EPA/600/R-15/280 | February 2016 | www.epa.gov/research
United States
Environmental Protection
Agency
Regional Monitoring Networks (RMNs)
to Detect Changing Baselines in
Freshwater Wadeable Streams
Support Clean Water Act programs
Detect trends attributable
to climate change
Establish
current
conditions
Detect trends in
high quality waters
Office of Research and Development
Washington, D.C.
-------
EPA/600/R-15/280
February 2016
Final Report
Regional Monitoring Networks (RMNs) to Detect Changing
Baselines in Freshwater Wadeable Streams
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
ABSTRACT
The United States Environmental Protection Agency (U.S. EPA) is working with its
regional offices, states, tribes, river basin commissions and other entities to establish Regional
Monitoring Networks (RMNs) for freshwater wadeable streams. RMNs have been established in
the Northeast, Mid-Atlantic, and Southeast, and efforts are expanding into other regions.
Long-term biological, thermal, hydrologic, physical habitat and water chemistry data are being
collected at RMN sites to document current conditions and detect long-term changes. Consistent
methods are being used to increase the comparability of data, minimize biases and variability,
and ensure that the data meet data quality objectives. RMN surveys build on existing state and
tribal bioassessment efforts, with the goal of collecting comparable data at a limited number of
sites that can be pooled at a regional level. Pooling data enables more robust regional analyses
and improves the ability to detect trends over shorter time periods. This document describes the
development and implementation of the RMNs. It includes information on selection of sites,
expectations for data collection, the rationale for collecting these data, data infrastructure and
provides examples of how the RMN data will be used and analyzed. The report concludes with a
discussion on the status of monitoring activities and next steps.
Preferred citation:
U.S. EPA (Environmental Protection Agency). (2016) Regional Monitoring Networks (RMNs) to detect changing
baselines in freshwater wadeable streams. Office of Research and Development, Washington, DC:
EPA/600/R-15/280. Available online at http://www.epa.gov/research.
11
-------
TABLE OF CONTENTS
LIST OF TABLES v
LIST OF FIGURES vi
LIST OF ABBREVIATIONS vii
PREFACE viii
AUTHORS, CONTRIBUTORS, AND REVIEWERS ix
EXECUTIVE SUMMARY x
1. INTRODUCTION 1-1
2. PROCESS FOR SETTING UP THE REGIONAL MONITORING NETWORKS
(RMNS) 2-1
3. REGIONAL MONITORING NETWORK (RMN) DESIGN 3-1
3.1. SITE SELECTION 3-1
3.2. METHODS FOR DATA COLLECTION 3-4
3.2.1. BIOLOGICAL INDICATORS 3-6
3.2.2. TEMPERATURE DATA 3-16
3.2.3. HYDROLOGIC DATA 3-18
3.2.4. PHYSICAL HABITAT 3-23
3.2.5. WATER CHEMISTRY 3-25
3.2.6. PHOTODOCUMENTATION 3-25
3.2.7. GEOSPATIALDATA 3-27
4. SUMMARIZING AND SHARING REGIONAL MONITORING NETWORK (RMN)
DATA 4-1
4.1. BIOLOGICAL INDICATORS 4-1
4.2. THERMAL STATISTICS 4-4
4.3. HYDROLOGIC STATISTICS 4-5
5. DATA USAGE 5-1
5.1. APPLICATIONS IN A 1-5 YEAR TIMEFRAME 5-1
5.2. APPLICATIONS IN A 5-10 YEAR TIMEFRAME 5-8
5.3. APPLICATIONS IN A 10+ YEAR TIMEFRAME 5-9
6. DATA MANAGEMENT 6-1
7. IMPLEMENTATION AND NEXT STEPS 7-1
8. LITERATURE CITED 8-1
in
-------
TABLE OF CONTENTS (continued)
APPENDIX A. POWER ANALYSIS A-l
APPENDIX B. CHECKLIST FOR STARTING A REGIONAL MONITORING
NETWORK (RMN) B-l
APPENDIX C. PRIMARY REGIONAL MONITORING NETWORK (RMN) SITES IN
THE NORTHEAST, MID-ATLANTIC, AND SOUTHEAST RMN REGIONS C-l
APPENDIX D. DISTURBANCE SCREENING PROCEDURE FOR RMN SITES D-l
APPENDIX E. SECONDARY REGIONAL MONITORING NETWORK (RMN) SITES
IN THE NORTHEAST AND MID-ATLANTIC REGIONS E-l
APPENDIX F.MACROINVERTEBRATE COLLECTION METHODS F-l
APPENDIX G. LEVEL OF TAXONOMIC RESOLUTION G-l
APPENDIX H. SUMMARIZING MACROINVERTEBRATE DATA H-l
APPENDIX I. MACROINVERTEBRATE THERMAL INDICATOR TAXA I-1
APPENDIX J. THERMAL SUMMARY STATISTICS J-l
APPENDIX K. HYDROLOGIC SUMMARY STATISTICS AND TOOLS FOR
CALCULATING ESTIMATED STREAMFLOW STATISTICS K-l
IV
-------
LIST OF TABLES
3-1. Main considerations when selecting primary sites for the regional
monitoring networks (RMNs) 3-2
3-2. There are four levels of rigor in the regional monitoring network (RMN)
framework, with Level 1 being the lowest and Level 4 being the
best/highest standard 3-6
3-3. Recommendations on best practices for collecting biological data at
regional monitoring network (RMN) sites 3-7
3-4. Recommendations on best practices for collecting macroinvertebrate data
at Northeast, Mid-Atlantic and Southeast regional monitoring network
(RMN) sites 3-9
3-5. Recommendations on best practices for collecting temperature data at
regional monitoring network (RMN) sites 3-17
3-6. Recommendations on best practices for collecting hydrologic data at
regional monitoring network (RMN) sites 3-21
-------
LIST OF FIGURES
1-1. States, tribes, river basin commissions (RBCs), and others in three RMN
regions (Northeast, Mid-Atlantic, and Southeast) have established regional
monitoring networks (RMNs) 1-2
3-1. Seasonal differences (spring vs. summer) in percentage cold water taxa
and individuals in the Mid-Atlantic 2014 data 3-12
3-2. Species accumulation curve based on the Mid-Atlantic 2014 data 3-13
3-3. Staff gage readings provide a quality check of transducer data 3-20
3-4. Photodocumentation of Big Run, WV, taken from the same location each
year 3-27
4-1. Spatial distributions of macroinvertebrate taxa, based on the National
Aquatic Resource Survey (NARS) data 4-3
5-1. RMN data can be used for multiple purposes, over short and long-term
timeframes 5-2
5-2. Proportion of cold/cool indicator taxa at RMN sites, based on preliminary
data from a subset of sites 5-4
5-3. The thermal tolerances of Sweltsa and Tallaperla match very closely with
brook trout 5-4
5-4. Connecticut Department of Energy and Environmental Protection (CT
DEEP) developed ecologically meaningful thresholds for three major
thermal classes (cold, cool, warm) 5-5
5-5. Salmon life cycle plotted in relation to yearly flow cycle (Ricupero, 2009) 5-6
5-6. RMN data will help us gain a better understanding of natural variability in
hydrologic conditions in small least disturbed streams, and will allow us to
investigate relationships between biological, thermal, and hydrologic
conditions 5-7
5-7. EPA and partners are conducting a broad-scale climate change
vulnerability assessment on streams in the eastern United States, based on
a scenario in which stream temperatures warm and the frequency and
duration of summer low flow events increases 5-10
5-8. Modeling results predict declines in species richness across much of the
Northeast by mid-century (2040-2069) 5-11
5-9. Comparison of macroinvertebrate density values at 10 stream sites in
Vermont before and after Tropical Storm Irene 5-12
7-1. Sampling has been underway at the Northeast, Mid-Atlantic and Southeast
RMNs for several years 7-1
VI
-------
LIST OF ABBREVIATIONS
BCG biological condition gradient
CT DEEP Connecticut Department of Energy and Environmental Protection
CWA Clean Water Act
E expected
ELOHA ecological limits of hydrologic alteration
EPT Ephemeroptera, Plecoptera, and Trichoptera
GIS geographic information system
GPS global positioning system
MA DEP Massachusetts Department of Environmental Protection
MD DNR Maryland Department of Natural Resources
MMI multimetric index
NARS EPA National Aquatic Resource Surveys
NLCD National Land Cover Database
NMDS nonmetric multidimensional scaling
NRSA National Rivers and Streams Assessment
NWQMC National Water Quality Monitoring Conference
O observed
QA/QC quality assurance/quality control
QAPP Quality Assurance Proj ect Plan
RBC river basin commission
RIFLS River Instream Flow Stewards Program
RMN regional monitoring network
SDM species distribution model
SOP standard operating procedure
SSN Sentinel Sites Network
SWPBA Southeastern Water Pollution Biologists Association
TNC The Nature Conservancy
USGS U.S. Geological Survey
VT DEC Vermont Department of Environmental Conservation
WQX Water Quality Exchange
WV DEP West Virginia Department of Environmental Protection
vn
-------
PREFACE
The U.S. Environmental Protection Agency (EPA) is working with states, tribes, river
basin commissions, and other organizations in different parts of the country to establish regional
monitoring networks (RMNs) to collect data that will further our understanding of biological,
thermal, and hydrologic conditions in freshwater wadeable streams and allow for detection of
changes and trends. This document describes the framework for the RMNs that have been
developed in the Northeast, Mid-Atlantic, and Southeast regions for riffle-dominated, freshwater
wadeable streams.
Vlll
-------
AUTHORS AND REVIEWERS
The National Center for Environmental Assessment, Office of Research and
Development, published this report. This document was prepared with the assistance of Tetra
Tech, Inc. under Contract No. EP-C-12-060, EPA Work Assignments No. 1-01 and 2-01. Dr.
Britta Bierwagen served as the Technical Project Officer, providing overall direction and
technical assistance.
AUTHORS
Center for Ecological Sciences, Tetra Tech, Inc., Owings Mills, MD
Jen Stamp, Anna Hamilton
U.S. EPA Region 3, Wheeling, WV
Margaret Passmore (retired)
Tennessee Department of Environment and Conservation
Debbie Arnwine
U.S. EPA, Office of Research and Development, Washington DC
Britta G. Bierwagen
Fairfax County Stormwater Planning Division, VA
Jonathan Witt
REVIEWERS
U.S. EPA Reviewers
Jennifer Fulton (R3), Ryan Hill, Ph.D. (ORISE Fellow within ORD), Sarah Lehmann (OW)
External Peer Reviewers
LucindaB. Johnson, Ph.D. (University of Minnesota), Kent W. Thornton, Ph.D. (FTN), Chris O.
Yoder, Ph.D. (Midwest Biodiversity Institute)
ACKNOWLEDGMENTS
The authors would like to thank the many partners who reviewed early versions of this
report for clarity and usefulness. Their comments substantially improved this document. Special
thanks to K. Herreman and D. Infante (Michigan State University), P. Morefield, C. Mazzarella,
and J. Fulton (U.S. EPA), former ORISE participant A. Murdukhayeva, and A. Olivero (The
Nature Conservancy) for their contributions to the disturbance screening process described in
Appendix D.
IX
-------
EXECUTIVE SUMMARY
The United States Environmental Protection Agency (EPA) is working with its regional
offices, states, tribes, river basin commissions and other entities to establish Regional Monitoring
Networks (RMNs) for freshwater wadeable streams. RMNs have been established in the
Northeast, Mid-Atlantic, and Southeast, and efforts are expanding into other regions. Long-term
biological, thermal, hydrologic, physical habitat and water chemistry data are being collected at
RMN sites to document current conditions and detect long-term changes. Consistent methods are
being used to increase the comparability of data, minimize biases and variability, and ensure that
the data meet data quality objectives. RMN surveys build on existing state and tribal
bioassessment efforts, with the goal of collecting comparable data at a limited number of sites
that can be pooled at a regional level. Pooling data enables more robust regional analyses and
improves the ability to detect trends over shorter time periods.
The goal of the RMNs is to provide data that can be used by biomonitoring programs for
multiple purposes, spanning short and long-term timeframes. Uses include:
Monitoring the condition of minimally and least disturbed streams
Detecting trends attributable to climate change
Supplementing Clean Water Act (CWA) programs and initiatives under Sections 303 and
305(b)
- Defining natural conditions/quantifying natural variability
- Informing criteria refinement or development
- Developing biological indicators for protection planning
Gaining a better understanding of relationships between biological, thermal, and
hydrologic data
Gaining a better understanding of ecosystem responses and recovery from extreme
weather events
Gaining insights into effects of regional phenomena such as drought, pollutant/nutrient
deposition and riparian forest infestations on aquatic ecosystems and bioassessment
programs
The need for RMNs stems from the lack of long-term, contemporaneous biological,
thermal, and hydrologic data, particularly at minimally disturbed stream sites. To help fill this
gap, efforts are underway to collect the following types of data from the RMN sites:
Biological indicators: macroinvertebrates
- Optional: fish and periphyton, if resources permit (fish are higher priority)
-------
Temperature: continuous water and air temperature (30-minute intervals)
Hydrological: continuous water-level data (15-minute intervals); converted to discharge
if resources permit
Habitat: parameters agreed upon by regional working group
Water chemistry: In situ, instantaneous water chemistry parameters (specific
conductivity, dissolved oxygen, pH), plus additional or more comprehensive water
chemistry measures agreed upon by regional working group
Photodocumentation
Geospatial data
The RMNs are designed to detect potentially small trends in biological, thermal,
hydrologic, physical habitat and water chemistry data at high quality sites in a decision-relevant
timeframe (e.g., within 5 years to inform criteria development; in 10-20 years to inform
changing baselines). The RMN design calls for sampling at least 30 sites with similar
environmental and biological characteristics in each region on an annual basis for 10 or more
years, using comparable methods. To help inform this design, EPA and partners performed
power analyses on an aggregated biomonitoring data set from a 2012 Northeast pilot study. The
power analyses suggest that significant trends in regional community composition can be
detected within 10-20 years if 30 or more comparable sites are monitored regularly. EPA and
partners also used literature and standard operating procedures (SOPs) from participating
organizations to help inform design decisions.
A generic Quality Assurance Project Plan (QAPP)1 has been developed for the RMNs
that details the core requirements for participation in the network, and outlines best practices for
the collection of biological, thermal, hydrologic, physical habitat, and water chemistry data at
RMN sites. The QAPP was written in a way that should be transferable across regions, with
region-specific protocols included as addendums. The QAPP is intended to increase the
comparability of data being collected at RMN sites, improve the ability to detect long-term
trends by minimizing biases and variability, and to ensure that the data are of sufficient quality to
meet data quality objectives. The ability of some participants to use the regional RMN methods
has been limited by resource constraints, so in some situations, there have been differing levels
of effort and differing methods across sites and organizations. While this is not ideal, the data
can still be used, just in more limited ways. The data management system that EPA and partners
are developing will contain metadata that will enable users to select data that meet their needs
(e.g., collected using certain methods and at certain levels of rigor).
Sampling efforts at the RMNs are concentrated at a core group of sites called "primary"
sites, where efforts are being made to collect the full suite of biological, thermal, hydrologic,
JThe QAPP (U.S. EPA, 2016. Generic Quality Assurance Project Plan for monitoring networks for tracking long-
term conditions and changes in high quality wadeable streams) is available online at
http://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=295758&inclCol=eco#tab-3.
XI
-------
physical habitat and water chemistry data. Efforts were made to select sites that had as many of
the following characteristics as possible:
Part of established, long-term monitoring networks (e.g., U.S. Geological Survey
[USGS], sentinel)
Low level of anthropogenic disturbance
Exhibit similar environmental and biological characteristics
Longevity (e.g., accessible [day trip], opportunities to share the workload with outside
agencies or organizations)
Located in watersheds protected from future development
Lengthy historical sampling record for biological, thermal, or hydrological data
Most primary RMN sites are minimally or least disturbed sites (per Stoddard et al.,
2006). High quality waters are being targeted because they are the standard against which other
bioassessment sites are compared. It is critical to document current conditions at high quality
sites and to track changes at these sites over time to understand how benchmarks may be shifting
in response to changing environmental and climatic conditions. Data from additional,
"secondary," sites are also being considered for the RMNs. These are sites at which a subset of
parameters are already being collected in accordance with RMN protocols as part of other
independent monitoring efforts. Data from secondary sites will increase the sample size and
range of conditions represented in the RMN data set, and may provide information about unique
or underrepresented geographic areas.
Data collection has been underway in the Northeast, Mid-Atlantic, and Southeast RMNs
for several years. This report describes the development and implementation of these pilot
RMNs. It includes information on selection of sites, expectations for data collection, the rationale
for collecting these data, and data infrastructure. The report also provides examples of how the
RMN data will be used and analyzed, and concludes with a discussion on the status of
monitoring activities. Currently, EPA and partners continue to build capacity and refine
protocols, indicator lists, analytical techniques and data management systems for the RMNs,
including working with regions to establish RMNs. Long-term data from RMNs can support
CWA programs, fill data gaps, and help detect trends attributable to climate change. The RMN
framework is flexible and allows for expansion to new regions, as well as to new stream classes
and waterbody types. The monitoring data being collected from these regional efforts will
provide important inputs for bioassessment programs as they strive to protect water quality and
aquatic ecosystems under a changing climate.
xn
-------
1. INTRODUCTION
The U.S. Environmental Protection Agency (EPA) has been working with states, tribes,
river basin commissions (RBCs), and other organizations in different parts of the United States
to establish regional monitoring networks (RMNs) to collect contemporaneous biological,
thermal, hydrologic, physical habitat, and water chemistry data from freshwater wadeable
streams. RMNs have been established in the Northeast, Mid-Atlantic, and Southeast2 (see
Figure 1-1), and efforts to establish new networks are expanding into other regions. The concept
of the RMNs stems from work that began in 2006 with pilot studies that examined long-term
climate-related trends in macroinvertebrate data from state biomonitoring programs in Maine,
North Carolina, Ohio, and Utah (U.S. EPA, 2012a). During these studies, a lack of long-term,
contemporaneous biological, thermal, and hydrologic data became apparent, particularly at
minimally disturbed (Stoddard et al., 2006) stream sites. These data gaps have been documented
elsewhere (e.g., Mazor et al., 2009; Jackson and Fureder, 2006; Kennen et al., 2011) and have
been recognized as important gaps to fill by the National Water Quality Monitoring Council
(NWQMC) (NWQMC, 2011).
The goal of the RMNs is to provide data that can be used by biomonitoring programs for
multiple purposes, spanning short and long-term timeframes. Uses include:
Monitoring the condition of minimally and least disturbed streams
Detecting trends attributable to climate change
Supplementing Clean Water Act (CWA) programs and initiatives
- Defining natural conditions/quantifying natural variability to support Section 305(b)
programs
- Informing criteria refinement or development under Section 303
- Developing biological indicators for protection planning for Section 303(d) programs
Gaining a better understanding of relationships between biological, thermal, and
hydrologic data
Gaining a better understanding of ecosystem responses and recovery from extreme
weather events
Gaining insights into effects of regional phenomena such as drought, pollutant/nutrient
deposition and riparian forest infestations on aquatic ecosystems and bioassessment
programs
2RMN regions are largely (but not exactly) based on EPA regions to help facilitate coordination and sharing of
resources. Differences include: New York (EPA Region 2), which joined the EPA Region 1 states in the Northeast
RMN; New Jersey (EPA Region 2), which joined the EPA Region 3 states in the Mid-Atlantic RMN; and
Mississippi and Florida, which did not join EPA Region 4 states in the Southeast RMN because they lack the
targeted habitat (medium to high gradient, cold, riffle-dominated streams).
1-1
-------
Northeast
Mid-ailantic
Southeast
Figure 1-1. States, tribes, river basin commissions (RBCs), and others in
three RMN regions (Northeast, Mid-Atlantic, and Southeast) have
established regional monitoring networks (RMNs).
The RMNs are designed to detect potentially small trends in biological, thermal,
hydrologic, physical habitat and water chemistry data at high quality sites in a decision-relevant
timeframe (e.g., 10-20 years to be relevant to climate change). Several states, tribes, RBCs, and
others are already collecting annual biological and continuous temperature data at targeted sites,
and to a lesser degree, hydrologic data. The goal is to supplement existing efforts like these, and
to collect comparable data at a limited number of sites to pool at a regional level. Pooling data
enables more robust regional analyses, improves the ability to detect trends over shorter time
periods and can inform on changes at a spatial scale similar to climatic changes.
The RMN design calls for sampling at least 30 sites with similar environmental and
biological characteristics in each region on an annual basis for 10 or more years, using
comparable methods. To help inform this design, EPA and partners performed power analyses
on an aggregated biomonitoring data set from the 2012 Northeast pilot study. The power
analyses suggest that significant trends in regional community composition can be detected
within 10-20 years if 30 or more comparable sites are monitored regularly. A detailed account of
these analyses can be found in Appendix A. Design decisions were also informed by literature
and standard operating procedures (SOPs) being used by the RMN participants. Efforts are being
1-2
-------
made to use consistent methods at RMN sites to increase the comparability of data and minimize
biases and variability. A Quality Assurance Project Plan (QAPP) has been developed for the
RMNs to ensure that participating entities understand the requirements and meet the data quality
objectives. Scientific considerations are balanced with practical considerations by participating
entities. The RMN framework needs to be flexible enough to tie into existing state and tribal
bioassessment efforts and must stay within the resource constraints of its participants.
Data collection in the Northeast, Mid-Atlantic, and Southeast RMNs has been underway
for several years, and EPA and partners are starting to use these data in initial evaluations and
data analyses. This report describes the development and implementation of these RMNs. It
includes information on selection of sites, expectations for data collection, the rationale for
collecting these data, and data infrastructure. The report also provides examples of how the RMN
data will be used and analyzed. It concludes with a discussion on the status of monitoring
activities and next steps.
1-3
-------
[This page intentionally left blank.]
-------
2. PROCESS FOR SETTING UP THE REGIONAL MONITORING
NETWORKS (RMNS)
The Northeast, Mid-Atlantic, and Southeast regions followed similar processes to
establish their RMNs. A regional, tribal, or state coordinator formed a working group of
interested partners to establish regional goals to determine basic survey bounds, such as selection
of a target population (e.g., freshwater wadeable streams with abundant riffle habitat). Working
groups selected RMN sites using consistent criteria (see Section 3.1), and selected appropriate
data-collection protocols and methodologies (see Section 3.2). As part of this process, working
groups considered the site selection criteria and methods being used in the other regions and tried
to utilize similar protocols where practical to generate comparable data. The groups then
identified logistical, training, and equipment needs and sought resources from agencies such as
EPA and the U.S. Geological Survey (USGS) to help address high-priority goals. The regional
working groups began implementation several years ago and are starting to use the RMN data in
initial evaluations and data analyses. EPA and partners recently developed a generic RMN QAPP
that details the core requirements for participation in the network, and outlines best practices for
the collection of biological, thermal, hydrologic, physical habitat, and water chemistry data at
RMN sites. The regional working groups are in the process of reviewing and approving the
QAPP. The EPA and partners are also developing a data management system that will allow
participating organizations and outside users to access data and metadata that are being collected
at RMN sites (see Section 6). Appendix B includes a step-by-step checklist on the process for
developing and implementing RMNs.
2-1
-------
[This page intentionally left blank.]
-------
3. REGIONAL MONITORING NETWORK (RMN) DESIGN
The RMN design calls for sampling at least 30 sites with similar environmental and
biological characteristics in each region on an annual basis for 10 or more years, using
comparable methods. In 2011-2012, EPA collaborated with seven states in the northeastern
United States on a pilot study that helped lay the groundwork for the RMNs. The goal of the
pilot was to design a monitoring network that could detect potentially small trends in biological,
thermal, hydrologic, physical habitat and water chemistry data at high quality sites in a
decision-relevant timeframe (e.g., 10-20 years to be relevant to climate change). EPA and
partners performed power analyses on an aggregated biomonitoring data set from the Northeast
to explore questions such as: How long will it take to detect trends in biological metrics? How
much of an effect does sampling frequency and classification scheme have on trend detection
time? The results suggest that detection times of 10-20 years (at 80% power) are possible for
some biological metrics if 30 or more sites with comparable environmental conditions and
biological communities are monitored regularly. These results are consistent with a study by
Larsen et al. (2004), which found that well-designed networks of 30-50 sites monitored
consistently can detect underlying changes of 1-2% per year in a variety of metrics within
10-20 years, or sooner, if such trends are present. The Northeast power analyses are described in
detail in Appendix A.
3.1. SITE SELECTION
Sampling efforts at the RMNs are concentrated at a core group of sites called "primary"
sites, where efforts are being made to collect the full suite of biological, thermal, hydrologic,
physical habitat and water chemistry data (see Section 3.2). The working groups selected 2 to
15 primary sites per state (depending on the size of the state and availability of resources), with
the overall goal of sampling at least 30 primary sites in each RMN region. The site selection
process takes into account numerous considerations, which are summarized in Table 3-1. Efforts
were made to select sites that had as many of the desired characteristics listed in Table 3-1 as
possible. Appendix C lists the primary RMN sites in each region as of September 2015.
5-1
-------
Table 3-1. Main considerations when selecting primary sites for the regional
monitoring networks (RMNs)
Consideration
Existing monitoring network
Disturbance
Equipment
Classification
Longevity
Sampling record
Potential for future disturbance
Desired characteristics at primary sites
Located in established long-term monitoring networks to
build upon data already being collected by states, tribes,
RBCs, and others.
Low level of anthropogenic disturbance.
Colocated with existing hydrologic equipment (e.g., USGS
gage, weather station).
Sites exhibit similar environmental and biological
characteristics, which minimizes natural variability across
sites, improves power for detecting long-term trends and
allows for pooling of data within and across regions.
Accessible (e.g., day trip), opportunities to share the
workload with outside agencies or organizations.
Lengthy historical sampling record for biological, thermal,
or hydrological data.
Located in watersheds that are protected from future
development.
Where feasible, organizations colocated RMN sites with existing stations like USGS
gages or in established long-term monitoring networks such as the sentinel networks of the
Vermont Department of Environmental Conservation (VT DEC), the Connecticut Department of
Energy and Environmental Protection (CT DEEP), Maryland Department of Natural Resources
(MD DNR), West Virginia Department of Environmental Protection (WV DEP), and Tennessee
Department of Environment and Conservation, continuous monitoring stations of the
Susquehanna River Basin Commission, and USGS networks, such as the Northeast Site Network
and the Geospatial Attributes of Gages for Evaluating Streamflow (GAGES-II) program. Some
of these sites have lengthy historical records, which are preferred for primary RMN sites.
Efforts were made to select minimally disturbed or least disturbed sites (per Stoddard
et al. 2006). High quality waters are being targeted because they are the standard against which
other bioassessment sites are compared. It is critical to document current conditions at high
quality sites and to track changes at these sites over time to understand how benchmarks may be
shifting in response to changing environmental and climatic conditions. EPA and partners
developed a standardized procedure for characterizing the present-day level of anthropogenic
disturbance and applied this across RMNs so that sites from all states and regions are rated on a
common scale (see Appendix D). Sites are screened for likelihood of impacts from land use
disturbance, dams, mines, point-source pollution and other factors.
5-2
-------
The selection criteria also prioritize sites that exhibit similar environmental and
biological characteristics, as this helps reduce natural variability across sites (which improves
power for detecting long-term trends) and allows for pooling of data within and potentially
across regions. The Southeast working group used ecoregions during the initial site selection
process because ecoregions dominate the reference-site-stratification approach used by many
programs for assessing streams (Carter and Resh, 2013). Most of the RMN sites in the Southeast
are located in ecoregions with hilly or mountainous terrain (e.g., Piedmont, Blue Ridge, Central,
and North Central Appalachians), where streams generally have higher gradients and more riffle
habitat. In the Northeast and Mid-Atlantic regions, size and gradient were key classification
variables (see Appendix A).
To further inform stream classification, EPA performed a broad-scale analysis on
macroinvertebrate survey data from the EPA National Aquatic Resource Surveys (NARS)
program.3 The data set included minimally disturbed freshwater wadeable stream sites from the
Northeast, Mid-Atlantic, and Southeast regions. A cluster analysis was performed, and sites were
grouped into three classes based on similarities in taxonomic composition. EPA then developed a
model based on environmental variables to predict the probability of occurrence of the three
classes in watersheds in the eastern United States. The three classes are referred to as: (1) small
to medium size, medium to high gradient, colder temperature; (2) small, low gradient; and
(3) warmer temperature, larger size, lower gradient. Most of the primary RMN sites that were
selected fall within the small to medium size, medium to high gradient, colder temperature
stream class. On average, sites in this stream class have higher numbers of cold water taxa,
which improves the likelihood of detecting temperature-related trends in this thermal indicator
metric over shorter time periods (see Appendix A).
There were several additional site selection considerations. Where feasible, sites with low
potential for future development were selected because future alterations could limit trend
detection power as well as the ability to characterize climate-related impacts at RMN sites.
Participants utilized what they felt were the best, most current data to assess potential for future
development. The Northeast utilized a spatial data set provided by The Nature Conservancy
(TNC)4 that showed public and private lands and waters secured by a conservation agreement.
Other RMN members contacted city planners and personnel from transportation and forestry
departments to obtain information about the likelihood of future urban and residential
development, road construction, and logging or agricultural activities.
Practical considerations were also important during the site screening process. For
example, organizations generally selected sites that could be sampled during a day trip and were
easy to access, which are factors that will likely increase the frequency at which sites can be
visited. More sites visits may improve the quality of data being collected (particularly the
3Data available at http://water.epa.gov/type/rsl/monitoring/riverssurvey/index.cfm.
4Secured lands data set available at
https://www.conservationgateway.org/ConservationByGeography/NorthAmerica/UnitedStates/edc/reportsdata/terre
strial/secured/Pages/defaultaspx.
3-3
-------
hydrologic data). Sites were prioritized if they were colocated with existing equipment, such as
USGS gages, or if there were opportunities to share the workload with outside agencies or
organizations. Efforts have been made to partner with national monitoring programs, such as the
EPA NARS, Long Term Ecological Research Network and the National Ecological Observatory
Network. For various reasons (e.g., sites are not revisited annually, sites are not located in same
stream class), sites that are being sampled for these programs have not been selected as primary
RMN sites, but EPA and partners are continuing to seek opportunities for collaboration with
these and other potential partners.
Data from additional, "secondary" sites are also being considered for the RMNs. These
are sites at which a subset of parameters are already being collected in accordance with RMN
protocols as part of other independent monitoring efforts. Data from secondary sites will increase
the sample size and range of conditions represented in the RMN data set, and may provide
information about unique or underrepresented geographic areas, such as the New Jersey Pine
Barrens or the Coastal Plain ecoregion. Appendix E lists the candidate secondary RMN sites in
each region as of September 2015.
3.2. METHODS FOR DATA COLLECTION
Efforts are being made to collect the following types of data (consistent with existing
programs and scientific literature) from RMN sites:
Biological indicators: macroinvertebrates
- Optional: fish and periphyton, if resources permit (fish are higher priority)
Temperature: continuous water and air temperature (30-minute intervals)
Hydrological: continuous water-level data (15-minute intervals); converted to discharge
if resources permit
Habitat: parameters agreed upon by regional working group
Water chemistry: in situ, instantaneous water chemistry parameters (specific
conductivity, dissolved oxygen, pH), plus additional or more comprehensive water
chemistry measures agreed upon by regional working group
Photodocumentation: photographs taken from the same locations during each site visit
Geospatial data: percentage land use and impervious cover, climate, topography, soils,
and geology, if resources permit
The goal is to use methods that will maximize the likelihood of detecting subtle changes
over as short a time period as possible, while staying within the resource constraints of
participating organizations. EPA and partners used results from the Northeast power analyses
(see Appendix A), literature and SOPs from participating organizations to help inform methods
decisions. Efforts are being made to use as consistent and comparable methods as possible since
5-4
-------
different methodologies may introduce biases in analyses and contribute to variability, which
reduces the sensitivity of indicators and increases trend detection times.
During the initial phases of RMN development, the regional working groups agreed upon
methods to use at primary RMN sites. These methods are summarized in Appendix F. EPA and
partners recently developed a generic RMN QAPP that details the core requirements for
participation in the network, and outlines best practices for the collection of biological, thermal,
hydrologic, physical habitat, and water chemistry data at RMN sites. The QAPP is available
online at http://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=295758&inclCol=eco#tab-3. The regional
working groups are in the process of reviewing and approving the QAPP, and are customizing it
for their regions via addendums.
The ability of some participants to use the regional RMN methods has been limited by
resource constraints, so there have been differing levels of effort and in some situations, differing
methods across sites and organizations. While this is not ideal, the data can still be used, just in
more limited ways. The data management system that EPA and partners are developing (see
Section 6) will contain metadata that will enable users to select data that meet their needs (e.g.,
collected using certain methods and at certain levels of rigor). To account for the differing levels
of effort across sites and organizations, EPA and partners have broken the sampling
methodologies down into different elements, and different levels of rigor are established for each
element. Examples of elements include type of habitat sampled, gear type, frequency of data
collection, level of taxonomic resolution, level of expertise of field and laboratory personnel, and
quality assurance/quality control (QA/QC) procedures. There are four levels of rigor in the RMN
framework, with Level 1 being the lowest and Level 4 being the best/highest standard (see
Table 3-2). Level 3 is the target for primary RMN sites. These elements and levels of rigor are
covered in more detail in Sections 3.2.1 through 3.2.7.
5-5
-------
Table 3-2. There are four levels of rigor in the regional monitoring network
(RMN) framework, with Level 1 being the lowest and Level 4 being the
best/highest standard. Level 3 is the target for primary RMN sites.
Level
Usability for RMNs
Data are usable under certain or limited circumstances. Data are not collected
and processed in accordance with methods agreed upon by the regional working
group, which severely limit the data's usefulness.
Data are usable under some, but not all circumstances. Only certain aspects of
sample collection and processing are done using the protocols that are agreed
upon by the regional working group, which limit the data's usefulness.
Data meet the desired level of rigor. They are collected in accordance with the
methods that are agreed upon by the regional working group. Where
methodological differences exist, steps have been taken to minimize biases, and
data are sufficiently similar to generate comparable indicators and meet RMN
objectives.
4 (optional)
Data exceed expectations. Data include optional high-quality data and meet or
exceed the desired level of rigor agreed upon by the regional working group.
3.2.1. BIOLOGICAL INDICATORS
Collection of multiple assemblages (macroinvertebrates, fish, periphyton) at RMN sites is
encouraged. At a minimum, macroinvertebrates should be collected at the primary RMN sites.
Collections from this assemblage are central to the RMNs because macroinvertebrates are
already collected by participating states, tribes, RBCs, and other agencies for a variety of other
purposes. For example, macroinvertebrates are crucial for quantifying stream condition because
(1) the assemblage responds to a wide range of stressors, (2) many (not all) are easily and
consistently identified, and (3) they have limited mobility, short life cycles, and are highly
diverse. Guidelines for collecting macroinvertebrates, fish, and periphyton can be found in
Sections 3.2.1.1, 3.2.1.2, and 3.2.1.3, respectively.
Data collection should be done by trained personnel (see Table 3-3) because formal
training can have a large impact on observer agreement and repeatability and can reduce
assessment errors (e.g., Herlihy et al., 2009; Haase et al., 2010). Repeatability is particularly
important for RMNs because data are gathered from multiple sources. Ideally, participating
organizations should adhere to the sample collection and processing protocols that are agreed
upon by the regional working group. Some of these guidelines include QA/QC procedures,
which improve data quality (Stribling et al., 2008; Haase et al., 2010). Example QA/QC
procedures include collecting replicate samples in the field, conducting audits to ensure that
crews are adhering to collection and processing protocols, replicate subsampling (meaning after
subsampling occurs, the subsample is recombined with the original sample and subsampled
again), and validating taxonomic identifications at an independent laboratory.
-------
Table 3-3. Recommendations on best practices for collecting biological data at regional monitoring network
(RMN) sites. The RMN framework has four levels of rigor for biological sampling, with 4 being the best/highest and 1
being the lowest. At primary RMN sites, RMN members should try to adhere to (at a minimum) the Level 3 practices,
which are in bold italicized text.
Component
1 (lowest)
4 (highest)
Expertise
Work is conducted by
a novice or apprentice
biologist or by
untrained personnel
Work is conducted by a
novice or apprentice
biologist under the
direction of a trained
professional
Work is conducted by a
trained biologist who has
some prior experience
collecting the assemblage
of interest
Work is conducted by a trained
biologist who has multiple years of
experience collecting the assemblage
of interest
Collection
and
processing
Some but not all of
the recommended data
are collected.
Not all aspects of
sample collection and
processing use
protocols agreed upon
by the regional
working group
All of the recommended
data are being collected,
but not all aspects of
sample collection and
processing use protocols
agreed upon by the
regional working group
All of the recommended
data are being collected.
All aspects of sample
collection and processing
use protocols agreed upon
by the regional working
group
In addition to the minimum
recommended data, optional data are
also being collected. All aspects of
sample collection and processing use
protocols agreed upon by the
regional working group
QA/QC
No QA/QC
procedures are
performed
Some but not all QA/QC
procedures agreed upon
by the regional working
group are performed
All of the QA/QC
procedures agreed upon by
the regional working group
are performed
QA/QC procedures that are more
stringent than those being used by
the regional working group are
performed
-------
3.2.1.1. Macroinvertebrates
Developing recommendations on macroinvertebrate sampling protocols is challenging
because organizations use different collection and processing protocols when they sample
macroinvertebrates, and each entity's biological indices are calibrated to data that are collected
and processed using these methods. When developing best practices at RMN sites, efforts were
made to accommodate differences in sampling methodologies within regions (see Appendix F)
while still providing data that are sufficiently similar that they can be used to generate
comparable indicators at the regional level, and to minimize variability where possible.
At primary RMN sites, macroinvertebrate sampling should be conducted at least once
annually (see Table 3-4). The Northeast power analyses showed that sampling frequency (1 vs. 2
vs. 5-year intervals) had a significant effect on trend detection time. Sampling
macroinvertebrates on an annual basis improves trend detection times, particularly if trends are
subtle (see Appendix A). Annual data are also important for quantifying temporal variability. As
discussed in Section 5, the data will help us to better understand how natural variability affects
the consistency of biological condition scores and metrics from year to year, and how this relates
to changing thermal and hydrologic conditions.
In the Northeast, Mid-Atlantic and Southeast RMNs, macroinvertebrate samples are
being collected in reaches with abundant riffle habitat (see Table 3-4). Cold water taxa, which
are of particular interest due to their potential vulnerability to climate change, typically inhabit
riffles. Furthermore, riffle habitat is being targeted because sample consistency is strongly
associated with the type of habitats sampled (Parson and Norris, 1996; Germ and Herlihy, 2006;
Roy et al., 2003). Recent methods comparison studies indicate that where abundant riffle habitat
is present, single habitat riffle, reach-wide, and multihabitat samples generally produce
comparable classifications and assessments, especially when fixed counts and consistent
taxonomy are used (e.g., Vinson and Hawkins, 1996; Hewlett, 2000; Ostermiller and Hawkins,
2004; Cao et al., 2005; Gerth and Herlihy, 2006; Rehn et al., 2007; Blocksom et al., 2008).
While sampling at RMN sites is focused primarily on riffles, other habitats are also of interest. In
the Southeast region, in addition to collecting quantitative samples from riffle habitat, some
organizations are also collecting qualitative samples from multiple habitats. They are keeping
taxa from the different habitats separate, which provides information on how changing thermal
and hydrologic conditions impact taxa in nonriffle habitats. For example, taxa in edge habitats
may show a greater response to extended summer low flow events than taxa in riffles because the
edge habitats are more likely to go dry.
-------
Table 3-4. Recommendations on best practices for collecting macroinvertebrate data at Northeast, Mid-Atlantic
and Southeast regional monitoring network (RMN) sites. The RMN framework has four levels of rigor for
macroinvertebrate sampling, with 4 being the best/highest and 1 being the lowest. At primary RMN sites, RMN
members should try to adhere to (at a minimum) the Level 3 practices, which are in bold italicized text.
Component
1 (lowest)
4 (highest)
Sampling
frequency
Site is sampled every 5 or
more years
Site is sampled every
2-4 years
Site is sampled annually
Site is sampled more than once
a year (e.g., spring and
summer)
Habitat
No riffle habitat
Multihabitat composite from
a sampling reach with scarce
riffle habitat
Abundant riffle habitat
Multihabitat sample with taxa
from each habitat kept separate
Time
period
Time period varies from
year to year, and
adjustments are NOT made
for temporal variability
Time period varies from year
to year, but adjustments are
made for temporal variability
Adherence to a single
time period
Samples are collected during
more than one time period
(e.g., spring and late
summer/early fall)
Fixed count
subsample
Presence/absence or field
estimated categorical
abundance (e.g., rare,
common, abundant,
dominant)
Fixed count with a target of
100 or 200 organisms
Fixed count with a target
of 300 organisms
Fixed count with a target of
more than 300 organisms
Processing
Organisms are sorted,
identified and counted in
the field
Samples are processed in the
laboratory by trained
individuals. Some but not all
aspects of sample processing
use methods that are agreed
upon by the regional working
group
Samples are processed in
the laboratory by trained
individuals and use
methods that are agreed
upon by the regional
working group
Samples are processed in the
laboratory by trained
individuals and use methods
that are more stringent than
those being used by the
regional working group
-------
Table 3-4. continued..
Component
1 (lowest)
4 (highest)
Sorting
efficiency
No checks on
sorting efficiency
Sorting efficiency checked
internally by a trained
individual
Sorting efficiency checked
internally by a taxonomist
Sorting efficiency checked by
an independent laboratory
Qualifications
Identifications are
done by a novice or
apprentice biologist
with no
certification
Identifications are done by an
experienced taxonomist
without certification
Identifications are done by a
trained taxonomist who has
the appropriate level of
certification
Identifications are done by a
certified taxonomist who is
recognized as an expert in
species-level taxonomy for one
or more groups
Taxonomic
resolution
Coarse resolution
(e.g., order/family)
Mix of coarse and genus-level
resolution (e.g., family-level
Chironomidae, genus-level
Ephemeroptera, Plecoptera,
and Trichoptera [EPT])
Mix of species and genus
level. Identifications are
done to the level of
resolution specified in
Appendix G
Species level for all taxa, where
practical
Validation
No validation
Taxonomic checks are
performed internally but not
by an independent laboratory.
The entire subsample (referred
to as a "voucher sample") is
retained for each site.
Taxonomic checks are
performed internally but not
by an independent
laboratory. The entire
subsample (referred to as a
"voucher sample") is
retained for each site as well
as a reference collection with
each unique taxon
Taxonomic checks are
performed by an independent
laboratory. The entire
subsample (referred to as a
"voucher sample") is retained
for each site, as well as a
reference collection with each
unique taxon verified by an
outside expert
-------
At primary RMN sites, macroinvertebrate sampling should occur during a consistent time
period to minimize the variability associated with seasonal changes in the composition and
abundances of stream biota and to allow for more efficient trend detection (Olsen et al., 1999).
At RMN sites, samples should be collected during the same time period (or periods) each year,
ideally within 2 weeks of a set collection date (see Table 3-4). If flooding or high water prevents
sample collection within the specified time period, samples should be taken as closely to the
target period as possible. In addition to taxonomic consistency, samples collected during the
same time period can be used to explore whether long-term changes in continuous thermal and
hydrologic measurements are occurring during the target period. For example, streams that were
once perennial may become intermittent during a late summer or early fall sampling period, or
changes in thermal and hydrologic conditions could result in lower abundances or replacement of
certain taxa, which could affect biological condition scores.
In the Northeast RMN, sampling is taking place during a summer/early fall
(July-September) index period because this range overlaps with existing state index periods and
because environmental conditions in the spring are generally not conducive to sampling (e.g.,
potential ice cover). In the Southeast RMN, macroinvertebrate samples are being collected in
April, with some states adding a September sample. States and RBCs in the Mid-Atlantic RMN
are currently collecting samples in both spring and summer, as resources permit. The spring
index period is being restricted to March-April and the summer index period to July-August
because this range overlaps with existing state and RBC index periods and reduces potential
temporal variability to a 2-month window. In the future, if only one collection is possible in the
Mid-Atlantic RMN, the spring index period is preferred because preliminary data suggest that on
average, assemblages are comprised of slightly higher proportions of cold water taxa and
individuals in the spring (see Figure 3-1).
3-11
-------
100
80
r eo
I
o
O
40
20
T
I
spring summer
Season
J/3
ro
^
;g
'>
T3
p
O
O
spring summer
Season
Median
25%-75%
I Non-Outlier Range
Figure 3-1. Seasonal differences (spring vs. summer) in percentage cold
water taxa and individuals in the Mid-Atlantic 2014 data.
When macroinvertebrate samples from primary RMN sites are processed, subsampling
should be performed in a laboratory by trained personnel. Participating organizations should
perform fixed counts with a target of 300 (or more) organisms to reduce sample variability and
ensure sample comparability (see Table 3-4). Consistent subsampling protocols are important
because sampling effort and the subsampling method can affect estimates of taxonomic richness
(Gotelli and Graves, 1996), taxonomic composition, and relative abundance of taxa (Cao et al.,
1997). The 300-organism target is larger than what is specified in some state, tribal, and RBC
methods. The purpose of using this larger fixed count is to increase the probability of collecting
cold water indicator taxa that are rarer and to improve the chances of detecting trends in richness
metrics over shorter time periods, as suggested in the Northeast pilot study (see Appendix A).
Having a 300-organism or higher target is further supported by the species accumulation curve
shown in Figure 3-2. The curve, which is based on preliminary 2014 data from the Mid-Atlantic
RMN, shows that the larger the subsample size, the higher the richness of the thermal indicator
taxa. If organizations normally use lower fixed targets (e.g., 100- or 200-count samples) for their
assessments, computer software can be used to randomly subsample 300- or higher-count
samples to those lower targets.
3-12
-------
0)
o
Cold/Cool - Warm
50
100
150 200
Subsample Size
250
300
Figure 3-2. Species accumulation curve based on the Mid-Atlantic 2014 data.
The larger the subsample size, the higher the richness of thermal indicator taxa.
Taxa collected at primary RMN sites should be identified to the lowest practical
taxonomic level (see Table 3-4). Research has shown that finer levels of taxonomic resolution
can discriminate ecological signals better than coarse levels (Lenat and Resh, 2001; Waite et al.,
2000; Feio et al., 2006; Hawkins, 2006). If this level of resolution is not possible, efforts should
be made to conform to the taxonomic resolution recommendations contained in Appendix G.
These call for genus-level identifications (where possible) for Ephemeroptera, Plecoptera,
Trichoptera, Chironomidae, and Coleoptera and specify certain genera within these taxonomic
groups that should be taken to the species level. These genera were selected because they are
believed to be good thermal indicators and have shown variability in thermal tolerances at the
species level (U.S. EPA, 2012a). Following these recommendations will increase the chances of
detecting temperature-related signals over shorter time periods at RMN sites, and will provide
important information about which taxa are most sensitive to changing thermal conditions. The
recommendations in Appendix G should be regarded as a starting point subject to revision as
better data become available in the future.
3-13
-------
High-quality taxonomy is a critical component of credible ecological research, and
taxonomic identifications for RMN samples should be done by a trained taxonomist who has the
appropriate level of certification (see Table 3-4). Analyses have shown that the magnitude of
taxonomic error varies among taxa, laboratories and taxonomists, and that the variability can
affect interpretations of macroinvertebrate data (Stribling et al., 2008). Sources of these errors
include incorrect interpretation of technical literature, recording errors, and vague or coarse
terminology, as well as differences in nomenclature, procedures, optical equipment, and handling
and preparation techniques (Stribling et al., 2003; Dalcin, 2004; Chapman, 2005). Experience
and training can prevent many of these errors (Haase et al., 2006; Stribling et al., 2008). A
reference collection of each unique tax on should be housed by each agency and made available
for verification or comparison. The entire fixed count subsample (referred to as "voucher
samples") for each primary RMN site should be preserved and archived. When a unique taxon is
removed from a voucher sample for the reference collection, it must be clearly documented.
Reference collections and voucher samples will be particularly important for RMN samples
because identifications often will be made by different taxonomists. If resources permit, a subset
of samples should be checked by a taxonomist from an independent laboratory to validate the
identifications and ensure consistency across organizations.
The collection of certain types of demographic or life history data could reduce the
amount of time needed to detect changes in biological indicators because these traits may
respond to climate change earlier than species richness and abundance (Sweeney et al., 1992;
Hogg and Williams, 1996; Harper and Peckarsky, 2006). Examples include rates of
development, size structure, timing of emergence, and voltinism. More importantly, the
frequency and occurrence of the traits themselves can be linked to environmental conditions and
used to predict vulnerability of other species (e.g., Townsend and Hildrew, 1994; Statzner et al.,
1994; Townsend et al., 1997; Richards et al., 1997; van Kleef et al., 2006; Poff et al., 2006). It is
also worth considering qualitative collections of adult insects to verify or assist in species
identification. At this time, the collection of these types of ancillary data at RMN sites is
optional, and any discussions of additional sampling should consider the costs and benefits of the
data for the states, tribes, or RBCs and RMN objectives.
When developing the macroinvertebrate methods for the RMNs, the intent was to balance
the need to generate comparable data that meets RMN objectives with generating data that has
value for individual RMN member's routine bioassessment programs. Without additional
resources and training, some organizations will not be able to attain these levels of rigor on a
consistent, long-term basis. For example, some organizations will not be able to follow the
regional protocols for the 300-organism count and species-level identifications. Instead, they will
likely follow their normal processing protocols, with counts of 100 or 200 organisms and
genus-level identifications. Reduced counts and coarser level identifications, in particular, are
likely to affect the richness metrics (Stamp and Gerritsen, 2009; also see Figure 3-2).
RMN members should collect each sample using the method agreed upon by the regional
working group and retain this sample, even if the organization lacks sufficient resources to count
3-14
-------
300 organisms and perform species-level identifications at this time, since funds may become
available at a future date to process samples in accordance with the RMN protocols. RMN
members should periodically refresh these samples with preserving agent so that specimens
remain in good enough condition to later be identified. In some cases, regional coordinators may
be able to obtain funding to cover the costs of macroinvertebrate sample processing and
species-level identifications at a common laboratory. For example, EPA Region 3 was able to
achieve this during the 2014 sampling season for the Mid-Atlantic RMN members. Even if this
can only be done for one year, it serves to establish valuable baseline information.
If the RMN protocols differ from those that are normally used by RMN members, EPA
and partners are exploring the possibility of conducting methods comparison studies at a subset
of sites. This could involve the collection of side-by-side samples using the different methods.
After the paired samples are processed using the respective methods, results would be compared
and differences between the methods could be quantified.
3.2.1.2. Fish
The collection offish at RMN sites is optional but encouraged. Fish are considered to be
a higher priority assemblage than periphyton at RMN sites because fish are routinely collected
by monitoring programs, are easily and consistently identified, and are often species of economic
and social importance. Further, the data can be obtained without a significant amount of further
sample processing, making this assemblage a cost-effective group to analyze, and the behavioral
and physiologic traits can be linked to environmental conditions. Many organizations have
strong interests in protecting fisheries, and numerous studies are being done to predict and
monitor how fish distributions will change in response to climate change (e.g., Clark et al., 2001;
Flebbe et al., 2006; Trumbo, 2010; Wenger et al., 2011). Best practices for fish collection at
RMN sites are shown in the following list.
• Participating organizations should follow the protocols that are agreed upon by the
regional working group. At this time, only the Southeast region is consistently collecting
fish data. Because fish sampling protocols are similar across organizations in this region,
the Southeast regional working group agreed to let organizations use their own standard
operating procedures. If organizations in other regions start to sample fish on a regular
basis, this topic should be revisited and the working groups should take an in-depth look
at the comparability offish sampling protocols within and across regions.
• There should be strict adherence to an index period (or periods).
• Species-level identifications should be done (where practical) by a trained fish
taxonomist.
• A reference collection of each unique taxon should be housed by each agency and be
made available for verification or comparison.
3-15
-------
3.2.1.3. Periphyton
The collection of periphyton at RMN sites is optional but encouraged, as periphyton are
important indicators of stream condition and stressors (Stevenson, 1998; McCormick and
Stevenson, 1998). At this time, the Southeast is the only region that has written guidelines for
periphyton collection. Their sampling protocols follow the Southeastern Plains instream nutrient
and biological response protocols (U.S. EPA, 2006) or equivalent. They strictly adhere to a
spring index period and have a subsampling target of 600 valves (300 cells). Species-level
identifications are being done (where practical) by a qualified taxonomist, and reference
collections of unique taxa are being retained. The protocols also recommend that the EPA rapid
periphyton survey field sheet or equivalent be completed (Barbour et al., 1999).
If organizations from other RMNs start to collect periphyton, they should follow the
protocols that are agreed upon by their regional working group. If standardized regional
protocols are not used, the methods that each entity uses should be detailed and well
documented. With periphyton, some programs have encountered problems with taxonomic
agreement among different laboratories and taxonomists, so steps should be taken to ensure
consistency in taxonomic identifications (e.g., send all samples to the same laboratory,
photodocument taxa in reference collections, conduct taxonomic checks with an independent
laboratory).
3.2.2. TEMPERATURE DATA
Some states, tribes, and RBCs have been early adopters of continuous temperature sensor
technology and have written their own protocols for deploying these sensors. In an effort to
increase comparability of data collection across states and regions, EPA and collaborators
published a document on best practices for deploying inexpensive temperature sensors
(U.S. EPA, 2014). The best practices for collecting temperature data at RMN sites closely follow
these protocols.
At primary RMN sites, both air and water temperature sensors should be deployed (see
Table 3-5). Together, the air and water temperature readings can be used to gain a better
understanding of the responsiveness of stream temperatures to air temperatures (also referred to
as thermal sensitivity), and provide insights into the factors that influence the vulnerability of
streams to thermal change (see Section 5). Air temperature readings are also used for quality
control (e.g., to determine when water temperature sensors are dewatered; Bilhimer and Stohr,
2009; Sowder and Steel, 2012).
3-16
-------
Table 3-5. Recommendations on best practices for collecting temperature data at regional monitoring network
(RMN) sites. The RMN framework has four levels of rigor for temperature monitoring, with 4 being the best/highest
and 1 being the lowest. At primary RMN sites, RMN members should try to adhere to (at a minimum) the Level 3
practices, which are in bold italicized text.
Component
1 (lowest)
4 (highest)
Equipment
No temperature
sensors
Water temperature sensor
only
Air and water temperature
sensors
Air temperature sensor plus
multiple water temperature
sensors to measure reach-scale
variability
Period of record
Single
measurement(s)
taken at time of
biological sampling
event
Continuous measurements
taken seasonally (e.g.,
summer only) at intervals
of 90 minutes or less
Continuous measurements
taken year-round at
30-minute intervals
Continuous measurements taken
year-round at intervals of less
than 30 minutes
Radiation shield
Not installed
Installed; the shield is
made using an untested
design (its effectiveness
has not been documented)
Installed; the shield is
made using a design that
has undergone some level
of testing to document its
effectiveness
Installed; the shield is made using
a design that has been tested
year-round, under a range of
canopy conditions
QA/QC—sensor
accuracy
No accuracy checks
are performed
No accuracy checks are
performed
Predeployment accuracy
check is performed, along
with any other QA/QC
checks that are agreed
upon by the regional
working group
In addition to the predeployment
accuracy check, the following
checks are also performed: initial
deployment, mid-deployment,
biofouling, and postdeploymenta
Tor more details, see the QAPP (U.S. EPA, 2016. Generic Quality Assurance Project Plan for monitoring networks for tracking long-term conditions and
changes in high quality wadeable streams), which is available online at http://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=295758&inclCol=eco#tab-3.
-------
Temperature measurements at RMN sites should be taken year-round at 30-minute
intervals (see Table 3-5). Year-round data are necessary to fully understand thermal regimes and
how these regimes relate to aquatic ecosystems (U.S. EPA, 2014). Radiation shields should be
installed for both water and air temperature sensors (see Table 3-5) to prevent direct solar
radiation from hitting the temperature sensors and biasing measurements (Dunham et al., 2005;
Isaak and Horan, 2011). The shields also serve as protective housings. Shield effectiveness varies
by design (Holden et al., 2013), so it is suggested that organizations use tested designs (see
Table 3-5). If a new design is used, organizations should test and document design performance.
This can be done using techniques like those described in Isaak and Horan (2011) and Holden
etal. (2013).
To ensure that data meet quality standards, at a minimum, predeployment accuracy
checks should be performed. In addition, participants are encouraged to perform initial
deployment, mid-deployment, biofouling and postdeployment checks. These types of QA/QC
checks are important because sensors may record erroneous readings during deployment for a
variety of reasons, such as being dewatered or buried in silt. The QA/QC checks improve data
quality and allow for data to be corrected (if needed). The QAPP contains more detailed
information on these checks.
3.2.3. HYDROLOGIC DATA
Many of the primary RMN sites are located on smaller, minimally disturbed streams with
drainage areas less than 100 km2. Monitoring flow in headwater and mid-order streams is
important because flow is considered a master variable that effects the distribution of aquatic
species (Poff et al., 1997), and small streams in particular play a critical role in connecting
upland and riparian systems with river systems (Vannote et al., 1980). These small upland
streams, which are inhabited by temperature sensitive organisms, are also projected to experience
substantial climate change impacts (Durance and Ormerod, 2007), though some habitats within
these streams will likely serve as refugia from the projected extremes in temperature and flow
(Meyer et al., 2007).
The USGS has been measuring flow in streams since 1889, and currently maintains over
7,000 continuous gages. This network provides long-term, high quality information about our
nation's streams and rivers that can be used for planning and trend analysis (e.g., flood
forecasting, water allocation, wastewater treatment, and recreation). Efforts have been made to
colocate RMN sites with active USGS gages, but many gages are located in large rivers that have
multiple human uses, so only a limited number meet the site selection criteria for the primary
RMN sites. As such, it is necessary to collect independent hydrologic data at most RMN sites.
A common way to collect hydrologic data at ungaged sites is with pressure transducers. If
installed and maintained properly, pressure transducers will provide important information on the
magnitude, frequency, duration, timing, and rate of change of flows. These devices can pose
challenges. For one, pressure transducers are more expensive than the temperature sensors,
which makes it more difficult for RMN participants to purchase the equipment. Then, if
3-18
-------
participants are successful at obtaining the transducers, they need the expertise to install and
operate the equipment, and also need resources to conduct QA/QC checks to ensure that the data
meet quality standards. Because of these challenges, some participating organizations have
adopted a "phased" approach, in which they start by installing pressure transducers at one or two
RMN sites (instead of all sites at once), and add more as they gain experience and as resources
permit.
When pressure transducers are installed at RMN sites, efforts should be made to follow
the recommendations in Table 3-6. These closely follow the protocols described in the recently
published EPA best practices document on the collection of continuous hydrologic data using
pressure transducers (U.S. EPA, 2014). Transducer measurements should be taken year-round5
(see Table 3-6). The transducers should be encased in housings to protect them from currents,
debris, ice, and other stressors. Staff gages should also be installed to allow for instantaneous
readings in the field, verification of transducer readings, and correction of transducer drift (see
Figure 3-3, Table 3-6).
When the pressure transducer is installed, the elevation of the staff gage and pressure
transducer should be surveyed to establish a benchmark or reference point for the gage and
transducer (see Table 3-6). This benchmark allows for monitoring of changes in the location of
the transducer, which is important because if the transducer moves, water-level data will be
affected and corrections will need to be applied (see Figure 3-3). While water-level
measurements alone yield information about streamflow patterns, including the timing,
frequency, and duration of high flows (McMahon et al., 2003), they do not give quantitative
information about the magnitude of streamflows or flow volume, which makes it difficult to
compare hydrologic data across streams.
If agencies have the resources to convert water-level measurements to streamflow (e.g.,
volume of flow per second), the most common approach is to develop a stage-discharge rating
curve. To develop a rating curve, a series of discharge (streamflow) measurements are made at a
variety of stages, covering as wide a range of flows as possible. The EPA best practices
document (U.S. EPA, 2014) contains basic instructions on how to take discharge measurements
in wadeable streams. More detailed guidance on this topic can be found in documents like Rantz
et al. (1982), Shedd (2011), or Chase (2005). After establishing a rating curve, discharge should
be measured quarterly. If resources don't permit quarterly measurements, discharge should be
measured at least once annually, and if possible, also after large storms and other potentially
channel-disturbing activities. In addition, elevation surveys should be performed annually or as
needed to check that the sensor has not moved.
5In places where streams become completely frozen during the winter, pressure transducers may be removed during
winter months if freezing will result in damage to the equipment.
3-19
-------
Transducer data
Staff gage readings
Dec Jan 2013
Date
Figure 3-3. Staff gage readings provide a quality check of transducer data. In
this example, staff gage readings stopped matching transducer readings in
November, indicating that the transducer or gage may have changed elevation.
3-20
-------
Table 3-6. Recommendations on best practices for collecting hydrologic data at regional monitoring network
(RMN) sites. The RMN framework has four levels of rigor for hydrologic monitoring, with 4 being the best/highest
and 1 being the lowest. At primary RMN sites, RMN members should try to adhere to (at a minimum) the Level 3
practices, which are in bold italicized text.
Component
1 (lowest)
4 (highest)
Equipment
Pressure transducer,
water only; no staff
gage
Pressure transducer,
water and air
(encased in
housings); no staff
gage
Pressure transducer, water and
air (encased in housings); staff
gage installed
Same as Level 3, plus a
precipitation gage or USGS gage
Type of data
to
Stage/water level
only; data are not
corrected for
barometric pressure
Stage/water level
only; data are
corrected for
barometric pressure
Stage/water level corrected for
barometric pressure. In
addition, a minimum of
5-10 discharge measurements
are taken at a variety of flow
conditions to develop a
stage-discharge rating curve.
The stage-discharge rating
curve is used to convert water
level to flow/discharge
Stage/water level corrected for
barometric pressure. In addition,
10 or more discharge
measurements that capture the
full range of flow conditions are
taken to develop a
stage-discharge rating curve. The
stage-discharge rating curve is
used to convert water level to
flow/discharge.
Period of
record
Discharge
measurements taken
with flow meter at
time of biological
sampling event
Continuous
measurements taken
seasonally (e.g.,
summer only)
Continuous measurements
taken year-round11
Continuous measurements taken
year-round51
Elevation
survey
Not performed
Performed once, at
time of installation
Performed annually
Performed more than once a year,
as needed (e.g., if a storm moves
the sensor and it has to be
redeployed)
-------
Table 3-6. continued...
Component
1 (lowest)
4 (highest)
QA/QC—sensor
accuracy
No accuracy checks
are performed
No accuracy checks
are performed
At least once annually, field
crews take a staff gage reading
or water depth measurement
over the transducer with a
stadia rod or other measuring
device and compare this to the
sensor reading
Multiple times per year, field
crews take a staff gage reading
or water depth measurement
over the transducer with a stadia
rod or other measuring device
and compare this to the sensor
reading
to
to
QA/QC—stage-
discharge rating
curve
After the rating curve
is established, no
checks are performed
to verify the
stage-discharge rating
curve
After the rating curve
is established, no
checks are performed
to verify the
stage-discharge
rating curve
After the rating curve is
established, discharge is
measured at least once
annually to verify the
stage-discharge rating curve,
and if possible, also after large
storms or any other potentially
channel-disturbing activities
After the rating curve is
established, discharge
measurements are taken
quarterly to verify the
stage-discharge rating curve,
and if possible, also after large
storms or any other potentially
channel-disturbing activities
QA/QC—
discharge
No discharge checks
are performed
No discharge checks
are performed
Periodically, duplicate
discharge measurements are
taken, ideally by different
people1'
Periodically, duplicate discharge
measurements are taken, ideally
by different people. Discharge
measurements are also
periodically compared to a
standard, such as a real-time
USGS gage, or to measurements
obtained by an experienced
hydrographer from the USGS or
another agency.
aln places where streams become completely frozen during the winter, pressure transducers may be removed during winter months if freezing will result in
damage to the equipment.
bFor more details, see the QAPP (U.S. EPA, 2016. Generic Quality Assurance Project Plan for monitoring networks for tracking long-term conditions and
changes in high quality wadeable streams), which is available online at http://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=295758&inclCol=eco#tab-3.
-------
As with temperature sensors, different types of errors can occur during deployment (e.g.,
the pressure transducers may become dewatered, buried in sediment, or fouled by algae).
Participants are encouraged to perform QA/QC checks to improve data quality and allow for data
to be corrected (if needed). For example, during site visits, field crews should take a staff gage
reading or water depth measurement over the transducer with a stadia rod or other measuring
device and compare this to the sensor reading. Periodically, the transducer measurements can
also be compared to a standard, such as a real-time USGS gage, or to measurements obtained by
an experienced hydrographer from the USGS or another agency. Additional information on
QA/QC checks for the hydrologic data can be found in the QAPP.
Because the collection of high quality hydrologic data is resource-intensive, states, tribes,
RBCs, and other participating organizations are encouraged to explore partnerships with the
USGS, universities, and other organizations (e.g., volunteer watershed groups). Some states have
been successful at forging such partnerships. For example, the Massachusetts Department of
Environmental Protection (MA DEP) has formed a partnership with the Massachusetts River
Instream Flow Stewards (RIFLS) program. MA DEP collects macroinvertebrate and temperature
data from the primary RMN sites, while the RIFLS program collects the flow data. New
Hampshire Department of Environmental Sciences has partnered with Plymouth State
University, who provided pressure transducers and helped with installations at New Hampshire's
primary RMN sites.
In the future, it would be valuable to start collecting precipitation data as well at the
primary RMN sites. Similar to air and water temperature relationships, these data can be used to
track responsiveness of stream flow to precipitation. Partnerships through groups, such as the
Community Collaborative Rain, Hail, and Snow Network (http://www.cocorahs.org/), can help in
this regard. Any discussions of additional sampling should consider the costs and benefits of the
data for the states, tribes, or RBCs and RMN objectives.
3.2.4. PHYSICAL HABITAT
During the first several years of data collection, EPA and partners considered the
biological, thermal and hydrologic data to be higher priority than the habitat and chemistry data,
so not all participants have been collecting habitat data. Of the entities that have been collecting
habitat data, most have been using qualitative assessments like EPA's Rapid Bioassessment
Protocol (Barbour et al., 1999). These qualitative assessments rate instream, bank, and riparian
habitat parameters using visual descriptions that correspond to various degrees of habitat
condition (e.g., optimal, suboptimal, marginal, and poor). With the proper training, skilled field
biologists can perform comparable and precise visual habitat assessments, and these data,
combined with photographs, can be used to qualitatively track habitat changes at RMN sites
through time.
The regional working groups are starting to reevaluate the habitat protocols. Many RMN
participants feel that quantitative measurements would be better suited for trend detection.
Quantitative habitat data would also be helpful for stream classification. EPA performed a
3-23
-------
broad-scale classification analysis on macroinvertebrate survey data in the eastern United States
and found that substrate (percentage sand, percentage fines, embeddedness), flow habitat
(percentage pools), and reach-scale slope were important predictor variables (see Appendix A).
Collecting these types of quantitative habitat data at RMN sites would improve the ability to
accurately classify sites and help inform decisions on how data from RMN sites could be pooled
together for analyses.
At this time, EPA and partners are encouraging RMN participants to collect the following
types of quantitative habitat data at RMN sites:
Geomorphol ogi cal
- Bankfull width (reach-wide mean or at an established transect)
- Bankfull depth (reach-wide mean or at an established transect)
- Reach-scale slope
Habitat
- Substrate composition (pebble counts to get percentage fines, percentage sand, etc.)
- Flow habitat types (percentage riffle, percentage pool, percentage glide, percentage
run)
- Canopy closure (measured with spherical densitometer, mid-stream and along bank)
There are several published methods, such as the EPA National Rivers and Streams
Assessment protocols (U.S. EPA, 2013a; Kaufmann et al., 1999), for making these
measurements. All of the methods require expertise and skill, and some can be time intensive. As
such, the regional working groups will decide which specific quantitative habitat methods to use
at RMN sites. The frequency with which quantitative habitat data should be collected from RMN
sites also warrants further discussion. It may not be necessary to collect these types of data on an
annual basis because channel forming flows that could change baseline geomorphological and
instream habitat features generally have 1-2 year return periods for bankfull events or 1-5 year
return periods for small flood events. However, specifying an exact timeframe for these
measurements is difficult because channel-forming flows are hard to predict and their impacts at
a given site can be highly variable. To help inform this discussion, one possibility would be to
conduct a pilot study in which RMN members collect quantitative data on an annual basis at a
subset of sites and then quantify how much the measurements vary from year to year and from
site to site. If this type of comparison is not feasible, another option would be to take quantitative
measurements less frequently but then also take measurements when visible geomorphic changes
are seen in the photodocumentation (see Section 3.2.6). This topic warrants further discussion
among RMN work group members and outside experts.
Also of interest are habitat measurements that are likely to be impacted by climate
change. Climate change could contribute to temporally and spatially complex fluvial adjustments
(Blum and Tornqvist, 2000). Some of the effects will be direct (e.g., changing precipitation
3-24
-------
patterns will alter hydrologic regimes, rates of erosion, and sediment yields). Other effects will
be indirect, such as increases in sediment yield, which may result from vegetation disturbances
that stem from changing thermal and hydrologic conditions (e.g., wildfire, insect/pathogen
outbreak, drought-related die off; Goode et al., 2012). Modeling studies from a range of different
environments suggest that the increases in rates of erosion could be on the order of 25-50%
(Goudie, 2006). Changes in the frequency or magnitude of peak flows could cause significant
channel adjustments, especially in higher order streams (Faustini, 2000), but channel adjustments
will vary according to many factors. For example, channel adjustments and changes in sediment
transport and storage can be greatly influenced by large woody debris dams and boulders that
increase roughness (Faustini and Jones, 2003). Climate-related changes in riparian vegetation
may also occur (e.g., Iverson et al., 2008; Rustad et al., 2012), which could in turn affect the
structure and composition of the benthic macroinvertebrate community (Sweeney, 1993; Whiles
and Wallace, 1997; Foucreau et al., 2013).
Monitoring the effects of climate change on physical habitat at RMN sites could be
greatly improved by adding carefully selected measurements of geomorphology and quantitative
habitat indicators. These measures could include indicators that directly or indirectly reflect
changes in hydrology and vertical or lateral channel adjustments (e.g., cross-sectional transects,
mean bankfull height throughout a study reach, bank stability, and pebble counts). Indices of
relative bed stability (Kaufmann et al., 2008; Kaufmann et al., 2009), measures of
embeddedness, or metrics derived from pebble counts (e.g., percentage fines) might be useful
measures in characterizing the effects of climate change if hydrological changes result in
changes to rates of erosion, channel geometry, slope, bank stability, or sediment supply.
However, more discussion among RMN work group members and outside experts is needed
before recommending additional habitat measurements.
3.2.5. WATER CHEMISTRY
In situ, instantaneous water chemistry parameters (specific conductivity, dissolved
oxygen, and pH) should be collected when RMN sites are visited for biological sampling. Some
participating organizations have also been collecting more complete water quality data (e.g.,
alkalinity, major cations, major anions, trace metals, nutrients). The regional working groups are
considering whether to require that a subset of these additional water quality parameters be
collected at primary RMN sites. If sufficient resources are available, these water chemistry
samples could potentially be collected multiple times per year during different flow conditions.
The purpose of collecting these data is to document whether water quality changes are occurring
that could potentially contribute to changes in biological assemblage composition and structure
over time.
3.2.6. PHOTODOCUMENTATION
Digital photographs should be taken when RMN sites are visited for biological sampling.
Photographs are important to document any changes to the monitoring locations, show the
3-25
-------
near-stream habitat where data are being collected, provide qualitative evidence of changes in
geomorphology (e.g., lateral and vertical channel stability), and to locate sensors during
subsequent visits (U.S. EPA, 2014). During each visit, the photographs should be taken from the
same location(s). Global positioning system (GPS) coordinates (latitude and longitude) should be
recorded for the location where the photographs are taken, as well as cardinal direction. The
coordinates should be recorded in decimal degrees, using the NAD83 datum for consistency. In
areas with good satellite reception, field personnel should wait until there is coverage from four
or more satellites before recording the coordinates. The accuracy of the coordinates should later
be verified in the office or laboratory by using software (e.g., Google Earth or geographic
information system [GIS] software) to plot the location on a map. If GPS coordinates are not
available on-site, the location (or locations) should be marked on a map and the coordinates
At least one set of photographs should be taken from a location at mid-reach. The photos
should be taken looking upstream and downstream from this location, and should include
specific and easily identifiable objects such as large trees, large stable boulders, large woody
debris, point bars, established grade control, and so forth (see Figure 3-4). In addition, field
personnel are encouraged to take photos of the riffles where macroinvertebrates are collected
and, for hydrologic data, the location where instantaneous discharge measurements are taken.
Photos of point bars (dominant substrate, extent and type of vegetation) and of banks at
established transects are also of interest to document any changes in physical habitat. The photos
should be archived yet easily accessible for future use.
3-26
-------
Figure 3-4. Photodocumentation of Big Run, WV, taken from the same
location each year.
Source: Provided by West Virginia Department of Environmental Protection (WV DEP).
3.2.7. GEOSPATIAL DATA
If resources permit, GIS software can be used to obtain land use and land cover data for
RMN sites based on exact watershed delineations for each site. Percentage land use and
impervious cover statistics should be generated from the most recent National Land Cover
Database (NLCD), and changes in these statistics should be tracked over time. For the RMNs,
the most current NLCD data set is preferred over other land use data sets because it is a
3-27
-------
standardized set of data that covers the conterminous United States and can be used with a
standardized disturbance screening process (see Appendix D). Drainage area should also be
calculated for each RMN site.
Having exact watershed delineations for RMN sites makes it possible to obtain a wide
range of additional geospatial data (e.g., climate, topography, soils, geology), and can also be
used to generate flow and temperature statistics (Carlisle et al., 2010; Carlisle et al., 2011; Hill
et al., 2013). For purposes of the RMNs, data that are available at a national scale from the
NLCD are preferred to landscape-level variables generated from sources that do not provide
nationwide coverage, in order to standardize disturbance screening for sites and facilitate other
comparisons and analyses. In addition, it would be valuable to examine aerial photographs of the
RMN sites for signs of past disturbance, because past land use can have lasting impacts on
stream biodiversity (Harding et al., 1998). The use of high resolution Light Detection and
Ranging data is also encouraged (if available) to delineate geomorphic features.
3-28
-------
4. SUMMARIZING AND SHARING REGIONAL MONITORING NETWORK (RMN)
DATA
This section contains recommendations on how to summarize the biological, thermal and
hydrologic data that are being collected at RMN sites.6 At a minimum, certain sets of metrics or
statistics should be calculated from the RMN data so that samples can be characterized and
compared in a consistent manner. A consistent set of summary metrics also helps in sharing data
across organizations. Metrics were selected that are:
• Relevant in the context of biomonitoring and to RMN members
• Straightforward to calculate and interpret
• Known or hypothesized to be most strongly associated with biological indicators
• Known or hypothesized to respond to climate change, and
• Limited in redundancy
These lists of metrics are intended to serve as starting points and should be reevaluated
after the first several years of data collection at RMN sites. Periodic literature reviews should
also be conducted to help inform parameter selection. Given the rapid pace of research, it is
important that the raw data collected at RMN sites be properly archived and stored so that
additional metrics can be calculated in the future.
4.1. BIOLOGICAL INDICATORS
Hundreds of different metrics could potentially be calculated from the biological data
being collected at RMN sites. When developing the list of recommended summary metrics for
the macroinvertebrate data, EPA and partners used a combination of published literature and best
professional judgment to narrow down the list. The list, which can be found in Appendix H,
contains both taxonomic and traits-based metrics. The list of taxonomic-based metrics includes
measures like total taxa richness and Ephemeroptera, Plecoptera, and Trichoptera (EPT) richness
and composition (Barbour et al., 1999), which are commonly used by biomonitoring programs
for site assessments. Traits-based metrics related to thermal and hydrologic conditions are also
included (e.g., functional feeding group, habit, thermal, and flow preference). Trait assignments
were obtained from the Freshwater Traits database7 (U.S. EPA, 2012b).
To derive the thermal preference metrics, methods described in Yuan (2006) were used to
estimate the optimal temperature values and ranges of occurrence (tolerances) for taxa that had a
sufficient distribution and number of observations to support the analysis. These data, along with
6The management and sharing of habitat and water chemistry data are discussed in Section 6 (Data Management).
7If time and resources permit, regional experts will review and edit the trait assignments and fill in data gaps where
possible.
4-1
-------
supplemental data provided by states and best professional judgment of regional experts, were
used to derive lists of cold and warm water taxa for the eastern states that are participating in the
current phase of RMN work (see Appendix I). Metrics known or hypothesized to be sensitive to
changing hydrologic conditions are also included in Appendix H. These metrics were selected
based primarily on literature review (e.g., Kerrigan and Baird, 2008; Chiu and Kuo, 2012;
U.S. EPA, 2012a; DePhilip and Moberg, 2013b; Conti et al., 2014). These thermal and
hydrologic traits-based metrics should be reevaluated periodically and refined as more data
become available and more is learned about relationships between biological, thermal and
hydrologic data.
In addition to the taxonomic and traits-based metrics, metrics of persistence and stability
are also being recommended. Persistence is a measure of variation in community richness over
time (Holling, 1973), while stability measures the variability in relative abundance of taxa in a
community over time (Scarsbrook, 2002; for formulas, see Appendix H). The persistence and
stability metrics can be used to quantify year-to-year variation in long-term data sets (Durance
and Ormerod, 2007; Milner et al., 2006). Quantifying natural variation in the occurrence and the
relative abundance of individual taxa allows biomonitoring programs to assess how this variation
affects the consistency of biological condition scores and metrics, and whether variation is linked
to specific environmental conditions. In addition, changes in the occurrence (i.e., presence or
absence) and the relative abundance of individual taxa should also be evaluated at RMN sites.
Spatial distribution maps like the one shown in Figure 4-1 can be created periodically (e.g., every
5 years) to track changes in species distributions over time.
Biological condition scores should also be calculated at RMN sites. Biological indices
often take the form of multimetric indices (MMIs) or predictive models like the River
Invertebrate Prediction and Classification System (Wright, 2000). MMIs are generally a
composite of biological metrics selected to capture ecologically important structural or
functional characteristics of communities, where poor MMI scores represent deviations from
reference condition (Karr, 1991; Barbour et al., 1995; DeShon, 1995; Yoder and Rankin, 1995;
Sandin and Johnson, 2000; Bohmer et al., 2004; Norris and Barbour, 2009). Predictive models
compare which reference site taxa are expected (E) to be present at a site, given a set of
environmental conditions, to which taxa are actually observed (O) during sampling, where low
O:E community ratios represent deviation from reference condition (Wright et al., 1984; Wright,
2000; Hawkins, 2006; Pond and North, 2013).
4-2
-------
Sweltsa
Eurytophelta
N = 163
Ablabesmyia
Stenacron
N = 296
N = 114
Figure 4-1. Spatial distributions of macroinvertebrate taxa, based on the
National Aquatic Resource Survey (NARS) data. These types of maps could be
created periodically to track shifts in spatial distributions of taxa over time. At
regional monitoring network (RMN) sites, particular attention will be paid to
changes in the thermal indicator taxa (in this example, the top two plots show
spatial distributions of two of the cold water indicators; the bottom two plots
show distributions of warm water indicators).
At this time, there are no plans to develop regional MMIs or O:E predictive models.
Rather the biological condition scores should be calculated in accordance with each entity's
bioassessment methods. Because different organizations use different techniques for calculating
biological condition scores, the index scores themselves may not be comparable across sites
sampled by different organizations. However, as discussed in Section 5 (Data Usage) the
direction of trends can be tracked over time. In some locations, it may be possible to obtain
comparable biological condition scores from regional Biological Condition Gradient (BCG)
4-3
-------
models. Section 5 (Data Usage) contains more detailed information on how the BCG could
potentially be used to summarize RMN data.
The biological data from the RMN sites will need to undergo some level of review,
formatting and standardization before it can be summarized and shared. For example, there will
be differences in nomenclature across entities that need to be resolved, as well as differences in
levels of taxonomic resolution. Users may need to develop operational taxonomic units (OTUs)
to address changes in taxonomic naming and systematics that have occurred over time (Cuffney
et al., 2007). EPA and partners are currently working on guidance, procedures and R scripts (R
Core Team, 2015) that will help facilitate the sharing of the biological data. As discussed in
Section 6, the biological data will be eventually be uploaded to a national water data system such
as the Water Quality Exchange (WQX). Users will need to be able to access metadata from the
data management system so that they can select data that meet their needs (e.g., collected using
certain methods and at certain levels of rigor). The raw biological data collected at RMN sites
should be properly archived and stored so that additional metrics can be calculated in the future.
4.2. THERMAL STATISTICS
Many metrics can be calculated from year-round air and water temperature
measurements. Summer temperature metrics are typically used in analyses with biological data
because summer captures a critical time period for most aquatic species' survival, and have been
found to predict macroinvertebrate distributions better than winter and summer temperature
metrics (Hawkins et al., 2013). Beyond this, there is currently limited information on which
temperature metrics are most ecologically meaningful in the context of biomonitoring.
When developing a list of potentially important temperature metrics for the RMN data,
EPA and partners sought input from organizations that have been collecting and processing
continuous stream temperature data for years, including MD DNR and the U.S. Forest Service
Rocky Mountain Research Station (Isaak and Horan, 2011; Isaak et al., 2012; Isaak and Rieman,
2013). They recommended a list of basic statistics that cover daily, monthly, seasonal, and
annual time periods, and basic percentage exceedance metrics (e.g., percentage of days that
exceed 20°C). The metrics are easy to calculate and capture various aspects of thermal regimes,
such as magnitude, frequency, duration, and variability. The list, which can be found in
Appendix J, should be regarded as a starting point. Other unlisted metrics also have promise,
including the use of more complex temperature exceedance metrics and moving average
calculations that are related to specific biological thresholds. Some studies have found that
moving average metrics such as 7-day mean and maximum are useful descriptors of thermal
regimes and often associate well with stream fish distribution patterns (Wehrly et al., 2003;
Nelitz et al., 2007). Other studies (e.g., Butryn et al., 2013) have found that additional metrics are
needed to sufficiently capture the variation caused by irregular and extreme events. As such, it is
important that the raw data collected at RMN sites is properly archived and stored so that
additional metrics can be calculated in the future.
4-4
-------
Before the temperature metrics are calculated, the continuous data should be run through
a series of quality assurance checks, which are described in the QAPP. EPA and partners are in
the process of developing R scripts and procedures to facilitate completion of quality control
activities, summarizing and sharing the temperature data (see Section 6). Other options are also
available for generating summary statistics, such as the ThermoStat software package (Jones and
Schmidt, 2012) and the Stream Thermal Version 1.0 R code package that was recently developed
by Tsang et al. (in review). In addition to calculating the summary statistics, the metadata for
each site should also be reviewed. Data should be interpreted with caution if no accuracy checks
were performed during the deployment period. For more information on data management, see
Section 6.
4.3. HYDROLOGIC STATISTICS
As with the thermal data, many different metrics can be calculated from daily hydrologic
data. Researchers have investigated which hydrologic metrics are most ecologically meaningful
in the context of state biomonitoring programs (e.g., Kennen et al., 2008; Chinnayakanahalli
et al., 2011) but a detailed understanding of how flow affects ecological conditions remains
elusive, in part because observed hydrologic data are unavailable for many biological sampling
sites. Also, due to the highly variable nature of hydrologic data, it takes a long period of record
to characterize hydrologic regimes. Richter et al. (1997) and Huh et al. (2005) suggest that at
least 20 years of data are needed to calculate interannual variability for most hydrologic
parameters, and that 30 to 35 years of data may be needed to capture extreme high and low
events (e.g., 5- and 20-year floods; Olden and Poff, 2003; DePhilip and Moberg, 2013b).
When developing the list of recommended hydrologic metrics for the RMN data, EPA
and partners used a combination of published literature and best professional judgment. The
literature included reports from TNC and several partners (states, RBCs, other federal agencies),
who developed ecosystem flow needs for some eastern and midwestern rivers and their
tributaries (e.g., the Susquehanna, the Upper Ohio, the Delaware, and the Potomac Rivers;
Cummins et al., 2010; DePhilip and Moberg, 2013a, 2013b; Buchanan et al., 2013). TNC and its
partners utilized components of the Ecological Limits of Hydrologic Alteration (ELOHA)
framework (Poff et al., 2010) to make recommendations on flows to protect species, natural
communities, and key ecological processes within various stream and river types. For the Upper
Ohio River, they recommended a list of flow statistics that capture ecologically meaningful
aspects of hydrologic regimes (see Appendix K; DePhilip and Moberg, 2013b). Research by
Olden and Poff (2003) and Hawkins et al. (2013; see Appendix K), which identifies hydrologic
metrics that capture critical aspects of hydrologic regimes and are ecologically meaningful in
different types of streams, also informed the list of metrics.
Appendix K contains the list of recommended hydrologic statistics to calculate for data
from RMN sites where water-level or flow data are being collected. The metrics are relatively
easy to calculate and include both summary statistics and measures of variability. As with the
thermal metrics, this list should be regarded as a starting point and should be reevaluated over
4-5
-------
time. The raw hydrologic data collected at RMN sites should be properly archived and stored so
that additional metrics can be calculated in the future.
Before the hydrologic metrics are calculated, the data should be run through a series of
quality assurance checks, which are described in the QAPP. EPA and partners are in the process
of developing R scripts and procedures to facilitate completion of quality control activities,
summarizing and sharing the hydrologic data (see Section 6). Other options may be available as
well, such as software like Indicators of Hydrologic Alteration (TNC, 2009) and Aquarius.8 In
addition to calculating the summary statistics, the metadata for each site should also be reviewed.
Data should be interpreted with caution if no accuracy checks (e.g., staff gage readings) were
performed during the deployment period, and if the elevations of the staff gage and pressure
transducer were not surveyed. The latter are especially important, because they can determine
changes in the location of the transducer. If the transducer moves, stage data will be affected and
corrections should be applied.
To supplement missing field data or provide estimates of streamflow at ungaged sites,
simulation models have been developed in some geographic areas. For example, the Baseline
Streamflow Estimator simulates minimally altered streamflow at a daily time scale for ungaged
streams in Pennsylvania. This freeware is publicly available, and has a user-friendly point-and-
click interface (Stuckey et al., 2012). Other examples of tools used to simulate flows are listed in
Appendix K. While these modeled data should not be regarded as a substitute for observational
data, participating organizations may want to take advantage of these additional resources to
supplement their monitoring efforts.
8http://aquaticinformatics.com/
4-6
-------
5. DATA USAGE
Biomonitoring programs can use RMN data for multiple purposes, spanning time periods
of 1-5, 5-10 and 10+ years (see Figure 5-1). Uses include:
• Monitoring the condition of minimally and least disturbed streams
• Detecting trends attributable to climate change
• Supplementing CWA programs and initiatives
- Defining natural conditions/quantifying natural variability to support Section 305(b)
programs
- Informing criteria refinement or development under Section 303
- Developing biological indicators for protection planning for Section 303(d) programs
• Gaining a better understanding of relationships between biological, thermal, and
hydrologic data
• Gaining a better understanding of ecosystem responses and recovery from extreme
weather events
• Gaining insights into effects of regional phenomena such as drought, pollutant/nutrient
deposition and riparian forest infestations on aquatic ecosystems and bioassessment
programs
5.1. APPLICATIONS IN A 1-5 YEAR TIMEFRAME
Many of the RMN sites are located on minimally or least disturbed streams (per Stoddard
et al. 2006), which are the standard against which other bioassessment sites are compared. It is
critical to document current conditions at these sites and to monitor how conditions change over
time, as this has implications for CWA programs. Monitoring high quality waters fits in with the
long-term vision and goals for a number of CWA programs, such as the Section 303(d) Program.
Historically, the 303(d) program has focused on the assessment and identification of waters that
are not meeting State water quality standards and on the development of Total Maximum Daily
Loads to inform restoration of those waters, but starting in 2016, protection planning priorities
that target high quality sites will also be incorporated into the reporting cycle (U.S. EPA, 2013b).
Monitoring high quality waters also ties into EPA's Healthy Watershed Initiative, in which State
and other partners identify high quality watersheds and develop and implement watershed
protection plans to maintain the integrity of those waters (U.S. EPA, 2011).
5-1
-------
Supplementing Clean Water Act (CWA) programs
• Informing criteria development or refinement (e.g.,
defining natural conditions/quantifying natural variability)
• Water Temperature
• Environmental/sustainable flows
• Section 303(d) Program - protection planning priorities
• Refining lists of biological, thermal and hydrologic indicators
1-2 years 3+years
Establish
current
I conditions
5-10 years
Continuation of CWA uses
Detecting trends in high
quality waters
• Track commonly-used
biological indicators such as:
0 EPT abundance
° Biological condition
scores
10+years
• Continuation of trend analyses
• Analyze climate-sensitive indicators
• Adaptive management
• Track effectiveness of adaptation
strategies
Figure 5-1. RMN data can be used for multiple purposes, over short- and
long-term timeframes.
To characterize current conditions at the RMN sites, the data, metrics, and summary
statistics described in Section 4 and Appendices H, J, and K will be compiled for each site and
sent to regional coordinators and EPA. Before this happens, the interim data infrastructure
systems described in Section 6 will be put into place. The procedures and R scripts (R Core
Team, 2015) that EPA and partners are currently working on will help facilitate these outputs.
For the macroinvertebrate data, the output will include metrics that are commonly used by
biomonitoring programs for site assessments (e.g., EPT metrics), as well as traits-based metrics
related to thermal and hydrologic conditions. At this time, there are no plans to develop regional
MMIs or O:E predictive models. Rather, assessments of overall biological condition will be
based on biological condition scores that are calculated in accordance with each entity's
bioassessment methods. In many cases, the index scores will not be comparable across sites
sampled by different organizations, but valuable information can be gleaned by monitoring the
direction of trends in biocondition scores across RMN sites, in addition to changes in the
biological metrics. Moreover, some programs may be able to use the biological data from RMN
sites to help calibrate or refine biological indices specific to their programs.
In some places, it may be possible to obtain comparable biological condition scores from
regional BCG models. The BCG is a conceptual, narrative model that describes how biological
5-2
-------
structure and function of aquatic ecosystems change along a gradient of increasing anthropogenic
stress (Davies and Jackson, 2006). The BCG model can be calibrated and applied to regional and
local conditions and puts biological condition on a common, quantifiable scale that can be
applied nationwide. BCG models are typically calibrated to six levels that reflect a continuum of
quality from pristine (BCG Level 1) to severely degraded (BCG Level 6; Davies and Jackson,
2006). Thus, a BCG Level 2 sample in one region is comparable to a BCG Level 2 sample in
another region because both assessments are dependent on comparisons to natural conditions. At
this time regional BCG models for macroinvertebrate or fish assemblages have been developed
for cold and cool streams in the Northern Forest region of the Midwest (Stamp and Gerritsen,
2009) and medium to high gradient streams in parts of New England (Stamp and Gerritsen,
2009). In addition, BCG models for fish and macroinvertebrate assemblages have been calibrated
for northern Piedmont streams of Maryland (Stamp et al., 2014), and are currently being
calibrated in Alabama, Illinois, and Indiana. These models can be applied to data collected from
RMN sites and BCG-level scores, as well as the component metrics of the BCG models (which
are typically related to tolerance of individual taxa), can be used to characterize biological
condition and track changes at sites over time.
Once the first several years of biological data become available, taxonomic composition
at the RMN sites will also be evaluated, along with the broad-scale macroinvertebrate
classification model developed by EPA and partners (see Section 2.2). Based on the broad-scale
model, which was developed using National Rivers and Streams Assessment (NRSA) data, most
of the primary RMN sites fall within the small to medium-size, colder temperature, faster water
stream class. EPA and partners will use nonmetric multidimensional scaling (NMDS) to evaluate
similarities and differences in taxonomic composition across RMN sites and to test the
performance of the classification model. Results will help inform if and how macroinvertebrate
data can be pooled for regional analyses. The ability to pool data could be particularly valuable
for biomonitoring programs that are trying to calibrate biocondition indices and develop numeric
biocriteria but only have limited numbers of high quality sites.
EPA and partners will also assess the number of cold/cool thermal indicator taxa at RMN
sites. Having higher numbers of cold/cool taxa at RMN sites will improve trend detection ability
(see Appendix A). Based on preliminary analyses, RMN sites have relatively high proportions of
cold/cool taxa (see Figure 5-2). The biological and continuous stream temperature data from
RMN sites can be used to refine the regional list of cold/cool and warm water macroinvertebrate
taxa, which were developed based on instantaneous stream temperature measurements (see
Appendix I). Biological indicator lists can also be used for protection planning. For example,
regulatory agencies in Maryland are currently assessing the accuracy of their current use
designations for cold water streams as part of their protection planning process. As part of these
efforts, MD DNR used continuous temperature data from its Sentinel Sites Network (SSN)9 to
develop a thermal indicator organism list for macroinvertebrates. Two stoneflies, Sweltsa and
9MD DNR has been collecting biological data and continuous temperature data from 27 high quality sites since 2000
as part of it SSN. Five of the SSN sites are primary RMN sites. For more information, see Becker et al. (2010).
5-3
-------
Tallaperla, meet obligate cold taxa requirements for Maryland streams and are being used in
combination with trout to help identify and protect cold water streams (see Figure 5-3).
% Cold/cool water taxa
Based on data
received to date
Figure 5-2. Proportion of cold/cool indicator taxa at RMN sites, based on
preliminary data from a subset of sites. More details on the cold/cool taxa list
can be found in Appendix I.
20
10 -
16
14
1!
10 -
I
s
4
2 -
0
Brook Trout and Cold water Berithic Taisa — Temperature Distribution
Non-tidal cold
water
10
15
20
Figure 5-3. The thermal tolerances of Sweltsa and Tallaperla match very
closely with brook trout. These two macroinvertebrate taxa are being used in
combination with trout to help identify and protect cold water streams in
Maryland. This figure was provided by the Maryland Department of Natural
Resources.
5-4
-------
The biological and stream temperature data from RMN sites can also be used to inform
criteria refinement or development and to help identify ecologically meaningful thresholds. For
example, some regulatory agencies are in the process of assessing whether their current
temperature criteria are adequately protecting designated uses related to cold water fisheries. In
Connecticut, Beauchene et al. (2014) used year-round temperature data and fish data to develop
quantitative thresholds for three major thermal classes at which there are discernible
temperature-related changes in stream fish communities (see Figure 5-4). This type of
information is very useful for fisheries management and can be used to help make criteria more
biologically meaningful and defensible.
Daily Mean V
3d-
25-
20-
0
$ 15-
10-
5
0-
month number
Vater Temperatures (CT DEEP TITAN Class July Mean)
1
K?
i
i
i
|T
'
YV
m
'
#
H
1*
1
1
1l *
23456 789 10 11 12
July mean
DEEP titjn
Dcdd
B Warm
Figure 5-4. Connecticut Department of Energy and Environmental
Protection (CT DEEP) developed ecologically meaningful thresholds for
three major thermal classes (cold, cool, warm). Outliers are shown with
asterisks. Temperature in these three classes differ most in the summer (figure
provided by Mike Beauchene, CT DEEP).
The hydrologic data being collected at RMN sites can be used for similar types of
analyses and applications. For example, Maine used biological and hydrologic data to develop
statewide environmental flow and lake level standards. The standards are based on thresholds
derived from principles of natural flow variation necessary to protect aquatic life and maintain
important hydrological processes (MDEP, 2007; see Figure 5-5; Ricupero, 2009). Other states
are also exploring the development of flow criteria, utilizing the ELOHA framework (Poffet al.,
2010). TNC and several partners (states, RBCs, other federal agencies) have used components of
the ELOHA framework that consider flow needs for sensitive species and key ecosystem
5-5
-------
processes to develop flow recommendations for some eastern and midwestern rivers (e.g., the
Susquehanna, the Upper Ohio, the Delaware, and the Potomac Rivers; DePhilip and Moberg,
2010; Cummins et al., 2010; DePhilip and Moberg, 2013a, 2013b; Buchanan et al., 2013). Data
from RMN sites can be used in similar ways to improve our understanding of these processes
and to help develop regionally informed standards and management strategies.
Salmon Lifecycle Calendar
Regional Yearly Flow Cycle
Fall Early Winter Winter Spring Early Summer Summer
Streamflow Cycle
— Median Yearly Hydrograiyi
— Ch 587 Ftaw Prescriptions
OS. Dec. FeE AprJuK
Ch. 587 mmlian median median median median median
flovg HQH tjjw BOW flow DflW
Salmon Calendar Fry seek shelter, smolt migrate to sea
Eggs overwinter in stream
Salmon migrate to spawning grounds
Salmon migrate to spawning grounds
Figure 4. Salmon life cycle plotted in relation to yearly flow cycle (Ricupero
2009).
Figure 5-5. Salmon life cycle plotted in relation to yearly flow cycle
(Ricupero, 2009).
At many RMN sites, the year-round thermal and hydrologic regimes are poorly
documented, so the first several years of continuous thermal and hydrologic data will be used to
start characterizing these regimes. The continuous data will provide robust data sets that capture
natural temporal patterns, episodic events and spatial variability, which may be missed by
limited numbers of discrete measurements (see Figure 5-6). The collection of these types of data
will also build capacity of biomonitoring programs that have limited experience with continuous
sensors and management of continuous data.
5-6
-------
350
- 2013
O Discrete
^ Continuous
2014
HI VIS 1/13 3,'I3
9,1* 10/t IC^Il 1VS 11/10 IOT U'17»VJ1
Figure 5-6. RMN data will help us gain a better understanding of natural
variability in hydrologic conditions in small least disturbed streams, and will
allow us to investigate relationships between biological, thermal, and
hydrologic conditions.
The continuous RMN data could also be used to help further the development of
ecologically relevant classifications of thermal and hydrologic regimes. For example, Maheu
et al. (2015) recently developed a thermal classification scheme for streams in the conterminous
United States based on magnitude and variability (amplitude and timing), with six classes: highly
variable cool, variable cold, variable cool, variable warm, stable cool, and stable cold. In another
study, Dhungel (2014) developed an empirical model and hydrologic classification scheme for
streams in the conterminous United States based on magnitude, timing, predictability, and
intermittency of flows, with eight classes: small steady perennial, large steady perennial, steady
intermittent, early intermittent, late intermittent, early flashy perennial, small flashy perennial,
and large flashy perennial. Efforts are also underway to develop a hierarchical classification for
natural flowing stream and river systems in the Appalachians (Olivero et al. 2015). These
existing models could be used to classify RMN sites based on thermal and hydrologic data.
Once the RMN sites have been classified, the continuous thermal and hydrologic data can
be used to help characterize baseline or reference conditions for the different stream classes.
Characterization (and quantification) of these regimes is important because some water quality
standards are based on comparisons with natural conditions. Thermal and hydrologic data from
stream segments altered by anthropogenic activities such as dams and land use changes could
potentially be compared to data from high quality RMN sites. Another potential application of
the RMN data is to improve or validate models and simulations that predict stream temperature
and flow. Several stream temperature models now exist for the conterminous United States,
including empirical models developed by Hill et al. (2013) and Segura et al. (2015).
5-7
-------
Advancements in flow models, such as the Variable Infiltration Capacity model (Liang et al.,
1994), continue to be made as well. These models could potentially be applied to RMN sites, and
the simulated stream temperatures and flows could then be compared to the observed values. If
the models provide good approximations, the models could potentially be used to predict climate
change effects on thermal and hydrologic regimes at RMN sites.
Another way in which RMN data can be used is to assess the "thermal sensitivity" of
each site. The "thermal sensitivity" of a stream is often quantified as the slope of the regression
line between air and stream temperature (Kelleher et al. 2012). Air temperature, which is
projected to increase due to climate change (Melillo et al. 2014), is known to be an important
predictor of water temperature (Kelleher et al., 2012; Hill et al., 2013). The relationship between
air and water temperature, however, varies depending on numerous modifying factors, such as
geographic location, stream size, and groundwater contributions (Kelleher et al., 2012; Hill et al.,
2013). Hildebrand et al. (2014) used the paired continuous air-water temperature data from MD
DNR's sentinel sites to explore thermal sensitivities in different regions of Maryland. They
found baseflow and riparian shading to be important modifying factors, along with discharge. In
Pennsylvania, Kelleher et al. (2012) found that stream size and groundwater contribution were
the primary controls of the sensitivity of stream temperature to air temperature. Relationships
between the air and stream temperature data can also be used to characterize groundwater
influence at local scales (Kanno et al., 2014; Snyder et al., 2015). The first several years of
paired air-stream temperature data can be used to assess the sensitivity of each RMN site to
rising air temperatures, and to gain a better understanding of factors that make some sites more
vulnerable to climate change than others. This information could be very useful for management
and conservation planning.
5.2. APPLICATIONS IN A 5-10 YEAR TIMEFRAME
Over the 5-10 year time period, RMN data will continue to be used to characterize
current conditions against which future climate influences can be assessed, to support CWA
programs in ways similar to those described in Section 5.1 and to evaluate and refine the lists of
recommended metrics and indicators provided in Section 4 and Appendices H, J, and K. In
addition, a variety of analyses can be performed to look for trends and patterns in the biological,
thermal, hydrologic, habitat, and water chemistry data. Scatterplots, simple correlation and
regression analyses, analysis of variance, NMDS ordinations, and other analytical tools will be
used to explore differences or trends over time, as well as relationships between the different
types of data. The analyses will be similar to those described in MD DNR's SSN report (Becker
et al., 2010) and the 2012 EPA pilot study in which long-term state biomonitoring data in Maine,
North Carolina, Ohio, and Utah were evaluated for climate-related trends (U.S. EPA, 2012a). In
both studies, there were examples of shifts in biological indicators occurring in association with
changing thermal or hydrologic conditions. In the Maryland study, the lowest Index of
Biological Integrity scores in the Coastal Plain—western shore region were recorded the year
after the lowest flow and rainfall conditions occurred (Becker et al., 2010), and in the EPA pilot
-------
study, a strong decline in EPT richness corresponded with a period of higher than normal
temperatures and lower than normal flows at one of the Utah sites (U.S. EPA, 2012a). RMN data
for the 5-10 year time period can also be used to assess temporal (year to year) variability, which
is not well documented at high quality sites (Milner et al., 2006). The RMN data facilitates a
better understanding of how natural variability affects the consistency of biological condition
scores and metrics from year to year, and how this relates to changing thermal and hydrologic
conditions.10
5.3. APPLICATIONS IN A 10+ YEAR TIMEFRAME
For the 10+ year timeframe, the data will continue to be used to characterize conditions
and temporal variability, support CWA programs, evaluate and refine the lists of recommended
metrics and indicators and perform trend analyses. In addition, climate change effects may start
to become evident. A number of climate projections are relevant to aquatic life condition,
including increasing temperatures, increasing frequency and magnitude of extreme precipitation
events, and increasing frequency of summer low flow events (Melillo et al. 2014). The long-term
data from high quality RMN sites will substantially enhance the ability to detect and characterize
trends attributable to climate change.
Many organizations are performing vulnerability assessments and developing hypotheses
about which organisms, community types, watersheds or stream classes are likely to be most
vulnerable to climate change. For example, the EPA and partners are conducting a broad-scale
climate change vulnerability assessment on streams in the eastern United States. They are
assigning vulnerability ratings to each watershed11 based on a scenario in which stream
temperatures warm and the frequency and duration of summer low flow events increases (see
Figure 5-7). The RMN data can be used to help test these types of hypotheses and to track
whether certain types of streams are showing greater resiliency to climate change effects than
others. This type of information can inform adaptation strategies and conservation planning.
10In order to be able to attribute variability in the biological data to 'natural' vs. other factors, anthropogenic factors
(such as changes in land use) also need to be tracked at each site over time.
"Watershed delineations are based on the NHDPlus v2 local catchment layer: http://www.horizon-
systems.com/NHDPlus/NHDPlusV2_data.php.
5-9
-------
• • ^ :\ -f-si^1
iVVE-•! i .-w W
. " -^&t
• Primary RMN sites (3/25/2014)
Scenario 1
Vulnerability rating
| least
moderate
^^| most
NA
Figure 5-7. EPA and partners are conducting a broad-scale climate change
vulnerability assessment on streams in the eastern United States, based on a
scenario in which stream temperatures warm and the frequency and
duration of summer low flow events increases. Vulnerability ratings (least,
moderate or most) are being assigned to each watershed.
The RMN data can also be used to track whether shifts in the distributions of biological
indicators (such as cold water taxa) are occurring as the climate changes, which would have
implications for bioassessment programs. In some locations, species distribution models (SDMs)
have been developed. For example, Zheng et al.12 (2014) generated models to predict how
species occurrence will change by mid-century in the Northeast under conditions of rising air
temperatures and changing precipitation patterns.13 Results suggest an overall decline in species
richness across much of the region (see Figure 5-8). SDM models have been generated for other
regions as well. For example, Hawkins et al. (2013) used biomonitoring data from the EPA's
2008-2009 NRSA to develop SDMs that predict how the distributions of individual
macroinvertebrate taxa and entire assemblages of taxa vary with stream temperature, flow, and
12Zheng, L., Stamp, I, Hamilton, A., Bierwagen, B. and J. Witt. 2014. Species Distribution Modeling in the
Northeast US - Impact of Climate Change and Taxa Vulnerability. Poster. Joint Aquatic Sciences Meeting. Portland,
OR.
13Mid-century (2040-2069) projections for air temperature, precipitation and moisture surplus were based on
average values from an ensemble of 15 GCMs, using the a2 (high) emissions scenario. Data were obtained from the
Climate Wizard website and are based on the WCPJ3 CMIP3 multimodel data set.
5-10
-------
other watershed attributes in the conterminous United States, for baseline (2000-2010) versus
late century (2090-2100) time periods. SDMs are also being developed for stonefly species in
the Midwest (Cao et al., 2013; DeWalt et al., 2013). If applied to RMN sites, SDMs could serve
as valuable tools for conservation planning.
Change in species richness (mid-century minus baseline)
Loss of 15 or more taxa
Loss of 10 to 14
Loss of 5 to 9
Loss of 1 to 4
-0.9 to 0.9
Gain of 1 or more taxa
Figure 5-8. Modeling results predict declines in species richness across much
of the Northeast by mid-century (2040-2069).14
Models have also been developed to predict climate change effects on stream temperature
and flow. For example, at the national scale, Hill et al. (2013) developed a stream temperature
model for the conterminous United States based on air temperature and watershed feature data
(e.g., watershed area and slope) from reference-condition USGS sampling sites, and applied the
model to simulate the effects of climate change on mean summer stream temperature (Hill et al.
2014). The model predicts a mean warming of 2.2°C for stream temperatures in the
14Zheng, L., Stamp, I, Hamilton, A., Bierwagen, B. and J. Witt. 2014. Species Distribution Modeling in the
Northeast US—Impact of Climate Change and Taxa Vulnerability. Poster. Joint Aquatic Sciences Meeting.
Portland, OR.
5-11
-------
conterminous United States by late century (2090-2099) relative to a 2001-2010 baseline period,
with values at individual sites ranging from 0°C to +6.2°C. In another study, Dunghel (2014)
developed statistical models to predict flow responses to projected changes in precipitation and
temperature. Results suggest that changes in flow attributes will be most evident in rain-fed
small perennial streams and intermittent streams in the central and eastern United States. These
models could potentially be applied to RMN sites, and the performance of the models could be
tracked over time.
RMN data may also provide insights on how organisms respond to and recover from
extreme weather events such as droughts and floods, which are projected to occur with greater
frequency in the future (Melillo et al., 2014). Impacts can be evaluated through comparative
analyses on the pre- and postevent data. For example, VT DEC performed these types of
analyses on macroinvertebrate data collected before and after flooding from Tropical Storm
Irene. Using data from 10 high-quality sites, VT DEC documented immediate decreases in
invertebrate densities of 69% on average, but also found that most sites recovered to normal
levels the following year (see Figure 5-9).15 The substantial decline in density and the rapid
recovery would have been missed if sampling had occurred at longer intervals, such as on a
5-year rotational sampling schedule. Moreover, the collection of continuous thermal and
hydrologic data will allow the magnitude, frequency and duration of the event to be documented.
Whether or not the RMN data can fully capture biological responses to extreme weather events
will depend on the timing of the event in relation to the RMN sampling period.
I
I
MOO
4000
3000
2000
1000
D
~ Pre-flood
• 1 month post-flood
D 1 year post-flood
Small Streams Medium Streams Large Streams
Figure 5-9. Comparison of macroinvertebrate density values at 10 stream
sites in Vermont before and after Tropical Storm Irene.16
15Moore, A. and S. Fiske. (2012) What's Left in Vermont's Streams after Tropical Storm Irene—Monitoring Results
From Long Term Reference Sites. Presentation. New England Association of Environmental Biologists Annual
Conference. Falmouth, MA.
16Ibid.
5-12
-------
6. DATA MANAGEMENT
The EPA and partners are developing a data management system that will allow
participating organizations and outside users to access data and metadata that are being collected
at RMN sites. The biological, habitat, and water chemistry data will be uploaded to a national
water data system such as the WQX. At this time not all RMN partners have the capacity to
upload these data into WQX, so the EPA and partners are working on an interim solution. Until
the interim system is in place, the individual organizations will be custodians and owners of
these data, and all data files will be backed up and stored in a centralized, secure location.
Because WQX cannot accommodate continuous data, the thermal and hydrologic data being
collected by RMN partners will be uploaded into a separate data management system. A
multiagency effort is underway to develop a data management system for the continuous data. In
the interim, EPA and partners are developing guidance for storing and managing the continuous
RMN data files. This includes the development of R scripts (R Core Team, 2015) and procedures
for performing QC on the data and for deriving standardized sets of summary outputs for desired
time periods. Until the permanent data management system is in place, the individual
organizations will be the custodians and owners of the continuous thermal and hydrologic
monitoring data. The EPA and partners are also developing procedures for joining the different
sets of data and facilitating the types of data analyses described in Section 5.
6-1
-------
[This page intentionally left blank.]
-------
7. IMPLEMENTATION AND NEXT STEPS
Implementation of the RMNs is underway in the Northeast, Mid-Atlantic, and Southeast
regions. Sampling efforts began in the Northeast in 2012, followed by the Southeast in 2013 and
the Mid-Atlantic in 2014. Currently there are 25 primary RMN sites in the Northeast, 27 in the
Mid-Atlantic and 38 in the Southeast (see Appendix C). More sites will be added in all regions as
resources permit. Efforts are also underway to develop RMNs in the Midwest, where sampling is
expected to start in 2016 (see Figure 7-1). A number of organizations in the western United
States have also expressed interest in setting up RMNs.
• Primary RMN sites (4/2/2014)
_ Sampling underway
Development in progress
Figure 7-1. Sampling has been underway at the Northeast, Mid-Atlantic and
Southeast RMNs for several years. RMNs are currently being developed in the
Midwest.
Efforts are being made to integrate RMN data collection flexibly within existing
monitoring programs to maximize available resources. Some RMN partners have taken a phased
approach to implementation, particularly with collecting hydrologic data at ungaged sites. For
example, some participating organizations have installed pressure transducers at one or two sites,
and are planning to add more as they gain experience and as resources permit. In many states, the
collection of macroinvertebrate data and year-round temperature measurements have been
feasible during the first year of implementation. However, it has been difficult for some
participating organizations to find resources to process the macroinvertebrates samples in
accordance with the regional methods. When this occurs, the preserved samples are being
retained so that they can be processed at a later time, when resources become available.
7-1
-------
To help build capacity and improve data quality, EPA and partners have held several
training workshops on field protocols for continuous temperature sensors and pressure
transducers, and wrote the EPA best practices document (U.S. EPA, 2014). In addition, the
regional coordinator in the Mid-Atlantic is planning to hold a training workshop on species-level
identifications for high priority taxa (see Appendix G). There are also plans to hold hands-on,
interactive workshops on data management at annual meetings (e.g., the New England
Association of Environmental Biologists conference, the Association of Mid-Atlantic Aquatic
Biologists Workshop, the Southeastern Water Pollution Biologists Association conference).
As discussed in Section 3, EPA has also developed a QAPP for the RMNs to address the
challenges of creating regionally consistent data sets. It is a generic QAPP that details the core
requirements for participation in the network, and outlines best practices for the collection of
biological, thermal, hydrologic, physical habitat, and water chemistry data at RMN sites. The
QAPP was written in a way that should be transferable to other regions, with region-specific
protocols included as addendums. The QAPP is intended to increase the comparability of data
being collected at RMN sites, improve the ability to detect long-term trends by minimizing
biases and variability, and to ensure that the data are of sufficient quality to meet data quality
objectives. Efforts will be made to finalize the QAPPs for the Northeast, Mid-Atlantic, Southeast
and Midwestern regions by 2016. The protocols in the QAPP will be reevaluated periodically
and updated as needed. EPA and partners are also exploring the possibility of conducting
methods comparison studies in regions where macroinvertebrate collection and processing
protocols differ across participating organizations. This could involve the collection of
side-by-side samples using the different methods. After the paired samples are processed using
the respective methods, results would be compared and differences between the methods could
be quantified.
In coming years, EPA and partners will continue to build capacity and refine protocols,
indicator lists, analytical techniques and data management systems for the Northeast,
Mid-Atlantic, and Southeast RMNs. Other regions have expressed interest in establishing RMNs
as well. They recognize how these types of long-term data can support CWA programs, fill data
gaps, and help detect trends attributable to climate change. The RMN framework is flexible and
allows for expansion to new regions, as well as to new stream classes and waterbody types. For
example, some parts of the Midwest lack the higher gradient, riffle-dominated cold water
streams that are being targeted in the Northeast, Mid-Atlantic, and Southeast RMN pilot studies.
Instead, these regions may focus their sampling efforts on low gradient, sandy-bottom, warm
water streams. A Midwestern working group has also been formed to explore the possibility of
setting up a RMN for inland lakes. The monitoring data being collected from these regional
efforts will provide important inputs for bioassessment programs as they strive to protect water
quality and aquatic ecosystems under a changing climate.
7-2
-------
8. LITERATURE CITED
Barbour, MT; Stribling, JB; Karr, JR. (1995) The multimetric approach for establishing biocriteria and measuring
biological condition, [p. 63-76]. In Davis, WS; Simon, TP (eds). Biological assessment and criteria: tools
for water resource planning and decision making. Boca Raton, FL: CRC Press.
Barbour, MT; Gerritsen, J; Snyder, BD; Stribling, JB. (1999) Rapid bioassessment protocols for use in streams and
wadeable rivers: Periphyton, benthic macroinvertebrates and fish, Second Edition. [EPA 841-B-99-002].
Washington, D.C: U.S. Environmental Protection Agency, Office of Water. Available online:
http://water.epa.gov/scitech/monitoring/rsl/bioassessment/index.cfm.
Beauchene, M; Becker, M; Bellucci, C; Hagstrom, N; Kanno, Y. (2014) Summer thermal thresholds offish
community transitions in Connecticut streams. J N Am Fish Manage 34(1):119-131.
Becker, AJ; Stranko, SA; Klauda, RJ; Prochaska, AP; Schuster, JD; Kashiwagi, MR; Graves, PH. (2010) Maryland
biological stream survey's sentinel site network: A multi-purpose monitoring program. Annapolis, MD:
Maryland Department of Natural Resources.
Bilhimer, D; Stohr, A. (2009) Standard Operating Procedures for continuous temperature monitoring of fresh water
rivers and streams conducted in a Total Maximum Daily Load (TMDL) project for stream temperature,
Version 2.3. Olympia, WA: Washington State Department of Ecology, Environmental Assessment
Program. Available online:
http://www.ecy.wa.gov/programs/eap/qa/docs/ECY_EAP_SOP_Cont_Temp_Monit_TMDL_v3_OEAP044.
pdf.
Blocksom, KA; Autrey, BC; Passmore, M; Reynolds, L. (2008) A comparison of single and multiple habitat
protocols for collecting macroinvertebrates in wadeable streams. JAWRA 44(3):577-593.
Blum, MD; Tornqvist, TE. (2000) Fluvial responses to climate and sea-level change: a review and look forward.
Sedimentology 47:2-48.
Bohmer, J; Rawer-Jost, C; Zenker, A. (2004) Multimetric assessment of data provided by water managers from
Germany: assessment of several different types of stressors with macrozoobenthos communities.
Hydrobiologia 516(l-2):215-228.
Buchanan, C; Moltz, HLN; Haywood, HC; Palmer, JB; Griggs, AN. (2013) A test of The Ecological Limits of
Hydrologic Alteration (ELOHA) method for determining environmental flows in the Potomac River basin,
U.S.A. Freshw Biol 58(12):2632-2647.
Butryn, RS; Parrish, DL; Rizzo, DM. (2013) Summer stream temperature metrics for predicting brook trout
(Salvelinus fontinalis) distribution in streams. Hydrobiologia 703:47-57.
Cao, Y; Williams, WP; Bark, AW. (1997) Effects of sample size (number of replicates) on similarity measures in
river benthic Aufwuchs community analysis. Water Environ Res 69(1): 107-114.
Cao, Y; Hawkins, CP; Storey, AW. (2005) A method for measuring the comparability of different sampling methods
used in biological surveys: Implications for data integration and synthesis. Freshw Biol 50(6): 1105-1115.
Cao, Y; DeWalt, RE; Robinson, JL; Tweddale, T; Hinz, L; Pessino, M. (2013) Using Maxent to model the historic
distributions of stonefly species in Illinois streams: The effects of regularization and threshold selections.
Ecol Model 259:30-39.
Carlisle, DM; Falcone, J; Wolock, DM; Meador, MR; Norris, HR. (2010) Predicting the natural flow regime:
models for assessing hydrological alteration in streams. River Res Appl 26(2): 118-36.
-------
Carlisle, DM; Wolock, DM; Meador, MR. (2011) Alteration of streamflow magnitudes and potential ecological
consequences: a multiregional assessment. Front Ecol Environ 9:264-270.
Carter, JL; Resh, VH. (2013) Analytical approaches used in stream benthic macroinvertebrate biomonitoring
programs of state agencies in the United States. [USGS Open-File Report 2013-1129]. Reston, VA: U.S.
Geological Survey, http://pubs.usgs.gov/of/2013/1129/.
Chapman, AD. (2005) Principles of data quality. Version 1.0. Report for the Global Biodiversity Information
Facility (GBIF). Global Biodiversity Information Facility, Copenhagen, Denmark: Global Biodiversity
Information Facility. Available online at http://www2.gbif. org/DataQuality.pdf.
Chase, R. (2005). Standard Operating Procedure: Streamflow measurement. [CN 68.0]. Massachusetts Department
of Environmental Protection, Division of Watershed Management.
Chinnayakanahalli, KJ; Hawkins, CP; Tarboton, DG; Hill, RA. (2011) Natural flow regime, temperature and the
composition and richness of invertebrate assemblages in streams of the western United States. Freshw Biol
56:1248-1265. doi: 10.1111/J.1365-2427.2010.02560.X.
Chiu, MC; Kuo, MH. (2012) Application of r/K selection to macroinvertebrate responses to extreme floods. Ecol
Entomol 37(2): 145-154.
Clark, ME; Rose, KA; Levine, DA; Hargrove, WW. (2001) Predicting climate change effects on Appalachian trout:
combining GIS and individual-based modeling. Ecol Appl 11(1):161-178.
Conti, L; Schmidt-Kloiber, A; Grenouillet, G; Graf, W. (2014) A trait-based approach to assess the vulnerability of
European aquatic insects to climate change. Hydrobiologia 721(1): 297-315.
Cuffney, TF; Bilger, MD; Haigler, AM. (2007) Ambiguous taxa: effects on the characterization and interpretation of
invertebrate assemblages. J N Am Benthol Soc 26(2):286-307.
Cummins, J; Buchanan, C; Haywood, HC; Moltz, H; Griggs, A; Jones, C; Kraus, R; Hitt, NP; Bumgardner, R.
(2010) Potomac large river ecologically sustainable water management report. [ICPRB Report 10-3].
Interstate Commission on the Potomac River Basin for The Nature Conservancy. Available online at
www.potomacriver.org/pubs.
Davies, SP; Jackson, SK. (2006) The Biological Condition Gradient: A descriptive model for interpreting change in
aquatic ecosystems. Ecol Appl 16(4): 1251-1266.
DePhilip, M; Moberg, T. (2010) Ecosystem flow recommendations for the Susquehanna River Basin. Harrisburg,
PA: The Nature Conservancy. Available online at
http://www.srbc.net/policies/docs/TNCFinalSusquehannaRiverEcosy stemFlowsStudyReport_NovlO_2012
0327_fsl35148vl.PDF.
DePhilip, M; Moberg, T. (2013a) Ecosystem flow recommendations for the Delaware River basin. Harrisburg, PA:
The Nature Conservancy.
DePhilip, M; Moberg, T. (2013b) Ecosystem flow recommendations for the Upper Ohio River basin in western
Pennsylvania. Harrisburg, PA: The Nature Conservancy.
DeShon, J. (1995) Development and application of the Invertebrate Community Index (ICI). [p. 217-243]. In Davis,
WS; Simon, TP (eds). Biological assessment and criteria: tools for water resource planning and decision
making. Boca Raton: CRC Press.
DeWalt, RE; Cao, Y; Robinson, JL; Tweddale, T; Hinz, L. (2013) Predicting climate change effects on riverine
stonefly (Insecta: Plecoptera) species in the Upper Midwest. INKS Technical Report 2013 (22). Prepared
for U.S. Fish and Wildlife Service.
8-2
-------
Dhungel, S. (2014) Prediction of climate change effects on streamflow regime important to stream ecology. All
Graduate Theses and Dissertations: Paper 3083. http://digitalcommons.usu.edu/etd/3083.
Dunham, J; Chandler, G; Rieman, B; Martin, D. (2005) Measuring stream temperature with digital data sensors: a
user's guide. [Gen. Tech. Rep. RMRSGTR-150WWW]. Fort Collins, CO: U.S. Department of Agriculture,
Forest Service, Rocky Mountain Research Station. 15 p. Available online at
http://fresc.usgs.gov/products/papers/143 l_Dunham.pdf.
Durance, I; Ormerod, SJ. (2007) Climate change effects on upland stream macroinvertebrates over a 25-year period.
Glob Change Biol 13:942-957.
Faustini, JM. (2000) Stream channel response to peak flows in a fifth-order mountain watershed. Corvallis, OR:
Oregon State University, Corvallis, OR.
Faustini, JM; Jones, JA. (2003) Influence of large woody debris on channel morphology and dynamics in steep,
boulder-rich mountain streams, western Cascades, Oregon. Geomorphology 51(1-3): 187-205.
Feio, MJ; Reynoldson, TB; Graca, MAS. (2006) The influence of taxonomic level on the performance of a
predictive model for water quality assessment. Can J Fish Aquat Sci 63(2):367-376.
Flebbe, PA; Roghair, LD; Bruggink, JL. (2006) Spatial modeling to project southern Appalachian trout distribution
in a warmer climate. Trans Am Fish Soc 135(5):1371-1382.
Foucreau, N; Piscart, C; Puijalon, S; Hervant, F. (2013) Effect of climate-related change in vegetation on leaf litter
consumption and energy storage by Gammarus pulex from Continental or Mediterranean populations.
PLoS ONE 8(10): e77242.
Gerth, WJ; Herlihy, AT. (2006) The effect of sampling different habitat types in regional macroinvertebrate
bioassessment surveys. J N AmBenthol Soc 25:501-512.
Goode, JR; Luce, CH; Buffington, JM. (2012) Enhanced sediment delivery in a changing climate in semi-arid
mountain basins: Implications for water resource management and aquatic habitat in the northern Rocky
Mountains. Geomorphology 139-140:1-15.
Gotelli, NJ; Graves, GR. (1996) Null models in ecology. Washington, DC: Smithsonian Institution Press.
Goudie, AS. (2006) Global warming and fluvial geomorphology. Geomorphology 79(3-4):384-394.
Haase, P; Murray-Bligh, J; Lohse, S; Pauls, S; Sundermann, A; Gunn, R; Clarke, R. (2006) Assessing the impact of
errors in sorting and identifying macroinvertebrate samples. Hydrobiologia, 566:505-521.
Haase, P; Pauls, SU; Schindehutte, K; Sundermann, A. (2010) First audit of macroinvertebrate samples from an EU
Water Framework Directive monitoring program: human error greatly lowers precision of assessment
results. JN AmBenthol Soc 29(4):1279-1291. 10.1899/09-183.1
http://www.jnabs.0rg/doi/abs/10.1899/09-183.l.
Harding, JS; Benfield, EF; Bolstad, PV; Helfman, GS; Jones III, EBD. (1998) Stream biodiversity: The ghost of
land use past. Proc Nat Acad Sci 95:14843-14847.
Harper, MP; Peckarsky, BL. (2006) Emergence cues of a mayfly in a high-altitude stream ecosystem: Potential
response to climate change. Ecol Appl 16:612-621.
Hawkins, CP. (2006) Quantifying biological integrity by taxonomic completeness: Its utility in regional and global
assessment. Ecol Appl 16(4): 1277-1294.
8-3
-------
Hawkins, CP; Tarboton, DG; Jin, J. (2013) Consequences of global climate change for stream biodiversity and
implications for the application and interpretation of biological indicators of aquatic ecosystem condition.
Final Report. [EPA Agreement Number: RD834186]. Logan Utah: Utah State University.
http://cfpub.epa. gov/ncer_abstracts/index.cfm/fuseaction/display.abstractDetail/abstract/9063.
Herlihy, AT; Sifneos, J; Bason, C; Jacobs, A; Kentula, ME; Fennessy, MS. (2009) An approach for evaluating the
repeatability of rapid wetland assessment methods: The effects of training and experience. Environ Manage
44:369-377.
Hewlett, R. (2000) Implications of taxonomic resolution and sample habitat for stream classification at a broad
geographic scale. J N AmBenthol Soc 19:352-361.
Hildebrand, RH; Kashiwagi, MT; Prochaska, AP. (2014) Regional and local scale modeling of stream temperatures
and spatio-temporal variation in thermal sensitivities. Environ Manage 54:14-22.
Hill, RA; Hawkins, CP; Carlisle, DM. (2013) Predicting thermal reference conditions for USA streams and rivers.
Freshw Sci 32(l):39-55.
Hill, RA; Hawkins, CP; Jin, J. (2014) Predicting thermal vulnerability of stream and river ecosystems to climate
change. Clim Change 125:399-412. doi:10.1007/s!0584-014-1174-4.
Hogg, ID; Williams, DD. (1996) Response of stream invertebrates to a global warming thermal regime: an
ecosystem-level manipulation. Ecology 77:395-407.
Holden, ZA; Klene, A; Keefe, R; Moisen, G. (2013) Design and evaluation of an inexpensive solar radiation shield
for monitoring surface air temperatures. Agric For Meteorol 180:281-286.
Holling, CS. (1973) Resilience and stability of ecological systems. Ann Rev Ecol Systems 4:1-23.
Horrigan, N; Baird, DJ. (2008) Trait patterns of aquatic insects across gradients of flow-related factors: a
multivariate analysis of Canadian national data. Can J Fish Aquat Sci 65: 670-680.
Huh, S; Dickey, DA; Meador, MR; Rull, KE. (2005) Temporal Analysis of the frequency and duration of low and
high streamflow: years of record needed to characterize streamflow variability. J Hydrol 310:78-94.
Isaak, DJ; Horan, DL. (2011) An assessment of underwater epoxies for permanently installing temperature in
mountain streams. N Am J Fish Manage 31:134-137. Available online:
http://www.treesearch.fs.fed.us/pubs/37476.
Isaak, DJ; Wollrab, S; Horan, D; Chandler, G. (2012) Climate change effects on stream and river temperatures
across the Northwest U.S. from 1980-2009 and implications for salmonid fishes. Clim Change
113:499-524.
Isaak, DJ; Rieman, BE. (2013) Stream isotherm shifts from climate change and implications for distributions of
ectothermic organisms. Glob Change Biol 19:742-751.
Iverson, L; Prasad, A; Matthews, S. (2008) Potential changes in suitable habitat for 134 tree species in the
Northeastern United States. Mitig Adapt Strat Glob Change 13:517-540.
http://treesearch.fs.fed.us/pubs/15295.
Jackson, JK; Fureder, L. (2006) Long-term studies of freshwater macroinvertebrates—A review of the frequency,
duration, and ecological significance. Freshw Biol 51:591-603.
Jones, NE; Schmidt, B. (2012) ThermoStat version 3.0: Tools for analyzing thermal regimes. Ontario Ministry of
Natural Resources, Aquatic Research and Development Section.
http://people.trentu.ca/nicholasjones/thermostat.htm.
8-4
-------
Kanno, Y; Vokoun, JC; Letcher, BH. (2014) Paired stream-air temperature measurements reveal fine-scale thermal
heterogeneity within headwater brook trout stream networks. River Res Applic 30:745-755.
Karr, JR. (1991) Biological integrity: A long-neglected aspect of water resource management. Ecol Appl 1:66-84.
Kaufmann, PR; Levine, P; Robinson, EG; Seeliger, C; Peck, DV. (1999) Quantifying physical habitat in wadeable
streams. [EPA/620/R-99/003]. Washington, DC: U.S. Environmental Protection Agency.
http://www.epa.gov/emap/html/pubs/docs/groupdocs/surfwatr/field/phyhab.html.
Kaufmann, PR; Faustini, JM; Larsen, DP; Shirazi, MA. (2008) A roughness-corrected index of relative bed stability
for regional stream surveys. Geomorphology 99(1-4): 150-170.
Kaufmann, PR; Larsen, DP; Faustini, JM. (2009) Bed stability and sedimentation associated with human
disturbances in Pacific Northwest streams. J Am Water Resour Assoc 45:434-459.
Kelleher, C; Wagener, T; Gooseff, M; McGlynn, B; McGuire, K; Marshall, L. (2012) Investigating controls on the
thermal sensitivity of Pennsylvania streams. Hydrological Processes 26(5):771-785.
Kennen, JG; Kauffman, LJ; Ayers, MA; Wolock, DM; Colarull, SJ. (2008) Use of an integrated flow model to
estimate ecologically relevant hydrologic characteristics at stream biomonitoring sites. Ecol Modell
211:57-76.
Kennen, JG; Sullivan, DJ; May, JT; Bell, AH; Beaulieu, KM; Rice, DE. (2011) Temporal changes in aquatic-
invertebrate and fish assemblages in streams of the north-central and northeastern U.S. Ecol Indie
18:312-329.
Larsen, D; Kaufmann, P; Kincaid, T; Urquhart, N. (2004) Detecting persistent change in the habitat of salmon-
bearing streams in the Pacific Northwest. Can J Fish Aqua Sci 61:283-291.
Lenat, DR; Resh, VH. (2001) Taxonomy and stream ecology—The benefits of genus- and species-level
identifications. J N Am Benthol Soc 20:287-298.
Liang, X; Lettenmaier, DP; Wood, EF; Burges, SJ. (1994) A simple hydrologically based model of land surface
water and energy fluxes for GSMs. J Geophys Res 99(D7): 14,415-14,428.
Maheu, A; Poff, NL; St-Hilaire, A. (2015) A classification of stream water temperature regimes in the conterminous
USA. River Research and Applications: doi: 10.1002/rra.2906.
MDEP (Maine Department of Environmental Protection). (2007) Chapter 587: In-stream flows and lake and pond
water levels. In Sustainable water use program. Augusta, ME: Maine DEP.
http://www.maine.gov/sos/cec/rules/06/096/096c587.doc.
Mazor, RD; Purcell, AH; Resh, VH. (2009) Long-term variability inbenthic macroinvertebrate bioassessments: A
20-year study from two northern Californian streams. Environ Manage 43:1269-1286.
McCormick, PV; Stevenson, RJ. (1998) Periphyton as a tool for ecological assessment and management in the
Florida Everglades. J Phycol 34(5):726-733.
McMahon, G; Bales, JD; Coles, JF; Giddings, BMP; Zappia, H. (2003) Use of stage data to characterize hydrologic
conditions in an urbanizing environment. J Am Water Resources Assoc 39:1529-1536.
Melillo, JM; Richmond, TC; Yohe, GW. (2014) Climate change impacts in the United States: the third National
Climate Assessment. US Global Change Research Program 841.
Meyer, JL; Strayer, DL; Wallace, JB; Eggert, SL; Helfman, GS; Leonard, ME. (2007) The contribution of headwater
streams to biodiversity in river networks. J Am Water Resour Assoc 43:86-103.
3-5
-------
Milner, AM; Conn, SC; Brown, LE. (2006) Persistence and stability of macroinvertebrate communities in streams of
Denali National Park, Alaska: implications for biological monitoring. Freshw Biol 51:373-387.
Nelitz, MA; Maclsaac, EA; Peterman, RM. (2007) A science-based approach for identifying temperature-sensitive
streams for rainbow trout. N Am J Fish Manage 27:405-424.
Norris, RH; Barbour, MT. (2009) Bioassessment of aquatic ecosystems. In: Likens, GE (Ed) Encyclopedia of inland
waters, volume 3. Oxford: Elsevier. pp. 21-28.
NWQMC (National Water Quality Monitoring Council). (2011) Establishing a collaborative and multipurpose
national network of reference watersheds and monitoring sites for freshwater streams in the United States.
U.S. Department of the Interior, U.S. Geological Survey.
http://acwi.gov/monitoring/workgroups/wis/National_Reference_Network_for_Streams.pdf.
Olden, JD; Poff, ML. (2003) Redundancy and the choice of hydrologic indices for characterizing streamflow
regimes. River Res Appl 19:101-121.
Olivero SA; Barnett, A; Anderson, MG. (2015) A Stream Classification for the Appalachian Region. The Nature
Conservancy, Eastern Conservation Science, Eastern Regional Office. Boston, MA.
Olsen, AR; Sedransk, J; Edwards, D; Gotway, CA; Liggett, W; Rathbun, S; Reckhow. KH; Young, LJ. (1999)
Statistical issues for monitoring ecological and natural resources in the United States. Environ Monitor
Assess 54:1-54.
Ostermiller, JD; Hawkins, CP. (2004) Effects of sampling error on bioassessments of stream ecosystems: application
to RIVPACS-type models. J N AmBenthol Soc 23:363-382.
Parson, M; Norris, RH. (1996) The effect of habitat specific sampling on biological assessment of water quality
using a predictive model. Freshw Biol 36:419-434.
Poff, NL; Allan, JD; Bain, MB; Karr, JR; Prestegaard, KL; Richter, BD; Sparks, RE; Stromberg, JC. (1997) The
natural flow regime: a paradigm for river conservation and restoration. BioScience 47(11):769-784.
Poff, NL; Olden, JD; Vieira, NKM; Finn, DS; Simmons, MP; Kondratieff, BC. (2006) Functional trait niches of
North American lotic insects: traits-based ecological applications in light of phylogenetic relationships. J N
AmBenthol Soc 25(4):730-755.
Poff, NL; Pyne, MI; Bledsoe, BP; Cuhaciyan, CC; Carlisle, DM. (2010) Developing linkages between species traits
and multiscaled environmental variation to explore vulnerability of stream benthic communities to climate
change. J N AmBenthol Soc 29(4): 1441-1458.
Pond, GJ; North, SH. (2013) Application of a benthic observed/expected-type model for assessing Central
Appalachian streams influenced by regional stressors in West Virginia and Kentucky. Environ Monit
Assess 185(11):9299-9320.
Rantz, SE; et al. (1982) Measurement and computation of streamflow, Volume I: Measurement of stage and
discharge and Volume II: Computation of discharge. [USGS Water Supply Paper 2175]. U.S. Department
of the Interior, U.S. Geological Survey. Available online at http://water.usgs.gov/pubs/wsp/wsp2175/.
R Core Team (2015) R: A language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria. Available online at http://www.R-project.org/.
Rehn, AC; Ode, PR; Hawkins, CP. (2007) Comparisons of targeted-riffle and reach-wide benthic macroinvertebrate
samples: implications for data sharing in stream-condition assessments. J N Am Benthol Soc
26(2):332-348.
Richards, C; Haro, RJ; Johnson, LB; Host, GE. (1997) Catchment and reach-scale properties as indicators of
macroinvertebrate species traits. Freshw Biol 37:219-230.
-------
Richter, BD; Baumgartner, J; Wigington, R; Braun, D. (1997) How much water does a river need? Freshw Biol
37:231-249.
Ricupero, K. (2009) Changing streamflow patterns in the New England region: Implications for ecosystem services,
water users and sustainable resources management. Master of Science thesis. Department of Civil and
Environmental Engineering, University of Maine, Orono, 212 pages.
Roy, AH; Rosemond, AD; Leigh, DS; Paul, MJ; Wallace, JB. (2003) Habitat-specific responses of
macroinvertebrates to land cover disturbance: biological consequences and monitoring implications. J N
AmBenthol Soc 22:292-307.
Rustad, L; Campbell, J; Dukes, JS; Huntington, T; Lambert, KF; Mohan, J; Rodenhouse, N. (2012) Changing
climate, changing forests: The impacts of climate change on forests of the Northeastern United States and
Eastern Canada. [General Technical Report NRS-99]. Newtown Square, PA: United States Department of
Agriculture, US Forest Service, Northern Research Station. Available online at
http://nrs.fs.fed.us/pubs/41165.
Sandin, L; Johnson, RK. (2000) The statistical power of selected indicator metrics using macroinvertebrates for
assessing acidification and eutrophication of running waters. Hydrobiologia 422/423:233-243.
Scarsbrook, MR. (2002) Persistence and stability of lotic invertebrate communities in New Zealand. Freshw Biol
47:417-431.
Segura, C; Caldwell, P; Sun, G; McNulty, S; Zhang, Y. (2015) A model to predict stream water temperature across
the conterminous USA. Hydrol Process 29(9):2178-2195.
Shedd, JR. (2011) Standard operating procedure for measuring and calculating stream discharge. Washington State
Department of Ecology, Environmental Assessment Program. Available online:
http://www.ecy.wa.gov/programs/eap/quality.html.
Snyder, CD; Hitt, NP; Young, JA. (2015) Accounting for groundwater in stream fish thermal habitat responses to
climate change. Ecological Applications 25(5):139--1419.
Sowder, C; Steel, EA. (2012) A note on the collection and cleaning of water temperature data. Water 4:597-606.
Stamp, J; Gerritsen, J. (2009) Methods and assessment comparability among state and federal biological monitoring
protocols. Lowell, MA: New England Interstate Water Pollution Control Commission.
Stamp, J; Gerritsen, J; Pong, G; Jackson, SK; Van Ness, K. (2014) Calibration of the Biological Condition Gradient
(BCG) for Fish and Benthic Macroinvertebrate Assemblages in the Northern Piedmont region of Maryland.
Prepared for U.S. EPA Office of Science and Technology and Montgomery County Department of
Environmental Protection.
Statzner, B; Resh, VH; Doledec, S. (1994) Ecology of the upper Rhone river: a test of habitat templet theories.
Freshw Biol 31:253-554.
Stevenson, RJ. (1998) Diatom indicators of stream and wetland stressors in a risk management framework. Environ
Monit Assess 51(1-2): 107-118.
Stoddard, JL; Larsen, DP; Hawkins, CP; Johnson, RK; Norris, RH. (2006) Setting expectations for the ecological
condition of streams: the concept of reference condition. Ecol Appl 16(4): 1267-1276.
Stribling, JB; Moulton, SR; Lester, GL. (2003) Determining the quality of taxonomic data. J N AmBenthol Soc
22:621-631.
Stribling, JB; Pavlik, KL; Holdsworth, SM; Leppo, EW. (2008) Data quality, performance, and uncertainty in
taxonomic identification for biological assessments. J N Am Benthol Soc 27(4):906-919. Available online
at http://www.jnabs.0rg/doi/abs/10.1899/07-175.l.
8-7
-------
Stuckey, MH; Koerkle, EH; Ulrich, J. (2012) Estimation of baseline daily mean streamflows for ungagged location
on Pennsylvania streams, wateryears 1960-2008. [USGS Scientific Investigations Report 2012-5142]. New
Cumberland, PA: U.S. Geological Survey. Available online at http://pubs.usgs.gov/sir/2012/5142/.
Sweeney, BW; Jackson, JK; Newbold, JD; Funk, DH. (1992) Climate change and the life histories and
biogeography of aquatic insects in eastern North America. [Pages 143-176]. In Firth, P; Fisher, SG (eds).
Global climate change and freshwater ecosystems. New York: Springer-Verlag.
Sweeney, BW. (1993) Effects of streamside vegetation on macroinvertebrate communities of White Clay Creek in
eastern North America. Proc Acad Nat Sci Phil 144: 291-340.
TNC (The Nature Conservancy). (2009) Indicators of hydrologic alteration version 7.1 user's manual. Available
online at
http://www.conservationgateway.org/ConservationPlanning/ToolsData/Tools/CommonlyUsedTools/Pages/
commonly-used-tools.aspx#IHA.
Townsend, CR; Hildrew, AG. (1994) Species traits in relation to a habitat templet for river systems. Freshw Biol
31:265-275.
Townsend, CR; Doledec, S; Scarsbrook, MR. (1997) Species traits in relation to temporal and spatial heterogeneity
in streams: a test of habitat templet theory. Freshw Biol 37:367-387.
Trumbo, BA. (2010) Sensitivity and exposure of brook trout (Salvelinus fontinalis) habitat to climate change.
[Masters Thesis]. Harrisonburg, VA: Department of Biology, James Madison University.
Tsang, Y; Infante, DM; Stewart, J; Wang, L; Tingly III, RW; Thornbrugh, D; Cooper, AR; Daniel, WM.
(submitted). StreamThermal: A software package for calculating thermal metrics from stream temperature
data (submitted to Fisheries Magazine).
U.S. EPA (Environmental Protection Agency). (2006) Southeastern plains in-stream nutrient & biological response
(SPINBR) study quality assurance project plan for state agency activities. [Project Number: 06-0267].
Region 4.
U.S. EPA (Environmental Protection Agency). (2011) Healthy Watersheds Initiative: National Framework and
Action Plan. Publication Number: EPA 841-R-l 1-005.
U.S. EPA (Environmental Protection Agency). (2012a) Implications of climate change for state bioassessment
programs and approaches to account for effects. [EPA/600/R-11/036F]. Washington, DC: Global Change
Research Program, National Center for Environmental Assessment.
http://cfpub.epa. gov/ncea/global/recordisplay.cfm?deid=239585.
U.S. EPA (Environmental Protection Agency). (2012b) Freshwater biological traits database (final report).
[EPA/600/R-11/038F]. Washington, DC: National Center for Environmental Assessment. Available online
athttp://cfpub.epa.gov/ncea/global/recordisplay.cfm?deid=241813
U.S. EPA (Environmental Protection Agency). (2013a) National rivers and streams assessment 2013-2014: Field
operations manual—wadeable. [EPA/841/B-12/007]. Washington, DC: U.S. Environmental Protection
Agency, Office of Water.
U.S. EPA (Environmental Protection Agency). (2013b) A Long-Term Vision for Assessment, Restoration, and
Protection under the Clean Water Act Section 303(d) Program.
http://www2.epa. gov/sites/production/files/2015-07/documents/vision_303d_program_dec_2013.pdf.
U.S. EPA (Environmental Protection Agency). (2014) Best practices for continuous monitoring of temperature and
flow in wadeable streams. Final report. [EPA/600/R-13/170F]. Washington, DC: Global Change Research
Program, National Center for Environmental Assessment. Available online at
http://cfpub.epa.gov/ncea/global/recordisplay.cfm?deid=280013.
-------
van Kleef, HH; Verberk, WCEP; Leuven, RSEW; Esselink, H; Van der Velde, G; Van Duinen, GA. (2006)
Biological traits successfully predict the effects of restoration management on macroinvertebrates in
shallow softwater lakes. Hydrobiologia 565:201-216.
Vannote, RL; Minshall, GW; Cummins, KW; Sedell, JR; Gushing, CE. (1980) The river continuum concept. Can J
Fish Aquat Sci 37(1): 130-137.
Vinson, MR; Hawkins, CP. (1996) Effects of sampling area and subsampling procedure on comparisons of taxa
richness among streams. J N Am Benthol Soc 15(3):392-399.
Waite, IR; Herlihy, AT; Larsen, DP; Klemm, DJ. (2000) Comparing strengths of geographic and nongeographic
classifications of stream benthic macroinvertebrates in the Mid-Atlantic Highlands, USA. J N Am Benthol
Soc 19(3):429-441.
Wehrly, KE; Wiley MJ; Seelbach, PW. (2003) Classifying regional variation in thermal regime based on stream fish
community patterns. Trans Am Fish Soc 132:18-38.
Wenger, SJ; Isaak, DJ; Luce, CH; Neville, HM; Fausch, KD; Dunham, JB; Dauwalter, DC; Young, MK; Eisner,
MM; Rieman, BE; Hamlet, AF; Williams, JE. (2011) Flow regime, temperature, andbiotic interactions
drive differential declines of trout species under climate change. Proc Nat Acad Sci USA
108(34):14175-14180.
Whiles, MR; Wallace, JB. (1997) Leaf litter decomposition and macroinvertebrate communities in headwater
streams draining pine and hardwood catchments. Hydrobiologia 353:107-119.
Wright, JF; Moss, D; Armitage, PD; Furse, MT. (1984) A preliminary classification of running-water sites in Great
Britain based on macroinvertebrate species and the prediction of community type using environmental data.
FreshwBiol 14:221-256.
Wright, JF. (2000) An introduction to RIVPACS. [pgs 1-24]. In Wright, JF; Sutcliffe, DW; Furse, MT (eds).
Assessing the biological quality of freshwaters: RIVPACS and other techniques. Ambleside, Cumbria, UK:
Freshwater Biological Association.
Yoder, CO; Rankin, ET. (1995) Biological criteria program development and implementation in Ohio. [pp.
109-144, Chapter 9]. In Davis, WS; Simon, T (eds.). Biological assessment and criteria: Tools for water
resource planning and decision making. Boca Raton FL: Lewis Publishers.
Yuan, L. (2006) Estimation and application of macroinvertebrate tolerance values. National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency,
Washington, DC; EPA/600/P-04/116F. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=154869.
8-9
-------
[This page intentionally left blank.]
-------
APPENDIX A.
POWER ANALYSIS
A-l
-------
A.l. BACKGROUND
In 2011-2012, EPA collaborated with seven states in the northeastern United States on a
pilot study that laid the groundwork for the Northeast Regional Monitoring Network (RMN).
The intent was to design a monitoring network that could detect potentially small trends in
biological, thermal, hydrologic, physical habitat, and water chemistry data at high quality sites in
a decision-relevant timeframe (e.g., 10-20 years to be relevant to climate change). The design
had to achieve a balance between scientific and practical considerations. It had to build on
existing state and tribal bioassessment efforts and not exceed resource limitations of the
biomonitoring programs.
To help inform the network design, EPA and partners performed a series of analyses on
an aggregated data set to explore the following questions:
1) How long will it take to detect trends in biological metrics?
2) How much of an effect do different design decisions, such as sampling frequency and
classification scheme, have on trend detection times?
A.2. METHODS
A.2.1. Data Preparation
EPA performed a series of analyses on a regional data set comprised of benthic
macroinvertebrate, habitat, and water quality data from participating state biomonitoring
programs in New York, Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and
Vermont. In addition, macroinvertebrate data sets covering the same northeastern study area
were obtained from the U.S. EPA Wadeable Streams Study (WSA; U.S. EPA, 2006), the New
England Wadeable Streams project (NEWS; Snook et al., 2007), and the U.S. Geological Survey
(USGS) National Water Quality Assessment (NAWQA). For purposes of the analyses, the data
set was limited to reference sites only, which were defined as locations with the least amount of
anthropogenic disturbance (Hughes et al., 1986). Table A-l lists the reference criteria and
Figure A-l shows the locations of the reference sites. The data set was further limited to include
samples collected from June-September1 and samples with 80 or more total individuals2.
Samples that had only family-level or coarser identifications were also removed from the data
set.
The final data set was comprised of 1,398 samples from 953 reference sites, with sample
years ranging from 1981 to 2010. Ten different methods were represented (see Table A-2). The
lrThis encompasses the general timeframe during which participating organizations collected samples for routine
assessments.
2We selected this threshold because it is 20% smaller than the smallest subsample target size among sampling
entities, and a target of ±20% is in keeping with data quality objectives for most subsampling routines (Barbour et
al., 1999). While some samples could have low densities due to natural conditions, others reasons, such as quality
assurance issues, may account for the low count.
A-2
-------
methods differ to varying degrees in sampling effort, sampling gear, habitats sampled, index
periods, subsampling/sample processing, and/or level of taxonomic resolution (see Table A-3).
Approximately 70% of the samples were collected using kick nets. The other sampling gear
consisted of artificial substrates or Surber samplers. Most samples were collected from riffle
habitats. Approximately 7% were collected from multiple habitats.
Table A-l. Reference site criteria
Variable
Percentage Natural land cover51 (NLCD 2001 upstream)
Landscape Disturbance Index (LDI)b
Percentage Imperviousness (2001 upstream)
Percentage Imperviousness (2006 1-km radius)0
NPDES major discharges
Dams
Dissolved oxygen
Conductivity
Reference criterion
>85%
<1.5 index units
<1%
<1%
>500 m from the site
>500 m from the site
>6mg/L
<200 uS/cm
aLand cover screenings were estimates based on data associated with the National Hydrography
Dataset Plus Version 1 (NHDPlusVl) catchments in which the sites are located
(http://www.horizon-systems.com/NHDPlus). Land use data were based on the 2001 National
Land Cover Database (NLCD; Homer etal., 2007) (http://www.mrlc.gov/nlcdOl_data.php). For
this exercise, natural land cover included open water, forest, wetlands, barren, and
grassland/herbaceous. Data were accumulated for the entire upstream catchment.
bThe LDI (Brown and Vivas, 2005) was calculated by associating land uses with a scale of
disturbance intensity and weighting the index score by percentage coverage of the 2001 land uses
in the upstream catchment of each site.
'Percentage impervious data for the 1-km radius were based on the 2006 National Land Cover
Database (NLCD; Fry et al., 2011) (http://www.mrlc.gov/nlcd06_data.php).
A-3
-------
Regional reference sites
Figure A-l. Reference site locations throughout New England and New York.
A-4
-------
Table A-2. Distribution of samples across states, sampling agencies and
methods
State
CT
MA
ME
NH
NY
RI
VT
Method
CT DEEP Kick Riffle
NEWS Kick Multihabitat
CT DEEP Kick Riffle
MA DEP Kick Riffle
NEWS Kick Multihabitat
USGS Surber Riffle
WSA Kick Multihabitat
ME DEP Rock Basket
NEWS Kick Multihabitat
WSA Kick Multihabitat
NEWS Kick Multihabitat
NH DES Rock Basket
USGS Surber Riffle
WSA Kick Multihabitat
NY DEC Kick Riffle
USGS Surber Riffle
WSA Kick Multihabitat
NEWS Kick Multihabitat
RIDEM Kick Riffle
NEWS Kick Multihabitat
USGS Surber Riffle
VT DEC Kick Riffle
WSA Kick Multihabitat
Total
# Samples
53
6
1
49
1
14
1
129
30
7
25
85
1
5
497
12
13
4
21
12
1
430
1
1,398
A-5
-------
Table A-3. Description of collection and processing methods included in the analytical data set
Entity
CT
DEEP
VT
DEC
ME
DEP
NH
DES
NEWS
Collection method
12 kick samples are taken
throughout riffle habitats within
the sampling reach
Kick samples taken from riffle
habitat in 4 different locations in
the sampling reach. Substrate
disturbed at each for ~
30 seconds, total active sampling
effort 2 minutes.
3 cylindrical rock-filled wire
baskets, placed in locations with
similar habitat characteristics for
28 ± 4 days.
3 cylindrical rock-filled wire
baskets in riffle habitat or at base
of riffles at depths that cover the
baskets by at least 5 inches, for 6
to 8 weeks.
A 0.2-m2 quadrat randomly
tossed in a particular mesohabitat
of stream reach; sampled for 1
minute. 20 total quadrats
collected per site in proportion to
existing habitat in reach.
Gear
Rectangular
net (18" wide
x 9 " high ),
500 um mesh
D-frame net
(18" wide x
12" high)
with 500 um
mesh
Contents
washed into
sieve bucket
with 600 um
mesh
Contents
washed into
sieve bucket
with 600 um
mesh
1/5 meter
square
quadrat. D-
frame net
with 500 um
mesh.
Habitat
Riffle
Riffle
Riffle/run
preferred.
Riffle/run
preferred.
Multihabitat
Composite
Sample
area
~2m2
~lm2
-0.3 m2
per
basket
-0.3 m2
per
basket
~4m2
Subsampling
200-organism minimum
count, randomly selected
from a 56 grid (5 cm x 5 cm
grids) subsampling tray
1/4 of the sample, with a
minimum of 300 organisms
(if less than 300 organisms
are found, 1 grid at a time is
picked until the target is
reached or the whole sample
is picked)
Subsampling rules are
difficult to briefly summarize
(see Davies and Tsomides,
2002). For this project, the
entire samples were
processed and identified.
Quarter of the sample with a
minimum of 100 organisms
(if less than 100 organisms
are found, then the entire
sample is processed)
200-organism minimum
count, randomly selected
from a Caton grid
Index
period
Sep
15-Nov
30
Sep- mid
Oct
Jul-Sep
30
late
Jul-Sep
Jul-Sep
>
-------
Table A-3. continued...
Entity
WSA
MA
DEP
RI
DEM
NY
DEC
USGS
Collection method
A 1 square foot area was sampled
for 30 seconds at a randomly
selected location at each of the 1 1
transects. The samples were
composited into one sample per
site.
10 kick-samples are taken in riffle
habitats within the sampling reach
and composited
Kick samples are taken from riffle
habitats along 100 meter reach
representative of the stream
sampled timed for a total active
sampling effort of 3 minutes.
Substrate is dislodged by foot,
upstream of the net for 5 minutes
and a distance of 5 meters. The
preferred line of sampling is a
diagonal transect of the stream
Semi quantitative sample,
composite of 5 discrete
collections from the richest
targeted habitat (typically riffle,
main-channel, coarse-grained
substrate habitat type).
Gear
Modified
D-frame net
(12" wide)
with 500 um
mesh
kick-net,
46 cm wide
opening,
500 um mesh
D-frame net
(0.3 m width)
with 500 um
mesh
Rectangular
net (9" x
18") with
800-900 um
mesh
Slack
sampler,
500-um nets
and sieves
Habitat
Multihabitat
Composite
Riffle/run
preferred
Riffle
Riffle
Riffle
Sample
area
~lm2
~2m2
100
liner
meters
2.5m
1.25m2
Subsampling
500-organism minimum
count, randomly selected
from a Caton grid
Count-based, 100-organism
randomized pick
100 organism minimum
count, grids randomly
selected from a 16-grid tray
until minimum is picked
100
300-organism target
Index
period
Jun- mid
Octa
Jul 1-Sep
30
Aug-Sep
Jul-Sep
Late
Jun-mid
Oct
>
aThe WSA index period was supposed to end in September but some samples were collected in October
-------
The analysis consisted of 7 different biological metrics. Three of the metrics (total taxa
richness and EPT richness and percentage composition) are commonly used in bioassessments,
and the other 4 (richness and percentage composition of cold and warm water taxa) are believed
to be climate-sensitive. The list of cold and warm water taxa, which can be found in Appendix I,
were derived from: (1) generalized additive modeling analyses (GAM; Yuan, 2006) on the
Northeast data set; (2) supplemental data provided by participating organizations in the
Northeast, Mid-Atlantic and Southeast regions; and (3) best professional judgment of regional
experts. Samples with larger numbers of individuals are likely to have greater numbers of taxa,
necessitating adjustments to account for differences in subsampling procedures before
calculating richness metrics (Gotelli and Colwell, 2001). Samples were randomly subsampled to
120 organisms, which is the upper end of the 20% target used for most subsampling routines
(Barbour et al., 1999). The subsampling routine was repeated 1,000 times, and metric values
were averaged across these 1,000 runs.
A.2.2. Classification
This study explored three different classification frameworks for the Northeast. The first
one is based on an analysis of the regional macroinvertebrate data set (for details, see the
supplemental data at the end of this appendix). The classification scheme is comprised of four
broad stream classes based on slope (NHDPlus flowline slope; unitless) and size (NHDPlus
cumulative drainage area; km2). The classes are broad enough to be represented in most states
(see Figure A-2) and similar enough biologically to justify combining the macroinvertebrate data
across the region. These analyses assessed the following 3 classes: high gradient, less than
100 km2 (HGL); moderate gradient, less than 100 km2 (MGL); and low gradient and/or greater
than 100 km2 (LGG). The fourth class ('other,' which are streams that are low gradient [<0.005]
with drainage areas <10 km2 or high gradient [>0.02] with drainage areas >100 km2) was
excluded from the analysis because it occurs infrequently in the data set.
A-8
-------
NHDPIus local catchments
Classification groups
High gradient, drainage area < 100 km2
Moderate gradient, drainage area < 100 km2
^^| Low gradient and/or drainage area >100 km2 (except 'Others' listed below)
j^B Other (low gradient, <10 km2; high gradient, >100 km2)
Size and slope data not available
Figure A-2. Spatial distribution of NHDPIus local catchments grouped by
size/slope class.
The other two stream classifications were: EPA Level 2 ecoregions (Omernik, 1995) and
the Northeast Aquatic Habitat Classification (NAHC) developed by The Nature Conservancy
(TNC; Olivero and Anderson, 2008). Ecoregions are delineated based on similarities in
characteristics such as geology, physiography, vegetation, climate, soils, land use, wildlife, and
hydrology, while the NAHC represents natural flowing-water aquatic habitat types using stream
size (seven classes; headwaters to great rivers), NHDPIus flowline slope (six classes; very low to
very high gradient), temperature (four classes: cold to warm water) and geology (three classes:
buffered to acidic). These analyses evaluated two of the EPA Level 2 ecoregions (the Atlantic
Highlands and Mixed Wood Plains, which encompass the majority of the study area) and two
groupings of the NAHC classes (Creek—Moderate to High Gradient—Moderately
Buffered—Cold [TNC 1]; and Creek—Moderate to High Gradient—Acidic—Cold [TNC 2]).
Table A-4 summarizes the number of samples in each classification group.
A-9
-------
Table A-4. Summary of the data sets on which the power analyses are based
Data set
Slope/size
EPA Level 2
Ecoregion
Northeast
Aquatic
Habitat
Classification
(Olivero and
Anderson
2008)
Class
HGL
MGL
LOG
ER5.3
ER8.1
TNC 1
TNC2
Description
High gradient (>0.02), drainage area <100 km2
Moderate gradient (0.005 to 0.02), drainage
area <100 km2
Low gradient (<0.005) and/or drainage area
>100 km2 EXCEPT low gradient (<0.005),
drainage area <10 km2 and high gradient
(>0.02), drainage area >100 km2
Atlantic Highlands (5.3)
Mixed Wood Plains (8.1)
Creek (10-100 km2), moderate or high
gradient (>0.005 and <0.05), cold, moderately
buffered or neutral
Creek (10-100 km2), moderate or high
gradient (>0.005 and <0.05), cold, low
buffered or acidic
Number
of
Samples
515
362
500
1,108
278
325
159
Years
covered
1986-
2010
1987-
2010
1981-
2010
1983-
2010
1981-
2009
1986-
2010
1988-
2010
A.2.3. Analytical techniques
Estimations of variance components and power analyses simulations can help assess
differences in trend detection ability between:
• climate-sensitive and traditional metrics
• sampling frequency (1 vs. 2 vs. 5-year)
• effect size3 (0.5, 1, and 2%)
• classification schemes (size/slope; Level 2 ecoregion; NAHC)
3Effect size refers to the annual rate of change (e.g., the 0.5% effect size represents the smallest rate of change while
the 2% effect size represents the largest rate of change).
A-10
-------
Power analyses are well-suited for experimental design because one can experiment with
different settings, such as power4 (e.g., 80% or higher), sample size (e.g., 25 vs. 50), and effect
size (0.5, 1, or 2%), to get a sense of how long it will take to detect a change of a given size at a
certain level of confidence. Results can also be used to provide information on design decisions
that minimize variability and increase the power to detect trends over shorter time periods.
The variance components analysis focused on three major components of variation:
among methods, among subbasins (using eight digit hydrologic units, or HUCSs), and residual
variation. Data were aggregated at the subbasin level instead of the site level because very few
sites had a significant number of revisits. Variance among methods was included because
methods strongly influence community structure (see Figure A-S1 in the supplemental material
at end of this appendix). Variance components were estimated for each metric and stream class
using mixed effects models of the form ytjk = /?0 + /?x * xtjk + eyfc, where xtjk is the year (k)
the sample was collected with a specific method (/') and within a specific subbasin (/'), and where
the error term is partitioned into three components, e^ = boi + b0j + rjijk. The models produce
linear equations of change in each metric over time, and estimate the variability among
collection methods, boi~N(Q, ffcoiiMeth), among subbasins, b0j~N(Q, ffHUC8), and unaccounted
for interannual variation, ilijk~N(Q, oresiduai). In each simulation run, samples were assigned to
an artificial subbasin and one often methods (3 samples per method). Each simulation was run
1,000 times, for a fixed number of years, ranging from 2 to 131 years, where power is the
percentage of runs with a significant slope effect (/?i,/> < 0.05). Variance components for each
model were estimated with REML using the LME4 package in R (Bates et al., 2011).
Power analyses were conducted for each combination of stream class (using the 7 classes
listed in Table A-4), invertebrate metric (7 metrics), effect size (0.5, 1.0, or 2.0%), and sampling
frequency (1 vs. 2 vs. 5-year) as categorical main effects. For each metric and stream class
combination, the power analyses calculated the number of years needed to reach 80% power by
creating simulated data sets using the above estimated error components at different effect sizes
(2.0, 1.0, or 0.5% annual rate of change) and sampling intervals (annual, biannual, or every five
years). It was assumed that 30 sites were sampled at each sampling frequency5. To summarize
these results, the number of years needed to exceed 80% power was analyzed using a linear
model that included the stream class, invertebrate metric, effect size, and sampling frequency as
categorical factors. Since it was expected that specific metrics would perform well in specific
stream classes (e.g., that cold-water taxa would have higher detection probabilities in classes that
included higher gradient, upland streams), a metric-class interaction was added into the model.
Similarly, because it was expected that higher sampling frequency would be more important for
low effect sizes, an effect size-sampling frequency interaction was also included in the model. To
4Power is the likelihood or probability of correctly detecting an outcome of a given size; for example, 80% power
means that there is an 80% probability that an outcome of a given size is correctly detected.
5EPA and partners performed an exploratory analysis to evaluate how much of a difference it would make if 30 vs.
50 sites were sampled. They found that trend detection times were very similar (detection times differed by
1-2 years, depending on the effect size; the smaller the effect size, the greater the difference; unpublished data).
A-ll
-------
meet model assumptions, exceedances were log transformed before analysis. When applicable,
follow-up multiple comparisons were conducted using a Bonferroni correction for multiple
testing.
A.3. RESULTS
Trend detection time was significantly (P < 0.0001) influenced by frequency of sampling,
type of metric, and stream classification (see Table A-5). The interactions included in the model
were significant as well, indicating that detection times associated with each metric depended on
the stream class, and that the relationship between sampling frequency and detection time
depended on the effect size (P < 0.001; see Table A-5).
Table A-5. Power analysis model output table, assuming 80% power and a
30-site sample size
Factor
Biological metric
Stream class
Biological metric x stream class
Effect size
Sampling frequency
Effect size x sampling frequency
Residual
DF
6
6
36
2
2
4
384
SS
92.32
0.92
3.95
62.8
15.87
0.04
0.83
MS
15.39
0.15
0.11
31.4
7.94
0.01
0.002
F
7,120.07
70.63
50.75
14,529.82
3,672.31
4.12
P
O.OOOl
O.OOOl
O.OOOl
O.OOOl
O.OOOl
0.0003
Results suggest that trends in biological indicators can be detected within 10-20 years (at
80% power) if 30 or more sites that have comparable environmental conditions and biological
communities are monitored regularly (see Figure A-3). As shown in Figure A-3, the 'traditional'
metrics (total taxa richness and EPT richness and percentage composition) had shorter trend
detection times than the climate-sensitive metrics, and the climate-sensitive richness metrics had
shorter trend detection times than the climate-sensitive percentage composition metrics (see
Figure A-3). Table A-6 contains a complete list of the number of years needed to exceed 80%
power for each metric and stream class combination with different effect sizes (0.5, 1, and 2%)
and sampling frequencies (1, 2 and 5-year). Overall, the total taxa richness metric had the
shortest trend detection time (13-15 years, depending on classification scheme), while the
warm-water percentage individual metric had the longest trend detection time (30+ years; see
Table A-6). As expected, the total taxa and EPT richness metrics had higher mean values than
the climate-sensitive richness metrics. As shown in Figure A-4, mean richness metric values are
A-12
-------
inversely related to the number of years needed to detect a trend with 80% power. Put more
simply, richness metrics that have higher mean values have shorter trend detection times.
There were differences in performance across the different stream classes (see
Figure A-3, Table A-6). The differences were most evident in the climate-sensitive metrics.
Overall, the Mixed Wood Plains (8.1) EPA Level 2 ecoregion generally had the longest trend
detection times, particularly with the traditional metrics (see Figure A-3, Table A-6). The NAHC
classes and 2 of the size/slope classes (high gradient/less than 100 km2 [HGL] and moderate
gradient/less than 100 km2 [MGL]) generally had the shortest trend detection times for the
traditional metrics and the cold water metrics, while the low gradient and/or greater than 100 km2
(LOG) class had the shortest trend detection times for the warm water metrics (see Figure A-3,
Table A-6). Generally speaking, the classifications that were built to separate out small to
medium-sized, moderate to high gradient, cold water streams versus large, low gradient warm
water streams tended to detect trends in the climate-sensitive metrics in the shortest time periods.
"-< _1
I t_> m —' rn *H —,—,_—.—
. < CQ <_> u u Q LJJ
Ł|^
"J S —h
100
90
80 -
70 -
60 -
&
§
»
1 50
u
i3 40 -
&
2 30
m
J-
c
n
-------
01
Q- <2
Cold Rich
EPT Rich
Total Rich
Warm Rich
"! f
10 15 20
Mean Metric Richness
25
30
Figure A-4. Relationship between mean metric richness values and number
of years to detect trends. Each dot represents the number of years needed to
exceed 80% power based on various combinations of the 4 richness metrics,
7 stream classes, 3 effect sizes, and 3 sampling frequencies. Richness metrics
are color-coded (cold water richness = black; EPT richness = red; total
richness = green; warm water richness = blue). This analysis assumes a
30-site sample size.
Sampling frequency (1 vs. 2 vs. 5-year) also had a significant effect on trend detection
times. The weaker the trend (and lower the effect size), the more of a difference the sampling
frequency made (see Figure A-5). Annual sampling had the shortest trend detection time across
all effect sizes. Annual sampling at 1.0 and 0.5% effect sizes was equivalent to sampling every
five years at 2.0 and 1.0% effect sizes, respectively (see Figure A-5).
A-14
-------
60
Every Every Every Every
Year (A) 2 Years (B) 5 Years (C) Year(C)
Every Every Every
2 Years (D) 5 Years (E) Year(E)
Every Every
2 Years (F) 5 Years (G)
-------
Table A-6. Number of years to detect trends with 80% power for metric and
stream class combinations with different effect sizes (0.5,1, and 2%) and
sampling frequencies (1, 2 and 5-year). This assumes a 30-site sample size.
Asterisks (a) indicate that trend detection times exceed 40 years
Metric
Number of
Total taxa
Number of
EPT taxa
Class
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8.1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic (TNC
2)
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8.1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic
(TNC 2)
2.0%
1
yr
7
7
7
7
7
6
6
9
9
8
9
9
8
7
2
yr
9
9
7
9
7
7
7
11
11
9
11
11
9
9
5
yr
11
11
11
11
11
11
11
11
16
11
11
11
11
11
1.0%
1
yr
11
11
11
11
10
10
9
13
14
13
13
14
12
11
2
yr
13
13
13
13
11
11
11
17
17
15
17
17
15
13
5
yr
16
16
16
16
16
16
16
21
21
21
21
21
21
16
0.5%
1
yr
17
18
16
17
15
15
15
21
23
20
21
21
19
17
2
yr
21
23
21
21
19
19
19
25
29
25
25
27
23
21
5
yr
26
26
26
26
26
26
21
31
36
31
36
36
31
26
A-16
-------
Table A-6. continued...
Metric
Percentage
EPT
individuals
Number of
Cold water
taxa
Class
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8. 1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic
(TNC 2)
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8. 1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic (TNC
2)
2.0%
1
yr
9
11
9
10
10
9
8
13
13
11
13
11
10
11
2
yr
11
13
11
11
11
11
9
15
15
15
17
13
13
13
5
yr
16
16
16
16
16
11
11
21
21
16
21
16
16
16
1.0%
1
yr
14
18
15
15
15
13
13
21
21
18
21
18
16
18
2
yr
17
21
17
19
19
15
15
25
25
21
25
21
21
21
5
yr
21
26
26
26
26
21
21
31
31
26
36
26
26
26
0.5%
1
yr
22
28
23
24
23
20
20
32
32
29
33
28
26
28
2
yr
29
33
29
31
29
25
25
39
a
35
a
35
31
35
5
yr
36
a
36
a
a
31
31
a
a
a
a
a
a
a
A-17
-------
Table A-6. continued...
Metric
Percentage
Cold water
individuals
Number of
Warm water
taxa
Class
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8. 1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic
(TNC 2)
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8. 1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic
(TNC 2)
2.0%
1
yr
18
17
14
21
17
14
14
17
17
17
14
16
18
15
2
yr
23
21
17
25
21
17
17
21
21
21
17
21
23
19
5
yr
31
26
21
31
26
21
21
26
26
26
21
26
31
26
1.0%
1
yr
28
26
22
32
26
22
21
27
26
28
23
26
29
24
2
yr
35
33
29
a
33
27
25
33
33
33
27
33
37
29
5
yr
a
a
36
a
a
36
36
a
a
a
36
a
a
a
0.5%
1
yr
a
a
35
a
40
34
33
a
a
a
34
a
a
38
2
yr
a
a
a
a
a
a
a
a
a
a
a
a
a
a
5
yr
a
a
a
a
a
a
a
a
a
a
a
a
a
a
A-18
-------
Table A-6. continued...
Metric
Percentage
Warm water
individuals
Class
Atlantic Highlands (ER 5.3)
Mixed Wood Plains (ER 8. 1)
High gradient, <100 km2
(HGL)
Low gradient and/or
>100km2a(LGG)
Moderate gradient, <100 km2
(MGL)
Creek, mod/high gradient,
cold, mod buffer/neutral
(TNC 1)
Creek, mod/high gradient,
cold, low buffer/acidic
(TNC 2)
2.0%
1
yr
25
26
7,5
22
29
31
32
2
yr
31
31
31
27
37
a
39
5
yr
a
a
a
36
a
a
a
1.0%
1
yr
39
40
39
34
a
a
a
2
yr
a
a
a
a
a
a
a
5
yr
a
a
a
a
a
a
a
0.5%
1
yr
a
a
a
a
a
a
a
2
yr
a
a
a
a
a
a
a
5
yr
a
a
a
a
a
a
a
aExcept low gradient (<0.005), drainage area <10 km2 and high gradient (>0.02), drainage area >100 km2
Table A-7 contains 3 sets of results from the variance components analysis: (1) overall
variability (estimates are averaged across all metric and stream class combinations); (2) mean
variability for each biological metric (estimates are averaged across the stream classes); and
(3) mean variability for each stream class (estimates are averaged across the biological metrics).
Overall, the residual variation (which represents variability that cannot be attributed to the trend
over time, the collection method, and HUC8) was largest (7.02), followed by collection method
(4.9) and HUC8 subbasin (3.3; see Table A-7). Percentage individual metrics had higher mean
variability than the richness metrics, particularly the percentage EPT and percentage cold water
individuals metric (see Table A-7). The mean variability estimates were fairly similar across
stream classes.
A-19
-------
Table A-7. Results from the variance components analysis. For the overall
average, estimates are averaged across all metric and stream class
combinations. For the biological metrics, estimates are averaged across the
stream classes and for the stream classes, estimates are averaged across the
biological metrics.
Set of results
Overall average
Mean CollMeth
4.9
MeanHUCS Mean Residual
3.3
7.0
Biological metrics
Total taxa richness
EPT richness
Cold water taxa richness
Warm water taxa richness
Percentage EPT individuals
Percentage Cold water individuals
Percentage Warm water individuals
5.7
2.0
1.8
0.8
14.1
7.0
3.0
2.6
1.8
1.1
0.7
6.8
6.7
3.3
5.1
3.5
2.3
1.3
16.5
13.0
7.4
Stream classes
Atlantic Highlands (5.3)
Mixed Wood Plains (8.1)
High gradient, <100 km2 (HGL)
Moderate gradient, <100 km2 (MGL)
Low gradient and/or >100 km2* (LGG)
Creek, mod/high gradient, cold, mod
buffer/neutral (TNC 1)
Creek, mod/high gradient, cold, low
buffer/acidic (TNC 2)
5.0
5.3
5.1
4.8
4.9
4.0
5.5
3.0
3.2
2.7
3.0
3.8
3.6
4.0
7.5
7.7
7.0
7.1
7.0
6.5
6.3
A.4. CONCLUSIONS
The analyses suggest that detection times of 10-20 years (at 80% power) are possible for
several of the biological metrics if 30 or more sites with comparable environmental conditions
and biological communities are monitored regularly. These results are consistent with other
research. For example, Larsen et al. (2004) found that well-designed networks of 30-50 sites
monitored consistently can detect underlying changes of 1-2% per year in a variety of metrics
within 10-20 years, or sooner, if such trends are present. Larsen et al. (2004) also emphasized
the importance of the duration of the survey, citing that trend detection capability increases
substantially with time.
A-20
-------
The analyses also show that richness metrics have lower variability than percentage
individual metrics, which means that there is a greater likelihood of detecting trends in the
richness metrics over shorter time periods. Because richness values increase with subsampling
effort, it is important to identify an adequate number of organisms, particularly for metrics with
low representation (e.g., cold-water taxa). Identifying a larger subsample of organisms at RMN
sites (e.g., 300 or more organisms) will improve the power to detect trends. While this may
increase costs in the short-term, the resulting increase in power lowers detection times, thereby
reducing overall costs of the long-term monitoring effort. Moreover, the earlier detection may
lead to management actions that could alleviate additional, costlier impacts.
The classification scheme also influences trend detection times. Thus, the Northeast
should be partitioned to minimize environmental variability and gradients. Size and gradient are
key variables to capture. Schemes that consider additional variables like thermal class and
geology (like TNC's NAHC) may further improve performance, but those variables are generally
not as readily available (or dependable), and in this study, did not make a large difference in
trend detection times. Classification is also an important consideration during site selection.
Since one of the objectives of the RMNs is to detect trends attributable to climate change,
tracking changes in the thermal indicator taxa (particularly cold water taxa) is of interest. As
discussed above, the more cold water taxa present at the RMN sites, the greater the chance of
detecting trends in the cold water richness metric over shorter time periods. Based on
unpublished analyses on the Northeast data set, sites in small to medium-sized, medium to high
gradient stream classes (MGL and HGL) have higher numbers of cold-water taxa, so selecting
sites in these stream classes will improve the chances of detecting climate-related trends over
shorter time periods.
Sampling frequency is another important consideration. The analyses show that sampling
macroinvertebrates on an annual basis improves trend detection times, particularly if trends are
subtle. If sites in the Northeast RMN are sampled annually over a long time period (e.g.,
10+ years), this will improve the chances of detecting trends. Other design recommendations,
which are described in Section 3.1.1, are based on literature and Standard Operating Procedures
(SOPs) commonly used by RMN partners. These include recommendations on time periods for
sample collection (also referred to as 'index periods'), collection protocols, type(s) of habitats
sampled, and level of taxonomic identification. The use of consistent methods will increase the
comparability of data, minimize biases and variability, and ensure that the data meet data quality
objectives of the Northeast RMN. These recommendations also can be applied to other regions
interested in climate change detection in streams.
A-21
-------
A.5. SUPPLEMENTAL MATERIAL
A.5.1. Derivation of the Size/Slope Classification Scheme
Prior to running the classification analysis on the aggregated Northeast data set, the data
set was reduced by randomly selecting one sample per site so that sites with multiple years of
data would not receive unequal weight (sample size = 689). The 689 sites were distributed across
the whole study region. For site classification, nonmetric multidimensional scaling (NMDS)
ordination was used to explore predominant types of reference samples based on taxa presence
and absence. The ecodist package in R was used to calculate Bray-Curtis similarity measures
(Goslee and Urban, 2007; R Core Development Team, 2012) to perform NMDS. Prior to running
the analysis, guidelines recommended in Cuffney et al. (2007) were used to develop Operational
Taxonomic Units (OTUs) as necessary to achieve an acceptable level of taxonomic consistency
across the data set. For interpretability, the NMDS was limited to four axes (Stress = 0.20,
r2 = 0.4).
Initial NMDS ordination using all data showed that sites group by method (see
Figure A-S1), indicating that method is an important determinant of taxonomic structure. These
differences appeared along all four axes, but were most pronounced in the first three. For
example, samples from Maine and New Hampshire (artificial substrate ellipsoid, see
Figure A-S1) overlapped and were largely distinct from New York and Vermont samples (riffle
D-net ellipsoid, see Figure A-S1). The strongest correlations (\r\ > 0.3) between taxa and
environmental variables included some combination of slope and drainage size in all four axes,
but also included longitude on axis 2, and elevation and maximum temperature on axis four (see
Table A-S1, left). Axis 1 can be interpreted as shifts in the macroinvertebrate community that
occurred as measures of slope increased and drainage area decreased (particularly slope and logic
drainage area), while axes two and three separated the two. Axis two was most strongly related
to mean slope (and longitude) and axis three to logic drainage area. Axis four is less easily
interpreted.
A-22
-------
9 _
9 _
(Q
9
-0.4
-0.2
0.0
0.2
0.4
0.6
Axis 2
Figure A-S1. Axes 1 and 2 from first NMDS ordination, using data from all
methods. All four axis correlations are confounded by method. Methods are
represented with different symbols and sampling devices are shown with two
rings (solid and dashed 95% confidence ellipsoids). WSA and NEWS samples
are also highlighted (dotted 95% confidence ellipsoid).
A-23
-------
Table A-S1. Correlations between environmental variables and NMDS axes
for all sites (left) and NEWS and WSA sites (right). Both analyses show
consistent ties between taxonomic axes and aspects of slope and drainage
area, r2 > 0.3 are bolded.
Environmental
Variable
Latitude
Longitude
Sinuosity
Slope
Mean Slope
Drainage Area
Logio(Dr Area)
Elevation
Stream Aspect
Max Temp
Min Temp
Length
All Sites
XI
-0.157
-0.173
-0.039
0.405
0.259
-0.245
-0.403
0.215
-0.070
0.041
0.027
0.082
X2
0.033
0.405
0.080
-0.112
-0.327
0.086
-0.117
-0.247
0.015
0.165
0.057
-0.118
X3
0.056
0.024
-0.002
-0.093
0.219
0.016
0.369
-0.073
-0.003
-0.049
0.013
0.143
X4
-0.178
-0.244
0.068
-0.462
-0.313
0.150
0.366
-0.308
0.053
0.305
0.263
-0.115
NEWS and WSA sites
XI
0.062
0.270
0.128
-0.205
-0.064
0.317
0.403
-0.203
-0.256
0.031
-0.086
0.102
X2
-0.123
-0.130
-0.041
0.174
0.028
-0.257
-0.339
0.278
0.112
-0.004
-0.064
-0.002
X3
-0.241
0.037
0.112
-0.506
-0.602
0.154
0.023
-0.351
0.060
0.416
-0.013
-0.309
X4
0.121
-0.001
-0.223
-0.320
-0.237
0.136
0.116
-0.095
0.094
-0.004
0.253
-0.324
To eliminate the confounding effects of method, a second NMDS ordination used data
only from agencies that used similar methods and sampled across the region. This included 64
sites from the NEWS and WSA studies (multihabitat ellipsoid, see Figure A-S1). The resulting
ordination also was limited to four axes for interpretability (Stress = 0.18, r2 = 0.53). This
ordination showed high overlap between the two studies on all four axes. Likewise, the strongest
correlations (\r\ > 0.3) between taxa axes and environmental variables included some
combination of slope or mean slope and drainage area or logio(drainage area) in all four axes, but
also included elevation and maximum temperature on axis 3, and NHDPlus flowline length on
axes 3 and 4 (see Table A-S1, right). Axes 1 and 2 reflected shifts in the macroinvertebrate
community that occur as measures of drainage area change, while axes 3 and 4 showed shifts
related to measures of slope.
Given the consistent appearance of slope and drainage area variables in both NMDS
results, these two variables were chosen as a simple way to identify stream classes. Thresholds
for both slope and drainage were selected to be consistent with the TNC's Northeast Aquatic
Habitat Classification (NAHC; Olivero and Anderson, 2008). The distributions of the biological
metrics used in this study were compared in multiple categories of slope and catchment area. In
streams with catchments <100 km2, slope was a dominant effect. Streams with similar slopes but
A-24
-------
differing catchment size (<10 or >10 km2) had similar metric distributions. In small rivers with
catchments >100 km2, slope was less of a factor and metric distributions resembled those in
lower gradient smaller streams.
The end result was the following four broad stream classes based on slope (NHDPlus
flowline slope; unitless) and size (NHDPlus cumulative drainage area; km2):
• High gradient, less than 100 km2 (HGL)
• Moderate gradient, less than 100 km2 (MGL)
• Low gradient and/or greater than 100 km2 (LGG)
• 'Other'—low gradient (<0.005), drainage area <10 km2 and high gradient (>0.02),
drainage area >100 km2
A.6. LITERATURE CITED
Barbour, MT; Gerritsen, J; Snyder, BD; Stribling, JB. (1999) Rapid bioassessment protocols for use in streams and
wadeable rivers: Periphyton, benthic macroinvertebrates and fish, Second Edition. [EPA 841-B-99-002].
Washington, D.C: U.S. Environmental Protection Agency, Office of Water. Available online:
http://water.epa.gov/scitech/monitoring/rsl/bioassessment/index.cfm
Bates, D; Maechler, M; Bolker, B. (2011) Ime4: Linear mixed-effects models using S4 classes. R package version
0.999375-42. Available online at http://CRAN.R-project.org/package=lme4.
Brown, MT; Vivas, MB. (2005) Landscape development intensity index. Environmental Monitoring and
Assessment 101(289-309).
Cuffney, TF; Bilger, MD; Haigler, AM. (2007) Ambiguous taxa: effects on the characterization and interpretation of
invertebrate assemblages. J N Am Benthol Soc 26(2):286-307.
Davies, SP; Tsomides, L. (2002). Methods for biological sampling and analysis of Maine's rivers and streams. (DEP
LW0387-B2002). August, Maine: Maine Department of Environmental Protection.
http://www.maine.gov/dep/water/monitoring^iomonitoring/materials/finlmethl.pdf
Fry, J; Xian, G; Jin, S; J.Dewitz, C.Homer, L.Yang, C.Barnes, N.Herold, and J. Wickham. (2011) Completion of the
2006 national land cover database for the conterminous United States. PE&RS 77(9): 859-864. Available
online at www.mrlc.gov/downloadfile2.php?file=September20HPERS.pdf
Goslee, S; Urban, D. (2007) The ecodist package for dissimilarity-based analysis of ecological data. J Stat Soft
22:1-19.
Gotelli, NJ; Colwell, RK. (2001) Quantifying biodiversity: procedures and pitfalls in the measurement and
comparison of species richness. Ecol Lett 4:379-391.
Homer, C; Dewitz, J; Fry, J; Coan, M; Hossain, N; Larson, C; Herold, N; McKerrow, A; VanDriel, JN; Wickham, J.
(2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. PE&RS
73(4):337-341. Available online at
http://www.asprs.0rg/a/publications/pers/2007journal/april/highlight.pdf
Hughes, RM; Larsen, DP; Omernik, JM. (1986) Regional reference sites: a method for assessing stream potentials.
Environmental Management 10(5):629-635.
A-25
-------
Larsen, D; Kaufmann, P; Kincaid, T; Urquhart, N. (2004) Detecting persistent change in the habitat of salmon-
bearing streams in the Pacific Northwest. Can J Fish Aqua Sci 61:283-291.
Olivero, AP; Anderson, MG. (2008) Northeast aquatic habitat classification. Denver, CO: The Nature Conservancy.
Available online at http://rcngrants.org/content/northeastern-aquatic-habitatclassification-project
Omernik, JM. (1995) Ecoregions: a spatial framework for environmental management. In: Davis, W; Simon, T; Eds
Biological assessment and criteria: tools for water resource planning and decision making, [pp 49-62].
Boca Raton, FL: Lewis Publishers.
R Core Development Team. (2012) R: A language and environment for statistical computing. Vienna, Austria.
Snook, H; Davies, S; Gerritsen, J; Jessup, B; Langdon, R; Neils, D;Pizutto, E. (2007) The New England Wadeable
Stream Survey (NEWS): development of common assessments in the framework of the biological
condition gradient. Prepared for USEPA Office of Science and Technology and Office of Wetlands,
Oceans and Watersheds, Washington, DC.
U.S. EPA (Environmental Protection Agency). (2006) Wadeable Streams Assessment: a collaborative survey of the
Nation's streams. Office of Water, Washington, DC; EPA 841/B-06/002.
Yuan, LL. (2006) Estimation and application of macroinvertebrate tolerance values. [EPA/60/P-04/116F].
Washington, DC National Center for Environmental Assessment. U.S. Environmental Protection Agency,
Washington, DC.
A-26
-------
APPENDIX B.
CHECKLIST FOR
STARTING A REGIONAL
MONITORING NETWORK
(RMN)
B-l
-------
1. Establish the regional working group.
• Coordinator (e.g., from an U.S. EPA Region or a state) volunteers to lead the regional
working group.
• The coordinator works with EPA and partners to create a contact list.
• EPA and the coordinator hold a kick-off webinar to brief potential partners on current
RMN efforts, describe the RMN framework and development process, and discuss a
potential timeline for implementation.
2. The coordinator requests candidate sites from each entity. During the site selection process,
the working group considers site selection criteria being used in other RMN regions and tries
to use similar criteria where practical. Desired site characteristics include:
Part of established, long-term monitoring networks
Low level of anthropogenic disturbance
Colocated with existing equipment (e.g., USGS gage, weather station)
Exhibit similar environmental and biological characteristics
Longevity (e.g., accessible [day trip], opportunities to share the workload with outside
agencies or organizations)
In watersheds that are protected from future development
Lengthy historical sampling record for biological, thermal or hydrological data
The regional coordinator compiles information on data collection and processing protocols
being used by each regional working group member, and EPA distributes the generic RMN
QAPP1. The regional working group reviews the generic QAPP and discusses appropriate
protocols for their region. The group attempts to use similar protocols to other RMNs. The
draft (region-specific) protocols are written up in an addendum to the QAPP.
4. EPA has been conducting research on screening, classification, and vulnerability analyses for
several pilot RMNs (additional documentation to conduct these steps are available from
EPA). Pending availability and funding, EPA may be able to assist with the following steps:
• Screening the candidate sites by running them through a disturbance screening process
similar to what is described in Appendix D. This may include developing criteria for
JU.S. EPA. 2015. Draft generic Quality Assurance Project Plan for monitoring networks for tracking long-term
conditions and changes in high quality wadeable streams. This document is available from EPA upon request.
Contact Britta Bierwagen (bierwagen.britta@epa.gov).
B-2
-------
"reference" sites in urban and agricultural areas. Disturbance ratings will be assigned to
the candidate sites.
Gathering information from the regional working group on existing classification
schemes in the region and performing analyses to explore regional classification. Sites
will be assigned to classification groups.
Gathering information on existing climate change vulnerability assessments and
performing broad-scale analyses to assess the vulnerability of the candidate RMN sites to
climate change.
5. The regional coordinator works with EPA and regional working group members to finalize
site selection and protocols. These are included as addendums to the QAPP.
6. The group identifies and seeks resources for implementation. High priority start-up items
typically include obtaining equipment and finding funds to process macroinvertebrate
samples.
B-3
-------
[This page intentionally left blank.]
-------
APPENDIX C.
PRIMARY REGIONAL
MONITORING NETWORK
(RMN) SITES IN THE
NORTHEAST,
MID-ATLANTIC, AND
SOUTHEAST RMN
REGIONS
c-i
-------
Table C-l. Site information for primary RMN sites in the Northeast (3/2015). Drainage area, slope, and
elevation are estimates based on NHDPlus vla local catchment data. Percentage forest is derived from the NLCD
2001b data layer and is based on the total watershed
Longitude
-73.27990
-71.83424
-72.83917
-72.16196
-73.03027
-72.96731
-72.04780
-72.38454
-69.64424
-71.35110
-71.24924
-71.36166
-71.29306
-71.87633
-73.54621
-74.26626
-71.61201
-71.63562
-72.88583
-72.66250
Latitude
41.92670
41.47482
41.94639
42.03448
42.66697
42.06555
42.39431
42.46471
44.95675
43.14410
44.21896
44.35426
43.89639
44.10563
41.49457
42.01954
41.83760
41.76482
43.87167
42.76389
State
CT
CT
CT
MA
MA
MA
MA
MA
ME
NH
NH
NH
NH
NH
NY
NY
RI
RI
VT
VT
Entity
CT DEEP
CT DEEP
CT DEEP
MADEP
MADEP
MADEP
MADEP
MADEP
MEDEP
NHDES
NHDES
NHDES
NHDES
NHDES
NY DEC
NY DEC
RIDEM
RIDEM
VTDEC
VTDEC
Station ID
CTDEP_2342
CTDEP_1748
CTDEP_1433
MADEP_BB01
MADEP_CR01AA
MADEP_HRCC
MADEP_PBCC
MADEP_WSR01
MEDEP_57229
NHDES_99M-44
USGS_0 1064300
NHDES_19-ISR
NHDES_98S-44
NHDES_WildAmmo
NYDEC_HAVI_01
NYDEC_LBEA_01
RIDEM_RMR03a
RIDEM_SCI01
VTDEC_135404000013
VTDEC_670000000 166
Water body name
Brown Brook
Pendleton Hill
West Branch Salmon
Browns
Cold River
Hubbard
Parkers Brook
West Branch Swift
East Branch Wesserunsett
Stream— Station 486
Bear
Ellis
Israel
Paugus
Wild Ammo
Haviland Hollow
Little Beaver Kill
Rush
Wilbur Hollow
Bingo
Green
Drainage area
(km2)
14.7
10.4
34.5
14.7
16.8
30.0
13.8
9.8
126.0
25.7
28.2
16.6
31.5
96.2
24.9
42.7
12.2
11.2
29.2
67.8
Slope
(unitless)
0.026
0.006
0.021
0.023
0.026
0.029
0.011
0.011
0.008
0.005
0.031
0.023
0.008
0.010
0.011
0.008
0.017
0.008
0.017
0.010
Elevatio
n(m)
286.4
55.2
169.35
253.5
592.4
359.8
244.9
209.9
207.2
138.9
686.7
544.7
264.2
481.0
202.9
393.3
118.2
124.3
458.5
293.3
Percentage
Forest (%)
90.2
71.7
81.6
87.3
89.3
86.5
79.5
91.5
83.4
81.5
88.6
92.5
97.8
96.7
85.7
90.3
72.6
74.5
97.3
89.9
p
to
-------
Table C-l. continued...
Longitude
-71.78528
-72.53705
-72.93194
Latitude
44.58417
44.43400
43.13833
State
VT
VT
VT
Entity
VTDEC
VTDEC
VTDEC
Station ID
VTDEC_21 1200000268
VTDEC_495400000161
VTDEC_03 3 50000008 1
Water body name
Moose
North Branch Winooski
Winhall
Drainage area
(km2)
59.0
29.1
43.8
Slope
(unitless)
0.015
0.014
0.017
Elevatio
n(m)
532.7
327.1
587.7
Percentage
Forest (%)
97.5
95.3
95.0
ahttp://www.horizon-sy stems. com/nhdplus/nhdplusvl_home.php
bhttp://www.mrlc.gov/nlcd01_data.php
O
-------
Table C-2. Equipment installed at primary RMN sites in the Northeast (3/2015)
State
CT
CT
CT
MA
MA
MA
MA
MA
ME
NH
NH
Entity
CT DEEP
CT DEEP
CT DEEP
MADEP
MADEP
MADEP
MADEP
MADEP
MEDEP
NHDES
NHDES
Station ID
CTDEP_2342
CTDEP_1748
CTDEP_1433
MADEP_BB01
MADEP_CR01AA
MADEP_ HRCC
MADEP_PBCC
MADEP_WSR01
MEDEP_57229
NHDES_99M-44
USGS_0 1064300
Water body name
Brown Brook
Pendleton Hill
West Branch Salmon
Browns
Cold River
Hubbard
Parkers Brook
West Branch Swift
East Branch Wesserunsett
Stream— Station 486
Bear
Ellis
Temperature
water
water and air
water
water and air
water and air
water and air
water and air
water and air
water3
water and air
water and air
Hydrologic
equipment
none
USGS gage
(01118300)
none
pressure
transducer
pressure
transducer/USGS
gage
USGS gage
(01187300)
pressure
transducer
USGS gage
(01174565)
USGS gage
(01048220)
pressure
transducer
pressure
transducer
Hydrologic
data type
none
discharge
none
stage and
discharge
stage and
discharge
discharge
stage and
discharge
discharge
discharge
stage
stage
Notes
gage located at biological
sampling site
flow rating curve
developed by MA RIFLS
flow rating curve
developed by MA
RIFLS; USGS gage is
now being installed at
this site
gage is downstream of
site but location looks
representative of stream
conditions
flow rating curve
developed by MA RIFLS
gage is downstream of
site but location looks
representative of stream
conditions
gage located at biological
sampling site
-------
Table C-2. continued...
State
NH
NH
NH
NY
NY
RI
RI
VT
VT
VT
VT
VT
Entity
NHDES
NHDES
NHDES
NY DEC
NY DEC
RIDEM
RIDEM
VTDEC
VTDEC
VTDEC
VTDEC
VTDEC
Station ID
NHDES_19-ISR
NHDES_98S-44
NHDES_WildAmmo
NYDEC_HAVI_01
NYDEC_LBEA_01
RIDEM_RMR03a
RIDEM_SCI01
VTDEC_135404000013
VTDEC_670000000 166
VTDEC_2 11200000268
VTDEC_495400000161
VTDEC_033500000081
Water body name
Israel
Paugus
Wild Ammo
Haviland Hollow
Little Beaver Kill
Rush
Wilbur Hollow
Bingo
Green
Moose
North Branch Winooski
Winhall
Temperature
water and air
water and air
water and air
water and air
water and air
Water
Water
water3
Water
water and air
(Wx station)
water
water3
Hydrologic equipment
pressure transducer
pressure transducer
pressure transducer
none
USGS gage (01362497)
USGS gage (01 1151 14)
USGS gage (01 115297)
none
USGS gage (01 170100)
pressure transducer
none
none
Hydrologic
data type
stage
stage
stage
none
discharge
discharge
discharge
none
discharge
stage
none
none
Notes
gage located at
biological sampling site
gage located at
biological sampling site
gage located at
biological sampling site
gage is downstream of
site but location looks
representative of stream
conditions
working on flow rating
curve
3not deployed year-round
-------
Table C-3. Site information for primary RMN sites in the Mid-Atlantic (3/2015). Most drainage area, slope, and
elevation measurements are estimates based on NHDPlus vla local catchment data. Percentage forest is based on
total watershed and is mostly derived from the NLCD 2001b data layer. Better data were used, where available
(e.g., MD DNR was able to provide information based on exact watershed delineations and the NLCD 2006° data
layer)
Longitude
-75.74869
-75.75587
-79.27980
-79.15566
-77.43406
-78.90556
-79.06689
-75.12664
-74.43437
-74.52972
-77.45100
-77.01929
-79.23750
-77.77068
-79.57152
-79.44821
-78.40670074
-80.57352
-78.26867
Latitude
39.74567
39.72995
39.64252
39.50363
39.60929
39.54581
39.59930
40.97143
41.10693
40.76500
39.89700
41.42653
40.00333
41.49970
41.69451
37.53920
38.48708676
37.37393
38.70296
State
DE
DE
MD
MD
MD
MD
MD
NJ
NJ
NJ
PA
PA
PA
PA
PA
VA
VA
VA
VA
Entity
DNREC
DNREC
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
NJ DEP/EPA R2
NJDEP
NJDEP
PADEP
SRBC
PADEP
SRBC
PADEP
VDEQ
ShenNP
VDEQ
ShenNP
Station ID
105212
105213
YOUG-432-S
SAVA-204-S
UMON-288-S
PRLN-626-S
SAVA-225-S
AN0012
AN0260
AN0215A
CR
SRBC_Grays
JMR/WQN_734
SRBCJCettle
WBC/WQN_873
2-HUO005.87
3-RAP088.21
9-LRY007.02
3-PIY003.27
Water body name
Tributary of White Clay
Tributary of White Clay
Bear Creek
Crabtree Creek
High Run
Mill Run
Savage River
Dunnfield Creek
Mossmans Brook
Primrose
Carbaugh Run
Grays Run
Jones Mill Run
Kettle
West Branch of Caldwell Creek
Hunting Creek
Rapidan River (upper)
Little Stony Creek
Piney River
Drainage
area
(km2)
2
2.2
22.7
43.9
3.3
2.0
138.3
9.5
10.0
0.01
15.5
51.2
12.8
210.3
50.7
10
12.7
48.0
10.0
Slope
(unitless)
0.023
0.018
0.011
0.041
0.075
0.108
0.018
0.048
0.009
0.014
0.022
0.014
0.019
0.000
0.005
0.047
0.030
0.061
0.047
Elevatio
n(m)
84.4
69.5
805.9
620.0
310.7
522.0
682.7
358.4
343.9
123.6
435.3
429.8
710.1
418.8
453.7
581.1
928
968.1
578.8
Percentage
Forest (%)
57.9
61.8
65.9
84.3
100.0
100.0
83.6
96.8
80.9
91.0
93.2
93.1
84.8
82.0
90.6
96
97.4
96.1
-------
Table C-3. continued...
Longitude
-79.34634
-80.33483
-81.75611
-79.60111
-79.56808
-79.67617
-79.48686
-80.30063
Latitude
38.32267
36.80553
36.62583
38.74322
38.62673
38.61844
38.84942
38.23512
State
VA
VA
VA
WV
WV
WV
WV
WV
Entity
VDGIF
VDEQ
TVA
WVDEP
WVDEP
WVDEP
WVDEP
WVDEP
Station ID
2-RAM007.29
4ARCC008.86
TVA_Whitetop
3593
6112
2571
8756
2039
Water body name
Ramseys Draft
Rock Castle Creek
Whitetop Laurel Creek
Big Run
Big Run
East Fork/Greenbrier River
Seneca Creek
South Fork/Cranberry River
Drainage
area
(km2)
20.0
20.6
145.3
10.4
36.0
28.0
42.5
36.3
Slope
(unitless)
0.020
0.020
0.012
0.031
0.027
0.011
0.024
0.004
Elevatio
n(m)
868.7
562.5
790.0
1,099.0
930.9
1,078.6
873.8
1,143.6
Percentage
Fores (%)t
94.0
90.0
91.1
98.3
96.3
93.5
98.3
97.5
O ahttp://www.horizon-systems.com/nhdplus/nhdplusvl_home.php
^ bhttp://www.mrlc.gov/nlcdO l_data.php
°http ://www. mrlc. gov/nlcd2006 .php
-------
Table C-4. Equipment installed at primary RMN sites in the Mid-Atlantic (3/2015)
State
DE
DE
MD
MD
MD
MD
MD
NJ
NJ
NJ
PA
PA
PA
PA
PA
VA
Entity
DNREC
DNREC
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
NJ DEP/
EPAR2
NJDEP
NJDEP
PA DEP
SRBC
PA DEP
SRBC
PA DEP
VDEQ
Station ID
105212
105213
YOUG-432-S
SAVA-204-S
UMON-288-S
PRLN-626-S
SAVA-225-S
AN0012
AN0260
AN0215A
CR
SRBC_Grays
JMRAVQN_734
SRBC_Kettle
WBCAVQN_873
2-HUO005.87
Water body
name
Trib White Clay
Trib White Clay
Bear Creek
Crabtree Creek
High Run
Mill Run
Savage River
Dunnfield Creek
Mossmans Brook
Primrose
Carbaugh Run
Grays Run
Jones Mill Run
Kettle
West Branch of
Caldwell Creek
Hunting Creek
Temperature
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
Hydrologic
equipment
USGS gage
(01597000)
USGS gage
(01596500)
USGS staff gage
(01378780)
pressure
transducer
pressure
transducer
Hydrologic
data type
discharge
stage
stage
discharge
occasional
stage
stage
stage
Notes
planning to install water and air temperature
sensors and pressure transducers in 2015
USGS gage (03076600) downstream of site; about
nine tributaries (including a major one) enter
between gage and site
gage is downstream of site but location looks
representative of stream conditions
applied for a grant to get a USGS gage installed
here
applied for a grant to get a USGS gage installed
here
Will try to install pressure transducer in 2015
Working on flow rating curve
Working on flow rating curve
p
oo
-------
Table C-4. continued...
State
VA
VA
VA
VA
VA
VA
WV
WV
WV
WV
WV
Entity
ShenNP
VDEQ
ShenNP
VDGIF
VDEQ
TVA
WVDEP
WVDEP
WVDEP
WVDEP
WVDEP
Station ID
3-RAP088.21
9-LRY007.02
3-PIY003.27
2-RAM007.29
4ARCC008.86
TVA_Whitetop
3593
6112
2571
8756
2039
Water body
name
Rapidan River
(upper)
Little Stony
Creek
Piney River
Ramseys Draft
Rock Castle
Creek
Whitetop Laurel
Creek
Big Run
Big Run
East
Fork/Greenbrier
River
Seneca Creek
South
Fork/Cranberry
River
Temperature
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
Hydrologic
equipment
gage
pressure
transducer
pressure
transducer
pressure
transducer
Hydrologic
data type
stage
stage
stage
Notes
Flow nearby though in another drainage (gage in
Staunton R)
planning to install pressure transducer in 2015
planning to install pressure transducer in 2015
planning to install pressure transducer in 2015
-------
Table C-5. Site information for primary RMN sites in the Southeast (3/2015). Most drainage areas are estimates
based on NHDPlus vla local catchment data. Where available, data from exact watershed delineations were
used. Slope and elevation are estimated based on NHDPlus vl local catchment data. Percentage forest is derived
from the NLCD 2001b data layer and is based on the total watershed
Longitude
-87.2862
-86.1330
-87.3991
-83.5716
-83.5166
-84.3851
-84.1512
-83.9039
-83.1924
-82.9940
-82.7916
-82.1014
-83.0728
-82.8089
-80.0303
-81.5672
-79.9906
-83.8552
-83.0793
-82.6477
Latitude
34.3307
34.9180
34.2856
34.9590
34.9520
34.9851
34.6020
37.4550
38.1311
37.0774
37.0666
35.7347
35.6672
35.2281
35.3792
35.5906
36.5355
35.3094
34.9235
35.0642
State
AL
AL
AL
GA
GA
GA
GA
KY
KY
KY
KY
NC
NC
NC
NC
NC
NC
NC
sc
sc
Entity
ALDEM
ALDEM
ALDEM
GADNR
GADNR
TVA
GADNR
KYDEP
KYDEP
KYDEP
KYDEP
NCDENR
NC DENR/TVA
NC/DENR/TVA
NCDENR
NCDENR
NCDENR
TVA
SC DHEC
SC DHEC
Station ID
BRSL-3
HURR-2
SF-1
66d-WRD768
66d-44-2
3890-1
66g-WRD773
DOW04036022
DOW06013017
DOW04055002
DOW02046004
CB6
EB320
EB372
QB283
CB192
NB28
10605-2
SV-684
S-086
Water body name
Brushy Creek
Hurricane Creek
Sipsey Fork
Charlies Creek
Coleman River
Fightingtown Creek
Jones Creek
Hughes Fork
Laurel Creek
Line Fork UT
Presley House Branch
Buck Creek
Cataloochee Creek
Cedar Rock Creek
Dutchmans Creek
Jacob Fork
Mayo River
Snowbird Creek
Crane Creek
Matthews Creek
Drainage
area (km2)
23.6
102.6
231.8
7.2
13.6
182.9
9.1
3.5
37.8
0.6
3.0
37.5
127.0
3.1
9.1
66.5
626.8
108.8
4.0
25.8
Slope
(unitless)
0.002
0.000
0.000
0.040
0.033
0.003
0.011
0.019
0.002
NA
0.093
0.011
0.010
0.042
0.014
0.001
0.010
0.007
0.078
0.003
Elevation
(m)
240.8
297.07
204.6
927.0
866.9
468.8
586.0
359.1
294.3
335.6
736.6
529.7
939.2
985.9
177.5
380.1
254.9
677.8
623.6
360.2
Percentage
Forest (%)
96.9
93.5
95.5
99.0
96.8
86.8
98.4
86.6
72.9
100.0
97.0
96.6
99.0
98.6
92.2
89.4
73.4
97.1
97.0
96.3
o
o
-------
Table C-5. continued...
Longitude
-82.5739
-82.2515
-87.5355
-84.1182
-82.5291
-84.0597
-85.9921
-85.9111
-82.9381
-87.7614
-84.6981
-83.5773
-84.9827
-84.4803
-84.6122
-83.8917
-82.9456
Latitude
35.1254
35.1831
35.4217
35.4548
36.1508
36.2136
35.9286
35.1155
36.5001
35.9806
36.5161
35.6533
36.1299
35.0539
35.0031
36.3436
35.9224
State
SC
sc
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
Entity
SC DHEC
SC DHEC
TNDEC
TVA
TNDEC
TNDEC
TNDEC
TNDEC
TNDEC/TVA
TNDEC
TNDEC
TNDEC
TNDEC/TVA
TNDEC
TNDEC
TNDEC
TVA
Station ID
S-076
B-099-7
ECO71F19
CITIC011.0MO
ECO66E09
ECO67F06
ECO71H17
ECO68C20
ECO6702
ECO71F29
ECO68A03
ECO66G05
MYATT005.1CU
ECO66G20
ECO66G12
ECO67F13
12358-1
Water body name
Middle Saluda River
Vaughn Creek
Brush Creek
Citico Creek
Clark Creek
Clear Creek
Clear Fork Creek
Crow Creek
Fisher Creek
Hurricane Creek
Laurel Fork Station
Camp Creek
Little River
Myatt Creek
Rough Creek
Sheeds Creek
White Creek
Wolf Creek
Drainage
area (km2)
16.0
12.0
33.3
118.1
23.8
7.2
38.1
47.7
30.0
177.6
15.3
81.2
12.4
15.5
14.8
8.0
28.5
Slope
(unitless)
0.042
0.008
0.004
0.010
0.017
0.014
0.005
0.006
0.003
0.003
0.014
0.029
0.016
0.020
0.031
0.009
0.014
Elevation
(m)
582.3
368.4
245.1
399.0
596.6
337.1
262.9
311.5
429.7
156.3
392.9
879.5
525.1
520.6
436.6
379.8
429.9
Percentage
Forest (%)
96.6
95.6
75.8
97.2
95.1
87.9
88.8
84.5
82.0
81.0
97.2
99.8
78.8
98.9
98.8
90.9
96.0
o
ahttp://www.horizon-systems.com/nhdplus/nhdplusvl_home.php
bhttp://www.mrlc.gov/nlcd01_data.php
-------
Table C-6. Equipment installed at primary RMN sites in the Southeast (3/2015). EPA R4 is planning to install
equipment at the sites in North and South Carolina as resources permit
State
AL
AL
AL
GA
GA
GA
GA
KY
KY
KY
KY
NC
NC
NC
NC
NC
NC
Entity
ALDEM
ALDEM
ALDEM
GADNR
GADNR
TVA
GADNR
KYDEP
KYDEP
KYDEP
KYDEP
NCDENR
TVA
NCDENR
NCDENR
NCDENR
NCDENR
Station ID
BRSL-3
HURR-2
SF-1
66d-WRD768
66d-44-2
3890-1
66g-WRD773
DOW04036022
DOW06013017
DOW04055002
DOW02046004
CB6
EB320
EB372
QB283
CB192
NB28
Water body name
Brushy Creek
Hurricane Creek
Sipsey Fork
Charlies Creek
Coleman River
Fightingtown Creek
Jones Creek
Hughes Fork
Laurel Creek
Line Fork UT
Presley House
Branch
Buck Creek
Cataloochee Creek
Cedar Rock Creek
Dutchmans Creek
Jacob Fork
Mayo River
Temperature
water and air
water and air
water
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
none
water
none
none
none
none
Hydrologic
equipment
pressure transducer
pressure transducer
USGS gage
(02450250)
pressure transducer
pressure transducer
pressure transducer
pressure transducer
pressure transducer
pressure transducer
pressure transducer
pressure transducer
none
USGS gage
(03460000)
none
none
USGS gage
(02143040)
USGS gage
(02070500)
Hydrologic
data type
stage
stage
discharge
stage
stage
stage
stage
stage
stage
stage
stage
none
discharge
none
none
discharge
discharge
Notes
water temperature is being measured at the
USGS gage
Inactive USGS gage (03560000)
water temperature is being measured at the
USGS gage
USGS gage downstream on Catheys Creek
(03440000)
inactive USGS gage (02123567)
precip is being measured at the USGS gage
o
to
-------
Table C-6. continued...
State
NC
sc
sc
sc
sc
TN
TN
TN
TN
TN
TN
TN
TN
TN
Entity
TVA
SC DHEC
SC DHEC
SC DHEC
SC DHEC
TNDEC
TVA
TNDEC
TNDEC
TNDEC
TNDEC
TN
DEC/TVA
TNDEC
TNDEC
Station ID
10605-2
SV-684
S-086
S-076
B-099-7
ECO71F19
CITIC011.0MO
ECO66E09
ECO67F06
ECO71H17
ECO68C20
ECO6702
ECO71F29
ECO68A03
Water body name
Snowbird Creek
Crane Creek
Matthews Creek
Middle Saluda
River
Vaughn Creek
Brush Creek
Citico Creek
Clark Creek
Clear Creek
Clear Fork Creek
Crow Creek
Fisher Creek
Hurricane Creek
Laurel Fork Station
Camp Creek
Temperature
water and air
none
none
none
none
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
water and air
Hydrologic
equipment
pressure
transducer
none
none
none
none
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
Hydrologic
data type
stage
none
none
none
none
stage
stage
stage
stage
stage
stage
stage
stage
stage
Notes
inactive USGS gage (03516000)
USGS gage (02162350) downstream of site but
unsure whether it is representative (some major
tributaries enter between site and gage); EPA
R4 will install equipment as resources permit
o
-------
Table C-6. continued...
State
TN
TN
TN
TN
TN
TN
Entity
TNDEC
TN
DEC/TVA
TNDEC
TNDEC
TNDEC
TVA
Station ID
ECO66G05
MYATT005.1CU
ECO66G20
ECO66G12
ECO67F13
12358-1
Water body name
Little River
Myatt Creek
Rough Creek
Sheeds Creek
White Creek
Wolf Creek
Temperature
water and air
water and air
water and air
water and air
water and air
water and air
Hydrologic
equipment
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
pressure
transducer
Hydrologic
data type
stage
stage
stage
stage
stage
stage
Notes
o
-------
APPENDIX D.
DISTURBANCE SCREENING
PROCEDURE FOR RMN SITES
D-l
-------
D.I. BACKGROUND
This project developed a screening procedure for candidate regional monitoring network
(RMN) sites to determine where the sites fall along a standardized disturbance gradient, using
data that are available nationwide and that are derived using common data sources and
methodologies. The first iteration of the screening process was developed during the Northeast
RMN pilot study in 2012, when over 900 sites were screened. Additional screening
considerations related to coal mining, shale gas drilling, atmospheric deposition and other
stressors were added during the Mid-Atlantic RMN site selection process.
The screening processes that were performed during the development of the pilot RMNs
have limitations. For one, there were insufficient resources to do exact watershed delineations for
all of the candidate sites. Instead, the land use data for many of the candidate sites were based on
data associated with the National Hydrography Dataset Plus Version 1 (NHDPlusVl) catchments
in which the sites are located (U.S. EPA, 2005. While this approach generally provides a good
approximation, sometimes there are discrepancies, which are described in Section D.2.1. To
address this issue, additional checks were performed to verify the accuracy of the results before
finalizing site selection (e.g., local experts who were familiar with the sites verified that results
were in keeping with their expectations). In the future, as resources permit, the RMN site
screening process will be further refined, and exact watershed delineations will be done for all of
the RMN sites.
D.2. METHODOLOGY
Candidate RMN sites were spatially joined with NHDPlusVl catchments (U.S. EPA and
USGS, 2005) using Geographic Information System software (ArcGIS 10.0). Each NHDPlusVl
catchment has a unique identifier called a COMID. Many data were linked to sites via this
COMID.
Three different types of disturbance screenings were performed:
1. Land use (see Section D.2.1);
2. Likelihood of impact from dams, mines, and point-source pollution sites (see
Section D.2.2); and
3. Likelihood of impact by the following other nonclimatic stressors:
• Roads (see Section D.2.3.1),
• Atmospheric deposition (see Section D.2.3.2),
• Coal (see Section D.2.3.3),
• Shale gas (see Section D.2.3.4),
• Future urban development (see Section D.2.3.5), and/or
• Water withdrawals (see Section D.2.3.6).
D-2
-------
These considerations are consistent with recent work performed by Michigan State
University (MSU) on the National Fish Habitat Action Plan (NFHAP; DFW MSU et al., 2011;
Esselman et al., 201 la). That work included the development of the cumulative disturbance
index (DFW MSU et al., 2011; Esselman et al., 201 Ib).
D.2.1. Land Use Disturbance
The first set of screening was done on land use and impervious cover data from the 2001
National Land Cover Database (NLCD) version 1 data set (Homer et al., 2007). The land use
disturbance screening was conducted at both the local catchment and total watershed scales
[important note: for purposes of this exercise, the total watershed scale will be referred to as the
"network" scale, in keepingwith the work done by DFW MSU et al. (2011)]. Local catchments
are defined as the land area draining directly to a reach, and network catchments are defined by
all upstream contributing catchments to the reach's outlet, including the reach's own local
catchment (see Figure D-l). GIS shapefiles with delineations of the local catchments were
downloaded from the Horizon-Systems website: http://www.horizon-
systems.com/NHDPlus/NHDPlusVl_data.php. The network-scale data were generated (and
graciously shared) by MSU.
A Local Catchments for B. Total Upstream Watershed
Reaches 20, 21, and 22 for Reach 20
Figure D-l. Land use data were evaluated at both the (A) local catchment
and (B) total watershed scales, using NHDPlusVl delineations (U.S. EPA and
USGS, 2005).
While these data generally provide good approximations of land use, they have
limitations. For one, there are biases and accuracy issues associated with the NLCD data set
(e.g., Novak and Greenfield, 2010; Wickham et al., 2013). Another limitation is that there is no
information on whether landscape disturbance mitigation measures are being applied in a given
D-3
-------
catchment, and if so, how effective those measures are. Thus, we had to assume that the impacts
associated with each land use type are equal.
As mentioned earlier, the preliminary land use screening was not based on exact
watershed delineations. Rather the data were associated with the entire catchment in which the
site is located, regardless of where the site falls within the catchment (it would have been
preferable to use data based on exact watershed delineations, but we lacked the resources needed
to do exact watershed delineations for all of the candidate sites). The estimates that were used
were readily available for all of the sites and generally provide a good approximation (especially
when sites are located at the downstream end of the catchment). However, sometimes
inaccuracies occur. An example is illustrated in Figure D-2. Maryland site UMON-288-S is
located about halfway up the catchment flowline. Urban and agricultural land uses are located
within this catchment, but are all downstream of the site. Because these land uses are in the
catchment, they are included in the land cover output for this site. An accurate output for that site
would only include forested land cover. To catch errors like these, an additional series of checks
were performed on the top candidate sites (e.g., we asked local experts to verify that results were
in keeping with their expectations and also performed additional desktop screening using aerial
photos from Google Earth).
D-4
-------
NHDPlusVl local
catchment delineation
Site UMON-288-S
Even though the
urban and agricultural
land use (color-coded
in pink, red, yellow
and brown) are
located downstream
from this site, they
are included in the
land cover output
associated with this
site; thus, we are
performing visual
checks and soliciting
feedback from
entities to ensure that
these land cover
estimates match with
expectations.
Figure D-2. Example of a situation in which the land use output for a site is
inaccurate.
Land use disturbance was assessed at both the local catchment and network scales. This
was done for the following four parameters (source: NLCD 2001 version 1 data set1):
1. Percentage impervious cover
2. Percentage urban (this includes low, medium, and high intensity developed—NLCD
codes 22 + 23 + 24)
3. Percentage cultivated crops (NLCD code 82)
4. Percentage pasture/hay (NLCD code 81)
The project developed a land use disturbance scale with six levels. Thresholds for each
parameter, which are guided by literature where available (e.g., King and Baker, 2010; Carlisle
et al., 2008), are listed in Table D-l. When rating a site, each parameter was first assessed
1http://www.mrlc.gov/nlcd01_data.php
D-5
-------
separately. If the parameter values at the local catchment and network scales differed, the
thresholds were applied to the maximum value. For example, if a site had 2% urban land cover at
the local catchment scale and 1% urban land cover at the network scale, the threshold was
applied to the maximum value (in this case, 2% or level 3 for urban land use). This was done for
each parameter. Then, sites were assigned an overall disturbance level. This was based on the
highest disturbance level assigned across parameters. For example, if a site was level 3 for
impervious, level 2 for urban, level 1 for crops, and level 2 for pasture/hay, it was assigned to
disturbance level 3. As a final step, local biologists familiar with the sites were consulted to find
out their thoughts on the degree of land use impact and also to inquire about the availability of
more detailed land use data to help better assess the potential degree of impact.
Table D-l. The thresholds used when assigning sites to the six levels of land
use disturbance. Each of the four parameters (impervious, urban, crops,
pasture/hay) were assessed separately. Then, sites were assigned an overall
disturbance level based on the highest level of disturbance across parameters
Level of land use disturbance
1
2
3
4
5
6
Percentage
Impervious (%)
<0.1
<1
<2
<5
<10
>10
Percentage
Urban (%)
0
<1
<3
<5
<10
>10
Percentage
Crops (%)
0
<1
<5
<15
<25
>25
Percentage
Pasture/hay (%)
0
<5
<15
<25
<35
>35
D.2.2. Likelihood of Impacts from Dams, Mines, and Point-source Pollution Sites
In the second set of screening, sites were flagged if they had a high likelihood of being
impacted by dams, mines, National Pollutant Discharge Elimination System (NPDES) major
discharges and/or Superfund National Priorities List (SNPL) sites. Both the proximity of these
stressors to the sites as well as the attribute data associated with each stressor were considered.
The attribute data are important because there are many site-specific factors, such as dam size
and storage capacity, that can greatly affect the degree of impact. Table D-2 contains a list of
data that were assessed, along with the sources of those data.
The following screening procedures were performed:
1. The data listed in Table D-2 were gathered.
D-6
-------
2. Using GIS software (ArcGIS 10.0), a 1-km buffer was created around the preliminary
RMN sites (this included both the upstream and downstream areas).
3. Using GIS software (ArcGIS 10.0), a procedure was performed to identify whether any
dams, mines, NPDES major discharges or SNPL sites were located within the 1-km
buffer.
4. If so, those sites were flagged and the likelihood of impact based on the following
considerations was assessed:
a. Location in relation to the site, assessed via a desktop screening with GIS software
(ArcGIS 10.0) and Google Earth.
b. Attributes of the stressors (e.g., dam size, storage capacity, size of NPDES major
discharge).
Best professional judgment was used to assign the flagged sites to one of three impact
categories:
• Unlikely impacted
• Likely impacted
• Unsure
Some examples of situations in which sites were assigned to the "unlikely impacted"
category are:
The site was flagged for an NPDES major discharge, but the discharge was relatively
small and was located hundreds of meters downstream from the site.
The site was flagged for a dam, but the dam was located on a different stream.
Some examples of situations in which sites were assigned to the "likely impacted"
category are:
• The site was flagged for a NPDES major discharge. It was a large discharge occurring
about 100 m upstream from the site.
• The site was flagged for a dam. It was a large dam located on the same stream, just
upstream from the site.
D-7
-------
Some examples of situations in which sites were assigned to the "unsure" category are:
• The site was flagged for a NPDES major discharge, but the site was located near a
confluence and it was difficult to determine which stream contained the discharge.
• The stressor was small- or medium-sized and was located 500 m or more from the site.
One additional check was performed to assess the potential for flow alteration at the sites.
We examined the type of NHDPlusVl flowline (FTYPE) located on the site (e.g., stream/river,
artificial pathway, canal/ditch, pipeline, connector; U.S. EPA and USGS, 2005). If the site was
located on a flowline designated as something other than a stream/river, the site was flagged.
As a final step, local biologists familiar with the sites were consulted to find out their
thoughts on the degree of impact and also to inquire about the availability of more detailed data
to help better assess the potential degree of impact.
D.2.3. Likelihood of impact from other nonclimatic stressors
In the third set of screening, sites were flagged if they had a high likelihood of being
impacted by:
Roads,
Atmospheric deposition,
Coal mining,
Shale gas drilling,
Future urban development, and/or
Water withdrawals.
D-8
-------
Table D-2. These data were assessed when screening for the likelihood of
impacts from flow alteration, mines, National Pollutant Discharge
Elimination System (NPDES) major discharges, and/or Superfund National
Priorities List (SNPL) sites
Stressor
Source
Dams
USGS (U.S. Geological Survey). (2014) Major dams
of the United States: National Atlas of the United
States. Reston, VA: United States Geological
Survey. Available online at:
http: //nati onal atl as. gov/atl asftp. html#dam sOOx
Mines
USGS (U.S. Geological Survey). (2005) Active
mines and mineral processing plants in the United
States in 2003. Reston, VA: USGS. Available online
at http://tin.er.usgs.gov/metadata/mineplant.faq.html
PASDA (Pennsylvania Spatial Data Access). (2013)
Data download -mine and refuse permits. Available
online at www.pasda.psu.edu.
National Pollutant Discharge Elimination
System (NPDES) major discharges from
the Permit Compliance System
Superfund National Priorities List
(SNPL) from the Compensation and
Liability Information System
U.S. EPA (U.S. Environmental Protection Agency).
(2013) Geospatial data download service -
Geospatial information for all publicly available
FRS facilities that have latitude/longitude data [file
geodatabase]. Available online at
http ://www. epa. gov/enviro/geo_data. html
Table D-3 contains a list of data that were gathered and assessed, along with the sources
of those data. There are a lot of site-specific factors that can greatly affect the degree of impact
from these stressors, which makes it difficult to set thresholds. For example, a site could be
exposed to high concentrations of atmospheric deposition but may not be impacted by acidity
because of site-specific mediating factors like calcareous geology. Another example is permit
activity associated with coal mining. Just because mining permits have been issued in an area
does not mean that mining activities are actually taking place. And even if mining activities are
taking place, impacts can vary greatly depending on site-specific factors such as the size and type
of mine.
D-9
-------
Table D-3. These data were assessed when screening for the likelihood of
impacts from roads, atmospheric deposition, coal mining, shale gas drilling,
future urban development, and water withdrawals. These data were
converted to relative measures, as described in the text.
Stressor
Parameters/description
Source
Roads
Length of roads, local catchment, and network
scales/catchment area
U.S. Census Bureau (2000)
from DFW MSU et al. (2011)
Number of road crossings, local catchment,
and network scales/catchment area
Atmospheric
deposition
s and SO4 concentrations, based on 2011
deposition grids
NADPa
The Nature Conservancy (TNC) geology class
Olivero and Anderson (2008)
Coal mining
Potential for development, based on:
• whether the site is located in a coal
field and/or the mountaintop removal
(MTR) region
• coal production by state
Coal fields (USGS, 2001)
MTR region [unknown source;
GIS layer was provided by
Christine Mazzarella
(U.S. EPA)]
Coal production by State [see
Table 6 in U.S. EIA(2012)]
Permit activity, based on number of permits
issued within 1 km of the site. Data type and
availability varied by state.
Alabama:
• Number of active coal mine permits
Pennsylvania:
• Anthracite permits
• Anthracite refuse
• Bituminous permits
• Bituminous refuse
West Virginia:
• WV_permitboundary
• WV_refuse
• WV_valleyfill
• WV_all_mining
Virginia:
• Surface mine permit boundaries
Alabama (Alabama Surface
Mining Commission, 2013)
Pennsylvania (PASDA, 2013)
West Virginia (WVDEP, 2013;
WVGES, 2014)
Virginia (VDEQ, 2013)
D-10
-------
Table D-3. continued...
Stressor
Shale gas
drilling
Future urban
development
Water
withdrawals
(county-level)
Parameters/description
Potential for development, based on whether
the site is located in the shale play region
Permit activity, based on the number of
unconventional permits issued within 1 km of
the site. These data were available for
Pennsylvania (file name:
PA UncPermits 05092013) and West Virginia
(file name: WV_Perm_05132013).
Potential for future urban development based
on projected change in percentage
imperviousness by 2050
Irrigation, total withdrawals, fresh (Mgal/day)
Total withdrawals, fresh (Mgal/day)
Total withdrawals, total (fresh + saline;
Mgal/day)
Source
U.S. EIA (2013)
Frac Tracker (20 13)
U.S. EPA (2011);
work performed by Angie
Murdukhayeva (U.S. EPA)
USGS (2010)
ahttp://nadp.sws.uiuc.edu/NTN/annualmapsbyyear.aspx
Because of these factors, a relative scale (vs. firm thresholds) was used to assess the
likelihood of impact. The relative scales were based on values found in NHDPlusVl catchments
across the entire study area. If a site had an elevated risk score (e.g., >75th percentile), it was
flagged for further evaluation (the thresholds [e.g., 50%, 75%] vary depending on the
distribution of data in each data set and best professional judgment). These thresholds should be
regarded as a starting point, and should be further refined in future iterations of the screening
procedure. Local biologists familiar with the sites were then consulted to find out their thoughts
on the degree of impact and also to inquire about the availability of more detailed data to help
better assess the potential degree of impact [e.g., is mining actually taking place? What are the
pH and acid neutralizing capacity (ANC) values at sites flagged for atmospheric deposition?].
The specific screening procedures that were followed for each stressor are described below.
D.2.3.1. Roads
Two aspects of potential road impacts were assessed:
• Length of roads and
• Number of road crossings
D-ll
-------
First the roads data listed in Table D-3 were gathered for both the local catchment and
network scales.
Next, to assess the likelihood of impact from length of roads, the following formulas
were used to normalize the data:
Local catchment scale = Length of roads in the local catchment (m) ^ Area of the local
catchment (km2)
Network scale = Length of roads in the network (m) + Area of the network
(km2)
Then, the following formula was used to convert these values to a scoring scale ranging
from 0 (no roads) to 100 (highest length of roads per area; note: the minimum and maximum
values used in this formula are based on the range of values found across the entire study area):
100 x (Value - Minimum) + (Maximum - Minimum)
If the parameter values at the local catchment and network scales differed, the maximum
score was used for the assessment. For example, if the local catchment score was 80 and the
network score was 50, the higher score of 80 was used for the assessment.
Sites were flagged for further evaluation if they received a score of >75%.
The same procedure was followed when assessing the likelihood of impact from road
crossings.
As a final step, we consulted with local biologists who were familiar with the sites to get
input on the degree of impact at flagged sites.
D.2.3.2. Atmospheric Deposition
Two aspects of atmospheric deposition were assessed:
• Concentrations of NOs
• Concentrations of 864
In addition, TNC geology class (Olivero and Anderson, 2008) was considered as a
potential mediating factor. First the data listed in Table C-3 were gathered. Using GIS software
(ArcGIS 10.0), the NCb and 864 deposition grid data (1-km resolution) were linked to the sites.
Next, the NCb and SO4 values were averaged. Then, the following formula was used to convert
D-12
-------
these values to a scoring scale ranging from 0 (no nitrogen and sulfate deposition) to 100
(highest average concentration of NCb and SO4; note: the minimum and maximum values used in
this formula are based on the range of values found across the entire study area):
100 x (Value - Minimum) + (Maximum - Minimum)
Sites were flagged for further evaluation if they received a score of >75%.
Geology can potentially mediate some of the effects of atmospheric deposition. To assess
this potential, GIS software (ArcGIS 10.0) was used to link the TNC geology class (Olivero and
Anderson, 2008) to the sites (note: at this time the TNC geology class data are only available for
Northeast and Mid-Atlantic regions).
Sites were scored as follows:
• Sites located in areas designated as "low buffered, acidic" received a score of 100.
• Sites located in areas designated as "moderately buffered, neutral" or "assume
moderately buffered (Size 3+ rivers)" received a score of 50.
• Sites located in areas designated as "highly buffered, calcareous" received a score of 0.
• Sites located in areas that lacked data or were designated as "unknown buffering/missing
geology" were not assessed.
Sites were flagged if they received a score of 100%.
As a final step, we consulted with local biologists who were familiar with the sites to get
input on the degree of impact at flagged sites, and to see if they had access to more detailed data,
such as pH and ANC measurements, to help us better assess the potential degree of impact.
D.2.3.3. Coal Mining
Two aspects of coal mining were assessed:
• Potential for mining
• Permit activity
First the data listed in Table D-3 were gathered.
To assess the potential for coal mining, the following were considered:
D-13
-------
• Whether the site is located in an area that has been designated as a mountaintop removal
(MTR) area and/or a coal field (USGS, 2001).
o If the site is located in a coal field, is it designated as "potentially minable" or is it
tagged for "other uses"?
• What the total coal production is for the state where the site is located [source: Table 6 in
the 2011 Annual Coal Report (U.S. EIA, 2012)].
The following steps were performed when assessing a site for mining potential:
1. First a coal field score was assigned, as follows:
• Using GIS software (ArcGIS 10.0), the coal field and MTR GIS layers were linked to
the sites.
• If the site is located in a catchment that has been designated as a "potentially
minable" coal field (USGS, 2001) and/or a mountaintop removal (MTR) area, it was
assigned a score of 1.
• If the site is located in a catchment that has been designated as a coal field with "other
uses" (USGS, 2001), it was assigned a score of 0.5.
• If the site is located in a catchment that is not part of a coal field or MTR area, it
received a score of 0.
2. Then a coal production score was assigned, as follows:
• Total coal production values for each state were taken from Table 6 in the 2011
Annual Coal Report (U.S. EIA, 2012).
• Those values were converted to a scale of 0 to 100 using this formula (note: the
minimum and maximum values used in this formula are based on the range of values
found in the states in our study area):
100 x (Value - Minimum) + (Maximum - Minimum)
• Sites were assigned scores based on what state they were located in. For example,
West Virginia had the highest total coal production of all of the states in the study
area, so any sites located in West Virginia received a coal production score of 100.
D-14
-------
3. To get the final score for mining potential, the coal field score was multiplied by the
coal production score. Scores ranged from 0 (no mining potential) to 100 (highest
potential for mining).
Sites were flagged for further evaluation if they received a score of >75%.
Permit data were not available for all the states, and where those data were available, data
type and quality varied, as did the attribute data. Therefore, permit activity was assessed on a
state-by-state basis. If sites were located in states where permit data were available, the following
steps were performed to assess the intensity of permit activity:
1. Gather the permit data listed in Table D-3.
2. Using GIS software (ArcGIS 10.0), create a 1-km buffer around the candidate RMN sites
(this included both the upstream and downstream areas).
3. Using GIS software (ArcGIS 10.0), perform a procedure to determine how many mining
permits had been issued within the 1-km buffer.
4. The following formula was used to convert those values to a scale of 0 to 100 (note: since
the type of data available for each state varied, the minimum and maximum values used
in this formula were based on the range of data found in each state):
100 x (Value - Minimum) + (Maximum - Minimum)
Sites were flagged for further evaluation if they received a score of >0.
As a final step, we consulted with local biologists who were familiar with the sites to get
input on the degree of impact at flagged sites, and to see if they had access to more detailed data
to help us better assess the potential degree of impact. Just because mining permits have been
issued in an area does not mean that mining activities are actually taking place. And even if
mining activities are taking place, impacts can vary greatly depending on site-specific factors
such as the size and type of mine.
D.2.3.4. Shale Gas Drilling
Two aspects of shall gas drilling were assessed:
Potential for drilling
Permit activity
D-15
-------
First the data listed in Table D-3 were gathered.
To assess the potential for shale gas drilling, the following screening procedure was
performed:
• Using GIS software (ArcGIS 10.0), the shale play GIS layer (see Table D-3) was linked
to the sites.
• If the site is located in a shale play region, it was assigned a score of 100 and flagged it
for further evaluation.
Permit data were only available for the states of West Virginia and Pennsylvania. The
following steps were performed to assess the intensity of permit activity at sites in those sites:
1. Gather the permit data listed in Table D-3.
2. Using GIS software (ArcGIS 10.0), create a 1-km buffer around the candidate RMN sites
(this included both the upstream and downstream areas).
3. Using GIS software (ArcGIS 10.0), perform a procedure to determine how many
unconventional permits had been issued within the 1-km buffer.
4. The following formula was used to convert those values to a scale of 0 to 100 (note: since
the type of data available for each state varied, the minimum and maximum values used
in this formula were based on the range of data found in each state):
100 x (Value - Minimum) + (Maximum - Minimum)
Sites were flagged for further evaluation if they received a score of >0%.
As a final step, we consulted with local biologists who were familiar with the sites to get
input on the degree of impact at flagged sites, and to see if they had access to more detailed data
to help us better assess the potential degree of impact. Just because drilling permits have been
issued in an area does not mean that drilling activities are actually taking place. And even if
drilling activities are taking place, impacts can vary greatly depending on site-specific factors.
D.2.3.5. Potential for Future Urban Development
EPA's ICLUS tools and data sets (Version 1.3 and 1.3.1; U.S. EPA, 2011) were used to
assess the potential that a site will experience future urban development. The ICLUS Tools were
used to project the percentage change in imperviousness in each NHDPlusVl local catchment by
D-16
-------
2050 based on high (A2) and low (Bl) emissions scenarios (note: the ICLUS data have a
resolution of 1-km).
GIS software (ArcGIS 10.0) was used to link sites with NHDPlusVl local catchments.
Sites were flagged for further evaluation if the following conditions occurred:
• The percentage impervious value in the NHDPlusVl local catchment where the site is
located is currently <10% (based on values derived from the 2001 NLCD version 1 data
set), and
• The future projection is for a positive value >0.5% (this is based on an average of the
high [A2] and low [Bl] emissions scenarios).
As a final step, we checked with local biologists who were familiar with the sites to find
out their thoughts on the potential for future development at flagged sites, and to see if they had
access to more detailed data to help us better assess the potential degree of impact.
D.2.3.6. Water Withdrawals
Three aspects of water use were assessed:
• Irrigation, total withdrawals, fresh;
• Total withdrawals, fresh only; and
• Total withdrawals, total.
First the data listed in Table D-3 were gathered. These data are based on 2005 water use
and are only available at the county-level (USGS, 2010). Then GIS software (ArcGIS 10.0) were
used to associate the county-level data with NHDPlusVl local catchments. Next sites were
linked with NHDPlusVl local catchments. For each parameter, the following formula was used
to convert the values to a scoring scale ranging from 0 (no withdrawals) to 100 (highest
withdrawals; note: the minimum and maximum values used in this formula are based on the
range of values found across the entire study area):
100 x (Value - Minimum) + (Maximum - Minimum)
Sites were flagged for further evaluation if they received a score of >50% for any of the
three parameters.
D-17
-------
As a final step, we checked with local biologists who were familiar with the sites to find
out their thoughts on the potential for future development at flagged sites, and to see if they had
access to more detailed data to help us better assess the potential degree of impact.
D.3. REFERENCES
Alabama Surface Mining Commission. (2013) Alabama coal mine geospatial data - permit boundaries (1983-
present). Jasper, AL: State of Alabama http://surface-mining.alabama.gov/.
Carlisle, DM; Hawkins, CP; Meador, MR; Potapova, M; Falcone, J. (2008) Biological assessments of Appalachian
streams based on predictive models for fish, macroinvertebrate, and diatom assemblages. J N Am Benthol
Soc 27(1): 6-37.
DFW MSU (Department of Fisheries and Wildlife, Michigan State University), Esselman, PC; Infante, DM; Wang,
L; Taylor, WW; Daniel, WM; Tingley, R; Fenner, J; Cooper, A; Wieferich, D; Thornbrugh, D; Ross, J.
(2011) National Fish Habitat Action Plan (NFHAP) 2010 HCI scores and human disturbance data for
conterminous United States linked to NHDPLUSV1. Denver, CO: National Fish Habitat Action Plan
(NFHAP). Available online athttps://www.sciencebase.gov/catalog/item/514afb90e4b0040b38150dbc
Esselman, PC; Infante, DM; Wang, L; Taylor, WW; Daniel, WM; Tingley, R; Fenner, J; Cooper, C; Wieferich, D;
Thornbrugh, D; Ross, R, (201 la) A landscape assessment offish habitat conditions in United States rivers
and their watersheds. Denver, Co: National Fish Habitat Action Plan (NFHAP). Available online at
http://fishhabitat.org/.
Esselman, PC; Infante, DM; Wang, L; Wu, D; Cooper, AR; Taylor, WW. (20 lib) An index of cumulative
disturbance to river fish habitats of the conterminous United States from landscape anthropogenic
activities. Ecol Restor 29:133-151.
Frac Tracker. (2013) Frac Tracker - data [web page], http://www.fractracker.org/downloads/
Homer, C; Dewitz, J; Fry, J; Coan, M; Hossain, N; Larson, C; Herold, N; McKerrow, A; VanDriel, JN; Wickham, J.
(2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. PE&RS
73(4):337-341. Available online at
http://www.asprs.0rg/a/publications/pers/2007journal/april/highlight.pdf
King, RS; Baker, ME. (2010) Considerations for identifying and interpreting ecological community thresholds. J N
Am Benthol Assoc 29(3):998-1008
Nowak, DJ; Greenfield, EJ. (2010) Evaluating the National Land Cover Database tree canopy and impervious cover
estimates across the conterminous United States: a comparison with photo-interpreted estimates. Environ
Manage 46:378-390.
Olivero, AP; Anderson, MG. (2008) Northeast aquatic habitat classification. Denver, CO: The Nature Conservancy.
Available online at http://rcngrants.org/content/northeastern-aquatic-habitatclassification-project
PASDA (Pennsylvania Spatial Data Access). (2013) Data download -mine and refuse permits. Available online at
www.pasda.psu.edu.
U.S. Bureau of the Census. (2000) Census 2000 TIGER/Line data. Available online at
http://www.icpsr.umich.edu/icpsrweb/ICPSR/themes/tiger/2000/
U.S. EIA (Energy Information Administration). (2013) Maps: exploration, resources, reserves, and production -
United States shale gas maps - Lower 48 states shale plays. Available online at
http://www.eia. gov/pub/oil_gas/natural_gas/analysis_publications/maps/maps.htm
D-18
-------
U.S. EIA (Energy Information Administration). (2012) Annual coal report 2011. http://www.eia.gov/coal/annual/
U.S. EPA/USGS (Environmental Protection Agency/U.S. Geological Survey). (2005). National hydrography dataset
plus, NHDPlus Version 1.0. (NHDPlusVl). Available online at http://www.fws.gov/r5gomp/gom/nhd-
gom/metadata.pdf
U.S. EPA (Environmental Protection Agency). (2011) ICLUS tools and datasets (Version 1.3 & 1.3.1). [EPA/600/R-
09/143F]. Washington, DC: U.S. Environmental Protection Agency. Available on line at
http://cfpub.epa. gov/ncea/global/recordisplay.cfm?deid=205305
U.S. EPA (U.S. Environmental Protection Agency). (2013) Geospatial data download service - Geospatial
information for all publicly available FRS facilities that have latitude/longitude data [file geodatabase].
Available online at http://www.epa.gov/enviro/geo_data.html
USGS (U.S. Geological Survey). (2001) Coal fields of the United States: National Atlas of the United States.
Reston, VA: USGS. Available online at http://nationalatlas.gov/atlasftp.htmMcoarfdp
USGS (U.S. Geological Survey). (2005) Active mines and mineral processing plants in the United States in 2003.
Reston, VA: USGS. Available online at http://tin.er.usgs.gov/metadata/mineplant.faq.html
USGS (U.S. Geological Survey) (2010). Water use in the United States: Estimated use of water in the United States,
2005. Reston, VA: USGS. Available online at http://water.usgs.gov/watuse/
USGS (U.S. Geological Survey). (2014) Major dams of the United States: National Atlas of the United States.
Reston, VA: United States Geological Survey. Available online at:
http://nationalatlas.gov/atlasftp.htmrMamsOOx
VDEQ (Virginia Department of Environmental Quality. (2013) Surface mine permit boundaries. Richmond, VA:
Division of Mined Land Reclamation. Available online at
ftp://ftp.dmme.virginia.gov/DMLR/downloads/permits/shape_files/nad83/
Wickham, JD; Stehman, SV; Gass, L; Dewitz, J; Fry, JA; Wade, TG. (2013) Accuracy assessment of NLCD 2006
land cover and impervious surface. Rem Sens Environ 130: 294-304.
WVDEP (West Virginia Department of Environmental Protection). (2013) Data download - mining permit
shapefiles. Charleston, WV: Technical Applications and GIS Unit (TAGIS). Available online at
http://tagis.dep.wv.gov/home/Downloads
WVGES (West Virginia Geological Ecomic Survey). (2014). West Virginia coal bed mapping [online mapper].
Available online at http://www.wvgs.wvnet.edu/www/coal/cbmp/coalims.html
D-19
-------
[This page intentionally left blank.]
-------
APPENDIX E.
SECONDARY REGIONAL
MONITORING NETWORK (RMN)
SITES IN THE NORTHEAST AND
MID-ATLANTIC REGIONS
E-l
-------
Table E-l. Secondary RMN sites in the Northeast (4/2/2014). At all of the VT DEC and CT DEEP sentinel sites,
macroinvertebrates are collected annually and water temperature sensors are deployed year-round. For this
exercise, natural land cover included open water, forest, wetlands, barren, and grassland/herbaceous. Data were
accumulated for the entire upstream catchment. Land cover screenings were estimates based on data associated
with the National Hydrography Dataset Plus Version 1 (NHDPlusVl) catchments in which the sites are located
(http://www.horizon-systems.com/NHDPlus). Land use data were based on the 2001 National Land Cover
Database (NLCD; Homer et al., 2007) (http://www.mrlc.gov/nlcd01_data.php).
Longitude
-72.7464
-72.9458
-71.6356
-72.7819
-73.2292
-72.9384
-72.3289
-72.3343
-72.82146
-73.2155
-72.5365
Latitude
43.7708
43.8556
44.7550
44.5036
44.2483
42.0356
41.4100
41.4603
41.93717
41.5575
41.6615
State
VT
VT
VT
VT
VT
CT
CT
CT
CT
CT
CT
Entity
VTDEC
VTDEC
VTDEC
VTDEC
VTDEC
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
Station ID
130000000324
135411000013
280000000003
493238200015
530000000037
1156
1236
1239
359
1468
2295
Water body name
White River
Smith Brook
Nulhegan River
Ranch Brook
Lewis Creek
Hubbard Brook
Beaver Brook
Burnhams Brook
West Branch Salmon
Weekepeemee River
Mott Hill Brook
Percentage
natural
(%)
92.4
99
90
96
64
96
89
85
86
72
92
Notes
VT DEC sentinel site
VT DEC sentinel site
VT DEC sentinel site
VT DEC sentinel site; USGS
gage (04288230)
VT DEC sentinel site
CT DEEP sentinel site;
colocated with USGS gage
(01187300)
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site;
colocated with USGS gage
(01203805)
CT DEEP sentinel site
w
to
-------
Table E-l. continued...
Longitude
-72.4226
-73.1214
-72.4338
-73.3200
-73.1679
-72.1509
-73.3678
-73.1745
-72.9630
-72.4640
-69.5933
-69.5313
Latitude
41.4283
41.9328
41.5623
41.9459
41.8646
41.7812
41.2931
41.5783
41.7807
41.8272
44.2232
44.3679
State
CT
CT
CT
CT
CT
CT
CT
CT
CT
CT
ME
ME
Entity
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
CT DEEP
MEDEP
MEDEP
Station ID
2297
2299
2304
2309
2312
2331
2346
2676
2711
345
MEDEP_56817
MEDEP_57011
Water body name
Hemlock Valley
Brook
Rugg Brook
Day Pond Brook
Flat Brook
Jakes Brook
Stonehouse Brook
Little River
Nonewaug River
Bunnell Brook
Tankerhoosen River
Sheepscot
River — Station 74
West Branch
Sheepscot
River— Station 268
Percentage
Natural
(%)
81
91
—
90
91
89
81
53
71
66
87
85
Notes
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site
CT DEEP sentinel site; USGS
gage (01 203 600)
CT DEEP sentinel site; USGS
gage (01 188000)
CT DEEP sentinel site
ME DEP long-term
monitoring site; USGS gage
(01038000)— water and air
temperature, discharge
ME DEP long-term biological
monitoring site
w
-------
Table E-l. continued...
Longitude
-68.2346
Latitude
44.3934
State
ME
Entity
MEDEP
Station ID
MEDEP_57065
Water body name
Duck Brook — Station
322
Percentage
Natural
(%)
83
Notes
ME DEP long-term biological
monitoring site
w
-------
Table E-2. Secondary RMN sites in the Mid-Atlantic (4/2/2014). At all of the MD DNR sentinel sites,
macroinvertebrates are collected annually and water and air temperature sensors are deployed year-round. At the WV
DEP sites, macroinvertebrates are collected annually and water temperature sensors may be deployed. At the SRBC
continuous monitoring sites, macroinvertebrates are collected annually and water temperature sensors are deployed
year-round; stage and precipitation data are also being collected at some sites (see Notes field). At the NFS—ERMN
sites (National Park Service sites that are in the Eastern Rivers and Mountains Network), macroinvertebrates are
collected every other year and efforts will be made to install temperature sensors at high priority sites. Land use data,
which were accumulated for the entire upstream catchment, were estimates based on data associated with the National
Hydrography Dataset Plus Version 1 (NHDPlusVl) catchments in which the sites are located (http://www.horizon-
systems.com/NHDPlus). Land use data were based on the 2001 National Land Cover Database (NLCD; Homer et al.,
2007)(http://www.mrlc.gov/nlcd01_data.php).
Longitude
-75.323216
-74.88980
-74.50486
-74.84479
Latitude
41.73465
40.77471
40.95164
40.75211
State
PA
NJ
NJ
NJ
Entity
DRBC
EPAR2
EPAR2
EPAR2
Station ID
MB_Dyberry
1
17
2
Water body
name
Middle Branch
Dyberry Creek
Unnamed
tributary to
Musconetcong
River
Hibernia Brook
Teetertown
Brook
Notes
long-term
monitoring
site — Jim
Kurtenbach
(U.S. EPA R2)
long-term
monitoring
site — Jim
Kurtenbach
(U.S. EPA R2)
long-term
monitoring
site — Jim
Kurtenbach
(U.S. EPA R2)
Percentage
Forest (%)
79.9
67.9
82.1
53.9
Percentage
Urban (%)
0.0
0.0
2.1
1.3
Percentage
Hay (%)
14.0
4.2
0.0
14.6
Percentage
Crop (%)
0.6
23.9
0.8
17.7
w
-------
Table E-2. continued...
Longitude
-77.54528
-76.09499
-78.45571
-76.04611
-76.69843
-76.71875
-76.86417
-76.69829
Latitude
39.65833
39.08754
39.68672
39.61055
39.43951
39.42925
39.44055
39.48052
State
MD
MD
MD
MD
MD
MD
MD
MD
Entity
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
Station ID
ANTI-101-S
CORS-102-S
FIMI-207-S
FURN-101-S
JONE-109-S
JONE-315-S
LIBE-102-S
LOCH-120-S
Water body
name
Unnamed
tributary to
Edgemont
Reservoir
Unnamed
tributary to
Emory Creek
Fifteen Mile
Creek
Unnamed
tributary to
Principio Creek
Unnamed
tributary to
Dipping Pond
Run
North Branch
of Jones Falls
Timber Run
Baisman Run
Notes
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
Percentage
Forest (%)
90.6
71.6
88.3
80.5
45.3
52.8
70.1
71.2
Percentage
Urban (%)
0.4
0.2
0.1
0.5
2.5
1.1
0.3
0.3
Percentage
Hay (%)
4.4
4.6
5.6
9.3
32.4
23.8
12.7
24.4
Percentage
Crop (%)
0.5
7.2
0.9
8.1
6.7
16.3
15.5
2.4
w
-------
Table E-2. continued...
Longitude
-76.21896
-77.09766
-77.08594
-75.49247
-75.46182
-76.76012
-77.02912
Latitude
39.19352
38.58225
38.48386
38.2495
38.26359
38.56392
38.51108
State
MD
MD
MD
MD
MD
MD
MD
Entity
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
Station ID
LOCR-102-S
MATT-033-S
NANJ-331-S
NASS-108-S
NASS-302-S
PAXL-294-S
PTOB-002-S
Water body
name
Swan Creek
Mattawoman
Creek
Mill Run
Millville Creek
Nassawango
Creek
Swanson Creek
Hoghole Run
Notes
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
Percentage
Forest (%)
20.8
48.3
72.5
8.9
18.8
67.3
74.2
Percentage
Urban (%)
0.0
12.8
0.2
0.8
0.4
0.2
1.0
Percentage
Hay (%)
0.0
1.8
1.8
0.6
4.4
2.4
0.7
Percentage
Crop (%)
17.7
7.5
9.9
8.0
17.0
15.3
12.1
w
-------
Table E-2. continued...
Longitude
-76.97198
-79.21349
-76.73717
-77.48935
-75.96062
-75.78362
-75.59259
Latitude
39.16949
39.54119
38.36662
39.58739
38.72408
39.28768
38.41408
State
MD
MD
MD
MD
MD
MD
MD
Entity
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
MDDNR
Station ID
RKGR-119-S
SAVA-276-S
STCL-051-S
UMON-119-
S
UPCK-113-S
UPCR-208-S
WIRH-220-S
Water body
name
Unnamed
tributary to
Patuxent River
Double Lick
Run
Unnamed
tributary to St.
Clements Creek
Buzzard
Branch
Unnamed
tributary to
Skeleton Creek
Cypress Branch
Leonard Pond
Run
Notes
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Highlands
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
MDDNR
sentinel
site — Coastal
Plain
Percentage
Forest (%)
31.8
92.1
69.8
97.5
23.9
40.3
26.7
Percentage
Urban (%)
0.5
0.0
0.0
0.0
0.3
0.2
8.0
Percentage
Hay (%)
44.0
6.6
2.2
0.2
10.7
4.0
6.3
Percentage
Crop (%)
14.2
0.0
15.8
0.0
29.4
26.7
30.3
w
oo
-------
Table E-2. continued...
Longitude
-76.90348
-78.45247
-78.51846
-79.92348
-79.58149
-79.93024
-80.97161
-78.48373
-81.02055
-79.59970
-75.14398
-74.98444
Latitude
38.49936
40.41597
40.43269
39.78393
39.81449
39.78248
37.58466
40.41876
37.53483
39.81014
40.97139
41.11381
State
MD
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
Entity
MDDNR
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
Station ID
ZEKI-012-S
DEWA.3001
DEWA.3002
Water body
name
Unnamed
tributary to
Zekiah Swamp
Run
Blair Gap
Run — Foot of
Ten
Blair Gap
Run—
Muleshoe
Dublin Run
Great Meadows
Run
Ice Pond Run
Little Bluestone
River
Milllstone Run
Mountain
Creek
Unnamed
tributary
(Scotts Run)
Caledonia
Creek 13
Van Campen
Creek 12
Notes
MDDNR
sentinel
site — Coastal
Plain
NPS— ERMN
high priority
Percentage
Forest (%)
89.8
92.9
96.1
-
75.2
-
75.7
97.0
75.9
95.9
78.0
74.2
Percentage
Urban (%)
0.0
1.4
0.4
-
2.4
-
1.5
0.3
1.6
0.0
1.0
2.4
Percentage
Hay (%)
0.6
1.8
0.5
~
9.4
~
15.3
0.8
14.9
0.0
8.0
0.1
Percentage
Crop (%)
5.3
0.2
0.1
~
3.5
~
0.9
0.0
0.6
0.0
1.5
0.1
-------
Table E-2. continued...
Longitude
-74.90309
-74.87464
-75.12652
-74.96252
-74.90598
-74.91831
-74.92372
-74.87711
-74.89043
-75.00533
-74.96505
-74.90343
-74.95916
-74.92673
Latitude
41.19744
41.22245
40.974
41.13729
41.1756
41.23772
41.09674
41.24882
41.2578
41.09383
41.07109
41.23052
41.12946
41.16889
State
PA
PA
NJ
PA
PA
PA
NJ
PA
PA
PA
NJ
PA
PA
PA
Entity
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
Station ID
DEWA.3003
DEWA.3004
DEWA.3005
DEWA.3006
DEWA.3007
DEWA.3008
DEWA.3010
DEWA.3011
DEWA.3012
DEWA.3013
DEWA.3014
DEWA.3015
DEWA.3018
DEWA.3020
Water body
name
Deckers Creek
03
Dingmans
Creek 05
Dunnfield
Creek 03
Toms Creek 20
Spackmans
Creek 08
Dingmans
Creek 57
Vancampens
Brook 95
Adams Creek
14
Adams Creek
33
Little Bushkill
Creek 01
Vancampens
Brook 43
Dingmans
Creek 39
Toms Creek 07
Mill Creek 25
Notes
Percentage
Forest (%)
82.0
71.3
96.8
80.2
91.1
69.8
93.8
86.9
82.8
76.9
95.3
69.8
80.2
83.5
Percentage
Urban (%)
0.0
2.0
0.0
0.6
0.5
2.0
0.0
0.9
1.2
0.5
0.0
2.0
0.6
0.3
Percentage
Hay (%)
0.3
0.1
0.3
0.0
0.7
0.1
0.0
0.1
0.1
0.0
0.0
0.1
0.0
0.4
Percentage
Crop (%)
2.4
0.0
0.0
0.0
0.4
0.0
0.0
0.1
0.1
0.0
0.0
0.0
0.0
0.4
-------
Table E-2. continued...
Longitude
-74.96279
-74.88573
-74.98445
-74.94059
-74.86975
-74.79550
-75.01434
-75.00528
-74.89481
-74.84545
-75.10517
-74.95645
Latitude
41.1415
41.23542
41.0647
41.08567
41.24147
41.29461
41.08235
41.03179
41.23067
41.2952
40.98337
41.12711
State
PA
PA
NJ
NJ
PA
NJ
PA
NJ
PA
PA
NJ
PA
Entity
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
Station ID
DEWA.3022
DEWA.3023
DEWA.3025
DEWA.3026
DEWA.3027
DEWA.3028
DEWA.3029
DEWA.3030
DEWA.3031
DEWA.3032
DEWA.3033
DEWA.3034
Water body
name
Toms Creek 25
Unnamed
tributary
Dingmans
Creek 07
Vancampens
Brook 22
Unnamed
tributary
Vancampens
Brook 05
Adams Creek
03
White Brook
15
Sand Hill
Creek 08
Yards Creek 07
Dingmans
Creek 30
Raymondskill
Creek 13
Dunnfield
Creek 26
Toms Creek 03
Notes
NPS— ERMN
high priority
NPS— ERMN
high priority
Percentage
Forest (%)
80.2
71.3
95.3
97.8
86.9
53.2
63.5
86.0
69.8
73.7
96.8
81.2
Percentage
Urban (%)
0.6
2.0
0.0
0.0
0.9
1.6
6.1
0.0
2.0
1.0
0.0
0.5
Percentage
Hay (%)
0.0
0.1
0.0
0.0
0.1
13.6
0.8
0.0
0.1
0.9
0.3
0.0
Percentage
Crop (%)
0.0
0.0
0.0
0.0
0.1
7.6
0.0
0.0
0.0
0.1
0.0
0.2
-------
Table E-2. continued...
Longitude
-74.89987
-74.92431
-74.94123
-74.88168
-81.09167
-81.01278
-81.02305
-80.91077
-81.04551
-81.04918
-81.02102
-81.02453
-80.90375
-81.03647
-80.97903
Latitude
41.19356
41.15917
41.09062
41.25185
37.9441
37.91956
37.88808
37.81927
37.87994
37.82895
38.03256
37.94417
37.714
37.87402
37.85864
State
PA
PA
NJ
PA
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
Entity
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
Station ID
DEWA.3035
DEWA.3036
DEWA.3038
DEWA.3039
NERI.3001
NERI.3005
NERI.3009
NERI.3011
NERI.3013
NERI.3016
NERI.3018
NERI.3021
NERI.3024
NERI.3025
NERI.3026
Water body
name
Hornbecks
Creek 15
Mill Creek 12
Vancampens
Brook 76
Adams Creek
21
Meadow Fork 1
Buffalo Creek
16
Slater Creek 20
Meadow Creek
17
Dowdy Creek
16
River Branch 4
Keeney Creek
10
Fire Creek 17
Big Branch 10
Dowdy Creek
30
Little Laurel
Creek 6
Notes
Percentage
Forest (%)
82.0
83.5
93.8
85.0
74.1
98.5
97.7
84.9
99.1
95.9
89.7
99.2
95.3
99.1
91.3
Percentage
Urban (%)
0.8
0.3
0.0
1.0
6.1
0.0
0.0
2.5
0.0
0.0
2.3
0.0
0.2
0.0
0.0
Percentage
Hay (%)
0.7
0.4
0.0
0.1
6.1
0.0
0.0
4.6
0.0
3.8
1.5
0.0
2.9
0.0
3.8
Percentage
Crop (%)
1.0
0.4
0.0
0.1
0.1
0.2
0.0
1.1
0.0
0.3
0.8
0.2
0.0
0.0
1.5
-------
Table E-2. continued...
Longitude
-81.08293
-81.04749
-81.02506
-81.06012
-80.94296
-81.00490
-81.08737
-81.01654
-81.01984
-80.95156
-80.89788
-81.03925
-80.92717
Latitude
38.04904
37.82782
37.98267
38.06032
37.74477
37.85802
37.96331
37.78795
37.8583
37.87324
37.83271
37.8512
37.80196
State
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
Entity
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
Station ID
NERI.3029
NERI.3032
NERI.3034
NERI.3035
NERI.3036
NERI.3037
NERI.3038
NERI.3040
NERI.3041
NERI.3042
NERI.3043
NERI.3044
NERI.3047
Water body
name
Wolf Creek 30
River Branch 6
Unnamed
tributary 21
New River 1
Fern Creek 1 1
Unnamed
tributary Fall
Branch 2
Laurel Creek
47
Arbuckle Creek
2
Polls Branch 14
Unnamed
tributary Laurel
Creek 3
Bucklick
Branch 3
Meadow Creek
39
Laurel Creek 8
Sewell Branch
2
Notes
WVDEP
reference site
Percentage
Forest (%)
56.6
95.9
69.3
88.1
97.8
93.0
47.7
70.0
93.0
98.8
84.3
93.0
71.0
Percentage
Urban (%)
6.5
0.0
2.2
2.9
0.0
0.6
24.9
1.2
0.6
0.0
2.6
0.6
0.4
Percentage
Hay (%)
25.5
3.8
20.8
3.1
1.4
1.5
12.2
20.5
1.5
0.2
4.8
1.5
22.0
Percentage
Crop (%)
0.1
0.3
0.7
0.4
0.1
0.6
0.3
0.1
0.6
0.0
1.1
0.6
0.1
-------
Table E-2. continued...
Longitude
-81.05710
-81.01080
-81.01287
-80.93170
-81.02849
-81.09031
-80.95197
-80.88025
-81.10399
-81.09510
Latitude
37.82369
37.91417
37.96168
37.74969
37.89156
37.96421
37.86122
37.83799
37.84261
37.94727
State
WV
WV
WV
WV
WV
WV
WV
WV
WV
WV
Entity
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
NFS— ERMN
Station ID
NERI.3048
NERI.3049
NERI.3050
NERI.3052
NERI.3053
NERI.3054
NERI.3058
NERI.3059
NERI.3064
NERI.3065
Water body
name
Slate
Fork— Mill
Creek 12
Unnamed
tributary
Buffalo Creek
6
Ephraim Creek
8
Fall Branch 7
Slater Creek 13
Arbuckle Creek
5
Richlick
Branch 17
Meadow Creek
58
BatoffCreek7
Meadow Fork 6
Notes
NFS— ERMN
high priority;
WVDEP
reference site
NFS— ERMN
high priority;
WVDEP
reference site
Percentage
Forest (%)
81.9
98.5
99.3
95.3
97.7
47.7
97.1
81.4
75.6
74.0
Percentage
Urban (%)
2.9
0.0
0.0
0.5
0.0
24.9
0.0
3.5
6.6
5.1
Percentage
Hay (%)
9.9
0.0
0.0
2.7
0.0
12.2
0.3
5.8
5.9
6.5
Percentage
Crop (%)
0.1
0.2
0.0
0.2
0.0
0.3
0.5
1.2
0.0
0.1
-------
Table E-2. continued...
Longitude
-81.02195
-80.90266
-81.05947
-81.05316
-81.01693
-80.98218
-81.08257
-81.05947
-80.93452
-76.91134
-76.15029
-77.60667
-76.72019
Latitude
37.91346
37.71391
37.88203
37.83172
38.03013
37.86476
38.04763
38.06101
37.74875
41.32519
42.06312
41.24694
42.04209
State
WV
WV
WV
WV
WV
WV
WV
WV
WV
PA
NY
PA
NY
Entity
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
NPS— ERMN
PADEP
SRBC
SRBC
SRBC
Station ID
NERI.3069
NERI.3072
NERI.3077
NERI.3080
NERI.3082
NERI.3085
NERI.3093
NERI.3099
NERI.3100
WQN_408
Apal
BAKRO.l
Baldwin
Water body
name
Buffalo Creek
4
Big Branch 9
Dowdy Creek 2
Slate
Fork— Mill
Creek 1
Keeney Creek
15
Laurel Creek
61
Wolf Creek 32
Fern Creek 12
Fall Branch 10
Loyalsock
Creek
Apalachin
Creek
Baker Run
Baldwin Creek
Notes
NPS— ERMN
high priority;
WVDEP
reference site
long-term data,
EV (protected)
precip gage
pressure
transducer
(real-time) and
precip gage
precip gage
Percentage
Forest (%)
98.5
95.3
99.1
86.3
89.7
91.9
56.6
88.1
95.3
81.6
69.5
97.0
73.0
Percentage
Urban (%)
0.0
0.2
0.0
2.0
2.3
0.8
6.5
2.9
0.5
0.2
0.6
0.0
0.1
Percentage
Hay (%)
0.0
2.9
0.0
7.9
1.5
1.7
25.5
3.1
2.7
6.3
23.1
0.0
15.4
Percentage
Crop (%)
0.2
0.0
0.0
0.1
0.8
0.7
0.1
0.4
0.2
4.4
1.5
0.0
5.0
-------
Table E-2. continued...
Longitude
-77.23044
-78.59258
-76.02756
-77.73670
-76.47508
-74.79921
-78.64757
-76.00931
-77.29313
-78.27008
Latitude
41.47393
40.26388
41.42725
42.31903
42.20472
42.70639
40.63052
42.01582
41.85752
41.52649
State
PA
PA
PA
NY
NY
NY
PA
NY
PA
PA
Entity
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
Station ID
BLOC
BOBS
BOWN
CANA
Catatonk
Cherry
CHEST
CHOC
CROK
Driftwood
Water body
name
Blockhouse
Creek
Bobs Creek
Bowman Creek
Canacadea
Creek
Catatonk Creek
Cherry Valley
Creek
Chest Creek
Choconut
Creek
Crooked Creek
Driftwood
Branch
Sinnemahoning
Creek
Notes
precip gage
pressure
transducer
(real-time) and
precip gage
precip gage
pressure
transducer
(stand-alone)
pressure
transducer
(stand-alone)
pressure
transducer
(real-time)
Percentage
Forest (%)
74.1
88.5
89.1
69.5
71.4
66.1
58.1
72.7
45.9
92.0
Percentage
Urban (%)
0.8
0.2
0.2
1.1
0.5
0.3
0.8
0.2
0.1
0.0
Percentage
Hay (%)
13.6
4.7
1.8
19.8
11.2
15.1
21.5
19.9
27.1
1.0
Percentage
Crop (%)
5.4
2.1
5.0
1.3
4.5
5.1
13.2
2.3
21.6
0.1
w
-------
Table E-2. continued...
Longitude
-77.91244
-76.34434
-76.07111
-77.58154
-76.91233
-78.25348
-78.17458
-76.24282
-75.47324
-77.18943
-78.40722
-76.06980
-76.64148
Latitude
41.57467
41.32261
41.78832
41.73642
41.99164
41.36235
41.45256
41.23366
41.68331
41.32739
40.97
41.58154
41.19353
State
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
PA
Entity
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
Station ID
East Fork
EBFC
EBWC
ELKR
HAMM
Hicks
Hunts
Kitchen
LACK
LARR
LCLFO.l
LMEHOOP
LMUN
Water body
name
East Fork First
Fork
Sinnemahoning
Creek
East Branch
Fishing Creek
East Branch
Wyalusing
Creek
Elk Run
Hammond
Creek
Hicks Run
Hunts Run
Kitchen Creek
Lackawanna
River
Larrys Creek
Little Clearfield
Creek
Little
Mehoopany
Creek
Little Muncy
Creek
Notes
precip gage
pressure
transducer
(real-time)
Percentage
Forest (%)
89.1
92.4
50.0
81.8
46.6
91.6
90.7
85.9
68.2
75.7
68.5
66.8
56.1
Percentage
Urban (%)
0.0
0.0
0.4
0.0
0.2
0.0
0.0
0.2
0.4
0.1
0.2
0.1
0.2
Percentage
Hay (%)
0.2
0.1
32.3
10.4
33.6
1.6
0.0
3.5
11.3
19.1
19.1
8.1
23.1
Percentage
Crop (%)
0.0
0.0
10.4
0.4
14.0
0.0
0.0
0.5
9.0
1.0
2.4
16.7
13.9
-------
Table E-2. continued...
Longitude
-77.55928
-77.36278
-76.33104
-77.60997
-77.41333
-75.98474
-76.05357
-77.76387
-78.46158
-77.45056
-76.92300
-78.22029
-75.50220
-75.77788
Latitude
41.76142
41.31
41.4588
41.06022
41.76306
41.61164
42.20426
41.79146
41.04564
41.64694
41.49143
41.51169
42.77596
41.55783
State
PA
PA
PA
PA
PA
PA
NY
PA
PA
PA
PA
PA
NY
PA
Entity
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
Station ID
Long
LPIN0.2
LYSK5.0
MARS
Marsh Tioga
MESH
Nanticoke
Ninemile
PA Moose
Pine
Blackwell
Pies
Portage
Sangerfield
SBTK
Water body
name
Long Run
Little Pine
Creek
Loyalsock
Creek
Marsh Creek
Marsh Creek
Meshoppen
Creek
Nanticoke
Creek
Ninemile Run
Moose Creek
Pine Creek
Pleasant Stream
Portage Creek
Sangerfield
River
South Branch
Tunkhannock
Creek
Notes
pressure
transducer
(real-time) and
precip gage
pressure
transducer
(stand-alone)
precip gage
pressure
transducer
(real-time)
Percentage
Forest (%)
82.0
82.7
85.5
83.7
72.0
47.1
62.2
84.4
90.3
80.5
87.6
91.9
35.8
53.4
Percentage
Urban (%)
0.0
0.4
0.1
0.7
1.3
0.2
0.4
0.4
-
0.4
0.0
0.3
0.4
3.4
Percentage
Hay (%)
10.4
8.4
0.6
7.9
15.5
18.5
24.9
3.8
~
7.7
1.4
2.5
17.0
5.9
Percentage
Crop (%)
1.9
3.5
0.7
3.4
5.0
27.3
7.8
2.8
~
2.0
1.0
0.1
14.0
24.1
-------
Table E-2. continued...
Longitude
-76.92222
-75.84137
-75.52351
-76.27436
-76.76835
-76.91416
-76.60723
-76.76011
-78.36118
-76.10589
-77.37918
-77.76123
-76.28083
Latitude
42.10278
41.92994
41.95946
41.62644
41.78974
41.70931
41.78132
41.65262
41.07359
42.59277
42.0752
41.79011
41.96661
State
NY
PA
PA
PA
PA
PA
PA
PA
PA
NY
NY
PA
PA
Entity
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
SRBC
Station ID
SING 0.9
SNAK
STAR
Sugar Run
SUGR
TIOG
TOMJ
TOWA
TROT
Trout Brook
Tuscarora
Upper Pine
WAPP
Water body
name
Sing Sing
Creek
Snake Creek
Starrucca Creek
Sugar Run
Sugar Creek
Tioga River
Tomjack Creek
Towanda Creek
Trout Run
Trout Brook
Tuscarora
Creek
Pine Creek
Wappasening
Creek
Notes
pressure
transducer
(stand-alone)
pressure
transducer
(real-time) and
precip gage
precip gage
pressure
transducer
(real-time)
pressure
transducer
(real-time)
Percentage
Forest (%)
61.5
67.1
73.1
64.4
44.9
83.0
42.7
54.9
90.4
62.7
42.9
75.1
63.4
Percentage
Urban (%)
3.1
0.2
0.2
0.1
1.0
0.0
0.3
0.8
0.0
0.5
0.3
0.0
0.0
Percentage
Hay (%)
9.9
23.0
15.9
9.9
34.3
1.7
35.9
25.5
0.4
24.8
35.1
11.4
30.2
Percentage
Crop (%)
10.9
2.8
1.1
18.6
13.2
2.9
15.6
13.1
0.0
5.3
13.6
4.0
1.8
-------
Table E-2. continued...
Longitude
-78.80331
-78.27484
-77.66985
-77.68520
-74.50528
-82.12353
-79.69583
-80.86781
-81.09958
Latitude
40.69289
41.49444
41.72483
41.40016
39.885
38.48514
38.73825
38.88133
39.22211
State
PA
PA
PA
PA
NJ
WV
wv
WV
wv
Entity
SRBC
SRBC
SRBC
SRBC
USGS
WVDEP
WVDEP
WVDEP
WVDEP
Station ID
WBSUS
West
WPIN
Young
USGS
01466500
11897
12455
12689
12690
Water body
name
West Branch
Susquehanna
River
West Creek
West Branch
Pine Creek
Young
Woman's Creek
McDonalds
Branch
Unnamed
tributary/Left
Fork river mile
1.69/Mill Creek
Laurel
Fork/Dry Fork
Long Lick Run
Unnamed
tributary/North
Fork river mile
22.26/Hughes
River
Notes
pressure
transducer
(stand-alone)
USGS gage in
Byrne State
Forest (Pine
Barrens)
Percentage
Forest (%)
68.7
83.8
87.0
96.8
89.5
-
Percentage
Urban (%)
3.5
0.5
0.0
0.0
0.0
-
Percentage
Hay (%)
15.8
6.5
0.5
0.1
0.1
~
Percentage
Crop (%)
3.0
0.4
0.0
0.0
0.1
~
w
to
o
-------
Table E-2. continued...
Longitude
-80.32127
-81.14683
-82.28014
-79.39594
-79.61147
-81.93119
-80.37117
Latitude
38.25981
37.50275
38.06845
38.97394
39.04225
38.38489
38.33544
State
WV
WV
WV
WV
WV
WV
WV
Entity
WVDEP
WVDEP
WVDEP
WVDEP
WVDEP
WVDEP
WVDEP
Station ID
2046
2359
4513
8255
8357
8482
9315
Water body
name
North
Fork/Cranberry
River
Mash Fork
Little Laurel
Creek
Red Creek
Otter Creek
Sams Fork
Middle
Fork/Williams
River
Notes
long-term
monitoring site
impacted by acid
ram
long-term
monitoring site
impacted by acid
ram
long-term
monitoring site
impacted by acid
ram
long-term
monitoring site
impacted by acid
ram
long-term
monitoring site
impacted by acid
ram
Percentage
Forest (%)
98.9
93.1
—
97.7
99.5
-
99.5
Percentage
Urban (%)
0.0
0.0
—
0.0
0.0
-
0.0
Percentage
Hay (%)
0.0
5.2
—
0.1
0.0
~
0.0
Percentage
Crop (%)
0.0
0.0
—
0.0
0.0
~
0.0
w
to
-------
E.I. REFERENCES
Homer, C; Dewitz, J; Fry, J; Coan, M; Hossain, N; Larson, C; Herold, N; McKerrow, A; VanDriel, JN; Wickham, J.
(2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. PE&RS
73(4):337-341. Available online at
http://www.asprs.0rg/a/publications/pers/2007journal/april/highlight.pdf
E-22
-------
APPENDIX F.
MACROINVERTEBRATE
COLLECTION METHODS
F-l
-------
Table F-l. Macroinvertebrate methods for medium-high gradient freshwater wadeable streams with abundant
riffle habitat and rocky substrate. These are the methods that were agreed upon by the Northeast, Mid-Atlantic,
and Southeast regional working groups during the pilot phase of the RMNs (before the QAPP was developed).
At this time (fall 2015), the regional working groups are working on the region-specific QAPP addendums. It is
possible that some updates will be made during this process.
Regional
network
Northeast
Mid-
Atlantic
Effort
Kick samples are taken
from riffle habitats in 4
different locations in the
sampling reach. At each
location the substrate is
disturbed for
approximately
30 seconds, for a total
active sampling effort of
2 minutes.
Data should be collected
with existing state or RBC
methods, or in such a way
that the data can be
rendered comparable to
historical state methods.
A minimum of 1 m2 is
collected using a
minimum of 4 separate
kicks in riffle habitats
throughout the 100-m
reach.
Reach
length
150m
100m
Gear
D -frame net
(46 cm wide *
30 cm high)
with 500-um
mesh
Varies by
entity (either
square frame
kick nets or
D-frame nets,
with mesh
size ranging
from
450-600 urn)
Habitat
Riffles
Abundant
riffles
Sampling
area
Approximately
1m2
Minimum of
1m2
Index
period
September-
mid-
October
Spring
(March-
April) and
summer
(July-
August)
Target #
organisms
300
300
Taxonomic
resolution
Lowest
practical
(species
whenever
possible)
Lowest
practical
(species
whenever
possible)
to
-------
Table F-l. continued...
Regional
network
Southeast
Effort
Semiquantitative: riffle
kick samples are taken
from 2 riffles or upper or
lower end of a large riffle
and composited; in
smaller streams, multiple
riffles may need to be
collected to achieve the
desired area
Qualitative: 3 "jabs" will
be collected from all
available habitats; taxa
from each habitat will be
kept in separate
containers (separate
species lists will be
generated for each
habitat)
Reach
length
100m
100m
Gear
Kick-net with
500-um mesh
Dip-net with
500-um mesh
Habitat
Riffles
Multihabitat
Sampling
area
Approximately
2m2
NA
(qualitative)
Index
period
April 20 13.
Subsequent
samples will
be collected
annually
within
2 weeks of
the original
collection
Target #
organisms
300± 10%
NA
(qualitative)
Taxonomic
resolution
Lowest
practical
(species
whenever
possible)
-------
Table F-2. Macroinvertebrate methods used by Northeastern states for routine sampling events in medium-high
gradient freshwater wadeable streams with riffle habitat and rocky substrate
Entity
CT
DEEP
VT
DEC
ME
DEP
Project or
stream type
Streams with
riffle habitat
Moderate to
high gradient
streams with
riffle habitat
Streams with
riffle and run
habitat
Effort
12 kick samples
are taken
throughout riffle
habitats within
the sampling
reach
Kick samples are
taken from riffle
habitats in 4
different
locations in the
sampling reach.
At each location
the substrate is
disturbed for
approximately
30 seconds, for a
total active
sampling effort
of 2 minutes.
3 cylindrical
rock-filled wire
baskets are
placed in
locations with
similar habitat
characteristics for
28 ± 4 days.
Gear
Rectangular
net (46 cm x
46 cm x
25 cm) with
500-um mesh
D-frame net
(46 cm
wide x 30 cm
high) with
500-um mesh
Contents are
washed into a
sieve bucket
with 600-um
mesh
Habitat
Riffles
Riffles
Riffle/run is
the
preferred
habitat.
Sampling
area
Approximately
2m2
Approximately
1m2
Approximately
0.3 m2 per
basket
Index
period
October 1-
November
30
September-
mid-
October
July 1-
September
30
Target #
organisms
200
300
Entire samples
are processed
and identified,
with
exceptions
Taxonomic
resolution
Lowest practical
(species whenever
possible)
Lowest practical
(species whenever
possible)
Lowest practical
(species whenever
possible)
-------
Table F-2. continued...
Entity
NHDES
RIDEM
NY DEC
Project or
stream type
Streams with
riffle and run
habitat
Routine
monitoring in
streams with
riffle habitat
Routine
monitoring in
streams with
riffle habitat
Effort
3 cylindrical rock-
filled wire baskets
are placed in riffle
habitats or at the
base of riffles at
depths that cover
the artificial
substrate by at
least 5 inches for
6 to 8 weeks.
Kick samples are
taken from riffle
habitats along
100-m reach
representative of
the stream
sampled timed for
a total active
sampling effort of
3 minutes.
Substrate is
dislodged by foot,
upstream of the
net for 5 minutes
and a distance of 5
m. The preferred
line of sampling is
a diagonal transect
of the stream
Gear
Contents are
washed into a
sieve bucket
with 600-um
mesh
D-frame net
(30-cm width)
with 500-um
mesh
Rectangular
net (23 cm x
46 cm) with
800-900-um
mesh
Habitat
Riffle/run
is the
preferred
habitat.
Riffle
Riffle
Sampling
area
Approximately
0.3 m2 per
basket
Within reach
(100 linear
meters)
2.5m2
Index
period
late July-
September
August-
September
July-
September
Target #
organisms
100
100
100
Taxonomic
resolution
Genus, except
Chironomidae
(family-level)
Mostly
genus-level.
Chironomidae are
identified to the
subfamily or
tribe-level
Lowest practical
(mostly genus- or
species-level, some
family-level [e.g.,
Gastropoda and
Pelecypoda])
-------
Table F-2. continued...
Entity
MADEP
Project or
stream type
Routine
monitoring in
streams with
riffle habitat
Effort
10 kick-samples
are taken in riffle
habitats within the
sampling reach
and composited
Gear
Kick-net,
46-cm wide
opening,
500-|o,m mesh
Habitat
Riffle/run
is the
preferred
habitat
Sampling
area
Approximately
2m2
Index
period
July 1-
September
30
Target #
organisms
100
Taxonomic
resolution
Lowest practical
level
-------
Table F-3. Macroinvertebrate methods used by Mid-Atlantic states and RBCs for routine sampling events in
medium-high gradient freshwater wadeable streams with riffle habitat and rocky substrate
Entity
NJDEP
DE
DNREC
PADEP
Project or
stream type
Riffle/run
Piedmont
Smaller
freestone
riffle -run
streams
(<25-50 mi2)
Limestone
spring
streams
Effort
10-20 kicks are
taken from riffle/run
areas and
composited
2 kicks composited
6 kicks are taken
from riffle areas and
composited
2 kicks are taken
from riffle-run areas
( 1 fast, 1 slow) and
composited
Gear
D-frame net
(30 cm) with
800 x 900-um
mesh
Kick-net (1-m2
area) with
600 um mesh
D-frame net
(30 cm wide x
20 cm high)
with 500-um
mesh
D-frame net
(30 cm wide
20 cm high)
with 500-um
mesh
Habitat
Riffle/run
Riffle
Riffle
Riffle-run (1 fast, 1
slow)
Sampling
area
10-20 net
dimensions
2m2
6 m2
2m2
Index
period
April-
November
October-
November
Year-round
January-
May
Target #
organisms
100 ± 10%
200 ± 20%
200 ± 20%
300 ± 20%
Taxonomic
resolution
Genus
Genus or
lowest
practical
Genus, except
Chironomidae,
snails, clams,
mussels
(family);
Nematoda,
^^1 f* m f* i*1~ f* 5\
IN ClllCl ltd,
rs V\7C\ 7C\ fl
LJL y \jŁ\jci
(phylum);
Turbellaria,
Hirudenia,
Oligochaeta
(class); water
mites
(artificial)
-------
Table F-3. continued...
Entity
MD
DNR
Project or
stream type
Maryland
Biological
Stream
Survey
(MBSS)
Effort
Approximately
20 kicks/jabs/sweeps
/rubs from multiple
habitats (sampled in
proportion to
availability in reach)
are composited
Gear
D-frame net
(about 30 cm
wide) with
450-um mesh
Habitat
Multihabitat (in order
of preference) riffles,
root wads, root
mats/woody
debris/snag, leaf
packs,
SAV/associated
habitat, undercut
banks; less
preferred = gravel,
broken peat, clay
lumps, detrital/sand
areas in runs; moving
water preferred to
still water; sampled in
proportion to
availability in reach,
ensuring all
potentially productive
habitats are
represented in sample
Sampling
area
About 2 m2
Index
period
March-
April
Target #
organisms
100 ±20%
Taxonomic
resolution
Genus (or lowest
practical);
crayfish and
mussels identified
to species
(sometime
subspecies?) in
the field along
with fish, reptiles,
amphibians, and
some invasive
plants
oo
-------
Table F-3. continued...
Entity
WV
DEP
VA
DEQ
Project or
stream type
Wadeable
streams
(WVSCI)
Wadeable
streams
(GLIMPSS)
— Mountain
and Plateau
Noncoastal
Plain (VSCI)
Effort
4 kicks composited
6 kicks from riffle
habitat (unless
absent, then
multihabitat) are
composited
Gear
Rectangular
kick net (50 cm
wide x 30 cm
high x 50 cm
deep) with
600-um net
mesh (595-um
sieve); D-frame
net (30 cm
wide) can be
iicpr| for
LlowLl -LU1
smaller streams
D-frame net
(50 cm wide x
30 cm high x
50 cm deep)
with 500 um
net mesh
Habitat
riffle-run
Riffle, unless absent,
then multihabitat
Sampling
area
1m2
1m2
2m2
Index
period
April 15-
October 15
Winter
(December-
mid-
February),
spring
r o
(March-
May)
— Plateau
only,
summer
(June-mid-
October)
Spring
(March-
May) and
fall
(September-
November)
Target #
organisms
200 ± 20%
200 ± 20%
110 ±10%
Taxonomic
resolution
Family (all
insects)
Genus (all insects
minus
Collembola)
Family (working
toward developing
a genus-level
index)
-------
Table F-3. continued...
Entity
SRBC
Project or
stream type
Aquatic
Resource
Surveys
Subbasin
Survey, Year
I/Interstate
Streams
Remote
Water
Quality
Monitoring
Network
Effort
6 kicks composited
or 5 minutes for a
distance of 5 m (PA
or NY)
2 kicks composited
6 kicks composited
Gear
D-frame
net/aquatic net
(30 cm x 20 to
23 cm x 46 cm
[PA or NY]);
500-um;
800 urn x
900 urn
(depending on
PA or NY)
Kick-net (1 m2)
with 600-um
mesh
D-Frame Net
(46 cm x
20 cm) with
500-um mesh
Habitat
Riffle -run
Sampling
area
6 m2or
distance of
5 m (PA or
NY)
2m2
6 m2
Index
period
Typically
late April
into May,
late June into
July, and
October
Year 1—
historically
spring-fall,
now
spring-May
30.
Interstate —
May (Group
3) or August
(Group 1 and
2); varies
depending
on site
classification
October
Target #
organisms
PADEP or
NYSDEC
protocol
200 ± 20%
200 ± 20%
Taxonomic
resolution
Genus, except
Chironomidae,
snails, clams
mussels (family);
Nematoda,
Nemertea,
Bryozoa
(phylum);
Turbellaria,
Hirudenia,
Oligochaeta
(class); water
mites (artificial)
-------
Table F-3. continued...
Entity
NFS
Project or
stream type
Eastern
Rivers and
Mountains
Network
Effort
A semiquantitative
sample consisting of
5 discrete collections
from the richest
targeted habitat
(typically riffle,
main-channel,
coarse-grained
substrate habitat
type) are processed
and combined into a
single composited
sample.
Gear
Slack sampler,
500-um nets
and sieves
Habitat
Riffle
Sampling
area
Each
discrete
sample =
0.25 m2
area; total
area
sampled =
1.25m2
Index
period
April-early
June
Target #
organisms
300
Taxonomic
resolution
Genus, except
Chironomidae,
snails, clams
mussels (family);
Nematoda,
Nemertea,
Bryozoa
(phylum);
Turbellaria,
Hirudenia,
Oligochaeta
(class); water
mites (artificial)
-------
Table F-4. Macroinvertebrate methods used by Southeast states for routine sampling events in medium-high
gradient freshwater wadeable streams with riffle habitat and rocky substrate
Entity
AL
DEM
GA
DNR
Stream
type
WMB-I
protocols
High
(riffle/run)
gradient
Effort
Several samples
are collected at a
site by stream
habitat type; each
sample is
processed
separately; the
taxa lists are
recombined after
standardizing
individual counts
to density units
20 jabs from
multiple habitats
are composited
Gear
Kick net, 2
A-frame nets, 2
#30 sieve buckets,
2 #30 sieves,
plastic elutriation
treys, 100%
denatured ethanol,
and plastic sample
containers
D -frame net
(30-cm width)
with 500-um net
mesh
Habitat
Riffle, rock-log,
Rootbank, CPOM,
sand, and
macrophytes
(macrophytes not
always available
and excluded from
index)
Multihabitat —
riffles, woody
debris/snags,
undercut
banks/rootwads,
leafpacks, soft
sediment/sandy
substrate, and
submerged
macrophytes (when
present)
Sampling
area
Approximately
4m2
20 jabs, each
for a linear
distance of
1m
Index
period
Late April-
early July
Mid-
September-
February
Target #
organisms
100
organisms
per habitat
200 ± 20%
Taxonomic
resolution
Genus or lowest
possible level
Lowest practical
level (generally
genus or
species)
to
-------
Table F-4. continued...
Entity
KYDEP
NC
DENR
Stream type
Wadeable,
moderate/hig
h gradient
streams
Headwater,
moderate/
high gradient
streams
Standard
qualitative
method for
wadeable
flowing
streams and
rivers
Effort
Combination of
quantitative
(composite of 4
riffle kicks) and
qualitative
(multihabitat)
samples
Composite of 2
kicks, 3 sweeps,
1 leaf pack
sample, 2 fine
mesh rock and/or
log wash
samples, 1 sand
sample and visual
collections from
habitats and
substrate types
missed or
under-sampled
by the other
collection
techniques
Gear
Quantitative
— kick net
(600-um
mesh);
qualitative — d
ip net, mesh
bucket,
forceps
r
600-um mesh
Multiple gear
types
(kick net with
600-um
mesh;
triangular
sweep net;
fine-mesh
samplers
[300-um
mesh]; sieve
bucket)
Habitat
Quantitative
samples are taken
from riffles;
qualitative are
taken from
multiple habitats
(undercut
banks/roots, wood,
vegetation, leaf
packs, soft and
rocky substrates)
Multihabitat
(riffles, bank areas,
macrophyte beds,
woody debris, leaf
packs, sand, etc.)
Sampling
area
1m2
(quantitative)
NA
(qualitative
only)
Index
period
Summer
(June-Septe
mber)
Spring
index period
(February-
May)
Year-round
Target #
organisms
300
Organisms
are field
picked
roughly in
proportion
to their
abundance.
Abundance
data are
recorded as
rare (1-2
specimens),
common
(3-9
specimens)
or abundant
specimens)
Taxonomic
resolution
Lowest practical
level (generally
genus or
species)
All of the
field-picked
organisms are
identified in the
laboratory to the
lowest practical
level (generally
genus or
species)
-------
Table F-4. continued...
Entity
SC
DHEC
TNDEC
Stream type
Ambient
monitoring
Streams with
riffles
Effort
Gear
Habitat
Sampling
area
Same as NC DENR
Single habitat,
semiquantitative ;
composite of
2 riffle kicks
Kick net
(1-m2,
500-um
mesh)
Riffle
2m2
Index
period
Feb 1 to
March 15:
Middle
Atlantic
Coastal
Plain
Ecoregion
(EPA Level
III 63); June
15 to Sept 1:
Statewide,
minus EPA
Level III
Ecoregion
63
Year-round
Target # Taxonomic
organisms resolution
Same as NC DENR
200 ± 20% Genus level
-------
Table F-5. Macroinvertebrate methods used in national surveys conducted by EPA and USGS
Entity
EPA
National
Aquatic
Resource
Surveys
USGS
Project or
stream type
WSAand
NRSA
NAWQA
Effort
A 0.1 -m2 area was
sampled for
30 seconds at a
randomly selected
location at each of
the 11 transects. The
samples were
composited into one
sample per site.
A semiquantitative
sample consisting of
5 discrete collections
from the richest
targeted habitat
(typically riffle,
main -channel,
coarse-grained
substrate habitat
type) are processed
and combined into a
single composited
sample.
Gear
Modified
D-frame
net (30 cm
wide) with
500-um
mesh
Slack
sampler,
500-um
nets and
sieves
Habitat
Multihabitat
Composite
Riffle
Sampling
area
Approximately
1m2
Each discrete
sample =
0.25-m2 area;
total area
sampled =
1.25m2
Index
period
June-
September
Late
June-mid-
October
Target #
organisms
500
300
Taxonomic
resolution
Genus level
Lowest
practical
level
-------
[This page intentionally left blank.]
-------
APPENDIX G.
LEVEL OF TAXONOMIC
RESOLUTION
G-l
-------
When possible, all taxa should be taken to the lowest practical taxonomic level (ideally
species level). If this is not possible, efforts should be made to identify the taxa listed in Table
G-l to the level of resolution described in the table. Ephemeroptera, Plecoptera, Trichoptera, and
Chironomidae that are not listed in Table G-l should be identified to at least the genus level,
where possible.
The taxa in Table G-l were selected based on differences in thermal tolerances that were
evident in analyses (U.S. EPA, 2012; unpublished Northeast pilot study) and from best
professional judgment. The list in Table G-l should be regarded as a starting point and should be
updated as better data become available in the future. Table G-2 contains a list of taxa that were
considered for inclusion in Table G-l but for various reasons, were not selected.
G-2
-------
Table G-l. At RMN sites, we recommend that the taxa listed below be taken
to the specified level of resolution, where practical. The Chironomidae
require a slide mount and a compound microscope to identify to the
species-level.
Order
Coleoptera
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Plecoptera
Family
Elmidae
Chironomidae
Chironomidae
Chironomidae
Simuliidae
Baetidae
Ephemerellidae
Perlidae
Genus
Promoresia
Eukiefferiella
Polypedilum
Tvetenia
Baetis
Ephemerella
Acroneuria
Level of
resolution
adults to
species
species
species
species
group
genus
species
species (as
maturity
allows)
species
Notes
Potential variability in thermal
preferences of P. tardella (cold)
and P. elegans (warm).
Potential variability in thermal
preferences of E. brevicalar, E.
brehmi, and E. tirolemis (cold);
and E. claripennis and E.
devonica (warm).
P. aviceps is generally regarded
as a cold water taxon.
T. vi trades is warm water
oriented in the Northeast.
General agreement that
Prosimilium is a cold water
indicator but there is potential
for variability within this genus
(e.g., P. mixtum vs. P. vernale),
and species-level systematics are
not well developed at this time.
Potential variability in thermal
preferences (e.g., B.
tricaudatus — cold; B.
intercalaris and B.
flavistriga — warm) .
Potential variability in thermal
preferences (e.g., E.
subvaria — colder); need mature
individuals (early instars are
difficult to speciate).
Potential variability in thermal
preferences of A. abnormis
(warmer) and A. carolinensis
(cooler).
G-3
-------
Table G-l. continued...
Order
Plecoptera
Plecoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Family
Perlidae
Pteronarcyidae
Brachycentridae
Hydropsychidae
Rhyacophilidae
Uenoidae
Genus
Paragnetina
Pteronarcys
Brachycentrus
Ceratopsyche
Rhyacophila
Neophylax
Level of
resolution
species
species
species
species
species
species
Notes
Potential variability in thermal
preferences of P. immarginata
(cold) and P. media and P.
kansanensis.
P. dorsata may be warmer water
oriented.
Potential variability in thermal
preferences in the Northeast.
Potential variability in thermal
preferences.
Most species are cold water, but
some variability has been
documented in the Northeast
(U.S. EPA, 2012, unpublished
data).
Some variability was noted in a
pilot study in North Carolina
(U.S. EPA, 2012).
G-4
-------
Table G-2. Taxa that were considered for inclusion in Table G-l
Order
Coleoptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Family
Elmidae
Chironomidae
Ceratopogonidae
Ephemerellidae
Ephemerellidae
Heptageniidae
Heptageniidae
Genus
Oulimnius
Micropsectra
Drunella
Eurylophella
Epeorus
Stenacron
Level of
resolution
species
species
species
species
species
species
species
Notes
O. latiusculus is regarded as a
cold-water taxon in Vermont, but
species-level IDs may not be
necessary for the larger region
because most of the taxa are O.
latiusculus.
General agreement that there is
variability in thermal preferences,
but the taxonomy for this genus
needs to be further developed.
General agreement that there is
variability in thermal preferences,
but the taxonomy for this family
needs to be further developed.
Variability in thermal tolerances
within this genus was noted in the
Utah pilot study, but in the
Eastern states, species are
believed to be all cold/cool water.
Some variability was noted in a
pilot study in North Carolina
(U.S. EPA, 20 12); could be
seasonal phenology vs. thermal
preference.
Some variability was noted in a
pilot study in Utah (U.S. EPA,
2012); can be difficult to speciate.
In the Mid-Atlantic region, some
regard S. interpunctatum as a
warm-water taxon and the others
as cooler/some cold. Taxonomy
may be tricky.
G-5
-------
Table G-2. continued...
Order
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Oligochaeta
Family
Goeridae
Hydropsychidae
Leptoceridae
Philopotamidae
Genus
Goera
Hydropsyche
Oecetis
Chimarra
Level of
resolution
species
species
species
species
family
Notes
Some variability was noted in a
pilot study in North Carolina
(U.S. EPA, 2012). The two species
found in Kentucky are associated
with cold water. In New Jersey,
this genus is found as often in the
coastal plain as in northern high
gradient streams and is currently
not taken to the species level.
Some variability was noted in a
pilot study in New England
(U.S. EPA, 2012, unpublished
data) but is generally considered to
be eurythermal (not sure which
species would be regarded as cold
water taxa).
Some variability was noted in a
pilot study in North Carolina
(U.S. EPA, 2012). The species
found in Kentucky are associated
with warm water. In New Jersey,
this genus is typically found in low
gradient coastal plain streams.
Some variability was noted in a
pilot study in New England
(U.S. EPA, 2012, unpublished
data) but most species were
warm-water oriented. C. obscura
and C. atterima predominate, but
tend to co-occur.
Enchytraeidae is regarded as a
cold-water family in Vermont. In
the Mid- Atlantic region, it is found
mostly in small streams. In New
Jersey, it is found throughout the
state.
G-6
-------
Table G-2. continued.
Order
Amphipoda
Amphipoda
Isopoda
Neoophora
Neoophora
Family
Gammaridae
Hyalellidae
Asellidae
Planariidae
Dugesiidae
Genus
Gammarus
Hyallela
Caecidotea
Dugesia
Cura
Level of
resolution
species
species
species
species
species
Notes
G. pseudolimnaeus is regarded as a
cold- or cool-water taxon in
Vermont (and is tolerant of
nutrients). Gammarus (assumed to
be pseudolimnaeus) is also
regarded as a cold-water indicator
in Minnesota (Gerritsen and
Stamp, 2012).
H. azteca is regarded as a cold/cool
water taxon in Vermont. In
Kentucky, Hyallela it is believed to
be a completely warm-water genus.
C. brevicauda has been noted as a
potential cold-water indicator in
the Midwest (Gerritsen and Stamp,
2012).
D. tigrina is regarded as a
warm-water taxon in Vermont, as
well as in New Jersey. Can be
difficult to speciate in speciose
regions.
C.formanii is regarded as a
cold-water taxon in Vermont. Can
be difficult to speciate in speciose
regions.
G-7
-------
G.I. LITERATURE CITED:
Gerritsen, J; Stamp, J. (2012) Calibration of the biological condition gradient (BCG) in cold and
cool waters of the upper Midwest for fish and benthic macroinvertebrate assemblages
[Final Report]. Prepared by Tetra Tech, Inc. for the USEPA Office of Water and USEPA
Region 5. Owings Mills, MD: Tetra Tech. Available online at http://www.uwsp.edu/cnr-
ap/biomonitoring/Documents/pdf/USEP A-BCG-Report-Final-2012.pdf
U.S. EPA (Environmental Protection Agency). (2012) Implications of climate change for
bioassessment programs and approaches to account for effects. [EPA/600/R-11/036F].
Washington, DC: Global Change Research Program, National Center for Environmental
Assessment. Available online at
http://cfpub.epa.gov/ncea/gl obal/recordisplay.cfm?deid=239585
G-8
-------
APPENDIX H.
SUMMARIZING
MACROINVERTEBRATE
DATA
H-l
-------
H.l. LIST OF CANDIDATE BIOLOGICAL INDICATORS
Table H-l. Recommendations on candidate biological indicators to
summarize from the macroinvertebrate data collected at regional monitoring
network (RMN) sites; many of these indicators are commonly used by
biomonitoring programs for site assessments.
Type of
indicator
Taxonomic-
based metric
Biological indicator
Total number of taxa
(richness)
Number of EPT taxa
(Ephemeroptera
[mayflies], Plecoptera
[stoneflies], and
Trichoptera [caddisflies])
Number of Ephemeroptera
(mayfly) taxa
Number of Plecoptera
(stonefly) taxa
Number of Trichoptera
(caddisfly) taxa
Percentage EPT
individuals
Percentage Ephemeroptera
individuals
Percentage Plecoptera
individuals
Percentage Trichoptera
individuals
Number of Odonata,
Coleoptera, and
Hemiptera (OCH) taxa
Percentage OCH
individuals
Expected response
Predicted to decrease
in response to
increasing
anthropogenic stress
Expected to be more
prevalent during
summer, low flow
(more pool-like)
periods
Source
Barbour et al., 1999
(compiled from DeShon,
1995; Barbour etal.,
1996; Fore etal., 1996;
Smith and Voshell,
1997); these metrics are
commonly used in
bioassessments
Bonada et al., 2007a
H-2
-------
Table H-l. continued...
Type of
indicator
Traits-based
metric related
to temperature
(for list of
thermal
indicator taxa,
see
Appendix I)
Traits-based
metric related
to hydrology
Biological indicator
Number of cold water taxa
Percentage cold water
individuals
Number of warm water
taxa
Percentage warm water
individuals
Collector filterer
Collector gatherer
Scraper/herbivore
Shredder
Predator
Swimmer
Rheophily — deposit! onal
Rheophily — erosional
Expected response
Predicted to decrease
in response to
warming
temperatures
Predicted to increase
in response to
warming
temperatures
Predicted to decrease
during low flow
conditions
Predicted to increase
during slow velocity
conditions
Predicted to increase
during conditions of
stable flow and
habitat availability;
decrease during
drought conditions
Expected to respond
to changing thermal
and hydrologic
conditions
Predicted to increase
during low flow
conditions
Predicted to comprise
higher proportion of
assemblage during
drier, harsher,
climatic conditions
Favor low flow/slow
velocity conditions
Favor high flow/fast
velocity conditions
Source
Lake, 2003; Hamilton
etal., 2010; Stamp etal.,
2010; U.S. EPA, 2012
Wills et al., 2006; Bogan
and Lytle, 2007; Walters
and Post, 2011
Heino, 2009
Richards et al., 1997;
McKay and King, 2006;
Wills et al., 2006;
Fenoglio et al., 2007;
Griswold et al., 2008;
Diaz et al., 2008
Richards et al., 1997;
Buzby and Perry, 2000;
McKay and King, 2006;
Foucreau et al., 2013
Bogan and Lytle, 2007;
Miller et al., 2007;
Walters and Post, 2011
Beche et al., 2006;
Bonada et al., 2007b;
Diaz et al., 2008
Richards et al., 1997;
Lake, 2003; Wills etal.,
2006; Poff etal., 2010;
Brooks etal., 2011
H-3
-------
Table H-l. continued...
Type of
indicator
Biological
condition
Individual taxa
Variability
Biological indicator
Bioassessment score (e.g.,
multimetric index,
predictive, biological
condition gradient)
Presence — absence
Relative abundance
Spatial distribution
Persistence (variability in
presence/absence; see
SectionH.2ofthis
appendix)
Stability (variability in
relative abundance; see
SectionH.2ofthis
appendix)
Expected response
Expected to worsen
in response to
increasing
anthropogenic stress
Hypotheses have
been developed for
some individual taxa
(e.g., the cold and
warm water taxa
listed in Appendix I)
Expect lower
persistence in
disturbed or
climatically harsh
environments
Expect lower stability
in disturbed or
climatically harsh
environments
Source
Barbour et al., 1995;
DeShon, 1995; Hawkins
et al., 2000; Davies and
Jackson, 2006
Becker etal., 2010
Rolling, 1973; Bradley
and Ormerod, 2001;
Milner et al., 2006;
Durance and Ormerod,
2007
Scarsbrook, 2002; Milner
et al., 2006
H.2. FORMULAS FOR CALCULATING PERSISTENCE AND STABILITY
Persistence between samples can be calculated using Jaccard's similarity coefficient (J):
J(AB) =
a + b-j
(H-l)
Herey is the number of taxa common to both years (or sites) A and B, while a and b are the
number of taxa in year (or site) A and B, respectively. It is interpreted as the proportion of taxa
common to both samples, such that values close to zero and one have low and high persistence,
respectively.
Stability, on the other hand, can be calculated using Bray-Curtis similarity (BC [Bray and
Curtis, 1957]):
H-4
-------
BC(AB)=\ - -
(H-2)
Here TIAI and risi are the number of individuals of taxa /' in year (or site) A and B, and NA and NB
are the total number of individuals in year (or site) A and B, respectively. It is interpreted as the
proportion of individuals (rather than taxa) common to both samples, such that values close to
zero and one have low and high stability, respectively.
As an example, we calculate persistence and stability using Jaccard and Bray-Curtis
similarities with the data in Table H-2:
BC(AB) = 1 -
= =0.60
|0 - 35| + |5 - 5| + |8 -
23 + 73
|0 -
_ 9 + 35 + 0 + 5 + 1 _ 50
23 +73 96
(H-3)
(H-4)
(H-5)
Table H-2. Sample data for calculating persistence and stability
Samples
Sample year (or site) A
Sample year (or site) B
TaxaV
10
19
TaxaW
0
35
TaxaX
5
5
TaxaY
8
13
TaxaZ
0
1
Sum
23
73
High persistence and stability are thought to occur where environmental conditions are similar or
relatively constant, or where change occurs incrementally. For additional background and an
example of these techniques applied to long-running surveys in Alaskan streams, see Milner
et al. (2006). At these sites, mean persistence and stability between study years ranged from 0.49
to 0.70 and from 0.29 to 0.44, respectively, which suggests that even among the most persistent
sites, substantial year-to-year shifts in relative abundances can occur.
H.l. REFERENCE:
Barbour, MT; Stribling, JB; Karr, J. (1995) Multimetric approach for establishing biocriteria and measuring
biological condition. In Davis, W.S. & T.P. Simon (eds). Biological Assessment and Criteria; Tools for Water
Resource Planning and Decision Making. Lewis Publishers. Boca Raton, FL: 63-77.
H-5
-------
Barbour, MT; Gerritsen, J; Griffith, GE; Frydenborg, R; McCarron, E; White, JS; Bastian, ML. (1996) A framework
for biological criteria for Florida streams using benthic macroinvertebrates. JN AmBenthol Soc 15(2):185-211.
Barbour, MT; Gerritsen, J; Snyder, BD; Stribling, JB. (1999) Rapid bioassessment protocols for use in streams and
wadeable rivers: Periphyton, benthic macroinvertebrates and fish, Second Edition. [EPA 841-B-99-002].
Washington, D.C: U.S. Environmental Protection Agency, Office of Water. Available online:
http://water.epa.gov/scitech/monitoring/rsl/bioassessment/index.cfm.
Beche, LA; McElracy, EP; Resh, VH. (2006) Long-term seasonal variation in the biological traits of benthic-
macroinvertebrates in two Mediterranean-climate streams in California U.S.A. Freshw Biol 51:56-75.
Becker, AJ; Stranko, SA; Klauda, RJ; Prochaska, AP; Schuster, JD; Kashiwagi, MR; Graves, PH. (2010) Maryland
biological stream survey's sentinel site network: A multi-purpose monitoring program. Annapolis, MD: Maryland
Department of Natural Resources.
Bonada, N; Rieradevall, M; Prat, N. (2007a). Macroinvertebrate community structure and biological traits related to
flow permanence in a Mediterranean river network. Hydrobiologia 589:91-106.
Bonada, N; Doledec, S; Statzner, B. (2007b). Taxonomic and biological trait differences of stream
macroinvertebrate communities between Mediterranean and temperate regions: implications for future climatic
scenarios. Global Change Biol 13:1658-1671.
Bogan, MT; Lytle, DA. (2007) Seasonal flow variation allows 'time-sharing' by disparate aquatic insect
communities in montane desert streams. Freshw Biol 52(2):290-304. doi: 10.1111/j.1365-2427.2006.01691.x
Bradley, DC; Ormerod, SJ. (2001) Community persistence among upland stream invertebrates tracks the North
Atlantic Oscillation. J Animal Ecol 70(6):987-996.
Brooks, A; Chessman, B; Haeusler, T. (2011) Macroinvertebrate traits distinguish unregulated rivers subject to
water abstraction. J N Am Benthol Soc 30(2):419-435.
Buzby, KM; Perry, SA. (2000). Modeling the potential effects of climate change on leaf pack processing in central
Appalachian streams. Can J Fish Aquatic Sci 57(9):1773-1783.
Davies, SP; Jackson, SK. (2006) The Biological Condition Gradient: A descriptive model for interpreting change in
aquatic ecosystems. Ecol Appl 16(4): 1251-1266.
DeShon, J. (1995) Development and application of the Invertebrate Community Index (ICI). [p. 217-243]. In Davis,
WS; Simon, TP (eds). Biological assessment and criteria: tools for water resource planning and decision making.
Boca Raton: CRC Press.
Diaz, AM; Suarez Alonso, ML; Vidal-Abarca Gutierrez, MR. (2008) Biological traits of stream macroinvertebrates
from a semi-arid catchment: patterns along complex environmental gradients. Freshw Biol 53:1-21.
Durance, I; Ormerod, SJ. (2007) Climate change effects on upland stream macroinvertebrates over a 25-year period.
Glob Change Biol 13:942-957.
Fenoglio, S; Bo, T; Cucco, M; Malacarne, G. (2007) Response of benthic invertebrate assemblages to varying
drought conditions in the Po river (NW Italy). Ital J Zool 74(2): 191-201.
Fore, LS; Karr, JR; Wisseman, RW. (1996) Assessing invertebrate responses to human activities: Evaluating
alternative approaches. J N Am Benthol Soc 15(2):212-231.
Foucreau, N; Piscart, C; Puijalon, S; Hervant, F. (2013) Effect of climate-related change in vegetation on leaf litter
consumption and energy storage by Gammarus pulex from Continental or Mediterranean populations. PLoS ONE
8(10): e77242.
H-t
-------
Griswold, MW; Berzinis, RW; Crisman, TL; Golladay, SW. (2008) Impacts of climatic stability on the structural
and functional aspects of macroinvertebrate communities after severe drought. Freshw Biol 53(12):2465-2483.
doi: 10.1111/j. 1365-2427.2008.02067.X.
Hamilton, AT; Stamp, J; Bierwagen, BG. (2010) Vulnerability of biological metrics and multimetric indices to
effects of climate change. J N AmBenthol Soc 29(4): 1379-1396.
Hawkins, CP; Norris, RH; Hogue, JN; Feminella, JW. (2000) Development and evaluation of predictive models for
measuring the biological integrity of streams. Ecol Appl 10:1456-1477.
Heino, J. (2009) Species co-occurrence, nestedness and guild- environment relationships in stream
macroinvertebrates. Freshw Biol 54(9): 1947-1959.
Holling, CS. (1973) Resilience and stability of ecological systems. Ann Rev Ecol Systems 4:1-23.
Lake, PS. (2003) Ecological effects of perturbation by drought in flowing waters. Freshw Biol 48:1161-1172.
McKay, SF; King, AJ. (2006) Potential ecological effects of water extraction in small, unregulated streams. River
Res Appl 22:1023-1037.
Miller, S; Wooster, D; Li, J. (2007) Resistance and resilience of macroinvertebrates to irrigation water withdrawals.
Freshw Biol 52:2494-2510.
Milner, AM; Conn, SC; Brown, LE. (2006) Persistence and stability of macroinvertebrate communities in streams of
Denali National Park, Alaska: implications for biological monitoring. Freshw Biol 51:373-387.
Poff, NL; Allan, JD; Bain, MB; Karr, JR; Prestegaard, KL; Richter, BD; Sparks, RE; Stromberg, JC. (1997) The
natural flow regime: a paradigm for river conservation and restoration. BioScience 47(11):769-784.
Richards, C; Haro, RJ; Johnson, LB; Host, GE. (1997) Catchment and reach-scale properties as indicators of
macroinvertebrate species traits. Freshw Biol 37:219-230.
Scarsbrook MR. (2002) Persistence and stability of lotic invertebrate communities in New Zealand. Freshw Biol
47:417-431.
Smith, EP; Voshell, Jr, JR. (1997) Studies of benthic macroinvertebrates and fish in streams within EPA Region 3
for development of biological indicators of ecological condition. Blacksburg, VA: Virginia Polytechnic Institute and
State University.
Stamp, J; Hamilton, A; Zheng, L; Bierwagen, B. (2010) Use of thermal preference metrics to examine state
biomonitoring data for climate change effects. J N Am Benthol Soc 29(4): 1410-1423.
U.S. EPA (Environmental Protection Agency). (2012) Implications of climate change for state bioassessment
programs and approaches to account for effects. [EPA/600/R-11/036F]. Washington, DC: Global Change Research
Program, National Center for Environmental Assessment.
http://cfpub.epa. gov/ncea/global/recordisplay.cfm?deid=239585.
Walters, AW; Post, D. (2011) How low can you go? Impacts of a low-flow disturbance on aquatic insect
communities. Ecol Appl 21:163-174.
Wills, TC; Baker, EA; Nuhfer, AJ; Zorn, TG. (2006) Response of the benthic macroinvertebrate community in a
northern Michigan stream to reduced summer streamflows. River Res Appl 22:819-836.
H-7
-------
H.2. REFERENCE:
Bray, JR; Curtis, JT. (1957) An ordination of the upland forest communities of southern Wisconsin. Ecol Monogr
27:325-349.
Milner, AM; Conn, SC; Brown, LE. (2006) Persistence and stability of macroinvertebrate communities in streams of
Denali National Park, Alaska: implications for biological monitoring. Freshw Biol 51:373-387.
-------
APPENDIX I.
MACROINVERTEBRATE
THERMAL INDICATOR
TAXA
1-1
-------
This appendix contains lists of macroinvertebrate taxa that are believed to have strong
thermal preferences based on analyses conducted by EPA (U.S. EPA, 2012; unpublished
Northeast pilot study) and state biomonitoring programs (Maryland Department of Natural
Resources [MD DNR], Pennsylvania Department of Environmental Protection [PA DEP],
Tennessee Department of Environment and Conservation [TN DEC], Vermont Department of
Environmental Conservation [VT DEC]). Best professional judgment from regional taxonomists
was also considered.
Table 1-1 contains a list of cold- and warm-water preference taxa (benthic
macroinvertebrates) for the eastern United States, based on Generalized Additive Models
(GAM)—full temperature range model rank results from the Northeast pilot study
(U.S. EPA, 2012) and the best professional judgment of regional experts.
Table 1-2 contains lists of taxa that have been identified as thermal indicators by VT DEC
(Steve Fiske, Aaron Moore, and Jim Kellogg, unpublished).
Table 1-3 contains the list of taxa that have been identified as cold water taxa by MD
DNR (Becker et al., 2010) and also contains information that was provided by PA DEP
(Amy Williams and Dustin Shull, unpublished data).
Table 1-4 contains a list of indicator taxa identified based on thermal tolerance analyses
(per Yuan, 2006) conducted on data from North Carolina (U.S. EPA, 2012), and also
contains information that was provided by Debbie Arnwine from TN DEC.
All of these lists are intended to be starting points, which can be revised as more data
become available.
1-2
-------
Table 1-1. List of cold- and warm-water preference taxa (benthic
macroinvertebrates) for the eastern United States, based on GAM—full
temperature range model rank results from the Northeast pilot study
(EPA, 2012) and the best professional judgment of regional experts
Order
Coleoptera
Coleoptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Odonata
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Family
Elmidae
Elmidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Dixidae
Psychodidae
Simuliidae
Ameletidae
Baetidae
Baetidae
Ephemerellidae
Ephemerellidae
Ephemerellidae
Heptageniidae
Heptageniidae
Heptageniidae
Leptophlebiidae
Gomphidae
Capniidae
Chloroperlidae
Nemouridae
Peltoperlidae
Genus
Oulimnius
Promoresia
Brillia
Eukiefferiella
Eukiefferiella
Heleniella
Parachaetocladius
Polypedilum
Dixa
Pericoma
Prosimulium
Ameletus
Baetis
Diphetor
Drunella
Ephemerella
Eurylophella
Cinygmula
Epeorus
Rhithrogena
Habrophlebia
Lanthus
Peltoperla
Final identification
Oulimnius
Promoresia tardella
Brillia
Eukiefferiella
brevicalcar
Eukiefferiella
claripennis
Heleniella
Parachaetocladius
Polypedilum tritum
Dixa
Pericoma
Prosimulium
Ameletus
Baetis tricaudatus
Diphetor
Drunella
Ephemerella
Eurylophella
Cinygmula
Epeorus
Rhithrogena
Habrophlebia
Lanthus
Capniidae
Chloroperlidae
Nemouridae
Peltoperla
Indicator
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
1-3
-------
Table 1-1. continued...
Order
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
-
Basommatophora
Coleoptera
Coleoptera
Diptera
Diptera
Family
Peltoperlidae
Perlodidae
Perlodidae
Pteronarcyidae
Taeniopterygidae
Taeni opterygi dae
Apataniidae
Brachycentridae
Glossosomatidae
Glossosomatidae
Hydropsychidae
Hydropsychidae
Hydropsychidae
Hydropsychidae
Hydropsychidae
Hydropsychidae
Hydroptilidae
Odontoceridae
Philopotamidae
Philopotamidae
Rhyacophilidae
-
Physidae
Elmidae
Hydrophilidae
Chironomidae
Chironomidae
Genus
Tallaperla
Isoperla
Malirekus
Pteronarcys
Taenionema
Taeniopteryx
Apatania
Brachycentrus
Agapetus
Glossosoma
Arctopsyche
Ceratopsyche
Ceratopsyche
Ceratopsyche
Diplectrona
Parapsyche
Palaeagapetus
Psilotreta
Dolophilodes
Wormaldia
Rhyacophila
-
Physella
Stenelmis
Berosus
Ablabesmyia
Cardiocladius
Final identification
Tallaperla
Isoperla
Malirekus
Pteronarcys
Taenionema
Taeniopteryx
Apatania
Brachycentrus
americanus
Agapetus
Glossosoma
Arctopsyche
Ceratopsyche alhedra
Ceratopsyche macleodi
Ceratopsyche ventura
Diplectrona
Parapsyche
Palaeagapetus
Psilotreta
Dolophilodes
Wormaldia
Rhyacophila
Turbellariaa
Physella
Stenelmis
Berosus
Ablabesmyia
Cardiocladius
Indicator
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Warm
Warm
Warm
Warm
Warm
Warm
1-4
-------
Table 1-1. continued...
Order
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Neotaenioglossa
Odonata
Odonata
Odonata
Odonata
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Veneroida
Family
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Chironomidae
Baetidae
Caenidae
Heptageniidae
Leptohyphidae
Hydrobiidae
Coenagrionidae
Coenagrionidae
Corduliidae
Corduliidae
Hydropsychidae
Hydroptilidae
Leptoceridae
Philopotamidae
Polycentropodidae
Pisidiidae
Genus
Dicrotendipes
Glyptotendipes
Nilotanypus
Nilothauma
Pentaneura
Polypedilum
Polypedilum
Stenochironomus
Tanytarsus
Tvetenia
Baetis
Caenis
Stenacron
Tricorythodes
Argia
Ischnura
Helocordulia
Macromia
Macrostemum
Hydroptila
Oecetis
Chimarra
Neureclipsis
Sphaerium
Final identification
Dicrotendipes
Glyptotendipes
Nilotanypus
Nilothauma
Pentaneura
Polypedilum convictum
Polypedilum flavum
Stenochironomus
Tanytarsus
Tvetenia vitracies
Baetis intercalaris
Caenis
Stenacron
Tricorythodes
Hydrobiidae
Argia
Ischnura
Helocordulia
Macromia
Macrostemum
Hydroptila
Oecetis
Chimarra obscura
Neureclipsis
Sphaerium
Indicator
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
aClass Turbellaria
1-5
-------
Table 1-2. Thermal indicator taxa that have been identified by VT DEC
(Steve Fiske, Aaron Moore and Jim Kellogg, unpublished data)
Order
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Plecoptera
Plecoptera
Plecoptera
Decapoda
Trichoptera
Diptera
Coleoptera
Coleoptera
Amphipoda
Amphipoda
Neoophora
Diptera
Diptera
Diptera
Genus
Polypedilum
Neostempellina
Tvetenia
Rhithrogena
Ameletus
Arctopsyche
Arctopsyche
Rhyacophila
Rhyacophila
Rhyacophila
Rhyacophila
Rhyacophila
Peltoperla
Tallaperla
Taenionema
Cambarus
Palaeagapetus
Eukiefferella
Oulimnius
Promoresia
Gammarus
Hyallela
Cura
Eukiefferella
Polypedilum
Tvetenia
Species
aviceps
reissi
bavarica grp
sp
sp
sp
ladogensis
Carolina
torva
nigrita
invaria
acutiloba
sp
sp
sp
bartoni
sp
brevicalar, brehmi, and tirolensis
latiusculus
tardella
pseudolimnaeus
azteca
formanii
claripennis
flavum
discoloripes, vitracies
Indicator
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold/cool
Cold/cool
Cold
Warm
Warm
Warm
1-6
-------
Table 1-2. continued...
Order
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Plecoptera
Plecoptera
Coleoptera
Neoophora
Genus
Leucotrichia
Rhyacophila
Rhyacophila
Rhyacophila
Neoperla
Taeniopteryx
Promoresia
Dugesia
Species
sp
mainensis
manistee
minora
sp
sp
elegans
tigrina
Indicator
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
1-7
-------
Table 1-3. Taxa that have been identified as cold or cool water indicators by
MD DNR (Becker et al., 2010) and/or PA DEP (Amy Williams and Dustin
Shull, unpublished data)
Type
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold (MD)/cool (PA)
Cold
Cold
Cold
Cold (MD)/cool (PA)
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Order
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Genus
Bittacomorpha
Dixa
Heleniella
Prodiamesa
Ameletus
Cinygmula
Diphetor
Drunella
Epeorus
Ephemera
Ephemerella
Eurylophella
Habrophlebia
Paraleptophlebia
Alloperla
Amphinemura
Diploperla
Haploperla
Isoperla
Leuctra
Malirekus
Peltoperla
Pteronarcys
Remenus
Sweltsa
MD
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
PA
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Occurrence in
PA DEP data set
Common
Common
Common
Common
Common
Common
Common
Rare
Common
Common
Rare
Rare
Common
Rare
Rare
Rare
Rare
Common
-------
Table 1-3. continued...
Type
Cold
Cold
Cold
Cold
Order
Plecoptera
Plecoptera
Trichoptera
Trichoptera
Genus
Tallaperla
Yugus
Diplectrona
Wormaldia
MD
Yes
Yes
Yes
PA
Yes
Yes
Yes
Occurrence in
PA DEP data set
Common
Rare
Common
1-9
-------
Table 1-4. Taxa that have been identified as cold, cool, or warm water
indicators based on thermal tolerance analyses (per Yuan, 2006) conducted
on data from North Carolina (U.S. EPA, 2012) and/or based on unpublished
data provided by Debbie Arnwine from TN DEC
Type
Cold (NC)/
cool (TN)
Cold (NC)/
cool (TN)
Cold (NC)/
cool (TN)
Cold
Cold
Cold
Cold
Cold
Cold (NC)/
cool (TN)
Cold
Cold
Cold
Cold
Cold (NC)/
cool (TN)
Cold (NC)/
cool (TN)
Cold
Cold (NC)/
cool (TN)
Cold (NC)/
cool (TN)
Cold
Order
Coleoptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Ephemeroptera
Odonata
Plecoptera
Genus
Promoresia
Antocha
Atherix
Cardiocladius
Diamesa
Dicranota
Eukiefferiella
Heleniella
Pagastia
Potthastia
Rheopelopia
Acentrella
Cinygmula
Drunella
Epeorus
Nixe
Rhithrogena
Lanthus
Amphinemura
NC
(U.S. EPA,
2012)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
TN
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Notes— TN
1-10
-------
Table 1-4. continued...
Type
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold
Cold (NC)/
cool (TN)
Cold
Cold
Cold
Cold
Cold
Cold/cool
Cold/cool
Cool
Cool
Order
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Plecoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Ephemeroptera
Trichoptera
Ephemeroptera
Plecoptera
Genus
Clioperla
Cultus
Diploperla
Isoperla
Malirekus
Peltoperla
Pteronarcys
Tallaperla
Zapada
Agapetus
Apatania
Arctopsyche
Dolophilodes
Glossosoma
Parapsyche
Ameletus
Lepidostoma
Habrophlebia
Alloperla
NC
(U.S.
EPA,
2012)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
TN
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Notes— TN
Uncommon in
TN data set
Uncommon in
TN data set
Uncommon in
TN data set
Uncommon in
TN data set
Uncommon in
TN data set
Mostly cool
or cold
Mostly cool
or cold
Uncommon in
TN data set
Uncommon in
TN data set
1-11
-------
Table 1-4. continued...
Type
Cool
Cool
Cool
Cool
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Order
Plecoptera
Plecoptera
Trichoptera
Trichoptera
Arhynchobdellida
Arhynchobdellida
Basommatophora
Coleoptera
Coleoptera
Decapoda
Diptera
Diptera
Diptera
Diptera
Diptera
Ephemeroptera
Ephemeroptera
Hemiptera
Isopoda
Odonata
Odonata
Odonata
Odonata
Genus
Sweltsa
Taenionema
Diplectrona
Wormaldia
Erpobdella
Mooreobdella
Physella
Berosus
Lioporeus
Palaemonetes
Nilothauma
Parachironomus
Pentaneura
Procladius
Stenochironomus
Diphetor
Tricorythodes
Belostoma
Caecidotea
Epicordulia
Helocordulia
Hetaerina
Ischnura
NC
(U.S.
EPA,
2012)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
TN
Yes
Yes
Yes
Yes
Yes
Notes— TN
Warm and
cold but
mostly cool
uncommon in
TN data set
Warm and
cold — more
common in
cool or cold
1-12
-------
Table 1-4. continued...
Type
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Warm
Order
Odonata
Odonata
Odonata
Rhynchobdellida
Rhynchobdellida
Trichoptera
Trichoptera
Trichoptera
Trichoptera
Unionoida
Genus
Macromia
Neurocordulia
Tetragoneuria
Helobdella
Placobdella
Chimarra
Macrostemum
Neureclipsis
Phylocentropus
Elliptic
NC
(U.S.
EPA,
2012)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
TN
Notes— TN
1.1. REFERENCES
Becker, AJ; Stranko, SA; Klauda, RJ; Prochaska, AP; Schuster, JD; Kashiwagi, MR; Graves, PH. (2010) Maryland
biological stream survey's sentinel site network: A multi-purpose monitoring program. Annapolis, MD: Maryland
Department of Natural Resources.
U.S. EPA (Environmental Protection Agency). (2012) Implications of climate change for bioassessment programs
and approaches to account for effects. [EPA/600/R-11/036F]. Washington, DC: Global Change Research Program,
National Center for Environmental Assessment. http://cfpub.epa.gov/ncea/global/recordisplay.cfm?deid=239585.
Yuan, L. (2006) Estimation and application of macroinvertebrate tolerance values. National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency,
Washington, DC; EPA/600/P-04/116F. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=154869.
1-13
-------
[This page intentionally left blank.]
-------
APPENDIX J.
THERMAL SUMMARY
STATISTICS
J-l
-------
Table J-l. Recommended thermal statistics to calculate for each year of
continuous temperature data at regional monitoring network (RMN) sites.
Timeframe
Daily
Monthly
Seasonal51
Thermal statistic
Daily mean
Daily maximum
Daily minimum
Daily difference
(maximum-minimum)
Daily variance
Monthly mean
Monthly maximum
Monthly minimum
Monthly difference
(maximum-minimum)
Monthly variance
Seasonal mean
Seasonal maximum
Seasonal minimum
Seasonal difference
(maximum-minimum)
Seasonal variance
Calculation
Mean temperature for each day
Maximum temperature for each day
Minimum temperature for each day
Difference between the maximum and
minimum temperatures for each day
Standard deviation for each day
Mean of the daily means for each month
Maximum value for each month
Minimum value for each month
Difference between the maximum and
minimum temperatures for each month
Standard deviation for each month
Mean of the daily means for each season
Maximum value for each season
Minimum value for each season
Difference between the maximum and
minimum temperatures for each season
Standard deviation for each season
J-2
-------
Table J.I. continued...
Timeframe
Annual
Thermal statistic
Annual mean
Annual maximum
Annual minimum
Mean annual difference
Maximum annual difference
Minimum annual difference
Variance of the annual mean
difference
Percentage exceedance
Calculation
Mean of the daily means for the year
(January 1 -December 31)
Maximum value for the year
(January 1 -December 31)
Minimum value for the year
(January 1 -December 31)
Mean of the daily difference
(January 1 -December 31)
Maximum of the daily difference
(January 1 -December 31)
Minimum of the daily difference
(January 1 -December 31)
Standard deviation of the daily difference
(January 1 -December 31)
([Number of measurements that exceed a
threshold13] + [total number of measurements
in a year]) x 100
aSeasons are defined as follows. Winter: December, January, February; Spring: March, April, May; Summer: June,
July, August; Fall: September, October, November.
bThresholds may vary by entity and location.
J-3
-------
[This page intentionally left blank.]
-------
APPENDIX K.
HYDROLOGIC SUMMARY
STATISTICS AND TOOLS
FOR CALCULATING
ESTIMATED
STREAMFLOW
STATISTICS
K-l
-------
K.l. LIST OF RECOMMENDED HYDROLOGIC SUMMARY STATISTICS FOR REGIONAL MONITORING
NETWORK (RMN) SITES
Table K-l. Recommended hydrologic statistics to calculate on each year of water-level or flow data from
regional monitoring network (RMN) sites. These provide information on high, seasonal, and low flow components to
maintain ecosystem flows. These candidate metrics were derived from DePhilip and Moberg (2013) for the Upper Ohio
River Basin and Olden and Poff (2003). Work that was done by Hawkins et al. (2013) was also considered. For more
details on these studies, see Sections K.2-K.4.
Timeframe
Daily
Metric
Daily mean
Daily median
Daily maximum
Daily minimum
Daily difference
(maximum-minimum)
Coefficient of variation
Calculation
Mean stage or flow for each day
Median stage or flow for each day
Maximum stage or flow for each day
Minimum stage or flow for each day
Difference between the maximum and minimum stage
Standard deviation for stage or flow for each day/mean
or flows for each day
daily stage or flow
to
-------
Table K-l. continued...
Timeframe
Monthly
Metric
Monthly mean
Monthly maximum51
Monthly minimumb
Monthly difference
(maximum-minimum)
High flow magnitude
(90th percentile)
Median magnitude (50th percentile)
Low flow magnitude
(25th percentile)
Low flow magnitude
(10th percentile)
Extreme low flow magnitude
(1st percentile)
Percentage high flow and floods
Percentage low flows
Calculation
Mean stage or flow for each month
Maximum stage or flow for each month
Minimum stage or flow for each month
Difference between the maximum and minimum stage or flow values for
each month
90th percentile of monthly stage or flow values; this represents high flows
and is similar to the Qio measurement used in DePhilip and Moberg (2013)
50th percentile of monthly stage or flow values; this represents the monthly
median
25th percentile of monthly stage or flow values; this represents low flows in
smaller streams (drainage areas <50 mi2, per DePhilip and Moberg [2013])
and is similar to the Q?s measurement used in DePhilip and Moberg (2013)
10th percentile of monthly stage or flow values; this represents low flows in
medium to larger-sized streams (drainage areas >50 mi2 per DePhilip and
Moberg [2013]) and is similar to the Qgo measurement used in DePhilip and
Moberg (20 13)
1st percentile of monthly stage or flow values; this represents extreme low
flows and is similar to the Q99 measurement used in DePhilip and Moberg
(2013)
Percentage of stage or flow measurements in each month that exceed the
monthly 90th percentile
Percentage of stage or flow measurements in each month that are between
the monthly 25th and 1st percentiles (similar to the Q?s and Q99 measurements
used in DePhilip and Moberg [2013])
-------
Table K-l. continued...
Timeframe
Seasonal
Annual
(January 1-
December 31)
Metric
Percentage typical
Percentage high flows and floods in
spring and fall
Annual mean
Annual maximum
Julian date of annual maximum
Annual minimum
Julian date of annual minimum
Mean annual difference
Maximum annual difference
Minimum annual difference
Variance of the annual mean
difference
Number of zero flow days
Calculation
Percentage of stage or flow measurements in each month that are between
the monthly 25th and 90th percentiles (similar to the Q?s and Qio
measurements used in DePhilip and Moberg [2013])
Percentage of stage or flow measurements in each month that exceed the
monthly 90th percentile in spring (March-May) and fall
(Septemb er-Novemb er)
Mean of the daily mean stage or flow
Maximum stage or flow
Julian date of annual maximum stage or flow
Minimum stage or flow
Julian date of annual minimum stage or flow
Mean of the daily difference
Maximum of the daily difference
Minimum of the daily difference
Standard deviation of the daily difference
Number of days having stage or flow measurements of 0
aln Olden and Poff (2003), mean maximum August flow and mean maximum October flow captured important aspects of high flow conditions.
bln Olden and Poff (2003), mean minimum April flow captured important aspects of low flow conditions.
-------
K.2. HYDROLOGIC METRICS DERIVED FROM DEPHILIP AND MOBERG (2013)
FOR THE UPPER OHIO RIVER BASIN
The Nature Conservancy and several partners (states, river basin commissions, other
federal agencies) have developed ecosystem flow needs for some Eastern and Midwestern rivers
and their tributaries (e.g., the Susquehanna, the Upper Ohio, and the Potomac Rivers) (DePhilip
and Moberg, 2010; Cummins et al., 2010; DePhilip and Moberg, 2013; Buchanan et al., 2013).
Table K-2 contains the lists of 10 flow statistics that were chosen to represent the high, seasonal,
and low flow components in the Upper Ohio River basin (DePhilip and Moberg, 2013). These
statistics were selected because they are easy to calculate, commonly used, and integrate several
aspects of the flow regime, including frequency, duration, and magnitude (DePhilip and Moberg,
2013). Diagrams like the one shown in Figure K-l can be generated for data from RMN sites.
Table K-2. Flow statistics that were selected to track changes to high,
seasonal, and low flow components in the Upper Ohio River basin. These are
flow exceedance values. For example, Qio equals the 10% exceedance probability
(Qio), which represents a high flow that has been exceeded only 10% of all days
in the flow period. This is a reproduction of Table 3.2 in DePhilip and Moberg
(2013)
Flow component
High flows
Annual/interannual (>bankfull)
Large flood
Small flood
Bankfull
High flow pulses ( monthly Qio in spring and fall
Monthly Qio
Monthly median
Area under monthly flow duration curve between Q?s
Qio (or some part of this range)
and
Area under monthly flow duration curve between Q?s
Q99
and
Monthly Q?s
Monthly Qgo
K-5
-------
loo-ao
Monthly Q75
-*^BIfcr
Monthly Q95
0 H
j i
Figure K-l. In the Upper Ohio River basin, monthly flow exceedance values
(Qex) were plotted against daily discharges to highlight specific portions of
the hydrograph and facilitate discussions about the ecological importance of
each portion (from DePhilip and Moberg, 2013).
K-6
-------
K.3. HYDROLOGIC METRICS DERIVED FROM OLDEN AND POFF (2003)
Olden and Poff (2003) did a comprehensive review of 171 hydrologic metrics, including Indicators of Hydrologic Alteration.
They provided recommendations on a reduced set of metrics that capture critical aspects of the hydrologic regime, are not overly
redundant, and are ecologically meaningful in different types of streams. Table K-3 contains a list of 34 metrics that, based on their
analyses, effectively capture different aspects of flow regimes in all stream types and have limited redundancy.
Table K-3. Based on analyses done by Olden and Poff (2003), these 34 hydrologic flow statistics effectively
capture different aspects of the flow regime in all stream types and have limited redundancy. This is a
reproduction of Table 3 (all streams) in Olden and Poff (2003)
Category
Magnitude —
average flow
conditions
Magnitude —
low flow
conditions
Metric
Skewness in daily flows
Mean annual runoff
Variability in daily flows 1
Spreads in daily flows
Baseflow index 1
Mean minimum April flow
Variability across annual
minimum flows
Variability in baseflow index 1
Description
Mean daily flows divided by median daily flows
Mean annual flow divided by catchment area
Coefficient of variation in daily flows
Ranges in daily flows (25th/75th percentiles) divided by median
daily flows
7-day minimum flow divided by mean annual daily flows
averaged across all years
Mean minimum monthly flow in April
Coefficient of variation in annual minimum flows averaged across
all years
Coefficient of variation in baseflow index (Ml 17)
Abbreviated
metric
Ma5
Ma41
Ma3
Mall
M117
M14
M121
M118
-------
Table K-3. continued...
Category
Magnitude —
high flow
conditions
Frequency of
flow
events — low
flow
conditions
Metric
High flow discharge
Mean maximum August flow
Mean maximum October flow
Median of annual maximum
flows
Frequency of low flow spells
Variability in low flow pulse
count
Low flow pulse count
Description
Mean of the 10th percentile from the flow duration curve divided
by median daily flow across all years
Mean maximum monthly flow in August
Mean maximum monthly flow in October
Median of the highest annual daily flow divided by the median
annual daily flow averaged across all years
Total number of low flow spells (threshold equal to 5% of mean
daily flow) divided by record length in years
Coefficient of variation in Fll
Number of annual occurrences during which the magnitude of
flow remains below a lower threshold. Hydrologic pulses are
defined as those periods within a year in which the flow drops
below the 25th percentile (low pulse) of all daily values for the
time period.
Abbreviated
metric
Mhl6
Mh8
MhlO
Mhl4
F13
F12
Fll
oo
-------
Table K-3. continued...
Category
Frequency of
flow
events — high
flow
conditions
Duration
Metric
High flood pulse count 2
Flood frequency
Flood frequency
Variability in high flood pulse
count
Number of zero flow days
Variability in low flow pulse
duration
Low flow pulse duration
Means of 30-day minimum
daily discharge
Means of 30-day maximum
daily discharge
Variability in high flow pulse
duration
High flow duration
High flow pulse duration
Description
Number of annual occurrences during which the magnitude of
flow remains above an upper threshold. Hydrologic pulses are
defined as those periods within a year in which the flow goes
above three times the median daily flow and the value is an
average instead of a tabulated count.
Mean number of high flow events per year using an upper
threshold of three times median flow over all years
Mean number of high flow events per year using an upper
threshold of seven times median flow over all years
Coefficient of variation in high pulse count (defined as 75th
percentile)
Mean annual number of days having zero daily flow
Coefficient of variation in low flow pulse duration
Mean duration of Fll
Mean annual 30-day minimum divided by median flow
Mean annual 30-day maximum divided by median flow
Coefficient of variation in Fhl
Upper threshold is defined as the 75th percentile of median flows
Mean duration of Fhl
Abbreviated
metric
Fh3
Fh6
Fh7
Fh2
D118
D117
D116
D113
Dhl3
Dhl6
Dh20
Dhl5
-------
Table K-3. continued...
Category
Timing of
flow events
Rate of
change
Metric
Constancy
Seasonal predictability of
nonflooding
Variability in Julian date of
annual minimum
Variability in reversals
Reversals
Change of flow
No day rises
Description
See Colwell( 1974)
Maximum proportion of the year (number of days/365) during
which no floods have ever occurred over the period of record
Coefficient of variation in Til
Coefficient of variation in Ra8
Number of negative and positive changes in water conditions from
1 day to the next
Median of difference between natural logarithm of flows between
2 consecutive days with increasing/decreasing flow
Ratio of days where flow is higher than the previous day
Abbreviated
metric
Tal
Th3
T12
Ra9
Ra8
Ra6
Ra5
-------
K.4. HYDROLOGIC METRICS USED BY HAWKINS ET AL. (2013)
Hawkins et al. (2013) used an iterative process to identify 16 streamflow variables that, in
their judgment, could characterize those general aspects of streamflow regimes relevant to
stream ecosystem structure and function. These variables are listed in Table K-4.
Table K-4. These 16 streamflow variables were selected by Hawkins et al.
(2013) to quantify aspects of hydrologic regimes believed to be important to
stream biota
Metrics
Extended low flow index (ELFI); this equals BFI—ZDF, where BFI is the baseflow index (ratio
of the minimum daily flow in any year to the mean annual flow) and ZDF is the zero day
fraction
CV of daily flows (DAYCV)
Contingency (M)
Number of low flow events (LFE)
Number of zero flow events (ZFE)
Mean 7-day minimum flow (Qmin7)
Mean daily discharge (QMEAN)
Mean bankfull flow (Q167)
Mean 7-day maximum flow (Qmax7)
Flow reversals (R)
Flood duration (FLDDUR)
Number of high flow events (FIFE)
Day of year of 50% of flow (T50)
Day of year of peak flow (Tp)
Predictability (P)
Constancy (C)
K-ll
-------
K.5. TOOLS FOR ESTIMATING STREAMFLOW AT UNGAGED SITES
Table K-5. Examples of tools for estimating streamflow and/or streamflow statistics at ungaged sites. A similar
tool is currently being developed for New York
Tool
Geographic area
Website
Description
USGS StreamStats1
Varies by state
http://water.usgs.gov/osw/streamstats/
Available for most but not all states in the
eastern United States. The types of output
statistics that are available vary by state. These
statistics represent long-term averages and do
not capture year-to-year variability.
BaSE (Stuckey et al.,
2012)
Pennsylvania
http://pubs.usgs.gov/sir/2012/5142/
This tool simulates minimally altered
streamflow at a daily time scale for ungaged
streams in Pennsylvania using data collected
during water years 1960-2008. It is free,
publicly available, and uses a point-and-click
interface.
MA SYE (Archfield
etal.,2010)
Massachusetts
http://pubs.usgs.gov/sir/2009/5227/
The MA SYE can estimate a daily time series
of unregulated, daily mean streamflow for a
44-year period of record spanning 1960 to
2004.
West Virginia DEP
7Q10 Report Tool
(Shank, 2011)
West Virginia
http: //tagi s. dep. wv. gov/streamfl ow/
This free, publicly available tool utilizes a
point-and-click interface. Seven Qio, annual
and monthly flow estimates are generated
when you click on a location.
USGS = U.S. Geological Survey; BaSE = Baseline Streamflow Estimator; MA SYE = Massachusetts Sustainable Yield Estimator; DEP = Department of
Environmental Protection
^ttpV/water.usgs.gov/osw/streamstats/
-------
K.6. REFERENCES
Archfield, SA; Vogel, RM; Sleeves, PA; Brandt, SL: Weiskel, PK; Garabedian, SP. (2010) The Massachusetts
sustainable-yield estimator: A decision-support tool to assess water availability at ungaged stream locations
in Massachusetts. [USGS Scientific Investigations Report 2009-5227]. Northborough, MA: Massachusetts-
Rhode Island Water Science Center, U.S. Geological Survey. 41 pp. Available online:
http://pubs.usgs.gov/sir/2009/5227/.
Buchanan, C; Moltz, HLN; Haywood, HC; Palmer, JB; Griggs, AN. (2013) A test of The Ecological Limits of
Hydrologic Alteration (ELOHA) method for determining environmental flows in the Potomac River basin,
U.S.A. Freshw Biol 58(12):2632-2647.
Colwell RK. (1974) Predictability, constancy, and contingency of periodic phenomena. Ecol. 55:1148-1153.
Cummins, J; Buchanan, C; Haywood, HC; Moltz, H; Griggs, A; Jones, C; Kraus, R; Hitt, NP; R. Bumgardner, R.
(2010). Potomac large river ecologically sustainable water management report. [ICPRB Report 10-3].
Interstate Commission on the Potomac River Basin for The Nature Conservancy. Available online at
www.potomacriver.org/pubs.
DePhilip, M; Moberg, T. (2010) Ecosystem flow recommendations for the Susquehanna River Basin. Harrisburg,
PA: The Nature Conservancy. Available online at
http://www.srbc.net/policies/docs/TNCFinalSusquehannaRiverEcosy stemFlowsStudyReport_NovlO_2012
0327_fsl35148vl.PDF
DePhilip, M; Moberg, T. (2013). Ecosystem flow recommendations for the Upper Ohio River basin in western
Pennsylvania. Harrisburg, PA: The Nature Conservancy.
Hawkins, CP; Tarboton, DG; Jin, J. (2013) Consequences of global climate change for stream biodiversity and
implications for the application and interpretation of biological indicators of aquatic ecosystem condition.
Final Report. [EPA Agreement Number: RD834186]. Logan Utah: Utah State University.
http://cfpub.epa. gov/ncer_abstracts/index.cfm/fuseaction/display.abstractDetail/abstract/9063
Olden, JD; Poff, ML. (2003) Redundancy and the choice of hydrologic indices for characterizing streamflow
regimes. River Res Appl 19:101-121.
Stuckey, MH; Koerkle, EH; Ulrich, J. (2012) Estimation of baseline daily mean streamflows for ungagged location
on Pennsylvania streams, wateryears 1960-2008. [USGS Scientific Investigations Report 2012-5142]. New
Cumberland, PA: U.S. Geological Survey. Available online at http://pubs.usgs.gov/sir/2012/5142/.
Shank, M. (2011) West Virginia DEP 7Q10 report tool [web page]. Accessed 1 August 2014. Available online at
http://tagis.dep.wv.gov/7qlO/.
K-13
-------
[This page intentionally left blank.]
-------
United States
Environmental Protection
Agency
National Center for Environmental Assessment
Office of Research and Development
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Printed on 100% recycled/recyclable paper
with a minimum 50% post-consumer
fiber using vegetable-based ink.
------- |