELLED: • PREMISE OF THE STUDY: An association between polyploidy and gender dimorphism has been noted in several plant lineages. Whereas the majority of Lycium species are diploid and have hermaphroditic flowers in cosexual populations, gender dimorphism (gynodioecy, dioecy) has been shown to be uniformly associated with polyploidy in previous studies. Preliminary field observations suggested that some populations of Lycium carolinianum were dimorphic, providing a test of this association.• METHODS: We assessed sexual systems and cytotype variation (to infer ploidy) across 17 populations of L. carolinianum. Comparison of flowers in cosexual and dimorphic populations were used to infer changes in reproductive morphology associated with the evolution of gynodioecy.• KEY RESULTS: The majority of populations were cosexual in gender expression, but dimorphism was present in the Yucatán and in some populations in Hawaii. Populations varied in ploidy and were either diploid or tetraploid. Floral sexual dimorphism was present in all gynodioecious populations, though the magnitude differed and was cryptic in some cases. Our results are consistent with the hypothesis that following the evolution of gynodioecy, flowers on hermaphrodites increased in size.• CONCLUSIONS: Dimorphic sexual systems have likely evolved convergently in L. carolinianum. In contrast to previous studies, dimorphism is not perfectly associated with polyploidy. Although our sample from the Yucatán was both tetraploid and dimorphic, all populations in Hawaii were diploid regardless of sexual system. Ongoing phylogeographic and mating system studies will contribute to our understanding of reproductive evolution in this widespread, polymorphic species.