The combined effects of scale and productivity on species richness

1 Although species-area curves are well documented, we know little about how they vary along productivity gradients, nor do we have a firm understanding of whether the richness-productivity relationship is independent of scale. In order to investigate these relationships, 10 species-area curves were constructed for 10 sets of nested quadrats (from 0.25 m(2) to 150 m(2)) in herbaceous marshes where biomass ranged from 128 g m(-2) to 1050 g m(-2). 2 Slopes (z) of the species-area curves (logS = logc + zlogA) were positively correlated with production (whether measured as biomass or soil organic content). A possible cause of increased slope was larger potential species pools at high biomass. The constants (c) were negatively associated with production, which concurs with the general richness-productivity relationship. 4 Linear models were used to predict both c and z using biomass and soil organic content. Biomass explained about 38% of the variation in both z and logc. Soil organic content was a better predictor. It explained 60% of the variation in z and 41% of the variation in c. 5 The slopes (z) and constants (c) of the species-area curves were negatively correlated. Because c is the expected species richness at 1 m(2) and z strongly affects larger-scale richness, the results suggest that small-scale richness is not a good indicator of larger-scale richness. There was no significant relationship between species richness in small-scale (0.25 m(2)) quadrats and large-scale quadrats (150 m(2)). 6 These results suggest that the conflicting data that have led to recent debate over the richness-productivity relationship may reflect both the effects of scale and the natural history peculiarities of the systems studied.