On the Relative Importance of Competition in Unproductive Environments

I Whether or not competition intensity increases or is similar along productivity gradients has been highly controversial for a number of years, but empirical results bearing on this question are quite variable and no consistent answer has yet emerged. We have developed a more general hypothesis that includes these contradictory predictions as special cases that apply under different types of resource dynamics and different types of interactions between the growth and survival components of fitness. 2 The two-phase resource dynamics hypothesis of plant interactions along productivity gradients is based on the fact that soil resources are usually supplied in pulses rather than continuously, as assumed by most formal theory for community dynamics of terrestrial plants. When soil resource supply is temporally variable, individual plants will experience two distinct phases of resource availability: pulse periods when resources are high and most growth and resource accumulation occurs, and interpulse periods when resources are too low for most plants to take up and most mortality due to resource deficits takes place. 3 Competitive effects on growth should occur during pulses at both high and low productivity. In productive environments, interpulse intervals should be relatively mild and infrequent and therefore competitive effects during pulses will usually be important for individual and population persistence. However, as productivity decreases, the frequency of pulses (as well as or in place of their magnitude) often decreases and the duration of interpulse periods increases. Therefore, we suggest that processes occurring during interpulse intervals become increasingly important for individual and population persistence as interpulse intervals become longer. Whether or not competition occurs under low productivity will then depend on (i) the extent to which the asymptotically low resource availability during interpulse periods is determined by plant uptake or by abiotic factors such as leaching, drainage, evaporation and volatization, and (ii) the extent to which decreased growth due to competition during pulses results in decreased survival during interpulse periods. 4 According to the two-phase resource dynamics hypothesis, Grime's hypothesis that competition is unimportant at low productivity will hold when soil resource availability between pulses in unproductive environments is controlled by abiotic factors and when survival during interpulse intervals is independent of or even negatively correlated with growth during pulse periods. In contrast, Newman's and Tilman's hypothesis that competition is equally important along productivity gradients will apply when either of these conditions is not true. We predict that the conditions for Grime's hypothesis to apply are more likely for productivity gradients driven by water than by mineral nutrients and when response to competition is measured for community structure or individual survival rather than for individual growth. 5 We tested the predictions about water vs. nutrient gradients and growth vs. survival or community structure responses using data from a literature survey on field and