MODELING FECAL COLIFORM CONTAMINATION IN THE RIO GRANDE1

This study examines sources of fecal coliform in Segment 2302 of the Rio Grande, located south of the International Falcon Reservoir in southern Texas. The watershed is unique because the contributing drainage areas lie in Texas and Mexico. Additionally, the watershed is mostly rural, with populated communities known as “colonias.” The colonias lack sewered systems and discharge sanitary water directly to the ground surface, thus posing an increased health hazard from coliform bacteria. Monitoring data confirm that Segment 2302 is not safe for contact recreation due to elevated fecal coliform levels. The goal of the study was to simulate the observed exceedences in Segment 2302 and evaluate potential strategies for their elimination. Fecal coliform contributions from ranching and colonia discharges were modeled using the Hydrologic Simulation Program‐Fortran (HSPF). Model results indicated that the regulatory 30‐day geometric mean fecal coliform concentration of 200 colony forming units (cfu) per 100 milliliters is exceeded approximately three times per year for a total of 30 days. Ongoing initiatives to improve wastewater facilities will reduce this to approximately once per year for 14 days. Best management practices will be necessary to reduce cattle access to streams and eliminate all exceedences. The developed model was limited by the relatively sparse flow and fecal coliform data.     

ESTIMATING NUTRIENT AND SEDIMENT THRESHOLD CRITERIA FOR BIOLOGICAL IMPAIRMENT IN PENNSYLVANIA WATERSHEDS1

ABSTRACT: This study employs a simple nonlinear statistical approach to establish nitrogen, phosphorus, and sediment concentration and unit area load thresholds to aid in the evaluation of aquatic biological health of watersheds within the state of Pennsylvania. Flow, nitrogen and phosphorus species, sediment, basin area, land cover, and biological assessment data were assembled for 29 Pennsylvania watersheds. For each watershed, rating curves depicting flow versus load relationships were developed using the U.S. Environmental Protection Agency's (USEPA's) storage and retrieval database (STORET) flow and concentration data, then applied to daily flow data obtained from U.S. Geological Survey (USGS) daily flow gauging stations to estimate daily load between 1989 and 1999. The load estimates and concentration data were then sorted into six sets of data: mean annual unit area nitrogen, phosphorus, and sediment loads; and average nitrogen, phosphorus, and sediment concentrations. Results of Mann‐Whitney tests conducted on each of the six datasets indicate that there is a statistically significant difference between the concentrations and unit area loads of nitrogen, phosphorus, and sediment in impaired and unimpaired watersheds. Concentration thresholds, calculated as the midpoint between the impaired and unimpaired watersheds’ 95 percent confidence interval for the median, were estimated to be 2.01 mg/L, 0.07 mg/L, and 197.27 mg/L for nitrogen, phosphorus, and sediment, respectively. Annual unit area load thresholds were estimated to be equal to 8.64 kg/ha, 0.30 kg/ha, and 785.29 kg/ha, respectively, for nitrogen, phosphorus, and sediment species.     

MODELING THE BUILDUP AND WASHOFF OF POLLUTANTS ON URBAN WATERSHEDS1

ABSTRACT: A model for urban stormwater quality was developed in this study. The basis for the model is the process by which pollutants build up on the watershed surface. For the wet climate of the study site, it was assumed that there exists an interval of time over which the pollutant buildup equals the pollutant washoff (no accumulation of pollutant). The buildup model was represented by a linear function of the antecedent dry time. The buildup function was then linked with a pollutant washoff model represented by a power function of the storm runoff volume. Various time intervals for no net accumulation were tested to calibrate the model. The model was calibrated to observed data for two small urban basins in Baton Rouge, Louisiana, and model results were used to analyze the behavior of phosphorus concentrations in storm runoff from these basins over a long period of time.     

MODELING SEDIMENT AND PHOSPHORUS LOSSES IN AN AGRICULTURAL WATERSHED TO MEET TMDLs1

ABSTRACT: This paper studies the effectiveness of alternative farm management strategies at improving water quality to meet Total Maximum Daily Loads (TMDLs) in agricultural watersheds. A spatial process model was calibrated using monthly flow, sediment, and phosphorus (P) losses (1994 to 1996) from Sand Creek watershed in south‐central Minnesota. Statistical evaluation of predicted and observed data gave r² coefficients of 0.75, 0.69, and 0.49 for flow (average 4.1 m³/s), sediment load (average 0.44 ton/ha), and phosphorus load (average 0.97 kg/ha), respectively. The calibrated model was used to evaluate the effects of conservation tillage, conversion of crop land to pasture, and changes in phosphorus fertilizer application rate on pollutant loads. TMDLs were developed for sediment and P losses based on existing water quality standards and guidelines. Observed annual sediment and P losses exceeded these TMDLs by 59 percent and 83 percent, respectively. A combination of increased conservation tillage, reduced application rates of phosphorus fertilizer, and conversion of crop land to pasture could reduce sediment and phosphorus loads by 23 percent and 20 percent of existing loads, respectively. These reductions are much less than needed to meet TMDLs, suggesting that control of sediment using buffer strips and control of point sources of phosphorus are needed for the remaining reductions.     

ENHANCING FECAL COLIFORM TOTAL MAXIMUM DAILY LOAD MODELS THROUGH BACTERIAL SOURCE TRACKING1

ABSTRACT: Surface water impairment by fecal coliform bacteria is a water quality issue of national scope and importance. In Virginia, more than 400 stream and river segments are on the Commonwealth's 2002 303(d) list because of fecal coliform impairment. Total maximum daily loads (TMDLs) will be developed for most of these listed streams and rivers. Information regarding the major fecal coliform sources that impair surface water quality would enhance the development of effective watershed models and improve TMDLs. Bacterial source tracking (BST) is a recently developed technology for identifying the sources of fecal coliform bacteria and it may be helpful in generating improved TMDLs. Bacterial source tracking was performed, watershed models were developed, and TMDLs were prepared for three streams (Accotink Creek, Christians Creek, and Blacks Run) on Virginia's 303(d) list of impaired waters. Quality assurance of the BST work suggests that these data adequately describe the bacteria sources that are impairing these streams. Initial comparison of simulated bacterial sources with the observed BST data indicated that the fecal coliform sources were represented inaccurately in the initial model simulation. Revised model simulations (based on BST data) appeared to provide a better representation of the sources of fecal coliform bacteria in these three streams. The coupled approach of incorporating BST data into the fecal coliform transport model appears to reduce model uncertainty and should result in an improved TMDL.     

SUSCEPTIBILITY OF INDIANA WATERSHEDS TO HERBICIDE CONTAMINATION1

ABSTRACT: An index of watershed susceptibility to surface water contamination by herbicides could be used to improve source water assessments for public drinking water supplies, prioritize watershed restoration projects, and direct funding and educational efforts to areas where the greatest environmental benefit can be realized. The goal of this study is to use streamflow and herbicide concentration data to develop and evaluate a method for estimating comparative watershed susceptibility to herbicide loss. United States Geological Survey (USGS) concentration data for five relatively water soluble herbicides (alachlor, atrazine, cyanazine, metolachlor, and simazine) were analyzed for 16 Indiana watersheds. Correlation was assessed between observed herbicide losses and: (1) a herbicide runoff index using GIS‐based land use, soil type, SCS runoff curve number, tillage practice, herbicide use estimates, and combinations of these factors; and (2) predicted herbicide losses from a non‐point source pollution model (NAPRA‐Web, an Internet‐based interface for GLEAMS). The highest adjusted R²value was found between herbicide concentration and the runoff curve number alone, ranging from 0.25 to 0.56. Predictions from the simulation model showed a poorer correlation with observed herbicide loss. This indicates potential for using the runoff curve number as a simple herbicide contamination susceptibility index.     

MODELING ACID MINE DRAINAGE ON A WATERSHED SCALE FOR TMDL CALCULATIONS1

ABSTRACT: The Cheat River of West Virginia is impaired by acid mine drainage (AMD). Fifty‐five of its river segments were placed on the 303(d) list, which required calculations of total maximum daily load (TMDL) to meet the water quality criteria for pH, total iron, aluminum, manganese, and zinc. An existing watershed model was enhanced to simulate AMD as nonpoint source load. The model divided a watershed into a network of catchments and river segments. Each catchment was divided into soil layers, which could contain pyrite, calcite and other minerals. A kinetic expression was used to simulate pyrite oxidation as a function of oxygen in the soil voids. Oxygen in the soil voids was consumed by pyrite oxidation and replenished by earth breathing. The by‐products of pyrite oxidation were calculated according to its mass action equations. Chemical equilibrium was used to account for the speciation of ferrous and ferric irons and precipitation of metal hydroxides. Simulated hydrology and water quality were compared to available data. The USEPA used the calibrated model to calculate the TMDLs in the Cheat River Watershed.     

DETERMINATION OF BEST TIMING FOR POULTRY WASTE DISPOSAL: A MODELING APPROACH1

ABSTRACT: Confined production of poultry results in significant volumes of waste material which are typically disposed of by land application. Concerns over the potential environmental impacts of poultry waste disposal have resulted in ongoing efforts to develop management practices which maintain high quality of water downstream of disposal areas. The timing of application to minimize waste constituent losses is a management practice with the potential to ensure high quality of streams, rivers, and lakes downstream of receiving areas. This paper describes the development and application of a method to identify which time of year is best, from the standpoint of surface water quality, for land application of poultry waste. The procedure consists of using a mathematical simulation model to estimate average nitrogen and phosphorus losses resulting from different application timings, and then identifying the timings which minimize losses of these nutrients. The procedure was applied to three locations in Arkansas, and three different criteria for optimality of application timing were investigated. One criterion was oriented strictly to water quality, one was oriented only to crop production, and the last was a combination. The criteria resulted in different windows of time being identified as optimal. Optimal windows also varied with location of the receiving area. The results indicate that it is possible to land-apply poultry waste at times which both minimize nutrient losses and maximize crop yield.     

STATEWIDE EMPIRICAL MODELING OF BACTERIAL CONTAMINATION OF SURFACE WATERS1

ABSTRACT: Bacterial contamination of surface waters is attributed to both urban and agricultural land use practices and is one of the most frequently cited reasons for failure to meet standards established under the Clean Water Act (CWA) (P.L. 92–500). Statewide modeling can be used to determine if bacterial contamination occurs predominantly in urban or agricultural settings. Such information is useful for directing future monitoring and allocating resources for protection and restoration activities. Logistic regression was used to model the likelihood of bacterial contamination using watershed factors for the state of Maryland. Watershed factors included land cover, soils, topography, hydrography, locations of septic systems, and animal feeding operations. Results indicated that bacterial contamination occurred predominantly in urban settings. Likelihood of bacterial contamination was highest for small watersheds with well drained and erodible soils and a high proportion of urban land adjacent to streams. The number of septic systems and animal feeding operations and the amount of agricultural land were not significant explanatory factors. The urban infrastructure tends to “connect” more of the watershed to the stream network through the creation of roads, storm sewers, and wastewater treatment plants. This may partly explain the relationship between urbanization and bacterial contamination found in this study.     

PREDICTING FECAL COLIFORM BACTERIA LEVELS IN THE CHARLES RIVER,MASSACHUSETTS, USA1

In Massachusetts, the Charles River Watershed Association conducts a regular water quality monitoring and public notification program in the Charles River Basin during the recreational season to inform users of the river's health. This program has relied on laboratory analyses of river samples for fecal coliform bacteria levels, however, results are not available until at least 24 hours after sampling. To avoid the need for laboratory analyses, ordinary least squares (OLS) and logistic regression models were developed to predict fecal coliform bacteria concentrations and the probabilities of exceeding the Massachusetts secondary contact recreation standard for bacteria based on meteorological conditions and streamflow. The OLS models resulted in adjusted R²s ranging from 50 to 60 percent. An uncertainty analysis reveals that of the total variability of fecal coliform bacteria concentrations, 45 percent is explained by the OLS regression model, 15 percent is explained by both measurement and space sampling error, and 40 percent is explained by time sampling error. Higher accuracy in future bacteria forecasting models would likely result from reductions in laboratory measurement errors and improved sampling designs.     

Estimating Potential E. coli Sources in a Watershed Using Spatially Explicit Modeling Techniques1

Riebschleager, K.J., R. Karthikeyan, R. Srinivasan, and K. McKee, 2012. Estimating Potential E. coli Sources in a Watershed Using Spatially Explicit Modeling Techniques. Journal of the American Water Resources Association (JAWRA) 48(4): 745‐761. DOI: 10.1111/j.1752‐1688.2012.00649.x Abstract: The Spatially Explicit Load Enrichment Calculation Tool (SELECT) was automated to characterize waste and the associated pathogens from various sources within a mixed land use watershed. Potential Escherichia coli loads in Lake Granbury watershed were estimated using spatially variable governing factors, such as land use, soil condition, and distance to streams. A new approach for characterizing E. coli loads resulting from malfunctioning on‐site wastewater treatment systems (OWTSs) was incorporated into SELECT along with the Pollutant Connectivity Factor (PCF) module. The PCF component was applied to identify areas contributing E. coli loads during runoff events by incorporating the influence of potential E. coli loading, runoff potential, and travel distance to waterbodies. Simulation results indicated livestock and wildlife are potential E. coli contributing sources in the watershed. The areas in which these sources are potentially contributing are not currently monitored for E. coli. The bacterial water quality violations seen around Lake Granbury are most likely the result of malfunctioning OWTSs and pet wastes. SELECT results demonstrate the need to evaluate each contributing source separately to effectively allocate site specific best management practices (BMPs) utilizing stakeholder inputs. It also serves as a powerful screening tool for determining areas where detailed investigation is merited.     

ADDRESSING TOTAL PHOSPHORUS IMPAIRMENTS WITH WATER QUALITY TRADING1

ABSTRACT: Water quality trading is a voluntary economic process that provides an opportunity for dischargers to reduce the costs associated with meeting a discharge limitation. Trading can provide a cost effective solution for point sources (i.e., wastewater treatment plants) to meet strict effluent limitations set in response to total maximum daily loads (TMDLs). A successful trading program often depends on first determining the trading suitability of a pollutant for a particular watershed. A simple technical approach has been developed to identify sub‐watersheds within the Raritan River Basin, New Jersey, where water quality trading could provide a cost effective and scientifically feasible method for addressing total phosphorus impairments. The methodology presented will serve as a model to conduct similar analyses in other watersheds. The Raritan River Basin was divided into 12 subwatershed‐based study areas. Point‐nonpoint source trading opportunities were examined for each study area by examining the point and nonpoint source total phosphorus loading to impaired water bodies. Of the 12 subwatersheds examined, four had a high potential for implementing a successful trading program. Since instream phosphorus concentrations are closely related to soil erosion, an additional analysis was performed to examine soil erodibility. Recommendations are presented for conducting an economic analysis following the feasibility study.     

AN ANALYSIS OF SEASONAL FECAL COLIFORM LEVELS IN THE TCHEFUNCTE RWER1

ABSTRACT: A model for estimating seasonal fecal coliform concentrations in the Tchefuncte River as a function of river discharge was developed. Data on fecal coliform concentration were obtained from the Louisiana Department of Health and Hospitals and were available for a period of 15 years (1975 through 1992) from three locations. Stream flow data were obtained from a gaging station of the U. S. Geological Survey at Folsom, Louisiana. These data were available for 49 years (1943 through 1991). The climate of the area is characterized by different precipitation/runoff mechanisms for the summer and winter seasons. The division for seasons used in this analysis was May through October (summer season), and November through April (winter season). Because of the combined effects of climatic mechanisms causing precipitation and the seasonal variation of evapotranspiration, runoff is greater in the winter season resulting in higher fecal coliform counts in the Tchefuncte River. Statistical analysis revealed a statistically significant relationship between fecal coliform concentration and discharge for each season, at each of three sites on the Tchefuncte River.     

ANALYSIS OF WASTE LOAD ALLOCATION PROCEDURES1

A study was made to analyze and modify procedures used for stream assimilation capacity and point source wasteload allocation calculations. This paper describes the sources and types of information collected and the analysis of alternative computation methods developed during the study. The calculation of stream assimilation capacity or Total Maximum Daily Load (TMDL), will depend upon assumed stream flows, quality standards, reaction rates, and modeling procedures. The “critical conditions” selected for TMDL calculations usually are low flows and warm temperatures. The complexity of water quality models used for TMDL and allocation calculations can range from simple, complete mixing to calibrated and verified mathematical models. A list of 20 wasteload allocation (WLA) methods was developed. Five of these WLA's were applied to an example stream to permit comparisons based on cost, equity, efficient use of stream assimilation capacity, and sensitivity to fundamental stream quality data. Based on insensitivity to data errors and current use by several states, the WLA method of “equal percent treatment” was preferable in the example stream.     

ATRAZINE AND METOLACHLOR OCCURRENCE IN SHALLOW GROUND WATER OF THE UNITED STATES, 1993 TO 1995: RELATIONS TO EXPLANATORY FACTORS1

ABSTRACT: Since 1991, the U.S. Geological Survey has been conducting the National Water Quality Assessment (NAWQA) Program to determine the quality of the Nation's water resources. In an effort to obtain a better understanding of why pesticides are found in shallow ground water on a national scale, a set of factors likely to affect the fate and transport of two herbicides in the subsurface were examined. Atrazine and metolachlor were selected for this discussion because they were among the most frequently detected pesticides in ground water during the first phase of the NAWQA Program (1993 to 1995), and each was the most frequently detected compound in its chemical class (triazines and acetanilides, respectively). The factors that most strongly correlated with the frequencies of atrazine detection in shallow ground‐water networks were those that provided either: (1) an indication of the potential susceptibility of ground water to atrazine contamination, or (2) an indication of relative ground‐water age. The factors most closely related to the frequencies of metolachlor detection in ground water, however, were those that estimated or indicated the intensity of the agricultural use of metolachlor. This difference is probably the result of detailed use estimates for these compounds being available only for agricultural settings. While atrazine use is relatively extensive in nonagricultural settings, in addition to its widespread agricultural use, metolachlor is used almost exclusively for agricultural purposes. As a result, estimates of agricultural applications provide a less reliable indication of total chemical use for atrazine than for metolachlor. A multivariate analysis demonstrated that the factors of interest explained about 50 percent of the variance in atrazine and metolachlor detection frequencies among the NAWQA land‐use studies examined. The inclusion of other factors related to pesticide fate and transport in ground water, or improvements in the quality and accuracy of the data employed for the factors examined, may help explain more of the remaining variance in the frequencies of atrazine and metolachlor detection.     

NONPOINT SOURCE DISCHARGES OF NUTRIENTS FROM PIEDMONT WATERSHEDS OF CHESAPEAKE BAY1

ABSTRACT: We measured annual discharges of water, sediments, and nutrients from 10 watersheds with differing proportions of agricultural lands in the Piedmont physiographic province of the Chesapeake Bay drainage. Flow-weighted mean concentrations of total N, nitrate, and dissolved silicate in watershed discharges were correlated with the proportion of cropland in the watershed. In contrast, concentrations of P species did not correlate with cropland. Organic P and C correlated with the concentration of suspended particles, which differed among watersheds. Thus, the ratio of N:P:Si in discharges differed greatly among watersheds, potentially affecting N, P or Si limitation of phytoplankton growth in the receiving waters. Simple regression models of N discharge versus the percentage of cropland suggest that croplands discharge 29–42 kg N ha^-1 yr^-1 and other lands discharge 1.2–5.8 kg N ha^-1 yr^-1. We estimated net anthropogenic input of N to croplands and other lands using county level data on agriculture and N deposition from the atmosphere. For most of the study watersheds, N discharge amounted to less than half of the net anthropogenic N.     

Using Spatially Detailed Water‐Quality Data and Solute‐Transport Modeling to Support Total Maximum Daily Load Development1

Walton‐Day, Katherine, Robert L. Runkel, and Briant A. Kimball, 2012. Using Spatially Detailed Water‐Quality Data and Solute‐Transport Modeling to Support Total Maximum Daily Load Development. Journal of the American Water Resources Association (JAWRA) 48(5): 949‐969. DOI: 10.1111/j.1752‐1688.2012.00662.x Abstract: Spatially detailed mass‐loading studies and solute‐transport modeling using OTIS (One‐dimensional Transport with Inflow and Storage) demonstrate how natural attenuation and loading from distinct and diffuse sources control stream water quality and affect load reductions predicted in total maximum daily loads (TMDLs). Mass‐loading data collected during low‐flow from Cement Creek (a low‐pH, metal‐rich stream because of natural and mining sources, and subject to TMDL requirements) were used to calibrate OTIS and showed spatially variable effects of natural attenuation (instream reactions) and loading from diffuse (groundwater) and distinct sources. OTIS simulations of the possible effects of TMDL‐recommended remediation of mine sites showed less improvement to dissolved zinc load and concentration (14% decrease) than did the TMDL (53‐63% decrease). The TMDL (1) assumed conservative transport, (2) accounted for loads removed by remediation by subtracting them from total load at the stream mouth, and (3) did not include diffuse‐source loads. In OTIS, loads were reduced near their source; the resulting concentration was decreased by natural attenuation and increased by diffuse‐source loads during downstream transport. Thus, by not including natural attenuation and loading from diffuse sources, the TMDL overestimated remediation effects at low flow. Use of the techniques presented herein could improve TMDLs by incorporating these processes during TMDL development.     

Estimating Contributions of Nitrate and Herbicides From Groundwater to Headwater Streams,Northern Atlantic Coastal Plain,United States1

Abstract: Groundwater transport often complicates understanding of surface‐water contamination. We estimated the regional flux of nitrate and selected herbicides from groundwater to nontidal headwater streams of the Atlantic Coastal Plain (New Jersey through North Carolina) based on late‐winter or spring base‐flow samples from 174 streams. Sampled streams were selected randomly, and flux estimates are based on resulting population estimates rather than on empirical models, which have been used previously for similar estimates. Base‐flow flux in the estimated 8,834 headwater streams of the study area are an estimated 21,200 kg/day of nitrate (as N) and 5.83, 0.565, and 20.7 kg/day of alachlor, atrazine, and metolachlor (and selected degradates), respectively. Base‐flow flux of alachlor and metolachlor is <3% of the total base‐flow flux of those compounds plus degradates. Base‐flow flux of nitrate and herbicides as a percentage of applications is typically highest in well‐drained areas and lowest in areas with abundant poor drainage and anoxic conditions. In Coastal Plain watersheds of Albemarle and Pamlico Sounds, <2% of applied nitrogen reaches headwater streams as base flow. On the Delmarva Peninsula part of the Chesapeake Bay watershed, however, more than 10% of such applications are transported through groundwater to streams, and base‐flow nitrate flux represents 70% of total nitrogen flux in headwater streams.     

REEXAMINING BEST MANAGEMENT PRACTICES FOR IMPROVING WATER QUALITY IN URBAN WATERSHEDS1

ABSTRACT: Municipalities will be implementing structural best management practices at increasing rates in their effort to comply with Phase II of the National Pollutant Discharge Elimination System (NPDES). However, there is evidence that structural best management practices (BMPs) by themselves may be insufficient to attain desired water quality standards. This paper reports on an analysis of the median removal efficiencies of structural BMPs and compares them to removal efficiencies estimated as being necessary to attain water quality standards in the Rouge River in Detroit, Michigan. Eight water quality parameters are reviewed using data collected from 1994 to 1999 in the Rouge River. Currently, five of the eight parameters in the Rouge River including bacteria, biochemical oxygen demand, and total suspended solids (TSS) exceed the required water quality standards. The reported analysis of structural BMP efficiencies reveals that structural BMPs appear capable of reducing only some of the pollutants of concern to acceptable levels.     

MODELING CONCENTRATION VARIATIONS IN HIGH-CAPACITY WELLS: IMPLICATIONS FOR GROUNDWATER SAMPLING1

ABSTRACT: High-capacity wells are used as a convenient and economical means of sampling groundwater quality. Although the inherent limitations of using these wells are generally recognized, little has been done to investigate how these wells actually sample groundwater. A semi-analytical particle tracking model is used to illustrate the influence of variable vertical contaminant distributions and aquifer heterogeneity on the composition of water samples from these wells during short pumping periods. The hypothetical pumping well used in the simulations is located in an unconfined, alluvial aquifer with a shallow water table and concentration gradients of nitrate-nitrogen contamination. This is a typical setting for many irrigated areas in the United States. The main conclusions are: (1) high-capacity wells underestimate the average amount of contamination within an aquifer; (2) shapes of concentration-time curves for high-capacity wells appear to be governed by the distribution of the contaminant and travel times to the well; (3) variables such as well construction, pumping rate, and hydrogeologic properties contribute to the magnitude of the concentration-time curves at individual high-capacity wells; and (4) a sampling strategy using concentration-time curves based on the behavioral characteristics of the well rather than individual samples will provide a much better framework for interpreting spatial contaminant distributions.