Occurrence of organic wastewater compounds in wastewater effluent and the Big Sioux River in the Upper Big Sioux River basin, South Dakota, 2003-2004

The U.S. Geological Survey (USGS) in cooperation with the East Dakota Water Development District conducted a reconnaissance study to determine the occurrence of organic wastewater compounds (OWCs) in wastewater effluent and the Big Sioux River at or near the cities of Watertown, Volga, and Brookings in the upper Big Sioux River Basin during August 2003 through June 2004. For each city, samples were collected from the wastewater treatment plant (WWTP) effluent and from Big Sioux River sites upstream and downstream from where the wastewater effluent discharges to the Big Sioux River. For Watertown and Brookings, samples were collected during a low-flow period (August 2003) and a runoff period (June 2004). For Volga, samples were collected during two low-flow periods (August 2003 and October 2003) and a runoff period (June 2004).

A total of 125 different OWCs were analyzed for and were classified into six compound classes-human pharmaceutical compounds (HPCs), human and veterinary antibiotic compounds (HVACs), major agricultural herbicides (MAHs), household, industrial, and minor agricultural compounds (HIACs), polyaromatic hydrocarbons (PAHs), and sterol compounds (SCs). Of the 125 different OWCs, 45 had acceptable analytical method performance, were detected at concentrations greater than the study reporting levels, and were included in analyses and discussion related to occurrence of OWCs in wastewater effluents and the Big Sioux River.

OWCs in all six compound classes were detected in water samples from sampling sites in the Watertown area. The Watertown WWTP discharged continuously to the Big Sioux River during both the low-flow August 2003 and runoff June 2004 sampling periods. Total OWC concentrations for Big Sioux River sites upstream from the Watertown WWTP discharge generally were small, less than 6 micrograms per liter (µg/L) for both sampling periods. SCs accounted for nearly all of the total OWC concentrations for upstream Big Sioux River sites for the low-flow August 2003 sampling period, and MAHs accounted for nearly all of the total OWC concentrations for the runoff June 2004 sampling period. Total OWC concentrations for samples collected from the Watertown wastewater effluent were relatively large for both sampling periods (estimated concentrations ranged from 20 to 41 µg/L), and primarily consisted of HIACs, SCs, and HVACs. Total OWC concentrations for Big Sioux River sites downstream from the Watertown WWTP discharge were relatively large for the low-flow August 2003 sampling period (estimated concentrations ranged from 6.9 to 19 µg/L) and smaller for the runoff June 2004 sampling period (estimated concentrations ranged from 3.3 to 6.5 µg/L), a pattern that probably reflects a greater fraction of the total flow of the Big Sioux River being derived from WWTP discharge during the low-flow sampling period. Major OWC classes contributing to total OWC concentrations for Big Sioux River sites downstream from the Watertown WWTP were HIACs, SCs, and HVACs. Total OWC concentrations decreased substantially between the two downstream Big Sioux River sites. Although confident conclusions could not be made primarily due to possible effects of non-Lagrangian sampling, OWC results for the Watertown area might indicate that (1) OWCs for upstream Big Sioux River sites probably were primarily contributed by nonpoint agricultural sources, with livestock agriculture accounting for most of the total OWC concentration for the low-flow August 2003 sampling period, and crop agriculture accounting for most of the total OWC concentration for the runoff June 2004 sampling period; (2) OWCs for downstream Big Sioux River sites were substantially influenced by contributions from the Watertown WWTP during both the low-flow and runoff sampling periods; and (3) contributions of OWCs that might be derived from nonpoint livestock agricultural sources increased in proportion for the most downstream site for both the low-flow and runoff sampling periods. Suspected endocrine-disrupting compounds (EDCs) were detected in all Big Sioux River samples in the Watertown area. For both the low-flow and runoff sampling periods, the numbers of EDCs detected, and EDC concentrations and loads generally were larger for downstream Big Sioux River sites than for upstream Big Sioux River sites. Combined EDC concentrations for downstream Big Sioux River sites consisted mostly of HIACs for the low-flow sampling period and both HIACs and MAHs for the runoff sampling period.

OWCs in all compound classes except PAHs were detected in samples from sites in the Volga area. The Volga WWTP was not discharging to the Big Sioux River during the low-flow August 2003 and runoff June 2004 sampling periods, but was discharging continuously to the Big Sioux River during the low-flow October 2003 sampling period. For the low-flow August 2003 sampling period, the upstream Big Sioux River site had larger total OWC concentrations and loads than downstream Big Sioux River sites, and SCs accounted for most of the total OWC concentration for all Big Sioux River sites. For the low-flow October 2003 sampling period, when the Volga WWTP was discharging to the Big Sioux River, total OWC concentrations and loads were larger for the downstream Big Sioux River site than for the upstream site, and the increase in load corresponded well with the load contributed by the Volga wastewater effluent discharge, especially for HIACs. HIACs and SCs accounted for most of the total OWC concentrations for Big Sioux River sites for the October 2003 sampling period. For the June 2004 runoff sampling period, the upstream Big Sioux River site had smaller total OWC concentrations and loads than downstream Big Sioux River sites, and MAHs accounted for most of the total OWC concentrations for all Big Sioux River sites. Although confident conclusions could not be made due to possible effects of non-Lagrangian sampling, the data might indicate that (1) for the low-flow August 2003 sampling period, nonpoint livestock agricultural and/or human wastewater sources might have been the primary contributors to occurrence of OWCs at Big Sioux River sampling sites; (2) for the low-flow October 2003 sampling period, nonpoint livestock sources and upstream human wastewater sources primarily contributed to occurrence of OWCs at Big Sioux River sampling sites; (3) for the runoff June 2004 sampling period, nonpoint crop agricultural sources primarily contributed to occurrence of OWCs at Big Sioux River sampling sites; (4) for the low-flow August 2003 and runoff June 2004 sampling periods, seepage of water from the Volga WWTP had little effect on downstream OWC concentrations; and (5) for the low-flow October 2003 sampling period, the Volga wastewater effluent discharge contributed to downstream OWC concentrations. EDCs were detected in all samples collected from sampling sites in the Volga area. For all sampling periods, total EDC concentrations generally were similar between upstream and downstream Big Sioux River sites and consisted of HIACs and MAHs. HIACs accounted for most of the total EDC concentrations for the low-flow August 2003 and October 2003 sampling periods, and MAHs accounted for most of the total EDC concentrations for the runoff June 2004 sampling period for all Big Sioux River sites.

OWCs in all compound classes except PAHs were detected in water samples from sampling sites in the Brookings area. The Brookings WWTP discharged continuously to the Big Sioux River during all sampling periods. For the low-flow August 2003 sampling period, the upstream site had slightly smaller total OWC concentrations and loads compared to the downstream Big Sioux River sites. SCs and HIACs accounted for most of the total OWC concentration in all Big Sioux River sampling sites, but the proportion of SCs increased at the most downstream site. For the runoff June 2004 sampling period, the upstream site generally had smaller total OWC concentrations and loads than downstream Big Sioux River sites. MAHs accounted for most of the total OWC concentration for all Big Sioux River sites, but the proportion of SCs increased at the most downstream site. Although confident conclusions could not be made due to possible effects of non-Lagrangian sampling, the data might indicate that (1) for the low-flow August 2003 sampling period, nonpoint livestock agricultural sources probably primarily contributed to occurrence of OWCs at all Big Sioux River sampling sites, and the Brookings WWTP wastewater effluent discharge contributed but did not have a substantial effect on concentrations at downstream sites; and (2) for the runoff June 2004 sampling period, nonpoint crop agricultural sources primarily contributed to occurrence of OWCs at all Big Sioux River sites, contributions of OWCs that might be derived from nonpoint livestock agricultural sources increased in proportion to other sources for the most downstream site, and the Brookings WWTP wastewater effluent discharge probably did not substantially contribute to total OWC concentrations at downstream sampling sites. EDCs were detected in all samples collected from sampling sites in the Brookings area. Total EDC concentrations for the upstream site consisted entirely of MAHs. Total EDC concentrations for downstream sites consisted of MAHs and HIACs. HIACs accounted for most of the total EDC concentrations for the low-flow August 2003 sampling period, and MAHs accounted for most of the total EDC concentrations for the runoff June 2004 sampling period for downstream Big Sioux River sites.

The city of Watertown is located near the upstream part of the Big Sioux River Basin, where the mean annual flow of the Big Sioux River is less than 100 cubic feet per second (ft³/s). Watertown WWTP discharges can account for a substantial part of the flow in the Big Sioux River, especially during low-flow periods. Effects of the Watertown WWTP wastewater effluent discharges on the occurrence of OWCs in the Big Sioux River downstream were apparent during both the low-flow and runoff sampling periods. For Volga and Brookings, which are farther downstream and where the mean annual flow of the Big Sioux River exceeds 400 ft³/s, wastewater effluent discharges from the Volga and Brookings WWTPs probably influenced the occurrence of OWCs in the Big Sioux River, but probably did not substantially contribute to total OWC concentrations, especially during the runoff sampling period.