Evaluation of Digital PCR (dPCR) for the Quantification of Soil Nitrogen Turnover Bacteria in Wetland Mesocosms in Response to Season, Fertilization, and Plant Species Richness

Excess nutrients from nonpoint sources are an ongoing problem that is expected to worsen as population and fertilizer usage rise. Conventional centralized treatment systems are not well suited to address nonpoint source pollution. More distributed best management practices (BMPs) like constructed wetlands are a promising alternative and have been widely implemented in the US since the 1970's. Constructed wetlands are multi-functional systems that can effectively store and transform harmful contaminants using primarily natural processes. However, the removal of pollutants like nitrogen by wetlands is highly variable, likely due to a combination of factors such as plant species-specific assimilation behavior, the effects of plant communities on microbial diversity and function, and variable nitrogen inputs. In this study, the effect of plant species richness (i.e., number of plant species in a system) and seasonal nutrient loading (i.e., nitrogen fertilization) on the microbial community responsible for regulating nitrogen turnover in wetland mesocosm soils was investigated. The chip-based QuantStudio 3D digital PCR (QS3D dPCR) system was used to quantify ammonia-oxidizing bacteria (AOB), nitrite-oxidizing bacteria (NOB), comammox, anammox, and denitrifiers. Principal component analysis (PCA) was used to identify dominant patterns in the microbial community and nitrogen species. Resampling-based analysis of variance (ANOVA) was used to assess statistical significance of any observed differences caused by nitrogen fertilization or plant species richness. Results indicated that fertilization or season, which was convolved with fertilization, was the dominant factor influencing the microbial community in the study environment (27% variance explained), as indicated by the disparate clustering of fall (fertilized) and spring (unfertilized) samples about principal component 1 (fall: negative PC1, spring: positive PC1). Because unplanted unfertilized controls sampled in November clustered within the season in which they were collected rather than with other unfertilized samples collected in May, season may have influenced microbial community shifts more than fertilization for unplanted systems. This finding should be interpreted cautiously, however, given the small number of unplanted unfertilized controls (N = 2) and the absence of similar controls in the planted systems. The most abundant bacterial groups detected in May (November) were AOB, nirK, anammox, and Nitrospira spp. NOB (AOB, anammox, Nitrospira spp. NOB, and nosZ). The effects of plant species richness were more nuanced, with greater richness significantly impacting the abundance of only a subset of bacterial groups (i.e., the nitrifying bacteria AOB, Nitrospira spp. NOB, and comammox, but not the denitrifying bacteria). Different relationships between richness and microbial abundance were observed in different seasonal nutrient loadings (i.e., interaction effects between richness and fertilization were detected for some bacterial groups).