Impacts of Long-Term Irrigation of Domestic Treated Wastewater on Soil Biogeochemistry and Bacterial Community Structure

Abstract

Freshwater scarcity and regulations on wastewater disposal have necessitated the reuse of treated wastewater (TWW) for soil irrigation, which has several environmental and economic benefits. However, TWW irrigation can cause nutrient loading to the receiving environments. We assessed bacterial community structure and associated biogeochemical changes in soil plots irrigated with nitrate-rich TWW (referred to as pivots) for periods ranging from 13 to 30 years. Soil cores (0 to 40 cm) were collected in summer and winter from five irrigated pivots and three adjacently located nonirrigated plots. Total bacterial and denitrifier gene abundances were estimated by quantitative PCR (qPCR), and community structure was assessed by 454 massively parallel tag sequencing (MPTS) of small-subunit (SSU) rRNA genes along with terminal restriction fragment length polymorphism (T-RFLP) analysis of nirK, nirS, and nosZ functional genes responsible for denitrification of the TWW-associated nitrate. Soil physicochemical analyses showed that, regardless of the seasons, pH and moisture contents (MC) were higher in the irrigated (IR) pivots than in the nonirrigated (NIR) plots; organic matter (OM) and microbial biomass carbon (MBC) were higher as a function of season but not of irrigation treatment. MPTS analysis showed that TWW loading resulted in the following: (i) an increase in the relative abundance of Proteobacteria, especially Betaproteobacteria and Gammaproteobacteria; (ii) a decrease in the relative abundance of Actinobacteria; (iii) shifts in the communities of acidobacterial groups, along with a shift in the nirK and nirS denitrifier guilds as shown by T-RFLP analysis. Additionally, bacterial biomass estimated by genus/group-specific real-time qPCR analyses revealed that higher numbers of total bacteria, Acidobacteria, Actinobacteria, Alphaproteobacteria, and the nirS denitrifier guilds were present in the IR pivots than in the NIR plots. Identification of the nirK-containing microbiota as a proxy for the denitrifier community indicated that bacteria belonged to alphaproteobacteria from the Rhizobiaceae family within the agroecosystem studied. Multivariate statistical analyses further confirmed some of the above soil physicochemical and bacterial community structure changes as a function of long-term TWW application within this agroecosystem.

FIG 1

Location of the spray-field agroecosystem in Tallahassee, FL. Groundwater in the area flows in a southerly direction toward the Gulf of Mexico. The right panel shows an aerial image of the wastewater-receiving plots (pivots) and the nonirrigated control sites sampled for this study. Aerial image data are from the U.S. Department of Agriculture, Farm Service Agency; the map was created using ArcGIS, version 10.1.

FIG 2

Box-and-whisker plots are shown for the soil biogeochemical parameters measured from the spray-field agroecosystem soils in Tallahassee, FL, analyzed as a function of land use and season. Box-and-whisker plots give the medians (horizontal lines inside the boxes), interquartile ranges (boxes), and outliers (small black dots). †, significant difference in values between seasons; *, significant difference in values between land use types. IR and NIR indicate data obtained from irrigated pivots and nonirrigated plots, respectively.

FIG 3

Box-and-whisker plots are shown for the gene copy numbers measured from the spray-field agroecosystem soils in Tallahassee, FL, analyzed as a function of land use and season. Box-and-whisker plots give the medians (horizontal lines inside the boxes), interquartile ranges (boxes), and outliers (small black dots). Differences between seasons were not significant. *, significant difference between values for land use types. IR and NIR indicate data obtained from irrigated pivots or nonirrigated plots, respectively.

FIG 4

(A) Bar plot showing abundances of the predominant phyla identified from the spray-field agroecosystem soils in Tallahassee, FL, as a function of land application of domestic TWW relative to those of nonirrigated soils. Identified taxa were categorized at the class level, with the exception of only Acidobacteria, which is shown at the phylum level. (B) Double-hierarchical dendrogram showing distribution at the class level of identified taxa from the spray-field agroecosystem soils in Tallahassee, FL. The phylogenetic tree was calculated using the neighbor-joining method, and the relationship among samples was determined by Bray-Curtis distance and the complete clustering method. The heat map depicts the relative percentage of each identified class (variables clustering on the y axis) within each sample (x-axis clustering). The relative Euclidean distance values for the bacterial classes identified are depicted by red and green, indicating low and high abundance, respectively, correlating with the legend at the bottom of the figure. Clusters based on the distance of samples along the x axis and the bacterial classes along the y axis are indicated in the top and left of the figure, respectively. Arrows point to the phyla/taxa that were clearly different in the irrigated pivots relative to those in the nonirrigated plots based on Euclidean distances.

FIG 5

Nonmetric multidimensional scaling ordination of the T-RFLP data for the nirK gene (A), nirS gene (B), and nosZ gene (C) obtained from the spray-field agroecosystem soils in Tallahassee, FL. Each data point is a mean of triplicate runs; the data are based on a Bray-Curtis similarity matrix. Open and filled blue squares and red circles represent the summer (S) and winter (W) seasons, respectively, for the IR and NIR sites, as indicated. Bray-Curtis similarity values between irrigated and nonirrigated sample bacterial communities are shown at the 20%, 40%, 60%, and 80% levels in the summer (S) and winter (W) seasons.

FIG 6

Phylogenetic analysis of the denitrifying bacteria obtained from the agroecosystem soils in Tallahassee, FL, based on nirK-type gene sequences. The neighbor-joining method was used to construct the phylogenetic tree with a bootstrap value of 1,000 iterations; only values above that were above 50 are shown at branch points. Accession numbers of the retrieved nirK-type gene sequences along with their closest phylogenetic relatives are shown in parentheses. The Escherichia coli formate dehydrogenase gene (fdoG) (GenBank accession number X87583) was used as an outgroup.

FIG 7

Biplot derived from canonical correspondence analysis (CCA) of the bacterial abundances, correlated with soil biogeochemical and environmental properties obtained from the spray-field agroecosystem in Tallahassee, FL. Percentages of variation are shown in parentheses on the x and y axes. Nonirrigated and irrigated sites and sampling seasons are indicated. The filled blue circles represent bacterial taxa/phyla identified over two seasons in IR and NIR soils.

FIG 8

(A) Nonmetric multidimensional scaling ordination plots of the bar-coded pyrosequences showing the bacterial community structure at the class level. Bray-Curtis similarity values between irrigated and nonirrigated sample bacterial communities are shown at the 50%, 60%, 70%, and 80% levels in the summer (S) and winter (W) seasons. (B) Dendrogram based on cluster analysis of the NMDS of the total bacterial community analyzed by 454 massively parallel tag sequencing. The relative abundance data were transformed by log(x + 1), and the bacterial communities were grouped using the complete linkage option.