Annual Water Quality and Benthic Algae Monitoring Results for the Clark Fork River Basin 2018

Home , Cabinet Gorge Dam, Clark Fork River, Lake Pend Oreille, Pend Oreille River

CLARK FORK RIVER WATER QUALITY MONITORING COMMITTEE

ANNUAL WATER QUALITY AND BENTHIC ALGAE MONITORING RESULTS FOR THE CLARK FORK RIVER BASIN 2018

6/30/2018

Prepared by: Clark Fork Coalition, P.O. Box 7593, Missoula, MT 59807 Table of Contents

1.0 Introduction ...... 2 2.0 History and Background ...... 2 3.0 Monitoring Program ...... 2 4.0 Data QA/QC Summary ...... 6 5.0 Nutrient Standards ...... 7 6.0 Nutrient Results ...... 8 6.1 Total Phosphorus ...... 10 6.2 Soluble Reactive Phosphorus ...... 10 6.3 Total Nitrogen ...... 13 6.4 Nitrate + Nitrite ...... 14 6.5 Ammonia ...... 17 7.0 Nitrogen – Phosphorus Ratios ...... 19 8.0 Benthic Algae Results ...... 20 9.0 Peak Flow Nutrient Monitoring and Loading to Lake Pend Oreille ...... 22 10.0 References ...... 24

Figures Figure 1: Clark Fork River Nutrient and Periphyton Monitoring Sites…………………………………………………..…..5 Figure 2: Hydrographs from USGS continuous monitoring stations……………………………………………….……..…9 Figure 3: Total phosphorous from summer (upper) and monthly (lower) monitoring stations…………..….12 Figure 4: Soluble reactive phosphorous for summer (upper) and monthly (lower) monitoring stations…13 Figure 5: Total persulfate nitrogen results for summer (upper) and monthly (lower) monitoring stations………………………………………………………………………………………………………………………………………………….15 Figure 6: Nitrate + Nitrite results for summer (upper) and monthly (lower) monitoring stations…………..16 Figure 7: Ammonia results from summer (upper) and monthly (lower) monitoring stations …….…………..18 Figure 8: Benthic algae chlorophyll-a and ash free dry weight results…………………………………………………...21 Figure 9: Estimated Lake Pend Oreille Nutrient Loads from Clark Fork River 1998-2017……………….……….23 TABLES Table 1: Monitoring Locations, Rationale, and Sampling Frequency………………………..……………………………...6 Table 2: SRP as a percentage of TP…………………………………………………………………………………….………….……...11 Table 3: Nitrate + nitrite as a percentage of total nitrogen…………………………………………………………….……..14 Table 4: Mass-based N:P rations for Total N:P (upper) and Dissolved N:P (lower)…………….……..……………20 Attachments QA/QC Report for Clark Fork River Monitoring (MDEQ) 2018 CFR-BASIN Data Tables (2018 Nutrient Results for the Clark Fork River Basin) HydroSolutions Tech Memo: Estimate of 2018 Nutrient Loading from Clark Fork River into Lake Pend Oreille

1 1.0 INTRODUCTION

This report presents 2018 nutrient and benthic algae monitoring results from the Clark Fork River basin collected by the University of Montana (UM), the City of Missoula, Avista Corporation (Avista), and the Clark Fork Coalition, and overseen by the Clark Fork River Water Quality Monitoring Committee (CFRWQMC). This report also summarizes and presents results of quality assurance and quality control analysis by the Montana Department of Environmental Quality (MDEQ), and annual nutrient loading analysis to Lake Pend Oreille by Avista. The purpose of the report is to present monitoring results and assess compliance with water quality standards.

Further analysis of annual results from this monitoring program is accomplished on a five-year schedule when the CFRWQMC contracts a statistical evaluation and trends analysis. The latest trends report covered the period 2013-2017 (HydroSolutions, 2018) and the next trends report will be presented in 2023, adding data through 2022.

2.0 HISTORY AND BACKGROUND

The current monitoring program is a continuation of the Montana portion of a program that was originally developed in 1998 by the Tri-State Water Quality Council (TSWQC). The TSWQC, a partnership of citizens, businesses, industry, tribes, government, and environmental groups, formed in 1993 to collaboratively address solutions to the problem of excess nutrients and algae identified in the Clark Fork-Pend Oreille Basin Management Plan (EPA, 1993) in Montana, Idaho and Washington. The monitoring committee of the TSWQC began a basin-wide targeted monitoring program in 1998, in part to assess the progress of a 10-year Voluntary Nutrient Reduction Program (VNRP) in Montana. The monitoring program employed a statistically-based sampling design that was intended to support sound, scientifically-based water management decisions. Nutrient and algal targets developed for the VNRP were later adopted as numeric standards in the Clark Fork basin in 2002 – the first numeric standards for nutrients and algae in Montana. Due to shrinking budgets, the TSWQC disbanded in 2012 and Montana DEQ assumed oversight responsibility for monitoring in the Clark Fork basin. Former members of the TSWQC formed the CFRWQMC, comprised of MDEQ, Avista, City of Missoula, UM Watershed Clinic, the Clark Fork Coalition, and Idaho Department of Environmental Quality (IDEQ), to continue the monitoring effort. The City of Missoula’s collected field samples through 2017 and provided lab analysis through 2018, but due to increased workload ended its participation in the project prior to the start of field season in 2019. The Tri-State Water Quality Council managed the previous 5 year monitoring programs from 1998-2002, 2003-2007, and 2008-2012, which provided the basis for a statistical analysis of water quality time trends for the Clark Fork River and Lake Pend Oreille Watershed. The CFRWQMC modified the monitoring plan to be as cost effective as possible while still maintaining continuity for statistical analysis. The 2013-2017 monitoring program represented the first 5 year monitoring program managed by the CFRWQMC. The 2018 data presented in this report represent the first year of CFRWQMC’s second 5 year period, 2018-2022.

3.0 MONITORING PROGRAM

The CFRWQMC monitoring program maintains two of the original four management goals established by the TSWQC: 1) Control nuisance algae in the Clark Fork River by reducing nutrient concentrations

2 2) Protect Lake Pend Oreille water quality by maintaining or reducing current rates of nutrient loading from the Clark Fork River.

To meet these goals, the CFRWQMC will: 1) Evaluate time trends in nutrient concentrations in the mainstem Clark Fork River and selected tributaries; 2) Evaluate time trends for periphyton (algae) standing crops in the Clark Fork River; 3) Monitor summer nutrient and periphyton target levels in the Clark Fork River; 4) Estimate nutrient loading rates to Lake Pend Oreille from the Clark Fork River.

The states of Idaho (ID) and Washington (WA) carry out additional monitoring and studies to address the remaining TSWQC goals and objectives specific to Lake Pend Oreille and the Pend Oreille River. These are described in the Quality Assurance Project Plan (QAPP) (MDEQ, 2018).

The 2018-2022 program, like the 2013-2017 program before it, consists of a basic monitoring component and several annual and periodic, rotational add-on elements are incorporated as needed. The basic monitoring program consists of the highest priorities for annual monitoring, while the add-ons represent options for additional monitoring that are contingent on annual funding availability (e.g., the lower Flathead River monitoring station was added in 2013 and additional sites for the Statewide Monitoring Program may be added in the future). All monitoring activities to meet the Montana objectives were performed in accordance with the “Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program from Headwaters to Below Cabinet Gorge Dam – Quality Assurance Project Plan (QAPP)” (MDEQ, 2018) which is updated annually.

The objectives are met by: 1. Monthly monitoring: Avista collects monthly nutrients and field constituents at the Clark Fork River below Cabinet Gorge Dam, site CFR-30 (excluding December, January, and February). During the months of July, August, and September, monitoring is also conducted at sites below Thompson Falls (CFR-28) and at Noxon Bridge (CFR-29). 2. Peak flow monitoring: In addition to monthly monitoring, Avista collects nutrient samples at the Clark Fork River below Cabinet Gorge Dam (CFR-30) during spring peak flow (six sampling events over a 1-month period in approximately May and/or June); 3. Summer monitoring: The Clark Fork Coalition (CFC), the City of Missoula, and the University of Montana Watershed Clinic collect nutrient samples and field constituents in summer at nine Clark Fork River sites and one site on the Lower Flathead River (six sampling events, July- September; coordinated between the CFC, the City of Missoula and UM); 4. Benthic algae monitoring: The UM Watershed Clinic collects summer benthic algae samples for chlorophyll-a and ash-free dry weight at seven Clark Fork River sites. In previous years, an eighth site was included on the lower Flathead River, but it was discontinued in 2018 due to persistently low densities of algae

Specifically, the CFRWQMC measures:  Nutrients: total phosphorus (TP), total persulfate nitrogen (TPN), dissolved nitrate + nitrite nitrogen (NO2+NO3-N), dissolved ammonia nitrogen (NH3+NH4-N), and soluble reactive phosphorus (SRP).  Field parameters: water temperature (˚C), dissolved oxygen (mg/l), pH (standard units), redox potential (mv), specific conductance (μs/cm), total dissolved solids (mg/l), and turbidity (NTU).  Benthic algae: chlorophyll-a (mg/m2) and ash-free dry weight (g/m2).

3 All nutrient samples were analyzed by the City of Missoula Wastewater Treatment Plant lab and benthic algae samples are analyzed by the UM Watershed Health Clinic. Sampling, QA/QC and analytical methods are described in the QAPP (MDEQ, 2018). The QA/QC Report for Clark Fork River Monitoring (May, 2019) is attached to this report and data qualifiers are noted in the 2018 CFR-BASIN Data Tables, attached. Monitoring stations locations are provided in Figure 1 and Table 1.

Figure 1: Clark Fork River Nutrient and Periphyton Monitoring Sites

5 Table 1: Monitoring Locations, Rationale, and Sampling Frequency

Station Name Rationale Sampling Frequency below mixing zone for Butte wastewater CFR 2.5 Silver Bow Creek at Opportunity S6 treatment plant upstream control site, start of Clark Fork CFR 07 Clark Fork below Warm Springs Creek S6 River upstream control site, upper river CFR 09 Clark Fork at Deer Lodge C20, S6 indicator site CFR 10 Clark Fork above Little Blackfoot River below mixing zone for Deer Lodge WWTP C20, S6 upper river site, between significant CFR 12 Clark Fork at Bonita (Beavertail Hill) C20, S6 tributaries below Blackfoot drainage, control site for CFR 15.5 Clark Fork above Missoula C20 S6 Missoula Clark Fork below Missoula (Tower CFR 18 below mixing zone for Missoula WWTP C20, S6 Street Conservation Area) lower river site, downstream of Missoula CFR 22 Clark Fork at Huson C20, S6 and Smurfit-Stone lower river site, upstream control for the CFR 25 Clark Fork above Flathead C20, S6 Flathead River Flathead River near Mouth (near Tribal CFR 26 Requested by EPA for TMDL purposes S6 Boundary) lower river site, downstream of Flathead CFR 28 Clark Fork below Thompson Falls N3 River and Thompson Falls WWTP CFR 29 Clark Fork at Noxon Bridge lower river site, reflects reservoir influence N3 estimation of nutrient loading to Lake CFR 30 Clark Fork below Cabinet Gorge Dam N15 Pend Oreille CFR = Clark Fork River and its tributaries N3 = Monthly monitoring for nutrients and field constituents, 3 monthly samples (July, August, and September). N15 = Monthly monitoring for nutrients and field constituents, 9 monthly samples (excludes December, January and February) and Peak flow monitoring for nutrients, 6 peak flow samples C20 = Benthic algae monitoring for Chlorophyll-a and ash-free dry weight, 20 replicates per site, August and September. In low-flow years stations 9 through 18 are also monitored in July. S6 = Summer monitoring for nutrients and field constituents, 6 samples during July, August and September

4.0 DATA QA/QC SUMMARY

All laboratory and field data were reviewed and validated per guidance in the QAPP (MDEQ, 2018). Montana DEQ analyzes and flags the monitoring data each year for quality assurance/quality control and provides the QA/QC Report for Clark Fork River Monitoring, 2018 that is appended to this report. This section briefly summarizes the results. Data flags from the QA/QC report are included in the 2018 CFR- BASIN Data Tables, also appended to this report.

All sites were sampled on schedule per the QAPP except for CFR-18 Below Missoula, which was not sampled during the late July monitoring event due to unsafe conditions associated with unusually high spring runoff.

6 Fifty-three data points were flagged ‘H’ for exceeding holding times (vs. 47 in 2017). In general, holding times were exceeded when initial analytical results fell outside of QA/QC requirements outlined in the project QAPP and the samples had to be reanalyzed, sometimes more than once. Forty-seven data points from blank samples had detects above the lower reporting limit and as a result, were flagged ‘B’ for blank contamination (vs. 12 in 2017). There were 14 data points flagged “D” because the reporting limit was increased due to sample dilutions (vs. 25 in 2017). There were 79 data points that had reported values between the method detection limit and the reporting limit and were flagged “J” (vs. 124 in 2017). One data point was rejected (flagged R) due to improper field blank preparation. The R- flagged results remain in our local database, but will not be available from the National STORET Warehouse and were not included in the figures and tables in this report.

There were no data flagged for laboratory issues. The overall project data quality objectives (DQOs) and data quality indicators (DQIs) were met as set forth in the QAPP. All laboratory QC was within limits specified by the specific analytical methods and the QAPP. Apart from the one R-flagged sample described above, no other data were determined to be unusable (rejected) in the validation process or lost due to sampler or laboratory error. All data were validated per guidance in the QAPP, approved for the intended use, and added to the database.

As a result of the QA review, the following are corrective actions items for 2019: • Rationale for any exceedance of holding times should be documented and communicated in reports to DEQ • Site visit forms and yearly sampling report should be submitted to DEQ as outlined in the QAPP • Final COC forms should be submitted to DEQ when the final EDD is submitted

The CFWQMC discussed ways to improve data quality and QA/QC reporting at their annual meeting, and the QAPP and SAPs were updated accordingly.

5.0 NUTRIENT STANDARDS

Nutrient standards for the Clark Fork River monitoring program fall under two authorities. For Silver Bow Creek at Opportunity, the standards are listed in MDEQ Department Circular DEQ-12A (2014). This circular contains information pertaining to the base numeric nutrients standards for wade-able streams (§75-5-301(2), MCA) and their implementation. Silver Bow Creek is located within the Middle Rockies ecoregion, thus the standards apply from July 1st to September 30th, and are as follows:  Total phosphorus as P: 30 µg/L  Total Nitrogen as N: 300 µg/L

For the mainstem Clark Fork River from below the Warm Springs Creek confluence (N46º11'17", W112º46'03") to the confluence with the Blackfoot River (N46º52'19", W113º53'35"), ARM 17.30.631 (2002) applies. The numeric water quality standards for Total Nitrogen, Total Phosphorus, and benthic algal chlorophyll a, applicable from June 21 to September 21, are as follows:  Total Phosphorus as P: 20 μg/L  Total Nitrogen as N: 300 μg/L  (Summer mean) - Benthic 100 mg/square meter algal chlorophyll a  (Maximum) - Benthic 150 mg/square meter algal chlorophyll a

7 For the mainstem Clark Fork River from the confluence with the Blackfoot River (N46º52'19", W113º53'35") to the confluence with the Flathead River (N47º21'45", W114º46'43"), ARM 17.30.631 (2002) applies. The numeric water quality standards for Total Nitrogen, Total Phosphorus, and benthic algal chlorophyll a, applicable from June 21 to September 21, are as follows:  Total Phosphorus as P: 39 μg/L  Total Nitrogen as N: 300 μg/L  (Summer mean) - Benthic 100 mg/square meter algal chlorophyll a  (Maximum) - Benthic 150 mg/square meter algal chlorophyll a

(History: 75-5-301, MCA; IMP, 75-5-301, MCA; NEW, 2002 MAR p. 2196, Eff. 8/16/02.)

6.0 NUTRIENT RESULTS

Streamflow conditions during spring runoff and summer months influence nutrient concentrations and algal densities. Years with less-than-average peak flows and early summer low flows typically see higher algal densities, and conversely, years with higher peak flows tend to produce less algal density. Figure 2 shows eight 2018 annual hydrographs (including the median daily flow for the period of record at each site) from stations in the study area, arranged upstream to downstream, to provide context for interpreting nutrient and algae results (USGS, 2018).

Peak flows during 2018 were some of the highest ever recorded in western Montana, and all station saw peaks well in excess of median values. In general, flows throughout the remainder of the year exceeded median values as well, though there were several exception, including winter flows in lower Silver Bow Creek and in the Clark Fork at Deer Lodge, and Drummond; and the falling limb and base flow period at the Flathead at Perma and Clark Fork at Plains USGS stations (Figure 2).

Figure 2: Hydrographs from USGS continuous monitoring stations (USGS, 2018).

6.1 TOTAL PHOSPHORUS

Results of total phosphorus (TP) monitoring are presented in Figure 3 and in the 2018 CFR-BASIN Data Tables, attached. TP was highest in Silver Bow Creek and all samples at this station exceeded the total phosphorus standard of 30 µg/L by 3 to 6 times. TP levels were noticeably lower than in 2016, when TP reached a high of 400 ug/L and were similar to concentrations in 2017. All samples at the Clark Fork below Warm Springs, above Little Blackfoot, and Bonita also exceeded the TP standard of 20 µg/L, as did all but the late September sample at Deer Lodge. The frequency of exceedance is notably higher than in 2017, particularly at Deer Lodge, where only the early July sample exceeded the standard in 2017. The frequency of exceedance above and below Missoula and at Huson was also higher than in 2017, when no concentrations of TP above 20 ug/L were detected at any of the sites below Bonita. In 2017, concentrations exceeded the standard twice above Missoula and below Missoula, and once at Huson. Concentrations of TP at the site below Cabinet Gorge Dam were also much higher in 2018 than in 2017, reaching levels almost double that of 2017 during peak flows of 2018. The relatively high concentrations of TP detected in 2018 are likely a function of the near record flows in the Clark Fork River watershed and an accompanying increase in suspended particulate matter, which is typically a source of phosphorus.

6.2 SOLUBLE REACTIVE PHOSPHORUS

Results of soluble reactive phosphorus (SRP) monitoring are presented in Figure 4 and in the 2018 CFR- BASIN Data Tables, attached. There are no numeric standards for SRP. SRP was highest in Silver Bow Creek (39-58 µg/L), but lower than in 2017 when it peaked at 155 ug/L in early July and was 80 ug/L or higher on 5 of the 6 monitoring events. Except for an anomalously low concentration of 6.3 ug/L below Warm Springs in early July, SRP followed a pattern in 2018 that was roughly similar to 2017, albeit at slightly higher concentrations, perhaps due to greater a greater availability of phosphorus containing particles created by the sustained high flows of 2018. From Missoula downstream, SRP concentrations were characteristically low, generally below 5 ug/L. The mean percentage soluble reactive phosphorus of total phosphorus is shown in Table 2.

10 Table 2: SRP as a percentage of TP

Site Mean Percentage SRP of Total Phosphorus Silver Bow Creek at Opportunity 42% Clark Fork below Warm Springs 55% Clark Fork at Deer Lodge 27% Clark Fork above Little Blackfoot 30% Clark Fork at Bonita (Beavertail Hill) 28% Clark Fork above Missoula 22% Clark Fork below Missoula 25% Clark Fork at Huson 14% Clark Fork above Flathead 18% Flathead River 26% Clark Fork below Thompson Falls 19% Clark Fork at Noxon Bridge 24% Clark Fork below Cabinet Gorge Dam 18% (Note: below detect values calculated at detection limit)

Figure 3: Total phosphorus for summer (upper) and monthly (lower) monitoring stations

Bars indicate 2-week sampling dates that range from early July to mid-September for Silver Bow to Flathead and monthly sampling dates from mid-March to mid-November for the lower Clark Fork. See 2018 CFR-BASIN Data Tables for exact dates. Samples below detection are shown at the lower reporting limit of 2 µg/L.

50 70 60 45 2018 Soluble Reactive Phosphorus Summer Sampling - Silver Bow to Flathead 50 40 40

35 30 ug/L SRP ug/L 20 30 10 25 0 Silver Bow 20 Creek at Opportunity

ug/L Soluble ReactiveP Soluble ug/L 15

0 Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Flathead River below River at River above River at River above River below River at River above River Warm Deer Lodge Little Bonita Missoula Missoula Huson Flathead Springs Blackfoot

8 2018 Soluble Reactive Phosphorus 7 Monthly Sampling Lower Clark Fork 6

3 ug/L Soluble Reactive P Reactive Soluble ug/L 2

0 Clark Fork River below Thompson Clark Fork River at Noxon Clark Fork River below Cabinet Gorge Falls Dam

Figure 4: Soluble reactive phosphorus for summer (upper) and monthly (lower) monitoring stations

Bars indicate 2-week sampling dates that range from early July to mid-September for Silver Bow to Flathead and monthly sampling dates from mid-March to mid-November for the lower Clark Fork. See 2018 CFR-BASIN Data Tables for exact 6.3dates. T SamplesOTAL belowNITROGEN detection are shown at the lower reporting limit of 2 µg/L.

13 6.3 TOTAL NITROGEN

Results of total nitrogen (TN) monitoring are presented in Figure 5 and in the 2018 CFR-BASIN Data Tables, attached. Total nitrogen concentrations were highest in Silver Bow Creek (range: 520 - 821 µg/L), continuing their steady decline since 2015, and down considerably from a high of 1215 ug/L in 2017, possibly due to Butte’s new wastewater treatment plant coming on line. Concentrations were still more than 2 times the TN standard (300 µg/L). TN concentrations were slightly higher at Warm Springs compared to 2017, with one exceedance of the 300 ug/L standard, but then dropped when compared to 2017 at the next 3 stations (Deer Lodge, Little Blackfoot, and Bonita) both in terms of total concentrations and number of exceedances. TN was in excess of 500 ug/L on 4 occasions at Deer Lodge and above the Little Blackfoot in 2017, but generally remained near or below 400 ug/L in 2018. Total nitrogen concentrations were below 300 ug/L at all stations on all occasions from Bonita downstream and in the Flathead River.

6.4 NITRATE + NITRITE

Results of nitrate + nitrite monitoring are presented in Figure 6 and in the 2018 CFR-BASIN Data Tables, attached. There are no numeric standards for nitrate + nitrite. As with TN, concentrations of nitrate+nitrite were highest in Silver Bow Creek at Opportunity, reaching a maximum concentration of 545 ug/L, almost 300 ug/L lower than in 2017. As was the case in 2017, late summer concentrations were relatively high at Deer Lodge and above the Little Blackfoot, but at levels typically 100 ug/L less than in 2017. Concentrations were lowest at Bonita and above Missoula, except for an unusual spike in concentration in late September at the site above Missoula. Concentrations increased at the below Missoula site but remained below 60 ug/L at all sites downstream of Missoula except on one occasion at Huson. Nitrate + nitrite as a percentage of total nitrogen is shown in Table 3.

Table 3: Nitrate + nitrite as a percentage of total nitrogen

Site Mean Percentage Nitrate+Nitrite of Total Nitrogen Silver Bow Creek at Opportunity 55% Clark Fork below Warm Springs 6% Clark Fork at Deer Lodge 15% Clark Fork above Little Blackfoot 9% Clark Fork at Bonita (Beavertail Hill) 1% Clark Fork above Missoula 2% Clark Fork below Missoula 4% Clark Fork at Huson 5% Clark Fork above Flathead 2% Flathead River 4% Clark Fork below Thompson Falls 0.4% Clark Fork at Noxon Bridge 1% Clark Fork below Cabinet Gorge Dam 7% (Note: below detect values calculated at detection limit)

14 800 1000 2018 Total Persulfate Nitrogen 800 700 Summer Sampling 600 Silver Bow to Flathead 400 600 TPN ug/L 200 0 500 Silver Bow Creek at 400 Opportunity Nitrogen Standard

300 ug/L Total Persulfate N Persulfate Total ug/L 200

100

0 Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Flathead River below River at River above River at River aboveRiver below River at River above River Warm Deer Lodge Little Bonita Missoula Missoula Huson Flathead Springs Blackfoot

300 2018 Total Persulfate Nitrogen 250 Monthly Sampling Lower Clark Fork

200

150

100 ug/L Total Persulfate N Persulfate Total ug/L

0 Clark Fork River below Thompson Clark Fork River at Noxon Clark Fork River below Cabinet Falls Gorge Dam

Figure 5: Total persulfate nitrogen results for summer (upper) and monthly (lower) monitoring stations

Tables for exact dates. Samples below detection are shown at the lower reporting limit of 50 µg/L.

15 160 600 2018 Nitrate + Nitrite 500 140 Summer Sampling 400 Silver Bow to Flathead 300 120

200 ug/L Nitrate ug/L 100 100 0 Silver Bow 80 Creek at Opportunity

60 ug/L Nitrate + Nitrite + Nitrate ug/L

0 Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Clark Fork Flathead River below River at River above River at River aboveRiver below River at River above River Warm Deer Lodge Little Bonita Missoula Missoula Huson Flathead Springs Blackfoot 70

60 2018 Nitrate + Nitrite Monthly Sampling Lower Clark Fork 50

ug/L Nitrate + Nitrite + Nitrate ug/L 20

0 Clark Fork River below Thompson Clark Fork River at Noxon Clark Fork River below Cabinet Falls Gorge Dam

Figure 6: Nitrate + Nitrite results for summer (upper) and monthly (lower) monitoring stations

16 6.5 AMMONIA

Results of ammonia monitoring are presented in Figure 7 and in the 2018 CFR-BASIN Data Tables, attached. Concentrations were generally below 40 ug/L, with a notable exception at Bonita in late September when the concentration reached 56 ug/L. The lowest concentrations were found at the lower three Clark Fork stations (74% at or below detection). Forty five percent of all samples were at or below the lower reporting limit of 10 µg/L. Concentrations in 2018 appear to be slightly higher than in 2017, with only one sample in excess of 30 ug/L in 2017 vs 7 in 2018.

Figure 7: Ammonia results for summer (upper) and monthly (lower monitoring stations

7.0 NITROGEN – PHOSPHORUS RATIOS

Since the observation of Redfield (1934 and 1958) that marine phytoplankton contains a molecular C:N:P ratio of 106:16:1 (40:7:1 by mass), the relative concentrations of N and P have been used to estimate which of these nutrients might be limiting, preventing additional primary production (algae growth) in aquatic ecosystems. Redfield also recognized that the ratio is an average with considerable variation by species, season, and environment. A departure from this ratio is assumed to imply nutrient deficiency such that by identifying which nutrient is responsible for enhanced algae growth, management actions can be directed toward the nutrient with the highest impact.

It is important to note that the C:N:P ratios in the above literature for benthic algae are for the internal contents of the algal matrix (cellular C:N:P concentration), not water column concentrations. The C:N:P of the benthic algal material is a much better estimator of nutrient limitation than water column TN:TP ratio. This is especially true for benthic algae; while water column total nutrients can be good estimators of optimal stoichiometry for phytoplankton (where suspended algal biomass is a large fraction of the total nutrients in the water column) benthic algae are more loosely coupled with the water column and respond only to bioavailable nutrients (from Kyle Flynn, MDEQ, personal communication).

Total nitrogen-phosphorus ratios (by mass) were calculated for 2018 results and are shown below in Table 4. The N:P Redfield ratio (by mass) is 7:1, and the color-coded thresholds in Table 4 are based on the following from Suplee and Watson (2013): “Studies of benthic algae show that it is necessary to move some distance above or below the Redfield ratio in order to be strongly convinced that a lotic waterbody is P or N limited (Dodds, 2003). When a benthic algal Redfield ratio (by mass) is <6, N limitation is suggested, and when it is >10 P limitation is indicated (Hillebrand and Sommer, 1999). Thus, there is a range of N:P values between about 6 and 10 where one can state, for practical purposes, that algal growth is co-limited by N and P.”

We also include dissolved N: P ratios (by mass) in Table 4 with caveats: the Redfield ratio is based on total N: P, but dissolved concentrations may better reflect nutrient limitation if total concentrations are dominated by particulates (including sediment particles and terrestrial material) which are not necessarily reflective of the condition of the benthic algae. The dissolved T:N ratios are simply presented for comparison.

For total N:P ratios, the uppermost 2 stations appear to be N-limited at times, while phosphorous generally becomes the limiting factor the further the stations are below Deer Lodge, with several exceptions in the sampling events above the Little Blackfoot and at Bonita. In past years, N-limitation in the upper river coupled with P-limitation moving downstream– a pattern previously reported by Suplee, et al, 2012 – was more pronounced than in 2018, perhaps because of the unusually high concentrations of TP in the river in 2018, presumably due to extended high flows. Dissolved N:P ratios indicate N limitation extending further downstream than limitations based on total N and P values.

19 Table 4: 2018 Mass-based N:P ratios for Total N:P (upper) and Dissolved N:P (lower)

Total N:P Clark Fork Clark Fork Clark Fork River River Clark Fork River Clark Fork Clark Fork Clark Fork Clark Fork below Silver Bow below River at above Clark Fork River River Clark Fork River River below Clark Fork Cabinet Creek at Warm Deer Little River at above below River at above Flathead Thompson River at Gorge Opportunity Springs Lodge Blackfoot Bonita Missoula Missoula Huson Flathead River Falls Noxon Dam 3.2 8.1 6.1 5.5 4.7 6.1 7.1 8.5 22.6 23.4 3.4 4.5 7.9 7.1 5.6 6.6 7.4 10.9 14.1 23.1 17.5 3.5 5.4 10.7 9.0 6.1 7.6 8.6 14.2 14.7 21.1 9.6 6.1 7.4 13.4 6.8 10.8 13.3 13.3 19.4 18.6 7.7 7.0 6.9 13.3 7.0 10.1 12.7 12.3 16.6 16.0 14.7 6.0 7.5 7.3 18.5 13.6 11.9 12.2 16.3 18.3 18.5 27.9 5.5 7.2 <6 indicates N-limited 7.0 >10 indicates P-limited 7.9 6 - 10 indicates either N or P may be limiting 10.4 11.4 12.0 12.3 20.5 14.3 13.8 20.4 21.0 15.1 19.7 21.1 Dissolved N:P 7.5 3.9 2.7 1.4 0.4 0.3 1.7 2.0 19.4 24.0 3.2 0.6 3.5 0.9 0.4 0.4 1.6 2.9 8.4 6.2 14.4 4.0 1.0 7.8 0.6 0.3 0.5 4.0 20.4 5.6 7.9 25.5 7.3 3.3 12.7 0.7 3.6 8.8 37.1 7.6 23.7 13.5 9.7 1.2 19.0 10.2 2.6 2.2 14.8 14.8 19.6 11.7 9.7 8.1 2.1 64.1 13.0 1.3 22.0 18.2 15.8 18.7 21.7 7.3 12.1 <6 indicates N-limited 11.5 >10 indicates P-limited 13.2 6 - 10 indicates either N or P may be limiting 12.5 3.4 5.9 4.8 7.1 17.1 28.7 17.1 16.1 16.5 15.1 33.5 8.0 BENTHIC ALGAE RESULTS

Benthic algae were sampled according to the QAPP two times at seven sites, in early July and early August. Averages for chlorophyll-a and ash free dry weight from each sample date are shown in Figure 8. Individual sample results for specific dates are included in the 2018 CFR-BASIN Data Tables, attached. Numeric standards for benthic algae chlorophyll-a are established in ARM 17.30.631 on the Clark Fork River below Warm Springs Creek to the confluence of the Flathead River from June 21 to September 21. The standard for summer mean algal chlorophyll-a is 100 mg/m2 and the standard for summer maximum density is 150 mg/m2. Algal density exceeded the summer maximum on only one occasion, in September at the above the Little Blackfoot site. The summer mean algal density standard was exceeded only 2 occasions, above the Little Blackfoot and below Missoula, both in September. Algal densities were slightly lower in 2018 than in 2017, when the maximum standard was exceeded 2 times and the mean 4 times. Algal levels were much lower in both 2018 and 2017 than there were in 2016, when the mean summer standard was exceeded in 12 of 21 samples across 3 sampling events, July- September.

Figure 8: Benthic algae chlorophyll-a and ash free dry weight results

Bars indicate average results for 2 sampling dates in (1) early August and (2) early September. See 2018 CFR-BASIN Data Tables for exact dates.

9.0 PEAK FLOW NUTRIENT MONITORING AND LOADING TO LAKE PEND OREILLE

This section summarizes the results of a technical memorandum to Avista on nutrient loading to Lake Pend Oreille in 2018 (HydroSolutions, 2019, attached). Peak flow nutrient monitoring is done below the Cabinet Gorge Dam to estimate nutrient loading from the Clark Fork watershed in Montana to Lake Pend Oreille in Idaho per the Montana and Idaho Border Nutrient Load Memorandum of Agreement (Border Agreement), established in 2002. Nutrient targets established in the Border Agreement were developed to maintain water quality in the open waters of Lake Pend Oreille from the mouth of the Clark Fork River to the Long Bridge (Highway 95). Nutrient targets are outlined in section VII of the Border Agreement as follows:  An area-weighted euphotic-zone average concentration of 7.3 μg/L total phosphorus (TP) for Lake Pend Oreille,  Total loading to Lake Pend Oreille of 328,651 kilograms per year (kg/year) total phosphorus,  259,500 kg/year total phosphorus from Montana (as measured at Clark Fork River below Cabinet Gorge Dam),  69,151 kg/year total phosphorus from Lake Pend Oreille watershed in Idaho,  Greater than 15:1 total nitrogen to total phosphorus ratio, by mass.

The Border Agreement establishes short-term and long-term exceedances of the established nutrient targets. As stated in the Border Agreement, an exceedance of the target exists when either of the following conditions is documented: 1) A short-term exceedance of the targets (three consecutive years of total phosphorus load increases at the border that are above the targets by greater than 10%). 2) A long-term exceedance of the targets (a ten year average total phosphorus concentration in the lake greater than 7.3 μg/L).

Every year the estimated annual TP load is evaluated against the Border Agreement’s nutrient load target for the Clark Fork River of 259,500 kg/year, and for short-term exceedance of this target. The loading estimation method is detailed in HydroSolutions, 2018. Results of the loading estimate for 2018 and the previous years are presented in Figure 9. In 2018 the estimated TP load was 441,100 kg/yr, well in excess of the target and the second largest TP since reporting began in 1998. The relatively high TP load in 2018 may, in part, have been a result of unusually high inflow to the lake, which was 125 percent of the long-term average. Since 1998, the estimated TP load has exceeded the allocated target load six times: in 2006, 2008, 2011, 2012, 2014, and 2018. Of those exceedances only estimated TP loads in 2011, 2012, and 2018 were greater than 110 percent of the target load, as defined in the Border Agreement in evaluating short-term exceedances. Based on this assessment there has not been a short term TP load exceedance in the past three years or any previous consecutive three year period since 1998 (HydroSolutions, 2019).

Figure 9: Estimated Lake Pend Oreille Nutrient Loads from Clark Fork River 1998-2018 (HydroSolutions, 2019).

10.0 REFERENCES

Dodds, W.K., 2003. Misuse of Inorganic N and Soluble Reactive P Concentrations to Indicate Nutrient Status of Surface Waters. Journal of the North American Benthological Society. 22(2): 171-181.

Hillebrand, H. and U. Sommer. 1999. The Nutrient Stoichiometry of Benthic Microalgal Growth: Redfield Proportions Are Optimal. Limnology and Oceanography. 44: 440-446.

HydroSolutions, 2019. Estimate of 2018 Nutrient Loading from Clark Fork River into Lake Pend Oreille, Technical Memorandum to Avista Corporation.

HydroSolutions, 2018. Clark Fork River Nutrient Water Quality Status and Trends Report, 1998—2017, Helena, MT.

Montana Department of Environmental Quality, 2014. Department Circular 12-A, Montana Base Numeric Nutrient Standards, Helena, MT.

Montana Department of Environmental Quality, 2018. Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program from Headwaters to Below Cabinet Gorge Dam – Quality Assurance Project Plan (QAPP).

Montana Department of Environmental Quality, 2019 QA/QC Report for Clark Fork River Monitoring, QAPP ID: WQPBQAP-10, Water Quality Planning Bureau, June 6.

Redfield A.C., 1934. On the proportions of organic derivatives in sea water and their relation to the composition of plankton. Liverpool University Press, Liverpool, p. 176–192

Redfield A.C., 1958. The biological control of chemical factors in the environment. Am Sci 46:205–221.

Suplee, M.W., Watson, V., Dodds, W., and C. Shirley, 2012. Response of algal biomass to large scale nutrient controls in the Clark Fork River, Montana, USA. Journal of American Water Resources Association 48:1008-1021.

Suplee, M.W., and V. Watson, 2013. Scientific and Technical Basis of the Numeric Nutrient Criteria for Montana’s Wadeable Streams and Rivers—Update 1. Helena, MT: Montana Dept. of Environmental Quality.

U.S. Environmental Protection Agency, 1993. Clark Fork-Pend Oreille Basin Water Quality Study: A Summary of Findings and a Management Plan, EPA Region 10, Seattle, EPA 910/R-93-0

U. S. Geological Survey, 2018. National Water Information System, Web Interface. Graphs created and downloaded May, 2018. https://waterdata.usgs.gov/mt/nwis/rt/

QA/QC REPORT FOR 2018 CLARK FORK RIVER MONITORING

May 2019

Prepared for:

Randy Apfelbeck, Clark Fork Monitoring Project Manager Montana Department of Environmental Quality Water Quality Planning Bureau Monitoring and Assessment Section

Prepared by: Michelle Hauer, Quality Assurance Officer and Jolene McQuillan, Water Quality Database Manager Montana Department of Environmental Quality Water Quality Division

QA/QC Report for Clark Fork River Monitoring 2018

TABLE OF CONTENTS

Table of Contents ...... i Acronym List ...... ii 1.0 Introduction ...... 1 2.0 Field Components ...... 1 Field Documentation...... 1 Chain of Custody Forms ...... 2 Sample Sites ...... 3 Frequency of Field Blanks and Field Duplicates ...... 3 3.0 Sample Handling ...... 3 Preservation and Holding Times from SAPs/QAPP ...... 3 Holding Times ...... 4 4.0 Analysis ...... 5 Required Analytical Methods ...... 5 Required Detection Limits ...... 5 Field Blanks ...... 6 Field Duplicates ...... 8 Rejected ...... 8 General Quality Checks ...... 8 Laboratory QC ...... 8 5.0 QC Summary ...... 9 Flagged Data ...... 9 Completeness...... 9 6.0 Recommended Corrective Actions ...... 9

5/1/2019 Final i QA/QC Report for Clark Fork River Monitoring 2018

ACRONYM LIST

Acronym Definition CFRWQMC Clark Fork River Water Quality Monitoring Committee COC Chain of Custody DEQ Department of Environmental Quality (Montana) DQI Data Quality Indicators DQO Data Quality Objective EPA Environmental Protection Agency HDPE High-Density Polyethylene QA Quality Assurance QAPP Quality Assurance Project Plan QC Quality Control SAP Sampling and Analysis Plan SRP Soluble Reactive Phosphorus TDS Total Dissolved Solids TP Total Phosphorus TPN Total Persulfate Nitrogen WQ Water Quality WQX EPA's Water Quality Exchange System

5/1/2019 Final ii 2018 Clark Form River Monitoring QA/QC Report

1.0 INTRODUCTION

A data quality control (QC) review has been completed on all data collected and submitted to DEQ in 2018 for the Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program. Monitoring activities were performed in accordance with the “Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program from Headwaters to Below Cabinet Gorge Dam-Quality Assurance Project Plan (QAPP)” (QAPP ID: WQPBQAP-10) and associated SAPs for Avista, the City of Missoula, and the University of Montana. The scope of the QC evaluation was to evaluate documentation associated with sampling and measurement (i.e., field logbooks and site visit forms) and laboratory analytical results to verify data quality. The QC evaluation included a review of the data quality objectives (DQOs) and data quality indicators (DQIs) as outlined in the QAPP and an assessment of compliance with the DEQ QA/QC process. The review also included:

• Review of field data sheets to verify calibration and to identify field notes that explain any deviations from the QAPP • Review of field notes and field data sheets for a data logic check and to identify any notes indicating deviations from the QAPP • Review of the sample delivery group to evaluate the overall quality of the data including reporting errors, data omissions, and suspect or anomalous values.

The QC review applies to the monthly and peak flow nutrient monitoring by Avista and the summer nutrient monitoring by the City of Missoula and the University of Montana.

2.0 FIELD COMPONENTS

FIELD DOCUMENTATION Avista Avista provided calibration logs and field forms as part of their data deliverable. There was a detailed calibration log for each monthly sampling date. The YSI meter was usually calibrated the same day as sampling, and the calibration sheets for each sampling event are very clear and include standard solution expiration dates. The only calibration deviations found were for August and September. Augusts calibration was done on 8/13/18, but the monthly samples were not collected until 8/14/18 and 8/15/18. Septembers calibration was done on 9/14/18, but the monthly samples were not collected until 9/15/18. The instruments were not calibrated on the dates of the six peak flow monitoring events below Cabinet Gorge Dam (CFRPO-30).

Avista’s field forms documented all the relevant field metadata including Sample IDs, Station IDs, site names, dates, field measurements, and samples collected. The April and May field forms are missing the Station ID, but it was not requested to add that field to the forms until the May 2018 annual meeting. All the forms had the required fields complete and they were all easy to read.

City of Missoula The City of Missoula provided calibration logs and field forms as part of their data deliverable. There was a detailed calibration log for each set of sampling events that were easy to read and included standard solution expiration dates. The YSI was calibrated up to 7 days before sampling. Many of the ‘%

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

Agreement’ values calculated on the calibration forms are opposite of their actual value. For example, if the standard is 7 and the meter read 7.5, the % Agreement should read 107%, not 93%.

The City of Missoula’s field forms were a consistent format and contained most of the relevant field metadata including Sample IDs, site names, dates, and samples collected. Although the Station ID is on the form, it is not the exact Station ID used to submit the data to DEQ. The time the samples arrived at the lab and the temperature of the cooler was not entered on the field forms even though the fields are highlighted yellow, which appears to indicate it’s required. There are also notes on various forms that were difficult to read.

University of Montana The University of Montana provided benthic algae field forms as part of their data deliverable. For the sites where UM collected water samples, City of Missoula field forms were used and the city submitted those forms as part of their deliverable. The algae field forms documented the collection date, personnel, Station ID, waterbody name, subsite coordinates, number of samples, additional comments, and the date samples were delivered to the UM lab. Samples were always delivered to the lab freezer the same day they were collected.

CHAIN OF CUSTODY FORMS Avista Avista submitted COC forms for each monthly nutrient and peak flow sampling date that included a relinquished signature by Avista field personnel. All of Avista’s monthly nutrient sampling COC forms indicate the samples were transferred from a cooler with ice directly to the freezer. The peak flow COC forms did not include this additional information. The monthly nutrient samples were relinquished by the Avista field personnel the same day to 4 days after the samples were collected. The peak flow samples were relinquished by the field personnel within 5-29 days after the samples were collected. Cooler temperature at the time of sending was not included on the COC. Completed COC forms including a receiving signature by City of Missoula were not received by DEQ. This was requested at the beginning of 2018, but not implemented. For 2019 sampling, Avista should submit all final COC forms to DEQ when the final EDD is submitted.

City of Missoula The City of Missoula did not fill out COC forms for the samples they collected since the samples did not change hands and were always in the City of Missoula’s custody. There is a section at the bottom of their field forms for arrival time to lab, sample handling at lab (refrigerate or freeze), and cooler temperature. Arrival time and cooler temperature was never filled out, and sample handling was always marked as “Freeze”.

University of Montana The University of Montana did not fill out COC forms for the benthic algae samples since the samples did not change hands and were always in the UM’s custody.

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

SAMPLE SITES Avista Avista’s field forms included site name, descriptions, and Station IDs that matched locations specified in the QAPP. The April and May field forms are missing the Station ID, but it was not requested to add that field to the forms until the May 2018 annual meeting.

City of Missoula City of Missoula’s field forms included site names, descriptions, and Station IDs that generally matched locations specific in the QAPP. The Station IDs were missing the CFR prefix, but it was easy to identify which Station ID it matched in the QAPP.

University of Montana UMs field forms included site names, descriptions, and Station IDs that matched locations specified in the QAPP.

FREQUENCY OF FIELD BLANKS AND FIELD DUPLICATES At least one field blank sample and one duplicate sample were collected for each nutrient monitoring event. For the peak flow monitoring below Cabinet Gorge Dam (CFRPO-30), one field blank sample and one field duplicate sample were collected. This frequency met the frequency outlined in the requirements as described in the QAPP.

3.0 SAMPLE HANDLING

PRESERVATION AND HOLDING TIMES FROM SAPS/QAPP • 500 ml sampling bottles with label (1 for each station per sampling event, field blank, and duplicate) • 250 ml sampling bottles (3 for each station per sampling event, field blank, and duplicate) • 1-60 cc syringe to collect water for filter • 0.45 μm disposable or acid rinsed filter (1-2 for each sampling event plus 1 for a field blank) • Field meter(s) and calibration solutions • Chain-of-Custody (COC)/site visit forms, pencils and waterproof pen • Cooler and ice • De-ionized water (I bottle)

The Missoula Laboratory will supply labeled bottles, syringes, filters, preservatives, coolers, field meters, and water chemistry COCs.

Table 3-1: Sampling Volumes, Containers, Preservation, and Holding Times Analytes Filtered? Bottle Size Container Preservation and Storage Holding Time Cool to ≤ 4 °C (on ice) in field, TPN N 250 ml HDPE Bottle 45 days then freeze solid Cool to ≤ 4 °C (on ice) in field, TP N 500ml HDPE Bottle 45 days then freeze solid (on ice)

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

Analytes Filtered? Bottle Size Container Preservation and Storage Holding Time Cool to ≤ 4 °C (on ice) in field, SRP Y 250 ml HDPE Bottle 45 days then freeze solid Dissolved NO2+NO3 Cool to ≤ 4 °C (on ice) in field, Y 250 ml HDPE Bottle 45 days and NH3+NH4 then freeze solid

HOLDING TIMES Analytical holding times were reviewed for Clark Fork River monthly, peak flow, and summer nutrient monitoring. The following results were H flagged for exceeding the method holding times, which are 45 days assuming the sample preservation was followed as described in the QAPP.

Table 3.2: Results H flagged for exceeding method holding time. Holding Sample Analysis Activity ID Characteristic Name Lab Method Time Date Date (days) CFR-18-092018-S Ammonia 4500-NH3(G) 9/20/2018 11/6/2018 47 CFR-28-091518-S Ammonia 4500-NH3(G) 9/15/2018 11/1/2018 47 CFR-29-091518-S Ammonia 4500-NH3(G) 9/15/2018 11/1/2018 47 CFR-30-091518-QC-FB Ammonia 4500-NH3(G) 9/15/2018 11/1/2018 47 CFR-30-091518-QC-FD Ammonia 4500-NH3(G) 9/15/2018 11/1/2018 47 CFR-30-091518-S Ammonia 4500-NH3(G) 9/15/2018 11/1/2018 47 CFR-22-091918-QC-FB Ammonia 4500-NH3(G) 9/19/2018 11/6/2018 48 CFR-22-091918-S Ammonia 4500-NH3(G) 9/19/2018 11/6/2018 48 CFR-25-091918-S Ammonia 4500-NH3(G) 9/19/2018 11/6/2018 48 FHR-26-091918-S Ammonia 4500-NH3(G) 9/19/2018 11/6/2018 48 CFR-10-091818-S Ammonia 4500-NH3(G) 9/18/2018 11/6/2018 49 CFR-12-091818-S Ammonia 4500-NH3(G) 9/18/2018 11/6/2018 49 CFR-15.5-091818-S Ammonia 4500-NH3(G) 9/18/2018 11/6/2018 49 CFR-7-091818-QC-FB Ammonia 4500-NH3(G) 9/18/2018 11/6/2018 49 CFR-9-091818-S Ammonia 4500-NH3(G) 9/18/2018 11/6/2018 49 CFR-30-031318-QC-FB Ammonia 4500-NH3(G) 3/13/2018 5/2/2018 50 CFR-30-052918-S-PF Ammonia 4500-NH3(G) 5/29/2018 7/19/2018 51 CFR-30-052218-QC-FB Ammonia 4500-NH3(G) 5/22/2018 7/19/2018 58 CFR-30-052218-QC-FD Ammonia 4500-NH3(G) 5/22/2018 7/30/2018 69 CFR-30-052218-S-PF Ammonia 4500-NH3(G) 5/22/2018 7/30/2018 69 CFR-28-091518-S Nitrate + Nitrite 4500-NO3(F) 9/15/2018 10/31/2018 46 CFR-29-091518-S Nitrate + Nitrite 4500-NO3(F) 9/15/2018 10/31/2018 46 CFR-30-091518-QC-FB Nitrate + Nitrite 4500-NO3(F) 9/15/2018 10/31/2018 46 CFR-30-091518-QC-FD Nitrate + Nitrite 4500-NO3(F) 9/15/2018 10/31/2018 46 CFR-30-091518-S Nitrate + Nitrite 4500-NO3(F) 9/15/2018 10/31/2018 46 CFR-30-052918-S-PF Nitrate + Nitrite 4500-NO3(F) 5/29/2018 7/18/2018 50 CFR-30-052218-QC-FB Nitrate + Nitrite 4500-NO3(F) 5/22/2018 7/18/2018 57 CFR-30-052218-QC-FD Nitrate + Nitrite 4500-NO3(F) 5/22/2018 7/18/2018 57 CFR-30-052218-S-PF Nitrate + Nitrite 4500-NO3(F) 5/22/2018 7/18/2018 57 CFR-28-091518-S Orthophosphate 4500-P-F 9/15/2018 10/31/2018 46 CFR-29-091518-S Orthophosphate 4500-P-F 9/15/2018 10/31/2018 46 CFR-30-091518-QC-FB Orthophosphate 4500-P-F 9/15/2018 10/31/2018 46 CFR-30-091518-QC-FD Orthophosphate 4500-P-F 9/15/2018 10/31/2018 46 CFR-30-091518-S Orthophosphate 4500-P-F 9/15/2018 10/31/2018 46

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

Holding Sample Analysis Activity ID Characteristic Name Lab Method Time Date Date (days) CFR-30-052918-S-PF Orthophosphate 4500-P-F 5/29/2018 7/18/2018 50 CFR-30-052218-QC-FB Orthophosphate 4500-P-F 5/22/2018 7/18/2018 57 CFR-30-052218-QC-FD Orthophosphate 4500-P-F 5/22/2018 7/18/2018 57 CFR-30-052218-S-PF Orthophosphate 4500-P-F 5/22/2018 7/18/2018 57 CFR-30-050118-S-PF Total Nitrogen, mixed forms 4500-N-C 5/1/2018 6/21/2018 51 CFR-30-060518-S-PF Total Nitrogen, mixed forms 4500-N-C 6/5/2018 8/1/2018 57 CFR-30-052918-S-PF Total Nitrogen, mixed forms 4500-N-C 5/29/2018 8/1/2018 64 CFR-30-052218-QC-FB Total Nitrogen, mixed forms 4500-N-C 5/22/2018 8/1/2018 71 CFR-30-052218-QC-FD Total Nitrogen, mixed forms 4500-N-C 5/22/2018 8/1/2018 71 CFR-30-052218-S-PF Total Nitrogen, mixed forms 4500-N-C 5/22/2018 8/1/2018 71 CFR-30-031318-QC-FB Total Nitrogen, mixed forms 4500-N-C 3/13/2018 5/30/2018 78 CFR-30-031318-QC-FD Total Nitrogen, mixed forms 4500-N-C 3/13/2018 5/30/2018 78 CFR-30-031318-S Total Nitrogen, mixed forms 4500-N-C 3/13/2018 5/30/2018 78 CFR-30-050118-S-PF Total Phosphorus, mixed forms 4500-P-F 5/1/2018 6/19/2018 49 CFR-30-060518-S-PF Total Phosphorus, mixed forms 4500-P-F 6/5/2018 7/24/2018 49 CFR-30-052918-S-PF Total Phosphorus, mixed forms 4500-P-F 5/29/2018 7/24/2018 56 CFR-30-052218-QC-FB Total Phosphorus, mixed forms 4500-P-F 5/22/2018 7/24/2018 63 CFR-30-052218-QC-FD Total Phosphorus, mixed forms 4500-P-F 5/22/2018 7/24/2018 63 CFR-30-052218-S-PF Total Phosphorus, mixed forms 4500-P-F 5/22/2018 7/24/2018 63

4.0 ANALYSIS

REQUIRED ANALYTICAL METHODS All requested parameters specified in the SAPs were reported. All analytical analyses were performed in accordance with the primary method or alternate methods as defined in the QAPP and SAPs.

Table 4.1: Analytical Methods Alternate Method in Parameter Method Reported Method in QAPP/SAP QAPP/SAP Total Phosphorus (TP) 4500-P-F 4500-P-H 4500-P-B & F Total Persulfate Nitrogen (TPN) 4500-N-C 4500-N-C 4500-N-C Nitrate + Nitrite-Nitrogen (NO2+NO3-N) 4500-NO3(F) 4500-NO3(I) 4500-NO3(F) Total Ammonia-Nitrogen (NH3+NH4-N) 4500-NH3(G) 4500-NH3(H) 4500-NH3(G) Soluble Reactive Phosphorus (SRP) 4500-P-F 4500-P-G 4500-P-F

REQUIRED DETECTION LIMITS The laboratory lower reporting limits (LRL) met the project-required detection limits defined in the QAPP and SAPs for all parameters except Ash Free Dry Weight (AFDW) and except where dilution increased the LRL. Although AFDW’s LRL did not meet the SAP and QAPP requirements, the method detection limit (MDL) did.

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

Table 4.2: Detection Limit Variations Lab Lower Lab Method Project Limit in Parameter Project Limit in SAP Reporting Limit Detection Limit QAPP Ash Free Dry Template – 4 g/m2 Template – 0.4 g/m2 Template – 0.4 g/m2 0.5 g/m2 Weight Hoop – 0.1 g/m2 Hoop – 0.01 g/m2

FIELD BLANKS The following field blanks had detects above the lower reporting limit. Field blank detected results are not B flagged.

Table 4.3: Field blanks with detects above the LRL. Activity ID Characteristic Name Result Value (ug/l) LRL (ug/l) MDL (ug/l) CFR-10-082118-QC-FB Ammonia 22.0 10 6.33 CFR-15.5-071018-QC-FB Ammonia 17.4 10 6.33 CFR-2.5-080818-QC-FB Ammonia 15.2 10 6.33 CFR-22-091918-QC-FB Ammonia 15.8 10 6.33 CFR-25-072518-QC-FB Ammonia 11.1 10 6.33 CFR-7-091818-QC-FB Ammonia 37.7 10 6.33 FHR-26-080318-QC-FB Ammonia 12.0 10 6.33 CFR-12-072418-QC-FB Nitrate + Nitrite 4.8 2 1.84 CFR-15.5-071018-QC-FB Nitrate + Nitrite 2.3 2 1.84 CFR-28-071718-QC-FB Nitrate + Nitrite 2.7 2 1.84 CFR-30-051518-QC-FB Nitrate + Nitrite 3.9 2 1.84 FHR-26-080318-QC-FB Nitrate + Nitrite 2.0 2 1.84 CFR-30-041718-QC-FB Total Nitrogen, mixed forms 158 50 39.62

B – Flags: Results that are associated with a field blank are B flagged if the result is <10x the detected blank value. A result is considered associated if it is the same parameter and collected on the same sampling trip. The following results were B flagged for being associated to a field blank detection.

Table 4.4: Results B flagged for being associated to a detected field blank. Activity ID Characteristic Name Result Value LRL (ug/l) MDL (ug/l) (ug/l) CFR-10-071018-S Ammonia 11.5 10 6.33 CFR-10-082118-QC-FD Ammonia 29.0 10 6.33 CFR-10-082118-S Ammonia 18.9 10 6.33 CFR-10-091818-S Ammonia 22.3 10 6.33 CFR-12-071018-S Ammonia 36.1 10 6.33 CFR-12-091818-S Ammonia 56.4 10 6.33 CFR-15.5-071018-QC-FD Ammonia 32.6 10 6.33 CFR-15.5-071018-S Ammonia 27 10 6.33 CFR-15.5-082118-S Ammonia 10.0 10 6.33 CFR-15.5-091818-S Ammonia 30.4 10 6.33 CFR-18-080818-S Ammonia 27.0 10 6.33 CFR-2.5-071018-S Ammonia 8.0 10 6.33 CFR-2.5-082118-S Ammonia 32.3 10 6.33 CFR-2.5-091818-S Ammonia 39.2 10 6.33 CFR-22-072518-S Ammonia 9.7 10 6.33

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

Activity ID Characteristic Name Result Value LRL (ug/l) MDL (ug/l) (ug/l) CFR-22-080318-S Ammonia 14.1 10 6.33 CFR-22-091918-S Ammonia 24.6 10 6.33 CFR-25-072518-S Ammonia 8.4 10 6.33 CFR-25-080318-S Ammonia 6.7 10 6.33 CFR-25-091918-S Ammonia 42.2 10 6.33 CFR-7-071018-S Ammonia 10.9 10 6.33 CFR-7-082118-S Ammonia 13.7 10 6.33 CFR-7-091818-QC-FD Ammonia 23.4 10 6.33 CFR-7-091818-S Ammonia 22.3 10 6.33 CFR-9-071018-S Ammonia 12.4 10 6.33 CFR-9-082118-S Ammonia 30.4 10 6.33 CFR-9-091818-S Ammonia 19.7 10 6.33 FHR-26-072518-S Ammonia 10.9 10 6.33 FHR-26-080318-QC-FD Ammonia 18.4 10 6.33 FHR-26-080318-S Ammonia 13.4 10 6.33 FHR-26-091918-S Ammonia 23.6 10 6.33 CFR-10-072418-S Nitrate + Nitrite 11.6 2 1.84 CFR-12-071018-S Nitrate + Nitrite 7.4 2 1.84 CFR-12-072418-QC-FD Nitrate + Nitrite 3.7 2 1.84 CFR-12-072418-S Nitrate + Nitrite 6.2 2 1.84 CFR-15.5-071018-QC-FD Nitrate + Nitrite 3.9 2 1.84 CFR-15.5-071018-S Nitrate + Nitrite 3.5 2 1.84 CFR-15.5-072418-S Nitrate + Nitrite 3.0 2 1.84 CFR-25-080318-S Nitrate + Nitrite 11.2 2 1.84 CFR-28-071718-QC-FD Nitrate + Nitrite 5.9 2 1.84 CFR-28-071718-S Nitrate + Nitrite 6.7 2 1.84 CFR-29-071718-S Nitrate + Nitrite 11.7 2 1.84 CFR-30-071718-S Nitrate + Nitrite 9.5 2 1.84 CFR-7-072418-S Nitrate + Nitrite 27.1 2 1.84 CFR-9-072418-S Nitrate + Nitrite 47.0 2 1.84 FHR-26-080318-QC-FD Nitrate + Nitrite 14.6 2 1.84 FHR-26-080318-S Nitrate + Nitrite 15.8 2 1.84

J – Flags: The following field blanks are J flagged for having result values between the MDL and LRL.

Table 4.5: Field blanks with detects between the MDL and LRL. Activity ID Characteristic Name Result Value (ug/l) LRL (ug/l) MDL (ug/l) CFR-12-072418-QC-FB Ammonia 6.7 10 6.33 CFR-30-041718-QC-FB Ammonia 7.0 10 6.33 CFR-30-091518-QC-FB Ammonia 9.3 10 6.33 FHR-26-090518-QC-FB Ammonia 7.5 10 6.33 CFR-15.5-071018-QC-FB Orthophosphate 0.8 2 0.66 CFR-22-091918-QC-FB Orthophosphate 0.8 2 0.66 CFR-7-091818-QC-FB Orthophosphate 1.7 2 0.66 CFR-9-090418-QC-FB Orthophosphate 0.7 2 0.66 FHR-26-090518-QC-FB Orthophosphate 0.7 2 0.66

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

FIELD DUPLICATES The following field duplicate was outside the data quality objective for relative percent difference (RPD). As specified in the QAPP, field duplicate RPD should be <25% for duplicate results that are >5 times the lower reporting limit (LRL). Field duplicate results, along with the parent duplicate, that exceed the objective are J flagged.

Table 4.6: Field duplicates with RPD <25% for results that are >5x the LRL. Activity ID Characteristic Result Value LRL MDL Relative Percent Difference Name (ug/l) (ug/l) (ug/l) CFR-25-082218-S Nitrate + Nitrite 15.2 2 1.84 RPD = 101.6% CFR-25-082218-QC-FD Nitrate + Nitrite 46.6 2 1.84

J – Flags: The following results are J flagged for being associated to a field duplicate that exceeded the objective for RPD.

Table 4.7: Results associated with field duplicates that exceeded their RPD objective. Activity ID Characteristic Name Result Value (ug/l) LRL (ug/l) MDL (ug/l) CFR-22-082218-S Nitrate + Nitrite 96.5 2 1.84 FHR-26-082218-S Nitrate + Nitrite 47.4 2 1.84

REJECTED The following result was rejected due to an unusually high detection in field blank. All other results in the field blank were below the LRL. Rejected results are R flagged.

Table 4.8: Rejected results Activity ID Characteristic Name Result Value (ug/l) LRL (ug/l) MDL (ug/l) CFR-30-041718-QC-FB Total Nitrogen, mixed forms 158 50 39.62

GENERAL QUALITY CHECKS The data was reviewed to make sure the total phosphorus results were greater than the orthophosphate (SRP) as well as that the total nitrogen results were greater than both ammonia and nitrate + nitrite combined. No result qualifiers were necessary.

LABORATORY QC The City of Missoula laboratory did not provide a quality control summary, but they did provide QC Reports for each parameter. The QC Reports contained the laboratory QC results including laboratory control blanks, laboratory control standards, and replicate and matrix spikes. All laboratory QC was within limits specified by the specific analytical methods and the QAPP.

5/1/2019 Final xi. 2018 Clark Form River Monitoring QA/QC Report

5.0 QC SUMMARY

FLAGGED DATA The overall project data had: • 1 result R flagged for having an unusually high detection in a field blank • 53 results H flagged for exceeding method holding time • 14 results D flagged for reporting limit increased due to sample dilution • 47 results B flagged for field blank contamination • 4 results J flagged for high field duplicate RPD • 79 results J flagged for having a result value between the MDL and LRL

There were no results flagged for laboratory QC issues. The overall project DQOs and DQIs were met as set forth in the QAPP.

COMPLETENESS There was 1 result that were determined to be unusable (rejected) due to an unknown field or lab issue. That rejected result had an insignificant effect on the overall project completeness rate, which was well above the 90% requirement set forth in the QAPP.

6.0 CORRECTIVE ACTIONS

5/1/2019 Final xi.

TECHNICAL MEMORANDUM

Date: June 20, 2019

To: Paul Kusnierz, Fisheries Biologist, Avista Corporation

From: Lucas Osborne, P.E., HydroSolutions Inc

Subject: Estimate of 2018 Nutrient Loads from the Clark Fork River into Lake Pend Oreille

Introduction HydroSolutions Inc (HydroSolutions) presents this technical memorandum under contract with Avista Corporation (Avista) as part of their water quality monitoring program. Avista’s water quality monitoring program is coordinated with the Clark Fork River Water Quality Monitoring Committee (CFRWQMC) through a Quality Assurance Project Plan (QAPP) and a Sampling and Analysis Plan (SAP) which directs the specific activities of the program.

Avista’s water quality monitoring program includes 1) “regular-monthly” water quality monitoring at three Lower Clark Fork River monitoring stations, and 2) additional “peak flow” monitoring at the furthest downstream monitoring station, Clark Fork River below Cabinet Gorge Dam (CFR- 30), near the Montana-Idaho border. Clark Fork River nutrient loading into Lake Pend Oreille considers both regular-monthly, and peak flow water quality results from station CFR-30. The current regular monthly monitoring includes collection of nine water quality sample sets per year (once per month March through November). Peak flow monitoring includes the collection of six water quality sample sets during the rising limb and peak flow of the runoff hydrograph, which generally occurs in May or June of each year. Note regular monthly monitoring at stations CFR- 28 and CFR-29 occurs once per month July through September.

The compilation of annual nutrient load estimates is specified by the Montana and Idaho Border Nutrient Load Memorandum of Agreement (Border Agreement). This technical memorandum presents the 2018 nutrient load estimates for total phosphorus (TP) and total nitrogen (TN). The TP loading estimate is compared to the nutrient target established in the Border Agreement. In 2018, nine monthly water quality sample sets and six peak-flow sample sets were collected at monitoring station CFR-30. The laboratory analytical results from these water quality samples were used to calculate the nutrient load estimates to Lake Pend Oreille. The CFRWQMC and Montana DEQ completed quality assurance review and validated the data prior to its use for computing the nutrient load estimate.

Montana and Idaho Border Nutrient Load Agreement The Border Agreement was established in 2002 with a goal to work collaboratively between states to protect Lake Pend Oreille from accelerated cultural eutrophication by managing and controlling nutrient loading into the lake. Nutrient targets established in the Border Agreement were developed to maintain water quality in the open waters of Lake Pend Oreille from the mouth of the Clark Fork River to the Long Bridge (Highway 95). Nutrient targets established in section VII of the Border Agreement include:

1. An area-weighted euphotic-zone average concentration of 7.3 micrograms per liter (μg/L) TP for Lake Pend Oreille 2. Total TP loading to Lake Pend Oreille of 328,651 kilograms per year (kg/year) 3. TP loading from Montana of 259,500 kg/year, as measured at the Clark Fork River at the Montana/Idaho state line, below Cabinet Gorge Dam 4. TP loading from Lake Pend Oreille watershed in Idaho of 69,151 kg/year 5. Greater than 15:1 TN to TP ratio. The Border Agreement established short-term and long-term exceedances of the nutrient targets. The Border Agreement states that an exceedance of the nutrient targets exist when either of the following conditions are documented: (a) A short-term exceedance of the nutrient targets consisting of three consecutive years of TP load increases at the border that are above the targets by greater than 10 percent (b) A long-term exceedance of the nutrient targets consisting of a ten-year average TP concentration in the lake greater than 7.3 μg/L. As directed in the CFRWQMC QAPP and in support of the Border Agreement, this technical memorandum assesses nutrient target number three, TP loading from Montana to Lake Pend Oreille. Additionally, an estimate of TN loading from Montana to Lake Pend Oreille is also provided. The estimated TP load is evaluated against the Border Agreement’s nutrient load target from Montana of 259,500 kg/year, and for the short-term exceedance of this nutrient target. Nutrient Loading Estimation Method Consistent with previous work, Clark Fork River nutrient loading (TP and TN) into Lake Pend Oreille was evaluated using the U.S. Army Corps of Engineers (USACE) Flux32 Load Estimation Software model (version 3.37). The Flux32 model is one of three USACE models that comprise the BATHTUB Eutrophication model (Walker 1999). The model uses grab-sample nutrient concentrations, corresponding average daily river discharge, and the complete annual average daily river discharge record to estimate annual mass loading. The Flux32 model provides six methods to synthesize the discharge-nutrient concentration relationship. The model imputes individual sample results onto the complete annual discharge record through stratification of the discharge record and regression analysis for each stratification. Data stratification of the input flows and nutrient concentrations serves to 1) adjust for differences in

2019.06.20 | Page 2 Avista Corporation | Clark Fork River Nutrient Loading 2018 the frequency distribution of sampled and unsampled flow regimes, 2) reduce or eliminate biases associated with some calculation methods and/or sampling program designs, and 3) reduce the standard error variances of the mean loading estimate (USACE 2014). Method 6, Regression Applied to Individual Flows, described in the Border Agreement Technical Guidance document (Tri-State Water Quality Council 2001), has been used in previous Clark Fork River annual water quality reports (Tri-State Water Quality Council 2009), and in five-year nutrient trend reports. Method 6 is used here for consistency and because the coefficient of variation reported for this method is generally low and is adequate for use in mass-balance modeling. Method 6, Regression Applied to Individual Flows, is defined by Walker (1999) and shown in the following equation:

Where:

W6 = estimated mean flux over N days, method 6 (kg/year) c = measured concentration in sample in milligrams per cubic meter (mg/m3) q = measured flow during sample

a = intercept of ln (c)versus ln (q) regression

b = slope of versus regression

3 Q j = mean flow on day j in cubic hectometers per year (hm /year)

= sum over N days in daily flow record j

SE = standard error of estimate for versus regression

For the loading analysis, the model converts nutrient concentrations from micrograms per liter (µg/L) to milligrams per cubic meter (mg/m3) and discharge values from cubic feet per second (cfs) to cubic hectometers per year (hm3/year). The number of stratified discharge regimes calculated by the model is limited by the number of sample results available to impute onto the complete discharge record. For this loading estimate, the following three data stratification were used: 1) less than one-half the mean daily flow (0 cfs to 13,736 cfs), 2) medium flow (13,736 cfs to 54,946 cfs), and 3) two times the mean daily flow (54,946 cfs to 110,604 cfs). For each stratum, a regression equation was applied to estimate a daily load for each flow regime. The sum of daily loads provides the annual estimate.

Consistent with previous work, nutrient loads are estimated using the record of mean daily discharge from U.S. Geological Survey (USGS) station 12392000, Clark Fork River at Whitehorse Rapids, which operated water years 1929 to 2014. This station was located about 2.75 miles downstream of Clark Fork River below Cabinet Gorge Dam (USGS station 12391950). Discharge measurements made by the USGS indicate that about 600 cfs

2019.06.20 | Page 3 Avista Corporation | Clark Fork River Nutrient Loading 2018 groundwater inflow (noted as seepage around Cabinet Gorge Dam) is contributed to the river between USGS station 12391950 and station 12392000 (USGS 2019). The USGS phased out discharge measurement and reporting at station 12392000, and no longer reports discharge there at all. To maintain consistency with previous work, a daily discharge record for station 12392000 is computed by adding 600 cfs to the published daily discharge flow record at USGS station 12391950. Nutrient Water Quality Sample Results and Annual Hydrograph Nutrient load estimates are related to sample timing, magnitude of stream flow, and nutrient sample result concentrations. Table 1 summarizes the data that was input into the model to estimate the annual nutrient load, including nutrient sample concentrations collected from monitoring station CFR-30; and the mean daily discharge at the Clark Fork River at Whitehorse Rapids for the dates nutrient samples were collected. Figure 1 shows the annual hydrograph of the Clark Fork River at Whitehorse Rapids from the computed mean daily discharge. Figure 1 also displays the sample collection dates and the magnitude of TP sample concentrations, in relation to the discharge hydrograph.

Table 1. Nutrient sample concentrations collected from Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program Station CFR-30, Clark Fork River below Cabinet Gorge Dam; and the computed mean daily discharge at the Clark Fork River at Whitehorse Rapids for the dates nutrient samples were collected.

Sample Total Phosphorus Total Nitrogen Mean Daily Discharge (cfs) Collection Date (µg/L) (µg/L) 3/13/2018 6.2 145 19,700 4/17/2018 9.9 173 33,100 5/1/2018 22.5 217 72,700 5/8/2018 27.0 208 96,500 5/11/2018 37.7 228 107,600 5/15/2018 44.8 246 100,600 5/22/2018 25.8 185 101,600 5/29/2018 24.7 173 108,600 6/5/2018 21.5 170 75,600 6/19/2018 15.8 165 56,100 7/17/2018 12.0 148 24,500 8/14/2018 10.7 148 10,110 9/15/2018 9.3 140 6,580 10/22/2018 6.5 128 8,310 11/13/2018 6.5 137 14,500

2019.06.20 | Page 4 Avista Corporation | Clark Fork River Nutrient Loading 2018

120,000 50.0

44.8 45.0 100,000 40.0 37.7 35.0 80,000

30.0 27.0 25.8 60,000 24.7 25.0 22.5 21.5

20.0 Discharge (CFS) Discharge 40,000 15.8 15.0 (µg/l) Phosphorus Total 12.0 10.7 9.9 9.3 10.0 20,000 6.2 6.5 6.5 5.0

0 0.0 1-Jan 1-Feb 4-Mar 4-Apr 5-May 5-Jun 6-Jul 6-Aug 6-Sep 7-Oct 7-Nov 8-Dec stream flow water quality sample TP concentration

Figure 1. Mean daily discharge at Clark Fork River at Whitehorse Rapids (USGS Station 12392000) in 2018, water quality sample collection dates and total phosphorus concentrations of Clark Fork River-Pend Oreille Watershed Water Quality Monitoring Program at Station CFR-30, Clark Fork River below Cabinet Gorge Dam

2019.06.20 | Page 5 Avista Corporation | Clark Fork River Nutrient Loading 2018

Nutrient Loading Results Results of the nutrient loading estimates for 2018 and the previous four years are presented in Table 2. Mean annual discharge (i.e. the average of the mean daily discharge for the year) in 2018 was greater than each year since 2011, at 125 percent of the long-term average discharge. The mean annual discharge in 2018 ranks as the second-greatest since water quality monitoring and reporting started for this program in 1998. Estimated TP and TN loads in 2018 are 441,1000 kg and 4,291,000 kg, respectively. This estimated nutrient load exceeds the target TP load of 259,500 kg/year by 70% and is the second-largest TP load estimated for this project since reporting began in 1998. In the last five years, the target TP load was exceeded in 2014 and 2018. Since 1998, the estimated TP load has exceeded the allocated target load a total of six times, including 2006, 2008, 2011, 2012, 2014, and 2018. Of those exceedances, estimated TP loads in 2011, 2012, and 2018 were more than 10 percent greater than the target load. However, a short-term exceedance has not been observed for TP loading because the nutrient target has not been exceeded by 10 percent for three consecutive years. No target load is established for TN loading to Lake Pend Oreille.

Table 2. Estimated Lake Pend Oreille nutrient loads from Clark Fork River and Clark Fork River mean annual discharge and inflow volumes, 2014 to 2018 Annual Mean Annual Inflow Discharge Volume TP Load TN Load Year ac-ft x cfs % of Average hm3 kg x 1000 lb x 1000 kg x 1000 lb x 1000 1000

2014 25,598 117% 22,799 18,483 260.5 574.3 3,720 8,200

2015 17,764 81% 15,865 12,862 132.3 291.6 2,179 4,803

2016 18,752 85% 16,794 13,615 141.2 311.3 2,355 5,193

2017 26,461 121% 23,633 19,160 258.5 569.8 3,347 7,379

2018 27,473 125% 24,537 19,892 441.1 972.4 4,291 9,460 Notes: cfs - cubic feet per second lb – pound mass TP – Total Phosphorus hm3 - cubic hectometer kg – Kilogram TN – Total Nitrogen ac-ft - acre feet -The average of published annual mean discharge 1929 to 2013 (based on complete calendar year) at Clark Fork River at Whitehorse Rapids near Cabinet Idaho (USGS gaging station 12392000) is 21,950 cfs. - The Montana and Idaho Border Nutrient Load Memorandum of Agreement established a target nutrient load of 259,500 kilograms per year TP from the Clark Fork River to Lake Pend Oreille.

Estimated TP and TN loads to Lake Pend Oreille from the Clark Fork River from 1998 to 2018 along with the annual mean discharge expressed as a percent of long-term average are presented in Figures 2 and 3, respectively. Flux32 model outputs of 2018 estimated TP and TN loads are provided in Attachment 1.

2019.06.20 | Page 6 Avista Corporation | Clark Fork River Nutrient Loading 2018

Estimated Loading into Lake Pend Oreille from Clark Fork River

600 160%

528 140% 500 441 120%

400 100%

308 312 300 285 80% 261 258 229 235 215 Target Load 259,500 Kg 60% 200 179 170 159 159 150 149 141 140 132 40% Phosphorus Load (Kg x 1,000) x (Kg Load Phosphorus 123 100 91

20% Average of percnet Discharge Mean Annual

0 0% 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018

Phosphorus Load Annual Mean Discharge

Figure 2. Estimated total phosphorus load from the Clark Fork River 1998 to 2018, annual mean discharge as percent of long-term average (1929 to 2013) at USGS Gaging Station 12392000 Clark Fork at Whitehorse Rapids, and Montana-Idaho Nutrient Border Agreement Clark Fork River allocated annual target load

2019.06.20 | Page 7 Avista Corporation | Clark Fork River Nutrient Loading 2018

Estimated Loading into Lake Pend Oreille from Clark Fork River

6,000 160%

140% 5,000 4,842

4,291 120% 4,050 3,966 4,000 3,720 100% 3,401 3,347 3,055 2,882 3,000 80% 2,525 2,587 2,482 2,355 2,254 2,279 2,179 60% 1,955 1,900 1,848 2,000 1,813 1,535

Nitrogen Load (Kg x 1,000) x (Kg Load Nitrogen 40%

1,000

20% Annual Mean Discharge percnet of Average of percnet Discharge Mean Annual

0 0% 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018

Nitrogen Load Annual Mean Discharge

Figure 3. Estimated total nitrogen load from the Clark Fork River 1998 to 2017 and annual mean discharge as percent long- term average (1929 to 2013) at USGS Gaging Station 12392000 Clark Fork at Whitehorse Rapids

2019.06.20 | Page 8 Avista Corporation | Clark Fork River Nutrient Loading 2018

Quality Control The Flux32 model uses the mean coefficient of variation (CV) as a measure of error. The CV equals the standard error of the estimate expressed as a fraction of the predicted value (e.g. a CV of 0.2 indicates that the standard error is 20 percent of the mean predicted value). Data stratification can serve to reduce the CV of the mean loading estimate (Walker 1999). According to Walker, CV values of less than 0.1 are usually adequate for use in mass balance modeling. CV values between 0.1 and 0.2 may be adequate for modeling purposes (Walker 1999). The CFRWQMC has not formally established data quality goals or objectives specific to nutrient load estimation or modeling.

Additionally, Walker (1999) provides a method to estimate the 95-percent confidence range for the mean estimated value, assuming that the errors are log-normally distributed around the predicted value. This method is shown by the following equation:

-2CV 2CV Yme < Y < Yme

Where:

Ym = predicted mean value

CV = error mean coefficient of variation

Y = 95-percent confidence range for mean value

A summary of model parameters (sample size and nutrient), error reporting, calculated 95 percent confidence minimum and maximum values, and the method corresponding to the lowest reported error, are presented in Table 3 for the previous five years.

Table 3. Summary of Flux32 model parameters, error reporting, and 95 percent confidence interval for estimated Lake Pend Oreille nutrient loads from Clark Fork River 2014 to 2018.

95% confidence interval of Mean Estimated Load Lowest Reported Nutrient Load Sample Mean Coef. of (kg x 1,000)* Error Method Year Parameter Size Variation (CV) Min. Max. (CV) TP 15 0.079 222.4 305.1 6 (0.079) 2014 TN 15 0.087 3,126 4,427 2, 3 (0.07) TP 9 0.129 102.2 171.2 2, 3 (0.07) 2015 TN 9 0.063 1,921 2,471 2 (0.035) TP 12 0.045 129.0 154.5 4 (0.043) 2016 TN 12 0.040 2,174 2,552 4, 5 (0.038) TP 15 0.051 233.4 286.2 3 (0.036) 2017 TN 15 0.090 2,796 4,007 3 (0.035) TP 15 0.097 363.3 535.5 4 (0.082) 2018 TN 15 0.036 3,993 4,612 4,6 (0.036) *assumes that the errors are log-normally distributed around the predicted (estimated mean) value

2019.06.20 | Page 9 Avista Corporation | Clark Fork River Nutrient Loading 2018

The 2018 TP loading estimate for method 6 had a CV of 0.097, and had a 95 percent confidence interval ranging from 363,285 kg/year to 535,493 kg/year. The CV for the 2018 TN estimate was 0.036, and had a 95 percent confidence interval ranging from 3,993,073 kg/year to 4,611,536 kg/year. The standard error for the 2018 TP and TN load estimates are less than 10 percent standard error, which is acceptable based on guidance from Walker (1999). Model load estimates from method 4 had the lowest standard error for the 2018 TP loading evaluation. Model load estimates from method 4 and method 6 had the lowest standard error for 2018 TN loading evaluation. Regardless, loading estimates from method 6 are used in this analysis to maintain consistency with previous estimates from previous years. The increases in reported standard error for TP in 2018 over the previous year may be due to increased flow variation over the year. However, it is noteworthy that the reported standard error for TN is considerably lower than the previous four year. As shown in Figure 1, it appears that sample collection was appropriately timed to coincide with the rising limb of the runoff and peak flow hydrograph, which is expected to improve model precision.

Cumulative frequency distribution plots of sample collection and discharge rates were developed by HydroSolutions in 2016 to help Avista understand how well their program’s sample collection covered the overall occurrence of stream discharge. The study concluded that sample collection could be improved to better match natural occurrences of flow from about 20,000 to 60,000 cfs, coinciding with the rising limb of the hydrograph at Station 30 (HydroSolutions Inc 2016). As shown in Figure 1 the Clark Fork River peaked in mid to late May 2018 and sample collection appears to have captured the rising hydrograph limb. Conclusions Nutrient loading of TP and TN from the Clark Fork River to Lake Pend Oreille varies from year to year in relationship to the mean annual discharge and total volume of inflow from the watershed. Since 1998, in most years when Clark Fork River mean annual discharge exceeds the long-term annual average discharge, the estimated TP load has exceeded the Border Agreement’s allocated target load of 259,500 kg per year. While the 2018 TP loading estimate follows this general trend, it is not a linear relationship. The mean annual discharge in 2018 was 125 percent of the long-term annual average and the estimated TP load was 170 percent of the target load established by the Border Agreement. This is in contrast to the 2017 estimated TP load which was at 99 percent of the target load and 121 percent of the long-term annual discharge. 2018 water quality results showed elevated levels of TP during the rising and falling limb of the runoff hydrograph which are primary inputs into the loading model. Since 1998, the estimated TP load has exceeded the allocated target load six times, while the mean annual discharge has exceeded the long-term mean annual discharge nine times. Since the monitoring program’s inception in 1998, 2018 is the year with the second highest estimated TP load and the second highest mean daily flow. The estimated TP load from the Clark Fork River to Lake Pend Oreille was greater than 110 percent of the target load in 2018, but it was less than the target load for the previous two years (2016 and 2017). Based on this assessment there is no short-term exceedance of the Border Agreement target TP load in the past three years or any previous consecutive three-year period since 1998.

2019.06.20 | Page 10 Avista Corporation | Clark Fork River Nutrient Loading 2018

Attachment 1 U.S. Army Corps of Engineers Flux32 Load Estimation Software model program outputs from, 2018 TP and TN load estimate

2019.06.20 | Page 11 FLOW AND LOAD SUMMARIES FOR TotalPhosphorus(ug/L)

Method: C/Q Reg3(daily) (6) DISTRIBUTION OF SAMPLES VS. DAILY FLOWS Daily Flow Smpl FlowTotalPhosphorus(ug/L) Flux SLOPE Stratum Flows Smpls Evnts Vol % (CFS) (CFS) (µg/L) (kg/y) LgC/LgQ R² p > C/Q 1 Flow < 1/2 Me 138 3 3 13.4 9703.478 8333.333 8.8333 84385 0.268 0.05 0.8486 2 Medium Flow 175 4 4 42.0 24041.14 22950 8.65 215782 0.6746 0.53 0.2714 3 Flow > 2x Mea 52 8 8 44.7 86178.85 89912.5 27.475 2145785 1.066 0.60 0.0231 Overall 365 15 15 100.0 27472.82 55740 18.727 441063 0.561 0.78 0.0001

STRATUM BOUNDARIES(CFS) ------STRATUM LOWER LIMIT UPPER LIMIT Flow < 1/2 Mean 0 13736.4 Medium Flow 13736.4 54945.6 Stratum 3 54945.6 110604

------

DAILY FLOW STATISTICS Daily Flow Duration 365 Days = 0.999 Years Daily Mean Flow Rate 27472.80 (CFS) Daily Total Flow Volume 24536.50 (Mega m³) Daily Flow Date Range 01/01/2018 to 12/31/2018 Samples Date Range 03/13/2018 to 11/13/2018

LOAD ESTIMATES FOR TotalPhosphorus(ug/L) Flw Wgted Method Mass(kg) Flux(kg/y) Flux Variance Conc.(µg/L) C.V. 1 Average Load 442822 443125 3.7379589E9 18 0.138 2 Flw Wghted Conc. 437603 437902 1.5161445E9 17.8 0.089 3 Flw Wghted IJC. 440698 441000 1.4947839E9 18 0.088 4 C/Q Reg1 427919 428212 1.2359431E9 17.4 0.082 5 C/Q Reg2(VarAdj) 434355 434652 1.6432848E9 17.7 0.093 6 C/Q Reg3(daily) 440761 441063 1.8152604E9 18 0.097 8 Time Series* 428927 429221 N/A 17.5 N/A

------*Time series estimates use residual interpolation. Maximum Interpolation Gap is set at 26.44 days FLOW AND LOAD SUMMARIES FOR TotalNitrogen(ug/L)

Method: C/Q Reg3(daily) (6) DISTRIBUTION OF SAMPLES VS. DAILY FLOWS Daily Flow Smpl FlowTotalNitrogen(ug/L) Flux SLOPE Stratum Flows Smpls Evnts Vol % (CFS) (CFS) (µg/L) (kg/y) LgC/LgQ R² p > C/Q 1 Flow < 1/2 Me 138 3 3 13.4 9703.478 8333.333 138.67 1234094 0.1132 0.11 0.7756 2 Medium Flow 175 4 4 42.0 24041.14 22950 150.75 3400360 0.27 0.89 0.0550 3 Flow > 2x Mea 52 8 8 44.7 86178.85 89912.5 199 15402145 0.283 0.20 0.2649 Overall 365 15 15 100.0 27472.82 55740 174.07 4291177 0.1649 0.72 0.0002

STRATUM BOUNDARIES(CFS) ------STRATUM LOWER LIMIT UPPER LIMIT Flow < 1/2 Mean 0 13736.4 Medium Flow 13736.4 54945.6 Stratum 3 54945.6 110604

------

LOAD ESTIMATES FOR TotalNitrogen(ug/L) Flw Wgted Method Mass(kg) Flux(kg/y) Flux Variance Conc.(µg/L) C.V. 1 Average Load 4214813 4217700 1.7908863E11 172 0.100 2 Flw Wghted Conc. 4255839 4258754 2.4991615E10 173 0.037 3 Flw Wghted IJC. 4273396 4276323 2.7998796E10 174 0.039 4 C/Q Reg1 4257519 4260435 2.2999564E10 174 0.036 5 C/Q Reg2(VarAdj) 4277654 4280584 2.2562177E10 174 0.035 6 C/Q Reg3(daily) 4288240 4291177 2.4225641E10 175 0.036 8 Time Series* 4232847 4235746 N/A 173 N/A

------*Time series estimates use residual interpolation. Maximum Interpolation Gap is set at 26.44 days Avista Corporation | Clark Fork River Nutrient Loading 2018

References HydroSolutions Inc. 2016. "Technical Memorandum - Peak Flow Monitoring Evaluation." Tri-State Water Quality Council. 2001. "Montana and Idaho Border Nutrient Load Agreement Technical Guidance." Montana DEQ Water Quality Planning Bureau Other Water Quality Links. January. Accessed May 6, 2014. http://www.deq.mt.gov/wqinfo/pdf/TechGuidanceFinal.pdf. Tri-State Water Quality Council. 2009. Water Quality Status and Trends in the Clark Fork-Pend Oreille Watershed: Time Trends Analysis for the 1984-2007 Period. Helena: PBS&J. USACE. 2014. Flux32 Load Estimation Software, Version 3.37, Help Version 05/20/2013. United States Army Corps of Engineers, November 21. USGS. 2019. USGS 12392000 Clark Fork at Whitehorse Rapids NR Cabinet ID. April 25. Accessed April 25, 2018. https://waterdata.usgs.gov/nwis/uv?site_no=12392000. Walker, W.W. 1999. Simplified Procedures for Eutrophication Assessment and Predicion: User Manual. Updated April, Instruction Report W-96-2. Vicksburg: U.S. Army Corps of Engineers Waterways Experimentation Station.

2019.06.20 | Page 12