Competitive ability: Definitions, contingency and correlated traits

Although the relationship between individual plant traits and competitive success in communities is an essential component of comprehensive models of the role of competition in structuring plant communities, three obstacles have stymied efforts to empirically examine such relationships. First, definitions of competitive ability are often inconsistent among bodies of theory and between theoretical predictions and empirical research. Much of the theoretical literature is for populations and often at equilibrium, while experimental work has been largely on individuals and short term. This situation is likely to continue, except for a few model systems, and therefore it is critical that individual-level surrogates for population level phenomena be found. I suggest that competitive response of seedlings to established vegetation may be an effective surrogate for estimating competitive success of populations at equilibrium and that competitive response of individuals with more similar-sized neighbours may be an effective surrogate for competitive success of populations earlier in succession or in non-equilibrium systems. Second, competitive ability may be contingent on many factors, such that it may not be an identifiable characteristic of any particular taxon and thus no broadly applicable relationships between traits and competitive ability may exist. However, a literature survey shows that both competitive response and competitive effect are generally, but not always, consistent regardless of identity of competing species, making the search for relationships with traits reasonable, at least within environments. Among environments, both competitive effect and competitive response are consistent in only about half the studies, making it unreasonable to assume a priori that competitive hierarchies will be similar under different conditions. The third obstacle is logistical; competitive ability is necessarily measured experimentally, and preferably in the field, making it difficult to obtain sufficient sample sizes (numbers of taxa) for rigorous analysis of relationships with traits. I suggest several simplifying assumptions and experimental approaches that could enable much more efficient assaying of competitive abilities of many species.