The Effects of Pollution Reduction

on a Wild Trout Stream

Baseline Studies Report: 2005

November 2005

Note: The 2005 & 2006 baseline report and the 2007 progress report are posted.

The Effects of Pollution Reduction on a Wild Trout Stream

Baseline Studies Report: 2005

Introduction

Spring Run is a unique aquatic resource in the Potomac Highlands region of West Virginia.  Unlike many small headwater streams that tend to go dry, it is fed by the largest spring in the region, with discharge typically ranging from 3000-3500 gallons per minute.  With a temperature of ~53 °F at the spring and a pH of ~8, aquatic conditions are ideal for trout and the aquatic insects they eat.  Spring Run flows about two miles from the spring source to its confluence with South Mill Creek, which is about four miles from the South Branch of the Potomac River.  Spring Run has no tributaries. Much of the stream is shallow, and does not provide the complex habitat that trout need - but that is not the case in a three-fourths mile section in the middle of the Run. 

Since the early 1960’s, landowner’s have issued permits for fly fishing, catch-and-release on three-fourths mile of Spring Run.  Landowners and other interested parties have installed and maintained various structures to form pools and overhead cover that provide hiding and feeding habitat for trout.  Spring Run is recognized as one of the best "wild" rainbow trout fisheries in West Virginia.  Friends of Springs Run’s Wild Trout, was formed in 1996 to restore structure to Spring Run following flooding in 1996.

In the last few years, however, fishermen have noted a decline in the fishery. Emergence of the mayfly, Ephemerellidae (sulfurs) disappeared in the late 1990s. The number of large trout (14” and above) has decreased and trout in the 11-13” range are also fewer. The population of trout is considerably lower in the lower reach of the three-fourths mile section.  Algae formation is heavy in the upper reach of the three-forth mile section, much heavier than in the past, and algae reforms soon after washout by high water.

There have long been plenty of nutrients in Spring Run, contributed largely in effluent from the Spring Run Trout Hatchery located about one-half mile upstream near the spring. (SRH is a rearing facility; trout are not spawned there). In recent years, however, SRH has been producing more rainbow and “golden trout” for stocking West Virginia streams, and it seems that the effluent stream now may be a problem for the health of Spring Run. WVDEP issued a citation for violation of the Spring Run Trout Hatchery NPDES permit in January 2004, specifically for discharging excess biochemical oxygen demand (BOD) and total suspended solids (TSS).  WVDNR, which operates SRH, is now preparing to install an effluent treatment process at the facility to meet their permit requirements. 

Installation of effluent treatment at SRH provides a unique opportunity to address a number of issues of both regional and national significance:

1. Will the hatchery effluent treatment process significantly reduce nutrient discharge?  Fish hatcheries throughout the country produce nutrient-rich effluents of concern to receiving waters.  This study will evaluate the downstream result of effluent reduction of BOD and TSS, as well as nutrients, from a small but high throughput point source. The results of renovation at SRH and this study will provide important information to the WV Potomac Tributary Strategy point source innovation process.

2. What are the biological impacts of Spring Run's high nutrient levels, and how is the biota affected by reductions in nutrients, TSS and BOD following hatchery upgrades?  This issue is of importance to the nutrient criteria development process that WV and the other 49 states are currently struggling through, as one of the key questions is: "what does nutrient impairment look like?"

3. Is the wild trout population in Spring Run being harmed by hatchery effluent, and does improvement in that effluent improve the trout fishery?

4. Is the benthic invertebrate population in Spring Run being harmed by hatchery effluent, and does improvement in that effluent improve diversity?  Spring Run fishermen have noted the loss in recent years of a certain family of mayflies, the Ephemerellidae (Spiny crawler mayfly) that used to emerge regularly in the springtime.  Also, WV DEP’s Tim Craddock did a benthic assessment of Spring Run in 2002, and found the lower part of the fly fishing section to be dominated by Chironomidae (midge) larvae, a group often indicative of pollution by organic waste.

5. Why do trout, especially larger fish, favor the upper part of the fly-fishing section?  Why has the density-center of the trout population moved upstream in recent years? Is there a relationship between distribution of benthic invertebrates in the stream and trout distribution?  If the Ephemerellidae mayfly rebound after the hatchery effluent is treated, will the trout population improve also? In particular, are trout avoiding areas they used to frequent that are now dominated by midge larvae? If upgrades to the hatchery reduce organics in the stream and also the midge populations, will trout return to those areas?  If that turns out to be true, and we could demonstrate that it is true, that would buttress public acceptance of benthic invertebrate stream assessments.

Overall, this project will have the potential to be used to address many questions beyond the five questions identified above.

Partners

Friends of Spring Run’s Wild Trout, Cacapon Institute (CI), the WV Conservation Agency (WVCA), WV Department of Agriculture (WVDA), WV Division of Natural Resources (WVDNR), WV Department of Environmental Protection (WVDEP), and the Freshwater Institute are partnering in this study. This project is funded primarily by West Virginia Conservation Agency’s participation through the Chesapeake Bay Program.  An associated sediment reduction project is funded through a Friends of Spring Run’s Wild Trout 2005 Stream Partners Grant.

WVDA, WVDEP and WVDNR are all contributing in-kind services to the project.  WVDA is collecting water samples, taking flow measurements, and performing field and laboratory water quality analyses.  WVDEP is participating in collections of benthic invertebrate and periphyton and helping to cover the costs of analysis.  WVDNR is performing fish surveys and Friends of Spring Run’s Wild Trout is providing information on size and location of trout caught and released by permitted fly fisherman.

The Freshwater Institute provided guidance to WVDNR on treatment methods for their effluent and is providing technical guidance for the project.  WVCA is acting as project coordinator.  Cacapon Institute has overall technical oversight for the project, will participate in field work, and will, in cooperation with partnering organizations, be responsible for data analysis and production of annual reports. 

Methods

The project has two experimental components, an upstream/downstream design in Spring Run, and a control/experimental design that includes Dumpling Run, another spring fed stream nearby.  Both streams have their origins in the same geology: limestone (Helderberg and Tonoloway/Wills Creek) and sandstone (Oriskany, McKenzie) formations.  Spring Run flows off the ridge to the northwest into South Mill Creek, a tributary of the South Branch of the Potomac River.  Dumpling Run flows east into the South Fork of the South Branch of the Potomac River.

The upstream/downstream part includes three sites in Spring Run: the first site is near the spring upstream of the hatchery; the second site is in the upper part of the fly fishing stream section; and the third is in the lower part of the fly fishing section. There are two sites on Dumpling Run, one just below the spring, the other some distance downstream.   Overall, this design allows within stream and between stream comparisons.  Under most conditions of flow the springs constitute the main source of water in both streams, but both streams also have periodic surface flow entering the main channel upstream of the spring.

Water chemistries are collected monthly from April through September, typically on Wednesday.  We chose to avoid collections on Mondays at the time of the hatchery cleanout because the "biosolids from the aquaculture effluent are notoriously patchy and difficult to characterize in sampling.  . . . my thoughts on the nutrients is to focus on the residual chronic impacts, not the pulse of the cleaning plume" (Joe Hankins, Freshwater Institute, personal communication). 

Water quality parameters include nitrogen in the forms of ammonia-nitrogen, nitrate/nitrite, total Kjeldahl nitrogen, total nitrogen (the sum of nitrate/nitrite and TKN), soluble reactive phosphorus, total phosphorus, total suspended solids (TSS), biochemical oxygen demand (BOD₅₎, and basic field parameters (pH, temperature, conductivity) (see Appendix 2 for laboratory methods).  Flow measurements are collected at the same time as water samples at one site in each stream.  This work is done primarily by the WVDA.

Benthic invertebrate and periphyton samples are collected twice each year at all sites, in May and August, according to the standard protocols in use by the WVDEP.  WVDEP format Rapid Bioassessment Protocol habitat analyses will be conducted once each year.  Primarily WVDEP and Cacapon Institute will do the fieldwork for this component.

WVDNR will conduct electro shocking fishery assessments, and the permitted fly fishermen of Spring Run have been enlisted to record information on size and location of trout caught and released.   

Since changes to the system may not occur rapidly, an assessment will be made at the end of the third year to determine if “out year” monitoring might be needed?

The methods used to analyze water quality data were graphical and statistical.  Data distributions were displayed using box plots (figure at right), which are useful for side-by-side visual comparisons of data distributions.  Because the data set is small (six data points at each site), traditional box plots with 25th and 75^th quartiles were deceptive.  Rather than use quartiles, the box boundaries are the 2^nd and 5^th data point in each series.  One way analysis of variance (ANOVA) was run on rank transformed data for comparison of median concentration distributions.  An alpha value of 0.05 was used as the threshold for statistical significance. If a significant difference among group medians was detected, Tukey’s multiple comparison test was used on the rank transformed data to determine where differences were located (Helsel and Hirsh, 1992).  Statistics were calculated using JMP Statistical Discovery Software (version 4.0.2).   Summary statistics and raw data are provided in Appendix XX.

Baseline Water Chemistry & Flow Data Results

Pre-treatment results and analysis the water quality data will focus on five questions:

1. How does the spring source water of the two streams compare?  It is assumed that the springs constitute the main source of water in both streams, certainly true at most conditions of flow.  Note: both streams have surface flow entering the main channel upstream of the spring.
2. How does the water in the control stream change as it flows downstream?
3. How does the water in the experimental stream change as it flows downstream?
4. Are there significant differences in water chemistry at any of the sites?
5. How did water quality vary over time?

Flow measurements were taken at the Dumpling Run Lower and Spring Run Bottom sites.   Flow in Dumpling Run ranged from about one third to one half of the flow in Spring Run (figure at left).  Water samples were collected on three days with fairly low water (June, July, and September), two moderate flow (April and May), and one high water (August).

Since we are most concerned with local effects in this study, concentration is the most relevant way to look at the data.  However, flow is necessary for interpretation of the time series data presented below. 

The flow stations are not suitable surrogates for flows at all of the stations.  This is particularly an issue in Spring Run, where a significant portion of the total stream flow is diverted at the springhouse to the trout hatchery and does not flow through the upper channel where samples are collected.  This means that we cannot reasonably estimate parameter loadings at any sites but those with flow measurements.

Correlation Analysis

The following four tables present simple correlation analysis on the un-transformed sample data for key parameters: total N, total P, TSS, BOD5, and flow.  The purpose of the four tables is to partition effects that might be due to different factors, such as point and non point sources of pollution.  The first table offers correlations on all sites, the second Spring Run only, the third Dumpling Run only, and the fourth excludes point source impacted sites in Spring Run.  More sophisticated approaches will be used in future reports when the size of the data set makes them more appropriate.