The contrasting roles of growth and architectural traits for plant performance in monocultures and mixtures

Plants are limited by relatively few resources, and therefore it is generally assumed that individual traits similarly affect intra- and inter-specific competition. However, species with high productivity in monocultures often are not the most successful in mixtures. Little work has been done to address this mismatch between inter- and intra-specific competitive abilities. In this study, we modeled effects of two groups of clonal plant traits on competitive abilities: i) growth traits (e.g., maximum ramet size), that directly affect productivity; and ii) architectural traits (e.g., branching probability) that control spatial resource allocation and thus affect productivity more indirectly.  For each of two species we parameterized a spatially explicit model and then generated four simulated “phenotypes”, differing only in a single trait that yielded differences in monoculture productivity.  We then examined whether monoculture performance of each phenotype predicted relative performance in two- or four-phenotype mixtures. For two-phenotype mixtures, the more productive phenotype in monoculture was almost always disproportionately more successful in mixture, regardless of trait type.   In four-phenotype mixtures, growth traits also generated the same phenotype ranking as monocultures, with the most productive phenotype always dominating.  However, for architectural traits, the most productive phenotype in monoculture was not necessarily more successful in four-phenotype mixtures.  Consequently, mixtures of four phenotypes that differed only in architectural traits tended to have greater evenness than mixtures of four phenotypes differing in growth traits.  We discuss the implications of these results for long-term coexistence of species that differ in growth vs. architectural traits.