Diatom-based TP and TN inference models and indices for monitoring nutrient enrichment of New Jersey streams

We evaluated the potential for using diatoms to assess and monitor nutrient enrichment in New Jersey streams and rivers, and propose inference models and indices for regulatory purposes. We assessed the relationship between benthic diatom and water chemistry samples (n = 101) collected from 45 sites in 3 ecoregions: Northern Piedmont, Northeastern Highlands and Ridge and Valley. Diatom assemblages were dominated by pollution-tolerant taxa. Multivariate analysis showed that nutrient concentrations explained significant proportions of the variation in diatom species composition. Weighted-averaging partial least square (WA-PLS) total phosphorus (TP) and total nitrogen (TN) inference models (n = 91) showed good predictive ability (TP model: r apparent 2 = 0.87 ; r boot 2 = 0.72 ; RMSEboot = 0.23 log10 μg L−1 TP; TN model: r apparent 2 = 0.88 ; r boot 2 = 0.58 ; RMSEboot = 0.23 log10 μg L−1 TN). Diatom TP and TN indices were created to simplify presentation of results for the general public by rescaling the inferred TP and TN values from 0 to 100. The obtained index scores were assigned to nutrient impairment categories for regulatory assessment purposes.