EPA/620/R-94/001
                             October 1993
  LOUISIANIAN PROVINCE
DEMONSTRATION  REPORT
 EMAP -  ESTUARIES - 1991
           J. Kevin Summers^,
           John M. Macaulev ,
           Virginia D. Engl^r,
            Gary T. Brooks ,
        P. Thomas Heltmullejr, and
            A  Matt Adams3
   U.S. Environmental Protection Agency
    Environmental Research Laboratory
          1 Sabine Island Drive
         Gulf Breeze, FL 32561
  U.S. Environmental Protection Ajjency, Environmental Research Laboratory, Gulf Breeze, FL1
           Technical Resources, Inc., Gulf Breeze, FL2
          Computer Sciences Corporation, Gulf Breeze, FL
                              Printed on Recycled Paper

-------
                                 DISCLAIMER
This report represents data from a single year of field operations of the Environmental
Monitoring and Assessment Program (EMAP).  Because the probability-based scientific design
used by the EMAP necessitates multiple years of sampling, there may be significant levels of
uncertainty associated with some of these data. This uncertainty will decrease as the full
power of the approach is realized by the collection of data over several years. Similarly,
temporal changes and trends cannot be reported, as these require multiple years of
observation. Please note that this report contains data from research studies in only one
biogeographic region (Louisianian Province) collected in a short index period (July-August)
during a single year (1991).  Appropriate precautions should be exercised when using this
information for policy, regulatory or legislative purposes.

Mention of trade names or commercial products does not constitute endorsement or
recommendation for us.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page it

-------
                                  PREFACE

This document is a final draft of the report on the demonstration project completed by EMAP-
E (Environmental Monitoring and Assessment Program - Estuaries) in the Louisianian
Province in 1991. It is being distributed at this time for discussion and modification to finalize
the summary of the non-base monitoring activities in 1991.
The appropriate citation for this report is:
Summers, J.K., J.M. Macauley, V.D. Engle, G.T. Brooks, P.t. Heitmuller, A.M. Adams,   1993.
Louisianian  Province Demonstration Report : EMAP - Estuaries - 1991. U.S. Environmental
Protection Agency, Office of Research and Development, Environmental Research Laboratory,
Gulf Breeze, FL. EPA/600/R-93/xxx.
 Demonstration Report, EMAP-E Louisianian Province -1991
Page Hi

-------

-------
                       LOUISIANIAM PROVINCE
                      DEMONSTRATION REPORT
                      EMAP - ESTUARIES -1991

                           Table of Contents
DISCLAIMER     ._•	•	.!!
PREFACE        	• • •	  !"
TABLE OF CONTENTS 	  "J
EXECUTIVE SUMMARY	• • • • ^

1 INTRODUCTION	" • •	5
   1 1 OBJECTIVES OF THE 1991 LOUISIANIAN PROVINCE DEMONSTRATION		5
   1.2 PURPOSE AND ORGANIZATION OF THIS REPORT	- -	6

2 INDICES OF ECOLOGICAL CONDITION	J
   2.1 BENTHIC INDEX	• • •	J
      2.1.1 SAMPLING METHODS	™
      2.1.2 ANALYSIS METHODS	}ป
         2 1.2.1  DEVELOPMENT OF TEST DATA SET  	11
         2122  ADJUSTMENT FOR HABITAT GRADIENTS	14
         2 1.2.3  DISCRIMINANT ANALYSIS METHODS	16
      2 1.3 DISCRIMINANT ANALYSIS RESULTS .	16
      2.1.4 APPLICATIONS OF BENTHIC INDEX TO LOUISIANIAN PROVINCE	 .  19
   2.2 FISH INDEX  	•	1?
      2.2.1 SAMPLING METHODS	 •	*\
      2.2.2 ANALYSIS METHODS 	-	21
         2 2.2.1  ADJUSTMENT FOR HABITAT GRADIENTS	22
         2.2.2.2 DISCRIMINANT ANALYSIS METHODS	23
         2.2.2.3 DISCRIMINANT ANALYSIS RESULTS	 23
         2.2.2.4 APPLICATION OF THE FISH INDEX TO LOUISIANIAN PROVINCE	24
                                                                      oo
 3 SENSITIVITY OF INDICATORS	fj
    3.1  ECOLOGICAL INDICATORS	-	*9
      3.1.1 BENTHIC COMMUNITY INDICATORS	3O
      3.1.2 FISH COMMUNITY INDICATORS	32
      3.1.3 DISSOLVED OXYGEN INDICATORS	• • • • •	33
      3.1.4 HUMAN USE INDICATORS	-	-34
    3.2 HABITAT INDICATORS	• • -	*|
      3 2 1 WATER COLUMN HABITAT INDICATORS  	35
      3.2.2 SEDIMENT CHARACTERISTICS	•	36
    3.3 EXPOSURE INDICATORS	™
      3 3.1 DISSOLVED OXYGEN CONCENTRATIONS	37
       3.3.2 SEDIMENTTOXICITY	• • • • • •--••-• •-••• •••••••• - • • •- 3'


  Demonstration Report. EMAP-E Louisianian Province -  1991            Page v

-------
                       Table of Contents (continued)
      3.3.3 SEDIMENT CONTAMINANTS	     38
   3.4 CONFOUNDING FACTORS AFFECTING SENSITIVITY ANALYSES 	        43
      3.4.1 GEOGRAPHICAL GRADIENTS	           46
      3.4.2 SOURCE ASSOCIATIONS 	47

4 RESEARCH INDICATORS  	  60
   4.1 INDICATOR TESTING AND EVALUATION (ITE) SAMPLING DESIGN	     60
   4.2 HISTOPATHOLOGY OF TARGET SPECIES	           62
   4.3 SPLENIC MACROPHAGE AGGREGATES	          " " 65
   4.4 VERTEBRAL ABNORMALITIES	           70
   4.5 BLOOD CHEMISTRY	              "72
   4.6 BILE FLORESCENCE	       '   "     73
   4.7 STABLE ISOTOPES RATIOS	^74

5 STATISTICAL ASSOCIATIONS BETWEEN RESPONSE AND EXPOSURE INDICATORS ...  81
   5.1 ASSOCIATIONS WITH THE BENTHIC INDEX	      81
   5.2 ASSOCIATIONS WITH SEDIMENT TOXICITY	89
   5.3 ASSOCIATIONS BETWEEN SEDIMENT CONTAMINANTS WITH ACID VOLATILE
      SULFIDES AND TOTAL ORGANIC CARBON	     90
   5.4 ASSOCIATIONS WITH DISSOLVED OXYGEN	91

6 EVALUATION OF SAMPLING DESIGN ATTRIBUTES	94
   6.1 COMPARISON OF INDEX SAMPLING AND RANDOM SAMPLING	     94
      6.1.1  BENTHIC RESPONSE INDICATORS	       95
      6.1.2 FINFISH RESPONSE INDICATORS	          99
      6.1.3 CONTINUOUS DISSOLVED OXYGEN	      100
      6.1.4 HUMAN USE INDICATORS	105
      6.1.5 HABITAT INDICATORS	         109
      6.1.6 SEDIMENT CONTAMINANTS  	120
   6.2 EFFECTS OF GRID DENSITY OF PARAMETER ESTIMATION IN LARGE ESTUARIES  130
   6.3 DEGREE OF SPATIAL AUTOCORRELATION FOR SITES SELECTED BASED
      ON THE GRID 	  134
      6.3.1  SPATIAL AUTOCORRELATION AMONG ALL PROBABILITY-BASED SITES ...  141
      6.3.2 SPATIAL AUTOCORRELATION WITHIN THE LARGE ESTUARINE CLASS	143
      6.3.3 SPATIAL AUTOCORRELATION IN SELECTED LARGE ESTUARIES	145
   6.4 NEED FOR REPLICATION OF BENTHIC GRABS	 .  148

7 CONCLUSIONS	155

8 REFERENCES	159
Demonstration Report, EMAP-E Louisianian Province -1991
Page vi

-------
                        EXECUTIVE SUMMARY
The Environmental Monitoring and
Assessment Program (EMAP) is a
comprehensive environmental monitoring
network designed to:

•  Estimate the current status and trends
   in the condition of the nation's
   ecological resources on a regional
   basis, with known statistical confidence;

•  Seek associations among
   anthropogenic stress and ecological
   conditions; and

•  Provide periodic statistical summaries
   and interpretive reports on ecological
   status and trends to resource managers
   and the public.

The first stage in implementing  EMAP
involves conducting demonstration projects
for each Resource Group (ecosystem type;
e.g., forests, estuaries, agroecosystems).
Demonstration projects provide an
opportunity to 'illustrate the kinds of
assessments that can be accomplished
using EMAP data (Summers et al. 1993)
and to evaluate program design and
indicator selection (this report).  The 1991
Louisianian Demonstration represents the
second demonstration project conducted by
the EMAP-Estuaries Resource Group.

This report provides an evaluation of:

•  Development of aggregate indicators
   like a benthic index of estuarine
   integrity;

•  Testing of the sensitivity of all indicators
   by a direct comparison of their values
   at known locations of good and poor
 :  environmental quality;

•  Efficacy of new, relatively untested
 ,  indicators as representative of
   environmental condition (e.g., bile
   florescence, skeletal abnormalities,
   splenic macrophage aggregates);

•  Appropriateness of the spatial scale
   used in sampling (i.e., dimensions of
   the sampling grid used in large
   estuarine sampling);

•  Strengths of random and index
   sampling sites for small estuaries;

•  Preliminary associations between
   response indicators and exposure
   indicators as well as among exposure
   indicators; and,

•  Need for sample replication for benthic
   indicators in the Louisianian Province.
The 1991 Louisianian Province
Demonstration involved sampling visits to
202 sites from the Rio Grande, TX to
Anclote Anchorage, FL, as part of an
overall probability-based sampling design
to assess the status of the ecological
condition of Louisianian Province estuaries.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 1

-------
 A series of core and developmental
 indicators of overall, or portions of,
 estuarine condition was collected at each
 site.  Additional research indicators were
 collected at selected sites in order to test
 their abilities to discriminate between
 known good and poor condition.  All
 indicators represented either biotic integrity
 or some parameter of a healthy, diverse,
 sustainable biological community; pollutant
 exposure or some chemical aspect of the
 environment; habitat characterization; or,
 aesthetics relating to human use of
 estuarine resources.

 This evaluation of indicators and design for
 the Louisianian Province is based on a
 single year of information; thus, it is subject
 to potential year-specific phenomena such
 as climate fluctuations (1991 was a high
 precipitation year), contaminant spills (a
 major oil spill occurred in Galveston Bay in
 late 1990), or year-class strengths (no
 deviations known).  This assessment is
 preliminary and its findings should be
 confirmed by subsequent years of sampling
 in the Louisianian Province.

 A companion report delineating a statistical
 summary of the 1991  results has been
 produced (Summers et al. 19931) and
 should be used if the reader is interested in
 the ecological status of the estuaries of the
 Louisianian Province.  The following
 conclusions have been drawn from the
 monitoring data collected from the
 Louisianian Province in 1991 with regard to
 indicators and design:

 Response Indicator Development

•  An index of benthic community
   structure has been  developed that
    effectively discriminated between sites
    of known hypoxia and sediment
    contamination and reference sites. The
    strength and validity of this index will be
    assessed using the 1992 monitoring
    data.

 Sensitivity of Indicators

 •   Benthic index values, benthic species
    diversity and number of species were
    sensitive indicators of ecological
    condition  relating to hypoxia and
    sediment contamination.

 •   Number of finfish species and
    abundance/trawl were sensitive
    indicators of ecological condition in
    estuaries.

 •   Human use indicators and tissue
    contaminants in fish tissue were not
    indicative of ecological condition in
    estuaries where condition was defined
    as extent of hypoxia and sediment
    contamination.

 •   Total alkanes, pesticides and  heavy
    metals were significantly associated
    with observed hypoxia and sediment
    contamination.

•   Only a few PAHs and PCBs were  '
    associated with observed hypoxia  and
    sediment contamination.

•   Significant longitudinal gradients (East
    to West) in the  Louisianian Province
    existed for the number of fish
    species/trawl, minimum dissolved
    oxygen concentration, Secchi  depth,
    acid volatile sulfides, total organic
    carbon in sediments, total and
Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 2

-------
   numerous specific alkanes, and several
   heavy metals.

Research Indicators

•  Number of observed external fish
   pathologies/trawls were 2 to 3 times
   higher in regions  of hypoxia and
   industrial contamination and a
   significant longitudinal gradient existed
   with western province fish having five
   times the pathologies observed in
   eastern province  fish.

•  The percent area occupied by splenic
   macrophage aggregates was 4-9 times
   greater in regions of hypoxia and
   sediment contamination than in
   reference areas for pinfish and Atlantic
   croaker.

•  The rate of vertebral deformities was an
   order of magnitude higher in western
   province estuaries than in eastern ones
   with rates that were 3 to 7 times higher
   in areas of hypoxia and high industrial
   discharges.

•  Selected blood chemistry compounds,
   including c-reactive proteins, were
   significantly higher in brown bullheads
   from heavily contaminated areas than
   in catfish from reference areas.

•  Stable isotope and nutrient analysis in
   hypoxic areas indicated the high
   potential for eutrophic conditions
    resulting from algal production and
   decay.

Associations

•  The benthic index was strongly
   associated with sediment contaminant
   levels and somewhat associated with
   dissolved oxygen concentrations,
   sediment toxicity, and habitat variation
   in Redox potential discontinuity depth
   and salinity.

Statistical Design

•  Most response and exposure indicators
   showed no differences in distribution
   functions at the estuaries class level
   between index and randomly-placed
   sites.

•  Significant spatial autocorrelation exists
   among many of the response and
   exposure indicators used in the 1991
   Demonstration.

•  No significant differences in the
 ,  estimates of response and exposure
   indicators were observed, at the large
   estuarine class-level, between the base
   grid  density and an  enhanced density
   increasing the sample size by a factor
   of four;  however local or estuary-
   specific estimates were significantly
   different.
1  Summers, J.K., J.M. Macauley, P.T.
Heitmuller, V.D. Engle, and G.T. Brooks.
1993.  Statistical Summary:  EMAP-
Estuaries Louisianian Province-1991. U.S.
Environmental Protection Agency, Office of
Research and Development Environmental
Research Laboratory,  Gulf Breeze, FL.
EPA/600/R-93/001.
 Demonstration Report, EMAP-E Louisianian Province -1991
                                Page 3

-------

-------
                                SECTION 1
                             INTRODUCTION
The Environmental Monitoring and
Assessment Program (EMAP) is a national
program initiated by EPA's Office of
Research and Development (ORD). EMAP
was developed in response to the need for
information about the degree to which
existing pollution control programs and
policies protect the nation's ecological
resources.  EMAP is an integrated federal
program; ORD is coordinating the planning
and implementation of EMAP with other
federal agencies including the Agricultural
Research Service (ARS),  the Bureau of
Land Management (BLM); the U.S. Fish
and Wildlife Service (FWS), the Forest
Service (FS), the U.S. Geological  Survey
(USGS), and the National Oceanic and
Atmospheric Administration (NOAA).
These other agencies and offices
participate in the collection and analysis of
EMAP data and will use it to guide their
policy decisions, as appropriate.

EMAP-Estuaries (EMAP-E) represents one
portion of EMAP's efforts  in near coastal
environments.  These efforts are designed
to provide a quantitative assessment of the
regional extent of coastal  environmental
problems by measuring status and change
in selected condition indicators. The
results of this effort, with regard to the Gulf
of Mexico, were documented in the 1991
Louisianian Province Statistical Summary
(Summers et al. 1993). In addition to the
statistical summary, the 1991
Demonstration was designed to assess the
sensitivity of selected ecological indicators,
test the efficacy of "new" indicators, and
evaluate the appropriateness of several
elements of the statistical design.  This
Demonstration Report represents the
results of those evaluations.
1.1 OBJECTIVES OF THE 1991
   LOUISIANIAN PROVINCE
   DEMONSTRATION
The Louisianian Province Demonstration
was conducted in the summer of 1991
(July-August) to show the utility of
probability-based regional monitoring
programs for assessing the condition of
estuarine resources.  The sampling was
conducted from 9 July through 30 August
spanning 202 sites whose selection was
based on a probabalistic design.  The
specifics of the planning activities,
sampling design and indicator selection for
the 1991 Louisianian Province
Demonstration are documented in
Summers et al.(1991).  Specifics related to
trie conduct of the field sampling in 1991
can be found in Summers et al. (1992),
while the summary of ecological status for
the Louisianian Province can  be found in
 Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 5

-------
Summers et al. (1993).

The objectives of the 1991 Louisianian
Province Demonstration were to:

1)  demonstrate the value of regional
monitoring using a statistically unbiased
sampling design as a basis for assessing
the condition of estuarine resources;

2)  evaluate the ability of a selected suite
of ecological and environmental indicators
to discriminate among polluted and
unpolluted sites over a regional scale;

3)  obtain data on Louisianian Province
specific variability in ecological parameters;

4)  develop and refine analytical
procedures for using regional-scale
monitoring data to assess the ecological
status of estuaries and apply these
procedures to establish the  baseline
conditions in  the Louisianian Province;

5)  evaluate potential design constraints
imposed by use of the unbiased sampling
design;  and

6)  identify and  resolve logistical problems
associated with sampling estuarine
resources in primarily shallow estuaries
spanning over 1800 miles of coastline
within a 4 to 6 week sampling period.
The field activities report (Summers et al.
1993b) addressed Objective #6 while the
Statistical Summary (Summers et al.
1993a) addresses Objectives #1, #3, and
#4. This report addresses Objectives #2
and #5.
1.2  PURPOSE AND
ORGANIZATION OF THIS REPORT

The purpose of this report is to evaluate
the utility of the indicators selected for use
in the EMAP-Estuaries program for the
Louisianian Province and to assess key
elements of the sampling design with
regard to scale and sample site location.
In addition, the development of integrated
indicators of estuarine condition is
examined.

This report is organized in sections
addressing the primary purposes of this
report. Section 2 provides an  evaluation of
the development of integrated benthic and
fish indices of estuarine condition.  These
composite indicators are discussed in detail
with regard to the discriminant analyses
performed to create them.

Section 3 discusses the sensitivity of
selected indicators to differentiate between
known good and  poor ecological
conditions. This  analysis has been
performed on all  the indicators used in the
1991 Louisianian Province Demonstration.

Section 4 discusses the efficacy of
research  indicators tested at selected sites
within the Demonstration.  These indicators
include fish blood chemistry, bile
fluorescence, skeletal abnormalities,
splenic macrophage aggregates,
histopathology, and stable isotopes of
carbon and nitrogen.

Section 5 examines statistical associations
associated with several selected ecological
indicators. These associations are the first
steps in attempting to ascertain the
probable  general  causes for observed
Demonstration Report, EMAP-E Louisianian Province -1991
                               Page 6

-------
ecological degradation.

Section 6 examines the design issues
evaluated in the 1991 Demonstration.
These elements include spatial scale, index
sampling, and spatial autocorrelation.

Section 7 summarizes the conclusions that
can be drawn from these special elements
of the 1991  Louisianian Province
Demonstration.

Section 8 lists the literature cited in this
report.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 7

-------

-------
                                 SECTION 2

            INDICES  OF ECOLOGICAL CONDITION
Response indicators are characteristics of
the environment that provide quantitative
evidence of the status of ecological
resources and biological integrity of the
sample site from which they are drawn
(Messer 1990). Ecosystems with a high
degree of biotic integrity (i.e., healthy
ecosystems) are comprised of balanced
populations of indigenous organisms with
species compositions, diversity, and
functional organizations comparable to
natural habitats (Karr and Dudley 1981,
Karr et al. 1986).  Response indicators
could include measurements of the kinds
and abundances of biota present, the
health of the individual organisms, and the
sustainability of critical ecological
processes.  Numerous individual measures
have been collected to characterize these
portions of ecosystem condition; however,
no single measurement can be made to
depict overall estuarine health.

We have combined several of these
individual measures into a single measure
of estuarine condition with regard to: (1)
benthic response indicators, and (2) fish
response indicators. All composite
elements of these indices were collected
during the 1991 Louisianian Province
Demonstration.
2.1 BENTHIC INDEX

Benthic organisms are invertebrates that
live in the bottom sediments of aquatic
habitats.  In estuaries they are a major
trophic link between primary producers and
higher trophic levels, including fish,
shellfish, birds, and other wildlife (Carriker
1967, Rhoads 1974). They are a
particularly important source of food for
juvenile fish and crabs (Chao  and Musick
1977, Bell and Coull  1978, Holland et al.
1989). Estuarine benthos also have
important roles in ecological processes that
affect water quality and productivity.  For
example, the feeding and burrowing
activities of macrobenthos affect sediment
depositional patterns and chemical
transformations (Carriker 1967, Rhoads
1974, Kemp and Boynton 1981).  Benthic
feeding activities can remove  large
amounts of particulate materials from
shallow estuaries, which may  improve
water clarity (Cloem  1982, Officer et al.
1982, Holland et al. 1989).

The study of benthic communities in the
Gulf of Mexico estuaries has historically
followed two paths: (1) the identification of
environmental factors which influence
benthic community structure or (2) the
evaluation of the health of benthic
communities as an indication  of
environmental perturbations of either
natural or anthropogenic origin.  The
conclusions of the former show that salinity
and sediment type are among the most
 Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 9

-------
important factors which determine benthic
infaunal relationships in estuaries of the
Gulf of Mexico (Flint and Kalke 1985;
Gaston et al. 1988; Rabalais 1990;
Rakocinski et al. 1991). The health or
biological integrity of benthic communities
has traditionally been characterized by
measures of abundance, diversity, or the
presence of pollution indicator species.
These factors have been used as
indicators  of hypoxic conditions (Rosenberg
1977; Harper et al. 1981; Jurat et al. 1983),
organic pollution (Cook and Brinkhurst
1973; Grizzle 1984; Reish 1986; Tsutsumi
et al.  1991), and toxic contamination
(Holland et al. 1973; Rygg 1986).  It has
been  argued, however, that no single  factor
is sufficient to distinguish environmentally
degraded  from undegraded  areas
(Rosenberg 1977; Pearson and Rosenberg
1978; McManus and Pauly 1990).

The objective of this section is to detail the
methodology used to develop a benthic
index of estuarine integrity and to present
the results of subsequent analyses using
this preliminary index. The  multivariate
techniques of stepwise and  canonical
discriminant analysis are utilized to select
and test a subset of parameters which
describes the benthic community and
discriminates between degraded and
undegraded habitats.  The components of
the resultant benthic index are supported
by the literature as individual indicators of
environmental condition with some
limitations if used alone.
Louisianian Province Demonstration during
July-August 1991.  Of these, 110 base
sites were probabilistically located in large
estuaries (> 250 km2), small estuaries
(< 250 km ), and the tidal portion of the
Mississippi River (from New Orleans to the
delta).  Fifty-two sites were systematically
located in the areas of sediment deposition
within small estuaries'and the Mississippi
River. In addition, 16 sites were
specifically selected as indicator testing
and evaluation (ITE) sites based on
historically documented conditions of high
or low concentrations of dissolved oxygen,
high or low agricultural runoff of pesticides,
and high or low levels of industrial
contamination. Finally, the remaining four
sites were randomly selected from existing
base sites and these were revisited as a
quality control measure and a measure of
inter-index period variability.

At each of these sites, at least 3 replicate
benthic samples were collected using a
young-modified Van Veen grab that
sampled a surface area of 400 cm2. A
small core (60cc) was taken from each
grab for sediment characterization (i.e.,
total organic carbon and  percent silt-clay).
The remaining sample was sieved through
a 0.5 mm screen, preserved in 10%
formalin-rose bengal solution, and stored
for at least 30 days prior to processing.  In
the laboratory, macrobenthic samples were
transferred from formalin to an ethanol
solution and sorted, identified to lowest
practical taxonomic level, and counted.
2.1.1  SAMPLING METHODS

A total of 182 stations throughout the Gulf
of Mexico were sampled as part of the
2.1.2  ANALYSIS METHODS

The general approach for the development
of a benthic index was originally described
 Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 10

-------
in Weisberg et al. (1992) as part of the
EMAP-E Demonstration in the Virginian
Province (Cape Cod, MA to Cape Henry,
VA). The first step in the development of a
benthic index of environmental quality was
to choose a test data set consisting of sites
with known environmental conditions.  Our
original inclination was to use the a priori
selected ITE sites which were located
specifically  due to their combinations of
environmental conditions. Unfortunately,
many of these sites were not characterized
by the dissolved oxygen concentrations
(i.e., hypoxic or not) or sediment
contaminant levels (i.e., high or low)
indicated by historical analysis or local
expert judgements.  Only 9 of the 16 ITE
sites conformed to original expectations.
These 9 sites were insufficient to conduct
the benthic index analyses.  The test data
set consisted of these sites combined with
a subset of the remaining 173 sites which
represented either clearly undegraded or
degraded environment conditions based on
established criteria.  Using data described
in Summers et al. (1993b) for sediment
contaminant concentrations, sediment
toxicity,  and dissolved oxygen levels, test
sites were chosen to represent extremes
within a range of environmental conditions
that would adversely affect benthos.

Hypoxic conditions (dissolved oxygen
concentrations < 2 ppm) can cause a
reduction in abundance and number of
benthic species (Harper et al. 1981; Gaston
1985). Although many benthic species are
resistant to short periods of hypoxia
(Rosenberg 1977), extended or recurrent
periods  of hypoxia or anoxia lead to
mortality of the benthic community (Boesch
1985). The concentration of heavy metals
in the sediment was chosen as evidence  of
toxic contamination because median
effects threshold levels have been
established for most heavy metals (Long
and Morgan 1990) and heavy metals are
lethal to many benthic species which have
no method of detoxification (Bryant et al.
1984; McClusky et al. 1986).  The results
of sediment bioassasy using the amphipod,
Ampelisca abdita, and the mysid,
tylysidopsis bahia, were also used to
determine if a test site was degraded.
Sediment bioassays have been shown to
be very effective in  identifying toxic
sediments in combination with sediment
chemistry and physical sediment
Characterization (Chapman 1989).
2.1.2.1  DEVELOPMENT OF TEST
   DATA SET

Sites were classified as reference
(undegraded) sites based on the absence
of any natural or anthropogenic stress if:
(1) the minimum dissolved oxygen value
over a 24-hour period was greater than  3.0
ppm (Summers and Engle 1992), (2)
sediment concentrations for any
contaminant did not exceed the minimum
effects concentration established by Long
and Morgan (1990) (equals the
concentration at which 10% of the collected
data demonstrated adverse biological
effects) and (3) the percent survival for
Ampelisca abdita (10-day) or Mysidopsis
bahia (96-hour) in acute sediment
bioassays was indistinguishable from
controls.  Degraded sites were required to
exhibit the cumulative impacts of low
dissolved oxygen stress and contaminated
sediment stress by using the following
criteria:  (1) the minimum dissolved
oxygen concentration over a 24-hour period
 Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 11

-------
    Habitat
    Olkjohaline
         (< 5 ppt)
    Mesohaline
         (5-18 ppt)
    Polyhaltne
        (18-35 ppt)
   Marine
        (> 35 ppt)
 Reference Sites
 High Dissolved Oxygen
 and Low Contaminants

 Lake Calcasieu, La
   29ฐ59.35    93ฐ20.08
 San Antonio Bay, TX
   28ฐ 18.25    96ฐ39.89

 Bayou Grande, FL
   30=22.21    87ฐ16.23

 Back of Biloxi Bay, MS
   30ฐ24.85    88ฐ53.11

 South Bay. TX
   26ฐ03.05    97ฐ10.96
Lake Pelto, LA
   29ฐ04.13   90ฐ44.40

Ciystal Bay, FL
   28ฐ54.73   82ฐ44.24

Grand Bay, AL
   30ฐ22.89   88ฐ20.31

Pelican Bay, AL
   30ฐ13.99   88ฐ05.68

Matagorda Bay, TX
   28ฐ35.60   96ฐ25.35

Laguna Madre, TX
   26ฐ08.09   97ฐ16.04
 Degraded Sites
 Low Dissolved Oxygen
 and High Contaminants

 Belle River, LA
  29ฐ50.25      91ฐ09.05

 Amite River, LA
  3CM7.77      90ฐ35.98

 Houston Ship Channel, TX
  29ฐ44.06      95ฐ08.13

 Choctawhatchee River, FL
  30ฐ23.99      86ฐ08.02

 Tensaw River, AL
  30ฐ41.21      88ฐ00;06

 Lake Pontchartrain, LA
  30ฐ02.71     90ฐ10.03

 Mobile Bay, AL
  30ฐ37.00     88ฐ00.00

 Perdido Bay, FL
  30ฐ20.55     87ฐ27.50

 Watsons Bayou, FL
 30ฐ08.59     85ฐ37.96

 Mobile Bay, AL
 30ฐ26.18     88ฐ03.99
Garden Island Bay, LA
 29ฐ01.69     89ฐ06.50

Arroyo Colorado, TX
 26ฐ20.74     97ฐ25.69
Tablo 2.1 Location* of test sites used to develop the benthlc Index of estuarlne condition.
was < 2 ppm, (2)  sediment concentrations
for at least one sediment contaminant
exceeded  Long and Morgan's (1990) ER-M
value for biological response (concentration
at which 50% of collected data
                              demonstrated adverse biological effects),
                              and (3) acute sediment bioassays yielded a
                              control-adjusted survival rate of < 80%.

                              Table 2.1 lists the locations of sites used in
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                                 Page 12

-------
the development of the benthic index.
These sites were chosen not only to
represent undegraded and degraded
environmental conditions as described
above but also to cover the range of
salinities (0-42 ppt), sediment types (mud,
muddy sand, sand) and biogeographical
locations (east and west of the Mississippi
delta) inhabited by  benthos in  Gulf of
Mexico estuaries.  Table 2.2 lists the
candidate measures used to develop the
benthic index.  These measures were
chosen to represent the major structural
attributes of benthic assemblages.  In order
to ensure the values of abundance and
proportion approximated a normal
distribution, these values were adjusted
using a Iog10(value+1) transformation for
abundances and an arcsine transformation
for proportions.
      Measures of Biodiversity/Species Richness

      Shannon-Wiener Diversity Index
      Pielou's Evenness Index
      Mean Number of Species
      Mean Number of Polychaete Species

      Measures of Abundance

      Mean Benthic Abundance per site

      Measures of Taxonomic Composition

      Mean abundance of amphipods per site
      Proportion of total benthic abundance as amphipods   	
      Mean abundance of decapods per site
      Proportion of total benthic abundance as decapods
      Mean abundance of bivalves per site
      Proportion of total benthic abundance as bivalves
      Mean abundance of gastropods per site
      Proportion of total benthic abundance as gastropods
      Mean abundance of molluscs per site
      Proportion of total benthic abundance as molluscs
      Mean abundance of polychaetes per site
      Proportion of total benthic abundance as polychaetes       ,
      Mean abundance of capitellid polychaetes per site
      Proportion of total benthic abundance as capitellid polychaetes;
      Mean abundance of spionid polychaetes per site
       Proportion of total benthic abundance as spionid polychaetes
       Proportion of total polychaete abundance as spionid polychaetes
       Mean abundance of tubiflcid oligochaetes per site
       Proportion of total benthic abundance as tubificid oligochaetes
  Table 2.2 List of candidate benthic measures used to develop the benthic Index.
  Demonstration Report, EMAP-E Louisianian Province -1991
                                    Page 13

-------
 2.1.2.2  ADJUSTMENT FOR
   HABITAT GRADIENTS

 Measures of estuarine benthic
 abundance and species richness have
 been linked to natural gradients in
 salinity and sediment characteristics
 (Jurat et al. 1983; Flint and Kalke 1985;
 Gaston et al. 1988; Rabalais 1990;
 Rakocinski et al. 1991). Because our
 initial purpose in developing a benthic
 index is to attribute spatial differences in
 benthic community structure to
 contaminant and low dissolved oxygen
 stress, natural differences due to salinity
 or sediment gradients would confound
 the analysis and reduce the
 effectiveness of an index.  Pearson
 correlations were performed between all
 candidate measures and salinity,
 longitude of sampling site (as a measure
 of geographical gradient), percent
 silt/clay, and total organic carbon content
 of sediments.  Although many of the
 correlations were statistically significant at
 p < .05 due to the large sample size, only
 three correlations accounted for at least
 20% of the variation  (Table 2.3).  All three
 correlations involved measures of species
 richness or diversity with salinity. Unless
 the natural variation attributable to salinity
 is partitioned from the original data set, any
 developed index that used species
 richness or diversity as a component would
 include salinity variation as part of its
 definition of an algorithm separating
 anthropogenically degraded sites from
 undegraded sites (Ke., assigned status).

We used the method described in
Weisberg et al. (1992) to adjust the
candidate measures that were significantly
related to the salinity gradient (i.e.,
Habitat
Variable
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
• Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Mean Salinity
Longitude
Longitude
% Silt-Clay
TOC
Var
Variable Prob >
Mean # Species

|R|
0001
Diversity-Grab 1 .0001
Diversity-Grab 2 .0006
Diversity-Grab 3 .0001
Amphipod Abundance .0173
Decapod Abundance .0003
Polychaete Abundance .0001
Capitellid Abundance .0003
Spionid Abundance .0001
Tubificid Abundance .0004
% Molluscs .0103
% Gastropods .0219
% Polychaetes .0001
% Spionid/Polychaetes .0361
% Tubificids
Amphipod Abundance
% Capitellids
Amphipod Abundance
Diversity-Grab 2
0006
0347
0121
0001
0027

R
.46
.47
.35
.45
.24
.36
.44
.36
.38
-.35
-.26
-.23
.41
.21
-.34
-.21
.25
-.42
-.33

R2
.21
.22
.12
.20
.06
.13
.19
.13
.14
.12
.07
.05
.17
.04
.12
.05
.06
.17 .
.09
Table 2.3 Summary of significant correlations between habitat
Indicators and candidate benthic measures.
       Shannon-Wiener Diversity Index, mean
       number of species, and mean number of
       polychaete species) in order to remove
       variation due to that gradient.  The method
       employed a two step process whereby: (1)
       the expected value of diversity or number
       of species at any given salinity is
       estimated, and (2)  the percent deviation of
       each observed value from that expectation
       is calculated.
      The expected value for diversity or number
      of species was estimated by first
      calculating the 90th percentile of observed
      diversity or species richness values for
      overlapping intervals of 5 ppt salinity (e.g.,
      salinity intervals of 0-5 ppt, 1-6 ppt,  2-7 ppt,
      ..., 39-42 ppt).  A third order polynomial
      was then fit through the 90th percentile
Demonstration Report, EMAP-E Louisianian Province - 1991
                                    Page 14

-------
values and the midpoints of the salinity
intervals. We used the data from the
randomly-selected base stations to obtain
as good a fit as possible.

The polynomials calculated for each of the
three variables are shown in Table 2.4.
The explained variation for these
relationships ranged  between 0.84-0.87.
This process assumes that the 90
percentile represents the number of
species that would occur at undegraded
reference sites.  This assumption proved to
be reasonable because calculated
expected values did  not differ from those
observed for the undegraded reference
sites; however,  the observed number of
species at degraded sites consistently fell
   Expected Diversity =

   0.754 + (0.008 S) + (0.0016 S2) - O.OO3 S3


   Expected Number of Benthlc Species =


   13.908 - 1.115 S + 0.1244 S2 - 0.0019 S3
                               *

   Expected Number of Polychaete Species =


   4.751 - 0.2262 S + 0.0435 S2 - 0.0007 5>3



   S = Salinity (ppt)
Table 2.4 Polynomials used to adjust benthlc parameters
significantly related to salinity.


far below the regression  line defining the
90th percentile for expected number of
species (Fig. 2.1a,b,c).

The salinity-adjusted candidate measures
(proportion of expected diversity, proportion
of expected number of species, and
  proportion of expected mean number of
  polychaete species) were calculated by
  dividing the observed value by the
  expected value for each measure.  These
  values were substituted for the original,
  unadjusted variables in the list of candidate
                                                                15   20   25   30
                                                                  Solinity (ppt)
                                                                                35
                                                                                    40
                                                                                        45
Figure 2.1 a  Benthlc measures and salinity for Shannon-
Wiener diversity Index. (• = reference sites, •*• = degraded
sites, • = base sites).
   80
     0    5    10   15   20   25   30   35   40   45
Figure 2.1 b Benthlc measures and salinity for mean number
of species. (•= reference sites, *•=degraded sites, • = base
sites).
 Demonstration Report, EMAP-E Louisianian Province -1991
                                  Page 15

-------
                IS   20   29   30
                  Salinity (ppt)
Figure 2.1 c Bonthlc measures and salinity for mean number
of potychaซte specie*. (• = reference sites, •*• = degraded
sites, • = base sites).
measures (Table 2.2).  These new
measures were no longer significantly
correlated with salinity.
2.1.2.3 DISCRIMINANT ANALYSIS
   METHODS

Once the test data set consisting of values
for the candidate measures at both
reference and degraded sites was
established, stepwise and canonical
discriminant analyses were applied.
Stepwise discriminant analysis selects a
subset of the candidate measures which
best discriminates between the reference
and degraded sites.  Applying canonical
discriminant analysis to this subset of
variables yields a linear combination of the
quantitative variables that shows the most
substantial  difference between degraded
and undegraded sites (Williams  1983; SAS
Institute 1989).  This analysis also
produces estimates of classification
efficiency whereby sites that have been
misclassified as reference or degraded are
enumerated.  The coefficients derived from
canonical discriminant analysis are then
standardized to a mean of 0 and a
standard deviation of 1 in order to calculate
a discriminant score. Because discriminant
scores may result in a distribution among
negative and positive scores, they are then
normalized to a range of 0 to 10 using the
test data set in order to be more easily
understood and graphically represented.
2.1.3  DISCRIMINANT ANALYSIS
   RESULTS

The first stepwise discriminant analysis
suggested that only two benthic measures
were required to discriminate between
degraded and  undegraded sites:  (1)
proportion of expected number of
polychaete species and (2)  mean
abundance of decapods (Table 2.5).
However, the polychaete species-decapod
model resulted in an 18% misclassification
rate of reference sites (i.e., false positives)
and a 17% misciassification  rate of
degraded sites (i.e., false negatives). A
total of 55% of the total variance was
explained by this model.

Evaluation of the results of the stepwise
discriminant analysis suggested that
several of the stations in the test data set
may be misclassified as reference or
degraded (i.e.,  the conditions were not
sufficient to represent extremes of the
degradation gradient).  The original
classification of sites, following the
established criteria, had some locations
with acceptable dissolved oxygen
concentrations near the degraded criteria
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 16

-------
  Analysis Sequence

  Analysis I-All Stations
  Analysis Il-Analysis I with
  Misclassified Stations
  Removed

  Analysis Ill-Analysis II with
  Excluded Proportion of Expected
  Number of Polychaete
  Species
   Analysis IV-Analysis II with
   all but the first three
   variables removed
Candidate Measures
Selected by Analysis

1) Proportion of expected
  number of polychaete species
2) Mean abundance of decapods

1) Proportion of expected
  number of polychaete species
1) Proportion of expected
   diversity,
2) Percent of total
   abundance as tubificids
3) Percent of total abundance as
   bivalves
4) Proportion of expected
   number of species
5) Percent of total abundance as
   capitellids
6) Mean abundance of capitellids
7) Mean abundance of tubificids
8) Mean abundance of bivalves

1) .Proportion of expected
   diversity          •'
2) Percent of total     ,
   abundance as tubificids
3) Percent of total abundance as
   bivalves
 Percent
 False
Positives

  18.2
                                                               12.5
                              0.0
 Percent
  False
Negatives

  16.7
                                                                         0.0
                                        0.0
Canonical
   r2

   .55
                       ,81
                                                  .99
                                                                0.0
                                                                          0.0
                                                                                    .90
Table 2.5  Sequence of stepwlse and canonical discriminant analyses conducted for combining candidate benthic
measures Into an index.
(e.g., 2-3 ppm) or with low Ampelisca
survival (75-85%) in the sediment toxicity
tests although no corresponding high
sediment contaminant concentrations were
observed.  Four sites, classified as
degraded sites, experienced low Ampelisca
mortalities or low contaminant levels that
exceeded the criterion by only a small
margin for one of the criteria. Because
these 7 sites did not conform to all three
levels of the criteria established and were
either degraded or undegraded  in terms of
dissolved oxygen concentrations, sediment
contaminants, and sediment toxicity,
                  simultaneously, they were considered
                  misclassified and removed from the
                  analysis (i.e., they were misclassified to
                  begin With).

                  The revised data set was  re-evaluated
                  using stepwise discriminant analysis.  The
                  results of this second analysis suggested
                  that only one candidate measure was
                  required to discriminate between degraded
                  and reference sites: proportion of expected
                  number of polychaete species (Table 2.5).
                  Use of this variable alone accounted for
                  81% of the variability observed in the test
 Demonstration Report, EMAP-E Louisianian Province -1991
                                                    Page 17

-------
data set.  While a viable model, we
decided that a model based solely on the
presence or absence of a single indicator
group such  as polychaetes would severely
limit the effectiveness of discriminating
among reference and degraded sites
(Pearson and  Rosenberg  1978). Once the
model was applied to a continuous gradient
of effects rather than the extremes
portrayed by the test data set, reliance on
a single indicator could result in poor
discriminant power.  Because proportion of
expected polychaete species was
significantly correlated with the proportion
of expected diversity (r2 = 0.58, p < .05)
and diversity has some historical support
as an indicator of estuarine integrity, the
discriminant analysis was  repeated
eliminating the expected number of
polychaetes as a potential variable. This
new model  resulted in an  eight parameter
model (Table 2.5) with an overall r2 = 0.99
which included proportion  of expected
diversity as  the primary contributor.
Subsequent regression analysis revealed
that all but the first three variables entering
the stepwise discriminant analysis exhibited
significant collinearity and contributed  little
to the overall model  r-square (i.e., < 1-2%
for each of the five minor variables).
Eliminating these five variables resulted in
a model using only the three remaining
variables  (i.e., proportion of expected
diversity, proportion of total benthic
abundance  as tubificid oligochaetes, and
proportion of total benthic abundance as
bivalves) accounting for 90% of the
observed variation in the test data set with
no misclassifications. Inspection of the
discriminant scores for this final model
showed all reference sites to have values
>0 (0.3-4.3) and all degraded sites to have
values < 0 (-3.8 to -1.8).  This distribution
  of discriminant scores provided a clear
  demarcation between undegraded and
  degraded sites.

  These indicators were first standardized
  and then combined to make a composite
  benthic index using the following algorithm:

  Score =  (2.38 x D)-(1.67 x T)+(0.67 x B)

  where:

  D =    Proportion of expected diversity
         (Shannon-Weiner) at observed
         salinity
  T =    Proportion of total benthic
         abundance as tubificids
  B =    Proportion of total benthic
         abundance as bivalves

  The final development of the benthic index
  involved  calculating the discriminant scores
  for all sample sites and normalizing the
  calculated scores to a scale of 0 to 10.
  This normalization step was used to ease
  interpretability and graphical display with
       (8)
        5   10   15   20   25 •  30   35  40  45
                  Salinity (ppi)
Figure 2.2a Relationships between the benthic Index and
salinity for the base sites. Correlations were not significant
at p < .05.
Demonstration Report, EMAP-E Louisianian Province -1991
                                Page 18

-------
       (b)
                         .
                              4
                             e
                           . f
     0   tO  20   30  40  SO  80  70   80  90  100
                Silt-Cloy Content (i)
Figure 2.2b Relationships between the benthlc Index and
percent silt/day for tha base sites. Correlations were not
significant at p < .05.
the original range of -3.75 to 4.32 being
normalized to 0 to 10 with the break point
between degraded and undegraded sites
being at 4.1 (i.e., corresponding to roughly
0.0 in the discriminant score).

One final check was made to assure that
the benthic index was not related to habitat
parameters (e.g., salinity, silt-clay content),
and the index was not significantly related
to these parameters (Fig. 2.2a,b). This
final step was important because the
benthic index was designed to be an
indicator of degradation experienced at a
site (whether anthropogenic or climatic)
rather than inherent differences in benthic
community structure due to salinity and
sediment variations.
2.1.4  APPLICATION OF BENTHIC
   INDEX TO LOUISIANIAN
   PROVINCE

As was described in the Statistical
Summary (Summers et al. 1993b), the
application  of the benthic index to the 182
sites from which benthic data were
collected in the Louisianian Province
showed that 31% ฑ10% (estimate ฑ 95%
confidence interval) of the sediments in the
Gulf of Mexico estuaries contained benthic
community structures similar to those seen
in areas of  environmental stress (Fig. 2.3).
The percentage of area degraded (i.e., 4.1
benthic index) varied among the estuarine
classes with 25% ฑ12% of the benthic
communities in large estuaries; 41%ฑ15%
in small estuaries, and 80%ฑ25% in the
Mississippi  River being degraded (Fig. 2.4).

Although Alabama, Mississippi, and Texas
estuaries showed the greatest proportion of
degraded benthic communities among the
Gulf States (Fig. 2.4), the largest area of
degraded benthos was found  in Louisiana
and Texas (6200 km2).

2.2  FISH INDEX

Fish have several  advantages as potential
indicators of estuarine condition.  Because
fish have long life-spans and dominate the
tipper end of the food web, their responses
integrate  many short-term and small-scale
environmental perturbations. They are
known to respond  to most of the major
environmental problems of concern  in
estuaries (NOAA 1988), including
eutrophication, habitat modification,  and the
presence of pathogenic or toxic
contaminants.  For example, eutrophication
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 19

-------
                                      Benthic Index
                        INDEX < 4.1
                             31.2ซ
                                  INDEX 4.1-8.1
                                     24.8i
                                                           INDEX > 6.1
                                                           44.OS
Figure 2.3 Distribution of benthlc Index throughout randomly sampled base sites In the Gulf of Mexico.
           <=
           82
           O
           03
          Q_
                     LR     SR      RR
                     Station Class
FL
AL     MS      LA
Station Class
   Figure 2.4 Distribution of degraded benthlc resources (benthlc Index < 4.1) throughout randomly sampled sites in
   the Gulf of Mexico a) by estuary class and b) by state.  (LR=Iarge estuary, SR=small estuary, RR=Mlsslsslppl River)
Demonstration Report, EMAP-E Louisianian Province - 1991
                         Page 20

-------
can affect fish adversely by diminishing
dissolved oxygen concentrations to below
critical levels for growth, survival, or
structural development.  Habitat
modification, such as the loss of
submerged aquatic vegetation, has been
linked to decreased fish productivity
through loss of important nursery areas.
Toxic and pathogenic contaminants can
decrease fish growth, reproduction, or
survival and can make fish unsafe for
human consumption.

Fish are valuable indicators because of
their importance in determining public
perception of estuarine quality.  Reports of
fishery closures due to chemical and viral
contamination alarm the public.  In
addition, the public has an economic
interest in estuarine and near coastal
fisheries.  More than seven billion dollars
are spent annually in this country on
saltwater recreational fishing, the vast
majority of which occurs in estuaries or
within three  miles of the coast. Combining
recreational  and commercial fishing,
estuaries and near coastal waters account
for 70% of U.S. landings (NOAA 1987).

The use  of indices of overall health of the
fish assemblage occurring at a site has
gained great favor in freshwater
environments, where the Index of Biotic
Integrity (IBI)(Karr 1981) has become the
standard measure in several states for
defining environmental quality (Plafkin et al.
1989). The  IBI incorporates measures for
the individual, population, and assemblage
level.  To date, a validated index of
estuarine environmental quality based on
information about fish assemblages that is
applicable over broad geographic areas
has not been developed.  The major
difficulty with accomplishing this is that the
mobility and migratory behavior of
estuarine fish may limit our ability to
delineate a response to environmental
conditions at the site of collection.
2.2.1  SAMPLING METHODS

A total of 182 stations throughout the Gulf
of Mexico were sampled as part of the
Louisianian Province Demonstration during
July-August 1991.  These locations are the
same as those described above for benthic
sampling.

At each of these sites, a single fish sample
was collected using a 16-ft otter trawl
pulled at approximately 1 m/s for 10
minutes. If the first trawl resulted in no or
few fish, a second trawl was taken. The
contents of the fish trawls were identified to
species, enumerated, and examined  for
external pathological disorders (e.g.,
lesions, swellings, scoliosis). A subset (up
to 30 fish) of each species in the catch was
measured for length to the nearest 0.1
millimeter.  All fish displaying external
pathologies were forwarded to a
histopathology laboratory and up to ten
individuals of selected target species were
forwarded for contaminant analysis of
edible fillets.
2.2.2  ANALYSIS METHODS

The general approach for the development
of a fish index was originally described by
Weisberg et al. (1992) as part of the
EMAP-E Demonstration in the Virginian
Province.  However, the fish index
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 21

-------
developed in the Virginian Province was
determined to be inadequate to
characterize the differences between
degraded and undegraded sites in the
Virginian Province. The development of a
fish index in the Louisianian Province from
its  1991 Demonstration data uses the same
test data set developed for the benthic
index. This data set is based on
occurrence of hypoxia, elevated levels of
sediment contaminants, and sediment
toxicity. Table 2.1 listed the locations of
the sites used in the test data set.  These
sites  represent undegraded and degraded
environmental conditions, a variety of
habitats (e.g., salinities, depth zones, open
water vs. nearshore).  Table 2.6 shows the
list of candidate measures used to develop
   Species Richness and Diversity
      Abundance
      Shannon-Wiener Index
      Number of Species
      Number of Species to comprise 90% of the Catch

   Community Composition
      % Engraulsdae (Anchovies)
      % Clupeidae (Herrings)
      % AriWae (Catfish)
      % Ponaoidae (Shrimp)
      % Carangidao (Jacks)
      % Sciaonidae (Drums)
      % Sparidae (Porgies)
      % Bothidae (Flounders)
      % Tetraodontidae (Puffers)
      Presence of Endangered Species

   Trophic Dynamics
      % Top Carnivores
      % Planktivores
      % Benthic Invertivoras
      % Planktonic Inverts/ores

   Health of Individual Organisms
      Number of Gross External Pathologies in Catch
the fish index.  These measures were
chosen to represent the major ecological
attributes of fish assemblages and heajth
characteristics of individual fish.  In order to
ensure that values of abundance and
proportion approximated a normal
distribution, these values were adjusted
using a Iog10(value+1) transformation for
abundances and an arcsine transformation
for proportions.
2.2.2.1  ADJUSTMENT FOR
   HABITAT GRADIENTS

Because our purpose in the construction of
the fish index is to attribute spatial
differences in fish community structure to
contaminant and low dissolved oxygen
stress, natural differences due to habitat
types or gradients would confound
development of the index.  Pearson
correlations were performed between all
candidate measures and bottom salinity,
longitude of the site, total organic carbon
content of the sediment, and percent silt-
clay content of the sediments. Many
correlations were significant (Table 2.7) but
none of the correlations accounted for
more than 20% of the observed variability
in the test data set.

Because no correlations accounted for
more than 25% of the observed variability
in the test data set, no adjustmentsjpr
natural environmental gradients were made
to the test data set.
Table 2.6 Candidate variables for the development of fish
Index of estuarine condition.
Demonstration Report, EMAP-E Louisianian Province -1991
                               Page 22

-------
Habitat Number of Significant Number of
Variable Correlations Correlations
(P < .05) (R2 > .10)
Bottom Salinity
Longitude of Site
Total Organic Carbon
Percent Silt Clay
Bottom Dissolved Oxygen
Water Dept
Habitat Type
ALL
Percent of All
Correlations (133)
6
6
3
5
0
3
4
27
20%
2
0
0
2
0
1
3
8
6%
Number of
Correlations
(R2 > .20)
0
o
0
0
0
'. 0
0
0
0%
Table 2.7 Results of Pearson correlation analyses between variables used In flsh Index
development and habitat variables.
2.2.2.2  DISCRIMINANT ANALYSIS
   METHODS

Stepwise and canonical discriminant
analysis was applied to the test data set in
the same manner as with the benthic
index.  The discriminant results were
evaluated for classification error and the
discriminant scores were normalized to a
scale of 0 to 10.
2.2.2.3  DISCRIMINANT ANALYSIS
   RESULTS

The stepwise discriminant analysis showed
that 11 fish community measures could be
included to account for 97% of the
variability observed in the test data set.
However, 7 of the variables accounted for
               < 1% of that total
               variability even though 6
               of the 7 variables were
               significant (Table 2.8).
               Thus, only four variables
               (number of species,
               proportion of total catch
               as shrimp, number of
               species comprising 90%
               of catch, and proportion  of
               total catch as puffers)
               accounted for 95% of the
               variability between
               degraded and undegraded
               sites (Table 2.9).
               However, number of
               species alone accounts
               for 78% of the differences
               between degraded and
               reference sites.  The
               combination of number of
               species with proportion of
total catch as shrimp accounts for 88% of
these difference.  There were no
misclassifications within  the test data set.

Discriminant scores ranged from -3.7 to 6.1
with all reference sites having scores > 0
and all degraded sites having scores < 0.
This distribution provided a clear
demarcation between degraded and
undegraded sites.  These indicators were
first standardized  and then combined to
make a composite fish index using the
following algorithm:

      Score = (2.28 x SP)+(0.94 x S)

where:

SP  =   Number of species comprising the
       catch
S =   Proportion of total catch as shrimps.
Demonstration Report, EMAP-E Louisianian Province • 1991
                             Page 23

-------
Analysis Number

#1 -All Variables












#2 - Analysis #1 but
Including Only
Variables Contributing
> .02 to Squared
Canonical Correlation
#3 - Analysis #1 but
Including Only
Variables Contributing
> .05 to Squared
Canonical Correlation
Variables Included

1) Number of Species
2) % Shrimp
3) Number of Species
Comprising 90% Catch
4) % Puffers
5) % Anchovies
6) % Carangids
7) % Catfish
8) % Sciaenids
' 9) Number of External
Pathologies/Trawl
10) Benthic Index
11)%Sparids
1) Number of Species
2) % Shrimp
3) Number of Species
Comprising 90% Catch
4) % Puffers
1) Number of Species
2) % Shrimp



Squared Canonical
Correlation
0.999












0.945




0.876




The final development of the fish index
involved calculating the discriminant scores
for all sample sites and normalizing the
calculated scores.
2.2.2.4 APPLICATION OF THE
   FISH INDEX TO LOUISIANIAN
   PROVINCE

The application of the fish index to the 182
sites from which the fish data was collected
in the Louisianian Province showed that
66% of the estuarine waters of the Gulf of
Mexico had fish communities with low
number of species as characterized by
trawls (Fig. 2.5). Inspection of the
                 subpopulation estimates
                 for large estuaries, large
                 tidal rivers and small
                 estuaries showed that
                 69%, 78%, and 60% of
                 the fish communities in
                 each of these classes,
                 respectively, were
                 degraded (Fig. 2.6).
                 Inspection of the data, as
                 a preliminary validation,
                 suggests that the index is
                 relatively weak when
                 applied overall to the
                 Louisianian Province data
                 set because apparently
                 "healthy" sites were
                 categorized as degraded
                 because relatively few
                 species were collected.
                 For example, a trawl at a
                 site could produce 200-
                 300 fish but of only one or
or a nan index.       ^Q specjes (e.g., pinfish,
                 menhaden, catfish).  This
 site would be assessed as degraded  and
 yet be highly productive.

 Even though the analysis accounted for
 significant portions of the variability
 between degraded and undegraded sites in
 the test data set, there is sufficient
 evidence to suggest that its overall
 application results in errors, particularly
 with regard to trawls in open, large
 estuarine sites.  This poor agreement
 between the observed conditions and the
 fish index in large estuarine sites
 suggested that either: (1) separate indices
 are required for different habitats (i.e., open
 water versus small shallow estuarine
 systems) or (2) that a multi-species
 compositional index like the IBI might
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 24

-------
                                       FISH INDEX
                     INDEX < 3.3
                          65.2*
                                                           INDEX => 3.3
                                                           33.8i
Figure 2.5  Distribution of fish index throughout randomly sampled base sites In the Gulf of Mexico.
      

                                            
o
                                             0>
                                            Q_
                                                                                      TX
 Figure 2.9  Distribution of degraded fish resources (fish Index < 3.3) throughout randomly sampled sites In the Gulf of
   Mexico (a) by estuary class and (b) by state (LR=large estuary, SR=small estuary, RR=Mlsslssippl River.
Demonstration Report, EMAP-E Louisianian Province -1991
                                     Page 25

-------
categorize sites more correctly.  Relatively
few degraded sites were located in large
estuarine resources; thus, there may be too
few sites within individual estuarine
classifications that are degraded to permit
a valid analysis.  The collection of
additional sites in 1992 will be combined
with the 1991 data in an effort to develop
fish indices for individual habitat types.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 26

-------
                                  SECTION 3

                   SENSITIVITY OF INDICATORS
One of the most difficult activities within the
development of EMAP demonstrations and
the eventual implementation of estuarine
monitoring is the selection of indicators.
Although EMAP has developed a rigorous
tiered selection process (Knapp et al.
1990), the true utility of individual indicators
cannot be determined until they are tested
at the geographic and temporal scales
EMAP require.  In addition, numerous
indicators may be redundant in that they
reflect similar portions of the variability
observed in the environment.

One of the major objectives  of EMAP in its
early demonstrations is to evaluate the
tiered selection process for indicators and
to assess the utility of "known" ecological
indicators. "Known" in this sense refers to
the belief by many ecologists that the
indicators in  question can differentiate, to
some degree, between good and poor
ecological conditions. The indicators used
in the 1991 Louisianian Province
Demonstration (Table 3.1) were selected
through a rigorous process.  Numerous
ecologists discussed, at workshops and
through individual contacts, the pros and
cons of using these indicators, as well as
those indicators discarded, to ascertain
environmental condition.  The selection
process focused on the perceived ability of
the indicators to differentiate between good
and poor ecological conditions, the
directness of their interpretation,  their
logistical feasibility in terms of
implementation in the field from small
vessels with no laboratory facilities nearby,
and their relationship to the endpoints of
concern for EMAP-E (i.e., estuarine
condition based on ecological integrity and
aesthetic value).

In order to provide an assessment of the
ability of the selected indicators to
differentiate between good and poor sites,
we have analyzed all indicators with
reference to their ability to differentiate
between a subset of the 1991 monitoring
sites representing extreme values of
contaminants and hypoxia described in
Section 2.  The sites comprising this data
set are shown in Table 3.2. A good site is
characterized by relatively high dissolved
oxygen conditions throughout the day (i.e.,
> 4 ppm at all times), low contaminant
concentrations  in  sediments (alkanes,
PAHs, PCBs, pesticides,  or metals), no
sediment toxicity as shown by Ampelisca
bioassays, and, in some cases, the
existence of healthy seagrass beds.  A bad
site is characterized by hypoxic conditions
(DO < 2 ppm) at least 20% of the time and
sediment contaminant concentrations
greater than the ERL-L levels (i.e., those
concentrations  producing significant
ecological effects in 10% of studies
reviewed) described by Long and Morgan
(1990)(Table 3.3), and known high levels of
industrial point  source discharges or heavy
agricultural applications of pesticides within
the estuary's watershed.
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 27

-------
    Indicator

    Biotic Condition
Category
                          Core
                          Developmental
                          Research
   Abiotic Condition
Core
                         Developmental
 Indicator

 Benthic Community Composition
 Benthic Abundance,
 Benthic Biomass
 Benthic Index
 Fish Community Composition
 Fish Abundance
 Fish Lengths
 Fish Index
 Fish Pathology
 Tissue Contaminants
 Continuous Dissolved Oxygen Concentration
 Relative Abundance of Large Bivalves

 Histopathology of Fish
 Skeletal Abnormalities
 Blood Chemistry
 Bile Florescence
 Stable Isotope Ratios
 Splenic Macrophage Aggregates
 Liver Contaminant Concentration
 Whole Body Contaminant Concentration

 Salinity
 Temperature
 pH
 Water Depth
 Redox Potential  Discontinuity Layer Depth
 Percent Silt-Clay
 Percent Total Organic Carbon

 Marine Debris
 Percent Light Transmittance
Secchi Depth
Sediment Contaminants
Sediment Toxicity
Instantaneous Dissolved Oxygen Concentration
Tablo 3.1  Ecological Indicators (measured and calculated) used in the 1991 Loulslanlan Province Demonstration.
3.1  ECOLOGICAL INDICATORS

EMAP-E focuses on response indicators to
characterize the ecological status of the
estuarine resources of the Louisianian
Province. Biotic condition indicators are
ecological characteristics that integrate the
responses of living resources to specific or
multiple pollutants and other stresses.
                           Abiotic condition indicators are ecological
                           characteristics that can be linked,
                           conceptually, to decreases in estuarine
                           condition.  Within EMAP-E these indicators
                           include:  (1)  measures of attributes of the
                           benthic and fish communities as they relate
                           to biotic condition, (2) continuous
                           dissolved oxygen concentrations as they
                           relate to in the eutrophication process, and
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                             Page 28

-------
GOOD ECOLOGICAL CONDITION (REFERENCE SITES)
Location
Calcasieu Lake Canal, LA
Laguna Madre, TX
San Antonio Bay, TX
Matagorda Bay, TX
Crystal Bay, FL
Grand Bay, AL
South Bay, TX
Bayou Grande, FL
BAD ECOLOGICAL CONDITION
Pendido Bay, FL/AL
Watsons Bayou, FL
Choctawhatchee River, FL
Arroyo Colorado, TX
Mobile Bay, AL
Belle River, LA
Tensaw River, AL
Amite River, LA
Latitude
29
26
28
28
28
30
26
30
59.38'
8.00'
18.30'
35.58'
53.24'
22.89'
2.98'
22.14'
Longitude
93
97
96
96
82
88
97
87
20.03'
16.00'
39.90'
25.46'
44.41'
20.33'
10.98'
16.23'
(AFFECTED SITES)
30
30
30
26
30
29
30
30
27.08'
8.59'
24.00'
20.00'
26.17'
50.25'
41.35'
17.84'
87
85
86
97
88
91
88
90
22.60'
38.001
8.00V
25.76'
3.99'
9.05'
0.00'
33.60'
ecological Indicators. Criteria for good and bad condition are described In
text.
(3) the presence of marine debris, water
clarity and contaminant concentrations in
edible fish/shellfish tissue as they relate to
human uses of estuaries.  Only core and
developmental response indicators will be
assessed here; research response
indicators will be discussed in Section 4.
3.1.1  BENTHIC COMMUNITY
   INDICATORS

Of the benthic indicators collected during
the 1991 Louisianian Province
Demonstration, only mean number of
species in a grab, the proportion of total
abundance as bivalves, the biodiversity
         associated with a grab (Shannon-
         Weiner Index), and the benthic
         index described in Section 2
         consistently differentiated
         between  "good" (reference)and
         "bad" (affected) sites (Table 3.4).
         Mean benthic abundance;
         proportion of total abundance as
         amphipods, decapods,
         polychaetes, or tubificid
         oligochaetes; or the abundance
         of large bivalves did not
         significantly discriminate among
         the good and bad sites.
         However, mean abundance and
         proportion of total abundance as
         amphipods did show higher
         values at good sites than at bad
         sites although the variability
         associated  with these variables
         was rather  high.  In addition, the
         proportion of total abundance as
         polychaetes, decapods, and
         tubificid oligochaetes showed
         higher values at bad sites than at
         good sites but again high
variability resulted in  an inability to
discriminate between the sites.

Mean number of benthic species at good
sites were generally seven times higher at
good sites than at bad sites (Table 3.5)
and accounted for 45% of the variability
between these sites.  Calculation of the
Shannon-Weiner Diversity Index for each
benthic grab produced values at good sites
that were five times greater than at bad
sites and diversity accounted for 67% of
the variability observed among these sites
(Table 3.5).  These two indicators are
strongly colinear; thus only biodiversity
enters into the construction of the benthic
index.  The proportion of total benthic
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 29

-------
CHEMICAL ANALYTE CRITERION
Trace Elements (ppm)
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Silver
Tin
Zinc

ButytUns (ppb)
Monobutyldn
Dibutyltin
Tributylb'n

Polycycllc Chlorinated Blphyenyls
Total PCBs
FOB Congeners

Chlorinated Pesticides (ppb)
DDT
DDE
ODD
Total DDT
Aldrin
alpha-BHC
beta-BHC
delta-BHC
alpha-Chlordane
gamma-Chlordane
Dieldrin
Endrin
Hexachlorobenzene
Hoptachlor Epoxide
LJndane
Mirex
cis-Nonachtor
trans-Nonachlor
Oxychtordane
Alkane* and Isoprenolds
C10-C34
Priatane
Phytane

NA
2
33
5
80
70
NA
35
NA
0.15
30
1
NA
120


NA
NA
NA

(ppb)
SO
NA


1
2
2
3
NA
NA
NA
NA
0.5
0.5
0.02
0.02
NA
NA
NA
NA
NA
NA
NA

NA
NA
NA
CHEMICAL ANALYTE
Polynuclear Aromatic Hydrocarbons
Acenaphthene
Acenaphthylene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Biphenyl
Chtysene
C1-Chrysene
C2-Chrysene
CSrChrysene,
Dibenzo(a,h)anathracene
Dibenzothio
C1 -Dibenzothio.
CZ-Dibenzothkx
C3-Dibenzothio
Fluoranthene
. Fluorene
C1-Fluorene
C2,-Ruorene
C3-Ruorene
Naphthalene
G1 -Naphthalene
C2-Naphthalene
C3-Naphthalene
C4-Naphtha!ene
Perylene
Phenanthrene
C1-Phenanthrene
C2rPhenanthrene
CS-Phenanthrene
C4-Phenanthrene
Pyrene
(i)1 ,2,3,c-,d-pyrene
1 -methylnaphthalene
1 -methylphenanthrene
2-methylnaphthalene
2,3,5-trimettiylnaphthalene
2,6-dimethylnaphthalene
Total PAHs

'-





CRITERION
(Ppb)
150
NA
230
400
NA
NA
NA
NA
400
NA
NA
NA
60
NA
NA
NA
NA
600
35
NA
NA
NA
340,
NA
NA
NA
NA
NA
225
NA
NA
NA
NA
350
NA
NA
NA
65
NA
NA
4000







Table 3.3 Criteria values used to characterize degraded sediments (from Long and Morgan 1990). NA = Not Available.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 30

-------
Indicator

Mean Number of Species
Mean Abundance
Percent Amphipods
Percent Decapods
Percent Bivalves
Percent Polychaetes
Percent Tubificids
Abundance - Large Bivalves
Biodiversity
Benthic Index
* p<0.05
** p<0.01
***p<0.001
Mean Significance
Good
83.8
95.0
4.3
4.4
19.0
50.5
0.4
0.0
1.0
8.3



Bad
3.4
60.0 :
0.0
12.5
2.4
56.5 !
7.8
O-.O
0.3
3.9 *** ;



Table 3.4  Results of sensitivity tests for benthic community Indicators.
Good refers to ecological reference sites with minimal Impacts and bad sites
refers to ecological sites Impacted by hypoxia and contaminants.
abundance as bivalves was nine times
greater a good sites than at bad sites
accounting for 26% of the variability.
              3.1.2  FISH COMMUNITY
                INDICATORS

              Of the fish community indicators
              collected during the 1991
              Demonstration, the number of
              fish species in a trawl, the
              abundance of fish in a trawl, and
              the fish index described in
              Section 2 could significantly
              differentiate between good and
              bad sites (Table 3.6). None of
              the proportional taxonomic
              groups could successfully
              differentiate between good and
              bad sites.

              The number of fish species in  a
              trawl was 3 times greater  at good
              sites than bad sites accounting
              for 47% of the variability among
These two indicators (diversity and percent
bivalves) combine with percent tubificids
(not significant as a univariate indicator) to
construct a benthic index that accounts for
89% of the variability observed between
good and bad sites.
Indicator Pr> |F|
Mean Number of Species .0048
Percent Bivalves .0416
Biodiversity .0001
Benthic Index .0001
R2
0.45
0.26
0.67
0.89
Indicator

Number of Species
Abundance
Percent Catfish
Percent Puffers
Percent Sciaenids
Percent Clupeids
Percent Bothids
Fish Index
* p<.05
** p<.01
***p<.001

Good
9.5
63.9
10.3
0.6
16.5
3.2
0.5
5.9



Mean
Bad
3.8
20.6
8.6
0.0
25.4
11.1
4.2
1.0



Significance

**
*





*



Table 3.5  Significance levels and Fr associated with
benthic  community  Indicators that  successfully
discriminated between site types.
Table 3.6  Results of sensitivity tests for fish community
indicators.  Good refers to ecological sites with minimal
impacts and bad sites refers to ecological sites Impacted by
hypoxia and contaminants.
Demonstration Report, EMAP-E Louisianian Province - 1991
                                    Page 31

-------
   Indicator

   Number of Species
   Abundance
   Fish Index
Pr>

.007
.021
.017
0.47
0.35
0.53
 Table 3.7  Significance levels and FT associated with fish
 community Indicators that successfully discriminated between
 ปIto types.

 sites (Table 3.7).  Fish abundance was
 also about 3 times higher at good sites but
 only accounted for 35% of the observed
 variability. Although there were some
 problems associated with the
 implementation of the fish index as
 described in Section 2, the index  was
 about 6 times greater at good sites and
 accounted for 53% of the observed
 variation.
3.1.3 DISSOLVED OXYGEN
   INDICATORS

Continuous monitors were deployed at all
sites during the 1991 Demonstration for a
24-hour period that recorded dissolved
oxygen concentrations and percent
saturation every 15 minutes. Earlier
studies (Summers and Engle 1992)
showed that selected data collected from
these continuous records can be used to
assess the dissolved oxygen conditions at
a site for the July-August index period.
These characteristics were minimum
dissolved oxygen concentration and the
percentages of time between 6:00 PM and
6:00 AM that concentrations were below 2
ppm and below 5 ppm.  In addition,
instantaneous measures of dissolved
oxygen concentration at the time of
sampling (i.e., generally  between 7:00 AM
and 5:00 PM) were taken every meter from
            surface to bottom.
Indicator
Minimum Concentration
Percent of Time with:
Concentration < 2ppm
Concentration < Sppm
Instantaneous Bottom
Concentration
* p<,05
*• p<.01
"*p<.001
Mean Significance
Good Bad
5.5 0.5
0.0 81.8
12.0 100.0
5.9 2.3
Table 3.8 Result* of sensitivity tests for dissolved oxygen indicators. Good
refer* to ecological site* with minimal Impacts and bad sites refers to ecological
•Ites Impacted by hypoxla and contaminants.
                                  All four of these dissolved
                                  oxygen measures could
                                  differentiate with varying
                                  levels of significance between
                                  good and poor sites (Table
                                  3.8).  Although the
                                  instantaneous dissolved
                                  oxygen measures were'
                                  significantly higher at good
                                  sites  (factor of 2), the
                                  instantaneous measures did
                                  not account for nearly all the
                                  variability as did the
                                  continuous measures.
                                  Minimum dissolved  oxygen
                                  concentrations were five times
                                  lower at bad sites and
                                  accounted for 92% of the
                                  variability in the test data set
Demonstration Report, EMAP-E Louisianian Province -  1991
                                                   Page 32

-------
   Indicator              Pr> |F|

   Minimum Concentration        <.001

   Percent of Time with:
    Concentration < 2 ppm       <.001

    .Concentration < 5 ppm       <.001

   Instantaneous Bottom
    Concentration             <.001
 R2

0.92



0.81


0.81



0.69
 Table 3.9  Significance levels and R* associated with dissolved
 oxygon Indicators that successfully discriminated between site
 types.
(Table 3.9).  The percentage of time that
dissolved oxygen concentrations were
below 2 ppm was about 80 times longer at
poor sites than at good sites and
accounted for 81% of the variation while
the percentage of time less than 5 ppm
was eight times more frequent at bad sites
accounting for 81% of the variability. Thus,
while instantaneous measures of dissolved
oxygen can differentiate between clearly
poor sites and reference areas, continuous
measures are necessary to identify areas
that have cyclic conditions characterized by
low dissolved oxygen conditions at night
(Summers et al. 1993).
3.1.4. HUMAN USE INDICATORS

Of the human use indicators measured in
the 1991 Louisianian Province
Demonstration, none could significantly
differentiate between the good and bad
sites (Table 3.10).  Even though these
indicators could not differentiate between
the sites, all human use variables showed
the expected distribution of values. Marine
Indicator

Percent Occurrence of
Marine Debris
Percent Surface Light
Reaching 1 meter
Secchi Depth (m)
Tissue Contaminants:
ODD
DDE
DDT
Aldrin
Chlondane
Dieldrin
Endosulfan
Endrin
Heptachlor
Heptachlor Epoxide
Hexachlono benzene
Lindane
Toxaphene
Trans-Nonachlor
Total PCBs
Aluminum -
Arsenic
Cadmium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Tin
Zinc
* p<.05
** p<.01
*"p<001

Good

29.0

24.1
1.8

8.9
0.0
2.7
0.0
0.6
0.6
0.3
0.5
1.8
0.3
434.5
0.0
0.0
0.8
64.6
8.0
0.7
0.0
4.8
0.0
0.0
0.7
0.6
0.1
1.4
54.0



Mean
Bad

50.9

19.1
1.1

104.7
1.2
5.7
0.0
0.2
1.3
0.9
0.3
0.0
0.9
36.1
0.0
0.0
0.0
52.5
3.2
0.9
0.1
0.6
0.0
0.0
0.6
0.4
0.1
1.0
56.2



         Table 3.10 Results of sensitivity tests for human use
         aesthetic Indicators. Good refers to ecological sites with
         minimal Impacts and bad sites refers to ecological sites
         impacted by hypoxla and contaminants.
Demonstration Report, EMAP-E Louisianian Province -1991
                                      Page 33

-------
 debris occurred twice as frequently at bad
 sites than good sites. Water clarity,
 measured as the proportion of surface PAR
 (photosynthetically active radiation)
 reaching a depth of one meter and Secchi
 depth, was about 25% better at good sites
 than bad sites. Fish contaminant
 concentrations in edible fish tissues in
 Atlantic croaker, marine catfish, and shrimp
 were an average of seven times higher at
 bad sites than at good sites. For two
 contaminants, DDE (an intermediate DDT
 breakdown product) and heptachlor, no
 residues were found at any good sites
 while an average of 1.2  ppb and 1.8 ppb
 occurred at the bad sites, respectively.
 The high degree of variability in these
 human use indicators makes them poor
 discriminators of ecological condition.
 However these indicators represent
 important factors to human uses of
 estuaries that vary in the expected manner
 (e.g., better water clarity and lower tissue
 concentrations at good sites).
3.2 HABITAT INDICATORS

A number of habitat indicators were
measured during the 1991 Demonstration,
including indicators of water column and
sediment characteristics.  These indicators
were selected as possible indirect
measures of ecological condition or as
possible covariates in the evaluation of
response indicators.  We tested the
sensitivity of these indicators in order to
evaluate their correlative strength with a
response indicator.  For example, degree
of stratification (bottom salinity-surface
salinity), may be related to dissolved
oxygen concentration and show this
relationship consistently at good and poor
 sites.
3.2.1  WATER COLUMN HABITAT
   INDICATORS

Five water column habitat indicators were
measured during the .1991  demonstration.
These were instantaneous water
temperature, salinity, pH, and dissolved
oxygen and the degree of stratification.
Degree of stratification was measured as
the simple difference between bottom and
surface salinity rather than as sigma-T as
no surface to bottom temperature
differences occurred.

As expected, instantaneous bottom
dissolved oxygen concentrations (as
described in Section 3.1.3) differentiated
between  good and bad sites  (Table 3.11).
Average  instantaneous dissolved oxygen
concentrations on the bottom were 2.3 ppm
at bad sites  and 5.9 ppm at good sites
accounting for 69% of the variation in  good
and bad sites. The only other water
column habitat indicator that could
differentiate  between test data sites was
stratification.  Degree of stratification was
different between site types with average
conditions showing well mixed water
columns at good sites (0.45 ppt difference)
and stratified conditions at bad sites (9.3
ppt difference) accounting for 42%  of
observed variability.

Water temperature showed virtually the
same means at good and bad sites; 29.9 C
and 29.3  C,  respectively. Similarly, ranges
of bottom salinities in the test data  set were
almost identical with 2.6-40.5  ppt at good
sites and 0.0-39.9 ppt at bad  sites.  The
mean values of bottom pH were identical at
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 34

-------
   Indicator
                            Mean      Significance
                       Good     Bad
   Bottom Water Temperature   29.9.     29.3

   Bottom Salinity           22.4     17.6

   Bottom pH               7.9      7.9

   Percent of Surface Light
    Reaching 1 meter        24.1     19.2

   Stratification
    (Bottom-Surface Salinity)     0.5      9.3

   Bottom Instantaneous
    Dissolved Oxygen         5.9      2.3
   *  p<.05
   " p<.01
   "*p<.001
Table 3.11 Results of sensitivity tests for water column habitat indicators.
Good refers to ecological sites with minimal Impacts and bad sites refers to
ecological sites Impacted by hypoxia and contaminants.
          sediment could successfully
          discriminate between good and
          bad sites (Table 3.12).  Good
          sites had an average TOG
          percentage of 0.9% while bad
          sites showed heavily enriched
          conditions with 3.2% TOG values.
          TOG content of the sediments
          only accounted for 25% of the
          observed variation in the test
          data set.  Acid volatile sulfides in
          sediments were three times
          higher at bad sites than good
          sites although this difference was
          not significant.
         3.3  EXPOSURE
         INDICATORS
7.9 for good and bad sites.

3.2.2 SEDIMENT
   CHARACTERISTICS

As with the water column habitat indicators,
sediment characteristics are primarily
measured for their potential use in
evaluating  replicate differences as indirect
measures of condition and in benthic
community data and sediment contaminant
concentrations at a site. As part of the
1991  Demonstration, percentage of total
organic carbon, percent silt-clay content,
concentration of acid volatile sulfides, and
the depth of the redox potential
discontinuity layer were measured for each
sediment grab or composite.  Only the
percent total organic carbon  (TOG) in the
         Several measures of the
         magnitude and extent of pollution
         exposure were collected at each
site during the 1991 Louisianian Province
Demonstration in order to ascertain some
preliminary links between observed
estuarine degradation and observed
pollution exposure.  While this may be the
primary purpose of this data, we examined
the ability of these indicators to
discriminate between good and bad sites
because most historical monitoring data in
the Louisianian  Province is of this type. If
an exposure indicator could be found that
differentiated between site types then some
potential to examine trends backward in
time could exist. The exposure indicators
examined were  dissolved oxygen
concentrations; sediment toxicity as
measured by bioassay; sediment
concentrations of 27 alkanes and
isoprenoids, 44  polynuclear aromatic
hydrocarbons (PAHs), 20 polycyclic
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 35

-------
   Indicator
   Acid Volatile Sulfides

   Percent Total Organic
    Carbon

   Percent Silt-Clay

   Mean Depth of Redox
    Potential Discontinuity
    Layer (mm)
   * p<05
   " p<,01
   "*p<.001
     Mean      Significance
Good    Bad

  1.0     3.1
  0.9     3.2

 73.1    72.7




 43.4    51.9
Table 3.12 Result* of sensitivity tests for sediment habitat indicators.
Good refers to ecological sites with minimal Impacts and bad sites refers
to ecological sites Impacted by hypoxla and contaminants.
3.3.2  SEDIMENT TOXICITY

Sediment bioassays are the most
direct measure for estimating the
potential for contaminant-induced
effects in biological communities.
Direct measures of sediment
contaminant concentrations can be
misleading  because many
chemicals are bound tightly to
sediment particles, are chemically-
complexed, or the contaminant
resulting in  toxicity may be present
but not analyzed for.  Two types of
sediment toxicity bioassays were
run using sediment from the 1991
Demonstration sites.  These were
a 10-day Ampelisca (amphipod)
test and a 3-day Mysidopsis
(mysid) test.
chlorinated biphenyls (RGBs), 3 butyltins,
21 chlorinated pesticides, and 14 heavy
metals.
3.3.1 DISSOLVED OXYGEN
   CONCENTRATIONS

All measures, instantaneous and
continuous, of dissolved oxygen
concentration differentiated between good
and bad sites (see Table 3.8). These data,
discussed earlier, provide information at all
three levels of indicators: response (as a
eutrophication endpoint), habitat (as a
barrier to fish movement), and exposure
(as a factor for benthic mortality, growth,
and reproduction).
                        Neither bioassay could successfully
                        discriminate between the site categories in
                        the test data set although the results of
                        both bioassays tended towards expected
                        directions (Table 3.13).  Average amphipod
                        corrected survival rates of 98% and 83%
                        for good and bad sites, respectively, were
                        not significantly different because the
                        survival rates at bad sites ranged from 11-
                        104%. While the range of survival  rates at
                        good sites was significantly narrower (86-
                        109%), the paucity of low survival sites
                        resulted in an inability to differentiate
                        between good and bad sites.  Without
                        further poor survival sites, this comparison
                        may be misleading. Like the amphipod
                        bioassay, the mysid test also failed to
                        discriminate between good and bad sites
                        with average survival rates of 96%  and
                        81%, respectively.  The ranges of survival
                        rates were similar to that observed  for
                        amphipod with good sites having a  rather
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                      Rape 36

-------
   Indicator
                        Mean   Significance
                   Good    Bad
   10-day Ampelisca Test
    Survival Rate       97.8
   3-day Mysid Test
    Survival Rate
   * p<.05
   ** p<.01
   "*p<.001
                    96.2
                           82.7
                           80.7
Table 3.13 Results of sensitivity tests for sediment toxiclty
exposure Indicators.  Good refers to ecological sites with
minimal impacts and bad sites refers to ecological sites
Impacted by hypoxla and contaminants.
narrow range (90-103%) and bad sites
displaying a wide range (0-111%).  As with
amphipod, the mysid test appears to be
varying in the appropriate direction to
distinguish between good and bad sites but
without additional low survival sites in the
test data set, this indicator will not clearly
discriminate between site categories and
the likelihood of assessing a site as a false
negative exists.
3.3.3  SEDIMENT CONTAMINANTS

A composite sediment sample from each
location in the 1991 Louisianian Province
Demonstration was analyzed for the
contaminants listed in Table 3.14. The
contaminants can be categorized into five
groups: alkanes and  isoprenoids, PAHs,
PCBs,  pesticides, and heavy metals.  All of
the constituents of each group was tested
for indicator sensitivity as well as a
combined measure (e.g., total alkanes). In
addition, both measured concentrations
 (observed) and aluminum-corrected
 concentrations of heavy metals were
 evaluated.

 Fifteen alkanes and total alkanes and
 isoprenoids successfully discriminated
 between good and bad sites (Table 3.15).
 While both types of sites (good and bad)
 had mean alkanes concentrations below
 the degraded criterion determined in
 Summers et al. (1993b) of 7000 ppb, the
 average increase in individual alkane
 concentrations from good to bad sites was
 a factor of 12.5 and was a factor of 9.4 for
 total alkanes.

 Four of the 44 measured PAHs proved  to
 be good discriminators of ecological
 condition: benzo(b)fluoranthene,
 benzo(e)pyrene, benzo(g,h,i)perylene, and
 (i)1,2,3,c-,d-pyrene (Table 3.16).  Although
 the concentrations of PAHs in the
 sediments in the test data set were
 generally below criteria levels, the average
 difference between good and bad sites for
 these four PAHs was a factor of 8.9.  Total
 PAHs showed no significant difference
 between good and bad sites.

 Although all concentrations  were low, 2 of
 the 20 PCB congeners showed significant
 differences between good and  bad sites
 (Table 3.17): PCB 128 and  187.  PCB 187
 is a composite of three PCBs that coelute
 (PCB 187, 182, and 159).  While the
 concentrations of PCBs 128 and 187 were
 negligible  at good sites (0.02 ฑ and 0.01 ฑ
 ppb, respectively), the concentrations of
these contaminants at bad sites were
Significantly higher (0.19 ฑ 0.06 ppb for
 PCB 128 and 0.30 ฑ0.12 ppb for PCB
 187.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 37

-------
Potynuctear Aromatic Hydrocarbons (ppb)
Aconaphthone
Aconaphlhylone
Anthracene
Benzo(a)anlhracone
Benzo(a)pyreno
Bsnzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Biphenyl
Chrysene
C1 -Chrysene
C2-Chrysene
C3-Chtyssne
C4 -Chrysene
Dibenzo(a,h)anathracene
Dibenzolhio
C1-Dibonzothio
C2-Dibenzothio
C3-Dibenzothk>
Ruoranthene
Ruorene
CI-Ruorene
C2-Ruorene
C3-Ruorene
Naphthalene
C1 -Naphthalene
C2-Naphtha!ene
C3-Naphthalene
C4-Naphthalene
Peiylene
Phenanthrene
Cl-Phenanthrene
C2-Phenanthrene
C3-Phenanthrene
C4 -Phenanthrene
Pyrane
([)1,2,3,c-,d-pyrene
1 -methylnaphthalene
2-methylnaphthalene .
1 -mothylphenanthrene
2,3,5-trimethylnaphthalene
2,6-dJmethylnaphthalene
Total PAHs







Trace Elements (ppm)
Aluminum
Antimony
Arsenic
Cadmium
Chrimium
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Tin
Zinc

PCBs

Total PCBs
20 Congeners

Alkanes

Total Alkanes and Isoprenoids
C10-C34
Phytane
Pristane

ButylUns

Monobutyltin
Dibutyltin
Tributyltin

Pesticides

ODD
DDE
DDT
Aldrin
BHC
Chlordane
Dieldrin
Endosulfan
Endrin
Heptachlor
Heptachlor Epoxide
Hexachloro benzene
Mirex
Toxaphene
Trans-Nonachlor
  Table 3.14 List of contaminants analyzed from sediments during the 1991 Loulsianlan Province Demonstration.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 38

-------
Indicator

Total Alkanes
C10
C11
C12
C13
C14
CIS
C16
C17
Phytane
CIS
Pristane
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
* p<.05
** p<.01
***p<.001
Mean
Good
607.3
4.2
6.2
6.5
3.7
9.9
24.3
19.8
58.1
32.0
17.5
38.9
25.0
19.6
28.1
12.4
21.7
13.6
31.9
10.9
36.5
13.3
60.6
16.4
59.9
14.1
20.4
1.7




Bad
5715.5
14.3
8.4
9.1
7.0
10.5
55.9
18.0
265.5
54.1 :
20.0
37.7
48.1
31.5
70.8
41.3
152.3 !
88.9
271.1 :
120.3
477.1
183.3
1291.0
230.8
1335.8
220.6
585.7
66.4


• .
Significance R2

0.40
0.26






0.39






0.36
** 0.40
0.44
0.42
0.47
0.41
0.46
0.33
0.36
* 0.34
0.40
0.37
0.34



Table 3.15 Results of sensitivity tests for sediment alkane exposure indicators. Good refers to ecological sites with
minimal impacted and bad sites refers to ecological sites Impacted by hypoxla and contaminants.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 39

-------
Indicator

Aconaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)lluoranthono
Benzo(e)pyrene
Banzo(g,h,i)perylene
Banzo(k)ftuoranthene
Biphonyl
Chtysena
C1-Chrysene
C2-Chiysene
C3-Chrysena
C4-Chtysene
Dibenzo(a,h)anathracene
Dibenzothio
C1-Dibenzothio
C2-Dibenzothk>
C3-Dibenzothio
Fluoranlheno
Ruorena
C1-Fluorene
C2-Fluorene
CS-Ruorene
Naphthalene
C1 -Naphthalene
C2-Naphtha!ena
Ca-Naphthalene
C4-NaphthaIene
Peiytane
Cl-Phonan throne
C2-Phenanthrene
C3-Phenanthrene
C4-Phonanthrene
Pyrene
(i)1 ,2,3,c-,d-pyrBne
1 -methylnaphthalene
1 -mothy Iprronanthrene
2-mothylnaphthaIene
2,3,5-trimothy!naphtha!ene
2,6-cfimothytnaphth alone
* p<05
" p<.01
— p<.001
Mean
Good
0.3
0.5
0.7
1.7
1.9
2.6
2.1
2.2
1.5

2.4
3.1
4.3
2.5
2.9
0.5
9.1
9.0
12.0
9.0
3.8
1.0
4.2
11.2
T6;8;
1.4
2.5
4.2
12.2
18.9
1.4
13.2
is.a
11.6
9.1
5.9
1.7
1.0
3.3
1.5
3.9
1.8



Significance R2
Bad
1.5
3.0
6.1
20.7 .
19.5
33.6 * 0.33
20.0 ' * 0.27
•12.2 * 0.38
15.1

23.6
19.1 -
15.1
5.5
6.0
2.6
1.3.
4.1
1-1.3
15.2
35.4
2.5
3.8
10.1:
.17.1
115.7
1 9.5
1:1.9
15:5
16.2
58.5
15.1
22.2
20.7
21.1
41.0
12.8 *• 0.37
4.3
3.5
6.2
3.3
4.3



Table 3.16  Result* of ปenซltivlty tests for sediment polynuclear aromatic hydrocarbon Indicator.  Good refers to
ecological site* with minimal Impacts and bad sites refers to ecological sites Impacted by hypoxla and contaminants.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 40

-------
Indicator

PCB Congener #
8
18
28
44
52
66
101
105
110&77
118,108, & 149
126
128
138
153
170
180
187,182, & 159
195
205
209
* p<.05
" p<.01
***p<.001
Mean
Good

0.20
0.01
0.01
0.01
0'.04
0.00
0.04
0.00
0.07
0.03
0.00
0.02
0.36
0.02
0.11
0.04
0.01
0.01
0.02
0.03



Significance
Bad

0.11
0.02
0.14
0.21
0.48
0.14
0.74
0.29
1.35
0.40
0.08
0.19
0.85
0.65
0.47 :
0.29
0.30
0.03
0.03
0.02



R2











'

0.36




0.29






Table 3.17 Results of sensitivity tests for sediment polycydlc chlorinated biphenyl exposure indicators. Good refers
to  ecological sites with minimal Impacts  and bad sites refers to ecological sites impacted by hypoxia and
contaminants.
The concentration of tributyltin showed no
pattern related to site type (Table 3.18).  In
fact, observed tributyltin concentrations
were slightly higher at good sites than
those seen at bad sites.  Five of the 21
pesticides tested showed significant
differences between the site categories in
the test data set: 2,4-DDD, 4,4-DDE, total
DDT, total chlordane, and total  E3HC
(lindane)(Table3.19). Total DDT
concentrations in sediments between good
and bad sites  differed by a factor of 125,
while the DDT degradation products DDE
and ODD differed by factors of 58 and 170,
respectively.  Although the mean
concentrations of total chlordanes at good
and bad sites was low (< 1 ppb), the mean
concentrations at bad sites was 93 times
higher than at good sites.  Similarly, the
average concentrations of total BHC at bad
sites was only 0.03 ppb but this
concentration was three times greater than
concentrations observed at good sites.
Four pesticides were not observed at any
of the test data sites (beta-BHC, alpha- and
beta-endosulfan, and toxaphene);
Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 41

-------
 therefore, the analysis does not test the
 ability of these pesticides to differentiate
 among site types.

 Heavy metal concentrations were evaluated
 from two perspectives: (1) the difference
 between predicted and observed
 anthropogenic concentrations corrected
 using aluminum, and (2) the ratio between
 observed concentrations and criteria levels.
 Differences between predicted
 anthropogenic concentrations and
 observed aluminum-adjusted
 concentrations for eight of the fifteen
 metals differentiated between good and
 bad sites (Table 3.20).  Metal
 concentrations at good sites averaged 2.5
 ppm below the predicted anthropogenic
 levels while concentrations at bad sites
 averaged 7.2 ppm above that level. When
 compared to criteria level concentrations,
 six of the tested metals described above
 were also successful discriminators of site
 type (cadmium, chromium, nickel, lead, tin,
 and zinc)(Table 3.21).  Three metals
 (arsenic, silver, and copper) could
 differentiate between good and bad sites
          only when compared to a criterion level.
          Metals at bad sites occurred at higher ratio
          (observed:  criterion levels) than at good
          sites by an average factor of 3.3.  Total
          metals were 16 times higher at bad sites
          than at good sites.
          3.4 CONFOUNDING FACTORS
             AFFECTING SENSITIVITY
             ANALYSES

          We have described the results of crude
          sensitivity analyses on all of the core and
          developmental indicators used in the 1,991
          Louisianian Province Demonstration. In
          general, many of the indicators used could
          differentiate between sites with ecological
          status at extreme ends of the  condition
          gradient. Only habitat indicators (as
          expected) and PAH indicators could not
          differentiate effectively between opposite
          ends of the ecological condition gradient.
   Indicator


   Tributyltin
   * p<.05
   " p<.01
   *"p<001
     Mean
Good     Bad
Significance
R2
  3.1
                                               2.8
Table 3.18 Results of senaiHvlty tests for sediment tributyltin indicator. Good refers to ecological sites with minima
Impacts and bad sites refers to ecological sites impacted by hypoxia and contaminants.
Demonstration Report, EMAP-E Louisianian Province - 1991
                                       Page 42

-------
Indicator

2,4'DDD
4,4'DDD
2.4'DDE
4,4'DDE
2,4'DDT
4.4'DDT
Aldrin
alpha-BHC
beta-BHC
delta-BHC
gamma-BHC
alpha-Chlordane
gamma-Chlordane
Dieldrin
Endosulfan 1
Endosulfan II
Endrin
Hexachlorobenzene
Heptachlor
Heptachlor Epoxide
Mirex
cis-Nonachlor
trans-Nonachlor
Oxychlordane
Toxaphene
Total DDT
Total BHC
Total Chlordane
* p<.05
" p<.01
"*p<.001
Mean
Giood
0.00
0.01
0.00
0.06
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.07
0.01
0.01
•



Significance
Bad
0.17
1.28
1.96
3.38 **
0.03
2.05
0.02
0.01
0.00
0.01
o.oi ;
0.29
0.29
0.08
0.00
0.00
0.03
0.03
0.01
0.00
0.01
0.08
0.13
0.00
o.oo :
8.87
0.03 *
0.81


'
R2

0.25


0.44





















0.35
0.29
0.26



Table 3.19 Results of sensitivity tests for pesticide Indicators.  Good refers to ecological sites with minimal Impacts
and bad sites refers to ecological sites Impacted by hypoxia and contaminants.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 43

-------
Indicator

Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Manganese
Mercury
Nickel
Lead
Silver
Tin
Zinc
Average Number of
Metals Exceeding
Criterion Value
*p <05 " p<.01 "*p<.001

Good
0.18
-0.59
-0.03
-7.46
-1.39
-0.22
-63.69
0.11
-2.81
-1.61
-0.02
-0.15
-7.72


0.13

Mean
Bad
0.28
4.02
0.12
13.57
4.20
0.48
5.63
0.31
3.97
5.70
0.06
0.56
32.83


2.13
•
Significance R2



0.41
0.30

0.25


0.28
0.35

0.48
0.37


0:45

Table 3.20 Result* of sensitivity tests for the difference between sediment heavy metal Indicator (C)BS) and the
predicted concentration based of background aluminum level. Good refers to ecological sites with minimal impacts
and bad sites refers to ecolog
Indicator
Antimony
Arsenic
Cadmium
Chromium
Copper
Mercury
Nickel
Lead
Silver
Tin
Zinc
Average Number of
Metals Exceeding 95%
Confidence Interval
* p<.05 "p<.01 *"p<.001
Mean
Good
0.34
0.12
0.01
0.30
0.08
1.03
0.24
0.27
0.08
0.28
0.23
0.13

Bad
0.48
0.35
0.06
0.85
0.25
2.50
0.80
0.73
0.21
0.74
0.90
1.00

Significance R2
0.29
0.50
0.40
0.46
0.39
0.54
0.29
0.45
0.51


Table 3.21  Results of sensitivity tests for sediment heavy metal indicator (OBS) a proportion of critical value (CV).
Good refers to ecological sites with minimal Impacts and bad
Demonstration Report. EMAP-E Louisianian Province -1991
Page 44

-------
3.4.1  GEOGRAPHICAL
   GRADIENTS

One possible confounding factor that could
affect this assessment would be the
existence of a significant longitudinal
gradient in the indicators.  The good and
bad sites in the test data set are comprised
of about 50% of locations in each class
being east of the  Mississippi delta.
Therefore, if significant differences exist
due to site position relative to the
Mississippi delta (i.e., east or west), the
inability to discriminate between site types
might be due to an existing east-west
gradient.  In this case, sufficient good and
bad sites would have to be located east
and west of the Mississippi delta and the
sensitivity recomputed.  In the event that a
east-west gradient existed and the indicator
could still differentiate between good and
bad sites, the observed sensitivity to site
types might be even stronger.

Table 3.22 displays a summary of all
indicators by indicator class and shows
those that portray strong east-west
gradients  coupled with  their ability to
differentiate between site types. An
asterisk in the longitudinal gradient column
alone suggests that the indicator should be
retested after being subsetted by region.
Unfortunately, sufficient bad sites do not
exist east of the delta to permit this re-
analysis.  Possibly, the re-analysis can be
completed by combining the 1991  and
1992 data sets.  Few indicators showed
strong east-west gradients and an inability
to assess site differences. One  response
indicator (proportion of total benthic
abundance as polychaetes) showed
significantly higher mean proportion pf
polychaetes the benthic community in
eastern areas (77%) than in western areas
(28%) with the gradient explaining 40% of
the observed variability.  Two other
response indicators (number of fish species
and minimum dissolved oxygen
concentration) showed strong longitudinal
gradients, but they also easily differentiated
between site types. The average number
of fish species per trawl was two times
greater in the west than in the east.
Similarly, minimum dissolved  oxygen
concentrations were, on the average, 2.7
times lower in estuaries east of the
Mississippi River than those to the west.

One sediment habitat indicator
(concentration of acid volatile sulfides)
showed concentrations four times higher in
eastern than in western sediments and an
inability to differentiate between good and
bad sites.  This AVS gradient may
correspond to similar gradients seen for
arsenic, chromium, lead, selenium, tin, and
zinc; all of which showed significantly
higher concentrations east of the
Mississippi delta than to the west. A
gradient exists for total organic carbon
content of sediments in which estuarine
sediments east of the delta have 4 times
the organic carbon content of sediments to
the west. The existence of this gradient
suggests that the discriminatory strength of
TOG would be stronger if the effects of the
. east-west gradient were removed.

Only two exposure indicators (both
sediment contaminants) showed strong
longitudinal gradients without the ability to
differentiate between site types.  One
pesticide (gamma-BHC) was  not found in
the west and found at only a  few locations
in the east. This apparent gradient
probably represents the low number of
Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 45

-------
 sites where this pesticide is found rather
 than a true longitudinal gradient.  However,
 the same cannot be said for the one PAH
 (i.e., perylene) which showed a strong
 east-west gradient.  Perylene, though
 generally low in concentration, showed
 concentrations about 4 times higher in the
 east than in the west. No PAHs other than
 perylene showed a significant east-west
 gradient. Nearly all the alkanes showed
. strong east-west gradients with eastern
 sites averaging 3 to 5 times higher
 concentrations than those in the west;
 however the alkanes already show the
 capacity to differentiate between good and
 bad sites. Most heavy metals also showed
 strong east-west gradients with higher
 concentrations in the east half of the
 Louisianian Province.  Like alkanes, heavy
 metals are a good indicator of ecological
 condition so that correction for a east-west
 gradient should only increase that
 differentiating power.

 In summary, longitudinal gradients (whether
 natural or anthropogenic) do not appear to
 affect significantly the ability of the selected
 indicators to differentiate* between extreme
 ecological conditions.
 3.4.2  SOURCE ASSOCIATIONS

 While the test data set has been
 constructed to reflect extreme conditions
 (i.e., hypoxia, high sediment contaminant
 concentrations, and sediment toxicity),
 some indicators may be related to only one
 of these criteria.  Therefore, its ability to
 separate good and bad sites will only occur
 along that gradient and its ability to
 differentiate among several contaminant
 gradients would be weakened.  We
evaluated these relationships by testing the
indicators abilities to differentiate along
three individual exposure gradients (Table
3.23):

•  Dissolved oxygen conditions (DO)
   categorized as having significant versus
   minor sources of organic material
   effluents in the vicinity of the site,

•  Sediment contaminant conditions due to
   point sources as categorized as having
   numerous versus few industrial outfalls
   near the site (IND), and

•  Sediment contaminant conditions due to
   non-point sources as categorized as
   having  high versus  low pesticide
   applications in the counties of the
   watershed (AGRO).

No response indicators showed an inability
to differentiate ecological conditions due to
the overriding effect of a point source or
non-point source gradients.  Only one
biotic condition indicator  (proportion of total
catch as marine catfish) showed a strong
gradient associated with  hypoxia being five
times lower under hypoxic conditions
without any relationships to industrial or
agricultural contaminants. Only four other
biotic condition variables showed
relationships to a dissolved oxygen
gradient (fish index, and  three dissolved
oxygen indicators).  The  fish index values
are 4 times greater under high dissolved
oxygen conditions than under hypoxic
conditions. As expected, minimum
dissolved oxygen concentrations are lower
and the percentage of time dissolved
oxygen is less than 2 ppm or 5 ppm are
higher under hypoxic conditions.
 Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 46

-------
Only one habitat indicator (bottom pH)
showed a strong relationship to the
dissolved oxygen gradient with lower pHs
co-occurring with hypoxic conditions but no
significant relationship to industrial or
agricultural contaminant sources.
As might be expected, seven PAHs, total
PAHs, three alkanes, and two rnetals
showed significant relationships to
industrial point sources. In all these cases,
areas receiving high industrial discharges
were characterized by significantly higher
sediment contaminants than sites receiving
low discharges.  Nine PAHs, total PAHs,
total DDT, and one alkane showed similar
relationships to dissolved oxygen gradients
portraying higher concentrations under
hypoxic conditions. Only manganese
showed lower concentrations under hypoxic
conditions.

In summary, with the possible exception of
selected PAHs, region-wide gradients were
associated with the location of industrial
discharges and areas of high agricultural
loadings.  However, regional gradients of
dissolved oxygen may mask the* ability of
some industrial contaminants (e.g., PAHs)
to differentiate between sites receiving high
industrial contaminant discharges and
those receiving relatively few discharges.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 47

-------
Indicator
Discriminatory Longitudinal Significance
Power Gradient

Response Indicators
Number of Benthlc Species
Benthic Abundance
Percent Amphlpods
Percent Bivalves
Percent Polychaetes
Percent Tubifidds
Abundance of Large Bivalves
Biodiversity
Benthte Index
Number of Fish Species
Rsh Abundance
Percent Catfish
Percent Puffers
Percent Sciaenlds
Percent Clupeids
Percent BothWs
Fish Index
Minimum Dissolved Oxygen
Percent of Time < 2ppm
Percent of Time < 5ppm
Instantaneous Bottom
Dissolved Oxygen
Marina Debris
Percent Light Transmittance
Socchi Depth
Tissue Contaminants
ODD
DDE
DDT
Atdrin '
Chtordane
Dlokkin
Endosulfan
Endrin
Heptachlor
Hoptachtor Epoxida
Hoxachlofobenzene
LIndano
Mirex
Toxa phone
Trans- Nonachlor
Total PCBs
Aluminum
Arsenic
Cadmium
Cop DOT
Lead
Mercury *
Nickel
East

6.0
44.3
5.7
3.4
76.9
0.3
1.9
0.4
5.7
4.4
47.6
7.5
0.0
26.8
15.3
0.0
2.9
1.5
31.9
60:1

4.7
0.2
23.5
1.3

26.2
4.6
13.0
0.6
4.4
1.4
0.6
0.9
1.2
0.6
15.2
2.2
4.7
83.3
4.1
60.9
24.1
0.3
0.0
0.6
0.0
0.1
0.7
West

112.8
42.9
6.6
14.1
28.5 *
0.0
0.0
0.6
6.6
8.9
100.0
11.0
0.3
25.7
12.7
0.0
5.4
4.0
14.4
49.0

4.9
0.2
12.8
0.6

45.2
1.1
7.7
0.2
1.4
1.8
2.5
0.3
0.0
2.9
294.2
0.0
7.7
362.1
1.2
36.7
4.7
0.6
0.1
0.3
0.0
0.1
0.6
Table 3.22 Summary of results of sensitivity tests for EMAP-E Indicators in the Louislanlan Province for discriminatory
power (* = significant at p < .05) and the co-occurrence of a significant longitudinal east-west gradient (*).
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 48

-------
Indicator Discriminatory
Power

Response Indicators
Selenium
Silver
Tin
Zinc
Habitat Indicators
Water Temperature
Salinity
PH
Light Transmittance
Secchi Depth
Stratjfiation
Instantaneous Bottom DO *
Acid Volatile Sulfides
Percent Organic Carbon *
Percent Silt-Clay
RPO Depth
Exposure Indicators
Ampelisca Bioassay
Mysid Bioassay
Alkanes
Total
C10
C11
C12
C13
C14
CIS
C16
C17
Phytane
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
Longitudinal Significance
Gradient
East

0.6
0.2
1.8

1
29.2
18.1
7.7
23.5
1.3
6.1
4.7
4.2
3.5
81.3
45.1

98.1
101.9

5006.4
19.9
16.7
7.2
8.4
11.6
48.7
17,4
177.4
29.4
19.0
34.0
26.6
78.3
46.8
132.5
92.6
249.5
1 19.4 ,
400.8
184.0
1 190.7
211.5
1 162.3
181.5
443.5
53.2
West

0.6
0.1
1.2


30.3
19.3
7.8
12.8
0.6
5.9
4.9
1.0
0.9
80.7
29.7

95.4
99.0

1 100.3
3.4
9.6
6.1
8.9
18.9
48.2
41.7
123.1
136.3
59.9
82.0
45.4
55.2
15.9
18.7
9.5
24.5
11.4
36.1
14.1
64.3
23.4
66.1
24.1
28.7
2.3
                   —ป -• ——•— -• --..-.w.nj tests for EMAP-E Indicators in the Loulsianian Province for
discriminatory power (* = significant at p < .05) and the co-occurrence of a significant longitudinal east-west gradient
Demonstration Report, EMAP-E Louisianian Province -1991
Page 49

-------
Indicator Discriminatory
Power

Exposure Indicators
PAHs
Acenaphthene
Acenaphlhylene
Anthracene
Benzo(a)anthracene
Bonzo(a)pyrone
Bonzo(b)nuoranthene *
Benzo(e)pyrene
Benzo(g,n,i)perylene
B0nzo(k)f!uor an then e
Blphonyl
Chrysene
C1-Chrysene
C2-Chrysene
C3-Chrysana
C4-Chrysene
Dibonzo(a,h)anathracene
Oibenzothlo
C1-Dibonzothlo
C2-Dibenzothio
C3-Dibenzothfo
Fluoranlhene
Fluorene
C1-FIuorene
C2-Fluorene ' ;
C3-Fluorene
Naphthalene
C1 -Naphthalene
C2-Naphthalene
OS-Naphthalene
C4-Naphthalene
Perytene
C1-Phenanthrene
C2-Phenanthrene
C3-Phenan throne
C4-Phonanthrene
Pyreno
(IJI.S.S.c-.d-pyrene *
1 -moth ylnaphth alone
1-methylphenanthrene
2-melhylnaphthalene
2,3,5-trimethylnaphthalene
2,6-dimothylnaphthalene
TributylBn
Longitudinal Significance
Gradient
'East


1.6
3.0
6.7
23.0
22.4
32.3
20.8
13.0
20.0

26.1
21.7
17.3
5.7
6.2
2.9
2.0
5.2
11.8
15.2
44.2
3.1
5.2
11.6
17.8
20.5
20.5
23.6
27.0
25.3
46.7
19.5
26.5
22.8
; 23.5
45.8
14.3
10.6
4.8
15.2
6.6
8.2
2.9
West


0.7
1.2
2.4
6.3
8.2
9.5
7.6 .
9.2 . !
5.9

9.2
9.4
13.0
8.5 • :•
5.8
1.6 ,
4.3 -
20.6
38.9
27.2
14.1
2.6
15.3
55.1
62.4
2.1
2.1
6.8
40.7
90.8
11.4
59.9
72.2
42.9
22.1
20.0
6.5
1.4 ' '
12.2
2.2
13.5
2.5
2.1
Table 3.22{cont) Summary of results of sensitivity tests for EMAP-E Indicators In the Loulslanlan Province for
discriminatory power (* = significant at p < .05) and the co-occurrence of a significant longitudinal east-west gradient
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 50

-------
Indicator Discriminatory
Power

Exposure Indicators
PCBs
Total
Congener 8
Congener 18
Congener 28
Congener 44
Congener 52
Congener 66
Congener 101
Congener 105
Congener 1 10/77
Congener 118/108/149
Congener 126
Congener 128 *
Congener 138
Congener 153
Congener 170
Congener 180
Congener 187/182/159
Congener 195
Congener 206
Congener 209
Pesticides
2,4' ODD
4,4' ODD
2,4' DDE
4,4' DDE
2,4' DDT
4,4' DDT
Aldrin
alpha-BHC
beta-BHC
delta-BHC
gamma-BHC
alpha-Chlordane
gamma-Chlordane
Dieldrin
Endosulfan I
Endosulfan II
Endrin
Hexachlorobenzene
Heptachlor
Heptachlor Epoxide
Mirex
cis-Nonachlor
trans-Nonachlor
Oxychlordane
Longitudinal Significance
Gradient
East


44.8
0.4
0.1
0.4
0.4
1.1;
0.7
2.2
0.9
3.3
1.6
0.0
0.4
2.1
1.6
0.4
0.3
0.3
0.0,
0.0
0.0

0.2;
1.2
0.9
2.6
0.0
0.3
0.0
0.0
0.0
0.0
0.1
0.1
0.1
0.1,
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.0
West


19.2
0.1
0.0
0.0
0.0
0.1
0.0
0.2
0.1
0.2
0.1
0.0
0.0
0.3
0.1
0.3
0.1
0.0
0.0
0.0
1.2

0.1
0.3
0.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.0
0.1
0.0
0.0
0.0
1.6
0.0
0.5
0.0
0.1
0.1
0.0
Table 3.22(cont) Summary of results of sensitivity tests for EMAP-E Indicators In the Loulslanlan Province for
discriminatory power (* = significant at p < .05) and the co-occurrence of a significant longitudinal east-west gradient
(*)•
Demonstration Report, EMAP-E Louisianian Province -1991
Page 51

-------
Indicator Discriminatory
Power

Exposure Indicators
Pesticides
Toxaphene
Total DDT
Total BHC
Total Chtordane *
Heavy Metals(OBS-P)1
Antimony
Arsenic
Cadmium *
Chromium *
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Tin
Zinc
Heavy Metals (%CV)2
Antimony
Arsenic. *
Cadmium *
Chromium *
Copper *
Lead
Mercury
Nickel
Silver
Tin
Zinc
1 Observed - Predicted concentrations based on aluminum
2 Percentage of criterion value
Longitudinal Significance
Gradient
East


0.0
5.2
0.0
0.4

0.2
7.44
0.0
23.6
3.8
1.0
6.0
-111.3
0.3
1.7
0.4
0.0
0.6
15.5

45.6
45.5
3.6
97.2
24.2
74.3
227.5
72.2
18.7
73.3
75.8
regression

West


0.0
1.4
0.0
0.7

0.1
0.3
0.0
-4.4
-0.1
-0.2
-1.5
-81.0
0.2
-1.1
0.0
0.0
-0.1
-9.1

35.0
20.2 •-••
2.7
50.0 "'•
14.0 ;
40.8
168.3
46.3
11.2
40.8
37.6
(positive value denoted anthropogenic sources)

Table 3.22(cont) Summary of results of sensitivity tests for EMAP-E Indicators In the Loulslanlan Province for
discriminatory power (* = significant at p < .05) and the co-occurrence of a significant longitudinal east-west gradient
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 52

-------
   Indicator
                            Discriminatory Power
   Response Indicators
   Number of Benthic Species
    Benthic Abundance
    Percent Amphipods
    Percent Decapods
    Percent Bivalves
    Percent Polychaetes
    Percent Tubificids
    Abundance of Large Bivalves
    Biodiversity
    Benthic Index

    Number of Fish Species
    Fish Abundance
    Percent Catfish
    Percent Puffers
    Percent Sciaenids
    Percent Clupeids
    Percent Bothids
    Fish Index

    Minimum Dissolved Oxygen
    Percent of Time < 2 ppm
    Percent of Time < 5
    Instantaneous Bottom
    Dissolved Oxygen

    Marine Debris
    Percent Light Transmittance
    Secchi Depth
    Tissue Contaminants
     DDD
     DDE
     DDT
     Aldrin
     Chlordane
     Dieldrin
     Endosulfan
     Endrin
     Heptachlor
     Heptachlor Epoxide
     Hexachlorobenzene
     Lindane
     Mirex
     Toxaphene
     Trans-Nonachlor
     Total PCBs
     Aluminum
     Arsenic
     Cadmium
 :   Source
 : Gradients
DO   IND   AGRO
Table 3.23  Summary of results of sensitivity tests for EMAP-E Indicators In the Louisianian Province significant
discriminatory power (*) and the co-occuirrence of significant source gradients (+).
Demonstration Report, EMAP-E Louisianian Province - 1991
                                      Page 53

-------
   Indicator
   Response Indicators
    Copper
    Load
    Mercury
    Nickel
    Selenium
    Silver
    Tin
    Zinc

   Habitat Indicators
    Water Temperature
    Salinity
    PH
    Light Transmittanco
    Secchi Depth
    Stratification
    Instantaneous Bottom DO
    Acid Volatile Sulfides
    Percent Organic Carbon
    Percent Silt-Clay
    RPD Depth
   Exposure Indicators
    Ampelisca Bioassay
    Mysid Bfoassay
    Alkanes
    Total
    C10
    C11
    C12
    C13
    C14
    C15
    C16
    C17
    Phytane
    C18
    Pristane
    C19
    C20
    C21
    C22
    C23
    C24
    C25
    C26
    C27
    C28
    C29
    C30
Discriminatory Power
   Source
  Gradients
DO   IND   AGRO
Table 3.23{cont) Summary of results of sensitivity tests for EMAP-E Indicators In the Loulslanlan Province significant
discriminatory power (*) and the co-occurrence of significant source gradients (+).
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                                 Page 54

-------
    Indicator
    Exposure Indicators
    C31
    C32
    C33
    C34
    PAHs
    Acenaphthene
    Acenaphthylene
    Anthracene
    Benzo(a)anthracene
    Benzo(a)pyrene
    Benzo(b)fluoranthene
    Benzo(e)pyrene
    Benzo(g,h,i)perylene
    Benzo(k)fluoranthene
    Biphenyl
    C1-Chrysene
    C2-Chrysene
    C3-Chrysene
    C4-Chrysene
    Oibenzo(a,h)anathracene
    Dibenzothio
    C1-Dibenzothio
    C2-Dibenzothio
    C3-Oibenzothk>
    Ruoranthene
    Fluorene
    CI-Ruorene
    C2-Fluorene
    C3-Fluorene
    Naphthalene
    C1-Naphthalene
    C2-Naphthalene
    C3-Naphthalene
    C4-Naphthalene
    Perylene
    Phenanthrene
    C1 -Phenanthrene
    C2-Phenanthrene
    C3-Phenanthrene
    C4-Phenanthrene
    Pyrene
    (i)1,2,3,c-,d-pyrene
    1-methylnaphthalena
    1 -methyl phenanthrene
    2-methylnaphthalene
    2,3,5-trimethylnaphthalene
    2,6-dimethylnaphthalene
    Total PAH
   Tributyltin
                            Discriminatory Power
    Source
   Gradients
DO  IND   AGRO
Table 3.23(cont) Summary of results of sensitivity tests for EMAP-E indicators In the LoulslanEan Province significant
discriminatory power (*) and the co-occurrence of significant source gradients (+).
Demonstration Report, EMAP-E Louisianian Province -1991
                                      Page 55

-------
       Indicator
       Exposure Indicators
       PCBs
       Total
       Congener 8
       Congener 18
       Congener 28
       Congener 44
       Congener 52
       Congener 66
       Congener 101
       Congener 105
       Congener 110/77
       Congener 118/108/149
       Congener 126
       Congener 128
       Congener 138
       Congener 153
       Congener 170
       Congener 180
       Congener 187/182/159
       Congener 195
       Congener 206
       Congener 209
      PesHddes
       2,4' ODD
       4.4' ODD
       2,4' DDE
       4.4' DDE
       2.4' DDT
       4.4' DDT
       AWrfn
       alpha-BMC
       beta-BHC
       dolta-BHC
       gamma-BHC
       alpha-Chtordane
       gamma-Chlordane
       Dtokfrln
       Endosulfan I
       Endosulfan II
       Endrin
       Hexachlorobonzene
       Heptaehlor
       Hoptachtor Epoxide
       Mirex
       c!s-Nonachlor
       bans-Nonachtor
       Oxychlcxdano
       Toxaphone
       Total DDT
       Total BHC
       Total Chlordane
                               Discriminatory Power
    Source
  Gradients
DO   IND   AGRO
   Table 3.23(cont) Summary of results of sensitivity tests for EMAP-E Indicators In the Loulslanlan Province significant
   discriminatory power (*) and the co-occurrence of significant source gradients (+).
Demonstration Report,  EMAP-E Louisianian Province -1991
                                  Page 56

-------
   Indicator


   Exposure Indicator* (Cont)

    Heavy Metals (OBS-P)1

    Antimony
    Arsenic
    Cadmium
    Chromium
    Copper
    Iron
    Lead
    Manganese
    Mercury
    Nickel
    Selenium
    Silver
    Tin
    Zinc
Discriminatory Power
   Source
1  Gradient*
DO  IND   AGRO
   1 Observed - Predicted concentrations based on aluminum regression (positive value denoted anthropogenic sources)


   Heavy Metals (%CV)1
    Antimony
    Arsenic
    Cadmium
    Chromium
    Copper
    Lead
    Mercury
    Nickel
    Silver
    Tin
    Zinc
   1 Percentage of criterion value
Table 3.23(cont) Summary of result* of itensltivlty tests for EMAP-E Indicators In the Loulslanlan Province significant
discriminatory power (*) and the co-occurrence of significant source gradients (+).
Demonstration Report,  EMAP-E Louisianian Province -1991
                                                                Page 57

-------

-------
                                 SECTION  4

                      RESEARCH  INDICATORS
A key element of the EMAP indicator
strategy is the identification and testing of
new indicators that may:

• Address an ecological issue not
  presently examined by one of the
  present indicators,

• Replace a present indicator with a more
  sensitive or direct indicator, or

• Allow a regional interpretation of the
  data because the new indicator does not
  respond to natural gradients that
  confound regional evaluations (e.g.,
  salinity, longitudinal).
Six research indicators were identified for
indicator testing and logistical evaluation
during the 1991 Louisianian Province
Demonstration. Indicator testing refers to a
determination of the scientific credibility of
the indicator and its ability to differentiate
between impacted and reference sites at
population rather than suborganismal
levels.  Logistical  evaluation refers to the
ability of EMAP-E crews to collect the
samples required for the indicator in a
manner consistent with the EMAP field
protocols.  A research indicator must be
compatible with both types of testing in
order to be considered for further
implementation within  EMAP as a
developmental indicator.  The indicators
examined in the 1991 Demonstration were:


• Histopathology of target fish (primarily
  catfish, croaker, pinfish, seatrout, and
  spot),

• Frequency and Density of Splenic
  Macrophage Aggregates in Atlantic
  croaker and pinfish,

• Frequency and type of vertebral
  abnormality

• Blood chemistry of fish

• Bile florescence in fish

• Stable carbon and nitrogen isotopes.
Information for these research indicators
was not collected at the base EMAP sites
but rather through a special study that
augmented the 1991  Demonstration.
4.1  INDICATOR TESTING AND
   EVALUATION (ITE) SAMPLING
   DESIGN

Sufficient information is not available to
verify the reliability of these research
indicators for the estuaries of the
Louisianian Province.  Therefore, a study to
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 59

-------
Site
Eastern Subreglon
Apalaohioola Bay, FL
Watsons Bayou, FL
Choclawhatchoe River, FL

Escambia Bay, FL
PertiWo Bay, AL/FL
Wolf Bay, AL
Mobile Bay. AL

Bayou Casotte. MS

Western Subreglon
Cateasieu Lake Canal, LA
Galveston Bay, TX
Houston Ship Channel, TX

Brazos River, TX

Lavaca Bay. TX
San Antonio Bay, TX
Laguna Madia, TX
Arroyo Colorado, TX

1 Sediment contamination due
2 Sediment contamination due
Sediment contamination due
Designation

Reference
Contaminated1
Hypoxia
Contaminated2
Contaminated3
Hypoxia
Contaminated2
Hypoxia
Contaminated3
• Hypoxia
Contaminated1

Hypoxia
Contaminated1
Hypoxia
Contaminated1
Hypoxia
Contaminated3
Contaminated3
Reference
Contaminated2
Hypoxia
Contaminated2
to industrial sources
to agricultural runoff
to industrial and agricultural sources
Latitude

2940.00'
30 8.50'
30 24.00'

3031.70'
30 27.08'
30 19.71'
30 37.00'

30 20.00'


29 59.38
2931.66'
29 43.96' ,

28 57.79'

28 38.30'
28 18.30'
26 8.00'
26 20.33




Longitude

84 56.65'
85 38.00'
86 8.00'

87 10.00'
87 22.60'
87 35.72'
88 0.00'

8830.71'


93 20.03'
94 56.90'
95 8.04'

95 22.83'

9632.41'
96 39.90'
97 16.00'
97 25.76'




Table 4.1 Locations of Indicator testing and evaluation (ITE) sites used In the 1991 Loulsianlan Province Demonstration.
Designation refer* to level of hypoxla and sediment contamination.
determine the reliability of these indicators
to discriminate between polluted 'and
unpolluted environments was conducted.
This study also, in some instances,
provides for the first field testing of these
indicators and the first assessments of the
accuracy and precision of these indicators
to discriminate ecological conditions on
large geographical scales.  Samples for the
ITE study were collected at 16 locations (8
east and 8 west of the Mississippi River
Delta)(Table 4.1). These 16 sites were
selected, based on historical information, to
represent combinations of environmental
situations related to dissolved oxygen and
sediment contaminant (industrial and
agricultural) conditions.  By combining
extreme conditions, several ITE sites were
located throughout the estuaries of the Gulf
of Mexico.  For example, conditions of
hypoxia without significant industrial
discharges or agricultural runoff were
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 60

-------
 examined  in Perdido Bay, FL for regions
 east of the delta. Hypoxic, highly
 industrially contaminated conditions
 receiving little agricultural runoff in the
 western Gulf were sampled in the Houston
 Ship Channel, TX.  These research
 indicators were collected in conjunction
 with the standard EMAP-E indicators
 (Section 3) at these sites during the period
 July 7-August 31, 1991.
 4.2  HISTOPATHOLOGY OF
   TARGET SPECIES

 While gross fish pathology is a response
 indicator used in EMAP-E for
 environmental condition, it may not provide
 insight into the potential cause of the
 pathology.  In addition, the existence of
 external pathologies may or may not be
 related to the occurrence of  internal
 pathologies of the liver, spleen, or gills.  To
 address this concern,  EMAP-E performed
 detailed h.istological examinations of all fish
 observed to have external pathologies
 (base and ITE sites) and randomly-selected
 individuals of target and non-target species
 at ITE sites. All individuals that "failed" the
 gross pathology examination and up to 25
 randomly selected individuals of the target
   Brown shrimp
   Atlantic croaker
   White shrimp
   Hardhead catfish
   Blue Crab
   Spot
   Pinfish
   Southern flounder
   Gafftopsail catfish
(Penaeus aztecus)
(Micropogonias undulatus)
(Penaeus setiferus)
(Anus felis)
(Callinectes sapidus)
(Leiostomus xanthurus)
(Lagodon rhomboidas)
(Paralichtys lethostigma)
(Bagre marinus)
Table 4.2. Target species collected at ITE site* during the
1991 Louisianian Province Demonstration.
 species (Table 4.2) that "passed" this
 examination at the ITE sites were
 subjected to an extensive histopathological
 examination.  In addition, up to 10
 randomly selected individuals of non-target
 species collected  from the ITE sites were
 examined.  Histopathology will be
 maintained at the research indicator level
 until it can be shown to discriminate clearly
 between polluted  and unpolluted sites.

 Representative tissue samples were taken
 from specimens and  processed for
 histological analysis.  Tissue samples were
 dehydrated in an ethanol gradient, cleared
 in a xylene solution, infiltrated, and
 embedded in paraffin. Sections were cut at
 Gum on a rotary microtome, stained with
 Harris'  hematoxylin and eosin and
 examined microscopically. The results of
 this microscopic examination were used to
 characterize the types of external/internal
 pathologies and to create a baseline for
 these features for the Louisianian Province.
 Based on these findings, a determination of
 whether or not to continue with
 histopathology as  a developmental
 indicator will be made.

 Table 4.3 lists the histopathologic
 conditions observed in fish collected at the
 ITE sites. These conditions are listed in
 terms of parasitic infection and specific
 histopathologic condition. Sample numbers
 beginning with the prefix "RF" refer to the
 reference fish collected (i.e., "passed" the
field examination for gross pathologies)
and "FP" refers to  those fish collected with
obvious external pathologies.  With the
exception of Atlantic croaker collected from
the Houston Ship Channel, the accuracy of
the field gross pathology tests were
Demonstration Report, EMAP-E Louisianian Province -1991
                                                             Page 61

-------
TซWป 4.3 FMtutt* of hlซtopซthology of flซh from
Itll Loulilantan Province Dwnommilon.





Site
Petdido Bay, AL


Eocarnbia Bay. FL

Watgora Bayou.
FL




Bayou Casotte,
MS
Cnoctawhaiehea
Rwef. FL








WoVBiiy.AL




Apatachieota Bay,
FL






Species
Anchoa mttcntllt
Brevocrta patronua
Crtforoocombrua
chrysurua
Carangidaa sp.
Lepisostaus oculatus
Lagodon rhombcidea

Leioatomua
xanthurua

Trachinotua fakatua
Harengula
penaacclae
Bravoortia patronua

Laioatomua
xanthurua

Leprsoitaua oculalua

Anchoa mitchiBf

Carartx hippoa
Bagre marinua
Bravoortia patronua
Lagodon rhomboidea
Laioatomua
xanthuma
Caranx latua

Chkmacombrua
cluyturus
Doroaoma pelenense
Micropogoniaa
undulatus

n
10
3
2

2
4
42

4

1

9

15

21

2

17

1

21
10
21
20

1

2

7
4

O
I
||
g. %r

ia

0
0
1

0
0
2

0

N

N

1

1

0
^
0

0

2
0
2
0

N

N

N
N

Histopathological Conditions

B =
t~

}'






2





3

1









1












?
a
o
3




i

12

2







2









1
3

1

t





I
1
3
f


1









2











2


1






1


o
2
a

i






31

3



1



6







1

18
3









m
•o
5

|
65'





13

1



4



3



















Parasites
-i
<ฃ.

CO
1






12

1

1

1





2














1
2

^
^

1






4











2





1




1






2.
ป

1




1
4
15







2

7







4
4
3
1

1






CD
Qป

13 ซ
S =5
3 3
(0





1



















1









z
3
::
i
w'


1





















4

2
2








<
ง. i

-------
Table 4.3 (Cent.) Results of Mttopethology of fish
from 1991 Loulslanlan Province Demonstration.



Site
Mobile Bay, FL

Calcasteu Lake, LA





Houston Ship
Channel, TX
Qalveston Bay, TX
Laguna Madre, TX

San Antonio Bay, •
TX

Lavaca Bay, TX

Brazos River, TX


Species
Anchoa mitchitti
Brsvoortia patronus
Caranx hippos
Chaetodipterus faber
Dorosoma petenense
Trachinotus falcatus
Trichiurus lepturua
Selene vomer
Micropogonias undulatus

Micropogonias undulatus
Lagodon rhomboidea
Micropogonias undulatus
Bagra marinua

Micropogonias undulatus
Bagre marinus
Micropogonias undulatus .
Leiostomus xanthurus
Micropogonias undulatus

n
2
7
2
7
IS
4
3
2
18

7
25
10
1

11
19
14
10
B
O
8
Sg


0
0
0
1
0
0
0
0
15

0
0
1
0

0
0
0
1
0
Histopathotogical Conditions
• Parasites
II
II
I*
ง
f




1

1
1




1





1

-o
5



2

2
1
2
1
3

6
1

1

5
4
3
2

Helminth
U)
'^
i


2
4

2
I
6

5
7
2




3

3
p
3











1

3
1

2

1
3
1
m
I
w




1
4
1
3

1

1
1



1

2


1
1.





2
3




5
1
3


8

4
4
7
I
S
5'




1
1










1


1

3.
S





3
1
1
1


1

1




1

1
if
1 ?





















I-






















fl
I \
1't




















I
a
1







2













2=
•i
•sr







2













Demonstration Report, EMAP-E Louisianian Province - 1991
Page 63

-------
 acceptable (i.e., accuracy rate = 97%).
 Atlantic croaker collected from the Houston
 Ship Channel all showed significant levels
 of fin erosion.  However, because this
 condition was common to all croaker
 collected at this site, the field crews had no
 pattern of reference to compare with for
 these fish and thus erroneously
 categorized them as without gross defects.

 ANOVA testing of the ITE pathology data
 showed that while the number of
 pathologies per trawl at contaminated sites
 was greater than those at reference sites
 (0.83 and 0.33, respectively), these
 differences were not significant given the
 variation in the estimates. No significant
 longitudinal patterns were detected
 although the number of pathologies per
 trawl in the western subregion (1.3) was
 more than twice that observed in the
 eastern subregion (0.5).  Although the
 number of pathologies found under hypoxic
 conditions (1.3) was three times those
 found in reference areas (0.4), this
 difference was not significant. The
 incidence of pathologies in fish collected
 from areas of sediment contamination were
 twice that seen at uncontaminated sites
 (1.2 versus 0.6 at industrial sites); however,
 this difference was also not significant.
 The high degree of variability might make
 use of pathology a difficult and insensitive
 indicator for EMAP.  However, the fact that
 incidence of pathology appeared to vary
 along the expected gradients for dissolved
 oxygen  and sediment contamination
suggests that the indicator might be
sensitive but that the significance of this
sensitivity will not be observed without
increased sample sizes. This situation was
more evident when examining rate of
pathology (# of pathologies/100 fish), which
 corrects for the variability in catch size.
 Rate of pathology was three times more
 frequent in areas with sediments
 contaminated by industrial discharges
 (2.2%) than in uncontaminated areas
 (0.6%).  However, this difference was not
 statistically significant. Although pathology
 rate varied as expected along the exposure
 gradients, the only significant difference
 observed from the ITE collections was a
 significant east-west gradient with western
 sites having five times (2.4%) the
 pathology rate as eastern sites (0.5%).
 The levels of pathology rate seen at
 uncontaminated sites agrees well with the
 overall background rates of pathology seen
 in the Louisianian Province
 (0.6%)(Summers et  al. 1993). Because
 histopathologic condition varies in the
 appropriate directions (i.e., increases in
 poor environments and decreases at
 reference sites) and in a consistent
 manner,  histopathology will be retained in
 the 1992 EMAP sampling in the
 Louisianian Province as a developmental
 indicator and examined at a broader
 gepgraphic scale. If the variability in this
 measure continues to mask its ability to
 differentiate between impacted and
 unimpacted sites based on this broader
 test, the pathology indicator will be
 reviewed for removal from the EMAP
 indicators.
4.3  SPLENIC MACROPHAGE
   AGGREGATES

Pigment-bearing macrophages are a
prominent feature of fish spleen, kidney,
and liver (Agius 1980) and in advanced
teleosts they form discrete aggregations
called macrophage aggregates
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 64

-------
   Indicator
   Number of
    Pathologies
   Frequencies of
    Pathologies
                 Gradient          Mean
 Site Type
  Good           0.33
  Bad            0.83

 Longitudinal
  East    '        0.56
  West           1.26

 Dissolved Oxygen
  Hypoxiri          1.26
  High            0.43

 Contamination (I)
  High            1:21
  Low            0.60

 Contamination (A)
  High            0.52
  Low            1.17
Site Type
  Good            0.64
  Bad             0.89

Longitudinal*
  East            0.50
  West            2.41

Dissolved Oxygen
  Hypoxic          1.51
  High            1.11

Contamination (I)
  High            2.16
  Low             0.69

Contamination (A)
  High             0.75
  Low             1.85
Table 4.4 ANOVA results of number and frequency of
pathologies using the ITE data set. (I = Industrial
Sources; A = Agricultural Runoff). (* = p < 0.05).
 (MAs)(Wolke et al. 1985). Suggested
 functions for these aggregates include the
 centralization of foreign materials and
 cellular debris for destruction,
 detoxification, and/or reuse (Ferguson
 1976; Ellis et al.  1976).  It has been
 demonstrated that MAs' occurrence may
 vary depending on the size, nutritional
 state, or health of a particular fish (Agius
 1979, 1980; Agius and Roberts 1981;
 Wolke et al. 1985) with the number and
 size of MAs increasing with age, starvation,
 and/or disease.   Recent studies suggest
 that MAs may be sensitive histological
 indicators of fish  health and environmental
 quality.  By comparing the MA number and
 percent area occupied by MAs among fish
 of the same age  and species from various
 sites of known environmental condition,  it
 may be possible  to determine their relative
 Conditions at those sites.

 Data on  MAs were collected from 6 jim
 histological sections of spleen from
 selected fish species of similar size.
 Sections were stained with Harris'
 hematoxylin and eosin or Perl's Prussian
 blue method (Luna 1968).  Occurrence of
 MAs are assessed by two  methods.  First,
 during initial histological evaluation, the
 occurrence and intensity of MAs were rated
 using a scale of 0 to 4, with 0 being no
 MAs present, 1 indicating minimal
 occurrence, and 2 through 4 indicating
 light, moderate, and heavy MA intensity,
 respectively. Secondly, the MA number
and individual MA area were estimated
from three random fields per spleen using
computer image analysis (MicroComp •
Integrated Image  Analysis System Particle
Analysis). These data, identified  by
individual and  site, were compiled and
analyzed (i.e., blind to the conditions at any
Demonstration Report, EMAP-E Louisianian Province -1991
                                                               Page 65

-------
site). These data were analyzed for
differences in number of MAs per mm2,
average MA area (urn2), and percent area
occupied by MAs (Table 4.5).  Analyses
were completed on a combination of all
data (i.e., all species combined) and by
selected species.

Because of the small number of sites from
which ITE samples were collected, the
analyses were completed on a data set
that combined fish species. The
percentage of area occupied by
macrophage aggegrates was significantly
different between sites with good ecological
condition (high dissolved oxygen, low
contaminants) and poor ecological
conditions (Table 4.6) regardless of
species. Areas experiencing hypoxic
condition, showed significantly higher
proportions of the spleen having
macrophage aggregates than under
unstressed dissolved oxygen conditions.
However, contaminated sites, whether from
industrial discharges or agricultural runoff
did not show any pattern associated with
macrophage aggregate concentrations.
This lack of a pattern associated with
contaminated sediments may be due to the
highly significant dissolved oxygen gradient
(p < 0.001) which might overshadow any
gradient due to contaminants because the
data set includes contaminated sites that
are hypoxic as well as having high
dissolved oxygen conditions. In order to
ascertain whether the dissolved oxygen
gradient is confounding the identification of
a contaminant gradient, the interaction of
hypoxia and contaminant level was
examined. Although the percentage of
area occupied by macrophage aggegrates
was higher under conditions of
contaminated sediments under high levels
of dissolved oxygen (1.3%) than in high
DO-uncontaminated areas (1.0%), the
difference was not significant.  The
macrophage aggregate also showed a
significant longitudinal gradient with eastern
sites (4.0%) covering about four times the
area of the spleen than western  sites
(1.2%).

Another possible confounding factor might
be that all species of fish collected were
analyzed as a combined data set.
Sufficient data existed so that analysis
could be performed on the species level for
only two species:  pinfish (Lagodon
rhomboides) and Atlantic croaker
(Micropogonias undulatus).  In pinfish,
macrophage aggregates covered a
significantly higher proportion of  the spleen
in the fish from sites with poor
environmental condition (6.0%) than in fish
from reference areas (1.5%)(Table 4.7).
Macrophage aggregates were  more dense
in fish from eastern sites (4.7%)  than those
from western sites (1.5%). Macrophage
aggregate density in pinfish was
significantly higher in areas experiencing
hypoxia (6.0%) and in areas receiving high
industrial discharges (6.0%) as compared
to reference sites  (1.8%).

Macrophage aggregate densities in Atlantic
croaker (Micropongias undulatus) were
significantly higher at sites characterized by
poor environmental conditions  but were not
significantly higher at any of the  other site
combinations described above although
sites with bad environmental conditions,
sites in the eastern subregion,  hypoxic
sites, and sites with high industrial
discharges had higher macrophage
aggregate densities than their counterpart
sites (Table 4.7).
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 66

-------
Table 4.5 Results of splenic macrophage
aggregate analysis for 1991 Loulslanlan
Province Demonstration.

Site


Perdido Bay, AL
Escambia Bay, FL
Watsons Bayou, FL





Choctawhatchee
River, FL





Wolf Bay, AL


Apalachicola Bay,
FL
Mobile Bay, AL
Caloasieu Lake, LA

Species


Brevoortia patronus
Carangidae sp.
Lagodon rhomboides

Leiostomus xanthurus

Trachinotus falcatus

Brevoortia patronus

Leiostomus xanthurus

Lepisosteus oculatus

Caranx hippos
Bagre marinus
Lagodon rhomboides
Leiostomus xanthurus
Micropogonias
undulatus
Brevoortia patronus
Trachinotus falcatus


Splenic Macrophage Aggregates
Histologically
determined
intensity
n

2
1
19
23
3
1

1
2
9
6
12
1
1
1
21
16
20

1
7
1
0













1
1








1



1 :




1









12


1

2

1
1
9
8
2




4

1


1
14
10
8

1
5
1
3

1

6
12




1
4
5
4



7
6



1

4



3
3
1
1


1
1
1
7










Computer
image
analysis
MAs/
mm2
24.4
14.2
41.7
44.6
28.8
37.6

5.1
60.0
47.2
39.3
39.1
0.0
0.0
12.2
18.4
27.3
7.6

23.4
23.7
7.9
AvgMA
area in um2
429.6
945.6
1531.0
1367.5
594.4
1997.6

357.2
1838.0
1284.2
2367.2
2408.1
—
—
540.0
1851.9
751.7
301.0

662.14
545.1
239.2
% area
occupied
1.025
1.346
6.433
5.582
1.642
7.515

0.182
10.486
5.929
8.819
9.473
—
—
0.659
3.639
2.013
0.221
	
1.549
1.254
1.900
Demonstration Report, EMAP-E Louisianian Province -1991
Page 67

-------
Tabla 4.5{cont) Results of splenic
macrophage aggregate analysis for 1991
Loulslanlan Province Demonstration.
Site
Galveston Bay, TX
Laguna Madre, TX

San Antonio Bay, TX

Lavaca Bay, TX

Brazos River, TX

Species
Micropogonias
undulatus
Lagodon rfiomboides
Micropogonias
undulatus
Bagra marinus
Micropogonias
undulatus
Bagre marinus
Micropogonias
undulatus
Leiostomus xanthurus
Micropogonias
undulatus
Splenic Macnophage Aggregates

n
2
1
9
1
8
19
3
6
1

0









Histologioally
determined
intensity
1
1

3
1
3
14
1
2

2
1
1
6

5
5
2
1

3







2
1
4







1

Computer
image
analysis
MAs/
rnm2
10.2
20.9
13.8
3.1
9.5
5.3
27.9
20.5
32.5
Avg MA
area in urn2
723.1
739.5
536.1
554.0
641.2
286.7
621.3
975.5
733.1
% area
occupied
0.757
1.546
0.731
0.169
0.590
0.147
i?
1.931
1.923
2.385
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 68

-------
   Indicator

   Percent Area
   Occupied by
   Macrophage
   Aggregates
Gradient          Mean

Site Type*
  Good           0.99
  Bad            5.25

Longitudinal*
  East            3.99
  West           1.21

Dissolved Oxygen*
  Hypoxic          4.35
  High            1.22

Contamination (I)
  High            2.59
  Low            3.25

Contamination (A)
  High            3.28
  Low            2.49
Indicator

Percent Area
Occupied by
Macrophage
Aggregates

1
!



1



,

Gradient

Site Type*pc
Good
Bad

Longitudinal
East
West
Dissolved Oxygen*p
Hypoxic
High
Contamination (l)*p
High
Low
Contamination (A)*p
High
Low
Mean
Pinflsh

1 .55%
6.01%


4.68%
1 .55%

6.01%
1.78%

6.01%
1 .78%

1 .78%
6.01%
Croaker

0.66%
5:98%


1 .54%
1 .28%

2.38%
1.11%

1 .69%
0.96%

1.68%
0.97%
Table 4.6 Results of the analysis of percentage of area Table 4.7 Results of the analysis of percentage of area occupied In the
occupied In the spleen by macrophage aggregates, spleen by macrophage aggregates for plnfleh (P) and Atlantic croaker (C).
(^Industrial, A=Agrlcultural). (* = p < 0.05).          (^Industrial, A=Agrlcultural). (* = p < 0.05).
Because of the ability of macrophage
aggegrates to discriminate between sites of
known good and poor environmental
condition for at least one target species,
this indicator will be elevated to
developmental status in  1992 and
macrophage aggregate information will be
developed for selected target species
throughout the Province.
4.4  VERTEBRAL
   ABNORMALITIES

Measurement of skeletal deformities in fish
has been proposed as a means of
monitoring pollution effects in marine
environments (Bengtsson 1979; Bengtsson
and Bengtsson 1983).  Likewise,
                                measurements of biochemical composition
                                and mechanical properties of vertebrae
                                have been shown to be indicators of bone
                                development in fish exposed to
                                contaminants in the laboratory (Hamilton et
                                al.  1981; Mayer et al. 1977), and in the
                                field (Mayer et al. 1988; Mehrie et al.
                                1982).  Skeletal abnormalities in fourhom
                                sculpin (Myoxocephalus quadricomis) have
                                been used to monitor the impacts of ore
                                smelter and pulp mill effluents in the Baltic
                                Sea (Bengtsson et al. 1985).

                                Effects of organic and inorganic
                                contaminants on  bone integrity are similar
                                in that vertebral anomalies are produced,
                                although they may develop through
                                different modes of action (Mayer et al.
                                1978). This similarity makes the use of
Demonstration Report, EMAP-E Louisianian Province -1991
                                                              Page 69

-------
   Site



   Apalachicola Bay. FL

   Walsons Bayou. FL

   Choclawtiatchee River, FL


   Escambia Bay, FL

   Pondido Bay, AL/FL

   WoH Bay, AL

   Mobile Bay. AL


   Bayou Casotte. MS


   Western Subreglon

   Cateasiou Lake Canal, LA

   Gatvoston Bay, TX
Designation



Reference

Contaminated
Hypoxia
Contaminated2

Contaminated3

Hypoxia

Contaminated2

Hypoxia
Contaminated3

Hypoxia
Contaminated1
Hypoxia

Contaminated1
   Houston Ship Channel, TX   Hypoxia
                         Contaminated
   Brazos River, TX


   Lavaca Bay, TX

   San Antonio Bay, TX

   Laguna Madre. TX

   Arroyo Colorado, TX
Hypoxia
Contaminated3

Contaminated3

Reference

Contaminated2

Hypoxia
Contaminated2
Percentage of
  Skeletal
 Deformities

    7.1

    4.7

    4.4



    0.0

    0.0

    0.0

    0.0


    0.0
    0.0

    25.0

    36.8



    0.0



    44.1

    35.3

    10.7

    NA4
    Sediment contamination due to industrial sources
   2 Sediment contamination due to agricultural runoff
   3 Sediment contamination due to industrial and agricultural sources
   4 No fish caught during trawling due to extreme hypoxia
Table 4.8 Condition of the vertebral column In fishes examined from ITE
•Ite* In the 1991 Loulslanlan Province Demonstration.
biochemical composition and
mechanical properties, as well as
vertebral deformities, conducive to
assessing the abuse and effects of
an array of contaminants on fish
health (Mayer et al. 1988).

All preserved fishes collected from
ITE sites during the 1991 Louisianian
Province Demonstration were x-rayed
laterally with a Hewlett Packard™
Faxitron Series X-ray System set at
50kVp for 20 to 50 seconds,
depending on the size  of the
specimen.  Kodak   Industrex M-2
film was used for all  radiographs and
they were developed for 5 minutes in
Kodak™ D-19 developer.  Vertebral
anomalies were determined from the
x-rays by light box and confirmed by
low-power light microscopy.
Deformities were classified according
to Bengtsson and Bengtsson (1983).

The observed incidences of vertebral
deformities at the ITE sites are
shown in Table 4.8.  In general,
vertebral deformities  were found only
at sites with contaminated sediments.
The exception to this observation is
the reference site in San Antonio
Bay, TX. All the vertebral deformities
found at this site were Atlantic
croaker with  anterior curvature and
only 12 specimens were examined.
While San Antonio Bay sediments
were confirmed by contaminant
analysis as having low
concentrations in the upper 2 cm of
sediments, deeper sediments show
that San Antonio Bay has had some
sediment contamination in the past.
A significant longitudinal  gradient was
Demonstration Report, EMAP-E Louisianian Province -1991
                                                          Page 70

-------
observed with western sites having a
frequency of vertebral deformities that was
eleven times greater (17.9%) than those
observed in the east (Table 4.9). With a
longitudinal gradient this strong, it would be
expected to be difficult to ascertain any
relationships from a mixed dataset.  Even
with the strong longitudinal gradients,
skeletal deformities were seven times more
prevalent in hypoxic areas and four times
more prevalent in areas  receiving high
industrial  discharges although the gradients
were not significant due  to the high
variability induced by a dataset including
the longitudinal gradient  (Table 4.9).
Regardless of the longitudinal gradient,
vertebral deformities in pinfish (Lagodon
rhomboides) and Atlantic croaker
(Micropogonias undulatus) were
significantly higher under hypoxic and high
   Indicator

   Percentage of Rsh
   with Vertebral
   Deformities
Gradient           Mean

Site Type
 Good            8.46
 Bad             1.19

Longitudinal*
 East             1.64
 West       ,     17.93

Dissolved Oxygen*
 Hypoxic          16.44
 High             2.32

Contamination (I)
 High             16.31
 Low             4.76

Contamination (A)
 High             7.77
 Low             10.70
Table 4.9 Results of the analysis of percentage of vertebral
deformities from ITE  sites during tho 1991 Loulslanlan
Province Demonstration. (^Industrial, A=Agricultural). (* = p
< 0.05).
industrial discharge conditions.

4.5 BLOOD CHEMISTRY

In both human and veterinary medicine,
clinical chemistry measurements and
routine hematology are used to assess the
health of individuals.  Altered values in
serum enzymes or other proteins,
electrolytes, or blood cells of fish can  be
indicative of tissue damage, tumors, or
impaired immunological functions.  Blood
was collected from all fish greater than 200
mm in total length at ITE sites. Blood,
removed by vacutainer, was stored on ice,
shipped as soon as possible to the
laboratory, and analyzed using a
Beckman™ Synch ron CX-5.

Unfortunately, only 7 of the 16 ITE sites  "-
visited provided fish of sufficient size to
extract blood. Further analysis of these
samples was not attempted and, based  on
logistical problems in providing fish of
adequate size, no samples will be  collected
in 1992 for blood chemistry.   However, the
funding provided for these analyses were
re-channelled into an evaluation of blood
chemistry in fish outside the Louisianian
Province for future applications in EMAP-
Estuaries.

Brown bullheads (Ictalurus nebulosus) were
collected from 4 freshwater sites: Old
Woman Creek, OH (reference); Niagara
River,  NY (moderately contaminated);
Buffalo River, NY (moderately
contaminated); and, Black River, OH
(heavily contaminated). Table 4.10 shows
the percentage of bullheads  having
excessive concentrations of the serum
constituents.  Excessive concentrations
were determined as any concentrations
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                         Page 71

-------
Serum
Constituent
Aspartata amlnotransferase
Alkaline phosphataso
Lactate dehydrogenose
Alan'mo aminotransferase
Total protein
Crsatinine
Blood urea nitrogen
Trigtycarides
C-Reactiva Proteins
1OWC = Old Woman Creek, OH
NR = Niagra River. NY
BR = Buffalo River, NY
BLR = Black River. OH
Site1
OWC NR BR BLR
0.0 4.0 14.0 66.0
0.0 0.0 2.9 11.0
0.0 37.0 9.0 25.0
0.0 59.0 26.0 78.0
0.0 0.0 10.0 67.0
0.0 0.0 0.0 20.0
0.0 97.0 66.0 50.0
0.0 97.0 38.0 62.0
0.0 100.0 — 100.0

Tablo 4.10 Percentage of brown bullheads exhibiting excessive serum
chemistry concentrations.
greater than the upper 95% confidence
limit observed at the reference site.  Clear
gradients from reference through heavily
contaminated sites were observed for AST,
ALP, TP, CREA, and ALT.  In addition, C-
reactive protein levels,  a protein associated
with liver tumors, were significantly higher
in bullheads taken from the Black (1300 ฑ
329) and the Niagara (1027 ฑ 384) River
than concentrations observed in bullheads
from the reference site (442 ฑ 20).

These data suggest that blood chemistry
could provide a strong  indicator along a
gradient  of sediment contamination. This
discriminatory power needs to be
demonstrated with estuarine species from
                                                    the Louisianian Province. Bullheads
                                                    are observed in many oligohaline
                                                    estuaries in the Virginian Province.
                                                    Therefore, if the logistical problems
                                                    of collecting the appropriate fish can
                                                    be solved in the Louisianian
                                                    Province, the potential of serum
                                                    chemistry to provide useful indicators
                                                    should be re-examined.
      4.6  BILE FLORESCENCE

      Organisms exposed to petroleum
      compounds often accumulate
      polynuclear aromatic hydrocarbons
      (PAHs). Tissue analysis for PAHs
      often show only trace concentrations,
      even after high-level exposures
      because enzymatic-mediated
      metabolism can rapidly reduce
      concentrations.  The exposure of fish
      to PAHs can be assessed by
      measuring the concentration of
      metabolites in bile.  The relative
      concentration of individual PAH
metabolites of benzo(a)pyrene,
phenanthrene, and naphthalene in bile
were determined using PHLC with
florescence detection. Bile was extracted
from all fish greater than 200 mm in total
length collected at ITE sites.

As with blood chemistry, only 7 of the 16
ITE sites visited provided fish of sufficient
size to extract blood.  These samples were
analyzed, but small sample sizes made it
difficult to find significant differences.
Based on logistical problems in providing
fish of adequate size, no samples will be
collected in 1992 for bile florescence.
However, the results of the analysis on fish
from the  7 ITE sites are reviewed below.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 72

-------
 The concentration of benzo(a)pyrene in bile
 from fish at sites of poor environmental
 condition (570 ppb) was nearly twice that
 found at reference sites (325 ppb) but this
 difference was not significant due to the
 large variability observed in individual fish
 (Table 4.11). Fish in the western
 subregion showed nearly ten times as
 much benzo(a)pyrene in bile (1113 ppb) as
 those eastern sites (196 ppb) but, again,
 this difference was not significant due to
 high variability. The pattern of contaminant
 concentrations in bile appeared to follow
 expected relationships with higher
 concentrations found in hypoxic and
 heavily industrialized areas.
                         Concentrations of naphthalene and
                         phenanthrene in bile of fish collected at ITE
                         sites showed similar patterns to that of
                         benzo(a)pyrene with the exception that
                         observed concentrations at good and bad
                         sites were about equal (Table 4.11).

                         4.7 STABLE ISOTOPES RATIOS

                         Stable isotopes, often  in combination with
                         elemental analyses, have been  used
                         traditionally to distinguish terrestrial and
                         marine sources of organic matter in
                         estuarine systems (Coffin et al.  1992).
   Indicator
   Concentrations of
   PAHs in Bile
                         Gradient
Site Type
  Good
  Bad

Longitudinal
  East
  West

Dissolved Oxygen
  Hypoxic
  High

Contamination (I)
  High
  Low

Contamination (A)
  High
  Low
   1 Benzo = Benzo(a)pyrene;
    Phen = Phenanthrene;
    Naph = Naphthalene
   2 Insufficient data to perform test
                           Mean Concentration1

                          Benzo   Phen    Naph
                                                   570
                                                   325
                                                   196
                                                   1113
16000   57500
13700   51250
NA2
NA
                                                   1080   55800  185833
                                                   500    22811  71111
                                                   1080   50367  159444
                                                   500    14467  53333
                                                   736    27342    92500
                                                   1216   70667   215000
Table 4.11 Results of the analysis of concentration of selected PAHs (ng/g) In the bile of fish collected
at ITE sites. ((^Industrial, A=Agrlcultural). (* = p < 0.05).
Demonstration Report,  EMAP-E Louisianian Province - 1991
                                                        Page 73

-------
In particular, isotopic analyses of carbon
(513C) differentiate C3 and C4 terrestrial
plants from algal sources of organic matter
(Fry and Sherr 1984, Coffin et al. 1992).
Stable nitrogen isotopes (815N) combined
with 813C can trace terrestrial, marine, and
in some cases, anthropogenic contributions
in estuarine systems (Fogel and Cifuentes
1992).

During the 1991  Louisianian Province
Demonstration, samples for stable isotope
analysis were collected from the 16 ITE
sites. The objective was to identify
estuaries that combined significant
anthropogenic inputs with net heterotrophic
activity, suggesting the existence of, or
potential for, oxygen depletion and
eutrophication. To accomplish this
objective, 813C and 815N was measured in
suspended particulate matter, humic acids,
and bacterial bioassays.

Particulate and dissolved suspended
matter were obtained from a depth of 1 m
at all  16 locations. Suspended particulate
matter (SPM) samples were collected by
pushing water through 47 mm GF/F filters
(heated to450 C for 2 hr) using a
Masterflex peristaltic pump.  Humic acids
(HA) were collected by a modification of
the method of Fox (1983). First,  1 L of
water was pre-filtered through a 47 mm
GF/F filter as described for SPM samples.
The filtrate was then acidified to pH 2 with
8N  H2SO4 to precipitate humic acids.
Other macromolecules, such as proteins
and mucopolysaccharides, will also
precipitate at low pH (Thurman 1985). This
precipitate was captured on a 47 mm GF/F
filter.

Approximately 30 ml of unfiltered water
was collected in QuorpakR bottles and
preserved with 2% HgCI2 for isotopic
analysis of dissolved inorganic carbon.
These samples were refrigerated at 4 C
prior to isolation of CO2 for isotopic
analysis by the in vacua acidification and
purging technique described in Grossman
(1984).  All samples, including the zeolite
with the exchanged ammonium, were
analyzed isotopically by a modified Dumas
combustion that converts organic carbon
and organic nitrogen to CO2 and N2,
respectively, for mass spectral analysis
(Macko 1981).

Uniformly high nutrient concentrations were
only  observed in the Houston Ship
Channel, with values similar to or higher
than  those reported for more "polluted"
estuaries (e.g., Delaware Estuary, Sharp et
al. 1982)(Table 4.12).  Sites selected for
high  agricultural runoff show high nutrient
concentrations. Brazos River,
Choctawhatchee River, and Escambia Bay
had significant nitrate+nitrite (NO3) and
ammonium (NH4) concentrations, but had
low PO4). Wolf Bay had high NH4 content,
but no NO3.  In contrast the reference sites
had low NH4.  Those sites with high
agricultural runoff but low nutrients showed
high  chlorophyll a levels. Arroyo Colorado
had the greatest algal biomass of 37.5  ug/l
chlorophyll a.  Wolf Bay, Mobile Bay,
Lavaca Bay, and Galveston Bay showed
high  chlorophyll a concentrations.

The range of carbon isotopes for SPM
measured at the ITE sites, -30.7 to -17.3
ppt (Table 4.13), was similar to that
reported for other estuaries that have been
extensively studied (Cifuentes et al. 1988;
Fogel et al. 1992).  Without further
analysis, it would appear that we
Demonstration Report, EMAP-E Louisianian Province - 1991
                              Page 74

-------
Station

Southern Laguna Madre
San Antonio Bay
Bayou Casotte
Watson's Bayou
Lavaca Bay
Wolf Bay
Arroyo Colorado
Brazos River
Perdido Bay
Apalachicola Bay
Galveston Bay
Lake Calcasieu
Choctawhatchee Bay
Houston Ship Channel
Mobile Bay
Escambia Bay
B.D. = Below Detection
Abbr.

SLM
SAB
BC
WBY
LB
WB
AC
BR
PB
AB
GB
LC
CB
HSC
MB
EB

Date

8/7/91
8/4/91
8/9/91
7/16/91
7/13/91
8/14/91
8/5/91
7/12/91
8/14/91
7/15/91
7/10/91
7/15/91
7/17/91
7/1 1/91
8/24/91
7/18/91

Salinity
(ppt)
36.0
12.5
26.4
23.6
12.0
15.9
15.5
9.5
10.6
10.0
12.0
2.5
0.0
7.5
7.4
18.0

Chlora
(H9/I)
2.86
14.69
11.27
2.52
6.53
16.32
37.54
.6.73
4.34
LOST
14.48
14.08
3.21
10.61
7.63
3.79

C:N
,
7.1
7.4
5.9
5.2
4.7,
6.7
5.3
6.0
5.3
4.8
4.6
5.4
4.8
6.8
4.0
5.8

PO4
(HM)
0.10
3.04
0.58
0.10
0.68
0.10
4.47
1.78
0.05
0.02
6.34
., 0.63
0.31
11.17
0.04
0.10

NO3
(HM)
0.24
0.24
0.24
0.24
0.24
0.24
0.24
38.64
0.24
2.22
0.24
4.59
10.13
137.60
0.63
8.55

NH4
OiM)
0.09
0.19
0.19
0.19
0.19
4.73
0.72
6.24
0.28
0.13
0.19
0.19
1.14
10.03
0.19
2.46

SIOH4
(HM)
20.50
160.60
32.80
33.60
132.00
98.30
191.80
93.80
90.20
86.90
119i70
89.30
83.60
140.20
62.30
135.20

DON
(HM)
47.6
158.8
11.0
17.5
113.2
22.4
87.8
67.3
18.2
16.8
72.4
62.8
14.6
131.5
15.0
11.5

DOC
(HM)
177.5
1232.5
185.0
352.5
365.8
369.2
505.0
254.2
351.7
1027.5
394.2
1070.0
403.3
730.0
225.8
640.8

Station

Southern Laguna Madre
San Antonio Bay
Bayou Casotte
Watson's Bayou
Lavaca Bay
Wolf Bay
Arroyo Colorado
Brazos River
Perdido River
Apalachicola Bay
G.alveston Bay
Lake Calcasieu
Choctawhatchee Bay
Houston Ship Channel
Mobile Bay
Escambia Bay
B.D. = Below Detection
Date

8/7/91
8/4/91
8/9/91
7/16/91
7/13/91
8/14/91
8/5/91
7/12/91
8/14/91
7/15/91
7/10/91
7/15/91
7/17/91
7/11/91
8/24/91
7/18/91

S13C
DEC
0.1
-2.6
-2.6
-2.1
-4.2
-2.4
-5.9
-8.7
-3.0
-4.6
-6.3
-9.5
-11.1
-10.4
-11.1
-10.5

813C
SPM
-17.3
-21.0
-21.5
-22.1
-22.8
-23.9
-24.5
-25.1
-25.8
-26.9
-27.3
-27.3
-27.4
-27.6
-28.1
-30.7

513C
HA
-22.4
-22.4
-22.9
-23.6
-22.3
-24.5
-24.4
-24.5
LOST
-24.9
-24.3
-25.4
-26.1
-26.4
-25.9
-26.1

813C
BA
-23.4
-22.9
-22.6
-24.2
-24.0
-23.8
.-24.2
-24.5
-25.2
-24.2
-23.7
-24.7
-24.1
-25.5
-24.2
LOST

A13C

-17.4
-18.4
-18.9
-20.0
-18.6
-21.5
-18.6
-16.4
-22.8
-22.3
-21.0
-17.8
-16.3
-17.2
-17.0
-20.2

515N
NH4
B.D.
B.D.
B.D.
B.D.
B.D.
7.2
B.D.
10.0
B.D.
B.D.
B.D.
B.D.
B.D.
32.6
B.D.
2.2

815N
NO3
B.D.
B.D.
B.D.
B.D.
B.D.
B.D.
B.D.
36.6
B.D.
2.1
B.D.
-2.4
3.3
7.8
B.D.
1.9

S15N
SPM
8.6
7.4
8.7
5.8
8.2
10.6
13.5
6.5
10.0
8.5
12.4
17.3
7.1
15.0
8.9
11.1

815N
HA
14.8
6.5
11.3
12.6
12.4
14.2
9.5
7.5
LOST
13.1
9.2
4.5
10.5
12.7
15.6
15.6

815N
BA
10.7
8.3
8.4
14.5
8.0
15.0
12.6
14.9
16.0
11.1
15.3
9.5
14.4
28.5
16.4
LOST

Table 4.13  Isotopic data collected from ITE sites during the 1991 Loulslanlan Province Demonstration.  AlaC Is the Isotopic
discrimination between suspended particulate matter and dissolved Inorganic carbon.
       Demonstration Report, EMAP-E Louisianian Province -1991
Page 75

-------
sampled waters spanning the range of
terrestrial and algal sources of organic
matter (Fig. 4.1), consistent with the fact
that these samples originated from both
fresh  and coastally-dominated waters.  The
most  positive value (-17.3 ppt) was from
Southern Laguna Madre where
seagrasses, which are relatively enriched
in 1 C, are known to contribute significant
quantities of organic matter (Fry et al.
1987). At the opposite end, Escambia Bay
SPM  had anomalously light 813C, -30.7
ppt, which is outside the range generally
reported for terrestrially-derived organic
matter in estuaries. The source of this
enriched material is discussed below.

When the stations are ordered from most
positive to negative 613C of SPM as in
Figure 4.1, the corresponding and expected
transition from terrestrial or sewage-derived
nitrogen  (815N of -2 to +4) to coastal
nitrogen  (515N of +8 to +12) was not
observed for SPM  (Fig. 4.2). Lavaca Bay,
Arroyo Colorado, Galveston Bay, and the
Houston Ship Channel all had significantly
   N-enriched values, resulting from either
degradation (Altabet and McCarthy 1986)
or from uptake of isotopically enriched
inorganic nitrogen (Mariotti et al. 1984;
Cifuentes et al. 1988):  Considering the
513C  ratios, the C:N ratios of SPM (Fig.
4.3) were generally in  the range  reported
for algae (7-10, Holligan et al. 1984) rather
than bacteria (3-5,  Lee and Furhman 1987)
or vascular plant material (>50, Hedges
and Mann 1979).  Some values were highly
enriched in nitrogen (C:N <5). These
values could result from algae growing in
nitrogen-enriched conditions, or from
extensive bacterial colonization of particles.
The combination of low C:N and negative
813C  in SPM, therefore, is more  likely the
result of algae growing on isotopically light
CO2 (Fogel et al. 1992).  The use of stable
isotopes as a developmental indicator will
be initiated in 1993.
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 76

-------
   Figure 4.1  5  C measured in suspended paniculate matter (SPM) from selected sites in Gulf of Mexico estuaries.
        St. Llfin \ttitl

        Sit Alttllt I

        Bt'ftl CllltU
        Luati ttj
                                           1^ SPU (ppt)
  Figure 4.2  519N measured In suspended paniculate matter (SPM) from selected sites In Gulf of Mexico estuarine
  systems.                                              ,
Demonstration Report, EMAP-E Louisianian Province -1991
Page 77

-------
      ft. LIIIII litfn
      III tiUilt Mr
      IซJFM
   Figure 4.3 Carbon to nitrogen ratios for suspended partlculate matter from selected sites In Gulf of Mexico estuarine
   systems.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 78

-------
                                SECTIONS

 STATISTICAL ASSOCIATIONS BETWEEN  RESPONSE
                  AND EXPOSURE  INDICATORS
One of the objectives of the EMAP-E
Louisianian Province is to ascertain
statistical associations for response
indicators (e.g., benthic index, water clarity,
fish tissue contaminant levels).  These
associations could relate,  statistically,
response variables with exposure
indicators (e.g., sediment  contaminants,
sediment toxicity, hypoxia) or stressor
indicators (e.g., human population,
discharge volumes in watersheds, pesticide
application rates).  One of the reasons for
collecting exposure indicators during EMAP
sampling is to allow the determination of
statistical associations. The existence of
these associations does not imply
causation but  rather denotes a statistical
correlation. The results of these
associations can be used  by researchers to
hone research hypotheses that can be
tested to determine causation. However,
this research is not a part of the monitoring
element of EMAP at present.

Several potential associations were
investigated relating to response indicators
used in the 1991 Louisianian Province
Demonstration. These included
examinations of the statistical relationships
between the benthic index and various
exposure indicators, between fish
contaminant concentrations and sediment
contaminant concentrations, and between
sediment toxicity and sediment
contaminants.  Many other associations
could be investigated but the purpose of
this section is to demonstrate how the
EMAP monitoring data can be used to
evaluate associations.  No statistical
relationships were investigated between
indicators and stressors, whether natural
(e.g., climate) or anthropogenic (e.g., total
watershed loadings) because the stressor
data will not be collated from existing data
bases until future years in EMAP.
5.1  ASSOCIATIONS WITH THE
BENTHIC INDEX

The methodology used to create the
benthic index was described in Section 2.
The benthic index accounts for 90% of the
variability observed in the 1991 subset of
sites exhibiting environmental extremes
and thus, the benthic index is assumed to
be representative of an integrative indicator
of ecological condition. While further
testing through Years 2-4 in the Louisianian
Province will be necessary to confirm this
assumption, it is worthwhile to evaluate the
potential associations between the benthic
index and exposure indicators. The results
of this analysis can be used to begin to
assess the causes of degraded benthic
communities in Gulf of Mexico estuaries
through long-term structured research.
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 79

-------
The associations between the benthic
index and exposure indicators were
determined using three statistical tools: (1)
univariate correlations between the benthic
index value at a site and corresponding
exposure indicators at that site, (2)
multivariate regressions between index
values and exposure indicator values, and
(3) multivariate regressions between
categorized index values (i.e., < 4.1
represents poor conditions and > 6.5
represents good conditions) and exposure
values. All of these assessments were
completed for the Louisianian Province as
a whole and for each of the three estuarine
classes (i.e., large estuaries, large tidal
rivers, and small estuaries).  Other spatial
divisions (Gulf States and individual
estuaries) could be evaluated after several
additional years of monitoring data are
collected (i.e., a single year's information
provides too small a sample size).

Univariate analysis  comparing the benthic
index to each exposure indicator showed
that few exposure indicators accounted for
significant portions of the variability in the
observed benthic index (Table 5.1).  While
44 individual sediment contaminants were
significantly related  to the benthic index
value, only 9 contaminants accounted for >
10% of the variability in the index (2
alkanes, 2 pesticides, 4 PAHs, and 1  PCB).
Of these sediment contaminants', only three
accounted for > 20% of the variation: two
alkanes (C10 and C13) and one pesticide
(4,4' ODD).  Univariate analysis of non-
sediment contaminant indicators revealed
only 3 indicators accounting for > 5% of
Sediment Contaminant
Indicator
Alkanes
C10
C12
C13
C14
C22
C23
C24 .
C25
C26
C34
Heavy Metals
Cadmium
Chromium
Copper
I nan
Lead
Manganese
Mercury
Nickel
Selenium
Tin
Zinc
Anthropogenic Metals
Mercury
Nickel
PCB Congeners
Congener #52
Congener #66
Congeners #11 0/#77
Pesticides
alpha-Chlordane
2,4'-DDD
4,4'-DDD
4,4'-DDE
4,4'-DDT
Dieldrin
trans-Nonachlor
R2 Significance


0.216
0.059
0.235
0.069
0.099
0.047
0.049
0.043
0.073
0.042

0.064
0.066
0.064
0.081
0.089
0.072
0.076
0.092
0.043
0.094
0.095

0.051
0.041

0.145
0.043
0.044- •

0.072
0.097
0.247
0.099
0.143
0.093
0.066
Table 5.1 Significant univariate relationships between the
benthic index and sediment contaminant Indicators.
(* = p < .05, " ซ p < .01, and "* ซ p < .001).
Demonstration Report, EMAP-E Louisianian Province -1991
                               Page 80

-------
Sediment Contaminant
Indicator
PAHs
Benzo(a)pyrene
Benzo(k)fluoranthene
Benzo(g,h,i)perylene
Biphenyl
C1 -Naphthalene
C2-Naphthalene
2,6 Dimethylnaphthalene
ldeno( 1 ,2,3,c,d)pyrene
1-methylnaphthalene
. 2-methylnaphthalene
Naphthalene
R2 Significance


0.043
0.046
0.052
0.088
0.147
0.091
0.127
0.050
0.158
0.134
0.110
Table 5.1 (Cont) Significant univariate relationships between
the benthic Index and sediment contaminant Indicators.
(* a p < .05, " = p < .01, and *** = p < .001).
Exposure
Indicator
Instantaneous Bottom
Dissolved Oxygen
Ampelisca Bioassay
Mysid Bioassay
Light Transmittance (PAR)
at 1 meter
RPD Depth
Light Transmittance (PAR)
at bottom
Bottom Salinity
Bottom pH
Percent Silt-Clay
Total Organic Carbon
Acid Volatile Sulfides
Bottom Temperature
R2 Significance
0.001
0.012
0.012
0.032
0.047
0.007
0.073
0.042
0.010
0.030
0.001
0.034
Table 5.2 Univariate relationships between the benthic Index
and exposure Indicators other than sediment contaminants.
(* * p < .05, ** = p < .01, and "* = p < .001).
 observed variability and no indicators
 accounted for > 10% of variation (Table
 5.2).

 Univariate statistics can be used to
 examine if there are any significant
 differences in how benthic communities in
 the three estuarine classes relate to
 exposure variables.  Benthic index values
 in all three classes, based on univariate
 correlations, showed associations with
 selected alkanes and heavy metals (Table
 5.3) while large estuaries and large tidal
 rivers had some relationships with
 sediment and water quality attributes.
 Benthic communities in small estuaries
 were the only benthic assemblages to
 show any association with sediment
 pesticide concentrations.  These analyses
 suggest that while some  differences exist
 among the three estuarine classes with
 regard to poor benthic assemblages as
 determined by the benthic index, benthic
 communities in all three estuarine classes
 are associated with similar environmental
 exposure indicators.

 We developed a stepwise multivariate
 regression model for the  province-wide
 benthic index. While a successful model
 was created,  it only accounted for 49% of
 the observed variation in  the distribution of
 benthic index values (Fig. 5.1).  The six
 variables that characterize this model  are
 all sediment contaminants suggesting that
 the only significant associations between
 the benthic index and exposure variables
 are sediment contaminants (2 heavy
 metals, 1 pesticide, 1  alkane, 1  PAH,  and 1
 PCB) with the degraded form of DDT  (4,4'-
 DDD) accounting for 25% of the variability.
As was shown in the univariate analyses,
some differences between the three
Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 81

-------
Exposure Variable

AHcanes
Total
CIO
C16
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
028
C30
C31
C32
C33
C34
Hปavy Metals
Arsonic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Tm
Zinc
PCS Congeners
Congener #28
Congener #126
Pesticides
4.4'-DDE
9

Largo
(n=48)
0.103*
0.054
0.000
0.133*
0,211"*
0.154"
0.102*
0.195"
0.191"
0.144"
0.196"
0.253"*
0.095*
0.010
0.232"*
0.219"*
0.284*"
0.209"
0.249"*

0.066
0.127*
0.076
0.172"
0.167"
0.195**
0.135*
0.039
0.191**
0.196"
0.108*
0.122*

0.148"
0.086*

0.022

R2
Tidal
(n=10)
0.132
0.444*
0.529*
0.144
0.105
0.000
0.000
0.029
0.005
0.010
0.004
0.000
0.339
0.485*
0.000
0.001
0.001
0.012
0.046

0.044
0.039
0.023
0.030
0.035
0.024
0.021
0.422*
0.037
0.027
0.078
0.021

0.029
0.017

0.014


Small
(n=42)
0.001
0.123*
0.060
0.038
0.025
0.000
0.003
0.083
0.096
0.093
0.109*
0.152*
0.023
0.009
0.052
0.020
0.028
0.016
0.049

0.288***
0.037
0.187"
0.121*
0.216"
0.169*
0.028
0.021
0.227**
0.057
0.210"
0.198"

0.023
0.024

0.158*

Exposure Variable

PAHs
Ruorene
Naphthalene
Phenanthrene
Water Quality
Bottom pH
Sediment Quality
Total Organic Carbon
Mean RPD Depth

Large

0.119
0.010
0.109*

0.041

0.276"*
0.011
R2
Tidal

0.007
0.017
0.030

0.534*

0.031
0.585***

Small

0.054
0.143*
0.043

0.040

0.060
0.010
                                                        Table 5.3 (Cont)  Unlvariate relationships between benthlc
                                                        Index values and  exposure  Indicators for large  estuaries
                                                        (Large), large tidal rivers (Tidal), and small estuaries (Small)
                                                        (* = p < 0.05, ** = p < 0.01, *•• = p < 0.001).
Table 5.3  Unlvariate relationships between benthlc Index
values and exposure Indicators for large estuaries (Large),
large tidal rivers (Tidal), and small estuaries (Small)
(• = p < 0.05, ** s p <0.01, *" * p <0.001).
                                                            Pestlcldtt
                                                                 24.7*
                                                                                                   Unknown
                                                                                                   51,0-
Figure 5.1 Associations between Province-wide benthic index
and exposure Indicators.  Percentage associated with portion
of pie portrays R2 of Indicator.
    Demonstration Report, EMAP-E Louisianian Province - 1991
                                        Page 82

-------
              Pmtlcldti
                         PAHซ
                         *.9i
                                 Hซgvy Httola
                                 5.9ซ
Figure 5.2   Associations between  benthlc Index and
exposure Indicators  for large  estuaries.  Percentages
associated with portion of pie portrays R2of Indicator.
                                  41.471
Figure 5.3   Associations between  benthic index  and
exposure Indicators for large tidal rivers.  Percentages
associated with portion of pie portrays R2of indicator.
estuarine classes may exist; therefore,
stepwise multivariate regressions were
developed for each estuarine class.

The multivariate regression for the large
estuary class increased the overall
accounted variability to 64% (Fig. 5.2) with
sediment contaminants again comprising
all of the explained variability (2 alkanes, 2
pesticides,  1  PAH, and 1 heavy metal).
The concentration of alkanes in  sediments
contributed to 40% of the observed
variability in the benthic index seen in large
estuaries.  Ninety-eight percent of the
variability in benthic index values in large
tidal rivers  in the Louisianian Province (i.e.,
Mississippi River) was associated with
sediment contaminants (1 alkanes,  1  PCB,
and 1 heavy metal). Alkanes and heavy
metals accounted for 54% and 41 % of the
variability in large tidal river benthic index
values (Fig 5.3).  Similarly, in small
estuaries, the benthic index is associated
with sediment contaminants that account
for 66% of  observed variability (Fig 5.4).
Unlike large estuaries and tidal rivers,
heavy metals in sediments accounted for
40% of the variability in the benthic index
with the remaining 26% of the explained
variability being associated with a PAH, a
PCB,  and a pesticide.

Because the benthic index represents an
evolving response indicator, we tested the
hypothesis that the index, based on a
sihgle year, only accurately discriminates
between poor benthic community structure
and good benthic community structure.
Under this  hypothesis, variability within
these two classes has not been developed
to the point to accurately represent the
ordering among the estuaries, although the
underlying  structure exists and will take
Demonstration Report, EMAP-E Louisianian Province - 1991
                               Page 83

-------
                                                               Unknown
                Heavy Metals
                      38.4*
                                                           Pssticidos
                                                           5.0s
                                   PAHs
                                   11.3*
PCBi
10.4i
 Figure 5.4 Associations between benthlc Index and exposure Indicators for small estuaries. Percentages
 associated with portion of pie portrays R2of indicator.
                                            TWofty
                                             2-1*   Unknown
                                                   J00%
        SKftnmt Oontanhsnts
                    7O3%
Rguro 5.5 Associations between Province-wide benthlc Index that has been categorized and exposure
Indicators. Percentages associated with portion of pie portrays R2 of Indicator.
Demonstration Report, EMAP-E Louisianian Province -1991
                                      Page 84

-------
 several years to ascertain.  In this analysis,
 we assigned two categories to the benthic
 index: poor < 4.1 and good > 6.1.  Using
 the categorical regression, 90% of the
 variability in these benthic index classes
 were associated with measured exposure
 indicators (Fig. 5.5).  Sediment contaminant
 concentrations and sediment toxicity
 accounted for 72% of the variability
 between good and poor benthic index
 values with heavy metals contributing the
 greatest portion of this association (42%).
 Alkanes contributed 17% with  other
 sediment contaminants (i.e., PCBs, PAHs,
 and pesticides contributing 11 %.  The
 degree of stratification, minimum dissolved
 oxygen concentration, and the percent of
 time dissolved oxygen  concentrations are
 less than 2 ppm  was associated with 5% of
 the differences between the index
 categories.  Other water quality habitat
                        UlkMIl
                         4-21    Ptitleiiti
                               a.4ซ
  Htifjr lit til i
       41.If
                                 Sซdlซปnt HoHtit
                                 U.S*
 Figure 5.6  Associations between benthic index that has
 been categorized  and exposure  Indicators  for large
 estuaries. Percentages associated with portion of  pie
 portrays r2 of indicator.
                                  SlJli.it Htlltlt
                                  17.li
Figure 5.7 Associations between benthic Index that has
been categorized  and  exposure indicators for  small
estuaries.  Percentages associated with portion of pie
portrays r2 of Indicator.
                                                  HIOTJT
                                                       18. 3ซ
                                                                Stdlient Hob Hot
Figure 5.8 Associations between benthic Index that has
been categorized and exposure Indicators for large tidal
rivers. Percentages associated with portion of pie portrays
r2 of indicator.
Demonstration Report, EMAP-E Louisianian Province -  1991
                                  Page 85

-------
Exposure Variable

Heavy Metals
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Nickel
Selenium
Tin
Zinc
PCBs
Congener #44
Congener #110
Congener #126
Congener #206
PAHs
C1-Chrysene
C2-Chrysene
C3-Chrysene
C4-Chrysene
C1 -Fluoranthene
C4-Phenanthrene
Dibezo(a,h)anthracene
Perytene
Pesticides
Aldrin
a!pha-BHC
alpha-Chlordane
cis-Nonachlor
4.4'-DDD
4.4'-DDT
Dioldrin
gamma-Chlordane
Mirex
Oxychtordane
Composite Totals
Total Alkanes
Alkanes > 7000 ppb1
Total Chtordanes
Number of Metals exceeding
Criteria Value

Ampelisca









-0.26"





-0.22*
-0.26"
-0.21*
-0.21*

-0.22*
-0.22*
-0.21*
-0.25*
-0.20*
-0.20*
-0.21*


-0.26**

-0.35***
-0.28"
-0.21*
-O.38"*
-0.26**
-0.38*"
-0.40"*
-0.37***


-0.22*
-0.33***


r
Mysids

-0.36"*
-0.25*
-0.26"
-0.34***
-0.27*
-0.38* *
-0.40* *
-0.35* *
-0.64* *
-0.40* *
-0.37* *
-0.38* *
-0.48* *



-0.33***









-0.46***


-0.22*
-0.51"*
-0.33***
-0.21*
-0.33***
-0.23*
-0.32**
-0.50***
-0.48***

-0.23*
-0.44***
-0.27**

.-0.47***
                                                     indicators (i.e., bottom pH, salinity,
                                                     temperature and light transmittance)
                                                     accounted for 12% of the variation.
                                                     When the categorical regressions were
                                                     developed by estuarine class, heavy
                                                     metals in sediments were associated
                                                     with major portions of the variation in
                                                     index categories for large estuaries
                                                     (49%) and small estuaries (65%) (Fig.
                                                     5.6 and 5.7, respectively).  Heavy
                                                     metals were less importantly
                                                     associated with the index in large tidal
                                                     rivers (18%) but alkanes were the
                                                     major association with the index in the
                                                     Mississippi River (26%).  However,
                                                     because the Mississippi River
                                                     contributes < 1% of the area of the
                                                     province, heavy metals were the
                                                     predominate association between the
                                                     benthic index and sediment
                                                     contaminants.

                                                     Regardless  of the statistical analysis
                                                     used, variation in the benthic index is
                                                     associated with sediment
                                                     contaminants, clearly, showing that
                                                     ecological integrity,  as measured by
                                                     the benthic index, is related to the
                                                     degree of sediment contamination.
                                                     While dissolved oxygen condition
                                                     plays a significant role, its association
                                                    with the benthic index accounts for a
                                                    small portion of the  observed
                                                    variability. Although the index was
                                                    adjusted for differences due to salinity,
                                                    several other habitat variables
                                                    apparently contribute to a small portion
                                                    of the variability observed in the index.
Table 5.4 Significant unlvarlate relationships between survival of
Ampellsca and mysld In bloassays and sediment contaminant
concentrations using Pearson correlation coefficient (r). (* = p < 0.05,
" ป p < 0.01, •" = p < 0.001).
Demonstration Report, EMAP-E Louisianian Province -  1991
Page 86

-------
   Ptttlctdit
       28.
 Hซaปjr Moto I a
       14.3>
                PAHa
                5.4s
PCBs
4.4i
Figure 5.9 Associations between Province-wide sediment
toxicity for (a) amphipods and (b) mysldes and sediment
contaminants.  Percentage associated with portion of pie
portrays R2 of Indicator.
  Hoayj Uซtal>
       44.8>
                                      Unknown
                                      27.61
                          Pittleldei
Figure 5.10 Associations between Province-wide sediment
toxicity for (a) amphipods  and (b) mysiids and sediment
contaminants.  Percentage  associated with portion of pie
portrays R2 of indicator.
                                                5.2  ASSOCIATIONS WITH
                                                SEDIMENT TOXICITY
The associations between the sediment
toxicity and sediment contaminant
concentrations were determined using two
statistical tools: (1) univariate correlations
between the bioassay survival rate at a site
and corresponding sediment contaminant
concentrations at that site, and (2)
multivariate regressions between index
values and exposure indicator values.
These assessments were completed for the
Louisianian Province. Other spatial
divisions (estuarine classes, Gulf States,
and individual estuaries) could be
evaluated after several additional years of
monitoring data are collected (i.e., a single
year's information provides too small a
sample size).

Table 5.4 shows the results of the
univariate analysis for metals, PCBs,
RAHs, pesticides, composite measures of
contaminants (e.g., total alkanes, total
PCBs). Ten-day bioassays with
Ampelisca abdita (amphipod) were most
strongly associated with pesticides
although the sediment concentration of any
individual pesticide only accounted for  no
more than 16% of the variability seen in
sediment toxicity.  While the explained
variance was small for any individual
sediment contaminant, all significant
correlations were negative showing the
increased concentrations of particular
contaminants were associated with
decreased survivorship  of amphipod in 10-
day bioassays.  Using multivariate
regressions confirmed the univariate
correlation results with 28.1% of the
Demonstration Report, EMAP-E Louisianian Province -1991
                                                         Page 87

-------
 variation in amphipod sediment toxicity
 being associated with pesticides (Fig. 5.9).
 Of the remaining 72% of the variation in
 sediment toxicity, 14% was associated with
 heavy metal concentrations and only 10%
 was associated with PAHs and PCBs
 combined.

 Table 5.4 shows the results of the
 univariate analysis for sediment
 contaminants and 3-day mysid bioassays.
 Mysidopsis (mysid) bioassays were most
 strongly associated with heavy metals.and
 pesticides.  Several heavy metals and
 pesticides were associated with greater
 than 20% of the observed variance in the
 3-day mysid tests (e.g., manganese, zinc,
 BHC, mirex, and oxychlordane).  Several
 composite contaminant values were
 strongly associated with mysid  bioassays.
 The number of heavy metals at a site
 exceeding critical values was associated
 with 22% of the variation in these
 bioassays while sites with total  alkanes >
 7000 ppb were strongly associated with
 mysid mortality (19%). About 72% of the
 variation in mysid survivorship was
 associated with sediment contaminants
 with higher concentrations of contaminants
 being associated with lower survival rates
 (Fig. 5.10).  Heavy metals and pesticides
 were associated with 45% and 20% of the
 variation in mysid survival rates,
 respectively.
 5.3  ASSOCIATIONS BETWEEN
    SEDIMENT CONTAMINANTS
    WITH ACID VOLATILE
    SULFIDES AND TOTAL
    ORGANIC CARBON

 An analysis was performed to assess the
 extent to which acid volatile sulfides (AVS)
 and total organic carbon (TOG) were
 associated with contaminant concentrations
. in sediment using Pearson correlations.
 Five of the 12 heavy metals (raw
 concentrations) were significantly
 correlated with AVS (Table 5.5). Nine of
 the 12 heavy metals were significantly
 associated with concentrations of sediment
 organic carbon. Most of the R2 values for
 individual metals were rather low (i.e., <
20%), the highest correlations were
between AVS and mercury (0.44) and
cadmium (0.43) and TOG ,and cadmium
(0.57), lead (0.49), selenium (0.45), and
mercury (0.43). Unlike raw concentrations,
aluminum-adjusted metal concentrations
were more strongly associated with TOG
than AVS. Aluminum-adjusted lead (0.61),
cadmium (0.58), selenium (0.44), mercury
(0.43), and chromium (0.43) were strongly
associated with TOG while only aluminum-
adjusted lead was strongly related to AVS
(0.47).

Many organic contaminants were
significantly associated with AVS and TOG
levels in sediments.  Nineteen of 28
alkanes  were significantly associated with
TOG concentrations while 10 alkanes were
associated with AVS. Correlations
between alkanes and TOG were 50-100%
stronger than those between AVS and
alkanes  (Table  5.5).  Fifty-eight percent of
PAHs were significantly associated with
Demonstration Report. EMAP-ELouisianian Province- 1991
                             Page 88

-------
 TOG and 34% of PAHs were associated
 with AVS. Unlike other organic
 contaminants, chlorinated pesticides were
 more strongly associated with AVS than
 with TOG (Table 5.5).

 These significant relationships suggest that
 AVS or TOG could be used as a possible
 covariate.  In addition, the relationships
 between heavy metal or organic
 contaminant concentrations and AVS or
 TOG should be investigated to determine if
 this relationship is related to bioavailabilily.
5.4 ASSOCIATIONS WITH
   DISSOLVED OXYGEN

The associations among dissolved oxygen
(DO) and habitat indicators were
investigated using univariate analysis.  Two
dissolved oxygen parameters
(instantaneous bottom DO concentration
and the minimum DO concentration) were
correlated with six habitat indicators (AVS,
TOG, mean RPD depth, temperature,
degree of stratification,  and surface area of
estuary).  Province-wide associations were
found between instantaneous bottom DO
and degree of stratification and bottom
temperature that accounted for 28% of the
variation in DO concentrations.  Minimum
DO concentrations were associated with
degree of stratification and AVS
concentrations accounting for 17% of
observed variation.

Examination of the three estuarine classes
showed that instantaneous bottom DO
concentrations were most strongly
associated with habitat  indicators in large
tidal rivers. Ninety-four percent of the
observed variability in instantaneous
 bottom DO was associated with degree of
 stratification, bottom temperature, and AVS
 concentrations. Only 31% of observed
 variability in bottom instantaneous DO in
 large estuaries was associated with a
 single habitat indicator (degree of
 stratification).  In small estuaries, 27% of
 variation in dissolved oxygen was
 associated with degree of stratification and
 bottom temperature.  Significantly smaller
 portions of overall variability in minimum
 DO were associated with habitat variables
 ip large and small estuaries (11% and
 22%, respectively).

 Evaluation of these results suggest that a
 portion of the hypoxic conditions observed
 during the 1991 Louisianian Province
 Demonstration could be related to physical
 factors (stratification, temperature).  Large
 portions of the observed variability in both
 instantaneous  and  minimum DO
 concentrations remained unexplained.
 Only DO concentrations in  the Mississippi
 River were predominately associated with
 physical factors.  If dissolved oxygen
 represents an endpoint for  estuarine
 eutrophication, much of the unexplained
variability in DO concentrations in large
 and small estuaries could be due to
eutrophic conditions (e.g., nutrient
concentrations, chlorophyll  concentrations).
As was shown in Section 4, stable isotope
analysis of carbon and nitrogen suggested
strongly eutrophic conditions
(phytoplankton sources of carbon) at
hypoxic sites.  Additional indicators would
have to be evaluated to assess the
potential for eutrophication  in Louisianian
Province estuaries.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 89

-------
Matafe
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
AI-AdJusted Metals
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Mercury
Selenium
Zinc
Alkano*
C10
C15
C17
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C30
C31
C32
C33
C34
Total
AVS
0.27"
0.43"*
0.30"
0.39"*
0.44*"

0.24*
0.43*"


0.47***
0.43*"
0.32"*
0.22*

0.28**
0.24*
0.35***


0.26**
0.25*
0.27"
0.28**
0.26**
0.26**


0.27**





r TOC
0.28**
0.57***
0.40***
0.33"*
0.49"*
0.43"**

0.26**
0.58"*
0.43"*
0.39***
0.33***
0.61*"
0.43*"
0.44***


0.47***
0.24*
0.49***
0.22*
0.31"
0.47***
0.53***
0.55***
0.58***
0.58"*
0.62***
0.33***
0.21*
0.47***
0.30**
0.33***
0.24*
0.32**
0.38***
Table 5.5  Significant Pearson correlations of sediment metal
concentration* with acid volatile sulffdes and total organic
carbon. (* ป p < .05, " * p < .01, and *** * p < .001).
PAHs
Acenaphthylene
Acenaphthene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(k)fluoranthene
Benzo(g,h,i)perylene
C1-Fluoranthene
C1 -Naphthalene
C2-Naphthalene
C3-Naphthalene
C4-Naphthalene
Chrysene
2,6-Dimethylnaphthalene
Fluorene
Fluoranthene
Ideno (1,2,3,c,d)pyrene
1 -Methylnaphthalene
2-Methylnaphthalene
1 -Methylphenanthrene
Naphthalene
Phenanthrene
Pyrene
Total PAHs
Pesticides
2,4'-DDD
2,4'-DDE
4,4'-DDE
2,4'-DDT
delta-BHC
cis-Nonachlor
Total DDT
PCBs
Total PCBs
AVS
0.35"*
0.27**
0.31"
0.52*"
0.48"*
0.61*"
0.54*"
0.58"*
0.27**
0.26*




0.55"*


0.55***
0.40***


0.38***


0.46***


0.37*"
0.30**
0.60***
0.35***


0.38*"

0.50***
r TOC
0.36***
0.29**
0.36***
0.48"*
0.45"*
0.51"*
0.48***
0.60*"
0.29**
0.26*
0.35*
0.35*
0.25*
0.24*
0.51***
0.32***
0.25**
0.53***
0.41*"
0.37***
0.32***'

0.41*"
0.27**
0.47***
0.30**

0.36***

0.45***
0.32*"
0,20*
0.28"
0.30**

0.45*"
Table S.S(Cont) Significant Pearson correlations of sediment
metal concentrations with acid volatile sulfldes and total
organic carbon. (* = p < .05, ** = p < .01, and *** = p < .001).
Demonstration Report,  EMAP-E Louisianian Province -1991
                                  Page 90

-------
                                 SECTIONS

   EVALUATION OF SAMPLING DESIGN ATTRIBUTES
The sampling design used in the 1991
Louisianian Province Demonstration is
described in detail in Summers et al.
(1993b). Only a brief description appears in
this section in order to facilitate discussion
of issues associated with the design.
These design issues involve:

•  Comparison of index sampling and
   random sampling in small estuaries;

•  Evaluation of the grid scale density (Is
   the present grid density sufficient to
   characterize estuarine characteristics?);

•  Examination of the degree of spatial
   autocorrelation within  the Louisianian
   Province dataset;

ซ  Evaluation of the degree of replication
   necessary to develop province- and
   class-wide descriptions.
6.1  COMPARISON OF INDEX
   SAMPLING AND RANDOM
   SAMPLING

During the planning of the statistical design
for the 1991 Louisianian Province
Demonstration, the method for locating the
sites within selected small estuaries and
within the segments of the large tidal rivers
were actively discussed.  Essentially two
alternatives existed: (1) random location of
sites within the bounds of the estuaries or
segments or (2) the location of index sites
believed to be representative of the
sampling space (i.e.,'small estuary or river
segment).  Basically, random locations are
probability-based with the assumption of
representativeness when sampling for
individual estuaries or they can be used as
individual points in the total space occupied
by the population of small estuaries or river
segments. Index sites are probability-
based only if enough knowledge exists to
select sites that are representative of the
sampling space being sampled. Generally,
index sampling requires fewer total
samples than random sampling (Overton,
personal communication) because of the
spatial variability of the resource.  Given
this fact, index sampling would have been
selected for sampling small estuaries.
However, it was difficult to develop criteria
for the selection of index sites (e.g., and
even more difficult to locate sites that fit
these criteria).  As a result, both index (i.e.,
located as well  as possible) and random
samples were located in each small
estuary and large tidal river segment
sampled.

All indicators were collected at both site
types. The resulting datasets were" tested
to determine if the paired random and
index sites are more alike than two
randomly selected sites.  If the sites are
similar then there is no advantage to index
sampling. If the sites are statistically
different, then index sampling could be
advantageous.  Four methods were used to
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 91

-------
 examine the question which should in ideal
 cases ali provide the same results. In
 marginal situations, differences between
 the four different tests should be helpful in
 discerning which test results are correct.
 The four methods were:

 •  Comparison of population-level
   cumulative distribution functions (CDFs)
   based on random sites and based  on
   index sites (similar CDFs would show no
   differences between population
   distributions based on random sites or
   index sites);

 •  ANOVA to determine significant
   differences between random and index
   sites based on paired sites as a factor
   (station effect) (if a station effect can  be
   shown then the pairs vary less than two
   randomly selected sites);

 •  Correlation analysis using Pearson's  r
   as a first order approximation to any
   relationship (if the slope is not different
   from 1.0 and the intercept is not different
   from 0, then the random  and index sites
   are not different); and

 •  Correlation analysis using Spearman's r
   to detect any consistent relationship
   based on ordered data.

 The types of indicators were discussed
 separately below.
6.1.1  BENTHIC RESPONSE
   INDICATORS

Benthic variables showed no differences in
the population-level distributions of number
of species, biodiversity, abundance,
, abundance of large bivalves, and
 proportional contributions of amphipods,
 decapods, bivalves, and polychaetes in
 large river segments.  Two of the CDFs
 corresponding to major variables
 comprising the benthic index biodiversity
 and proportion of total abundance as
 bivalves, are shown in Figure 6.1 and 6.2.
 Thus, at a population level (class) for
 benthic variables, there are no differences
 in index and random sampling.  However,
 ANOVA which shows paired consistencies
 rather than total distribution consistencies
 showed only 3 of the  8 benthic variables
 were similar at index and random sites in
 river segments (Table 6.1).  Similar results
 were seen using correlational analyses
 where only percent decapods and percent
 polychaetes showed no differences
 between index  and random sites.  In  large
 tidal river segments, higher mean number
 of benthic species, higher mean benthic
 abundance, and higher mean benthic
 diversity occurred at index sites (Table
 6.1).  Lower mean abundance and mean
 proportion of bivalves were seen at index
 sites.  While CDFs of  population
 distributions of these variables are not
 different between index and random sites
 in rivers (i.e, class-wide estimates are not
 significantly different), the sites do not
 occur at the same locations along the
 distribution function.  Thus, for large tidal
 rivers, random sites are not representative
 of the segments they are supposed to
 represent and multiple sites would be
 required to represent individual segments
 of the rivers.

 Benthic variables showed no differences in
 the population-level distributions of number
 of species, biodiversity, abundance,
 abundance of large bivalves, and
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 92

-------
100
90-
SO
ซ•• ซป
o 70
ซ 60
* so

| 30
o
20
10
0-
0

'*'*)
sS
/
y.

i
i
•0 0.2 0.4 0.6 0.8 1.0 1.2








—~ Large River Index
~ ~ Large River Rondos








• i 	 1 	 f
1.4 1.8 1.8 2.0
ShonBon-fiener Divereitjr Index
Figure 6.1 Cumulative distribution functions for biodiversity as bivalves for random and Index sites In large tidal rivers.
100


90

80
oป __
i 60
I"
•5 4"
| 30
20-
10-
.
OH

r-




;
.

,'




• i i i 	 1 	
0 200 400 SOO 800
Mean Abundance













— Large River Index
~ ' Large River Randon














	 1 	 1 	 r
1000 1200 1400

Figure 6.2 Cumulative distribution functions for proportion of total abundance as bivalves for random and Index sites in large
tidal rivers.
Demonstration Report, EMAP-E Louisianian Province- 1991
Page 93

-------
 proportional contributions of
 amphipods.decapods, bivalves, arid "
 polychaetes in small estuaries/small tidal
 rivers.  Two of the CDFs corresponding to
 major variables comprising the benthic
 index are shown  in Figure 6.3: biodiversity
 and Figure 6.4: proportion of total
 abundance as bivalves.  Thus, at a
 population level (class) for dissolved
 oxygen variables, there are no differences
 in index and random sampling. Similarly,
 the results of the ANOVA testing showed
 no significant differences for any benthic
 response variables except for abundance
 of large bivalves  (Table 6.2).   Pearson and
 Spearman correlations generally confirmed
 the lack of significant differences between
 random and index samples for benthic
 response variables (13 of 16  pairs were
 significantly correlated).  The three
 correlative analyses (mean abundance,
abundance of large bivalves, and
proportion of bivalves showed relatively
small differences [mean abundance (98.7
organisms versus 93.6 organisms), large
bivalve abundance (6.1  versus 5.9), and
percent bivalves (18.0% versus 15.1%)].
These small differences reflect the low
variability seen in  the small estuarine class
whether using randorp or index sampling.
Thus, random sampling js adequate to    ;
represent small estuaries at the class level
or at reduced levels (e.g:, small estuaries
in individual states or individual estuaries).
Varfabfo
Moan number
olspodos
Mean abundance
Diversity
Largo btvafva
abundance
% Amphipods
% Decapods
% Bivalves
% Potychaets
Benthic Index
Anova Pearson Spearman
Pr>F r p r p
0.299 0.132 NS 0.134 MS
0.691 -O.280 MS -0.030 NS
0.116 0.355 NS 0.411 NS
0.032
0.521
0.007 0.997 .0001 0.745 .0133
0.616 -0.048 NS -0.102 NS
0.000 0.872 .0010 0.908 .0003
0.930 -0.508 NS -0.444 NS
Mean
Random ' Index
2.53
38.67
0.22 ;
4.75
3.27
0.11
23.09
31.89
-1.29
3.00
63.60
0.26
f.so
b.oo
0.30
5.23
27.77
-2.28
Table 6.1 Results of Index and random sampling comparisons for benthic response variables In large tidal rivers using ANOVA
(p < 0.1 a no significant difference) and Pearson and Spearman correlations (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province -  1991
                              Page 94

-------
Variable
Mean number
of species
Mean abundance
Diversity
Large bivalve
% Amphipods
% Decapods
% Bivalves
% Polychaetes
Anova
Pr>F

0.000
0.000
0.050
0.481
0.000
0.041
0.000
0.003
Pearson
r P

0.714
0.204
0.687
056
0.757
0.669
0.240
0.381

.0001
NS
0001
NS
.0001
.0001
NS
.0106
, Spearman
r P

0.658
0.520
0.633
0.334
0.568
6.353
0.340
0.364

.0001
.0003
.0001
.0375
.0001
. .0187
.0237
.0150
Mean
Random Index

11.11
98.67
0.63
6.13
4.48
2.74
18.03
42.44

14:89
93.56
0.75
5.86
5.61
3.40
15.08
45.49
rivers using ANOVA (p < 0.1 = no significant difference), Pearson and Spearman correlations (p < 0.1 = no significant
difference)
       100


        90


        80
    €>
    "•  i~
    ซ>   70
    -••*

    S   60


    |   50

    |   40


    ง   30
    o
        20


        10

         o^
-~~ SnolI  Estuary  Index
          Ettuory  Random
          0.0     0.2     0.4     0.8     fl.8     1.0     1.2     1.4

                                   Shannon-Wiener  Diversity Index
            1.6
1.8
2.0
Figure 6.3 Cumulative distribution functions for biodiversity as bivalves for random and Index sites In small estuaries/small
tidal rivers.	-	- 	 - 	-	 	 -	
Demonstration Report, EMAP-E Louisianian Province -1991
                       Page 95

-------
100-
90-
SO-
5 70-
S BOJ
• 1
I 40
| 30-
0
20-
10-
0-
(

/rf--~ " 	 "*"
r^i'
j /
/



) 200 400 600 800
(lean Abundance








— Snail Ettuary Index
•"•Snail Eftuary Randon








1000 : 1200 1400

Figure 6,4 Cumulative distribution functions for proportion of total abundance as bivalves for random and Index sites in
•mall estuaries/small tidal rivers.
6.1.2 FINFISH RESPONSE
   INDICATORS

Finfish community response indicators
showed no differences in the population-
level distributions of number of species,
abundance, and proportional contributions
of marine catfish, puffers, sciaenids,
clupeids, and bothids in large river
segments. Two of the CDFs of number of
finfish species and abundance/trawl are
shown in Figure 6.5 and 6.6. Thus, at a
population level (class) for finfish variables,
there are no differences in index and
random sampling. ANOVA and correlation
analyses,  however, showed significant
differences between random and index
sampling for all finfish indicators in large
tidal rivers (Table 6.3).  Major differences
revealed, at random versus index sites,
significantly higher proportions of catfish
(70.2% versus 48%), higher proportions of
bothids (flounders)(7.8% versus 0.9%), and
lower proportions of clupeids (1.3% versus
11,4%). Like benthic response indicators,
fish indicators would be better to represent
individual segments  but class-level
population distributions based on random
sites are no different than  index sites.

Finfish community response indicators
showed no differences in the population-
level distributions of  number of species,
abundance, and proportional contributions
of marine catfish, puffers, sciaenids,
clupeids, and bothids in small
estuaries/small tidal rivers.  Two of the
CDFs of number of finfish  species and
abundance/trawl are shown in Figures 6.7
and 6.8. ANOVA and correlation analyses
Demonstration Report, EMAP-E Louisianian Province • 1991
                             Page 96

-------
 generally showed no differences between
 random and index sites in small estuary
 with the exception of puffers (Table 6.4).
 Due to the low proportional  abundance of
 puffers (0.6% and 0.4% in random and
 index sites, respectively), the statistical
 difference does not represent any
 ecological difference.  Random sampling
 for finfish indicators is adequate to
 represent small estuaries at the class level
 or at reduced levels (e.g., small estuaries
 in individual states or individual estuaries).
6.1.3  CONTINUOUS DISSOLVED
   OXYGEN

Continuous dissolved oxygen parameters
(i.e., minimum concentration, percent of
time with DO < 2 ppm, and percent of time
with DO < 5 ppm) were collected at
random and index sites only in the small
estuary class. The above dissolved
oxygen variables showed no differences in
the population-level distributions in small
estuaries/small tidal rivers.  The CDFs
corresponding to these variables are shown
      100
       90
       80

   JF  70
   o  60

   i"
   i  40

   !  *
   o
       20-
       10-
        0-
              Lorge River Index
              Large River Random
   8      10      12      14
Number  of  Fish Species
                     16
                                                                          18
20
Figure 6.5 Cumulative distribution function* for number of fish species/trawl for random and index sites in large tidal rivers,
Demonstration Report,  EMAP-E Louisianian Province - 1991
                             Page 97

-------
100-
90-
BO-
S' 70-
•*-ซ
S 60-
^ 50-
= 40-
1 30-
O
20-
10-
o-

	 -""












•"-" Large River Index
— Large River Random

| I I 1 1 1 1 1 1
] 50 100 150 200 250 300 350 400 450
Fish Abundance (no. /trail)













500
Figure 6.6 Cumulative distribution functions for abundance/trawl for random and Index sites In large tidal rivers.
Variable
# Fish species
Fish abundance
Fish index
% Catfish
% Puffers
% Scianids
% Clupeids
%Bothids
Anova
Pr>F
0.043
0.562
0.045
0.234

0.174
0.593
0.649
Pearson
r P
0.587 .09
-0.053 NS
0.625 .07
0.462 NS

0.370 NS
-0.136 NS
•0.178 NS
Spearman
r P
0.558
0.555
0.607
0.492

0.204
0.219
-0.244
NS
NS
.08
NS

NS
NS
NS
Mean
Random Index
3.44
38.00
1.88
70.20
0.00
12.49
1.28
7.80
4.70
24.60
2.82
48.03
0.00
16.67
11.35
0.82
Table 6.3 Results of index and random sampling comparisons for flnflsh response variables In large tidal rivers using ANOVA
(p < 0.1 = no significant difference) and Pearson and Spearman correlations (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 98

-------
             100


              90


              BO


           ?  70
          •*•ป
           c
           S  60


          !  50


          |  +0

           I  30
          C*

              20


              10


               0
                        — SซalI  Estuary Indix
                        ~" Snail  Eituory Randoi
                                     6      6      10      12      14
                                          (timber of Fish Species
16     18
                                                 20
Figure 6.7   Cumulative distribution functions for number of flsh species/trawl  for random and Index sites In small
estuaries/small tidal rivers.
                                                              - SnalI  Estuary Index
                                                              - Snail  Estuary Randon
                              100
150     200     250     300    350    400    450    500

   Fish Abundance (no./trawl)
Figure 6.8 Cumulative distribution functions for abundance/trawl for random and Index sites In small estuaries/small tidal
rivers.     	    	  ~	  ~	"	"" '  ""	   ' 	'' "	'
Demonstration Report, EMAP-E Louisianian Province -1991
                                                 Page 99

-------
Variable
# Rsh spp
Rsh abundance
% Catfish
% Puffers
% Scianids
% Clupeids
%Bothids
Anova
Pr>F
0.019
0.001
0.008
0.487
0.006
0.009
0.063
Pearson .
r p -
0.357
0.442
0.390
-0.106
0.310
0.194
0.251
.0257
.0034
.0142
NS
.0551
NS
NS
Spearman Mean
r p Random Index
, 0.318
0.360
0.234
• -0.143
0.343
0.448
0.519
.0484
0192
NS
NS
.0325
.0043
.0007
6.08
73.21
17.29
0.56 ,
23.36
10.92
0.93
7.38
70.19
, 14.96
•0.35
• 23.47
8.05
1.48
 	• • • "————• .••--ซ"..-• u.!—ซ..• oaiiipiiny.^MMjjaiu.uiio mi HiiMDii itmpufisu vanaDies in smaii Qsiuanes/small Bdal rivers
 uซlng ANOVA (p < 0.1 * no significant difference) and Pearson and Spearman correlations (p < 0.1 = no significant difference).
 in Figures 6.9-6.11. Thus, at a population
 level (class) for benthic variables, there are
 no differences in index and random
 sampling. ANOVA and correlational
 analyses confirmed this similarity between
 random and index sites for continuous
dissolved oxygen indicators (Table 6.5).
Random sampling for dissolved oxygen
indicators is adequate to represent small
estuaries at the population level or at
reduced spatial scales (e.g., state small
estuaries).
                 100

                 90

                 BO
              •
              5  70
              6
              Z  so

              f  s<
              >•
              ^  40

              1  30
              o
                 20-

                 10-

                  0-

       —""Sซal I  Eituarjr Indtx
       — "Siall  Eituary Rondon
                                            458
                                           Hlnlnni D.O. (ppn)
                             10
Figure 6.9 Cumulative distribution functions for minimum dissolved oxygen concentration for random and index sites in small
ostuades/smali tidal rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 100

-------
              100

               90

               80


           |  70
           e
           S  so



           •  50

           |  40

           ง  30
           o
               20

               10
~— Snol I  Ejtunry Index
— • Snail  Estuary Random
                100     90     80     70     60     50     40

                                             x D.O. < 2.0 ppn
                                          —i—

                                           30
 Figure 6.10 Cumulative distribution functions for percent of time at concentrations < 2 ppm for random and Index sites In small
 estuaries/small tidal rivers.

100
90
80
ซป
ฐ* -.
.s 70
c
ป 60
! 50-
'-i ซ
1 30^
0
20-
10-
0-
1


— Snol 1 Estuary Index
- - Snal 1 Estuary Randon

—/ 	 ^

10 90 BO 70



1
i
i
i
i
i
^ .-.--""
*

' ' i i 	 1 	 1 	 r
60 50 40 30 20 10 0
x D.O. < 5.0 ppn






'
Figure 6.11 Cumulative distribution functions for percent of time at concentrations < 5 ppm for random and Index sites In small
estuaries/small tidal rivers.                         	•••-             -••--             	    	
Demonstration Report. EMAP-E Louisianian Province -1991
                                                             Page  101

-------
• Anova Pearson
Variable Pr>F r p
Minimum/Do 0.000 0.541 .0002
% Time DO < 2 0.001 0.626 .0001
% Time DO < 5 0.000 0.437 .0030
Spearman ' Mean
r p Random Index
0.464 .0015 4.81 4.77
0.628 .0001 7.14, 3.43
0.358 .0170 26.50 27.36
Table 6.5 Result* of Index and random sampling comparisons for dissolved oxygen response variables In small estuaries/small
tidal river* using ANOVA and Pearson and Spearman correlations (p < 0.1 = no significant difference).
6.1.4  HUMAN USE INDICATORS

Human use response indicators showed no
differences in the population-level
distributions of presence of water clarity,
marine debris and contaminants in edible
fish fillets in large river segments.
Cumulative distribution functions for water
clarity, total PCB and mercury
concentrations in catfish are shown in
Figures 6.12-14.  Thus, at a population
level (class) for human use variables, there
are no differences in index and random
sampling. ANOVA and correlation
analyses show differences between
              100-

              90-

              80

           I 70
           e
           S  60


           5  "
           |  ซ

           |  30
           ซ
              20

              10H
               o-j
         ~"~ Largs Rivtr ludtx
         — Large Rlw Randon
                0.0        0.1         0.2        0.3        0.4
                                  Photosynthetical ly Actiปซ Radiation
                   0.5
0.6
Figure 6.12 Cumulative distribution functions for percent surface light transmlttance at 1 m for random and Index sites In large
tidal rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
                              Page 102

-------
                  100

                  10

                  BO

               I 70
               e
               S  10

               i  "
               f  40

               g  30
               o
                  20

                  10

                   0
                                    1 Larjt Riปtr Indtx
                                    ' Largt Riปปr Raซ4o*
                                    100              200
                                              Total PCBt (ng/j)
                                       500
        400
 Figure 6.13 Cumulative distribution functions for concentrations of total PCBs In catfish flilete for random and Index sites In
 large tidal river*.
              100

               90

               80

            ?  70
            .**
            2  60

            •  50
            |  40

            |  30
            u
               20

               10
                                   — Largi Rivซr Indsx
                                   - • Large River Rondoi
                 0.0
0.1
0.2         0.3
       Mercury
                                                                0.4
0,5
                                                           0.6
 Figure 6.14 Cumulative distribution functions for concentrations of mercury In catfish fillets for random and Index sites in large
" tidal rivers. "  '  "  '	"""'""	""""	•"    	         	"	 " "" "" ".    	
 Demonstration Report,  EMAP-E Louisianian Province -1991
                                                          Page 103

-------
 random and index sites for tissue
 contaminant concentrations and presence
 of marine debris but no significant
 differences for water clarity (Table 6.6).
 Therefore, a single random sample for
 human use indicators  may not be adequate
 to represent large tidal rivers at reduced
 spatial scales (e.g., segments).

 Human use response  indicators showed no
 differences in the popuFation-level
 distributions of presence of water clarity,
 marine debris and  contaminants in edible
 fish fillets in small estuaries/small tidal
 rivers.  Cumulative distribution functions for
water clarity, total PCB and mercury
concentrations in catfish are shown in
Figures 6.15 through 6.17.  Thus, at a
population level (class) for human use
Variables, there are no differences in index
and random sampling.  ANOVA and
correlation analyses show no differences   <
between random and index sjtes for human
use indicators (Table 6.7). Therefore,
random sampling for human use indicators
is adequate to represent small
estuaries/small tidal rivers at the population
level or at reduced spatial scales (e.g.,
states).
Variabfo
PAR
Aldrin
Chtoidane
2.4'-DDD
4.4'.DDD
2.4'-DDE
4.4'-DDE
2.4'-DDT
4,4'-DDT
Oiotdrin
Endosulfan
Endrin
Gamma BHC
HCB
Hopta Epox
Heptachlor
Mirex
Tot PEST
Total PCB
Toxaphene
Transnona
Anova Station Peareon
P-Value r p
0.420
0.718
0.578
0.638
0.795
0.920
0.928
0.842
0.004
0.689
0.514
0.787

0.733
0.901
0.934
0.259
0.732
0.950
0.847
0.913
.02
0.700
0.764
-0.073
•0.183
-0.383
-0.313
0.498
-0.563
-0.346
0.399
-0.202

0.461
-0.461

-0.270
0.091
-0.375
-0.192
0.634
NS
.08
.05
NS
NS
NS
NS
NS
NS
NS
NS
NS

NS
NS

NS
NS
NS
NS
NS
Spearman
ซ• P
.05
0.291
0.618
0.144
0.218
-o:i 27
-0.315
0.432
-0.468
0.036
0.221
0.164

0.530
-0.360

-0.162
0.468
-0.306
-0.254
0.559
NS
NS
NS
NS
NS
NS ,.
NS
NS
NS
NS
NS
NS

NS
NS

NS
NS
NS
NS
NS
Mean
Random Index
.21
2.14
8.97
75.73
11.17
9.11
1.97
33.01
22.98
49.46
2.36
5.94
0.00
6.30
8.26
0.00
26.41
267.81
133.07
558.33
6.95
.22
2.01
8.51
80.44
1^.85
5.03
6.34
17.00
32.17
16.12
0.86
3.22
0.00
3.34
4.90
2.27
9.43
216.83
157.70
791.67
8.98
	 ___	.	... __...r....j comparisons for human use response variables In large tidal rivers using
ANOVA (p < 0.1 = no significant difference) and Pearson and Spearman correlations (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 104

-------
Anova Station Pearson
Variable
PCB8
PCB18
PCB28
PCB44
PCB52
PCB66
PCB77
PCS 99
PCB 101
PCB 105
PCB 118
PCB 126
PCB 128
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
- PCB 209
Silver
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Nickel
Lead
Selenium
Tin
Zinc
P-Value
0.850
0.850
0.866
0.981
0.990
0.971
0.919
0.873
0.919
0.886
0.829
0.922
0.978
0.000
0.531
0.596
0.826
0.393
0.285
0.659
0.702
0.317
0.728
0.391
0.897
0.901
0.672
0.363
0.155
0.616

0.311
0.545
0.268
r


0.306
-0.380
-0.630
0.093
-0.263
-0.272
-0.143
-0.920
0.172
-0.335
0.247
0.510
0.361
0.028
0.119
0.482
0.611
0.442
0.244
0.028
•0.721
0.328
•0.456
•0.620
•0.107
•0.485
-0.345
•O.299

0.748
•O.134
0.435
P


NS
NS
NS
NS
NS
NS
NS
.01
NS
NS
NS
NS
NS
NS
NS
NS
.06
NS
NS •
NS
NS
NS
NS
NS
NS
NS
NS
NS

NS
NS
NS
Spearman
r

<
0.273
-0.394
-0.491
i-0.164
0.432
-0.288
-0.198
-0.691
0.631
-0.270
-0.108
0.455
0.468
0.054
0.000
0.512
0.444
0.430
0.178
-0.100
-O.900
0.600
-0.456
-0.574
0.200
rO.100
k>.100
0.083

0.632
0.100
0.500
P


NS
NS
NS
NS
NS
NS
NS
.09
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
....- NS
NS
.03
NS
NS
NS
NS
NS
NS
NS

NS
NS
NS
Mean
Random
0.00
0.00
1.14
1.73
5.48
5.60
10.63
12.87
9.67
7.14
13.54
7.44
2.13
0.67
24.60
6.47
14.14
4.88
3.19
4.38
6.76
0.24
8.92
4.84
0.08
0.11
12.78
44.88
0.22
0.34
0.00
0.48
2.51
53.65
Index
0.61
0.80
1.14
2.54
3.96
5.72
11.70 t
12.71
9.23
13.75
11.94
10.74
2.89
5.28
28.89
10.11
20.58
5.20
2.Q3
3.83
9.96
0.30
41.08
2.09
0.04
0.18
6.60
• 38.52
0.10
0.21
0.00
0.33
0.45
34.05
Table 6.6 (Cent) Results of Index and random sampling comparisons for human use response variables In large tidal rivers
using ANOVA (p < 0.1 = no significant difference) and Pearson and Spearman correlations (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province - 1991
Page  105

-------
 6.1.5  HABITAT INDICATORS

 Habitat indicators showed no differences in
 the population-level distributions of
 temperature, salinity, pH, instantaneous
 bottom dissolved oxygen concentration,
 degree of stratification, RPD depth, acid
 volatile sulfide concentration, percent total
 organic carbon and percent silt-clay
 presence in large river segments.
 Cumulative distribution functions for bottom
 pH, stratification, and bottom salinity are
 shown in Figures 6.18-6.20.  Thus, at a
 population level (class) for habitat
 variables, there are no differences in  index
 and random sampling.  ANOVA and
 correlation analyses showed significant
 differences for bottom dissolved oxygen,
 degree of stratification, bottom  pH, bottom
 salinity, mean RPD depth, acid volatile
 sulfides, total organic carbon, and percent
               silt-clay (Table 6.8).  However, the mean
               differences are rather small with 0.2 ppm
               for dissolved oxygen differences, 0.8 ppt
               for stratification differences, <0.1  pH units,
               1.2 ppt salinity, 0.4 ppb AVS, 0.1% TOG,
               and < 1% silt-clay. While the small
               variability observed in the random and
               index sites results in significant differences,
               these differences do not appear to be
               ecologically meaningful.  Only the
               differences observed for mean RPD depth
               seems a "real" statistical  difference with
               index sites having RPD depths about 19
               mm deeper than those at random sites.
               However, the difference between 58 mm at
               random sites and 76 mm at index sites
               does not appear to be ecologically
               significant. Random  sampling is  adequate
               to represent habitat variables in large tidal
               rivers.
                                                       — Snail  Est/Riv  lndปr
                                                       ~ " Snail  Est/Rit  Randoi
                         0.1
   0.2        0.3        0.4
Photosynthetical ly Active Radiation
                                                                   0.5
0.6
Figure 6.15 Cumulative distribution functions for percent surf ace light transmittance at 1 m for random and index sites in small
estuaries/small tidal rivers.
Demonstration Report,  EMAP-E Louisianian Province - 1991
                                            Page 106

-------
              100

               90

               80

               70

               SO

               SO

               40

               30

               20

               10

                0
— Snail Ettuory  Indix
—Snail Estuary  Rondos
                                        100
 —i—
  200
                                                                                      300
                                             Total PCBs (ng/g)
Figure 6.16 Cumulative distribution functions for concentrations of total PCBs In catfish fillets for random and Index sites In
small estuaries/small tidal rivers.
           3  60
                                                              — Snol 1  Estuary  lndปx
                                                              ~ * Snal 1  Estuary  Randoi
                                               3

                                               Mtrcury
Figure 6.17 Cumulative distribution functions for concentrations of mercury In catfish fillets for random and Index sites In small
estuaries/small tidal rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
                        Page 107

-------
 Habitat indicators showed no differences in
 the population-level distributions of .
 temperature, salinity,  pH, instantaneous
 bottom dissolved oxygen concentration,
 degree of stratification, RPD depth, acid
 volatile sulfide concentration, percent total
 organic carbon and percent silt-clay
 presence in small estuaries.  Cumulative
 distribution functions for  bottom pH,
 stratification, and bottom salinity are shown
 in Figure 6.21-23. Thus, at a population
 level (class) for habitat variables, there are
 no differences in index and random
 sampling. AN OVA and correlation
 analyses showed significant differences for
 only bottom pH and total organic carbon
 (ANOVA test only). Correlation analyses
 were showed index and random sites to be
 similar for TOC and the mean differences
 between random and  index sites were 0.1
 pH units and < 0.1% TOC.  These
 differences in small estuarine environments
                                      do not represent ecological differences;
                                      therefore, random sampling is adequate for
                                      representing habitat variables in small
                                      estuaries at the population level and at
                                      reduced scales.
   100

   ง0

   BO

jr 7ฐ

S  ซo-
•
^  50-

~  40-
3
SI

   20-

   10-
              0-
                                                        —"Large River Index
                                                        — Large River Rondo!
                                                                   10
                                                                   11
                                        pH  at thi Button
Figure 6.18 Cumulative distribution functions for bottom pH for random and index sites In large tidal rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
                                                                  Page 108

-------
Variable
PAR
CROAKER
Aldrin
Chlordane
2,4'-DDD
4,4'-DDD
2,4'-DDE
4,4'-DDE
2,4'-DDT
4,4'-DDT
Dieldrin
Endosulfan
Endrin
Gamma BHC
HCB
Hepta Epox
Heptachlor
Mirex
Tot. PEST
Total PCB
Toxaphene
Transnona
PCB 8
PCB 18
PCB 28
PCB 44
PCB 52
PCB 66
PCB 77
PCB 99
PCB 101
PCB 105
PCB 118
PCB 126
PCB 128
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
PCB 209
Silver
Aluminum
Anova Station Pearson
P-Value r p
.001

0.030
0.001
0.032
0.780
0.108
0.954
0.001
0.000
0.000
0.101
0.135
0.000
0.000
0.000
0.254
0.492
0.000
0.192
0.000
0.301
0.883
0.112
0.030
0.057
, 0.001
0.282
0.141
0.002
0.365
0.556
0.309
0.092
0.323
0.114
0.266
0.000
0.570
0.000
0.064
0.013
0.520
0.035
0.660
.64

-0.138
-0.188
0.585
-0.172
-0.166
-0.246
-0.316
0.302
0.161
0.073

-0.064
0.160
0.120
-0.110
0.296
0.026
-0.072
-0.032
-0.067

0.366
0.293
0.224
-0.110
-0.215
-0.111
-0.237
-0.183
-0.249
0.651
-0.294
-0.067
0.378
0.518
-0.012
0.489
-0.276
-0.026
0.170
0.466
0.257
.001

NS
NS
.017
NS
NS
NS
NS
NS
NS
NS

NS
NS
NS
NS
NS
NS
• NS
NS
NS

NS
NS
NS
NS
NS
NS
NS
NS
NS
.006
NS
NS
NS
.040
NS
.050
NS
NS
NS
NS
NS
Spearman
r P
.75

-0.228
.-0.127
0.310
-0.136
-0.287
-0.272
-0.231
, 0.639
0.179
0.310

-0.091
0.244
i 0.424
0.164'
0.336
0.519
-0.081
0.011
-0.067

0.443
0.318
-: 0.015
0.196
-0.294
0.098
!-0.114
-0.316
0.138
0.443
-0.214
0.023
0.438
0.322
0.099
0.619
i-Q.336
0.164
0.191
6". 423
0.189
.001

NS
NS
NS
NS
NS
NS
NS
.007
NS
NS

NS
NS
NS
NS
NS
.04
NS
NS
NS

.09
NS
NS
NS
NS
NS
NS
NS
NS
.09
NS
NS
.09
NS
NS
.01
NS
NS
NS
NS
NS
Mean
Random Index
.16

1.13
1.65
8.97
0.66
1.39
0.60
1.42
15.08
3.45
0.00
1.99
0.21
20.89
1.96
0.60
9.47
50.26
38.36
1700.00
1.27
1.69
2.22
0.65
1.02
1.40
1.71
1.37
1.92
1.75
2.11
2.28
1.71
0.83
0.36
5.74
1.06
2.16
2.01
1.71
3.53
5.21
0.26
6.74
.20

0.64
3.53
11.59
1.41
1 .64
0.69
2.00
44.42
2.02
1.03
1.29
0.00
0.49
5.68
1.73
12.57
91.38
37.16
100.00
1.62
0.21
0.00
0.72
0.88
4.54
2.13
1.13
2.31
2.94
7.33
5.70
1.85
1.03
0.63
3.05
1.67
3.72
0.90
1.12
0.88
2.32
0.16
4.94
river* using ANOVA (p < 0.1 = no significant difference), Pearson and Spearman correlations (p < 0.1 = no significant
difference).
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 109

-------
Anova Station Pearson
Variable
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Nickel
Lead
Selenium
Tin
Zinc
SHRIMP
Aldrin
Chtordane
2,4'-DDD
4.4'-DDD
2.4--DDE
4,4'-DDE
2.4'-DDT
4.4'-DDT
Dioldrin
Endosulfan
Endrin
Gamma BHC
HCB
Hepta Epox
Heptachlor
Mirox
Tot PEST
Total PCB
Toxaphene
Transnona
PCB 8
PCB 18
PCB 28
PCB 44
PCB 52
PCB 66
PCB 77
PCB 99
PCB 101
PCB 105
PCB 118
PCB 126
PCB 128
P-Value
0.024
0.423
0.439
0.047
0.685
0.047
0.003
0.000
0.091
0.007
0.219

0.849
0.850
0.052
0.057
0.993
0.345
0.768
0.603
0.759



0.067
0.024

0.500
0.381
0.465

0.041


0.987
0.057
0.001
0.282
0.141
0.068
0.208
0.625
0.720
0.522
0.559
r
0.233
-0.167
-0.175
•O.040
-0.211
0.320
0.676
0.262
0.294
0.192
0.130


0.086
0.164
0.738
-0.189
0.661

0,593
0.127



0.259
0.819

0.629
0.485
0.287

0.468


-0.189
0.388
0.131
0.448
0.614
0.495
0.597
0.499
-0.227
-0.115
0.291
P
NS
NS
NS
NS
NS
NS
.03
NS
NS
NS
NS


NS
NS
.02
NS
.05

.09
NS



NS
<.01

.06
NS
NS

NS


NS
NS
NS
NS
.03
NS
.09
NS
NS
NS
NS
Spearman
r
0.210
-0.167
-0.010
-0.094
-0.088
0.292
0.722
0.100
0.266
0.363
0.319


0.312
0.141
0.586
-0.189
0.661

0.339
0.344



0.435
0.802

0.068
0.734
0.295

0.713


-0.189
0.317
0.388
0.855
0.783
0.700
0.851
0.614
0.135
0.169
0.423
P
NS
NS
NS
NS
NS
NS
.02
NS
NS
NS
NS


NS
NS
.09
NS
.05

NS
NS



NS
<.01

NS
.02
NS

.03


NS
NS
NS
<.01
<.01
.03
<.01
.08
NS
NS
NS
Mean
Random
1.92
0.01
0.14
1.05
28.30
0.20
0.35
0.18
0.60
0.72
27.96

0.18
1.10
3.66
0.99
0.46
0.68
0.08
24.04
0.76
0.00
0.00
0.00
2.13
3.12
0.00
19.24
57.31
16.60
0.00
0.84
0.00
0.00
0.17
1.02
1.40
1.71
1.37
1.22
0.75
1.18
1.62
0.71
0.32
Index
1.70
0.01
0.22
1.11
24.44
0.03
0.23
0.50
0.38
0.59
28.32

1.16
0.72
2.12
1.09
1.02
0.90
0.32
59.78
0.69
0.00
0.00
0.00
1.95
2.67
0.00
18.65
92.35
34.11
0.00
1.23
0.00
0.00
0.65
1.12
3.64
2.70
1.44
1.15
0.92
1.71
1.52
1.02
0.32
Table 6.7(Cont) Results of Index and random sampling comparisons for human use response variables In small estuaries/small
tidal river* utlng ANOVA (p < 0.1 * no significant difference), Pearson and Spearman correlations (p < 0.1 = no significant differ
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 110

-------
Anova Station Pearson
Variable
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
PCB 209
Silver
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Nickel
Lead
Selenium
Tin
Zinc
CATFISH
Aldrin
Chlordane
2,4'-DDD
4,4'-DDD
2,4'-DDE
4,4'-DDE
2,4'-DDT
4,4'-DDT
Dieldrin
Endosulfan
Endrin
Gamma BHC
HC8
Hepta Epox
Heptachlor
Mirex
Tot. PEST
Total PCB
Toxaphene
Transnona
P-Value

0.612
0.421
0.093
0.991
0.813
0.534
0.845
0.644
0.951
0.258
0.279
0.983
0.808
0.710
0.772
0.985
0.709
0.492
0.700
0.899

0.241
0.104
0.551
0.251
0.034
0.932
0.009
0.240
0.018
0.381
0.689
0.000
0.006
0.427
0.002
0.091
0.591
0.046
0.491
0.102
r

0.150
0.574
0.565
-0.250
0.829
0.784
-0.083
-0.235
0.493
-0.307
0.598
0.585
-0.032
0.074
0.598
0.401
0.410
0.321
-0.023
0.372

0.105
0.360
0.140
0.187
0.568
-0.081
0.650
0.004
0.558

0.003

0.487
0.180

-0.003
0.048
0.364

0.403
P

NS
NS
NS
NS
<.01
<.01
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

NS
NS
NS
NS
<.01
NS
<.01
NS
.01

NS

.03
NS

NS
NS
NS

.07
Spearman
r

0.581
0.071
0.371
-0.246
0.594
0.284
-0.034
-0.181
0.051
-0.154
0.583
0.676
0.051
0.103
0.564
0.631
0.663
0.359
-0.051
0.410

-0.057
0.149
0.239
0.338
0.154
0.014
0.289
0.050
p.572
1
0.111

0.837
0.258

0.244
0.303
0.527

0.359
P

NS
NS
NS
NS
.09
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

NS
NS
NS
NS
NS
NS
NS
NS
<.01

NS

<.01
NS

NS
NS
<.01

NS
Random
0.00
1.85
1.15
0.76
0.20
0.88
1.81
3.05
0.15
107.58
3.28
0.10
3.37
17.98
73.93
0.21
4.87
0.01
0.25
1.30
48.21

1.93
3.45
32.67
2.27
2.38
1.04
3.63
24.45
1.31
0.81
1.28
0.30
1.71
1.54
0.72
14.75
96.81
59.86
0.00
3.32
Mean
Index
0.00
1.81 '
2.45
0.78
0.20
5.10
1.71
13.89
0.15
73.62
2.78
0.04
0.18
10.36
63.23
0.22
0.48
0.08
0.16
0.84
41.66

0.96
4.27
85.22
4.37
3.24
0.91
4.07
26.66
1.65
0.00
0.87
0.00
1.31
2.80
0.00
12.87
154.00
83.66
103.45
.4.73
Table 6.7 (Cont)  Results of index and random sampling comparisons for human use response variables in small
estuaries/small tidal rivers using ANOVA (p < 0.1 = no significant difference), Pearson and Spearman correlations (p < 0.1
= no significant differs
Demonstration Report, EMAP-E Louisianian Province -1991
Page  111

-------
Variable
PCB 8
PCB 18
PCB 28
PCB 44
PCB 52
PCB 66
PCB 77
PCB 99
PCB 101
PCB 105
PCB 118
PCB 126
PCB 128
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
PCB 209
Silver
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Mercury
Nickel
Lead
Selenium
Tin
Zinc
Anova Station Pearson
P-Value r p
0.000
0.539
0.955
0.992
0.717
0.105
0.000
0.002
0.021
0.426
0.011
0.129
0.272
0.204
0.618
0.243
0.095
0.295
0.101
0.948
0.501
0.970
0.928
0.000
0.144
0.000
0.580
0.800
0.803
0.657
0.960
0.550
0.780
0.829

-0.206
-0.137
-0.192
0.209
0.266
0.466
0.312
0.336
0.489
0.434
0.086
0.616
0.142
0.242
0.394
0.096
0.316
-0.106
0.086
-0.065
0.208
-0.116
0.392
0.211
0.244
0.254
-0.243
0.281
-0.100
"-0.064
-0.054
-0.406

NS
NS
NS
NS
NS
.03
NS
NS
.03
.05
NS
<.01
NS
NS
.09
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Spearman
r P

-0.268
-0.165
-0.065
0.393
0.367
0.617
0.309
0.356
0.523
0.586
-0.212
0.661
0.169
0.322
0.408
0.304
0.199
0.128
0.014
0.220
-0.147
•0.321
0.524
0.173
0.400
0.200
0.207
0.254
-0.100
0.338
-0.032
-0.132

NS
NS
NS
.09
NS
<.01
NS
NS
.01
<.01
NS
<.01
NS
NS
.07
NS
NS
NS
NS
NS
NS
NS
.09
NS
NS
NS
NS
NS
NS
NS
NS
NS
Mean
Random Index
0.59 0.00
0.00
0.86
1.39
2.33
1.83
2.24
5.03
2.12
4.97
4.80
3.03
0.83
0.15
12.39
2.67
4.95
2.53
1.77
3.03
7.26
0.28
35.70
2.67
0.03
0.19
1.41
47.65
0.90
0.27
0.08
0.40
0.99
66.93
0.21
0.58
0.76
1.81
2.24
4.81
6.83
2.64
8.14
5.75
5.31
1.45
0.88
20.18
3.95
7.43
2.96
2.71
5.07
7.30
0.05
47.67
7.19
0.04
0.46
4.14
48.69
0.10
0.43
0.10
0.25
0.92
106.83
Table 6.7 (Cent)  Results of Index and random sampling comparisons for human use response variables In small
estuaries/small tidal rivers using ANOVA (p < 0.1 = no significant difference), Pearson and Spearman correlations (p < 0.1
no significant dlffero
Demonstration Report, EMAP-E Louisianian Province -1991
Page  112

-------
             100

              90

              80
           ซ>
           oป  __
           o  70

           S  60


           •  50
          I  40-

           !  30
          o
              20

              10-

              0-
               -2
                    — Lorgซ Riปซr  lndซx
                    " ' Urgซ Rivซr  Rondo*
T  -    I	!	1	1—

 6      8     10    12

    Stratification
                                                               14
16
—i—
 18
                                                                                 20
Figure 6.19 Cumulative distribution functions for degree of stratification for random and index sites In large tidal rivers.
100
90
80
*
o* 70
' 60
^ 50
1 30
o
20-
10-
0-



1
(\
II
D


/

f


2 "4 S 8
Total Organic Carbon






~~ Large Riปtr lndซx
~ " Largs Riปir Ra'ndon

10 12












14

Flgure 6.20 Cumulative distribution functions for total organic carbon for random and Index sites in large tidal rivers.
Demonstration Report, EMAP-E Louisianian Province - 1991
                                        Page 113

-------
Variable
STRATIFICATION
BOTTOM SALINITY
BOTTOM pH
BOTTOM
TEMPERATURE
SURFACE SALINITY
Anova Station Pearson
P- Value r . p
.07
.054
.36
.014
.049
.42 NS
.48 NS
.07 NS
.61 .06
.81 <.01
Speaiman
r P
.94 <.01
94 < 0 1
.37 NS
.65 .04
.77 .01
Mean
Random Index
2.54
2.76
9.57
30.43
0.22
3.32
4.07
9.65
30.42
0.75
Table 6.8 Results of Index and random sampling comparisons for habitat variables in large tidal rivers using ANOVA Pearson
and Spoarman correlations and random versus Index means (p < 0.1 = no significant difference).
Variable
STRATIFICATION
BOTTOM SALINITY
BOTTOM pH
BOTTOM TEMPERATURE
SURFACE SALINITY
Anova Station Pearson
P- Value r p
.02 .31 .051
<.01 .95 <.01
.14 16 NS
<.01 .60 <.01
<.01 .93 <.01
Spearman
r P
.31 .05
.92 <.01
.49 <.01
.68 <.01
.90 <.01
Random
1.79
15.23
8.05
29.93
12.82
Mean
Index
2.77
17.27
7.91
29.67
14.56
Table 6.9 Result* of Index and random sampling comparisons for habitat variables In small estuaries/small tidal rivers using
ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province -1991
                                                                                    Page 114

-------
              100


               90

               80


               70


               60


               50


               ซ<>

               30


               20 i

               10

                0
                 — Sill I  Eปt/Riv Index
                 - ' Siflll  Eซt/Riป Random
                                                                           10
                                                                                       11
                                             pH at the  Botton
Figure 6.21 Cumulative distribution functions for bottom pH carbon for random and index sites In small estuaries/small tida
rivers.                                               :
             100


              90


              80


          |  70
           C •
           o  60
           ซ

          :  ซ

          I  40

           |  30
          O

              20


              10
                -2
                —Small Eat/Riv  Index
                "•Snail Est/Riv  Random
	1	1	1	1	1	1

    8      10     12     14     16    18

 Strati ficatlon
                                                                                       20
Figure 6.22 Cumulative distribution functions for degree of stratification for random and index sites In small estuaries/small
tidal rivers.  "	"	;	""	'	"     	
Demonstration Report, EMAP-E Louisianian Province -1991
                                        Page 115

-------
    o
    L.
    m
    Ou
                                                         — Snail  Eit/RIv  Index
                                                         ~ -Snail  Eit/Rfv  Random
                                                     8
10
—i—

 12
—r

 H
                                      Total  Organic Carbon
Figure 6.23 Cumulative distribution functions for total organic carbon for random and Index sites In email estuaries/small tidal
rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
             Page 116

-------
 6.1.6  SEDIMENT CONTAMINANTS

 Sediment contaminant exposure indicators
 showed no differences in the population-
 level distributions of heavy metals, alkanes,
 PAHs, PCBs, and pesticides in large river
 segments. Cumulative distribution
 functions for mercury, total alkanes, total
 PAHs, total PCBs, and 4,4'-DDT are shown
 in Figures 6.24-6.28.  Thus, at a population
 level (class) for sediment contaminant
 variables, there are no differences in index
 and random sampling. ANOVA and
 correlation analyses showed differences
 between random and  index sites in large
 tidal rivers for all heavy metals (Table
 6.10), all alkanes (Table 6.11), all PAHs
 (Table 6.12), all PCBs except PCB
 congener #206 and #209 (Table 6.13), and
 all pesticides with measurable
 concentrations (Table 6.14).
 Concentrations of heavy metals  between
 random and index sites in large tidal rivers
 generally differed by < 2 ppm for all metals
 except chromium  (6 ppm), manganese (37
 ppm), and zinc (7 ppm).  Alkane
 concentrations between random and index
 sites differed by 1-143 ppb while mean
 concentrations of  PAHs between random
 and index sites in large tidal rivers differed
 by only 1-20 ppb. These differences
 represent statistical differences due to
 small variability and small sample size but •
 they do not represent significant ecological
variability.  Differences in PCB and
 pesticides concentrations between random
and index sites were < 1 ppb except for
total DDT (< 2ppb).

Sediment contaminant exposure indicators
showed no differences in the population-
level distributions of heavy metals, alkanes,
PAHs, PCBs, and pesticides in small
 estuaries/small tidal rivers. Cumulative
 distribution functions for mercury, total
 alkanes, total PAHs, total PCBs, and 4,4'-
 DDT are shown in Figures 6.29-6.33.
 thus, at a population level (class) for
 sediment contaminant variables, there are
 no differences in index and random
 sampling. ANOVA and correlation
 analyses showed no differences between
 random and index sites for heavy metals
 (Table 6.15) except for mercury whose
 mean concentration difference was < 0.02
 ppm showing low variability rather than
 significant ecological difference. Table
 6.16 shows no differences for individual
 alkanes except for C11 for which the mean
 difference was about 1 ppb; again not
 ecologically significant.  Unlike metals  and
 alkanes in small estuaries, several
 significant differences in PAH, PCB, and
 pesticide concentrations existed between
 random and index sites (18 of 40 PAHs, 17
 of 20 PCB congeners,  and 21 of 25
 pesticides) with random sites having the
 greater concentrations for PAHs (Table
 6.17) and no consistent pattern  of
 differences for PCBs (Table 6.18) and
 pesticides (Table 6.19). Although these
 differences are not large in terms of
 concentrations, significant variability exists
 within individual small estuaries for PAHs
 and PCBs. All mean concentrations, are
 low so  that the differences probably have
 little ecological significance.  Because of
the significant statistical variability in many
 PAH and PCB concentrations, multiple
samples within an estuary would be
required to characterize an individual
estuary but random samples appear
adequate to represent the population
distribution of PAHs and PCBs in small
estuaries.
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 117

-------
   100

    90

    80


J"  70
c=
S   60
*ป
n.
ซP

•5   40-


o
    20-

    10-
                                                              Lorgt Rivซr Indix
                                                              Large River Rondoi
              D.O
                               0.5
                                               1.0
                                           Mercury (ppn)
                                                     —i—
                                                      1.5
	r
 2.0
 tidal riven.
         Cumulative distribution functions for sediment concentrations of (a) mercury for random and index sites in laraf


01
o
e
o
a.
^
o


100
90
80
70
80-
50-
40-
30'
20-
10-
0-

il
i J
i /
i /
il
il
\ I
1 1
1 1
i 1
i 1
i 1
il
i /
il
t /
i 1
i I
t /






— Largซ Rlปซr Indsx
— ' Larg* River Random

i i i i i i i i 	 1 	 1 	
1 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
Totfll Alkanea












22000

Figure 6.25 cumulative distribution functions for sediment concentrations of total alkanes for random and Index sites in larae
tidal river*.
Demonstration Report, EMAP-E Louisianian Province -  1991
                                                                     Page 118

-------
            100

             90

             80

             "

             60-
             30-

             20

             10-

              0-
— Large River  lndซx
~ ' Large River  Randon
                       1000      2000      3000      4000

                                            Total  PAH
  5000
6000
                     7000
 Igure 6.26 Cumulative distribution functions for sediment concentrations of total PAHe for random and Index sites In large
tidal rivers.

100
90-
80

ซ>
5 70-
e
S 80-
^
ซ 50-

*
i 40-

ง 30
O
20-
10-

0-
i ^r
i /
i /
> f
' /
' /
it
if
il
il
i 1
t I
il
i 1
i 1
1
I
t
1


0 10 20 30 40 50 60 70 80
Totol PCB


















~r~ Large River Index
— • Large River Randan

	 1 	 1 	 1 	 1 —
90 100 110 120
































~~~









	 r
130

Figure 6.27 Cumulative distribution functions for sediment concentrations of total PCBs for random and Index sites in large
Hdal rivers. 	 	 	 	 - -
Demonstration Report, EMAP-ELouisianian Province- 1991
                     Page 119

-------
                                                               ~~ Urge RJver  Index
                                                               ~~ Urge Rfver Random
                                         -i	1	1	r-
                                          5678
                                             4.4-DDT (ppb)
10
11
12
—r
 13
 Figure 6.28 Cumulative distribution functions for sediment concentrations of 4,4'-DDT for random and Index sites In larae
 tidal river*.

Variable
Silver
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Manganese
Nickel
Lead
Antimony
Selenium
Tm
Zinc
Anova Station Pearson
P-Value
0.77
0.37
0.14
0.06
0.14
0.17
0.60
0.10
0.18
0.09
0.33
0.20
0.03
0.11
r
-0.18
0.08
0.37
0.49
0.33
0.29
-0.15
0.39
0.28
0.43
0.12
0.26
0.57
0.38
P
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Spearman
r
-0.23
0.07
0.15
0.45
0.45
0.19
-0.40
0.16
0.16
0.35
0.15
0.23
0.59
0.47
P
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Random
0.17
5.10
6.75
0.31
58.50
13.43
0.26
771.10
22.73
18.64
0.91
0.31
1.87
78.10
Mean
Index
0.24
5.31
7.73
0.35
52.60
14.86
0.33
733.70
24.83
19.04
0.95
.380
2.11
85.50
Table 6.10 Results of Index and random sampling comparisons for sediment heavy metals concentrations In large tidal rivers
using ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Demonstration Report,  EMAP-E Louisianian Province -1991
           Page 120

-------

Variable
C10
C11
C12
C13
C14
C15
C16
C17
PrisTane
C18
Phytane
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
Alkanes
Anova Station Pearson
P-Value
0.91
0.17
0.16
0.21
0.16
0.33
0.36
0.22
0.45
0.48
0.32
0.42
0.50
0.09
0.052
0.28
0.29
0.50
0.46
0.96
0.90
0.95
0.47
0.63
0.64
0.77
0.57
0.74
r
-0.66
0.35
0.35
0.28
0.33
0..19
0.06
0.28
0.05
-0.02
0.13
0.04
-0.07
0.41
0.47
0.19
0.17
0.03
0.05
-0.63
-0.67
-0.63
0.04
-0.07
-0.08
-0.23
-0.07
-0.24
P
.04
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
.05
.03
.05
NS
NS
NS
NS
NS
NS
Spearman
r
-0.64
0.44
0.48
0.32
0.31
0.42
0.18
0.22
0.19
-0.05
0.16
0.21
0.18
0.19
0.16
0.07
0.03
o:os
-0.04
-0.50
-0.39
-0.48
0.13
-0.15
ฃ.22""
-0.41
-0.18
-0.08
P
.04
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Random
20.08
17.26
20.78
19.71
24.18
23.24
19.81
35.74
37.9
17.6
26.5
22.8
36.6
37.5
34.7
50.5
39.8
79.5
40.2
165.3
63.3
365.2
47.3
289.4
56.2
149.8
15.2
108
Mean
Index
15.17
22.89
25.27
22.13
29.45
34.11
21.61
52.28
56.8
22.5
34.1
28.9
35.4
42.2
37.7
37.7
47.7
105.4
52.2
177.4
57.9
387.0
63.6
432.1
37.3
88.4
10.0
168
               f Index and random sampling comparisons for sediment alkane concentrations In large tidal rivers using
ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province -1991
Page 121

-------
Anova Station Pearson
Variable
Acenaphlhene
Aconaphthylene
Anthracene
Bonzo(a)anthracene
Benzo{a)pyrone
Benzo(b)fluoranthena
Benzo(e)pyrene
Bonzo(k)rkioranthone
Bonzo(f.h,i)poryleno
Biphonyl
C1-chrysene
C2-chrysene
C3-chrysene
C4-chrysene
Cl-dbonzothio
C2-dibenzothio
Ca-dbonzothio
C1 -fluoranthpyrene
Cl-ftuorene
C2-fluorene
C3-fluorene
C1 -naphthalene
C2-naphthalene
C3 -naphthalene
C4-naphthaIene
Cl-phenanthrene
C2-phonanthrene
C3-phonanthrene
C4-phenanthrene
Chrysone
Dibonzo(a,h)anlhracene
Dibonzothio
2,6-cfimothylnaphthaJene
Ruorena
FHioranthene
(i)1.2,3"C,d-pyrene
2.3,5-trimethylnaphthalene
1 -methylnaphthalene
2-mothylnaphthalen9
1 -mothylphonanthrene
TOT PAHS
P-Value
0.38
0.17
0.53
0.47
0.46
0.37
0.37
0.23
0.40
0.37
0.43
0.36
0.44
0.26
0.53
0.51
0.37
0.48
0.38
0.57
0.54
0.32
0.37
0.45
0.50
0.56
0.52
0.40
0.27
0.47
0.50
0.37
0.44 '
0.26
0.27
0.48
0.49
0.29
0.34
0.48
.41
r
0.20
0.33
0.04
0.10
0.10
0.16
0.15
0.36
0.13
0.17
0.14
0.19
0.16
0.28
0.04
0.06
0.21
0.04
0.13
0.01
0.02
0.21
0.12
0.15
0.11
-0.05
0.05
0.19
0.32
0.08
0.09
0.22
0.08
0.32
0.31
0.08
0.12
0.23
0.19
0.12
.16
P
NS
NS
NS
NS
NS
NS.
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Spearman
r
0.03
-0.10
0.13
0.04
0.08
0.30
0.19
0.38
0.08
0.30
0.07
0.02
0.03
0.13
0.02
0.28
0.44
0.13
0.26
-0.07
0.04
0.33
0.10
0.02
0.15
0.13
0.26
0.28
0.58
0.24
0.18
0.01
0.20
0.14
0.24
0.18
0.01
0.36
0.24
0.28
.20
P
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Mean
Random
2.45
2.03
5.12
17.05
19.21
22.81
17.232
14.15
14.24
3.39
22.96
21.89
9.43
8.87
4.84
12.35
16.16
34.53
4.19
9.07
13.90
22.46
23.64
24.81
18.68
25.65
31.79
31.72
24.82
24.81
3.68
1.68
8.78
3.64
30.36
12.89
5.76
9.39
13.07
4.58
731.05
Index
4.24
2.81
11.68
30.37
31.60
36.25
26.16
24.23
21.55
4.79
38.66
34.58
16.43
13.88
7.55
19.75
25.65
49.01
6.85
13.94
21.08
31.21
31.74
36.80
27.68
36.58
51.03
50.67 ,
37.75
41.91
6.53
2.72
12.26
6.39 ,
51.13
13.87
8.92
12.97
18.24
8.36
1139.00
Table 6.12 Results of Index and random sampling comparisons for sediment PAH concentrations In large tidal rivers using
ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 122

-------
Anova Station Pearson
Variable
PCB 8
PCB 18
PCB 28
PCB 44
PCB 52
PCB 66
PCB 101
PCB 105
PCB 110
PCB 118
PCB 126
PCB 128
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
PCB 209
TOTAL PCBs
P-Value
0.495
0.542
0.277
0.086
0.479
0.216
0.235
0.716
0.602
0.418
0.641
0.299
0.476
0.159
0.418
0.344
0.134
0.682
0.039
0.006
0.294
r


0.514
0.837
-0.001
0.326
0.394
0.178
-0.034
0.091
-0.167
0.133
-0.010
0.757
0.113
0.157
0.451
-0.199
0.594
0.792
0.328
P


NS
<01
NS
NS
NS
NS
NS
NS
NS
NS
NS
.01
NS
NS
NS
NS
.07
<.01
NS
; Spearman
r


0.536
0.861
0.292
0.437
0.337
0.184
-0.215
0.168
; -0.1 67
0.131
0.012
0.685
0.000
0.107
0.179
0.048
0.596
0.750
0.280
P


NS
<.01
NS
NS
NS
NS
NS
NS
NS
NS
NS
.02
NS
NS
NS
NS
.06
.01
NS
Moan
Random
0.0
0.0
0.05
0.07
0.50
0.14
0.18
0.01
0.81
0.14
0.01
0.11
0.68
0.20
0.12
0.33
0.12
0.02
0.04
0.04
20.66
Index
0.02
0.01
0.17
0.18
0.71
0.25
0.33
0.15
1.43
0.22
0.00
0.13
0.80
0.59
0.27
0.47
0:21
0.01
0.03
0.08
25.34
Table 6.13 Results of Index and random sampling comparisons for sediment PCB concentrations In large tidal rivers using
ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Anova Station Pearson
Variable
Aldrin
Alpha BHC
Alpha Chlordane
Beta BHC
Cis-nonachlor
2,4'-DDD
4,4'-DDD
Delta BHC
2,4'-DDE
4,4'-DDE
2,4'-DDT
4,4'-DDT
Dieldrin
Endosulfan 1
Endosulfan 2
Endrin
Gamma BHC
Gamma Chlordane
HCB
Heptaohlor Epoxide
Heptachlor
Mirex
Oxychlordane
Total BHC
Total Chlordane
Total DDT
Toxaphene
Transnonachlor
P-Value
0.095
0.148
0.620
0.711
0.095
0.578
0.411

0.495
0.454
0.353
0.548
0.610


0.495
0.344
0.507
0.756
0.366


0.517
0.400
0.272
0.411

0.674
r
0.382
0.325
-0.078
-0.242
0.444
0.024
0.069


0.027
0.092
0.016
-0.104



0.125
0.016
-0.274
0.816


-0.111
0.029
0.278
0.089

-0.174
P
NS
NS
NS
NS
NS
NS
NS


NS
NS
NS
NS



NS
NS
NS
<.01


NS
NS
NS
NS

NS
Spearman
r
0.383
0.364
-0.201
-0.248
0.727
0.047
0.042


0.103
0.042
0.310
-0.018

'

; 0.028
-0.144
-0.249
0.581 "


-0.111
0.219
0.253
0.152

; -0.236
P
NS
NS
NS
NS
.01
NS
NS


NS
- NS
NS
NS



NS
NS
NS
.07


NS
NS
NS
NS

NS
Mean
Random
0.01
0.05
0.11
0.03
0.07
0.07
1.41
0.00
0.05
0.90
0.02
0.48
0.38
0.00
0.00
0.00
0.01
0.17
0.38
0.00
0.00
0.00
0.00
0.09
0.47
2.96
0.00
0.20
Index
0.02
0.03
0.17
0.04
0.07
0.12
1.97
0.00
0.00
1,18
0.03
0.83
0.55
0.00
0.00
0.02
0.02
0.26
0.46
0.07
0.00
0.00
0.00
0.10
0.73
4.15
0.00
0.14
Table 6.14 Results of Index and random sampling comparisons
using ANOVA, Pearson and Spearman correlations and random
for sediment pesticide concentrations In large tidal rivers
versus Index means (p < 0.1 •- no significant difference).
Demonstration Report, EMAP-E Louisianian Province -1991
                                 Page 123

-------
   100

    90

    80
•
O*  —ft
o   70
c
o   60
ซ
O.
•

^   *(H


I   30
    20-

    10-
     0.0
                                                                Sun If Eit/Riv Indix
                                                                Snot 1 Est/Rivi Rondo*
                                 O.S
     1.0

Mercury (ppn)
                                                          1.5
                                                                                     2.0
Figure 6.29 Cumulative distribution functions for sediment concentrations of mercury for random and Index sites In small
estuaries/small tidal rivers.
                                                           — Snail  Eซt/Rlv lndซx
                                                           — •Snail  Est/Riv Randon
              OH
                     2000   4000  6000   8000   10000  12000  14000  16000  18000  20000  22000
                                             Total Alkanfls
Rguro 6.30 Cumulative distribution functions for sediment concentrations of total alkanes for random and Index sites In
small estuaries/small tidal rivers.
Demonstration Report, EMAP-E Louisianian Province -1991
                                                                          Page 124

-------

10U"
90
80
f 70
' 60
: 50-
"•i 4o-
1 30-
o
20-
10-
0-
•


^— ^
^/— --/
fi
1
J
ij
y
i
a
]
f
0 1000 2000 3000 4000
Total PAH










— Snail E*t/R1v Index
— •Snail Ett/Riv Random

i — ,_ 	 ji
5000 6000 70(













)0

 Figure 6.31 Cumulative distribution functions for sediment concentrations of total PAHs for random and Index sites In small
 estuaries/small tidal rivers.

1UU-
90
80
ฃ 70-
S 80-
: so
i 40-
1 30-
CJ
20-
10-
0-














0 10





i







t

'
fT
i*




i




10 30 40 50 60 70 8
Total PCB










— Snail Eซt/Rlv lndซx
.— 'Snail Est/Riv Randan

0 90 100 110 120













13













0

Figure 6.32 Cumulative distribution functions for sediment concentrations of total PCBs for random and Index sites In small
estuaries/small tidal rivers.               	         —	 		
Demonstration Report, EMAP-E Louisianian Province -1991
Page 125

-------
        100

         so

         80
     •

     j*   7tt
     IE
     ฃ   6Q



     >•
     =   40
     0
20

10

 OH
                                                             — Snoll Et.t/Riซ  Index
                                                             -- Snail Eit/Riv  Randon
               23456789

                                    4,4-DDT (ppb)
                                                                       10    11    12     13
Rgura 6.33 Cumulative distribution functions for sediment concentrations of 4,4'-DDT for random and Index sites In small
e*tuarie*/smali tidal rivers.
Anova Station Pearson
Variable
Silver
Aluminum
Arsenic
Cadmium
Chromium
Copper
Iron
Hg
Manganese
Nickel
Lead
Antimony
Selenium
Tin
Zinc
P-Value
.012
.060
.001
.080
.001
.001
.001
.660
.001
.001
.005
.010
.030
.001
.001
r
0.39
0.57
0.32
0.28
0.49
0.64
0.52
-0.07
0.46
0.48
0.36
0.38
0.31
0.53
0.49
P
.01
<.01
.04
NS
<.01
<.01
<.01
NS
<.01
<.01
.02
.01
.04
<.01

Spearman Mean
r
0.55
0.59
0.48
0.53
0.59
0.61
0.57
0.31
0.64
0.53
0.56
0.51
0.35
0.51
0.51
p Random
<.01 .14
<.01 4.06
<.01 5.35
<.01 .19
<.01 46.56
<.01 10.42
<.01 1.82
.04 .11
<.01 358.2
<01 14.6
<.01 16.7
<.01 .66
.02 .29
<.01 1.4
<.01 60.4
Index
.12
3.80
6.30
.14
40.76
9.13
1.76
.09
352.6
13.8
13.8
.54
.29
1.2
51.6
Table 6.15 Results of Index and random sampling comparisons for sediment heavy metals concentrations In small estuaries
using ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
Demonstration Report, EMAP-E Louisianian Province - 1991
                                                                          Page 126

-------
Anova Station Pearson
Variable
C10
C11
C12
C13
C14
C15
C16
C17
Pristane
C18
Phyjane
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
Alkanes
P-Value
.001
.280
.001
.010
.010
.001
.002
.001
.001
.001
.001
.001
.001
.030
.010
.001
.003
.001
.001
.001
.001
.002
.001
.001
.001
.002
.001
.001
r
0.51
0.49
0.09
0.33
0.43
0.47
0.70
0.53
0.63
0.60
0.57
0.58
0.55
0.36
0.29
0.64
0.40
0.59
0.47
0.71
0.78
0.91
0.43
0.65
0.47
0.69
0.40
0.67
P
.001
.001
NS
.030
.004
.001
.001
.001
.001
.001
.001
.001
.001
.020
NS
.001
.010
.001
.001
.001
.001
.001
.003
.001
.001
.001
.010
.001
Spearman
: r
0.36
0.61
0.49
0.60
0.67
0.54
0.81
0.65
0.67
0.71
0.79
0.75
0.71
0.46
0.51
0.54
0.63
0.51
0.52
0.53
0.53
0.55
0.47
0.54
0.47
0.53
0.44
0.54
P
.020
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.002
.002
.001
.001
.001
.001
.001
.001
.001
.001
.010
.001
.001
.003
.001
Mean
Random
8.28
3.08
9.01
5.62
15.76
50.61
58.69
216.80
202.8
70.6
200.6
103.0
60.0
79.0
28.4
75.3
47.8
126.1
52.3
234.8
90.4
611.8
96.5
402.0
67.0
149.5
15.9
233
Index
7.55
6.97
6.49
4.47
11.08
35.92
38.53
158.48
129.0
48.1
127.2
70.4
47.4
56.9
22.0
65.6
36.5
108.1
40.3
182.5
66.5
430.0
69.7
316.8
93.8
199.8
19.1
317
        Results of Index and random sampling comparisons for sediment alkane concentrations In small estuaries using
ANOVA, Pearson and Spearman correlations and random versus index means (p < 0.1 = no significant difference).
6.2  EFFECTS OF GRID DENSITY
   OF PARAMETER ESTIMATION
   IN LARGE ESTUARIES

Although some historical sediment
contaminants data (O'Connor 1990) was
examined to estimate the sample sizes
necessary to estimate contaminant
concentrations with program objectives, is
was unknown whether this sample size
would be relevant to large geographic
areas.  As in the Virginian Province
(Weisberg et al. 1992), the necessary
sample size corresponded to a systematic
grid with a density creating 280 km2
sampling spaces for the Louisianian
Province.  In order the test the
appropriateness of this  spatial scale .for the
systematic grid,  sampling was conducted
at a grid scale four times denser in Mobile
Bay, AL for all indicators.  This
supplemental sampling  data can be used
to address two questions:

• Would sampling at this increased scale
  improve the estimates of the sampled
  indicators  in terms of accuracy or
  precision for the large estuary class or
  the Louisianian Province?
Demonstration Report, EMAP-E Louisianian Province - 1991
                           Page 127

-------
Anova Station Pearson
Variable
Aconaphlhone
Acanaphthylene
Anthracene
Benzo{a)anthracono
Benzo(a)pyreno
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(k)fluoranthene
B0nzo(f,h.i)perylene
Biphenyl
C1-chrysene
C2-chrysene
C3-chrysone
C4-chrysene
C1
-------
Anova Station Peareon
Variable
PCB 8
PCB18
PCB28
PCB44
PCB52
PCB66
PCB 101
PCB 105
PCB 110
PCB 118
PCB 126
PCB 128
PCB 138
PCB 153
PCB 170
PCB 180
PCB 187
PCB 195
PCB 206
PCB 209
PR>F
0.002
0.088
0.167
0.643
0.015
0.517
0.329
0.401
0.190
0.381!
0.000
0.240
0.446
0.412
0.259
0.389
0.335
0.400
0.152
0.212
r
0.400
0.115
0.460
-0.104
0.322
-0.018
0.152
0.134
0.190
0.141
0.589
0.038
0.321
0.121
0.209
0.150
0.153
0.158
0.697
0.595
P
.0071
NS
.0017
NS
.0328
NS
NS
NS
NS
NS
.0001
NS
.0337
NS
NS
NS
NS
NS
.0001
.0001
Spearman
r
0.430
0.267
0.496
. -0.062
0.432
0.241
0.554
0.471
0.356
0.440
0.337
0.271
0.444
: 0.500
0.426
1 0.633
0.570
0.137
0.180
0.500
P
.0035
.0795
.0006
NS
.0034
NS
.0001
.0013
.0177
.0028
.0253
.0750
.0025
.0005
.0086
.0001
.0001
NS
NS
.0005
Mean
Random
0.09
0.02
0.15
0.04
0.24
0.07
0.32
0.11
0.48
0.20
0.02
0.11
0.75
0.56
0.51
0.33
0.25
0.05
0.08
0.24
Index
0.08
0.02
0.03
0.02
0.15
0.02
0.08
0.03
0.19
0.05
0.25
0.02
0.31
0.09
0.22
0.07
0.04
0.01
0.01
0.03
 • uuia w. iป naouiio ui mum ซnu ituiuuiii v*ปii|jiiiig comparisons ror sediment rue concentrations in small estuaries
 ANOVA, Pearson and Spearman correlations and random versus Index means (p < 0.1 = no significant difference).
For the first question, sampling for the
Mobile Bay increases the sample size for
the large estuarine class from 56 to 69
(+23%).  If the supplements are better
characterizing the indicators within the
large estuarine class, an increase of 23%
in the sample size should result in a
cumulative distribution function that is
significantly different than the CDF based
only on the large estuarine base samples.
For the second question, the increase in
sample size from 3 to 13 (+333%) with the
supplemental  samples should result in
significantly different CDFs if the
supplements better characterize the EMAP-
E indicators.

Although all indicators were examined with
regard to these questions, only selected
indicators of each indicator type are
discussed here; namely, benthic
biodiversity (biological response), water
clarity (human use response), minimum
dissolved oxygen (water quality), total
organic carbon (habitat), and total DDT
(exposure).  Comparison of province-wide
CDFs for benthic biodiversity (Fig. 6.34),
light transmittance to 1  m (Fig. 6.35),
bottom minimum dissolved oxygen
concentrations (Fig. 6.36), TOG (Fig. 6.37),
and total sediment DDT concentration (Fig.
6.38) show that the addition  of the
supplemental sites do not significantly alter
the province-wide distributions of these
indicators.  Unlike the province-wide
distributions, the addition of the
supplemental samples does  significantly
alter the indicator CDFs observed for
Mobile Bay (Figures 6.39-43).  Thus, the
present spatial scale for the systematic grid
Demonstration Report, EMAP-E Louisianian Province - 1991
                             Page 129

-------
Anova Station Pearson
Variable
Aldrin
Alpha BHC
Alpha Chlordano
Bola BHC
Cis-nonachlor
2,4'-DDD
4.4'-DDD
Delta BHC
2,4'-DDE
4,4'-DDE
2.4--DDT
4.4'-DDT
Dieldrin
Endosulfan 1
Endosulfan 2
Endrin
Gamma BHC
Gamma Chlordane
HCB
Hoptachlor Epoxide
Hoptachlor
Mirox
Oxychlor
Total BHC
Total Chlordane
Total DDT
Toxaphene
Transnonachlor
PR>F
0.383
0.008
0.079
0.575
0.042
0.071
0.561
0.005
0.174
0.472
0.058
0.428
0.396


0.293
0.037
0.000
0.288
0.253
0.438
0.468
0.438
0.092
0.221
0.319

0.052
r
0.017
0.186
0.336
-0.067
0.393
-0.026
0.221
0.354
0.202
0.237
-0.024
0.045
0.028


0.312
0.290
0.212
0.974
0.516

-0.023

0.172
0.247
0.054

0.489
P
NS
NS
.0257
NS
.0084
NS
NS
.0186
NS
NS
NS
NS
NS


.0391
NS
NS
.0001
.0004

NS

NS
NS
NS

.0008
Spearman
r
0.262
0.469
0.512
-0.073
0.497
0.215
0.588
0.383
0.588
0.472
0.220
0.346
0.433


0.476
0.310
0.221
0.652
0.200

-0.023

0.381
0.526
0.419

0.339
P
.0854
.0013
.0004
NS
.0006
NS
.0001
.0103
.0001
.0012
NS
.0213
.0033


.0011
.0409
NS
.0001
NS

NS

.0108
.0002
.0046

.0246
Mean
Random
0.00
0.01
0.06
0.01
0.04
0.03
0,17
0.00
0.02
0.52
0.00
0.04
0.04
0.00
0.00
0.00
0.00
0.13
0.73
0.11
0.00
0.00
0.00
0.03
0.42
0.79
0.00
0.05
Index
0.00
0.01
0.04
0.00
0.01
0.01
0.19
0.00
0.07
0.36
0.00
0.35
0.08
0.00
0.00
0.03
0.00
0.03
0.39
0.01
0.00
0.00
0.00
0.03
0.12
0.99
0.00
0.01
Table 6.19  Results of Index and random sampling comparisons for sediment pesticide concentrations in small estuaries
using ANOVA, Pearson and Spearman correlations and random versus index means (p < 0.1 = no significant difference).
corresponding to sampling spaces of 280
km2 represents the indicator as well as the
reduced 70 km2 sampling spaces for
province-wide estimates (as well as for
large estuary distributions which comprise
75% of the province). However, if the
objective is to characterize a specific
estuary, the 280 km2 is inadequate to
characterize estuaries the size of Mobile
Bay (about 600 knrr).
Demonstration Report, EMAP-E Louisianian Province -1991
                             Page 130

-------
6.3  DEGREE OF SPATIAL
   AUTOCORRELATION FOR SITES
   SELECTED BASED ON THE GRID

During the 1991 Louisianian Province
Demonstration, multiple stations were often
sampled within large estuaries (i.e., > 250
krrr)(e.g., Apalachee Bay, Galveston Bay,
Lake Pontchartrain). Because the initial
sampling spaces (280 km2 hexagons) were
organized based on the random placement
of a systematic sampling grid, some spatial
relationship among proximally located
stations might exist.  The CDF for the
distances among proximal stations is
shown in Figure 6.44 and illustrates that
distances range from 6 km to 39 km.
Because sampling points in the Louisianian
Province were not equidistant, a spatial
                           autocorrelation analysis was completed to
                           determine if an interdependence of
                           parameter values existed due to proximity.
                           The analysis uses autocorrelation statistics
                           that are basically descriptive in nature and
                           can be used in tests of hypotheses.  The
                           spatial autocorrelation statistics are
                           functions of both the data values and a
                           weighing function which assigns values to
                           pairs of sites to represent their geographic
                           arrangement. The choice of weighing
                           function determines the hypothesis tested
                           and for one analysis presented here,
                           examination of several tests simultaneously
                           is used to ascertain the maximum distance-
                           of influence for selected parameters. The
                           statistic employed for this spatial analysis
                           is Moran's I (Odland 1988).

                           Most pertinent indicators were examined
                           using a weighing function equal to the
                                                     •—• Large Estuary
                                                     — • Large Estuary + Suppl
             0.0
C.2
0.4
                   -i	r
0.6    0.8    1.0    1.2    1.4
 Shannon-Wiener Diversity Index
                                                               1.8
Figure 6.34 Cumulative distribution functions for benthic biodiversity Incorporating supplemental sampling for Louisianian
Province.
Demonstration Report, EMAP-E Louisianian Province -1991
                                                       Page 131

-------
100-
90-


SO-
a
? 70-
"*E
•k
S 60
*~
*
^ 50
| 40
I 30-
o

20-
10-
0-














S'
t''
i
i
f
s

0.0 0.2 0.4 0.6 0.8 1.0



















—~ Large Estuary Random
~ • Large Estuary Supplenent


















„


1.2 1.4 1.6 1.8 2.0
Shannon-Wiener Diversity Index
Figure 6.35 Cumulative distribution functions for benthlc biodiversity Incorporating supplemental sampling for Mobile Bay,
   100



    90



    80



    70!



|   60





O

ฃ   40
a.


    30



    20



    10



    0
                                                           ~~ Large Estuary
                                                           "'Large Estuary + Suppl
5678

Vininun  D.O. (ppn)
10
                                                                 11
                                                                                12    13
Figure 6.36 Cumulative distribution functions for minimum dissolved oxygen concentrations Incorporating supplemental

sampling for Loulslanlan Province.
Demonstration Report. EMAP-E Louisianian Province - 1991
                                                                          Page 132

-------
              100


              90


              80


              70


           I  Sฐ

           •ฃ  50
           O
           ฃ  40
           Q.

              30


              20


              10-


               0-
~~ Large Estuary
~ 'Lorgs Estuary + Suppl
                                               3         4
                                         Total Organic Carbon (ซ)
 Figure 6.37 Cumulative distribution functions for total organic carbon In sediments Incorporating supplemental sampllnc
 for Loulslanlan Province.                                                                         K.
m.
90-
80
70
• 60
•ฃ 50-
0
5 40-
Q-
30-
20
10
0


	 	 	 : 	 : 	 : 	 ^—. 	 : 	 : 	 = — ^— = 	
/*^~^
I'
\
I
I





D 5 10 U
Total DDT (ppb)









~~ Large Estuary
"'Large Estuary + Suppl

20 25














Figure 6.38 Cumulative distribution functions for total DOT concentrations in sediments Incorporating supplemental sampling
for Louislanlan Province.    	  -	-	t— "  "  	"	 ~	
Demonstration Report, EMAP-E Louisianian Province - 1991
                        Page  133

-------
                                                             - Larga Estuary
                                                             • Large Estuary + Suppl
               0.00
                          0.15
                                                                                    0.90
 Figure 6.39 Cumulative distribution functions for percentage of surface light reaching a depth of 1 m incorporating
 supplemental sampling for Loulslanlan Province.
   100

    90

    80
•
01  -A
^  70

I  80


t  50

i  ซ•

ง  30
o
    20

    10-

    0-
              0.20
                                                             — Alflbana: Lorgs Est.
                                                             — - Alabama: Supplinant
                0.25
0'.30
0.35

PAR
                                                             0.40
                                                               0.45
                                                                                    0.50
Figure 6.40  Cumulative distribution functions for percentage of surface light reaching a depth of 1 m incorporating
•upplomantal sampling for Mobile Bay, AL
Demonstration Report, EMAP-E Louisianian Province -1991
                                                                           Page 134

-------
100
90
80
5 70
5 60
| 30
o
20-
10-
0-


^
^*
~~~~~~ /
_...-•-" /
I
J
;
1 1 2 34 5

~**'S^
' ^T





— Lorjซ Ettuary Rondos
'•Large Estuarjr Sypplentnt








i i 	 1 	 1 	 f
8 7 8 9 10
Uininun D.O. (ppn)
Figure 6.41 Cumulative distribution functions for minimum dissolved oxygen concentrations Incorporating supplemental
sampling for Mobile Bay, AL.
100
90
80
*

S 60-
5 50
ป
= 40-
| 30-
o
20-
10-
0-

y ,,-•—-
/ /
/ i
/ t
s •
i
i
i
i
ป
i •
~'~~
ป i 23
Total Organic Carbon

._- — ^--— -."






— Alobana: Large Ett.
- ~ Alabama: Supplement

i 	 f
4 5








Figure 6.42 Cumulative distribution functions for total organic carbon In sediments incorporating supplemental sampling
for Mobile Bay, AL.
Demonstration Report, EMAP-E Louisianian Province -1991
Page 135

-------
100
90.-
80-
; ™ 70-
; | 6tt-
f 50,
j-> ifQ-
n
•S JJQ. •
C3
2Q-
10-
0-

/ ^ 	 	 	 	 "
/ *•'
f f
•.ft.
f e
/ f
ft
I t-
/••
i
r
t
e
s
f
t r^—^— —
/ ~~Arobflmo: Larj,e Eat.,
"-• Alabama: Supplement








T~ •• i.i. i 	 	 : 	 T
Q 5 10 15 20 25
Totol DDT
Rgure 6.43 Cumulative distribution functions for total DDT concentrations In sediments Incorporating supplemental sampling
for Mobile Bay.
inverse of the distance in kilometers
between the two random base sites. This
weighing function allows every random site
to impact on all other random  sites unless
a maximal  influence distance is defined.
For example, if site A is twice  the distance
to a site C  as site B, the farther site will
have  1/2 the spatial impact of the closer
site.  The hypothesis tested is, "The values
are randomly distributed geographically."
versus the  alternative, '"Nearer sites have
related values.". Three sets of spatial
autocorrelative analyses were  completed:

•  Analysis examined the interdependency
   of all random base sites regardless of
   estuarine class,
Analysis examining the interdependency
of all sites within the large estuarine
class (sites randomly located within
hexagonal sampling space) with the
assumption of no dependency existing
after 20, 25, or 30 km except for bottom
dissolved oxygen, bottom temperature,
and degree of stratification which also
included distances of 10 and 15 km,

Analysis of five selected water bodies in
the large estuarine class within the
Louisianian Province (i.e., Galveston
Bay, Lake Pontchartrain, Mobile Bay,
Chandeleur Sound, and Mississippi
Sound) to determine  if patterns seen
overall for the province were evident in
individual systems.
Demonstration Report, EMAP-E Louisianian Province -1991
                          Page 136

-------
             100
              90
              BO
          f  70
          e
          3  to
          !ซ
          >>
          |  40
          I  30
          o
              20
              10
              0
                                10       15       20      25       30
                                     Hid  Distune*  Bitaocn  LR Sitti
35
        40
Figure 6.44 Cumulative distribution function for distance (km) between probability-based sampling sites for large estuaries
(LR) In the Loulslanlan Province In 1991.
Demonstration Report, EMAP-E Louisianian Province -1991
      Paae 137

-------
 6.3.1 SPATIAL  .
   AUTOCORRELATION AMONG
   ALL PROBABILITY-BASED
   SITES

 The results of the spatial autocorrelation
 analysis among all probability-based sites
 regardless of estuarine class in shown in
 Table 6.20. Several benthic response
 indicators showed significant spatial
 autocorrelation including: mean benthic
 abundance, mean number of benthic
 species, and percent of total abundance of
 polychaetes and amphipods.  However, the
 benthic index showed no spatial .
 autocorrelation.  This lack of spatial
 dependency may be the result of the
 removal of the effects of salinity on
 biodiversity, a statistic related to
 abundance and  number of species (both
 spatially dependent).  No spatial
 dependency was seen in  the mean
 abundance of fish per trawl possibly
 confirming  the effects of finfish mobility.
 Regardless, the  results with several
 response indicators (other than the
 calculated indices) suggest that spatial
 autocorrelation will have to be adjusted for
 in any long-term analyses.

 As with response indicators, several habitat
 indicators portrayed spatial autocorrelation
 in the Louisianian Province.  Water clarity,
 bottom salinity, and  bottom temperature
were all significantly spatially dependent.
Sediment contaminant exposure indicators
exhibited a heavy spatial dependency.
Eighty five percent of alkanes, 87% of
heavy metals, 80% of PAHs, and 70%  of
pesticides showed significant spatial
dependency within the Louisianian
Indicators
Response Indicators
Benthic Abundance
Benthic Species
Percent Polychaetes
Percent Amphipods
Benthic Index
. Fish Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom 'Salinity
Stratification
Bottom Temperature
Exposure Indicators
Alkanes
Total Alkanes
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
C28
C29
C30
C31
C32
C33
C34
Phytane
Pristane
Moran's IP-value

0.055
0.110
0.102
0.043
0.027
0.009

0.001
0.104
0.249
0.017
0.042


0.037
0.068
-0.031
0.020
0.046
0.058
0.018
0.069
0.054
0.074
0.084
0.104
0.074
0.061
0.071
0.049
0.073
0.088
0.054
0.039
0.020
0.059
0.078
0.069
0.078
0.065
0.076
0.068

0.0042*
<0.0001*

-------
Indicators
Exposure Indicators
Pesticides
Dieldrin
2,4'-DDT
4,4'-DDT
Heavy Metals
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Tin
Zinc
PAHS
Total PAHs
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Biphenyl
Moran's 1


0.037
0.001
0.061

0.064
0.139
-0.026
0.068
0.090
0.113
0.094
0.090
0.120
-0.007
0.127
0.040
0.033
0.113
0.093

0.085
0.047
0.030
0.041
0.028
0.044
0.023
0.033
0.045
0.082
0.041
P-value


0.0270*
0.3261
0.0019*

0.0014*
<0.0001*
0.7463
0.0009*
•dO.0001*
<0.0001*
<0.0001*
<0.0001*
<0.0001*
0.4537
<0.0001*
0.0213*
0.0411*
<0.0001*
^0.0001*
.
0.0001*
0.0097*
0.0535
0.0199*
0.0619
0.0143*
0.0910
0.0388*
0.0127*
0.0001*
0.0182*
Table  6.20 (Cont) Spatial autocorrelation based on
Moran's I  for all probability-based  sites within the
Louisianian Province. (* p < 0.05 = spatial dependency
exists).
Indicators
Exposure Indicators
Chrysene
C1-Chrysene
C2-Chrysene
C3-Chrysene
C4-Chrysene
Dibenzo(a,h)anthracene
Dibenzothio
C1-Dibenzothio
C2-Dibenzothio
C3-Dibenzothio
Fluoranthene
Fluorane
C1 -Ruorene
C2-Fluorene
C3-Fluorene
Naphthalene
C1 -Naphthalene
I C2-Naphthalene
; C3-Naphthalene
: C4-Naphthalene
Petylene
Phenanthrene
C1-Phenanthrene
.C2-Phenarithrene 	
C3-Phenanthrene
C4-Phenanthrene
Pyrene
ldeno( 1 ,2,3,o,d)Pyrene
1 -Methylnaphthalene
1 -Methylphenanthrene
2-Methylnaphthalene
2,3,5-Trimethylnaphthalene
2,6-Dimethylnaphthalene
Moran's 1

0.031 ,
0.008
0.007
0.009
0.007,
0.015
0.061
0.074
0.087
0.076
0.038
0.067
0.078
0.090
0.106
0.049
0.067
0.071
0.063
0.070
0.054
0.066
0.077
0.090
0.080
0.025
0.053
0.053
0.072
0.063
0.062
0.061
0.070
P-value

0.0460*
0.2356
0.2434
0.2240
0.2405
0.1552
0.0018*
0.0003*
<0.0001*
0.0002*
0.0254*
0.0009*
0.0002*
<0.0001*
<0.0001*
0.0082*
0.0009*
0.0005*
0.0016*
0.0006*
0.0044*
0.0010*
0.0002*
<0.0001*
0.0001*
0.0754
0.0052*
0.0052*
0.0004*
0.0015*
0.0017*
0.0020*
0.0005*
Table 6.20 (Cont)  Spatial autocorrelation based on
Moran's  I for all probability-based sites within  the
Louisianian Province. (* < 0.05 = spatial dependency
exists).
Demonstration Report, EMAP-E Louisianian Province - 1991
                                 Page  139

-------
 6.3.2  SPATIAL
   AUTOCORRELATION WITHIN
   THE LARGE ESTUARINE
   CLASS.

 Selected indicators were used to evaluate
 spatial dependency within the large
 estuarine class and to assess the distance
 at which this dependency becomes
 minimal. Several distances were used in
 this analysis  as "cut-off" distances (i.e.,
 distance beyond which no dependency is
 assumed to exist). In this analysis, a non-
 reflexive weight was used so. that a value
 of 1 was assigned to the distance
 associated with the nearest neighbor to a
 site and 0 was assigned to all other
 distances to restrict the influence to only
 the closest site.

 As the cut-off distance increases, and the
 spatial influence is still present, the test p-
 value should tend to decrease.  If the
 distance goes beyond the real sphere of
 influence then unrelated values will be
 given a positive weight and the p-value will
 tend to increase.  The benthic response
 indicators: benthic index, mean benthic
 abundance, and mean numbers of benthic
 species collected at a large estuarine site
 showed strong spatial dependence display
 a tendency to be most strongly related
 spatially to their nearest neighbor (Table
 6.21).  However, significant spatial
 autocorrelation for these response
 indicators exists to a distance of at least 30
 km. As a result, corrections for spatial
 autocorrelation in long-term analyses for
these indicators will have to include more
than simple nearest-neighbor adjustments.
 No spatial autocorrelation was observed for
 mean abundance of fish/trawl regardless of
 distance (Table 6.21).

 Instantaneous bottom dissolved oxygen
 concentration's p-value decreased from
 0.80 at 10 km to 0.67 at 30 km but was
 only 0.37 for the nearest neighbor test.
 Although  never significant, the value at the
 nearest site appeared to be the most
 closely related (Table 6.21).  The optimal
 distance for spatial dependency of
 stratification appeared to be about 20 km.
 P-values for bottom temperature, water
 clarity, and bottom salinity were still
 decreasing at 30 km suggesting a strong
 extended (> 30 km) spatial dependency.
 Within large estuaries, spatial
 dependencies were different for different
 heavy metals (Table 6.21).  Arsenic,
 mercury, selenium, and silver showed no
 spatial dependency suggesting that these
 concentrations might simply represent point
 phenomena or, at worst, spatial
 dependencies of < 20 km.  Aluminum,
 cadmium, chromium, copper, iron, lead,
 manganese, nickel, tin, and zinc showed
 increasing p-values with distance indicating
"spatial dependencies ranging farther than
 nearest neighbor. However, in all these
 spatial dependent metals, dependency was
 strongest on the site closest to the
 sampling point.  Total PAHs showed a
 significant and consistent spatial
 dependency within the range of 0-30 km.
 Dieldrin and 4,4'-DDT were spatially
 dependent in large estuaries.  However, for
 dieldrin, this dependency was stronger
 within the distance range 25-30 km than
 closer suggesting a patchy  but related
 structure of occurrence for this  pesticide.
 2,4'-DDT was not spatially dependent.
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 140

-------
Indicators

Response Indicators
Benthic Abundance



Benthio Species



Benthic Index



Fish Abundance



Habitat Indicators
Instantaneous DO





Percent Light at 1 m



Bottom Salinity



Stratification





Bottom Temperature





Distance
(
-------
Indicators

Exposure Indicators
Iron



Lead



Manganese



Mercury



Nickel



Selenium



Silver



Tin



Zinc



Distance
(
-------
Indicators
Response Indicators
Benthic Abundance
Benthic Species
Benthic Index
Fish Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom Salinity
Stratification
Bottom Temperature
Exposure Indicators
Total Alkanes
Dieldrin
Total PAHs
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Silver
Tin
Zinc
Moran's 1

-0.617
-0.381
-0.393
-0.671

-0.308
-0.310
-0.315
-0.311
-0.311

-0.627
-0.636
-0.694
-O.321
-0.388
-0.454
-0.390
-0.337
-0.306
-0.307
-0.308
•0.695
-0.335
-0.466
-0.350
-0.403
-0.307
P-value

0.5890
0.4096
0.4183
0.6293

0.3554
0.3572
0.3609
0.3578
0.3578

0.5969
0.6037
0.6461
0.3652
0.4148
0.4645
0.4159
0.3768
0.3544
0.3547
0.3559
0.6465
0.3755
0.4742
0.3861
0.4260
0.3547
Table 6.22 Spatial autocorrelation based on Moran's
I for all probability-based sites within Galvoston Bay,
TX. (p < 0.05 = spatial dependency exists).
Indicators
Response Indicators
Benthic Abundance
Benthic Species
Percent Polychaetes
Percent Amphipods
Benthic Index
' Fish Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom Salinity
Stratification
Bottom Temperature
Exposure Indicators
Total Alkanes
Dieldrin
Total PAHs
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
, Silver
Tin
Zinc
Moran's 1

•-0.206
-0.102
-0.212
-0.266
-0.054
-0.155

-0.073
-0.292
-0.127
-0.042
-0.249

-0.225
-0.163
-0.157
-0.220
-0.311
-0.068
-0.07O
-0.263
-0.080
-0.204
-0.133
-0.173
-0.261
-0.168
•fl.234
-0.154
-0.202
P-value

0.5118
0.3216
0.5219
0.6220
0.2458
0.4162

0.2742
0.6677
0.3653
0.2284
0.5922

0.5477
0.4298
0.4188
0.5370
0.6992
0.2666
0.2702
0.6176
0.2855
0.5071
0.3759
0.4497
0.6127
0.4399
0.5632
0.4141
0.5042
Table 6.23 Spatial autocorrelation based on Moran's
I  for  all  probability-based sites  within Lake
Pbntchartrain, LA. (p < 0.05 = spatial dependency
exists).
Demonstration Report, EMAP-E Louisianian Province -1991
                                Page 143

-------
Indicator*
Response Indicators
Banthic Abundance
Bonthic Species
Percent Polychaetes
Percent Amphipods
Bonthic Index
Rsh Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom Salinity
Stratification
Bottom Temperature
Exposure Indicators
Total Alkanes
Dioldrin
Total PAHs
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Silver
Tm
Zinc
Moran's 1

-0.111
-0.134
-0.266
-0.142
-0.370
-0.259

-0.260
0.021
-0.229
-0.247
-0.347

-0.108
-0.087
-0.203
•0.350
-0.320
-0.308
-0.279
-0.343
-0.299
-0.346
-0.067
-0.172
-0.120
-0.350
-0.114
•O.302
-0.336
P-value

0.3541
0.3902
0.6096
0.4027
0.7638
0.7166

0.5997
0.1763
0.5493
0.5782
0.7317

0.3491
0.3162
0.5042
0.7359
0.6940
0.6752
0.6301
0.7264
0.6625
0.7315
0.2875
0.4528
0.3678
0.7368
0.3578
0.6660
0.7166
Tafala 6.24  Spatial autocorrelation based on Moran's I for all
probabnity-baซed cites within Chandeleur Sound, LA. (p < 0.05
a spatial dependency exists).
Indicators
Response Indicators
Benthic Abundance
Benthic Species
Percent Polychaetes
Percent Amphipods
Benthic Index
Fish Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom Salinity
Stratification
Bottom Temperature
Eposure Indicators
Total Alkanes
Dieldrin
Total PAHs
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Silver
Tin
Zinc
Moran'* 1

0.231
0.406
-0.099
-0.135
0.092
-0.200

-0.356
0.131
-0.011
-0.123
0.032

0.258
0.022
0.216
0.144
-0.258
-0.214
-0.068
0.078
0.097
0.031
0.158
-0.020
-0.090
0.049
0.139
0.202
0.070
P-value

0.0312*
0.0027*
0.4463
0.5219
0.1284
0.6523

0.8871
0.0901
0.2754
0.4951
0.2051

0.0226*
0.2209
0.0373*
0.0799
0.7561
0.6799
0.3824
0.1447
0.1227
0.2079
0.0691
0.2905
0.4272
0.1810
0.0836
0.0437*
0.1540
Table 6.25 Spatial autocorrelation based on Moran's I for all
probability-based sites within Mississippi Sound, MS/AL (* p
< 0.05 = spatial, dependency exists).
Demonstration Report, EMAP-E Louisianian Province - 1991
                                Page 144

-------
Indicators
Response Indicators
Benthic Abundance
Benthic Species
Benthic Index
Fish Abundance
Habitat Indicators
Instantaneous DO
Percent Light at 1 m
Bottom Salinity
Stratification
Bottom Temperature
Exposure Indicators
Total Alkanes
Dieldrin
2,4'-DDT
4,4'-DDT
Total PAHs
Aluminum
Antimony
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Silver
Tin
Zinc
Moran'g 1

-0.167
-0.034
-0.140
0.001

-0.304
-0.037
-0.195
-0.292
0.203

-0.026
•0.061
-0.089
-0.145
-0.176
-0.045
-0.057
-0.071
-0.073
-0.074
-0.077
-0.090
-0.098
-O.099
-0.037
-0.084
-0.069
-0.106
-0.094
P-value

0.6787
0.3237
0.6106
0.2396

0.9227
0.3293
0.7458
0.9102
0.0172*

0.3240
0.4284
0.5176
0.6904
0.8122 ,
0.3783
0.4168
0.4612
0.4672
0.4712
0.4783
0.5223
0.5465
0.5504
0.3541
0.5010
0.4544
0.5733
0.5360
Table 6.26 Spatial autocorrelation based on Moran's I for all
probability-based sites within Mobile Bay, Al_  (* p < 0.05 =
spatial dependency exists).
 6.4  NEED FOR REPLICATION OF
    BENTHIC GRABS

 During the 1991 Louisianian Province
 Demonstration, replicates were used to
 potentially reduce site variation for all
 benthic response indicators (i.e., 3
 grabs/site), RPD depth (3 measures per
 site), and fish response indicators (2
 trawls/site at ITE stations and 2-3 trawls at
 sites exhibiting low abundance). A set of
 analyses were completed to assess the
 need for replication of benthic response
 indicators using total abundance and
 number of species as test cases.

 A province-wide CDF for total benthic
 abundance is shown in Figure 6.45
 depicting individual distributions for each
 replicate. With the exception of one outlier
 in grab 2 that extends abundance from
 about 900 to 1400 (i.e., one site), the CDFs
 for the replicates appear similar.  This
 observation only means that the overall
 province-wide distribution does not change.
 It cannot be interpreted to mean that
 significant replicate differences at a site do
 not occur.  Examination of the results of
 province-wide testing for significance of
 replicate differences for total abundance
 and the abundances of key benthic
 taxonomic groups showed that no replicate
 differences occurred in total abundance,
 but significant station differences were
 observed (Table 6.27). Of all the
 taxonomic groups examined, only
 amphipods showed a significant replicate
 effect. Significant station effects were
 Observed for all taxonomic groups.
 Similarly no replicate effects were observed
for total benthic abundance or taxonomic
abundance in the three estuarine classes:
large estuaries (Table  6.28), large tidal
Demonstration Report. EMAP-E Louisianian Province - 1991
                             Page 145

-------
 rivers (Table 6.29) or small estuaries
 (Table 6.30).  In large tidal rivers, station
 differences were not observed for the
 abundances of amphipods or gastropods
 suggesting that multiple sites would not be
 required to estimate these abundances for
 the large tidal river class.

 Abundance only represents part of the
 assessment of the benthic community.
 However, if number of species also shows
 no replicate effect then the collection of
 multiple samples is not required to make
 province-wide or class-wide assessments.
 A province-wide CDF of number of benthic
 species by grab is shown  in Figure 6.46
 and  depicts no differences in the number of
 benthic species due to replicate number.
                 This observation only means that the
                 overall province-wide distribution of number
                 of species does not change and cannot be
                 interpreted to mean that significant
                 replicate differences at a site do not occur.
                 Examination of the results of province-wide
                 testing for significance of replicate
                 differences for total number of species and
                 the number of species of key  benthic
                 taxonomic groups showed that no replicate
                 differences occurred in total number of
                 benthic species. However, significant
                 station differences were observed (Table
                 6.31).  Of all the taxonomic groups
                 examined, only amphipods showed a
                 significant replicate effect. Significant
                 station effects were observed  for all
                 taxonomic groups. Similarly no replicate
         
-------
 effects were observed for total benthic
 abundance or generally for taxonomic
 abundance in the three estuarine classes:
 large estuaries (Table 6.32), large tidal
 rivers (Table 6.33) or small estuaries
 (Table 6.34).  There were replicate
 differences among the grabs in large
 estuaries for the number of species of
 amphipods.  In large tidal rivers, station
 differences were not observed for the
 abundances of amphipods or gastropods
 suggesting that multiple sites would not be
 required to estimate these abundances for
 the large tidal river class.

 Evaluations of the redox potential
 discontinuity layer for each of the grabs
 showed no differences due to replicates but
 significant differences were observed for
 Stations (Table 6.35). No replicate
 differences were seen for any of the three
 estuarine classes.
Benthic Indicator
Total Abundance

Percent Abundance as:
Amphipods

Bivalves
j
Capitellids

Decapods

Gastropods

Molluscs

Polychaetes

Spionids

Tubiflcids

Test Variable
Replicate Grab
Station

Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
P-Value
0.621
<0.001*

0.049*
<0.001*
0.367
0.001*
0.367
<0.001*
0.783
<0.001*
0.404
<0.001*
0.220
<0.001*
0.483
<0.001*
0.961
<0.001*
0.576
<0.001*
         Results of ANOVA testing for differences In total
benthic abundance and abundance by taxonomic group for
benthic replicate grabs and stations for the Louisianian
Province. (* p < 0.05 = significant difference due to test
variable).
: Benthic Indicator
Total Abundance
'
Percent Abundance as:
Amphipods

Bivalves

Capitellids

Decapods

Gastropods

Molluscs

Polychaetes

Spionids

Tubificids

Test Variable
Replicate Grab
Station

Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
P-Value
0.406
<0.001*

0.272
0.002*
0.432
0.001*
0.387
<0.001*
0.366
0.004*
0.108
<0.001*
0.140
<0.001*
0.542
<0.001*
0.980
<0.001*
0.497
<0.001*
benthic abundance and abundance by taxonomic group for
benthic replicate grabs and stations for the large estuarine
class.(*p < 0.05 = significant difference due to test variable).
Demonstration Report. EMAP-E Louisianian Province -1991
                              Page 147

-------
Benthlc Indicator
Total Abundance
Percent Abundance as:
Amphipods
Bivalves
Gapitellids
Decapods
Gastropods
Molluscs
Polychaetes
Spionids
Tubificids
Test Variable
Replicate Grab
Station

Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
P-Value
0.638

-------
Benthlc Indicator
Total Number of Species

Number of Species:
Am phi pods

Bivalves

Capitellids

Decapods

Gastropods

Molluscs

Polychaetes

Spionids

Tubificids
Test Variable
Replicate Grab
Station

Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grab
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
Station
Replicate Grabs
P-Value
0.574
<0.001*

0.019*

-------
   Estuarine Claซซ

   Louisiantan Province


   Largo Estuaries


   Large Tidal Rivera
   Small Estuaries/Small
     Tidal Rh/ora
Test Variable

Replicate Grab
Station

Replicate Grab
Station

Replicate Grab
Station

Replicate Grab
Station
P-Value

0.232
<0.001*

0.566
<0.001*


0.737
<0.003*


0.322
<0.001*
Table 6.35 Results of ANOVA testing for differences In RPD depth (mm) for benthlc replicate
grabs and stations for the Loulslanlan Province and the three estuarlne classes. (* p < 0.05 =
significant difference due to test variable).
Demonstration Report,  EMAP-E Louisianian Province • 1991
                                                   Page 150

-------
                                  SECTION  7

                               CONCLUSIONS
 This evaluation of indicators and design for
 the Louisianian Province is based on a
 single year of information;  thus, it is subject
 to potential year-specific phenomena such
 as climate fluctuations (1991 was a high
 precipitation year), contaminant spills (a
 major oil spill occurred in Galveston Bay in
 late 1990), or year-class strengths (no
 deviations known). This assessment is
 preliminary and its findings should be
 confirmed by subsequent years of sampling
 in the Louisianian Province.

 A companion report delineating a statistical
 summary of the 1991 results has been
 produced (Summers et al.  1993) and
 should be used if the reader is interested in
 the ecological status of the estuaries of the
 Louisianian Province. The following
 conclusions have been drawn from the
 monitoring data collected from the
 Louisianian Province in 1991 with regard to
 indicators and design:

 Response Indicator Development

 •  An index of benthic community structure
   has been developed that effectively
   discriminated between sites of known
   hypoxia and sediment contamination
   and reference sites. The strength and
   validity of this index will  be assessed
   using the 1992 monitoring data.

•  A preliminary index of fish community
   structure has been developed that
   discriminated between sites of known
   hypoxia and sediment contamination
   and reference sites.  While appearing
   statistical strong, this index produces
   considerable vague ecological
   conclusions. Significant additional effort
   is needed before this index is ready for
   general use.

Sensitivity of Indicators

•  Benthic index values, benthic species
 ;  diversity and number of species were
   sensitive indicators of ecological
   condition  relating to hypoxia and
   sediment  contamination.

•  Number of finfish species and
   abundance/trawl were sensitive
   indicators of ecological  condition in
 ,  estuaries.

•  Instantaneous and 24-hr continuous
   measures of bottom dissolved oxygen
   concentrations are indicative  of hypoxic
   condition throughout the index period.

•  Human use  indicators and tissue
   contaminants in fish tissue were not
   indicative  of ecological condition in
   estuaries where condition was defined
   as extent of  hypoxia and sediment
   contamination.

•  Degree of stratification and percent
   organic carbon content  of sediments
   were habitat indicators that were
   strongly associated with ecological
Demonstration Report, EMAP-E Louisianian Province -1991
                            Page 151

-------
   condition.

 •  While mortalities in sediment bioassays
   averaged 15-20% higher in areas of
   hypoxia and sediment contamination
   than in reference sites, this difference
   was not statistically significant.

 •  Total alkanes, pesticides and heavy
   metals were significantly associated with
   sites having observed sediment
   contamination.

 •  Most PAHs and PCBs were not
   associated with observed sediment
   contamination.

 •  Most sediments that were contaminated
   had heavy metal contents exceeding the
   expected concentrations based on
   crustal aluminum.

 •  Significant longitudinal gradients (East-
   West) in the Louisianian Province
   existed for the number of fish
   species/trawl, minimum dissolved
   oxygen concentration, Secchj depth,
   acid volatile sulfides, total organic
   carbon in sediments, total and numerous
   specific alkanes, and several heavy
   metals.

 Research Indicators

•  Number of observed external fish
   pathologys/trawls were 2 to 3 times
   higher in regions of hypoxia and
   industrial contamination and a significant
   longitudinal gradient existed with
   western province fish  having five times
   the pathologies observed in eastern
   province fish.
 •  The percent area occupied by splenic
   macrophage aggregates was 4 to 9
   times greater in regions of hypoxia and
   sediment contamination than reference
   areas for pinfish and Atlantic croaker.

 •  The rate of vertebral deformities was an
   order of magnitude higher in western
   province estuaries than in eastern
   Louisianian Province with rates that
   were 3 to 7 times higher in areas of
   hypoxia and high industrial discharges.

 •  Selected blood chemistry compounds
   including c-reactive proteins were
   significantly higher in brown bullheads
   from heavily contaminated than in
   catfish from reference areas.

 •  Stable isotope and nutrient analysis in
   hypoxic areas indicated the high
   potential for eutrophic conditions
   resulting from algal  production and
   decay.

 Associations

 •  The benthic index was strongly
   associated with sediment contaminant
   levels  and somewhat associated with
   dissolved oxygen concentrations,
   sediment toxicity, and habitat variation in
   RPD depth and salinity.

•  When  the benthic index was
   characterized as categories above or
   below  the 4.1 criteria value, heavy
   metals were strongly associated with
   index classes in large  and small
   estuaries; however,  total alkanes were
   the primary contaminants associated
   with low index values in large tidal
   rivers.
Demonstration Report, EMAP-E Louisianian Province - 1991
                            Page 152

-------
  •  Presence of pesticides and selected
    PCBs, PAHs and heavy metals was
    associated with amphipod toxicity while
    the presence of heavy metals and
    pesticides was strongly associated with
    mysid toxicity.

  •  Significant associations exist between
    sediment contaminant concentrations
    and total organic carbon and acid
    volatile sulfides in the sediment.

  Statistical Design

  •  Most response and exposure indicators
    showed no  differences in distribution
    functions at the estuaries class level
    between index and randomly-placed
    sites.

 •  Paired comparisons showed many
    significant differences in response and
    exposure indicators between index and
    random sites in small estuaries and
    large tidal river segments.

 •  Significant spatial autocorrelation exists
    among values for benthic abundance,
    number of benthic species, water clarity,
    and bottom salinity.

 • Significant spatial autocorrelation  exists
   for most sediment contaminants.

 •  No significant differences were observed
   in benthic abundance or total number of
   species among three replicate benthic
   grabs for province-wide and estuaries
   class distributions.

 1  No differences were observed among
   three replicates of RPD depth for
   province-wide and estuarine class
   distributions.

•  No differences were observed for finfish
   abundance or number of fish species in
   replicate trawls for province-wide or
   estuarine class distributions.

•  No significant differences in the
   estimates of response and exposure
   indicators were observed at the large
   estuarine class-level between the base
   grid density and an enhanced density
   increasing the sample size by a factor of
   four; however, local or estuary-specific
   estimates were significantly different.
Demonstration Report, EMAP-E Louisianian Province - 1991
                           Page 153

-------

-------
                                 SECTION 8
                               REFERENCES
Agius, C.  1979.  The role of melano-macrophage centers in iron storage in normal and
   diseased fish.  J. Fish Dis.  2:337-343.

Agius, C.  1980.  Phylogenetic development of melano-macrophage centers in fish.  J. Zool
   London  191:111-132.

Agius, C. and Roberts, R.J. 1981.  Effects of starvation on the melano-macrophage centers in
   fish.  J. Fish Biol.  19:161-169.

Altabet,  M.A. and J.J. McCarthy.  1986.  Vertical patterns .in 15N natural abundance in PON
   from the surface waters of several warm-core rings in the Sargasso Sea.  J. of Mar. Res
   44:185-201.

Bell, S.S. and B.C. Coull. 1978.  Field evidence that shrimp predation regulates meiofauna.
   Oecologia35:141-148.

Bengtsson, A. and B. E. Bengtsson.  1983.  A method to registrate spinal and vertebral
   anomalies in fourhom sculpin, Myoxocephalus quadricornis  L. (Pisces). Aquilo Ser. Zool.

Bengtsson, B.E.  1979.  Biological variables, especially skeletal deformities in four fish for
   monitoring marine pollution.  Philos. Trans.  Soc. London. 286:457-464.

Bengtsson, B.E., A. Bengtsson and  M. Himberg. 1985. Fish deformities and pollution in
   some Swedish waters. Ambio 14:32-35.

Bryant, V., D.S. McClusky, K. Roddie, and D.M. Newberry. 1984.  Effect of temperature and
   salinity on the toxicity of chromium to three estuarine invertebrates (Corophium volutator.
   Macoma balthica.  Nereis diversicolor).  Marine Ecology - Progress Series.  20:137-149.

Carriker, M.R. 1967.  Ecology of estuarine benthic invertebrates:  A perspective. Pages 442-
   487.  In: G.H. Lauff, ed. Estuaries, Publ. No. 83, 442-487.  American Association for the
   Advancement of Science, Washington, D.C.

Chao, LN. and J.A. Musick. 1977.  Life history, feeding habitats, and functional morphology
   of juvenile scianid fishes in the York River estuary.  Fish. Bull. 75:657-702.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 155

-------
Chapman, P.M. 1989.  Current approaches to developing sediment quality criteria.
   Environmental Toxicology and Chemistry.  8:589-599.

Cifuentes, LA., J. H. Sharp and M. L Fogel 1988. Stable carbon and nitrogen isotope
   biogeochemistry in the Delaware estuary.  Limnol. Oceanogr.  33:1102-1115.

Cloern, J.E.  1982.  Does the benthos control phytoplankton biomass in South San Francisco
   Bay?  Mar. Ecol.  Prog. Serv. 9:191-202.

Coffin, R.B., L.A. Cifuentes, and P. Eldridge.  1992. The use of stable carbon isotopes to
   study microbial processes in estuaries. Stable Isotopes in Ecology (in press).

Cook, D. G. and R.O. Brinkhurst.  1973. Marine flora and fauna of the Northeastern  United
   States, Annelida: Oligochaeta. NOAA Technical Report NMFS CIRC-374.

Ellis, A.E., Munro, A.L.S., and Roberts,  R.J. 1976.  Defense mechanisms in fish. I.  A study
   of the phagocytic system and the fate of intraperitoneally injected particulate material in
   place (Pleuronectes platessa).  J. Fish Biol.  8:67-78.

Ferguson, H.W.  1976.  The relationship between ellipsoids and melano-macrophage centers
   in the spleen of turbot (Scopthalmus maximus). J. Comp. Path. 86:377-380.

Flint, R.W. and R. D. Kalke. 1985.  Benthos structure and function in a south Texas estuary.
   Contributions in Marine Science.  28:33-53.

Fogel, M.L and LA. Cifuentes. 1992.  Isotopic fractionation during primary production. ]nj
   M. Engel and S.A. Macko - Eds. Organic Geochemistry.  Plenum Press (in press).

Fogel, M.L, LA. Cifuentes, D.J. Velinski, and J.H. Sharp.  1992.  The relationship of carbon
   availability in estuarine phytoplahkton to isotopic composition.  Mar. Ecol. Prog. Series (in
   press).

Fox, LE.  1983. The removal of dissolved humic substances during estuarine mixing.  Estuar.
   Coastal Mar. Sci. 16:431 -440.

Fry, B. and E.B. Sherr. 1984.  13C Measurements as Indicators of Carbon Flow in Marine  and
   Freshwater Ecosystems. Cont. Mar. Sci. 27:13-47.

Fry, B., S.A. Macko and J.C. Zieman. 1987. Review of Stable Isotopic Investigations of Food
   Webs in Seagrass Meadows, Florida Marine Research Publications.
 Demonstration Report, EMAP-E Louisianian Province - 1991
Page 156

-------
Gaston, G.R., P.A. Rutledge, and M.L Walther. 1985. The effects of hypoxia and brine on
   recolonization by macrobenthos off Cameron, Louisiana (USA). Contributions in Marine
   Science.  28:79-93.

Gaston, G.R. and J.C. Nasci.  1988. Trophic structure of macrobenthic communities in the
   Calcasieu Estuary, Louisiana.  Estuaries.  11:201*211.

Grizzle, R.E. 1984. Pollution indicator species of macrobenthos in a coastal lagoon.  Marine
   Ecology - Progress Series.  18:191-200.

Grossman, E.O. 1984.  Carbon isotopic fractionation in live benthic foraminifera-comparison
   with inorganic precipitate studies.  Geochim. Cosmochim. Acta 48:1505-1512.

Hamilton, S.J., P.M. Mehrie, F.L. Mayer and J.  R. Jones.  1981.  Mechanical properties of
   bone in channel catfish as affected by vitamin C and toxaphene. Trans. Am. Fish Soc
   110:718-724.

Harper, D.E., L.D. McKinney, R.R. Salzerand R.J. Case.  1981.  The occurrence of hypoxic
   bottom water off the upper Texas coast and  its effects on the benthic biota.  Contributions
   in Marine Science. 24:53-79.

Hedges, J.I. and D.C. Mann. 1979. The characterization of plant tissues by their lignin
   oxidation  products.  Geochim. Cosmochim. Acta 43:1803-1807.

Holland, J.S., N.J. Maciolek, and C.H. Oppenheimer.  1973. Galveston Bay benthic
   community structure as an indicator of water quality. Contributions in Marine Science.
   17:169-188.
Holland, A.F., AT. Shaughnessy, L.C. Scott, V.A. Dickens, J. Gerritsen, J.A. Ranasinghe.
   1989.  Long-term benthic monitoring and assessment program for the Maryland portion of
   Chesapeake Bay:  Interpretive report. CBRM-LTB/EST-2. Prepared for Maryland
   Department of Natural Resources, Power Plant Research Program. Versar, Inc., ESM
   Operations, Columbia, MD.

Holligan,  P.M., R.P. Harris, R.C.Newell, D.S. Harbour, R.N. Head, E.A.S. Linley, M.I. Lucas,
   P.R.G. Tranter and C.M. Weekley.  1984. Vertical distribution and partitioning of organic
   carbon in mixed, frontal and stratified waters of the English Channel. Mar. Ecol. Prog. Ser.
   14:111-127.

Jurat, J.A.,  M.A. Poirrier, and T.M. Soniat.  1983.  Effects of saltwater intrusion from the Inner
   Harbor Navigation  Canal on the benthos of Lake Pontchartrain, Louisiana.  Gulf Research
   Reports.  7:247-254,
Demonstration Report, EMAP-E Louisianian Province -1991
Page 157

-------
 Karr, J.R. and D.R. Dudley. 1981. Ecological perspective on water quality goals.
   Environmental Management.  5:55-68.

 Karr, J.R., K.D. Fausch, P.L. Angenmeier, P.R. Yant, and I.J. Schlosser.  1986. Assessing
   biological integrity in running waters:  a method and its rationale.  Special Publication 5,
   Illinois  Natural History Survey, Champaign, IL.

 Kemp, W.M. and W.R. Boynton. 1980.  Influence of biological and physical processes on
   dissolved oxygen dynamics in an estuarine system: Implications for measurements of
   community metabolism. Estuarine and Coastal Marine Science 11:407-431.

 Knapp, C.M., D.R. Marmorek, J.P.  Baker, K.W. Thornton, J.M. Klopatek, and D.F. Charles.
   1990.  The indicator development strategy for the Environmental Monitoring Assessment
   Program. EPA 600/3-91/023. U.S. EPA Office of Research and Development,
   Washington, D.C.

 Lee, S.H.  and J.A. Fuhrman. 1987. Relationships between biovolume and biomass of
   naturally derived marine bacterioplankton.  Appl. Environ. Microbiol. 53:1298-1303.

 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.  US Department of Commerce, National Oceanic and
   Atmospheric Administration, National Ocean Service, Rockville, MD.

 Luna, L.G. 1968. Manual of Histologic Staining Methods of the Armed Forces Institute of
   Pathology, 3rd Edition.  McGraw-Hill, New York.

 Macko, S.A. 1981.  Stable nitrogen isotope ratios as tracers of organic geochemical
   processes.  Ph.D.  thesis, Univ. of Texas at Austin.  181 p.

 Mariotti, A., C. Lancelot, and G. Billen.  1984. Natural isotopic composition of  nitrogen as a
   tracer of origin for suspended organic matter in the Scheldt estuary. Geochim Cosmochim.
   Acta 48:549-555.

 Mayer, F.L., Jr., P.M. Mehrie, Jr. and W.P.  Dwyer.  1977. Toxaphene: Chronic toxicity to
   fathead minnows and channel catfish. Ecol. Res. Ser. No. EPA-600/3-77-069.  U.S.
   Environmental Protection Agency, Duluth, MN.

McClusky, D.S., V. Bryant and R. Campbell.  1986.  The effects of temperature and salinity on
   the toxicity of heavy metals to marine and estuarine invertebrates.  Oceanography  and
   Marine  Biology Annual Review. 24:481-520.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 158

-------
McManus, J.W. and D. Pauly.  1990.  Measuring ecological stress:  variations on a theme by
  R.M. Warwick.  Marine Biology. 106:305-308.

Mehrle, P.M., Jr., T.A. Haines, S.J. Hamilton, J.L. Ludke, F.L. Mayer, Jr., and M.A. Ribick.
  1982.  Relationship between body contaminants and bone development in east-coast
  striped bass.  Trans. Am. Fish. Soc. 111:231-241.

Messer, J.J. 1990. EMAP indicator concepts. Chapter 2. In, C.T. Hunsakerand D.E.
  Carpenter (eds) Ecological indicators for the Environmental Monitoring and Assessment
  Program. EPA 600/3-90/060.  U.S. Environmental Protection Agency, Office of Research
  and Development, Research Triangle Park, NC.

NOAA (National Oceanic and Atmospheric Administration).  1987.  NOAA estuarine and
  coastal ocean  science framework.  U.S. Department of Commerce, National Oceanic and
  Atmospheric Administration,  EEstuarine Program Office, Washington, D.C.

NOAA.  1988.  Federal plan for ocean pollution research development, and monitoring: Fiscal
  years 1988-1991.  Prepared by the National Ocean Pollution Program Office for the
  National Ocean Pollution Policy Board, Rockville, MD.

Officer, C.B., T.J. Smayda, and R. Mann.  1982.  Benthic filter feeding:  A natural
  eutrophication  control. Mar. Ecol. Prog. Ser. 9:203-210.

Overton, W.S., D.L Stevens, and D. White.  1991.  Design  report for EMAP, Environmental
  Monitoring and Assessment Program.  Document in Review.  U.S. EPA, Environmental
  Research Laboratory, Corvallis, OR.

Pearson, T.H. and R. Rosenberg.  1978.  Macroberithic  succession in relation to organic
  enrichment and pollution of the marine environment.  Oceanography and Marine Biology
  Annual Review. 16:229-311.

Plafkin, J.L, M.T. Barbour, K.D. Porter, S.K. Gross, and R.M. Hughes.   1989. Rapid
  bioassessment protocols for use in streams and rivers:  Benthic macroinvertebrates and
  fish. EPA/444/4-89-001. U.S. Environmental Protection  Agency, Assessment and
  Watershed Protection Division, Washington, D.C.

Rabalais, N.N. 1990. Biological communities of the south Texas continental shelf. American
  Zoologist. 30:77-87.

Rakocinski, C., R.W. Heard, T. Simons and D. Gledhill.  1991. Macroinvertebrate associations
  from beaches of selected barrier islands in the Northern Gulf of Mexico: important
  environmental  relationships. Bulletin of Marine Science.  48:689-701.
Demonstration Report, EMAP-E Louisianian Province • 1991
Page 159

-------
Reish, DJ. 1986. Benthic invertebrates as indicators of marine pollution:  35 years of study.
  Oceans 86. 3:885-888.

Rhoads, D.C. 1974.  Organism-sediment relations on the muddy sea floor. Oceanogr. Mar.
  Biol. Ann. Rev. 12:263:300.

Rosenberg, R. 1977.  Benthic macrofaunal dynamics, production, and dispersion in an
  oxygen-deficient estuary of West Sweden.  Journal of Experimental Marine Biology and
  Ecology. 26:107-133.

Rygg, B. 1986.  Heavy-metal pollution and log-normal distribution of individuals among
  species in benthic communities.  Marine Pollution Bulletin.  17:31-36.

SAS Institute, Inc. 1989. SAS/STATฎ  User's Guide, Version 6, Fourth Edition, Volume 1.
  SAS Institute,  lnc.:Cary,  NC.  943 pp.
 #     *
Sharp, J.H., Culberson, C.H., and Church, T.M. 1982. The chemistry of the Delaware
  Estuary. General considerations. Limnol. Oceanogr. 27:1015-1028.

Summers, J.K. and V.D. Engle.  1992. Evaluation of sampling strategies to characterize
  dissolved oxygen conditions in northern Gulf of Mexico estuaries.  Environmental
  Monitoring and Assessment.  In press.

Summers, J.K., J.M. Macauley, and P.T. Heitmuller. 1992.  Environmental Monitoring and
  Assessment Program - Estuaries Component: Louisianian Province 1991 Demonstration
  Field Activities Report.  EPA/ERL-GB No. SR-188. U.S. Environmental  Protection Agency,
  Environmental Research Laboratory, Gulf Breeze, FL

Summers, J.K., J.M. Macauley, P.T. Heitmuller, V.D. Engle, A.M. Adams, G.T. Brooks. 1993.
  Annual Statistical Summary:  EMAP-E Louisianian Province. 1991. U.S. Environmental
  Protection Agency, Office of Research and Development, Environmental Research
  Laboratory, Gulf Breeze, FL EPA/600/R-93/001.

Thurman, E.M.  1985. Organic Geochemistry of Natural Waters. Martinus Nijhoff/Dr. W. Junk
  Publishers. 497 pp.

Tsutsumi, H., T. Kikuchi, M. Tanaka, T. Higashi, K. Imasaka, and M. Miyazaki. 1991. Benthic
  faunal succession in a cove organically polluted by fish farming.  Marine Pollution  Bulletin.
  23:23-238.
Demonstration Report, EMAP-E Louisianian Province - 1991
Page 160

-------
Weisberg, S.B., J.B. Frithsen, A.F. Holland, J.F. Paul, K.J. Scott, J.K. Summers, H.T. Wilson,
   R. Valente, D.G. Heimbuch, J. Gerritsen, S.C. Schimmel, and R.W. Latimer. .1992.  EMAP-
   Estuaries Virginian Province 1990 Demonstration Project Report. EPA/600/R-92/100.  U.S.
   Environmental Protection Agency. Environmental Research Laboratory, Narragansett, Rl.

Williams, B.K.  1983. Some observations on the use of discriminant analysis in ecology.
   Ecology.  64:1283-1291.

Wolke, R.E., Murchelano, R.A., Dickstein, C.D., and George, C.J. 1985. Preliminary
   evaluation of the use of macrophage aggregates (MA) as fish health monitors.  Bull.
   Environ. Contam. Toxicol. 35:222-227.                                             /
 Demonstration Report, EMAP*E Louisianian Province -1991
Page 161
                                                    U.S. GOVERNMENT PRINTING OFFICE: 1994 - 550-001/80333

-------

-------