Partitioning stoichiometric components of epilithic biofilm using mixing models
Limnol. Oceanogr. Methods 9:185-193 (2011) | DOI: 10.4319/lom.2011.9.185
ABSTRACT: Epilithic biofilms are an important and complex food resource in shallow aquatic ecosystems. Standard methods of sampling these biofilms typically yield a combination of algal cells, other organic material, and inorganic sediment. Physical methods of isolating algal fractions rely on differences in density among biofilm components, and clean separation is not always possible. Here, we explore the application of linear mixing models to the problem of estimating nutrient content of biofilm components. This method is based on the assumption that variation in bulk sample elemental composition within a sampling site is due to differences among samples in relative composition. By using standard methods to quantify organic content and algal biomass within biofilm samples, the mixing model produces estimates of nutrient content of each end member. To test this method, we analyzed the phosphorus (P) content of algal cells, non-algal organic material, and inorganic sediment associated with biofilm samples that increase in P-content 4-fold across a natural P gradient in lowland streams in Costa Rica. Our analyses show that nearly all of the observed increase in biofilm P from high-P streams was due to elevated P-content of algal cells, rather than P sorbed to inorganic sediments or in Prich heterotrophic bacteria. Results from the mixing model were supported by results from density fractionation. This linear mixing model approach complements empirical methods of separating complex food resources and may lead to a better understanding of trophic interactions in aquatic food webs.