was incorrectly coded in such a way that bji replaced bij. While we originally reported that our model had symmetric interactions among species (p. 105), the error results in a model with asymmetries in competition among consumer species. These asymmetries cause resource species to go extinct in the models with two trophic levels. For the model originally reported in the paper in which we assumed symmetric interactions among species, resource species do not go extinct for the parameter values used to produce Figures 2–5. As a result, the relationship between consumer diversity and consumer density is identical for systems with one and two trophic levels.
While the programming error changed Figures 2–5, it does not change the general conclusions described in the article. We concluded that when resource species go extinct with increasing numbers of consumer species, the relationship between consumer diversity and consumer abundance will differ between systems having one vs. two trophic levels, whereas when there is no extinction, the relationship is the same. Comparing the asymmetrical model caused by the programming error and the correct, symmetrical model substantiates this claim.