United States
Environmental Protection
Agencf
Developing Biological
Indicators: Lessons Learned
from Mid-Atlantic Streams
*-* -f -,- i--
-V:'-^*76!
_
-------
-------
EPA/903/R-03/003
March 2003
Developing Biological Indicators:
Lessons Learned from
Mid-Atlantic Streams
Prepared for:
Wayne Davis
U.S. Environmental Protection Agency
Environmental Science Center
701 Mapes Road
Ft. Meade, MD 20755-5350
Prepared by:
Leska S. Fore
Statistical Design
136NW40thSt.
Seattle, WA 98107
leska@seanet. com
Under subcontract from:
Technology Planning & Management Corporation
Wharf Plaza, Suite 208
Scituate, MA 02066
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
NOTICE
This document has been reviewed and approved in accordance with U.S. Environmental Protection
Agency policy. Mention of trade names, products, or services does not convey and should not be
interpreted as conveying, official USEPA approval, endorsement, or recommendation for use.
The appropriate citation for this report is:
Fore, Leska S. 2003. Developing Biological Indicators: Lessons Learned from Mid-Atlantic
Streams. EPA 903/R-003/003. U.S. Environmental Protection Agency, Office of
Environmental Information and Mid-Atlantic Integrated Assessment Program, Region 3, Ft.
Meade, MD.
This document can be downloaded from EPA's website for Biological Indicators of Environmental
Health:
http://www.epa.gov/bioindicators/
in
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
ACKNOWLEDGMENTS
This document builds on the work of hundreds of people involved in the sampling design, field
work, laboratory analysis, data management and statistical analysis for the Mid-Atlantic Integrated
Assessment project. Much of the analysis described here was derived from the workshop
discussions and publications of the participants. Editorial reviews by E. Chu and J. Scott improved
this report. Additional helpful reviews were provided by K. Blocksom, W. Davis, P. Larsen, L.
Reynolds, and J. Stoddard. Funding was provided by the U.S. Environmental Protection Agency
under U.S. Department of Commerce, Commerce Information Technical Solutions Contract No.
50-CMAA-900065 with Technology Planning and Management Corporation.
IV
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
TABLE OF CONTENTS
NOTICE iii
ACKNOWLEDGMENTS iv
FIGURES vi
TABLES vii
ABSTRACT ix
I. INTRODUCTION 1
II. SAMPLING DESIGN 3
A probabilistic sampling design was the best choice for MAIA 4
Free information is cost-effective 5
Reference sites did not always meet criteria for reference condition 6
III. THE PERILS OF DATA MANAGEMENT 9
Different "names" for the same site caused confusion 9
Original data must be archived 10
File structure mattered 10
Simple files were best 13
IV. LINKING HUMAN DISTURBANCE TO BIOLOGICAL CHANGE 15
Addressing concerns about circular reasoning 15
Metric testing included safeguards against circular reasoning 18
Patterns of human disturbance were complex 19
Integrated measures of disturbance were better predictors of index values 20
V. METRIC TESTING 23
Simple criteria were used first to eliminate potential metrics 24
Statistical precision was no substitute for correlation with disturbance 24
Watershed features were confounded with metric response to disturbance 26
Metrics from different assemblage types were eliminated for different reasons 26
VI. DEVELOPMENT AND APPLICATION OF MULTIMETRIC INDEXES 29
Biological criteria depend on the definition of reference sites 29
Patterns of index variability were similar across assemblage types 29
Invertebrate and diatom index values were comparable for
pool and riffle samples 33
Assemblages differed in their sensitivity to disturbance types 33
VII. CONCLUSIONS 35
REFERENCES 37
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
FIGURES
FIGURE 1 COMPARISON OF REFERENCE SITE SELECTON CRITERIA 8
FIGURE 2 EXAMPLE DATA FILE FOR DIATOM SAMPLES 12
FIGURE 3 EXAMPLE OF A HORIZONTAL FILE STRUCTURE 14
FIGURE 4 MULTIMETRIC INDEX VALUES BY DISTURBANCE TYPE 17
FIGURE 5 RANGES OF VALUES FOR POOL DEPTH AND EMBEDDEDNESS 25
FIGURE 6 VARIANCE COMPONENTS FOR THE INVERTEBRATE INDEX AND ITS METRICS 31
FIGURE 7 PERCENTAGE CHANGE THAT MULTIMETRIC INDEXES COULD DETECT 32
VI
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
TABLES
TABLE 1 SPEARMAN'S CORRELATION OF MULTIMETRIC INDEXES
AND HUMAN DISTURBANCE MEASURES 18
TABLE 2 SPEARMAN'S CORRELATION MATRIX FOR MEASURES OF HUMAN DISTURBANCE .... 21
TABLE 3 CANDIDATE METRICS TESTED FOR FISH AND INVERTEBRATE
MULTIMETRIC INDEXES AND REASON FOR EXCLUSION 27
TABLE 4 COMPONENTS OF VARIANCE FOR DIATOM, INVERTEBRATE,
AND FISH MULTIMETRIC INDEXES 31
TABLE 5 BIOLOGICAL METRICS INCLUDED IN THE FISH, INVERTEBRATE,
AND DIATOM INDEXES 34
Vll
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
ABSTRACT
As part of the U.S. Environmental Protection Agency's (EPA) Environmental Monitoring and
Assessment Program (EMAP), a survey of water chemistry, land use, riparian condition, and
channel morphology was conducted to understand how human influence alters fish, invertebrate
and periphyton assemblages. During 1993-1996, 296 sites were sampled for fish, 583 for
invertebrates and 317 for periphyton. A primary goal of the Mid-Atlantic Integrated Assessment
(MAIA) study was to define biological indicators for each assemblage that could be used to assess
stream condition at the regional level.
During the course of the project, researchers working independently derived different approaches
to data analysis and reported different results regarding the relationship between human influence
and biological change. To build consensus among the scientists involved, EPA sponsored a series
of workshops to create a consistent approach for testing and selecting biological indicators for fish,
invertebrates and periphyton. This document presents some of the issues from those workshops
that were most challenging to resolve.
The probabilistic sampling approach was the most efficient method for obtaining an unbiased
estimate of regional condition and is a cornerstone of EMAP. The coarse level of resolution has
made this design somewhat controversial for application at the state level, but alternative sampling
designs fail to yield data that can be applied across the region. Although much attention was given
to database design, data management issues often drove the agenda. Major issues not anticipated
during the design phase arose in the course of analyzing the data. Final results were delayed, for
example, by lengthy discussions of how to count invertebrate taxa with incomplete phytogeny and
how to combine periphyton data from soft algae and diatom counts that were derived from
different laboratory methods.
One of the greatest challenges for the MAIA project was selecting the best measure of human
disturbance from among the hundreds of potential variables. Through discussion and consensus, a
set of variables was selected that summarized human influence at multiple spatial scales and
included measures related to specific types of disturbance and measures that integrated across the
landscape. During the process of testing biological attributes for their association with disturbance,
concerns repeatedly arose regarding whether the results would be robust in new contexts. As a
consequence, safeguards were introduced at many steps of the analysis to avoid circular reasoning.
Similar statistical tests and criteria were used to select from among the list of candidate metrics for
fish, invertebrates and periphyton. All three assemblages showed strong and consistent
associations with general measures of disturbance, but showed different sensitivities to individual
stressors. Diatoms were more sensitive to water chemistry and invertebrates were more sensitive to
riparian condition. Multimetric indexes for each assemblage showed a similar ability to detect both
differences in sites and change through time (trend). The indexes for each assemblage represent
statistically reliable and biological meaningful monitoring tools for assessing and reporting the
biological condition of wadeable streams in the Mid-Atlantic.
IX
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
I. INTRODUCTION
The Environmental Monitoring and Assessment Program (EMAP) of the U.S. Environmental Protection
Agency (EPA) was implemented to answer questions about the status and trends of natural resources and
to connect observed changes in biological condition to stresses associated with human influence (Stevens,
1994). The program's emphasis on biological indicators of resource condition and its probabilistic
sampling design distinguish EMAP from other national monitoring programs (Olsen et al., 1999). One of
EMAP's fundamental mandates is to use the probabilistic sampling design to generalize conclusions
drawn from a sample of sites to an entire region with a known level of statistical confidence (Stevens,
1994; Urquhart et al., 1998).
EMAP was conceived to include all of the nation's natural resources. To match that scope, resources were
partitioned into groups: agro-ecosystems, rangeland, forests, the Great Lakes, estuaries, inland surface
waters, and wetlands. To demonstrate how monitoring and assessment at the regional scale could be
achieved, EPA implemented the Mid-Atlantic Integrated Assessment (MAIA) pilot study for surface
waters. Data were collected from wadeable streams in the eastern United States, from portions of
Pennsylvania, Maryland, Delaware, Virginia, West Virginia, and southeastern New York (Herlihy et al.,
2000). During the project's first four years (1993-1996), hundreds of sites were sampled for water
chemistry, riparian condition, channel morphology, land cover, and fish tissue contaminants; samples
were also taken offish, invertebrate, and periphyton assemblages (Davis and Scott, 2000).
Although EMAP's goals have shifted somewhat since the program began in the late 1980s, its
commitment to assessing resource condition in terms of resident biota has been constant (NRC, 1995).
For surface waters, one purpose of EMAP is to support the goals of the Clean Water Act (Paulsen et al.,
1998; Hughes et al., 2000). Under the act, states are required to assess their surface waters periodically
and report the condition of those waters to Congress (Karr, 1991; Ransel, 1995). For biological
monitoring and reporting, this mandate has typically been met by assessments based on multimetric
biological indexes (Hughes et al., 1998; Hughes et al., 2000; O'Connor et al., 2000; Jackson et al., 2000;
EPA, 2002). Such indexes consist of biological metrics, defined as attributes of the biological assemblage
that respond in predictable and measurable ways to human disturbance (Karr et al., 1986; Karr and Chu,
1999; Barbour et al., 1999). Metrics describe and measure a sampled population in terms of its taxonomic
composition, community structure, trophic structure, and presence of tolerant and intolerant taxa.
One of the primary outcomes expected from the MAIA study was a set of biological indicators for fish,
invertebrates, and periphyton that could be used to assess the biological condition of surface waters across
the region, either by EPA or by states within the region. Using the MAIA data, numerous studies have
examined the relationships among biological assemblages, human land use, and geography (Pan et al.,
1996; O'Connell et al., 1998; Kaufmann, et al.1999; Pan et al., 1999; Hill et al., 2000; Hughes et al.,
2000; Pan et al., 2000; Waite et al., 2000, Hill et al., 2001; McCormick et al., 2001; Fore, 2002b; Klemm,
et al., 2002; Blocksom, 2003). Overtime, the methods for selecting biological indicators have evolved
from individual approaches associated with different taxonomic assemblages and research groups to a
general approach, developed during a series of EPA-sponsored workshops, which could be applied more
or less uniformly across different assemblages and would yield similar results when applied by different
investigators. Through discussion and meetings, a consistent approach emerged for selecting biological
indicators. Results of this process have been documented for fish and invertebrates and will also be
applied to periphyton (Davis and Scott, 2000; McCormick et al., 2001; Klemm et al., 2003). The process
included identifying potentially meaningful metrics, selecting from among the hundreds of measures of
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
site condition and human disturbance for metric testing, choosing appropriate statistical tests for metric
analysis, and constructing a multimetric index.
Lessons learned from the Mid-Atlantic pilot study were related to sample design and data management as
well as data analysis. Random site selection is a fundamental feature of any EMAP project, yet has been
controversial as a sampling design at the state level. Other national monitoring programs and their
sampling designs were considered by the EMAP designers but failed to meet the needs of a regional
assessment program (Olsen et al., 1999). Alternative sampling designs cannot reliably provide unbiased
estimates of regional condition, the statistics of greatest concern for EPA. During the MAIA project, data
management and storage decisions both supported and hindered the scientific analysis at every step. The
best formats for different types of data were not obvious from the outset, and several versions of the
taxonomic data files were created, corrupted and corrected over a period of months or years before the
data were finalized and analysis could proceed.
Linking human disturbance to biological response generated concerns about circular reasoning, which
were addressed at multiple stages. The complexity of human land use and the difficulty involved in
quantifying site condition were recurrent issues. The enormous amount of data available for evaluating
site condition did not necessarily make quantifying human disturbance any easier. During the workshops,
much discussion centered on defining statistical criteria and logical decision rules for selecting metrics.
The consensus was that metrics should be correlated with several different types of disturbance measured
at multiple spatial scales. Finally, the development and application of the multimetric indexes revealed the
robust ability of these indicators to detect change through time.
Multimetric indexes for fish and invertebrates were finalized before this report was completed, but
analysis for the periphyton index is not yet complete (McCormick et al., 2001; Klemm et al., 2002).
Diatoms represent a subset of the periphyton assemblage. Where appropriate, results from a previous
analysis of diatom metrics are presented (Fore, 2002b).
This document presents issues, discussed in the MAIA workshops, that were controversial, confusing, or
time-consuming in the development of biological monitoring tools for fish, invertebrates, and periphyton.
It is aimed at agency scientists or managers tasked with implementing regional monitoring programs. The
hope is that the lessons learned in the Mid-Atlantic pilot study may serve as guideposts in the
development of indicators or management plans. The document's structure reflects the process of
developing the biological monitoring tools. Major headings represent general concepts, and subheadings
represent lessons learned from the Mid-Atlantic. Although the focus of the workshops and all the
scientific analysis that they inspired was the development of biological monitoring tools, the reality was
that the sampling design and data management issues defined the boundaries of what was possible at
every step of the analysis (see, for example, Ward et al., 1986). Consequently, I consider them first below.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
II. SAMPLING DESIGN
The only way to know the true condition of a regional resource is to continuously monitor every site. This
is impractical, and a subset of sites must be selected. A major strength of EMAP's monitoring strategy
was the conscious choice of a sampling design before data were collected (Ward et al., 1986). The
strengths and weaknesses of EMAP's probabilistic sampling design are debatable, but that controversy is
secondary to the fact that the program's goals along with the expected outcomes and products were
defined in advance, and the sampling design was chosen to support those goals (Stevens, 1994; Olsen et
al., 1999). In short, there was a plan from the beginning for summarizing and applying the sample data to
the larger region.
EMAP is a federal program with a national perspective and a mandate to estimate resource condition over
a large spatial area (Stevens, 1994). A probabilistic survey sampling design randomly selects sites from
all possible sites. An important advantage of random site selection is that results can be applied to similar
sites in the region that were not sampled. In addition, estimates of regional condition derived from a
probabilistic survey will have a known level of uncertainty associated with them. Without the error bars
around a summary statistic, there is no way to know whether an observed change represents simple
variability or a true change in resource condition. This type of information allows management actions to
be prioritized according to, for example, regions or resource types that show the greatest decline or
poorest status.
EMAP's pilot studies along with numerous regional EMAP, or REMAP, studies conducted at smaller
spatial scales were designed to demonstrate how a probabilistic survey design could support emerging
state monitoring programs that had previously lacked adequate assessment programs. Because EMAP was
envisioned as a template for states, it is somewhat ironic that a defining aspect of the program,
probabilistic sampling, has become controversial at the state level. State resource agencies are expected to
manage surface waters at the local level, not just assess the general condition of the resource (Yoder and
Rankin, 1998). Therefore, from a localized viewpoint, the coarse resolution of data associated with sites
selected at a regional scale may not provide information on particular areas of interest. Moreover, sites of
particular interest, such as permitted sites or sites with known problems, are unlikely to be selected
randomly. Thus, very degraded sites may avoid the TMDL process which is designed to rehabilitate those
sites by identifying their causes of degradation (NRC, 2001). From a state's point of view, then,
probabilistic sampling can seem like an additional monitoring requirement rather than a design to simplify
current monitoring.
Alternative sampling approaches select sites according to other specific (i.e., nonrandom) criteria. These
approaches may target sites associated with specific management activities or types of disturbance. Many
state programs adopt this approach over probabilistic sampling. Other states combine the two approaches
by using a probabilistic design to identify problem areas and a targeted design to focus within those areas
(EPA, 2003). A similar conflict between assessments at the regional scale vs. evaluation of a set of sites
with known properties arose for the MAIA study within the context of testing biological indicators.
Random sampling tends to miss sites with little or no human disturbance as well as sites with very high
levels of human disturbance; therefore, the concern arose that randomly selected sites would not provide a
broad enough range of conditions to test biological indicators. As a consequence, targeted sampling of
additional reference and impaired sites was used to supplement the MAIA probabilistic data. In the end,
this supplemental sampling was largely unnecessary for MAIA because many of the targeted reference
sites failed to meet the criteria for reference condition while the probabilistic survey included a sufficient
number of minimally disturbed and severely degraded sites for metric testing.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
A probabilistic sampling design was the best choice for MAIA
Given the size of the sampling area and the scope of the questions asked for the MAIA study, most
agreed that randomization of site selection was a necessity. The MAIA pilot was designed to
answer questions about the overall condition of surface waters within a large geographic area.
These types of questions are of greatest interest to EPA for planning programs and allocating funds
to support state and regional programs. Other sampling designs were not appropriate for an EMAP
project because they fail to yield an unbiased estimate of conditions across the entire region (Chart
1). Unless one samples every site (census sampling), selecting sites randomly is the only method for
inferring regional condition from a smaller set of sites (Overton and Stehman, 1996).
Chart 1: Types of sampling.
Census sampling: samples every site, no inference needed.
Probability sampling: randomly selects sites from the set of all possible sites
within a region.
Targeted sampling: selects sites based on some known aspect, for example sites
might be sampled downstream of a point source to evaluate the extent of its
influence.
Judgment sampling: similar to targeted sampling, sites selected according to
some scientific or observational criteria. (See example in Stoddard et al., 1998).
Convenience sampling: samples selected according to where sampling is
possible, e.g., due to private access issues. (See Olsen et al. [1999] for further
discussion).
Synoptic sampling: typically implies sampling within a short period of time,
perhaps to assess variables that are time-dependent, for example, chemical
concentration along a stream after a toxic release. Also implies breadth across a
large area or concurrent sampling of multiple measures.
Haphazard sampling: Selection of sampling units may be casually described as
"random"; however, if the selection process is not formally random, that is, based on
the use of random numbers table or randomization algorithm, the method of
selection is actually haphazard.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Census sampling, that is, visiting every site, was not possible on as large a scale as the MAIA study area.
This is true of most regional monitoring programs. Selecting sites according to scientific criteria or
professional judgment was not relevant because regional, rather than local, estimates of all indicators and
stressors were desired. Any design that limited the set of sample sites to easily accessible locations
(convenience sampling) or sites with known human impacts (targeted sampling) risked producing a
biased picture of current regional condition and were rejected (Stevens, 1994; Olsen et al., 1999).
From a statistical point of view, any sampling approach other than census or random sampling will be
technically "haphazard," that is, potentially biased and not representative of the larger unsampled
population. Haphazard or targeted sampling designs are not inherently bad or wrong. Their primary
drawback for monitoring is that the results from sampled sites cannot be extended to any unsampled sites.
Free information is cost-effective
Why sample randomly? The key point is free information. If site selection is random and all sites
are sampled with an equal or known probability, then information from the sampled sites can be
used to infer the condition of sites not sampled. Thus, results based on a random sample of sites can
be scaled up to the entire population of sites within a region, as long as each site in the region could
have been included in the sample. The knowledge obtained through random sampling about the
unsampled sites is essentially free information. The only other way to know something about every
site in a region would be to sample each one. Sampling all the sites is much more expensive than
randomly sampling a few.
If sites are selected according to any method other than randomization, then statistics and
conclusions derived for those sites can be safely applied only to those sites sampled. Generalizing
from a nonrandom sample to the larger unsampled population of sites typically yields biased
conclusions. The statistical truth of this fact may not be compelling; in fact, it seems somewhat
counterintuitive that a large enough sample size would fail to be representative of general
conditions in a region. Nonetheless, numerous large-scale studies based on real data have
demonstrated the inevitable bias in nonrandom sampling schemes, even when hundreds of sites
were sampled (Paulsen et al., 1998; Stoddard et al., 1998; Peterson et al., 1999).
In the controversy over how probabilistic sampling fails to meet state needs, the coarse scale of
sampling is often emphasized as inadequate at the state level (White and Merritt, 1998; Yoder and
Rankin, 1998). One relevant concept that is often missed in the discussion is that the scale of
probabilistic sampling can be altered to fit a state's more local assessment needs. Both EPA and
state programs ask the same types of questions, but they ask them at different scales. EPA wants to
know if surface waters in a state or region are getting better or worse. A state manager or citizen
group wants to know if a specific stream, or basin, is getting better or worse. Same question,
different scale. The EMAP concept of random sampling and free information could be applied at a
different spatial scale such as a basin or a watershed to infer the condition of other sites or reaches
not sampled within the basin or watershed. The EMAP method used to define and select sampling
units is flexible and was intentionally designed to be applicable at different spatial scales including
the conterminous U.S., geographic regions covering multiple states, or smaller regions defined
within state boundaries (Herlihy et al., 2000).
Our statistical ability to detect change in regional condition is also independent of the size of the
region when regional condition is summarized in terms of the percentage of sites meeting (or failing
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
to meet) specific criteria. For example, suppose in the first year of sampling that 50 sites are
randomly selected and sampled. Of these, 50% (25 sites) fail to support their designated uses and
are considered impaired. If 50 new sites are sampled the next year, a change of greater than 12% in
either direction represents a significant change (at 90% confidence) in stream condition within the
sample region (EPA, 2003). These types of summary statistics based on proportions along with
their estimates of precision would apply to a probabilistic survey of any 50 sites, whether sampled
within a county or within a state.
Reference sites did not always meet criteria for reference condition
Expectations for biological indicators such as multimetric indexes are based on observed conditions
at undisturbed or minimally disturbed locations. These reference sites represent a standard for what
the biological assemblage would look like in the absence of human influence (Hughes, 1995).
Given a probabilistic sampling design, the concern developed during the course of the MAIA pilot
that only moderately disturbed sites would be included in the data because a moderate level of
human influence was typical throughout the region. As a consequence, examples from the ends of
the spectrum, that is, sites with minimal human influence and extreme degradation, would be
missing from the data. Therefore, to supplement the EMAP design data, local biologists helped
select reference and impaired sites based on their experience and knowledge of the region, i.e.,
according to their best professional judgment (Gerritsen et al., 1994). The idea was to use these
"hand-picked" sites to test metric response to disturbance and then use the probabilistic sites to
estimate regional status and trends.
In the meantime, researchers developed a concept of reference condition and defined specific
criteria for the Mid-Atlantic (Hughes, 1995; Waite et al., 2000; Klemm et al., 2002). Ironically,
most of the hand-picked reference sites (44 out of 60, or 73%) failed to meet the independently-
established criteria for reference condition. The lesson learned was that best professional judgment
should always be confirmed with objective criteria.
The criteria for reference condition included acid-neutralizing capacity (ANC) > 50 |ieq/L, total P <
20 |ig/L, total N < 750 |ig/L, chloride (CL-) < 100 |ieq/L, SO4 < 400 |ieq/L, and mean RBP habitat
scores >15 (based on EPA's Rapid Bioassessment Protocol; Barbour et al., 1999). Reference sites
had to satisfy all criteria. Five of the six criteria were based on water chemistry and selected for
their close association with specific human activities in the watersheds. Chloride increased along
with development, total N and P were associated with agricultural intensity, low levels of ANC
indicated acid rain, and SO4 was related to mine drainage (Herlihy et al., 1990; Herlihy et al., 1993;
Herlihy et al., 1998). Therefore, extreme values of these chemical measures also indicated other
potential stressors and disturbances associated with these human activities.
The "hand-picked" reference sites selected according to best professional judgment included sites
with strong indications of intense human disturbance (Figure 1). For example, values for total N >
3000 |ieq /L are usually considered high and are typical of urban or agricultural land use. Chloride
values > 300 |ieq /L also represented some of the highest values in the entire data set and were
associated with relatively high levels of urban development. Similarly RBP values below 12
represented obvious signs of human influence nearby or within the sample reach.
Although the first years of random site selection may not have included a large enough sample of
reference sites, successive years of sampling did provide a broad enough range of human influence
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
to test and develop biological indicators. In fact, overtime, researchers moved away from the idea
of a simple comparison of reference vs. impaired sites and adopted a more sophisticated approach
that evaluated metric response across multiple gradients of human disturbance. Thus, the
probabilistic survey design yielded an adequate range of conditions for testing metrics.
Furthermore, objective definitions of site condition and impairment represented a more reliable
approach for testing and selecting biological indicators than did definitions of site condition based
on best professional judgment.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
I 6
z
4
2
0
BPJ
Criteria
I
100 250 500 750 1500 2000 3000
t
Total Nitrogen (|ig/L)
10 25 50 75 100 200 300
t
Chloride (ueq/L)
18
16
14
| 12
£ 10
o
I 8
I 6
4
2
1
20
16
12
nnH
12 16 20 24 28 32 36
t
Total Phosphorus (|ig/L)
10 12 15 17
Mean RBP score
Figure 1. Comparison of reference sites selected on the basis of "best professional judgment" (BPJ;
light bars) with sites selected according to reference condition criteria for total N, total P, chloride, and
mean RBP score (dark bars). Arrows indicate criteria defined for reference condition; higher values to
the right of the arrows indicate greater disturbance; for RBP scores higher values indicate better
condition. The broad range of values for BPJ sites indicate that highly disturbed sites were initially
included as reference sites.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
III. THE PERILS OF DATA MANAGEMENT
Every scientific analysis is founded on the format and reliability of its underlying data. Initial decisions
about how to store and format data will influence how or if data are ever used for scientific analysis. What
agency doesn't have data languishing on diskettes in file drawers? The lesson learned from the MAIA
pilot was that time spent on data management was never wasted and that the time required was usually
more than expected.
During the course of the MAIA study, several hundred sites were sampled for hundreds of variables over
six years. Some variables were sampled once per site (e.g., land cover and use); other variables were
sampled on every possible occasion (e.g., water chemistry); and still others were sampled on only some
site visits (e.g., fish and habitat). For invertebrates and algae, two samples were collected during many
site visits: one from pools and the other from riffles. Most sites were visited only once, but some sites
were sampled repeatedly, either during one year or in successive years, in order to estimate the variability
associated with site scale data. As a result, the data set was huge and complex, and data management was
an issue that was carefully considered from the beginning (Hale et al., 1998; Hale et al., 1999). However,
in spite of careful planning, data analysis was slowed by months due to mishandling of data files that
resulted in incomplete and corrupted data.
Many agencies struggle with data management choices made early in their programs that restrict their
current ability to analyze and interpret those data. Although many state water monitoring programs are
smaller in scale than MAIA, the amount of data collected increases each year and databases are much
larger than they were ten or even five years ago. Data management takes on a life of its own when more
than about 200 sites are sampled for more than 50 variables. Most state sampling programs now exceed
this amount of data for monitoring under the Clean Water Act. Data management lessons learned in the
Mid-Atlantic will become increasingly relevant to state programs as data accumulate.
Different "names" for the same site caused confusion
The MAIA data set was large and complex; therefore, it was not possible to put all the data in a
single file. Multiple files of related data meant that site names were extremely important for
matching data across files. Unfortunately, different types of data required different site information
to uniquely identify data from a single site visit. Thus, it was not possible to create a single unique
code for matching site visits across files that would correctly match all variables.
For land use data, only the site name was needed because satellite data were only recorded once for
each site during the course of the project. For fish and chemistry data, the site name, year, and visit
number within year were needed to identify data from a unique site visit. For diatom and
invertebrate data, the site name, year, visit number and habitat type (pool vs. riffle) were needed.
As a consequence, files were easily corrupted when smaller files were merged without using all the
variables necessary to match site visits correctly. An incorrectly merged file could have twice as
many entries for some variables or data from one site visit assigned to another. This issue will be
complicated for any monitoring program when the same site has different types of data associated
with it and the data are collected at different times. The lesson learned was that more information
should have been included with the data to identify unique sampling occasions.
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Original data must be archived
Biological data evolve from field sheets for fish or laboratory bench sheets for invertebrates and
algae with coded taxonomic identifiers; to files with taxa names, counts and natural history
information; to calculated metrics. With each new generation of the data file, the original files may
appear crude in retrospect and the tendency was to lose track of original files with confusing
formats when newer versions were created. For the MAIA pilot, original files were the only way to
catch major errors in later versions of the data. The lesson learned was to archive the original field
or bench sheets along with the first generation of electronic data in a way that the data will not be
changed or lost.
Different research groups tended naturally to follow different paths in evolving their data files.
When the different researchers came together, their results sometimes disagreed either due to errors
or different decision rules. Checks against the first generation of the data were often the only way to
resolve discrepancies between researchers. Errors crept into data sets, for example, when files were
merged incorrectly. Other problems arose when large files were truncated without warning by a
spreadsheet or statistical program because the number of variables exceeded the number of columns
permitted by the software.
During the course of the project, rules for combining invertebrate taxa evolved independently
among different researchers. Similarly for periphyton, rules for calculating the relative abundance
of soft algae and diatoms also evolved. In both cases, the first generation data files were required to
create a correct master file and to calculate metric values consistently. For invertebrates,
discrepancies can arise if one mayfly in a sample is identified to genus while another mayfly can
only be identified to the family that includes that same genus (either due to damage or being an
early instar). Should the two mayflies be counted as two distinct taxa or only one? If the family
typically has only one genus, then it's an easy decision, they should be counted as one taxon. But if
the family has many genera or the genus identified is known to be rare, then perhaps they should be
counted as two taxa. For the MAIA data, much time was spent defining consistent rules and
creating a master file that used the taxonomic rules to combine taxa. For periphyton, soft algae and
diatoms are identified with different laboratory methods; consequently, the taxa counts were
originally saved in separate files. In order to combine the taxonomic data for both types of algae,
the relative abundance of each type in the original sample had to be considered. Bench data related
to sample volumes was retrieved from the earliest data files to combine the information.
Without access to the original archived data, much of the data would have been irretrievably lost.
Although the archived data prevented this, a larger lesson learned from MAIA was that an
information manager should be an integral part of any scientific team for projects of this scale.
Furthermore, the hours dedicated in the beginning of the project to ensure data were correctly
stored were dwarfed by the hours spent after the fact trying to repair or retrieve corrupted data.
File structure mattered
Three key points contributed to the usefulness of the MAIA data and the successful analysis of the
data in so many publications. Simple files were shared via an Internet server, file formatting was
consistent across files, and metadata files were provided to describe the variable names. In addition,
the data were organized into files that lent themselves to statistical analysis. File structure logically
anticipated the type of analysis that was likely to be applied to a particular type of data.
10
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
When selecting the file structure for different data types, EMAP database designers distinguished
between "vertical" and "horizontal" files (Hale and Buffum, 2000). A horizontal file, has a single
row, or case, for each unique site visit and a single column for every variable recorded for that site
visit. In contrast, a vertical file lists different types of data in a single column and may use multiple
rows for a single site visit. Vertical files save space when many of the variables contain no
information for a particular site visit. Both are appropriate in different situations.
For the MAIA study, vertical files were used to list fish, invertebrate, diatom, and soft (non-diatom)
algal taxa collected at each site (Figure 2). The taxon code and name were listed along with the
number of individuals in that taxon. A horizontal file structure would have headed each column
with the name of one taxon and yielded a file with most of the cells blank, because not all taxa are
found at a particular site. For diatoms, the number of taxa averaged 29 per site out of approximately
950 possible. Most spreadsheet programs do not allow so many columns, so rather than split the
data across multiple files, a vertical file structure was used. Taxonomic information was also too
cumbersome to include in the data file and was instead recorded in a companion file of metadata
that listed phylum, class, order, family and other taxonomic details once for each species code.
11
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Data: List of taxa found at each site
SITE CODE
DE750S
DE750S
DE750S
DE750S
DE750S
DE750S
DE750S
MD003S
MD003S
MD003S
MD003S
MD003S
ETC.
VISIT
NO
1
1
1
1
1
1
1
1
1
1
1
1
YEAR
1994
1994
1994
1994
1994
1994
1994
1995
1995
1995
1995
1995
SAMPLE
TYPE
Pool
Pool
Pool
Pool
Pool
Pool
Pool
Riffle
Riffle
Riffle
Riffle
Riffle
TAXA
CODE
BAACBIO
BAACMNU
BAANVIT
BACBAMP
BAEUBIL
BASYRURU
BATAFLO
BAACMNGR
BACBAMP
EASY NAN
BASYULUL
BATAFLO
TAXON
Achnanthes bioreti
Achnanthes minutissima
Anomoeneis vitrea
Cymbella amphicephaia
Eunotia bilunaris
Synedra rumpens v rumpens
Tabellaria flocculosa
Achnanthes minutissima v gracillima
Cymbella amphicephaia
Synedra nana
Synedra ulna v ulna
Tabellaria flocculosa
COUNT
7
130
2
2
2
16
94
5
1
1
5
4
Metadata: Taxonomic nomenclature and authority for each taxon
TAXA CODE
BAAC
BAACBIA
BAACBIO
BAACBITH
BAACDAU
BAACDEAL
BAACDEF
BAACDEL
BAACDESE
BAACDIS
BAACEXA
BAACEXEL
BAACEXI
BAALPEL
BAAMOVA
BAAMPED
BAAMSUB
ETC.
PHYLUM
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
Bacillariophyta
GENUS
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Amphipleura
Amphora
Amphora
Amphora
SPECIES
spp.
Biasolettiana
bioreti
biasolettiana
daui
deflexa
deflexa
delicatula
delicatula
distincta
exigua
exigua
exilis
pellucida
oval is
peducilus
submontana
Vorf
v
v
spp
V
SUBSP
thienemannii
alpestris
septentrional
elliptica
AUTHORITY
(Kutzing) Grunow
Germain
(Hustedt)
Lange-Bertalot
Lowe & Kociolek
Reimer
(0estrup)
Lang-Bertalot
Messikommer
Grunow
Hustedt
Kutzing
(Kutzing) Kutzing
(Kutzing) Kutzing
(Kutzing) Grunow
Hustedt
Figure 2. Example data file for diatom samples from MAIA sites and companion metadata file with
taxonomic details keyed by taxa codes. The upper table is an example of a vertical file, with more than
one row per site visit and taxa names collapsed into a single column rather than one column per taxon.
12
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Once calculated, metrics were stored as horizontal files where a single row of data contained all the
information for each site visit (Figure 3). Each data file also had a companion file of metadata that
explained the meaning, units, and derivation of the codes naming each variable, represented by one
column of data, in the file. Other horizontal files were created that grouped sets of variables by
topics such as fish metrics, water chemistry, habitat measures, and stressors (Hale et al., 1998).
Although the metadata files were somewhat minimal, they were a key component to the usefulness
of the data across projects and research groups (Hale et al., 2000). Unfortunately, additional
important information regarding how data were collected was not archived with the data. This
caused some confusion at the time of analysis because sampling protocols changed through time. In
short, all the attention paid to file structure and data organization was well worth the effort; in fact,
more metadata and data description would have been useful.
Simple files were best
Data analysis for MAIA involved multiple institutions and investigators using different statistical
software. Posting files on an Internet server was the most practical approach to sharing files
between so many remote users. Hosting a searchable relational database that included all the data
was an option, but these are typically slow and difficult for the host to maintain (Hale et al., 2000).
Because researchers were typically interested in a subset of data, smaller, simpler files with
variables grouped according to topic worked best.
Relational database programs (e.g., Access and Oracle) provide the opportunity to develop
complicated data structures composed of multiple linked files. However, this approach assumes that
users are working in a similar software environment, that is, using the same programs to analyze
data or working from the same central database. Relational databases typically link multiple vertical
files, thereby saving space by not repeating information that is identical for each entry. Although
relational databases avoid some repetition within files, the unseen program code required to support
the relationships between files typically takes up more space than a simple flat file with redundant
information. Another advantage of a relational database is that data can be stored in one place and
relationships between variables and files are encoded in the program so that individual users are not
required to remember or derive these details; however, these relationships can be very complicated
to code correctly and to maintain.
The MAIA data had to be accessible to many remote users with the intention of manipulating the
data within a variety of software environments. Most statistical software packages expect flat files
and cannot import the program information that a relational database uses to link variables across
files. Flat files have all the information for a site visit in the same place, that is, in one file on the
same row (Hale and Buffum, 2000). Thus, rather than create a complicated database structure from
which data would have to be exported for analysis, data files were kept simple from the beginning
so that they could be easily downloaded from the EMAP Internet site and quickly entered into the
user's own statistical software.
13
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Data: Example of a horizontal data file structure
STRM ID
DE750S
MD003S
MD005S
MD006S
MD008S
MD507S
MD508S
MD510S
MD511S
MD512S
MD513S
MD750S
MD751S
MD752S
MD753S
MD754S
MD755S
MD756S
MD757S
NY001S
NY002S
ETC.
VISIT
#
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
DATE
COL
5/17/94
5/15/95
5/15/95
5/22/95
5/23/95
5/25/93
5/27/93
6/1/93
6/10/93
6/9/93
6/7/93
5/16/94
5/19/94
5/31/94
5/23/94
5/18/94
5/24/94
5/23/94
5/18/94
6/22/95
6/22/95
TEAM
ID
3
3
3
3
3
1
1
1
1
1
1
3
3
1
3
3
5
5
6
2
2
PHSTVL
6.17
6.91
6.66
7.42
6.39
6.82
7.14
7.08
7.11
7.06
8.44
7.12
7.18
6.72
7.55
6.66
7.15
7.93
7.24
7.4
7.6
PHEQ
7.38
7.57
7.15
7.9
6.92
7.48
7.75
7.47
7.73
7.82
8.66
8.3
7.93
7.77
8.49
7.45
7.68
8.23
8.4
7.83
8.03
ANC
115
207.2
82.2
435.2
37.4
154
317
153
286
391
3550
1090
412
295
1670
145
246
895
1450
385.5
576
COND
102.5
90.6
42.4
99
31.9
53.2
68.8
62.2
121.8
90.8
441
298.5
171.2
142.3
251.5
62.7
78
142.3
279.3
115
129
COLOR
8
13
30
13
6
3
4
7
7
6
5
17
8
2
11
5
8
8
17
5
9
CA
303.4
528.9
181.6
391.7
124.3
272
340.8
237.5
367.3
409.2
3263.5
1462.1
588.8
673.7
1556.9
183.6
327.3
728.5
1117.8
408.2
623.8
MG
213.1
147.2
80.6
231.2
73.2
107.8
176
179.3
261.6
283
1299.7
804.5
418.7
380.9
479.6
167
154.6
394.8
519.1
168.6
193.3
NA
266.2
84
82.2
263.6
37.4
53.1
85.3
69.6
394.5
102.2
97.4
418.5
425.9
124
391.5
151.4
160.1
226.6
648.2
416.3
305.4
K
67
20.2
15.6
27.4
27.6
20.7
34
35.5
47.8
48.1
43
67.8
54.5
50.9
35
26.3
23.5
27.9
82.6
22.2
39.4
NH4
2.8
0
0
0
0
2.8
0.6
0
0
0
0
2.2
0
0
1.1
0
0
0
232.1
1.1
1.1
CL
229
44
85
310
30
79
36
102
429
54
204
696
663
32
515
166
166
242
327
469
326
NO3
303
30.3
9.7
72.3
1.6
44.3
50.9
37.4
16.5
7
575
375
248
7
61.4
7.2
28.4
60.3
15.1
4.7
79.7
SO4
132
443
160
77
180
200
243
234
323
376
418
401
135
851
172
224
228
168
862
156
191
ALTD
54
36
47
12
19
10
12
5
8
6
6
12
30
1
20
20
12
19
30
13
15
Metadata: Description of column headings
Variable Name
STRMJD
VISiTJJO
DATE_COL
TEAMJD
TRANSECT
PHSTVL
PHEQ
ANC
COND
COLOR
CA
MG
Description
ID
Visit Number
MMDDYY
Sampling crew
Transect ID
Closed System pH
Air-equilibrated pH
Gran Acid Neutralizing Capacity (ueq/L)
Specific Conductance (uS/cm)
f PCU)
Calcium (peq/L)
Magnesium (peq/L)
Variable Name
NA
K
NH4
CL
NO3
SO4
ALTD
ALPS
ALOR
DIC
ere.
Description
Sodium (jieq/L)
Potassium (jieq/L)
Ammonium (y,eq/L)
(peq/L)
Nitrate (ueq/L)
Sulfate (jieq/L)
Total Dissolved aluminum (u\g/L)
PCV reactive (monomeric) aluminum (ug/L)
Nonexch. PCV (organic) aluminum (|,ig/L)
Dissolved Inorganic Carbon (mg/L)
Figure 3. Example of a horizontal file structure for water chemistry data. Stream ID, visit number and
date identify a unique sampling event represented by a single row in the data file. A companion
metadata file lists each variable name, its description, and units of measure.
14
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
IV. LINKING HUMAN DISTURBANCE TO BIOLOGICAL CHANGE
Like any scientific inquiry, understanding how human disturbance influences, alters, and degrades
biological processes is an iterative process. Hypotheses are proposed on the basis of current
understanding; theories are modified in the wake of results; and new insights drive subsequent hypothesis
testing. As knowledge accrues, connections between predictions and results tighten, and new studies
present fewer surprises. A paradigm forms.
The development of biological monitoring tools has followed this type of scientific process. After 20
years of testing, consistent patterns have emerged (see literature review in Barbour et al., 1999; Karr and
Chu, 1999; Karr et al., 2000). The same biological measures tend to correlate with human disturbance in
very different geographic settings, e.g., the number of taxa decline, tolerant taxa dominate, and taxa with
unique habitat requirements disappear. Yet when results closely match expectations, concerns arise
regarding "circular reasoning." An argument based on circular reasoning is one in which the conclusion is
embedded in the premise, as for example, in the statement, "decline in mayfly taxa richness is a good
indicator of biological disturbance because we find many types of mayflies at undisturbed places."
Without a direct causal link, the concern is that metrics may be selected as indicators of human
disturbance simply because they are correlated with human disturbance (Suter, 2001). That is, biological
indicators may be unrelated to specific biological or societal values.
Because biological systems are complex and human disturbance is multidimensional, single causes and
mechanisms of impairment are difficult to isolate; as a result, much of the evidence for human
degradation of natural resources is correlative. In such situations, although the path to causality is blocked
by the inability to perform controlled experiments and use statistical inference, logical argument (or
weight of evidence) constructed according to a recognized set of rules can be used instead (Beyers, 1998).
In fact, this approach typically yields a stronger case because researchers consider alternative
explanations explicitly, rather than assuming they do not operate. Results from the Mid-Atlantic illustrate
how a causal argument can be constructed to support the idea that human disturbance causes biological
change.
Addressing concerns about circular reasoning
The data for MAIA were not collected with the intention of demonstrating a causal link between
human disturbance and biological degradation; however, the expectation of a cause and effect
relationship is implicit in EMAP's survey design and project goals. The process used to test and
select metrics did, however, support the type of structured logical argument reviewed by Beyers
(1998) for establishing a causal connection between human influence and biological change. In fact,
results from the MAIA pilot supported six of Beyers' ten criteria (Chart 2).
Researchers in the field of epidemiology face a similar challenge in defining causality when linking
a specific disease with its infectious agent. Rules developed within that field can be applied in an
environmental context because the situations are parallel (Beyers, 1998). In epidemiology, patients
develop a disease; an investigator cannot randomly infect patients with a variety of infectious
agents to see which one causes the disease. Analogously for environmental studies, human
development cannot be "applied" at will to see how a place will respond. Furthermore, treatments
cannot be replicated in either case. Each patient is unique in terms of medical history and life style
as is each watershed unique in terms of its geological formation, hydrological structure, size, and
climate (Hurlbert, 1984; Heffner et al., 1996).
15
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Chart 2. Ten criteria for constructing causal arguments (modified after Beyers, 1998).
1. Strength: a large proportion of sampling units are affected in exposed
areas compared with reference areas
2. Consistency: the association has been observed at other times and places
3. Specificity: the effect is diagnostic of exposure
4. Temporality: exposure must precede the effect in time
5. Dose-response: the intensity of the observed effect is related to the
intensity of the exposure
6. Plausibility: a plausible mechanism links cause and effect
7. Evidence: a valid experiment provides strong evidence of causation
8. Analogy: similar stressors cause similar effects
9. Coherence: the causal hypothesis does not conflict with current knowledge
10. Exposure: indicators of exposure must be found in affected organisms
Six of Beyers' ten criteria were relevant for constructing a logical argument for the causal
connection between human disturbance and biological decline in Mid-Atlantic streams.
First, the association between proposed cause and effect was strong. The majority of sites with
human disturbance in the watershed had lower values for biological indexes than did the reference,
or minimally disturbed, sites (Figure 4). In addition, indexes were significantly correlated with
independently derived measures of human disturbance (Table 1).
Second, the observed association with biological metrics and indexes was consistent with the
results observed by other scientists in similar situations. Klemm et al. (2002) found that many of
the same invertebrate metrics associated with disturbance at the regional level had also been
selected for their response to disturbance by state programs at a more local level. McCormick et al.
(2001) related fish metrics selected for Mid-Atlantic streams to functionally similar metrics
selected in other regions.
Third, because evidence of human disturbance tends to persist, it was reasonable to conclude that
exposure to disturbance preceded biological change.
Fourth, graphed dose-response relationships illustrated that the biota changed in proportion to the
intensity of disturbance.
Fifth, the hypothesis that human disturbance causes biological degradation does not conflict with
existing knowledge or experimental evidence (see examples in Hudson and Cibrowski, 1996;
Wallace et al., 1996; Lemly, 2000; Richardson and Kiffney, 2000; Mebane, 2002).
Sixth, tissue analysis found chemicals associated with human development, e.g., DDT and
mercury, in fish. These contaminants were also correlated with an increase in the number offish
species tolerant of chemical pollution.
16
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
100
80
60
40
20
0
o
8
IILI
lm
i] Fish
D Invertebrate
T1 Diatom
Figure 4. Multimetric index values for fish, invertebrates, and diatoms as a function of types of
human disturbance. All index values were higher at reference sites. Diatom index values were
lower for sites with high nutrients. Diatom index values were higher than invertebrate index values
for sites with acid deposition.
The remaining four criteria (numbers 3,6,7 and 8 in Chart 2) require that an observed effect (metric
value) be diagnostic of exposure (human disturbance); a plausible mechanism of action exists to link
cause and effect; controlled experiments support causation; and analogous responses are associated with
similar stressors. A broader survey of the literature could add further examples under these headings to
the overall argument for causality. For example, Yoder and DeShon (2002) used metrics to diagnose
disturbance type, and Richardson and Kiffney (2000) present experimental evidence for the effect of
heavy metals on invertebrates. Nonetheless, the examples above serve to illustrate how the logical
construction of an argument for causality represents an alternative to doubts regarding circular reasoning.
17
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Table 1. Spearman's correlation of three multimetric indexes (based on fish, invertebrates, and diatoms)
with selected measures of human disturbance. All correlation coefficients were significant; only values
> 0.3 (or < -0.3) are shown. Measures of human disturbance were related to (1) nutrients (total nitrogen
[N], total phosphorus [P], ammonia [NH4]); (2) acidity (acid neutralizing capacity [ANC] and sulfate [SO4]);
(3) sediment (turbidity [Turb], percentage of sand and fine sediments [%S_F], and pebble size corrected
for stream power [PbSz]); and (4) channel and riparian condition (riparian vegetation [RVeg], sum of all
disturbance types within the riparian area weighted by proximity to the stream [RDist], and average of
measures from a rapid habitat protocol [RBP]). Measures of general disturbance included chloride (CL);
Bryce et al.'s [1999] disturbance categories derived from watershed and riparian measures of disturbance
(Bryce); and the sum of urban, agricultural, and mining land use within the watershed (%Dist).
Measure
N
P
NH4
ANC
SO4
Turb
%S_F
PbSz
RVeg
RDist
RBP
CL
Bryce
%Dist
Total
Fish
-0.45
-0.33
-0.34
-0.45
-0.39
-0.33
6
Invertebrate
-0.32
-0.36
-0.33
-0.33
-0.54
0.43
-0.35
0.42
-0.31
-0.57
-0.40
11
Diatom
-0.54
-0.61
-0.32
-0.53
-0.39
-0.39
0.33
0.30
0.36
-0.55
-0.54
-0.54
12
Metric testing included safeguards against circular reasoning
One risk associated with using correlative data to demonstrate connections between human
disturbance and biological degradation is that the observed correlation may be due to spurious
correlation with another underlying cause, such as elevation or watershed size that drives both
biology and patterns of human settlement. Or, the observed correlation may be due to some
additional factor that was not considered. When developing biomonitoring tools, the goal is to
select biological indicators that vary only as a function of human disturbance and are immune to
variability associated with natural physical or geographic features. Unfortunately, humans are biota,
too, and their land use patterns tend to follow landscape features, thus confounding human activities
with physical features. The challenge for the MAIA project was to demonstrate that human
disturbance was the most likely agent of biological change.
18
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
A variety of safeguards helped reduce the probability of drawing unsubstantiated conclusions from
the MAIA data analysis. Five approaches were used to isolate the relationship between human
disturbance and biological change from other confounding influences.
First, site selection was randomized across a large geographic area to ensure that the sample was
representative of the entire population of possible sites. Unbiased selection of sites provides some
protection against confusing the effects of human disturbance with other, natural features
(Stewart-Oaten et al., 1986; 1992).
Second, measures of disturbance were selected independently of the biological metrics. Bryce et
al. (1999) present an integrated definition of human disturbance for the Mid-Atlantic region
derived without consideration of biological indicators.
Third, metrics were tested in multiple years, or part of the data set was reserved to test the final
indexes as an independent test (McCormick et al., 2001; Fore, 2002b; Klemm et al., 2003). In this
way the observed relationships were demonstrated to be consistent across years.
Fourth, all metrics were tested for correlation with multiple gradients of human disturbance rather
than for their simple ability to distinguish between one set of minimally disturbed, or reference,
sites and severely degraded, or impaired, sites. Thus, the metrics selected were consistently
associated with different aspects of human disturbance.
Fifth, potential confounding factors such as watershed area or elevation that could underlie
patterns of both human influence and biological condition were explicitly tested. Where necessary
these effects were removed, for example, when fish taxa richness metrics correlated with
watershed area. In this way, metrics were selected for their association with disturbance rather
than other natural features.
Patterns of human disturbance were complex
The MAIA pilot had the luxury (and the curse!) of data for practically any variable that has ever
been recorded to evaluate water resources. Dozens of variables related to water chemistry, metals,
nutrients, fish tissue contaminants, habitat, channel morphology, geographic features, human census
data, satellite land cover and use, and specific point sources were included in the data set. Hundreds
more were derived from the data collected. The hope was that such a complete record of human
activity would provide a clear picture of human influence and disturbance within a watershed. The
reality was that the different measures of disturbance tended to tell their own story; that is, different
measures were associated with specific types of human activity. Therefore, disturbance measures
were not necessarily correlated with each other because not all activities were present in every
watershed. As a consequence, one of the primary challenges for the MAIA pilot study was to
determine which variables most accurately characterized human influence.
During the process of developing biological indicators, much discussion surrounded the choice of
appropriate measures of site condition for metric testing. A primary lesson learned in the Mid-
Atlantic was that no simple method existed to quantify human influence in such a complex
landscape with such a long and varied history of human activity (Herlihy et al., 1998; Bryce et al.,
1999, McCormick et al., 2001). A missing piece from this project was a comprehensive study
linking the types of human activities (e.g., mining or agriculture) with their specific stressors (e.g.,
SO4 or nutrients). Given the tendency to find multiple types of disturbance in each watershed, a
19
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
clear picture may not have been possible. Nonetheless, a better understanding of which of the many
measures of disturbance tended to vary together along with a better understanding of which
measures were related to natural geographic or landscape features would have helped clarify metric
response to disturbance. For example, diatom metrics were correlated with elevation but so was
disturbance because towns and farms tended to be found at lower elevations (Fore, 2002b).
Examination of a correlation matrix of all the variables related to site condition (too large to show
here) revealed patterns of correlation among related variables. Sets of variables could be identified
that seemed to measure similar or underlying processes. Variables related to water chemistry,
nutrients, and water quality tended to be significantly correlated with each other. Other sets of
variables related to channel structure, fish cover, or the condition and extent of riparian (streamside)
vegetation showed similar patterns of higher correlation among related sets of variables. In contrast,
some groups of variables showed little correlation across groupsfor example, measures of
riparian vegetation did not tend to correlate with measures of water chemistry.
Integrated measures of disturbance were better predictors of index values
In general, specific stressors tended to be more highly correlated with integrative measures of
human disturbance than they were with similar measures that measured only a single aspect of
disturbance. For example, turbidity, percentage of sand and fine, pebble size, riparian vegetation
condition and riparian disturbance were correlated with one or two of each other, but all five were
correlated with a habitat index developed to integrate measures of site condition at the reach scale
(Table 2). A similar pattern was observed for water chemistry measures such as total N, total P,
NH4, and SO4 that showed fewer significant correlations with each other than with integrative
measures that summarized human influence at the watershed scale. Bryce et al.'s (1999) condition
classes were derived from an analysis of patterns of human land use within the watershed and the
observed riparian condition at 102 sites. All the listed measures correlated with Bryce et al.'s index
of disturbance. The percentage of disturbed land was the sum of land in the upstream watershed
used for agriculture, urbanization or mining; all but one of the uni-dimensional measures correlated
with this measure.
Similarly for biological indicators, multimetric indexes for all three assemblages showed a higher
correlation with integrative measures of disturbance than with specific stressors (see Table 1 on
page 18). One chemical measure, chloride, was a strong indicator of general disturbance and also
highly correlated with all three biological indexes (Herlihy et al., 1998).
Thus, measures of disturbance that integrated measures of site condition over multiple spatial scales
tended to better capture the cumulative effects of human influence. This result supports the idea that
much of the scatter observed in plots of biological measures against human disturbance gradients is
in fact associated with the x-axis: one-dimensional measures of disturbance simply fail to capture
the cumulative influence of human activities on the biota (Karr et al., 2000).
20
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Table 2. Spearman's correlation matrix for measures of human disturbance that were used to test
biological metrics for the MAIA study; only correlation coefficients > 0.3 (or < -0.3) are shown. (See Table
1 for description of variables.)
Measure
N
P
NH4
ANC
SO4
Turb
%S_F
PbSz
RVeg
RDist
RBP
CL
Bryce
%Dist
Total
N
0.50
0.34
0.39
0.51
0.38
0.67
6
P
0.50
0.36
0.31
0.57
0.44
-0.34
0.35
0.30
0.50
9
NH4
0.34
0.36
0.41
0.36
0.32
5
ANC
0.39
0.31
0.35
0.33
0.45
0.47
0.54
7
SO4
0.35
0.45
0.39
3
Turb
0.57
0.47
-0.38
-0.32
0.32
0.33
6
%S_F
0.44
0.47
-0.75
-0.39
0.55
0.49
6
PbSz
-0.34
-0.38
-0.75
0.31
-0.36
-0.34
6
RVeg
-0.65
0.60
-0.60
3
RDist
0.33
-0.65
-0.35
0.45
0.30
5
RBP
-0.32
-0.39
0.31
0.60
-0.35
-0.77
-0.41
7
CL
0.51
0.35
0.41
0.45
0.45
0.58
0.68
7
Bryce
0.38
0.30
0.36
0.47
0.39
0.32
0.55
-0.36
-0.60
0.45
-0.77
0.58
0.47
13
%Dist
0.67
0.50
0.32
0.54
0.33
0.49
-0.34
0.30
-0.41
0.68
0.47
11
21
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
V. METRIC TESTING
Candidate metrics are selected for inclusion in a multimetric index if they are biologically meaningful,
consistently associated with human disturbance, not redundant with other metrics, and reliably and easily
quantified from field samples (Karr and Chu, 1999; Jackson et al., 2000). Typically the approach selected
for metric testing is limited by the type and variety of data available for quantifying human disturbance.
For the MAIA study there were virtually no limitations because of the variety of variables measured.
For any study relating biological change to the degradation associated with human activities, the most
difficult step is identifying the independent measure of human disturbance. Independent measures of
disturbance are needed to test for a consistent biological response across the range of possible conditions.
Measures of human disturbance must be independently derived from biological data to avoid simply
choosing aspects of disturbance or measures of biology that match our expectations. Because human
influence is complex and human activities are multidimensional, the challenge revolves around how to
integrate disparate measures of human influence into a single axis of human disturbance for metric
testing.
Statistically, it is simplest to compare measures taken at impaired sites with the same measures taken at
minimally disturbed reference sites (Barbour et al., 1996). When only a few measurements of disturbance
are made, statistical testing may involve simple tests of differences in means (ANOVA) or association
between one-dimensional measures of disturbance (regression or correlation; Miltner and Rankin, 1998).
When multiple, related measures of disturbance are made, multiple regression may be used test the
disturbance measures together. If, however, these measures are themselves correlated with each other
(which they often are), multiple regression's assumption of independence is violated, and the results may
not be robust (Loftis et al., 1991; Wang et al., 1998; Olden and Jackson, 2000). When many measures of
disturbance are collected, measures may be combined using principal components analysis (Hughes et al.,
1998; Norton et al., 2000). Other projects have used a ranking system to summarize information about
human disturbance at different spatial scales (Bryce et al., 1999; Fore and Grafe, 2002).
For the MAIA pilot, the wealth of measures associated with human influence necessitated a larger
discussion among the researchers involved. The disturbance measures used for metric testing were
selected during a series of workshops sponsored by the EPA. Through discussion and consensus,
researchers derived a list that captured multiple aspects of human influence at different spatial scales
(McCormick et al., 2001; Klemm et al., 2003). Reference and test conditions were defined to include a
subset of least disturbed and most degraded sites (Waite et al., 2000). Measures of nutrient concentration
were included to reflect the influence of agriculture and urban land uses; acidity was included to capture
the effects of acid rain deposition and acid mine drainage; variables related to sediment and turbidity
measured erosion; and measures of riparian condition summarized the physical disturbance near the site.
In addition, Bryce et al. (1999) developed an integrative measure of human disturbance at the watershed
and reach scale. The percentage of disturbed land in the watershed was also used as a summary measure.
Redundant testing of metrics against multiple measures of disturbance ensured that metrics were selected
for their biological meaning rather than statistical chance. With a large list of candidate metrics and a
single test for each, candidates may meet the criteria for metric selection because of chance alone. The
probability of such chance selection equals the p-value selected for the test. Multiple tests against
measures of different aspects of human disturbance avoided this pitfall and insured that the metrics
selected represented meaningful and reliable indicators of biological change associated with human
influence.
23
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Simple criteria were used first to eliminate potential metrics
For each assemblage, a large number of candidate metrics were identified for testing (Stevenson
and Bahls, 1999; McCormick et al., 2001; Klemm et al., 2003). Simple statistical rules were
developed to shorten the long list of candidate metrics to a smaller number that were then
considered more carefully.
During the first round of elimination, metrics were evaluated for their range of values. Metrics
calculated on the basis of a small number of species might not yield an adequate range of values for
calculating taxa richness or percentage relative abundance. For example, taxa richness metrics that
could take on values of only 0 or 1 or percentage metrics that only had a range of 10% had too few
potential values to distinguish between different levels of human disturbance. These candidate
metrics were eliminated in favor of metrics with a broader range of values. For Mid-Atlantic
streams, candidate metrics such as number of native cottid species and percentages ofCorbicula
were eliminated because their ranges of values were simply too small.
Statistical precision was no substitute for correlation with disturbance
If the variability of a candidate metric within individual sites is higher than its variability between
all sites, then the measure is unlikely to detect differences in biological condition among sites (or
differences at sites that change through time). Signal-to-noise ratios estimate a measure's ability to
distinguish differences among sites from differences observed within individual sites. In this
context, "signal" is defined as variability of a metric value across all sites and "noise" as variability
over repeated visits to the same site during a single year (Kaufmann et al., 1999).
Although most metrics incorporated into multimetric indexes have high signal-to-noise ratios, i.e.,
small within-site variability compared to large between-site variability, a high signal-to-noise ratio
alone does not guarantee that a candidate metric will be a meaningful indicator of biological
condition. Metric values can be highly repeatable at individual sites but still be unrelated to human
disturbance. Consider, for example, pool depth and embeddedness, two candidate metrics described
by Kaufmann et al. (1999) in their assessment of habitat measures for MAIA. Pool depth is often
considered an indicator of good fish habitat. It is expected to decline as erosion, dredging, and
sedimentation fill pools, creating a homogeneous channel profile. Embeddedness was defined as an
average of several substrate measures at the stream reach scale; it represented the proportion of the
reach filled with sand and fine sediments.
Certainly statistical precision is a desirable property of a good metric, but statistical precision alone
does not guarantee a predictable association with human disturbance. For the MAIA study, pool
depth was very precise, with signal-to-noise ratio equal to 16. In contrast, embeddedness was more
variable for repeat site visits with a signal-to-noise ratio of 1.9, which failed to meet the authors'
suggested minimum value of 2. Embeddedness, however, showed a strong correlation with human
disturbance. In contrast, pool depth was precise but not related to human disturbance.
Embeddedness, though less statistically precise, was the better indicator of biological condition
(Figure 5). Thus, statistical precision alone was not a good criterion for metric selection.
The most important signal for biomonitoring is a metric's response to human disturbance. The term
"signal" in this case may, in fact, be misleading. The "signal" part of the signal-to-noise ratio is
24
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
actually just a measure of the observed range of values across all sites. In terms of biological
assessment, the actual signal we are interested in detecting is the change in biological condition
associated with human disturbance. In short, evaluation of the statistical properties of metrics
should never substitute for actual metric testing against disturbance. These are two separate, though
complementary, tests.
30
25
20
o
o
51 15
10
5
0
-R
| I Min-Max |
I
II IJ
-L
n1 T- I*.
112233334455555
100
80
Q 60
LU
m
S
X 40
20
I
I
I-,-
11 22 3333 44 55555
Figure 5. Ranges of values for mean residual pool depth (RP100, upper panel) and embeddedness
(XEMBED, lower panel) for 15 sample sites sorted along the x-axis by disturbance class from least (1) to
most (5) disturbed (Bryce et al., 1999). Vertical lines span the range of values recorded for two to six
repeat visits to each site. Repeat visits to the same site yielded more similar values for RP100 than for
embeddedness indicating greater precision (shorter vertical lines); however, embeddedness consistently
increased with greater human disturbance while RP100 did not.
25
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Watershed features were confounded with metric response to disturbance
A good monitoring tool must correlate with human disturbance and show little or no association
with natural features. When human disturbance itself is associated with natural features, isolating
the biological change associated with human disturbance can be tricky.
Watershed area was an example of this situation. Some fish taxa richness metrics tended to be
correlated with watershed area because larger streams have more different types of habitat that
support more species. To address this problem, metrics that correlated with watershed area were
first regressed against watershed area using only reference sites. The residual values from this
regression, that is, the metric values with the influence of watershed area removed statistically,
were used to define a new version of the metric that was independent of watershed size. When the
"corrected" metrics were significantly correlated with human disturbance, concerns about an
underlying spurious correlation with watershed size were eliminated and the metrics were retained
as good indicators of human influence. Watershed size was most important for fish, somewhat
important for invertebrates, and did not influence diatom assemblages. The different sensitivity
across assemblages to watershed area may reflect the relative range size of the organisms. Fish may
travel throughout the watershed, invertebrates tend to stay within a reach or local stream, and
diatoms may pass their lives on the same rock.
Metrics from different assemblage types were eliminated for different
reasons
The list of plausible metrics proposed for testing in Mid-Atlantic streams was much shorter for fish
(58), a bit longer for invertebrates (120) and much longer for periphyton (240). The initial list was
shortest for fish because the greatest amount of metric testing has occurred for fish; invertebrates
place a close second and periphyton a distant third. For periphyton, most of the candidate metrics
represented untested hypotheses. Metrics were selected and eliminated for different reasons across
assemblages.
The majority of candidate fish metrics were eliminated because they failed to correlate with
disturbance (30 metrics; Table 3). In contrast, a larger percentage of candidate invertebrate metrics
were consistently associated with disturbance. For the invertebrate index, many metrics that were
significantly correlated with multiple measures of disturbance and that demonstrated good
statistical properties were excluded because their correlation with one another exceeded 0.7.
Approximately 25 metrics were eliminated for this reason, leaving relatively few (7) in the final
index (Klemm et al., 2003). Because index precision increases as a function of the number of
metrics, some caution may be warranted in eliminating good biological signal on the basis of
statistical correlation (Fore, unpublished data). Results are not complete for periphyton, but based
on preliminary results for a subset of diatom metrics, there will likely be many metrics to choose
from that are significantly correlated with disturbance (Fore, 2002b). Additional criteria related to
the type of environmental processes measured by the metrics and metric redundancy will probably
be used to select metrics for inclusion in the final multimetric index.
26
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Table 3. Numbers of candidate metrics tested for MAIA's fish and invertebrate multimetric indexes and a
summary of the reasons for which they were eliminated. Starting with the total number of candidate
metrics, the number of candidates listed in each column were eliminated because their values spanned
an insufficient range, their signal-to-noise ratios were low (indicating low precision), they were redundant
with other metrics, they failed to correlate with human disturbance, or their correlation with watershed size
could not be corrected. This winnowing process resulted in fewer than 10 metrics included in the final
indexes.
Total number of candidate metrics
Insufficient range
Poor signal/noise
Redundant
Fail to correlate with disturbance
Persistent correlation with watershed area
Number of metrics in final index
Fish
58
13
2
3
30
1
9
Invertebrates
120
20
66
25
2
0
7
27
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
VI. DEVELOPMENT AND APPLICATION OF MULTIMETRIC INDEXES
The Clean Water Act requires all states to define water quality standards for their surface waters (Ransel,
1995). The standards include designated uses, criteria to protect those uses, and a prohibition against
degradation of existing uses. States must assess the condition of water bodies and determine whether they
support or fail to support their designated uses. Water bodies that fail are then listed as impaired (NRC,
2001). Currently, most states have narrative biological criteria in place, but are mandated to develop
numeric criteria for biological condition (Karr, 1991; EPA, 2002).
Multimetric indexes were created to fill this role as numeric assessment tools and are used by most states
for this purpose (EPA, 2002). The relevance and defensibility of the decision to list a site as impaired
depends on both the biological meaning of index values and the statistical precision of the index. Decline
in biological condition is a continuous process, and drawing a line of impairment, or defining biological
criteria, at any single point along its range will inevitably be somewhat arbitrary. The most common
approach for drawing this line defines impairment in terms of deviation from values observed for
reference sites (Hughes, 1995; Southerland and Roth, unpublished data). A second approach uses
statistical power analysis to calculate the change in biological condition an index can detect for a given
statistical model and defines impairment in terms of that measurable change (Fore, 2002a). Other authors
have recommended that better definitions of beneficial use should drive this process and that definitions
of impairment should be tied to societal values, such as the support of salmonid spawning (NRC, 2001).
Biological criteria depend on the definition of reference sites
Many states currently define biological impairment in terms of multimetric index values observed
at reference, or minimally disturbed, sites. Sites of unknown condition are then sampled and judged
against this standard. Defining a set of reference sites may be arbitrary, either in terms of which
sites are included or which criteria are used to define reference condition. For the MAIA study, we
have seen that hand-picked reference sites did not necessarily match objective criteria for reference
condition. This lesson learned in the Mid-Atlantic is relevant to states as they develop criteria for
reference condition and rules for defining impairment. Currently, states vary both in the way they
characterize reference condition and the way they define deviation from reference condition.
Furthermore, the line of impairment ranges from the index value that corresponds to the 5th to the
95th percentile of reference sites (Southerland and Roth, unpublished data.}.
For the MAIA fish index, three methods were used to define reference sites that represented
increasingly stringent criteria for reference condition in that more conditions had to be satisfied
(McCormick et al., 2001). The least restrictive criteria were based on chemical criteria and RBP
habitat measures; the moderately restrictive criteria also included measures based on watershed
land use; and the most restrictive criteria included all these as well as Bryce et al.'s condition class.
The value of the fish index that represented the 25th percentile for reference sites was selected as the
line of impairment for each method. For the three sets of reference sites, the values of the fish index
were very similar and their average was used to define reference condition for fish assemblages.
Patterns of index variability were similar across assemblage types
Statistical precision is an important feature of any monitoring tool because it determines the ability
of an indicator to detect change should it occur. A highly variable indicator must show a large
29
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
change in value before the change is statistically significant. Lack of sensitivity translates into an
inability to sound an alarm that will protect resources from degradation. Statistical power analysis
can be used to estimate the magnitude of change that an indicator can detect. Two commonly used
statistical models for power analysis are the t-test and regression (Peterman, 1990; Carlisle and
Clements, 1999; Fore et al., 2001). Given EMAP's focus on estimating trends through time, a
regression model with index regressed against year is arguably the more relevant statistical model
for power analysis (Stevens, 1994; Larsen et al., 2001; Hughes et al., 1998; Urquhart et al., 1998).
Results from the two approaches indicate that the MAIA multimetric indexes had adequate
precision to distinguish between two and five categories of biological condition and could detect
between 1.5% and 2.5% change per year after five years of monitoring.
Both approaches for estimating statistical power to detect change use estimates of variance
components derived from ANOVA. For power analysis based on a t-test to compare two sites with
three replicates each, within-year variance is used to calculate the minimum detectable difference
for index values at two sites (MDD; Zar, 1984). By dividing the range of the index by the MDD,
one can calculate the number of categories of biological condition that the index can detect (Fore et
al., 1994; Fore et al., 2001; Fore 2002a; Blocksom, 2003). The regression model uses the within-
year variance as well, but also uses estimates of variance associated with site x year interaction and
year-to-year variability to calculate statistical power (Larsen et al., 1995; Urquhart et al., 1998).
The relative magnitude of the variance components illustrates which temporal influences were
relatively more important for each index. The percentage of total variance associated with repeat
visits to the same site, that is, all sources of variance besides that associated with site differences
were approximately similar across assemblages (13-20%). For fish and diatom assemblages, the
multimetric index tended to have less variance associated with repeat visits than did its component
metrics (Figure 6; McCormick and Peck, 2000; Fore, 2002b). EPA guidelines for biological
indicators recommend that the overall variability associated with repeat sampling within a single
year, or the error component, not exceed 10% (McCormick and Peck, 2000). Only the fish index
met this target with 6% of total variance associated with error; at 17%, the diatom index was
furthest from the mark and the invertebrate index was close with 13% (Table 4).
Indexes for each assemblage differed in the way that the "nuisance" variance was allocated to each
of the different sources, i.e., year-to-year differences, site x year interaction, and repeat sampling
within year (measurement error). For the diatom index, most of its variance was associated with
repeat visits to a site while year-to-year variance was nearly zero. In contrast, fish and invertebrate
indexes did tend to change together across years (year-to-year variance). This difference may be
related to longer life cycles for fish and invertebrates compared with diatoms.
Based on statistical power calculations for the t-test model, the fish index was most precise and
could detect approximately 5.5 categories of biological condition, the invertebrate index could
detect 4 categories, and the diatom index 2.4. The diatom index had the largest percentage of its
total variance associated with error; therefore, it follows that it had the lowest statistical power to
detect differences.
30
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
100
80-
CD
o
sz
05
05
JS
2 40
60-
o
^
20-
0J
Figure 6. Variance components for the invertebrate index and its metrics. Variance
associated with repeat visits to the same site during the same year ("error") and from year to
year ("year") was lower for the index than for its metrics. Approximately 17% of the index's
variability was associated with repeat visits (see Table 3).
Table 4. Components of variance expressed as a percentage of the total
variance for diatom, invertebrate, and fish multimetric indexes. Variance
associated with site differences, year-to-year differences, site x year
interaction, and repeat visits within years are shown for each index.
Site
Year
Site x year
Error (repeat visits)
Diatom
80.4
0
2.7
16.9
Invertebrate
83.3
2.1
1.6
12.9
Fish
86.8
1.5
5.6
6.2
31
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
For the trend (regression) model, the statistical power of the fish index was highest and the
invertebrate index lowest (Figure 7). The differences were small after the first few years of
monitoring, however, and the indexes had very similar statistical power over the long term. EPA's
performance objectives for trend detection specify that an indicator should be capable of detecting a
2% change per year after approximately five years of sampling 30-50 sites assuming a type I error
rate of 0.1 and a type II error rate of 0.2 (McCormick and Peck, 2000). On the basis of five years of
monitoring at 40 sites, the diatom index could detect a 1.5% change per year, the fish index 2.1%,
and the invertebrate index 2.5%.
Thus, the two statistical models ranked the three indexes differently in terms of their statistical
power; these differences are entirely an artifact of the statistical model used to calculate power.
Because the year-to-year component of variability is relatively more important for detecting trends
through time, higher annual variability of the invertebrate index caused it to have the lowest power
for the trend (regression) model (Larsen et al., 1995; Urquhart et al., 1998). These results illustrate
that when comparing statistical power, identical statistical models must be used because results
depend entirely on the statistical model selected and its underlying assumptions (Blocksom, 2003).
O Diatom
-0- Invertebrate
-»- Fish
4 5 6 10
Number of years
20
Figure 7. The percentage change per year that would represent a statistically significant
change in multimetric index values decreases as the number of years of sampling
increases. A smaller percentage change indicates a more precise index and greater
statistical power to detect change. Thus, the invertebrate index was the least precise
although all three indexes were very similar. Based on repeat visits to the same 40 sites
each year.
32
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Invertebrate and diatom index values were comparable for pool and riffle
samples
Stream resources are naturally varied and complex but index values used for assessment must be
independent of natural features and mean the same thing whether the assessment is from a low or
high elevation site or a wide or narrow reach. In this way, an index must be applicable to all types
of streams in a region.
The tendency when developing multimetric indexes is to select similar sites for metric testing, and
eliminate data from small or unique ecoregions, unusual sites, or different habitat types. The
purpose is to identify a set of homogeneous sites with similar underlying physical and geographic
features so that metric response will be stronger in the absence of other confounding factors.
In the Mid-Atlantic region, a homogeneous set of sites could not be easily defined because there
were so many choices for criteria including watershed size, ecoregion, and lowlands vs. uplands. A
key lesson learned from the MAIA pilot was that including as much data as possible in the metric
testing process was more efficient and simpler than testing multiple sets of sites. In addition, unique
sites were not 'orphaned' from the assessment process.
Riffles and pools within a reach represent different habitats to invertebrates and diatoms and
sampling in these areas typically yield different taxa. The MAIA sampling protocol kept the
samples separate when both habitats occurred at a sample site. For invertebrates, pool and riffles
were different: some metrics calculated for pools tended to indicate poorer conditions. Fortunately,
similar metrics were correlated with disturbance in each habitat. To compensate for these natural
differences, metrics were simply adjusted when they were converted to unit-less scores. In this way,
the final index was comparable for samples from a pool or a riffle (Klemm et al., 2003). For
diatoms, similar adjustments were not necessary for pool and riffle habitats because index values
were similar from each habitat and contributed no measurable source of variability to the index
(Fore, 2002b).
Assemblages differed in their sensitivity to disturbance types
When multiple assemblages are monitored, potentially conflicting assessments must be reconciled.
What if a stream reach falls below the "impaired" threshold according to its fish index value but not
according to its invertebrate index value? Several studies have found that biological indexes based
on different assemblages generally agree in terms of assessment condition, that is, indexes for
different assemblages are highly correlated (Lammert and Allan, 1999; O'Connor et al., 2000).
Nonetheless, assemblages may disagree at specific sites and the conflict must be resolved to
determine whether the site is impaired. Some of the disagreement may be associated with
differential sensitivities of assemblages to specific types of disturbance (Bryce and Hughes, 2002;
Norton etal., 2002).
The lesson learned from Mid-Atlantic streams was that any of the three assemblages could be used
to monitor stream condition because multimetric indexes for all three assemblages could reliably
distinguish degraded sites from sites with little or no human influence. However, different
assemblages showed different sensitivities associated with their natural history and assessments
based on different assemblages provide more information about the type and source of disturbance.
33
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Multimetric indexes for all three assemblages could distinguish reference sites from sites disturbed
by nutrient enrichment, mixed impacts of development and agriculture, and acid mine deposition
(see Figure 4 on page 17). Overall, the indexes agreed, with lower values for mine drainage than for
other types of disturbance. The indexes disagreed somewhat on the relative magnitude of
degradation associated with nutrient and acid deposition, in that diatoms indicated greater
degradation associated with nutrients, probably because several diatom metrics were related to
different aspects of nutrient enrichment (e.g., organic vs. inorganic sources; Table 5). In contrast,
acid deposition sites looked more like reference sites from the diatom point of view. Because acid
deposition sites were typically nutrient poor and steeply sloped, they lacked diatom taxa that were
associated with alkalinity, nutrients, or sediment.
The different assemblages tended to respond differently to specific stressors. Invertebrate and
diatom indexes were more highly correlated with specific stressors measured at the reach scale,
such as nitrogen or riparian condition than the fish index (see Table 1 on page 18). In contrast, the
fish index was less significantly correlated with chemical measures and measures at the reach scale,
possibly because fish are not limited to the smaller spatial scale of a reach, as are invertebrates and
diatoms.
Table 5. Biological metrics included in the fish, invertebrate, and diatom indexes. Percent sign (%)
denotes the percentage of individuals belonging to a given group out of the total number of sampled
individuals.
Metric type
Taxonomic composition
Tolerance & intolerance
Assemblage structure
Autecological guild
Trophic guild
Morphometric guild
Reproductive guild
Diatom
% intolerant
% very tolerant
% salt tolerant
% tolerant of low
oxygen
% eutrophic
% N heterotrophs
% polysaprobic
% alkaliphilic
% very motile
Invertebrate
No. Ephemeroptera taxa
No. Plecoptera taxa
No. Trichoptera taxa
Tolerance index
% non-insects
% dominance
No. collector/filterertaxa
Fish
No. cyprinid taxa
% cottid
No. sensitive taxa
% tolerant
No. benthic taxa
% exotics
% piscivore/invertivore
% macro-omnivore
% gravel-spawning taxa
34
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
VII. CONCLUSIONS
The goal of the MAIA study was to demonstrate how a probabilistic sampling design could be used to
develop monitoring protocols and assess surface waters at a regional scale. The purpose of this document
was to distill the critical lessons from the Mid-Atlantic experience and present approaches and
conclusions that will be relevant to states as they move forward with their own monitoring programs.
Although EMAP's objectives are to support states in their assessment of surface waters by developing
survey design methods, biomonitoring tools, and biocriteria, these tools will not necessarily be adopted by
states. The MAIA pilot was a demonstration; and states in the region are not obligated to adopt EMAP
methods. Currently, monitoring and assessment under the Clean Water Act is each state's responsibility,
although the EPA may intervene if a state's program or reporting is inadequate. Whether states in the
region of this pilot will adopt or adapt EMAP protocols is unclear. Nonetheless, the initial MAIA pilot
study has metamorphosed into a larger collaborative effort that coordinates data and information across
multiple agencies responsible for water resources in this region.
Whether or not the specific methods and protocols developed by EMAP are appropriate for state
monitoring programs, the issues related to sampling design, data management, and the development of
assessment tools are relevant to any emerging monitoring program. For sampling design, the primary
issue relates to sample site selection. Resolution of this issue depends on a clear understanding of how the
data will be used before they are collected (Ward et al., 1986). The way data are entered into the
computer seems remote from data analysis and interpretation, but when data management issues are left
unresolved, they can derail or corrupt any analysis.
For the MAIA study, objective criteria were used to establish links between human disturbance and
biological response. To ensure that biological monitoring tools were both meaningful and reliable, they
were evaluated for their correlation with independent measures of human disturbance and for their
variability through time. Multimetric indexes developed for fish, invertebrates and diatoms showed a
strong and consistent correlation with multiple aspects of human disturbance measured at a variety of
spatial scales. For all biological indexes, the more integrative the measure of human disturbance, the
higher the correlation. All three indexes were capable of detecting small changes in biological condition
within a few years of sampling (2.5% or less per year). Both their strong correlation with disturbance and
their statistical precision support multimetric indexes in their role as biological monitoring tools for
protecting water resources within the legal framework of the Clean Water Act.
The size of the MAIA project created some of its own new challenges; hundreds of people have been
involved in creating the sampling design and collecting, managing and analyzing the data. As state
programs grow to meet the requirements of the Clean Water Act, the scale of this project appears less
daunting (Ransel, 1995; NRC 2001). In fact, many states already assess hundreds of sites each year.
Consequently, the process and outcomes of the MAIA pilot serve to highlight persistent issues or sticking
points that states will continue to grapple with as they develop adequate biological monitoring programs
(Yoder and Rankin, 1998).
35
-------
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
REFERENCES
Barbour M.T., Gerritsen J., G.E. Griffith, R. Frydenborg, E. McCarron, J.S. White, and M.L. Bastian.
1996. A framework for biological criteria for Florida streams using benthic macroinvertebrates.
Journal of the North American Benthological Society 15: 185-211.
Barbour M.T., Gerritsen J., Snyder B.D. and Stribling J.B. 1999. Rapid bioassessment protocols for use in
streams and wadeable rivers: Periphyton, benthic macroinvertebrates, and fish. Second edition.
EPA 841-B-99-002. US Environmental Protection Agency, Office of Water, Washington, D.C.
Beyers, D.W. 1998. Causal inference in environmental impact studies. Journal of the North American
Benthological Society 17:367-373.
Blocksom, K.A. 2003. A performance comparison of metric scoring methods for a multimetric index for
Mid-Atlantic Highlands streams. Environmental Management, impress.
Bryce, S.A., D.P. Larsen, R.M. Hughes, and P.R. Kaufmann. 1999. Assessing relative risks to aquatic
ecosystems: A mid-Appalachian case study, Journal of the American Water Resources Association,
35(1), 23-36.
Bryce, S.A. and R.M. Hughes. 2002. Variable Assemblage Responses to Multiple Disturbance Gradients:
Oregon and Appalachian, USA, Case Studies. Biological Response Signatures: Indicator Patterns
Using Aquatic Communities (Ed T.P. Simon), pp. 539-560. CRC Press LLC, Boca Raton, FL.
Carlisle D.M. and Clements W.H. 1999. Sensitivity and variability of metrics used in biological
assessments of running waters. Environmental Toxicology and Chemistry, 18, 285-291.
Davis, W.S. and Scott. 2000. Mid-Atlantic Highlands Streams Assessment: Technical Support Document.
EPA/903/B-00/004. USEPA Mid-Atlantic Integrated Assessment, Ft. Meade, Maryland.
U.S. Environmental Protection Agency (EPA). 2002. Summary of Biological Assessment Programs and
Biocriteria Development for States, Tribes, Territories, and Interstate Commissions: Streams and
Wadeable Rivers. EPA-822-R-02-048. U.S. Environmental Protection Agency, Office of
Environmental Information and Office of Water, Washington, DC.
U.S. Environmental Protection Agency (EPA). 2003. Aquatic Resource Monitoring Web Site.
www.epa.gov/nheerl/arm/index.htm
Fore L.S., Karr J.R. and Conquest L.L. 1994. Statistical properties of an index of biotic integrity used to
evaluate water resources. Canadian Journal of Fisheries and Aquatic Sciences, 51, 212-231.
Fore L.S., Paulsen K., K. O'Laughlin. 2001. Assessing the performance of volunteers in monitoring
streams. Freshwater Biology, 46, 109-123
Fore L.S. 2002a. Biological assessment of mining disturbance on stream invertebrates in mineralized
areas of Colorado. Biological Response Signatures: Indicator Patterns Using Aquatic Communities
(Ed T.P. Simon), pp. 445-480. CRC Press LLC, Boca Raton, FL.
Fore L.S. 2002b. Response of diatom assemblages to human disturbance: development and testing of a
multimetric index for the Mid-Atlantic Region (USA). Biological Response Signatures: Indicator
Patterns Using Aquatic Communities (Ed T.P. Simon), pp. 347-370. CRC Press LLC, Boca Raton,
FL.
Fore, L.S. and C. Grafe. 2002. Using diatoms to assess the biological condition
of large rivers in Idaho (USA). Freshwater Biology 47:2015-2037..
Gerritsen, J.G., J. Green, and R. Preston. 1994. Establishment of regional reference conditions for stream
biological assessment and watershed management. Proceedings of Watersheds '93: A National
Conference on Watershed Management, Arlington, Va., March 21-24, 1993, U.S. Environmental
Protection Agency, EPA-804-R-94-002, p. 797-801.
Hale, S.S., L.H. Bahner and J.F. Paul. 2000. Finding common ground in managing data used for regional
environmental assessments. Environmental Monitoring and Assessment 63:143-157.
37
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Hale, S.S. and H.W. Buffum. 2000. Designing environmental databases for statistical analyses.
Environmental Monitoring and Assessment 64:55-68.
Hale, S.S., M.M. Hughes, J.F. Paul, R.S. McAskill, S.A. Rego, D.R. Bender, N.J. Dodge, RL. Richter
and J.L. Copeland. 1998. Managing scientific data: The EMAP approach. Environmental
Monitoring and Assessment 51: 429-440.
Hale, S.S., J. Rosen, D. Scott, J. Paul, and M. Hughes. 1999. EMAP Information Management Plan:
1998-2001. EPA/620/R-99/001a. National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency.
Heffner, R.A., M.J. Butler IV, and C.K. Reilly. 1996. Pseudoreplication revisited. Ecology 77:2558-2562.
Herlihy, A.T., P.R. Kaufmann, and M.E. Mitch. 1990. Regional estimates of acidic mine drainage impact
on streams in the Mid-Atlantic and Southeastern United States. Water, Air and Soil Pollution
50:91-107.
Herlihy, A.T., P.R. Kaufmann, M.R Church, P.J. Wigington, Jr., J.R. Webb, and M.J. Sale. 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., D.P. Larsen, S.G. Paulsen, N.S. Urquhart, and B.J. Rosenbaum. 2000. Designing a
spatially balanced, randomized site selection process for regional stream surveys: the EMAP Mid-
Atlantic pilot study, Environmental Monitoring and Assessment, 63, 95-113.
Herlihy, A.T., J.L. Stoddard, and C.B. Johnson. 1998. The relationship between stream chemistry and
watershed land cover data in the Mid-Atlantic region, U.S. Water, Air and Soil Pollution, 105, 377-
386.
Hill, B.H., A.T. Herlihy, P.R. Kaufmann, R.J. Stevenson, F.H. McCormick, and C. Burch Johnson. 2000.
Use of periphyton assemblage data as an index of biotic integrity, 19(1), 50-67.
Hill, B.H., RJ. Stevenson, Y. Pan, A.T. Herlihy, P.R. Kaufmann, C.B. Johnson. 2001. Comparison of
correlations between environmental characteristics and stream diatom assemblages characterized at
genus and species levels. Journal of the North American Benthological Society 20:299-310.
Hudson, L.I. and J.J.H. Cibrowski. 1996. Teratogenic and genotoxic responses of larval Chironomus
salinarius group (Diptera: Chironomidae) to contaminated sediment. Environmental Toxicology
and Chemistry 15: 1375-1381.
Hughes, R.M. 1995. Defining acceptable biological status by comparing with reference conditions. In
Davis, W.S. and T.P. Simon (eds.), Biological Assessment and Criteria: Tools for Water Resource
Planning and Decision Making. Lewis Publishers, Boca Raton, FL:31-47.
Hughes, R.M., P.R. Kaufmann, A.T. Herlihy, T.M. Kincaid, L. Reynolds, and D.P. Larsen. 1998. A
process for developing and evaluating indices offish assemblage integrity, Canadian Journal of
Fisheries and Aquatic Sciences, 55, 1618-1631.
Hughes, R.M., S.G. Paulsen, and J.L. Stoddard. 2000. EMAP-Surface Waters: A national,
multiassemblage, probability survey of ecological integrity, Hydrobiologia, 423, 429-443.
Hurlbert S.H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological
Monographs, 54, 187-211.
Jackson, L.E., J.C. Kurtz, and W.S. Fisher (Eds.). 2000. Evaluation Guidelines for Ecological Indicators.
EPA/620/R-99/005. U.S. Environmental Protection Agency, Office of Research and Development,
Research Triangle Park, NC. 107 pp.
Karr J.R. 1991. Biological integrity: a long-neglected aspect of water resource management. Ecological
Applications, 1, 66-84.
Karr J.R. and Chu E.W. 1999. Restoring Life in Running Waters: Better Biological Monitoring. Island
Press, Washington, DC.
38
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Karr J.R., Fausch K.D., Angermeier P.L., Yant P.R. and Schlosser I.J. 1986. Assessment of biological
integrity in running water: a method and its rationale. Illinois Natural History Survey Special
Publication Number 5, Champaign, IL.
Karr J.R., Allan J.D. and Benke A.C. 2000. River conservation in the United States and Canada. Global
Perspectives on River Conservation: Science, Policy, Practice (Eds P.J. Boon, B.R. Davies and
G.E. Petts), pp. 3-39. J. Wiley, Chichester, UK.
Kaufmann, P.R., P. Levine, E.G. Robison, C. Seeliger, and D.V. Peck. 1999. Quantifying physical habitat
in wadeable streams. EPA/620/R-99/003. U.S. Environmental Protection Agency, Washington,
DC.
Klemm, D.J., K.A. Blocksom, F.A. Fulk, A.T. Herlihy, R.M. Hughes, P.R. Kaufmann, D.V. Peck, J.L.
Stoddard, W.T. Thoeny, M.B. Griffith, and W.S. Davis. 2003. Development and evaluation of a
macroinvertebrate biotic integrity index (MBII) for regionally assessing Mid-Atlantic Highlands
streams. Environmental Management 31(5): 656-669.
Klemm, D.J., K.A. Blocksom, W.T. Thoeny, F.A. Fulk, A.T. Herlihy, P.R. Kaufmann, S.M. Cormier.
2002. Using macroinvertebrates as indicators of ecological conditions for streams in the Mid-
Atlantic Highlands region. Environmental Monitoring and Assessment 78:169-212.
Lammert, M. and J.D. Allan. 1999. Assessing biotic integrity of streams: Effects of scale in measuring the
influence of land use/cover and habitat structure on fish and macroinvertebrates, Environmental
Management, 23(2), 257-270.
Larsen, D.P., N.S. Urquhart, and D.L. Kugler. 1995. Regional scale trend monitoring of indicators of
trophic condition of lakes. Water Resources Bulletin 31:117-140.
Larsen, D. P., T. M. Kincaid, S. E. Jacobs, and N. S. Urquhart. 2001. Designs for evaluating local and
regional scale trends. BioScience 12:1069-1078.
Lemly, A.D. 2000. Using bacterial growth on insects to assess nutrient impacts in streams. Environmental
Monitoring and Assessment 63:431-446.
Loftis, J.C. C.H. Taylor, A.D. Newell, and P.L. Chapman. 1991. Multivariate trend testing of lake water
quality. Water Resources Bulletin 27:461-473.
McCormick, F.H., R.M. Hughes, P.R. Kaufmann, D.V. Peck, J.L. Stoddard, A.T. Herlihy. 2001.
Development of an index of biotic integrity for the Mid-Atlantic Highlands region. Transactions of
the American Fisheries Society 130:857-87.
McCormick, F.H. and D.V. Peck. 2000. Application of the indicator evaluation guidelines to a
multimetric indicator of ecological condition based on stream fish assemblages. Chapter 4 in
Evaluation Guidelines for Ecological Indicators (L.E. Jackson, J.C. Kurtz, and W.S. Fisher, eds.),
EPA/620/R-99/005. U.S. Environmental Protection Agency, Office of Research and Development,
Research Triangle Park, NC.
Mebane, C.A. 2002. Effects of metals on freshwater macroinvertebrates: a review and case study of the
correspondence of multimetric index, toxicity testing, and copper concentrations in sediment and
water. Biological Response Signatures: Indicator Patterns Using Aquatic Communities (Ed T.P.
Simon), pp. 287-312. CRC Press LLC, Boca Raton, FL.
Miltner, R.J. and E.T. Rankin. 1998. Primary nutrients and the biotic integrity of rivers and streams.
Freshwater Biology 40:145-158.
National Research Council (NRC). 1995. Review of EPA's Environmental Monitoring and Assessment
Program: Overall Evaluation. National Academy Press, Washington, DC.
National Research Council (NRC). 2001. Assessing the TMDL Approach to Water Quality Management.
National Academy Press, Washington, DC. 109 pp.
Norton, S.B., S.M. Cormier, M. Smith, R. Christian Jones. 2000. Can biological assessments discriminate
among types of stress? A case study from the Eastern Corn Belt Plains ecoregion. Environmental
Toxicology and Chemistry 19:1113-1119.
39
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
O'Connell, T.J., L.E. Jackson, and R.P. Brooks. 1998. A bird community index of biotic integrity for the
mid-Atlantic highlands, Environmental Monitoring and Assessment, 51(1-2), 145-156.
O'Connor, R.J., T.E. Walls, and R.M. Hughes. 2000. Using multiple taxonomic groups to index the
ecological condition of lakes. Environmental Monitoring and Assessment 61:207-228.
Olden, J.D. and D. A. Jackson. 2000. Torturing data for the sake of generality: how valid are our
regression models? Ecoscience 7:501-510.
Olsen, A.R., J. Sedransk, D. Edwards, Gotway, C.A., Liggett, W., Rathbun, S., Reckhow, K.H. and
Young, L.J. 1999. Statistical issues for monitoring ecological and natural resources in the United
States. Environmental Monitoring and Assessment 54: 1-45
Overton, W.S. and Stehman, S.V. 1996. Desirable design characteristics for long-term monitoring of
ecological variables. Environmental and Ecological Statistics. 3: 349-361.
Pan, Y.R.J. Stevenson, B.H. Hill, A.T. Herlihy, and G.B. Collins. 1996. Using diatoms as indicators of
ecological conditions in lotic systems: a regional assessment. Journal of the North American
Benthological Society 15: 481-495.
Pan Y., Stevenson R.J., Hill B.H. and Herlihy A.T. 2000. Ecoregions and benthic diatom assemblages in
Mid-Atlantic Highlands streams, USA. Journal of the North American Benthological Society, 19,
518-540.
Pan, Y.R.J. Stevenson, B.H. Hill, P.R. Kaufmann, and A.T. Herlihy. 1999. Spatial patterns and ecological
determinants of benthic algal assemblages in Mid-Atlantic streams, USA. Journal of Phycology 35:
460-468.
Paulsen, S.G., R.M. Hughes, and D.P. Larsen. 1998. Critical elements in describing and understanding
our nation's aquatic resources. Journal of the American Water Resources Association 34:995-1005.
Peterman R.M. 1990. Statistical power analysis can improve fisheries research and management.
Canadian Journal of Fisheries and Aquatic Sciences, 47, 2-15.
Peterson, S.A, N.S. Urquhart, and E.B. Welch. 1999. Sample representativeness: a must for reliable
regional lake condition estimates. Environmental Science and Technology. 33: 1559-1565.
Ransel K.P. 1995. The sleeping giant awakes: PUD No. 1 of Jefferson County v. Washington Department
of Ecology. Environmental Law, 25, 255-283.
Richardson, J.S. and P.M. Kiffney. 2000. Response of a stream macroinvertebrate community from a
pristine, southern BC stream to metals in experimental mesocosms. Environmental Toxicology and
Chemistry 19:736-743.
Stevens, D.L. Jr. 1994. Implementation of a national environmental monitoring program. Journal of
Environmental Management 42:1-29.
Stevenson R.J. and Bahls L. 1999. Chapter six: periphyton protocols. Rapid bioassessment protocols for
use in streams and wadeable rivers: Periphyton, benthic macroinvertebrates, and fish, 2nd edn. (Eds
M.T. Barbour, J. Gerritsen, B.D. Snyder and J.B. Stribling) EPA 841-B-99-002. US Environmental
Protection Agency, Office of Water, Washington, D.C.
Stewart-Oaten, A., W.W. Murdoch, and K.R. Parker. 1986.Environmental impact assessment:
"Pseudoreplication" in time? Ecology 67:929-940.
Stewart-Oaten, A., J.R. Bence, and C.W. Osenberg. 1992. Assessing effects of unreplicated perturbations:
no simple solutions. Ecology 73: 1396-1404.
Stoddard, J.L., C.T. Driscoll, J.S. Kahl, and J.H. Kellogg. 1998. Can site-specific trends be extrapolated
to a region? An acidification example for the Northeast. Ecological Applications 8: 288-299.
Suter, G.W. II. 2001. Applicability of indicator monitoring to ecological risk assessment. Ecological
Indicators 1:1010-112.
Urquhart, N.S., S.G. Paulsen, and D.P. Larsen. 1998. Monitoring for policy-relevant regional trends over
time. Ecological Applications 8:246-257.
40
-------
Developing Biological Indicators:
Lessons Learned from Mid-Atlantic Streams
Waite, I.R., A.T. Herlihy, D.P. Larsen, D.J. Klemm. 2000. Comparing strengths of geographic and
nongeographic classifications of stream benthic macroinvertebrates in the Mid-Atlantic Highlands,
USA. Journal of the North American Benthological Society 19:429-441.
Wallace, J.B., J.W. Grubaugh and M.R. Whiles. 1996. Biotic indices and stream ecosystem processes:
results from an experimental study. Ecological Applications 6: 140-151.
Wang, L., J., Lyons, P. Kanehl. 1998. Development and evaluation of a habitat rating system for low-
gradient Wisconsin streams. North American Journal of Fisheries Management 19:775-785.
Ward, R.C., Loftis, J.C., and G.B. McBride. 1986. The "data rich but information poor" syndrome in
water quality monitoring. Environmental Management 10:291-297.
White, J. and G. Merritt. 1998. Evaluation of R-EMAP techniques for the measurement of ecological
integrity of streams in Washington state's coast range ecoregion. Environmental Monitoring and
Assessment 51:345-355.
Yoder, C.O. and E.T. Rankin. 1998. The role of biological indicators in a state water quality management
process. Environmental Monitoring and Assessment 51: 61-88.
Yoder, C.O. and J.E. DeShon. 2002. Using biological response signatures within a framework of multiple
indicators to assess and diagnose causes and sources of impairments to aquatic assemblages in
selected Ohio rivers and streams. Biological Response Signatures: Indicator Patterns Using Aquatic
Communities (Ed T.P. Simon), pp. 23-82. CRC Press LLC, Boca Raton, FL.
Zar J.H. 1984. Biostatistical Analysis, 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, NJ.
41
-------
xvEPA
Agencf
Office of Environmental Information
and Office of Water
Washington, DC
Official Business
Penalty for Private Use
$300
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE
detach, or copy the cover, and return to the address in
the upper left-hand corner.
PRESORTED STANDARD
& PAID
EPA
PERMIT No. G-35
EPA/903/R-03/003
March 2003
------- |