Background/Question/Methods Variable species interactions and environmental heterogeneity are often invoked as explanations for coexistence of plant species. Yet, even closely related species may coexist within communities at small spatial scales. Because of their shared ancestry, related species are similar in their phenotype and ecology, and are expected to compete for shared resources. However, evidence for the pattern that species growing with close relatives (e.g. congeners) perform worse than when in mixtures with more distant relatives is lacking, particularly in a multi-species, community context. I tested this prediction experimentally with seven species of perennial, herbaceous Asteraceae that frequently co-occur in old-field communities of northeastern North America: three goldenrods (Solidago), three asters (Symphyotrichum), and the related Euthamia graminifolia. I constructed field mesocosms containing confined, equal-density (21 individuals) mixtures of three types: goldenrods only, asters only, and a confamilial mixture of goldenrods and asters; in all cases Euthamia was included as a phytometer species. The mixtures were protected from vertebrate and insect herbivory, and fully crossed with a watering treatment (with or without water supplementation) to test for the effects of environmental (drought) stress on species interactions. In mid autumn, I harvested the senescing above-ground biomass as a measure of plant performance.

Results/Conclusions On average, both goldenrods and asters reached similar average biomass in their respective congeneric and confamilial mixtures, not supporting the prediction of more intense competition when grown with close relatives than in more phylogenetically diverse mixtures. Surprisingly, total above-ground biomass was nearly twice as high in the aster mixtures as in the mixtures with goldenrod (old-field dominant species), in part due to species-specific growth-form and phenological differences. Euthamia performed equally well in all mixtures, but its biomass was more variable within the confamilial mesocosms. Two of the goldenrod species with only a small proportion of plants that flowered by the end of the season were more likely to flower in their congeneric mixtures. The watering treatment had only minor effects on the performance of the mixtures or its component species. These results suggest that initial idiosyncratic responses of species may over-ride the overall effects of relatedness on competition and community-level outcomes even at high densities, and highlight the need to include long-term interactions among these clonal perennials to explain their patterns of coexistence.