Soil pH is an important determinant of microbial community composition and diversity, yet few studies have characterized the specific effects of pH on individual bacterial taxa within bacterial communities, both abundant and rare. We collected composite soil samples over two years from an experimentally maintained pH gradient ranging from 4.5 to 7.5 from the Craibstone Experimental Farm (Craibstone, Scotland). Extracted nucleic acids were characterized by bacterial and group-specific denaturing gradient gel electrophoresis (DGGE) and next-generation sequencing of bacterial 16S rRNA genes. Both methods demonstrated comparable and reproducible shifts within higher taxonomic bacterial groups (e.g., Acidobacteria, Alphaproteobacteria, Verrucomicrobia, and Gammaproteobacteria) across the pH gradient. In addition, we used non-negative matrix factorization (NMF) for the first time on 16S rRNA gene data to identify positively interacting (i.e., co-occurring) operational taxonomic unit (OTU) clusters (i.e., "components"), with abundances that correlated strongly with pH, and sample year to a lesser extent. All OTUs identified by NMF were visualized within principle coordinate analyses of UniFrac distances and subjected to taxonomic network analysis (SSUnique), which plotted OTU abundance and similarity against established taxonomies. Most pH-dependent OTUs identified here would not have been identified by previous methodologies for microbial community profiling and were unrelated to known lineages. This article is protected by copyright. All rights reserved.