Subwatershed Prioritization of the Watershed Using Baseflow Water Quality Monitoring Data

Bradley J. Austin, Brina Smith, and Brian E. Haggard Eutrophication

Process by which excess nutrients entering aquatic systems promotes increased aquatic plant and algal growth reducing water quality.

• Point Source: Single identifiable source from which pollutants are discharged – Pipe or outflow from a waste water treatment plant (WWTP) Eutrophication

Process by which excess nutrients entering aquatic systems promotes increased aquatic plant and algal growth reducing water quality.

• Nonpoint Source: Diffuse sources of pollutants from an altered landscape (agricultural and urban land use) that enter waterways during runoff events Lake Wister Watershed

Lake Wister is listed on: ’s 303 D List of Impaired Waters • Total Phosphorus • Color • Turbidity • pH • Chlorophyll a • Mercury PVIA’s Strategic Plan to Improve Water Quality and Enhance the Lake Ecosystem 3 Zones of action:

1. Lake Wister 2. Quarry Island Cove – near PVIA’s intake 3. The watershed PVIA’s Strategic Plan to Improve Water Quality and Enhance the Lake Ecosystem 3 Zones of action:

1. Lake Wister 2. Quarry Island Cove – near PVIA’s intake 3. The watershed The goal of this study was to collect water quality data that could be used by PVIA to prioritize where to invest resources to improve water quality in the watershed and ultimately Lake Wister. Field Monitoring

• 26 sites selected representing 23 HUC 12 subwatersheds flowing into Lake Wister. • Sampled monthly during base flow conditions for a year. • Samples were analyzed for: • Total Phosphorus • Soluble Reactive • Chlorophyll a • Total Nitrogen Phosphorus • Fluoride • Nitrate + Nitrite • Turbidity • Chloride • Ammonia • Total Suspended Solids • Sulfate • Geometric means of the constituents were compared to a Human Development Index (HDI) – This is just the total percentage of agriculture and urban land use in a watershed.

• • • spec No seasonalpatterns for N ch No seasonalpattern in summer. s SRPBoth TP and were lightly elevated in the elevatedlightly the in • lorophyll Variability ies. T Possiblyto due elevated Seasonal SS in the summer.SS in

a

NO -N (mg L-1) -1 3 -1

TSS (mg L ) 0.001 TN (mg L ) 0.01 0.01 100 0.1 0.1 0.1 10 10 1 1 1 C A E Annual pSu Sp Fa Wi

-1 -1 SRP (mg L ) -1

0.001 TP (mg L ) Chl a (mg L ) 0.001 0.01 0.01 100 0.1 0.1 0.1 10 1 1 1 D F B Annual pSu Sp Fa Wi Regression Analysis looks at how water quality parameters respond to increasing human development in the watershed

Nutrients, sediment, and algal biomass increase with human development (HDI) in the watershed

2.0 0.30 A r2 = 0.783 B r2 = 0.707 p < 0.001 0.25 26 p < 0.001 1.5 23

) y = 0.022x + 0.018 y = 0.003x - 0.004

) 23 0.20 -1 -1

1.0 0.15

2 1 22 8 TP (mg L TN (mg L 20 0.10 8 26 24 19 0.5 21 222 19 3 21 1 20 25 17 16 24 3 9 0.05 7 15 25 17 16 9 13111210 18 7 18 1445 6 1445 61311121015 0.0 0.00

35 15 C r2 = 0.584 D r2 = 0.594 30 23 p < 0.001 p < 0.001 23 )

) 25 y = 0.377x - 0.986 y = 0.204x - 0.225 2 -1

-1 10 8

20 g L m (

15 a 24 22 19 1 5

TSS L (mg 22 10 Chl 3 21 2 3 26 19 26 24 8 7 17 21 20 25 20 16 5 16 1 25 17 15 18 9 71215 9 6 10 1445 6131110 18 1445 131112 0 0 0 10 20 30 40 50 0 10 20 30 40 50

HDI (% Land Use) 2.0 10 0.25 1 A B Across State Lines 1 0.1 0.1 0.20 1.5 ) ) 0.01 0.01 -1 ARK HUC 12 OK HUC 12 -1 0.15 ARK HUC 12 OK HUC 12 • Black Circles: 1.0 0.10 TP (mg L • LWW in Oklahoma TN (mg L 0.5 (2016-2017) 0.05

0.0 0.00 • Gray circles: 0 20 40 60 80 100 0 20 40 60 80 100 0.5 1 0.05 1 C • Poteau Watershed in D 0.1 0.1 0.01 0.4 0.04

) 0.001 ) (2011-2012) -1 0.01 -1 0.0001 0.3 ARK HUC 12 OK HUC 12 0.03 ARK HUC 12 OK HUC 12

-N (mg L 0.2 0.02 • Slightly greater values 3 SRP (mg L NO reported for Arkansas 0.1 0.01

streams than Oklahoma 0.0 0.00 streams. 0 20 40 60 80 100 0 20 40 60 80 100 60 100 35 100 E F • Except for SRP 30 50 10 10

) 25

40 1 -1 1 ARK HUC 12 OK HUC 12 ARK HUC 12 OK HUC 12 • Likely due to greater human 20 30 development in Arkansas . 15 20 TSS L (mg 10 Turbidity (NTU) Turbidity • Data from both studies fit 10 5 0 0 well together across the HDI 0 20 40 60 80 100 0 20 40 60 80 100

gradient. HDI (% Land Use) Changepoint Analysis allows us to see if and where a shift occurs in how water quality parameters respond to increasing human development in the watershed. For all of the parameters there is a shift or change point at around 20−30% human development, where sites above the changepoint have greater mean values and variability than the sites below this changepoint.

2.0 0.30 A B 0.25 CP = 28% 26 1.5 23 r2 = 0.585 )

) 23 0.20 CP = 21% -1 p = 0.001 -1 r2 = 0.534 1.0 0.15 p = 0.001

2 1 22 8 TP (mg L TN (mg L 20 0.10 8 26 24 19 0.5 21 222 19 3 21 1 20 25 17 16 24 3 9 0.05 7 15 25 17 16 9 13111210 18 7 18 1445 6 1445 61311121015 0.0 0.00

35 15 C D 23 30 23 CP = 22% ) CP = 28% ) 25 2 2 -1 10 2 -1 r = 0.411 r = 0.411 8 20 p = 0.017 g L p = 0.003 m (

15 a 24 22 19 1 5

TSS L (mg 22 10 Chl 3 21 2 26 19 20 26 24 3 8 7 17 16 21 25 20 5 16 1 25 17 15 18 9 71215 9 6 10 1445 6131110 18 1445 131112 0 0 0 10 20 30 40 50 0 10 20 30 40 50

HDI (% Land Use) Option 1: If we do not have WQ data, or data is missing for a portion of the watershed, then HUC 12s can be prioritized based on the amount of human development within their watershed 90th percentile confidence interval about the changepoint

0.30 A 0.25 26 Preservation Low Medium High

) 23 0.20 -1

0.15 Changepoint

TP (mg L 0.10 8 222 19 21 1 20 0.05 24 3 25 17 16 9 7 18 1445 61311121015 0.00 0 10 20 30 40 50 HDI (% Land Use)

• Preservation: HUC 12s with %HDI < 90th percentile confidence interval • Low priority: HUC 12s with %HDI within the 90th percentile confidence interval but less than the changepoint th • Medium priority: HUC 12s with %HDI within the 90 percentile confidence interval but greater than the change point • High priority: HUC 12s with %HDI > 90th percentile confidence interval

Option 1: If we do not have WQ data, or data is missing for a portion of the watershed, then HUC 12s can be prioritized based on the amount of human development within their watershed

0402 0401 0404 0503 0406 0508 0405 0507 0403 0409 0303 0408 0501 0506 0305 0304 0407 0502 0505 0203 0504 0206 0205

0204 0202 • Priorities for nonpoint source management range 0201 from preservation (lightest) to high priority (darkest). • Based on this method, the subwatersheds along the main stem of the Fourche Maline and the stand out as the highest priorities. • Using this method, prioritization of subwatersheds is not influenced by subwatersheds upstream.

Option 2: When water quality data is available

0.30 B High 0.25 26

) 23 0.20 -1 Average above 0.15 threshold

TP (mg L 0.10 8 222 19 Medium 21 1 20 24 3 0.05 Average below 25 17 16 9 7 18 Low threshold + 2 1445 61311121015 0.00 standard 0 10 20 30 40 50 deviations HDI (% Land Use)

• Low priority: HUC 12s with measured values less than the horizontal dashed line.

• Medium priority: HUC 12s with constituent concentrations greater than the horizontal dashed line but less than upper solid line

• High Priority: HUC 12s with constituent concentrations greater than upper solid line. Option 2: When water quality data is available

• We can use this method TP to prioritize subwatersheds based TN on specific water quality TSS parameters. Chl-a • Or, priorities for each parameter can be 0402 added up to create a 0401 cumulative rank for 0404 0406 0503 0405 0508 each HUC 12 0403 0409 0507 0501 0305 0303 0408 0506 0304 Cumulative 0407 0502 0505 0203 0504 0206 0205 • With this approach we must be mindful of how 0204 0202 water quality in the upstream HUC 12s might 0201 influence the water quality in the downstream HUC 12s

Regression analysis can be useful in setting realistic targets for improving water quality.

0.30 C 0.25 26

) 23 0.20 -1 Realistic 0.15 target

TP (mg L 0.10 Unrealistic 8 222 19 target 21 1 20 0.05 24 3 25 17 16 9 7 18 1445 61311121015 0.00 0 10 20 30 40 50 HDI (% Land Use)

• The regression line represents the average conditions for a given level of human development. • The goal should be to improve water quality at high priority watersheds so that they fall below this regression line.

Based on these potential methods of HUC 12 prioritization we selected 4 HUC 12 subwatersheds for further investigation.

Fourche Maline F1 Poteau River

F2

F3

F4 P2 1 F5 2 P1 24 3 P4 P3

25 20 9 24 21 19 23 22 16

7 17 3 2 1

18 4 8 Oklahoma 6 15 5 B1 B4 12

B2 14 13 Arkansas B3 23 11 B5 26 10 13 S2 S3 26 S1 S4 Bandy Creek Shawnee Creek Special Studies

• Both Shawnee and Bandy Creek watersheds had poor water quality due to known WWTPs in their watersheds.

“Are the WWTPs the only source of nutrients to these watersheds?”

• In the Fourche Maline and Poteau River watersheds we wanted to know:

“How far does high nutrient and sediment concentrations extend into the headwaters of the Fourche Maline Watershed?”

“Is Arkansas the only source of sediment and nutrients to the Poteau River, or are Oklahoma tributaries sources of sediment and nutrients as well?”

• Additional sites within each of these watersheds were sampled following the same methods used before. Are the WWTPs the only source of nutrients Upstream Downstream 0.6 to these watersheds? A 0.5

) 0.4 -1 Shawnee Creek 0.3

TN (mg L 0.2 13 0.1

0.0

WWTP 0.30 S2 26 B 0.25 S3 &S4

) 0.20 S1 -1

0.15

TP (mg L 0.10

0.05

0.00

8 C

6 ) -1 • Yes! As you move from upstream to downstream there is an 4 TSS L (mg increase in both nutrients and sediment below the WWTP. 2 • The effect from the WWTP was localized, as can be seen by 0 lower concentrations at the most downstream site. S1 S2 S3 S4 26 13 Site Number • • to to Arethe WWTPs theonly source of nutrients i However,two sites upstreamB2) of the had WWTP(B1 and NutrientsTSS and were greatestWilburton’s below WWTP reflected greater human development in their catchment. ncreased nutrient sediment and concentrations that these watersheds?these • and sediments.and po Installationbestmanagement of practices this (BMPs)in rtion ofthe watershed may be helpful in reducing nutrients Bandy Creek Bandy Watershed B1 B1 B5 B5 B2 B2 B3 B3

B4 B4 WWTP

Site 23 Site

-1 -1 TP (mg L ) TN (mg L-1) TSS (mg L ) 0.00 0.05 0.10 0.15 0.20 0.25 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 10 15 20 25 30 35 0 5 Upstream B A C 3B 1B B4 B2 B1 B5 B3

Site Number Downstream 23

How far does high nutrient and sediment concentrations Upstream Downstream

0.8 extend into the headwaters of the Fourche Maline A

Watershed? 0.6 ) -1

0.4 TN (mg L 0.2 F1

0.0

0.07 F2 B 0.06

0.05 ) -1 F3 0.04 0.03 TP (mg L 0.02

F4 0.01

F5 0.00 24 8 C

6 ) The headwaters of the Fourche Maline have relatively low -1 • 4 nutrient and sediment concentrations. TSS L (mg • Both nutrient and sediment concentrations begin to 2

increase in the lower portion of the watershed as %HDI 0 increases. F1 F2 F3 F4 F5 24 • BMPs installed in the lower portion of the watershed are Site Number likely to have the greatest effect. Upstream Downstream Is Arkansas the only source of sediment and nutrients Main Stem Poteau River Sites 1.0 to the Poteau River, or are Oklahoma tributaries A 0.8

sources of sediment and nutrients as well? ) -1 0.6 0.4 TN (mg L

0.2

WWTP 0.0 1 0.14 P4 P2 B 2 P1 0.12 3 0.10 P3 ) -1 0.08

0.06 TP (mg L 0.04

0.02

0.00

10 C 8

• Nutrient and sediment concentrations were greatest along ) -1 the main stem of the Poteau River. 6 • However, site P4 also had elevated TP and TSS relative to 4 TSS L (mg the other tributaries. 2

• This was likely driven by effluent discharge out of 0 Heavener, Oklahoma. 1 P1 P2 2 P3 P4 3 Site Number Summary

• Routine baseflow water quality monitoring was useful: – For developing potential frameworks for prioritizing the HUC 12s within the Lake Wister Watershed, both with and without water quality data. – In developing linear relationships between water quality parameters and human development, that can be used to set realistic targets for assessing water quality improvements in the watershed. • Following these sampling methods within individual HUC 12 subwatersheds can be useful in pinpointing specific areas in need of NPS best management practices. Future Directions

• 2nd year of LWW water quality monitoring – Examine: • Interannual variability • Seasonal variability over multiple years – Compare LWW water quality in Oklahoma to Poteau sub-basin in Arkansas

Poteau Sub-basin Monitoring

• ANRC 319h funded project to monitor water quality within the Poteau sub-basin of Arkansas. – Baseflow water quality monitoring. – Stage to discharge rating curve development. – HUC 12 load estimation.

Acknowledgements

• This project was funded by Poteau Valley Improvement Authority and Oklahoma Conservation Commission through the 319(h) EPA grant # C9-996100-17. • Laboratory Technicians – Keith Trost – Jennifer Purtle Questions?

• Digital copies of the MSC report can be found at: • https://arkansas-water-center.uark.edu/ – Click the publications link – Click the MSC Technical Reports link – Select MSC385. • You can also e-mail me to request a digital copy: [email protected]