Continuous Streamflow and Gage Height Measurements

Continuous (1-hour maximum interval) streamflow was measured at the Highway 50 and Sedgwick sites and continuous (1-hour maximum interval) gage height (stream stage) was measured at the Upstream ASR site. Streamflow has been measured at the Highway 50 and Sedgwick sites since May 1995 and November 1993, respectively. Gage height was measured at the Upstream ASR site since April 2011. Streamflow and gage height were measured using standard USGS methods (Sauer and Turnipseed, 2010; Turnipseed and Sauer, 2010; Painter and Loving, 2015).

Continuous Water-Quality Monitoring

Continuous (1-hour maximum interval) water-quality physicochemical properties were measured at the Highway 50, Sedgwick, and Upstream ASR sites. Water-quality physicochemical properties measured included water temperature, specific conductance, pH, dissolved oxygen, turbidity, and nitrate plus nitrite. Water-quality monitors were installed near the centroid of the stream cross section at the Highway 50 and Sedgwick sites and on the right streambank (facing downstream) at the Upstream ASR site. The Highway 50, Sedgwick, and Upstream ASR sites were equipped with a YSI Incorporated 6600 Extended Deployment System water-quality monitor (YSI Incorporated, 2012a) in May 1998, April 1998, and April 2011, respectively, to continuously (60-minute interval) measure water temperature, specific conductance, pH, dissolved oxygen (YSI Clark cell or optical dissolved oxygen sensors), and turbidity (YSI 6026 or 6136 turbidity sensors). Nitrate sensors (HACH Nitratax plus sc; HACH Company, 2014) were installed at the Highway 50, Sedgwick, and Upstream ASR sites in February 2017, March 2012, and March 2016, respectively.

Some equipment was upgraded throughout the duration of the project. YSI 6026 turbidity sensors were initially installed at the Highway 50 and Sedgwick sites in October 1998 and September 1998, respectively. YSI 6136 turbidity sensors were installed at the Highway 50, Sedgwick, and Upstream ASR sites in January 2007, July 2004, and April 2011, respectively. A YSI EXO2 water-quality monitor (YSI Incorporated, 2012b) equipped with water temperature, specific conductance, pH, dissolved oxygen, and YSI EXO turbidity sensors was installed in January 2017 at the Highway 50 site, in September 2014 at the Sedgwick site, and in October 2015 at the Upstream ASR site. YSI 6026 and YSI 6136 turbidity sensors were deployed concurrently at the Highway 50 and Sedgwick sites during July 2004 through August 2005 and July 2004 through January 2006, respectively. YSI 6136 and YSI EXO turbidity sensors were deployed concurrently at the Highway 50, Sedgwick, and Upstream ASR sites during January 2017 through May 2019, September 2014 through March 2015 (and again March 2017 through July 2019), and May 2017 through July 2019, respectively, to compare turbidity sensor measurements. Water-quality monitors were maintained following standard USGS procedures (Wagner and others, 2006; Pellerin and others, 2013; Bennett and others, 2014; Rasmussen and others, 2014).

Discrete Water-Quality Data Collection

Discrete surface-water-quality samples were collected at the three study sites across a range of streamflows. Samples were collected using primarily depth- and width-integrated sample collection techniques at the Highway 50 and Sedgwick sites and either single vertical or grab-dip techniques at the Upstream ASR site (U.S. Geological Survey, 2006; Rasmussen and others, 2014). Discrete water-quality constituent datasets used in this report were selected based on the following criteria:

Discrete water-quality samples were analyzed for primary ions (hardness, dissolved solids, calcium, magnesium, potassium, sodium, bromide, chloride, fluoride, silica, and sulfate), nutrient and carbon species (ammonia plus organic nitrogen, nitrate plus nitrite, total phosphorus, and dissolved and total organic carbon), sediment (suspended solids and suspended-sediment concentration), indicator bacteria (E. coli and fecal coliforms), trace elements (iron and arsenic), pesticides (including deethylatrazine, atrazine, and metolachlor), and absorbance and color at all study sites. Total nitrogen values were computed by summing ammonia plus organic nitrogen and nitrate plus nitrite values. Discrete samples collected at the Sedgwick site included additional nutrients (ammonia, nitrate, nitrite, orthophosphate, and dissolved phosphorus), strontium, and pesticides (including aminomethylphosphonic acid [AMPA] and glyphosate). Collection and analyses for dissolved and total suspended solids, suspended-sediment concentration, primary ions, nutrients, total organic carbon, arsenic, bacteria, and pesticides followed methods described by Ziegler and Combs (1997), Ziegler and others (1999, 2010), Stone and others (2012, 2016, 2019), Tappa and others (2015), and Stone and Klager (2022). Indicator bacteria analyses were done using methods described by the U.S. Environmental Protection Agency (2000a, 2006a, 2006b) and Myers and others (2014). AMPA and glyphosate were analyzed following methods described in Meyer and others (2009).

Primary ions, nutrients, organic carbon, and trace element samples were analyzed by the Wichita Municipal Water and Wastewater Laboratory and the USGS National Water Quality Laboratory (NWQL; Denver, Colorado). Suspended-sediment concentrations were analyzed at the USGS Iowa Sediment Laboratory (Iowa City, Iowa) following methods described in Guy (1969). Indicator bacteria samples were analyzed by the USGS Kansas Water Science Center (Lawrence, Kans.). Pesticides were analyzed by the USGS NWQL, with the exception of AMPA and glyphosate, which were analyzed by the USGS Kansas Organic Geochemistry Research Laboratory.

Continuous and Discrete Water Chemistry Data Analysis

Duration curves were used to summarize continuously measured or select computed Little Arkansas River water-quality data. Duration curves are cumulative distribution functions and were constructed using hourly values to evaluate and compare frequency and magnitude characteristics at the Highway 50, Sedgwick, and Upstream ASR sites during the study period. Duration curves are indicative of the percentage of time that specified conditions were equaled or exceeded, or the frequency of exceedance (Searcy, 1959). The Weibull formula (Weibull, 1939; Helsel and others, 2020) was used for plotting position.

Some water-quality constituents had censored values. Summary statistics for constituents that had censored values were calculated using regression on order statistics (Helsel and Cohn, 1988) for datasets with sample sizes less than (<) 50 and having as much as 80 percent censored data and for datasets with sample sizes greater than or equal to 50 and having as much as 50 percent censored data (Helsel, 2005; Bolks and others, 2014). Summary statistics for constituents that had censored values were calculated using maximum likelihood estimation (Helsel and Cohn, 1988) for datasets that had sample sizes greater than or equal to 50 and having between 50 and 80 percent censored data (Helsel, 2005; Bolks and others, 2014). Summary statistics were not computed when greater than 80 percent of data were censored. Statistical summaries were computed using the censtats function in R (version 4.1.3) programming language (R Core Team, 2021).

Surrogate Regression Model Development

Surrogate models were newly developed for constituents of interest at the three study sites and included models for bromide (all sites), dissolved organic carbon (all sites), total organic carbon (Upstream ASR site), 2-chloro-4-isopropylamino-6-amino-s-triazine (deethylatrazine; all sites), acetochlor (Upstream ASR site), AMPA (Sedgwick site), atrazine (Upstream ASR site), glyphosate (Sedgwick site), and metolachlor (all sites). Surrogate models for atrazine at the Highway 50 and Sedgwick sites were updated from previously published models (Stone and Klager, 2022). Surrogate models were not developed for water-quality constituents of interest that had datasets with greater than 20 percent left-censored data. Additional streamflow-based or gage-height-based models were developed to compute estimates of concentrations or densities during periods when concomitant continuous surrogate measurements were unavailable.

Regression models were developed using ordinary least squares (OLS) estimation for constituents that had datasets without left-censored data (“<” values). Tobit regression methods were used for fitting linear models for constituents that had datasets with left-censored data using absolute maximum likelihood estimation (AMLE; Hald, 1949; Cohen, 1950; Tobin, 1958; Helsel and others, 2020). Discrete datasets containing left-censored data included bromide, deethylatrazine, acetochlor, and metolachlor.

Model datasets and modeled constituents included available concomitant real-time physicochemical properties as explanatory variables during model development. Potential explanatory variables were evaluated individually and in combination and included available concomitant continuously recorded streamflow, water temperature, specific conductance, pH, dissolved oxygen, YSI EXO turbidity, and nitrate plus nitrite measurements. Periodic functions (seasonal components as sine and cosine variables) also were evaluated as potential explanatory variables using day of the year. Explanatory variables were interpolated within the continuous record based on discrete sample time. The maximum time span between two continuous data points used for interpolation was 2 hours.

Potential linear regression models were evaluated based on diagnostic statistics (coefficient of determination [R²], or adjusted R² for OLS-estimated models; pseudo-R² for AMLE-estimated models; Mallow’s C_p (Mallows, 1973) for OLS-estimated models; root mean square error for OLS-estimated models; prediction error sum of squares for OLS-estimated models; and residual standard error for AMLE-estimated models), patterns in residual plots, and the range and distribution of discrete and continuous data (Helsel and others, 2020). Models with the following criteria were selected:

If either a sine or a cosine seasonality variable was included in the model, both variables were included in the model. A bias correction factor was calculated for models with logarithmically transformed response variables because transformation of estimates to original units results in a low biased estimate (Duan, 1983; Helsel and others, 2020).

Potential outliers were identified following Rasmussen and others (2009) and Helsel and others (2020). Studentized residuals, leverage, Cook’s D (Cook, 1977), and difference in fit values were used to identify influential data points for OLS-estimated models; and leverage and Cook’s D values were used to identify influential data points for AMLE-estimated models. Studentized residuals were used to identify outliers with high leverage. Cook’s D is a combination of each observation’s leverage and residual value (large values indicate influential observations). Difference in fits is the product of the studentized residual and leverage (large values indicate influential observations). Removing data points that were based only on outlier criteria may overestimate the certainty of the model. Data points that exceeded Cook’s D and difference in fits thresholds for OLS-estimated models and Cook’s D thresholds for AMLE-estimated models were removed from model datasets to avoid erroneous inflation of model-computed values at the upper range of surrogate relations.

Continuous Water-Quality Constituent Concentration and Load Computation

Daily concentrations of selected primary ions (hardness, dissolved solids, calcium, sodium, bromide, chloride, and sulfate), nutrient and carbon species (ammonia plus organic nitrogen, total phosphorus, dissolved organic carbon, and total organic carbon), indicator bacteria (E. coli and fecal coliform bacteria), arsenic, sediment (suspended solids and suspended-sediment concentration), and pesticides (deethylatrazine, acetochlor, AMPA, atrazine, glyphosate, and metolachlor) were computed using previously developed (Christensen and others, 2003; Rasmussen and others, 2016; Stone and Klager, 2022) and newly developed models during 1998 through 2021. Daily mean values were used to compute water-quality constituent concentrations because of missing instantaneous streamflow data and periods of missing water-quality data.

Daily concentrations of hardness, dissolved solids, calcium, sodium, chloride, sulfate, and arsenic were computed at the Highway 50 and Sedgwick sites using surrogate models developed by Stone and Klager (2022). Water-quality constituent models that used turbidity measurements as a surrogate were developed using different turbidity sensors (YSI 6026, YSI 6136, or YSI EXO) and included ammonia plus organic nitrogen, total phosphorus, total organic carbon, E. coli bacteria, fecal coliform bacteria, suspended solids, and suspended-sediment concentration. Christensen and others (2003) developed models using continuous YSI 6026 turbidity measurements, Rasmussen and others (2016) developed surrogate relation models using continuous YSI 6136 turbidity measurements, and Stone and Klager (2022) developed surrogate relation models using continuous YSI EXO turbidity measurements at the Highway 50 and Sedgwick sites. Computed continuous concentrations or densities of ammonia plus organic nitrogen, total phosphorus, E. coli bacteria, fecal coliform bacteria, suspended solids, and suspended-sediment concentrations at the Highway 50 and Sedgwick sites were developed using a combination of previously published models by Christensen and others (2003; YSI 6026 turbidity sensor), Rasmussen and others (2016; YSI 6136 turbidity sensor), and Stone and Klager (2022; YSI EXO turbidity sensor) that used continuous turbidity sensor measurements. Computed continuous total organic carbon concentrations at the Highway 50 and Sedgwick sites were developed using a combination of previously published models by Rasmussen and others (2016; YSI 6136 turbidity sensor) and Stone and Klager (2022; YSI EXO turbidity sensor) according to continuous turbidity sensor measurements (Christensen and others [2003] did not include a total organic carbon model). Continuous total organic carbon at the Upstream ASR site was computed using the newly developed surrogate model and continuous YSI EXO turbidity sensor data. Continuous bromide, deethylatrazine, atrazine, and metolachlor concentrations were computed at the Highway 50, Sedgwick, and Upstream ASR sites using newly developed models. Continuous acetochlor concentrations were computed for the Upstream ASR site, and continuous AMPA and glyphosate concentrations were computed for the Sedgwick site. Previously published (Stone and Klager, 2022) and newly developed streamflow-based or gage-height-based surrogate models were used to compute continuous water-quality constituent concentrations when continuous water-quality data were unavailable. The streamflow-based or gage-height-based computed constituent concentrations were shifted to the primary water-quality values computed using a surrogate regression model directly preceding and following the streamflow-based or gage-height-based computed concentrations following Porterfield (1972) to align computed concentration differences among model types.

Loads were computed at the Highway 50 and Sedgwick sites for selected water-quality constituents, including those that have TMDLs. Chloride, nitrate plus nitrite, ammonia plus organic nitrogen, total phosphorus, dissolved organic carbon, total organic carbon, suspended solids, suspended-sediment concentration, E. coli bacteria, fecal coliform bacteria, and atrazine loads were computed during 1999 (the first full year of water-quality monitor measurements) through 2021. Water-quality constituent loads, with the exception of indicator bacteria, were calculated in tons per day. E. coli and fecal coliform bacteria were calculated in trillions of colony-forming units per day. Loads were computed by multiplying regression model-computed daily concentrations or densities (in milligrams per liter, micrograms per liter, or colony forming units per 100 milliliters) by daily streamflow (cubic feet per second) and a unit conversion factor of 0.0027 (for constituents measured in concentrations of milligrams per liter), 2.7x10⁻⁶ (for constituents measured in concentrations of micrograms per liter), or 2.45x10⁻⁵ (for constituents measured in densities of colony forming units per 100 milliliters). Annual loads were computed by summing the daily loads. Annual yields (in pounds or trillions of colony-forming units per square mile) were calculated by dividing the annual loads by the contributing drainage area at the Highway 50 (685 mi²) and Sedgwick (1,165 mi²) sites.

Water-Quality Constituent Trends

Water-quality constituent trends were analyzed at the Highway 50 and Sedgwick sites for discrete datasets that did not have more than 50 percent censored data (Hirsch and De Cicco, 2015) and were not associated with strong seasonality (for example, pesticides [Gilliom, 2007; Ryberg and York, 2020]) and included primary ions (hardness, dissolved solids, calcium, magnesium, potassium, sodium, bromide, chloride, fluoride, silica, and sulfate), nutrient and carbon species (ammonia [Sedgwick site only], nitrate plus nitrite, total nitrogen, nitrate [Sedgwick site only], nitrite [Sedgwick site only], orthophosphate [Sedgwick site only], dissolved phosphorus [Sedgwick site only], total phosphorus, dissolved organic carbon, and total organic carbon), sediment (suspended solids and suspended-sediment concentration), indicator bacteria (E. coli and fecal coliform bacteria), and trace elements. Water-quality constituent trends were analyzed using the R package Exploration and Graphics for RivEr Trends (EGRET; Hirsch and De Cicco, 2015). The Weighted Regressions on Time, Discharge, and Season model (WRTDS; Hirsch and others, 2010; Hirsch and De Cicco, 2015) was run using the EGRET package for long-term flow-normalized water-quality concentration and flux computations and trends. Annual water-quality concentrations and fluxes were estimated using WRTDS with Kalman filtering (WRTDS–K; described in Zhang and Hirsch, 2019). WRTDS–K generally produces more accurate concentration and flux results than the standard WRTDS method (Zhang and Hirsch, 2019). Lee and others (2019) tested 13 methods for computing annual water-quality constituent loads and determined WRTDS–K was the most accurate.

The Exploration and Graphics for RivEr Trends Confidence Intervals (EGRETci) package was used to quantify WRTDS model uncertainty using a WRTDS bootstrap test (WBT; Hirsch and others, 2015). Hirsch and others (2015) developed WBT trend likelihood descriptive statement definitions (table 2) to provide a method to classify water-quality trends into categories ranging from highly unlikely to highly likely. Water-quality concentrations, flow-normalized concentrations, fluxes, and flow-normalized fluxes were computed annually. Changes in annual flow-normalized concentrations and fluxes were computed at the Highway 50 and Sedgwick sites for three periods:

Ions

Primary dissolved ions come from decomposing rocks and soils and generally include calcium, magnesium, sodium, potassium, bicarbonate, carbonate, chloride, and sulfate ions (Hem, 1992). Primary ions are a constituent of concern in drinking water because large concentrations of ions may cause physiological effects, unpalatable mineral tastes, and greater costs owing to corrosion or additional treatment needs (U.S. Environmental Protection Agency, 2009). The EPA established National Secondary Drinking Water Regulations that set nonmandatory water-quality standards. The EPA does not enforce SMCLs because they were established as guidelines to assist public water-supply systems in managing their drinking water for aesthetic considerations such as taste, color, and odor (U.S. Environmental Protection Agency, 2009). Several primary ions have EPA SMCLs and include dissolved solids, chloride, fluoride, and sulfate (U.S. Environmental Protection Agency, 2009).

The EPA dissolved solids SMCL is 500 mg/L (U.S. Environmental Protection Agency, 2009), and discrete samples exceeded the SMCL at all three study sites (table 4). Dissolved solids concentrations exceeded the SMCL in 45, 29, and 23 percent of samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively. Mean dissolved solids concentrations during the study period ranged from 344 mg/L at the Upstream ASR site to 463 mg/L at the Highway 50 site (table 4). Dissolved solids ranges at the Highway 50 and Sedgwick sites did not exceed the ranges previously reported (Tappa and others, 2015; Stone and Klager, 2022; table 4). The dissolved solids minimum at the Upstream ASR site was 26 percent less than the minimum previously reported (Stone and others, 2016) and were in May 2019 (table 4).

Bromide is a constituent of concern for the Little Arkansas River. Bromide is not a risk to human health; however, during drinking-water decontamination, bromide reacts with natural organic matter present in source water and chemical disinfectants to create brominated disinfection byproducts, including bromate. The EPA has classified bromate (Federal MCL of 0.01 mg/L) as a probable human carcinogen (U.S. Environmental Protection Agency, 2009). Bromate is formed when ozone used to disinfect drinking water reacts with naturally occurring bromide in source water (U.S. Environmental Protection Agency, 2001b). The city of Wichita uses a threshold of 0.3 mg/L for bromide as a treatment standard. Bromide concentrations exceeded the treatment threshold of 0.3 mg/L at all three study sites (table 4). Bromide concentrations exceeded the treatment threshold in 38, 19, and 17 percent of samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively. Mean bromide concentrations at the Highway 50, Sedgwick, and Upstream ASR sites during the study period were 0.32, 0.18, and 0.17 mg/L, respectively (table 4). Bromide ranges at the Highway 50 and Sedgwick sites did not exceed the ranges previously reported (Tappa and others, 2015; Stone and others, 2019; table 4).

The EPA SMCL for chloride and sulfate is 250 mg/L (U.S. Environmental Protection Agency, 2009). Mean chloride concentrations during the study period at the Highway 50, Sedgwick, and Upstream ASR sites were 129, 67, and 61 mg/L, respectively (table 4). Little Arkansas River chloride concentrations exceeded the Federal SMCL at the Highway 50 and Sedgwick sites, but the SMCL was not exceeded at the Upstream ASR site during the study (table 4). Chloride concentrations exceeded the SMCL in 14 and <1 percent of samples at the Highway 50 and Sedgwick sites, respectively. Sulfate concentrations exceeded the Federal SMCL at the Highway 50 site in <1 percent of samples but not the Sedgwick or Upstream ASR sites during the study (table 4). Chloride concentration ranges during this study at the Highway 50 and Sedgwick sites did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022). The chloride minimum at the Upstream ASR site was less than the range previously reported (Stone and others, 2016) and were in May 2019 (table 4). Sulfate concentration ranges at the Highway 50 and Sedgwick sites did not exceed the previously reported range (Tappa and others, 2015). The sulfate minimum at the Upstream ASR site was less than the range previously reported (Stone and others, 2016) and were in May 2019 (table 4).

Nutrients and Carbon Species

Nutrients, such as nitrogen and phosphorus species, in water are closely related to agricultural activities because of their presence in fertilizers and animal waste. Nutrients are a pollutant of concern in the Little Arkansas River (Kansas State University Research and Extension and others, 2018). Nitrogen is present as ammonia, nitrate, nitrite, and as part of organic compounds. The EPA level III ecoregion 27 guideline for ammonia plus organic nitrogen (organic nitrogen; as nitrogen) is 0.52 mg/L (U.S. Environmental Protection Agency, 2001a) and was exceeded at all study sites (table 4). Ammonia plus organic nitrogen concentrations exceeded the guideline in 92, 90, and 98 percent of samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively. During the study period, mean ammonia plus organic nitrogen (as nitrogen) concentrations ranged from 1.4 mg/L (Sedgwick site) to 1.8 mg/L (Highway 50 site; table 4). Ammonia plus organic nitrogen values at the Highway 50 site did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022). Ammonia plus organic nitrogen values at the Sedgwick and Upstream ASR sites exceeded previously reported ranges (Stone and others, 2016; Stone and Klager, 2022; table 4) by 3 and 18 percent, respectively, and maximum values at both sites were in February 2020. Mean nitrate plus nitrite (as nitrogen) concentrations ranged from 0.87 mg/L (Highway 50 site) to 1.2 mg/L (Sedgwick site; table 4). Nitrate plus nitrite values during the study period at the Highway 50 and Sedgwick sites did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022), but the Upstream ASR site minimum did (Stone and others, 2016; table 4) in May 2019.

The EPA level III ecoregion 27 guideline for total phosphorus of 0.09 mg/L (U.S. Environmental Protection Agency, 2001a) was exceeded in all samples at all three study sites with the exception of one sample at the Sedgwick site (table 4). Discrete sample total phosphorus (as phosphorus) means were 0.70 at the Highway 50 and Sedgwick sites and 0.80 mg/L at the Upstream ASR site for the study period (table 4). Total phosphorus values did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022) at the Highway 50 and Sedgwick sites, but did exceed previously reported minimum and maximum values (Stone and others, 2016) by 0.20 and 0.04 mg/L, respectively, at the Upstream ASR site (table 4). The minimum total phosphorus value (0.29 mg/L) at the Upstream ASR site was in December 2019 and the maximum (1.5 mg/L) was in May 2021.

An important artificial recharge concern is the introduction of potentially reactive organic material into an aquifer (Aiken, 2002). Mean dissolved organic carbon concentrations were 7.9, 6.6, and 7.7 mg/L at the Highway 50, Sedgwick, and Upstream ASR study sites, respectively, during the study period (table 4). Dissolved organic carbon values did not exceed previously reported ranges (Tappa and others, 2015; Stone and others, 2019) at the Highway 50 and Sedgwick sites, but did exceed previously reported minimum and maximum values (Stone and others, 2016) by 0.80 and 0.70 mg/L, respectively, at the Upstream ASR site (table 4). The minimum dissolved organic carbon value (2.9 mg/L) was in January 2018 and the maximum (14 mg/L) was in June 2016 at the Upstream ASR site. Mean total organic carbon concentrations were 13 mg/L at the Highway 50 site and 12 mg/L at the Sedgwick and Upstream ASR sites (table 4). Total organic carbon concentrations during the study period at the Highway 50 site did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022). The maximum total organic carbon value (36 mg/L) at the Sedgwick site exceeded the range of previously reported values (Tappa and others, 2015; Stone and Klager, 2022) by 4.0 mg/L and were in February 2020. The minimum and maximum total organic carbon values at the Upstream ASR site exceeded the range of previously reported values (Stone and others, 2016) by 1.6 and 5.4 mg/L, respectively, and were in August 2019 and January 2018, respectively (table 4).

Sediment

Sediment is a naturally occurring constituent that is also a pollutant of concern in the Little Arkansas River (Kansas State University Research and Extension and others, 2018). Suspended solids and suspended-sediment concentration are two analytes typically used to describe concentrations of suspended solid-phase material in surface water. Suspended-sediment concentration is a better measure for natural water than suspended solids (Gray and others, 2000), but suspended solids are required for TMDLs. Suspended solids represent suspended solids material and may consist of organic or inorganic material. Suspended solids originate from sources that include algae, decaying vegetation, runoff, discharges (for example, industrial or wastewater), and physical degradation of geologic formations. The amount of suspended solids in a sample is equal to the dry weight of organic and inorganic solids filtered from a subsample of the original sample. Suspended-sediment concentration is the measure of the dry weight of the organic and inorganic sediment in a full sample volume of a water-sediment mixture (Guy, 1969). Suspended sediment may consist of clay, silt, sand, or organic material. Mean Little Arkansas River suspended-solids concentrations were 264, 219, and 184 mg/L at the Highway 50, Sedgwick, and Upstream ASR sites, respectively, during the study period (table 4). Suspended-solids concentrations during the study period did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022) at the Highway 50 and Sedgwick sites. The maximum suspended-solids value at the Upstream ASR site exceeded the previously reported value (Stone and others, 2016) by 357 mg/L and was in May 2021 (table 4). Mean suspended-sediment concentrations at the Highway 50, Sedgwick, and Upstream ASR sites were 400, 232, and 246 mg/L, respectively (table 4). Suspended-sediment concentration during the study period at the Highway 50 and Sedgwick sites did not exceed previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022). The maximum suspended-sediment concentration at the Upstream ASR site exceeded the previously reported value (Stone and others, 2016) by 706 mg/L and was in May 2021 (table 4).

Indicator Bacteria

E. coli and fecal coliform bacteria are types of coliform bacteria that are generally specific to fecal material from humans and other homeotherms and are commonly used as pathogen indicators (Myers and others, 2014). E. coli and fecal coliform bacteria presence indicates potentially contaminated water by human or animal wastes and presence of other harmful bacteria or viruses (Dufour and others, 1981; Dufour, 1984). The State of Kansas E. coli criteria for publicly accessible (Class B) Kansas streams with flows of at least 1 ft³/s require that the geometric mean of at least five samples collected during separate 24-hour periods within a 30-day period not exceed 262 colony forming units per 100 milliliters (cfu/100 mL) for primary contact during April 1 through October 31 of each year and 2,358 cfu/100 mL during November 1 through March 31 (Kansas Department of Health and Environment, 2011, 2017, 2018). Little Arkansas River E. coli and fecal coliform medians ranged from 540 to 640 cfu/100 mL and 590 to 910 cfu/100 mL, respectively, at the three study sites during the study period (table 4). E. coli densities exceeded the primary contact Class B criteria of 262 cfu/100 mL at all sites (table 4). E. coli densities exceeded the primary contact Class B criteria of 262 cfu/100 mL in 67, 60, and 62 percent of samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively. E. coli densities at the Highway 50 site were within ranges previously reported (Tappa and others, 2015; Stone and Klager, 2022); the maximum E. coli densities at the Sedgwick and Upstream ASR sites exceeded previously reported ranges (Stone and others, 2016; Stone and Klager, 2022) by 1,300 and 11,000 cfu/100 mL, respectively, and were in October 2021 (table 4). Fecal coliform densities at the Highway 50 and Sedgwick sites during the study were within previously reported ranges (Tappa and others, 2015; Stone and Klager, 2022) during the study (table 4). The maximum fecal coliform density at the Upstream ASR site exceeded the maximum previously reported value (Stone and others, 2016; table 4) by 16,000 cfu/100 mL and were in June 2016.

Trace Elements

Dissolved concentrations of particular interest in the Little Arkansas River include iron and arsenic. Iron in water is derived from rocks and soils (Hem, 1992). Excessive concentrations of iron in water cause unpalatability because of odor, a metallic taste, and rusty color. The EPA Federal SMCL for iron is 300 µg/L (U.S. Environmental Protection Agency, 2009) and was exceeded in 5, 7, and 9 percent of samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively, during the study period (table 4). Mean study site iron concentrations ranged from 63 to 137 µg/L (table 4). Iron concentrations at the Highway 50 and Sedgwick sites did not exceed the ranges previously reported (Tappa and others, 2015; Stone and others, 2019). The maximum iron concentrations at the Upstream ASR site exceeded the previously reported range (Stone and others, 2016) by 807 µg/L and were during August 2019 (table 4).

Arsenic is present naturally in clay layers associated with iron sulfide minerals (Hem, 1992) and is a health concern in drinking water because it causes skin damage, affects the circulatory system, and increases the risk of cancer (U.S. Environmental Protection Agency, 2009). The EPA Federal MCL for arsenic is 10 µg/L (U.S. Environmental Protection Agency, 2009) and was exceeded in 10 percent of samples at the Highway 50 and Sedgwick site and 25 percent of samples at the Upstream ASR site (table 4). Mean arsenic concentrations were 5.4, 5.8, and 7.7 µg/L at the Highway 50, Sedgwick, and Upstream ASR sites during the study period (table 4). Arsenic concentrations were within the ranges previously reported (Tappa and others, 2015; Stone and others, 2019) at the Highway 50 and Sedgwick sites (table 4). The minimum arsenic concentration was 1.6 µg/L less than the range previously reported (Stone and others, 2016; table 4) at the Upstream ASR site and was in May 2019.

Pesticides

Pesticides enter streams from field application and through irrigation return flow and surface runoff in agricultural landscapes. Pesticides that were detected in greater than one-half of the collected samples at any of the three study sites included 2-chloro-4-isopropylamino-6-amino-s-triazine (deethylatrazine, a degradation product of the atrazine molecule), acetochlor (applied herbicide), AMPA (a chemical that results from the breakdown of glyphosate), atrazine (applied herbicide), desulfinylfipronil (a degradation product of the fipronil [applied insecticide] molecule), glyphosate (applied herbicide), metolachlor (applied herbicide), and prometon (applied herbicide; table 4). Mean deethylatrazine concentrations during the study period were 0.350, 0.394, and 0.297 µg/L at the Highway 50, Sedgwick, and Upstream ASR sites, respectively, and maxima did not exceed previously reported ranges (Stone and others, 2019; table 4). Mean acetochlor concentrations at the Highway 50, Sedgwick, and Upstream ASR sites were 0.321, 0.380, and 0.982 µg/L, respectively (table 4). Acetochlor maxima exceeded the ranges previously reported (Stone and others, 2016, 2019) by 3.01, 12.2, and 6.14 µg/L at the Highway 50, Sedgwick, and Upstream ASR sites, respectively. AMPA and glyphosate were analyzed for the Sedgwick site only, were detected in all of the samples collected during the study, and mean concentrations were 1.44 and 0.970 µg/L, respectively (table 4). The AMPA maxima exceeded the range previously reported (Stone and others, 2019) by 1.6 µg/L, but glyphosate concentrations were within previously reported ranges (Stone and others, 2019; table 4). Mean study period metolachlor concentrations were 2.08 µg/L at the Highway 50 site and 1.77 µg/L at the Sedgwick and Upstream ASR sites (table 4). Metolachlor maxima at the Highway 50, Sedgwick, and Upstream ASR sites exceeded the ranges previously reported (Stone and others, 2016, 2019; table 4) by 24.4, 0.9, and 5.28 µg/L, respectively.

Atrazine is an herbicide commonly used on corn and sorghum, which are crops grown in the study area. Atrazine can cause cardiovascular system or reproductive problems in humans (U.S. Environmental Protection Agency, 2009). Previous (2010 through 2014 and 2015 through 2019) NPDES permits (Kansas permit number: I-LA24-PO01; Federal permit number: KS0099694) for the ASR phase II treatment facility required that atrazine be monitored monthly while the facility was operating. The EPA MCL for atrazine is 3.0 µg/L (U.S. Environmental Protection Agency, 2009) and was exceeded in 34, 36, and 37 percent of discrete samples at the Highway 50, Sedgwick, and Upstream ASR sites, respectively (table 4). Mean atrazine concentrations were 3.89, 3.55, and 3.47 µg/L during the study period at Highway 50, Sedgwick, and Upstream ASR sites, respectively. Atrazine concentrations during the study were within previously reported ranges (Tappa and others, 2015; Stone and others, 2016; Stone and Klager, 2022) at all study sites (table 4).

Primary Ions

The primary components of dissolved solids come from decomposing rocks and soils. Concentrations of dissolved solids in streams may increase because of atmospheric deposition, sewage inputs, industrial effluent, and agricultural and urban runoff (Hem, 1992). Dissolved solids often are used as a general indicator of salinity or water quality, and large concentrations are undesirable in drinking water because of possible physiological effects, strong mineral tastes, increased treatment costs, and corrosion of plumbing (U.S. Environmental Protection Agency, 2009). Dissolved solids commonly exceeded the EPA SMCL in the Equus Beds aquifer (Ziegler and others, 2010; Tappa and others, 2015; Stone and others, 2019). Computed primary components of dissolved solids (hardness, calcium, sodium, bromide, and chloride), with the exception of sulfate, were larger at the Highway 50 site compared to the Sedgwick site (table 6; fig. 3A–D). Computed dissolved solids concentrations during the study period ranged from 60 to 1,592 mg/L at the Highway 50 site and 53 to 929 mg/L at the Sedgwick site; mean computed dissolved solids concentrations were 546 and 447 mg/L at the Highway 50 and Sedgwick site, respectively (table 6). During the study period, the EPA SMCL for dissolved solids of 500 mg/L was exceeded 57 percent of the time at the Highway 50 site and 38 percent of the time at the Sedgwick site (fig. 3A).

Bromide is a constituent of interest because of its potential for conversion to brominated organics of concern through drinking-water treatment processes. Bromate, a known carcinogen, can form during the disinfection of bromide-containing water that involves an oxidation process to kill pathogens (Xie, 2003; U.S. Environmental Protection Agency, 2009). Mean continuously computed bromide concentration during the study period was 0.44, 0.23, and 0.22 mg/L at the Highway 50, Sedgwick, and Upstream ASR sites, respectively (table 6). Computed bromide concentrations were larger at the Highway 50 site and exceeded the city of Wichita treatment threshold of 0.3 mg/L about 70, 21, and 19 percent of the time at the Highway 50, Sedgwick, and Upstream ASR sites, respectively (fig. 3B).

Chloride is of particular concern in the Equus Beds aquifer because of the chloride plume upgradient from the city of Wichita’s well field, which was estimated to be moving toward the well field at a rate of as much as 400 feet per year from 1990 through 2008. The plume is expected to continue moving toward the well field regardless of pumping in the area (Klager and others, 2014). Chloride concentrations significantly increased in two shallow monitoring wells near the phase II recharge basin after the onset of artificial recharge, and the increases likely were caused by treated and artificially recharged surface water (Stone and others, 2016). Large chloride concentrations (100 to 500 mg/L) are common in the western part of the study area and along the Arkansas River (fig. 1; Ziegler and others, 2010; Klager and others, 2014; Tappa and others, 2015; Stone and others, 2019). Mean Arkansas River chloride concentration was about 600 mg/L during 1988 through 1991 (Myers and others, 1996) and about 500 mg/L during 1997 through 2006 between Hutchinson and Maize, Kans. (Kansas Department of Health and Environment, 2006a). Little Arkansas River chloride sources include contamination from past oil and gas activity near McPherson, Kans., and industrial and municipal waste-water discharges (Leonard and Kleinschmidt, 1976; Kansas Department of Health and Environment, 2006a; Schmidt and others, 2007; Whittemore, 2007). Continuously computed chloride concentrations during the study period were larger at the Highway 50 site (mean: 156 mg/L) compared to the Sedgwick site (mean: 81 mg/L; table 6; fig. 3C). Chloride concentrations exceeded the EPA SMCL of 250 mg/L about 16 percent of the time at the Highway 50 site and did not exceed the SMCL at the Sedgwick site (table 6; fig. 3C).

Natural sources of sulfate in surface water and groundwater are rock weathering, oxidation of sulfide minerals, and biological processes (Hem, 1992). Anthropogenic sources of sulfate in water include atmospheric deposition from coal and petroleum combustion products (Hem, 1992) and irrigation return flows. Mean sulfate concentrations during the study period were 42 and 55 mg/L at the Highway 50 and Sedgwick sites, respectively (table 6). Computed sulfate concentrations never exceeded the EPA SMCL of 250 mg/L (table 6; fig. 3D).

Nutrients and Carbon

Nutrients, particularly nitrogen and phosphorus, have been identified as a primary cause of water-quality and biological degradation in the Midwest and the Nation (U.S. Environmental Protection Agency, 2000b, 2006c; Kansas Department of Health and Environment, 2004; Dubrovsky and others, 2010; Munn and others, 2018). Dissolved and particulate organic carbon are primary food sources for aquatic food webs. Organic carbon is of interest for ASR operation because of its role in THM formation during the water treatment process. THMs are disinfection byproducts that are formed when naturally existing inorganic and organic materials in water react with the disinfectants chlorine and chloramine (U.S. Environmental Protection Agency, 2005). THMs can cause liver, kidney, and central nervous system problems and are carcinogens (Pyne and others, 1996; U.S. Environmental Protection Agency, 2005). Excessive organic carbon in recharge water may lead to arsenic mobilization in groundwater owing to decreased oxygen from microbial respiration of organic carbon (U.S. Environmental Protection Agency, 2023).

Computed ammonia plus organic nitrogen concentrations were similar at the Highway 50 and Sedgwick sites during the study (table 6; fig. 4A). Mean computed ammonia plus organic nitrogen concentrations were 1.10 mg/L at the Highway 50 site and 1.14 mg/L at the Sedgwick site during the study period (table 6). Ammonia plus organic nitrogen concentrations exceeded the EPA level III ecoregion 27 guideline of 0.52 mg/L about 85 percent of the time at the Highway 50 site and 93 percent of the time at the Sedgwick site (fig. 4A). Total phosphorus concentrations were generally larger at the Sedgwick site (table 6; fig. 4B). Mean continuously computed total phosphorus concentrations were 0.47 mg/L at the Highway 50 site and 0.61 mg/L at the Sedgwick site (table 6). Total phosphorus concentrations during the study period exceeded the EPA level III ecoregion 27 guideline of 0.09 mg/L nearly 100 percent of the time at the Highway 50 site and 100 percent of the time at the Sedgwick site (fig. 4B). Continuous total organic carbon concentrations were generally larger at the Highway 50 site (table 6, fig. 4C). Mean total organic carbon concentrations were 9.2 mg/L at the Highway 50 site, 8.3 mg/L at the Sedgwick site, and 8.1 mg/L at the Upstream ASR site (table 6).

Sediment

Continuously computed suspended solids and suspended-sediment concentration ranges and means were similar between the Highway 50 and Sedgwick sites during the study (table 6; fig. 5). Mean suspended solids concentrations were 101 mg/L at the Highway 50 site and 94 mg/L at the Sedgwick site (table 6). Mean suspended-sediment concentrations were 134 and 129 mg/L at the Highway 50 and Sedgwick sites, respectively (table 6).

Indicator Bacteria

Continuously computed E. coli and fecal coliform bacteria density means were larger at the Sedgwick site during the study period (table 6, fig. 6A). Median study period E. coli densities were 330 cfu/100 mL at the Highway 50 site and 230 cfu/100 mL at the Sedgwick site (table 6). E. coli densities exceeded the Kansas Department of Health and Environment primary contact criterion for Class B recreation of 262 cfu/100 mL during April 1 through October 31 (Kansas Department of Health and Environment, 2017) 84 percent of the time at the Highway 50 site and 62 percent of the time at the Sedgwick site during the months of April through October (fig. 6B). E. coli densities exceeded the Kansas Department of Health and Environment primary contact criterion for Class B recreation during November 1 through March 31 and secondary contact criterion for Class A recreation of 2,358 cfu/100 mL 5 percent of the time at the Highway 50 site and 6 percent of the time at the Sedgwick site during the months of November through March (fig. 6C). Median fecal coliform bacteria densities were 420 and 310 cfu/100 mL at the Highway 50 and Sedgwick sites, respectively (table 6).

Arsenic mobility is generally controlled by adsorption and desorption reactions and solid-phase precipitation and dissolution reactions (Hem, 1992; Hinkle and Polette, 1999; Smedley and Kinniburgh, 2002; McMahon and Chapelle, 2008); these processes are affected by pH, oxidation/reduction reactions, and competing anion presence, all of which could be altered because of artificial recharge activities. The EPA Federal MCL for arsenic is 10 µg/L (U.S. Environmental Protection Agency, 2009). Arsenic is a constituent of concern in the Equus Beds aquifer and commonly is present in concentrations that exceed the MCL (Ziegler and others, 2010; Tappa and others, 2015; Stone and others, 2019). Dissolved arsenic concentrations in two shallow monitoring wells near the ASR phase II recharge basin increased significantly after the onset of phase II ASR activity (Stone and others, 2016). Continuously computed arsenic concentrations were generally similar between the Highway 50 and Sedgwick sites during the study period (table 6; fig. 7A). Mean arsenic concentrations were 6.8 µg/L at the Highway 50 site and 6.4 µg/L at the Sedgwick site (table 6). Continuous arsenic concentrations exceeded the Federal MCL 15 percent of the time at the Highway 50 site and 9 percent of the time at the Sedgwick site (fig. 7A).

Atrazine and glyphosate are pesticides that are commonly detected at the Little Arkansas sites (Ziegler and others, 2010; Tappa and others, 2015; Stone and others, 2016, 2019; Stone and Klager, 2022). The EPA MCL is 3.0 µg/L for atrazine and 700 µg/L for glyphosate, and both compounds can cause reproductive difficulties in humans (U.S. Environmental Protection Agency, 2009). Continuously computed atrazine concentrations were largest at the Upstream ASR site (table 6; fig. 7B). Mean computed atrazine concentrations at the Highway 50, Sedgwick, and Upstream ASR sites were 1.10 µg/L, 1.12 µg/L, and 1.49 µg/L, respectively (table 6). Atrazine concentrations exceeded the MCL 10 percent of the time at the Highway 50 and Sedgwick sites and 14 percent of the time at the Upstream ASR site during the study (fig. 7B). Computed glyphosate at the Sedgwick site ranged from 0.108 to 5.81 µg/L during the study period and never exceeded the MCL (table 6; fig. 7C).

Primary Ion Trends

Little Arkansas River flow-normalized primary ion concentrations during 1995 through 2021 (fig. 9A–K) generally had downward trends based on categories defined by Hirsch and others (2015; table 2); percent decreases in flow-normalized concentrations were generally larger at the Highway 50 site (table 8). Dissolved solids, potassium, sodium, chloride, and fluoride flow-normalized concentrations (fig. 9B, E, F, H, I) were very likely or highly likely to have downward trends at both sites (table 9). Dissolved solids mean annual concentrations decreased by 30 percent (Highway 50 site) and 15 percent (Sedgwick site); sodium concentrations decreased by 49 percent (Highway 50 site) and 32 percent (Sedgwick site); and chloride concentrations decreased by 54 percent (Highway 50 site) and 43 percent (Sedgwick site) during the study period (table 9). Hardness, calcium, and magnesium flow-normalized concentrations (fig. 9A, C, D) had downward trends at the Highway 50 site and no trend at the Sedgwick site for the entire study period (table 9). Bromide and silica flow-normalized concentrations (fig. 9G, J) exhibited no trend at the Highway 50 site and downward trends at the Sedgwick site from 1995 through 2021 (table 9). Bromide concentrations decreased by 68 percent at the Highway 50 site and 45 percent at the Sedgwick site during the study period (table 9). Flow-normalized sulfate concentrations (fig. 9K) during the entire study period exhibited a downward trend (and a decrease of 10 percent) at the Highway 50 site and an upward trend (and an increase of 7 percent) at the Sedgwick site (table 9).