Grazers, producer stoichiometry, and the light : nutrient hypothesis revisited.

The stoichiometric light : nutrient hypothesis (LNH) links the relative supplies of key resources with the nutrient content of tissues of producers. This resource-driven variation in producer stoichiometry, in turn, can mediate the efficiency of grazing. Typically, discussions of the LNH attribute this resource–stoichiometry link to bottom-up effects of light and phosphorus, which are mediated through producer physiology. Emphasis on bottom-up effects implies that grazers must consume food of quality solely determined by resource supply to ecosystems (i.e., they eat what they are served). Here, we expand upon this largely bottom up interpretation with evidence from pond surveys, a mesocosm experiment, and a model. Data from shallow ponds showed the ‘‘LNH pattern’’ (positive correlation of an index of light: phosphorus supply with algal carbon : phosphorus content). However, algal carbon: phosphorus content also declined as zooplankton biomass increased in the ponds. The experiment and model confirmed that this latter correlation was partially caused by the various bottom-up and top-down roles of grazers: the LNH pattern emerged only in treatments with crustacean grazers, not those without them. Furthermore, model and experiment clarified that another bottom-up factor, natural covariation of nitrogen : phosphorus ratios with light : phosphorus supply (as seen in ponds), does not likely contribute to the LNH pattern. Finally, the experiment produced correlations between shifts in species composition of algae, partially driven by grazing effects of crustaceans, and algal stoichiometry. These shifts in species composition might shape stoichiometric response of producer assemblages to resource supply and grazing, but their consequences remain largely unexplored. Thus, this study accentuated the importance of grazing for the LNH; deemphasized a potentially confounding, bottom-up factor (covarying nitrogen : phosphorus supply); and highlighted an avenue for future research for the LNH (grazer-mediated shifts in producer composition).We thank T. Darcy-Hall, A. Downing, P. Geddes, and N. Howe for help with sampling the cattle tank experiment and natural ponds. We analyzed the C:N samples in the Robertson lab at the W. K. Kellogg Biological Station (KBS) with the help of A. Corbin and T. Darcy-Hall. Thanks also go to G. Mittelbach, N. Consolatti, A. Tessier, and P. Woodruff for technical support. Finally, S. Diehl kindly provided us with data from his experiments. Primary funding came from NSF DEB 98-15799, DEB 02-35579, and funds from Indiana University. This is KBS contribution number 1277