Inorganic Constituents

Nitrate was the inorganic constituent most prevalent at concentrations greater than a regulatory benchmark; nitrate concentrations exceeded its EPA MCL (10 milligrams per liter as nitrogen [mg-N/L]) in 17 percent of domestic groundwater resources (fig. 2). Nitrate also was detected at moderate and low concentrations in 9 and 46 percent of domestic groundwater resources, respectively. Nitrate is a nutrient that exists naturally in groundwater but is detected at higher concentrations in developed areas because of fertilizer applications, feedlots, and septic systems. Fluoride was detected at moderate concentrations relative to the State of California MCL of 2 mg/L in 6 percent of domestic groundwater resources. The trace elements barium, selenium, and chromium were each detected at concentrations exceeding their respective MCLs (EPA MCL 1,000 micrograms per liter [μg/L] for barium, EPA MCL 50 μg/L for selenium, and State of California MCL 50 μg/L for chromium) in 4 percent of domestic groundwater resources. Hexavalent chromium, Cr(VI), a carcinogenic valence state of chromium, was not analyzed in 18 percent of domestic groundwater resources, and the proportion of domestic groundwater resources exceeding its State of California MCL (10 μg/L) is therefore not included because it may not be spatially representative. Chromium is composed of Cr(VI) and trivalent chromium and thus can be considered a maximum potential concentration of Cr(VI); a statewide assessment of chromium in groundwater found that almost all chromium in groundwater is present as Cr(VI) (Izbicki and others, 2015). The trace elements arsenic and uranium (natural) were detected at moderate concentrations in 7 and 13 percent of domestic groundwater resources, respectively.

Constituents with secondary maximum contaminant levels (SMCLs; aesthetic-based and technical-based benchmarks) are not included in the overview of water-quality charts (inset on fig. 1) and are considered separately to highlight factors affecting aesthetic and technical properties of groundwater, such as taste, color, scaling, or odor (total dissolved solids [TDS], manganese, and iron on fig. 2). Total dissolved solids are a measure of groundwater salinity and have recommended and upper State of California SMCLs of 500 milligrams per liter (mg/L) and 1,000 mg/L, respectively, which were used as moderate and high thresholds for this study. Total dissolved solids were detected at high and moderate levels in 24 and 52 percent of domestic groundwater resources, respectively (fig. 2). Chloride (Cl⁻) and sulfate (SO₄⁻²) are major ionic components of TDS and have lower and upper State of California SMCL threshold values of 250 and 500 mg/L, respectively. Chloride and sulfate were greater than the upper State of California SMCL in 4 and 9 percent of domestic groundwater resources, respectively. Manganese, a trace metal that can affect the taste and color of groundwater and can precipitate in pipes, was present at concentrations greater than its State of California SMCL of 50 µg/L in 28 percent of domestic groundwater resources (fig. 2). Manganese also has a SWRCB defined notification level (500 µg/L; California State Water Resources Control Board, 2022) that was exceeded in 2 percent of domestic groundwater resources. Iron, a trace metal that can affect taste and color of groundwater, was detected at moderate concentrations relative to its State of California SMCL of 300 µg/L in 7 percent of domestic groundwater resources (fig. 2). Boron has a non-regulatory, health-based benchmark (EPA lifetime health advisory level of 5,000 mg/L) that was used as the high threshold for this study; boron was present at high concentrations in 7 percent of domestic groundwater resources (fig. 2).

Organic Constituents

No VOC or pesticide constituents were present at concentrations that exceeded regulatory MCL benchmarks in this study area. The VOCs with regulatory benchmarks that were detected in the study area included a fumigant (1,2-dichloropropane [1,2-DCP]), two gasoline components (tert-butyl alcohol and benzene), a solvent (1,1-dichloroethane), and a disinfection byproduct (trichloromethane). Each VOC was detected at low concentrations in less than 10 percent of domestic groundwater resources. No pesticide constituents with regulatory benchmarks were detected.

Microbial Indicators

Microbial indicator constituents can be used to evaluate the potential for bacterial contamination of groundwater from wastewater sources. In this study, groundwater samples were tested for the presence or absence of total coliform bacteria, enterococci, and Escherichia coli. Total coliform bacteria and enterococci were detected in 4 percent of domestic groundwater resources. E. coli bacteria were not detected. Coliform bacteria are common and naturally in soil and surface water and are generally harmless to humans. E. coli and enterococci are more specific indicators of fecal contamination because they are in human and animal wastes (California State Water Resources Control Board, 2019).

Constituents of Special Interest—Per- and Polyfluoroalkyl Substances

Per- and polyfluoroalkyl substances (PFAS) are a family of human-made chemicals that commonly are used as surfactants in a variety of industrial processes and are persistent in the environment (Evich and others, 2022). Of the 28 PFAS constituents analyzed, 10 were detected at concentrations greater than laboratory reporting levels, including 5 that have EPA MCL benchmarks (fig. 3; Faulkner and Dupuy, 2024). At least one PFAS constituent was detected at high concentrations (greater than an EPA MCL) in 10 percent of domestic groundwater resources, and an additional 4 percent of the resource had at least one PFAS constituent detected at moderate concentrations (greater than one-tenth of an EPA MCL). The constituents PFOA and PFOS were detected at concentrations exceeding their EPA MCLs (both 4 μg/L) at the same sites, corresponding to 6 percent of domestic groundwater resources. The constituent PFHxS was detected at a concentration exceeding its EPA MCL of 10 μg/L in 4 percent of domestic groundwater resources. The sites with high concentrations of PFOS, PFOA, or PFHxS also had detections at concentrations greater than laboratory reporting levels as much as 3 of the PFAS constituents comprising the Hazard Index EPA MCL (PFHxS, PFNA, or PFBS). There were no detections of the fourth component of the Hazard Index, HFPO-DA. The site with a high concentration of PFHxS also exceeded the Hazard Index EPA MCL, which is set at 1 (unitless) for mixtures containing two or more of PFHxS, PFNA, PFBS, or HFPO-DA (U.S. Environmental Protection Agency, 2024a).

Agricultural Factors

Nitrate was present at moderate or high concentrations in 26 percent of domestic groundwater resources (fig. 2). Nitrate concentrations greater than the moderate threshold (5 mg-N/L) are commonly associated with agricultural land use (Dubrovsky and others, 2010), and in this study, nitrate was positively correlated with the percentage of agricultural land within 500 m of each well (beige on fig. 4; Spearman’s rho=0.44, p-value=0.01 for land use in 1974; Faulkner and Dupuy, 2024; Dupuy and others, 2025). This correlation implies that the source of moderate and high nitrate in domestic groundwater resources may be related to the application of fertilizers on agricultural land, consistent with the Salt and Nutrient Management Plans for the Llagas and North San Benito subbasins (Todd Groundwater, 2014a, 2014b). Nitrate also was positively correlated with the percentage of urban land within 500 m of each well (pink on fig. 4; Spearman’s rho=0.63, p-value less than 0.001 for land use in 1974; Faulkner and Dupuy, 2024; Dupuy and others, 2025), highlighting the effect of legacy agriculture in the study area (Grossinger and others, 2008). Previous studies of nitrogen and oxygen isotopes of nitrate concluded that applied fertilizer is the primary source of nitrogen in the Gilroy-Hollister Valley groundwater basin (Moran and others, 2005; Visser and others, 2020). Agricultural land use was 23 percent in the study area in 1974 (beige on fig. 4) and slightly decreased to 21 percent in 2012 (Falcone, 2015). Within this same period, nitrate fertilizer application rates on agricultural land in Santa Clara and San Benito counties increased by 51–69 percent (Gronberg, 2012). Because most of the agricultural land use in those two counties is in the Gilroy-Hollister Valley, we assumed that the nitrate fertilizer applications rate in the study area increased by a similar amount. Vegetables and other annual crops accounted for 70 percent of the mapped crop types grown in the study area (Grossinger and others, 2008; California Department of Water Resources, 2018); these crop types account for most of the nitrogen fertilizer application in California (Rosenstock and others, 2013).

High nitrate concentrations (greater than 10 mg-N/L) are more prevalent in the North San Benito subbasin than in the Llagas subbasin (fig. 4). The median nitrate concentrations in the North San Benito subbasin and Llagas subbasin were 7.1 mg-N/L and 3.7 mg-N/L, respectively. Nitrate concentrations are positively correlated with distance from the Gilroy-Hollister Valley groundwater basin boundary (green and blue outline on fig. 4; Spearman’s rho=0.53, p-value=0.02; California Department of Water Resources, 2021; Faulkner and Dupuy, 2024). This correlation indicates that wells in the middle of the groundwater basin are more likely to have recharge containing higher nitrate than wells near the groundwater basin margins and surrounding foothills. Most wells sampled in the North San Benito subbasin are farther from the groundwater basin margins (median distance to margin is 2,529 m) than wells sampled in the Llagas subbasin (median distance to margin is 401 m). Wells in the middle of the basin may be more likely to receive recharge with higher nitrate from upgradient agricultural areas. Similarly, wells closer to groundwater basin margins may receive more natural recharge with lower nitrate from creeks and streams draining the adjacent foothills. This finding is consistent with a previous study of nitrate in the Llagas subbasin, which determined that higher nitrate concentrations are in the downgradient parts of the subbasin (Moran and others, 2005) and a 2008 study of groundwater quality in the aquifer used for public supply where nitrate was moderate or high in most (7 of 12) wells in the Llagas subbasin that are downgradient from the wells in this study (Mathany and others, 2009; Parsons and others, 2014). Although nitrate concentrations in this study were low or not detected in wells with the top of the screened interval deeper than 21 m below land surface in the Llagas subbasin, the 2008 study of groundwater quality in the aquifer used for public supply reported that moderate and high concentrations in public supply wells are at greater depths (in wells with the top of the screened interval as deep as at least 110 m below the land surface). These results indicate that although nitrate has reached greater depths in the downgradient parts of the subbasin, it has not reached greater depths in the upgradient parts, and that position of a well relative to the groundwater basin margins has a greater effect on nitrate concentrations than the depth of the screened interval. The domestic supply wells farther from the groundwater basin margins may be more likely to contain moderate or high nitrate concentrations.

The distribution of groundwater age in a sample can be used to assess the degree of susceptibility to nitrate contamination (Tesoriero and others, 2007). Most (85 percent) domestic groundwater resources contain water recharged since the mid-1970s (based on Table_1_AgeInterpretations: Rpt_Age_C1_yrs of Faulkner and Jurgens, 2025). Within the Gilroy-Hollister Valley groundwater basin, groundwater age generally increases along lateral groundwater flow paths. Nitrate concentrations are positively correlated to the proportion of modern water captured by the well (Table_1_AgeInterpretations: Rpt_FracModern of Faulkner and Jurgens, 2025; Spearman’s rho=0.59, p-value less than 0.001), indicating that wells with higher proportions of modern water are more likely to contain higher concentrations of nitrate. Nitrate was not detected in domestic groundwater resources with premodern groundwater ages, which are primarily in the upland areas surrounding the Gilroy-Hollister Valley groundwater basin.

When redox conditions are anoxic, denitrification can remove dissolved nitrate in groundwater (Payne and Grant, 1981; Dubrovsky and others, 2010). As oxygenated water in equilibrium with the atmosphere moves into the subsurface, oxygen can be consumed by microbes, which leads to anoxic conditions. In this study, samples with dissolved oxygen (DO) greater than 0.5 mg/L were classified as “oxic,” and samples with DO less than or equal to 0.5 mg/L were classified as “anoxic” (fig. 5). About half (48 percent) of domestic groundwater resources in the study area were anoxic (Faulkner and Dupuy, 2024). Nitrate concentrations in groundwater will decrease during denitrification as nitrate is microbially transformed to nitrogen gas. This additional dissolved nitrogen gas exceeds the solubility of nitrogen in water in equilibrium with the atmosphere and is referred to as “excess nitrogen gas.” Excess nitrogen gas was quantified for 26 samples from measurements of dissolved noble gases using the computer program Dissolved Gas Modeling and Environmental Tracer Analysis (DGMETA; Jurgens and others, 2020; Table_2_DissolvedGasModeling: DG_ExcN2_mgL of Faulkner and Jurgens, 2025). The sum of nitrate and excess nitrogen gas (total nitrogen) is the theoretical nitrate concentration prior to the partial transformation to nitrogen gas. Half of the samples had minor amounts of excess nitrogen gas (less than 2 mg-N/L) that accounted for less than half of the total nitrogen. Total nitrogen results indicated that if denitrification did not happen, an additional 9 percent of domestic groundwater resources could have had high nitrate concentrations or 26 percent of domestic groundwater resources overall (fig. 6A). The correlation between total nitrogen and the percentage of agricultural land (Spearman’s rho=0.47, p-value=0.005 for land use in 1974) is stronger than the correlation between measured nitrate concentrations and the percentage of agricultural land (Spearman’s rho=0.44, p-value=0.01 for land use in 1974; Faulkner and Dupuy, 2024; Dupuy and others, 2025). Excess nitrogen gas accounted for more than 50 percent of the total nitrogen in 12 samples, indicating denitrification is an important process mitigating nitrogen inputs from the land surface (fig. 6A).

Excess nitrogen gas also can originate from deep mantle fluids moving through faults in the area (Mikhail and Sverjensky, 2014). Four samples that had high excess nitrogen gas (greater than 2 mg-N/L) also had high terrigenic helium and helium isotope ratios (greater than 1.65x10⁻⁶; Table_3_ComputedTracerConcentrations: TRC_He_terr_ccpg and 3He_4He_s of Faulkner and Jurgens, 2025), indicating that mantle fluids migrating through faults could be contributing excess nitrogen gas in these samples. However, the excess nitrogen gas in these samples was within the range of excess nitrogen gas in samples with low terrigenic helium and helium isotope ratios, indicating that denitrification was more likely the primary source of excess nitrogen gas.

Salts (TDS) can be concentrated in the water that recharges the aquifer as it evaporates at the surface in transportation or after application for irrigation, uptakes agricultural amendments, accumulates salts in the vadose zone, and increases dissolution and weathering of geologic sources. The relative proportions of TDS loading from agricultural and natural sources varies, where relative agricultural contribution generally decreases from northwest to southeast in the Llagas and North San Benito subbasins (Todd Groundwater, 2014a, 2014b). TDS concentrations are not significantly correlated with the percentage of agricultural land within 500 m of each well. Elevated concentrations of nitrate and TDS are frequently detected together in the North San Benito subbasin, where higher TDS is correlated with higher total nitrogen (Spearman’s rho=0.77, p-value=0.01; fig. 6B; Faulkner and Dupuy, 2024), indicating that agricultural recharge could be a source of TDS in the subbasin. Plots of the stable isotopes of water (deuterium [δ²H] and oxygen-18 [δ¹⁸O]) in the study area indicate a shift in isotopic values away from the global meteoric water line, denoting either a local meteoric water line, differences in climatic or recharge conditions, or the effect of evaporation (fig. 6C). Some of the surface water used for engineered recharge is evaporated, imported water, which can account for a shift off of the global meteoric water line; however, the stable isotope signature of the imported water (δ¹⁸O range from −10.6 to −9.9 per mil; Moran and others, 2005) is more depleted than in domestic groundwater resources (fig. 6C) and is therefore not considered a major contributor of TDS in domestic groundwater resources at the time of sampling.

The degree of evaporation was characterized by deuterium excess (d-excess, defined as δ²H−8δ¹⁸O; Dansgaard, 1964), where lower d-excess values indicate more fractionation through evaporation. TDS were negatively correlated with d-excess (Spearman’s rho=−0.8, p-value less than 0.001) in the North San Benito subbasin, which indicates that TDS had been affected by evaporation. Moreover, d-excess had a negative, although slightly weaker, correlation with TDS for all samples (Spearman’s rho=−0.57, p-value less than 0.001; fig. 6D), indicating that evaporated groundwater recharge has affected groundwater quality to varying degrees in domestic groundwater resources. Because premodern groundwater (containing more than 50 percent premodern water based on Table_1_AgeInterpretations: Rpt_FracModern of Faulkner and Jurgens, 2025) should reflect groundwater recharged under natural conditions, groundwater with d-excess values less than that of premodern groundwater (average d-excess value of 6.28 in domestic groundwater resources; dashed lines on fig. 6C and fig. 6D) are more likely to have been affected by evaporation related to agricultural practices with a modern, evaporative component. Most (88 percent) of the samples with high TDS concentrations had d-excess values less than that of premodern water in domestic groundwater resources. The uptake of agricultural amendments and salts from the vadose zone during irrigation and the addition of salt from imported water used for irrigation (Hansen and others, 2018; Pauloo and others, 2021) also contribute to TDS, particularly in the samples with moderate and high total nitrogen results (fig. 6D). Most (53 percent) of the samples with a modern, evaporative stable isotope signature had low or not-detected total nitrogen results, indicating that the dissolution and weathering of geologic sources likely contribute more TDS to domestic groundwater resources than agriculture (fig. 6D).

Geologic Factors

In the Gilroy-Hollister Valley groundwater basin and surrounding areas, the geology strongly affects groundwater TDS, redox conditions, and trace elements. Valley-fill sediments are derived from the surrounding Diablo Range, Gabilan Range, and Santa Cruz Mountains and contain abundant soluble marine sediments (Kilburn, 1973; Powell and others, 2007; fig. 7A). As water infiltrates through these sediments and recharges the aquifer, it can become naturally elevated in TDS greater than moderate levels (greater than 500 mg/L; Fram, 2017). In this study, 76 percent of domestic groundwater resources had moderate (greater than 500 mg/L) or high (greater than 1,000 mg/L) TDS concentrations (fig. 7A; Faulkner and Dupuy, 2024). Because groundwater TDS in the study area can be naturally elevated, evaporation can cause groundwater to reach higher TDS levels (mentioned earlier in the text; fig. 7A; Faulkner and Dupuy, 2024). The relations between TDS and the major anions (Cl⁻, bicarbonate [HCO₃⁻], and SO₄⁻²) that constitute TDS can be used to understand the mechanisms controlling groundwater chemistry (Piper, 1944). Moderate concentrations of TDS in many of the HCO₃⁻ dominated samples indicated that geology and the dissolution of carbonate from marine sediments is the primary factor contributing to TDS concentrations in the study area (fig. 6B). The higher relative concentrations of both Cl⁻ and SO₄⁻² in most of the high TDS samples indicate that a combination of evaporation and agricultural inputs further increases TDS in domestic groundwater resources (fig. 6B).

Natural geochemical processes can mobilize concentrations of barium, selenium, boron, and chromium from aquifer sediments (Chapman and others, 2013; Izbicki and others, 2015). Each of these constituents was detected at high concentrations in less than 10 percent of domestic groundwater resources and were not detected together, indicating these constituents were mobilized locally. Barium, selenium, and boron are associated with the marine sediments that are dominant in the study area (Deverel and Millard, 1988; Gonneea and Paytan, 2006) and sediments derived from the ultramafic rocks in the southern Coast Range are a dispersed source for chromium (Oze and others, 2003; Izbicki and others, 2015; fig. 7A).

Anoxic redox conditions in groundwater can attenuate nitrate, but anoxic conditions also can mobilize constituents like manganese, iron, and Cr(VI) from aquifer sediments (Tesoriero and others, 2007). Manganese and iron have SMCLs of 50 and 300 µg/L, respectively. Two-thirds of domestic groundwater resources had excess nitrogen gas (presumably from denitrification) greater than 0.5 mg/L, and 28 percent had manganese concentrations greater than the SMCL. One sample had a manganese concentration more than five times the notification level (500 µg/L; Faulkner and Dupuy, 2024); this single very high concentration does not indicate a regional issue. Mixed redox conditions are indicated by an oxic well with a high manganese concentration and by an anoxic well with low manganese and high nitrate concentrations (fig. 8). Of the 19 percent of domestic groundwater resources with chromium concentrations greater than the State of California MCL for Cr(VI) (10 µg/L), most (80 percent) were oxic; chromium is present as Cr(VI) under oxic redox conditions, which is more soluble in groundwater. As these results illustrate, redox reactions can have a positive effect on groundwater quality by attenuating nitrate concentrations through denitrification and a negative effect by mobilizing other trace elements.

Urban Factors

PFAS constituents have a broad range of applications and are persistent in the environment, which has led to their widespread distribution and the label “forever chemicals” (California State Water Resources Control Board, 2024c). Sources of PFAS in the study area are mainly collocated with urban land use (fig. 9; Falcone, 2015; Skinner and Wise, 2019; U.S. Environmental Protection Agency, 2024c; U.S. Geological Survey, 2024). Urban land use in the study area increased from 10 to 15 percent between 1974 and 2012 (Falcone, 2015). PFAS was detected in 19 percent of domestic groundwater resources, and the number of PFAS constituents detected in a well, including those with and without EPA MCL benchmarks, was significantly correlated to the percentage of urban land within 500 m of the well (Spearman’s rho=0.55, p-value=0.001 for land use in 2012; Faulkner and Dupuy, 2024; Dupuy and others, 2025). This moderate correlation reflects the nature of point-source contaminants in urban landscapes, where PFAS detection may only be associated with wells that are downgradient from PFAS contamination sources and other wells may be unaffected. The well with the highest number of PFAS constituents detected (S16-GH-08; n=6; app. 1, fig. 1.1) contained high concentrations of PFOA and PFOS, moderate concentrations of PFHxS and PFNA, and detections of PFBS and PFBA (Faulkner and Dupuy, 2024). This well is surrounded by predominantly agricultural land with 25-percent urban land (Dupuy and others, 2025). In this study, the detection frequency of PFAS was 55 percent (6 of 11) in wells that were surrounded by at least 25-percent urban land.

The primary sources of PFAS in drinking water are fire response, industrial use, and waste handling (landfills, wastewater treatment plants, and biosolids; California State Water Resources Control Board, 2024c). There are no known PFAS sources within 500 m of any of the wells that have PFAS detections (fig. 9; Skinner and Wise, 2019; U.S. Environmental Protection Agency, 2024c; U.S. Geological Survey, 2024). There are potentially many unknown point sources for PFAS constituents given the broad range of applications and widespread distribution of PFAS. Modern groundwater in a well can indicate greater susceptibility to anthropogenic contamination and a faster response to surface activities compared to wells that do not contain modern groundwater. Most (82 percent) samples, including all samples with PFAS detections, contain modern groundwater recharged after the proliferation of PFAS in manufacturing in the 1940s (U.S. Environmental Protection Agency, 2021; Faulkner and Jurgens, 2025). The number of PFAS constituents detected is weakly correlated with the proportion of modern water captured by the well (Table_1_AgeInterpretations: Rpt_FracModern in Faulkner and Jurgens, 2025; Spearman’s rho=0.4, p-value=0.02), reflecting that although all the wells with PFAS detections (n=6) contain almost completely modern water, half of the wells without PFAS detections do as well (n=13 of 26). Urban land use, locations of PFAS sources, and presence of modern water are important indicators of potential PFAS contamination, but site-specific characterization of groundwater flow paths and identification of additional potential PFAS sources are needed for a comprehensive understanding of PFAS detection in groundwater.