We investigated relationships between diversity (measured as species richness) and plant productivity with nine different consumer-resource competition models based on competition for two essential resources. The models share a common analytical framework, but each uses a unique combination of mechanistic assumptions regarding plant population growth, resource consumption, and resource supply. We regarded both peak community biomass and community biomass at equilibrium as measures for community productivity. Using MATLAB, we generated numerical data showing the relationships between initial species richness and (1) peak biomass, (2) equilibrium biomass, (3) the difference between peak and equilibrium biomass, (4) unconsumed resources at equilibrium, (5) time at which community biomass reaches its peak, (6) time at which the community reaches equilibrium and (7) the time lapse between peak biomass and equilibrium. For each model and each initial species richness (one through ten competing species) we calculated 1000 trial competition processes with species randomly drawn from an unlimited species pool. Furthermore, we closely analyzed one explicit competition trial in which we applied the same parameter values to each of the nine different models.

Results/Conclusions

We found that different mechanistic assumptions, all else being equal, can lead to radically different predictions about competition outcomes.  Despite these differences, however, all nine models predict a positive correlation between initital species richness and average peak biomass.  Seven out of the nine models show a positive correlation between initial species richness and average equilibrium biomass.  All models show, on average, a fuller utilization of resources for more diverse communities. The average time lapse between peak biomass and equilibrium, as well as the average difference between peak biomass and equilibrium biomass, were shown to be positively correlated with species richness for all models. These results suggest that harvest yields (for example for cellulosic biofuel production) could possibly be maximized by managing for more diverse ecosystems and harvesting periodically before the plant community reaches equilibrium. Since our findings are robust for different combinations of mechanistic assumptions regarding population growth, resource supply, and resource consumption, they may apply to a broad range of specific ecosystems.