EPA/903/B-00/004
                                                                              October 2000

                                Compiled and Edited by:

                                     Wayne Davis
                         EPA Office of Environmental Information
                            Environmental Analysis Division

                                         and

                                      John Scott
                      Science Applications International Corporation
                                       Region 3
                          Office of Research and Development
                       Mid-Atlantic Integrated Assessment Program
                          U.S. Environmental Protection Agency
                              Ft. Meade, MD 20755-5350
Printed on chlorine free 100% recycled paper with
100% post-consumer fiber using vegetable-based ink.

-------
                                            Contents

1.0    Overview/Introduction
       1.1     Overview of the Projects that Constitute MAIA Field Sampling from
               1993-1998: EMAP, REMAP, and TIME	1.1
               1.1.1   Mid-Atlantic Highlands Assessment Project	1.1
               1.1.2   Temporal Integrated Monitoring of Ecosystems Project	1.2
               1.1.3   Mid-Atlantic Integrated Assessment Program	1.2
       1.2     Physical/Geographic Setting of the Mid-Atlantic Highlands	1.2
       1.3     Assessment Questions	1.5
       1.4     General Objectives of the MAHA State of Streams Report	1.6
2.0    Program Design	2.1
       2.1     Overall Basis  of EMAP Design	2.1
       2.2     EMAP/MAHA Sampling Design	2.1
               2.2.1   Basic EMAP Mid-Atlantic Grid Design	2.1
               2.2.2   First-Stage and Second-Stage Sample Identification	2.2
               2.2.3   Selecting Population of Interest-lst-3rd Order Streams	2.2
               2.2.4   Intensifying Sample Density	2.3
               2.2.5   Estimates of Uncertainty	2.3
       2.3     Sites Selected for Sampling	2.4
       2.4     Identification  of the Sampling Site and Layout of the Sampling Reach	2.4
       2.5     Indicator Selection	2.7
       2.6     Reference Conditions	2.7
       2.7     Temporal Sampling Frame	2.8
3.0    Fish Assemblage	3.1
       3.1     Sample Collection and Processing	3.1
       3.2     Historical Perspective	3.1
               3.2.1   Overview of Human Disturbance and  Potential Impacts to Fish Populations ....3.1
               3.2.2   Estimation of pre-Selllemenl Fish Assemblage Condition	3.2
               3.2.3   Conceptual Model of Fish Assemblage Response to Stressors	3.2
       3.3     Identification  of Candidate Metrics	3.3
       3.4     Analysis and Testing of Candidate Metrics	3.5
               3.4.1   Test of Range of Metric Values	3.5
               3.4.2   Signal to Noise Ratio Test	3.5
               3.4.3   Relationship to Watershed Size and Correction Procedure	3.6
               3.4.4   Test of Redundancy	3.9
               3.4.5   Metric Responses to Disturbance	3.9
       3.5     Metrics Selected and Metric Scoring	3.13
               3.5.1   Metrics Selected	3.13
               3.5.2   Metric Scoring	3.14
       3.6     IBI Validation and Threshold Development	3.16
               3.6.1   IBI Validation	3.16
               3.6.2   Development of IBI Thresholds and Estimation of Condition	3.19
       3.7     Non-Native Species Issue 	3.20

-------
                                      Contents (continued)
4.0    Benthic Macroinvertebrate Assemblage	4.1
       4.1     Sample Collection and Processing	4.1
       4.2     Metric Selection and Testing	4.4
       4.3     Index Testing	4.4
       4.4     Reference Condition	4.5
5.0    Physical Habitat	5.1
       5.1     Data Collection	5.1
       5.2     Metric Selection and Testing	5.4
       5.3     Index Calculation Incorporating Reference Condition	5.6
               5.3.1   Index of Riparian Habitat Quality	5.6
               5.3.2   Channel Sedimentation Index	5.6
               5.3.3   Fish Cover from Large Woody Debris	5.8
               5.3.4   Channel and Riparian Disturbance Index	5.8
               5.3.5   Watershed Quality Index	5.8
               5.3.6   Watershed, Riparian, and Channel Habitat Complexity Index	5.8
               5.3.7   Channel Habitat Quality	5.9
6.0    Rapid Habitat and Visual Stream Assessment (EPA RBP)	6.1
       6.1     Data Collection	6.1
       6.2     Metric Selection and Testing	6.1
       6.3     Index Calculation and Testing	6.3
7.0    Watershed Disturbance	7.1
       7.1     Watershed Risk Index	7.1
               7.1.1   Watershed Disturbance Metrics	7.1
               7.1.2   Index Computation	7.1
               7.1.3   Testing of the Watershed Risk Index	7.4
       7.2     Watershed Disturbance Index	7.6
               7.2.1   Watershed Disturbance Metrics	7.6
               7.2.2   Index Computation	7.8
               7.2.3   Testing of the Watershed Disturbance Index	7.8
8.0    Fish Tissue Contaminants	8.1
9.0    Water Chemistry	9.1
10.0   Strcssor Identification	10.1
11.0   Classification for Reporting Results	11.1
       11.1    General Classification Approach	11.1
       11.2    Watershed Classification	11.1
       11.3    Bcoregion Classification	11.1
12.0   Information Management	12.1
13.0   References	13.1

Appendix Table 1.      Complete set of assessment questions for MAHA streams.
Appendix Table 2.      Sample site characteristics and parameters measured.

-------
                                             Tables
Table 1-1.      Organization and content of the MAIIA State of the Streams Report.

Table 2-1.      Number of samples sites visited parameters measured in the RMAP, R-RMAP, and
                      TIMEprograi-nsiilheMid-Atlanlicl 993-1994.

Table 3-1.      Metrics Selected.
Table 3-2.      Criteria for definition of Reference and Test sites.
Table 3-3.      Estimates of stream condition (% stream length) based upon Fish IBI.

Table 5-1.      Precision of Physical Habitat Metrics in the Mid-Atlantic region.

Table 6-1.      Precision of Rapid Bioassessment Protocol (RBP) habitat quality metrics in the
               Mid-Atlantic region.

Table 7-1.      Types of information obtained from data sources for incorporation in a Watershed
               Disturbance Risk Index.
Table 7-2.      Stressor matrix showing criteria and progression of risk index scores for six sites in the
               Ridge and Valley ecoregion.
Table 7-3.      Thresholds for watershed disturbance metrics classifying streams as in good or poor
               condition.

Table 8-1.      Analytical methods for metals analysis in fish.
Table 8-2.      Analytical methods for pesticides analysis in fish.
Table 8-3.      Analytical methods for PCB congeners analysis in fish.

Table 9-1.      Field measurement methods for water chemistry.
Table 9-2.      Laboratory analytical methods for water chemistry.

Table 11-1.     Percent of stream miles in good (a), fair (b), and poor (c) condition or affected by potential
               stressors for the Mid-Atlantic  Highlands, four Highland ecoregions, three watersheds, and
               two states.
Table 11 -2.     Number of stream samples for each medium in the Mid-Atlantic Highlands, four Highland
               ecoregions, three watersheds,  and two states.

-------
                                            Figures
Figure 1-1.     Ecoregions of the MAHA region.

Figure 2-1.     Distribution of Stage 1, 1/16 sampling area 40 km2 hexagons.
Figure 2-2.     Distribution of intensified sample design using 13 km2 hexagons.
Figure 2-3.     Effort-return curve of fish species richness versus length of stream sampled.
Figure 2-4.     Sampling reach features.

Figure 3-1.     Conceptual model offish assemblage indicators and types of stressors.
Figure 3-2.     Approach to watershed size correction using expected reference value and regression
               residuals.
Figure 3-3.     Approach to watershed size correction by normalization to 100 km2 reference values.
Figure 3-4.     Responsiveness of the metric number of intolerant taxa (adjusted for watershed area) to
               chemical and habitat disturbance factors.
Figure 3-5.     Response of the metric number of intolerant taxa (adjusted for watershed size) to inte-
               grated measures of habitat disturbance and watershed condition class.
Figure 3-6.     Derivation of maximum and minimum metric scores at  Reference and Test sites for
               number of tolerant taxa metric (adjusted).
Figure 3-7.     Derivation of maximum and minimum metric scores at Reference and Test sites for
               number offish collected metric (adjusted).
Figure 3-8.     Fish IBI scores at reference and degraded sites and compared to Watershed Disturbance
               Index from the validation data set.
Figure 3-9.     Comparison of the Fish IBI to Watershed Condition  Class from the calibration data set.
Figure 3-10.    Relationship of number of fish collected to habitat volume and habitat volume to water-
               shed size.
Figure 3-11.    Calculation of good-fair-poor thresholds of condition based upon the Fish IBI.

Figure 4-1.     Modified kick net for benthic macroinvertebrate sampling.
Figure 4-2.     Index sampling design for benthic macroinvertebrates.
Figure 4-3.     Laboratory sample analysis scheme for benthic macroinvertebrales.
Figure 4-4.     KPT Taxa Index at hand-picked Reference and Test sites.
Figure 4-5.     Comparison of EPT taxa metric with modified SBII.
Figure 4-6.     Cumulative distribution of EPT Taxa Index scores for all probability sites and filtered
               Reference  sites.

Figure 7-1.     Relationship of watershed risk index to ionic strength and chloride.
Figure 7-2.     Comparison of the watershed risk index to biotic condition with normal score and those
               adjusted for natural variability.

-------

-------
1.0    Overview/Introduction

       1,1     Overview of the Projects that Constitute MAIA Field Sampling from 1993-1998:
               EMAP, REMAP, and TIME

The EPA Environmental Monitoring and Assessment Program (EMAP) was initiated in the late 1980s
in response to its Science Advisory Board's (SAB) report that encouraged the Agency to quantitatively
determine the effectiveness of its regulatory programs. The SAB recommended the implementation of a
program to monitor ecological status and trends that would identify emerging environmental problems
before they reach crisis proportions (Science Advisory Board 1988). EMAP became a multi-agency
activity to evaluate the ecological status of terrestrial and aquatic ecosystems. The following three
objectives have guided the EMAP research activities since that time (Lazorchak et al.  1998):

*    Estimate the current status, extent, changes and trends in indicators of the condition of the nation's
     ecological resources on a regional basis with known confidence.
*    Monitor indicators of pollutant exposure and habitat condition and seek associations between
     human-induced stresses and ecological condition.
«    Provide periodic statistical summaries and interpretive reports on ecological status and trends to
     resource managers and the public.

       1.1.1   Mid-Atlantic Highlands Assessment Project

The stream sampling component of EMAP-SW was initiated in 1993 in the mid-Appalachian region of
the eastern  United States;  it specifically focused on the all of the Highlands in Region 3 west of the
Blue Ridge Mountains. It was carried out in conjunction with a Regional-EMAP (R-EMAP) project that
emphasized the Ridge and Valley regions and the TIME program (see below) to address acid-sensitive
systems in the Appalachian spine. The designs of these three projects were blended into one assessment
program for 1993 and 1994 that is known as the Mid-Atlantic Highlands Assessment study (MAHA), that
was carried out over a 4-year period. The MAHA project was designed to test the EMAP approach in a
few of the most heavily impacted ecoregions of Region 3, the mid-Appalachians, the Ridge and Valley,
and the Central Appalachians (Lazorchak et al.  1998).

The Region 3 R-EMAP project was designed to answer the following questions:

       What are biological reference conditions for the Central Appalachian Ridge and Valley
            Eco region?
       Do biological communities differ between subregions?
       What is the status of Mid-Atlantic Highlands stream biota?
       Can linkages be established between impairment and possible causes of impairment?

During the  MAHA study,  550 wadeable stream sites predominately in the western two-thirds of EPA
Region 3 (DE, MD, VA, WV, PA) and the Catskill Mountains of New York were visited and sampled
using the field protocols being developed by EMAP. Streams were sampled each year during a 10-week
index period from April to July by field crews from EPA, the U.S. Fish and Wildlife Sendee, State,
and contract personnel.
             Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
       1.1.2   Temporal Integrated Monitoring of Ecosystems Project
A special interest component of RMAP-SW is the Temporal Integrated Monitoring of Ecosystems Project
(TIME). The purpose of the TIME project is to assess the changes and trends in chemical condition
in acid-sensitive surface waters (lakes and streams) of the northeastern and eastern U.S. resulting from
changes in acidic deposition caused by the 1990 Clean Air Act Amendments.

Components of this program were included in the 1993-1994 MAHA program. The TIME project has
three goals:

       Monitor current status and trends in chemical indicators of acidification in acid-sensitive
           regions of the U.S.
       Relate changes in deposition to changes in surface water conditions.
       Assess the effectiveness of the Clean Air Act emissions reductions in improving the
          acid/base status of surface waters.

       1.1.3   MM-Atlantic Integrated Assessment Program

From 1995 to the present, the EMAP Surface Waters Program became a collaborator with R-EMAP and
TIME, and the partnership was called the Mid-Atlantic Integrated Assessment  (MAIA) project, which
is attempting to produce an assessment of the condition of surface water andcstuarinc resources. The
MAIA project represented a follow-up to the MAHA study, with an expanded geographic scope (southern
New York to northern North Carolina, with more sites located in the Piedmont  and Coastal Plain
ecoregions) and a different index period  (July-September). In  1997, the first year of the MAIA study,
approximately 200 sites (150 wadeable sites, 21 repeated wadeable sites, and approximately 30 riverine
sites) were visited for sampling.

1.2     Physical/Geographic Setting of the Mid-Atlantic Highlands
The focus of the MAHA Streams report is on the
condition of first, second, and third-order streams
which constitute approximately 89% (72,000 miles)
of all streams in the Highlands. The Mid-Atlantic
Highlands contain parts of eight distinct Level III
ecorcgions (sec Figure 1-1). For the MAHA State of
the Streams report, similar Level III ecoregions were
combined into four ecoregions to generate sufficient
sample sizes to make estimates of stream  condition.
The four ecoregions are (1) Valley ecoregion, (2)
Ridge and Blue Ridge ecoregion, (3) North-Central
and Central Appalachian ecoregion, and (4) Western
Appalachian ecoregion. The following descriptions of
these four ecoregions are excerpted from Woods et al.
(1999).

Valley Ecoregion: The Valley ecoregion extends
from eastern Pennsylvania southwesterly through
southwestern Virginia. It is characterized by
                                       Figure 1-1. Ecoregions of the MAHA region.
1.2
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
agricultural valleys that are elongated, folded and faulted which alternate with the ridges of the Ridge
and Blue Ridge ecoregion. Local relief varies from approximately 50 to 500 feet. The ecoregion narrows
toward the south and is generally bordered by the higher Blue Ridge Mountains and the higher and
less deformed Allegheny and Cumberland plateaus. The ecoregion is underlain largely by Paleozoic
sedimentary rocks that have been folded and faulted. Sandstone, shale, limestone, and dolomite are the
predominant rock types. Lithological characteristics often determine surface morphology. Valleys tend to
be created on weaker strata, including limestone and shale. Inceptisols and Ultisols are common and were
developed on noncarbonate rock. Alfisols and Ultisols are found in the limestone valleys.

The valleys vaiy in microtopography and agricultural potential. Those derived from limestone and
dolomite are smoother in form and have a lower drainage density than those developed in shale. Shale
valleys often display a distinctive rolling topography. Soils derived from limestone are fertile and well
suited to agriculture, while those derived from shale have a much lower agricultural potential unless they
are calcareous. The nutrient rich limestone valleys contain productive agricultural land and tend to have
few streams, and stream flows have little association with the sizes of the watersheds. In contrast, the
shale valleys are generally less productive, more irregular, and have greater densities of streams. Most of
the streams in the limestone valleys are colder and flow all year, whereas those in the shale valleys tend to
lack flow in dry periods. Poultry operations are locally common and economically important.

Many of the stream networks are trellised; topography dictates that the swift, actively down-cutting
streams which run off steep ridges join the gentle valleys perpendicularly into gentler gradient, warmer,
more meandering streams. Partially as a result of the latitudinal extent of the ecoregion, aquatic habitat
diversity is good.

Climate varies significantly, and generally, both growing season and precipitation increase southward.
The frost-free period varies from less than 120 days to more than 180 days and the precipitation varies
from 36 to 50 inches. Locally, however, relief and topographic position have significant effects on the
microclimate. The Valley ecoregion is significantly lower than the Central Appalachians, which results
in less severe winters, considerably warmer summer temperatures, and lower annual precipitation due
to a rain shadow effect.

Ridge and Blue Ridge Ecoregion: The Ridge and Blue Ridge ecoregion is a narrow strip of
mountainous ridges that are mostly forested at elevations from approximately 1,000 to 5,700 feet. Local
relief varies up to 1,500 feet. This ecoregion contains high gradient, cool, clear streams occurring over
mostly sandstone and shale bottoms.

The Blue Ridge portion of the ecoregion to the east is a narrow strip of mountainous ridges that are
forested and well dissected. Local relief is high and both the side slopes and the channel gradients are
steep. Streams are cool and clear and have many riffle sections; they support a different, less diverse
fish assemblage than do the streams of the valleys below, which are warmer, lower in gradient, and
more turbid.

The Blue Ridge Mountains are underlain by resistant and deformed metavolcanic, igneous, sedimentary,
and metasedimentary rock. Inceptisols, Ultisols, and Alfisols have developed on the Cambrian, Paleozoic,
and Precambrian rock. They can be divided into northern and  southern parts at the Roanoke River. North
of the river, just three different rock types form the crest and the effects of differential erosion partially
determine their local altitude. South of the Roanoke River, the Blue Ridge Mountains become higher
and lithologically complex.
             Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
Climate varies significantly. Generally, both growing season and precipitation increase southward. The
frost-free period varies from less than 150 days to more than 175 days, and the precipitation varies
from 39 to 49 inches. Locally, however, relief and topographic position have significant effects on the
microclimate.

The natural vegetation varies from north to south. North of a transitional area near the Roanoke River, it
is predominantly Appalachian Oak Forest (dominated by white and red oaks). South of the transitional
area, a mix of Appalachian Oak Forest, Oak-Hickory-Pine Forest (dominants: hickory, longleaf pine,
shortleaf pine, loblolly pine, white oak and post oak) grows, and, in higher areas, Northern Hardwoods
(dominants: sugar maple, yellow birch, beech, and hemlock). On the foothills, a mix of loblolly and
shortleaf pines occur and are mixed with Appalachian Oak Forest.

The ecoregion does not contain any major urban areas and has a low population density.  However, due in
large part to the close proximity of metropolitan  areas in the Coastal Plain and Piedmont regions to the
east, recreational development  in the ecoregion has increased considerably in recent years.

North-Central and Central Appalachian Ecoregion: The North-Central and Central Appalachians in
northern and central Pennsylvania and central West Virginia are a vast elevated plateau of high hills, open
valleys, and low mountains with sandstone, siltstone, and shale geology and coal deposits. To the north
(North-Central Appalachians),  it is made up of plateau surfaces, high hills, and low mountains, and was
only partly glaciated. Both the  southwest and the glaciated cast are low in comparison to the  central
section, which rises to a general elevation of about 2,300 feet on erosion resistant sandstones. The climate
can be characterized as continental, with cool summers and cold winters. Average annual precipitation
is from 33 to 50 inches and there can be as few as 100 days without killing frost, the shortest period
in Pennsylvania. Soils are often frigid and are derived from sandstone, shale, and till; they are low
in nutrients, and support extensive forests. The original vegetation was primarily Northern Hardwoods
(dominants: sugar maple, yellow birch, beech, and hemlock), but scattered Appalachian  Oak Forest
(dominants: white and red oaks) and isolated highland pockets of spruce/fir forest also occurred. Land
use activities arc generally tied to forestry and recreation but some coal  and gas extraction occurs in
the west.

The southern portion of this ecoregion (Central Appalachians) includes parts of south central
Pennsylvania, eastern West Virginia, western Maryland, and southwestern Virginia. It is  a high, dissected,
and rugged plateau made up of sandstone, shale, conglomerate, and coal of Pennsylvanian and
Mississippian age. The plateau is locally punctuated by a limestone valley and a few anticlinal ridges. Its
soils have developed from residuum and are mostly frigid and mesic Ultisols and Inccptisols. Local relief
varies from less than 50 feet in mountain glades to over 1,950 feet in watergaps where
high-gradient streams are common. Crestal elevations generally increase towards the east and range from
about 1,200 feet to 4,600 feet. Elevations can be  high enough to insure a short growing season, a great
amount of rainfall, and extensive forest cover. In lower, less rugged areas,  more dairy and livestock farms
occur, but they are still interspersed with woodland. Bituminous coal mines are common and associated
stream siltation and acidification have occurred.

Much of the eastern part of the ecoregion is farmed and in pasture, with hay and grain for dairy cattle
being the principal crops. There also are large areas containing oak and northern hardwood forests.  Land
use activities are generally related to forestry and recreation, but some coal and gas extraction occurs in
the west. The southern part of the ecoregion in West Virginia is primarily a forested plateau composed of
sandstone and shale geology and coal deposits. Due to the rugged terrain,  cool climate, and infertile
1.4          Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
soils, this area is more forested and contains much less agriculture than docs the Valley and Western
Appalachian ecoregions. Coal mining is a major industiy in this region and acid mine drainage and stream
sill alion associaled with coal mining is common.

Western Appalachian Ecoregion: The Western Appalachian ecoregion extends from southwestern
Pennsylvania into western West Virginia. The hilly and wooded terrain of this ecoregion is less rugged and
not as forested as are the ecoregions to the east. Much of this region has been mined for bituminous coal.
Once covered by a maple-bccch-birch forest, this region is now largely in farms,  many of which arc dairy
operations. This ecoregion is characterized by low rounded hills and extensive areas of wetlands.

The Western Appalachian ecoregion is a mostly unglaciated, dissected plateau with 200 to 750 feet of
local relief and  crestal elevations of less than 2,000 feet. The region is composed of horizontally bedded
sedimentary rock. Soils have developed from residuum  and support a potential natural vegetation of
Appalachian Oak Forest (dominants: white and red oaks) and, especially in the south. Mixed Mesophytic
Forest. Land use and land cover is a mosaic of forests, urban-suburban-industrial activity, general farms,
dairy and livestock farms, pastures, coal mines, and oil-gas fields. Urban and industrial activity is common
in valleys along the major rivers. Bituminous coal mining is widespread and has diminished water quality
and reduced fish diversity; recent stream quality improvements have occurred in some rivers including the
Allegheny, Monongahela, Youghiogheny, and Ohio Rivers.

The western Appalachians are less forested, warmer, and lower than the North-Central Appalachians. Its
border with the Central Appalachians approximates a break in elevation and forest density. It is lower,
warmer, less steep, and less densely forested than the Central Appalachians and is underlain by less
resistant rock.

1.3    Assessment Questions

Chesapeake Bay and its watershed historically have been a primary focus of EPA Region III and the
states because of its environmental and sociocconomic importance to the Mid-Atlantic region. With the
emergence of regional issues of acidic deposition, climate change, habitat alteration, and loss of biotic
diversity, there has been an increased emphasis on other geographic areas within the Mid-Atlantic by
EPA and the slates. Other environmental issues affecting aquatic ecosystems are mine drainage, nutrient
loading, and fish tissue contamination have been identified through biennial state water quality assessment
reports required under Section 305(b) of the  Clean Water Act.

The Mid-Atlantic Highland State of the Streams report describes the biological condition of streams
throughout the Mid-Atlantic Highland  area and documents potential stressors to these stream ecosystems.
Geographic patterns in both biological conditions and potential stressors are presented and potential
management options are discussed. The later section of the Highland report presents an overview of
Highland streams within the Mid-Atlantic region, and within four aggregated ecoregions, by discussing
their condition with respect to three levels of potential stressors: acceptable levels, warning levels or levels
of concern, and unacceptable levels. Potential management options are then discussed for these three
categories of potential stressors.

Preliminary assessment questions were first formulated in  1992 prior to  the development of the sampling
design. The following three questions were identified:

*      What is the biological condition of streams in the Mid-Atlantic Highlands (any patterns
       to this condition)?
             Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
*       What is the relative magnitude of the stressors impacting aquatic systems (any patterns
        lo this relative ranking)?

*       What is the acidification status of sensitive streams in the Mid-Atlantic?

Once the study design was developed and indicators chosen, a group was formed in 1994 to outline more
detailed question that could be addressed with data in hand. A complete set of questions is found in
Appendix A-l. These questions have been refined over the succeeding years and used to guide the data
analysis and assessment process for Highland streams.

1.4     General Objectives of the MAHA State of Streams Report

The Highland  Streams report had five objectives:

        1.      Assess the ecological condition of streams in the ecoregions and watersheds
               of the Mid-Atlantic Highlands,
        2.      Use biological indicators with physical and chemical indicators to describe
               the condition and characteristics of Highland streams,
        3.      Produce an objective report on the ecological condition of streams in the
               Highlands that can contribute to state and regional 305(b) reports,
        4.      Identify potential stressors that affect stream condition, and
        5.      Influence state monitoring  design and reports in assessing stream condition.

The report was written for an audience of senior administrators, managers, decision makers and informed
lay public. The report was not written for a scientific audience so it does not discuss scientific
concepts, indicator or index development, techniques, or data analysis procedures. This Technical Support
Document presents the underlying scientific basis for the report and the conclusions reached in the
Highland Streams report.  It draws upon and complements material found in the peer-reviewed literature
and,  as such, is not intended to contain all the information available on the MAHA program. A companion
Technical Feasibility Study on biocriteria, which will be based in part on the MAHA effort, has the
objective to further explore the data and analysis methods, and their application to state water quality
programs. The content and organization of the Highlands Streams report is shown in Table  1-1.
 1.6          Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
Table 1-1. Organization and content of the MAHA State of the Streams Report.
                                    Table of Contents
    Introduction	1
            Purpose	1
            Background	1
            Stream Condition	3
            Regional Statistical Surveys	5
            The Highlands Stream Population	6

    Ecological Condition of Streams	8
            Fish Assemblages	9
            Aquatic Insect Assemblages	11
            Comparison of Fish and Aquatic Insect Scores	12

    Potential Stressors	14
            Acidification of Streams	15
            Nutrient Runoff.	17
            Physical Habitat Alteration	IS
            Fish Tissue Contamination	20
            Watershed Disturbance	22
            Non-Native Fish: Stressor or Success Story	23
            Summary of Ranking of Potential Stressors	24

    Ecorcgions, Watersheds, and States: Another Perspective on Stream Condition	26
            Western Appalachian Plateau	2S
            North-Central Appalachian Plateau	30
            Watersheds	33
            States	37

    Developing a Scorecard: Summarizing Stream Condition	40

    Management Implications	43

    Where Do We Go From Here	43
    Appendix A: Additional Readings	50
    Appendix B: Stream Population Estimates	53
    Appendix C: Glossary	%
             Mid-Atlantic Highlands Streams Assessment:  Technical Support Document         1.7

-------

-------
2.0    Program Design

2.1    Overall Basis of EMAP Design

The EMAP Statistical Surveys are designed to collect probability samples that result in the following:
1.      Eveiy member of the population has a known probability of being included in the sample;
2.      The sample is drawn by some method of random selection consistent with these probabilities, and
3.      These probabilities of selection arc taken into account in making population estimates from the
       samples (Snedecor and Cochran 1967).

Using a probabilistic design, samples are collected in direct proportion to their occurrence in the
population or resource. The probability of selection does not have to be equal for all members of the
population; it is simply sufficient that the probabilities be known. The EMAP stream survey design takes
advantage of the attribute of unequal selection of samples as described in later sections. A key feature of
probability samples is that the standard error of the estimate, and confidence limits for the true population
value, can be computed.  If probability samples are collected, it is possible, therefore to determine the
accuracy of the  estimates and provide estimates of uncertainty (or certainty).

The spatial dispersion of the sample is controlled by using a spatially explicit grid, typically a triangular
grid, but rectangular or square grids have also been used. The spatial control of the samples ensures
there is adequate spatial  coverage across the resource and reduces clumping or aggregations of samples
in  space. Variable spatial density and nested subsampling permit different sampling intensities to occur
within a population, such as sampling first order streams with lower density that higher order streams to
ensure a more equitable  distribution of samples across stream sizes. In addition, certain areas of interest
such as the Ridge and Valley ecoregions can be sampled with greater density, but within the same grid
structure used to sample streams across the Mid-Atlantic region (Stevens and Olsen 1999).

EMAP resource sampling typically has occurred within a discrete temporal frame referred to as an index
period, but there arc no statistical constraints to sampling at any interval. Logistical issues such as time
and personnel usually constrain  sampling to once or twice per year during index periods. Index periods
correspond to a period when sampling  can be used to characterize the population or resource to answer a
specific set of questions. Different index periods might be selected based on the specific questions being
asked. Index sampling is not intended to describe the processes or dynamics of a system over time, but
rather to characterize the important attributes of the population or resource and describe the distribution of
attributes over the population. Each site is important only as it represents a portion of the population, not
because it describes the dynamics at the site.

2.2    EMAP/MAHA  Sampling Design

       2.2.1   Basic EMAP Mid-Atlantic Grid Design

The elements of the probabilistic design for streams is described in Ilerlihy et al. (2000). The EMAP grid
design was used as the basis of the selection of sample sites. This design is represented by a randomly-
placed triangular grid of points draped over the continental U.S. and fit within a global framework. The
grid points are spaced 27 km apart and, when contiguous hexagons  are scribed around each point, a
hexagonal sample area of 635 km2 results. Since this represented a very large sampling area,  a liner grid
scale was used that all owed for a search area of 40 km2 (1/16 of the area).
             MM-A tlantic Highlands Streams Assessment:  Technical Support Document          2.1

-------
The hexagonal grid selection for the Mid-Atlantic was based on the original consideration of a national
four-year stream survey which would have sampled about 800 sites. To ensure enough sites would be
accessed and sampled, the Mid-Atlantic area (EPA Region 3) was allocated 100 sites (instead of the 80
for each region), and this comprised the base EMAP sample. These 100 sites (actually 102) are the only
ones to have all EMAP parameters sampled.

                                            2.2.2   First-Stage and Second-Stage Sample
                                            Identification
                                            The EMAP hexagonal grid (40km2) was used along
                                            with the EPA Reach File 3 (RF3) representing
                                            the hydrography network. The area within this grid
                                            consisting of all of the RF3 stream traces is referred
                                            to as a "First-Stage Sample". This corresponded to a
                                            1/16 area sample evenly spread across the Mid-Atlantic
                                            region (Figure 2-1). Based upon the EMAP four-year
                                            rotating design, 1/4 of the hexagons were chosen
                                            for sampling in 1993, and another 1/4 in 1994;
                                            sample allocation as described below was accomplished
                                            separately for each year. The first-stage sample is
                                            represented by the identification of all 1st to 3rd order
                                            streams contained in the 40 km2 hexagons.
Figure 2-1. Distribution of Stage 1, 1/16 sam-
pling area 40 km2 hexagons.
The second-stage sample was accomplished using GIS and the digital RF3 data. Within each hexagon,
all of the digital stream lengths, as stream fragments in the reach file, were identified and mechanically
placed in random order along a single continuous line representing all of the stream traces within the
hexagon. Fragments within one continuous stream of the same order were kept together. To assure that
samples were spread out evenly across the region of interest, the hexagons were placed into spatial
clusters, such that each cluster contained approximately the same stream length. Then the hexagons within
each cluster were arranged in random order, and finally, the clusters were arranged in random order. In
this manner, a random selection of sites along the trace could be made, but it also ensured that sites would
not be "clumped" together within certain portions of the Mid-Atlantic region.

       2.2.3    Selecting Population of Interest — I8t-3rd Order Streams

Based on the subset of streams of interest, the design was modified for different purposes. For MAH A, the
design focused on the 1st, 2nd' and 3rd order wadeable streams. The goal was to sample an equal number of
sites from each order stream, however, sampling of the stream traces would yield 60-70% of the samples
in the 1st order streams based upon their abundance. Since it  was thought that the 1st order streams had
a higher probability of being dry or non-target, the 1st order streams were over-sampled and allocated
50% of the  sample  sites. The 2nd and 3rd order streams were each allocated 25% of the sample sites.
Adjustments to the  continuous line of stream traces were made to "stretch" streams in each order until the
appropriate ratios as a proportion of all stream miles by order in the hexagon were met.

Once the desired factors were applied, a single continuous stream trace was partitioned to randomly select
the individual sites  to be sampled. For 1993, the length of this "stretched" stream trace (5,090.24 km)
2.2
           MM-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
was divided by (he number of sites for (he original base EMAP sample which was 100. The first site was
located randomly in the first 50.9 km interval, with each successive site located further down on the trace
an distance equal to the interval size.

       2.2.4   Intensifying Sample Density

Another modification to the design was
intensifying the sample density for the acid
deposition stream monitoring (TIME) and
the regional-EMAP (R-EMAP) study in the
Highlands.  A set of six additional hexagons
were identified in relation to each of the
base 40 km2 hexagons (see Figure 2-2). This
resulted in six additional, but smaller hexagons
(13 km2) which were then used as the frame
to extract the stream traces for the first-stage
intensified sample sizes. These intensified sites
were to be sampled in only certain areas of
the region, and thus, the first-stage sample only
clipped stream traces  from 13 km2 hexagons in
areas of interest. Second-stage sampling was
accomplished in the same manner as described
above to allocate 150  samples to the intensified
design from a 4,638.8 km total intensified
stream length (i.e., sample sites identified on a
30.9 km interval).

       2.2.5   Estimates of Uncertainty
Figure 2-2. Distribution of intensified sample design
using 13 km2 hexagons.
The variance or error in statistical surveys is influenced, primarily, by two factors: the sample size
(i.e., the number of samples collected) and the proportion of the samples in selected categories such
as acceptable/unacceptable condition. In general, the confidence interval is halved for each four fold
increased in sample size. For example, the confidence interval associated with a sample size of 100 when
50% of the population is affected is approximately ± 10%. When the sample size increases to 400 (i.e., 4
fold increase), the confidence interval decreases to approximately ± 5%. "1"he proportion of the population
in one of two binary categories also affects the confidence interval with smaller confidence intervals
associated with  the tails of the distribution and larger confidence intervals associated with the central
portion of the distribution.

Confidence limits of estimates of the proportion of stream length exhibiting specified conditions (e.g.,
proportion of stream length with no fish, proportion stream length degraded) were calculated using the
Horvitz-Thompson estimation procedure. For the MAHA data set, region-wide estimates of condition
with a sample size of approximately 500 would exhibit 90% confidence limits in the 6-10% range.
Large (n-100) and small (n-30) subpopulations had confidence limits in the 7-12% and 12-20% ranges,
respectively (Herlihy, personal communication). Population estimates in the central  portion of the
distribution would have higher confidence limits within these ranges.
             MM-A tlantic Highlands Streams Assessment: Technical Support Document
                                              2.3

-------
2.3     Sites Selected for Sampling

In 1993 and 1994, the MAHA region was sampled at 448 sites. The number of samples collected on the
EMAP design grid is designated as Target samples and are shown with parameters measured in Table
2-1. A number of hand-picked sites thought to be in good and bad condition are also shown as Reference
and Test sites, respectively. Sample locations, site descriptors and the parameters measured are detailed in
Appendix Table A-2 and the complete data set can be found at the MAI A web site streams homepage at:

                 http://www.epa.gov/emap/html/datal/surfwatr/data/mastreams/
         Table 2-1. Number of samples sites visited parameters measured in the EMAP,
         R-EMAP, and TIME programs in the Mid-Atlantic 1993-1994.
Parameter
Macroinvertebrate Assemblage
Fish Assemblage
Fish Tissue
Physical Habitat
Rapid Bioasscssmcnt
Stream Chemistry
Dissolved Oxygen/Temperature
Watershed Characteristics
Target
378
222
78
101
378
378
101
380
Reference
58
58
0
58
58
58
58
58
Test
10
9
0
0
10
10
0
10
Total
446
289
78
159
446
446
159
448
2.4
MM-A tlantic Highlands Streams Assessment: Technical Support Document

-------
2.4    Identification of the Sampling Site and Layout of the Sampling Reach

In order to get a representative picture of the ecological community, most of the biological and habitat
structure measures require sampling a certain length of a stream. A critical aspect of obtaining a
representative sample of the fish assemblage under the proposed plot design was determining the length
of stream that must be sampled at each site. For the fish indicator, it was necessary to collect a sample
of the assemblage from a single pass through a prescribed length of stream (Karr ct al. 1986). Repeated
sampling of a stream reach was neither practical nor representative. Thus, to determine the optimal length
of stream thai should be sampled lo maximize the number of different species collected, a small pilol
study on a few selected streams was conducted. The results are presented in Figure 2-3. Based on this
study, a stream length equal to 40 times the mean channel width was selected as  the area to be sampled.
This length of stream was sufficient to obtain approximately 90 percent of the fish species inhabiting
the reach. Sampling additional lengths of streams did not substantially increase the number of species
obtained. This approach was adopted to define the sample reach for all parameters measured in the
program.
   100
           10   20   30    40    50    60   70   80   90
                     Length (Channel Width Units)

Figure 2-3. Effort-return curve of fish species richness
versus length of stream sampled (McCormick and
Peck 1999).
                                                          Stream sampling points were chosen
                                                          from the "blue line" stream network
                                                          represented on 1 : 1 00,000- scale IJSGS
                                                          maps,  following a systematic randomized
                                                          selection process developed for EMAP
                                                          stream sampling described above. Sample
                                                          sites were then marked with an "X"
                                                          on liner-resolution l:24,000-scale USGS
                                                          maps.  This spot is referred to as the
                                                          "index site" or "X-site". Figure 2-4
                                                          illustrates the principal features of the
                                                          established sampling reach, including the
                                                          location of 1 1  cross-section transects
                                                          used for physical habitat characterization,
                                                          and  specific sampling points on each
                                                          cross-section transect for later collection
                                                          of periphyton samples andbenthic
                                                          macroinvertebrate samples.
            MM-A tlantic Highlands Streams Assessment:  Technical Support Document
2.5

-------
                                        X-site
                                                      ©
                                                   ,-' li

                          ..^       V   X®---""'    ;   determined at random
                                     .           : • Subsequent points         in
     Distance       trar»3ects=4                    ;   or^er \_ Q R
     mean      width at X-site                      :          '

    Figure 2-4. Sampling reach features.
Some conditions required adjusting the reach about the X-site (i.e., the X-site was no longer located at
the midpoint of the reach) to avoid features that should not be sampled. These features included upstream
lower order streams or downstream higher order streams. When these were encountered, the loss of reach
length was made up by moving ("sliding") the other end of the reach an equivalent distance away from
the X-sitc. Similarly, lakes, reservoirs, or ponds were avoided. In any case, the X-sitc always remained
within the sampling reach. If sliding caused the X-site to fall outside the sampling reach, the site was
classified as non-target and not sampled.

The full complement of field data and samples were not collected from streams that are categorized as
"Dry Channel" or "Intermittent." Physical habitat information was collected in all streams. Intermittent
streams had some cross-sections with biological measurements and some with none. No biological
sampling was collected from totally dry channels. Samples and measurements for water chemistry were
collected at the X-site (even if the reach has been adjusted by "sliding" it). If the X-site was dry, the
sample and chemical measurements were taken from a location having water with a surface area greater
than 1 in2 and a depth greater than 10 cm. All data for the physical habitat  indicator were collected
from all streams, regardless of the amount of water present in the channel  or at the transects. Depth
measurements along the deepest part  of the channel (the "thalweg") were obtained along the entire
sampling reach for  all target streams, whether they were dry, intermittent, or completely flowing. Other
measurements associated with characterizing riparian condition, substrate  type, etc. were collected to help
infer conditions in the stream when water is flowing.
2.6
MM-A tlantic Highlands Streams Assessment: Technical Support Document

-------
2.5     Indicator Selection

Indicators were selected based upon a framework for indicator interpretation that identified
environmental values for streams, relationships to assessment questions, the primary environmental
stressors, and critical ecosystem components. The overall process for selecting EMAP indicators is
presented in Barber (1994).

The streams program has emphasized biological integrity as the primary environmental value which
should be used to describe stream condition. Stressors that potentially affect this condition are deposition
of nutrients and chemical contaminants from anthropogenic emissions, alteration of stream physical
habitat, contamination offish, and introduction of exotic species. In the MAHA streams report, biological
integrity is represented quantitatively by the macroinvertebrate and fish indices of biotic integrity. Acid-
neutralizing capacity (ANC) and concentrations of nitrogen and phosphorus were used as indicators
of mine drainage, acidic deposition, and eutrophication. Indices of riparian habitat quality and channel
sedimentation were developed to address the extent of habitat alteration. A watershed risk index was
applied to integrate  all identifiable stressors that might be affecting wadeable streams. Direct measures of
metal and organic contamination in fish and presence of non-native species also were made.

'EPA recently has published evaluation guidelines for ecological indicators (Jackson et al. 2000) that
specify the criteria an indicator or index must meet in order to perform effectively. Evaluation of stream
indicators presented here and in the streams report, according to these performance criteria, is ongoing.

2.6     Reference Conditions

Identification of reference conditions is a critical element in the evaluation of biotic integrity. Reference
conditions are expectations on the status of biological communities in the absence of any human
disturbance, i.e., the biota exist under ideal, and solely natural, conditions (Plafkin et al.1989, Gerritsen et
al. 1994). However, since there are few if any waters not influenced by human activities, other methods
for estimating reference conditions, including historical records, best professional judgement,  and/or
identification of minimally impaired sites, must be employed.

Biological characteristics may be derived from historical records made prior to any human disturbance;
this information usually is contained in museum/university collections, water resource agency
documents, or the published literature. It is unlikely, however, that biotic condition could be reconstructed
from a single complete record and multiple sources of information would be required. A drawback
to a historical reconstruction of biotic condition is that multiple information sources likely had
multiple objectives and sampling procedures that may not be contemporaneous with methods in current
evaluations.

Minimally impaired sites are commonly employed to define reference biotic  condition. Often these sites
are selected, hand-picked, based on expert  opinion, best professional judgement or local knowledge on
biotic condition. These sites also can be identified and evaluated as to their unimpaired status based
upon measurements of all stressor characteristics that may affect biotic integrity;  these are necessary to
confirm that stressors do not exceed levels known to cause biological or ecological effects. Because of the
pervasiveness of atmospheric deposition and habitat alteration in the MAHA region, sufficient numbers
of unimpaired, pristine sites may not exist.  In this instance, reference sites can be established as those that
are minimally impaired, i.e., they meet relaxed standards of stressor characteristics. It is important
             MM-A tlantic Highlands Streams Assessment: Technical Support Document         2.1

-------
to regard the interpretation of biotic condition in this circumstance as less than the ideal and more as a
relative measure of impacts. In extreme cases, where minimally impaired sites are lacking, the best sites
available are employed to define a best attainable reference condition. This condition generally has no
relation to true reference condition.

Another approach to defining reference biotic condition, particularly when undisturbed sites are not
available, is to model biotic responses relative to a disturbance gradient in the form of a dosc-rcsponsc
curve. Estimates of biotic responses then can be made under minimally disturbed, reference conditions.

Regardless of the approach used, reference condition is classified in such a way that natural factors
affecting biotic assemblages are taken into account. Reference conditions specific to ecoregions are the
most common form of this classification.

 2.7    Temporal Sampling Frame

Stream sample collections and observations reported in the MAHA State of the Streams report were made
in 1993 and 1994. EMAP employs an approach whereby samples collected within a multi-year program
are taken at the same time each year which is termed the index period. The EMAP stream indicator
workgroup concluded that the appropriate time for collection of biotic information was during low flow
conditions after leaf out and not following flood events (Hughes  1993).

The index period for sampling Mid-Atlantic streams from 1993 through 1996 was spring base flow.
Spring base flow should include contributions from both point and nonpoint sources for nutrients,
sediment, and organic loading. This index period also should capture both episodic and chronic sources
of acidity from acidic deposition and mine drainage. This period was selected to occur after the streams
had started to warm and there was increased biological activity in periphyton, benthos and fish, including
collecting spring spawning fish species. Finally, there would be sufficient flow in the streams to collect
water samples during a spring index period.
2.8          MM-A tlantic Highlands Streams Assessment: Technical Support Document

-------
3.0    Fish Assemblage

Development of the fish assemblage metrics and IBI described in this section are after McCormick et
al. (2001) and a summary of an IBI workshop held in Corvallis, Oregon 26-28 January 2000 (Stoddard
2000), unless otherwise noted.

3.1    Sample Collection and Processing

All methods for MAHA field sample collections are provided in Lazorchak et al. (1998). Relevant
excerpts from these methods are provided below.

Fish were collected according to time and distance criteria using pulsed DC backpack electrofishing
supplemented by seining. The reach length was equivalent to 40 times the average channel wetted width
at the midpoint of the site and consisted of an approximate minimum to maximum distances of 150 to
500 m. The sample interval was no shorter than 45 minutes and did not last more than 3  hours. Transects
were established every 10 channel widths or 15 m. Sampling was initially estimated at a maximum of 3
hours to determine the maximum amount of time that should be  spent fishing an area. Due to habitat and
structural complexities, actual shock time could be 50-75% of the sampling time. Seining was used to
supplement electrofishing if it was felt that the electrofishing may have under represented some species,
or if the stream was too deep or turbid for optimal electrofishing efficiency.

Fish were identified in the field to species and were also examined for external anomalies, measured
for length of some specimens, and voucher specimens were prepared for taxonomic confirmation and
archival. Voucher collections of up to 25  individuals of all species were made, with the smaller and harder
to identify species collected more often, with only a few larger species in the voucher samples.

3.2    Historical Perspective

       3.2.1  Overview of Human Disturbance and Potential Impacts to Fish Populations

McCormick et al. (2001) have summarized the long history of human impact on the landscape, streams,
and fish assemblages of the region (Denevan  1992). Streams in the region have been subjected to stresses
from acid deposition,  mining, logging, agriculture, and development (Railz et al. 1984; Jones et al. 1997).
Settlement of the Highlands did not begin in earnest until the 1700's as German, Irish, and English
immigrants spread from Pennsylvania into Virginia and West Virginia. In the mid-1800's, the advent of
rail transportation in the region (1830-1860) and discoveries of anthracite and bituminous coal and oil
and gas (1850's) opened the region to major industrial development by  the coal,  oil, and steel industries.
Devastating floods and fires occurred in the watersheds  of the Allegheny and Monongahela Rivers around
the turn of the century. Clear-cutting allowed the deep humus layer covering the  forest floor to dry
out, resulting in fires that, in some cases, exposed the underlying bedrock. Agriculture and clear-cutting
of highland and valley forests exacerbated soil erosion and sedimentation (U.S. DA 1996). In a recent
estimate, active and abandoned coal mining resulted in mine  drainage that affected 4,000 km of streams
(U.S.  EPA 1995). Extensive areas of the Ridge, Blue Ridge, and Appalachian plateaus have poorly
buffered soils and steep slopes, which have also made streams draining these areas susceptible to acid
precipitation (Herlihy et al. 1993).
            Mid-Atlantic Highlands Streams Assessment: Technical Support Document         3.1

-------
Stocking of brown trout (Salino trutta), rainbow trout (Oncorhynchus mvkiss), common carp (Cvprinus
carpio), and large warmwater species (Micropterus, Lepomis and Ameiurm spp.) was conducted by the
United States Fish Commission and state agencies (Courtenay et al. 1986; Jenkins and Burkhead 1994).
Hatcheries were established in the 1870's to culture trout and warmwater game fishes in response to the
loss of native species and public demand for augmented sport fisheries. Other introductions, particularly
those of forage fish, occurred to support sport fisheries or as bait bucket transfers (Nico and Fuller
1999).  Nonindigcnous species constitute as much as 33% of the fish fauna of the Potomac drainage and
48% of the fish species in the upper Kanawha (New) River drainage (Hocutt et al. 1986; Jenkins and
Burkhead 1994).

       3.2.2    Estimation of pre-Settlement Fish Assemblage Condition

The entire MAHA landscape is assumed to have been forested with old growth interspersed with the
occasional openings caused by fire, beaver-clearing, blow downs, and hurricanes. The streams flowed
clearly, with minimal stream channelization and incision. Because of a greater channel complexity,
storage of sands and silts into well sorted homogeneous patches was likely greater than at present. Large
woody debris in and around streams were abundant and caused a heterogeneity of channel slope, cross
section, and stream flow. These all contributed to greater habitat complexity and patchiness. Mountain
streams were  stepped by fallen trees; valley streams meandered and had extensive wetlands, braiding, and
logjams. Beaver were abundant and provided openings  and nutrients in low-gradient streams and flats of
higher  gradient streams; therefore, smaller streams were normally heavily shaded.

The fish inhabiting these stream habitats required clear  cool/cold waters in the mountains and cool/warm
waters  in the valleys. Non-native species were absent. Long-lived fishes attained large sizes because
human predation was minimal and large persistent pools existed. Brook trout, sculpin, and dace inhabited
cold headwater streams, which contained up to eight species in larger first order systems. Additional dace,
sucker, and darter species would be found in larger cold/cool streams contributing to a species count of up
to 15. Warm headwater streams supported dace in smaller systems and chub, sculpin, and shiners in larger
systems. Suckers, shiners, sunfish, darters, and bullhead would have been found in large warm streams,
which would have contained up to 20 species.

       3.2.3    Conceptual Model of Fish Assemblage Response to  Stressors

Component metrics offish  assemblage condition are expected to exhibit hypothesized responses to
stressors, which can be monitored at different scales. These metrics also incorporate information from
different levels of biological organization. Possible causes of poor condition as determined by the
assemblage response can be identified (although specific cause-effect relationships cannot be  ascertained)
by examining correlative relationships between specific indicators or component metrics and various
measures of ecosystem stress (measurement variables or multi-component indicators).

Basic relationships between major structural components and processes of a stream ecosystem and
general sources of anthropogenic stressors have been documented. Fish assemblages can be used to
demonstrate those stressor-response relationships and to assess condition both in the water  column and
bottom habitats and to  provide information on multiple trophic levels. Specific information on stressors
and their relationship to the indicators is presented in Figure 3-1. This graphical approach conceptualizes
the hypothesized relationships between stressors and component metrics. This approach is based on a
more generalized model originally conceived by Karr et al. (1986). The model has been modified to
3.2         Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
organize it by types of major stressors (following terminology presented in U.S. EPA 1997). The figure
provides a means to show direct linkages between individual metrics arid each type of stressor and
illustrates the diagnostic capability of the fish assemblage indicators. Low scores for certain component
metrics are associated with responses to certain groups of stressors.
              CHEMICAL
                                                             PHYSICAL HABITAT ALTERATIONS
                    Disturbance
                                    Metric Response
                                                       Stressor
                                                                      Disturbance
                                                                                     Metric Response
                                                                                    i| # Water Column spp. "}
                                                                                      Trophic Strategies |
                                                                                    I* % Carnivores
                                                                                    t| % tnvertivores
                                                                                    I1 % Tolarartt Spawners |
   Stressor
             HYDFIOLOQIC ALTERATIONS
                   Disturbance          Metric Response
                                                       BIOLOGICAL ALTERATIONS
                                                 Stressor        Disturbance        Metric Response
 Figure 3-1. Conceptual model of fish assemblage indicators and types of stressors (McCormick and
 Peck 2000),
3.3
Identification of Candidate Metrics
The fish metric analysis arid IBI development contained in the MAHA State of the Streams Report
and as described herein are after McCormick et al. (2001). The metric selection and testing, and IBI
calculation were developed using a calibration data sel which consisted of 177 1993-1996 sites where
fish and quantitative physical habitat data were collected. The IBI was tested on 119 remaining sites,
which were set aside and not used in IBI development.
             Mid-Atlantic Highlands Streams Assessment: Technical Support Document
                                                                                         3.3

-------
Fish were classified into taxonomic, habitat, tolerance, trophic, and reproductive categories for
compulation of metrics. The classifications of species in an assemblage was limited after Karr (1981)
and Karr et al. (1986) in order that neither sensitive nor tolerant species comprised more than 10% of
the ichthyofauna. As is common practice (Simon and Lyons 1995), non-native species were retained in
the calculation of proportional habitat and trophic metrics but excluded from the richness metrics so as
to not artificially desirable attributes. Of the 139 species identified at the drainage basin level, 45 or 32%
were considered as non-native, including brown trout  and smallmouth bass. The resultant 57 candidate
metrics are shown below; 27 are richness metrics and 30 are proportional metrics. Note that each metric
is preceded by the data base identifier.
                               Fish Assemblage Variables
Number of:
NAT IV HAM
NREPROS
NSANGU
NSATHER
NSDENT2

NSCATO
NSCATO2
NSCHNT
NSCOLU
NSCOTT
NSCYPR2
NSDART
NSDRUMX
NSESOXX
NSFUND
NSGAMB
NSICTA
NSINTOL
NSLAMP
NSPERCO
NSPPER
NSSALM
NSUMBR
NTROPH
NUMHISH
NUMNATSP
NUMSPEC
  families represented
  reproductive guilds
  anguilla species
  atherin species
  native benthic invertivore species
  minus 3 tolerant taxa**
  sucker species
  native intolerant Catostomids
  sun fish species
  number of water column species
  sculpin species
  intolerant cyprinid species
  darter species
  drum species
  esox species
  fundulus species
  gambusia species
  ictalurid species
  intolerant species
  lamprey species
  percopsis species
  perch species
  trout species
  umbridae species
  trophic guilds
  individuals in sample
  native species
  total fish species
Proportion of:
PANOM
PATNCi
BCLN

PBCST
PBENT
PBENTSP
PC ARM
PCGBU
PCOLD1
PCOLD2
PCOLSP
PCOTTID
PC Y PIT,
PEXOT
PGRAVEL
PHERB
PINSK
PINVERT
PMACRO
PMICRO
PMICRO2
PNEST
PNTGU
POMNI
POMNI	H
PPISC
PPISCIN2
PPISCINV
PTOLE
PTREPRO
individuals with anomalies
individuals as attacher non-guarder
individuals as broadcast spawners
clear substrate
individuals as broadcast spawners
fish as benthic insectivores
benthic habitat species in native species
piscivore and invertivore
individuals as clear gravel buryers
cold water individuals
cold and cool water individuals
column species in native species
individuals as cottids
individuals as tolerant cyprinids
individuals as introduced
simple lithophils
individuals as herbivores
individuals as native insectivores
invertivores
macro-omnivorcs
micro-omnivores
micro-omnivores minus RIHNATRO
individuals as nest associates
individuals as nester guarder
omninore individuals (pmicro+pmacro)
omni-herbivores (pmicro+pmacro+herbiv)
individuals as carnivores
piscivore-inscctivorc minus SEMOATRO
piscivore-insectivores
individuals as tolerant
tolerant reproductive guild individuals
        White Sucker (CATOCOMM), Blachiose Dace (RHINATRA), Black Bullhead (AMEIMELA),
        Yellow Bullhead (AMEINATA), Brown Bullhead (AMEINEBU) were excluded.
3.4
Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
3.4    Analysis and Testing of Candidate Metrics

The 57 metrics were evaluated in a step-wise process that was designed to: assess the effective range of
response, evaluate the repeatability of measurements (signal to noise), determine relationship to watershed
area and adjust if necessary, identify metrics that provided redundant information, and finally, assess
the discriminatory ability of the metrics to disturbance. The following sections describe results  of this
evaluation.

       3.4.1   Test of Mange of Metric Values

All richness metrics were subjected to a range test to determine if they had sufficient breadth of values
to contribute sufficient information to an fish IBI, i.e., meaningful differences could be detected between
reference and impaired sites. The following 13 metrics were eliminated from the list because they only
had observed values of 0, 1, or 2:

              NSANGU      number anguilla species
              NSATHER     number alherin species
              NSCATO2     number native intolerant catastomids
              NSDRUMX    number drum species
              NSESOXX     number esox species
              NSFUND      number fundulus species
              NSGAMB     number gambusia species
              NSICTA       number ictalurid species
              NSLAMP      number lamprey species
              NSPERCO     number percopsis species
              NSPPER       number perch species
              NSSALM      number trout species
              NSUMBR     number umbridae species

       3.4.2   Signal to Noise .Ratio Test

Repeated measurements of each metric at the same site were evaluated for the remaining 44 metrics to
determine signal to  noise ratio. An effective metric should exhibit higher between site variance  than within
site variance. Two metrics, NTROPII-number trophic guilds and PNEST-proportion of individuals as nest
associates, were removed from further evaluation because their signal to noise ratios (between to within
site variance) were less than 3.
            Mid-Atlantic Highlands Streams Assessment:  Technical Support Document         3.5

-------
       3.4.3   Relationship to Watershed Size and Correction Procedure

Species richness metrics are known to be related to the size of the watershed drainage area (Fausch
et al. 1984). It was determined that if a relationship to watershed size exists, the metric should be
correcled before ils discriminatory ability was evaluated at a later step in the metric testing framework.
The following 17 metrics exhibited a strong relationship to watershed area:

              NATIVFAM           number families
              NREPROS2           number reproductive guilds
              NSBENT2            number native benthic invcrtivorc species minus 3 tolerant taxa
              NSCATO             number sucker species
              NSCENT             number sunfish species
              NSCOLU             number water column species
              NSCYPRA2           number intolerant cyprinid species
              NSDART             number darter species
              NSINTOL            number intolerant species
              NUMFISH            number of individuals
              NUMNATSPEC       number native species
              N U MS PEC           number total spec ies
              PATNG               proportion individuals as allacher non-guarders
              PBENT               proportion benthic habitat species in native species
              PCARN              proportion piscivore and invertivore
              PINSE               proportion individuals as native insectivores
              PINVERT            proportion invertivores

These seventeen metrics were normalized by regression to a watershed area of 100 km2 according to the
following process. First, the regression for each metric value against watershed size (loglo) in predefined
reference sites was  calculated. This regression was then used at all sites to calculate a residual value for
each site. Figure 3-2 demonstrates these steps using the number of benthic species metric as  an example.
Next, the expected metric value at 100 km2 was estimated. This value was then applied to the residuals
for all sites such that each  site/metric value was normalized to the expected value at 100 km2. Figure 3-3
illustrates this example. Use of this approach is thought to maximize the correction for watershed size
without eliminating disturbance factors to which the metrics are responding.
3.6         Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
         J)

         55
                                                       o
                                                       o
                                                       o
            00
                      CD
                                         C\J
                                                   O
                       £                               "

t_ *
2 ^ "w
03 0 «>
11 §
03 ® E
0 &J>
T- 2 £
0) ,
0 _ J
C "js
®
cxi 2w
k
1
!w.
CD 5
4«^
JO j)
S w
C»? *"%
D 1
. -
                                                (0
         CD
         0)
         •few**

         CO

         0)
         o
         c
         
N
3
"O
a>

f

75
60
13
®

E
5
                                                                      m
                                                                      CD

                                                                      c
                                                                      g
                                                                     'co
                                                                      en
                                                                      CD

                                                                      O)
                                                                      CD

                                                                     TJ
                                                                      C
                                                                      CO
                                                                     T3
                                                                      a>
                                                                      0
                                                                      CL
o
CD

O
o
CD
_N
"cn
T5
CD
jC
CO

_CD
"5


o
         Q.
         CL
                                                                     O
                                                                     CO

                                                                      £
                                                                      3
                                                                     .5?
                                                                     LL
Mid-Atlantic Highlands Streams Assessment: Technical Support Document
                      3.7

-------
                                                            o
                                                            o
















:.J "" /-: J-/ /"; ,'-> ..-'.'.-",-•-' / •• /' •' ..-
'-'{'//£ •-•;,-9//" -"ry ::'',:;-'/

*. „. ... .--^' !>'-: •'% •*%- A — . ™'"' ,.i-.r" '..,* '^ ...,
..»'= !-* ;^'. _^" , V /*.'" """ '— ' ,^% .:""r" '-
----- '^;r.l/ ^"' .- ~"V-'
-,' ' ..-.; *^' _ -.^/' '. ..- . -' -"' .». '-»•?"
.'" ( ,-*\ '"". — *" ; .'*"" '" ""
. 7 I-/ '-„ ' f.' >: £-"
l'"> .-*" ,~. .' .-" ~~
Q '~S':.:~r~ O '.;;"_;"

'*•«•' """ ,< J/ '"" / —
;"='. ;";•"-> ,*" "• ..-. T


1 	 ' 	 • 	 " 	 ' 	 ' 	 T 	
a
5

T
a
. 0 •§
"•"* t.
8

"5
3





T*™
OO>Q3fN»COtf)"'4"COC\iT-O'»— C\JCO
1— ' '
UJOJ|
1_
o
® C8~ "
® 1 •§ ^
ffl _ CD g
™3 m O ^"^
« 1 | "S
. X *m ft)
T- ® c en ,,



"O
«£ C'

*Q^ CO -S Q 0) S3 p
S m *o rt ^ "^ 5 -
<£ g .if jg ^ C Q Q
^- 0 £ Lj 0 5 0 .N
„ CMOif: oocc/5-r-®
T
























\ °
\.
™\
"•™'f\
. ) ' " •' \ «""*


•1^ r>
'«*.t
'„
„
j-:'
••„-•'












• •
-..-
\ -~- .•••-. r"-.
,™* \— • "-*
'— \_ -.., ,~:

i '": ~"~^\
-ii- . ,-i
-^
v f ,. f

\ - --. ;'

-...-•' X
- \ -~:
\
i".-.
... \ :
\




/•""I
^^
**^ ^a
^~~ CM
1

^«
a
IS
5
T
- o a
T— J^
W
V
:'*»
a
"5
S





CO
CD

To
CD
0
c.
CD
42
"5
C\l
E
o
o
o
c
o
CO
_N
"co
O
JZ
JZ
_g
1
o
o
CD
N
"co

CD
JZ
CO
0
ns
^
o
JC
o
cd
o

Q.
Q.
*t
CO
CO
£
3
[JL
00 C0 ^ <5f CM O
U5

£ snuiuu) 'ON
3.8
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
       3.4.4   Test of Redundancy

All remaining 42 metrics, were subjected to a correlation analysis to determine their degree of
independence from one another. Two pairs of metrics had Pearson Correlation coefficients greater than
0.75. Proportion cold water individuals (PCOLD1) was redundant with proportion of cold and cool
water individuals (PCOLD2) and the latter was removed from further testing. Similarly, proportion of
individuals as broadcast spawners on clean substrate (PBCLN) was retained in favor of its redundant
partner, the proportion macro-omnivores (PMACRO).

       3.4.5   Metric Responses to Disturbance

The remaining 40 metrics were evaluated as to their responsiveness in a positive or negative manner to
habitat disturbance factors. Habitat disturbance was characterized by the following 18 measures which
include, physical, chemical,  and catchment parameters:

Chemical       pIT
               Sulfate concentration
               Total nitrogen concentration
               Total phosphorus concentration
               Chloride concentration
               Disturbance class
Physical        Percent sands  and fines (PCT_SAFN)
               Bed stability (LRBS	BW4)
               Density of large woody debris (XFC_LWD)
               Fish cover
               Riparian disturbance (Wl  11 ALL)
               Channel and riparian disturbance index
               Channel habitat quality index
               Reach slope (XSLOPH)
Catchment     Watershed quality index
               Watershed and riparian condition index
               Watershed, riparian, and channel habitat quality index
               Bryce  watershed condition class

Derivation of physical  habitat measures is after Kaufmann et al. (1999) and are summarized in Section
5. Condition class is from Bryce et al. (1999) and is described in Section 7. Chemical classification
of disturbance  class is provided by Herlihy (A. Herlihy, personal communication) and was derived as
follows. Sample sites were divided into four classes by water chemistry using a scheme similar to that
used by Herlihy et al. (1990, 1991) in previous Mid-Atlantic assessments:

       1.  Acidic Deposition — ANC < 25 ueq/L AND sulfate < 400 ueq/L
       2.  AMD Impacts — (ANC < 25 ueq/L AND sulfate > 400 ueq/L) OR sulfate > 1000 ueq/L
       3.  Mixed Impacts —ANC > 25 ueq/L AND (400 100 ueq/L
       4.  Least Disturbed — ANC < 25 ueq/L AND sulfate < 400 ueq/L AND chloride < 100 ueq/L
            Mid-Atlantic Highlands Streams Assessment: Technical Support Document
3.9

-------
All sites with an ANC below 25 ucq/L were assumed to be acid impacted and assigned to cither the
acidic deposition or AMD Impacts class using sulfate concentration. Streams with ANC below 25 ueq/'L
are either chronically acidic (no acid neutralizing capacity; ANC < 0) or usually transiently acidic (ANC
0-25). The dominant acid anion in both acidic deposition and acid mine drainage is sulfate. In the
Mid-Atlantic, streamwater sulfate concentrations based on evapoconcentration of sulfate in deposition
are expected to be around 150-250 ueq/'L. Streams with sulfate below 400 ueq/'L have sulfate anion
composition dominated by deposition sources. Similarly, streams with sulfate above 400 ueq/L are
dominated by internal watershed  sources (mining) of sulfate.

Using data from the National Stream Survey (NSS), Herlihy et al. (1990) found that very few acidic .NSS
stream samples had sulfate concentrations between 250 and 500 ueq/L. Thus, the selection of an arbitrary
cutoff value in this range has only a small impact on interpreting the chemical classification scheme. In
most acidic streams, the dominant source of sulfate was clearly either atmospheric (stream sulfate less
than 250 ueq/L) or from watershed sources (stream sulfate greater than 500 ueq/L).

Dissolved iron or manganese concentrations were not used as a screening factor in the AMD classification
because the less acidic mine drainage impacted streams had very  low iron concentrations. Sulfate is a
better indicator of AMD than Fe because sulfate is a much more conservative ion. Very few processes act
to remove sulfate from solution in stream water. On the other hand, iron and manganese rapidly precipitate
out of solution (e.g., iron hydroxides or "yellow boy") as streamwater pH increases downstream from
the AMD source. Sulfate concentrations were also used to identify mine drainage impacts in non-acidic
streams. Non-acidic streams with sulfate concentrations above  1000 ucq/L in the Appalachian Plateau
were classified as non-acidic, mine drainage impacted. All the EMAP sites in the Appalachian Plateau
with sulfate greater than 1000 ueq/L had evidence of mining activity in their watersheds on 7.5" USGS
maps and/or in the crew field notes. In general, acidic streams are more severely impacted by mine
drainage than non-acidic streams  because the water itself is toxic  to many organisms due to  low pH
and high metal concentrations. While the water in the non-acidic, mine drainage impacted streams is not
necessarily toxic, these sites are often impaired by sedimentation, armoring, sediment metals, and physical
habitat alteration due to mine drainage. The high sulfate concentrations in these sites serves as an excellent
indicator of mine drainage impacts in the watershed.

In Ihe Mid-Atlantic, stream chloride concentrations are a good  indicator of human disturbance in a
watershed (Herlihy et al. 1998). Streams with both low chloride and sulfate concentrations and that were
not acidic were considered "Least Disturbed" for purposes of this assessment. Chemistry at these sites and
visual examination of site maps and field notes indicate that these sites are those with the least human
impacts in the region and they could be considered good condition or reference sites. Streams that had
chemical signatures too high to make the least disturbed class but not high enough to be considered
AMD or acidic deposition  impacted were classified as a "Mixed Impacts" class. The streams in the mixed
impacts class could be influenced by a number of factors such as roads, point sources,  agriculture as
well as weak mine drainage.

Nutrient disturbance was deemed high if total phosphorus was  > 30 ug/'L or total nitrogen was
>1000ug/L.

Scatterplots and box and whisker plots of each metric against each disturbance factor were visually
examined as to response. An example of responsiveness for the metric number of intolerant taxa is
illustrated in Figures 3-4 and 3-5.
3.10         Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
Figure 3-4. Responsiveness of the metric number of intolerant taxa (adjusted for
watershed area) to chemical and habitat disturbance factors. Plots outlined in bold
illustrate good metric response.
       Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
                   3.0


                   25

                to
                «  2.0

               "c
                to
               J5  1-5
                Q
               "c.
                6  1.0
               2

                   0.5


                   0.0
                   3,0
                   2.5 -J
                   0.0 --
                                                         "3T
                                                          *
                                       234
                           Acid      Mine      Mixed    Nutrients  Reference
                           Rain    Drainage   impacts
                       Figure 3-5. Response of the metric number of intolerant
                       taxa (adjusted for watershed size) to integrated measures
                       of habitat disturbance and watershed condition class.
3.12
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
3.5    Metrics Selected and Metric Scoring

       3.5.1   Metrics Selected

The 10 most responsive metrics were selected for calibration and scoring using the calibration data set
(n=l 19) as follows. Scattcrplots of each metric against each of the 15 individual disturbance metrics
(chemistry and habitat), and box and whisker plots of the two integrated measures of disturbance
condition  class, disturbance class) were examined. Any metrics that showed relationships with two or
fewer of these disturbance gradients were discarded. Of the metrics that passed this test, the final metric
suite retained for the 1BI was composed such that one or more metrics were responsive to each type of
disturbance. The selected metrics, listed in Table 3-1, include four proportional metrics and six richness
metrics. All richness metrics are adjusted for watershed size.
 Table 3-1. Metrics Selected.
Metric Class
Tolerance

Abundance
Reproductive
Habitat


Alien
Trophic

Metric Name
NSINTOL4
PTOLE
MUMPISH
PG RAVEL
PCOTTID
NSBENT23
NSCYPR3
PEXOT
PMACRO
PPISCIN2
Description
Number Intolerant Taxa
Proportion Tolerant Taxa
Number of Fish
Proportion Simple Lithophils
Proportion Cottids
Number Benthic Species
Number Cyprinid Species
Proportion Introduced
Individuals
Proportion Macro-omnivores
Proportion Piscivore/
Invertevores
Response Class
Chemistry, Channel Habitat,
Watershed Condition
Chemistry, Channel Habitat,
Watershed Condition
Nutrients
Channel Habitat
Nutrients, All Habitat
measures
Disturbance Classes
Condition Classes
Introduced Species
Nutrients
All Habitat measures
            Mid-Atlantic Highlands Streams Assessment: Technical Support Document
3.13

-------
        3.5.2   Metric Scoring

The 10 metrics were scored on a scale of 0-10, with 10 representing the median value of the metric at
reference sites, and 0 representing the 10th percent! le of the metric values from test (disturbed) sites, all
of which were taken from the calibration data set (N=177). Definition of reference and test (disturbed)
sites are shown below in Table 3-2. To be classified as a Reference site, all listed criteria must be met
and to be defined as a Test (disturbed site), at least one of the criteria must be met.
              Table 3-2. Criteria for definition of Reference and Test sites.
Stressor Criterion
ANC (ueq/L)
PH
Total Phosphorus (ug/L)
Total Nitrogen (ug/L)
Chloride (ueq/L)
Sulfate (ueq/L)
Mean RBP Score
Habitat Quality Metrics (QTPH1 ,
QCPH1,QW1,QWR1)
Watershed Condition Class
Reference
>50

<20
<750
<100
<400
>15
>0.5

Test

<5
>100
>5000
>1000
>1000
<10
<0.3
5
Figure 3-6 demonstrates the derivation of maximum and minimum scores from Reference and Test site
equivalent to 10 and 0, respectively, using the number of tolerant taxa metric as an example. In this case, a
metric score of 1.5 is equivalent to 10 and values above 1.5 are set to 10. A score of approximately 0.25 is
equivalent to 0 and scores lower than 0.25 arc set to 0.

In the process of calculating metric scores, one metric, number offish collected, could not be  calibrated,
i.e., the median reference value was not different from the test site median. Thus, if scored in a way
similar to the other nine metrics, about one-half of the test sites would score a 10. Although the metric
passed all other tests, its information content was low, predominantly because of a high degree of
variability (Figure 3-7) and increased abundance was not necessarily associated exclusively with either
good or impaired condition. This metric was dropped from the IBI suite.
3.14
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
IT
                                                   reference sites = 10

                                                   10!hPercentile
                                                   from test      = 0
                           Non-Reference
Figure 3-6. Derivation of maximum and minimum
metric scores at Reference and Test sites for number
of tolerant taxa metric (adjusted).
                                                                      10
                                                    10thPercentiie
                                                         test      = 0
       Reference
                  Non-Reference
 Figure 3-7. Derivation of maximum and minimum
 metric scores at Reference and Test sites for number
 offish collected metric (adjusted).
Mid-Atlantic Highlands Streams Assessment:  Technical Support Document
                                                                 3.15

-------
3.6     IB I Validation and Threshold Development
        3.6.1    IBI Validation
The IBI was calculated as the sum of nine [or 10 for some of the examples] metrics. The IBI scores
were evaluated against reference and disturbed sites in the validation data set and against the Watershed
Disturbance Index from the same data set. In the first comparison, the IBI clearly and statistically
distinguished reference from disturbed sites. It also clearly identified high versus low sites as identified
by the Watershed Disturbance Index (Figure 3-8). The IBI also was compared to Watershed Condition
Class (Biyce et al. 1999) and it demonstrated gradient of response from pristine to degraded condition
(Figure 3-9).
                                                       Site
                  s
                  &
                                 Degraded
                                                 Tsst p =
                                             Reference
                     100
                     90
                     80
                     70
                  8  60
                  S  50
                     40
                     30
                     20
                     10
                       0
                                 Using Watershed
     oo
                                                    D -- Q.OOS
                                                              Hitjil
                         Figure 3-8. Fish IBI scores at reference and
                         degraded sites and compared to Watershed
                         Disturbance Index from the validation data set.
3.16
Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------





£
o
o
05
5




100
90
80
70
60
50
40
30
20
10
0
                                   234
                                  Watershed Condition


               Figure 3-9. Comparison of the Fish IBI to Watershed Condition Class
               from the calibration data set.
During this validation phase, it was observed that fishlcss site, which scored 0, were distributed along the
disturbance gradient from low to high. This condition is analogous to that described earlier for the metric,
number offish. Number of individuals collected at a site may be low for two reasons:

(1)  severe disturbance means the site cannot support many fish; or

(2)  sites are naturally low in productivity.

It became apparent that the number of fish collected were directly related to habitat volume, which in
turn, is related to watershed size (Figure 3-10). These data indicated that the probability of finding fish
at Habitat Volume < 0.4 was veiy low and furthermore, that these low habitat volumes were all found in
watersheds < 2 km2 (494 acres). Because of this limitation, all fishless sites in watersheds < 2 km2 in size
were excluded from the analysis of condition.
            Mid-Atlantic Highlands Streams Assessment:  Technical Support Document
3.17

-------

TJ
Oj
u
41
_
U
To
3
T3
]>
TJ
C
V4_
O
5
JQ
e
lot.
3
"Z.









X
V
—
E
3
P
^
nj
25
ro
X






I \J\JU
1400 -


1200 -

1000


800 -


600 -


400 -


200 -
0 >
0

1 n -
i r\j

0.9

0.8

0.7
0.6

0.5

0.4

0.3

0.2
0.1
n n
U -U
t




•>.&=< .


C :
O j
^ i


/•"•I, j
t !' :
|
I ,••'", '""'' f'j ^ !
.""! <"'"• *'*"' ,(""•-. f'.'L,'~' '•
^'"^.L. ,-*-., "" ...J :
,,-.C:'l' v""' .-• . ../- O — ^
! .' "-" f*"Ii, t' *K 	 '* ).-«-' 	 f i:
•r">r;i. ,-~v"Si"".'Q ';.-.;"T'>'"rr<'..'\. ;
.— . ':•''•. '•— •'•-!• ^'^''-^''•'^''••ff^-^"" •S?*'V
'=';"' — , ™="; 	 ''! "'-- 	 '•I™X/""*™i._.".="'-L;;"' y ":''~" =>"'-.-^i..-^ — ^— *" .'^ 	 {/^tldo^l, 	 H^X^^A.-"*'-^: 	 j
.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1,0


1 :
' ,_% 0 Q
I C f~J '"' a-^ I
- .-. ""' '^.y' i
j O ..-, ,- vv w ,'•-, O "^ i
'-*.•. *-<"' / " fc- j™"". /,";. . °^> "^* . *'•"*" _m_ \
%f ,^ ."^p" ;":i;"x4cP^ % o.-, HD oi
P. .-T"'%>-Nc' _ 'c ° ^ '^ I
'^11 V\ -d""' — "^"^ ^
rx^Cf'"*-. .;"T '••—-"' '•— ' I
...^fi^'5'i^ /'•....' ^-'- i

™^: i
-< j !
i i
cq |
| |
X 	 L 	 , 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 _, 	 J
) 5 10 15 20 2S
Watershed Size {km')
Figure 3-10. Relationship of number of fish collected to habitat volume and
habitat volume to watershed size.
3.18
Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
       3.6.2   Development of IBI Thresholds and Estimation of Condition

The objective of threshold development was to derive IBI values that could be used to categorize stream
condition as good, fair, or poor. The distribution of IBI scores at reference sites was used to set these
thresholds in the following manner:

                      IBI > 25th percentile of reference scores = Good
                  5th < IBI < 25th percentile of reference scores = Fair
                      IBI <  5th percentile of reference scores = Poor

Three separate reference conditions were identified that ranged from the least restrictive with the most
sites included (n=27) to the most restrictive with the least number of sites included. Chemical and RBP
habitat criteria were used at all sites (the least restrictive), quantitative habitat filters were added to create
the medium level of restriction (n=23), and the most restricted (n=l 2) further added watershed condition
class. In order to acknowledge the uncertainty associated with each of the reference approaches, all three
were used. The mean of the 25th percentiles and 5th percentiles were calculated to derive the thresholds
as shown in Figure 3-11.
                                                Set
      100
                                                                   1   >K

                                                                    1  FAIR
       50
                                                                               Meanof 5lh
                                                                     POOR
                  Low             Medium             High
                      Restrictions on Reference Definition

          Figure 3-11. Calculation of good-fair-poor thresholds of condition
          based upon the Fish IBI.
I
            Mid-Atlantic Highlands Streams Assessment:  Technical Support Document
                    3.19

-------
Using these thresholds, stream condition was estimated for the MAHA region; these are shown in
'Fable 3-3. Sites with less than 10 fish observed and watershed area less than 2 km2 were not included in
the assessment and are noted in the "insufficient data" category.

          Table 3-3. Estimates of stream condition (% stream length) based upon Fish IBI.
Region
Western Appalachians
North-Central Appalachians
Ridge and Blue Ridge
Valleys
Entire MAHA
Good
3
15
28
23
17
Fair
32
32
44
37
36
Poor
30
43
14
31
31
Insufficient
Data
35
10
15
10
17
3.7    Non-Native Species Issue

The objective of the 1972 Clean Water Act is to "restore and maintain the chemical, physical and
biological integrity of the Nation's waters." To achieve this goal, the Act calls for the formal designation
of beneficial uses such as drinking water supply, primary contact recreation (e.g., swimming), and
aquatic life support (e.g., fish) for each stream. Each designated use has a unique set of water
quality requirements or criteria that must be met for the use to be attained. Some states have created
subcategories of aquatic life use for specific types of fisheries, such as cold water fisheries or warm
water fisheries, because the public wanted to develop and manage specific fisheries such as brown trout,
rainbow trout, or smallmouth bass fisheries in cold and cool water streams. In many streams these fish
are not native to the stream or watershed, but rather have been artificially introduced.  In these streams,
non-native fish have been stocked and are managed by the states for sport fisheries. Presence of non-
native fish does not necessarily imply poor stream condition in terms of habitat or water quality, but it
does mean the stream does not have a natural fish community which is of interest when assessing biotic
integrity and overall ecological condition.

Non-native fish species also can be a potential stressor on the aquatic resource. These introduced species
have been known to replace native fish by  direct predation or by out-competing them for available
habitat or food or both. In the Highlands, approximately 34% of the stream miles have non-native fish
species in  the fish communily. It is important to note thai this is a "presence/absence" criterion, and
may not represent a level at which stressor effects from introduced species occur. One may also wish to
assess effects at different thresholds of non-native individuals proportions (e.g., 10% or 50%).

The definition of biotic integrity used to develop the fish Index of Biotic Integrity reported in this
assessment considers the stream to be of lower quality or condition if non-native fish species arc present
in the stream because it is not the "natural" condition  for the stream. Among the purposes of a report like
the MAHA report is to simply present quantitative information on topics that ultimately will be debated
and decided by society. Many argue that non-native species and their introduction are a serious sign
3.20
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
of biological impairment and have significant economic impacts. Others argue that non-native sport
fish are highly prized and have an equal economic benefit. The MAHA report presents data from both
perspectives. Ultimately, society will be required to make an informed decision on what we want in
our streams and rivers.
            Mid-Atlantic Highlands Streams Assessment: Technical Support Document         3.21

-------

-------
4.0     Bcnthic Macroinvcrtcbratc Assemblage

4.1     Sample Collection and Processing

Benthic macroiinvertebrates were collected with a modified kick net (Figure 4-1) at each of nine cross-
section transects of the sampling reach (approximately 40 times the mean width) (Figure 4-2). Samples
were collected from a rectangular area 0.5 m2 area in front of the net (one net width and two net widths
long) by dislodging organisms with a 20-second kick followed by a hand-picking of any larger rocks
remaining in the 0.5 in2 area.  Samples for riffle/run and pool/glide were kept separate as individual
composites and preserved with ethanol to approximately a 70% final solution. No subsampling in the
field was conducted. Figure 4-2 depicts the sampling and compositing design.

                                 s         pm


                          End         V"-""

                        j •^f-*'""  ...--•"**
30cm

               Figure 4-1. Modified kick net for benthic macroinvertebrate sampling.


Of the sites visited for macroinvertebrate sampling, more than 90% had riffles and 40% were pools.
Data were collected from a total for 446 sites in 1993 and 1994. Benthic macroinvertebrates were
not identified or subsampled in the field. Preserved composite pool and riffle samples were sorted,
enumerated, and invertebrates identified to the lowest possible taxonomic level using specified standard
keys and references. Analytical methods are based on standard limnological practices. Figure 4-3 portrays
the steps in the laboratory analysis.
            Mid-Atlantic Highlands Streams Assessment: Technical Support Document
                     4.1

-------
                                       TRANSECT        (1 per transect)
                              Sampling point of 9ich transect (114,1/2, 3/4) selected at random
                                  Comtsirm »JF hick n»l
                                 samples collected from
                                    rifflos and runs
                                                  1
                                            Combifs® all McU net
                                           samples col So c ted from
                                                 pools

            Figure 4-2. Index sampling design for benthic macroinvertebrates.
4.2
Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------

                                 STREAM INDEX SAMPLES
                                    SAMPLE RECEIPT
                           »Inspect samples and complete tracking form
                           9 Store samples between 4 and 20°C
                                SUBSAMPUNG AND SORTING
                             »Grid squares selected at random
                             * All organisms removed from squares
                              until 300 organisms have (seen removed
                             * Organisms sorted into vials of major
                              taxonomic groups
     /Update refereoceV
     I     c0l!@cti®n     I
       Send specimens
     I      out for      J
     \.  confirmation  f
Figure 4-3. Laboratory sample analysis scheme for benthic macroinvertebrates.
      Mid-Atlantic Highlands Streams Assessment:  Technical Support Document
4.3

-------
4.2    Metric Selection and Testing

Stream condition as represented by macroinvertebrate assemblages was assessed using an EPT index, i.e.,
number of EPT taxa. This index reflects the number of species found in three orders of aquatic insects, the
mayflies (Ephemeropterd), stoncflics (Plecopterd), and caddis flics (Trichoptera). Insects in these three
orders are known to be sensitive to pollution and stream disturbance.

A mulli-metric index of biotic integrity using macroinvertebrate data is under development. For the
purposes of the MAHA State of the Streams report, an EPT metric was used as representative of the
metrics being tested and developed. The 46 macroinvertebrate metrics that were evaluated included:
10 richness measures; 22 trophic measures; 13 composition measures;  and three tolerance measures.

These metrics were evaluated for inclusion into a multimetric index by using the following procedures:
       box plots
       correlations with stressors
       relationships to watershed size
       extent of redundancy
       PCA  with chemistry and physical habitat parameters

The following eight metrics were  selected for inclusion in a Stream Benthos Integrity Index (SBII):

       total number of taxa
       modified HBI
       % Plecoptera taxa
       % Oligochaetes/leeches
       % non-insects
       % Chironomid taxa
       % intolerant taxa
       number of EPT taxa
                                                  Reference
                                                        Test
                                         Figure 4-4. EPT taxa index at hand-picked
                                         Reference and Test sites.
4.4
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
                                               4.3
                                                          Index Testing
                                               The 'EPT Index employed in the MAH'A streams
                                               report was responsive to stressor conditions
                                               as represented by conditions at hand-picked
                                               Reference and Test sites (Figure 4-4). It also
                                               compared favorably with the modified Stream
                                               Biotic Integrity Index (SBII) described above. That
                                               comparison, shown in Figure 4-5, indicates quile
                                               good agreement between the single EPT metric
                                               and a multimetric index.
                                               4.4
                                                          Reference Condition
              10
                       20
                  EPT Tax Index
                                 30
                                           40
                                                 The reference condition considered for benthic
                                                 macroinvertebrate EPT Index was the "minimally-
                                                 impaired" condition as there is no basis (e.g.,
                                                 museum records, publications) on which to
                                                 develop a historical reference condition.
Figure 4-5. Comparison of EPT taxa metric with
modified SBII.
                                                 A reference site approach was taken for the benthic
                                                 macro-invertebrates, which examined a subset of
                                                 the overall number of sample sites based upon
various physical and chemical parameters measured during collection. This approach was considered
feasible for the macroinvertebrates due to the large number of samples collected and analyzed. The
reference sites were developed (filtered) following Waite et al. (2000) using the following reference
criteria:

              Acid Neutralizing Capacity (ANC) > = 50 ueq/L (ca. = 2.5 mg/1 CaCO3)
              Chloride (Cl) < 100 ueq/L~(ca.	3.5 mg/1 Cl)
              Sulfate (SO42-) < 400 ueq/L (ca. = 19.2 rng/L SO421
              Total P < 20 ug/L
              Total N < 750 ug/L
              Mean RBP Metric Score > 15 (the mean score of all 12 metrics
              computed for the site, each ranging from 0-20)
           Mid-Atlantic Highlands Streams Assessment: Technical Support Document
                                                                                            4.5

-------
                        100-
                         80-
                      a,
                      :,
                      e
                      3
                      u
                            0 2468  10 !2 14 16 18 20 22 24 26 28 30 32
                                         EPTTaxa Richness

                          Figure 4-6. Cumulative distribution of EPTTaxa
                          Index scores for all probability sites and filtered
                          Reference sites (from J. Stoddard, unpublished).
Once the reference sites were selected, a 25th percentile score from those sites was selected as the cutoff
for "good" and "marginally impaired". The results for HPT Tax a Richness showed that in riffles the
25th percentile value was 17 and, in pools, it was 6. Figure 4-6 shows the cumulative distribution of
the reference site and all sites for the EPT Taxa Index and the derivation of the good and marginally
impaired threshold.

The following criteria for stream condition based upon the EPT Taxa Index were set:
                                          Riffles
                           Good
                           Marginal
                           Poor
                             >=17
                               9-16
                               0-8
4.6
Mid-Atlantic Highlands Streams Assessment: Technical Support Document

-------
5.0    Physical Habitat

Physical habitat in streams includes all those physical attributes that influence organisms within the
stream. Stream physical habitat varies naturally, as do biological characteristics; thus, expectations differ
even in the absence of anthropogenic disturbance. Within a given physiographic-climatic region, stream
drainage area and overall stream gradient are likely to be strong natural determinants of many aspects
of stream habitat.  This is due to their influence on discharge, flood stage, and stream power (the product
of discharge multiplied by gradient). Summarizing the results of a workshop conducted by EMAP on
stream monitoring design, Kaufmann (1993) identified seven general physical habitat attributes important
in influencing stream ecology:

       Channel Dimensions
       Channel Gradient
       Channel Substrate Size and Type
       Habitat Complexity and Cover
       Riparian Vegetation Cover and Structure
       Anthropogenic Alterations
       Channel-Riparian Interaction

All of these attributes may be directly or indirectly altered by anthropogenic activities. Nevertheless,
their expected values tend to vary systematically with stream size (drainage area) and overall gradient (as
measured from topographic maps). The relationships of specific physical habitat measurements described
in this section to these seven attributes are discussed by Kaufmann (1993). Aquatic macrophytes, riparian
vegetation, and large woody debris are included in physical habitat assessments because of their role in
modifying habitat structure and light inputs, even though they are actually biological measures.

5.1    Data Collection

The procedures were employed on a sampling reach length 40 times its low flow wetted width, as
described earlier in Section 2. Measurement points were systematically placed to statistically represent
the entire reach. Stream depth and wetted width were measured at very lightly spaced intervals, whereas
channel cross-section profiles, substrate, bank characteristics and riparian vegetation structure were
measured at larger spacings. Woody debris was tallied along the full length of the sampling reach, and
discharge was measured at one location. The tightly  spaced depth and width measures allowed calculation
of indices of channel structural complexity, objective classification of channel units such as pools, and
quantification of residual pool area, pool volume, and total stream volume.

There are live different components of the EMAP physical habitat characterization, including stream
discharge. The thalwegprofile is a longitudinal survey of depth, habitat class, and presence of soft/small
sediment at 100 equally spaced intervals (150 in streams less than 2.5 in wide) along the centerline
between the two ends of the sampling reach. "Thalweg" refers to the flow path of the deepest water
in a stream channel. Wetted width was measured at 21  equally spaced intervals. Data for the second
component, the woody debris tally, were recorded for each of 10 segments of stream located between the
11 transects. The third component, the channel and riparian characterization, includes measures and/or
visual estimates of channel dimensions, sinuosity, and morphometric complexity.
             Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
Stream Discharge: Stream discharge is equal to the product of the mean current velocity and vertical
cross sectional area of flowing water. Discharge measurements are critical for assessing trends in
streamwater acidity and other characteristics that are very sensitive to streamflow differences. Discharge
was measured at a suitable location within the sample reach that was as close as possible to the location
where chemical samples were collected (typically the X-site as described in Section 2). No single method
for measuring discharge was applicable to all types of stream channels. The preferred procedure for
obtaining discharge data was based on "velocity-area" methods (e.g., Rantz 1982; Lindslcy ct al. 1982).
For streams that were too small or too shallow to use the equipment required for the velocity-area
procedure, two alternative procedures were employed. One procedure is based on timing the filling of
a volume of water in a calibrated bucket. The second procedure is based on timing the movement of a
neutrally buoyant object (e.g., an orange) through a measured length of the channel, after measuring one
or more cross-sectional depth profiles within that length.

Thalweg Profile: The thalwcg profile is a longitudinal survey of maximum depth and  several other
selected characteristics at 100 or 150 equally spaced points along the centerline of the stream between
the two ends of the stream reach. Data from the thalweg profile allowed calculation of indices of residual
pool volume, stream size, channel  complexity, and the relative proportions of habitat types such as
riffles and pools.

Large Woody Debris Tally: Methods for large woody debris (LWD) measurement was a simplified
adaptation of those described by Robison and Beschta (1990). This component of the EMAP physical
habitat characterization allowed quantitative estimates of the number, size, total volume and distribution
of wood within the stream reach. LWD was defined here as woody material with a small end diameter
of al least 10 cm (4 inches) and a length of at least 1.5 m (5 ft). Generally, the extent  of large woody
debris is directly related to the extent of habitat complexity through development of obstructions and
diversions within the stream flow.

Slope and Bearing: The slope, or gradient, of the stream reach is useful in three different ways. First,
the overall stream gradient is one of the major stream classification variables, giving  an indication of
potential water velocities  and stream power, which are in turn important controls on aquatic habitat and
sediment transport within the reach. Second, the spatial variability of stream gradient is a measure of
habitat complexity, as  reflected in the diversity of water velocities and sediment sizes within the stream
reach. Lastly, using methods described by Stack (1989) and Robison and Kaufmann (1994), the water
surface slope allowed the computation of residual pool depths and volumes from the multiple depth and
width measurements taken in the thalwcg profile. Compass bearings between cross section stations, along
with the distance between stations, all owed the estimation of the sinuosity of the channel (ratio of the
length of the reach divided by the straight line distance between the two reach ends).

Substrate Size and Channel Dimensions: Substrate size is  one of the most important determinants of
habitat character for fish and macroinvertebrates in streams. Stream bottom characteristics are often cited
as major controls on the species composition of macroinvertebrate, periphyton, and fish assemblages
in streams (e.g., Hynes 1972; Cummins 1974; Platts et al. 1983). Along with bcdform (e.g., riffles
and pools), substrate character influences the hydraulic roughness and consequently the range of water
velocities in the channel. It also influences the size range of interstices that provide living space and cover
for macroinvertebrales, salamanders, and sculpins. Substrate characteristics are often sensitive indicators
of the effects of human activities on streams (MacDonald et al. 1991). Decreases in the mean substrate
size and increases in the percentage of fine sediments, for example, may destabilize channels and
indicate changes in the rates of upland erosion and sediment supply (Dietrich et al. 1989; Wilcock 998).
Consequently, changes in substrate size distributions are  often indicative of catchment and streamside


5.2          Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
disturbances that alter hillslope erosion or mobilize sediment. Accumulations of fine substrate particles
also fill the interstices of coarser bed materials, reducing habitat space and its availability for benthic
fish and macroinvertebrates (Platts et al. 1983; Hawkins et al. 1983; Rinne 1988). In addition, circulation
of well-oxygenated water is impeded when fine particles embed coarser, more permeable substrates.
Most practitioners (e.g., Platts et al. 1983; Bauer and Burton 1993), including the EMAP field protocols
(Kaufmann and Robison 1998) employ a systematic "pebble count," as described by Wolman (1954), to
quantify the substrate size distribution, with visual assessments of substrate embcddcdness as described
by Platts et al. (1983). Substrate size and embeddedness were evaluated at each of the 11 cross-section
transects using a combination of methods adapted from those described by Wolman (1954), Bain et al.
(1985), Platts et al. (1983), and Plafkin et al. (1989).

Bank Characteristics:  Bank and channel dimension measurements included bank angle and bank
undercut distance determined on the left and right banks at each cross section transect. Other features
that were measured included  the wetted width of the channel, the width of exposed mid-channel bars of
gravel or sand, estimated incision height, and the estimated height and width of the channel at bankfull
stage. The "bankfull" or "active" channel was defined as the channel that is filled by moderate-sized flood
events that typically occur eveiy one or two years. Such flows do not generally overtop the channel banks
to inundate the valley floodplain, and are believed to control channel dimensions in most streams.

Canopy Cover Measurements: The importance of riparian vegetation to channel structure, cover,
shading, nutrient inputs, large woody debris, wildlife corridors, and as a buffer against anthropogenic
perturbations is well recognized (Naiinan et al. 1988; Gregory et al. 1991). Riparian canopy cover over
a stream is important not only in its role in moderating stream temperatures through shading, but also
as an indicator of conditions that control bank stability and the potential for inputs of coarse and fine
particulate organic material (MacDonald et al. 1991). Organic inputs from riparian vegetation become
food for stream organisms and structure to create and maintain complex channel habitat. Canopy cover
over the stream is determined at each of the 11 cross-section transects. A Convex  Spherical Densiometer
(model B) was used (Lemmon 1957).

Riparian Vegetation Structure:  Visual estimation procedures were used to supplement previous
measurements with a semi-quantitative evaluation of the type and amount of various types of riparian
vegetation. These data were used to evaluate the health and level of disturbance of the stream corridor.
They also  provide an indication of the present and future potential for various types of organic inputs
and shading. Observations to  assess riparian vegetation apply to the riparian area upstream 5 m and
downstream 5 m from each of the 11 cross-section transects. They included the visible area from the
stream back a distance of 10 m (30 ft) shoreward from both the left and right banks, creating a
10 m x 10m riparian plot on  each  side of the stream. The riparian plot dimensions were estimated,
not measured. Riparian vegetation structure was measured by visual estimates of the areal cover and
type of vegetation in three layers (canopy, mid-layer, and ground cover), distinguishing evergreen from
deciduous vegetation, and woody trees and shrubs from herbaceous vegetation.

Instream  Fish Cover, Algae, and Aquatic Macrophytes: This portion of the EMAP physical habitat
protocol was a visual estimation procedure that semi-quantitatively evaluated the type and amount of
important  types of cover for fish and macroinvertebrates. Alone and in combination with other metrics,
this information was used to assess habitat complexity, fish cover, and channel disturbance. Estimates
were made of the areal cover  of all of the fish cover and other features that were in the water and on the
banks 5 m upstream and downstream of the cross-section. The areal cover classes of fish concealment and
other features were the same  as those described for riparian vegetation.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
Human Influence:  The field evaluation of the presence and proximity of various important types of
human land use activities in the stream riparian area was used in combination with mapped watershed land
use information to assess the potential degree of disturbance of the sample stream reaches. For the left
and right banks at each of the 11 detailed Channel and Riparian Cross-Sections, the presence/absence and
the proximity of 11 categories of human influences was evaluated. This assessment included the frequency
and extent of both in-channel and near-channel human activities and disturbances. In-channel disturbances
include channel revetment, pipes, straightening, bridges, culverts, and trash (e.g., car bodies, grocery
carts, pavement blocks, etc.). Near-channel riparian disturbances include buildings, lawns, roads, pastures,
orchards, and row crops. The observations and proximity evaluations were related to the stream and
riparian area within 5 m  upstream and 5 m downstream from the station.

5.2    Metric Selection and Testing

Eighteen metrics from the measurement suite described in Section 5.1 were selected for inclusion in seven
separate indices that describe physical habitat condition in the MAHA streams report. These indices and
composite metrics are defined in Section 5.3. The metrics selected are derived from channel morphology,
substrate, fish cover, riparian vegetation, riparian human disturbance, pool habitat, and riparian canopy
cover features as found in Kaufmann et al. (1999).

For each metric, an analysis of variance (ANOVA)  model was used to estimate variances among streams,
the signal, and those associated with repeat visits in the same year, which is referred to here as
measurement noise. The latter variance estimate includes measurement error, and combined effects of
within  season habitat variation, information collection by separate field crews, and ability to relocate
revisit samples. Three tests of precision were employed: a measure  of the repeat visit variance, i.e.,
residual mean square error in the ANOVA model; the coefficient of variation (CV), i.e., repeat visit
variance divided by the grand mean across sites as percent; and the signal to noise ratio which is the ratio
of the metric variance across the entire region to the repeat visit metric variance. Precision of a metric
will increase as repeat visit variance and CV decrease and the signal to noise ratio increases. The higher
the S/N ratio is for a metric, the more that metric is able to discern changes in single or multiple sites.
The number of streams evaluated and number of repeat visit data used in the analysis were  169 and
50, respectively.

The metrics selected for  inclusion in the seven habitat indices and results  of precision testing of 15
physical habitat variables are presented in Table 5-1. Precision testing for the following three metrics
were unavailable:

       * Percent of substrate as concrete (PCX RC)
       • Bed Stability (LRBS_BW4)
       « Mean Bed Shear Stress Index (LDMB	BW4)
5.4          Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
Table 5-1. Precision of Physical Habitat Metrics in the Mid-Atlantic region (N=169 streams with 50
repeat visits in 1993-1994), (after Kaufmann et al. 1999),

Thalweg mean depth, cm (XDEPTH)
Thalweg depth standard deviation, cm (SDDEPTH)
depth, cm (RP100)
channel gradient, % (XSLOPE)
16mm % (PCT_SFGF)
Embedded of midchannel and margin, % (XEMBED)
Areal cover of filamentous proportion (XFC_ALG)*
cover macrophytes, proportion {XFC__AGM)*
cover woody debris, proportion (XFC_LWD)
Areal cover of all types summed, proportion (XFC_ALL)
Canopy cover at bank by densitometer, % (XCDENBK)
Woody cover in proportion (XCMGW)
index compi
Riparian human disturbance from channel revetment
(W1_WALL)
Riparian Human Disturbance Index (W1_HALL)
Rcwisit
Var.
6.4
1.7
1.6
0.8
7.5
15
0.067
0.031
0.040
0.22
8.0
0.25

0.02
0.51
cv
(%)
22
13
17
42
17
27
224
102
142
46
10
28
162
20
41
Signal/
Noise
7.3
16
16
18
11
1.9
0.08
4.7
0.2
0.8
7.3
2.3
3.4
185
3.3
            Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
5.3    Index Calculations

       5.3.1   Index of Riparian Habitat Condition

The importance of riparian vegetation to channel structure, cover, shading, nutrient inputs, large woody
debris, wildlife corridors, and as a buffer against anthropogenic perturbations is well recognized (Naiman
ct al. 1988; Gregory et al. 1991 j. Riparian canopy cover over a stream is important not only for its role in
moderating stream temperatures through shading, but also as an indicator of conditions that control bank
stability and the potential for inputs of coarse and fine particulate organic material (MacDonald et al.
1991).  Organic inputs from riparian vegetation become food for stream organisms and provide structure
that creates and maintains complex channel habitat. Land use, buildings, and other evidence of human
activities in the stream channel and its riparian zone may, in themselves, serve as habitat quality
indicators; they may also serve as diagnostic indicators of anthropogenic stress. The EMAP wadeable
stream field methods (Kaufrnann and Robinson 1998) evaluate channel shading (using canopy densimeter
measurements) and riparian vegetation structure by visual estimates of the areal cover and type of
vegetation in three layers (canopy, mid-layer, and ground cover), distinguishing evergreen from deciduous
vegetation, and woody trees and shrubs from herbaceous vegetation. They assess the frequency and extent
of both in-channel and near-channel human activities and disturbances. In-channel disturbances include
channel revetment, pipes, straightening, bridges, culverts, and trash (e.g., car bodies, grocery carts,
pavement blocks, etc.). Near-channel riparian disturbances include buildings, lawns,  roads, pastures,
orchards, and row crops.

Aspects of riparian vegetation cover, riparian vegetation structural complexity,  and the intensity of human
disturbances were incorporated into the index of Riparian Habitat Quality used in the MAHA State
of Streams. Based on historic literature and the judgment of experts, the "pre-Columbian" reference
condition for riparian vegetation in the Mid-Atlantic Highlands was assumed to be a multi-storied
corridor of woody vegetation (XCMGW approaching 2.0), with bankside canopy density (XCDENBK)
generally complete (85%-100%) along wadeable streams. The reference condition was assumed to lack
the types of riparian human activities identified by the EMAP Physical Habitat field methods, which are
typical of an agro-industrial society. Kaufrnann et al. (1999) calculate the proximity-weighted sum of
human activities in the stream and riparian corridor as the variable Wl	HALL. To express the  combined
Riparian Habitat Quality imparted by Riparian vegetation, the variables XCMGW, XCDENBK, and
W 1_HALL were scaled from 0 (poor quality) to 1.0 (excellent quality) and combined by multiplication,
and application of the cube-root  of the product to avoid extreme skewness in the resultant index
(termed QWR1). A riparian habitat quality index value <0.50 denotes "Poor" condition, >0.50 to <0.63
"Marginal" condition, and values >0.63 indicate "Good" riparian condition.

       5.3.2   Channel Sedimentation Index

Stream bottom characteristics are often cited as major controls on the species composition of
macroinvertebrate, periphyton, and fish assemblages in streams (e.g., Ilynes 1972; Cummins 1974; Platts
ct al. 1983). Along with bcdform (e.g., riffles and pools), substrate size influences the hydraulic roughness
and consequently the range of water velocities in a stream channel. It also influences the size range of
interstices that provide living space and cover for macroinvertebrates, salamanders, sculpins, and darters.
Substrate characteristics are often sensitive indicators of the effects of human activities on streams
(MacDonald et al. 1991). Decreases in the mean substrate size and increases in the percentage of fine
sediments, for example, may destabilize channels and indicate changes in the rates of upland erosion and
sediment supply (Dietrich et al. 1989). Consequently, changes  in substrate size distributions are often
5.6          MM-Atlantic Highlands Streams Assessment: Technical Support Document

-------
indicative of catchment and streamside disturbances that alter hillslope erosion or mobilize sediment.
Accumulations of fine substrate particles also fill the interstices of coarser bed materials, reducing habitat
space and its availability for benthic fish and macroinvertebrates (Platts et al. 1983; Hawkins et al. 1983;
Rinnc 1988). In addition, circulation of well-oxygenated water is impeded when fine particles embed
coarser, more permeable substrates. Most practitioners (e.g., Platts et al. 1983; Bauer and Burton 1993),
including the EMAP field protocols (Kaufmann and Robinson 1998) employ a systematic "pebble count,"
as described by Wolman (1954), lo quantify the substrate size distribution, with visual assessments of
substrate embeddedness as described by Platts et al. (1983).

Stream bed substrate size distributions and their percentage of fine particles vary naturally among streams
of different sizes, slopes, and natural rates of upslope erosion. For the MAIIA State of Streams Stream
Sedimentation assessment, substrate reference condition assumptions are based on Section 3.2.7 of
Kaufmann et al. (1999). Stream sedimentation was defined as an increase or excess in the amount of fine
substrate particles relative lo an expected reference value that is based on the region and the sediment
transport capability (bankroll streambed shear stress) of each sample stream reach. Bankroll streambed
shear stress was estimated in this case by the variable LDMB BW4 (see discussion in Kaufmann et
al. 1999), which incoiporates physical habitat data on channel slope, bankfull dimensions, large woody-
debris, and channel cross-section irregularities. Stream channels undergo a long-term adjustment to
a region-specific rate of sediment supply delivered by erosion processes under a natural disturbance
regime. The size distribution of streambed particles is dependent upon the relationship between sediment
supply and stream sediment transport capability. We hypothesize  that, given a natural disturbance
regime, sediment supply in watersheds  not altered by human disturbances may be roughly in long-term
equilibrium with stream sediment transport. The relationship between bed particle size and stream
transport capability in  streams draining watersheds relatively undisturbed by humans should tend toward
a characteristic value typical to the region. The largest positive deviations in the amount of fine substrate
from predicted values were assumed to be in streams with high sediment input rates, and these augmented
rates are generally related to disturbance from human activities. This is born out from relating values of
observed/expected substrate diameter to watershed disturbances (see Kaufmann el al. 1999).

In the MAHA  State of Streams Sedimentation assessment, predicted values were approximated by
regressing PCT_SFGF (i.e., % substrate smaller than 16 mm diameter) on a measure of stream bed shear
stress (LDMB  BW4).  This procedure yields a range of deviation values above and below the regional
mean, which includes  contributions from streams over a wide range of disturbance. The lowest residuals
(i.e., negative residuals) from the prediction equation are from streams that do not have an excessive
amount of fine particles relative to expectations, and tend to be relatively undisturbed streams. Those with
the highest residuals are those streams with excess sedimentation, and these tend to drain basins with
relatively intensive and extensive human activities.

Values of excess fines percentage were  established in the following manner.  Streams with a PCX  SFGF
at least 10% below the predicted value were rated to be in "Good" condition relative to the sedimentation
criteria. Those with PCX SFGF  10% below to 20% above the predicted value were rated "Marginal".
Those with PCX_SFGF more than 20% above regional mean expectations were rated "Poor".

        5.3.3    Fish Cover from Large Woody Debris

This metric is the mean areal percent cover in the stream channel that is provided by woody debris with
diameter >0.3  m, as estimated by field crews. The variable name used here was XFC_LWD as described
by Kaufmann et al. (1999).
             MM-Atlantic Highlands Streams Assessment: Technical Support Document          5.7

-------
       5.3.4   Channel and Riparian Disturbance Index

This disturbance index is a proximity-weighted index of the extent and intensity of human activities within
the channel, riparian, and near the riparian, as visible to field crews working at the sample stream reach.
The index is calculated as the proximity-weighted sum of 11 categories of human disturbances, including
buildings, roads, mining activities, lawns and parks, pastures and grazing, row crops, dams and bank
revetments, influent and  effluent pipes, trash and landfills, land clearing, and silvicultural activities. It is
referred to by the variable name Wl  HALL in Kaufmann et al. (1999).

       5.3.5   Watershed Quality Index

This is an integrated index that combines information on the land cover, land use, road density, and
human population density in the contributing drainage area upstream from each sample stream reach.
The measure of natural land cover is the sum of percent arcal cover of "non-human" land cover (Forest
+ wetland + rock outcrop + open water from LUDA Land cover/Land use GIS coverage). Human
disturbance information  includes LUDA GIS cover for % Urban Land use, % Agricultural Land use, and
% Mining Land use. Road density is from "TIGER" GIS data, and human population density is from the
U.S. Census Bureau. Each land cover, land use type is given a separate modeled response shape describing
the relative contribution  (or degradation) to watershed quality as the percentage of the land cover/land use
type increases incrementally from zero to 100%, or the density of roads or human population increase
from zero to high values. The variable name used in the streams assessment was QW1.

       5.3.6   Watershed, Riparian, and Channel Habitat Complexity Index

This index, denoted as variable QWRC2, also is an integrated measure that combines the index of
in-channel habitat quality (QCPH2) with the same watershed and riparian quality measures for Watershed
Quality and Riparian Habitat Condition Indices described above. The in-channel measures exclude habitat
volume indicators, but include measures of five major aspects of channel habitat quality (the variable
names below are from Kaufmann et al. 1999):

       Velocity and  Stream Power:
       • Mean channel slope (XSLOPE)
       * Mean bed shear stress index (LDMB_BW4)

       Substrate Quality:
       • % embedded substrate (XEMBED)
       * % substrate <16mm diameter (PCT_SFGF)
       * % filamentous algae cover (PCT	ALG)
       • % aquatic macrophyte cover (PCT_AQM)

       Channel Alteration:
       • % substrate concrete (PCT_RC)
       • % revetted banks (Wl WALL)
       * % of channel stops with influent or effluent pipes (W 1_PIPE)
       * Bed Stability,  measured as a deviation of substrate mean diameter
         from that predicted from channel hydraulics (LRBS_BW4)
       * Deviation of residual pool depth (RP100) from that predicted
         from watershed area and channel slope
5.8         Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
       Channel Spatial Complexity:
       « Coefficient of Variation in Thalweg depth [100(SDDEPTH/XDEPTH)J

       Cover for Fish:
       * Sum of cover from all types of concealment features (boulders/ledges, undercuts, LWD,
         brush, overhanging vegetation, and artificial structures (XFC  ALL)
       * Cover diversity (number of different types of cover) — so far applied only in Reg 7
         Cover from brush + overhanging vegetation (XFC BRS + XFC OHV)
       * Cover from rock-related elements (XFC	RCK)
       • Undercut bank cover (XFCJJCB)
       * Large woody debris cover (XFC LWD)

       5.3.7   Channel Habitat Quality

This index also is an integrated measure of in-channel physical habitat quality that excludes habitat
volume indicators, but includes measures of five major aspects of channel habitat quality: Velocity and
Stream Power, Substrate Quality, Channel Alteration, Channel Spatial Complexity, and Cover for Fish.
The variables  used to quantify these five aspects of channel habitat quality are described  above, as they
contribute to the channel portion of the Watershed, Riparian, and Channel Habitat Complexity Index
QWRC2. It is referred to by the variable name QCPII2.
            MM-Atlantic Highlands Streams Assessment: Technical Support Document         5.9

-------

-------
6.0    Rapid Habitat and Visual Stream Assessment (EPA RBP)

6,1    Data Collection

This habitat assessment protocol was adapted from EPA's "rapid" bioassessmenl protocols (Plafkin et
al. 1989; Barbour et al. 1999), and has been refined from various applications across the country. The
approach focuses on integrating information from specific parameters on the structure of the physical
habitat. The objective of the visual stream assessment is to record field team observations of catchment
and stream characteristics that arc useful for data validation, future data interpretation, ecological
value assessment, development of associations, and verification of stressor data. The observations and
impressions of field teams are extremely valuable.

Each stream was classified as either "Riffle/run" or "Pool/glide" prevalent based on visual impression
of the dominant habitat type. For each  prevalent habitat type, twelve characteristics of habitat were
considered and evaluated as part of the rapid habitat assessment. These parameters include: instream
fish cover; bcnthic invertebrate cpifaunal substrate; cmbeddedncss; velocity and depth regimes; channel
alteration; sediment deposition; frequency of riffles; channel flow status; condition of banks; bank
vegetative protection; grazing or disruptive pressure; and riparian vegetated zone.

Most of the parameters were evaluated similarly for both types of prevalent habitats. In four cases,
the same parameter was evaluated differently, or a different (but ecologically equivalent) parameter
was evaluated in riffle/run prevalent versus pool/glide prevalent streams. Epifaunal substrates were
evaluated differently in riffle/run and pool/glide prevalent streams. Substrate cmbeddedncss was evaluated
in riffle/run prevalent streams, while pool substrate composition was evaluated in pool/glide prevalent
streams. The presence of four potential types of microhabitat types based on combinations of depth and
current velocity was evaluated in riffle/run prevalent streams, while the presence of four potential types of
pool microhabitat based on  depth and area were evaluated in pool/glide prevalent streams. The frequency
of riffles was evaluated in riffle/run prevalent streams, while channel sinuosity was evaluated in pool/glide
prevalent streams.

6.2    Metric Selection and Testing

As discussed in above, data were collected on 12 visual habitat metrics. These parameters include
the following:

    instream fish cover                                  frequency  of riffles (or channel sinuosity)
    benthic invertebrate epifaunal substrate               channel flow status
    embeddedness (or pool substrate characterization)      condition of banks
    velocity and depth regimes (or pool variability)        bank vegetative protection
    channel alteration                                   grazing or disruptive pressure
    sediment deposition                                 riparian vegetated zone

                    (Note that pools  and riffles were evaluated slightly differently.)
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document          6.1

-------
Each of these channel and riparian habitat metrics were scored by the field surveyors from poor
(score	0) to excellent (score	20).

The ANOVA model described above was again used by Kaufmann et al. (1999) to estimate the precision
of the RBP habitat metrics using the residual variance associated with repeat visits, the Coefficient of
Variation of that variance, and the signal to noise ratio. The precision test data for the 12 RBP metrics are
shown in Table 6-1 for a total of 459 stream samples with 36 repeat visits in the years 1993-1994.

Subcomponent metric repeat variance values ranged from 2.0 to 4.3 points (out of 20) and the CVs ranged
from 12 to 30%. Signal to noise ratio ranged from 0 to 4.2. There was general agreement among metrics
in all three values. Higher precision was associated with channel alteration, sediment deposition, riffle
frequency, bank condition, and grazing (or "other pressures") metrics. Lower precision was exhibited by
the instream cover, epifaunal substrate embeddedness,  and bank vegetation metrics. The highest S/N ratio
was found with riparian vegetative zone width, which had moderate values for the other two precision
estimates.
Table 6-1. Precision of Rapid Bioassessment Protocol (RBP) habitat quality metrics in Mid-Atlantic region
(N=459 streams with 36 repeat visits in 1993-1994) [after Kaufmann et al, 1999].
MBP Habitat Metrics
Instream Cover (Fish)
Epifaunal Substrate
Embeddeness (or Pool Substrate*)
Velocity/Depth Regime (or Pool Variability*)
Channel Alteration
Sediment Deposition
Riffle Frequency (or Channel Sinuosity*)
Channel Flow Status
Bank Condition
Bank Vegetative Protection
Grazing or Other Disruptive Pressure
Riparian Vegetative Zone Width
RBP Habitat Quality Total Score
Revisit Var.
3.7
4.3
3.6
3.2
2.0
2.5
2.8
3.2
2.5
3.7
2.3
2.9
23
CV (%)
28
30
28
25
12
19
18
22
18
25
15
24
14
Signal/Noise
0.7
0
0.6
0.9
2.0
2.5
1.1
0.8
1.8
0.4
3.3
4.2
1.6
 " Repeat visits were not made to measure these low gradient stream habitat assessment features.
6.2
Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
It should be noted that, in general, S/N ratios were substantially lower that those described for most of the
fish metrics. The RBP metrics associated with flow-related parameters are expected to exhibit ihe greatest
variability. Some of these do have the lower S/N ratios.

6.3    Index Calculation and Testing

The RBP Habitat Quality score is based upon the sum of the individual 12 metric score of 0-20, which
when summed can have a total score of 240. Tests of its precision also are found in Table 6-1. With repeat
variance and CV values of 23 and 14%, respectively, Kaufmann et al. (1999) conclude thai these values
are relatively small compared to the potential range of variation and the overall mean. They also found
that the RBP indicate a good potential to identify among-stream variation and change in habitat quality
over time. However, it also was observed that the low S/N ratio of 1.6 is indicative of either a true
lack of variation in habitat quality among streams or a failure of the RBP metric to be responsive to
habitat quality variation.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document          6.3

-------

-------
7.0    Watershed Disturbance

MAHA stream condition was evaluated by two independent measures of watershed disturbance. The
first was the Watershed Risk Index developed by Btyce et al. (1999) that classified streams into five
condition classes. A second measure defined as the Watershed Disturbance Index (Burch-Johnson, in
preparation) classified streams into good, fair, and poor categories. These indices and associated metrics
arc described below.

7.1    Watershed Risk Index

A watershed disturbance risk index was developed by Bryce et al. (1999) that incorporates landscape
features at the watershed level in order to identify the human activities that pose risks to stream
ecosystems. This index was used to evaluate 102 stream reaches and their watersheds that were otherwise
sampled in the MAHA program in 1993 and 1994. The watersheds were stratified by ccorcgion and
respective reference conditions that was defined as those sites minimally altered by human activity. In
general, these conditions were most often associated with mature second growth forests with roads absent
from the riparian zone and minimal human activity in the watershed.

       7.1.1    Watershed Disturbance Metrics

Three types of information were evaluated to identify metrics to be used in the risk index computation.
Watershed physical characteristics, population distributions, and farm/forest land use estimates were made
from U.S. Geological Survey 1:24,000 topographic maps. Aerial photographs taken from 1989 to 1993 at
1:40,000 scale by the National Aerial Photography Program (USDA-ASCS) were used to update USGS
maps and provide more detail on land use and land cover. Site visit data were reviewed to provide stream
reach physical habitat and riparian zone information. All identifiable human alterations were recorded,
particularly as they would influence vegetative cover, channel morphology,  sedimentation, and chemical
loading. Some of the predominant human activities included agriculture, silviculture, mining, urban and
residential development, and stream channelization. Table 7-1  lists the information obtained from each
of the noted sources.

       7.1.2    Index Computation

Regional, watershed,  and stream reach scale information gathered as noted in Table 7-1 were consolidated
into a strcssor matrix  for each of the  102 reach sites. Ecorcgional factors related to local climate,
lithology, soil credibility, stream density, and runoff were considered in developing expectations relative
to streamside and upland uses. Individual components of the stressor matrix were assigned a weight of
+, 0, or — depending on whether that condition preserved "naturalness", had a neutral effect, or was
detrimental to naturalness,  respectively. Not all stressors were  applied to each site; therefore the stressor
matrix for each site was somewhat unique. A risk index score of 1 to 5 was  assigned to characterize the
range of risk from minimal to highest risk of impairment. Table 7-2 offers an example of how six streams
were scored using the stressor matrix.
             Mid-A tlantic Highlands Streams Assessment:  Technical Support Document         1.1

-------
Table 7-1. Types of information obtained from data sources for incorporation in a Watershed Disturbance
Risk Index (Bryce et al. 1999),
       Topographic Maps
           (1:24,000)
                         Aerial Photographs
                              (1:40,000)
         Visit
         (1993-1994)
 Regional location
 Watershed size
 Elevation
 Drainage pattern
 Wetland
 Population pattern
 Relative area cleared
 Mines, gravel pits
 Oil and gas wells
 Road density
 Powerline corridors
 Protected areas
 Public land
                    Update % cleared
                    New development
                    Logging pattern
                    Riparian vegetation pattern
                    Relative forest age class
                    Rowcrop agriculture
                    Grazing (estimated)
                    Feedlots
                    Reclaimed mines
                    Channelization
Riparian age class
Canopy structure
Shoreline habitat complexity
Woody debris
Shoreline development
Farm type
Visible point sources
Visible recreation pressure
Presence of aquatic vegetation
Substrate types
Sedimentation
Impressions of biodiversity
Aesthetic appeal
Anecdotal information
Watersheds that were generally forested with low road and residential densities received a score of 1
or 2. A watershed with these characteristics would receive a 2 score due to a number of disqualifying
factors, such as a road paralleling a stream or presence of sedimentation. The highest risk score of 5
was reserved for those sites that exhibited a majority of conditions thought to negatively impact stream
condition. The presence of mitigating factors, such as mine reclamation, would lower that score to
the 4 categoiy. The final score integrated quantifiable aspects of watershed condition with  qualitative
interpretations of degree of impact. The repeatability of the scoring process was evaluated by Bryce et
al. (1999) with the result that two individuals scored 12 of the 13 evaluated watersheds alike.
7.2
Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
Table 7-2. Stressor matrix showing criteria and progression of risk index scores for six sites in the Ridge
and Valley ecoregion (Bryce et al, 1999),
STREAMED
RISK ATTRIBUTES
Protected area or trail access
Completely forested
Low instream sediment (0-30% area)
Complex instream fish habitat (>40% est. area of 10m w. trans.)
Large riparian trees (>0.3 dbh)
Few residences upland or streamside
Mostly forested (<30% cleared)
>18m forested riparian zone
Moderate streamside residential
Road density 5-15m/ha
Road parallels stream
Watershed cleared (30-60%), moderate agriculture and logging
Moderate sediment (30-50% area affected)
Trash, odor, surface film present
High watershed cleared (>60%)
Near stream agriculture, grazing, and logging
High bank erosion (50-60%)
Little instream fish habitat (<10% est. area of 10m wide trans.)
Road density >15 m/ha
High sediment (>50% area affected)
Minimal riparian buffer
High streamside industrial, urban, or rural point source (feedlot)
Channelization, dredging, rip-rap present
Oil and gas wells, pipes
Strip/underground mines, mine drainage

WT.
+
+
+
+
+
+
0
0
0
0
0
0
0
-
-
-
-
-
-
-
-
-
-
-
-

w
V
7
5
4
X
X
X
X
X




















1
p
A
7
5
6


X
X
X
X
X
X

X
X














2
M
D
7
7
2


X

X
X

X

X

X

X

X
X








3
W
V
7
5
6








X
X
X
X
X
X

X




X

X

X
4
P
A
5
2
2













X
X
X

X
X
X
X
X
X


5
P
A
5
3
2










X

X
X
X
X


X

X
X
X
X
X
5
            Mid-A tlantic Highlands Streams Assessment:  Technical Support Document
7.3

-------
        7.1.3   Testing of the Watershed Risk Index
The responsiveness of the watershed risk index was evaluated by comparison to chemical factors and
benthic macroinvertebrate measures collected synoptically in the same streams. PC-A was used to capture
nutrient richness (total P, total N, nitrate and ammonia-N) and ionic strength (eight major anions and
cations) gradients. The PCA on nutrient richness revealed two axes (PCA I and II) that accounted for 59
and 24% of the variability, respectively. Similarly, two axes of the PCA on ionic strength accounted for
61 and 14% of the variability. In general, the gradient in chemistry values in each ecoregion corresponded
with a gradient in risk scores, i.e., higher ionic strength and nutrient richness values were associated with
higher risk scores. For example, the first PCA axis for ionic strength shows a linear increase relative to
risk index scores; a similar relationship was found in a comparison to chloride  content (Figure 7-1).

Comparisons also were made to the biotic stream measures Ililsenhoff Biotic Index (IIBI) and % EPT
    a)
   0.07-

   0.05-

£ 0.03-
CB
c
£ 0.01-
5
O -0.01-
'c
— -0.03-
    O
   -0.05-

   -0.07-

   -0.09-

   -0.11
                                        b)
                 approximate
                   interval
                     T    I
                     2345
                   Risk Index Score
                                                       300
                                                       250'
                                                       200
                                        ^150'
                                        m
                                        TJ
                                        _o 100
                                        JC
                                        O
                                           50'
                                                        -50'
                                                                    i
                                                        2345
                                                      Risk Index Score
   Figure 7-1. Relationship of watershed risk index to ionic strength and chloride.
taxa. Figure 7-2 a andb shows fairly good agreement between improved biotic condition and watershed
risk. After adjustment of these two biotic measures for shear stress and elevation by regression analysis,
a weaker but identifiable relationship to watershed risk scores still existed, thus indicating that the risk
index has the ability to capture anthropogenic effects in spite of corrections for natural variability.
7.4
Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
a)
b)
                                                   80!

7-

-
m c

4-

3-

2











3fcf qu8rtii@ JL
m&disn mpm
fsfgi/wf/fe LJ

I I
JU TJ
4pJi Y



1 2





a








I

k
¥



3



I

f


J










4 5


3 6°"
UJ 40-
"c
O
e •
o
°~ 20-


0-


i


m
••






1
Risk Index Score
c)
q

2-

m
5"
a
1 o-
o
DC
1
9














1 1 ^^T1
TSr 1^
I







i






I

M
T

d)
JA



ai

e
){



J



I
30-
•
20-
UU 10-
T3 o-
2
W -10-
0
u.
-20-
-30-
_>in





J-
••




12345 1
Risk


!
i i




M 1

T








2



ij








1 i




8 1

T

345
Risk Index Score




'(








ft








)
I




1 J
1 B


L
B 1


1
I
2345
Risk Index Sc»re
 Figure 7-2. Comparison of the watershed risk index to biotic condition with normal score and
 those adjusted for natural variability.
         Mid-A tlantic Highlands Streams Assessment:  Technical Support Document
                                       7.5

-------
7.2    Watershed Disturbance Index

Research regarding appropriate thresholds or criteria for classifying individual stream watersheds is
continuing. Therefore, this approach should be viewed as the current status in the development process
rather than a finished product.

The EMAP-Surface Water classification scheme for these disturbance metrics was deliberately restricted
to watershed-level data derived from available sources (i.e., USGS Land Use/Land Cover and Census
Bureau data) using GIS techniques. Influences of watershed land use/land cover on aquatic ecosystems
have been widely reported in the literature (see Richards et al. 1996; Allan et al. 1997). However,
until recently, many investigations focused only on chemical contaminants, nutrient enrichment, a single
drainage basin, or one land use type.

       7.2.1   Watershed  Disturbance Metrics

A wide range of natural and  anthropogenic data are available for EMAP watersheds. Principal component
analyses (PCA) on Northeast lake data identified forest, urban, and agriculture percentages, human
population density, and road density as primary variables for watershed disturbance (Whittier et al. 1997).
In the Mid-Atlantic region, forest and agriculture percentages are strongly, inversely related to watershed
condition (Burch Johnson et al., in review) therefore, the forest variable was dropped. The percentage
of mines/quarries was added to the variable list because mining activities are an important stressor
in the Highlands. Threshold  values for each variable were determined by literature recommendations,
professional judgment, and experimentation. When using EMAP data to determine a cut-off, generally
the data were split by sampling year (93-94) and one half were restricted from the development process
for later testing. Often the experimental thresholds were first examined against the "condition class"
variable developed and documented by Bryce et al. (1999), and then applied to the entire data set.

The EMAP urban percentage criterion for the "poor" category was set progressively lower as more
information became available and more experimentation was done. The MDNR used a value of 50%
urban as part of the "degraded" criteria. Maxted and Shaver (1996) reported in a study of 38 Delaware
watersheds that stormwater management pond facilities did not attenuate the impacts of urbanization
once 20% impervious  cover  was reached. Further, about 90% of the sensitive macroinvertebrates were
generally eliminated at 10-15% impervious cover in the watershed. Asa "rale of thumb", the runoff
coefficient from highly developed urban areas is 0.8 - 0.9, given a particular rainfall amount and land
area (Corvallis Public Works Department, personal communication). A rough estimate of the percent
urban area was calculated as U = 1/0.8; where U = urban % and I = impervious surface %. Thus, serious
macrobenthos effects occurring at 10-20% impcrviousncss translates to roughly 12.5-25% urban. Wang
et al. (1997) reported a similar threshold of 10-20% urban land use beyond which IB1 scores were
consistently low for 134 stream sites in Wisconsin. Although these thresholds are a good starting point,
they represent areas of higher urbanization than in the Mid-Atlantic region. About 48% of the Delaware
project was in urban land while 7% of the Wisconsin watersheds were more than 20% urban (3% urban
for entire state). On average, the 368 EMAP-SW stream watersheds were only 1% urban, based on
classified thematic mapper data (Ilerlihy et al. 1998). When using the USGS Land Use/Land Cover
(LULC) data, the average  for the watersheds was about 2-4% urban. Of course, the percent of urban land
varies by ecoregion; ranging from 1.9% to 6.7% urban for the entire "Blue Ridge/Ridge" and "Valley"
ecoregions, respectively.
7.6          MM-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
For hydrologic units, the percent of urban land ranges from approximately 3.2% to 4.5%. With these
averages as guidelines, the cut-off for the "poor" category was set at 3% urban. In ihe "good" category, the
urban percentage criterion is zero. This does not mean that there is a total absence of "urban" features in
the watersheds because scattered residences and narrow commercial/residential developments along roads
or lake shorelines may not have met mapping criteria for either TM or LULC.

Mines/quarries comprise only a small portion of the total land cover in all ccorcgions (0.2% mines/
quarries in the Valleys up to 1.5% in the Northern and Central Appalachians). Few EMAP stream
watersheds have substantial amounts of mining (>10% mines). However, mining is a significant aquatic
stressor when present in a watershed. The thresholds were set at zero for "good" sites and at the 1994
sample mean of 0.6% for "poor".  Like the urban data, a zero percent mining value does not necessarily
imply a complete absence of mines/quarries in the watershed. The age of the LULC data and the difficulty
of detecting and mapping subsurface mines from high-altitude imagery may affect the percentages
reported. Because mines and quarries are classed together in the database, different effects cannot be
distinguished.

The thresholds for agriculture percentage are tentative. In Wisconsin, where 73% of the watersheds
studied were >50% agriculture, Wang et al. (1997) detected obvious declines in habitat quality and IBI
scores only after agricultural land exceeded 50%. However, some sites with more than  80% agriculture
retained good quality and biotic integrity. Bryce et al. (1999) used 30-60% cleared land (agriculture and/or
logging) to define "moderate" impacts and >60% cleared for the "highly disturbed" class when ranking
102 Mid-Atlantic watersheds. If calculated by MAFIA ecoregions, agriculture ranges from 13% in the
Blue Ridge/Ridge region to 57% in the Valleys. The mean agriculture percentage for all 1994 watersheds
was approximately 24% while the median was 15%. As a starting point, the agriculture thresholds were
set to the median of 15% for the "good" category and 45% for the "poor" class (i.e., 3 times 15%;
roughly equal to Wang et al.).

Although the literature frequently identifies roads as a watershed stressor, particularly in terms of chloride
in lakes or streams, few investigations try to quantify road density effects. McGurk and Fong (1995)
used an "equivalent roaded area" (ERA) index, developed by the USDA Forest Service, to assess the
effects of forest management in California's Sierra Nevada and Klamath mountain ranges. The method
does not separate road effects from other disturbances but standardizes management and natural activities
(clear-cuts, prescribed burns, wildfires) in terms of equivalent roaded acres based on coefficients. Road
cut-and-fill areas have a disturbance coefficient of 1.00 while a tractor clearcut has coefficients of 0.2
- 0.3. Equivalent Roaded Area values less than 5% were not associated with changes in aquatic insect
diversity, whereas higher values were associated with declines. Although this index cannot be used directly
for MAHA sites due to differences in purpose and road type or usage,  it does suggest that thresholds exist
and are likely to be low. By sorting Ihe  1994 EMAP data by road density, it appeared that the percent of
urban lands and condition classes were higher (indicating more disturbance) above the  mean road density
of 15 ni/ha, thus that became the cut-off for the "poor" class. Road densities between 10 and 15 m/ha were
most frequent, so 10 was set as the "good" threshold.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document          7.7

-------
The human population density thresholds were the result of some literature information, statistical
distributions, and professional judgment. Because detailed studies of lake water quality often include data
on the number, age, and septic systems of dwellings around the lake, it seemed that individual residences
are important "stressor" units. Further, each building is mapped on USGS 1:24,000-scale topographic
maps when it can be done legibly (USGS 1991). A "locale" is defined as a place at which there is or
was relatively minor human occupation or activity (i.e., farm, camp, ghost town, junction, railway station,
etc). Populated places arc classified by population and labeled using distinctive type sizes. A "compact
community" consists of 5-40 houses. According to EMAP watershed population and housing estimates
derived from 1990 Census data, the number of persons per household is most often 2-3. Also, the
frequency distribution of population density decreases almost exponentially; Ql 	2.99, median	8.09,
Q3 = 19.26, mean = 32.12, and maximum = 2,625.36. Because the first quartile translates to about one
dwelling with average occupancy, the threshold for "good" sites was set to 3. The Q3 and mean values
would be roughly equivalent to 6-16 houses, or "compact communities" as defined by USGS mapping
standards. Thus, the threshold for "poor" sites was set at  15 to connote a small community in a watershed.
These values are not definitive and will likely change when better information becomes available.

        7.2.2   Index Computation

The disturbance metrics were used to  define classes of increasing anthropogenic disturbance, such that
good < marginal < poor. All good criteria must be met (AND) to be classified in good condition and
cxcecdancc of any poor (OR) criteria will designate poor condition.  Streams not classified as good or
poor are in the fair category.
Table 7-3. Thresholds for watershed disturbance metrics classifying streams as in good or poor condition.
Watershed Metric
Urban Land Use (% cover)
Mines/Quarries (% cover)
Agriculture (% cover)
Road Density (m/ha)
Population Density (persons/km2)

0
0
<15
3
>0.6
>45
>15
>15
 7.8
Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
       7.2.3   Testing of the Watershed Disturbance Index

The 1993 and 1994 EMAP watersheds were classified using the above criteria. Some preliminary one-way
ANOVAs were conducted with the resulting watershed condition variable (wscond) and selected chemical,
physical habitat, and macrobenthos metrics. In general, the differences in means were more pronounced
for the chemistry variables than for habitat or benthos variables. The relationship with chloride (L CL)
was particularly strong. In many cases, variability was largest for streams in the "poor" category. Most
analyses showed significant differences of the means for at least the good and poor classes. The watershed
condition variable was calibrated to some of the qualitative condition class values (assigned by Bryce),
therefore a strong relationship was expected. However, this step seemed important lo make the watershed
condition variable a predictive '"screening" tool for the sites not yet assigned condition classes and to
efficiently identify candidate reference sites.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document          7.9

-------

-------
8.0
Fish Tissue Contaminants
Specimens offish species that commonly occurred throughout the region of interest, and that were
sufficiently abundant within a sampling reach were retained for analysis offish tissue contaminants. If
possible, two types of composite samples offish were prepared at each site. One composite sample was
prepared using individuals of a Primary Target Species, which included species of fish whose adults
are small (e.g., small minnows, sculpins, or darters). The second composite sample was prepared using
individuals of a Secondary Target Species, which were those whose adults are of larger size (e.g., suckers,
bass, trout, sunfish, carp).

At the analytical  laboratory, the fish were composited, processed, and analyzed by the methods
summarized in Table 8-1 for metals, Table 8-2 for pesticides, and Table 8-3 for PCB congeners. Maximum
holding times for frozen whole fish have not been established; all EMAP fish tissue samples were
analyzed within one year of date of collection.
  Table 8-1. Analytical methods for metals analysis in fish.
Analyte (CAS No.)'
Aluminum (7429-90-5)
Arsenic (7440-38-2)
Cadmium (744(M3=9)
Chromium (7440-47-3)
Copper (7440-50-8)
Iron (7439-89-6)
Lead (7439-92-1)
Nickel (7440-02-0)
Selenium (7782-49-2)
Silver (7440-22-4)
Tin (7440-3 1-5)
Zinc (7440-66-6)
Mercury (7439-97-6)
Detection
Limit
(neW*
10
2.0
0.2
0.1
5.0
50.0
0.1
0.5
0.1
0.01
0.05
50.0
0.01
Summary of Method
Digestion with hot HNO3 and H2O2.
Analysis by graphite furnace atomic
emission spectrometry (GFAAS) or
inductively coupled plasma (ICP)
Digestion with hot HNO3 and H2O2.
Analysis by cold vapor atomic ab-
sorption spectrometry
References
EPA 200.3 (rev. 1); EPA
200. 11 (EPA, 199 la);
McDaniel, 1990; EPA, 1989b;
CLP (EPA, 1991b); APHA,
1989; EPA 7000 series (EPA,
1990a)
EPA 200.3 (rev. 1),EPA
245.6 (rev. 1)
   " Chemical Abstract Sen-ices (CAS) registration number.
   6 Units are ng/g fresh tissue weight.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document
                                                                                     8.1

-------
Table 8-2. Analytical methods for pesticides analysis in fish.
Analyte No.f
Aldrin (309-00-2)
Chlordane-c;s (5103-71-9)
Chlordane-frans (5103-74-2)
2,4'-DDD (53-1 9-0)
4,4'-DDD (72-54-8)
2,4'-DDE (3424-82-6)
4,4'-DDE (72-55-9)
2,4'-DDT (789-02-6)
4,4'-DDT (50-29-3)
Dieldrin (60-57-1)
Endosulfan-I
Endosulfan-Il (33213-65-9)
Endrin (72-20-8)
Heptachlor (76-44-8)
Heptachlor Epoxide (1024-57-3)
Hexachlorobenzene [Gamma-BHC/Lindane] (58-89-9)
Mirex
frans-Nonaehlor (3765-80-5)
cfe-Nonachlor (51 03-73-1 )
Oxychlordane (27304-13-8)
Detection
Limit
1



















Summary of Method
Soxhlet extraction into
hexane/methylene
chloride; analysis by
gas chromatography/
electron capture
detection (GC/ECD)
recommended













References
EPA 608
(NOAA, 1988);
EPA 682
(NOAA, 1988);
CLP
(EPA, 19910)














      Chemical Abstract Services (CAS) registration number.
      Units are ng/'g fresh tissue weight.
8.2
Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
Table 8-3. Analytical methods for PCB congeners analysis in fish.
Analyte No.f
2,4-DichIorobiphenyl #8 (34883-43-7)
2,2',5-Trichlorobiphenyl #18 (37680-65-2)
2,4,4'-TrichlorobiphenyI #28 (7012-37-5)
2,2',5,51-Tetrachlorobiphenyl #52
2,2',3,51-Tetrachlorobiphenyl #44 (41464-39-5)
2,31,4,41-Tetrachlorobiphenyl #66 (32598-10-0)
2,2',4,5!5'-Pentachlorobiphenyl #101 (37680-73-2)
2,3',4,4',5-PentachIorobiphenyl #118 (31508-00-6)
2,3,3' ,4,4'-Pentach!orobiphenyl #105 (32598-14-4)
2,2',4,41,5,51-Haxachlorobiphenyl #153 (35065-27-1)
2,2',3,4,4',5-Hexachlorobiphenyl #138
2,2',3,4',5,5',6-Heptachlorobiphenyl #187
2,2',3,3114141-Hexachlorobiphenyl #128 (38380-07-3)
2,21,3,4,4',5p5'-HeptachlorobiphenyI #180
2,2',3,3l,4,4l,5-Heptachlorobiphenyl #170 (35065-30-6)
2,2',3,3',4,4',5,5-OctachIorobiphenyl #195 (52663-78-2)
2,2l,3,31,4,4',5>5l,6-Nonach]orobipheny] #206 (40186-7-2-9)
Decachlorobiphenyl #209 (2051-24-3)
3,3',4,4'-TetrachIorobiphenyl #77° (32598-13-3)
3,3',4,4',5-Pentachlorobiphenyl #126° {??)
3,3',4,41I5I51-Hexachlorobiphenyl #169° (32774-16-6)
Detection
Limit
1




















Summary of Method
Soxhlet extraction into
hexane/methylene
chloride; analysis by
gas chromatography/
electron capture
detection (GC/ECD)
recommended














References
EPA 682
(NOAA, 1988);
(EPA,
1990)

















            Mid-A tlantic Highlands Streams Assessment: Technical Support Document
8.3

-------

-------
9.0
Water Chemistry
The primary purposes of the water samples and the field chemical measurements are to determine:

       * Acid-base status
       * Trophic condition (nutrient enrichment)
       * Chemical Strcssors
       * Classification of water chemistry type

A 4-1, bulk sample was collected at the X-site for measurement of the major cations and anions, nutrients,
total iron and manganese, turbidity and color. Syringe samples also were collected from the same location
for analysis of pH, dissolved inorganic carbon, and monomeric aluminum species. In situ and streamside
measurements were made using field meters for specific conductance (or conductivity), dissolved oxygen
(DO), and temperature. DO and temperature were only collected at sites where sediment oxygen demand
was measured and these usually were those included in the physical habitat assessment.

Table 9-1 describes methods for field measurements and Table 9-2 indicates analytical methods for
laboratory measurements.
    Table 9-1. Field measurement methods for water chemistry.
Variable or
Measurement
Temperature, in situ
Dissolved oxygen,
in situ
Conductivity, field
Summary of Method
Measured at mid-channel using thermistor
probe.
Measured at mid-channel (streams) using
membrane electrode and meter.
Conductivity meter; reading corrected to
25 C
References
EPA 150.6; Chaloud
et al. (1989)
EPA 360,1; Chaloud
et al. (1989)
EPA 360.1
            Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
Table 9-2. Laboratory analytical methods for water chemistry.
Analyte
pH, closed system
pH, equilibrated
Acid Neutralizing
Capacity (ANC)
Carbon, dissolved*
inorganic (DIG), closed
system
Carbon, dissolved organic
(DOC)
Conductivity
Aluminum, total dissolved
Aluminum, monomeric
and organic monomeric
Summary of
Sample collected and analyzed without exposure to atmosphere;
electrometrlc determination (pH meter and glass combination electrode)
Equilibration with 300 ppm CO2 for 1 hr prior to analysis;
Electrometric determination (pH meter and glass combination electrode)
Acidlmetric titration to pH 3.5, with modified Gran plot analysis
Sample collected and analyzed without exposure to atmosphere;
acid-promoted oxidation to CO2, with detection by infrared
spectrophotometry
UV-promoted persulfete oxidation, detection by infrared
spectrophotometry
Electrolytic (conductance cell and meter)
Atomic absorption spectroscopy (graphite furnace)
Collection and analysis without exposure to atmosphere. Portion of
sample passed through a cation exchange column before analysis to
obtain estimate of organic-bound fraction. Colorimetric analysis
(automated pyrocatechol violet).
References
EPA 150.6 (modified);
U.S. EPA (1987)
EPA 150.6 (modified);
U.S. EPA (1987)
EPA 31 0.1 (modified);
U.S. EPA (1987)
U.S. EPA (1987)
EPA 41 5.2,
U.S. EPA (1987)
EPA 120.6,
U.S. EPA (1987)
EPA 202.2,
U.S. EPA (1987)
APHA 3000-A1 E.;
APHA(1989),
U.S. EPA (1987)

Calcium, Magnesium,
Sodium, Potassium
Ammonium
Atomic absorption spectroscopy (flame)
Colorimetric (automated phenate)
EPA 200.6,
U.S. EPA (1987)
EPA 350.7;
U.S. EPA (1987)
Unions
Chloride, Nitrate, Sulfate
Silica, dissolved
Phosphorus, total
Nitrogen, total
True Color
Turbidity
Total Suspended Solids
(TSS)
Ion chromatography
Automated Colorimetric (moiybdate blue)
Acid-persulfate digestion with automated Colorimetric determination
(moiybdate blue)
Alkaline persulfate digestion with determination of nitrate by
cadmium reduction and determination of nitrite by automated
colorimetry (EDTA/sulfanilimide).
Visual comparison to calibrated glass color disks
Nephelometric
Gravimetric
EPA 300.6;
U.S. EPA (1987)
EPA 370.1 (modified)
U.S. EPA (1987)
USGS 1-4600-78;
SkDugstadetal.(1979),
U.S. EPA (1987)
EPA (modified);
U.S. EPA (1987)
EPA 100.2 (modified),
APHA 204 A.;
U.S. EPA (1987)
APHA 21 4 A.,
EPA 180.1;
U.S. EPA (1987)
EPA 160.3;
APHA (1989)
           " for D1C. "dissolved" is defined as that portion passing through a 0.45 m nominal pore size filter.
           For other analytes, "dissolved" is defined as that portion passing through a 0.4 m pore size filter
           (Nudeopore or equivalent).
9.2
Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
10.0    Stressor Identification

Stressors were identified based on the 305(b) EPA Region and state Report to Congress, input from EPA
and state personnel, and knowledge of emerging issues in the Mid-Atlantic region. The focus was on
stressors that effected stream ecosystems. There was an emphasis on including not only chemical, but also
physical and biological stressors. Habitat indicators and metrics were selected so that potential stressors
to both riparian and instream habitat might be determined. Non-native fish were included as potential
stressors in the development of earlier work on fish IBI indices (Karr 1981, 1991; Karr et. al. 1986;
McCormicket al. 2001). The definition of biotic integrity, as used by Karr (1991) indicates that non-
native fish detract from the biotic integrity of stream ecosystems. There has been considerable research
on competitive and predatory interactions of non-native game fish on native fish species (see Nico et al.
1999), which indicates non-native fish can be stressors on native fish species. Considering non-native
fish a potential stressor on stream ecosystems, therefore, was not unreasonable and can be scientifically
justified. The issue of non-native game fish species as potential stressors revolves around sociopolitical
designations of uses in stream ecosystems and the subsequent management to achieve these designated
uses.  Presenting information on the proportion of stream miles with non-native species permits an
informed discussion on whether these species are considered stressors or success stories (see the
Highlands Streams Report).

The Highlands Streams Report refers to potential stressors because the linkage between stressors and
effects in Highland stream ecosystems has not been determined. Statistical association and regression
analyses are in progress, including exploratory analyses using multivariate statistical procedures such as
cluster, principal component, and factor analysis. Within stream association analyses are being conducted
to evaluate the relationships among habitat (e.g., instream and riparian indicators, metrics, and indices),
chemical (e.g., nutrient concentrations, SOD), and biological indicators and metrics with fish and benthic
assemblages. Similar analyses are being conducted to evaluate the relationships  among land use/land
cover indicators and instream indicators. These analyses were not included in the Highlands Stream
Report and,  therefore, are not included in this Technical Support Document. Subsequent reports will
provide results and supporting documentation for these analyses.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------

-------
11.0   Classification for Reporting Results

11.1   General Classification Approach

To compute population estimates with reasonable confidence intervals generally requires about 50 samples
per reporting unit (see Section 2.1 EMAP Design). The confidence limits for a sample size of 30 and 50
(proportion of the streams in poor condition < 25%), are about ±18 and 12%, respectively. Reporting units
with sample sizes less than 30 are not recommended. The sample size for many of the desired reporting
units in the Highlands (e.g., Level III or Level IV ecoregions, 8-digit HIJC watersheds, slates) ranged
from 6 samples to 83 samples per reporting unit. There were differential numbers of samples collected by
media, which further limited the number of samples available for each reporting unit. For example, there
were 448 sites sampled for stream chemistry and 446 for benthos across the Mid-Atlantic in 1993-94; 289
sites were sampled for fish, and 159 sites sampled for physical habitat (Table 2-1). The decision made for
the Highlands Streams Report was to use the lowest common denominator in determining the aggregation
needed to have about 30-50 sites per reporting unit for any of the media indicators. It would have been
confusing to the reader if some media were omitted because of insufficient sample sizes for reasonable
estimates in that reporting unit. This would have eliminated comparisons across reporting units for all
media. The decision was made to include all the media and aggregate smaller reporting units until the
sample size was appropriate for making reasonable population estimates. Therefore, both Level III and IV
ecoregions and 8-digit IIUC watersheds were aggregated to achieve the desired sample size. The results
reported in the Highlands Streams Report, by indicator type and aggregated reporting unit, arc shown
in Table 11-1 a, b, and c.

The number of samples by media by Level III and IV ecoregions and 8-digit HUC watershed reporting
units are listed in Table 11-2.  It is possible to make population estimates for some indicators in selected
media with these non-aggregated reporting units and still have reasonable confidence limits.

11.2   Watershed Classification

Watershed aggregations were based on the larger drainage basins into which the aggregated watersheds
contributed. The  Susquehanna had a sufficient number of samples for all media so aggregation was
unnecessary. The Allegheny and Monongahela watersheds were  aggregated because these two rivers join
in Pittsburgh to form the Ohio River. The Kanawha and Upper Ohio watersheds were aggregated because
these both drain into the Ohio River.

11.3   Ecoregion Classification

Ecoregion aggregations were based on conversations with J. Omernik, author of the Level III and Level
IV ecoregions for the U.S. (Omernik 1987, 1995). As indicated in Section 1, ecoregions were aggregated
to ensure there were adequate sample sizes in each aggregated ecoregion to make population estimates
with reasonable confidence limits.
             Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
             CO
            -a
             CD
            ±
             o
            ^o

             CO

             o
             CO
             CO
             CD
             c
             0)
             o
             a.
            ~o
            s
             o
             CD
 c
_g

TJ
 c
 o
            := -C
             C CO


             11
             CD •*•*
            i: CD
 CD c
 o o
 CD 'en
Q_ CD
            ja
                     i
           S
          e S
                       3;
Blu
          o
         Ic
         J .5

         if
         s
Ti
                                           8
abi
C
                                                                        Q 3  ts
11.2
Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
        -8
        £2

        _cg

        JZ
        I

        o

        "E
        _to


        

        ^3  <0
        •r- TJ

        T-  C


        « -55
          I

                 e S
K S
  m
11
C "ss

ll
t 8.
II

*«
C
C
                                                                Ci

                                                                8
                                                                JA
S  A

?  fe
.£  t>H
                                                              ^ Qj  ^

Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------




=3
_o
ta
"c
]E
D>
I
1=
CO
5

T5
2

CD

to
CO

"^
0>
o
Q,
jQ
-a
CD
«
tu
to
o
o CD
s "S
o o
o JS
O -D
o c
a. co
-— of
ra ~o
E ^
E £
« s
CD >
^3 0
W) 0

0 £
C m
CB c
0 ^o
CD "CD
Q, 0
t> O
T- °
•r- ~O
T- C


*s
i!
E 3
m
11
Netrth-Ce
Appalaeh
Il
s
s





c
*f
u

3



R



5




m



rl







0



m



2


5
m








| Fish mr

in



R



so




(N



^







m



«



2


(N
l>
(N








| EFT Index1

«



5



O
CN




so



C*l







ON



0



2


so
^








1 Non-native Fish3











































IM

1
1
1 Fish Tissue Cont

-,



2



Q-,




f~i



m







t~



«



«


— '
o






„
s
1

o



J^



o




^*



Tf







w



0



o


oo
Tt








0}
s

2



s



CN




o



m







s



0



o


s
•^si-








Mine Drainage4

2



2



p-




"



^







0



M



00


s
jzj







*fl
I Acidic Dqjositic

*-,



GO



^H




es



t*-







S



2



—


CN
'^1








| Total Nitrogen*

«



o%



2








CN







S



rO



-


*"*
w°i







»
| Total Phosphor™

S



C-l



O




0©



t— -1







OG



S



•n


cn
"*•
C^l








S
^
I

DO



2



O









^
.

Srf
•S
&~.
Mj

\J
s
o
fc!
$3
"° ^
^ S
Csj ^S
s 5/j
s
> g ttj
•§ 1 §
~S ^ QJ
^| ^ \)
g ^g vj "^ ^_
."S .> § O~ ^
1 ? 1 A A?
!< S ^ -~ - ~~
9 § ft S K ^

a, ^; ^s u -s >a:
•S 's '2 ^ '5; "^

^ ^ •! !> -S M
^ S £ g S g
1 S S g £ g
S Q ^ ^ 5; ??
K su y> a> g $o
E? s ~ a v 
-------





CD
CD
.c
co"
c:
o
'as
CD
O
O
CD
TJ
C
CO
JC
O>
I
=5
T5
JO
D)
I
_O

_cg
TJ
CD
JC
C
E
=3
T5
CD

O
CO
CD
Jj
CO
CD
Q,
CO
m
£
CO
CD (0
4= CD
03 ffl
**= i •
o ro
5— o
CD >
=5 "c
Z CO
"•8
T- <»
m «
IS
is i




u
1








s
i
UJ
1












O
C9
HI
E
o
O


















2
"Is "s
S 5
5
.1
1
£
I
• O
P
SJ{ t
I 8.
5 °^

—
5 a
= K
< 0






™ j™
||
<
1

s
si
2 a
03
™
S «
«d jg
II
i 8.
o a
z <
«
1,3
? ™
"O SE
ii




|
IS
1



q!


oo



0




c5



«





n
m





ON

VO
CM


2




§





S







1



g


E*.



S




s



f-





s

o


tN
OO




2





in







Mine Drainage



8


o-



s




s



s





CO

o


CN
OO




2





«n






rt
Acidic Depositioi



s


g



ON




S



S





CN

O


CN
oo




QO





m







Total Nitrogen



§


ft



o\




s



s





en

o


fN
QO




QQ





W1







Total Phosphoras



m


£



O




a



en





oo
CN

en
CN


eN
CN




en





O







Riparian Habitat



m


P



0




a



en
en





03

en
CN


CN
CN




a





o







1
1


m
0


o
CN



3




P



OO





8

m


oo




so





0^






1
Watershed Condi'



















































Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------

-------
12.0    Information Management

A description of information management practices for EMAP are found in U.S. EPA (1999). The
collection of streams monitoring data in the EMAP and MAIA programs by EPA and non-EPA
participants is coordinated by the EPA Western Ecology Division (WED - Corvallis, Oregon) under the
direction of the Surface Waters Principal Investigator, John Stoddard. Raw data are transferred to WED
and then are forwarded to researchers acting as indicator leads. These individuals are responsible for
coordination of indicator development and assessment of ecological condition in the Mid-Atlantic region.
The indicator leads for the data presented in the MAHA Streams Report arc as follows:

        « Macroinvcrtcbratcs:
               Donald Klemm
               'EPA National Exposure Research Laboratory, Cincinnati
        • Fish:
               Frank McCormick
               EPA National Exposure Research Laboratory, Cincinnati
        • Physical Habitat:
               Philip Kaufmann
               EPA National Health and Environmental Effects Laboratory, Corvallis
        « Watershed Risk:
               Robert Hughes
               EPA National Health and Environmental Effects Laboratory, Corvallis

Upon completion of indicator research, raw and summarized data are maintained by WTED Information
Management Team (POC: Marlys Cappaert) in SAS and Arc/Info on a Unix server.

Metadata for all data sets are produced in EMAP data catalog format and are provided along with
station-specific data on the EMAP public web site:

                 http://www.epa.gov/emap/html/dataFsurfwatr/data/niastreams/

Metadata and data sets currently residing on this site are:

        • Benthic macroinvertebrate counts and metrics
        * Fish assemblage counts, metrics, and identification codes
        * Fish tissue contaminants for metals and organics
        * Watershed characteristics
        * Physical habitat metrics
        * Sample site information
                  and
        • Stream chemistry measurements

These data are downloadable  in the form of comma-delimited text (.txt) files. WED personnel may be
contacted for access to these and other MAIIA data products in electronic or printed form.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document         12.1

-------

-------
13.0   References

Allan, ID. and Johnson, L.B. (1997). "Catchment-scale analysis of aquatic ecosystems." Freshwater
       Biology 37: 107-111.

American Public Health Association. (1989). Standard methods for the examination of water and
       wastewater. Seventeenth Edition. American Public Health Association, Washington, DC.

Baine, M.B, Finn, IT. and Booke, H.E. (1985). "Quantifying stream substrate for habitat analysis studies.'
       North American Journal of Fisheries Management 5: 499-500.

Barber, C.M. (1994). "Environmental Monitoring and Assessment Program: Indicator development
       strategy." EPA/620/R-94/022, U.S. Environmental Protection Agency, Office of Research and
       Development, Research Triangle Park, NC.

Harbour, M.T., (jerritsen, I, Snyder, B.D. and Stribling, IB. (1999). "Rapid Bioassessment Protocol for
       use in rivers and streams: periphyton, benthic macroinvertebrates, and fish, 21"1 edition,"
       EPA/841/B-99/002, U.S. Environmental Protection Agency, Office of Water, Washington, DC.

Bauer, S.B. and Burton, T.A. (1993). "Monitoring protocols to evaluate water quality effects of grazing
       management on western rangeland si reams. "EPA 910/9-91-001, U.S. Environmental Protection
       Agency, Region X, Seattle, WA, 166 p.

Bryce, S.A., Larsen, D.P., Hughes, R.M. and Kaufmann, PR. (1999). "Assessing relative risks to aquatic
       ecosystems: A Mid-Appalachian case study." Journal of American Water Resources Association
       35: 23-36.

Burch Johnson,  C.B., Weaver, G.W. and D.I1. Larsen. (In review). "A comparison of land use and land
       cover data in watersheds of the Mid-Atlantic region." Submitted to Environmental Management.

Chaloud, D.C., Nicholson, J.M., Baldigo, B.P., Hag ley, C.A. and Sutton, D.W. (1989). "Handbook of
       methods for acid deposition studies: Field methods for surface water chemistry."
       EPA/600/4-89/020, U.S. Environmental Protection Agency, Washington, DC.

Cochran, W.G. (1977). Sampling techniques. 3rd ed. John Wiley & Sons, New York.

Courlenay, W.R., Hensley, D.A., Taylor, IN. and McCann, .T.A. (1986). "Distribution of exotic fishes in
       North America." In Hocutt, C.H. and E.O. Wiley (eds.) The Zoogeography of North American
       Freshwater Fishes. John Wiley and Sons, Inc., 675-698.

Cummins, K.W. (1974). "Structure and function of stream ecosystems." Bioscience 24: 631 -641.

Denevan, W.M.  (1992). "The pristine myth: The landscape of the Americas in 1492!' Annals of the
       Association of American Geographers 82: 369-385.
            Mid-A tlantic Highlands Streams Assessment:  Technical Support Document         13.1

-------
Dietrich, W.Ii, Kirchner, J.W., Ikeda, H. and Iseya, F. (1989). "Sediment supply and the development of
       the coarse surface layer in gravel bed rivers." Nature 340: 215-217.

Fausch, K.D., Karr, J.R. andYant, P.R. (1984). "Regional application of an index of biotic integrity
       based upon stream fish communities." Transactions of the American Fisheries Society 113:
       38-55.

Gerritsen, J.G., Green,.!., and Preston, R. (1994). Establishment of regional reference conditions for
       stream biological assessment and watershed management: Watersheds '93, A National
       Conference on Watershed Management, Arlington, VA, March 21-24, 1993. EPA/804/R-94/002.
       U.S. Environmental Protection Agency, Office of Water, Washington, D.C.

Gregoiy, S.V., Swanson, F..!., McKee, W.A. and Cummins, K.W. (1991). "An ecosystem perspective of
        riparian zones" Bioscience 41: 540-551.

Hawkins, C.P., Murphy, M.L. and Anderson, N.J. (1983). "Density offish and salamanders in relation
       to riparian canopy and physical habitat in streams of the northwestern United States." Canadian
       Journal of Fisheries and Aquatic Science 40: 1173-1186.

Herlihy, A.T., Kaufmann, PR., Mitch, M.E.  and Brown, D.D. (1990). "Regional estimates of acid mine
       drainage impact on streams in the Mid-Atlantic and Southeastern United States." Water, Air and
       Soil Pollution 50: 91-107.

Herlihy, A.T., Kaufmann, P.R. and Mitch, M.E. (1991). "Chemical characteristics of streams in the
       Eastern United States: 11. Sources of acidity in acidic and low A.NC streams." Water Resources
       Research 27: 629-642.

Herlihy, A.T., Kaufmann, P.R, Church, M.R., Wigington, P.J., Jr., Webb, J.R. and Sale, M.J. (1993).
       "The effects of acidic deposition on streams in the Appalachian Mountain  and Piedmont region
       of the Mid-Atlantic United States." Water Resources Research 29: 2687-2703.

Herlihy, A.T., Stoddard, J.C. and Burch Johnson, C. (1998). "The relationship between stream chemistry
       and watershed land use in the Mid-Atlantic region, U.S." Water, Air, and Soil Pollution
        105: 377-386.

Herlihy, A.T, Larsen, D.P., Paulsen, S.G, Urquhart, N.S. andRosenbaum,B.J. (2000). "Designing
       a spatially balanced, randomized site selection process for regional stream surveys: The EMAP
       Mid-Atlantic pilot study" Environmental Monitoring Assessment 63(1): 95-113.

Hocutt, C.H., Jenkins, R.E. and Stauffer, J.R. Jr. (1986). "Zoogeography of the fishes of the
       Central Appalachians and Central Atlantic Coastal Plain." in C.H. Hocutt and E.O. Wiley, (eds.)
       The zoogeography of North American freshwater fishes. John Wiley and Sons, Inc., 161-211.

Hynes, H.B.N. (1972). Ecology of running waters. University of Toronto Press, Canada. 555p.
13.2         Mid-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
Hughes, R.M. (1993). "Stream indicator and design workshop." EPA/600/R-93/138, U.S. Environmental
       Protection Agency, Office of Research and Development, Corvallis, OR.

Jackson, L., Kurtz, J. and Fisher, W. (2000). "Evaluation guidelines for ecological indicators."
       EPA/620/R-99/005, U.S. Environmental Protection Agency, Office of Research and
       Development, Gulf Breeze, FL.

Jenkins, R.E. and Burkliead, N.M. (1994). Freshwater fishes of Virginia. American Fisheries Society,
       Bethesda, MD.

Jones, K.B., Riittcrs, K.H., Wickham, J.D., Tankcrslcy, R.D. Jr., O'Neill, R.V., Chaloud, D.J., Smith,
       E.R. and Neale, A.C. (1997). "An ecological assessment of the United States Mid-Atlantic
       region: A landscape atlas." EPA/6007R-9'77130. U.S. Environmental Protection Agency, Office of
       Research and Development, Washington, DC.

Karr, J.R. (1981). "Assessment of biotic integrity using fish communities." Fisheries 6: 21-27.

Karr, J.R. (1991). "Biological integrity: A long neglected aspect of water resource management."
       Ecological Applications 1: 66-84.

Karr, J.R., Fausch, K.D., Angermeier, PL., Yant, PR., and Schlosser, I..T.  (1986). "Assessing biological
       integrity in running waters:  A method and its rationale." Special Pub. 5. II, Natural History
       Survey. Urbana.

Kaulmann, PR. (1993). "Physical habitat." In R.M. Hughes (ed.) Stream indicator and design workshop.
       EPA/600/R-93/138, U.S. Environmental Protection Agency, Office of Research and
       Development, Corvallis, OR, 59-69.

Kaufmann, PR, Levine, P., Robison, E.G., Seeliger, C. and Peck, D.V. (1999). "Quantifying physical
       habitat in wadeable streams." EPA/620/R-99/003, U.S. Environmental Protection Agency, Office
       of Research and Development, Washington, D.C.

Kaufmann, PR. and Robison, E.G. (1998). "Physical Habitat Characterization." In J.M. Lazorchak, D.J.
       Klemm and D.V. Peck (eds.). Environmental monitoring and assessment program — surface
       waters: field operations and methods for measuring the ecological condition of wadeable
       streams. EPA/620/R-94/004F, U.S. Environmental Protection Agency, Office of Research and
       Development, Washington, DC, 77-118.

Lazorchak, J.M., Klemm, D.J., and Peck, D.V. (1998). "Environmental Monitoring and Assessment
       Program — Surface Waters: Field Operations and Methods for Measuring the Ecological
       Condition of Wadeable Streams " EPA/620/R-94/004F. U.S. Environmental Protection Agency,
       Washington, DC.

Lemmon, P.E. (1957). "A new instrument for measuring forest overstory density." Journal of Forestry
       55: 667-669.
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document         13.3

-------
Lindsley, R.K., Kohler, M.A. and Paulhus, J.L.H. (1982). Hydrology for engineers. McGraw Hill,
       New York.

MacDonald, L.H., Smart, A.W. andWissmar, R.C. (1991). "Monitoring guidelines to evaluate effects
       of forestry activities on streams in the Pacific Northwest and Alaska." EPA/910/9-91/001, U.S.
       Environmental Protection Agency, Region X, Seattle, WA. 166 p.

Maxtcd, J. and Shaver, E. (1996). "The use of retention basins to mitigate stormwater impacts on
       aquatic life." In: L. A. Roesner [ed.] Effects of watershed development and management on
       aquatic ecosystems. Proceedings of an Engineering Foundation Conference, Snowbird, Utah,
       American Society of Civil Engineers, NY.

McCormick, F.H. and Peck, D.V. (2000). "Application of the indicator evaluation guidelines to a
       multimetric indicator of ecological condition based on stream fish assemblage." Chapter 4, In
       Jackson, L., J. Kurtz,  and W. Fisher,  (eds.) Evaluation guidelines for ecological indicators. EPA/
       620/R-99/005, U.S. 'Environmental Protection Agency, Office of Research and Development,
       Research Triangle Park, NC, 107p.

McCormick, F.H, Hughes, R.M., Kaufmann, P.R., Peck, D.V. Stoddard, J.L, and Herlihy, A.T. (1999).
       "Development of an index of biotic integrity for the Mid-Atlantic Highlands region." Transactions
       of the American Fisheries Society 130(5): 857-877.

McDaniel, W. (1990). Method 200.3. Sample preparation procedure for spectrochemical analyses of
       total recoverable elements in biological tissues. Revision 1. U.S. Environmental Protection
       Agency, Cincinnati, OIL

McGurk, B.J. and Fong, D.R. (1995). "Equivalent roadecl area as a measure of cumulative effect of
       logging." Environmental Management 19: 609-621.

Naiman, R.J., Decamps, II, Pastor, J. and Johnston, C.A. (1988). "The potential importance of
       boundaries to fluvial ecosystems." Journal of North American Benthological Societ\! 1: 289-306.

Nico, E.G. and Fuller, PL. (1999). "Spatial and temporal patterns of non-indigenous fish introductions
       in the United States." Fisheries 24: 16-27.

NOAA. (1988). "Determination of extractable organic compounds in sediments and tissues using a high
       performance liquid chromatography extract clean-up and gas chromatography with an electron
       capture detector." NMFSF/NWC-153, Environmental Conservation Division, National Marine
       Fisheries Service,  National Atmospheric  and Oceanographic Administration, Seattle, WA.

Omernik, J.M. (1987). "Ecoregions of the conterminous United States." Annuls of the Association of
       American Geographers 11: 118-125.

Omernik, J.M. (1995). "Ecoregions: A spatial framework for environmental management." In W.
       Davis and T. Simon (eds.) Biological Assessment and Criteria: Tools for Water Resource
       Planning and Decision Making. Lewis Publishers, Boca Raton, FL, 45-62.
 13.4        Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
Plafkin, J.L., Barbour, M.T., Porter, K.D., Gross, S.K. and Hughes, R.M. (1989). "Rapid bioassessment
       protocols for use in streams and rivers: Benthic macroinvertebrates and fish." EPA/440/4-89/001,
       U.S. Environmental Protection Agency, Assessment and Watershed Protection Division,
       Washington, DC.

Platts, W.S., Megahan, W.F. and Minshall, G.W. (1983). "Methods for evaluating stream, riparian and
       biotic conditions." General Technical Report INT-138. U.S. Forest Service, Intermountain Forest
       and Range Experiment Station, Ogden, LIT.

Raitz, K.B., Ulack, R. and Leinbach, T.R. (1984). Appalachia: A regional geography. Weslview Press.
       Boulder, CO.

Rantz, S.E. (1982). Measurement and computation of stream/low:  Volume 1. Measurement of stage and
       discharge. U.S. Geological Survey, Water-Supply Paper 2175.

Rohde, F.C., Arndt, R.G., Lindquist, D.G. and Parnell,  J.F. (1994). Freshwater fishes of the Carolinas,
        Virginia, Maryland, and Delaware. University  of North Carolina Press, Chapel Hill, NC.

Richards, C., Johnson, L.B. and Host, G.E. (1996). "Landscape-scale influences on stream habitats and
       biota." Canadian Journal of fisheries and Aquatic Science 53: 295-311.

Rinne, J. (1988). "Effects of livestock grazing exclosure on aquatic macroinvertebrates in a montane
       stream, New Mexico." Great Basin Naturalist 48: 146-153.

Robison, E.G. and Beschta, R.L. (1990). "Characteristics of coarse woody debris for several coastal
       streams of southeast Alaska, USA." Canadian Journal of Fisheries and Aquatic Science
       47: 1684-1693.

Robison, E.G. and Kaufmann, PR. (1994).  "Evaluating two objective techniques to define pools in
       small streams." In R.A. Marston and V.A. Hasfurther (eds.) Effects of human induced changes
       on hydmlogic systems. Summer Symposium Proceedings, American Water Resources
       Association. Jackson  Hole, WY, 659-668.

Robison, H.W. and Buchanan, T.M. (1988). The fishes of Arkansas. University of Arkansas Press,
       Fayetteville, AR.

Science Advisory Board. (1988). Future risk: Research strategies for the 1990s. SAB report to L.M.
       Thomas, EPA Administrator.

Simon, T.P. and Lyons, J. (1995). "Application of the index of biotic  integrity to evaluate water resource
       integrity in freshwater ecosystems." In W.S. Davis and T.P Simon (eds.) Biological assessment
       and criteria:  Tools for water resource planning and decision-making. Lewis Publishers, Boca
       Raton, FL, 245-262.

Snedecor, G.W. and Cochran, W.G. (1967). Statistical Methods. 6th ed.  Iowa State University Press
       Ames, IA.
            Mid-A tlantic Highlands Streams Assessment: Technical Support Document         13.5

-------
Stack, B.R. (1989). "Factors influencing pool morphology in Oregon coastal streams. M.S. Thesis,
       Oregon State University, 109p.

Stevens, D.L. and Olsen A.R. (1999). "Spatially restricted surveys over time for aquatic resources."
       Journal of Agriculture, Biological and Environmental Statistics 4(4): 328-330.

Stoddard. (2000). Personal communication.

U.S. Department of Agriculture. (1996). America's private land, a geography of hope. United States
       Department of Agriculture. National Resources Conservation Service.

U.S. EPA. (1987).  "Handbook of methods for acid deposition studies: Laboratory analyses for surface
       water chemistry" EPA/600/4-87/026, U.S. Environmental Protection Agency, Office of Research
       and Development, Washington, DC.

U.S. EPA. (1990). Extraction and analysis oforganics in biological tissue, Method OB 8/90. U.S.
       Environmental Protection Agency, Environmental Services Division, Region IV, Analytical
       Support Branch, Athens, GA.

U.S. EPA. (1991a). "Methods for determination of metals in environmental samples."
       EPA/600/4-91/010, U.S. Environmental Protection Agency, Office of Research and Development,
       Environmental Monitoring Systems Laboratory, Cincinnati, OH.

U.S. EPA. (1991b). "Contract laboratory program — statement of work for inorganic analysis, multi-
       media, multi-concentration." Document ILMO2.0-ILMO2.1, U.S. Environmental Protection
       Agency.

U.S. EPA. (1991c). "Contract laboratory program — statement of work for organic analysis multi-media,
       multi-concentration." Document OLMO1.0-OLMO1.8, U.S. Environmental Protection Agency.

U.S. EPA. (1995).  "Stream fisheries impacted by acid mine drainage in Maryland, Ohio, Pennsylvania,
       Virginia and West Virginia." U.S. Environmental Protection Agency, Region 3, Philadelphia,
       PA.

U.S. EPA. (1997).  Environmental Monitoring and Assessment Program: Research plan 1997. U.S.
       Environmental Protection Agency, Office of Research and Development, Washington, DC.

U.S. EPA. (1999).  "EMAP information management plan: 1998-2001." EPA/62Q/R-99/QQla, U.S.
       Environmental Protection Agency, Office of Research and Development, Washington, DC.

U.S. Geological Survey National Mapping Division. (1991). Technical Instructions: Standards for
       1:100,000-scale quadrangle maps. U.S. Department of the Interior. Part 3, 28-37.

WTaite, I.R., Ilerlihy, A.T., Larsen, D.P., Klemm, D.J. (2000). "Comparing strengths of geographic and
       non-geographic classifications of stream benthic macroinverebrates in the Mid-Atlantic
       Highlands, USA." Journal of the North American Benthological Society 19(3): 429-441.
13.6        MM-Atlantic Highlands Streams Assessment:  Technical Support Document

-------
Wang, L., Lyons, I, Kanehl, P. and Gatti, R. (1997). "Influences of watershed land use on habitat
       quality and biolic integrity in Wisconsin streams." Fisheries 22: 6-12.

Whitticr, T.R., Halliwcll, D.B. and Paulscn, S.G. (1997). "Cyprinid distributions in Northeast USA.
       lakes: Evidence of regional-scale minnow biodiversity losses." Canadian Journal of Fisheries and
       Aquatic Science 54: 1593-1607.

Wilcock, PR. (1998). "Two-fraction model of initial sediment motion in gravel-bed rivers." Science
       280:410-412.

Wolman, M.G. (1954). "A method of sampling coarse river-bed material." Transactions of the American
       Geophysical Union 35: 951-956.

Woods, A.J., Omernik, J.M. and Brown, D.D. (1999). Level III and IV ecoregions of Delaware,
       Maryland, Pennsylvania, Virginia, and West Virginia. U.S. Environmental Protection Agency,
       Office of Research and Development, Corvallis, OR.
            Mid-A tlantic Highlands Streams Assessment: Technical Support Document        13.7

-------

-------
Appendix Table A-1. Assessment questions for the Mid-Atlantic Highland streams.
                                          (MAHA)
                                             Set of
                                     (Population)
 Category - Physical
  I,     How many stream miles are estimated to be in MAHA? Ecoregions?
  2,     How many stream miles of wadable streams are estimated to be in MAHA? Ecoregions?
               Where?
  3.     How many stream miles of each stream order are estimated to be in MAHA? Ecoregions?
        States?
  4,     How many stream miles in MAHA? Ecoregions? States? Are estimated to be remote?
  5.     What % of streams in MAHA, Ecoregions, states, had water in them (i.e., were not dry) at the
        time of sampling?
  6.     What % of streams in MAHA, Ecoregions, states, have gravel bottoms? Mud bottoms?
  7.     What % of stream miles in MAHA, Ecoregions, states are estimated to be in
        1.      Public ownership
        2.      Private ownership
  8.    What % of stream miles in MAHA, Ecoregions, states have buffer strips (i.e., trees, shrubs,
        vegetation - not cultivated, pasture or asphalt)?
  9.    What % of stream miles in MAHA, Ecoregions,      have in-stream obstructions?
 10.    What % of stream miles have bank revetment or artificial banks?
          -
 11.     What are the distributions of stream ANC, pH, SO4, AL, conductivity values in MAHA,
        Ecoregions, states?
 12.    What are the distributions of stream TP, NO3, and TSS concentrations in MAHA, Ecoregions,
        states?
 13.    What is the distribution of stream DO, % saturation values in MAHA, Ecoregions, states?
 14.    What is the distribution of stream SOD in MAHA, Ecoregions, states?
            Mid-A tlantic Highlands Streams Assessment: Technical Support Document        A1.1

-------
Appendix Table A-1 (con't). Assessment questions for the Mid-Atlantic Highland streams.
                                           (MAHA)
                                 Preliminary Set of
 Category -           Attributes:
 15.    Which fish species (and assemblages) are most ubiquitous in MAHA, Ecoregions, states?
 16.    What is the spatial distribution of the species listed above?
 17.    What % of stream miles in MAHA, Ecoregions, states, have exotic fish species?
 18.     What is average number of fish species/site for:
        1.      Ecoregion
        2.      State
        3.      MAHA
 19.    What is cumulative fish species richness for Ecoregion, MAHA, states?
 20.    What % of stream miles in MAHA, Ecoregions, states had fish with observed abnormalities?
 21.    What % of stream miles in MAHA Ecoregions      have threatened and endangered species?
 Category -           Attributes: FishabiHty
 22.    What % of stream miles in MAHA Ecoregions states have game fish?
 23.    What % of stream miles in MAHA, Ecoregions, states have legal     game fish?
 24.    What % of stream miles in MAHA Ecoregions      are cold vs warm water        as
        determined by the fish species?
 25.    What % of stream miles in MAHA, Ecoregions, states have size-distributions indicative of
        natural reproducing game fish populations?
        1.     Specific fish assemblages of interest
        2.     Cold water
        3.     Cool water (i.e., small mouth bass)
        4.     Warm water
 26.    What % of stream miles have fish tissue contaminant residue levels exceeding human health
        or wildlife criteria?
 Category -           Attributes:
 27.    What is the distribution of the total number of Benthic species/site in streams in MAHA,
        Ecoregions, states?
 28.    What is the distribution of stream E-P-T scores for MAHA, Ecoregions, states?
A1.2        Mid-A tlantic Highlands Streams Assessment:  Technical Support Document

-------
Appendix Table A-1 (con't). Assessment questions for the Mid-Atlantic Highland streams.
                                  Preliminary Set of
           -
 29.     What % of the    In MAHA Ecoregions states are in the following land use categories:
         1.      Agriculture
         2.      Forest
         3.      Urban
         4.      Wetlands (includes lakes, streams)
  30.     What is the distribution of the area of the above land use categories in watersheds, by stream
         order?
  31.     What % of stream miles have Superfond sites in the watershed?
  32.     What % of stream miles have point sources in the watershed?
  33.     What % of watersheds have gypsy moth infestations in the watershed that have been sprayed?
  34.     What % of watersheds have had pesticide or nutrient applications in the watershed?
  35.     What % of stream miles receive storm water discharge?
  36.     Where are the minimally impacted streams (reference conditions) and what are their
         landuse/landscape characteristics?
  37.     What % of stream miles are associated with heavily disturbed watersheds?
  38.     What is the distribution of connectivity (shape - complexity, dominance) indices for
         watersheds in MAHA, Ecoregions, states?
  39.     What are the changes for each questions above from 1970-1990?

 Category - Bioflc Integrity
 40.     What % of steam miles in MAHA, Ecoregions states have fish IBI scores indicating good, fair,
         and poor stream conditions?
         1.      Species Richness
         2.      % mtolerants
         3.      Cumulative Index, IBI (Ecoregion, WS)
                [Note:  Want similar scale across the region - vary metrics and scores, but use the
                       same process]
            Mid-A tlantic Highlands Streams Assessment: Technical Support Document       A1.3

-------
Appendix Table A-1 (con't). Assessment questions for the Mid-Atlantic Highland streams.
                                  Preliminary Set of Questions
 41.    What % of stream miles in MAHA, Ecoregions states in study area have EPT scores that
        indicate good, fair, and poor stream conditions?
        1.      E-P-T
        2,      Sampling richness - % dominance
        3.      Summary index, e.g., HBI
 42.    What % of stream miles in MAHA, Ecoregions states have periphyton assemblages that
        indicate nutrient enrichment?
        1.      Sampling richness
        2.      Biomass (i.e., - chl/cm2)
        3.      % Abundance or filamentous forms
                [Note: May be an index period question - calibrate?]
 Category -
 43.    What % of steam miles in MAHA, Ecoregion,      have riparian habitat scored in good, fair,
        or poor condition?
 44.    What % of stream miles in MAHA, Ecoregion,      have aesthetically-pleasing habitat?
 Category -        Acidity
 45.    What % of chronically acidic stream miles in MAHA, Ecoregions, states are associated with
        AMD or acidic deposition as measured by: ANC, pH, SO4, conductivity.
 46.    What % of stream miles in MAHA, Ecoregions, states are susceptible to acidic deposition?
           -                    -
 47.    What % of stream miles in (MAHA) Ecoregions states with degraded biotic integrity are
        associated with:
        1.      AMD
        2.      Acidic deposition
        3.      Eutrophication
        4.      Habitat degradation
        5.      Exotic
 48.    What % of stream miles with degraded biotic integrity are associated with specific chemical
        stressors such as metals (Zn, Cr, Cd), organics (TCDD, PCB's, etc.)?
 49.    What is the association of biotic integrity with different geologic types?
A1.4         Mid-A tlantic Highlands Streams Assessment: Technical Support Document

-------
Appendix Table A-1 (con't). Assessment questions for the Mid-Atlantic Highland streams.
                               Mid-Atlantic
                                           (MAHA)
                                  PreMmlnary Set of
  50.     What Is the association between biotic integrity and elevation?
  51.     What % of stream miles in MAHA, Ecoregions, states would be expected to have brook trout
         (mussels, endangered species, etc.) if:
         1.      Acidity
         2.      Eutrophication
         3.      Toxics
         4.      Habitat degradation
         were not impacting the stream system?
  52.     What are potential recovery times for         systems following improvement?
  53.     What % of stream miles in MAHA, Ecoregions,       with degraded biotic integrity are
         associated with:
         1.      % Agric - Till/No-Till
         2.      % Forest - Forest mgt. Practices (clear-cuttmg/selective)
         3.      Width of buffer strips
         4.      Erosion potential
         5.      Number of animal (i.e., poultry) production units
         6.      % urban
         7.      Interaction among stressor - land use - biotic responses
  54.     What % of stream miles in MAHA Ecoregions, states with degraded biotic integrity are
         associated with landscape indices such as:
         1.      Connectivity
         2.      Shape - complexity
         3.      Dominance
  55.     What        have occurred In the %       miles in MAHA, Ecoregions,       with
         degraded biotic integrity that are associated with changes in landscape indices?
         1970-1980-1990
  56.     What % of stream miles in MAHA, Ecoregions, states have degraded biotic integrity that is
         associated with indicators of condition from other EMAP/REMAP Resources (e.g., Forest
         canopy index, Agricultural Lands erosion potential indices)?
  57.     What % of stream miles in MAHA, Ecoregions, states have biotic integrity values that
         indicate cumulative impacts from different land uses in the watershed?
             Mid-A tlantic Highlands Streams Assessment: Technical Support Document        A1.5

-------

-------
     w o


     1 *S :



      r



        >->~>->->~>->->~>>->~>>->~>>->~>->->~>->->~>->->>->^




      ^^^^^^^^^^^^^.^^^.^^^.^^^.^^^.^^^.^^^.^^^^^.^^^.^^^.^^^.^^^.^^^^^^^
      ~l ;


        !i2EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE







        jkpggjSf*---"---*"-tt"-°ip*Th'--'0  •*'"'''"

         I  SO (*, [•«. >af T- !
        . j  (O *- CD (S n- I
        1 fc  S ? K 5 ?: i
      r^ tA  lA 1C) tfl lA ffl <
"i^J    J w  S KS W t*3 S*) •
 CD
 0)


 2
 co
 Q.

"D
 D)

"co
 h-E^!*-r~N.f^r~l
                                          iiSiK^S^^iS^iS
                                                                                             j
                                                      j LU UJ UJ LU UJ
                                                                                                        J   LU lil UJ



                                                                                                         E E E E
        z% K K % X   STQSa
        3ajp»KfflQ«™aD:5[Jj
        ^c^HttciUSLuPgo!
                        : to co h- :
                                    OoUJZM|o!{j|uu«og
                                                       ^iS:
                                                                                    g ga *o   K
                                                                                    *• O 2 ° O K ° F
                                                                                    uj¥>ri-¥K>'
i °>
J =! i?
   a
                                         i u. 0_ S W JO I
                                                                                   c 3 8 « a 2 S
                                                                                   3 < z o J z o
                                                                                   : 5 
-------
        :>~>~>->~>~>-zzzzz>->~>">-zzzzzzzzzz:
                                                                                                     -zzzzzizzz:
                                                                                                                     c\i

                                                                                                                     OJ

           CM C*J v CM T- •*-
                                                                         . M v~ v {*3 f*3 •
                                 ffiJ
                                                           8s*-  I
                                                                 ?   1
                                                   tf
                 gSiipgiii!
                            g|

                            II   .
                            K &   P

                            §1   P   S
                            S ±"  ^   O
                                                                          I  03 Z ffl LU I
                                                                               g g
                                                                                 i8lagll8|i§S2!

                                                                                 P=lgttigilQol35i:
csi
_x

T3
 C
 ®
 a.
 a.
       I
u_ u. y. u. is
ill tu LU in M
K a? K o' «S
                 tfc!
                                                                                          i ti tii fc tu uj I

-------
T3

 £
 Zi
 (0
 CO
 ->->->->->->-z>->-



             >zz>>>~zzzzzzzzzz>zz>>>>>>>zz;>>z;>-zz;zzzzzzzzzz>>>->>>->>->'''>''
                       it t !
                       liii

                    i SD lit W SO <


              S'^^^
                ^^^
                                   loej'^-^-osoffi^-m'^'CN'—wi
                                   jh-BSwir-^-OTife^BigmrjiN!

                                   ]^s^ioffis^o^58i3i'*oi!r'*j!

                  333
                                               O O

                                                                                                                   uj  ui  Hi
                                                                                                                   OJMCO
                                                                                                                   tEffilDC
                                                                                                                   UJ  111  LU  f
                                                                                m m ™t LL, u. i
                                                                                                                                                          ca o o *? u_  !
               Z  Z  2  2 Z Z '
Q,Q.Q,                                           i
{0ffl!

                                                                                                                                   ;<£<<<<<<<•
                                                                                                                                                                                    CM
                                                                                                                                                                                    <

-------
T5

£
zs
ra
CO


E

oj



"5


CO

CO
Q.

T3


CO
D)

"(0
g
'co

£
o
o
of
c
o

"5
o
_g
OT

csi
_x

T3
C
®
o.
Q.
            :>>•>-






       > > >





      ^ooGolrlrooaoGaoooofflooffloooooooooocDooeDooaooaooaoaooffloofflolrffl
      ^2sz2^52zz2zz2zz^zs2zs2zszzszzszzs:z^s:z^s:z^'^^"^'^5"^'^=I"^p^=1"^-^












        'toKSS^S^m^^^SiCTnSra^Ki^STO^i^E^^ooiOCTOfflcoBS^e^^.






       eQ
-------
T3

 £
 Zi
 (0
 CO
 -z>->->->-z>->->z>->->-z>zzzzzzzz:zzz:zzzz
                                            •ptlfflt'r-ir-SNJ'r-T-PiJi
                                 o o o o o o >
                                 f~ h. n -#• s m >

                                 t? « i2 to S m '>
                                                                                      J(0DIQK(Q«3K5PWa:§Q{QI
                                                                                                      I OS 85 «S 03 Q3 Q Q
                                                                                                     '
                          ' "* * •* "tf ** 't '
             Z
            _, UJ UJ SJJ

           ;|SSS!
                             LU :
                        iL
                                                                                                  x x x;


                                                                                                 - Q 9 Q
         oiJrio^oooossrj
         U-UJOUJLLOMCOro^SwO
                             111!
                             ) O O I
                               » w sh ^ » ^ (h *
       I 03 i
            ssgsllls!
                                     5 LU I

                                    11!
           : ? S 2 Bu c £   '
           • IK s u- co o S D; i
           I ffl UJ [^ 03 g ^ Q ,


           ' te a o   ^ ~' ^ '

           igjgigSi:
i o 3 3 on LU i
                      :SF!
                                      •m
                                                                                        r S Z S 5 S O I
i   uj     t£
'   OK   °
I   I O   C3


!   £3   oo


:   K H   p jjg
i ^ yj LU a: in v.
• —' S S O 3S S



I 5 i i Si i H
                                                                       Q. Q. 0_ Q. Q, 0_ Q. I

                                                                                                             iy uj LU lu
                                                                                                             S 5 S 5
                                                                                                             F F F F
                                                        :UJiiJLUUJiiJlJUllliiJlJUlJJiiJlJjljJiiliIjlJJiiliIjUJiillJ^
                                                                                                                          CM
                                                                                                                          <

-------
         !zzzzzzzzzzzzzzzzzzzzz>>->->:»->->:»->->:»->->>-;»->>-;»-zzzzzzzzzzzzzzz
                                                                                                                                               OJ
                                                                                                                                               <
                            • z > z > z z >- ;
      is.
      si'
                                                                                                                     :  z z z z z z :
T5

 £
 ZS

 ra
 CO



 E

 OJ




"5



 CO

 CO

 Q.

T3


 CO



 D)

"(0
 o
'co

 £
 o
 o
 of
 c
 o

"5
 o
_o
o: o; i
o o ;
                                                                                 ; z " z
                                                                                 ! I z I
GO

csi
                                                                            n   ft. 0. fl, ft. I
                                                                                                                   JLUUJUJiU
                                                                                                        ! ! 1 !
                                                                                                        ppilJl
_x

T3
 C
 ®
 o.
 OL
                                    :pPPPPPPPPPPPPPPPPPPPl

-------
T3
CO

-------
T5

£
zs
ra
CO
->->->->->->->->->-zzzzzzzzzzzzzzzzzzzz

       IJLIJ.Ii.IJLIJLIi.IJLIJLIi.Ii.lJLIi.Ii.LLIi.Ii.LLI^
                                 •W'«~T-(»)«i«'«-CM w «e <
                                                               "O "D «


                                                               ffl ra <
                                                                     J Cfl 05  CO I
                                                         S5 ff
                                                        . a. a.
                                                                  z: z z s:


                                                                  M to « cp
                                                                              ££0 OT 03 M CO
                                                                              0,0.0,0,0,
                                       ilssillisiiiiiissoisgiisiilill
                                       'UJUJ^IJJLJJUJIJJlJj!>lJJLU!>^UJ^tJJUJtJJlJJIlJLUlJJl^
                                        oo    ooooo    oo       ogoo  ooooooo
LU f£ > ',
> HI Q i

i F 5 '
co o o i

                                        3   ~ I K
                                        - z g = o

                                         sSlis
                                                    K „ C   S
                                                   , o i o z a
             BoziiKBzaioa:
                                         i|||l||;



                                         L.lL.ffi....'!Jr....*L.'d...ffi....:
                              !g^l!l|
                              lozzszZS,
              ! 8 S 8 8 !
        SS5SSSSS5S
                           SSSSESSESSE:
                           FFFFFFFFFFFi
       isslssl!

                            lOGGOOGOraGOOOOfflOOOOOtSOOOOOOOG



                                                           t__l_J_t~-l~h~l~h~h~l~h~h~l~h~h~l~h~
                                                                                                  OJ
                                                                                                  <

-------
13

 C

 41

 Q.

 Q.
                                                                                                                                                              CM

                                                                                                                                                              <
         fzzzzzzzzzzzzzzzz
T3
 CO
 
-------