Tropical Fishes as Biological Bulldozers: Density Effects on Resource Heterogeneity and Species Diversity

We examined whether grazing fishes exert density-dependent effects on the spatial heterogeneity of resources and benthic species diversity in a tropical Andean stream. We hypothesized that bulldozer grazers (i.e., all or nothing grazing) can be important sources of spatial heterogeneity and that a unimodal relationship should exist between grazer density and resource heterogeneity. We reasoned that organismally generated heterogeneity would be minimal at low densities if grazers are ineffective, and at high densities if grazers thoroughly remove resources. In contrast, spatial heterogeneity should be highest at intermediate densities as bulldozer grazers maintain a dynamic mosaic of resource states. We posited that a corresponding unimodal relationship should exist between grazer density and benthic diversity, if resource heterogeneity is an important mechanism maintaining species diversity. We carried out observational and experimental studies to test our hypotheses. First, we quantified natural spatial patterns of benthic resources across transects using fish feeding scars as surrogates. We then analyzed data using a set of landscape indices that capture different components of heterogeneity. A large degree of spatial heterogeneity of algal and sediment resources was apparent both within and among pools. Second, we manipulated densities of the common grazing fish Parodon apolinari (Parodontidae) in 8-m2 pens, and quantified spatial heterogeneity and corresponding benthic diversity within enclosures. Parodon exerted strong density-dependent effects on the spatial heterogeneity of resources. Initially (at 7 d) we observed a unimodal relationship between grazer density and spatial heterogeneity along our density gradient (1.25–10.0 Parodon/m2), as predicted. However, at 14 d, our lowest fish density (1.25 Parodon/m2) was sufficient to generate substantial spatial heterogeneity. Although grazers were major sources of heterogeneity, we found little support for the hypothesis that organismally generated heterogeneity is an important mechanism for maintaining species diversity. Interestingly, the greatest invertebrate richness occurred in high density treatments, which had the most spatially uniform resource distributions. These results suggest that at the within-pool scale, grazers can have important effects on both spatial heterogeneity and benthic diversity, yet their generation of resource patchiness does not appear to be the underlying mechanism driving strong fish effects on species diversity.