This data release and accompanying spreadsheet illustrate the ecosystem scale selenium model for Lake Koocanusa described in USGS Open-File Report 2020-xxxx, "Understanding and Documenting the Scientific Basis of Selenium Ecological Protection in Support of Site-Specific Guidelines Development for Lake Koocanusa, Montana, U.S.A., and British Columbia, Canada." All report sections, equations, and tables referenced in this data release and spreadsheet are from that report, referred to hereafter as “the OFR.” Two models are described in the OFR. Results for several different “runs” of this model with different model input parameters and assumptions are included in this data release. The spreadsheet calculations that produced those results are also included with this data release. Model 1, IFM = Insect-to-fish model, summarized by Eq. 11: protective C(Se)dissolved = tissue guideline/TTFfish /[(TTFinvert1*invert fraction1) + (TTFinvert2*invert fraction2)]/SPM bioavailability fraction/(Kd/1,000). Model2, TFM = Trophic-fish model, summarized by Eq. 12: protective C(Se)dissolved = tissue guideline/TTFfishTL4/TTFfishTL3/[(TTFinvert1*invert fraction1) + (TTFinvert2*invert fraction2)]/SPM bioavailability fraction/(Kd/1,000). Model results are presented for the four representative IFM food webs described in Table 6 of the OFR, and for three representative TFM food webs described in Table 7 of the OFR. “Generic” IFM and TMF modeling spreadsheets (consistent with Tables 8 and 9 of the OFR) are provided that implement the two models such that a user can establish their own modeling assumptions and view the results of the models with those assumptions.Finally, four alternative modeling scenarios and results specifically requested by the Montana Department of Environmental Quality, the cooperator in this work, are also presented here.

This Data Release presents multi-agency data for selenium concentrations in ecosystem media that includes water column, suspended particulate material, zooplankton, invertebrates, and fish. Because the data are compiled from multiple sources, the significant figures used to report contaminant concentrations and other metrics may not be internally consistent. These data will serve as the basis for ecosystem-scale modeling of Lake Koocanusa, a bi-national reservoir in Montana and British Columbia. Spreadsheets are ordered in a food-web format to facilitate modeling that emphasizes spatially and temporally paired data. Selenium concentrations are species-specific for fish and taxa-specific for invertebrates to address required specificity for biodynamic dietary modeling. Phytoplankton, zooplankton, and invertebrate densities or biomass are compiled, in addition to fish catches, to help elucidate productivity and identify which groups, taxa, or species are abundant on a seasonal basis. For water quality context, the historical record of selenium concentrations is given, with emphasis on the primary selenium loading site near where the Elk River enters Lake Koocanusa. Spreadsheets with the term "annex" in their file name address a competing toxin, mercury, for fish. Recent high frequency monitoring of selenium concentrations at the international border and at a gaging station below Libby Dam provide a perspective on future selenium data availability. Methodologies are described as federal, provincial, and state agencies transition to a consistent set of protocols to ensure consistency in monitoring for locations on both sides of the border.

The U.S. Geological Survey is studying Lake Koocanusa, a border reservoir between British Columbia in Canada and Montana, in collaboration with the Montana Department of Environmental Quality and the U.S. Environmental Protection Agency (Region 8). This work is part of a multi-agency project to apply an ecosystem-scale selenium modeling methodology (Presser and Luoma, 2010) to the lake to support development of a site-specific selenium guideline for the protection of aquatic life. This Data Release makes available measurements taken during the periods of May, 2015 to October, 2015 and April, 2016 to September, 2016. Spatially and temporally matched sets of dissolved and suspended particulate material selenium concentrations were collected mainly at two locations (forebay and international boundary) and at two depths (epilimnion, 3 meters below lake surface; hypolimnion, 3 meters above lake bottom). These data allow calculation of an environmental partitioning factor (i.e., Kd) between dissolved and suspended particulate material selenium concentrations as an initial modeling step. This Data Release contains 1) a location map; 2) a modeling schematic; 3) collection methods, analytical methods, and quality control information; 4) spreadsheets and graphical displays of selenium data; and 5) hydrological parameters related to sampling conditions at the lake. This Data Release will be updated and modified as additional sampling and analysis takes place.

The trace element selenium is an essential element with a narrow window between concentrations needed to support life and those that cause toxicity to egg laying organisms. Selenium bioaccumulation in aquatic organisms is primarily the result of trophic transfer through food webs and is poorly predicted by dissolved concentrations in freshwater bodies. To better understand the hydrologic and biological dynamics that control selenium accumulation into fishes of the Lower Gunnison River Basin (Colorado), ecosystem scale selenium accumulation models were developed from data collected between June 2015 and October 2016.

The trace element selenium is an essential element with a narrow window between concentrations needed to support life and those that cause toxicity to egg laying organisms. Selenium bioaccumulation in aquatic organisms is primarily the result of trophic transfer through food webs and is poorly predicted by dissolved concentrations in freshwater bodies. To better understand the hydrologic and biological dynamics that control selenium accumulation into fishes of the Lower Gunnison River Basin (Colorado), ecosystem scale selenium accumulation models were developed from data collected between June 2015 and October 2016.

The U.S. Geological Survey is studying Lake Koocanusa, a border reservoir between British Columbia in Canada and Montana, in collaboration with the Montana Department of Environmental Quality and the U.S. Environmental Protection Agency (Region 8). This work is part of a multi-agency project to apply an ecosystem-scale selenium modeling methodology (Presser and Luoma, 2010) to the lake to support development of site-specific selenium guidelines for the protection of aquatic life. A conceptual model framework was published in 2017 (Jenni, Naftz, and Presser, 2017) to serve as a coherent and consistent structure for organizing relevant scientific information, to provide an appropriate context for interpreting new information as datasets and site parameters are developed, and to identify data and science gaps that limit understanding of the implications of alternative Se criteria. This Data Release makes available measurements taken in May, July, and October, 2015; April, May, July, and September, 2016; and monthly from April through October, 2017. Spatially and temporally matched sets of dissolved and suspended particulate material selenium concentrations were collected mainly at two locations (forebay and international boundary) and at two depths (epilimnion, 3 meters below lake surface; hypolimnion, 3 meters above lake bottom). In 2017, samples were additionally collected during June, July, and September at a site south of the Elk River inflow by the British Columbia Ministry of Environment & Climate Change Strategy. These data allow calculation of an environmental partitioning factor (i.e., Kd) between dissolved and suspended particulate material selenium concentrations as an initial modeling step. This Data Release contains 1) a location map; 2) a modeling schematic; 3) collection methods, analytical methods, and quality control information; 4) spreadsheets and graphical displays of selenium data; and 5) hydrological parameters related to sampling conditions at the lake. Also included is an example sampling matrix that illustrates the types of data needed within Lake Koocanusa to support robust ecosystem-scale selenium modeling. For this 2017 update of the Data Release, the same methodologies were followed as in 2016, except sampling occurred over seven consecutive months. This allowed documentation of a more detailed view of the dissolve and particulate selenium dynamics of Lake Koocanusa to help identify trends and connect to other ecological and management variables. This Data Release will be updated and modified if additional sampling and analysis takes place.

The U.S. Geological Survey is studying Lake Koocanusa, a border reservoir between British Columbia in Canada and Montana, in collaboration with the Montana Department of Environmental Quality and the U.S. Environmental Protection Agency (Region 8). This work is part of a multi-agency project to apply an ecosystem-scale selenium modeling methodology (Presser and Luoma, 2010) to the lake to support development of site-specific selenium guidelines for the protection of aquatic life. A conceptual model framework was published in 2017 (Jenni, Naftz, and Presser, 2017) to serve as a coherent and consistent structure for organizing relevant scientific information, to provide an appropriate context for interpreting new information as datasets and site parameters are developed, and to identify data and science gaps that limit understanding of the implications of alternative Se criteria. This Data Release makes available measurements taken in May, July, and October, 2015; April, May, July, and September, 2016; and monthly from April through October, 2017. Spatially and temporally matched sets of dissolved and suspended particulate material selenium concentrations were collected mainly at two locations (forebay and international boundary) and at two depths (epilimnion, 3 meters below lake surface; hypolimnion, 3 meters above lake bottom). In 2017, samples were additionally collected during June, July, and September at a site south of the Elk River inflow by the British Columbia Ministry of Environment & Climate Change Strategy. These data allow calculation of an environmental partitioning factor (i.e., Kd) between dissolved and suspended particulate material selenium concentrations as an initial modeling step. This Data Release contains 1) a location map; 2) a modeling schematic; 3) collection methods, analytical methods, and quality control information; 4) spreadsheets and graphical displays of selenium data; and 5) hydrological parameters related to sampling conditions at the lake. Also included is an example sampling matrix that illustrates the types of data needed within Lake Koocanusa to support robust ecosystem-scale selenium modeling. For this 2017 update of the Data Release, the same methodologies were followed as in 2016, except sampling occurred over seven consecutive months. This allowed documentation of a more detailed view of the dissolve and particulate selenium dynamics of Lake Koocanusa to help identify trends and connect to other ecological and management variables. This Data Release will be updated and modified if additional sampling and analysis takes place.

Pollutant loads have been increasing over time in the Elk River, B.C. due to coal mining operations and runoff from associate spoil piles. The Elk River is a tributary to the Kootenay/Kootenai River and Lake Koocanusa. Extensive water chemistry monitoring has been conducted in Lake Koocanusa to assess the impacts from the Elk River, however, this is not the case for the Kootenai River downstream of Lake Koocanusa, downstream of Libby Dam (http://deq.mt.gov/DEQAdmin/LakeKoocanusa). This study generated data on selenium and nutrient concentrations and loads in the Kootenai River (Libby Dam to Canadian border), which will help to differentiate between local loads and loads transported via Lake Koocanusa. Working in cooperation with the U.S. Environmental Protection Agency (USEPA), the U.S. Geological Survey (USGS) collected and analyzed water-column samples in the Kootenai River watershed downstream of Libby Dam. Field data collection also included discharge measurements. The following are key design components of water quality sample collection in the Kootenai River watershed downstream of Libby Dam: Sampling locations: (1) USGS gages on the mainstem Kootenai River: Kootenai River below Libby Dam, MT (USGS site identification number 12301933), Kootenai River at Leonia, ID (12305000), Kootenai River at Tribal Hatchery near Bonners Ferry, ID (12310100), and Kootenai River at Porthill, ID (12322000); (2) USGS gages on tributaries to the Kootenai River: Fisher River near Libby MT (12302055), Yaak River near Troy MT (12304500); (3) Ungaged tributaries: Moyie River near the mouth (12307750). Water quality analyses: (1) Low-level nutrients, total water (3 sampling events) (2) Total selenium, dissolved water (3 sampling events) (3) Selenium speciation, dissolved water (2 events): selenite, Se(IV); selenate, Se(VI); selenocyanate, SeCN; selenomethionine, SeMet; and methylseleninic acid, MeSe(IV). Total dissolved selenium and selenium speciations analyses were performed by Brooks Applied Labs, Bothell, WA. The nutrient samples were analyzed by the USGS National Water Quality Laboratory (NWQL), Lakewood, CO. Secondary, total dissolved selenium analyses were also conducted by the NWQL for a subset of samples for interlaboratory comparisons. The three sampling events were: (1) September 2018, a period of low and stable flows in both the Kootenai River and tributaries; (2) December 2018, a period of low, stable winter flows in the tributaries but unstable and increasing flows in the Kootenai River due to dam releases in preparation for spring runoff (3) May 2019, a period close to annual peak runoff from snowmelt in tributaries plus high dam releases of snowmelt originating in the Canadian Rockies. Results: Results of the selenium sampling are compiled and summarized in this data release. These data and the nutrient data are separately published through the USGS National Water Information System (https://doi.org/10.5066/F7P55KJN). No samples exceeded USEPA’s national recommended aquatic life criteria for selenium, expressed as a water concentration, of 3.1 µg/L. Selenium concentrations in the four mainstem Kootenai River sites had low variability across sites and sampling events, ranging from 0.628 to 1.17 µg/L for the 12 samples, for an average concentration of 0.91 µg/L (± 0.17 µg/L, standard deviation). In all measured concentrations, selenate was the dominant chemical species, accounting for 87% of the total selenium, on average (range 79-94%). Selenium concentrations were less than the detection limit of 0.037 µg/L in all of the tributary samples, indicating that selenium concentrations in the Kootenai River cannot be attributed to tributary sources. Daily average selenium loads in the Kootenai River were estimated by assuming that the width and depth integrated water samples collected were representative of average conditions for that day and were multiplied by daily average discharge volumes of the Kootenai River at those locations. In

Pollutant loads have been increasing over time in the Elk River, B.C. due to coal mining operations and runoff from associate spoil piles. The Elk River is a tributary to the Kootenay/Kootenai River and Lake Koocanusa. Extensive water chemistry monitoring has been conducted in Lake Koocanusa to assess the impacts from the Elk River, however, this is not the case for the Kootenai River downstream of Lake Koocanusa, downstream of Libby Dam (http://deq.mt.gov/DEQAdmin/LakeKoocanusa). This study generated data on selenium and nutrient concentrations and loads in the Kootenai River (Libby Dam to Canadian border), which will help to differentiate between local loads and loads transported via Lake Koocanusa. Working in cooperation with the U.S. Environmental Protection Agency (USEPA), the U.S. Geological Survey (USGS) collected and analyzed water-column samples in the Kootenai River watershed downstream of Libby Dam. Field data collection also included discharge measurements. The following are key design components of water quality sample collection in the Kootenai River watershed downstream of Libby Dam: Sampling locations: (1) USGS gages on the mainstem Kootenai River: Kootenai River below Libby Dam, MT (USGS site identification number 12301933), Kootenai River at Leonia, ID (12305000), Kootenai River at Tribal Hatchery near Bonners Ferry, ID (12310100), and Kootenai River at Porthill, ID (12322000); (2) USGS gages on tributaries to the Kootenai River: Fisher River near Libby MT (12302055), Yaak River near Troy MT (12304500); (3) Ungaged tributaries: Moyie River near the mouth (12307750). Water quality analyses: (1) Low-level nutrients, total water (3 sampling events) (2) Total selenium, dissolved water (3 sampling events) (3) Selenium speciation, dissolved water (2 events): selenite, Se(IV); selenate, Se(VI); selenocyanate, SeCN; selenomethionine, SeMet; and methylseleninic acid, MeSe(IV). Total dissolved selenium and selenium speciations analyses were performed by Brooks Applied Labs, Bothell, WA. The nutrient samples were analyzed by the USGS National Water Quality Laboratory (NWQL), Lakewood, CO. Secondary, total dissolved selenium analyses were also conducted by the NWQL for a subset of samples for interlaboratory comparisons. The three sampling events were: (1) September 2018, a period of low and stable flows in both the Kootenai River and tributaries; (2) December 2018, a period of low, stable winter flows in the tributaries but unstable and increasing flows in the Kootenai River due to dam releases in preparation for spring runoff (3) May 2019, a period close to annual peak runoff from snowmelt in tributaries plus high dam releases of snowmelt originating in the Canadian Rockies. Results: Results of the selenium sampling are compiled and summarized in this data release. These data and the nutrient data are separately published through the USGS National Water Information System (https://doi.org/10.5066/F7P55KJN). No samples exceeded USEPA’s national recommended aquatic life criteria for selenium, expressed as a water concentration, of 3.1 µg/L. Selenium concentrations in the four mainstem Kootenai River sites had low variability across sites and sampling events, ranging from 0.628 to 1.17 µg/L for the 12 samples, for an average concentration of 0.91 µg/L (± 0.17 µg/L, standard deviation). In all measured concentrations, selenate was the dominant chemical species, accounting for 87% of the total selenium, on average (range 79-94%). Selenium concentrations were less than the detection limit of 0.037 µg/L in all of the tributary samples, indicating that selenium concentrations in the Kootenai River cannot be attributed to tributary sources. Daily average selenium loads in the Kootenai River were estimated by assuming that the width and depth integrated water samples collected were representative of average conditions for that day and were multiplied by daily average discharge volumes of the Kootenai River at those locations. In

The U.S. Geological Survey (USGS), in cooperation with the Bureau of Reclamation, completed a review of dissolved selenium data collected from the Selenium Management Program network during each water year (WY) in the lower Gunnison River Basin, in western Colorado. The data tables include dissolved-selenium concentrations, selected streamflow data, and computed loads as well as dissolved-selenium regression model calibration and output tables. Concentration and streamflow data are compiled from the USGS National Water Information System database. Tables include monitoring location identification information, data collection time frames, concentration data, as well as calculations of mean, median, select percentiles, and percentage reduction in loads. Tables are organized by each year, with a zipped folder for WY 2018 (October 1, 2017 through September 30, 2018), WY 2019 (October 1, 2018 through September 30, 2019), WY 2020 (October 1, 2019 through September 30, 2020), WY 2021 (October 1, 2020 through September 30, 2021), and WY 2022 (October 1, 2021 through September 30, 2022). Within each zipped folder is a metadata .xml file describing each comma-separated value (.csv) data file contained therein. The methods used to complete this review are available in Henneberg (2018) and Mayo and Leib (2012). Version 3.0 contains additional zipped folders for WY 2021 and 2022, which were not included in the original data release. While previous versions are available from the author, all records in previous versions can be found in version 3.0. First posted - April 23, 2021 (available from author) Revised - June, 2022 (version 2.0) Revised - July, 2023 (version 3.0)