CBP/TRS 262J02
EPA 903-R-02-004
June 2002
Development Of Diagnostic Approaches To
Determine Sources Of Anthropogenic Stress
Affecting Community Condition In The
Chesapeake Bay
Chesapeake Bay Program
A Watershed Partnership
-------
Development Of Diagnostic Approaches To Determine
Sources Of Anthropogenic Stress AffectingBenthic
Community Condition In The Chesapeake Bay
Principal Investigators: Daniel M. Dauer l
Michael F. Lane l
Roberto! Llanso2
1- Department of Biological Sciences
Old Domini on University
Norfolk, VA 23529-0456
2- Contract # CA0201/074307223734
Prepared for:
Chesapeake Bay Program
A Watershed Partnership
Chesapeake Bay Program
410 Severn Avenue, Suite 10S
Annapolis, Maryland 21403
1-800-YOUR-BAY
www.chesapeakebay.net
f'rimed'for the Chesapeake 2 Program iy tfit Smironmental Protection Agerxy
£.ecycledt£*cyclal>le - Printed with Vegetable Oil Sased fnki on Recycled Paper 3Q%Postc0nsiMer
-------
Table of Contents
List of Figures ii
List of Tables iii
List of Appendices v
I. Introduction 1
II. Methods 2
A. Overview 2
B. Database 2
C. Candidate Metrics 3
D. Data Aggregation 4
E. Spatial Scales and Analytical Scenarios 5
F. Discriminant Function Calibration and Validation .. • 6
G. Salinity Corrections 7
H. Variable Reduction Approaches 7
III. Results 8
A. Description of Database 8
B. Within Habitat Type Scenarios 10
C. Within Salinity Regime Scenarios 10
D. Baywide Scenarios 12
IV. Discussion 13
A. Overview of Results 13
B. Usage Constraints 14
C. Technical Approaches to Implementation 15
D. Recommendations 16
VI. Literature Cited 17
Figures 22
Tables 24
Appendices 47
-------
List of Figures
Figure 1. Discriminant function classification efficiencies for individual habitat types for
classifying Mesohaline severely degraded sites (excluding Low D.O. sites) into the
Contaminant and Other stress groups. Numbers above the bars indicate the number
of observations within each habitat type 22
Figure 2. Discriminant function classification efficiencies for individual habitat types for the
Baywide discriminant function for classifying severely degraded and degraded sites
(including Low D.O. sites) into the Contaminant and Other stress groups. Numbers
above the bars indicate the number of observations within each habitat type. .. 23
-------
List of Tables
Table 1. Data aggregation schemes used in analyses 24
Table 2. Candidate metrics used for analytical tool development 25
Table 3. Habitat types for the Chesapeake Bay B-IBI as defined by Weisberg et al.
(1997) 26
Table 4. ERM guidelines for 24 trace metals (ppm dry wt) and organic compounds (ppb, dry
wt) as defined from Long et al. (1995) 26
Table 5. Number of sampling location/date combinations for each monitoring program within
Chesapeake Bay 27
Table 6. Frequency and percentage of sites and mean B-IBI for sites within each status
classification category 27
Table 7. Frequency of sites classified as severely degraded and degraded within habitat
types 28
Table 8. Frequency of sites classified as severely degraded and degraded within habitat
types 28
Table 9. Frequency of sites classified as severely degraded and degraded within each habitat
and effect type for each of the sediment contaminant classification schemes ... 29
Table 10. Classification efficiencies of linear discriminant functions developed for the Within
Habitat Type scenarios for all available stress groups 30
Table 11. Classification efficiencies of linear discriminant functions developed for Within
Habitat Type scenarios for discriminating between the Contaminant and Other stress
groups 33
Table 12. Classification efficiencies of linear discriminant functions developed for classifying
Polyhaline sites into one of the four stress groups 36
Table 13. Classification efficiencies of linear discriminant functions developed for classifying
Polyhaline sites into the Contaminant, Combined and Unknown stress groups.. 36
Table 14. Classification efficiencies of linear discriminant functions developed for classifying
Polyhaline sites into the Contaminant and all Other stress groups with and without
Low D.O. sites 37
in
-------
Table 15. Classification efficiencies of linear discriminant functions developed for classifying
Mesohalme sites into one of the four stress groups 38
Table 16. Classification efficiencies of linear discriminant functions developed for classifying
Mesohaline sites into the Contaminant, Combined and Unknown stress
groups 38
Table 17. Classification efficiencies of linear discriminant functions developed for classifying
Mesohaline sites into the Contaminant and all Other stress groups with and without
Low D.O. sites 39
Table 18. Classification efficiencies of linear discriminant functions developed for classifying
Tidal Freshwater and Oligohaline sites into the Contaminant and all Other stress
groups 40
Table 19. Classification efficiencies of linear discriminant functions developed for Baywide
scenarios to discriminate between four potential stress groups for both uncorrected
and salinity corrected data 41
Table 20. Classification efficiencies of linear discriminant functions developed for Baywide
scenarios to discriminate between the Contaminant, Combined and Unknown stress
groups for both uncorrected and salinity corrected data 42
Table 21. Classification efficiencies of linear discriminant functions developed for Baywide
scenarios to discriminate between the Contaminant and all Other stress groups with
Low D.O. sites for both uncorrected and salinity corrected data 43
Table 22. Classification efficiencies of linear discriminant functions developed for Baywide
scenarios to discriminate between the Contaminant and all Other groups without Low
D.O. sites for both uncorrected and salinity corrected data 44
Table 23. Classification efficiencies of discriminant functions developed for selected scenarios
after application of the stepwise discriminant and ANOVA variable reduction
procedures 45
Table 24. Coefficients and cutoff values for the Baywide linear discriminant function for
classifying severely degraded and degraded sites into the Contaminant and Other
stress groups using "uncorrected" data 46
IV
-------
List of Appendices
Appendix A. List of species classified as epifaunal 47
Appendix B. List of species classified as deep deposit feeders. 48
Appendix C. List of species classified as suspension feeders 49
Appendix D. List of species classified as interface feeders 50
Appendix E. List of species classified into the carnivore/omnivore feeding group category. . 51
Appendix F. Number of contaminants exceeding the Effects Range Median concentration (ERM
Cone.), the mean Sediment Quality Guidelines (SQG) quotient, the number of
missing analytes, and a listing of missing analytes for each station date combination
classified as severely degraded or degraded. Habitat type is based on Weisberg et al.
(1997) 52
Appendix G. Correlations between benthic bioindicators and salinity 63
Appendix H. Regression relationships for salinity corrections of selected benthic
bioindicators 64
-------
I. Introduction
Benthic macrofauna are an important component of estuarine ecosystems. These organisms are a
food source for higher trophic levels including fishes and crabs (Virnstein 1977, 1979; Holland et
al. 1980; Dauer et al. 1982) and birds (Botton 1984; Quammen 1984). Benthic macrofauna affect
both the physical and chemical properties of the sediment and the overlying water column (e.g.,
Rhoads and Young 1970; Rhoads 1973; Aller 1978, 1980), influence nutrient cycling (Rowe et al
1975; Zeiteschel 1980; Flint and Kamykowski 1984), and are capable of directly controlling
phytoplankton biomass in the water column (Cleorn 1982; Officer et al 1982; Cohen et al. 1984;
Nichols 1985). Because of these characteristics, monitoring of the benthos provides important
information for making management decisions in marine systems (Bilyard 1987). Also, the
relatively long life span and sedentary nature of these organisms make them good indicators of water
quality and the effects of man-made disturbances on aquatic systems (Reish 1973; Pearson and
Rosenberg 1978; Bilyard 1987).
Numerous studies have documented the effects of pollution and other anthropogenic activities on
macrofaunal communities within estuaries (e.g., Boesch 1972; Brown et al. 1987; Beukema 1991;
Gaston and Young 1992; Dauer et al. 1992, 1993; Dauer 1993, 1997; Dauer and Alden 1995).
Investigators attempting to describe the effects of pollution on benthic macrofaunal communities
have often experienced the problem of distinguishing the natural variation in these communities due
to habitat type (i.e., salinity regime, sediment type, depth, etc.) from the effects caused by pollution.
These problems have resulted in the development of multi-metric indices that allow for the
characterization of benthic biological condition within and between habitat types. This approach has
been used primarily in freshwater ecosystems and is typically referred to as the index of biotic
integrity (IBI) approach (see reviews by Davis and Simon 1995; Karr and Chu 1999). Recently.a
benthic index of biotic integrity (B-IBI) was developed for the Chesapeake Bay and its major
tributaries (Weisberg et al. 1997). This index compares the deviation of community metrics from
values at reference sites that are assumed to be minimally impacted by anthropogenic activities. This
index has been successfully used to describe the status of and long-term trends in benthic community
conditions within the Chesapeake Bay and its tributaries in relation to water quality characteristics
(Dauer et al. 1998; 1999) and is correlated to measures of land use and nutrient loads within the
Chesapeake Bay watershed (Dauer et al. 2000). However, one of the major limitations of this index
is its inability to directly identify the source of stress that is the cause of degraded benthic
community condition.
The objective of this study was to develop analytical tools that are capable of classifying regions m
Chesapeake Bay identified as having degraded benthic communities into categories distinguished
by the type of stress experienced by those communities. Sediment contaminants and bottom low
dissolved oxygen concentrations were identified as the primary sources of anthropogenic stress on
benthic communities and an attempt was made to develop multivanate statistical tools that could
distinguish between these sources of stress. Ultimately, environmental managers could use these
tools to make recommendations for analytical chemistry studies to confirm the sources and levels
of contaminants in predetermined regions of concern and to develop management plans for
controlling contaminant effects.
-------
II. Methods
A. Overview
The objective of this study was to develop statistical diagnostic tools that would allow environmental
managers to identify potential sources of anthropogenic stress to benthic communities within
Chesapeake Bay. To accomplish this task, data characterizing benthic community condition were
aggregated at different spatial scales and with a variety of defined stress groups (Table 1). Three
spatial scales of aggregation were identified: (1) a Within Habitat scale as defined by Weisberg et
al.(1997), (2) a Within Salinity Regimes scale, and (3) on a Baywide scale. At each spatial scale,
stress categories were defined. Four types of stress groups were defined 1) a Contaminant stress
group, 2) a low dissolved oxygen (Low DO) stress group, 3) a combined contaminant and low
dissolved oxygen (Combined) stress group, and 4) an Unknown stress group. For some scenarios,
the Low DO stress group was excluded for two reasons: (1) regions affected by low dissolved
oxygen stress, particularly associated with a stratified water column, are fairly well known; and (2)
benthic community condition due to contaminant stress might be less unique and, therefore, less
distinguishable from other sources of stress when including benthic community data affected by low
DO stress. For each spatial scale and stress groups combination tested, three Contaminant stress
groups criteria and two levels of benthic community degradation were appl ied to each data set (Table
1).
Linear discriminant analysis was used to develop diagnostic tools to differentiate between stress
groups as defined for each scenario (Kachigan 1991; Huberty 1994). Linear discriminant analysis
is a procedure that uses a set of predictor variables from a calibration data set to create a multivariate
discriminant function for assigning observations into one of two or more mutually exclusive
qualitative groups. Once developed, the discriminant function can be used to assign new
observations into the groups defined in the calibration data set (Kachigan 1991; Huberty 1994).
Classification of new observations into the groups is accomplished by one of two methods. The
discriminant scores calculated for each observation can be compared to a predetermined cutoff value
or values that determine group membership or posterior probabilities of group membership
calculated during the analysis are examined and new observations are assigned to the group with the
highest probability of group membership. For this study, linear discriminant functions for stress
group classification were developed using bioindicators calculated from a subset of data compiled
from existing and historical monitoring programs conducted within Chesapeake Bay. A second
subset of this data set was used to validate the discriminant functions developed A similar approach
has been used to differentiate between degraded and reference benthic communities in the Gulf of
Mexico (Engle et al. 1994; Engle and Summers 1999; Paul et al. 2001) and more recently to identify
stress source within a specific habitat type in Chesapeake Bay (Christman and Dauer, 2002)
B. Database
The analytical tools were calibrated and validated using data collected within Chesapeake Bay that
were used previously to develop the Chesapeake Bay Program's Benthic Index of Biotic Integrity
-------
(B-IBI) and USEPA's Mid-Atlantic Integrated Assessment (MAIA) Program's Benthic Index data
(Weisberg et al. 1997; Llanso, In Review). Additional data from sites monitored as part of
probability-based sampling regime for the Chesapeake Bay Program's Benthic Monitoring Program
(Dauer 1999; Dauer and Rodi 1999,2001 jVersar Inc. 2001) and for the Chesapeake Bay Program's
Ambient Toxicity Program (McGee et al. 2001) were also included. Data used in these analyses met
the following criteria: (1) all samples were collected within the geographic boundaries of
Chesapeake Bay and its tributaries, (2) all benthic biological samples were collected using a Young
grab with a sampling area of 0.0440 m2, (3) all benthic biological samples were collected during the
period of July 15 through September 30, (4) measurements of dissolved oxygen were collected
concurrently with the biological data, and (5) sediment contaminant data were collected during the
same year as the biological samples. Finally, only sites classified as either degraded (B-D31 < 2.6
but > 2.0) or severely degraded (B-IBI s2.0) were retained for subsequent analysis.
C. Candidate Metrics
Table 2 provides the list of candidate metrics used for the analyses that included measures of
abundance, richness, proportional abundance, species diversity and dominance of species in various
taxonomic, life history, and trophic categories. Additional metrics included total community
biomass, the ratio of epifaunal species abundance to mfaunal abundance and the ratio of total
biomass to total abundance. Abundance metrics were calculated as the total count of individuals for
each metric category per replicate. Richness metrics were calculated as the number of taxa for each
metric category per replicate. Proportional abundance metrics were calculated as the value of the
total count of individuals per replicate for each metric category divided by the total count of infaunal
individuals per replicate. Species diversity metrics were estimated using the Shannon-Wiener
diversity index (H') which is calculated as follows:
H'= ~ X
1=1
where/?, is the proportion of the ith species and S is the number of species. Dominance metrics were
estimated using Pielou's evenness index (J) which is calculated as follows:
H1
\QgiS
where //' is the diversity index and 5 is the total number of species collected. Diversity and
dominance metrics were calculated only for the total infaunal and epifaunal life history categories.
The assignment of species to life history and feeding categories was based on designations used for
the development of the B-IBI for Chesapeake Bay (Weisberg et al. 1997). Appendix A provides a
list of species designated as epifaunal while Appendices B-E provide lists of species belonging to
each feeding group used. Taxonomic category metrics were calculated using only infaunal species.
-------
D. Data Aggregation
1. Spatial Scales
Three spatial scales of aggregation were identified: (1) a Within Habitat scale, (2) a Within Salinity
Regimes scale, and (3) on a Baywide scale. Since estuarine macrobenthic community structure
varies in relation to salinity and sediment type, all sites were first classified into the seven habitat
types defined for the Chesapeake Bay by Weisberg et al. (1997)(Table 3) for the Within Habitat
scale for spatial aggregation. Our a priori expectation was that benthic community indicators
allowing discrimination between stress groups would be more effective if they were developed
separately for each major habitat type. For example, the tidal freshwater and polyhalme regions
have no species in common and higher level metrics based upon community characteristics might
also have better discriminatory abilities at the habitat spatial scale. The other two spatial aggregation
scales increased the number of samples available for developing any discriminant function.
2. Stress Categories
Sites were classified into stress groups using four aggregation schemes (Table 1). The maximum
number of stress groups was four: (1) a contaminant effect stress group (Contaminant), (2) a low
dissolved oxygen effect stress group (Low DO), (3) a combined contaminant and low dissolved
oxygen effect stress group (Combined), and 4) and a stress group of unknown source(s) (Unknown).
The criteria for inclusion in the Contaminant stress group was based on sediment quality guidelines
established for a suite of organic and metal contaminants known to adversely affect benthic
invertebrates. Three different criteria, presented below, were used and separately analyzed for
discriminant function development. A site was classified into Low DO stress group if dissolved
oxygen concentration at the time of collection was £.2 ppm. A site was classified into the Combined
stress group if it met both the Low DO criterion and the Contaminant criterion. Sites not classified
into either the Contaminant, Low DO or Combined stress groups were assigned to the Unknown
group.
3. Contaminant Stress Category Criteria
Three different sediment quality guideline (SQG) schemes were used. Each of the classification
schemes was based on sediment quality guidelines established for a suite of organic and metal
contaminants known to adversely affect benthic invertebrates. The first contaminant stress group
criterion used the Effects Range Median (ERM) values developed to represent concentrations at or
above which adverse toxic effects occur frequently (Long and Morgan 1990; Long et al. 1995). In
the first classification scheme, referred to as the ERM Exceedance classification scheme, a site was
assigned to the contaminant stress group if any of a suite of 24 sediment contaminants (Table 4)
detected at the site exceeded the ERM concentration for the contaminant as specified by Long et al.
(1995). Several of the analytes originally listed by Long et al. (1995) were not used in this study
because they were not measured at a large number of sites.
-------
The second and third classification schemes were based on mean sediment quality guideline (SQG)
quotients. This approach involves the calculation of the mean of ratios of individual contaminant
concentrations relative to their corresponding ERM values. The mean SQG quotients are then
compared to thresholds established for specific geographic regions (Hyland etal. in preparation).
For this study, two mean SQG quotients were used.
One SQG quotient value (SQV) was developed for the EMAP Virginian province which includes
all estuarine locations from Chesapeake Bay to Cape Cod. We used the median SGV value derived
from a frequency distribution plot. The plot included all sites where the benthic community
condition was declared degraded and at which no low dissolved oxygen effects occurred. The
Virginian Province median SQV value was 0.098. This threshold represents median SQG quotient
at or above which there is a high risk that benthic communities will be degraded within the Virginian
province (Hyland et al. in preparation).
The second SQG quotient value (SQV) was developed for the region encompassing the EMAP
Lousianian, Carolinian, and Virginian provinces combined and has a value of 0.044. The region
includes samples from the Gulf of Mexico and the eastern U.S. estuaries from north Florida through
Cape Cod. This threshold represents the median SQV value at or above which there is a high risk
that benthic communities will be degraded within all three provinces combined (Hyland et al., in
preparation) and was used to assign sites into the contaminant stress group for the classification
scheme referred to as All Province.
4. Level of Benthic Community Degradation
Only sites classified as either degraded (B-IBI < 2.6 but > 2.0) or severely degraded (B-IBI s2.0)
were retained for subsequent analysis. Our a priori expectation was that the most severely degraded
benthic community conditions might allow better discrimination between stress groups.
Consequently for each spatial scale and stress group combination discriminant functions were
developed for data using only the severely degraded sites and also using both severely degraded and
degraded sites. The latter data aggregation increased the number of samples available for
developing any discriminant function.
E. Spatial Scales and Analytical Scenarios
1. Within Habitat Scale
The first set of analytical scenarios, referred to as Within Habitat Type scenarios, were intended to
develop discriminant functions for six of the seven separate habitat types as defined by Weisberg
et al. (1997). For each habitat type, functions were created to discriminate between the Four Stress
Groups combination and for the Two Stress Groups combination of a Contaminant stress group and
all other stress groups combined. No attempt was made to develop functions for the polyhaline sand
habitat type because no sites within this habitat type were classified into the Contaminant stress
group regardless of the stress group classification scheme used.
-------
2. Within Salinity Regime Scale
The second set of analytical scenarios, referred to as Within Salinity Regime scenarios, were
intended to develop discriminant functions for three salinity regimes: (1) polyhaline (> 18 ppt), (2)
mesohaline (5-18 ppt), and (3) tidal freshwater/oligohalme combined (< 5 ppt). For the polyhaline
and mesohaline salinity regimes functions were created to discriminate: (1) between the Four Stress
Groups combination, (2) between the Three Stress Groups combination (Contaminant, Combined
and Unknown stress groups), (3) between the Two Stress Groups combination with a Contaminant
stress group and all other stress groups combined, and 4) between the Two Stress Groups
combination but without the Low DO stress group. For the tidal freshwater/oligohalme regime
discriminant functions were created only for Two Stress Group combination between the
Contaminant stress group and all other stress groups combined. Other scenarios for the tidal
freshwater/oligohaline salinity regime were not conducted because most sites were classified into
either the Contaminant or Unknown stress groups regardless of the classification scheme used.
3. Bay wide Scale
The final group of scenarios were attempts to develop discriminant functions that were applicable
to any habitat within Chesapeake Bay regardless of salinity regime or sediment type. Discriminant
functions for these scenarios were developed to discriminate between: all four possible stress groups;
the Contaminant,Unknown, and Combined stress groups; and, the Contaminant stress group and all
other stress groups combined both with and without the Low DO stress group.
When conducting a discriminant analysis if the number of variables approaches or exceeds a value
of n-\, where n is the total number of observations, the pooled sample variance-covariance matrix
will be singular and the resulting functions developed may not reliable (Khattree and Naik 2000).
For a number of the scenarios attempted, the number of variables relative to the total number of
samples in the calibration data set surpassed this theoretical limitation. Despite this problem, all
scenarios except those listed above were conducted in order to identify scenarios that could be
potentially useful if future studies generate sufficient data to produce more reliable discriminant
functions.
F. Discriminant Function Calibration and Validation
Linear discriminant function development and calibration procedures were conducted on a randomly
selected subset of each classified data set comprising two thirds of the total number of observations
for a given scenario. The number of discriminant functions required for the classification of
observations into stress groups was dependant upon the number of stress groups being classified for
each of the analytical scenarios. All discriminant analyses were conducted assuming proportional
prior probabilities of group membership. If the total percentage of correctly classified observations
was less than 75% for the calibration data set the discriminant functions developed were considered
inapplicable for the scenario. If the percentage of correctly classified observations for any of the
individual stress groups within the calibration data set was less than 70%, the discriminant function
-------
was considered inapplicable for the scenario.
Validity of the linear discriminant functions were tested by classifying the remaining third of the
data set into stress groups. Percentages of observations classified into each stress group using the
functions were compared with the known percentages in each stress category of the validation data
set. If the total percentage of correctly classified observations was less than 75% fora given scenario
then the discriminant function was considered inapplicable for the scenario. If the percentage of
correctly classified observations for any of the individual stress groups was less than 70%, the
discriminant function was considered inapplicable for the scenario. If the validation data set lacked
data for one or more of the stress groups, the discriminant functions developed were considered
inapplicable for the scenario under consideration.
G.. Salinity Correction
Salinity is an important environmental stressor that affects the composition and distribution of
benthic communities in estuaries. In an attempt to improve classification efficiency of the
discriminant functions, two additional runs of the Baywide scenarios were conducted using indicator
values from which the effect of natural variation due to salinity was removed. Pearson's correlation
coefficient was used to identify significant relationships between salinity and all of the indicators.
If a significant correlation between salinity and a given indicator was indicated (p <;0.01) and the
absolute value of the Pearson correlation coefficient was ;> 0.50 (Paul et al. 2001), a linear regression
analysis was employed to remove variance in the indicator due to salinity. For each of these
indicators, a linear regression equation was developed and predicted values for each indicator were
estimated based on the observed salinity. These predicted indicator values were subtracted from the
observed indicator values to obtain salinity corrected residuals. These residuals were then
substituted for the original values in the indicators data set and the discriminant function analysis
for the Baywide scenarios were rerun.
Significant relationships withr values aO.50 were found forpolychaetespecies richness,proportional
abundance of polychaetes, oligochaete species richness, proportional abundance of oligochaetes,
tubificid species richness, proportional abundance of tubificids, and species richness of deep deposit
feeders (Appendix G). Regression relationships developed for salinity correction of these parameters
are presented in Appendix H. Plots of residuals for two of these parameters, oligochaete species
richness and tubificid species richness, indicated a potential polynomial relationship with salinity.
Polynomial relationships for these two parameters are also provided in Appendix H.
H. Variable Reduction Approaches
Classification efficiencies of discriminant functions can be adversely affected if the number of
variables is large relative to the number of observations in the data set used (Huberty 1994; Khattree
and Naik 2000). For those scenarios considered applicable to a given management scenario, an
attempt was made to simplify the function and improve classification efficiencies by reducing the
number of variables used. A variety of techniques are typically employed to select variables for
-------
linear discriminant function analysis; however, there is little agreement in the literature as to the
validity and relative efficacy of different approaches (McLachlan 1992; Huberty 1994; Khattree and
Naik 2000).
For this study, two separate variable selection approaches were attempted. The first approach
involved the use of a stepwise discriminant analysis using a stepwise selection method with an F-test
selection criterion of 0.15 (Khattree and Naik 2000). Applicable scenarios were conducted again
using this reduced variable set. The second approach involved testing for variables that were
significantly different between stress groups using an ANOVA. Applicable scenarios were
conducted again using only those variables which were significantly different between stress groups
at p s 0.05. Similar approaches have been effectively used as a variable reduction technique
(Huberty 1994).
All statistical analyses were conducted using SAS/Base® and SAS/Stat® v. 8.1 statistical software.
Correlation, linear discriminant function and regression analyses were conducted using the CORR,
DISCRIM, and GLM procedures, respectively (SAS Institute 1990a,b). Stepwise discriminant
analyses and ANOVA's were conducted using the STEPDISC and ANOVA procedures (SAS
Institute 1990a,b).
III. Results
A. Description of Database
A'total of 608 sampling event/location combinations were compiled from 1,450 replicate
biologicalsamples collected throughout Chesapeake Bay and its tributaries. Most of these data were
generated by the EPA's EMAP and MAIA Programs (Table 5). Thirteen of these sampling
event/location combinations were repeat visits to the same location. A total of 268 (44%)
observations were classified as either degraded or severely degraded based on the mean B-IBI
values. Approximately 45% were classified as meeting benthic restoration goals and approximately
11% were classified as marginal. The mean B-IBI value across all sites was 2.76 and ranged from
1.58 at severely degraded sites to 3.61 at sites that met benthic restoration goals (Table 6). Of the
observations classified as degraded or severely degraded, 12 were eliminated due to a lack of
sufficient dissolved oxygen and/or contaminant concentration data leaving a reduced databaseof 256
observations for all subsequent analyses (Table 5).
More than 30% of the sites m the reduced database were found in high mesohaline muds while
polyhalme sands had the fewest number (*4%) of sites. The polyhaline mud, oligohaline and tidal
freshwater habitat types had approximately equal numbers of sites. For most habitat types, the
number of severely degraded sites was greater than the number of degraded sites (Table 7).
s,
The number of contaminants exceeding the ERM concentration across all sites was 0.19 ± 0.68
(mean ± standard deviation) with a maximum of six contaminants exceeding ERM concentrations
at a single site. The two contaminants with the highest number of observations exceeding the ERM
8
-------
were zinc and total DDTs which were higher than the ERM concentration at twelve and nine sites,
respectively. A total of 13 contaminants including arsenic, copper, acenaphthene, acenaphthylene,
anthracene, benzo[a]anthracene, benzo[a]pyrene, chrysene, dibenz[a,h]anthracene, fluoranthene,
2-methy (naphthalene, naphthalene, phenanthrene, did not exceed ERM concentrations at any of the
severely degraded and degraded sites. The mean SQV quotient for severely degraded and degraded
sites ranged from 0.002 to 2.87 with an average mean SQV quotient of 0.111 ± 0.204 (mean ±
standard deviation). Appendix F provides a listing of number of contaminants exceeding the ERM
concentration, the mean SQG quotient, and the number missing analytes for each station date
combination in the reduced database.
Based on the ERM classification scheme, nearly 75% of sites were classified into the Unknown
stress group. Most of the remaining sites were classified into either Contaminant or Low D.O. effect
sites (Table 8). The majority of sites in each habitat type was classified into the Unknown stress
group (Table 9). The highest number of Contaminant stress group sites was found in the oligohaline
habitat type while the highest number of Low D.O. stress group sites occurred in high mesohaline
muds. The maximum number of Combined stress group sites was found in the low mesohaline
habitat type. No sites in the high mesohaline and polyhalme sand habitat types were classified into
the Contaminant or Combined stress groups. No Low D.O. sites were identified in the oligohaline
and tidal freshwater habitat types.
Using the Virginian Province SQV, nearly 59% of sites were classified into the Unknown stress
group (Table 8). Most of the remaining sites were classified into the contaminant stress group. The
majority of sites in each habitat type was classified into the Unknown stress group except for the low
mesohaline and oligohaline habitat types (Table9). The maximum number of Contaminant stress
group sites was found in the oligohaline habitat type while no Contaminant stress group sites were
found in the polyhaline sand and high mesohaline sand habitat types. The maximum number of Low
DO stress group sites was found in the high mesohaline mud habitat type while the oligohaline and
tidal freshwater had no Low DO stress group sites. The maximum number of Combined stress group
sites was found in the low mesohaline habitat type while no Combined stress group sites were found
in the polyhalme sand, high mesohaline sand and oligohaline habitat types. The high mesohaline
mud habitat type had the highest number of Unknown stress group sites.
Using the All Province SQV resulted in an increase in the number of Contaminant and Combined
effect sites, primarily as a result of a decrease in Unknown effects sites (Table 8). The majority of
sites in each habitat type was classified as Contaminant effect sites except for the high mesohaline
sand and polyhaline sand habitat type where the majority of sites were classified as Unknown effect
sites (Table 9). The maximum number of Contaminant effect sites was found in the high mesohaline
mud habitat type while no Contaminant effect sites were found in the polyhaline sand habitat type.
The maximum number of Low DO sites across habitat types was three, and three habitat types (low
mesohaline, oligohaline, and tidal freshwater) had no Low DO effect sites. The maximum number
of Combined effect sites was found in the low mesohaline habitat type while the polyhalme sand,
high mesohaline sand and oligohaline habitats had no Combined effect sites. The high mesohaline
sand habitat type had the most Unknown effect sites.
-------
B. Within Habitat Type Scale
None of the Within Habitat Type scenarios had a sufficient sample size for discriminant function
development. The High Mesohaline Mud habitat type, when using both degraded and severely
degraded sites, had the highest number of samples available for the calibration data set - 57 sites;
however, 63 sites were necessary. The next highest sample number was the Low Mesohaline
habitat type with 31 samples. Correct classification rates are presented below even though the
sample size was inadequate.
1. All Four Stress Groups
None of these scenarios met criteria for applicability based upon correct classification rates (Table
10) due to low classification efficiencies in the validation data sets and missing values in individual
stress groups. No attempts were made to reduce variable sets for these scenarios. Use of the
discriminant functions developed for these scenarios is not recommended.
2. Contaminant vs All Other Stress Groups
Overall classification efficiencies for the calibration data sets for these scenarios were 100%.
Overall classification efficiencies for the validation data sets exceeded 75% for several scenarios
(Table 11) but only two had high (a75%) stress group specific classification efficiencies and had
more than one observation in the Contaminant stress group for the validation data sets. These two
scenarios included the All Province Polyhaline Mud scenario for severely degraded and degraded
sites; and the All Province High Mesohaline Mud scenario for severely degraded sites (Table 11)
However, the number of variables relative to sample size exceeded the theoretical limitation for
discriminant analysis and as a result the classification efficiencies obtained may be unrealistic and
these discriminant function may not accurately classify new observations. Variable reduction
procedures were attempted for these scenarios but resulted in lower overall classifications
efficiencies for validation data sets (Table 23). Although the use of the discriminant functions for
these scenarios is not recommended at present, the high classification efficiencies for the calibration
data obtained for some of these scenarios suggest that the use of additional data could result in
discriminant functions that might be applicable.
C. Within Salinity Regime Scale
Only the Mesohaline salinity regime had a sufficient sample size for discriminant function
development. For the Polyhaline and combined Tidal Freshwater and Oligohaline regimes the
maximum number of available sites for the calibration data set were 34 and 49, respectively when
using both degraded and severely degraded sites. The minimum number of sites was 63. Correct
classification rates are presented below even when the sample size was inadequate.
1. Polyhaline
10
-------
a. All Four Stress Groups
None of these scenarios met criteria for applicability due to low classification efficiencies or missing
values in some of the stress groups in the validation data set (Table 12). No attempts were made to
reduce variable sets for these scenarios. Use of the discriminant functions developed for these
scenarios is not recommended.
b. Three Stress Groups with no Low DO sites - Contaminant, Combined and Unknown
None of these scenarios met criteria for applicability due to low classification efficiencies or missing
values in some of the stress groups in the validation data set (Table 13). No attempts were made to
reduce variable sets for these scenarios. Use of the discriminant functions developed for these
scenarios is not recommended.
c. Contaminant vs All Other Stress Groups
Although overall classification efficiencies for the calibration data sets were 100% and overall
classification efficiencies for the validation data sets exceeded 75% for 7 out of 12 of these
scenarios, classifications for individual stress groups were generally low with one exception: the All
Province scenario with the Low D.O. stress group for severely degraded and degraded sites which
had classification efficiencies of 80% for both stress groups (Table 14). The number of variables
relative to sample size exceeded the theoretical limitation for discriminant analysis and as a result
the classification efficiencies obtained may be unrealistic and the discriminant function may not
accurately classify new observations. Variable reduction approaches resulted in a decrease in
classification efficiency for this scenario (Table 23). The use of this discriminant function is not
recommended at present; however, the high classification efficiencies obtained suggest that the use
of additional data could result in a discriminant function that might be applicable to this scenario.
2. Mesohaline
a. All Four Stress Groups
Overall classification efficiencies for the calibration data sets were high ranging from 92% to nearly
99% but none of the scenarios had overall classification efficiencies above 58% for the validation
data sets. As a result, none of these scenarios met criteria for applicability (Table 15). No attempts
were made to reduce variable sets forthese scenarios. Use of the discriminant functions developed
for these scenarios is not recommended.
b. Three Stress Groups with no Low DO sites - Contaminant, Combined and Unknown
Although overall classification efficiencies for the calibration data sets for these scenarios ranged
from 93% to 100%, overall classification efficiencies for the validation data sets were low ranging
from approximately 19% to a maximum of 66% (Table 16). Implementation of discriminant
11
-------
functions Tor these scenarios is not recommended.
c. Contaminant vs. All Other Stress Groups
Calibration data set overall classification efficiencies ranged from 93% to 100% for the scenarios
with Low DO sites and from 96% to 100% for the scenarios without Low DO sites (Table 17). The
discriminant function for the All Province SQV without Low DO sites scenario and for severely
degraded sites had the highest overall classification efficiency for the validation data set (79%).
Stress group specific classification efficiencies were 82% and 75% for the Contaminant and Other
stress groups, respectively. Validation data set classification efficiencies within habitat type forthis
function were > 80% for the High Mesohaline Mud and Low Mesohaline habitat types but < 30%
for the High Mesohaline Sand habitat type (Figure 1). Both variable reduction approaches for this
scenario resulted in a decrease in overall and within stress group classification efficiencies (Table
23). The function for this scenario met the criteria for applicability and could be implemented.
3. Tidal Freshwater/Oligohaline
Although overall classification efficiencies for the calibration data sets for these scenarios were
always at or above 90%, overall classification efficiencies or stress group specific classification
efficiencies for the validation data sets were too low to meet the criteria for applicability (Table 18).
Poor classification efficiencies of the validation data set were probably the result of low numbers
of observations for these scenarios. Implementation of the discriminant functions developed for
these scenarios is not recommended.
D. Baywide Scale
1. All Four Stress Groups
Although the calibration data set overall classification efficiency for these scenarios ranged from
78% to 96%, overall classification efficiencies for the validation data set did not meet criteria for
applicability ranging from 39% to 66% (Table 19). Neither of the salinity correction approaches
used resulted in classification efficiencies that met the criteria for applicability (Table 19).
Implementation of discriminant functions developed for these scenarios is not recommended.
2. Three Stress Groups with no Low DO sites - Contaminant, Combined and Unknown
Overall classification efficiencies for the calibration data sets ranged from nearly 82% to nearly 98%
(Table 20). However, the overall classification efficiencies for the validation data sets were less than
70% for all scenarios except two: (1) the ERM Exceedance scenario for degraded and severely
degraded sites, and (2) the All Province SQV scenario for severely degraded and degraded sites.
Although overall classification efficiencies were above 70% for these scenarios, classification
efficiencies for some stress groups were less than or equal to 50%. Salinity correction procedures
did not improve and generally reduced the classification efficiencies of the discriminant functions
12
-------
for these scenarios (Table 20). Implementation of the discriminant functions developed for these
scenarios is not recommended.
3. Contaminant vs. All Other Stress Groups
a. With Low DO Sites
Overall classification efficiencies for the calibration data sets ranged from 78% to 100% while
overall classification efficiencies for the validation data sets ranged from approximately 49% to just
over 83% (Table 21). The All Province SQV scenario for severely degraded and degraded sites had
an overall classification efficiency of 75% and the best classification efficiencies for individual stress
groups (82% for the Contaminant stress group and 68% for the Other stress group). Classification
efficiencies within habitat types for this scenario were > 75% for five of the seven habitat types
(Figure 2). Salinity correction procedures did not improve overall classification efficiencies forthe
calibration or validation data sets for any of the scenarios (Tables 21). Neither of the variable
reduction approaches improved the classification efficiencies of this function (Table 23). The All
Province SQV scenario for severely degraded and degraded sites without salinity correction met the
criteria for applicability. This discriminant function could be implemented to identify potentially
contaminated sites.
b. Without Low DO Sites
Overall classification efficiencies for the calibration data sets ranged from 90% to 100% while
within stress group classification efficiencies were > 75% (Table 22). Although overall
classification efficiencies for the validation data sets for half the scenarios were above 70%,
classification efficiencies for at least one stress group were always less than 70% (Table 22).
Salinity correction procedures did not improve overall classification efficiencies for the calibration
or validation data sets for any of the scenarios attempt.
IV. Discussion
A. Overview of Results
Regardless of the spatial scale under consideration, discriminant functions developed for more than
two separate stress groups had very poor classification efficiencies for either the validation data sets
or both the calibration and validation data sets. As a result, none of discriminant functions developed
to discriminate between three or four potential stress groups should be implemented. Poor
classification efficiencies for these scenarios were due primarily to low numbers of observations
within individual stress groups.
The only Within Salinity Regime discriminant function that met criteria for applicability was for the
Mesohaline salinity regime, using two stress groups and severely degraded sites only (excluding
Low DO sites) and using the All Province SQV contaminant classification scheme (Table 17).
13
-------
Implementation of this discriminant function is not recommended unti 1 functions for the other habitat
type combinations can be successfully validated.
The discriminant function for one Baywide scenario met the criteria for use in identifying potential
sources of stress: the Contaminant versus Others stress groups (with Low DO sites) using the All
Province SQV contaminant criterion for severely degraded and degraded sites without a salinity
correction. This particular function is capable of discriminating contaminated sites from sites
affected by all other potential sources of stress in any of the seven habitat types.
B. Usage Constraints
The characteristics of the data sets used in this study and statistical techniques employed put certain
constraints on how the tool should be used and how results of subsequent classification analyses
should be interpreted. The diagnostic tool developed provides a means to assign new observations
to one of two groups of potential sources of stress and assign a probability of group membership to
each new observation. The discriminant function coefficients used to make these assignments were
developed based on the distributional, variance-covarianceand correlation structure of the predictor
variables in calibration data set. In effect, new observations are assigned to stress groups based on
their similarity to observations in the two stress categories in the calibration data set.
The calibration data set was taken from benthic biological data sets collected under a set of specific
conditions which affects the underlying data structure of the predicator variables. As a result, new
observations can be classified into stress categories only if they meet these conditions. Since the
functions were developed using samples collected within Chesapeake Bay and its tnbutaries,
samples collected outside of these geographical boundaries should not be classified using these
functions. Since the functions were developed using samples collected with a Young grab and
different sampling gear have inherent properties that affect estimates of various biological variables
(Word 1975,1976; Ewmg et al. 1988), samples collected using any gear type other than a Young
grab cannot be classified using these functions. All observations used in this study were collected
during the B-IB1 index period (July 15 through September 30). No attempt should be made to
classify into stress groups new observations that are not collected during the index period. The
calibration data set contained only observations that had been previously classified as either
degraded or severely degraded using the Chesapeake Bay Program Index of Biotic Integrity. No
attempt should be made to classify into stress groups new observations that have not been previously
classified as degraded or severely degraded by the B-IBI.
It is possible that characteristics of Contaminant stress group in the calibration data do not reflect
the characteristics of all of the potentially contaminated sedimentary environments found in
Chesapeake Bay. The number of contaminants used in contaminant classification schemes was
limited to a total of 8 metals and 16 organic compounds. As a result, the Contaminant stress group
for the calibration data sets may not include some samples that were, in fact, affected by
anthropogenic contaminants not included in the list used by this study. Therefore, it is possible that
a new observation could be classified into the Other category despite the presence of anthropogenic
14
-------
sediment contaminants. Assigning group membership to new observations using discriminant
function is always accompanied by the risk of mis-classifying the new observations. For the case
of the diagnostic tools developed for this study, the classification efficiencies of the validation data
sets can be used to estimate the risk of mis-classifying new observations. For the Baywide
diagnostic tool, the risk of mis-classifying a newobservation wouldbe approximately 25%. Because
of these limitations the diagnostic tool developed cannot be used to definitively assign new
observations to the contaminant stress group or not without independent and direct measurement of
sediment contaminant concentrations. The tool developed should be used exclusively as a screening
tool to identify sites or regions with a high probability of sediment contamination that should be
targeted for further study. Posterior probabilities of group membership could be used to prioritize
sites with respect to the need for conducting additional studies to identify and quantify sediment
contaminants. Sites with the highest posterior probability of group membership in the Contaminant
stress group would warrant the highest priority for additional investigations.
C. Technical Approaches to Implementation
From a technical standpoint, discriminant functions could be implemented using a variety of
techniques. The simplest method would be to create a spreadsheet containing formulae to multiply
the linear discriminant coefficients with values for each of the bioindicators for each observation
being classified. The resulting transformed values would be summed together to produce the
discriminant score for each observation. These discriminant scores would then be compared to the
cutoff value for the function. The primary advantage to this approach is that users would not be
required to have specialized computer programming skills to use the functions. The disadvantage
is that entry of formulae and bioindicators into spreadsheets would be tedious, labor intensive and
prone to data entry errors. In addition, this approach does not provide posterior probabilities of
stress group membership for new observations. Table 24 provides the linear discriminant
coefficients for the function recommended for implementation along with the cutoff values used to
determine stress group membership. Values below the cutoff values are classified into the
Contaminant stress group while values above the cutoff are classified into the Other stress group for
the Baywide function.
The use of SAS statistical programming language would appear to be the most efficient means to
implement the diagnostic tool provided the user is familiar with this application. To classify new
sites into stress groups using SAS would require the user to: (1) have access to copies of the original
calibration data sets used for this study, (2) create a SAS format data set containing the new
observations with the same format as that of the calibration data sets, and (3) be familiar with and
able to interpret output from the SAS DISCRIM procedure. A copy of the calibration data set along
with SAS programs for conducting a discriminant analysis are provided on the diskette attached to
this report to assist users in implementing the diagnostic tools. Using SAS programs would combine
relative ease of use in combination with the detailed output provided by this statistical package.
Other programming languages such as Visual Basic or C" could be used to create programs for
calculating discriminant scores and comparing them to the cutoff values and for calculating posterior
15
-------
probabilities. Such programs could be written to perform the same operations as SAS programs but
the user would be required to have not only computer programming language skills but would need
an extensive knowledge of multivariate statistics. A typical user would find this approach time
consuming and difficult to implement.
D. Recommendations
Prior to implementation, it is recommended that operational effectiveness of the diagnostic tools be
further tested using additional validation data sets. A variety of benthic community data sets exist
that do not include sediment contaminant data and, therefore, could not be included in our calibration
and validation data sets. For example, since 1996, the entire tidal Chesapeake Bay has been sampled
using a stratified random procedure (Llanso et al. 2002). The Bay is divided into ten strata and
within each stratum 25 random locations are sampled for a total of 250 random locations each year.
Sites with degraded benthic community condition could be putatively placed into stress categories
for further validation. In addition, this large random data set could be reviewed to generate
additional data to (1) attempt to develop discriminant functions including additional stress groups,
e.g., a Low DO stress group and (2) possibly provide an adequate sample size for discriminant
function development for some of the spatial scales below the Baywide scale. Other data sets from
areas known to have sediment contaminant problems but not meeting our data inclusion criteria
could provide additional validation data sets. For example, Dauer and Llanso (2002) present data
from 125 randomly selected locations sampled for benthic community condition in 1999 in the
Elizabeth River watershed.
All diagnostic tools implemented should be periodically "re-calibrated" as new benthic biological
data sets with associated contaminants data become available. Two of the Within Habitat Type and
two of the Within Salinity Regime functions showed promise and efforts to update and validate these
functions should be attempted if additional data become available. If and when the diagnostic tools
described are implemented for regular use by the Chesapeake Bay management community, they
should be employed with all usage constraints as described above.
VI. Literature Cited
Aller, R.C., 1978. Experimental studies on changes produced by deposit feeders on pore water,
sediment and overlying water chemistry. Amer. J. Sci. 278:1185-1234.
Aller, R.C., 1980. Relationships of tube-dwelling benthos with sediment and overlying water
chemistry, pp. 285-308, In Marine Benthic Dynamics, K..R. Tenore and B.C. Coull, eds., University
of South Carolina Press, Columbia SC.
Beukema, J.J., 1991. Changes in composition of bottom fauna of a tidal flat area during a period of
eutrophication. Mar. Biol. 111:293-301.
Bilyard, G.R., 1987. The value of benthic infauna in marine pollution monitoring studies. Mar.
16
-------
Poll. Bull. 18:581-585.
Botton, M.L., 1984. Effects of laughing gull and shorebird predation on the intertidal fauna at Cape
May, New Jersey. Estuar. Coast. Shelf Sci. 18:209-220.
Boesch, D.E., 1972. Species diversity of marine macrobenthos in the Virginia area. Ches. Sci.
13:206-211
Brown, J.R., R.J. Gowen and D.S. McClusky, 1987. The affect of salmon farming on the benthos
of a Scottish sea loch. J. Exp. Mar. Biol. Ecol. 109:39-51.
Christman, C.S. and D.M. Dauer 2002. Development of a diagnostic tool to determine the cause of
benthic community degradation in the Chesapeake Bay. Environmental Monitoring and Assessment.
IN PRESS.
Cleorn, J.E., 1982. Does the benthos control phytoplankton biomass in south San Francisco bay?
Mar Ecol. Prog. Ser., 9:191-202.
Cohen, R., P. Dresler, E. Philips, and R. Cory, 1984. The effect of the Asiatic clam Corbicula
fluminea, on phytoplankton of the Potomac River, Maryland. Limnol. Oceanogr., 29:170-180
Dauer, D. M., 1993. Biological criteria, environmental health and estuanne macrobenthiccommumty
structure. Mar. Poll. Bull. 26:249-257.
Dauer, D.M., 1997. Dynamics of an estuanne ecosystem: Long-term trends in the macrobenthic
communities of the Chesapeake Bay, USA (1985-1993), Oceanologica Acta, 20: 291-298.
Dauer, D.M. 1999. Baywide benthic community condition based upon 1997 random probability
based sampling and relationships between benthic community condition, water quality, sediment
quality, nutrient loads and land use patterns in Chesapeake Bay. Final report to the Virginia
Department of Environmental Quality. 18 pages plus Appendix.
Dauer, D. M. and R. W. Alden, 1995. Long-term trends in the macrobenthos and water quality of
the lower Chesapeake Bay (1985-1991). Mar. Poll. Bull. 30:840-850.
Dauer, D.M., R.M. Ewing, G.H. Tourtellotte, W.T. Harlan, J.W. Sourbeer, and H. R. Barker Jr.
1982. Predation pressure, resource limitation and the structure of benthic infaunal communities.
Internationale Revue der gesamten Hydrobiologie 67: 477-489.
Dauer, D.M., M. F. Lane, H.G. Marshall, K.E. Carpenter and R.J. Diaz. 1999. Status and trends in
water quality and living resources in the Virginia Chesapeake Bay: 1985-1998. Final report to the
Virginia Department of Environmental Quality. 65 pages.
17
-------
Dauer, D.M. and R. J. Llanso. 2002. Spatial scales and probability based sampling in determining
levels of benthic community degradation in the Chesapeake Bay. Environmental Monitoring and
Assessment. In Press.
Dauer, D.M., M.W. Luckenbach, M.W. and A.J. Rodi, Jr., 1993. Abundance biomass comparison
(ABC method): effects of an estuarine gradient, anoxic/hypoxic events and contaminated sediments,
Marine Biology, 116:507-518.
Dauer, D.M., H.G. Marshall, K.E. Carpenter, M. F. Lane, R.W. Alden, III, K.K. Nesius and L.W.
Haas. 1998. Virginia Chesapeake Bay water quality and living resources monitoring programs:
Executive Report, 1985-1996. Final report to the Virginia Department of Environmental Quality.
28 pages.
Dauer, D.M., J.A. Ranasinghe, and Rodi, Jr., 1992. Effects of low dissolved oxygen levels on the
macrobenthos of the lower Chesapeake Bay, Estuaries, 15:384-391.
Dauer, D.M. and A.J. Rodi, Jr. 1999. Baywide benthic community condition based upon 1998
random probability based sampling. Final report to the Virginia Department of Environmental
Quality. 126pp.
Dauer, D.M. and A.J. Rodi, Jr. 2001. Baywide benthic community condition based upon 1999
random probability based sampling. Final report to the Virginia Department of Environmental
Quality. 154 pp.
Dauer, D.M, Weisberg S.B. and J.A. Ranasinghe, 2000. Relationships between benthic community
condition, water quality, sediment quality, nutrient loads, and land use patterns in Chesapeake Bay.
Estuaries, 23:80-96.
Davis, W.S. and T.P. Simon, 1995. Biological Assessment and Criteria, tools for water resource
planning and decision making. Lewis Publishers, New York, NY. 415pp.
Engle, V.D., J.K. Summers, and G.R. Gaston, 1994. A benthic index of environmental condition
of Gulf of Mexico estuaries. Estuaries 17:372-384.
Engle, V.D. and J.K.. Summers, 1999. Refinement, validation, and application of a benthic condition
index for the northern Gulf of Mexico. Estuaries 22:624-635.
18
-------
Ewmg R.M., J.A. Ranasinghe and D.M. Dauer. 1988. Comparison of five benthic sampling devices.
Applied Marine Research Laboratory Technical Report. Report to the Virginia State Water Control
Board 84 pp.
Gaston, G.R. and J.C. Young, 1992. Effects of contaminants of macrobenthiccommumties in the
upper Calcasieu Estuary, Louisiana. Bull. Environ. Contam. Toxicol. 49:922-928.
Flint, R.W. and D. Kamykowski, 1978. Benthic nutrient regeneration in south Texas coastal waters.
Estuar. Coast. Shelf Sci. 18:221-230.
Holland, A., N. Mountford, M. Heigel, D. Cargo, and J. Mihursky, 1980. Influence of predation on
infaunal abundance in the upper Chesapeake Bay. Mar. Biol., 57:221-235.
Huberty, C.J., 1994. Applied discriminant analysis. Wiley-Interscience. New York, N.Y. pp 446.
Hyland, J.L., W.L. Balthis, V.D. Engle, E.R. Long, J.F. Paul, J.K. Summers, and R.F. Van Dolah.
In Press, Incidence of stress in benthic communities along the U.S. Atlantic and Gulf of Mexico
coasts within different ranges of sediment contamination from chemical mixtures. Environmental
Monitoring and Assessment Special Issue on EMAP Symposium 2001: Coastal Monitoring
Through Partnerships, April 24-27, 2001, Pensacola, FL.
Johnson, R.A. and D.W. Wichern, 1998. Applied multivariate statistical analysis. 4lh Edition.
Prentice Hall, Upper Saddle River, N.J. 816 pp.
Karr, J.R. and E.W. Chu, 1999. Restoring life in running waters. Island Press Washington D.C. 206
pp.
Kachigan, S.K., 1991. Multivariate statistical analysis. 2nd Edition. Radius Press. New York, N.Y.
303 pp.
Khattree, R. and D.N. Naik, 2000. Multivariate data reduction and discrimination with SAS
software. SAS Institute Inc. Cary,N.C. 574pp.
Llanso, R.J., D.M. Dauer, J.H. Volstad, and L.S. Scott. 2002. Application of the Benthic Index of
Biotic Integrity to environmental monitoring in Chesapeake Bay. Environmental Monitoring and
Assessment. In Press.
Llanso, R. J., In Review. An estuarine benthic index of biological integrity for the Mid-Atlantic
region of the United States. II. Index development. To be submitted to Estuaries.
Long, E.R., D.D. MacDonald, S.L. Smith and F.D.Clader, 1995. Incidence of adverse biological
effects within ranges of chemical concentration in marine and estuarine sediments. Environ. Man.
19:81-95.
19
-------
Long, E.R., and L.G. Morgan, 1990. The potential for biological effects of sediment-sorbed
contaminants tested in the National Status and Trends Program. NOAA Technical Memorandum
NOS OMA 52, 175 pages, Two appendices. United States Department of Commerce, National
Oceanic and Atmospheric Administration, Seattle, Washington.
McGee, B.L., D.J. Fisher, J. Ashley, and D. Velinsky, 2001. Using the sediment quality triad to
chaaracterize toxic conditions in Chesapeake Bay (199). Report to the U.S. Environmental
Protection Agency. IN PRESS.
McLachlan.G.J. 1992. Discriminant analysis and statistical pattern recognition. Wiley-Interscience,
New York, N.Y. 526 pp.
Nichols F., 1985. Increased benthic grazing: An alternative explanation for low phytoplankton
biomass in Northern San Francisco Bay during the 1975-1977 drought. Estuar. Coast. Shelf Sd.
21:379-388
Officer, C.B., A.J. Smayda and R. Mann, 1982. Benthic filter feeding: A natural eutrophication
control. Mar. Ecol. Prog. Ser., 9:203-210.
Paul, J.F., K.J.Scott, D.E. Campbell, J.H. Gentile, C.S. Strobel, R.M. Valente, S.B. Weisberg, A.F.
Holland, J.A. Ranasinghe, 2001. Developing and applying a benthic index of estuarine condition
for the Virginian Biogeographic Province. Ecological Indicators. 1:83-99.
Pearson, T. H. and R. Rosenberg, 1978. Macrobenthic succession in relation to organic enrichment
and pollution of the marine environment. Oceanogr. Mar. Biol. Ann. Rev. 16:229-311.
Quammen, M.L., 1984. Predation by shorebirds, fish, and crabs on invertebrates in mtertidal
mudflats: An experimental test. Ecology, 65:529-537.
Reish, D., 1973. The use of benthic animals in monitoring the marine environment. J. Environ.
Plan. Poll. Cont. 1:32-38.
Rhoads, D.C., 1973. The influence of deposit-feeding benthos on water turbidity in nutrient cycling.
Amer.J.Sci. 271:1-22.
Rhoads, D.C. and D.K.. Young, 1970. The influence of deposit-feeding organisms on sediment
stability and community trophic structure. J. Mar. Res. 28:150-178.
Rowe, G.T., C.H. Clifford, K..L. Smith and P.L. Hamilton, 1975. Benthic nutrient regeneration and
its coupling to primary productivity in coastal waters. Nature 255:215-217.
SAS Institute Inc., 1990a. SAS/STAT® User's Guide, Volume 1, .ACECLUS-FREQ. SAS Institute
Inc., Gary N.C. pp. 890.
20
-------
SAS Institute Inc., 1990b. SAS/STAT® User's Guide Volume 2 GLM-VARCOMP. SAS Institute
Inc., Gary N.C. pp. 1686.
Versar Inc., 2001. Chesapeake Bay Water Quality Monitoring Program, benthic monitoring
component quality assurance project plan: 2001-2002. Versar Inc., Columbia MD. Report to the
Maryland Department of Natural Resources Tidewater Resource Assessment. Annapolis MD.
Vimstein, R.W., 1977. The importance of predation by crabs and fishes on benthic infauna m
Chesapeake Bay. Ecology, 58:199-217.
Virnstein, R.W., 1979. Predation on estuarine infauna: Response patterns of component species.
Estuaries 2:69-86.
Word.J.Q., 1975. A comparison of grab samplers. California Coastal Water Research Project 1975
Annual Report, pp. 63-66.
Word, J.Q., 1979. An evaluation of benthic invertebrate sampling devices for investigating feeding
habits of fish. Proceedings of the 1SI Pacific Northwest Technical Workshop, pp. 43-55.
Weisberg, S.B., J.A. Ranasinghe, D.M. Dauer, L.C. Schaffner, R.J. Diaz and J.B. Fnthsen. 1997.
An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries, 20: 149-158.
Zeitzschel, B., 1980. Sediment-water interactions in nutrient dynamics, pp 195-218. In Marine
Benthic Dynamics, K.R. Tenoreand B.C. Coull, eds., University of South Carolina Press, Columbia
SC.
21
-------
Figures
-------
All Province Contaminant Classification Scheme
Severely Degraded Mesohaline Sites Only
Contaminant vs. Others - Without Low D.O. Sites
36
c
OJ
u
High Mesohaline Mud
High Mesohaline Sand
Habitat Type
Low Mesohaline
Figure 1.
Calibration j Validation
Discriminant function classification efficiencies for individual habitat types for
classifying Mesohaline severely degraded sites (excluding Low D.O. sites) into
the Contaminant and Other stress groups. Numbers above the bars indicate the
number of observations within each habitat type.
22
-------
100.00-1
II.IHI
All Province Contaminant Classification Scheme
Severely Degraded and Degraded
Contaminant vs. Others - With Low D.O. Sites
Tidal Freshwater Olignhulinr l.»n Mf.solialinc High Mcsohalinc Ilittli Mcsolialinc Polyhalinc Mud I'olylialino Sand
Mud Sand
Habitat Type
Calibration 3 Validation
Figure 2. Discriminant function classification efficiencies for individual habitat types
for the Baywide discriminant function for classifying severely degraded and
degraded sites (including Low D.O. sites) into the Contaminant and Other
stress groups. Numbers above the bars indicate the number of observations
within each habitat type.
23
-------
Tables
-------
Table 1 Data Aggregation schemes used in analyses. For definition of habitat types see Table
2. Stress categories are defined in the text.
A. Spatial Scale
Within Habitat
Tidal Freshwater
Oligohaline
Low Mesohaline
High Mesohaline Sand
High Mesohaline Mud
Polyhalme Sand
Polyhaline Mud
Within Salinity Regime
Tidal Freshwater/Oligohaline
Mesohaline
Polyhaline
Baywide
B. Stress Categories
Four Stress Groups
Contaminant
Low DO
Combined
Unknown
Three Stress Groups
Contaminant
Combined
Unknown
Two Stress Groups
Contaminant
Others
C. Contaminant Stress Group Criterion
ERM Exceedance
Virginian Province ERM quotient
All Province ERM quotient
D. Level of Benthic Community Degradation
Severely degraded and degraded B-IBI < 2.6
Severely degraded B-IBI s 2.0
24
-------
Table 2. Candidate metrics used for analytical tool development. An asterisk indicates that
a given metric for the category listed was included in the analytical tools.
Relative Species
Metric Categories Abundance Richness Abundance Diversity Dominance Biomass
Taxonomic Categories
Isopoda ***...
Amphipoda ***...
llaustorndae ***..-
Ampeliscidae ***...
Corophndae ***...
Mollusca ***-.-
Bivalvia ***..-
Gastropoda ***..-
Polychaeta ***...
Spionidae ***...
Capitellidae *••---
Nereidae ••*..-
Oligochaeta ***-.-
Tubificidae • ••---
Life History Categories
Infaunal species **.**.
Epifaunal species *****.
Infaunal and epifaunal species _.---*
Trophic Categories
Deep Deposit feeder •**...
Suspension feeder ***.--
Interface feeder **•-..
Carnivore/Omnivore ***.--
25
-------
Table 3. Habitat types for the Chesapeake Bay B-IBI as defined by Weisberg et al. (1997).
(N/A: Not applicable)
Bottom Salinity Silt/Clay (<63n)
Habitat (ppt) Content by Weight (%)
Tidal Freshwater
Oligohaline
Low Mesohalme
High Mesohalme Sand
High Mesohalme Mud
PolyhalmeSand
Polyhalme Mud
0-05
05-5
5-12
12-18
12-18
> 18
> 18
N/A
N/A
N/A
0-40
>40
0-40
>40
Table 4. ERM guidelines for 24 trace metals (ppm dry wt) and organic compounds (ppb, dry
wt) as defined from Long et al. (1995).
Effects Range Median
Concentration
Trace Metals
Arsenic 70
Cadmium 9 6
Chromium 370
Copper 270
Lead 218
Mercury 0 71
Silver 3 7
Zinc 410
Organic Compounds
Acenaphthene 500
Accnaphthylenc 640
Anthracene 1100
Benzo[a]anthracene 1600
Benzo[a]pyrene 1600
Chrysene 2800
Dibenz[a,h]anthracene 260
Fluoranthene 5100
Fluorene 540
2-Methylnaphthalene 670
Naphthalene 2100
Phenanthrene 1500
Pyrene 2600
Total PCBs ISO
4,4'-DDE 27
Total DDTs 46.1
26
-------
Table 5. Number of sampling location/date combinations for each monitoring program within
Chesapeake Bay and the number of location date combinations retained for
discriminant analysis. An asterisk indicates that contaminants data were collected
separately as part of the Ambient Toxicity Program.
Monitoring Program
EMAP Virginian Province
Mid-Atlantic Integrated Assessment Program
CBP Long-term Benthic Monitoring Program (Maryland)*
Tidal Freshwater Goals Program
CBP Long-term Benthic Monitoring Program (Virginia)*
Ambient Toxicity Program (Maryland)
Ambient Toxicity Program (Virginia)
Years of
Collection
1990-93
1997-98
1997
1996
1997
1999
1999
Sampling
Locations
290
121
48
47
46
36
20
Total=608
Samples
109
67
17
22
17
11
13
Total=256
Table 6. Frequency and percentage of sites and mean B-IBI for sites within each status
classification category. Values in parentheses represent one standard deviation in the
B-IBI within each classification category.
Number of
Status Sites
Meets Goals 272
Marginal 69
Degraded 1 10
Severely Degraded 158
Overall 609
% of Sites
44.66
11.33
18.06
25.94
Mean B-IBI
3.6(0.5)
2.8(0.1)
2.4(0.1)
1.6(0.4)
2.8(1.0)
27
-------
Table?. Frequency of sites classified as severely degraded and degraded foreach habitat type.
Total
Severely Degraded
Degraded
Habitat Number Percentage Number Percentage Number Percentage
Polyhalme Mud
Polyhaline Sand
High Mesohaline Mud
High Mesohaline Sand
Low Mesohaline
Oligohaline
Tidal Freshwater
35
9
78
26
42
32
34
13.67
3.52
30.47
10.16
16.41
12.50
13.28
19
2
51
16
33
15
17
7.42
0.78
19.92
6.25
12.89
5.86
6.64
16
7
27
10
9
17
17
6.25
2.73
10.55
3.91
3.52
6.64
6.64
Table 8. Frequency of sites classified as severely degraded and degraded for each stress group.
Stress Group
Contaminant
Low D.O.
Combined
Unknown
ERM Sediment
Contaminant
Classification
Number Percentage
23 8.98
34 13.28
9 3.52
190 74.22
VA Province
Mean SQG Quotient
Sediment Contaminant
Classification
Number Percentage
63 24.61
24 9.38
19 7.42
150 58.59
All Province
Mean SQG Quotient
Sediment Contaminant
Classification
Number Percentage
140 54.69
10 3.91
33 12.89
73 28.52
28
-------
Table 9. Frequency of sites classified as severely degraded and degraded within each habitat and effect type for each of the sediment
contaminant classification schemes.
VA Province
ERM Sediment Mean SQG Quotient
Contaminant Classification Sediment Contaminant Classification
Habitat
High Mesohahne Mud
High Mesohahne Mud
High Mesohahne Mud
High Mesohahne Mud
High Mesohahne Sand
High Mesohahne Sand
High Mesohahne Sand
High Mesohahne Sand
Low Mesohahne
Low Mesohaline
Low Mesohaline
Low Mesohaline
Oligohaline
Ohgohalme
Oligohaline
Ohgohalme
Polyhahne Mud
PolyhalmeMud
Polyhahne Mud
PolyhalmeMud
Polyhahne Sand
Polyhahne Sand
Polyhahne Sand
Polyhahne Sand
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Stress Group
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Combined
Contaminant
Low D O
Unknown
Total
2
3
12
61
0
0
3
23
6
5
9
22
0
8
0
24
0
3
7
25
0
0
3
6
1
4
0
29
Severely
Degraded
2
2
11
36
0
0
3
13
5
5
9
14
0
5
0
10
0
1
7
11
0
0
2
0
1
3
0
13
Degraded
0
I
1
25
0
0
0
10
1
0
0
8
0
3
0
14
0
2
0
14
0
0
1
6
0
1
0
16
Total
4
14
10
50
0
0
3
23
14
14
1
13
0
16
0
16
0
5
7
23
0
0
3
6
1
14
0
19
Severely
Degraded
3
9
10
29
0
0
3
13
13
13
1
6
0
9
0
6
0
2
7
10
0
0
2
0
1
6
0
10
Degraded
1
5
0
21
0
0
0
10
1
1
0
7
0
7
0
10
0
3
0
13
0
0
1
6
0
8
0
9
All Province
Mean SQG Quotient Sediment
Contaminant Classification
Total
12
46
2
18
0
3
3
20
IS
20
0
7
0
26
0
6
5
21
2
7
0
0
3
6
1
24
0
9
Severely
Degraded
1 1
31
2
7
0
3
3
10
14
17
0
2
0
14
0
1
5
II
2
1
0
0
2
0
1
10
0
6
Degraded
1
15
0
11
0
0
0
10
1
3
0
5
0
12
0
5
0
10
0
6
0
0
1
6
0
14
0
3
29
-------
Table 10 Classification efficiencies of linear discriminant functions developed for the Within Habitat Type
scenarios for all available stress groups Shown are the percentages of correctly classified
observations for each stress group and the total percentage of observations correctly classified by the
discriminant function Values in parentheses are the total number observations for each stress group
Polyhahne Mud Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Polyhaline Mud
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded On|y
Combined
10000(3)
10000(4)
Contaminant
10000(2)
100.00(1)
10000(4)
100.00(1)
10000(18)
10000(9)
Low DO
10000(5)
10000(5)
10000(5)
10000(5)
10000(2)
10000(1)
Unknown
10000(21)
10000(9)
10000(19)
10000(9)
10000(5)
100.00(1)
Total
10000
10000
100.00
10000
10000
10000
Validation Data Set
Combined
000(1)
Contaminant
000(1)
000(1)
000(1)
66 67(3)
5000(2)
High Mesohalme Sand Calibration Data Set
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
-
Contaminant
10000(3)
10000(3)
Low DO
10000(2)
10000(2)
10000(2)
10000(2)
10000(1)
Low DO
10000(3)
10000(2)
10000(3)
10000(2)
10000(3)
10000(2)
Unknown
25.00(4)
50 00(2)
0.00(4)
10000(1)
0 00(2)
Unknown
10000(16)
10000(12)
10000(16)
10000(12)
10000(13)
10000(9)
Total
3661
6786
1786
93.33
5769
39.29
Total
100 00
10000
10000
10000
10000
10000
High Mesohalme Sand Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
-
Contaminant
-
Low D O
10000(1)
10000(1)
10000(1)
Unknown
100 00(6)
10000(1)
10000(5)
10000(1)
100 00(5)
10000(1)
1 OtQl
10000
10000
100.00
10000
10000
10000
30
-------
Table 10
Continued
High Mesohalme Mud Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
10000(1)
10000(2)
50 00(2)
33 33(3)
10000(7)
10000(7)
Contaminant
10000(2)
10000(2)
10000(12)
10000(9)
10000(37)
10000(23)
Low DO
62 50(8)
57.14(7)
10000(7)
10000(6)
100.00(2)
10000(2)
Unknown
10000(46)
100.00(28)
10000(36)
10000(21)
10000(11)
10000(7)
Total
94.74
9271
9825
9487
10000
10000
High Mesohaline Mud Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
10000(1)
0 00(2)
60 00(5)
50 00(4)
Low Mesohaline Calibration
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Low Mesohaline
Data Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and degraded
Severely Degraded Only
Combined
80 00(5)
50 00(2)
10000(12)
10000(10)
10000(13)
10000(11)
Validation
Combined
000(1)
0 00(3)
50 00(2)
66 67(3)
50 00(2)
66.67(3)
Contaminant
000(1)
0 00(2)
II 11(9)
75 00(8)
Data Set
Contaminant
100.00(2)
10000(3)
10000(9)
10000(9)
10000(13)
10000(13)
Data Set
Contaminant
0 00(3)
0 00(2)
0 00(5)
0 00(4)
42 86(7)
25 00(4)
Low DO
0 00(4)
25 00(4)
66.67(3)
75 00(4)
Low DO
10000(8)
10000(9)
10000(1)
10000(1),
Low DO
000(1)
Unknown
46 15(13)
75 00(8)
1667(12)
75 00(8)
0 00(5)
Unknown
10000(16)
10000(12)
10000(9)
10000(6)
10000(5)
10000(2)
Unknown
20 00(5)
50.00(2)
66 67(3)
10000(1)
Total
3900
6500
1871
7500
15 11
6917
Total
9677
96.15
10000
10000
10000
10000
Total
1032
3529
4000
3509
5507
44.10
31
-------
Table 10
Continued
Oligohalme Calibration Data Set
Classification Scheme Data Used
ERM Exceedancc Severely Degraded and Degraded
ERM Exceedance Severely Degraded Only
VA Province Severely Degraded and Degraded
VA Province Severely Degraded Only
All Province Severely Degraded and Degraded
All Province Severely Degraded Only
Oligohaline
Classification Scheme Data Used
ERM Exceedance Severely Degraded and Degraded
ERM Exceedance Severely Degraded Only
VA Province Severely Degraded and Degraded
VA Province Severely Degraded Only
All Province Severely Degraded and Degraded
All Province Severely Degraded Only
Combined Contaminant Low D 0
10000(6)
10000(4)
- 10000(10)
10000(5)
- 10000(19)
- 10000(10)
Validation Data Set
Combined Contaminant Low D O
10000(2)
000(1)
50 00(4)
5000(2)
10000(5)
50 00(2)
Tidal Freshwater Calibration Data Set
Classification Scheme Data Used
ERM Exceedance Severely Degraded and Degraded
ERM Exceedance Severely Degraded Only
VA Province Severely Degraded and Degraded
VA Province Severely Degraded Only
All Province Severely Degraded and Degraded
All Province Severely Degraded Only
Combined Contaminant Low D.O
10000(1) 10000(3)
10000(1) 10000(1)
10000(1) 10000(10)
10000(1) 10000(5)
10000(1) 10000(19)
10000(1) 10000(8)
Unknown Total
10000(19) 10000
10000(7) 10000
10000(15) 10000
100.00(6) 10000
10000(6) 10000
10000(1) 10000
Unknown Total
66 67(3) 74 67
000(1) 000
10000(1) 8000
5000
10000
5000
Unknown Total
10000(23) 10000
10000(12) 10000
10000(16) 10000
10000(8) 10000
10000(7) 10000
10000(5) 10000
Tidal Freshwater Validation Data Set
Classification Scheme Data Used
ERM Exceedance Severely Degraded and Degraded
ERM Exceedance Severely Degraded Only
VA Province Severely Degraded and Degraded
VA Province Severely Degraded Only
All Province Severely Degraded and Degraded
All Province Severely Degraded Only
Combined Contaminant Low D O
000(1)
50 00(2)
25.00(4)
000(1)
0 00(5)
0 00(2)
Unknown Total
66 67(6) 58.97
10000(1) 96.15
3333(3) 30.13
50.00(2) 30 77
0 00(2) 0.00
0.00(1) 0.00
32
-------
Table II Classification efficiencies of linear discriminant functions developed for Within Habitat Type
scenarios for discriminating between the Contaminant and Other stress groups. Shown arc the
percentages of correctly classified observations for each stress group and the total percentage of
observations correctly classified by the discriminant function Values in parentheses are the total
number observations for each stress group
Polyhalme Mud Calibration Data Set
Classification
Scheme
ERM Exceedancc
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Set Used Contaminant
Severely Degraded and Degraded 1 00 00(2)
Severely Degraded 1 00.00( 1 )
Severely Degraded and Degraded 1 00 00(4}
Severely Degraded 10000(1)
Severely Degraded and Degraded 1 00 00( 1 8)
Severely Degraded 1 00 00(9)
Other
10000(26)
10000(14)
10000(24)
10000(14)
10000(10)
10000(6)
Total
10000
10000
10000
10000
10000
10000
Polyhalme Mud Validation Data Set
Classification
Scheme
ERM Exceedance
ERM Exceedancc
VA Province
VA Province
All Province
All Province
Data Set Used Contaminant
Severely Degraded and Degraded 0 00( 1 )
Severely Degraded
Severely Degraded and Degraded 0 00( 1 )
Severely Degraded 0 00( 1 )
Severely Degraded and Degraded 1 00 00(3)
Severely Degraded SO 00(2)
Other
66 67(6)
10000(4)
66 67(6)
10000(3)
75 00(4)
10000(2)
Total
6190
10000
57 14
9333
9107
7000
High Mesohaline Sand Calibration Data Set
Classification
Scheme
All Province
All Province
Data Set Used Contaminant
Severely Degraded and Degraded 1 00 00(3)
Severely Degraded 1 00 00(3)
Other
10000(16)
100.00(1 1)
Total
100.00
10000
High Mesohaline Sand Validation Data Set
Classification
Scheme
All Province
All province
Data Set Used Contaminant
Severely Degraded and Degraded
Severely Degraded
Other
10000(5)
100 00(2)
Total
10000
10000
High Mesohaline Mud Calibration Data Set
Classification
Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Data Set Used Contaminant
Severely Degraded and Degraded 100 00(2)
Severely Degraded 100 00(2)
Severely Degraded and Degraded 1 00 00( 1 2)
Severely Degraded 100.00(9)
Severely Degraded and Degraded 1 00 00(37)
Severely Degraded 1 00 00(23)
Other
10000(55)
10000(37)
10000(45)
10000(30)
10000(20)
10000(16)
Total
10000
10000
10000
10000
10000
10000
High Mesohaline Mud Validation Data Set
Classification
Scheme
ERM Exceedance
ERM Exceedancc
VA Province
VA Province
All Province
All Province
Data Set Used Contaminant
Severely Degraded and Degraded 0 00( 1 )
Severely Degraded
Severely Degraded and Degraded 0 00(2)
Severely Degraded
Severely Degraded and Degraded 44 44(9)
Severely Degraded 75 00(8)
Other
8333(18)
8333(12)
7059(17)
91 67(12)
6000(10)
10000(4)
Total
8041
8333
5573
10000
4990
8526
33
-------
Table 11
Continued
Low Mesohalme Calibration Data Set
Classification
Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
ALL Province
ALL Province
Data Set Used
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Contaminant
10000(2)
10000(3)
10000(9)
10000(9)
10000(13)
10000(13)
Other
100.00(29)
10000(23)
10000(22)
10000(17)
10000(18)
100.00(13)
Total
10000
10000
10000
10000
10000
10000
Low Mesohalme Validation Data Set
Classification
Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
ALL Province
ALL Province
Classification
Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
ALL Province
ALL Province
Classification
Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
ALL Province
ALL Province
Data Set Used
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Oligohalme Calibration
Data Set Used
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Uligohalme Validation
Data Set Used
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and Degraded
Severely Degraded
Contaminant
0 00(3)
0.00(2)
40 00(5)
25 00(4)
42 86(7)
25 00(4)
Data Set
Contaminant
10000(6)
10000(4)
10000(10)
10000(5)
10000(19)
10000(10)
Data Set
Contaminant
10000(2)
000(1)
50 00(4)
50 00(2)
10000(5)
50 00(2)
Other
57 14(7)
80 00(5)
8000(5)
66 67(3)
66 67(3)
66 67(3)
Other
10000(19)
10000(7)
10000(15)
100.00(6)
10000(6)
10000(1)
Other
66 67(3)
000(1)
10000(1)
Total
53.46
7077
68.39
5224
5668
4583
Total
10000
10000
10000
10000
10000
10000
Total
7467
000
80.00
5000
10000
5000
34
-------
Table 11
Continued
Tidal Freshwater Calibration Data Set
Classification
Scheme
Data Set Used
Contaminant
Other Total
ERM Exceedance
ERM Exceedance
VA Province
VA Province
ALL Province
ALL Province
Severely Degraded and Degraded
Severely Degraded
Severely Degraded and
Severely Degraded
Severely Degraded and
Severely Degraded
Degraded
Degraded
10000(3)
10000(1)
10000(10)
10000(5)
100.00(19)
100 00(8)
10000(24)
100.00(13)
10000(17)
10000(9)
10000(8)
10000(6)
100.00
10000
10000
10000
10000
10000
Tidal Freshwater Validation Data Set
Classification
Scheme
Data Set Used
Contaminant
Other Total
ERM Exceedance Severely Degraded and Degraded 10000(1) 8333(6) 85.19
ERM Exceedance Severely Degraded 5000(2) 10000(1) 9643
VA Province Severely Degraded and Degraded 5000(4) 3333(3) 3951
VA Province Severely Degraded 5000(1) 5000(2) 67.86
ALL Province Severely Degraded and Degraded 40.00(5) 000(2) 2815
ALL Province Severely Degraded 5000(2) 000(1) 2857
35
-------
Table 12 Classification efficiencies of linear discrimmantfunctionsdeveloped forclassifying Polyhalme sites
into one of the four stress groups Shown are the percentages of correctly classified observations for
each stress group and the total percentage of observations correctly classified by the discriminant
function Values in parentheses are the total number observations for each stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined Contaminant
10000(2)
- 10000(1)
- 10000(4)
- 10000(1)
10000(4) 100.00(16)
10000(5) 100.00(9)
Low DO
10000(8)
10000(7)
10000(8)
10000(7)
10000(4)
10000(2)
Unknown
10000(24)
10000(9)
10000(22)
10000(9)
10000(10)
10000(1)
Total
10000
10000
10000
10000
100.00
10000
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined Contaminant
000(1)
- 10000(1)
000(1)
10000(1) 10000(5)
50 00(2)
Low DO
50 00(2)
10000(2)
50 00(2)
10000(2)
10000(1)
10000(2)
Unknown
57 14(7)
100.00(2)
57 14(7)
10000(1)
000(3)
Total
52 10
10000
6050
94 12
7059
5909
Table 13 Classification efficiencies of linear discnminantfunctionsdeveloped forclassifying Polyhalme sites
into the Contam mant, Comb med and U nknow n stress groups Shown are the percentages of corrcc tly
classified observations for each stress group and the total percentage of observations correctly
classified by the discriminant function Values in parentheses are the total number observations for
each stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined Contaminant Unknown
10000(4)
10000(4)
10000(3) 10000(24)
10000(1) 10000(10)
10000(5) 10000(22)
10000(2) 10000(9)
10000(15) 10000(12)
10000(9) 100.00(1)
Total
10000
10000
10000
100.00
10000
10000
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined Contaminant Unknown
000(1)
000(1)
8571(7)
- 10000(1)
- 7143(7)
- 10000(1)
6667(6) 000(1)
50.00(2)
Total
85 71
10000
71.43
10000
3226
3462
36
-------
Table 14 Classification efficiencies of linear discriminant functions developed forclassifying Polyhaline sites
into the Contaminant and all Other stress groups with and without Low D 0 sites Shown are the
percentages of correctly classified observations for each stress group and the total percentage of
observations correctly classified by the discriminant function Values in parentheses are the total
number observations for each stress group
With Low DO stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant
100 00(2)
10000(1)
10000(4)
100.00(1)
10000(16)
10000(9)
Other
100.00(32)
10000(16)
10000(30)
10000(16)
10000(18)
10000(8)
Total
10000
10000
10000
10000
10000
100.00
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant
000(1)
100.00(1)
0.00(1)
80 00(5)
10000(2)
Other
77.78(9)
10000(4)
66 67(9)
10000(3)
80 00(5)
50 00(2)
Total
7320
10000
7059
9412
80.00
7647
Without Low DO stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant
10000(3)
10000(1)
10000(5)
10000(2)
10000(15)
10000(9)
Other
10000(24)
10000(10)
10000(22)
10000(9)
100.00(16)
10000(5)
Total
10000
10000
10000
10000
10000
10000
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant
66 67(6)
10000(2)
Other
8571(7)
10000(1)
71.43(7)
10000(1)
50 00(2)
000(1)
Total
8571
10000
7143
10000
5806
6429
37
-------
Table 15 Classification efficiencies oflmeardiscriminantfunctionsdevelopedforclassifying Mesohalinc sites
into one of the four stress groups Shown are the percentages of correctly classified observations for
each stress group and the total percentage of observations correctly classified by the discriminant
function Values in parentheses are the total number observations for each stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded On(y
Combined
57 14(7)
50 00(6)
10000(15)
10000(12)
95 45(22)
10000(20)
Contaminant
10000(6)
10000(4)
10000(22)
10000(16)
97 87(47)
97 22(36)
Low DO.
9474(19)
10000(19)
6364(11)
6923(13)
75.00(4)
100 00(5)
Unknown
96.92(65)
97 83(46)
95.92(49)
94.12(34)
91 67(24)
10000(14)
Total
9381
9467
9381
9200
9485
9867
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
10000(1)
000(1)
66.67(3)
75 00(4)
10000(5)
60 00(5)
Contaminant
0 00(2)
0 00(3)
66 67(6)
33 33(6)
45 45(22)
5333(15)
Low DO
40.00(5)
75.00(4)
0 00(3)
000(1)
000(1)
Unknown
61 11(36)
1765(17)
40 63(32)
57 14(14)
5625(16)
0 00(5)
Total
5600
2982
4595
4502
5862
4457
Table 16 Classificattonefficicnciesoflineardiscriminantfunctionsdcveloped forclassifying Mesohalme sites
into the Contam inant, Comb ined and U nknow n stress groups Shown arc the percentages of correc tly
classified observations for each stress group and the total percentage of observations correctly
classified by the discriminant function. Values in parentheses are the total number observations for
each stress group.
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
10000(6)
10000(6)
9333(15)
10000(14)
92 00(25)
10000(17)
Contaminant Unknown
10000(5) 9733(75)
10000(5) 10000(47)
9130(23) 9444(54)
10000(17) 9697(33)
10000(47) 9565(23)
100.00(40) 9333(15)
Total
9767
10000
9348
9844
9684
9861
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
10000(2)
000(1)
66.67(3)
10000(2)
10000(2)
87.50(8)
Contaminant Unknown
33 33(3) 57 69(26)
10000(2) 1250(16)
80 00(5) 59 26(27)
4000(5) 0.00(15)
45.45(22) 5882(17)
6364(11) 2500(4)
Total
5923
1875
6565
3250
6305
61 22
38
-------
Table 17 Classification efficiencies of linear discrimmantfunctionsdeveloped forclassify ing Mesohalinc sites
into the Contaminant and all Other stress groups with and without Low D 0 sites Shown are the
percentages of correctly classified observations for each stress group and the total percentage of
observations correctly classified by the discriminant function Scenarios with the best overall and
within stress group classification efficiencies a re highlighted in bo Id Values in parentheses arc the
total number observations for each stress group
With Low DO stress group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Contaminant
10000(91)
10000(4)
95 45(22)
100.00(16)
89 36(47)
97 22(36)
Other Total
10000(6) 10000
10000(71) 10000
96 00(75) 95 88
0 00(59) 98 67
96 00(50) 92 78
10000(39) 9867
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Contaminant
000(2)
0.00(3)
66 67(6)
33 33(6)
50 00(22)
5333(15)
Other Total
76 19(42) 71 48
72 73(22) 68 85
73 68(38) 72 09
7368(19) 6508
8182(22) 6640
7000(10) 6200
Without Low D.O Stress Group
Calibration Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Contaminant
10000(5)
10000(5)
91 30(23)
10000(17)
93 62(47)
100.00(40)
Other Total
10000(81) 10000
10000(53) 10000
97 10(69) 95 65
10000(47) 10000
97 92(48) 95 79
100.00(32) 100.00
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Severely Degraded and Degraded Only
Severely Degraded Only
Contaminant
33 33(3)
10000(2)
8000(5)
40 00(5)
4091(22)
81.82(11)
Other Total
78 57(28) 75 94
7059(17) 7312
66 67(30) 70 00
7647(17) 6678
8421919) 6279
75.00(12) 78.79
39
-------
Table 18 Classification efficiencies of linear discriminant functions developed forclassi Tying Tidal Freshwater
and Oligohalme sites into the Contaminant and all Other stress groups Shown are the percentages
of correctly classified observations for each stress group and the total percentage of observations
correctly classified by the discriminant function Values in parentheses are the total number
observations for each stress group
Calibration Data Set
Classification Scheme Sites Used
Contaminant
Other Total
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
10000(7)
10000(6)
9444(18)
10000(10)
94 59(37)
10000(18)
9767(43) 9800
10000(19) 100.00
10000(32) 9800
10000(15) 10000
7692(13) 9000
100.00(7) 10000
Validation Data Set
Classification Scheme Sites Used
Contaminant
Other Total
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
40 00(5)
50 00(2)
5000(10)
33.33(3)
4545(11)
50.00(4)
55 56(9) 53 38
100.00(3) 8800
000(4) 1800
000(2) 1333
3333(3) 4230
000(1) 3600
40
-------
Table 19 Classification efficiencies of linear discriminant functions developed for Baywidc scenarios to
discriminate between four potential stress groups for both unconnected and salinity corrected data
Shown are the stress group specific and total percentages of correctly classified observations foreach
discriminant function Values in parenthesesare the total number observations for each stress group
Without Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
57 14(7)
100.00(8)
8667(15)
10000(15)
72 00(25)
87.50(16)
Calibration Data Set
Contaminant
8462(13)
10000(12)
58 33(36)
88 46(26)
88 64(88)
95 16(62)
Low DO
70 83(24)
7368(19)
5625(16)
8333(12)
10000(6)
10000(5)
Unknown
92 13(127)
96 83(63)
87 50(104)
97 96(49)
6731(52)
8421(19)
Total
8723
93 14
7836
9612
8012
9216
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Onry
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
0 00(2)
50 00(4)
10000(2)
62 50(8)
7333(15)
Linear Regression Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
Va Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
57 14(7)
10000(8)
9333(15)
10000(15)
72 00(25)
95 45(22)
Contaminant
6000(10)
23 08(4)
40.00(25)
4545(11)
78 00(50)
59.09(22)
Low DO
7000(10)
25.00(13)
25 00(8)
1818(11)
0 00(4)
0 00(4)
Unknown
69 64(56)
36 73(32)
6341(41)
40 00(25)
43 75(16)
50 00(8)
Total
66.11
3881
5371
4569
62.58
5673
Calibration
Contaminant
9231(13)
10000(12)
63 89(36)
88 46(26)
8851(87)
91 23(57)
Low DO
70 83(24)
7368(19)
6250(16)
91 67(12)
83 33(6)
60 00(5)
Unknown
9237(118)
91 53(59)
90 53(95)
91 11(45)
75 00(44)
8571(14)
Total
8765
8980
82 10
91.84
82 10
8980
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
0 00(2)
50 00(4)
10000(2)
62 50(8)
77 78(9)
Polynomial Regression Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
42.86(7)
100.00(8)
93.33(15)
10000(15)
72 00(25)
95 45(22)
Contaminant
3000(10)
0.00(4)
32 00(25)
4545(11)
73.47(36)
40 74(27)
Low D O.
7000(10)
1538(13)
62 50(8)
909(11)
0 00(4)
0.00(4)
Unknown
66 67(54)
37 50(32)
69 23(39)
40 00(25)
4000(15)
44 44(9)
Total
61 34
3450
5851
4685
5997
4571
Calibration
Contaminant
7692(13)
9167(12)
66 67(36)
84.62(22)
8851(87)
91 23(57)
Low DO
70 83(24)
6842(19)
6250(16)
7500(12)
83 33(6)
60 00(5)
Unknown
91.53(118)
93.22(59)
88 42(95)
95 56(45)
79 55(44)
8571(14)
Total
85.19
88.78
8148
9082
8333
8980
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
0 00(2)
50 00(4)
10000(2)
62.50(8)
77 78(9)
Contaminant
3000(10)
75 00(4)
32 00(25)
4545(11)
6531(49)
37 04(27)
Low DO
6000(10)
1538(13)
37 50(8)
4545(11)
0 00(4)
25 00(4)
Unknown
72.22(54)
3750(12)
61.54(39)
44 00(25)
4000(15)
44 44(9)
Total
63.90
3783
51 53
53 14
5558
4663
41
-------
Table 20 Classification efficiencies of linear discriminant functions developed for Baywide scenarios to
discriminate between the Contaminant, Combined and Unknown stress groups for both uncorrcctcd
and salinity corrected data Shown are the stress group specific and total percentages of correctly
classified observations for each discriminant function Values in parentheses are the total number
observations for each stress group
Without Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
75 00(4)
10000(8)
84.62(13)
10000(13)
77 27(22)
7895(19)
Calibration Data Set
Contaminant Unknown
9167(12) 9688(128)
10000(9) 9718(69)
8205(39) 8911(101)
9167(24) 9286(56)
8901(91) 7000(50)
94 83(58) 90 00(20)
Total
9583
9773
86.93
9355
8160
9072
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
40 00(5)
50 00(6)
10000(4)
5455(11)
8333(12)
Linear Regression Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
75.00(4)
100.00(8)
8462(13)
10000(13)
81.82(22)
91 67(24)
Contaminant Unknown
909(11) 87.27(55)
28 57(7) 62.50(24)
27 27(22) 63 64(44)
3846(13) 3333(18)
8511(47) 5000(18)
61 54(26) 57 14(7)
Calibration
Contaminant Unknown
9167(12) 9833(120)
10000(9) 9706(68)
8462(39) 9011(91)
9167(24) 9057(53)
93 33(90) 73 81(42)
96.23(53) 81 25(16)
Total
78.96
5868
5321
4398
70.21
6490
Total
9706
9765
88 11
9222
8636
9247
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
40 00(5)
66 67(6)
10000(4)
5455(11)
57 14(7)
Polynomial Regression Salinity Correction
Cl assi tication Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
75 00(4)
10000(8)
8462(13)
10000(13)
81.82(22)
91 67(24)
Contaminant Unknown
909(11) 8462(52)
14 29(7) 60 87(23)
22 73(22) 48 84(43)
2308(13) 2353(17)
8261(46) 5882(17)
6452(31) 7143(7)
Calibration
Contaminant Unknown
8333(12) 9833(120)
10000(9) 9706(68)
7692(39) 9011(91)
9167(24) 9245(53)
9333(90) 7381(42)
9623(53) 8125(16)
Total
76 64
5542
4334
3445
72 II
6380
Total
9632
9765
8601
9333
8636
9247
validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Combined
40 00(5)
50 00(6)
10000(4)
5455(11)
57 14(7)
Contaminant Unknown
18 18(11) 8077(52)
1429(7) 6522(23)
22 73(22) 58 14(43)
1538(13) 2353(17)
8043(46) 5882(17)
5806(31) 7143(7)
lotal
74 05
5926
4774
3240
7084
60 13
42
-------
Table 21 Classification efficiencies of linear discriminant functions developed for Baywide scenarios to
discriminate between the Contaminant and all Other stress groups with Low 00. sites for both
uncorrected and salinity corrected data Shown are the stress group specific and total percentages
of correctly classified observations for each discriminant function Values in parenthcscsare the total
number observations for each stress group
Without Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Calibration Data Set
Contaminant Other
8462(158) 9620(13)
100.00(12) 10000(90)
5000(36) 9481(135)
84 62(26) 98 68(76)
81.82(88) 73.49(83)
91.23(57) 9333(45)
Total
95.32
10000
8538
95 1
77.78
92 16
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Linear Regression Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant Other
4000(10) 8676(68)
25 00(4) 73 33(45)
48.00(25) 79 25(53)
4545(11) 7105(38)
82.00(50) 67.86(28)
40 74(27) 59 09(22)
Calibration
Contaminant Other
8462(13) 9597(149)
100.00(12) 9884(86)
61 11(36) 9365(126)
80 77(26) 97 22(70)
8621(87) 7733(75)
9298(57) 9024(41)
Total
8321
6765
7267
6453
75.14
4884
Total
9506
9898
8642
9286
82 10
91 84
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Polynomial Regression Salinity Correction
Class! tlcaiion Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant Other
3000(10) 8485(66)
5000(4) 71 11(45)
4000(25) 7647(51)
4545(11) 1105(38)
8163(49) 6667(18)
5185(27) 5909(22)
Calibration
Contaminant Other
7692(13) 9597(149)
9167(12) 10000(86)
5833(36) 9365(126)
8077(26) 9861(72)
87 36(87) 77 33(75)
9298(57) 9024(41)
Total
8045
6853
6837
6426
7470
5488
lotal
9444
9898
8580
9388
82.72
91 84
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant Other
3000(10) 89.39(66)
75.00(4) 66 67(45)
3200(25) 7647(51)
4545(11) 7368(38)
67 35(49) 66 67(27)
51.85(27) 5909(22)
lotal
8463
6769
6659
66 19
6703
5488
43
-------
Table 22 Classification efficiencies of linear discriminant functions developed for Baywide scenarios to
discriminate between the Contaminant and all Other groups without Low DO sites for both
uncorrected and salinity corrected data Shown are the stress group specific and total percentages
of correctly classified observations for each discriminant function Values in parentheses arc the
total number observations for each stress group.
Without Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
• Severely Degraded Only
Calibration Data Set
Contaminant Other
9167(12) 9848(132)
10000(9) 10000(79)
7949(39) 9386(114)
87.50(24) 95 65(69)
8462(91) 75.00(72)
92 45(53) 88 64(44)
Total
9792
10000
9020
9355
8037
9072
Validation Data Set
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Linear Regression Salinity Correction
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Polynomial Regression
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Salinity Correction
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant Other
909(11) 9167(60)
28 57(7) 83 33(24)
22 73(22) 68 00(50)
3846(13) 4545(22)
8723(47) 5517(29)
5806(31) 5000(14)
Calibration
Contaminant Other
9167(12) 10000(124)
10000(9) 10000(76)
8205(39) 9519(104)
9167(24) 9394(66)
88.89(90) 71 88(64)
92 45(53) 85 00(40)
Validation
Contaminant Other
909(11) 8947(57)
1429(6) 8261(19)
2727(22) 5918(49)
2308(13) 3333(21)
84 78(46) 64 29(28)
5806(31) 4286(14)
Calibration
Contaminant Other
8333(12) 10000(124)
10000(9) 10000(76)
8205(39) 9423(104)
9167(24) 9697(66)
8889(90) 71.88(64)
94.34(53) 87.50(40)
Total
84 79
7773
5646
4365
7307
5441
Total
99.26
100
91 61
9333
81.82
89.25
Total
8238
7537
5048
3060
7626
51 52
Total
9853
10000
9091
9556
81 82
9140
Validation
Classification Scheme
ERM Exceedance
ERM Exceedance
VA Province
VA Province
All Province
All Province
Sites Used
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Severely Degraded and Degraded
Severely Degraded Only
Contaminant Other
18 18(11) 8421(57)
1429(7) 7391(23)
2727(22) 6122(49)
2308(13) 3810(21)
8261(46) 6429(28)
6452(31) 5714(14)
lotal
78 38
6760
51.96
3409
7499
61 34
44
-------
Table 23 Classification efficiencies of discriminant functions developed for selected scenarios after application
of the stepwise discriminant and AN OVA variable reduction procedures
Across Habitat Severely Degraded and Degraded Stepwise Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province
All Province
Validation
Validation
63 33(90)'
68 00(50)
7093(86)
6667(30)
6705
6735
Across Habitat Severely Degraded and Degraded ANOVA Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Validation 71.11(90) 6860(86) 3011
All Province Validation 6j°0^0) _ 7000(:??) 3102
Polyhaline Mud Severely Degraded and Degraded With Low D O Sites Stepwias Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 8889 8000 8571
All Province Validation 10000 25 00 7321
Polyhaline Mud Severely Degraded and Degraded With Low D.O Sites ANOVA Variable Reduction ~
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 10000 10000 10000
All Province Validation 66.67 .50.°°... „ .6061
High Mesohaline Mud Severely Degraded and Degraded With Low D O Sites Stepwise Variable Reduction
Classification Scheme •
Data Set Contaminant Other Total
All Province Calibration 8919 7143 8276
All Province Validation _ 44_41 „. ..*4.5.5 „ .48.'°
High Mesohaline Mud Severely Degraded and Degraded With Low D O Sites ANOVA Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 8647 7500 8246
All Province Validation 7778 6000 71 54
Polyhaline Severely Degraded and Degraded With LowD O Sites Stepwise Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 9375 8333 8824
All Province Validation I000?._ .40_°° 6824
Polyhaline Severely Degraded and Degraded With Low D O. Sites ANOVA Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 87.50 10000 9412
All Province Validation 10000 4000 68.24
Mesohaline Severely Degraded Only With Low D U Sites Stepwise Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 7500 9487 8533
All Province Validation 53 85 6667 6051
Mesohaline Severely Degraded Only With Low D O Sites ANOVA Variable Reduction
Classification Scheme
Data Set Contaminant Other Total
All Province Calibration 86T1 88lO 8718
All Province Validation 4615 5385 5030
45
-------
Table 24 Coefficients and cutoff values for the Baywide linear discriminant function for classifying severely
degraded and degraded sites into the Contaminant and Other stress groups using "uncorrccted" data
Variable
Coefficient Variable
Coefficient
Bivalvia Abundance
Deep Deposit Feeder Species Richness
Haustorndae Abundance
Carnivore-Omnivore Species Richness
Epifaunal Species Richness
Spionidae Species Richness
Interface Feeder Species Richness
Polychaeta Proportional Abundance
Suspension Feeder Species Richness
Corophndae Species Richness
Deep Deposit Feeder Abundance
Isopoda Species Richness
Gastropoda Abundance
Ohgochaeta Proportional Abundance
Infaunal Species Evenness
Amphipoda Proportional Abundance
Ampeliscidae Abundance
Corophndae Proportional Abundance
Amphipoda Species Richness
Gastropoda Proportional Abundance
Nereidac Abundance
Mollusca Proportional Abundance
Ampeliscidae Species Richness
Epifaunal Species Diversity
Haustorndae Species Richness
Tubificidae Proportional Abundance
Ratio of Biomass to Abundance
Tubificidae Abundance
Bivalvia Proportional Abundance
Spionidae Abundance
Nereidae Species Richness
-5 7758 Mollusca Abundance 3.4078
-49318 Deep Deposit Feeder Proportional Abundance 2 9854
-19847 Infaunal Species Richness 28957
-19341 Haustomdae Proportional Abundance 21790
-16869 Corophndae Abundance 20845
-16390 Tubificidae Species Richness 18683
-1.5044 Ohgochaeta Species Richness 17297
-13688 Interface Feeder Proportional Abundance 14703
-12402 Interface Feeder Abundance 14380
-12291 Capitellidae Species Richness 13813
-1.1099 Epifaunal Species Richness 13278
-0 9923 Suspension Feeder Abundance I 2508
-09463 Infaunal Species Diversity 12457
-09326 Isopoda Abundance 12194
-0 7874 Ratio of Epifaunal to Infaunal Abundance 1 1108
-0 7706 Spionidae Proportional Abundance 0 9090
-0.6400 Total Biomass 08661
-06079 Ohgochaeta Abundance 08107
-0 4602 Polychaeta Species Richness 0 7996
-04197 Nereidae Proportional Abundance 06975
-04029 Bivalvia Species Richness 06930
-03791 Mollusca Species Richness 06648
-0 3257 Ampeliscidae Proportional Abundance 0 6035
-0 2631 Suspension Feeder Proportional Abundance 0 5728
-0.2470 Amphipoda Abundance 05295
-02319 Carmvore-Ommvore Proportional Abundance 05292
-01790 Carmvore-Ommvorc Abundance 05151
-01686 Gastropoda Abundance 02762
-01372 Isopoda Proportional Abundance 02669
-01167 Polychaeta Abundance 01674
-00909 Total Infaunal Abundance 01516
Capitellidae Proportional Abundance 0 1383
Capitellidae Abundance ' 0.1211
CutoffValuc=024ll
46
-------
Appendices
-------
Appendix A.
List of species classified as epifaunal.
Turbellaria Gastropoda
Stylochus elhplicus Gomobasis virgimca
Turbellaria spp Gyraulusspp
Hydrobia spp
Polychaeta llydrobndae spp
Dipolydora commensahs Hydrobudae sp Y Moms
Filogramnae sp A Moms Hydrobudae sp Z Moms
Harmothoe extenuala Kurtziella atrostyla
Harmoihoe spp Littoridmops tenmpei,
Hydroides dianthus Melanella spp
Hydroides protulicola Nudibranchia
Hydroides spp Ocloslomia engoma
Lepidonotus sublevis Odoslomia spp
Lepidonotus variabilis Physidac spp
Polydora websteri Planorbidae spp
Polynoidae spp Pleurocera spp
Sabellana vulgaris Pyramidella Candida
Serpulidae spp Pyramidellidae spp
Sayella chesapeakea
Hirudinea Tiirbomlla spp
Hirudmea spp Tumdae sp A Mountford
Bairacobdella phalera Urotalpinx cinerea
Helobdella spp Valvala sincera
Vitnnellidae spp
Gastropoda Vivipandae spp
Ammcola limosa
Anachis lafresnayi Bivalvia
Anachis obesa Anomia simplex
Anachis spp Anomia spp
Asiyris lunata Crassosirea virgimca
Bittium allernatum Ceukensia demtssa
Boonea bisuluralis Ischadium recurvum
Boonea impressa Modiolus spp
Boonea lemmuda Mytihdae spp
Cincinnati winkleyi Mytilopsis leucophaeata
Columbella spp Mynlus edulis
Columbellidac spp
Crauispira ostrearum Chelicerata
Cralena pilata Limulus polyphemus
Crepidula convexa-fornicata
Crepidula maculosa Cladoeera
Crepidula plana Cladoeera spp
Crepidula spp
Cylichnella bidenlala Cirripedia
Doridella obscure Balanus itnprovisus
Epitomum greenlandicum Balanus spp
Epilonium humphreysi
Epitomum rupicola Mysidae
Epilonium spp Americamysis almyra
Eupleura caudala Americamysis bigelowi
Fargoa bushiana Americamysis spp
Ferrissia rivulans Heteromysts formosa
Gastropoda spp. Mysidae spp.
Mysidae Decapoda
Neomysis amencana Panopeus herbstn
Pcnaeidae spp
Isopoda Pinnotheres ostreum
Edotea tnloba Processa vicina
Erichsonella attenuata Rhithropanopeus harrisn
ErichsoneHafiliformis Trachypenaeus conslrictus
Paracereis caudala Xanthidac
Sphaeroma quadndentatum
Cassidimdea ovalis Insecta
Brachycercus spp.
Amphipoda Cuenis spp
Ampilhoe longimanna Coenagnonidae
Apocorophium lacuslre Odonata spp
Apocorophium simile Curculionidae
Bated catharmensis Dubiraplua spp
Caprella andreae Elmidae
Caprella penanns Gynmdae
Caprella spp Stenelmis spp
Caprellidae spp Cyrnellus fraternus
Cerapus tubularis Hydroptilidae
Corophium spp. Deceits spp
Cymadusa compta Tnchoptera
Dulichiella appendiculata
Elasmopus laevis Bryozoa
Ericthomus brasiliensis Alcyontdtum spp
Gammaropsis sutherlandi Angumella palmata
Cammarus daiben Callopora craticula
Gammarus fasciatus Membrampora tenuis
Gammarus spp.
Gilanopsis spp Ascidiacea
Melita mtida Ascidiacea spp
Microprotopus raneyi Molgula arenala
Monocorophium acherusicum Molgula manhattenin,
Monocorophium insidiosum Perophora virtdis
Monocorophium tuberculalum
Mucrogammarus mucronatus
Paracaprella tenuis
Parametopella cypris
Photis pugnator
Stenolhoe mmuta
Stenothoe spp
Decapoda
Callmectes sapidus
Crangon septemspmosa
Decapoda spp
Dissodactylus mellitae
Eurypanopeus depressus
Hexapanopeus angustifrons
Pagurus longicarpus
Pagurus spp
Palaemonetes pugio
47
-------
Appendix B.
List of species classified as deep deposit feeders.
Polychaeta
Amastigos caperaius
Capitella capitata complex
Capitelhdae spp
Clymenella torquata
Heterom astus filiformis
Leitoscolo plos fragilis
Leiloscobplos robustus
Leiloscoloplos spp
Macro clymene zonalis
Maldanidae spp
Media mastus ambiseta
Notomaslussp A Ewing
Notomaslus spp
Orbinia risen
Orbmudaespp.
Pectinana gouldn
Sabaco elongatus
Scalibregma mflatum
Scoloplos rubra
Travisia sp A M orris
Oligochaeta
Aulodrilus limnobius
Aulodrilus paucichaeta
Aulodrilus pigueti
Aulodrilus pluriseta
Branclnura sowerbyi
Bratulavia umdeniata
Dero digitata
Dero spp
Haber cf spcciosus
Homochaela naidina
llyodrilus temple ton i
Isochaetides freyi
Limnodrilus cervix
Limnodrilus claparedianus
Limnodrilus hoffmeisteri
Limnodrilus spp
Limnodrilus udekemianus
Naididae spp
Nais pardalis
Nais pseudo blusa
Nais varia bills
Oligochaeta spp
Piguetiella michiga nensis
Prisimella jenkinae
Prislmella osbomi
Quisladrilus multisetosus
Specana josinae
Oligochaeta
Stephensoniana spp
Slephensoniana tandyi
Stephensoniana trivandrana
Telmatodrilus vejdovskyi
Tubificidac with capiliform chaetac
Tubificidae without capiliform chaetae
Tubificouies heterochaelus
Tubtficoides spp
Bivalvia
Nucula annulata
Nucula proximo
Nucula spp
Solemya velum
Yoldia lim alula
Enteropneusta
Enteropneusta spp
48
-------
Appendix C. List of species classified as suspension feeders.
Polychaeta
Chaelopterus variopedatus
Demonax microphthalmus
Sabelhdae spp
Bivalvia
Aligena elevata
Anadara avails
Anadara transversa
Anodonta spp
Borneo truncata
Corbicula flummea
Donax variabilis
Ensis directus
Gemma gemma
Lyonsia hyalma
Lyonsia spp
Mactndae spp
Mercenaria mercaiana
Mulinia lateralis .
Musculium spp
Mya arenaria
Mysella planulata
Pandora spp
Parvilucina multilineata
Periploma marganlaceum
Petricola pholadiformis
Paidium spp
Pilar morrhuanus
Rangia cuneata
Sphaendae spp
Spisula solidissima
Tagelus divuus
Tageliu plebeitu
Tagelus spp
Unionidae spp
Amphipoda
Ampeluca abdita-vadorum complex
Ampehsca spp
Ampelisca verrilli
Phoronida
Phoronis spp
Cephalochordata
Branchwsloma caribaeum
49
-------
Appendix D.
List of species classified as interface feeders.
Polychaeta
Ampharetidae spp.
Amph itrite ornala
Apopnonospio pygmaea
Aricidea cathermae
Aricidea wassi
Asabellides oculata
Boccardiella ham ala
Boccardiella ligenca
Carazziella hobsonae
Caulleriella sp B (Blake)
Cirratuhdae spp
Cirriformia grandis
Cirrophorus spp
Dipolydora socia Its
Dispio uncinata
Enoplobranchus sangumeus
Hobsoniaflonda
Levinsen la gracilis
Loimia medusa
Magelona spp
Manayunkia aestuanna
Marenzelleria vmdis
Melinn a macu lata
Monticellina baptisteae-dorsobranchialis
Monticellina spp
Owenia fusiform is
Owenndae spp
Paraonis fulgens
Paraprionospio pinnala
Pista cristata
Pista spp
Polycirrus spp
Polydo ra cornu la
Polydora spp
Polydora/Boccardiella spp
Polygordius spp
Pnonospio heterobronchia
Prionospio perkinsi
Pnonospio spp
Pseudopolydora spp
Scolelepis bousflddi
Scolelepis spp
Scolelepis squam ala
Scolelepis lexana
Spio setosa
Spiochaelopterus costarum
Spionidae spp
Spiophanes bom byx
Streblospio benedicti
Terebellidae spp.
Tharyx sp. A Morris
Bivalvia
Macoma balthica
Macoma mitchelli
Macoma lenta
Tellina agilis
Telhmdae spp
Cumacea
Almyracuma proximo cull
Bodotria sp A Morns
Cyclaspis varians
Leu con americanus
Mancocuma stellifera
Oxyurostylis smith i
Tanaidacea
Hargeria rapax
Tanaidacea spp
Tanaissus psammophilus
Amphipoda
Acanthohaustorius millsi
Acanthohaustorius similis
Americheltdium americanum
Ameroculodes species complex
Amphipoda spp
Bathypore ta parkeri
Corophium lacustre
Eobrolgus spmosus
Haustorndae spp
Leptdactyhts dytiscus
Leptocheirus plumulosus
Listriella barnardi
Lisiriella clymenellae
Listriella s mil hi
Listriella spp
Monoculodes edwardsi
Parahaustorius longimerus
Phoxocephalidae spp
Protohaustonus cf deichmannae
Protohaustorius wigleyi
Rhepoxynius hudsoni
Unciola dissimilis
Unciola irrorata
Unciola serrala
Unciola spp
Insecta
Stictochironomous spp.
Sipuncula
Microphiopholis atra
Ophiuroidea
Ophiuroidea spp
Holothuridea
Havelockia scabra
Holothuroideaspp
Leplosyn apla lenu is
Peniamera pulcliernma
Enteropneusta
Saccoglossus kowalevskn
50
-------
Appendix E.
List of species classified into the camivore/omnivore feeding group category.
Anthozoa
Anthozoa spp
Edwardsta elegans
Nemertea
Amphiporus bioculatus
Carmoma tremaphoros
Micrura leidyi
Ncmertmea
Nematoda
Nematoda spp
Polychaeta
Aglaophamus vernlli
A ncistrosylhs hartmanae
Ancistrosyllis jonea
Arabella incolor-mullidentala
Arabellidac spp
Autolytus spp
Bhawania helerosela
Brama clavata-swedmarki
Brama spp
Brama welljleelensis
Cabira incerta
Diopalra cuprea
Drilonereis tonga
Eteonefohosa
Eleone heteropoda
Eleone spp.
Eumida sanguined
Exogone dispar
Exogonespp
Clycera americana
Glycera dibranchwla
Glycen spp
Glyccridae spp
Glycinde solitaria
Gomadidae
Gyplis crypto
Hesionidae
Laeonereis culvert
Lepidametria commensahs
Lumbnnendae spp
Malmgremella taylori
Marphysa sanguined
Microphthalmus aberrans
Microphthalmus sczeScowu
Microphthalmus similis
Microphthalmus spp
Polychaeta
Neanthes arenaceodentata
Neanthes succmea
Nephtyidae
Nephtys bucera
Nephtys cryptomma
Nephtys incisa
Nephtys picta
Nephtys spp
Nereididae
Nereis grayi
Onuphidae
Onuphis eremita
Parahesione luteola
Paranaitis speaosa
Parapionosyllis longicirrata
Parougia caeca
Phyllodoce arenas
Phyllodoce spp
Phyllodocidae
Pilargidac
Pionosyllis spp
Podarke obscura
Podarkeopsis levifuiana
Protodnloides chaetifer
Pseudeurythoe paucibranchiata
Scoleloma tenuis
Sigambra bassi
Sigambra spp.
Sigambra tentaculata
Sphaerosyllis aciculata
Sphaerosyllis taylori
Slhenelais boa
Sthenelais spp
Slreptosyllis arenas
Streptosyllis pettiboneae
Syllidae spp
Sylhdes spp
Syllides vernlli
Oligochaeta
Chaetogaster spp
Gastropoda
Acteocma canaliculata
Bithynia tentaculata
Busycon spp
Caecidac spp
Caecum regulare
Caecum sp A Mountford
Gastropoda
Gastropoda sp A Mountford
Hammoea solitaria
llyanassa obsoleta
Lymnaeidae spp
Nassarius spp
Nassarius trivittatus
Nassarius vibex
Natica pusilla
Naticidae
Rictaxis punctoslnalus
Copepoda
Argulus spp
Stomatopoda
Squilla empusa
Isopoda
Amakiuanlhura magnijica
Ancmiu depressus
Chindotea almyra
Chiridolea caeca
Cyathura burbancki
Cyathura pohta
Cyathura spp
Ptilanthura tenuis
Decapoda
Alpheus heterochaelis
Automate sp A Williams
Callianassa setimanus
Euceramus praelongus
Libinia spp
Ogyrides alphaerostru
Ovalipes ocellatus
Pinmxa chaetopterana
Pinnixa retinens
Pmnixa spp
Polyonyx gibbest
Thalassmidea
Upogebia afflnis
Insecta
Ephemendae
Hexagenia limbata
Hexagema spp
Bezzia spp
Ceratopogonidae spp
Chaoborus albatus
Insecta
Chaoborus punctipennis
Chaoborus spp.
Diptera spp
Ablabesmyia annulata
Axarus spp.
Chironomidae spp
Chironommi spp.
Chironomus spp
Cladopelma spp
Cladotanytarsus spp
Clmotanypus pmguis
Chnotanypus spp
Coelolanypus spp
Cricolopus spp.
Cricotopus/Onhocladius spp
Cryptochironomus fulvus
Cryptocluronomus spp
Cryptolendipes spp
Demicryptochironomus spp
Dicrotendipes spp
Endochironomus spp
Epoicocladws spp
Glyplotendipes spp
Harmschia spp.
KiefferuliK spp
Microchironomus spp
Nanocladiwi spp
Orthocladunae
Parachironomus spp
Paracladopelma spp
Paralauterbormella spp
Phaenopsectra spp
Polypedilum halterale group
Polypedilum spp
Procladius spp.
Procladius subleltet
Pseudocluronomus spp
Kheotanytarsus spp
Tanypodmac
Tanypus spp.
Tanytarsmi
Tanytarsus spp
Echinoidea
Echmoidea spp
Mellita quinquiesperforatt
51
-------
Appendix F Number of contaminants exceeding the Effects Range Median concentration (ERM Cone ), the mean Sediment QualilyGuidelmes (SQG) quotient, the number of missing analytes, and a listing of missing
analytcs for each stall on date combinan on classified as severely degraded or degraded Habitat type is based on Wcisbcrg ct al (1997)
Station
CP94084
AR4
VA90-088
VA90-088
VA92-494
VA90-090
VA90-090
VA90-090
VA90-I40
VA9 1-090
VA90-081
VA92-483
MET06424
MET06425
VA90-089
VA92-521
VA9 1-306
VA91-3I2
VA92-452
VA90-091
VA92-519
VA90-050
VA90-056
VA90-059
Date
07/12/94
08/26/98
07/08/90
08/26/90
07/29/92
07/08/90
07/26/90
09/05/90
08/15/90
09/05/91
08/27/90
08/15/92
09/09/99
09/09/99
08/07/90
08/28/92
07/28/91
07/28/91
08/09/92
08/14/90
08/05/92
07/20/90
08/19/90
07/22/90
Estuary
Albcmarlc-Chcsapcakc Canal
Anacostia River
Anacoslia River
Anacostia River
Aquia Creek
Back River
Back River
Back River
Back River
Back River
Bear Creek
Big Anncmcsscx River
Bohemia River
Bohemia River
Bohemia River
Bohemia River
Breton Bay
Breton Bay
Broad/Lmkhorn Bay
Bush River
Bush River
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Number of
Contaminants
Exceeding Mean SQG
Habitat ERM Cone quotient
Low Mcsohalmc
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Oligohalme
Oligohalme
Oligohalme
Oligohalme
Oligohalme
Low Mesohalme
Low Mcsohalmc
High Mcsohalmc Mud
Oligohalme
Oligohalme
Oligohalme
Oligohalme
High Mcsohalmc Sand
High Mcsohalmc Mud
Polyhalinc Mud
Tidal Freshwater
Oligohalme
High Mcsohalmc Sand
Polyhalinc Mud
Polyhalinc Mud
0
4
1
1
0
3
3
3
6
3
3
0
2
0
0
2
0
0
0
0
0
0
0
0
0018
0405
0237
0237
0 100
0449
0449
0449
0723
0451
0578
0035
0437
0034
0068
2867
0036
0065
0091
0 109
0231
0022
0038
0029
Number of
Missing
Analytes Missing Analytes
0
3
1
1
0
1
1
1
1
1
1
0
5
5
1
0
0
0
0
1
0
1
1
1
AC, Total PCBs,2-Melhylnaphlhalene
AS
AS
AS
p.pDDE
AS
Accnaphthcnc,Acenaphthylcnc,Dibcnz(a,h)anthraccnc,2Mcthylnaphthalcnc,Naphthalcne,
Accnaphlhcnc,Accnaphthylcnc,Dibcnz(a,h)anlhracenc,2Mcthylnaphihalcnc,Naphthalcnc,
AS
AS
AS
AS
AS
52
-------
Appendix F
Continued
Station
VA90-062
VA90-062
VA90-062
VA90-065
VA90-066
VA90-080
VA9 1-050
VA9 1-282
VA9 1-283
VA9 1-303
VA9 1-325
VA9 1-426
VA92-050
VA92-058
VA92-455
VA92-482
VA92-497
VA92-500
VA93-OSO
VA93-050
VA93-617
VA93-622
VA93-626
VA93-630
VA93-644
Date
07/05/90
08/24/90
09/07/90
09/07/90
08/24/90
08/16/90
07/11/91
08/12/91
08/23/91
08/27/91
08/15/91
07/10/91
08/03/92
08/30/92
08/08/92
08/30/92
08/14/92
08/30/92
07/29/93
08/26/93
08/22/93
08/07/93
09/03/93
08/04/93
09/02/93
Estuary
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Number of
Contaminants
Exceeding
Habitat Type ERM Cone
High Mesohalme Mud
Polyhalme Mud
Polyhaline Mud
High Mesohalme Sand
High Mesohalme Mud
Polyhaline Mud
High Mesohalme Mud
Polyhaline Mud
Polyhaline Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
Low Mesohalme
Polyhaline Sand
Polyhaline Mud
Polyhaline Mud
Polyhaline Sand
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
Polyhalme Sand
High Mesohalme Mud
Polyhaline Mud
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Mean SQG
quotient
0054
0054
0054
0002
0082
0073
0049
0049
0043
0088
0160
0047
0037
0020
0006
0056
0083
0018
0010
0077
0049
0010
0013
0052
0046
Number of
Missing
Analytes Missing Analytes
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
AS
AS
AS
p,PDDE
p.pDDE
p,pDDE
p.pDDE
p.pDDE
53
-------
Appendix F
Continued
Station Date
VA93-647 08/05/93
VA93-650 08/29/93
VA93-6S3 08/27/93
VA93-657 08/25/93
MMS-04508 09/17/97
MMS-04512 09/16/97
MMS-045 15 09/02/97
UPB-04613 09/03/97
UPB-04621 08/26/97
VBY-04M1408/04/97
VBY-04M1608/11/97
VBY-04M22 08/1 2/97
VBY-04M24 08/1 2/97
VBY-04M30 08/1 2/97
CR59 09/10/98
CR61 09/10/98
VA93-661 08/05/93
CH10 09/15/99
CH9 09/15/99
VA90-082 08/27/90
VA93-620 08/08/93
VA93-730 08/08/93
MA9S-I02I 08/27/98
MA9S-I022 08/29/98
MA9S-1023 08/27/98
Estuary
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay
Chesapeake Bay Mamstem
Chesapeake Bay Mamstem
Chesapeake Bay Mamstem
Chesapeake Bay Mamstem
Chesapeake Bay Mamstem
Chesapeake Bay Mamstem
Chesapeake Bay Mamstcm
Chesapeake Bay Mamstcm
Chesapeake Bay Mamstcm
Chesapeake Bay Mamstem
Chester River
Chester River
Chester River
Choptank River
Choptank River
Colgate Cove
Corrotoman River
Corrotoman River
Eastern Bay
Eastern Bay
Eastern Bay
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
Tidal Freshwater
Polyhalmc Sand
Polyhalinc Mud
Polyhalmc Mud
Polyhalinc Mud
Polyhalme Sand
High Mesohalme Mud
High Mesohalme Sand
Low Mesohalme
Ohgohalme
Low Mesohalme
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Sand
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0094
0006
0 135
0 137
0007
0075
0 101
0138
0060
0003
0026
0029
0044
0026
0035
0015
0 135
0022
0026
0236
0069
0054
0011
0063
0006
Number of
Missing
Analyles Missing Analytes
0
0
0
0
3
2
2
0
1
2
2
2
2
2
3
3
0
4
4
1
0
0
2
2
3
Total PCBs, p.pDDE, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs,
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
AGJotal PCBs 2-Methylnaphthalene
AG.Tolal PCBs 2-Methylnaphthalene
AC, Total PCBs, Total DDTs,
AC, Total PCBs, Total DDTs,
AS
Total PCBs, Total DDTs
Total PCBs, Total DDTs
HG, Total PCBs, Total DDTs
2-Methylnaphthalene
2-Methylnaphthalene
54
-------
Appendix F
Continued
Station Dale
MA98-I028 08/26/98
MA98-1029 08/26/98
MA98-1030 08/26/98
VA90-086 08/01/90
VA90-086 09/13/90
VA9 1-308 08/29/91
VA9 1-286 08/1 1/91
VA91-290 08/11/91
JAM-04J01 08/2S/97
JAM-04J05 08/25/97
JAM-04J26 08/21/97
JAM06J17 08/03/99
JAM06J23 08/03/99
VA90-208 08/22/90
VA90-210 07/23/90
VA9 1-273 08/04/91
VA9 1-275 08/05/91
VA92-464 08/17/92
VA93-602 08/13/93
VA93-609 08/15/93
VA93-610 08/16/93
VA93-728 08/15/93
MMS-045 14 09/02/97
VA91-322 08/15/91
VA9 1-323 08/15/91
Estuary
Eastern Bay
Eastern Bay
Eastern Bay
Elizabeth River
Elizabeth River
Fishing Bay
Great Wicomico River
Great Wicomico River
James River
James River
James River
James River
James River
James River
James River
James River
James River
James River
James River
James River
James River
James River
Little Choptank River
Little Choptank River
Little Choptank R iver
Number of
Contaminants
Exceeding Mean SQG
Habnat Type ERM Cone quotient
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
Polyhalme Mud
Polyhalme Mud
High Mesohalme Mud
Polyhalme Mud
High Mesohalme Mud
Polyhalme Mud
Polyhalme Mud
Oligohalmc
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Polyhalme Mud
Tidal Freshwater
Tidal Freshwater
Oligohalme
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0056
0040
0038
0342
0342
0038
0061
0085
0 129
0050
0085
0244
0 120
0034
0010
0 115
0080
0061
0098
0 104
0029
0110
0006
0025
0037
Number of
Missing
Analytes Missing Analytes
2
2
2
1
1
0
0
0
1
2
1
5
5
1
1
1
0
0
0
0
0
0
2
1
1
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
AS
Total PCBs
Total PCBs, TotalDDTs
Total PCBs
Accnaphthcnc, Accnaphthylcnc, Dibcnz[a,h]anthraccnc, 2-Mcthylnaphthalcnc, Naphthalene
Accnaphlhcnc, Accnaphthylcnc, Dibcnz[a,h]anthraccnc, 2-Mcthylnaphthalcnc, Naphthalene
AS
AS
p.pDDE
Total PCBs, Total DDTs
p.pDDE
p.pDDE
55
-------
Appendix F
Continued
Station Date
MWT06309 09/08/99
MWT063IO 09/08/99
VA92-136 08/04/92
VA92-I36 08/29/92
VA93-136 08/03/93
VA93-136 08/30/93
VA91-330 08/17/91
VA9 1-331 08/16/91
VA92-466 08/21/92
VA92-45I 08/10/92
VA90-134 07/07/90
VA90-134 08/15/90
VA90-134 09/06/90
PXR-04216 09/05/97
PXR-04223 09/12/97
PXR06207 08/31/99
VA9I-280 08/09/91
PMR-04101 09/15/97
PMR-04102 09/15/97
PMR-04104 09/15/97
PMR-04108 09/15/97
PMR-04110 09/16/97
PMR-041II 09/16/97
PMR-04112 09/16/97
PMR-04115 09/16/97
Estuary
Magoihy River
Magothy River
Middle River
Middle River
Middle River
Middle River
Miles River
Miles River
Mobjack Bay
Nanscmond River
Patapsco River
Patapsco River
Patapsco River
Patuxent River
Patuxent River
Patuxent River
Piankatank River
Potomac River
Potomac River
Potomac River
Potomac River
Potomac River
Potomac River
Potomac River
Potomac River
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
Low Mesohalme
Low Mesohalme
Oligohalme
Oligohalme
Oligohalme
Low Mesohalme
High Mesohalme Mud
High Mcsohalmc Mud
Polyhalme Mud
High Mesohalme Mud
Low Mesohalme
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
2
0
0
0
2
0
0
0
0
0
1
1
1
0
0
3
0
0
0
0
0
0
0
0
0
0410
0034
0077
0307
0268
0 132
0051
0056
0048
0081
0210
0210
0210
0066
0046
0617
0052
0070
0010
0082
0081
0008
0095
0090
0091
Number of
Missing
Analyles Missing Analytes
5
5
0
0
0
0
1
1
0
0
1
1
1
2
2
5
0
2
2
2
2
2
2
2
2
Acenaphthene, Acernphthylene, Dibenz[a,h]anthracene, 2-Methylnaphthalene, Naphthalene
Acenaphthene, Acernphthylene, Dibenz[a,h]anthracene, 2-Methylnaphthalene, Naphthalene
p.pDDE
p.pDDE
AS
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Acenaphthene, Acernphthylene, Dibenz[a,h]anthracene, 2-Methylnaphthalene, Naphthalene
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
56
-------
Appendix F
Continued
Station Date Estuary
PMR06I06 09/20/99 Potomac River
VA90-I80 08/16/90 Potomac River
VA90-I82 08/06/90 Potomac River
VA90-188 08/26/90 Potomac River
VA9 1-302 07/28/91 Potomac River
VA92-I88 07/27/92 Potomac River
VA92-489 07/30/92 Potomac River
VA93-637 08/11/93 Potomac River
VA93-645 08/10/93 Potomac River
RAP-04R01 08/28/97 Rappahannock River
RAP-04ROS 08/28/97 Rappahannock River
RAP-04RI2 08/28/97 Rappahannock River
RAP-04RIS 08/28/97 Rappahannock River
RAP-04RI7 08/28/97 Rappahannock River
RAP-04R2S 09/17/97 Rappahannock River
RP1 08/11/99 Rappahannock River
RP2 08/11/99 Rappahannock River
RP3 08/11/99 Rappahannock River
RP4 08/11/99 Rappahannock River
RPS 08/11/99 Rappahannock River
RP6 08/11/99 Rappahannock River
RP8 08/10/99 Rappahannock River
RP9 08/10/99 Rappahannock River
VA90-084 08/14/90 Rappahannock River
VA90-I90 08/15/90 Rappahannock River
Number or
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
High Mesohalinc Mud
High Mesohalme Sand
Low Mesohalme
Tidal Freshwater
High Mesohalme Sand
Tidal Freshwater
Low Mesohalme
High Mesohalme Mud
Oligohalme
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0247
0016
0046
0 138
0012
0 118
0086
0080
0 I2S
0033
OOSS
0058
0056
0058
0049
0043
0047
0043
0040
0041
0015
0029
0048
0040
0035
Number of
Missing
Analytes Missing Analyles
5
1
1
1
0
0
0
0
0
2
3
3
2
3
2
4
4
4
4
4
4
4
4
1
1
Acenaphthene, Acemphthylene, Dibenz[a,h]anthracene, 2-Methylnaphlhalene, Naphthalene
AS
AS
AS
Total PCBs, Total DDTs
Total PCBs, p.pDDE, Total DDTs
Total PCBs, p,pDDE, Total DDTs
Total PCBs, TotalDDTs
Total PCBs, p.pDDE, TotalDDTs
Total PCBs, Total DDTs
AG, Total PCBs, TotalDDTs, 2-Mahylnaphthalene
AC, Total PCBs, TotalDDTs, 2-Methylnaphthalene
AC, Total PCBs, TotalDDTs, 2-Mclhylnaphlhalene
AG. Total PCBs, TotalDDTs, 2-Mahylnaphthalene
AG, Total PCBs, TotalDDTs, 2-Methylnaphthalene
AG, Total PCBs, TotalDDTs, 2-Methylnaphthalene
AG, Total PCBs, TotalDDTs, 2-Methylnaphthalene
AG, Total PCBs, TotalDDTs, 2-Mahylnaphthalene
AS
AS
57
-------
Appendix F
Continued
Station Dale
VA90-I92 07/06/90
VA90-I92 09/07/90
VA90-I96 08/05/90
VA9 1-294 07/30/91
VA9 1-298 07/30/91
VA92-477 08/04/92
VA92-481 08/06/92
VA93-628 08/19/93
VA92-S04 08/06/92
VA9 1-304 07/24/91
VA92-486 08/28/92
VA9 1-351 07/30/91
MMS-04511 09/17/97
VA92-045 08/02/92
VA93-627 08/09/93
VA93-652 08/28/93
VA9I-332 08/16/91
VA93-729 08/28/93
YRK-04Y02 08/26/97
YRK-04YI4 08/26/97
YRK-04Y23 09/16/97
YRK06Y16 08/10/99
YRK06Y18 08/04/99
YRK06Y2I 08/04/99
MA97-0061 07/27/97
Estuary
Rappahannock River
Rappahannock River
Rappahannock River
Rappahannock River
Rappahannock River
Rappahannock River
Rappahannock River
Rappahannock River
South River
Si Clements Bay
St Marys River
Susquchanna River
Tangier Sound
Tangier Sound
Tangier Sound
Tred Avon River
Wye River
York River
York River
York River
York River
York River
York River
York River
Unknown
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
Ohgohalme
Oligohalme
Tidal Freshwater
Oligohalme
Tidal Freshwater
High Mesohalme Mud
Oligohalme
Oligohalme
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
Tidal Freshwater
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Mud
Low Mesohalme
High Mesohalme Mud
Low Mesohalme
Polyhalme Mud
High Mesohalme Sand
High Mesohalme Sand
Low Mesohalme
Low Mesohalme
Oligohalme
Polyhalme Sand
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0050
0050
0037
0050
0062
0228
0061
0067
0137
0061
0051
0085
0006
0015
0038
0055
0032
0031
0051
0036
0047
0082
0 167
0 194
0012
Number of
Missing
Analytes Missing Analytes
1
1
1 AS
0
0
0
0
0
0
1 p.pDDE
0
0
3 Total PCBs, p.pDDE, TotalDDTs
0
0
0
1 p.pDDE
0
2 Total PCBs, Total DDTs
2 Total PCBs, Total DDTs
2 Total PCBs, Total DDTs
5 Acenaphlhene, Aceraphthylene, Dibenz[a,h]anthracene, 2-Methylnaphthalene, Naphthalene
5 Acenaphlhene, Aceraphthylene, Dibenz[a,h]amhracene, 2-Methylnaphthalene, Naphthalene
5 Acenaphlhene, Aceraphthylene, Dibenz[a,h]anthracene, 2-Methylnaphthalene, Naphthalene
2 Total PCBs, Total DDTs
58
-------
Appendix F
Continued
Station Date
MA97-0062 07/26/97
MA97-0063 07/26/97
MA97-0064 07/27/97
MA97-0065 07/26/97
MA97-0068 07/26/97
MA97-0069 07/27/97
MA97-007I 07/29/97
MA97-0076 07/31/97
MA97-0084 08/30/97
MA97-0090 08/26/97
MA97-0096 07/30/97
MA97-OIIO 08/04/97
MA97-0112 08/OS/97
MA97-OII3 08/06/97
MA97-01I4 08/07/97
MA97-0116 08/06/97
MA97-OII7 08/08/97
MA97-OI18 08/07/97
MA97-01I9 08/09/97
MA97-0120 08/09/97
MA97-0121 08/08/97
MA97-OI22 08/08/97
MA97-0124 08/15/97
MA97-0125 08/09/97
MA97-0126 08/14/97
Estuary
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
Polyhalme Mud
Polyhalme Sand
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
Tidal Freshwater
Polyhalme Mud
Polyhalme Sand
High Mcsohalinc Mud
High Mcsohalinc Mud
Low Mcsohalinc
Low Mcsohalinc
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
High Mesohalme Mud
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0043
0008
0047
0053
005S
0043
0059
0049
0005
0308
0083
0 175
0 107
0203
0188
0182
0 ISO
0262
0076
0272
0 196
0015
0016
0 154
0 152
Number of
Missing
Analytes Missing Analytes
2
2
2
2
2
2
2
2
2
1
1
2
2
1
2
2
1
2
2
1
2
2
2
2
2
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
59
-------
Appendix F
Continued
Station Dale
MA97-OI28 08/10/97
MA97-OI29 08/12/97
MA97-OI31 08/15/97
MA97-OI32 08/11/97
MA97-OI34 08/11/97
MA97-OI37 08/13/97
MA97-0138 08/19/97
MA97-0141 08/19/97
MA97-0142 08/18/97
MA97-0144 08/19/97
MA97-0145 08/18/97
MA97-0146 08/22/97
MA97-0147 08/16/97
MA97-0148 08/16/97
MA97-01S2 OS/23197
MA97-0153 08/25/97
MA97-0159 08/21/97
MA97-OI63 08/21/97
MA97-0177 07/28/97
MA97-0228 08/01/97
MA97-0229 08/01/97
MA97-0230 08/03/97
MA97-023I 08/01/97
MA97-0232 07/31/97
MA97-0233 08/03/97
Estuary
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
Low Mesohalme
High Mesohalme Mud
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Mud
Low Mesohalme
Low Mesohalme
Low Mesohalme
High Mesohalme Mud
Low Mesohalme
Low Mesohalme
Low Mesohalme
Low Mesohalme
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Sand
Oligohalme
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
Polyhalme Mud
0
0
0
0
0
0
0
1
1
0
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0 190
0146
0 116
0166
0 182
0 171
0219
0 198
0308
0159
0243
0228
0068
0135
0 148
0 126
0008
0019
0078
0069
0 121
0051
0059
0049
0057
Number of
Missing
Analytes Missing Analytes
2
2
0
2
2
1
1
1
2
2
1
2
2
2
2
1
2
3
2
1
1
1
1
2
1
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total DDTs
Total DDTs
Total PCBs, Total DDTs
Total PCBs. Total DDTs
Total DDTs
Total PCBs. Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs, Total DDTs
Total PCBs. p.pDDE. Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs
Total PCBs
Total PCBs
Total PCBs, Total DDTs
Total PCBs
60
-------
Appendix F
Continued
Station Date
MA97-0234 07/31/97
MA97-0236 08/03/97
MA97-0237 08/29/97
MA97-0238 08/27/97
MA97-024I 08/27/97
MA97-0242 08/29/97
MA97-0243 08/28/97
MA97-0244 08/28/97
MA97-0245 08/28/97
MA97-0246 08/28/97
OL-01 08/27/96
OL-08 08/29/96
OL-09 08/29/96
OL-ll 09/15/96
OL-12 09/12/96
OL-14 09/15/96
OL-15 09/12/96
OL-20 09/12/96
TF-03 09/19/96
TF-04 09/19/96
TF-06 09/19/96
TF-08 09/19/96
TF-16 09/05/96
TF-18 09/15/96
TF-19 09/18/96
Estuary
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Number of
Contaminants
Exceeding Mean SQG
Habitat Type ERM Cone quotient
Polyhalme Mud
Polyhalme Sand
High Mesohalme Mud
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Sand
High Mesohalme Mud
High Mesohalme Sand
Oligohalmc
Oligohalmc
Oligohalmc
Oligohalme
Ohgohalme
Oligohalmc
Oligohalme
Oligohalme
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
Tidal Freshwater
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0051
0005
0052
0054
0018
0022
0049
0013
0069
0015
0047
0075
0076
0032
0 137
0117
0035
0 135
0044
0041
0015
0072
0050
0074
0044
Number of
Missing
Analytes Missing Analytes
2
2
2
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, Total DDTs
Total PCBs, 2-Mclhylnaphthalcnc
Total PCBs, 2-Mclhylnaphthalcnc
Total PCBs, 2-Mcfliylnaphihalcnc
Total PCBs, 2-Melhylnaphthalene
Total PCBs, 2-Melhylnaphthalene
Total PCBs, 2-Melhylnaphthalene
Total PCBs, 2-Mclhylnaphthalene
Total PCBs, 2-Melhylnaphihalcne
Total PCBs, 2-Meftiylnaphihalene
Total PCBs, 2-Mediylnaphthalene
Total PCBs, 2-Mediylnaphthalene
Total PCBs, 2-Methylnaphthalene
Total PCBs, 2-Melhylnaphthalene
Total PCBs, 2-Melhylnaphthalene
Total PCBs, 2-MeSiyInaphthalene
61
-------
Appendix F
Continued
Station
Dale
Estuary
Habitat Type
Number of
Contaminants
Exceeding Mean SQG
ERM Cone quotient
Number of
Missing
Analytes Missing Analytes
TF-20
TF-21
TF-22
TF-23
TF-24
TF-25
TF-28
09/25/96
09/19/96
09/20/96
09/19/96
09/19/96
09/20/96
09/11/96
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Tidal Freshwater 0 0115 2 Total PCBs, 2-Melhylnaphthalene
Tidal Freshwater 0 0115 2 Total PCBs, 2-Mclhylnaphthalcnc
Tidal Freshwater 0 0 034 2 Total PCBs, 2-Melhylnaphthalene
Tidal Freshwater 0 0121 2 Total PCBs, 2-Mcftiylnaphthalene
Tidal Freshwater 0 0081 2 Total PCBs, 2-Melhylnaphthalene
Tidal Freshwater 0 0167 2 Total PCBs, 2-Mefliylnaphthalene
Tidal Freshwater I 0174 2 Total PCBs, 2-Mefliylnaphthalene
62
-------
63
-------
Appendix G. Correlations between benthic bioindicators and salinity. Shown are the p values for
the statistical test and Pearson's correlation coefficients r values for each biomdicator. Values in gray
and bold face type are those selected for salinity correction.
Isopoda
Amphipoda
Ilaustonidae
Ampehscidae
Corophndae
Mollusca
Bivalvia
Gastropoda
Polychaeta
Spiomdae
Capitellidae
Ncrcidac
Ohgochaeta
Tubificidae
Deep Deposit Feeder
Suspension Feeder
Interface Feeder
Camivorc/Omnivore
Total Infauna
Epifauna
Abundance
p value r valui
00317 -014
00309 -014
0 0976 0 1 1
003S3 013
07316 -002
03157 006
0 3628 0 06
OOOOI 027
00011 021
0 0855 0 1 1
00019 020
00304 014
OOOOI -035
-------
Appendix H Regression relationships for salinity corrections of selected benthic biomdicators
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
Source
Model
Error
Corrected
D.F.
1
243
244
DF
1
243
244
DF
1
243
244
DF
3
241
244
DF
1
243
244
DF
1
243
244
DF
3
241
244
DF
1
243
244
DF
1
243
244
DF
1
243
244
Polvchaete Soecies Richness (Linear Relationship
Sum of Mean
Squares Square F Value Prob > F R-Squarcd Equation
567146 567146 10126 F R-Squared Equation
9759 9759 12733 <00001 034 004l+0027*Salimty
18624 0077
28384
Oligochaete Species Richness (Linear Rdationship)
Sum of Mean
Squares Square F Value Prob > F R-Squared Equation
351491 351.492 18043 O0001 043 3 13-0 1623*Salimty
473 374 1 948
824 866
Oligochaete Species Richncss(Polynomial Rdationship)
Sum of Mean
Squares Square F Value Prob. > F R-Squared Equation
476414 158805 10983 <00001 058 4 143-0 733*Sal+00463*Sal2-0 OOl'Sal5
348 452 1 446
824 866
Proportional Abundance of Oljgochaetes (Linear Relationship)
Sum of Mean
Squares Square F Value Prob > F R-Squared Equation
12076 12076 18170 <00001 043 0 624-0 030*Salimty
16149 0066
28225
Tubiticid Species Richness (Linear Relationship)
Sum ot Mean
Squares Square F Value Prob > F R-Squared Equation
364483 36448 204.01 <00001 045 2 865-0 165*Salimty
434 132 1787
798614
lubiticid Species Richness (Polynomial Relationship)
Sum ot Mean
Squares Square F Value Prob > F R-Squared Equation
511690 170563 143.26 <00001 064 3 958-0 786*Sal+00497*SalM) 00 l*SalJ
286924 1 191
798614
Proportional Abundance ot I ubihcids (Linear Relationship)
Sum of Mean
Squares Square F Value Prob > F R-Squared Equation
13539 13539 23919 <00001 050 0 561-003l9*Salmity
13755 0057
27294
Richness ot Deep Deposit Feeders (Linear Relationship)
Sum ot Mean
Squares Square F Value Prob > F R-Squared Equation
14675 146722 $359 <0006l 6 l83 061-0 104'Salinity
665 266 2 738
811 99
Richness ot Deep Deposit feeders (Polynomial Kciationsmp)
sum ot Mean
Squares Square F Value Prob > F R-Squared Equation
30265 10088 41/173 <00001 0.3V4 18-0737*Sal+0050*Sal'-0001*SalJ
509 34 211
811 99
65
-------
Chesapeake Bay Program
A.
Chesapeake Bay Program
410 Severn Ave. Suite 109
Annapolis Ml) 21403
------- |