TABLE OF CONTENTS
I. Introduction
Project Objective and History
Harpeth River Watershed Association
Significance of the Study
General Watershed Description
II. Methods and Materials
Visual Assessment Survey Protocol
Training
Site Selection
Data Management
Visual Stream Assessment Database
Quality Assurance
Quality Control
III. Results and Discussion
Specific Findings
Land Use
Erosion
Stream Visual Assessment Scores
Opportunities for Water Quality Improvements
IV. Recommendations, Suggestions for Further Work, and Conclusions
Modifications to the Visual Stream Assessment Protocol
Next Steps
V. Outputs
VI. References
Appendices:
Appendix A-Protocol: Instructions and Sample Field Data Sheet
Appendix B-Quality Control Methods and Results
Appendix C-Visual Stream Assessment Data
Photographs of each site
Topographic maps showing site locations
Table of Streamside land use at each site
Table of Stream Visual Assessment Scores
Appendix D- CD-ROM with Database and Photographs
I. INTRODUCTION
Project Objective and History
In August of 2001, the Harpeth River Watershed Association (HRWA) received a $19,000 grant from the Tennessee Department of Agriculture’s Non-Point Source Program from funds via the Environmental Protection Agency (EPA) to conduct a volunteer site-specific visual stream assessment. The goal of the project was to identify specific sites along 303(d)/305(b) listed streams that may be contributing pollutants associated with the Tennessee Department of Environment and Conservation’s (TDEC) findings of impairment. The objective was that this assessment would help the NPS program, other federal and state agencies, and other entities prioritize where to focus efforts to implement best management practices (BMPs) with the overall goal to improve the water quality of these stream segments to the point that these segments can be removed from the 303(d)/305(b) list.
Harpeth River Watershed Association
The HRWA was founded in late 1998, and currently has
approximately 100 members, and a full-time paid Executive Director as of July
2001. The organization’s mission is to
preserve and re-build the ecological balance of the
Significance of the Study
This study is significant for several reasons. 1) This is
the first time volunteers have been used to identify specific sites causing
impairment along 303(d) listed streams in
General Watershed Description
The
Historically, the majority of land in the watershed has been
used for agricultural purposes.
However, the landscape and uses are shifting dramatically toward
residential and commercial development along two interstate corridors
(Cumberland River Compact, 2000). Many
cities exist within the watershed including
II. METHODS AND MATERIALS
This study involved designing a standard methodology, called
the visual stream assessment protocol, for volunteers to use to visually assess
stream habitat. The study focused
initially on the 303(d)/305(b) listed streams in the
The Project Director is a freshwater water biologist with expertise in volunteer data collection and stream monitoring. The Director designed the protocol, conducted the training sessions, conducted the QA/QC analysis, analyzed the data, and prepared the bulk of the final report. The Project Manager, an environmental engineer, managed the design and compilation of the database, provided analytical support, prepared Appendix C, and contributed to the final report. The Executive Director is a conservation policy specialist with over 15 years of professional experience on conservation issues, including the design and management of volunteer scientific monitoring projects. For this project the E.D. set the overall design of the study with the NPS program, discussed the project with NRCS, TDEC, and other agency and municipal staff, identified the project staff, recruited and managed the volunteer core, prepared the final report, and handled all grant administration.
TDEC staff in the Division of Water Pollution Control
involved in watershed management and monitoring were contacted at the beginning
of the study to get their input and advice on how best to gather data for their
various purposes and to conduct the study.
Similarly each District Conservationist (DC) of NRCS in which the watershed resides was
contacted, in this case with a formal letter and copy of the protocol. The DCs were specifically consulted for
advice on how to train volunteers to discuss the project with a landowner and
avoid trespassing. The NRCS was very
willing to express support of the study to any landowner that wished to contact
them about it. Also, various
municipalities, such as
Visual Stream Assessment Survey Protocol: (copy in Appendix A)
The Visual Assessment Survey protocol used in this study was developed from a number of sources including the Tennessee Valley Authority (TVA) Clean Water Initiative Volunteer Stream Monitoring Methods Manual, U.S. Environmental Protection Agency (USEPA)’s Volunteer Stream Monitoring: A Methods Manual, and U.S. Department of Agriculture (USDA)’s Stream Visual Assessment Protocols. Portions of the field data sheets from the above sources were used to develop the visual stream assessment protocol for use in this project (see Appendix A for a copy of the protocol).
The visual assessment survey protocol included both qualitative and quantitative data collection relative to physical water characteristics and habitat components. The qualitative portion included information on land use, litter, erosion and special problems such as algal blooms. The quantitative sections included information on riparian zone, bank stability, canopy cover, invertebrate habitat, sedimentation, water appearance, nutrient enrichment, and channel
The protocol is included in Appendix A and was designed with eight sections. The sections include 1) general information such as name and contact information of the volunteers, 2) weather data, 3) land uses, 4) other information including litter, erosion and special problems, 5) pictures and descriptions, 6) comments, 7) sketch of site, and 8) stream visual assessment scores. The latter includes eight characteristics that were scored 1 through 5 for the riparian zone and 1 through 4 for all the other characteristics.
The Stream Visual
Assessment Score (SVAS) in the protocol was patterned after the USDA’s Stream Visual Assessment Protocols. This was modified in an effort to reduce
and/or eliminate the need for professional judgment. This was accomplished primarily by converting
scoring criteria into quantifiable measures.
For example, riparian zone was scored based on width, while benthic
aquatic habitat was scored based on the number of habitat types present.
Training
Project staff designed a training protocol that included both classroom and field components. The training consisted of a five-hour session in which project goals and objectives, safety, and methodology for the visual stream protocol were covered. Methods for respecting private property and to avoid trespassing were covered also. In addition, volunteers and trainers went into the field and sampled two sites utilizing the visual stream assessment protocol. The sites were chosen to represent both high and low habitat quality ratings. Lastly, the project director and project manager remained available throughout the duration of the project to assist volunteers and to address any questions or concerns that arose.
Site Selection
Rivers and streams surveyed within the watershed were
chosen based on the 303(d)/305(b) list prepared, by TDEC’s Division of Water
Pollution Control. Volunteers drove,
walked and/or canoed sections of the watershed to find specific sites. General guidance was provided during training
and included two basic ideas. These were
to 1) plan a route through the watershed along roads, trails etc. prior to the
survey and 2) view the stream and a particular site from as many different
locations or angles as possible.
Volunteers were given a set of applicable USGS 7.5 minute series
topographical maps (USGS topo maps), and a county road map. Sites were identified and marked on the USGS
topographic maps. Because there has
been so much development and new road systems built in areas of the watershed
since the USGS topo maps were prepared, volunteers also were provided with
county road maps.
Data Management
Data management was
accomplished utilizing the Microsoft AccessÔ
Database program. Each field data sheet
for a site was numbered prior to data entry and entered into the database as
soon as possible. The data entry screens
were designed to mirror the field data sheets to minimize input errors and to
simplify the overall process. In
addition, 10 % of the data were checked to assure accuracy of input. Finally, data were imported in Microsoft
ExcelÔ for statistical analysis and for preparation of
charts and graphs.
Visual Stream Assessment Database
Volunteers recorded
data on Field Data Sheets (see Appendix A), in photographs, and on USGS topo
maps. Data was transferred directly into
the database. Information from data
sheets was typed into the database just as it was recorded on the sheets.
Occasionally, volunteers gave separate scores for sites where the environment
varied significantly from upstream to downstream. In that case, the lowest
score was used in the database and this was noted in the corresponding comment
field. Photographs were scanned into the computer or copied digitally. Volunteers located each site on a USGS topo
map. If the volunteer did not record the
latitude and longitude of the site on the data sheet, the project manager used
the map to find the latitude and longitude on www.topozone.com.
Quality Assurance
Quality Assurance (QA) generally refers to the management activities designed to ensure data meet established standards with a given level of confidence. The QA team consisted of the project director, project manager, and the Executive Director of the HRWA. In addition, staff from TDEC, Tennessee Department of Agriculture (TDA), and U.S. Geological Survey (USGS) were consulted to assist in the design of the overall protocol.
Quality Control
Quality Control (QC) addresses error control in the collection of field data, laboratory analysis and data management. QC defines specific techniques that will be used to control for or identify errors in the data set. The sites (n=181) were stratified by volunteer, and ten percent from each were randomly chosen for duplicate sampling by the project director. The QC sampling included only streamside land use, qualitative measures for erosion and the SVAS. QC measures calculated from this data included measures of precision, and accuracy. In addition, percent comparability was calculated. See Appendix B-Quality Control Methods and Results, for a description of each QC measure.
Quality assurance measures included training volunteers to use the visual stream assessment protocol and data management activities. Quality control was accomplished utilizing internal duplicate sampling. This was accomplished with built in redundancy and by professional assessment of 10 % of sites. From this data measures of precision, accuracy and percent comparability were calculated (see below).
III. RESULTS and DISCUSSION
Twenty-one volunteers surveyed a total of 217 sites in the
The major product of this study is the Visual Stream
Assessment Database (Appendix D on a CD-ROM).
All the information gathered during the site visits, including the
photographs, are in the Visual Stream Assessment Database. All the original data is housed in the HRWA
office. Appendix C contains
representative pictures of each site, USGS topo maps showing the location of
each site, a table of land use at each site, and a table of the Stream Visual
Assessment Scores for each site. Sites
in Appendix C are grouped by stream, and streams are in order from the
headwaters to the mouth of the
The data from this study was both provided to TDEC during
the course of the project and in this final format for several reasons. Some Field Data Sheets and comments from
volunteers appeared to document possible permit violations. These were submitted to the
Though over 30 volunteers expressed interest in this study, not all of them had the time to commit. The volunteers were recruited primarily from the HRWA membership using electronic mail and telephone follow-up by the E.D. A core of the volunteers were already volunteers in the sediment study that is a joint project designed and managed by the Cumberland River Compact. Yet, volunteers came from the membership who had not participated in a project before. Also, volunteers came from contacting the membership of other conservation and outdoor groups, such as the Middle TN Fly Fishers Association, the TN Scenic Rivers Association, and Friends of Radnor Lake, and via posting an announcement on several email listservs.
Specific Findings:
The data indicate several land use practices and stream characteristics that appear to be contributing to TDEC’s findings of water quality impairment in 303(d)/305(b) listed streams. The project director conducted the Quality Control Analysis using measures of precision (RPD), accuracy, and percent comparability (PC) to document the quality of the data and to help guide resource managers relative to how much confidence they may place in the data. See Appendix B for the QC methods and results.
Land Use
The survey identified seven specific types of land use and one catchall category “other” for three different distances from the stream: streamside, within ¼ mile, and beyond ¼ of a mile from the stream bank edge. Data on streamside land use found agriculture and forestry at 109 (36 %) sites, while residential land use was noted at 86 (28 %) sites (see Table 1). “Other” was the next most frequently cited land use at 33 (11 %). Volunteers identified parkland as the streamside land use at 24 (8 %) sites and commercial land use at 27 (9 %) sites. Development and construction were identified at 14 (5 %) sites, while industry and mining land uses were reported at 8 (3 %) and 3 (1 %) sites, respectively.
Erosion
Volunteers identified stream bank erosion with both qualitative and quantitative means. As Table 2 shows, no stream bank erosion was noted at 64 sites (32 %), while occasional stream bank erosion was identified at 103 sites (52 %). Common stream bank erosion was noted at 32 sites (16 %), and seven sites were identified that had sediment and/or point bars, along with sediment along the margins of streams.
Stream Visual Assessment Scores (SVAS)
The Stream Visual Assessment Scores (SVAS) are summed in Table 3 and consist of eight stream characteristics. Forty-five sites scored between 36 and 42 (42 was the maximum score), while 67 sites had a total score between 31 and 35. Forty-six sites scored between 26 and 30; 19 sites had total SVAS scores between 21 and 25, and 14 sites scored between 16 and 20. Finally, 17 and 4 sites scored between 11 and 15, and 6 and 10, respectively.
One of the most alarming sets of data from the SVAS was the numeric scores for
the riparian zone as shown in Table 4.
One hundred twenty-one right bank and 117 left bank sites scored in the
lowest category for the riparian zone.
This finding is consistent with the findings of TDEC’s Division of Water
Pollution Control (Denton et al., 2000).
The majority (128 right bank, 129 left bank) of bank stability scores were in the excellent range; however 42 (left bank) and 36 (right bank) scored in the good range, 29 and 36 in the fair range and 10 and 9 in the poor range as shown in Table 5.
Bank stability data coupled with the qualitative stream bank erosion data and
verified with a picture indicate that perhaps 32 sites are significant
contributors or causes of impairment when related to sediment (See Table
12). TDEC indicates that 36 out of 42
stream segments are impaired, at least, in part as a result of sediment (Denton
et al., 2000).
Canopy cover was also assessed utilizing the SVAS. As shown in Table 6, 54 sites scored excellent, 39 scored good, 52 scored fair, while 65 scored poor. This data must be viewed in relation to the size of the stream, as the main stem of the Harpeth would not be expected to have as large of a percent shading (canopy cover) as a small headwater stream.
Overall, the invertebrate habitat was rated as excellent, with 131 sites scoring four or excellent on the SVAS. This is shown in Table 7. Thirty-seven sites scored good, 12 sites scored fair and 16 scored poor.
Riffle and pool sedimentation results are presented in Table 8. One hundred twenty-five sites scored
excellent, while 36 sites scored good.
However, 13 sites scored fair and 6 sites scored poor. This data coupled with the qualitative stream
bank erosion information, and the quantitative bank stability data indicate
that nine sites are adding significant amounts of sediment and may be contributing
to impairment (See Table 12).
One hundred thirty-three sites scored excellent on the SVAS for water
appearance, while 43 sites scored good as shown in Table 9. Thirteen sites scored fair, while three sites
scored poor for water appearance/characteristics. Data for nutrient enrichment are presented in
Table 10. One site scored in the poor
range for nutrient enrichment, while 15 sites scored in the fair range, 50 in
the good range and 118 in the excellent range.
This data, coupled with the pictures, indicate that 15 sites are
problematic regarding nutrient enrichment (See Table 12).
Finally, data for channel condition are presented in Table 11. Twelve sites were rated in the poor range, while 31 sites were rated in the fair range. Volunteers assigned 68 sites to the good range and 98 sites to the excellent range.
Opportunities for Water Quality Improvements:
Volunteers surveyed 217 sites along 303(d)/305(b) listed
streams and unassessed streams. They identified as many as 48 sites that appear
to be obviously contributing pollutants to the associated aquatic systems. These 48 sites, listed in Table 12, represent
sites of clear water quality related problems and priority opportunities for
the implementation of BMPs. Riparian
zone impacts appear to play a large part in water quality impairment, and bank
instability was identified as an important source of sediment in the
Table 12 includes sites in the survey database with both
common stream bank erosion and a bank stability score of less than four. In addition, the table includes sites with
nutrient enrichment scores less than four. A site was only included in Table 12
if the photographs taken on site confirm scores for bank stability or nutrient
enrichment. This table includes five sites from the survey of erosion and
habitat quality on the main stem of the
|
Table 12-Opportunities for Water Quality Improvements |
||||
|
Form
Number |
Stream
Name |
Point
on Map |
Problem
ID |
Potential
cause |
|
33 |
Arrington Creek |
16 |
Erosion,
Nutrient Enrichment, Riparian Impairment |
Livestock
Access |
|
45 |
Big Turnbull Creek |
1 |
Nutrient
Enrichment, Riparian Impairment |
Mining,
Agriculture |
|
108 |
Burns Branch |
3 |
Erosion,
Riparian Impairment |
Development |
|
143 |
Cartwright Creek |
1 |
Erosion,
Nutrient Enrichment |
Lawn
management |
|
144 |
Cartwright Creek |
2 |
Erosion,
Riparian Impairment |
Not
Apparent |
|
145 |
Cartwright Creek |
3 |
Erosion,
Riparian Impairment |
Lawn
management |
|
146 |
Cartwright Creek |
4 |
Erosion,
Riparian Impairment |
Lawn
management |
|
174 |
Cheatham Branch |
3 |
Nutrient
Enrichment, Riparian Impairment |
Agriculture,
Vegetation Management |
|
115 |
Donelson Creek |
5 |
Erosion,
Riparian Impairment |
Utility
line, Development |
|
62 |
Five Mile Creek |
5 |
Nutrient
Enrichment, Riparian Impairment |
Vegetation
Management |
|
70 |
Flat Creek |
9 |
Erosion,
Riparian Impairment |
Not
Apparent |
|
74 |
Flat Creek |
8 |
Nutrient
Enrichment, Riparian Impairment |
Vegetation
Management |
|
167 |
Franklin Area Tributary |
10 |
Erosion,
Riparian Impairment |
Not
Apparent |
|
127 |
Jones Creek |
3 |
Erosion |
Volume
and Velocity |
|
128 |
Jones Creek |
|||