Correlations of sedimentary diatoms with watershed land use and limnological conditions in Northern New Jersey Lakes

Sedimentary diatoms were collected from 62 New Jersey lakes and correlated with limnological and land use variables. Four diatom attributes (species richness, percentage dominance by a single diatom species, percentage acidobiontic species, percentage eutraphentic species) and 4 diatom indices, Centrales:Pennales (C/P), Araphidinaea:Centrales (A/C), Generic Diatom (GI), and diatom index of biotic integrity (DIBI), were compared to the environmental variables. Lakes were classified as forested lakes (±80% the watershed in forests), mixed land use lakes (21–79% forests), or deforested lakes (≤20% forests) based on the satellite imagery. Forested lakes had significantly lower specific conductance and pH, greater Z_SD and hypo limnetic DO, and less urban land use within their watersheds than either mixed land use or deforested lakes. Forested lakes also had greater species richness, lesser dominance by single species, more acidobiontic diatoms, fewer eutraphentic diatoms, lower C/P, and higher DIBI scores. There were no significant differences among lake classes for A/C or GI indices. The 4 diatom attributes and 4 diatom indices were correlated with each other, though A/C was not significantly correlated with any of the diatom attributes or indices. The remaining 3 diatom indices were highly correlated with taxa richness, percentage dominance, and percentage acidobiontic species. The attributes and indices were variously correlated with the 13 limnological and watershed land use variables, with alkalinity, specific conductance, pH, Z_SD, and percentage of the watershed in forests and urban land uses exhibiting the strongest correlations. Canonical correlation analysis revealed significant correlations between the diatom attributes and indices and the environmental axis related to lower alkalinity and pH, and higher Z_SD and percentage of the watershed in forests. Our results suggest that diatom attributes and indices are useful indicators of the ecological condition of lakes and that these attributes and indices respond to human disturbances within a lake's watershed. Of the indices tested, the DIBI appears to be the most useful, significantly correlating with stressors and limnological variables of interest, especially the loss of forests from watersheds. The results support the use of sedimentary diatom attributes and indices for monitoringlake water-quality and watershed land use changes relative the management of these resources.