Abstract
Increasing nitrogen (N) loads present a threat to estuaries, which are among the most heavily populated and perturbed parts of the world. N removal is largely mediated by the sediment microbial process of denitrification, in direct competition to dissimilatory nitrate reduction to ammonium (DNRA), which recycles nitrate to ammonium. Molecular proxies for N pathways are increasingly measured and analyzed, a major question in microbial ecology, however, is whether these proxies can add predictive power around the fate of N. We analyzed the diversity and community composition of sediment nirS and nrfA genes in 11 temperate estuaries, covering four types of land use in Australia, and analyzed how these might be used to predict N removal. Our data suggest that sediment microbiomes play a central role in controlling the magnitude of the individual N removal rates in the 11 estuaries. Inclusion, however, of relative gene abundances of 16S, nirS, nrfA, including their ratios did not improve physicochemical measurement-based regression models to predict rates of denitrification or DNRA. Co-occurrence network analyses of nirS showed a greater modularity and a lower number of keystone OTUs in pristine sites compared to urban estuaries, suggesting a higher degree of niche partitioning in pristine estuaries. The distinctive differences between the urban and pristine network structures suggest that the nirS gene could be a likely gene candidate to understand the mechanisms by which these denitrifying communities form and respond to anthropogenic pressures.