Diatom–chemistry relationships in Yellowstone Lake (Wyoming) sediments: Implications for climatic and aquatic processes research

We analyzed sediment elemental composition (C, N, P, and Si) and planktonic diatom fossils from the top 32 cm of a short core (55 cm) from Yellowstone Lake (Wyoming), tracking deposition from about 1650 to 1998. Associations between fossil communities and sediment chemistry were measured by rank-order correlation to gauge decadal scale relationships between lake chemistry and phytoplankton community composition. Strong associations were found between the abundance of individual diatom species and the chemical composition of sediments, suggesting that sediment chemistry can directly track the elemental composition of seston. Among three fossil measures (relative frustule abundance, relative biovolume, and absolute biovolume), associations between sediment chemistry and relative biovolume were strongest. We then compared measures and trends in sediment elemental composition to in-lake seston chemistry to assess transport processes that alter sediment composition relative to source material. On the basis of elemental stoichiometry, sediments were enriched in P and depleted in C and N compared with seston. Trends of increasing C and N accumulation along with decreasing Si accumulation were consistent with patterns of increasing productivity, increasing lake N levels, and decreasing dominance of diatoms, respectively. A sharp increase in sediment C : N ratio, with a decrease in diatom absolute biovolume and absolute C and N levels, was associated with a prolonged drought in the 1930s. These trends suggest alteration of ecosystem properties by a combination of climatic variation and increasing N availability, adding complexity to larger goals of climatic reconstruction by diatom–nutrient chemistry–climate transfer functions in large lakes of the region. Lacustrine sediments hold a wealth of natural history that allows the reconstruction of changes in aquatic and ecosystem processes with both direct and indirect metrics. In particular, diatom fossil records alone are most often used to reconstruct lake chemistry. Transfer functions between diatoms and/or chrysophytes found in surface sediment communities and lake chemistry have been constructed using weighted averaging regression and calibration models that allow correlation of diatom community composition with water chemistry in a defined regional series of lakes (e.g., Dixit et al. 1992). The fossil stratigraphy in those and similar lakes can then be assessed by means of the calibration model to reconstruct historical water chemistry. These models make the general assumption that the chemistry–community relationship has remained within the range of the modern calibration set throughout the historical period recorded in the lake sediments. The reconstructed aquatic conditions then allow us to infer environmental change on a broader scale. For instance, transfer functions developed between diatom communities and