Illegal dumping of oil and gas wastewater alters arid soil microbial communities

Abstract

The Permian Basin, underlying southeast New Mexico and west Texas, is one of the most productive oil and gas (OG) provinces in the United States. Oil and gas production yields large volumes of wastewater with complex chemistries, and the environmental health risks posed by these OG wastewaters on sensitive desert ecosystems are poorly understood. Starting in November 2017, 39 illegal dumps, as defined by federal and state regulations, of OG wastewater were identified in southeastern New Mexico, releasing ~600,000 L of fluid onto dryland soils. To evaluate the impacts of these releases, we analyzed changes in soil geochemistry and microbial community composition by comparing soils from within OG wastewater dump-affected samples to unaffected zones. We observed significant changes in soil geochemistry for all dump-affected compared with control samples, reflecting the residual salts and hydrocarbons from the OG-wastewater release (e.g., enriched in sodium, chloride, and bromide). Microbial community structure significantly (P < 0.01) differed between dump and control zones, with soils from dump areas having significantly (P < 0.01) lower alpha diversity and differences in phylogenetic composition. Dump-affected soil samples showed an increase in halophilic and halotolerant taxa, including members of the Marinobacteraceae, Halomonadaceae, and Halobacteroidaceae, suggesting that the high salinity of the dumped OG wastewater was exerting a strong selective pressure on microbial community structure. Taxa with high similarity to known hydrocarbon-degrading organisms were also detected in the dump-affected soil samples. Overall, this study demonstrates the potential for OG wastewater exposure to change the geochemistry and microbial community dynamics of arid soils.IMPORTANCEThe long-term environmental health impacts resulting from releases of oil and gas (OG) wastewater, typically brines with varying compositions of ions, hydrocarbons, and other constituents, are understudied. This is especially true for sensitive desert ecosystems, where soil microbes are key primary producers and drivers of nutrient cycling. We found that releases of OG wastewater can lead to shifts in microbial community composition and function toward salt- and hydrocarbon-tolerant taxa that are not typically found in desert soils, thus altering the impacted dryland soil ecosystem. Loss of key microbial taxa, such as those that catalyze organic carbon cycling, increase arid soil fertility, promote plant health, and affect soil moisture retention, could result in cascading effects across the sensitive desert ecosystem. By characterizing environmental changes due to releases of OG wastewater to soils overlying the Permian Basin, we gain further insights into how OG wastewater may alter dryland soil microbial functions and ecosystems.

Keywords: contaminants; dryland soils; microbial communities; microbial ecology; oil and gas; produced water; soil microbiology; wastewater.

FIG 1

Map of the OG-wastewater dump sites in southeastern New Mexico. A total of five sites were studied, and soil samples were collected from dump-affected and unaffected (control) zones at each location. Brown squares and lines are OG well pads and roads, respectively. Basemap sources: Esri, DigitalGlobe, GeoEye, i-cubed, USDA FSA, USGS, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo, and the GIS User Community.

FIG 2

Comparisons of soil properties and selected solute concentrations in soil-water extracts between dump and control soils across all sampling sites. Soil percent (%) moisture (A), % carbon (B), % nitrogen (C), and HEM (D) were measured on bulk samples. Na (E), Cl (F), Sr (G), and sulfate (H) concentrations in soil-water extracts were selected as they are known markers of OG wastewaters; concentrations of additional inorganic solutes were determined in soil-water extracts and are presented in Fig. S1 and reference (83). The lower and upper hinges correspond to the 25th and 75th percentiles, respectively, and the whiskers correspond to minimum and maximum values, with the middle line being the median. Significant differences were observed for all analytes, as indicated by P < 0.05 based on a Mann-Whitney U test.

FIG 3

Differences in alpha diversity in dump and control soil microbial communities were assessed by (A) observed OTU richness, (B) Chao1 index, (C) Inverse Simpson diversity index, and (D) Shannon diversity index. Each sample is represented as a dot on the box and whisker plot. P-values correspond to the nonparametric pairwise Mann-Whitney rank-sum test. The lower and upper hinges correspond to the 25th and 75th percentiles, respectively, and the whiskers correspond to the minimum and maximum values. Across all four measures, a significant decrease in alpha diversity was observed in dump soil communities. OTU richness was determined using raw counts of a rarefied data set.

FIG 4

NMDS of (A) weighted UniFrac distances among control and dump-affected soil microbial communities and (B) with significant (ENVFIT, P < 0.05) co-occurring geochemical vectors fitted. Ellipses represent 95% confidence intervals. Significant differences in community structure were determined using PERMANOVA with soil condition as a factor and 999 permutations. Vectors are various soil physicochemical factors that significantly co-occurred with changes in community structure. The stress value for the NMDS ordination is 0.096.

FIG 5

Relative abundance (%) at the phylum-level for taxa detected at >1% in dump-affected and control soil samples. The lower and upper hinges correspond to the 25th and 75th percentiles, respectively, and the whiskers correspond to minimum and maximum values, with the middle line being the median. Proteobacteria are presented at the class level; note that Gammaproteobacteria are presented on a separate scale.

FIG 6

The differential abundance of microbial taxa at the family level that had a significant (P < 0.01) log₂ fold change (doubling) ± standard error. Class-level taxonomy is denoted via square brackets or in parentheses. For the DESeq2 analysis, only the 14 taxa with significant log₂ fold changes that were found in >1% relative abundance are shown. An additional 64 taxa had significant log₂ fold changes (Table S9). Bars on the left indicate a significant decrease in abundance in dump soils compared to controls, while bars on the right indicate an increase in dump soils.