Virginian Province Macroinfaunal
Community Structure: PCA-H Analyses
and an Assessment of Pollution
Degradation Indices
30 -
25
£20
I15
<0
10
5 -
0.5
Eugene D. Gallagher
Environmental, Coastal, and Ocean Sciences
100 Morrissey Boulevard
University of Massachusetts Boston
Boston MA 02125-3393
J. Frederick Grassle
Institute of Marine and Coastal Sciences
Rutgers University
New Brunswick NJ 08903-0231
''< • $& • " iSS ' . : '
-1 -1
Final Report Submitted to the
United States Environmental Protection Agency
Atlantic Ecology Division (AED)
Narragansett, Rl 02882
EPA PROJECT OFFICER: Brian D. Melzian
0.5
- :• • :.=• :...="•: -.:'. ••:••••-.-. •.. •
-------
Virginian Province Macroinfaunal
Community Structure: PCA-H
Analyses and an Assessment of
Pollution Degradation Indices
Eugene D. Gallagher
Associate Professor
Department of Environmental, Coastal, and Ocean Sciences
University of Massachusetts at Boston
Boston MA 02125
(World Wide Web URL: http://www.es.umb.edu/edgwebp.htm)
J. Frederick Grassle
Professor and Director
Institute of Marine and Coastal Science
Rutgers University
PO Box 231
New Brunswick NJ 08903-0231
A final report submitted to the
United States Environmental Protection Agency (EPA)
Atlantic Ecology Division (AED)
Narragansett, Rhode Island 02882
EPA PROJECT OFFICER: Brian D. Mebtian
February 11,1997
-------
EMAP-E VP COMMUNiTY STRUCTURE
TABLE OF CONTENTS
EXECUTIVE SUMMARY 1
INTRODUCTION 2
STATISTICAL METHODS USED IN THIS STUDY 4
How the EMAP-E VP benthic data were collected 4
Preparing the EMAP-E Virginian Pro ince data for community analysis 4
Methods to analyze community structure 7
RESuLTs & DiscussIoN 8
The What, Why and Where of benthic monitoring 8
Analysis of the EMAP-E VP Benthic Degradation Indices 12
Community Structure Analyses of Virginian Province 23
CONCLUSIONS 37
On the central roleof salinity 37
Suggestions for improving EMAP-E VP 37
Have the EMAP Goals and Objectives Been Met 9 ... 43
REFERENCES 44
APPENDIX I METHODS FOR ANALYZING COMMUNITY STRUCTURE 54
Diversity indices 54
Rarefaction, CNESS and Principal Components Analysis of Hypergeometric probabilities
(PCA-H) 56
APPENDIX II TERMS AND DEFINITIONS 66
APPENDIX HI FULL EMAP-E VIRGINIAN PROVINCE MODui v FAUNAL LIST 70
APPENDIX IV SAMPLE CLUSTER ANALYSiS 90
APPENDIX V SPECIES CLUSTERS FOR ALL 551 EMAP-E VP TAXA 107
-------
GALLAGHER & GRASSLE
EXECUTIVE SUMMARY
This report describes research done under a cooperative agreement between the EPA and the University of
Rhode Island and Rutgers University. The lead author performed this work while on sabbatical leave at
Rutgers University in the 1994- 1995 academic year. Dr. J. Fred Orassle hosted and assisted Dr. Gallagher in
the analysis of the EPA’s “Environmental Monitoring and Assessment Program-Estuaries Virginian Province
“(EMAP-E VP) data.
The EMAP program set as a goal the identification of degraded biological conditions in the nation’s
ecosystems. Biotic indicators of degraded conditions were to be developed rather than focusing on the
concentrations of pollutants (Messer er at. 1991, p. 70-8 1). Each EMAP program was to design biotic
indices to detennine the extent of impacted biological conditions. Then, the assessment portion of the EMAP
program was to determine whether these adverse effects could be attributed to the effects of pollution. The
EMAP program was designed as a multi-decade program. The EMAP-Estuaries program was to assess the
status of our nation’s near-shore coastal zone, including estuaries. The Virginian and Louisianian Provinces
were chosen to be the demonstrations for the EMAP-E program. The EMAP-E Virginian Province (EMAP-E
VP) program developed four different indices of benthic degradation. We review each of these indices and
offer our assessment of their adequacy as descriptions of impacted benthic communities. The first three
benthic indices are no longer used in the EMAP-E VP program. We provide additional reasons to reject these
indices. We provide an analysis of the latest EMAP-E VP index, the 1990-1993 index. This index may
discriminate between a subset of stations regarded as degraded and non-degraded, but it should not be used as
a general index of degraded marine benthos. When reduced to simple terms, this index states that roughly
7000 spionid polychaetes or 17000 oligochaetes per square meter indicates benthic degradation. This index
will classify large portions of pristine benthic areas in the Virginian Province and throughout the world as
degraded.
We review one of the central assumptions of the development of EMAP-E VP benthic degradation indices:
that sediment pollutant concentrations, overlying dissolved oxygen concentrations, and amphipod toxicity are
necessary and sufficient conditions for creating test data sets of degraded and reference stations. We do not
agree with this assumption. It is possible to have relatively unmipacted benthic sites be classified as
degraded using the EMAP-E VP criteria for degradation. Conversely, many impacted sites would fit the
EMAP-E VP criteria that define natural or reference sites. The EMAP-E program is based on circular logic.
Instead of devising independent criteria to assess the effects of contsminants and low dissolved oxygen on
benthic communities, the EMAP-E program has a priori assumed that sites having concentrations of
contaminants in excess of published thresholds, amphipod survival less than a threshold or low dissolved
oxygen in the overlying water must be degraded. Once these criteria were established, two test data sets were
created and an equation derived which would separate these two groups. All other EMAP-E VP sites are
scored using this flmction. The EMAP-E VP program did not evaluate whether any of these sites really
showed biological impacts indicative of degradation.
We present new analyses of the patterns of benthic community structure in the Virginian Province. The
effects of pollution on benthic communities must be assessed relative to natural patterns of variation in
community structure. We use both classification and ordination analysis to describe the major patterns in
cozmnunity structure in the EPA’s EMAP-E VP benthic data. These analyses use the metric faunal distance
metric CNESS, short for Chord-normalized expected species shared (Trueblood et aL 1994). This index is a
metric version of Grassle and Smith’s (1976) NESS or Normalized Expected Species Shared faunal similarity
index. The ordination method based on CNESS is called PCA-H, short for principal components analysis of
hypergeomethc probabilities. We conclude that salinity is the overriding factor controlling natural patterns of
Virginian Province community structure. It is the major factor controlling the maximum number and type of
-------
2
EMAP-E VP COMMUNITY STRUCTURE
species chat occur in an area. Species richness in the Virginian Province is a strong function of salinity, with
very low salinity habitats having roughly eight species in three replicate samples and high salinity areas
having over forty species.
The EMAP-E VP benthic degradation indices have attempted to account for this habitat factor. Salinity
strongly affects degradation indices based on species richness and the abundance of opportunistic taxa,
variables which strongly covary with salinity. The use of opportunistic or pollution-indicating taxa is
problematic because almost all opportunistic taxa used in marine ecology are natural components of medium
and low salinity habitats.
The existing EMAP-E indices indicate that tidal-river habitats have the highest percentage of degraded area,
approximately 40%. Tidal river systems, often near urban sources of pollution, are more likely to have
degraded benthic communities than large and s mall estuaries. However, these sites are also the ones most
likely to be misclassified as degraded if the effects of salinity are not properly taken into account
We review the extensive literature on benthic pollution indices. Much of this literature has attempted to fmd
an index to classify pollution-affected vs. natural benthic conununities. The Chesapeake Bays Program
developed a Restoration Goals Index for the Chesapeake Bay. This index is different from the EMAP-E
index. An index similar to the Chesapeake Bay Restorations Goals Index was developed for the Regional
EMAP program in the New YorklNew Jersey harbor system.
INTRODUCTION
The Environmental Monitoring and Assessment-Estuaries (EMAP-E) program was designed to be a long-
term, multi-decade monitonng plan for the nation’s estuaries. EMAP-E is a subset of the nationwide EMAP
program, which was to provide answers to the following questions (Weisberg e aL 1993):
• What. is the status, extent, and geographical distribution of the nation’s ecological resources?
• What proportion of these resources is declining or improving? Where? At what rate?
• Whatfactorsarelikelytobe
contributing to declining
conditions?
• Are pollution control, reduction,
mitigation, and prevention
programs achieving overall
improvement in ecological
condition?
The Virginian Province was selected as the
demonstration area to implement the sampling and
analytic procedures for the EMAP-E program. As a
result, it has been sampled over four different years
(1990-1993). The Louisianian Province was
sampled in 1991. These provinces are shown in
Figures 1 and 2.
EMAP-E sampling stations (sites) were chosen
using Overton et aL’s (1990) probability-based
FigureL The EMAP-E Provinces. The Virginian
Province was sampled each year from 1990-
1993 and the Louisianian Province was first
sampled in 1991. (Figure from the EMAP-E
World-Wide Web page.J
-------
GALLAGHER & GRASSLE
3
sampling procedure Probability sampling was
applied to threc strata in the Virginian Province v ia prsvinc. sauipunq Sliss
large estuaries, small estuaries and tidal rivers A -
random selection of small estuaries was selected for I C
probability-based sampling, and all large tidal -
rivers and large estuaries were sampled. -
Approximately two thousand benthic grab samples
were processed during the four-year duration of
EMAP-E VP.
The EMAP-E VP program used patterns of benthic
community structure to assess the status of the
Virginian Province’s near-shore coastal zone.
Benthic ecologists throughout the world use benthic
community structure to assess the effects of
anthropogenic pollution (e.g., Boesch and
Rosenberg 1981, Bloom 1980, Chapman eta!.
1987, Field ci a!. 1982, Gray 1976, 1979a & b,
1989, Jumars 1981, Warwick 1993). Sozucreasons
for the efficacy of monitoring changes in benthic community structure are that benthic populations are
relatively sedentary (i.e., they can’t migrate away from a pollution source or source of disturbance), and their
typical monthly to annual generation times arc such that the populations are adapted to short-term
fluctuations in environmental variables but are capable of a strong numerical response to significant long-
term environmental changes. Moreover, the populations are sensitive enough to respond to relatively low
levels of toxic substances. For example, Grassle ci aL 1981 observed pronounced community responses to
90 nglg of #2 diesel oil in the MERL ecosystem tanks). The recovery time of benthic populations is short
enough that changes in community structure can be detected in a matter of months, but long enough that the
community structure is to some extent a response to the integrated habitat quality over the previous months or
even years.
The EMAP-E VP program developed benthic degradation indices to determine the proportion of the
Virginian Province with degraded benthic communities. The EMAP-E VP program selected a set of degraded
stations and a set of undegraded, or “reference” stations, and then produced an equation based on biological
variables that disaiminated between these two groups. The Chesapeake Bay Restorations Goals Index [ ROll
(Ranasinghe ci aL 1993) used a similar approach. The EMAP-E VP and RGI indices differ in the statistical
methods used to classify samples into degraded and undegraded classes. The EMAP-E VP uses a parametric
linear disaiminant function. The RGI index is based on an ordinal ranking of biological variables into the
groupings 1 (below expected), 2 (expected), and 3 (greater than expected). A degraded station in the ROl
index is one that has an average rRnking aaoss biological variables of less than 2.
To demonstrate the utility of alternate methods for assessing the effects of pollution on benthic communities,
this report contains detailed analyses of patterns of community structure in the four-year Virginian Province
data. These analyses are performed using methods based on CNESS and PCA-H (Gallagher cia!. 1992,
Trueblood eta!. 1994). CNESS or the chord-normalized expected species shared, is a metric for assessing
the faunal similarity among samples. PCA-H, short for principal components analysis of hypergeometric
probabilities, is an ordination technique based on CNESS. CNESS is the metric equivalent of Grassle and
Smith’s (1976) Normalized Expected Species Shared or NESS index. Using either NESS or CNESS, the
entire 1918-sample EMAP-E VP data can be clustered using COMPAH96 in about 10 minutes using a
desktop PC. COMPAH96 is the latest version of Boesdi’s (1977a) clustering program. It is available for
Figure 2. Virginian Province EMAP-E sample
locations from 1990-1992(1993 sample
locations not shown, Figure downloaded from
the EMAP-E web pa gel
-------
4
download with documentation on the lead author’s World Wide Web page. Boesch’s (1977a) EPA report
describes many of the methods available in COMPAH and applies th in to benthic data fiDm Chesapeake
Bay.
STATISTICAL METHODS USED IN THIS STUDY
How the EMAP-E VP benthic data were collected
Samples are tak within the Virginian Province using a probability-based sampling design within pre-
defined strata. Weisberg et a!. (1993, p.2-4 to 2-5) describe the strata: Large estuaries, Large Tidal Rivers,
and Small Estuarine Systems. The specific application of probability sampling to each strata was different.
A systematic random two-dimensional grid was applied to the large estuaries. Overton’s (1989) sampling
grid, designed for the nationwide EMAP program, was scaled down to niake these grids. The sampling
points were the centers of the hexagonal gilds. The five large tidal rivers (i.e., Hudson, Potomac, James,
Delaware, and Rappahnannock) were sampled using a one-diiw tcional analog of the two-dimensional gild
used for the large estuaries. A list frame was used to select 32(23%) of the 137 small estuarine systems
during the 1990 sampling. All small estuaries were ranked by latitude and grouped into groups of four. One
small estuary out of each group was randomly chosen. In the 1990 sampling, Delaware Bay and the
Delaware River were sampled more intensively. Subsequent EMAP-E VP sampling used the same basic
approach. Strobel a a!. (1995, p. A-2) summarize all four years of EMAP-E VP sampling. There were 446
sites sampled with probability-based sampling. Several EMAP-E VP sampling sites were sampled repeatedly
over the four-year peziod (e.g., twenty samples from Indian River Bay site No. 150, and eighteen benthic
samples from the Potomac River site No.188). Most sites were sampled only once, with three replicate
grabs. A subset of sampling sites from the 1990 sampling, called Index sites, were chosen because they are
‘located in depositional environm its, where there is a high probability of sediment cont min tion or low
dissolved oxygen conditions.’ There were 86 index sites in the EMAP-E VP dM hase. There were a select
number of 1990 sampling sites chosen for long-term sampling, the Long-term trend sites. There were twelve
Long-term trend (LTT) sites. Strobel a a!. (1995) describe additional sites that were sampled repeatedly
during the EMAP-E VP program.
Typically three Ted Young nxxlifled van Veen grab samples (0.044 m 2 ) were collected at each sampling
station (site) from the Virginian Province (Figure 2). Additional grabs were t k i and composited for
sediment chemistry, grain size, and amphipod toxicity tests. The bembic infauna retained a 500 jim mesh
sieve were identified. Most of the sorting and identification in the EMAP-E VP program was done by Cove
Associates. Most of the individuals were identified to species. The taxonomy of samples collected in areas
with bottom water salinity less than 5 psu (short for practical salinity units, formerly called parts per
thousand or abbreviated % ) is handled differently than the rem ining samples. Oligochaetes and
chironomids were classified to levels finer than the taxonomic level Class. Many of the oligochaetes and
chironomids collected from sites with bottom salinities less than 5 psu were identified to species. The
chironomids and oligochaetes from samples collected in areas with salinities above 5 psu were lumped at the
level of Family and Class, respectively.
Preparing the EMAP-E Virginian Province data for community analysis
The unedited EMAP-E VP data cannot be used for traditional comnumity structure analysis. There are 868
taxonomic categories in the full EMAP 4s.thbase. We dropped many taxa and nu ged others to form a much
smaller set of valid taxonomic categories. Some of the EMAP taxa arc invalid, some are redundant , and
many refer to epif nn l taxa. We arrived at a final list of 551 taxa. We discuss the reasons for dropping and
merging categories below. This provide the full and edited list of EMAP-E VP taxa in Appendix IL
-------
GALLAGHER & GRA T..E
Drop and Merge Rules
There are over 868 taxononiic designations listed in the four-year Virginian Province data set. Two of
these ‘taxa’ include:
NOORGPRS No Organisms Pr seni
POLYCHAE Polychaeta Other - Unidentified & fragments
Of the remaining 866 species codes in the EMAP-E VP database, more than three hundred must be
dropped or merged for community structure analyses. We use 551 taxonomic designations (most are
species) for our analyses. For more detailed analyses of community structure, we would probably drop
many or merge many additional EMAP-E VP taxa. For example, much of the break in community
structure at the 5 psu mark is because chironomids and oligochaetes were sorted to levels finer than family
and class in the low Salinity habitats, but not in the higher salinity habitats. To assess whether 5 psu really
marks a key transition in community structure, these taxa would have to be merged or dropped. The
existing EMAP-E VP data set flags only a subset of the invalid taxa with the database flag SPEC_ION
(Ignore this species in calculating total species per event).
The authors of this study developed the first list of “valid EMAP taxa” using two DROP rules and three
MERGE rules:
The drop rules
Drop Rule 1. Drop all strictly epifaunal, meiofaunal, and pelagic taxa.
Drop Rule 2. Drop all general taxonomic designations at the generic, familial and higher
taxonomic levels if there are more than two valid lower-level designations for that
group. For example, there ale many species of the family Spionidae identified in
the EMAP data and three species of the genus Spio. Therefore, both the familial
level EMAP taxon SPIONIDA and the generic level taxonomic category SPIO
must be dropped. Failure to implement this “drop” rule would have the
unfortunate effect of greatly enhancing the faunal similarity of samples along
environmental gradients. Samples would appear more similar than they really are.
Only a limited number of non-specific designations escaped the drop nile. These
are listed below. Several of these higher level taxonomic designations should be
dropped in future analyses designed to assess biogeographic patterns (e.g.,
oligochaetes , Tubificidae with capiliform chaetae, Tubificidae without capiliform
chaetae).
Acanthohaustoruis app. Dolichopodidae
Amphitntinac Flabclhgendae
Aphelochae:a app. Laonzce app.
Bewa app. Magelona app.
Bucctnidae Microchironomus app
Capuelia app. Nemertinea
Chuonomidac Oligochacta
Ch:ronomus app. Ophryotrocha app.
Cladosanytarsus app. Owerna app.
CoeIotan pus app. Palpomysa app.
Crjp:ochironomus app. Phoronis app.
Demicryptochirononuas app. Ptsidiwn app.
Dicrotendapes app. Polygordius app.
Dipicra Pc I ypedilwn app.
-------
6 EMAP-E VP COMMUNITY STRUCTURE
ProcLadlus spp. Sphaeromias app.
Procladius (Holotanypus) spp Sdctochirc,wmus spp.
Protodrilus app. Tanypus app.
Protthaustorius spp. Tanytarsux app.
Pseudochironomus app. Th I2 , inidea
Sipuncula Tubificidae without capiliform chaezae
Solecurtidac Tubificidac with capilifonn chaezac
S tharmdompsis app.
The merge rules
Merge Rule 1. Merge all taxa that can not be adequately distinguished taxonomically. The
EMAP-E VP data set contains hundreds of taxonomic designations at
taxonomic levels higher than the species. For example, in the full EMAP
species list, the following eight ampeliscid amphipod categories are found:
AMPEABDI AnqielLaca abd Ira
AMPEABVA Ampellrca abdita-vadonan complex
AMPELISC Asnpelisca app.
AMPEVADO Ampelisca vadorum
AMPEAGAS AmpelLaca agauizi
AMPEVERR Antpelisca verrilli
AMPHIPOD Amphipoda Other
We consulted with the taxonomists at Cove Associates (Tim Morris and
Nancy Mountford) to detemilne whether Ampelisca spp. and ‘Amphipoda:
Other’ were indeed different from the juvenile stages of Ampelisca abdita and
A. vadorum, which can not be identified to species. They stated these
categories did not refer to either A. abdita or A. vadorwn, therefore both
higher level taxa were dropped using DROP RULE 2 (above). Using Merge
Rule I, the AMPEABDI, AMPEABVA, and AMPEVADO designations were
fused, reducing the original seven taxonomic categories used to describe
ampeliscid amphipods to three:
AMPEABVA Anipelisca a dita.vadoruin complex
AMPEAGAS Anipelisca agassizi
AMPEVERR Ampelisca verrilhi
Merge Rule 2. If there are a pair of taxonomic designations indicated by Genus A species x
and Genus A spp., and there is a high probability that the individuals
identified only as Genus A spp. are indeed Genus A species x, then merge the
two taxa. This merge usually occurs when there is only one species in
addition to the higher level taxonomic designation. Note that Drop Rule 2
would force the deletion of Genus A spp. if there were more than one species
of Genus A in the data set. Our main justification for merging these taxa is
that the taxonomists were better able to distinguish species in the later years of
the EMAP-E VP sampling.
The following groups of EMAP taxa refer to single species and should be
merged:
LUMBHEBE Scole:onia lithe, (These refer to the same species, but
SCOLHEBE Scoletoma hebes were listed as separate species in the
EMAP data)
-------
fl . , £a flV £ I DAIL
7
MELINNA Melinna spp.
MELIMACU Melinna ,naadwa
NOTOMAST Nowmam&s app.
NOTOSPA Nowniassas ap. A
Ewing
OWENFUSI OweniafusÜformis
OWENIA Owenia app.
PECFGOUL Pecunaria gout d i i
PECTINAR Pectinaria spp.
ASYCELON
ASYCHIS
SABAELON Sahaco elonganu
(All 3 designanons refer to the same
maldanid polychaete species, but
different codes were used in different
years in the EMAP database (The genus
name had changed from Asychis to
SabacoJ)
EUDOPUSI Eudorellapusila
EUDORELL Eudorella spp.
MICRATRA Mlcrophiopholis
wra
OPHIUROI Op luuroidea
MUSC RAN MUSCUIIVZn
transverswn
MUSCULIU Musczduun app.
ORCHMINU Orchomenella
msnuta
ORCHOMEN Orchomenella app.
PANDGOUL Pandora
gouldiana
PANDORA Pandora app.
PANDORID Pandondae
TELLAGIL Tellina agiis
TELLINA Telhna spp.
(Note: the EMAP taxon TELLINID
Tcllinidae is dropped)
YOL.DIA Yoldia app.
YOLDLIMA Yoldia limatuta
For example. the EMAP species list contains the following three taxa:
DROP Rule 2 would dictate that LE1TOSCO (Leiroscoloplos spp.) should be
dropped. However, Cove Associates is reasonably ceitain that the individuals
identified as LEITOSCO are Leitoscoloplos robustus, so these two categories
are merged. forming 2 valid taxa
Converting EMAP-E VP data from SASTh to MatlabTM
The EMAP-E VP data are stored in a SAS database. We adapted a SAS program, provided by S.
Weisberg and A. Ranasinghe (VERSAR, Columbia MD) to convert the SAS EMAP-E VP database to
PHERAFFI
PHERUSA
Phenasa affinis
Pjienssa app.
Sabaco elonganas
SOLEMYA
SOLEMYID
SOLEVELU
mya app.
Solemyidae
Solemya veiwn
Merge Rule 3.
On occasion, a merge could occur between a higher level category and a
species, even if more than one species in a genus were present. If the
taxonomists at Cove Associates were reasonably certain that individuals
identified as Genus spp. belonged to a valid species designation, these taxa
were fused rather than dropping the higher level designation.
LEITFRAG
LEITOSCO
LEITROBU
LeuoscoloplosfragilLr
Leitoscoloplos app.
Leitoscoloplos robustus
LEITFRAG
LEJTROBU
Methods to analyze community structure
Leiroscolop losfragilis
Leaoscoloplos robustus
-------
8
COMPAH input format COMPAH ieads data in a number of accepted formats and will convert to
other formats, including the binary form used by MA1i.AB 1 ’.
Most of the analyses performed in this report were done with MATLABTM programs written by the lead
author. Many of these programs ale available on the lead author’s World-Wide Web page.
Diversity analyses
We used the following diversity indices to analyze the EMAP-E VP data: Bnllouin’s H, Shannon’s
H’, Huribert’s E(S), Pielou’s .1’ Evenness, Simpson’s diversity, total number of species, and
Gleason’s D. The first six indices are described in Pielou (1969, 1975, 1977), Peet (1974) and
Magurran (1988). Smith and Grassle (1977) describe the statistical properties of Hurlbert’s (1971)
E(S,) and Simpson’s unbiased diversity indices. Hurlbert’s (1971) E(S ) is based on Sanders’ (1968)
rarefaction method for analyzing species diversity. Gleason’s D diversity index, the number of species
divided by the logarithm of number of individuals, is described in Washington (1984).
Cluster analysis
COMPAH was used to cluster samples and species. Both the sample and species cluster analyses
follow methods described in Trueblood et aL (1994). Sample clustering uses CNESS (rn=25) as the
distance measure and unweighted pair group (UPGMA) sorting. For clustering species, we used
Pearson’s r of the normalized hypergeometric probability matrix with single-linkage clustering. The
lead author distributes full documentation, source, and executable codes for COMPAH on his web
page.
FCA-H Analysis
Trueblood et aL (1994) describes the methods used to perfonn an ordination using CNESS faunal
distances. This ordination uses a principal components analysis of hypergeometric probabilities, which
is abbreviated as PCA-H. Programs to perform PCA-H are available on the lead author’s web page.
Appendix I provides background information on CNESS and PCA-H.
RESULTS & DISCUSSION
The What, Why and Where of benthic monitoring
What is monitoring and assessment?
Chapman et aL (1987a) provided this definition of monitoring:
“Monitoring consists of repetitive data collection for the purpose of determining
trends in the parameters Isici monitored.”
According to Chapman et aL (1987a), monitoring must be based on three questions:
• What beneficial uses should be protected?
• What water-quality problems have been identified in the past or at present that
need to be monitored?
-------
GAL lAGHER & Ga. ssi.E 9
• What major natural and anthropogenic factors affect the ecosystem?
The first question is difficult to answer. Chapman et aL (1987a) urge ecologists to consider which
changes ate meaningful ecologically or for regulatory purposes. O’Connor and Dewling (1986) argued
that ecological significance is different from statistical significance. An ecologically significant, or
perhaps environmentally significant, result is one that is important enough for ecologists to warn
regulators about and important enough for environmental regulators to consider regulatory action.
Using O’Connor and Dewling’s (1986) criteria, we might assess whether the EMAP-E VP benthic
degradation indices be used now to assess whether estuarine sediments ate sufficiently ‘clean’ for
oceanic dredge disposal.
Green (1979, p. 68) divides the broad field of ecological survey sampling into three categories:
baseline studies, monitoring studies, and impact studies. A baseline study is a sampling program that
determines the present state of the system (e.g., estimates of biological and chemical variables). An
impact study assesses the effects of an impact such as an oil spill. In a monitoring study, the goal is
merely to detect change from the present state. Baseline data must be available in an impact study to
provide a standard against which to detect a change.
One of the goals of the EMAP-E VP program is not only to establish baseline monitoring data, but also
to measure a wide variety of habitat and sediment pollutant variables. The habitat factors include
water depth, temperature. salinity, pH. stratification, total suspended solids, water clarity, and sediment
grain size. The pollutant variables include most of the EPA priority pollutants including heavy metals,
pesticides, PCB’s, PAH’s and pesticides. The assessment portion of the EMAP-E VP program, based
on analyses of the covariation of physical and biological variables, should allow the assessment of
which factors might be responsible for changes in community structure.
Statistics and sampling designs
A monitoring plan should be based on established principles of statistics. All variables, hypotheses
and statistical models should be specifled in advance. The use of sample statistics in the broad sense
should be an essential part of almost all monitoring plans. Unfortunately, despite token references to
the contrary, hypothesis testing using valid sampling designs is rarely incorporated in most
monitoring studies. A strong case could be made for the view that hypothesis testing need not be an
essential feature of monitoring. Just as museum collections of bird shells provided essential baseline
data for documenting the effects of DDT in the I 960s. some might feel that data collection per se has
intrinsic value. However, when funds for monitoring are scarce and the potential array of variables
that might be monitored is large. data collection without a rigorous sampling design can no longer be
justified. A monitonng program %hould attempt to link changes in the environment with the variables
that account for that change.
There are three types of error involved in an experimental or survey design. The first two are well
known: Type I and Type II error. The’c well-known statistical errors in a monitoring program involve
finding change when there is none and not detecting a change in the environment. Underwood (1981)
called a third major source of error model misspecification. Model misspecification is caused by using
an inappropriate statistical model to pertorm the analyses. Model misspecification can result in either
failing to detect important patterns in the data, or detecting patterns and attributing the result to the
wrong cause. This third source of error subsumes much of what Hurlbert (1984) has called
pseudoreplication. Monitoring should attempt to minimize this third source of error by insuring that
the assumptions of the tests being used are met, and that the appropriate covariates are either measured
or randomized out of the design.
-------
10
Type I & II error and model mlsspedfication
The following table shows the relationship between Type I and Type 1!
Null Hypothesis
True
False
Decision Based on
Statistical Test
Reject H
Type I error
Correct decision
“Science Advances”
Accept
Correct Decision
“No Advance”
Type H error
In environmental monitoring, the traditional null hypothesis is that of’no ithnnge ’ (e.g., H 1 : pi 1 = 1 . 12).
The probability of Type I error, symbolized as I, is the probability of rejecting a true null hypothesis;
its magnitude is set through the choice of the critical value of the underlying statistical distribution
against which the value of the test statistic is to be judged. Conventionally, the probability of Type I
error is set at 0.05 or 0.01, so that the odds of rejecting a true null hypothesis are only I in 20 or 1 in
100. The choice of a significance level is merely convention, and there is justification for choosing a
relatively large a-level (e.g., Probability (Type I error)’O. 10) for environmental monitoring studies.
Committing a Type II error by accepting a false null hypothesis of no change in the environment may
have serious regulatory consequences. For example, the depletion of atmospheric ozone is of such
immediate world-wide concern that sampling programs for Antarctic ozone levels should be designed
to minimize the probability of Type II error. If the null hypothesis is ‘the percentage of benthic area
containing few animals, e.g., less than 1000 per square meter) is not changing, the environmental
consequences of a large Type II error could be extremely serious. For a given sample size, increasing
the probability of Type I error (e.g.. testing at the a=0. 10 rather than a=0.05 level) leads to a decrease
in the probability of Type II error. Increasing sample size reduces the probabilities of both Type I and
Type II error. However, if sample size can not be increased, then many scientists would argue for a-
levels larger than the conventional 0.05 level to reduce the probability of Type II error.
The EMAP-E VP program is one of the few benthic monitoring programs that has included explicit
power analyses in its statistical summaries. The EMAP-E VP program is designed so that a 2% annual
change in the percentage of area classified a degraded can be detected with high statistical power over
a 10-year period.
Trend analysis
Chapman et aL (1987) state that the goal of monitoring is not merely to detect a change in
environmental variables, but a trend. Trend usually implies a non-random temporal pattern in the
data. It would be possible to take enough samples so that a change in an environmental variable could
be detected between each and every sampling period; roughly half of these changes would be positive
and half negative. It is not sufficient to merely find significant results in a monitoring program. If that
was the goal, by increasing the sample size and program cost, a monitoring program could detect even
slight changes in variables. If there was no long.term trend in the variable, half of these significant
changes would be positive and half negative.
-------
GALLAGHER & GRAasLE 11
As O’Connor and Dewling stress, applied ecologists must move beyond minimizing Type I and Type
II error and determine what magnitude of change is significant ecologically. Year-to-year fluctuations
about the mean can be detected with a large enough sample size (Le., more replicate samples), but
these fluctuations are not good indicators of the long-term changes in the ecosystem. However, the
demonstration of a significant trend in an environmental variable cannot be achieved by simply
increasing the number of samples and solves this dilemmn Increasing the number of replicates will
not produce a trend if none exists.
The EMAP-E VP program is designed to detect trends in environmental degradation over the decade
time scale. The EMAP-E VP sampling program has been curtailed after only four sampling periods.
The detection of a significant temporal trend in a variable (at an a-level of 0.05) requires at least six
sampling periods. The probability of observing five straight increases (or decreases) in an
environmental variable sampled six times is (0.5) or 0.03125. In order to be assured of detecting a
trend, far more sampling periods are needed. In trend analyses there is one null hypothesis, ‘No trend’,
but there are many alternate hypotheses:
• A short-period cycle plus a unidirectional trend.
• Umdirectional trend.
• Unidirectional trend confounded with a long-term cyclic trend.
U Cyclic trend.
Distinguising between increasing or decreasing trends and cycles requires long time series. Nichols
(1988) discovery of a long-term cyclic trend with an apparent twenty-year period in the benthic
infaunal community structure of the 200-rn main basin of Puget Sound should give pause to any
benthic ecologist who assumes that a 4-year trend is due to a degradation of the marine environment.
The deepest part of Puget Sound, at 200-rn depth in Elliot Bay was first sampled by Ulf Lie in the
early 1960s (Lie 1968, Lie and Evans 1973). Fred Nichols sampled this station in the late 1960s for
his masters and doctoral dissertations (Nichols 1975). During the 1960s and 1970s, this site (“the 100-
fathom hole”) was dominated by a subsurface deposit-feeding polychaete Pectinaria californiensis.
Nichols continued to monitor this area every year throughout the 1970s and 1980s. The Pec:inaria
populations which had reached abundances of 1000 per m 2 1969, had nearly disappeared from the
main Basin of Puget Sound by 1976. The new numerical and biomass dominants were surface deposit
feeding bivalves and polychactes. Nichols (1985) thought that the changes he observed in benthic
community structure at his 200-rn station were due to degradation of the Sound environment due to
anthropogefliC pollutants (Seattle METRO’s West Point Sewer Outfall). The pattern fit Pearson and
Rosenberg’s (1978) paradigm which predicts that organic enrichment will lead to the replacement of
subsurface feeders by surface deposit feeders. Shortly after Nichols published his 20-year data set in
1985, P. cal iforniensis returned to the main basin of Puget Sound. Annual monitoring shows that
Pecnnaria abundances are now as high as they were in the early 1960s. Nichols (1988) rejected his
hypothesis of a long-term unidirectional trend (alternate 2 above) in favor of 20-year cyclic trend
driven by long-term hydrographic changes in the Sound. The flushing charactersitics of the deep
portion of Puget Sound change on a twenty-year time scale, and this pattern may have led to changes
in Larval recruitment to the site. Gray and Christie (1983) document other long-term trends in benthic
populations driven presumably by allogenic trends, especially hydrography.
Mistaking long-term trends and cycles in benthic communities for changes in pollutant Loading is an
example of model misspecification. Detecting significant ecological changes in benthic communities
is not the problem. The major problem is connecting those changes to changes in pollutant loading.
-------
12 EMAP-E VP COMMUNiTY STRUCTURE
Analysis of the EMAP-E VP Benthic Degradation Indices
The 1990,1991-1992, and 1990-1993 Benthic Degradation Indices
To date, them have been at least four different benthic degradation indices applied to EMAP-E VP
benthic community structure daxa the 1990 index (Weisberg et aL 1993). the 1991 Louisianian Index:
(Summers et aL 1992), the 1991 Virginian Province index (Schimmel et a!. 1994), and the 1990-1993
Index (Strobe! et aL 1995). Strobel et aL (1994) used the 1991 Index to summarize the 1992 EMAP-E
VP Virginian Province data. All of these indices are based on a one-dimensional discriminant
function, developed to distinguish between a selected group of degraded and non-degraded stations.
The abiotic data from the EMAP-E VP program are used to classify benthic samples into degraded and
non-degraded “reference” groups. Weisberg et aL (1993), Summers et aL (1992), Schimmel et aL
(1994), Strobe! a aL (1994), and Strobe! et aL (1995) describe the protocol for the creation of these
indices. A set of degraded stations is chosen based on low dissolved oxygen, sediment contaminant
concentrations, and amphipod survival. The absolute values of these thresholds have changed from
year to year. For example, in 1990, to be classified as “degraded” due to low dissolved oxygen,
oxygen concentrations had to be less than 0.3 mgfL at any time, or 10% of measurements less than 1
mg/L or 20% of continuous measurements <2 rngfL or less than 2 mg L for 24 consecutive hours.
Strobe! eta!. (1995, p. A-12) state that a site could be classified as a degraded test site if the bottom
dissolved oxygen was less than 2 mgfL. The 1990-1993 index used the recent Long eta!. (1995) ER-
L and ER-M values to divide samples into degraded and reference data sets. These stations were
selected from three salinity regimes (<5 psu, 5-18 psu, and >18 psu).
Reference, or non-degraded, stations were also selected. None of these stations could have significant
Ampelisca mortality, DO less than I mgIl, and no pollutants could exceed Long and Morgan’s (1990)
ER-M value. Strobel a aL (1995) used the newer Long a al. (1995) ER-L and ER-M contaminant
thresholds. In the latest 1990-1993 index, to be considered a “reference” site, no more than three
sediment contaminant concentrations could exceed the Long a aL (1995) ER-L value and none could
exceed the ER-M concentration.
Once these two groups of test stations were chosen, Student’s t-tests were used to find variables that
significantly differed between groups. In all years, the effects of habitat factors on species richnes
were evaluated. In the 1990 index, the EMAP-E VP diversity measure, Total Species per Event, was
calculated relative to the expected species richness at each salinity. A similar analysis was performed
for the 1991 index, including organic carbon as well, but salinity was not included in the final 1991
index, becasue the “salinity-normalized” variable was poorer at discriminanting degraded and
undegraded sites than the unnormalized species number. These variables, including those adjusted for
salinity, are then entered into a set of step-wise discriminant analyses to find the linear combination of
variables that best discriminates between groups. The 1990 Benthic Index from Weisberg a al. (1993)
is shown below:
-------
GAUAGHU & GaAS5LE
13
This 1990 Benthic Index was based on only
the first year of EMAP-E VP sampling. The
expected species included normalization for
salinity. Weisberg et aL (1993) found a very
strong positive correlation between the
number of species collected in a grab and
salinity. At the lowest salinities in estuaries,
only about eight to ten species will be found
per grab (Figure 3). With each additional psu
salinity, about one additional species is added.
Weisberg et al. (1993) used a three-point
running average to fit a polynomial curve to
the top of the cluster of points in a plot of total
species sampled vs. salinity.
Weisberg et aL (1993) compared their 1990
Benthic Index with Rhoads & Germano’s
(1986) Organism-Sediment Index. The OSP
is based on the photographs of the sediment-
water interface. Figure 4 shows the values of
the 1990 Benthic Index and the OSFM for a
subset of the 1990 samples. The two indices
are weakly correlated. However, if the indices
are converted to their binary “degraded-
nondegraded” form, the statistical association
is no greater than one might expect by chance
alone.
I
a.
C
C
U
a
U
a
a
0
I-
0 5 10 15 20 35 30 35
Salinity
Figure 3. Total species per sampling event (species
in oil 3 replicate grabs) is strongly correlated with
salinity. At 0 psu salinity, 8.5 species are expected
and roughly I species is added for ever, psu of
salinity. The variance increases at saiinities greater
than 15 psu. This plot is similar to Weisberg et aL
(1993, Fig. 4-2) but is based on an additional three
years of data and a much reduced list of valid axa.
The R 2 is 48.3%. 95% Confidence limits for the
mean value of the dependent variable are based on
three replicates.
Schimmel et aL (1994) tested the 1990 Biotic Index on the EMAP-E VP data collected in 1991.
Following the procedures developed by Weisberg et aL (1993), they identified a set of degraded and
“reference” sites based on pollutant concentration (pollutants > Long & Morgan’s ER-M), amphipod
toxicity, and dissolved oxygen concentration. New reference and degraded stations were added to the
list of stations used in 1990. Schimmel et aL (1994) identified thirteen new stations in the 1991
dataset as being degraded using their established criteria. The 1990 index classifed 7 of these 13
(54%) as degraded and 6(46%) as non-degraded. Based on this high rate of misclassification of
presumably degraded stations, Schimmel et aL (1994) developed the 1991 index (shown below). In
1990 Benthic Index
Weisberg et aL (1993)
BI=
• 0.011 • Perwu Eqected Species (mewv ,u.,sber)
• 0.817 a Nw,’ber of amphjpods per 0.044 - i a 2 g
• 0.671 • Percent of total abundance es bltalves
• 0.465 a Number of capiteilidsper 0.044 m2 g
• 0.577 * Average weight per Individual polychoete.
BI < 3.40 INDICATES DEGRADED.
-------
14
EMAP-E VP COMMUNiTY STRUCTURE
developing the 1991 index, Schimmel et aL
(1994) analyzed the effect of salinity on
species richness. They peiformed their
discriminant analyses using bedi a
unnorinalized species number and species
number normalized by both organic carbon
and salinity. Figure 5 plots the second order po1y4ii
S
I
equation, used to normalize simultaneously for
organic carbon and salinity (2 minor
typographic errors in the original equation on
Page B -3 of Schimmel etaL 1993 have been
corrected):
Schinimel a aL (1993, B-3) divided the
observed number of species by the expected
number of species predicted by this equation.
Schimmel a aL (1994) created a test data set
S.
I
N I
N
I
+
.:
:+
+.
+ . +
.
?
+ .
t
+
..... ....
++
+
45 4 4 .4 4 0 2 4 S S 10
S.dlm.nt Profil. Imags (OSPM)
Figure 4. The wzlue of the 1990 E fAP-E VP
benthic index vs. that calculated using Rhoads and
Germano ‘s(1986) Sediment-Profile-image
Organism-Sediment index The thresholds between
degraded and non-degraded are indkated The two
indices are correlated (Kendall’s r=0.326,
probcO.044). However, when these ordinal data are
converted to nominal “good-bad” or “degraded-
nondegraded” classes, the concordance between
indices is non-significant.
and ran a series of discriminant analyses. The TOC/salinity adjusted species richness measure was less
effective at discriminating between degraded and reference conditions than unadjusted species
number.
S
. 5
41
I
I
Expected number of species
8.25 +3.87x10 4 (70C)
-1.9x10 4 (7W )2
+ 0.784(sallnity) -0.00125 (salinity) 2
-2.031x1O 5 (7VC) (salinity).
0 2 3 4
T OsVsn CsIbsn(uWg) x i0
Figure 5 The salinity and Total Organic Carbon
norinali zatlon used by Schimmel a aL (1994) in
developing the 1991 EMAP.E biotic index is plotted
The expected number of species continues to decline,
reaching about -150 expected species at 8 x 10 ’ pg/g
(8%) Total Organic Carbon.
-------
a’ .. £ t v antV
15
Schinimel et aL ‘s (1994) EMAP-E VP 1991 benthic degradation index is shown below. The only
negative term in the 1991 benthic index is the abundance of a set of oppoitunistic species. The list of
opportunistic species is not provided.
Strobel et al. (1994) used the 1991 Benthic Index to analyze analyze the 1992 EMAP-E VP data. The
only change made in the 1992 Benthic Index was to add 0.5 so that a BI score less than 0 indicated
degraded.
Summers et al. (1993) developed a benthic index
for the EMAP-E Louisianian Province (EMAP-E
LP). Their index included a correction for the
effects of salinity on species richness
Strobel et aL (1995) reevaluated Schimmel et aL’s _________________________________________
(1994)1991 Index in their statistical analysis of all
four years of EMAP-E VP benthic data. Strobel et
al. ‘s (1995) rejected the
Schimmel etal. (1994) benthic
index. Their new 1990-1993
Index, shown at the right is based
on all four years of EMAP-E VP
sampling. It is based on a set of
thirty degraded and thirty
reference sites. They do not list
these sites in their report.
Presumably, many of these
degraded and reference sites are
the same as those listed in
Schimmel et a!. (1994).
The 1990-1993 index includes
two forms of salinity
normalization. Salinity must be
entered in psu for the polynomial
fit (e.g., 0 to 30 psu) and must be
entered in decimal form for the
tubificid normalization (0 to
0.030). Gleason’s D diversity is
LYYI isenriuc inaex
Schinvnel et a!. (1994)
•BI=
- 0.68 Mean Abundance of Opportunistic Species
+ 0.36 * Biomass / Abundance Ratio for all Species
+ 1.14 * Mean Number Infaunal Species per Grab.
R1 -05 JNDICATFS DFGRADF1)
EMAP-LP ‘s 1992 Benthic Index
DI .. 2.3841 • Proportion of Erpected Dr,ersir
- 16728 s Percent of Tubtflcfd Abundance
• 0.6683 s Percent of Bnralve Abundance.
BI < 4.1 INDICATES DEGRADED.
1990-1993 Benthic Index
Strobe! et al.(1995)
BI=
+ 1.389 * % Expected Gleason’s D - 51.5
28.4
- 0651 • Normalized tubificid abundance - 28.2
119.5
- 0.375 • Spionid abundance - 20.0
45.4
where,
% expected Gleason’s D
Gleason’s D
(4.283-0498 .sali,utyi .0.0542*saliiuty 2_O.00103 ‘salinity 3 ) * 100.
Gleason’s D
In N
S - Number of species.
N = Number of individuals.
Normalized tubificid abundance
Tubific ids - 500*e I5 •
B! 0 INDICATES DEGRADED.
-------
1
EMAP-EVP COMMUNiTY STRUCtURE
fit to a polynomial equation. FIgure 6 shows a display of the Gleason’s D diversity pa’ three grabs vs.
salinity. The lack of fit isn’t too surprising. We deleted hundreds of invalid taxa front the EMAP-E
VP data set, including perhaps one hundred taxa included in the EMAP-E VP analyses. Also, it is not
clear from Strobel at aL (1995) how they were calculating Gleason’s D. Gleason’s D can be calculated
several different ways.
z
U
a
C
0
0
N
0
Are these benthic indices adequate?
Are the benthic communities of the Virgmns n Province being properly assessed by the EMAP-E VP
degradation indices? In the four major suts iscaI summanes of tbe EMAP-E VP data, there are
statistical summaries of the pev entage of the are.a in the province that is degraded. Table 1 shows
these values with 95% confidence limits
C) 5 10 15 20 25 30 35
Salinity
Figure 6 Gleason’s D species ditersiry vs. salinity plotted with Strobe! at aL ‘s (1995)
salinity normalization used in the 1990.1993 EMAP-E VP degradation index. The solid
line is a linear regres.swn fit: the dotted line is the 2nd order polyno nianl fit from
Strobel et al. (1995). The R 2 for the linear regression is 47.2%. Note that we use only a
subset of species used by Strobe! et il. (1995) and we deleted many individuals used by
StrobeletaL (1995).
-------
fl I A#!!IVD I. I DACCVR
17
Table 1. Percentage of the Virginian Province classfied as “degraded” using biotic indices. Also
shown for the 1990-1993 data are percentages of the Province with low bottom dissolved oxygen
(2 mglL), toxicity (ainpbipod survival less than 80% of controls), and sediment contaminant levels
(any inoraanic ‘ ‘ pi , ,t 1,nin 1nt - -M. usjng Lona et aL 1995).
INDEX
LARGE
ESTUARIES
LARGE
TIDAL
RIVERS -
SMALL
ESTUAIUNE I
SYSTEMS
OVERALL
1990 Index
WeisbergetaL
(1993, p. 5-16)
20±8
46±32
23±14
23±7
l99lIndex
(Combined 1990-1991 data)
Sthimniel et aL (1994, p. 18,71)
6±7
27± 14
32± 17
14±6
,
1992 Index
Strobel et aL (1994, p. 59)
10±10
37±22
23±12
14±6
1990-1993
Summary
Strobel et aL
(1995)
BloticindexcO
18±4
33±14
35±6
23:3
Bottom DO2
mglL
6±2
10:6
0.2±1.3
5±2
Toxicity
(<80% control)
10±3
3±4
12±6
10±2
Any analyte
(organic or
inorganic)>
ER-M
5±2
14±6
5±2
6±2
How are we to interpret the results in Table 1? First, a comparison of the final four rows indicates that
the biotic condition index is identifying degraded conditions in a much higher percentage of the
Province than the abiotic condition indicators (DO, toxicity, and sediment contaminants). Also, the
major biotic condition indices have produced significantly different estimates of the proportion of the
Province that has degraded benthos. For example, Schimmel et aL (1994) concluded using the first
two years of EMAP-E VP data that 14 ± 6% of the Virginian Province was degraded. Strobel et aL
(1995) used the same data plus two additional years of data to conclude that 23±3 % of the Virginian
Province was degraded. What could account for this large, significant increase in degradation in such
a short period? As noted by Strobe! et aL. (1995), the indices used to evaluate the 1990-1991 data are
very different from those used to evaluate the 1990-1993 data seL The definition of ‘degradation’
hasn’t changed much in any of the EMAP-E VP indices. Strobel et aL (1995) used the newer Long et
a!. (1995) ER-M levels to define degradaded stations for the test data set and noted that the new ER-M
values were higher for metals resulting in a significant reduction in the percent area of the Province in
exceedence. The differences in the estimates must be based on the differences in the equations used to
detennine degradation. The success of the EMAP-E program must be judged, in part, on the accuracy
of the biotic indices. Is nearly one quarter of the Province’s benthos degraded, even though no more
than 8% of the Province has contaminant levels in excess of Long et aL ‘s (1995) ER-M values?
There are major problems with the three major benthic indices developed in the EMAP-E VP program.
We list the problems and discuss them in the following subsections. These problems are:
-------
18 EMAP-E VP COMMUNiTY STRUCIURE
The validation problem. The indices appear to work with the test data sets on which they are
based, but fail when new sets of degraded and inference stations are added.
The “good-bad” dichotomy. The degradation indices am based on a linear discriminant
function derived from test data sets that are supposed to represent degraded and reference
benthic communities. Some of the degraded benthic stations may not be degraded and some
of the reference benthic stations may be impacted. The EMAP-E VP benthic degradation
indices begs one of the major questions that the entire EMAP program was designed to assess:
Is impacted benthic community smicwre a result of pollution?
Inconsitencies with basic benthic ecology. Many of the variables used in the EMAP-E VP
degradation indices are inconsistent with basic principles of benthic ecology.
Inadequate data and documentation. All but the 1990 benthic index are inadequately
documented. The list of opportunistic taxa used in the 1991 and 1992 biotic indices is not
provided. One of the three terms in the 1990-1993 biotic index is based on a taxon which was
not identified in most of the EMAP-E VP data.
The validation problem
The first problem, noted in EMAP-E VP reports, is that the indices developed for one year of data
misclassify too high a percentage of new stations that are added in subsequent years. The 1990 index
failed to discriminate between inference and degraded sites sampled in 1991. The 1991-1992 indices
failed to successfully discriminate between reference and degraded sites identified in the full four-year
EMAP-E VP data. The only EMAP-E VP index that was used in more than one annual statistical
report was the 1991 EMAP-E VP biotic index (Schimmel et aL 1993). This index was used by Strobel
et al. (1994) to describe the 1992 EMAP-E VP data. Strobel et aL (1995) rejected the 1991 index in
their analysis of all four years of EMAP-E VP data. The only index that has not been rejected within
the EMAP-E VP program is the Strobel et aL (1995) 1990-1993 biotic index.
The “good-bad” dichotomy.
The major weakness in the EMAP-E VP macroinfaunal analyses is the reliance on the assumption that
benthic samples can be unambiguously classified using only two classes: degraded and reference. A
review panel, convened by the Estuarine Research Foundation (Schubel et aL 1992, p. 5), raised this
concern at an early stage of the EMAP-E program, stating that the EMAP-E program needed to
modifly “the ‘black’ and ‘white’, ‘good’ and ‘bad’, binazy characterization of
ecological/environmental conditions.” This review panel concluded:
“We are concerned that EMAP-E may have unecessarily compromised its ability to
achieve its goals by an overly simplistic (binaiy) approach to defining environmental
quality as either “good” or “bad.” Environmental quality is a continuum and
society’s definitions of “good” and “bad”, “acceptable” and “unacceptable “,
“nominal” and “subnominal” may -- and indeed do — change as knowledge of
natural conditions increases, as management approaches and philosophies become
more sophisticated and as society’s priorities change.” Schubel et aL (1992, p. 28)
The EMAP-E VP degraded vs. reference dichotomy is poorly defined. An EMAP-E VP site is
degraded if it is more similar to a set of degraded sites than it is to a set of reference sites. This
similarity is based on the value of three variables in the latest 1990-1993 benthic index.. The EMAP-E
VP program selected degraded and non-degraded reference sites using three criteria: dissolved oxygen
concentration, sediment contaminant level, and short-term amphipod survival. As Table I shows, even
-------
GALIAGUER & GRAasLE
though abiotic indices (ER-M, DO, and amphipod survival) were used to define the “degraded” test
data set, the biotic indices identified a much larger percentage of the Virginian Province as being
degraded than the abiotic indices. The major reason for this is the black-white dichotomy inherent in
the linear discriminant function. The 1990-1993 index chose an equal number of impacted and
reference sites for the test data set (Strobel et aL 1995, p. A-12) and the classification of other sites was
based on their ‘nearness’ to one of these two endpoints. If only 10% of the Virginian Province had
pollutant concentrations or low dissolved oxygen that could impact benthic communities, and yet 50%.. -
of the test data set was composed of these stations, then the discriminant function would tend to
classify too many sites as being degraded even if the assumptions of the discriminant analysis were set.
For example, a discriminant analysis could be performed with one group consisting of thirty NBA All-
Stars, and another group of equal size drawn randomly from the general population. The EMAP-E VP
discriminant function was based on twenty-eight variables measured from the two groups (Weisberg er
a!. 1993, p. 4-17; Schimmel et aL 1994, p. B-3). If a similar number of variables were measured on
the NBA All-Stars and on the general public, a linear discriminant function would undoubtedly
identify some of the properties necessary to play in the NBA (e.g., height, jumping ability, reflexes,
and big hands). This discriminant function could then be used to divide the US population into two
groups: NBA caliber, and not-NBA caliber. It is obvious that such a good-bad approach would
produce a tremendous overestimate of the percentage of the US population that might be of NBA
caliber. A similar problem may exist in the EMAP-E VP linear discriminant function. If the degraded
stations are greatly overrepresented in the test data sets, then the function will overestimate the
percentage of area that is impacted. The EMAP-E VP investigators can set the a priori Bayesian
expectations for the classification frequencies expected from the discriminant analysis to reduce the
probability of misclassification, but there is no documentation that they have done so.
The major assumption of discrirninant analysis is the equality of variance-covariance matrices. What
does this mean? It means that the variables used in the discriminant analysis should have the same
scale of variation in the groups being discriminated. Discriminant analysis is designed to classify by
differences in the mean values of discriminating variables, not their variance. Many of the variables
used in the EMAP-E VP indices clearly violate this assumption. In the 1990-1993 index the first term
is expected Gleason’s D. adjusted for salinity. As shown in Figure 6, the variance in this species
richness index will be much higher in the high salinity areas of the EMAP-E VP province.
There is another senous problem with the existing EMAP-E VP indices. There is no independent
evidence that the degraded and reference groups in the EMAP-E VP data are really degraded or non-
degraded. Sediment contaminant level is judged by ER-M and ER-L concentrations (Long & Morgan
1990, Long eta!. 1995). These ERM concentrations, which represent the concentrations at which
50% of the studies demonstrated some adverse biological effect are tabulated for a wide variety of
organic and inorganic pollutants. There was little theoretical justification for these ER-M and ER-L
concentrations. Long and Morgan (1990) and Long a aL (1995) determine their ER-M values by
simply sorting existing environmental and biological data to find the median pollutant concentration at
which any biological effect was observed. They considered a number of biological effects, including
changes in benthic community structure. Thus, only half the samples containing a given pollutant at or
above the ER-M level showed some kind of biological effect indicative of degradation. Long and
Morgan (1990) did not assess the covanation among environmental variables or formally analyze the
causal connection between a given pollutant variable and the biological effect observed. Other
branches of the EPA and state agencies have been reluctant to codify the ER-M levels into regulations.
DiToro eta!. (1990. 1991, 1992) review alternate approaches for establishing sediment pollutant
criteria Strobel a aL (1995, p. 48-51) provide a nice summary of alternate approaches to estimated
sediment contaminant levels. If the sediment acid volatile sulfide concentrations or organic carbon
concentrations are high. a metal in excess of the ER-M level may have little significant biological
-------
20 EMAP-E VP COMMUNiTY STRUCTURE
effect Thus, only about 50% of the sites in the Virginian Province with pollutant concentrations in
excess of ER-M are predicted to have significant biological effects, and this percentage might be lower
if the SEMIAVS ratios are used (Strobel etaL 1995, p. 48).
The EMAP-E VP program does not provide an assessment of degraded community structure due to
pollutants independent of the abiotic indicators. One of the stated goals of the EMAP program in all
ecosystems was to develop indices of impacted community structure or biological effects independent
of toxic loads or abiotic stressors. Ecologists were to identify samples or areas showing impacted
patterns of community structure, and the assessment portion of the EMAP program was to determine
whether these impacts could be due to pollution. The EMAP-E VP program begged this vital
question, by assuming that sites with sediment contaminants in excess of ER-M, low dissolved oxygen,
or significant amphipod toxicity, must have degraded benthic communities. The EMAP-E VP
program did not develop an independent set of criteria to determine whether the patterns of community
structure in the degraded sites were indeed degraded. It is entirely possible, and likely, that sites in the
Virginian Province could have multiple sediment contaminants in excess of the ER-M thresholds and
yet have no significant departures from ‘reference’ patterns of benthic community structure. The
contaminants might be unavailable (e.g., bound to sulfides or organic carbon), and dissolved oxygen
concentrations of 2 mg L in the overlying water might be more than adequate for benthic respiration.
More importantly the EMAP-E VP investigators chose “reference” sites using dissolved oxygen
concentration, ER-L concnetrations and amphipod toxicity. Again, toxicologists and environmental
regulators, including the EPA have been reluctant to embrace the Long and Morgan ER-L levels as
indicators of ‘clean’ sediments. The dissolved oxygen concentration one meter above the bottom does
not mean that the sediments are adequately oxygenated. In fact, due to the physics of the benthic
boundary layer, anoxic sediments can be associated with very high oxygen levels even centimeters
above the bed. Finally, benthic communities with adequate dissolved oxygen one meter above the bed
and low pollutant concentrations can still have patterns of community structure indicating recent
disturbance or impact. Physical disturbance by other processes — storms, predators, disruption by
fishing nets, and red tides, drifting macroalgae — can all produce patterns in community smicture that
indicate recent disturbance. Many of these patterns of disturbance mimic the effects of pollution. For
example, the frequencies of opportunistic taxa increase, species richness declines, species evenness
declines, and infaunal abundance levels change. The EMAP-E VP program developed no independent
criteria to detect these ‘natural’ impacts.
Inconsitencies with basic benthic ecology
The third problem with the indices is that they are inconsistent with basic benthic community ecology.
The 1990 index (Weisberg e: aL 1993) predicts that the more capitellid polychaetes, the more likely a
site is to be classified as non-degraded. Members of the genus Capitelia spp. are generally regarded as
pollution indicators (e.g., Grassle & Grassle 1974, Pearson and Rosenberg 1978, Grassle and Grassle
1985). Grassle and Grassle (1976) showed that the species formerly known as Capuelia capitata is a
sibling species complex. Most if not all of these sibling species are found in areas high concentrations
of utilizable organic carbon. However, the genus Capiteila is rare in the EMAP-E VP data. Capiteila
is found in only about ninety of the nearly two thousand EMAP-E VP samples. Moreover, it is not
particularly abundant in any of those samples.
One reason why ‘capitellid polychaetes’ may have been important in the 1990 index is that the
capitellid Mediomastus anzbiseta is the most abundant and wide-spread taxon in the entire Virginian
Province. Mediomastus ambiseta can be found in salinjues ranging from 10 psu through 35 psu. Diaz
and Schaffner’s (1990) summarized the benthic species characteristic of different salinity and grain
-------
GALL4CHER & Ga. ssLE 21
size habitats in Chesapeake Bay. They found that M. ambiseta reaches peak abundance in high
mesohaline (10-18 psu salinity) muds and mixed mud and sand habitats. However, M. ambiseta is
often the numerical dominant in high salinity coastal waters. Fuller et aL (1988) found M. ambiseta
was the numerical dominant in Buzzards Bay, reaching abundances of 720,000 m 2 in outer New
Bedford Harbor. This species would be a poor choice as an indicator of pristine, unpolluted sites.
Grassle & Grassle (1985) review the use of M. ambLseta as a pollution indicator. Grassle eta!. (1986)
observed M. a,nbLceta increase dramatically in response to eutrophication. Grassle ef aL (1988) added -
sewage-sludge to the MERL tanks and again observed substantial increases in the abundance and
frequency of M. ambLceta. Grassle eta!. (1981) observed that M. ambiseta was the taxon that declined
most sharply with the addition of 90 ng/g #2 fuel oil in the MERL ecosystem tanks. Using M.
wnbise:a as an indicator of high petroleum hydrocarbon concentrations in sediments is unwise, since
M. ambiseta increased dramatically in abundance in heavily oiled offshore stations after the West
Falmouth oilspill (Grassle and Smith 1976, Sanders et aL 1980). It is very difficult to provide any
ecological basis for assuming that more capitellids indicates a lower probability of degradMion.
There are two major problems with the Schimmel et aL (1994)1991 biotic index. They used the
mean number of species per grab as a discriminating variable. As Weisberg eta!. (1993, Fig. 4-2)
showed, salinity has a profound effect on species richness. We show the strong effect of salinity on
species richness in Figure 3 (p. 13) Any index based on species richness which fails to take salinity
into account will classify many non-degraded low-salinity sites as degraded. Strobel et aL (1995, p. A-
12) noted this, stating that the 1991 Index is “highly correlated with salinity and appared to misclassify
good sites in the oligohaline [ <5 psu] and impacted sites in the meso- [ 5-18 psu] and polyhaline [ >18
psu].”
The 1991 and 1992 biotic indices failed to incorporate a correction for the effects of salinity on species
richness because they combined salinity with total organic carbon concentration (TOC) in the
normalization. This was not a proper way to assess the covariation of salinity and diversity. Figure 5
(p. 14) shows the expected species at different TOC concentrations predicted by that function at three
different salinities (1 psu. 18 psu. 30 psu). All EMAP-E VP samples with greater than 3.5% TOC are
expected to have less than 0 species. Dividing observed by expected species number can produce high
adjusted species richness values in the range 3 to 3.5% or negative species richness numbers if TOC is
greater than 3.5%. Simultaneously normalizing TOC and salinity together was not a good idea.
Total organic carbon concentration in sediments is one of the best indicators of both contamination and
eutrophication. It is not a natural environmental factor like salinity, temperature, or depth. Wallace et
a!. (1991) have shown that sediment total organic carbon concentration is tightly coupled with
sediment heavy metal concentrations in Boston Harbor. The correlation is often 0.9 or higher. In
Boston’s Inner Harbor. sediments ‘ ith TOC concentrations of about 3.3 % are either anoxic in the late
summer or contain about six species per grab. Figure 5 (p. 14) shows that these sites would have
100% or more of the expected species nchness expected at these organic carbon concentrations.
Strobel et aL (1995, Figure 3-27) shows that TOC in the Virginian Province ranged from 0 to 7%, with
roughly 10% of the Province having TOC concentrations greater than 3% organic carbon. By not
adding a proper normalization for salinity, the 1991 and 1992 indices would be highly likely to classify
non-degraded oligohalme sites as degraded, as Strobel eta!. (1995) noted.
The 1990-1993 Index ,shown on p. IS. was developed by Strobel et aL (1995). This index has three
terms: salinity-normalized Gleason s D cpecies richness, spionid abundance, and salinity-adjusted
tubificid oligochaete abundance. We will discuss each term of the equation.
-------
EMAP-E VP COMMUN Y STRUCTURE
The 1990-1993 index, like Weisberg at aL’s (1993) Index, normalizes species diversity by salinity.
Species diversity is calculated using Gleason’s D (Gleason 1922). We show the relationship between
Gleason’s D and salinity in Figure 6 (p. 16). Washington (1984, p. 661) included Gleason’s D for
historical purposes only, noting that it lies not been used extensively in the recent literature, being
replaced by Maiplef’s.” Gleason’s D has some very poor statistical properties (as does Margalef’s
index). Trn gine a situation where all of the individuals in the sediments are independently Poisson
distributed (Le., randomly placed). Gleason’s D, the number of species divided by the natural log of
the number of individuals, is one of the only diversity indices listed in Washington (1984) that declines
as larger samples (either by area or numbers) are collected. Margalef’s index also declines with
increasing sample size. Most other indices either remain the same (H, H’, E(S)) or increase (total
species) with increasing sample size. Even with a fixed sample area, this sample-size dependence is a
serious problem. In two grabs containing an identical number of species, with identical frequencies,
Gleason’s D will find that the sample with the lowest abundance has the highest diversity.
The final two terms of the 1990-1993 index
are based on spiomd abundance and wbificid
oligochaete abundance, the laner scaled by
salinity. It is difficult to assess this index
using the existing EMAP-E VP publications.
Oligochaetcs were never identified below
Class Oligochacta in samples collected from
areas where bottom salinity was greater than 5
psu. We are assuming that Strobel et a!.
(1995) are following Weisberg et aL (1993, p.
2-12) in assuming that most oligochactes in
habitats with salinities >5 psu are tubificids.
This is not true, but given the assumption that
most marine oligochaetes are tubiticids, we
can calculate how many tubilicids and
spionids per m 2 are sufficient to produce a
classification of “degraded” if the Gleason’s D
diversity is 100% of the expected value.
Figure 7 plots this relationship. If a sample
has 100% of the expected Gleason’s D
diversity, then observing more than 70(X)
spionids per m 2 at 30 psu salinity indicates
that a sample is degraded. Observing 170(X)
tubificids per m 2 at 30 psu indicates
degradation. Fewer tubificids are required to indicate degradation at lower salinities.
Spionid polychaetes are a species-rich succesful family of polychacte worms. There are a handful of
spionid polychacte species that have been used as pollution indicators, most notably Streblospio
benedicti and Polydora cornutu. However, both of these species are natural components of shallow or
mesohaline habitats in the Virginian province. Dauer et aL (1981) reviews the factors controlling the
distribution and abundance of the six major spionid species in Chesapeake Bay. Members of the
polychaete family Spionida are among the most important members of the Chesapeake Bay benthos in
all sediment types and most salinities. The family Spionidae cannot be used as a pollution indicator.
Spionids are among the most abundant surface deposit and suspension feeding organisms in habitats
ranging from the intertidal zone to the deep sea. The abundance of individuals belonging to this
polychacte family does not indicate either pollution or disturbance. The 1990-1993 index sets a 5000-
cli gochaetes and spionid polychaetes necessary to
equal 100% of the expected Gleason’s D diversity at
salinities from 0:035 PSU (plotted in Fig. 6, p. 16).
A sample with 100% Gleason ‘sD diversity that plots
to the right or above this veil would be classified as
degraded. A: 30 psu, 7000 spionids per m 2 or
17.000 tubificids per m 2 indicates degradation.
1 1000
4o
0
Figure 7 The 1990-1993 EMAP-E VP index (p. 15)
was recast to predict the number of tubçficid
-------
GAUAGHER & GL4SSLE
23
7000 spionid per m 2 threshold for degradation (Fig. 7) and would classify many pristine areas as
degradeci
Inadequate taxonomic data and documentation
One of the three terms in the 1990-1993 Index is based on tl abundance of tubificid oligochaetes.
This family of organisms was never identified in samples collected in areas with more than 5 psu. The
only variable in the EMAP-E VP for these samples is the number of individuals of Class Oligochaeta.
In evaluating this index, we assumed that Strobel etaL (1995) followed Weisberg eta!. (1993) in
arguing that most oligochactes in marine waters arc tubificids. This assumption may not be true. If
only individuals properly identified as tubificids are used, then all marine samples automatically have
o tubiflcid abundance and after scaling, the second term in the 1990-1993 benthic degradation index
(p. 15) becomes positive. We would not recommend basing one of the three terms of the biotic index
on a variable that was not measured in the majority of stations.
It is difficult to evaluate the 1991 index, because it is based on the abundance of opportunistic
polychactes. The list of species considered to be opportunistic is not provided in any of the EMAP-E
reports.
Community Structure Analyses of Virginian Province
The 1918 samples in the full EMAP data set were analyzed using COMPAH and PCA-H. A subset of
the full data set was created. This subset contained all samples for which salinity data existed and
which had infaunal abundances greater than 25 individuals. Replicate grabs from each sampling event
were summed and used only if the maximum CNESS distance among the three replicates was less than
0.7. We used only base sampling sites represented by three replicate grabs in this analysis. A CNESS
value of 0.7 indicates tremendous differences in community structure. Only 371 ‘sampling events’ in
the full EMAP data set met both the CNESS cutoff and salinity criteria.
Diversity analyses
In this section we will evaluate the
correlations among diversity indices using the
EMAP-E VP data, and the covariation of
diversity with salinity. Of the diversity
indices that might be used in the EMAP-E VP
program, the two with the richest body of
theory are Shannon’s H’ and Huribert’s
expected number of species E(SR). E(SI ) is
the expected number of species if 10
individuals are drawn at random from a
sample. Peet (1974) and Smith et aL (1979a)
showed that Huribert’s E(S 10 ) is highly
correlated with the Shannon’s H’ diversity
index. Figure 8 shows this relationship.
Either E(S 10 ) or H’ would provide an
assessment of diversity of EMAP-E VP
samples. However, both of these indices are
I
t
a,
S
C.
U.
4 6 6
Hialbsit’s !(S ,,)OIvursIty
Figure 8. Ata random size of 10 individuaLs, the
Sanders-Hurl bert expected number of species E(S,,)
is highly correlated with Shannon’s H’ diversity.
The above shows the association with EMAP-E data
(all 4 years, replicates combined). The R 2 is 97.8%.
The 95% confidence limits are based on 3 replicate
1 2 3
7
$
a
grabs.
-------
24
EMAPEVP COMMUNiTY STRUCTURE
sensitive to the evenness component of
diversity and neither one is particularly
sensitive to species richness . At a larger
random sample size, E(S ) becomes
increasingly sensitive to species richness. The
relationship between E(S ) arid total species S
per event is shown in Figure 9. A nearly identical
paLiern (but with a different scale) is found
when Shannon’s H’ is plotted versus total
species. E(S ,) is corelatcd with Total species Z
per event (Figure 9), but the two diversity
indices are obviously measuring different
components of species diversity. It is possible
to have a sampling site with over 60 species
collected in 3 grabs that would have a species
diversity much lower than the median
(E(S 10 )<5). This pattern would be expected if
one or a few species in&le up most of the
individuals in a sample.
Neither E(S 10 ) nor Shannon’s H’ are strongly
correlated with Gleason’s D diversity (Figure
10).
We have performed an analysis that shows at a
glance the correlations among the major
diversity indices and Huribert’s E(Sa) With
increasing sample size. The diversity of the
371 EMAP-E VP event data set was analyzed
using four different diversity indices. The
nonpararnetric correlation between the ranked
diversity was compared with the E(Sa)
diversity with n ranging from 2 to 200. E(S,)
is Simpson’s diversity+1. At larger random
sample sizes, E(S 1 ) becomes more strongly
correlated with both Gleason’s D and total
number of species per sampling event (Figure
11). Figure II clearly shows thatthere is no
universal index of diversity. Gleason’s D
clearly falls in the class of indices that are
more heavily influenced by species richness.
S
$
0 40 80
Total sp.cI.s In 3 grabs
Figure 9. Sanders-Hurlbert diversity E(SI) at n=1O
is only weakly correlated with species per sampling
event (R =23.1%). The 95% confidence limit is
based on 3 replicates.
-a -
1 2 3 4 5 6 7 6 9
Hurlbsrt’s E(S,)
Figure 10 Gleason’s D species diversity vs.
Hurl bert’s E(S, 0 ). The 95% confidence limits for 3
replicates are shown.
Shannon’s H’ is sensitive to species richness, but it is also strongly influenced by species evenness. In
a later section we will review the effects of pollution on diversity. Pollution and disturbance can affect
both the richness and diversity components of diversity. One viitue of analyzing the effects of
pollution in the EMAP-E Virginian Province using the evenness component of diversity is that it
shows virtually no statistical association with salinity (FIgure l2).
Sanders-Huribert E(Srn) is positively correlated with salinity (FIgure 13), but this relationship is much
weaker than either the correlation between total species and salinity (Figure 3, p. 13) or Gleason’s D
0
a
a
.
a
S
.2
0
++
++
+
++
+
+
+
++
+
+ +
++
+ +
+ + +
+
+
.‘
+
+
+
+
12
+
+ +
a
++
4
+
+
+
I -
+
-------
GALLAGHER & GR LE
25
and salinity (Figure 6, P. 16). At this random sample size, E(S 10 ) is sensitive to both the species
richness and evenness components of diversity.
C
C
I.
Ii
0
.
0.
I-
C
C
.
J.
200
Figure 12. The nonparametric Kendall’s rcorrelation
between Sanders-Huribert diversity E(SJJ at various
random sample sizes, n, is plotted versus other diversity
indices. Brillouin ‘s H (not shown) plots just below the
line for Shannon’s H’. E(S , ) is the largest n shown. At
small n, E(S,) is sensitive to both species richness and
evenness. At high n, E(S,) is more strongly associated
with species richness. E(S,) is highly correlated with H’
at n=1O (see Figure 8).
Salinhy
Figure 11. There is virtually no statistical association
between species evenness, as estimated by I’, the
evenness measure for Shannon’s H’, and salinity
(R 2 =O.2%).
E(S 3 . .)
I %
Gisason’s D
H —
III .II._ I1. *
— — — —————4.
- — —. — — — — — E(S, )
I
I
40 80 120 160
Random sampi. siz. (n) for E(S,)
-------
26
EMAP.EVPCOMMUNTrY STRUCflJR
.
w
0
Salinity
Figure 13. E(S, 0 ) is only weakly correlated with salinity.
Cluster analysis
Sample clusters
The large sample cluster analysis of all four years of EMAP-E VP data is found in Appendix IV. The
major pattern in the data set is the clear break between all samples taken in the 0-5 psu salinity regime
from the other mesohaline and oligohaline samples (5 to 35 psu). There is a further break among the
higher salinity samples corresponding to a salinity of roughly 15 psu (not the 18 psu used in the
stratification of degraded and clean reference sites in the creation of EMAP-E VP degradation indices).
There is a considerable amount of within-site, among-season and among-year variation in the EMAP-E
VP data. In Appendix IV, two degraded and two clean reference sites are colored to show the extent
of this variation. On a qualitative level, there appears to be as much variation among samples taken at
the same site during different seasons (e.g., New Bedford Harbor 099 sampled on August 15, 1990
and September 4, 1990) as there is between any difference between degraded vs. clean (using the
criteria in Weisberg et aL 1993 and Schimmel et aL 1994). There appears to be as much variation at
New Bedford site 099, sampled during the same year but 3 weeks apart, as between any pair of
estuarine stations (salinity >5 psu) in the entire data set. It was impossible to analyze the full extent of
within site variation relative to the degraded vs. non-degraded dichotomy since most of the ‘degraded’
samples, especially in the oligohaline and mesohaline portions of the Virginian Province, had fewer
than 25 individuals per grab and were dropped from the analysis.
Species clusters
Clustering of species reveals three distinct groupings of species (Figure 14), roughly corresponding to
the oligohaline, mesohaline, and euhaline habitats in the Virginian Province. All species that
contributed at least 0.5% of the variation in the EMAP-E VP CNESS distances among stations are
0 5 10 15 20 25 30 35
-------
GAU.AGHER & GRASSLE 27
shown. Within each of these large salinity-controlled groupings ale groups that are characteristic of
different grain sizes, biogeographic regions, and depths.
The low salinity species assemblage in Figure 14 consists of two sub groupings. The first subgrouping
(TUBLFWI to cHIKONOM) axe mainly taxa which could only be found in samples from 0-5 psu
salinity, since they involve taxonomic designations not used for samples taken from areas higher than 5
psu salinity. The most important contributor to CNESS distances in the 1990-1993 EMAP-E VP data
is the taxon TUBIFIWI, which is the EMAP-E VP dode for Tubificidae with capiliform chaetae
(see Appendix 11 for translations for the EMAP-E VP species codes). This group probably does not
represent a single species. The fifth most important taxon is LIMNHOFF, Limnodrilus hoffineiswri.
TUBWWO, the 28th most important contributor is another composite taxon; the EMAP-E VP species
code stands for Tubificidae without capiliform chaetae.
The oligohaline species group also includes some very important macroinfaunal species.’ 1he spionid
polychaete Marenzelleria viridis (MAREVIRI), formerly called Scolecolepides viridis, is the eighth
most important contributor to CNESS distances in the Virginian Province. It tends to occur with the
amphipod Leptocheirus plumulosus (LEPTPLUM) and the isopod Cyathura polita (CYATPOLI).
These oligohaline to low mesohaline taxa tend to be geographically wide-spread in the Virginian
Province. These latter three taxa are also abundant throughout the Gulf of Maine region (EMAP-E
VP’s Arcadian Province).
-------
EMAE VP COMMUNITY STRUCTURE
Figure 14. auster analysis of the 48 species that contribute most to CNESS (m=25) variation.
-p
— aa_
______ 1
as
. aa
U
30
U
n,m I
CThTI .Z 13
31
. —-——- — as
30
43
as
— a
. — 47
a _ ii
.3c1 Z 15
13 101 15
33
. ——‘ 30
53
— • — ‘ 10
— . 37
L472 40
P xO47fl 35
ai I 30
T J.5IXL 13
I AIIPA 37
• — — U 31
1, I 33
P3 41
47 3
4
?T1XL 43
LZT 17
P 1 15
34
IAIV S
!QSPA 43
T.ZZAT1 7
. — 33
.. .Z 14
an .
—ma
as
10
Pearson’s r
0.I 5.73 4.13 0.33 4.33
-.47
-.4’
OLIGOHALINE
TAXA
MESOHALINE
TAXA
a’
I I
-------
GALIAGHE& & GRA&SLE 29
The euhaline species assemblage, shown in FIgure 14, consists of two subgroupings. The most
impoitant species in the first grouping and the second most important taxon contributing to CNESS
faunal distances in the entire Virginian Province is the capitellid polychaete Mediomastus ambiseta
(MEDIAMBI). This head-down subsurface deposit feeder is also the most abundant taxon in the
entire Virginian Province.
Included within the second euhaline species assemblage is the Nucula annzdata (NUCUANNU),
Nephzys incisa (NEPHINCI) group. These two taxa, tenth and thirty-seventh most important
contributors toCNESS distance, are the classic indicator species for both Long Island Sound and
Buzzards Bay infaunal communities (Sanders 1956, 1960). Sanders (1956), in his Yale Ph.D.
dissertation, described the Long Island Sound benthos using the Petersen-Thorson descriptive
designations as a Nucula proxima - Nephtys incisa community. Buzzards Bay was described as being
a Nucula proxima - Yotdia limatula - Nephtys incisa community. Yoldia limatula (YOLDLIMA), a
larger but less abundant protobranch bivalve, was not among the top forty-eight species contributing
the CNESS distances, but is the first taxon to cluster with Nucula annulata in the species cluster
analysis of all 551 EMAP-E VP taxa (Appendix V). Subsequent to Sanders’ surveys of Long Island
Sound and Buzzards Bay, Hainpson (1971) found that there are two Nucula sibling species in the
region: Nucula proxima dominates in nearshore fine sands and Nucula annulata dominates offshore
muds. The two species have nearly allopatric distributions, and can be distinguished by the locations
of the abductor muscles on the inside of the shells. Both have probably been combined in the EMAP-
E VP designation NUCUANNU since the widespread and abundant taxon Nucula proxima is not
among the EMAP-E VP species.
Within the second euhaline species assemblage are the tellimd bivalve Tellina agilis (TELLAGIL),
the cirratulid polychaete Thwyx sp. A Morris (THARSPA), the paraonid polychaete Arricidea
cathennae (ARICCATH), the ampeliscid amphipod Ampelisca verrilhi (AMPEVERR), and the
spionid polychaete Spiophanes bombyx (SPIOBOMB).
The mesohaline species assemblage, shown in Figure 14, also consists of two sub groupings. The
mesohaline taxa include some of the most abundant and wide-spread taxa in the Virginian and
Arcadian provinces. All of these taxa can be found in both the shallow subtidal zone (usually in
intermediate salinities) and the intertidal zone. The spionid polychaete Streblospio benedicti Webster
(STREBENE) is a typical numerical dominant in intertidal zones throughout the Virginian and
Arcadian Provinces (e.g., 1982, Trueblood et aL 1994, Diaz and Schaffner 1990). This spionid is the
third most important contributor to the variance in CNESS distances among Virginian province
samples. In the subtidal, S. benedicti dominates in areas of intermediate salinities ( ‘ 15-25 psu). The
EMAP-E VP taxon with the highest affinity to this spionid is the class Oligochaeta (OLIGOCHA).
This composite EMAP-E VP taxon is only used for samples having salinities greater than 5 psu, at
lower salinities the oligochaetes are further divided into species and designations such as Tub ficoides
with capiliform setae (TUBIFIWI. see above). Also associated with this Streblospio benedicti -
Oligocbaete assemblage are other taxa characteristic of the intertidal zone, but which also can be very
abundant in shallow subtidal mesohaline environments. These include the spionid polychaete
Polydora cornuta. formerly called P. higni, (POLYCORN), the orbiniid polychaete Leitoscoloplos
robustus (LEITROBU), the isopod Edozea triloba (EDOTTRIL), and the venend bivalve Gemma
gemma (GEMMGEM1 ’1). Each of these taxa is abundant in intertidal and shallow subtidal
mesohaline zones throughout the Virginian and Arcadian provinces.
Included in the first mesohaline species assemblage is the Ampehisca abdua-Ampehisca vadorum
complex (AMPEABVA). This sibling species group is the third most abundant taxon in the Virginian
province (1st in the NY/NJ REMAP data) and the sixth most important contributor to the variance in
-------
30 EMAP.E VP COMMUNITY STRUCFURE
CNESS distances among stations. Ampelisca abdita is the species used in the EMAP-E VP amphipod
toxicity assays. AmpelLsca abdita and A. vadorum constitute a sibling species complex (Mills 1967).
Since the juveniles cannot be distinguished, the adults of both species are pooled with the juveniles in
the EMAP-E VP taxon (AMPEABVA). This sibling species complex co-occurs with the capitellid
polychaete Notomastus sp. A Ewing (NOTOSPA). The opportunistic machid bivalve Mulinw
lazeralis (MULILATE) is the seventh most important contributor to CNESS distances. It is weakly
associated with the cumacean Leucon americanus (LEUCAMER).
The following mesohaline taxa form a well-defined assemblage. characteristic of slightly lower
salinities than the remaining mesohaline taxa the capitellid polychaete He:eromasuufi1 formLs
(HETEFIL1), the opportunistic tellinid bivalve Macoma baithica (MACOBALT), the nereid
polychaete Neanthes succinea (NEANSUCC), the tellinid bivalve Macoma mitchelli (MACMITC),
and members of the class Nemertinea (NEMERTIN). Each of these taxa can be found in intertidal
zones throughout the Virginian and Arcadian provinces, and thrive in lower mesohaline salimties in
the lower subtidal.
Notable for its absence from the species cluster analysis in Figure 14 is the sibling species complex
Capitella. Figure 14 contains only the most important contributors to CNESS distance, and Capitella
is a minor component of the Virginian Province communities, being found in only eighty nine samples.
As shown in Appendix V , the pollution-indicating Capiteila sibling species complex appears to be just
another relatively rare euhaline taxon. Surprisingly, the species with the highest affinity to Capi:eila in
the full species cluster analysis (Appendix V) is the poitunid crab Ovalipes oceilatu.c (OVALOCEL).
Of the forty eight taxa shown in Figure 14, Capitelia is associated most closely with the cirratulid
polychaete Tharyx sp. A Moms (THARSPA).
PCA-H analysis
The first step in the PCA-H analysis was to determine a random sample size, or NESSm, for the
CNESS faunal distance index. This random sample size should produce a faunal distance index that is
sensitive to the contribution of both rare and abundant species in the community. A NESSm of
approximately 20-25 produces an index that is highly conalated (tau > 0.8) with both CNESS
(NESSm=1 or Orloci’s chord distance) and CNESS (NESSm=l00).
Figure 15 shows the Gabriel covariance biplot of species (Gabriel 1971). This is a different way of
plotting the species data that were clustered in Figure 14. This figure shows the major species groups
in the entire Virginian Province. Only those species that contributed at least 1% to the variation in
community structure (measured by CNESS) are plotted. This figure shows the major gradient in
species distributions as a function of salinity in the Virginian province. The cluster of species vectors
at about 4 o’clock consists of species that were only identified in samples taken from less than 5 psu
salinity areas. Oligochaetes , Mulinia lateralis, and Streblospio benedicti are all characteristic of
intermediate salinities in the Virginian Province (about 15 psu). The spionid polychaete Marenzelleria
viridLs, the isopod Cyathura polita, and the aorid amphipod Leptocheirus plumulosus are characteristic
of slightly lower salinities (5-10 psu). Ampelisca abdita and A. vadorum are associated with higher
salinities. The key species indicating euhaline conditions in the Virginian province is the capitellid
polychaete Mediomastus wnbiseta. The final three species vectors represent Howard Sanders’ classic
Long Island Sound and Buzzards Bay Nucula-Yoldia-Nephtys incisa community (Sanders 1956 and
1960).
-------
CAr1ACR R & GR. LE
31
EDOTTRIL
NEANSUCC
MACOMITC
STREBENE MACOBALT
LEPTPI
OLIGOCHA MAREVIRI
NEMERTIN
CYATPOLI
LEU
MU LILA
LEITROBU
GLYCSOLI
PARAPINN
ACTEPUNC
AMPEABVA
PA
SPIOBOMB
AMPEVERR NUCUANNU
LOIMMEDU
ENT YOLDLIMA
TELLAGIL
Figure 15 This is the Gabriel covariance biplot corresponding to the species clusters shown in Figure
14. Each of the 48 moSt important species is indicated by a vector (arrow), and the cosine of the
angle between vectors is a measure of whether species are likely to be found in the same samples. A
remarkable feature of this plot is that the major species groupings in different salinity regimes can be
read sequentially by moving from 4 oclock counterclockwise to 7 o’clock. The species codes can be
found in Appendix 111. Some ke species discussed in the text are bolded.
Figure 16 shows the Gabriel Euclidean distance biplot for the 1736 samples containing morn than 25
individuals (out of 1918 total samples). The most important caxon accounting for CNESS variation
among samples is the oligochacle taxon ‘Tubificoides with capiliform chaetae’ (TUBIFIWI in Figures
14 and 15). This taxon alone accounts for 5% of the total CNESS variation among samples. The
samples in the upper left portion of this plot are all characterized by having high frequencies of
Tubificoides with capiliform chactac and Limnodrilus hoffmeisten. The relative frequency of these
key species can be determined by projecting each sample onto the species vectors at right angles.
As one moves counterclockwise around Figure 16 from the upper left, the samples are distributed
according to salinity. The key species contnbuting to the position of samples in the first two PCA-H
axes are the oligohaline isopod Cvazhura polita (CYATPOLI in Figures 14-16), the spionid
polychaete Marenzelleria viridis (MAREVIRI). the aorid amphipod Leptocheirus plumulosa
(LEPTPLUM), the capitellid polychaete Heteromastusfilifornus (HETEFILI), the spionid
(
RANGCUNE
TUBIHETE
GAMMDAIB
CHIRALMY
-------
32
EMAP-EVP COMMUNiTY STRUCTURE
polychacte Streblospio benedicti (STREBENE), and members of the class oligochacta (OLIGOCHA
- an EMAP-E VP designation used only for oligochaetes collected from samples with salinity >5 psu).
The upper right euhalinc portion of the biplot is controlled by two euhaline taxa the capitellid
polychaeteMediomastus ainbiseta (MEDIAMBI), and the gastropod Acteocina canaliculata
(ACTECANA).
c ,1
C,
a.
0.5
0
-0.5
-0.5
PCA-H Axis 1 (15%)
Figure 16. A Gabriel Euclidean distance biplo: of the full (1918 sample,
551 species) EMAP Virginian Province data at m=25. All species
vectors are plotred, only those accowuing for 2% of the variation in the
first 2 dimensions are labele€L The ten most important species
contributing to CNESS distances are Tub jflcoides with capilliform
chaerae (55), M. ambiseta. S. benediciL oligochaetes (4%), Ampelisca
abdita-vadorum complex (3%, not shown), Mullinia lateralis (3%, not
shown), Maren7elleria (formerly Scolecolepides) viridLc (3%), and
Nucula annulata (2%, not shown). The important species not shown in
this plot are important contributors to the third PCA-H axis.
The 1736 samples in Figure 16 were pooled if salinity data existed for the samples and if the CNESS
distances among replicates were less than 0.7 (a large CNESS distance, see Methods). The biplot
prooduced for these 320 samples are analyzed in Figures 17-23.
Figure 17 shows the PCA-H ordination of samples, with all samples in 5 psu increments being
surrounded with different colored convex hulls. This figure shows clearly that PCA-H axis 1 serves
largely to separate samples with salimties less than 5 psu from the remaining samples. Undoubtedly,
some of this clear demarcation among samples is due to the use of different taxonomic designations for
Acteocuia canahculata
Lknnodn!us hoffmelsterl
rubificldae with capUltorm chaetae
.
.;. .
..
‘ ! •S
_ .
. •.• S • •
‘a •.
Cyathura
polita
Marenzelleria
viridis
Leptocheirus
plumulosus
0 0.5
-------
GALLAGHER & GL ssLE
33
0
A
(3
EMAP-E VP samples taken from areas with less than 5 psu. It is also clear that there is a second, but
less distinct break in community stnkture between the 5-15 and 15-35 psu samples.
In the first two PCA-H axes, the 15-35 psu samples plot on top of each other, but they are separable in
the plot of the 2 vs. 3 PCA-H axes, which are shown In FIgure 18.
30
25
20
U
15
10
5
-0.5 0 0.5
PCA-H AxIs 1(16%)
Figure 17. Convex hulls containing all samples within 5psu salinity
ranges are plotted vs. PCA-HAxes 1 and 2. PCA-H Axis 1 is controlled
almost entirely by the contrast between 0-5 and 5-35 psu salinity.
0.5
A
S
(3
a.
-0.5
0
-0.5 0 0.5
PCA-H AxIs 2 (10%)
Figure 18. Convex hulls are drawn around all samples within each 5 psu
salinity range vs. PCA-H Axes 2 and 3. Salinity is still a major
determinant of PCA-H scores in the second and higher dimensions.
-------
34
EMAP-E VP COMMUNITY STRUCTURE
In the two dimensional PCA-H displays in Figure 17, the salinity groupings appear to form a horse-shoe
(often called a Kendall’s horseshoe) or U shape. This is the expected low-dimensional projection of
salinity-controlled coenodine. Salinity continues to control conuminity structure in the 3rd and higher
PCA-H dimensions (Fig. 18).
The strong effect of salinity on PCA-H site scores in the first two PCA-H dimensions is graphically
shown in Figure 19. The salinity of each sampling event is plotted in the third dimension vs. the first two
PCA-H site scores (shown in Figure 17). The samples shown in the contour plot above the figure are
distributed in a counter-clockwise fashion from the lower left of the plot to the upper right. The enhaline
portion of the estuarine coenodline is shown in the upper right. The two arrows in the mesohaline (10-15
psu) portion of the plot mark two Hudson River samples (VA9O-177 and VA9O-198), which had
measured salinities of 735 and 9.55 psu, but which had species compositions characteristic of salinities 3
or 4 psu higher. These two samples produced a ‘hole’ in the contour plot (marked with the red and green
contours for 15 and 10 psu, respectively). This pattern could be produced in a tidal river system when
the salinity decreased from 11 to 14 psu to 7 to 10 psu a few weeks before the benthic grab samples were
taken. Within the oligohaline and mesohailne portions of the Virginian province, species composition is
a very good predictor of salinity (but not vice versa).
The upper right euhalinc portion of Figure 19 reveals a rugged mountainous topography. Samples within
this region have species compositions characteristic of salinities greater than 15 psu, but there isn’t a
clear one-to-one correspondence between salinity and community structure.
Figure 20 shows the total number of species per sampling event vs. PCA-H axes 1 and 2. At salinities
less than 15 psu, the number of species per sampling event is closely coupled to salinity. At salinities
greaterthan lspsu,thereisagreatdealofvariationinthetotalSpeciesperSamPlingeveflt.
Undoubtedly some of this variation is due to the effects of disturbance, and some is due to depth, grain-
0.5
0
-0.5
.- r’-..
0.5
a
4
-0.5 pGk
-1 -1
Figure 19. Salinity is plotted against PCA-H Axis 1 and 2, showing that salinity
is a major factor controlling community structure. The arrows on the right
indicate the two Hudson River samples discussed in the text.
-------
GALLAGHER & GRASSLE
35
0
iO
6O
0
I!
U
.
I
0.5
Figure 20. Total number of species per sampling event (3 grabs) is plotted
against PCA-H Axes I and 2. Total species in 3 grabs in the oligohaline
habits (lower left, see Figs. 17 and 19) is uniformly low. Total species in 3
grabs in the euhaline habitats (upper right, see also Figs. 17 & 19) in the
Virginian province is much higher. with a much higher variance.
size and biogeographic factors. All of these factors are secondary to the dominant salinity effect.
0.5
0
80.
0.5
0
-0.5
-1 —1
0
I
-0.5 pCfr
? 3.5
U
a
a
1.5
0.5
‘0
* -0.5
pCk
—1 —1
Figure 21. Log, 0 (Total individuals in 3 0.044 m 2 grabs) is plotted vs. PCA-
H Axes 1 and 2. Infaunal abundances are very low in the oligohaline ( <10
psu) portion of the Virginian Province. Abundances are more variable in the
euhaline estuarine habitats.
-------
36
EMAP-E VP COMMUNITY STRUCTURE
Figure 21 shows that the oligobaline and mesohaline portions of the Virginian province are also
associated with relatively low population abundances (< 1000 individuals per 30.044 in 2 grab samples).
We plot the log 10 of abundance, because abundance in 3 grabs varies from 25 (the lower cutoff for our
analyses) to over three thousand.
Figure 22 shows the plot of the Sanders-Huribert E(S 1 ) diversity index vs. PCA-H axes 1 vs. 2. This
diversity index exhibits a very high correlation with Shannon’s H’ (see Fig. 8 p. 23). Shannon’s if is far
less sensitive to salinity effects than Total species per sampling event’. Shannon’s if, while often used as
a species richness index, is very sensitive to species evenness, or the relative distribution of individuals
among species. The ‘potholes’ in FIgure 22 indicate samples which have roughly the same salinity-
controlled species composition as adjacent samples, but which have very low species evenness. One of
the predicted effects of pollution is a drastic reduction in species evenness in a sample. Pollution or
disturbance could produce the ‘potholes’ in Figure 22.
Figure 22. E(S 10 ) is plotted vs. PCA-H Axes 1 and 2. E(S 10 ) is not as sensitive to salinity effects as
total species. ‘Potholes’ in the landscape may pinpoint the degraded areas within each salinity regime.
Lu
8
7
6
5
4
3
2
0.5
0.5
—1 —1
Ck 14 :xIs f
-------
37
CoNcLUSIONS
On the central role of snlinity
It is apparent from our analyses that salinity is the dominant factor controlling EMAP-E VP
community structure. Part of this is due to the differences in taxonomy used to sort EMAP-E VP
samples from low-salinity regions. More than that, salinity is a major determinant of species nchness
and community composition. This comes as no surprise. Sanders et aL (1965) reached this conclusion
in their survey of the Pocasset River in Massachusetts They concluded that the variance in salinity as
well as its absolute value restricts the number of species present. Boesch (1977b) and Diaz and
Schaffner (1990) have stressed the central role played by salinity in controlling benthic community
structure in Chesapeake Bay. To our knowledge, the analyses in this report provide one of the clearest
demonstrations of the role of salinity as the major factor controlling community structure in a broad
biogeographic region.
Weisberg eta!. ‘s (1993) 1990 EMAP-E VP benthic degradation index (Weisberg et aL 1993)
explicitly accounted for the effects of salinity on species richness. Salinity effects were not properly
assessed in the 1991 and 1992 indices, but Strobel eta!. (1995) again designed their 1990-1993 index
to remove the strong effects of salinity. The sampling properties of Gleason’s D diversity are not well
worked out. Gleason’s D diversity is not among the diversity indices that ecologists use. We would
strongly recommend that future analyses use Shannon’s H’ or Huribert’s E(S ). Hurlbert’s E(S ) has
the advantage that it can be made more sensitive to species richness by increasing the sample size.
Total species per grab is an acceptable index of diversity. However, as our analyses show, it cannot be
used without careful removal of the overriding effects of salinity. Salinity is strongly correlated with
the total number of species (Figure 3, p. 13). Without explicitly removing the effects of salinity, a
degradation index based on species richness would have a tendency to identify oligohaline and
mesohaline habitats as degraded. This tendency would be compounded if the list of opportunistic taxa
used in several EMAP-E VP indices included species that are the natural dominants of oligohaline and
mesohaline environments.
Suggestions for improving EMAP-E VP
Sampling
Two major problems appeared in our preparation of the EMAP-E VP data for community structure
analysis. First, many of the replicate grabs from a site were extremely heterogeneous (CNESS > 0.7).
Approximately 1/4 of the sampling sites were discarded because the three replicate grab samples were
too heterogeneous to be considered true replicates. If the goal of the analysis is to characterize
community structure alone, then this heterogeneity is important. However, in comparing sediment
chemistry data collected from a different set of grabs than the benthic community structure data, it is
essential that the benthic samples exhibit adequate replication. A CNESS distance of 0.7 or above
indicates drastic changes in community structure, comparable to sampling sand vs. mud or sampling
the same community in early spring and late summer (see Trueblood e: a!. 1994). We suspect that the
small-scale heterogeneity was such that different sediment types were sampled with the three replicate
biology grabs. If the three benthic grab samples indicate differences in community structure
comparable to sampling sand vs. mud, then we felt that we could not trust the corresponding chemical
data from the site. Would it correspond to the sandy benthic samples or the muddy ones? The extent
-------
38
of the inter-replicate within site variance can be obtained by examining the clustering patterns of the
sample cluster analysis in Appendix l v.
A second troubling feature of the EMAP-E VP analysis is shown in Appendix IV. In the 1990
sampling, triplicate box cores were taken from a site (STA 099) in New Bedford Harbor on August 15
and September 4, 1990. The differences in community stuctwe at this single site with samples taken
three weeks apart is as great as that observed between ‘degraded’ Chesapeake Bay dnh2line samples
and ‘Clean’ Nantucket Sound samples. Additional seasonal sampling should be carried out in the
Virginian province to establish the extent of seasonal variability in the benthic communities.
Taxonomic issues
The EMAP-E VP program should have archived taxonomic reference material. The taxonomy of even
the most wide-spread taxa in the Virginian province changes on the decadal time scale. Future
ecologists will never know what species were really sampled in the EMAP-E VP. This is not the fault
of the consulting firm which analyzed the EMAP-E VP data (Cove Associates). Cove Associates uses
the latest taxonornic keys and certifies their species identifications with experts. Unfortunately, every
year brings new taxonomic revisions. Currently, the status of the most abundant taxon in the EMAP-E
VP data is in doubt. Mediomasrus ambisera was first described on the East Coast only in the early
1970s by Hobson (1971). Before that time, all Mediomastus were probably incorrectly identified as
the capitellid llereromasrusfiliformis, another Virginian province taxon which thrives at much lower
salinity.
Mediomasrus is not just another worm. It is the most abundant taxon in the entire Virginian province.
There is a dramatic break in the distribution of Mediomasrus ambiseta and M. californiensis at Cape
Cod. Tim Morris at Cove Associates, who identified the capitellids in the EMAP-E VP program, was
responsible for correcting a decades-old misclassification of Mediomastus in the Gulf of Maine.
Throughout the 1960s and 1970s, the dominant Mediomastus species in Massachusetts Bay and Cape
Cod Bay was misidentified as Capitella capitata, Heteromastusfil!formis, or Mediomastus ainbiseta.
When Cove Associates was hired by the MWRA to process benthic samples in Massachusetts Bay in
the mid 1980s, Moms identified the dominant Mediomastus species as Mediomastus c c i jforniensis.
Since Moms’s revelation, not a single Mediomasrus ambiseta has been positively identified North of
Cape Cod. All Mediomastus appear to be M. californiensis. Mediomastus californiensis was
originally described from a California mudflat, and is the numerical dominant at the Los Angeles
sewer outfall (Swartz et at. 1986), and in many areas of Puget Sound (Lianso, unpublished Puget
Sound Ambient Monitoring data). Swartz et a!. (1986) concluded that M. californiensis was an
indicator of mild organic enrichment. In 1973, Day published a key to the polychaetes of the North
Carolina region, listing Mediomasrus cal(forniensis as being common off Beaufort North Carolina in
10-20 meters depth. Day (1973) did not list M. ambiseta, and his description of M. californiensis lists
the presence of the key characteristic that separates M. ainbisei’a and M. cahforniensis (abdominal
neuropodia with hooks). in 1981, Ewing and Dauer published a key to the Chesapeake Bay capitellids
that included both M. wnbise:a and M. caljforniensis. Mediomastus ca1 forniensis is the only
Mediomastus species found in the offshore Minerals Management Survey of the Gulf of Mexico
(Ewing 1984).
The only Medionzastus species identified in the first three years of the EMAP-E VP program was
Mediomastus ambiseta. Before the processing of the final year of EMAP-E VP data, the 1993
sampling, Moms (pers. comm.) identified M. ca1 fonziensis for the first time within the Virginian
province. This key species was found in only thirteen samples in the 1993 sampling. Cove Associates
keeps archive samples of the Mediomasrus collected from Massachusetts Bay and the Gulf of Maine,
-------
GAUAGHER & Ga LE 39
and Morris has been examining the individuals of this key genus. Virtually all of the Mediomasrus
individuals from the first three years of EMAP-E VP sampling were destroyed in order to estimate
biomass. This is very unfortunate. Reference material from the EMAP-E VP program would have
been very valuable in determining the broad-scale distribution of this key genus. The problem is even
mom serious because the extensive Minerals Management Service Survey of the Georges Bank region
(Neff et aL 1989, Battdlle & WHOI, 1985) identified a third Mediomastus species: Mediomastus
fragilis. This identification was confirmed by Linda Warren, who has recently revised the genus
Mediomastus (Warren et al. 1994). Mediomastusfragilis, first identified in Northern Europe, would
key out as Mediomastus cahforniensis using North American polychaete keys.
Many of the other numerically dominant taxa in the EMAP-E VP data set also require taxonomic
work. Oligochaetcs are the most important group of species in distinguishing oligohaline and
mesohaline benthic habitats. The most important taxon controlling CNESS distances among samples
is the odd species group ‘Tubificidae with capiliform chaetae’. None of the oligochae in the
EMAP-E program from areas with more than 5 psu salinity were archived future taxonomic analysis.
We will never know how many mesohaline oligochaete species were present in the Virginian Province
from 1990-1993. The 1990-1993 benthic index cannot be properly evaluated because we do not know
the proportion of oligochaetes that axe members of the oligochaete family Tubificidae. The true extent
of species turnover along the Virginian Province salinity gradient can never be properly assessed with
the EMAP-E VP data because of the lack of taxonomy of the oligochaetes collected in areas having
more than 5 psu salinity.
Nucula proxima and Nucula annulata are among the numerically dominant taxa in the Northern
Virginian province. Nucula annulaza is the tenth most important taxon contributing to CNESS
distances in our analysis. Howard Sanders (1956, 1960) described the Long Island Sound and
Buzzards Bay communities inhabiting subtidal muds as Nucula proxbna-Nephlys ijzcisa communities.
George Hampson (1971) showed that the offshore mud-dwelling species was Nucula annulata, while
the inshore sand-dwelling species in Buzzards Bay was Nucula proxima. Nucula proxima and N.
annulata can be easily confused. but their habitats are very different N. proxima thrives in nearshore
fine sands, but N. annulata is found in offshore muds. The two species show little overlap in their
distributions. In the EMAP-E VP data, N. annulata is found in both nearshore sand and offshore mud
habitats, and N. proxinia is not included in the EMAP-E VP species list. We must assume that N.
proxima was called N. annulata in these analyses. If preserved samples were available, it would be a
simple matter to obtain samples for nearshore sandy environments to determine whether the dominant
protobranch bivalve was N. proxima and not N. annulata. Dr. Robert Prezant, a malacologist at
Indiana University of Pennsylvannia. has been sampling the Assateague/Chincoteague area of Virginia
for many years. He has sampled many Nucula from this shallow sandy area, that the EMAP-E
program uses as one of its undegraded reference stations. The EMAP-E program lists only Nucula
annulata from this shallow sandy area. Dr. Prezant has closely examined the
Chincoteague/Assateague Nucula All are Nucula proxima, in accordance with Hampson’s (1971)
description of the distribution of Nuru!a species. In analyzing the EMAP-E data, it is vital to be able
to distinguish the dominant species from inshore sandy areas from offshore muds. The fact that N.
proxima was missed in the four-year EMAP-E VP program is the type of slipup that could have been
resolved readily if archive material were available. Unfortunately, all bivalve individuals collected
during the EMAP-E VP program were destroyed to determine bivalve biomass for the EMAP-E VP
degradation indices.
The cost for archiving benthic samples is high, but most other federal benthic surveys have managed
this cost. Archiving samples is not incompatible with measuring biomass. The Minerals Management
Service (MMS) insists that bromass be determined, but they also require that all samples be archived.
-------
EMAP-E VP COMMUNITY STRUCTURE
Only subsets of the individn2k need to be analyzed to produce biomacs esnm tes. The remaining
matenal is archived at the Smithsonian. This MMS material is an important resource in the ongoing
taxonomic revisions of our nation’s marine communities. The EMAP-E VP was designed to serve as
a multi-’1ecw monitoring and assessment program. Archiving benthic samples should be a
requirement of all such long-term monitoring plans.
Data analysis
Warwick (1993) asked whether there is an ‘absolute’ measure of benthic community structure that can
determine whether a site was affected by pollution. After assessing indices similar to the EMAP-E VP
degradation index, and several other indices of degradation, he answered ‘No.’ There is no clear-cut
index that can determine whether an observed distribution of species in a sample is the result of
pollution. Only marine sites that are nearly azoic or dominated by members of the genus Capitella can
be unambiguously defined as degraded. Warwick (1993) even discounts the use of Capiteila as
pollution indicator, regarding it more of an indicator of organic enrichment than toxicity. It is equally
risky to assign a site with high species richness or abundance as being non-degraded. In the West
Falmouth oil spill (Grassle and Grassle 1974, Sanders e: al. 1980), the Amoco Cadiz oil spill (Cabioch
et aL 1982) and even the MERL eutrophication experiments (Smith et aL 1979b), species diversity is
insensitive to drastic changes in toxic loadings. However, species composition is sensitive to very low
changes in pollutant input (e.g., Grassle ef a!. 1981).
Dubious indices and statistics
In soft-bottom benthic ecology, there has been a plethora of dubious indices proposed to assess
community structure. Most of these indices and approaches share a few common features:
• They are less expensive than standard benthic sampling which relies on the
identification of all indi iduals to species.
• They rarely require taxonomic identifications beneath the level of family or
competent taxonomists.
• Some require no benthac sampling or a reduced number of replicate samples.
While some of these indices may have merit, all should be viewed with caution, and none should be
made the sole basis of a monitoring program Some indices, such as Jack Word’s (1978, 1980 a & b)
infaunal trophic index (published only in the SCCWRP biennial reports), can be ruled out almost
immediately as being too seriously fla cd to merit discussion as a monitoring tool. Word’s index
forces improper polychaere irtiphic guild cLassifications into an improperly constructed mathematical
formula to produce an index ol pollution cflccts Fenaro et aL (1989) provide one of the few
published uses of Word’s infaunal rrophu unik
The following is a list of indices that ha’c t cn proposed to assess marine benthic degradation. We
will divide them into two groups The first group is composed of seriously flawed indices.
Seriously flawed:
• The nematode/copepod ratio (Rziffaelli & Mason 1981, Warwick 1981,
Raffaelli 1981. 1982 & 1987. A mjad & Gray 1983, criticized by Lambshead
et aL 1983 & 1984 Shaw c i aL 1983).
-------
GAUACHER & GRAssLE 41
• BRAT the benthic resource assessment technique (Lunz & Kendall 1982).
Benthic individuals are ranked by their presence in the guts of important
bottom feeding fish species. Benthic habitats composed of “good” fish food
are giving a higher BRAT ranking than habitats containing species that are
rarely fed on by bottom-feeding fish. This index, as originally proposed had
merit The index was later simplified to matching the size composition of
infauna in fish stomachs with the size composition of benthic communities.
• Jack Word’s infaunal trophic index (ITI) (Word 1978, 1980 a, b, & c, Ferraro
etal. 1989)
Indices that are worth investigating:
• Organism-sediment index (OSPM) (Rhoads & Gennano 1986). This index
was applied to the 1990 EMAP-E VP data (see Figure 4, p 14). Difficulty in
obtaining samples in coarse sediments greatly reduces the utility of sediment-
profile imaging for EMAP-E VP scale monitoring. Grizzle and Pennimen
(1991) found a close correspondence between the OS1 and traditional
benthic community structure analysis in the a polluted New Jersey habitat.
O’Connor et ci. (1989) applied the OSI” to British benthic communities.
• Warwick’s species abundance-biomass comparison (the ABC method)
(Clarke 1990, Warwick and Clarke 1991, Warwick eraL 1987, 1989, 1990,
Essink and Beukema 1986, reviewed by Beukema 1988, McManus and
Pauly 1990,). Dauer et al (1993) found that the ABC method was a poor
predictor of degraded conditions in the Elizabeth River in Chesapeake Bay.
• Caswell’s (1976, 1983) neutral model as a pollution or disturbance index.
The infinite alleles model of population genetics predicts a logarithmic
distribution of species frequencies. When applied to the benthos, Caswell’s
neutral model is assessing the fit of the log-series to benthic abundance data.
Lambshead et ci. (1983) and Lainbshead and Platt (1985) introduced this
method for assessing disturbance pollution. Goldman and Lambshead (1989)
wrote an improved version of Caswell’s program for assessing the effects of
disturbance on benthic communities. Warwick etaL (1990) applied the
neutral model to a pollution study in Bermuda and Warwick (1993) applied
the neutral model to a well-characterized pollution gradient in a Norwegian
fjord. Both studies found the lack of fit to the neutral model to be a very
poor predictor of degradation.
• Departures from the log-normal distribution of individuals among species
[ Gray & Mirza 1979, Gray 1979a & b, Stenseth 1979, Gray 1980, Mirza &
Gray 1981, Gray 1982, 1983, 1989, Gray & Pearson 1982, Gray & Christie
1983, Pearson et at. 1983, Bonsdorf and Kovisto 1982, Nelson 1987; poor fit
found by Rygg (1986)1
• The variance in species frequencies among replicates at a site. Warwick
(1993) argues that polluted sites have a higher variance in faunal similarity
than non-degraded communities.
World-wide, no ‘degradation index’ has yet been found to classify single benthic samples into
degraded and non-degraded categories (Warwick 1993). While environmental regulators might need
this information desperately (e.g., O’Connor and Dewling 1986), benthic ecologists have been unable
to identify a single index that is reliable. No benthic ecologist has developed an index that can be
applied for pollution assessment over large geographic areas. Rhoads and Germano’s (1986)
-------
42
organism-sediment index (OS ) has been one of the most widely used and highly regarded. It is
disturbing that the one test of the OSI index vs. the EMAP-E VP index in Weisberg et aL (1983)
found viitually no statistical association between the two sets of degraded-nondegraded classifications
(see Figure 4, p. 14).
The EMAP-E VP analyses of pollution effects has been hindered by not directly incorporating analyses
of pollutant concentration, grain size, depth, and salinity in the biotic Integrity indices. Weisberg et a!.
(1993) and Strobel etaL (1995) do include salinity normalization of expected species diversity, but far
more use could have been made of the extensive physical and chemical data in developing the benthic
index. Coats (1995) has recently applied Gallagher’s PCA-H method to establish a monitoring
baseline for the Massachusetts Water Resource Authority. He produced PCA-H diagrams similar in
style to those presented in this report. Coats then determined that grain size was the dominant factor
controlling community structure in Massachusetts Bay. He developed a modification of the PCA-H
method which he calls detrended PCA-H which removes the dominant grain-size effect from the major
patterns in community structure. Coats (1995) then perfonned a full power analyses of the four years
of MWRA Massachusetts Bay data, showing how pollution effects that might result from the MWRA
sewage effluent outfall could be detected and assessed.
Coats (1995) DPCA-H method has great potential. However, Coats applied this method to only one
small, approximately 100 k& region of Massachusetts Bay. This is smaller than many of the estuaries
measured in the EMAP-E VP program.
In addition to the PCA-H method, techniques which assess the distribution of individuals among
species might be a valuable tool in determining which samples in the EMAP-E program might exhibit
patterns indiating disturbance or pollution. John Gray introduced the “log-normal” plotting technique
as a indicator of pollution or disturbance. In Gray’s method, the distribution of individuals are fit to
the log-normal distribution. Recent disturbance or pollution produces a characteristic break in the
normally linear plot of the number of species vs. the number of individuals, when plotted on normal
probability paper. John Lambshead, at the Natural History Museum in London, has developed another
method which is based on Fisher a aL’s (1943) log-series. According to Lambshead, undisturbed
communities have distributions which are close to the log-series. Lambshead tests his fit to the log
series, using Caswell’s neutral model, and Goldman and Lambshead (1989) have written a program
called CASVAR which tests benthic community structure data to see how closely it conforms to the
log series.
The lead author of this report has recently applied a modified form of these methods to the EMAP-E
VP data. They seem to work quite well. Samples from the EMAP-E VP reference test data set (listed
in Schimmel a a!. 1994) seem to conform to the log-series expectations. Most of the samples from the
“degraded” test data set reveal two types of departure from log-series expectation. Most of the
degraded stations exhibit very low evenness, compared to log-series expectations. However, in
samples with very low infaunal abundances and very few species, the evenness is much higher than
log-series expectations. This failure to recognize departures on either side of the log-series expectation
may have produced some of the negative critiques of the Log-series method (e.g., Warwick 1993).
Considerable work needs to be done on fitting the log-series to the EMAP-E VP data. Some of the
degraded test data sites in Schimmel a a!. (1994) have distributions of individuals among species
which axe close to the log-series expectation. For example, the three grabs from the Kill van Kull
degraded site (VA91-373, sampled on 8/3/91) have log-series distribution patterns. There is no sign
of impact. Analysis of the species composition of this site indicates that two of the three grabs have
benthic populations typical of shallow depths and salinities of 28 psu. This site was flagged as
-------
GAL lAGHER & GRAasLE
degraded by Schimmel et aL (1994) because four sediment contaminants exceeded the 1990 Long and
Morgan ER-M concentrations (silver, mercury, nickel, and zinc). However, using the Long et aL
(1995) revised ER-M values, only mercury and nickel exceed ER-M. The Long et aL (1995) ER-M
for mercury is 0.71 g/g, and the Kill van Kull site had 1.76 pg/g. The ER-M for nickel was 51.6 pg/g
and the Kill van Kull site had 57.6 pg/g. The organic carbon concentration at this site was 1.8%. It is
entirely possible that the relatively high metal and organic pollutant concentrations at this site could be
bound to organic carbon or acid-volatile sulfide phases (see Di Tore etal. 1990, 1991, 1992), making-
the mercury unavailable to the biota. The Kill van Kull appears to have a ‘reference’ benthic
community, despite having relatively high sediment contaminant concentrations. Even within the three
replicate grab samples from this site, there are some striking differences. Replicate grab three has a
distribution of individuals among species that is least similar to the log series. This same grab contains
a few species like Mulinia lateralis and Streblosplo benedicti that are usually regarded as
opportunistic.
Further work of this sort must be done to assess whether nearly one quarter of the nearshore area in the
the Virginian Province is truly degraded (Table 1, p. 16). The EMAP-E VP dataset with its synoptic
measurements of benthic community structure, sediment contaminant levels, and abiotjc variables will
be an invaluable resource in establishing the causal connection between sediment contaminant
concentrations and altered patterns in community structure.
Have the EMAP Goals and Objectives Been Met?
The EMAP-E VP has created a benthic data set that may be unmatched in the world in terms of broad-
scale geographic coverage. With the physical and chemical data available for each sampling event,
these data should prove to be a rich resource for benthic ecologists for decades to come. The Virginian
Province EMAP-E VP program has collected the necessary data to assess the scale and pattern of
variability in benthic communities. Understanding this variability, and separating the natural and
anthropogenic contributors to it, should be the goal of the Assessment portion of EMAP.
The EPA can be proud that it has created this outstanding database. There is nothing to match it in
scale or scope. The EMAP-E VP program has not been succesful in developing a benthic degradation
index. This is not surpnsing. Benthic ecologists have been searching for indices that will reliably
indicate pollution or disturbance for the past three decades, if not longer. No index has been found to
work reliably.
The EMAP-E VP indices are flawed, but the obstacles faced by the EMAP-E investigators were
massive. The EMAP-E VP program attempted to find an index that would work across a very large
biogeographic province in habitats as diverse as sandy fresh water habitats in the upper Hudson River
to marine subtidal muds like Buzzards Bay, Massachusetts. It is not surprising that the indices
developed for the EMAP-E VP program have flaws.
We would encourage the EMAP-E VP program to critically evaluate the latest index: Strobel et aL’s
(1995) 1990-1993 index. It would be very unfortunate if government agencies began using 7000
spionids per m 2 as a regulatory guideline indicating degradation.
The EMAP-E VP data set could become a rich resource for ecologists and regulators alike if it were
made more accessible and in a wider variety of formats. It is difficult to access the EMAP-E VP SAS
abundance, species name, chemistry and waler column data files in order to do most analyses. It
doesn’t have to be this way. As a model data base management system, EPA should look to NOAA’s
Coastal Zone Color Scanner data distribution network. CZCS data is distributed in a wide variety of
-------
44
formats, from the beautiful World-Wide Web graphic displays of chlorophyll a concentration
throughout the world to the detailed pixel-by-pixel reflectance data needed by investigators interested
in perfecting the chlorophyll a algorithnL The lead author will provide the community structure data
used to produce this report in ASCII or MA1tAB format to any investigator that might request it.
If future programs like EMAP-E VP are implemented, I would strongly recommend that a subset or
perhaps all of the benthic animals identified to species be archived. Many of the species, especially the
numerical dominmits in the EMAP-E VP program, are taxonomically difficult. These taxa include the
Mediomashis mbsera-callfornien.sis-fragihs complex, the Nucula annulaa- proxima-wacellana
complex, oligochaetes, the Ampelisca abdUa-vadoru,n complex, and the Capitella sibling species
complex. Each of these numerically dominant groups has undergone massive taxonomic revision over
the last thirty years, the time scale of the planned EMAP-E VP program. Without archival material,
future ecologists and regulators will never know what species were really present in the period 1990-
1993 in the Virginian Province.
Acknowledgments
David Shull provided valuable editorial assistance and commentary in producing this report. Ana
Ranasinghe and Steve Weisberg (Versar, Columbia Maryland) provided invaluable assistance in
analyzing the EMAP-E VP data. The lead author travelled to Columbia Maryland in February 1995,
where the final species list used in this report were agreed upon. Ranasinghe and Weisberg helped us
identify many undocumented features of the EMAP-E VP tbitiicet, informed us when our SAS data
were in error, and in late June 1995 provided a SAS program that converts EMAP-E VP data from its
raw form to the form that we needed for our analyses. Without their assistance and Aria’s SAS
program, this work could not have been completed. Nancy Mountford and Tim Morris (Cove
Associates) were very helpful in addressing the taxonomic issues involved in preparing the EMAP-E
VP species abundance data for the community structure analyses discussed in this report. Joel
O’Connor and Darvene Adams were very helpful in providing data for the NY/NJ REMAP program in
the form that I needed it.
REFERENCES
Amjad, S. and J. S. Gray. 1983. Use of the nematode-copepod ratio as an index of organic pollution.
Maxine Pollution Bulletin 14: 178-181.
Battelle New England & Woods Hole Oceanographic Institution. 1985. Georges Bank Benthic
infauna monitoring program. Final Report for Third Year of Sampling, Volume 3:
Appendices. Prepared for U. S. Dept. of the Interior, Minerals Management Service.
Beukema, J. J. 1988. An evaluation of the ABC method (abundance/biomass comparison) as applied
to macrobenthic communities living on tidal flats in the Dutch Wadden Sea. Marine Biology
: 425-433.
Boesch. D. F. 1977a. Application of numerical classification in ecological investigations of water
pollution. EPA-600/3-77-033. US. EPA Environmental Laboratory, Corvallis OR.
Boesch, D. F. 1977b. A new look at the zonation of benthos along the estuarine gradient. Pp. 245-266
in B. C. Coull, ed., The ecology of maxine benthos. University of S. Carolina Press, Columbia
SC. 467 pp.
-------
GALLAGHER & GR asLE 45
Boesch, D. F. and R. Rosenberg. 1981. Response to stress in marine benthic communities. Pp. 179-
200 in G. W. Banett and R. Rosenberg, edr., Stress effects on natural ecosystems. John Wiley
& Sons, NY.
Bonsdoff, E. and V. Kovisto. 1982. The use of the log-normal distribution of individuals among
species in monitoring zoobenthos in the northern Baltic Archipelago. Mar. Poll. Bull. 13: 324.
Cabioch, L, J.-C. Daurin, C. Retiere. V. Rivain, and D. Archambault. 1982. Evolution de
peuplements de Rescoff, perturbes par les hydrocarboures de l’Amoco Cadiz. Neth 1. Sea Res.
16; 491-501. In French, English Summaty.
Caswell, H. 1976. Community structure: a neutral model analysis. Ecol. Monogr. 46:327-354.
Caswell, H. 1983. Reply to a comment by Ugland and Gray. Ecology 64: 605-606.
Chapman, P. M. R. N. Dexter and L. Goldstein. 1987. Development of monitoring programmes to
assess the long-term health of aquatic ecosystems. Marine Pollution Bulletin 18: 52 1-528.
Clarke, K. R. 1990. Comparisons of dominance curves. J. exp. Mar. Biol. Ecol. 138:143-157.
Coats, D. A. 1995. 1994 Annual soft-bottom benthic monitoring Massachusetts Bay outfall studies.
MWRA Enviro. Quality Dept. Tech. Rpt series 95-20. Massachusetts Water Resources
Authority, Boston MA. 194 pp.
Dauer, D. M., C. A. Maybury. and R. M. Ewing. 1981. Feeding behavior and general ecology of
several spionid polychaetes from the Chesapeake Bay. J. exp. mar. Biol. Ecol. 54: 21-38.
Dauer, D. M., M. W. Luckenbach. and A. J. Rodi. 1993. Abundance biomass comparison (ABC
method): Effects of an estuanne gradient, anoxic/hypoxic events and contaminated sediments.
Marine Biology jJ. : 507.5 18.
Diaz, R. J. and L. C. Schaffner. 1990. The functional role of estuarine benthos. Pp. 25-56 in M.
Haier and E. C. Krome. eds.. Perspectives on the Chesapeake Bay, 1990. Advances in
estuarine sciences. Chesapeake Research Consortium. Gloucester Pt., Virginia.
Digby, P.O. N. and R. A. Kempion. 1987. Multivariate analysis of ecological communities. Chapman
& Hall, New York.
DiToro, D. M., J.D. Mahony, D. i Hansen, K. J. Scott, M. B. Hicks, S. M. Mayr, and M. S. Redmond.
1990. Toxicity of cadmium in .ediments: the role of acid volatile sulfide. Environ. Sci.
Technol. 9: 1487-1502.
DiToro, D. M., C. Zarba. D. J. Hnasen. W. Berry, R. C. Swartz, C. E. Cowan. S. P. Pavlou, H. E.
Allen, N. A. Thomas and P R. Paquin. 1991. Technical basis for establishing sediment
quality criteria for nontonic organic chemicals using equilibrium partitioning. Environ. Sci.
Technol. 10: 1541-1583.
DiToro, D.. .1. D. Mahoney, D. .1. Hansen. K. i. Scott, A. R. Carison and G. T. Ankey. 1992. Acid
volatile sulfide predicts the acute toxicity of cadmium and nickel in sediments. Environ. Sci.
Technol. 26: 96-101.
-------
46
Essink, K. and J. J. Beukema. 1986. Long-term changes in intertidal flat macrozooDenmos as an
indicator of stress by organic pollution. Hydrobiologia 142: 209-215.
Ewing, R. M. 1984. Capitdllidac. Chapter 14 in W. Uebelacker and P. G. Johnson, eds. Taxononuc
guide to the polychaetes of the noithern Gulf of Mexico. Bariy A. Vittor & Associates, Report
prepared for the Dept. of the Interior Minerals Manngement Service.
Ewing, R. M. and D. M. Dauer. 1981. A new species of Amastigos (Polychaeta Capitellidac) from
the Chesapeake Bay and Atlantic Coast of the United States with notes on the Capitellidae of
the Chesapeake Bay. Proc. Biol. Soc. Wash. 94:163-168.
Fager, E. W. 1972. Diversity: a sampling study. Amer. Natur. 106: 293-3 10.
Ferraro, S. P., F. A. Cole, W. A. DeBen, and R. C. Swartz. 1989. Power-cost efficiency of eight ..
macrobenthic sampling schemes in Puget Sound, Washington, USA. Can. J. Fish. Aquat. Sci.
- 46:2157-2165.
Field, .1. G., K. R. Clarke, and R. M. Warwick. 1982. A practical guide for analysing multispecies
distribution patterns. Mar. Ecol. Prog. Ser. : 37-52.
Fisher, R. A., A. S. Corbet, and C. B. Williams. 1943. The relation between the number of species
and the number of individuals in a random sample of an animal population. J. Anim. Ecology
12: 42-58.
Fuller, C. M., C. A. Butman and N. M. Conway. 1988. Periodicity in fecal pellet production by the
capitellid polychaete Mediomazzus ambiseza throughout the day. Opheia 22: 83-91.
Gabriel, K. R. 1971. The biplot graphic display of matrices with application to principal component
analysis. Biometrika 58: 453-467.
Gallagher, E. D., P. A. Jumars. and D. D. Trueblood. 1983. Facilitation of soft-bottom benthic
succession by tube builders. Ecoogy 64: 1200-1216.
Gallagher. E. D., G. B. Gardner. and P. A. Jumars. 1990. Competition among the pioneers in a
seasonal soft-bottom benthic succession: field experiments and analysis of the Gilpin-Ayala
competition model. Occologia 83. 427-442.
Gallagher, E. D., G. T. Wallace and R. P. Eganhouse. 1992. The effects of the Fox Point CSO on
chemical and biological pollution indices in Boston Harbor’s Dorchester Bay. 130 pp.
Gleason, H. A. 1922. On the relation between species and area. Ecology 3: 158.
Goldman, N. and P. J. D. Lamb ’ head. 1989 Optimization of the EwenslCaswell neutral model
program for community diversity analysis. Mar. Ecol. Prog. Ser. 50 255-261.
Gower, .1. C. 1987. Introduction to ordination techniques. Pp. 3-64 in Legendre, P. and L. Legendre,
eds., Developments in numerical ecology. Proc. NATO Advanced Research Workshop on
Numerical Ecology. NATO AS! Senes G: Ecological Sciences 14, Springer Verlag.
-------
GALLAGHER & GRASSLE 47
Grassle, I. F. and 3. P. Grassle. 1974. Opportunistic life histories and genetic systems in marine
benthic polychaetes. J. Mar. Res. 32: 253-284.
Grassle, 3. F. and W. Smith. 1976. A similarity measure sensitive to the contribution of rare species
and its use in investigation of variation in marine benthic communities. Oecologia 25: 13-25.
Grassle, I. F., R. Elmgren, and 3. P. Grassle. 1981. Response of benthic communities in MERL
ecosystems to low level chronic additions of No.2 Fuel Oil. Mar. Environ. Res. 4: 279-297.
Grassle, J. F., 3. P. Grassle, L. S. Brown-Leger, R. F. Petrecca and N.J. Copley. 1986. Subtidal
macrobenthos of Narragansett Bay. Field and mesocosm studies of the effects of
eutrophication and organic input on benthic populations. Pp. 421-433 in Marine Biology of
Polar regions and effects of stress on Marine Organisms. J. S. Gray and M. E. Chnstinasen,
eds., John Wiley and Sons.
Grassle, 3. P. and 3. F. Grassle 1976. Sibling species in the marine pollution indicator Capitella
(Polychacta). Science 192: 567-569.
Grassle, I. P. and J. F. Grassle 1985. The utility of studying the effects of pollutants on single species
populations in benthos of mesocosms and coastal ecosystems. Pp. 621-642 in H. H. White,
ed., Concepts in Marine Pollution Measurements. Maryland Sea Grant Publication, College
Park MD.
Gray, I. S. 1976. Are baseline surveys worthwhile? New Sci. 2Q: 219-221.
Gray, 3. S. 1979. Pollution-induced changes in populations. Phil. Trans. R. Sco. B. 286: 545-561.
Gray, I. S. 1979. The development of a monitoring programme for Norway’s coastal marine fauna.
Amblo 8: 176-179.
Gray, I. S. 1980. The measurement of effects of pollutants on benthic communities. Rapp. P.-v. Re’un
Cons. penn in Explor. Mer. 179: 188-193.
Gray, 3. S. 1982. Effects of pollutants on marine ecosystems. Neth. J. Sea Res. 16:424-443.
Gray, 3. S. 1983. Use and misuse of the log-normal plotting method for detection of effects of
pollution - a reply to Shaw a aL Mar. Ecol. Prog. Ser. 11: 203-204.
Gray. J. S. 1989. Do bioassays adequately predict ecological effects of pollutants? Hydrobiologia
188/189: 397-402.
Gray, 3. S. 1989. Effects of environmental stress on species-rich assemblages. Biol. I. Linnaen Soc.
37: 19-32.
Gray, J. S. and F. B. Mirza. 1979. A possible method for the detection of pollution-induced
disturbance on marine benthic communities. Marine Pollution Bulletin 10: 142-146.
Gray, 3. S. and T. H. Pearson. 1982. Objective selection of sensitive species indicative of pollution-
induced change in benthic communities. I. Comparative methodology. Mar. Ecol. Frog. Ser. 9:
111—119.
-------
48 EMAP-E VP COMMUNITY STRUCFURE
Gray, J. S. and IL Christie. 1983. Predicting long-term changes in marine benthic communities. Mar.
Ecol. Prog. Ser. 13: 87-94.
Gray, I. S., K. R. Clarke, R. M. Warwick and G. Hobbs. 1990. Detection of initial effects of pollution
in marine benthos, an example from the Ekofisk and Eldfisk oilflelds, Noith Sea. Mar. Ecol.
Prog. Ser. 66: 285-300.
Green, R. H. 1979. Sampling design and statistical methods for environmental biologists. John
Wiley & sons, New York.
Greenacre, M. J. 1984. Theoi7 and application of correspondence analysis. Academic Press, London.
364 pp.
Grizzle, R. E. and C. A. Penmman. 1991. Effects of organic enrichment on estuarine macrofaunal
benthos: a comparison of sediment profile imaging and traditional methods. Mar. Ecol. Prog.
Ser. 74: 249-262.
Hampson, G. R. 1971. A species pair of the genus Nucula (Bivalvia) from the eastern coast of the
United States. Proc. malac. Soc. Land. 39:333-342.
Heltshe, J. R. and N. E. Forrester. 1985. Statistical evaluation of the jackknife estimate of diversity
when using quadrat samples. Ecology 66: 107-111.
Hobson, K. D. 1971. Polychaeta new to New England, with additions to the description of Aberranta
enigma2ica Hartman. Proc. Biol. Soc. Wash. 84: 245-252.
Huribert, S. H. 1971. The non-concept of species diversity: a critique and alternative parameters.
Ecology 52: 577-586.
Huribert, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological
Monographs 54: 187-211.
Jumars, P. J. 1981. Limits in predicting and detecting benthic community response to manganese
nodule mining. Marine Mining 3: 213-228.
Kenkel, N. C. and L. Orloci. 1986. Applying metric and nonmetric multidimensional scaling to
ecological studies: some new results. Ecology 67: 919-928.
Lambshead, P. J. D. 1984. The Nematode/Copepod ratio-some anomalous results from the Firth of
Clyde. Mar. Poll. Bull. 15: 256-259.
Lambshead, P. J. D. 1986. Sub-catastrophic sewage and industrial waste contamination as revealed by
marine nematode faunal analyses. Mar. Ecol. Prog. Ser. 29: 247-260.
Lambshead, P. J. D., H. M. Plan, and K. M. Shaw. 1983. The detection of differences among
assemblages of marine benthic species based on an assessment of dominance and diversity. I.
Nat. Hist. 17: 859-874.
-------
GAL lAGHER & GRIasLE 49
Laznbshead, P.1. and H. M Platt. 1985. Structural patterns of marine benthic assemblages and
relationship with emperical statistical models. Pp. 371-380 in P. E. Gibbs, ed., Proc. 19th Eur.
Marine Biol. Symposium, Cambridge University Press, Cambridge.
Lambshead, P.1. D. and H. M. Plait. 1988. Analyzing disturbance with the EwenslCaswell neutral
model: theoretical review and practical assessment. Mar. Ecol. Prog. Ser. 43: 31-41.
Legendre, L & P. Legendre 1983. Numerical Ecology. Elsevier, Amsterdam.
Lie, U. 1968. A quantitative study of benthic infauna in Puget Sound, Washington, U.S. A., in 1963-
1964. Fiskdir. Skr. Ser. Havunders. 14: 229-256.
Lie. U. and R. A. Evans. 1973. Long-term variability in the structure of subtidal benthic
communities in Puget Sound, Washington, USA. Marine Biology 21: 122-126.
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. 175 pp.+2
Appendices.
Long, E. R., D. D. MacDonald, S. L. Smith and F. D. Calder. 1995. Incidence of adverse biological
effects within ranges of chemical concentrations in marine and estuarine sediments.
Environmental management 19: 81-97.
Lunz, J. D. and D. R. Kendall. 1982. Benthic resources assessment technique: a method for
quantifying the effects of benthic community changes on fish resources. Oceans: 1021-1027.
Magurran, A. E. 1988. Ecological diversity and its measurement Princeton University Press.
Princeton.
May. R. M. 1975. Patterns of species abundance and diversity. Pp. 81-120 in M. L. Cody and J. M.
Diamond, eds., Belknap Press, Cambridge.
McManus, J. W. and D. Pauly. 1990. Measuring ecological stress: variations on a theme by R. M.
Warwick. Marine Biology 106: 305-308.
Messer, J. 1., P.. A. Linthurst. and W. S. Overton. 1991. An EPA program for monitoring ecological
status and trends. Environmental Monitoring and Assessment 17: 67-78.
Mirza., F. B. and .1. S. Gray. 1981. The fauna of benthic sediments from the organically enriched
Oslofjord. Norway. 3. exp. mar. Biol. Ecol. 54: 181-207.
Neff, .1. M., M. H. Bothner, N. 3. Maciolek and 3. F. Grassle. 1989. Impacts of exploratory drilling for
oil and gas on the benthic environment of Georges Bank. Marine Env. Res. 27: 77-114.
Nichols, F. H. 1975. Dynamics and energetics of three deposit-feeding benthic invertebrate
populstions in Puget Sound, Washington. Ecol. Monogr. 45: 57-82.
-------
50 EMAP.E VP COMMUNiTY STRUCTURE
Nichols, F. H. 1985. Abundance fluctuations among benthic inveitebrates in two Pacific estuaties.
Estuaries. 8:136-144.
Nichols, F. H. 1988. Long-term changes in a deep Puget Sound benthic community: local or basin-
wide? Proc. 1st Ann. meeting on Puget Sd. Res. 1:65-71.
O’Connor, B. D. S., I. Costelloe, B. F. Keegan, and D. C. Rhoads. 1989. The use of REMOTS
technology in monitoring coastal enrichment resulting from mariculture. Mar. Poll. Bull. 8:
384-390.
O’Connor, J. S. and R. T. Dewling. 1986. Indices of marine degradation: their utility. Environmental
Management 10: 3 35-343.
Orloci, L. 1978. Multivariate analysis in vegetation research. Junk, the Hague.
Overton, W. S. 1989. Design report of the environmental monitoring and assessment program. U.S.
EPA, Environmental Research Laboratory, Corvallis, OR.
Overton, W. S., D. White, and D. L. Stevens. 1990. Design report for EMAP: Environmental
Monitoring and Assessment Program. EPA/600/3-91/053. U.S. Environmental Protection
Agency, Office of Research and Development, Washington, D.C.
Pearson, T. H. and R. Rosenberg. 1978. Macrobenthic succession in relation to organic enrichment
and pollution of the marine environment. Oceanogr. Mar. Biol. Ann. Rev. 16: 229-311.
Pearson, T. H., J. S. Gray, and P. J. Johannessen. 1983. Objective selection of sensitive species
indicative of pollution-induced change in benthic communities. 2. Data analyses. Mar. Ecol.
Prog. Ser. 12: 237-255.
Peet, R. K. 1974. The measurement of species diversity. Ann. Rev. Ecol. Syst. 5: 285-307.
Pielou, E. C. 1969. An introduction to mathematical ecology. Wiley-Interscience, New York.
Pielou, E. C. 1975. Ecological diversity. Wiley, New York.
Pielou, E. C. 1977. Mathematical ecology. John Wiley and Sons, New York..
Pielou, E. C. 1984. The interpretation of ecological data. John Wiley & Sons, New York. 263 pp.
Plan, H. M. N. and P. J. D. Lambshead. 1985. Neutral model analysis of patterns of marine benthic
species diversity. Mar. Ecol. Prog. Ser. 24: 75-81.
Raffaelli, D. 1981. Monitoring with meiofauna - a reply to Coull, Hicks and Wells (1981) and
additional data. Mar. Poll. Bull. 12: 381-382.
Raffadili, D. 1982. An assessment of the potential of major meiofauna groups for monitoring organic
pollution. Marine Environmental Research 7: 151-163.
Raffaelli. D. 1987. The behaviour of the nematode/copepod ratio in organic pollution studies. Mar.
Env. Res. 2 : 135-152.
-------
51
Raffaelli, D. G. and C. F. Mason. 1981. Pollution monitoring with meiofauna, using the ratio of
nematodes to copepods. Mar. Poll. Bull. 12: 158-163.
Ranasinghe, 1. A., S. B. Weisberg, D. M. Dauer, L C. Schaffner, R. J. Diaz and J. B. Frithsen. 1993.
chesapeake Bay Benthic Community Restoration Goals. December 1993 report for the EPA
and Maryland Governor’s Council on Chesapeake Bay research.
Rhoads, D. C. and J. D. Gennano. 1986. Interpreting long-term changes in benthic community
structure: a new protocol. Hydrobiologia 142: 29 1-308
Rygg, B. 1986. Heavy metal pollution and log-normal distribution of individuals among species in
benthic communities. Marine Pollution Bulletin 17: 31-36.
Sanders, H. L. 1956. The biology of marine bottom communities. Bull. Bingham OceaIlbgraphic
Collection 15: 345-414.
Sanders, H. L. 1960. Benthic studies in Buzzards Bay ifi. The structure of the soft-bottom
community. Limnol. Oceanogr. 5: 138-153.
Sanders, H. L. 1968. Marine benthic diversity: A comparative study. Amer. Natur. 102: 243-282.
Sanders, H. L., P. C. Mangelsdorf and G. R. Hampson. 1965. Salinity and faunal distribution in the
Pocasset River, Massachusetts. Limnol. Oceanogr. 10:216-220.
Sanders, H. L., J. F. Grassle, G. R: Hampson, L. S. Morse, S. Gamer-Price, and C. C. Jones. 1980.
Anatomy of an oil spill: long-term effects from the grounding of the barge Florida off West
Falmouth, Massachusetts. .1. Marine Research 38: 265-380.
Schimmel, S. C, B. D. Meizian. D. E. Campbell, C. 3. Strobel, S. 3. Benyi, J. S. Rosen and H. W.
Buffum. 1994. Statistical Summary: EMAP-Estuaries Virginain Province -1991. Us
Environmental Protection Agency, Office of Research and Development, Environmental
Research Laboratory, Narragansett, RI EPA/620/R-94/005.
Schubel, 1. R., D. F. Borsch, T. M. Church. J. W. Farrington, S. L. Rathbun, S. Saila, and K. Tenore.
1992. Review of the U. S. Environmental Protection Agency’s Environmental Monitoring and
Assessment Program-Estuanes (EMAP-Estuaries). A report of review panel convened by the
Estuanne Research Foundation on 30-31 March 1992.
Sedgewick, R. 1992. Algorithm% an C++. Addison-Wesley, Reading MA. 656 pp.
Shaw, K. M., P.3. D. Lambshead and H. M. Plait. 1983. Detection of pollution-induced disturbance
in marine benthic assemblagc with special reference to nematodes. Mar. Ecol. Prog. Ser. 11:
195-202.
Simberloff, D. 1972. Properties of the rarefaction diversity measurement. Amer. Natur. 106:414-418.
Simberloff, D. 1979. Rarefaction as a distribution-free method of expressing and estimating diversity.
Pp. 159-176 in 3. F. Grassle. G. P. Paul. W. Smith, and C. Tallie, eds., Ecological diversity in
theory and practice. International Cooperative Publishing House, Fairland, MD.
-------
52 EMAP-E VP COMMUNiTY STRUCTURE
Smith, W. and I. F. Grassle. 1977. Sampling properties of a family of diversity measures. Biometrics
33: 283-292.
Smith, W., I. F. Giassle, and D. Kravitz. 1979a. Measures of diversity with unbiased estimates. Pp.
177-191 n J. F. Grassle, G. P. Pall, W. Smith, and C. Tallie (eds.), Ecological diversity in
theory and practice. International Cooperative Pub1i hing House, Fairland, MD.
Smith, W., V. R. Gibson. L. S. Brown-Leger and J. F. Grassle. 1979b. Diversity as an indicator of
pollution: cautionary results from microcosm experiments. Pp. 269-277 in 3. F. Grassle, G. P.
Paul, W. Smith, and C. Taffie, eels, Ecological diversity in theory and practice. International
Cooperative Publishing House, Fairland, MD.
Stenseth, N. C. 1979. Where have all the species gone? On the nature of extinction and the Red
Queen hypothesis. Oikos 33: 196-227.
Strobel, C. 1. S. J. Benyi, D. 1. Keith, H. W. Buffum, and E. A. Petrocdlli. 1994. Statistical Summary:
EMAP-Estuaries Virginian Province - 1992. U. S. Environmental Protection Agency, Office
of Research and Development, Environmental Research Laboratory, Narragansett, RI
EPA/620/R-94/O 19.
Strobel, C. 3., H. W. Buffum, S. J. Benyi, E. A. Petrocelli, D. R. Reifsteck, and D. 3. Keith. 1995.
Statistical Sumary: EMAP-Estuaries Virginian Province -1990 to 1993. U. S. Environmental
Protection Agency, National Health and Environmental Effects Research Laboratory, Atlantic
Ecology Division, Naragansett. RI. EPA/6201R-941026.
Summers, 3. K., 3. M. Macauley. P. T. Heitmuller, V. D. Engle, A. M. Adams, and G. T. Brooks.
1992. Annual Statistical Summary: EMAP-Estuanes Louisianian Province - 1991. U. S.
Environmental Protection Agenc.. Office of Research and Development, Environmental
Research Laboratory. Gulf Breeze. FL EPA/600/R-93/O01.
Swartz, R. C., F. A. Cole, D. W. Schults and W. A. DeBen. 1986. Ecological changes in the Southern
California Bight near a large sewage outfall: benthic conditions in 1980 and 1983. Mar. Ecol.
Prog.Ser.31: 1-13.
ter Braak, C. 3. J. 1983. Pnncipal component’ biplots and alpha and beta diversity. Ecology 64:454-
462.
Trueblood, D. D., E. D. Gallagher. and I) M Gould. 1994. Three stages of seasonal succession on
the Savin Hill Cove mudflat. Boston Harbor. Limnol. Oceanogr. 39: 1440-1454.
Ugland, K. I. and 3. S. Gray. 1983. Rean.il. i. of Caswell’s neutral model. Ecology 64: 603-605.
Underwood, A. 3. 1981. Techniques of an.d’.si. of variance in experimental marine biology and
ecology. Oceanogr. Mar Biol. Ann Rc 19 513-605.
Wallace, 0. T., C. F. Krahforst. L. C. Puts. .1 P Shine. M. M. Studer, and C. R. Bollinger. 1991.
Assessment of the chemical composition of the Fox Point CSO effluent and associated
subtidal and intertidal environments Analysis of CSO effluents, reciving water and surface
sediments for trace metals prior to CSO modification. Final Report to MA Dept. of
Environmental Protection. Office of Research and Standards. 225 pp.
-------
GAUACIIE* & GRAanZ 53
Warren, L. M., P. A. Hutchings, and S. Doyle. 1994. A revision of the genus Mediomasus Haitman
1944 (Poychaeta Capitellidae). Records of the Australian Museum 46: 227-256.
Warwick, R. M. 1981. The Nematode/Copepod ratio and its use in pollution ecology. Mar. Poll. Bull.
12: 329-333.
Warwick, R. M. 1993. Environmental impact studies on marine communities: pragmatical
considerations. Aust. J. Ecology 18:63-80.
Warwick, R. M. 1988. Effects on community structure of a pollution gradient -SUMMARY. Mar.
Ecol. Prog. Ser. 4 : 207-211.
Warwick, R. M., T. H. Pearson, and Ruswahyuni. 1987. Detection of pollution effects on marine
macrobenthos: further evaluation of the species abundance/biomass method. Marine Biology
95: 193-200.
Warwick, P.. M. and K. R. Clarke. 1991. A comparison of some methods for analysing changes in
benthic community structure. J. Mar. Biol. Assoc. U. K. 21:225-244.
Warwick, R. M., H. M. Platt, K. R. Clarke, I. Aghard. and i. Gobin. 1990. Analysis of macrobenthic
and meiobenthic community structure in relation to pollution and disturbance in Hamilton
Harbour. Bermuda. J. exp. Mar. Biol. Ecol. 138: 119-142.
Washington, H. G. 1984. Diversity, biotic and similarity indices: a review with special relvance to
aquatic ecosystems. Water Res. 18:653-694.
Weisberg, S. B,!. B. Frithsen, A. F. Holland, J. F. Paul, K. J. Scott, J. K. Summers, H. T. Wilson, R.
M. Valente, D. G. Heimbuch, G. Gerritsen, S. C. Schimmel and P.. W. Latimer. 1993.
EMAP-Estuaries, Virginian Province 1990 Demonstration Project Report. EPA/620/R-
93/006. Narragansett, RI: US Environmental Protection Agency, Environmental Research
Laboratory, Office of Research and Development.
Word, J. Q. 1978. The infaunal trophic index. Annual Report 1978. Southern California Coastal
Water Research Project. El Segundo, California. pp 19-39.
Word, J. Q. 1980a. Classification of benthic invertebrates into infaunal trophic index feeding groups.
Annual Report 1980. Southern California Coastal Water Research Project. El Segundo,
California. Pp. 103-121.
Word, .1. Q. 1980b. Effects of screen size and replication on the infaunal trophic index. Annual Report
1980. Southern California Coastal Water Research Project. El Segundo, California. p. 123-
130.
Word,!. Q. I 980c. Extension of the infaunal trophic index to a depth of 800 meters. Annual Report
1980. Southern California Coastal Water Research Project El Segundo, California.
-------
54 EMAP.E VP COMMUNITY STRUCTURE
APPENDIX I METHODS FOR ANALYZING COMMUNITY STRUCTURE
Diversity indices
This section will survey the major diversity and similarity (or dissimilarity measures) in use in
community ecology today. Magurran (1988) provides a review of the history of diversity measures in
community ecology. She ignores the most common method used to assess diversity in marine benthic
communities: Sanders-Huribert rarefaction.
Sanders (1968) introduced the rarefaction method for assessing species diversity. Single samples are
plotted as rarefaction curves representing the number of species observed in the sample and the
number of species expected from randomly drawn subsamples of the total sample. Unfortunately,
Sanders method for calculating rarefaction curves was wrong. This error was pointed out by Fager
(1972), and corrections were published by Hurlbert (1971) and Simberloff (1972). Hurlbert (1971)
criticizes the use of diversity indices and introduces a correction to Sanders’ (1968) rarefaction
method. Simberloff (1972) introduced an identical correction for Sanders’ rarefaction method, but
Hurlbert (1971) gets priority. Simberloff (1979) introduced a variance estimator for E(S ), but Smith
and Grassle (1977) proposed a better measure earlier. This modified rarefaction method is now
routinely described as Huribert’s E(S ). Peet (1974) and Pielou (1969, 1977) review the use of H’ and
other measures of diversity.
Species richness indices
While there are indices specifically designed to assess species evenness or equitability, there are no
unbiased estimators of species richness per Se. An unbiased estimator is a statistic that has an expected
value equal to the true value. The EMAP-E VP program used the total number of species per
sampling event (usually three replicate samples) as an index of species richness in the 1990, 1991, and
1992 biotic indices. While this is a straightforward measure of species richness, it is also a biased
statistic. Bias means that the expected value of the statistic will be a strong function of sample size.
Total number of species per event will vary strongly with the number of samples taken. This bias is
not alleviated by calculating the mean number of species per grab. Even with samples drawn from the
same theoretical distribution, the total number of species will vary strongly with the number of
individuals sampled.
The information content measures of diversity, Brillouin’s H and Shannon’s H’ are often used as
measures of species richness. Shannon’s H’ is the older and more widely used:
H’ = - Pt log Pt’ (6)
where, Pk is the frequency of species k in the sample, and S is the number of species. Pielou (1977)
recommends Brillouin’s information content, appreviated H, should be used to calculate the diversity
of fully enumerated samples:
Brillouins H = I * log N ! (7)
N N 1 ! N 2 !
-------
GALLAGHER & Ga LE 55
where, N is the total number of individuals in a sample, and N is the abundance of species i. Both H
and H’ can be calculated using Naparien logarithms, log 10 or log 2 . H’ assumes an infinite population,
and if used it should include an estimate of the variance. Both are highly sensitive to differences in
species evenness. Pielon (1977) regarded this feature as desirable. Deep-sea benthic ecologists often
use the Sanders-Huribert rarefaction index E(S 1 ) as an index of species richness. While at large n, the
index becomes mote sensitive to species richness, it is still very sensitive to differences in species
evenness.
Magurran (1988), following May (1975), advocites using the log-series cc as a measure of species
richness. A MATLAB m.flle, logseries.m, that calculates the log-series cc and its variance estimator is
available from the senior author. Magurran (1988, p. 11) describes the Menhinick species richness
index, which is simply the number of species divided by v’(number of individuals). This index is
similar to Margalef’ s species richness index which is (Total species-I)! v’(number of individuals).
Species evenness indices
Two samples containing the same number of species can differ in their species diversity. Community
ecologists, following Pielou (1977) regard samples with a more equitable distribution of individuals
among species as being more diverse. Hurlbert (1971) introduced statistics to measure the evenness
component of Brillouin’s H. called V, and Shannon’s H’, called J’. An alternate measure of the
evenness component for Bnllouin’s H is called E. If the individuals are equally distributed among S
species, the maximum value for H’ is log(S). Both V and .1’ range from 0 to 1. A sample containing
just one species has an undefined evenness.
Simpson’s diversity, known as Cmi diversity in genetics, is often used as a species richness index, but
it is very sensitive to species evenness. Simpson’s diversity index has both biased and unbiased
estimators. Smith and Grassle (1977) showed that the unbiased estimator for Simpson’s diversity is
one minus the Sanders-Hurlbert expected number of species at n=2 (Simpson’s diversity=E(S 2 )-1). At
such a low rarefaction sample size. Simpson’s diversity and E(S 2 ) are influenced strongly by both
species richness and evenness. As shown by Peet (1974) and Smith et aL (1979a), there is usually a
very strong correlation (r>0.95) between Shannon’s H’ and E(S 10 ). This holds for the EMAP-E VP
data, but does not hold for species-rich deep-sea communities. Hurlbert’s E(S ) index can be made
less sensitive to species evenness component of diversity by increasing n, but there is no set rarefaction
sample size n at which point E(S ) can be regarded as a species richness index. In some deep-sea
benthic samples. E(S 1 ) is strongly correlated with 3’, but not with any of the other indices regarded as
species richness indices (e.g.. H. H, total species per sample, and log-series a).
It would be nice if there were an c ipected species evenness for a given community. Departures from
this expected evenness might indicate disturbance. Caswell (1976) borrowed the Ewens infinite alleles
model from population genetics to establish a ‘null model’ for the expected H’, given the number of
individuals and species in a sample. The lack of fit to the Caswell neutral model is being used as an
index of pollution. especially by Lambshead and co-workers in Great Britain. Goldman and
Lambshead (1989) describe a program CASVAR.FOR which they use to fit Caswell’s neutral model
to benthic data. The Ewens infinite alleles model fits data to a log series, and the results of the neutral
model tests are similar to that obtained by fitting the log-series to community structure data. May
(1975) was the first to show that rarefaction curves generated from shallow-water data seem to
conform to the expectations produced by the log-series (Fisher et aL 1943). An unfortunate
consequence of applying the neutral model and fitting the log-series to benthic data is that highly
impacted, species-poor. benthic communities often depart from the log-series and neutral model
expectations in having too equitable a distribution of individuals among species. Warwick (1993)
-------
56
tested the neutral model using data from a known pollution gradient and concluded that it was a poor
predictor of benthic degradation. John Gray proposed that both undisturbed and highly impacted
benthic communities should confonn to the log-normal distribution. Gray and Mirza (1979) proposed
that departures from the log-normal distribution of individuals among species could be used to assess
transitions from one ‘stable state’ to another. Stenseth (1979) provided a mathematical model for
Gray’s empirical resulL Lambshead et aL (1983) analyzed the distributions of individuals among
species in a number of published benthic data sets, finding that they conformed neither to the log-
normal (canonical or otherwise) or log-series. Hughes (1984, 1986) provides analyses and a model
showing that the expected number of individuals among species in shallow-water benthic communities
is even more inequitable than log-series expectations.
Jackkn fed diversity indices
All diversity indices are biased to some extent. Bias corrections exist for H’ (Peet 1974). This
jackknife method can be applied to any diversity index. Smith and Grassle (1977) showed that E(S ) is
a minimum variance unbiased estimator of diversity, but only if the underlying populations are
independently Poisson distributed. Heltshe and Forester (1985) present the jackknife bias correction
and variance estimator for the Brillouin and H’ diversity estimators. The lead author has adapted this
jackknife bias correction for E(S,.) and programmed the algorithm in MATLAB. Organic enrichment
often dramatically increases the number of benthic individuals in a sample. Using a strongly biased
species richness index, such as ‘Total species per event’, greatly weakens the utility of diversity as an
indicator of pollution effects.
Rarefaction, CNESS and Principal Components Analysis of Hypergeometric
probabilities (PCA-H)
The clear and concise description of benthic community structure is fundamental to the analysis of
both basic and applied benthic ecological problems. We define community structure as “the variation
and covariation of species abundances in time and space.”
Sanders’ (1968) rarefied species diversity, modified by Huribert as E(Sj, and Grassle & Smith’s
(1976) faunal similarity index NESS are both based on the sample x spp. matrix of hypergeometric
probabilities (H). These hypergeometric probabilities are simply the probability of sampling species k
in sample i with a random draw of m individuals:
-------
$7
Ha,,, = I —
=1— (8)
Total 1 = the sample total.
xk = the abundance of species k in sample i.
m=NESSm Number of individuals to be drawn at random.
= a factorial.
Any sample by species matrix of counts can be conveited quickly to a hypergeometric probability
matrix H. In the original calculation of NESS and Huribert’s E(S), fractional abundances, that might
arise from calculating the mean species abundances in replicate samples, were rounded to integers
prior to the calculation of hypergeometric probabilities using factorials. However, this rounding is not
a good idea. In all of The lead author’s programs for calculating E(Sn), NESS. NNESS, and CNESS,
factorials are calculated using the natural log of the r (gamma) distribution since F(n+I)=n!. The F
distribution is continuous, and since it does not require integer values, E(S,,), NNESS, and CNESS can
be calculated using non-integer data. The senior author provides FORTRAN and Matlab programs
(with documentation) for calculating hypergeometric probabilities. Hurlbert’s E(Sj, NESS, and
CNESS.
E(S,) , or the rarefied species diversity for sample i with a random draw of n individual is simply the
row sum of the H matrix. It is defined as:
( w.Nk )
E(S,,) = El - _____
where, n random sample size.
(N’
hinonual coefficient. (9)
No. of ways to sample N objects, n at a time.
N!
- (N-n)! •
N = Total individuals in sample.
Nk = Individuals of species k.
S = Number of species.
Totaig!
m! * (TOTALs - m)!
-------
58 EMAP-E VP COMMUNiTY STRUCTURE
Smith and G i sslc (1977) determined that E(S) was a minimum variance unbiased estimator (MVUE)
of diversity and i,resented equations to estimate the variance of E(S).
NESS, NNESS, and CNESS
The NESS faunal similarity index was described by Grassle and Smith (1976). Trueblood et aL
(1994) correct a flaw in the original index, calling this new version NNESS. They also proposed a
metric verson of NNESS, called CNESS. The equations for NNESS and CNESS are shown below:
NNESSgp,, = _______________
+ ES4.,,,)
I 4 O)
CNESS = 12 1 - 4fri
bn ,/ESS, * ES
NESS and CNESS are families of similarity and dissimilarity indices. NNESS at its upper and lower
sample sizes converges to the Sorensen binary and Morisita similarity indices. NNESS at its upper and
lower sample sizes converges to Sorensen’s index and the Morisita-Horn similarity index. At a sample
size of 1, CNESS is Orloci’s (1978) chord distance. Kenkel and Orloci (1986) showed that the chord
distance analyzed with non-metric multidimensional scaling (NMDS) was the best of eight procedures
tested for recovering the patterns in complex simulated ecological data. No one has apparently
described a binary similarity index corresponding to CNESS at m= .
Principal Components Analysis of Hypergeometric Probabilities (PCA-H)
There is a major advantage of CNESS over NNESS. While both indices have a straightforward
geometric interpretation, CNESS is a metric but NNESS is only a semimetric. CNESS is the
Pythagorean distance between the intersection of sample vectors, with positions determined by the H
matrix, and the unit hypersphere. These intersection points with the unit hypersphere axe calculated
through a row normalization of the H (i.e.. the sum of squared elements in each row is 1). Because
these distances are chords on the hypersphere. they are called chord distances. CNESS is the chord
distance between sample vectors at a distance of one unit from the origin. These Pythagorean
distances are calculated in S-dimensional ordination space, where S is the number of species. For
samples containing more than 3 species. the human mind cannot perceive the distribution of samples
in ordination space. We can perceive the di%tances among samples only in 2- or 3-dimensional
displays.
Principal component analysis projects the major %ources of variation in a complex swarm of points in
S-dimensional space in fewer dimenslon%. oFten only 2 or 3. The first step in a principal components
analysis of hypergeometric probabilities, called PCA-H here, is to mw-normalize the H matrix so that
the sum of the squared elements on each mu. i’ I. This is the mathematical equivalent of projecting
sample points onto the unit hypersphere. CNESS is the Pythagorean distance among sample points
projected onto the hypeisphere. Geometrically, these distances are the lengths of chords between
points on the hypersphere. hence the name chord distance (reserved for CNESS 1 I). This row-
normalized H matrix is then centered by column, so that the mean of each column isO. This new
matrix, called XR, contains all of the information necessary to calculate CNESS, which is simply the
-------
GALLAGHER & GRASSLE 59
Pythagorean distance among samples with coordinates specified by the rows of XR. Principal
components analysis identifies which linear combination of species is most important in determining
the CNESS distances among samples. PCA creates a low-dimension projection of XR by transforming
the original S axes to new axes, now called principal components, that are linear functions of the old.
The number of principal components is the minimum of the number of samples or species in the
original data. Pythagorean distances between samples piotted with respect to these new axes will be
the same as those calculated with the original axes in XR. However, these new axes or principal
components are derived so that the first axis represents the largest source of variation in CNESS
distances among samples. The second and higher principal components reflect less important sources
of variation in CNESS distances among samples, and are orthogonal to all previous ordination axes.
These principal components are also normed so that the sum of the squared elements of each principal
component equals 1. If V is the matrix of principal components, with the number of rows equal to
number of species and the number of columns equal to the number of components, the first principal
component is represented by the first column of the matrix, V(:, I). Orthogonal and normal principal
components (=orthonormal) implies that V(:j)’*V(:,i) = 0 V i’j and V*V’=I, where I is the identity
matrix. The positions of each sample in this new multidimensional space, defined by the principal
components, are contained in the sample x principal component score matrix Y, which can be
calculated by: Y=XR*V. The Y matrix of principal component scores is identical to that obtained by
a principal coordinates analysis of the original CNESS matrix.
The relative amounts of variation explained by each principal component are provided by the sum of
the squared principal component scores for each axis, divided by the sum of the squared coordinates
for all samples in the original XR matrix. If an Eigenanalysis is used to calculate principal
components, the percentage of variation explained by each component is the eigenvalue corresponding
to that component divided by the sum of the eigenvalues for all components.
There are at least a dozen different algorithms to calculate the principal components of a given data
set. These will be summarized in a later section on the matrix algebra of PCA-H. These algorithms
can differ greatly in the amount of computer memory required, processing speed and numerical
accuracy. The large size of the EMAP-E data set required using some non-standard, but still highly
accurate, methods to calculate principal components.
PCA-H retains the information on species frequencies. A simple metric scaling or Principal
Coordinates analysis or a Non-metric multidimensional scaling of CNESS distances does not retain the
information on which species control contribute to differences in faunal composition among samples.
Gabriel’s (1971) graphical biplot can retneve this species frequency data and show the relative
importance of each species to the CNESS distances among samples. The biplot reveals those species
that account for the major sources of variation in CNESS distances among samples. A graphical biplot
of PCA-H results shows the relative CNESS distances among samples and the species that account for
the distances. In the biplot, species are represented by vectors (arrows). The terminus for the arrows
are the elements of V. which may be called the species loadings. Since the square of these coordinates
for each axis sum to I, the relative lengths of arrows in a 2-dimensional biplot shows the relative
importance of each species in controlling the CNESS distances among samples. The sample positions
in the first 2 principal components can be calculated using Y(:,1:2)=XR*V(:,1:2). The longer a
species’ arrow, the more important that species is in controlling the position of samples in a two-
dimensional display. The length and direction of the arrow away from the origin are important. The
cosine of the angle with each axes is directly related to the principal component loading for that
species on that axis. The relative species composition of a sample can be determined by projecting the
sample points at right angles onto the longest species vectors. Digby and Kempton (1987) provide a
clear description of the use of Gabriel (1971) graphical biplot. Note that the asymmetric display used
-------
60 EMAP.E VP COMMUNITY STRUCTURE
by Digby and Kempton, and in this report, is different from and preferable to that originally described
by Gabriel.
Ter Braak (1983) came very close to describing the PCA-H method. In his discussion of the geometric
relationship between faunal similarity and diversity, he stated that samples might be plotted according
to their hypergeometric probabilities. The distances between samples i and j in his proposed
ordination would be /(ESS, 1 +ESS ,-2ESS ). This distance measure would have the unfortunate
consequence of having no set upper limit and would be heavily dependent on the diversity of samples.
The full set of MATLAB programs needed to perform all PCA-H analyses in Trueblood et a!. (1994)
are now on the lead author’s web page.
Choosing the appropriate sample size, m
The lead author developed a non-parametric procedure using Kendall’s non-parametric rank order
correlation coefficient (v=tau) to find a value for m that yielded a distance index that was highly
correlated with both the CNESS I and CNESSm , ,i matrices. This procedure is described in
Trueblood et al. (1994), and the program that performs it (findcnm.m) is provided on the lead author’s
web page.
The largest sample size m for which NNESS or CNESS can be calculated is set by the minimum
sample total in the data set. The original NESS algorithm was even more restrictive, with the largest m
being half the minimum sample total. One can extend the range of m by transforming the H matrix to
l’s and zeros (i.e., the Boolean transform option in COMPAH) and then calculating the chord distance
among samples. This procedure is equivalent to calculating CNESS at m= .
The m size that is sensitive to both the rare and abundant species varies depending upon the
distribution of individuals among species. For most soft-bottom benthic data, m=l0-20 is an
appropriate sample size. Trueblood et aL (1994) found that m=15 was appropriate for an intertidal
benthic community. If the species distributions are heavily dominated by one or a few species, then m
should be increased above 10. A NESSm value of 20-25 appears optimal for the EMAP-E data.
Interpreting the graphical displays:
Graphical biplots
The biplot produced by PCA-H is the asymmetrical Euclidean distances biplot. The plot presents a
low-dimensional projection of the samples, representing the best least-squares fit of the original
CNESS distances. For this reason, samples will always be less than ‘2 units apart.
Species are plotted as vectors. The relative frequencies of a species in a sample can be estimated by
projecting the sample orthogonally onto the appropriate species vector. Only one tail of the species
vector is shown, but each species vector projects in the opposite direction as well. The origin
represents the mean frequencies of a species in the normalized H matrix. The length of the species
vectors indicates the importance of that species in that 2-dimensional projection.
Digby and Kempton (1987) provide a particularly clear explanation of how the Euclidean distance
biplot method can be used to recover the basic structure of the original data matrix.
-------
a—_.. .,‘.- , .n 2,floAC IV
61
This biplot method is different from the dual display biplot in correspondence analysis (see Greenacre
1984). That biplot is a symmetric biplot in which both samples and species are plotted in scaled
coordinates. In that display, the angles between samples and species vectors determine the association
between a given species and sample.
The asymmetric Gabriel Euclidean distance biplot display should not be used to examine the statistical
association among species (i.e., the R-mode ordination). To analyze the association of species among
samples, the species should be plotted in scaled coordinates such that the sum of the squared species
coordinates in each dimension equals the eigenvalue for that dimension (Legendre and Legendre
1983). This scaling, called the covariance biplot, is performed using the MA1’LAB m.file rmode.m.
Non-significant species vectors are not labeled. The data can also be clustered using the same
similarity measure, the angle between species vectors. COMPAH can perform this clustering usirig the
following steps:
1) Calculate the hypergeometric probability matrix, H
2) Standardize the H matrix by station (the sum of squared H elements for each
station will sum to 1)
3) Cluster using Pearson’s r (the equivalent of clustering based on the angular
cosines among species vectors in the r-mode plot)
Some applications of the Gabriel biplot method rescale the species vectors uniformly to aid in the
projection of samples onto species vectors. As noted by Gower (1987), this is not a good idea. The
endpoint of a species vector indicates the maximum value for a species. If a sample point is plotted at
a greater distance than the tip of a species vector, then its position is determined by a combination of
species.
The Geometry of Sanders-Huribert E(S,.) and CNESS
A simple 3-species, 3-sample data set is used
to demonstrate the geometric interpretation of
the Sanders-Huribert E(Sn) diversity index
and CNESS. Both are related to distances
among samples and the origin when sample
points are plotted using the hypergeometric
probability matrix H.
Figure 22 shows the position of three sample
points, A-C, in a 3-dimensional space
determined by the abundances of Species 1-3. “
The Euclidean distances among sample points
in this species space is a very poor indicator of 2
faunal similarity. One of the more troubling
aspects of straight Euclidean distance is the
‘double-zero’ problem. Samples sharing no
species will appear similar because they both
have sample coordinates near the origin.
Figure 23. Three samples are plotted in species
space. The Pythagorean distance among samples is
a poor, unbounded, distance measure offaunal
similarity. Note the greatly compressed vertical
scale.
-------
62
EMAP-E VP COMMUNITY STRUCTURE
Plotting samples by the relative frequencies of
species in samples (Figure 24) is an excellent
basis for ordination. Using such a plot, ter
Braak (1983) describes the geometric
interpretation of faunal similarity and Gini-
Simpson diversity index. Unfortunately,
Euclidean distances among sample points
determined solely by species frequencies are
often very insensitive to the rarer species in a
community.
Figure 25 shows the Sanders-Hurlbert
rarefaction curves for the three samples.
Sample C has the same three species as
sample A but has greater evenness. Sample A
is as species-rich as sample C, but the
equitability among species frequencies is low.
I
There is a straightforward geometnc
interpretation of the rarefaction curves shown
in Figure 25. The Sanders-Hurlbert E(S ) is
the City-Block distance (=Manhattan metric)
from the origin to the sample point plotted
using hypergeometric probabilities. This city block distance is shown in Figure 26.
U)
UI
0
U
0
a.
U I
0
0
z
.
U
0
a.
UI
0.5 1)
00
Figure 26. The Sanders-Huribert diversity from
the previous picture is the city-block distance
between the origin and the sample point, with
points plotted using hypergeometric
probabilities. The Euclidena distance to the
origin is /ESS, 1 . At m=1, the distance between
a sample point and the origin is Simpson’s
diversity; the closer to the origin, the higher the
diversity (ter Braak 1983).
1
1
00
0.5 j)
(SP
pro”
Figure 24.The same three stations can be plotted
using the probability that they will be sampled with a
random draw of 1, 2, or m individuals (the
hypergeometric probabilities). The H(m= 1) sample
coordinates are shown.
180
0 40 80 120
Rarefied sample size (n)
Figure 25. The Sanders-Hurlbert rarefaction
curves for the three samples shown in the
previous figures. While A and C have the same
species rihcness (all 3 species), sample C has the
greater evenness.
-------
GALLAGHER & GR ssLE
63
Figure 27 shows the position of sample points plotted using hypergeometric probabilities at NESSm
values from 1 to 40. Both the Pythagorean distance and city-block distance between the origin and
sample points increases monotonically. The city block distance is E(S ), and it is this distance that is
plotted in rarefaction curves (e.g., Figure 25). With increasing NESSm, the probability of sampling at
least one individual of the abundant ta.xa approaches 1.0. That is why both CNESS and NNESS
become relatively insensitive to changes in the abundance of dominant species at large random sample
sizes.
c_____-...
Ctordbis ancës
. -(1 tinltfronvorigin)’
Figure 27. CNESSm 10 is the Euclidean distance between sample
points I unit from the origin on the vectors connecting the origin
and sample positions set with The original species
frequencies are plotted as + ‘s.
Figure 27 shows the chord distances among sample vectors at a distance one unit from the origin.
These coordinates are from the row-normalized H matrix. The distances among these points are the
CNESS faunal distances. These distances also serve as chords along the unit hypersphere and are also
known as chord distances.
. 0 0.5
00
-------
64
EMAP-E VP COMMUNITY STRUC1’URE
Figure 28 shows the same configuration of
points as in Figure 27 after subtracting the
mean value for each species (centering).
Centering the normalized H matrix leaves the
distances among points unaltered but rigidly
translates the data points so that the centroid is
at the origin. If the data are not centered prior
to PCA, then the distances among stations are
reflected in the second and higher PCA axes.
The first PCA axis serves only to center the
data.
Figure 29 shows the results of the metric
scaling of the three sample points using
NESSm=1 (identical to performing a Principal
coordinates analysis of Orloci’s chord distance
or a PCA-H with NESSm=1) and NESSm=1O.
A Procrustes rotation (Digby and Kempton
1987) was used to rigidly rotate the PCA-H
(NESSm=1) ordination to fit the PCA-H
(NESSm=1 0) ordination.
I
C l )
(0
S.
C
a
0
-0.2
02
CNESS 1
02
0.1
) .....
CNESSCAI IO .
CNESSBCIII _IO
Figure 28. After centering the normalized data by
species, the CNESS distances from the previous plot
can be seen more clearly. Principal components
analysis will show the planar proj ection of this 3-d
configuration.
-0.2 0 0.2
PCA-H Axis 1 (72%)
Figure 29. The CNESS distances among sample points a:
NESSm= I (Orloci ‘s chord distance) (dotted lines I and CNESS
(NESSm=IO) (solid lines I. These plots are identical to those
produced using principal coorndinates analysis of Orloci ‘s chord
distance (=PCA-H with NESSm=I).
-------
GALLAGHER & GRASSLE
65
The Gabriel Euclidean distance biplots for the 0.8
two metric scalings shown in Figure 29 are
shown in Figures 30 and 31. Species 1, the
rarestofthethreespe cieSiflthedata 04
contributes virtually nothing to the CNESS
distances among samples (note that only 3%
of the CNESS variation is expressed on the
2nd PCA-H axis). 0
With NESSm=l0, Species I becomes an
important contributor to CNESS distances
(Figure 31)
Figure 32 shows the Gabriel covariance plot
of species vectors. The R-mode clustering of
the normalized H matrix using Pearson’s r is
mathematically equivalent to clustering
species using the cosine of the angles among
species vectors in the covariance biplot.
0.8
‘U ,
c’i
0.4
.2
0 0.
C
C.) .5
04
0
C.)
-S
2
3
B
A
-0.5 0 0.5
PCA-H Axis 1 (97%)
Figure 30. The Gabriel Euclidean distance biplot
showing the species contribution to CNESS distance
at NESSm=1. Species I contributes little to CNESS
distance). Only 3% of the variation is on Axis 2.
1
Sp. 2
0 0.5
PCA-H Axis 1 (72%)
Figure 31. The Gabriel Euclidean distance
biplot for CNESS (m=1O). Species 1, the rarest
species, now contributes much more to CNESS
distances among samples and PCA-H axis 2
explains 28% of the variance in CNESS.
Covariance plot Axis 1 (70%)
Figure 32. The covariance plot showing the
inve, se relationship between the frequencies of
Species 2 and 3. The cosine of angles among
species vectors can be clustered for na R-mode
analysis (see Trueblood et a!. 1994.)
-------
EMAP-E VP COMMUNITY STRUCTURE
APPENDIX II TERMS AND DEFINITIONS
Arch Effect (=Kendall’s horseshoe) The horse-shoe like shape produced when coenocline data are
analyzed using Q-mode ordination. The arch can be observed in most forms of PCA and non-
metric multidimensional scaling. The arch has at least two and perhaps three causes and no
clear-cut solutions other than plotting using the appropriate full dimensionality:
(1) In principal coordinate analysis and non-metric multidimensional scaling, the
similarity or dissimilarity values may “bottom out” so that samples sharing no species
cannot be ranked. Williamson’s step-across procedure has been proposed as a solution
to this problem.
(2) non-linearity in the data. Most Eigenanalysis procedures fit a linear, additive model to
the data. If species abundances are not linearly related to each other, the arch
phenomenon occurs.
(3) Inherently, high-dimension data. For example, when the frequency of heterozygotes,
homozygous recessive, and homozygous genotypes is analyzed by CA, a 2-
dimensional arched structure is produced.
Centered data Data presented as deviations from their mean value.
Centered SSCP matrix In standard PCA, data are usually (but not always standardized by the
mean (i.e., centered). For Q-mode analysis, the mean of each species is usually subtracted
from each species’ cell. The centered SSCP matrix is the sum of squares and cross products
matrix formed by multiplying the data matrix by its transpose.
Correlation A standardized form of covariance obtained by dividing the covanance of two
variables by the product of the standard deviations of x and y.
Covariance a measure of association between 2 variables; covariance is the mean of the cross
products of the centered data; expected value of the sum of cross products between 2 variables
expressed as deviations from their respective mean. The covariance between z-transformed
variables is also known as correlation.
DPCA-H Coats’ (1995) term for detrended Principal components analysis of hypergeometric
probabilities.
Eigenanalysis The process of finding the eigenvalue-eigenvector pairs of a square matrix A. The
eigenvalues are the elements of the diagonal matrix L and the eigen vectors are the columns of
U where A=U’LU.
eigenvalues (=characteristic values, latent values) a set of real or even imaginary scalars which can
be used with their associated eigenvectors as an alternate description of a square matrix A. An
N x N matrix A is said to have an eigenvector u and corresponding eigenvalue X if
Au = 2 u.
Every square, full-rank matrix A can be decomposed into a product of the diagonal eigenvalue
matnx L and eigenvector matrix U such that: A=U’LU, where U’ is the transpose of the U
matrix.
-------
GALLAGHER & GRASSLE 67
eigenvectors a column vector associated with its respective eigenvalue; Normalized eigenvectors of
unit length (sum of squares of elements equal 1.0) are the principal components. Right
eigenvectors UR satisfy:
AUR = AuR (12)
where, A is the eigenvalue associated with the eigenvector U, and A is a square matrix, left
eigenvectOrS 11 L satisfy:
ULA = ).UL (13)
Every left eigenvector is the transpose of a right eigenvector of the transpose of A. The left
and right eigenvalues are identical.
Graphical biplot Legendre and Legendre (1983) review this technique, introduced by Gabriel
(1971). Greenacre addresses the graphical biplot, or joint display in Correspondence analysis.
There are 3 types of graphical biplots. In the first, the variable loadings for the R-mode PCA
are normalized so that the sum of squares of loadings equal the eigenvalue for the ails. The
site scores are normalized so that the sum of squared PCA scores on each is are one. This is a
covariance biplot. In this scaling, the angle between arrows of each pair of species, plotted as
vectors provides an approximation of their pair-wise correlation, i.e., r=cosø. The orthogonal
projection of sites onto species vectors indicates the rank order of sites with respect to that
species.
In the second form of graphical biplot called the Eucidean distance plot, the eigenvectors
(species loadings) are standardized to unit sums of squares and the site scores are standardized
so that the sums of squares equals the eigenvalue of each axis (a normalized eigenvector times
the vector of observations will produce site scores with a sum of squares = A). This plot is
intended to preserve the Euclidean distances between sites and is called a Eudlidean distance
plot.
Greenacre (1984) calls both of these plots asymmetric, since the sites and variables are scaled
differently. The third type of graphical biplot is the symmetric biplot where sites and variable
vectors are scaled so that the sum of squared elements equals the eigenvalue.
Loadings The elements of the eigenvectOrS are also the weights or loadings of the various
original descriptors. If the eigenvectors have been normalized to unit length (i.e., the sum of
the squared loadings for a variable across factors equals 1.0), then the elements of the
eigenvector matrix (the loadings) are direction cosines of the angles between the original
descriptors and the principal axes. So that if the element of the U vector (the loading for a
variable) is .8944, the angle is cos’ (.8944)=arc cos(.8944)=26° (Legendre and Legendre
1983).
normalization A term often misused by environmental scientists. Normalization refers to the
standardization of an n-dimensional vector to unit length (i.e., a projection of a data point onto
the unit hypersphere). The etymology of normalization is from norm, the length of a vector.
There are an infinite number of eigenvectors associated with each eigenvalue. PCA and FA
normalize these eigenvectOrS to either the unit length or the square of the eigenvalue.
-------
68 EMAP.E VP COMMUNiTY STRUCTURE
Ordination Direct ordination The process of arranging sites (or species) in relation to one or
more environmental (or successional) gradients or to abstract axes representing such gradients.
Indirect ordination a collective term for continuous multivariate techniques which arrange
objects (e.g., sites or species) along axes, regardless of the interpretation of the axes.
Pielou (1984): Ordination is a procedure for adapting a multidimensional swarm of data
points in such a way that when it is projected onto a two-space (e.g., a sheet of paper) any
intrinsic pattern the swarm may possess becomes apparent.
orthogonal factors factors that are not correlated with each other.
Orthogonal matrix A square matrix that when used as a transformation matrix, causes a rigid
rotation of the data swarm without any change of scale. The product of an orthogonal matrix
and its transpose is the identity mathx (Pielou, 1984, p. 253): A’A=AA=I, where I is the
identity matrix.
PCA-H Principal component analysis of hypergeometric probabilities (Gallagher et a!.
1992, Trueblood et al. 1994.
principal component method Developed by Hotelling. PCA is simply the rotation of the original
system of axes in the multidimensional space. The principal axes are orthogonal and the
eigenvalues measure the amount of variance associated with each principal axis. PCA is used
to summarize in a few important dimensions the greatest part of the variability of a dispersion
matrix of a large number of descriptors (R mode) or cases (Q-mode).
principal component scores the value of a principal component for individual points, hence the
new coordinates of data points measured along axes created by the principal component
method. A principal component score can be regarded as an additional variable for each case,
this variable is a linear function of the original variables.
principal coordinates analysis An ordination based on a metric similarity or dissimilarity matrix.
Q-mode, R-mode Legendre & Legendre (1983, p. 172). The measurement of dependence
between two descriptors (variables) is achieved my means of coefficients like Pearsons
product-moment correlation, r. This type of study of the data matrix is therefore called an R
analysis. In contrast, a study of an ecological data matrix based upon the relationship between
objects is called Q analysis. Many authors (e.g., Pielou 1984) reverse this conventional usage.
SSCP the sum-of-squares-and-cross-Products matrix, the matrix formed by multiplying a matrix
times its transpose. The (i,i)th element is the sum of squares of the ith variable. The (h,i)th
element is the sum of cross-products of the h’th and ith variables.
-------
GALLAGHER & GLIS5LE 69
standardization- an algebraic operation (e.g., x 1 /(standard deviation of x ) performed on a
variable or site vector to achieve a desired property (e.g., non-dimensionality, common
variance). Standardization requires calculation of the row or column sums of a data matrix.
Data measured on different scales must be standardized prior to analysis. Norm
standardization is dividing each element by f (x 1 ) 2 . If the data have been previously
centered, then dividing the centered variables by the norm is equivalent to dividing the original
variable by the standard deviation.
transformation A transformation can be performed without knowledge of the row or column
sums of a data matrix. A standardization requires such knowledge.
variance a measure of the dispersion of a variable; defined as the sum of squared deviations
from the mean divided by the number of cases or entities.
C
-------
70
EMAP-E VP COMMUNITY STRUCTURE
APPENDIX ifi FULL EMAP-E VIRGINIAN PROVINCE MOD wthD FAUNAL LIST
This appendix shows the full EMAP-E VP faunal list. It also shows which EMAP-E VP faunal
groupings must be dropped and those that must be pooled to replicate the community structure
analyses in this report.
EMAP-E VP SPEcIES CODES
VALiD TAXA
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
I
AMPHARCT
mpharee arctica
Annelida
Polychacta
Ampharetidac
2
ANOBGRAC
nobothrus gracilis
Annelida
Polychacta
Ampharetidac
3
ASABOCUL
sabellides oculata
Annelida
Po lychaeta
Ampharetidae
4
HOBSFLOR
robsoniaflorida
Annelida
Po lychaeta
Ampharetidae
5
MELIMACU
felinna maculasa
Annehda
Polychaeta
Ampharetidae
6
PSEUPAUC
. ta
Annelida
Po lychaeta
Amphinomidae
7
ARABSPEA
rabeilidae sp. A Morris
Annelida
Polychacta
Arabe l lidae
8
DRILLONG
)rilonere:s longa
Annelida
Po lychaeta
Arabel lidae
9
DRILSPEB
B
Annelida
Po lychaeta
Arabellidac
10
II
NOTOSPIN
Vorocsrrus spin jferus
Annelida
Po lychaeca
Arabellidae
AMASCAPE
masngos caperatus
Annelida
Polychacta
Capitellidac
12
HETEFILI
eteromastusfilifornus
Annelida
Polychaeta
Capuellidac
13
MEDIAMBI
ledioniastus ambiseta
Annelida
Polychacla
Capitel lidae
14
MEDICALI
I thomt1stUS
ai iforn,ensis
Annelida
Po lychaeta
Capite l lidae
15
NOTOLOBA
fotomasrus lobazus
Annelida
Po lychaeta
Capite l lidae
16
NOTOLURI
fotomastus luridus
Annelida
Polychacta
Capite l lidae
17
NOTOSPA
foromastus sp A Ewing
Annelida
Polychacta
Capite llidae
18
CHAEVARI
ariopedatus
Annelida
Polychaeta
Chaetoptendae
19
SPIOCOST
:: t0Pt
Annelida
Po lychaeta
Chaetopteridae
20
BHAWHETE
hawan :a heteroseta
Annelida
Po lychaeta
Chrysopeta lidae
. . L.
CAULBIOC
aulieriella cf bioculata
Annelida
Po lychaeta
Cirratulidae
22
CAULSPEB
aulieriella sy. B Blake
Annelida
Po lychaeta
Cirratulidac
23
CIRRGRAN
irriform :a grand :s
Annelida
Po lychaeta
Cirratu lidae
24
THARACUT
‘harx acutus
Annelida
Polychaeta
Cirratulidae
25
TIIARSPA
‘harvx sp. A Morris
Annelida
Po lychaeta
Cirratulidae
26
COSSSOYE
:ossura longocirrata
Annelida
Polychacca
Cossuridae
27
DORVRUDO
)orvillea rudolph:
Annehda
Po lychaeca
Dorvilleidae
28
DORVSPEA
)orv:lieidae sp. A Hilbig
Annelida
Polychaeca
Dorvilleidae
29
MEIOSPEA
le,odorv:llea sr. A
lorris
Annelida
Po lychaeza
Dorvi l leidae
30
PAROCAEC
‘arougia caeca
Annelida
Polychacia
Dorvi l leidae
31
PROTKEFE
‘rotodorvillea keferstein
Annelida
Po lychaeta
Dorvi lleidae
32
MARPBELL
4arphysa belli
Annelida
Polychacia
Eunicidae
33
MARPSANG
farphvsa sanguinea
Annelida
Polychaeia
Eunicidae
34
BRAD VILL
frada viliosa
Annelida
Polychaeta
Flabelligeridac
-------
( At ,Amlvg & GRASSLE 71
No
CODE
GENUS SPP.
PHYLYUM
CLAss
FAMILY
35
PHERAFFI
herusa t ffims
Annelida
Polychacta
Flabelligendae
36
GLYCAMER
lycera americana
Annelida
Polychaeta
Glycendac
GLYCDIBR
;& era dibranchiata
Annelida
Polychaeta
Glyceiidae
38
GLYCROBU
;frera robusta
•
Annelida
Polychaeta
Glycendac
39
HEMIROSE lemipodus roseus
Annelida
Polychaeta
Glycendae
40
GLYCSOLI
ilvcinde solitaria
Annelida
Polychacta
GONIGRAC
on:adeila gracilis
Annehda
Polychacta
Goniadidae
42
OPHIGIGA
)phioglicera gigantea
Annelida
Polychaeta
Goniadidac
43
GYPTV1TT
vpris crypta
Annelida
Polychacta
Hesionidae
44
MICRABER
imtu
Annelida
Polychacta
Hesionidae
45
46
47
48
MICRFRAG
licropluhalmusfragilis
Annelida
Po lychaeta
Hesionidae
MICRSCZE
.MICRSIMI
htcrophthatmus
dicrophthalmus similis
Annelida
Annelida
Po lychaeta
Polychaeta
Hesionidae
Hesiorndae
PARALUTE
‘arahesione luteola
Annelida
Polychacta
Hesionidae
49
PODAOBSC
‘odarke obscura
Annelida
Polychaeta
Hesionidae
50
PODALEVI
‘odarkeopsis Iev:fiiscina
Annelida
Polychaeta
Lumbrrneridae
51
NINONIGR
inoe nigripes
Annelida
Po lychaeta
52
LUMBACIC
coletoma acicularum
Annelida
Polychaeta
Lumbnneridae
53
SCOLBEBE
ëoletoma hebes
Annelida
Polychaeta
54
LUMBTENI
coletoma :enuis
Annehda
Polychacta
Lumbnnendae
Ma ldanidae
55
CLYMTORQ
ivmenella zorguata
Annelida
Po lychaeta
Maldanidae
56
MACRZONA
.iacroclvmene zonalis
Annelida
Polychaeta
Maldanidac
57
SABAELON
abaco elongatus
Annelida
Po lychaeta
58
AGLACIRC
glaophasnus circinata
Annelida
Polychaeta
Nephtyidae
59
AGLAVERE
glaophamus verrilli
Annelida
Polychaeta
Nepbtyidae
60
NEPHBUCE
Iephrys bucera
Annelida
Po lychaeta
Nephtyidae
61
NEPHCRYP
iephrvs crvptomma
Annelida
Polychaeta
Nephtyidae
62
NEPHINCI
‘ephzvs incisa
Annelida
Po lychaeta
63
NEPHPICT
‘ephtvs picta
Annelida
Polychacta
Ncreididae
CERAIRRI
erazonereis irritabslis
Annehda
Po lychae ta
Nereididae
65
LAEOCULV
aeonereis culvert
Annelida
Po lychaeta
66
NEANAREN
e°c jentata
Annelida
Po lychaeta
Nercididae
Nereididac
67
NEANSUCC
leant he: succinea
Annelida
Po lychaeta
Nereididae
68
NEANVIRE
leanthes wrens
Annelida
Polychacta
Nereididae
69
NEREGRAY
leress gravE
Annelida
Po lychaeta
Nereididae
70
PLATDUME
‘latvnereis dumerilis
Annelida
Polychaeta
Onuphidae
71
DIOPCIJPR
)iopatra cuprea
Annelida
Polychacta
Onuphidae
72
ONUPEREM
)nuphts eremzla
Annelida
Polychaeta
73
OPHEBICO
phelia bicornis
Annelida
Po lychaeta
Opheliidae
74
OPHEACUM
ina acuminata
Annelida
Polychacta
Ophe liidac
75
TRAVSPEA
ravisia 5,1. A Morris
Annelida
Polychacta
76
TEA VSPEB
ravissa sp B Morris
Annelida
Polychae ta
Orbiniidae
77
LEITFRAG
eztoscoloplosfragilis
Annelida
Polychaeta
—
Orbiniidae
78
LEITROBU
.eizoscoloploS robustus
Anne lida
Polychacta
.22.
ORBIRISE
)rbinsa risen
Annelida
Po lychae ta
Orbiniidae
Orbiniidae
..!
ORBIS WAN
)rbinea .cwani
Annelida
Polychacta
Orbiniidae
IL
SCOLCAPE
colonlos car,ensis
Annelida
Polvchaeta
-------
72 ___ EMAP.E VP COMMUNITY STRUCTURE
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
82
SCOLRUBR
coloplos rubra
Anneida
Po lychaeta
Orbiniidae
83
MYRIOCUL
aIa:howenia oculata
Annelida
Po lychaeta
Oweniidae
84
OWENFUSI
wenzafiss:7ormis
Annelida
Po lychaea
Owcniidae
85
ARICCATH
ricidea cathennae
Annelida
Po lychaeta
Paraonidae
86
ARICCERU
#cdea cerrutti
Annelida
Potychaeta
Paraonidae
87
ARICFRAG
ricideafragilLs
Annelida
Polychacla
Paraonidae
88
ARJCWASS
ricidea wass:
Annelida
Po lychaeta
Paraonidae
89
CIRRSPEA
irrophorus sp. A Morris
Annelida
Po lychaeta
Paraonidae
90
CIRROSPB
irrophorus sp B Morris
Annelida
Polychaeza
Paraorndae
91
LEVIGRAC
evinsen :a gracths
Annelida
Polychaeta
Paraonidae
.2 ..
LEVISPEA
evinsenia sp A Morris
Annelida
Polychacta
Paraonidae
93
PARADSPB
aradoneis sp B Morris
Annehda
Po lychaeta
Paraonidae
94
PARAFULG
aroonisfulgens
Annelida
Polychaeta
Paraonidae
95
PARAPYGO
araonis pygoentgniaflcc
Annelida
Po lychaeta
Paraonidae
.2 ..
PECTGOUL
ect:naria gouldil
Annelida
Polychacta
Pectinanidae
97
EUMISANG
um:da sanguinea
Annelida
Polychacta
Phyllodocidac
98
HESIELON
fesionura elongata
Annelida
Polychaeta
Phy llodocidae
99
ETEOFOLI
rypereteonefohosa
Annelida
Polychacta
Phyllodocidae
100
ETEOHETE
Ivpere:eone heteropoda
Annelida
Polychaeta
Phyllodocidae
101
HYPELONG
Tvperereone longa
Annehda
Polychaeta
Phyl lodoc idae
102
PARASPEC
aranaitis speciosa
Annelida
Polychaeta
Phyllodocidae
103
PHYLAREN
hvllodoce arenae
Annelida
Po lvchaeta
Phy llodocidae
104
PHYLMACU
hvllodoce maculata
Annelida
Polychaeta
Phy llodocidae
105
PHYLMUCO
hvllodoce mucosa
Annelida
Polychaeta
Phyllodocidac
106
ANCIIIART
ncistrosllis hartmanae
Annelida
Polychaeta
Pi larg idae
107
ANCUONE
nc:stros’ylhsjones :
Annelida
Polychaeta
Pi largidae
108
CABUNCE
ab,ra ncerra
Annelida
Polychaeta
Pi larg idae
109
SIGABASS
‘igwnbra bass:
Annelida
Polychaeta
Pilargidae
110
SIGATENT
igambra tentaculala
Annelida
Polychacta
Pilargidae
III
PISIREMO
isione reinota
Annelida
Polychacta
Pisionidac
112
HARMEXTE
larn,othoe extenuata
Annelida
Polychaeta
Po lynoidae
113
HARMIMBR
armothoe :mbr :ca:a
Annelida
Polychaeta
Polynoidae
114
HARMMACG
Varmothoe macgimne:
Annelida
Po lychaeta
Po lynoidae
115
HARTMOOR
rarzmwua moore:
Annelida
Po lychae la
Polynoidae
116
LEPICOMM
“ ‘
ommensalis
Annelida
Polychaeta
Po lynoidae
117
LEPISQUA
ep:donotus squamatus
Annelida
Polychacta
Po lynoidae
118
LEPISUBL
ep:donozus sublevis
Annelida
Polychaeta
Po lynoidae
119
LEPIVARI
epidonotus variab :I,s
Annelida
Polychaeta
Po lynoidae
120
MALMSPA
1aImg h1a SP• A
Veston
Annelida
Polychaeta
Polynoidac
121
MALMSPEB
falmgreniella P B
Veston
Annehda
Polychacta
Po lynoidae
122
PROTCRAE
rotodriosdes chae (er
Annelida
Polychaeta
Protodrilidae
123
SABEVULG
àbeliana vulgar :s
Annelida
Polychaeta
Sabe l lariidae
124
CHONINFU
hone :nfund:bulifornns
Annelida
Polychacta
Sabe l lidae
125
DEMOMICR
)emonax
ucrophihalmus
Annelida
Polychacta
Sabe l lidae
126
EUCHELEG
uchone elegans
Annelida
Polychaeta
Sabe l lidae
127
EUCHINCO
uchone incolor
Annehda
Polychaeta
Sabellidae
-------
f AI T £flU P Rr Cv&cc 73
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
128
LAONKROY
aonome kroeyeri
Annelida
Po lychaeta
Sabe l lidae
129
MANAAEST
1anavunlcia aestuarina
Annelida
Po lychaeta
Sabe l lidae
130
MYXIINFU
yxicoIa infundibulum
Annelida
Po lychaeta
Sabe l lidae
131
132
j
I’SEURENI
SCALINFL
s1 ,ulla
calibregma inflatum
Annelida
Annelida
Po lycbaeta
Po lychaeta
Sabe l lidae
Sca1ibre matidae
PHOLMINU
holoe sninura
Annelida
Polychacta
Sigalionidae
134
SIGAAREN
igalion arenicola
Annelida
Po lychaeta
Sigalionidae
135
STENBOA
thenelais boa
Annelida
Po lychaeta
Sigalionidae
136
STHELThII
thenelais hm coIa
Annelida
Polychacta
Siga lionidae
137
APOPPYGM
poprionosplo pygmaea
Annelida
Polychacta
Spionidae
138
BOCLUAMA
roccardiella haniata
Annelida
Po lychaeta
Spionidae
139
BOCCLIGE
Ioccardiella ligerica
Annelida
Po lychaeta
Spionidae
140
CARAHOBS
arawella hobsonae
Annelida
Po lychaeta
Spionidae
141
DISPUNCI
)ispio uncinata
Annelida
Po lychaeta
Spionidae
MAREVIRI
larenzeileria viridis
Annelida
Po lychaeta
Spionidae
143
PARAPINN
‘araprionospio pinnara
Annelida
Polychaeta
Spionidae
144
POLYAGGR
o1vdora aggregata
Annelida
Po lychaeta
Spionidae
145
POLYCAUL
‘olydora caullerw
Annehda
Po lychaeta
Spionidac
146
POLYCORN
‘olvdora cornuta
Annelida
Po lychaeta
Spionidae
147
POLYGIAR
‘olydora giard
Annelida
Polychaeta
Spionidae
148
POLYQUAD
olvdora guadrilobata
Annelida
Polychacta
Spionidae
149
POLYSOCI
olvdora socialis
Annelida
Po lychaeta
Spionidae
150
POLY WEBS
oI vdora webster,
Annelida
Polychaeta
Spiorndae
151
PRIOIIETE
IIOSPIO
:e:erobranchia
Annelida
Po lychaeta
Spionidae
152
PRIOPERK
rionospio perkinsi
Annelida
Polychaeta
Spionidae
153
PRIOSTEE
‘rionospto steenstrupi
Annelida
Polychacta
154
PYGOELEG
‘ygospio elegans
Annelida
Polychaeta
Spionidae
155
SCOLBOUS
colelepis bousfieldi
Annelida
Polychaeta
156
SCOLQUAD
co1elepis guadrilobata
Annelida
Po lychaeta
157
SCOLSQUA
co1e1ep:s squamara
Annelida
Po lychaeta
Spionidae
158
SCOLTEXA
colelepis lexana
Annelida
Polychacta
159
SPIOFILI
piofihicornis
Annelida
Polychaeta
160
SPIOLIMI
pio liirncola
Annelida
Polychacta
Spionidae
161
SPIOSETO
pzo setosa
Annelida
Po lychaeta
Spionidae
162
SPIOBOMB
piophanes bombvx
Annelida
Polychaeta
163
STREBENE
reblosp:o benedict,
Annelida
Po lychaeta
164
STERSCUT
:ernaspis sculatus
Annelida
Po lychaeta
Syllidac
AUTOSPEA
utolvtus sp. A Glasby
Annelida
Polychaeta
Sy lhdae
166
BRANWELL
fransa wehlfleetensss
Annelida
Po lychaeta
167
EXOGDISP
xogone dispar
Annelida
Po lychaeta
Sy II,dae
368
EXOGHEBE
xogone hebes
Annelida
Polychaeta
169
EXOGSPEA
xogone sp. A Glasb
Annelida
Polychacta
170
EXOGVERU
xogone verugera
Annelida
Polychacta
Syllidac
17.!
ODONFULG
donrosy1lisfr1gurans
Anne lida
Polychaeta
Sy l lidae
172
PARALONG
ongicirrata
Annelida
Polychacta
Sy l lidae
173
PIONSPEA
‘ionosvllis sp. A Glasby
Annelida
Polychacta
Syl lidae
PIONSPEB
ionosv1hs si’ B GIasbv
Annelida
Po lychaeta
-------
74 EMAP.E VP COMMUNiTY STRUCTURE
No
175
CODE
GENUS SPP.
PIIYLYUM
CLAss
FAMILY
PROCCORN
roceraca cornuta
Annelida
Po lychaeta
Syllidac
176
SPIIAACIC
phaerosyllis ac cuIata
Annelida
Po lychaeta
Sy l lidae
177
SPIIATAYL
phaeroryllis taylori
Annelida
Polychacta
Syllidac
178
STREAREN
-eptosy1lis arenae
Annelida
Po lychaeta
Syllidac
179
STREPETT
reptosy1lispettiboneae
Annelida
Polychaeta
Syflidac
180
STREVARI
weptosyllis varians
Annelida
Po lychaeta
Syflidac
181
SYLLCONV
vilides convolusa
Anneida
Po lychaeta
Syllidac
182
SYLLVERR
vilides verriili
Annelida
Polychacta
Syllidac
183
AMPHORNA
rnphftrite ornata
Annelida
Polychaeta
Terebe l lidae
184
ENOPSANG
flOPlObrwIChUS
Annelida
Polychaeta
Terebelhdae
185
LOIMMEDU
otmia medusa
Annelida
Polychaeta
Terebe l lidae
186
NICOZOST
Vicolea zostericola
Annelida
Polychacta
Terebeltidae
187
PISTCRJS
Lua cristata
Annelida
Po lychaeta
Terebel lidae
188
PISTPALM
istapalmata
Annelida
Po lychaeta
Terebe l lidae
189
POLYHAEM
OIYCZfl•US
aematodes
Annelida
Polychacta
Terebellidac
190
POLYEX!M
olvcirrus exunius
Annelida
Polychaeta
Terebe llidae
191
POLYMEDU
olvcimis medusa
Annelida
Polychaeta
Terebel lidae
192
TERESTRO
erebellides su-oemi
Annelida
Polychaeta
Trichobranchidae
193
TROCMULT
rochochaeta multisetow
Annelida
Po lychaeta
Trochochaetidae
194
POLYSPEA
olchaera sp A Arcuri
Annelida
Polychaeta
Unidentified
195
POLYSPEB
olvchaeta sp B Arcur,
Annelida
Po lychaeta
Unidentified
196
AMPEAGAS
mpelisca aRassizi
Arthropoda
Amphipoda
Ampe liscidae
197
AMPEVERR
ntpelisca verr,lii
Anhropoda
Amphipoda
Ampe liscidae
198
BYBLSERR
lyblis serrwa
Arthropoda
Amphipoda
Ampe lisc idae
199
AMPILONG
mpithoe longunanna
Arthropoda
Amphipoda
Ampiihoidae
200
AMPIVALI
mp,thoe vahda
Arthropoda
Aniphipoda
Ampithotdae
201
CYMACOMP
‘vmadusa compta
Arthropoda
Amphipoda
Ampithoidae
202
LEMBSM1T
embos smith:
Arthropoda
Amphipoda
Aoridae
203
LEMB WEBS
.embos websteri
Arthropoda
Amphipoda
Aorzdae
204
LEPTPING
£prochelrus pinguis
Arthropoda
Amphipoda
Aoridae
205
LEPTPLUM
eptocheirusplumulosus
Arthropoda
Amphipoda
Aoridae
206
MICRANOM
(icrodeutopus anomalus
Arthropoda
Amphipoda
Aoridae
207
MICRGRYL
Arthropoda
Amphipoda
Aondae
208
PSEUOBLI
‘seudunciola obliguua
Arthropoda
Amphipoda
Aoridae
209
RUDINAGL
udi1emboides nagle,
Anhropoda
Amphipoda
Aoridae
210
UNCIDISS
Inciola dissimilis
Arthropoda
Amphipoda
Aoridae
UNCIINER
Inciola inerm:s
Arthropoda
Amphipoda
Aoridae
212
UNCIIRRO
Inciola irrorata
Arthropoda
Amphipoda
Aondae
213
UNCISERR
Inciola serrata
Arthropoda
Amphipoda
Aondae
214
ARIGHAMA
lr,gissa hamaupes
Arthropoda
Amphipoda
Argissidac
215
BATECATH
atea catharinensis
Arthropoda
Amphipoda
Bateidae
216
CALLLAEV
‘a11iopius laevsusculus
Arthropoda
Amphipoda
Ca lhopidae
217
COROACHE
orophium acherusicum
Arthropoda
Amphipoda
Corophiidae
218
COROACUT
orophium acutum
Arthropoda
Amphipoda
Corophiidae
219
COROBONN
orophium bonelli:
Arthropoda
Amphipoda
Corophiidae
220
COROCRAS
orophium crassicorne
Arthropoda
Amphipoda
Corophiidae
221
COROINSI
;oroyhium ,ns,diosuni
Arthro oda
Amohinoda
Coronhiidae
-------
I’ • ‘tI fl 1,
75
No CODE_
222 COROLACIJ o
GENUS SPP.
PHYLYUM
Arthropoda
CLASS
Amphipoda
Corophzidae
rophiwn lacustre -
Aithropoda
Amphipoda
Corophiidae
22 COROSEXT. o
rophium sextoni
Azthropoda
Amphipoda
Corophildae
224
COROSIMI ‘orophiwn simile —
tuberculaiwn
AithrOVOd a
Amphipoda
Corophiidae
225
COROTUBE orophium
Arthropoda
Ajnphipoda
Dcxaminidae
226
DEXATIIEc )examine thea
—
—
Azthropoda
Amphipoda
Gammaridae
227 GAMMANN!J ammarus annul atus
Arthropoda
Amphipoda
Gammandae
228 GAMMDAIB
229 GAMMFASC
ammarusdaiberi
ammarusfasciatus
oceanicus
— Arthmpoda
Arthropoda
Amphipoda
Amphipoda
Gammaridae
Gammandae
230 GAMMOCEA ammaruS
Amphipoda
Gammaridae
231 MUCRMUCR
232 ACANMILk canthohaustorus mdlxi
233 ACANSIMI.. canthohauStOrius simik
234 BATIIPARK_ a:hyporeia parkeri —
Aithropoda
— Arthropoda
— Arthropoda
— Arthropoda
Arthropoda
— Amphipoda
Amphipoda
Amphipoda
Amphipoda
Hauscoriidae
Haustoriidae
Haustonidae
Haustorudae’
235
LEPIDYTI p dactvlus dvnscus_
—
Amphipoda
Haustoriidae
236
237
PARAATFE
PARAHQI M..
arahaustoriUS holmesi
Arthropoda
— Arthr opOda
Amphipoda —
- Haustoriidae -
Hausionidac
238
240
PARALNGI —
FROTDEIC rot ohaustorius cf
PILOTWIGL rotohaustortus wsgleyi
Arthropoda
Arthropoda
Arthropoda
Amphipoda
Amphipoda
AmphipOda
Haustorudae
Haustonidae
Hausionidae
241
242
243
2
PSEUBORE
PSEUCARO
GAMMSTJTII
MICRRANE
—
seudohaUStOrIUS
ammaropsls su:herlal!4
ticrOPrOtOPUS ranevL
dentata
Ailhrop oda —
Arthropoda —
Arthropoda —
— ArthropOda
Anh iopOda
Amphipoda
Amphipoda
— Amphipoda
Amphipoda
Haustoriidae
Isacidac
isacidac —
Isacidac
245
PHOTDENT
hons
—
Arihropoda
Amphipoda
Isacidac
246
PHOTPOLk
hotis pollex
‘holts
Arthropoda —
Amphipoda
isacidac
247
PHOTPUGN_
pugnator
tubularis
Arthropoda
Arnphipoda
1sch ocer1dae
248
CERATUBU_
erapus
—
brasiltensis
Arthropoda
—
Amphipoda —
Ischyroccr idac
249
Q
ERICBRAS_
ERICFASC
r,c:honsus
ricrhoniusfasC latus
Arthropoda —
Arthropoda
— Amphipoda —
Amphipoda
ischyrocciidae
Ischyroceridae
ISCHANGQ
tchvrocerus anguipes
marmorata
Arthropoda
Amphipoda
Ischyroccndac
253
JASSMARM
LISTBA L.
assa
.,strsella barnarth
clvmenellae
Arthro pOda
ArthropOda
Amphipoda
Amphipoda
L i i 1 eborgiidae
Lii jeborgiidac —
254 LISTCLYM.
—
sirte1la smithi
—
Arthro pOda
—
Amphipoda
Liljebcxgiidae —
255 LISTSMIT_
Arthropoda
Amphipoda
Lysianassidac
256 ANONLILJ_ nonrx lilleborRi
serratut
—
ArihropOda
Amphipoda
Lysianassidac
257 HIPPSERR .. lippomedon
alba
Arthropoda
Amphipoda
— Lysianassidae
258 LYSIALBA
tsIanopSlS
)rchomenella minwa
ArthropOda
—
Amphipoda —
LysianasSidac —
259 ORCHMINtL
Amph ipoda Mdliuidac
260 DULL4PPE a:a
261 ELASLAEV lasniopus laevis
nuida
Anhropoda
ArthropOda
Arthropoda
Amphipoda
Amphipoda
—
Meiiudae
— Me litidae —
262 MELINITI_
—
Ocdiccrouidac
MONOSPE1
Arthropoda Amphipoda
—
-------
76 EMAP-E VP COMMUNITY STRUCTURE
No
264
CODE
GENUS SPP.
PHYLYuM
C lAss
FAMILY
SYNCAMER
Azthropoda
Amphipoda
Oediceroüdae
265
EOBRSPIN
brolRus 5PUIOSILc
Ailhiwoda
Ainphipoda
Phoxocephalidac
266
HARPPROP
Iarpinsaproputqua
Arthropoda
Amphipoda
Phoxocepha lidae
267
PIIOXHOLB
hoxocephalus holbolli
Azthropoda
Amphipoda
Phoxoccphalidae
268
RHEPEPIS
hepoxynius epistomus
Arthropoda
Amphipoda
Phoxocephalidac
269
RHEPHUDS
‘Jieporynius hudsoni
Arthropoda
Amphipoda
Phoxocephalidac
270
PARAAEST
‘arapleustes aestuaruis
Arthropoda
Amphipoda
P leusudae
271
PLEUGLAB
‘leusynues Richer
Arthropoda
Amphipoda
P leustidae
272
STENGRAC
enopleustes Rradii.c
Anhropoda
Amphipoda
P leustidae
273
STENINER
enop1eus:es inernus
Anhropoda
Amphipoda
Pleustidac
274
DYOPMONA
)yopedos monacanthu.s
Arthropoda
Amphipoda
Podocendae
275
PARACYPR
‘arametopella cyprs
Arthropoda
Amphipoda
Stenothoidae
276
STENMINU
enothoe muiuta
Azthropoda
Amphipoda
Stenothoidac
277
STENVALI
enothoe valida
Ailhrcpoda
Amphipoda
Stenothoidae
278
HUTCMACR
Arthropoda
Cephalocanda
Hutchrnsonie l lida
279
ABLAPARA
blabesmy :a parajanta
Arthropoda
Chironomidae
Tanypodinae
280
PROCSUBL
rocladius subleuei
Arthropoda
Chironomidae
Tanypodinae
281
BODOSPEA
odo:ria sp. A Morris
Arthropoda
Cumacea
Bodotrudae
282
CYCLVARI
yclaspis vartans
Arthropoda
Cumacea
Bodotnidae
283
MANCSTEL
fancocuma srellifera
Arthropoda
Cumacea
Bodoiriidae
284
BODOTRII
seudoleptocuma minor
Arthropoda
Cumacea
Bodotnidae
285
PSEUMINO
seudoleptocuma minor
Arthropoda
Cumacea
Bodotriidae
286
DIASQUAD
)iasrvlis guadnsp,nosa
Arthropoda
Cumacea
Diastyl idae
287
DIASSCUL
iasrv1is scuipta
Arthropoda
Cumacea
Diastyl idae
288
OXYUSMIT
)zvurosiylis smith:
Arthropoda
Cumacea
Diaslylidae
289
EUDOPUSI
udoreila pusilla
Arthropoda
Cumacea
Leuconidae
290
LEUCAMER
eucon wnericanus
Arthropoda
Cumacea
Lcucomdae
291
ALMYPROX
lmyracumaprox :moculi
Arthropoda
Cumacea
Nannastacidae
292
ALPHHETE
Ipheus heterochaelis
Arthropoda
Decapoda
Alpheidae
293
AUTOMSPA
utomwe sp. A Williams
Arthropoda
Decapoda
A lpheidae
294
CALLSETI
1Iianassa setimanus
Arthropoda
Decapoda
Ca l lianassidae
295
CRANSEPT
ranRon sep:emsp:nosa
Arthropoda
Decapoda
Crangonidae
296
LIBIEMAR
ibi,ua emargmata
Arthropoda
Decapoda
Majidae
297
OGYRALPJI
) yrides alphaeros:ns
Arthropoda
Decapoda
Ogyrididae
298
PAGUACAD
‘agurus acadianus
Arthropoda
Decapoda
Paguridac
299
PAGUANNU
agurus annulipes
Arthropoda
Decapoda
Pagundae
300
PAGTJLONG
agurus long:carpus
Arthropoda
Decapoda
Pagundae
301
PAGUPOLL
ORUFUS pollicar,s
Arthropoda
Decapoda
Pagundae
302
EUCEPRAE
uceramus pradongus
Arthropoda
Decapoda
Porcel lanidae
303
POLYGIBB
olyonyx gibbesi
Arthropoda
Decapoda
Porcel lanidae
304
OVALOCEL
vaIipes acellatus
Arthropoda
Decapoda
Potiunidac
305
PROC VICI
rocessa vicina
Arthropoda
Decapoda
Processidae
306
UPOGAFFI
fpogebia affinis
Arthropoda
Decapoda
Upogebiidae
307
NEOPSAYI
yspanopeus saw
Anhropoda
Decapoda
Xanthidae
308
HEXAANGU
Arthropoda
Decapoda
Xanthidae
309
PANOHERB
anopeus herbs:::
Arthrcpoda
Decapoda
Xanthidae
310
RHITHARR
h:rhropanopeus harr:s,i
Arthronoda
Deca oda
Xanthidae
-------
GALLAGHER & GIUSSLE - 77
No
311
312
313
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
APANMAGN
CYATBURB
rnMuswuhUni
vathura burbancki
Arthropoda
Arthropoda
Isopoda
Isopoda
Anthuridac
Anthuridae
CYATPOLI
yathura polila
Azthropoda
Isopoda
Anthuridae
314
PTILTENU
‘rilanthura tenuis
Arthropoda
Isopoda
Anthuridae
315
POLIPOLI
o1lio1wza polita
Aiihropoda
Isopoda
Cirolanidac
316
CHIRA.LMY
luridotea almyra
Arthropoda
Isopoda
ldoteidae
317
CHIRCOEC
hindo:ea coeca
Azthropoda
Isopoda
ldoteidae
318
EDOTTRJL
4otea inloba
Arthropoda
Isopoda
Idoteidac
319
ERICATTE
richsonella attenuwa
Anhropoda
Isopoda
Idoceidac
320
ERICFfl.I
r,chsoneliafiliformis
Arthropoda
Isopoda
Idoteidae
321
IDOTBALT
dotea baUheca
Arthropoda
Isopoda
Idoteidae
322
IDOTPIIOS
dozea phosphorea
Arthropoda
Isopoda
Idotcidae
323
JAERMARI
!aem marina
Arthropoda
Isopoda
Janiridae
324
PLEUINER
‘leurogontum inerme
Arthropoda
[ sopoda
Munnidae
325
326
327
PLEUSPIN
ANCIDEPR
ncinus depressus
Arthropoda
Arthropoda
Isopoda
Isopoda
Munnidae
Sphaeromatidae
CASSOVAL
assidinidea ovalis
Arthropoda
Isopoda
Sphaeromaudae
328
PARACAUD
aracerceis caudaza
Arthropoda
Isopoda
Sphaeromatidae
329
SPHAQUAD
atum
Arthropoda
Isopoda
Sphaeromatidae
330
CALLBREV
a11ipaliene brevirostris
Arthropcda
Pvcnogonida
Callipallenidae
331
ANOPPETI
noplodacrdus penolaw
Arthropoda
Pvcnogonida
Phoxichi lidiidae
332
TANYORBI
anvsiylum orbiculare
Arthropoda
Pycnogonida
Tanystylidae
333
NANNGRAY
(annosguilla gray:
Arthropoda
Scomatopoda
Nannospuillidae
334
SQUIEMPU
guilla empusa
Arthropoda
Stomatopoda
Sguillidae
335
LEPTDUBI
eptochelia dubia
Arthropoda
Tanaidacea
Nototanaidac
336
TANAPSAM
anaissus psammophilus
Anhropoda
Tanaidacea
Nototanaidae.
337
HARGRAPA
larger,a rapax
Arthropoda
Tanaidacea
Paratanaidae
338
TANASPEA
SP• A
Arthropoda
Tanaidacea
Tanaidacea
Polycentropodidae
Cenanthidae
339
CYRNFRAT
vrnellusfrazernus
Arthropoda
Trichoptera
340
CERIAMER
tw ae0 1 51 5
inericanus
Cnidana
Anthozoa
341
CAUDAREN
audina arenata
Echinodermata
Holothuroidea
- Caudirndae
342
STERUNIS
lereoderma unisemita
Echinodermata
Holothuroidea
Cucumariidae
343
HAVESCAB
avelock,a scabra
Echinodermata
Holothuroidea
Phy l lophoridae
344
PENTPULC
enlamera pulcherrinta
Echinodermata
Holothuroidea
Phil lophondac
345
LEPTTENU
eptosvnapta tenuis
Echinodermata
Holothuroidea
Synaptidac
346
SACCKOWA
àccoglossus kowalevskv
Hemichordata
Hemichordata
Harrimarnidae
347
STERCAND
tereobalanus candensis
Hemichordata
Hemichordata
Hammanndae
348
ANADOVAL
nadaraovalis
Mollusca
Bivalvia
Arcidac
349
ANADTRAN
nadara vransversa
Mollusca
Bivalvia
Arcidae
350
ARCTISLA
rclzca islandtca
Mollusca
Bivalvia
Arcticidae
351
ASTACAST
Lstare castanea
Mollusca
Bivalvia
Astanidae
352
ASTACREN
sfarte crenata
Mollusca
Bivalvia
Astartidae
353
ASTASPEA
starre sp. A Mountford
Mollusca
Bivalvia
Astartidae
ASTAUNDA
I ciarre undata
Mollusca
Bivalvia
Astartidae
-------
78 - EMAP.E VP COMMUNITY STRUCTURE
No
355
CODE
GENUS SPP.
PHYLYUM
CLAss
FAMILY
CERAPINN
‘ “°
Mollusca
Bivalvia
Cardudae
356
357
LAEVMORT
aevicardiwn mortoni
Mollusca
Bivalvia
Cardiidae
CYCLBORE
.‘yclxardia borealis
Mollusca
Bivalvia
Cardiudae
358
CORBFLUM
orbiculaflummea
Mollusca
Bivalvia
Corbiculidae
359
CORBCONT
orbula contracta
Mollusca
Bivalvia
Corbulidac
360
DONAVARI
)onac variabilis
Mollusca
Bivalvia
Donacidae
361
ALIGELEV
Lilgena elevata
Mollusca
Bivalvia
Ke lliidae
362
PARVMULT
ar#ducina ,nultilineaia
Mollusca
Bivalvia
Lucinidae
LYONAREN
,vons,a arenosa
Mollusca
Bivalvia
Lyonsiidae
LYONIIYAL
.vons :a hvahna
Mollusca
Bivalvia
Lyonsiidae
365
MULILATE
ulinia lateral is
Mollusca
Bivalvia
Mactr idae
366
RANGCUNE
angia cuneata
Mollusca
Bivalvia
Mactridae
367
SPISSOLI
pisula solidissuna
Mollusca
Bivalvia
Mactridac
368
MYAAREN
fi a arenaria
Mollusca
Bivalvia
Myidae
369
YOLDLIMA
‘old:a limatula
Mollusca
Bivalvia
Nucu lanidae
370
NUCUANNU
luculaannulata
Mollusca
Bivalvia
Nucu lidae
371
NUCUDELP
lucuLa deiphinodonta
Mollusca
Bivaivia
Nuculidae
372
CRASVIRG
>assos:rea virginica
Mollusca
Bivalvia
Ostreidae
373
PANDGOUL
‘andora gouldiana
Mollusca
Bivalvia
Pandondae
374
AEQUIRRA
Lrgopecten irradians
Mollusca
B,valvia
Pecunidac
375
PERIMARG
‘eriploma margaritacea
Mollusca
Bivalvia
Perip lomaudae
376
PETRPIIOL
‘e:ricolapholad(form:s
Mollusca
Bivalvia
Petricotidae
377
TAGEDIVI
agelus d:visus
Mollusca
Bivalvia
Solecurtidae
378
TAGEPLEB
agelus plebetus
Mollusca
Bivalvia
So lecurndae
379
SILICOST
iigua costata
Mollusca
Bivalvia
Solemyidae
380
SOLEVELU
olemya velum
Mollusca
Bivalvia
Solemyidae
381
ENSIDIRE
ns:s directus
Mollusca
Bivalvia
Solenidae
38
MUSCTRAN
fusculiu,n :ransversum
Mollusca
Bivalvia
Sphaeriidae
383
MACOBALT
facoma baithica
Mollusca
Bivalvia
Te llinidae
384
MACOMITC
facoma mitchell:
Mollusca
Bivalvia
Tellinidac
385
MACOTENT
facoma tenta
Mollusca
Bivalvia
Tellinidae
386
TELLAGIL
ellina agilts
Mollusca
Bivalvia
Tellinidac
387
ASTHHEMP
s:henothaerus hemphill :
Mollusca
Bivalvia
Thraciidae
388
BUSHELEG
ushia elegans
Mollusca
Bivalvia
Thraciidae
389
BIVASPEA
ivalvia sp A Mouatford
Mollusca
Bivalvia
Unidentified
390
ELLICOMP
Thptw complania
Mollusca
Bivalvia
Unionidac
391
CEMMGEMM
emma gemma
Mollusca
Bivalvia
Veneridae
392
MERCMERC
fercenaria mercenar:a
Mollusca
Bivalvia
Venendae
393
PITAMORR
ivar morrhuanus
Mollusca
Bivalvia
Veneridac
394
ACTEPUNC
:ctaxis punctostriatus
Mollusca
Gastropoda
Acteonidae
395
LAEVFUSC
aevapexfuscus
Mollusca
Gastropoda
Ancyhdae
396
BITHTENT
izhyn :a zentaculata
Mollusca
Gasuopoda
Bithynudae
397
CAECJOHN
aecumjohnson :
Mollusca
Gastropoda
Caecidae
398
CAECREGU
aecum regulare
Mollusca
Castropoda
Caecidae
399
CAECSPEA
aecum sp A Mountford
Mollusca
Gastropoda
Caccidac
400
CAECSPEB
aecum sp B Moun ford
Mollusca
Gastropoda
Caecidae
401
402
CALYPSPA
BITTALTE
alyp:rwdae sip. A
uz :um algernatum
Mollusca
Mollusca
Gasiropoda
Gastronoda
Calyptraetdae
Ccrithiidae
-------
GALLAGHER & G ssLE
79
No
CODE
GENUS SPP.
PIIYLYUM
CLAss
FAMILY
403
SEILADAM
eila adamsi
Mollusca
Gasiropoda
Cerithiopsidae
404
ANACLAFR
nachis lafresnayi
Mollusca
Gasirupoda
Columbcllidae
405
ANACOBES
nachLr obesa
Mollusca
Gastropoda
Columbellidac
406
ASTYLUNA
srnslwiata
Mollusca
Gasiropoda
Columbellidac
407
DORIOBSC
)oridella obscura
Moihisca
Gastropoda
Corambidae
408
CYLIBIDE
. ‘vIichneila bidentoia
Mol lusca
Gastropoda
Cy lichnidae
409
EPITGREE
uuonium greenlandicur
Mollusca
Gastropoda
Epitonaidae
410
EPITHUMP
rj,iton:um hwnphreysi
Mollusca
Gasiropoda
Epiconiidae
411
EPITRUPI
rpuonium rupicola
Mollusca
Gastropoda
Epitoniidae
412
CRATPILA
>azenapdoia
Mollusca
Gastropoda
Face linidae
413
414
41
GASTSPEA
HAMISOLI
4oun ford A
ianiinoea soh:ar:a
Mollusca
Mollusca
Gastropoda
Gastropoda
Gastropoda
Haminocidac
AMNILIMO
Imnicola IimO5a
Mollusca
Gastropoda
Hydrobiidae
416
CINC WINK
incinnana winkleyi
Mollusca
Gastropoda
Hydrobiidae
417
HYDRTRUN
fydrobia truncola
Mollusca
Gastropoda
Hydrobiidae
418
LrrrIENU
4ttondinops zenulpes
Mollusca
Gasiropoda
Hydrobiidae
419
LACUVINC
acuna vrncza
Mollusca
Gastropoda
Lacunidae
420
EIJPLCAUD
upleura caudaza
Molluscs
Gastropoda
Muncidac
421
UROSCINE
rosa1pinx cinerea
Mollusca
Gastropoda
422
ILYAOBSO
‘vanassa obsoleza
Mollusca
Gastropoda
Nassarudae
423
NASSTRIV
assarius trivinolus
Mollusca
Gastropoda
424
NASSVIBE
assaruu vibex
Mollusca
Gastropoda
Nassanidae
425
NATIPUSI
alica pusilla
Mollusca
Gastropoda
Naucidae
Naucidae
426
POLIHERO
o1inices heros
Mollusca
Gastropoda
P leurocendae
427
GON1VIRG
;oniobarn virginica
Mollusca
Gastropoda
428
BOONBISU
Ioonea bisuturalis
Mollusca
Gastropoda
Pyramidellidae
Pyramidel lidae
429
BOONIMPR
oonea impressa
Mollusca
Gasiropoda
430
BOONSEMI
oonea seminuda
Mollusca
Gastropoda
Pyrain ide l lidae
431
ODOSSULC
. Odostomsa sulcosa
Mollusca
Gastropoda
Pyramide l lzdae
432
FARGBART
argoa barischi
Mollusca
Gastropoda
Pyramidcllidae
433
FARGBUSH
argoa bushiana
Mollusca
Gastropoda
Pyraznidel lidae
434
FARGGIBB
argoa gibbosa
Mollusca
Gastropoda
Pyrairnde llidae
435
ODOSENGO
doszomia engonia
Mollusca
Gasiropoda
- Pyramidellidac
436
437
438
000SSPEA
SAYECHES
A
àvella chesapeakea
Mollusca
Mollusca
Gastropoda
Gastropoda
- Pyramide l lidae
Pyranudelhdae
TURBINTE
urbonilia inzerrup:a
Mollusca
Gasuopoda
Pyramide l lidae
439
TURBSPEB
B
Mollusca
Gastropoda
Pyramide l lidae
440
TURB?AEQ_
‘urbonilla ‘aegualis
Mollusca
Gastropoda
Pyramiddilidae
44
ACFECANA
kteocina canaliculaza
Mollusca
Gastropoda
Scaphandridae
442
ACTEORYZ
Lczeocsna ort’Za
Mollusca
Gastropoda
Scaphandridac
Turridae
443
KURTATRO
urtz,el1a azrostvla
Mollusca
Gastropoda
Tumdae
444
TURRSPEA
urridae sp A Mounifon
Mollusca
Gastropoda
445
VALVSINC
‘alvata sincera
Mollusca
Gastropoda
Valvattdae
446 VALVTRIC
‘ -“ •‘i tricariflala
Mollusca
Gastroooda
9 str,nella floridana
Mollusca
Gastropoda
Vitnnellidae
Valvatidae
447 VITRFLOR
-------
80 - EMAP-E VP COMMUNITY STRUCTURE
No I CODE I GENUS SPP. 1 FHYLYUM I CLASS I FAMILY
— Higher-Level Valid Taxa
448
OLIGOCHA
)ligochaeta
Annelida
Oligoch ta
Unidentified
449
ACRO FAM
crocirridae
Annelida
Po lychaeta
Acrocirridac
450
ARABIRMIJ
rabeila
rgcolor-multidentaia
omplex
Annelida
Po lychaeta
Arabe l lidae
451
CAPITELL
apitelia spp.
Annelida
Polychacta
Capire lhdae
452
hJ’HELOCH
kphelochaeta spp.
Annelida
Po lychaeta
Cirratulidac
453
DODECACE
)odecaceria s
Annelida
Polychacta
Cirratu lidae
454
MONTBPDS
4onticeliina
Armelida
Po lychaeca
Cirratulidac
apUs:eae-dorsobrwich
455
OPHRYOTR
)phr,otrocha spp.
Annelida
Po lychaeta
Dorvilleidac
45
MAGELONA
1aRelona spp
Annelida
Po lychaeta
Mage lonidae
457
POLYGORD
‘olvRordius spp
Annelida
Po lychaeta
Po lygordiidae
458
PROTODRI
rotodrdus spp
Annelida
Polychacta
Protodrilidae
459
SPHAEROD
phaerodoropsis spp
Annelida
Polychacta
Sphaerodondae
460
LAONICE
aonice .cpp
Annelida
Polychaeta
Spionidae
461
BRANCLSW
rania
Annelida
Polychacta
Syllidac
lavata-swedmarki
omplex
462
POAL 1
yposyll:s a lternala-sp ‘
omplex
Annelida
Polychaeta
Syllidae
463
AMPEABVA
orum complex
Azihropoda
Ainphipoda
Ampeliscidac
464
G1TANOPS
;iW, opsIs . cpp
Arthropoda
Amphipoda
Amphi lochidae
465
THALASSI
7ialassinzdea
Arthropoda
Crustacea
Thalassindea
466
TRICORYT
r,corv:hodes .cpp
Arthropoda
Ephcmaoptera
Tncoeythidae
467
OPHIUROI
)phwrosdea
Echinodermata
Ophiuroidea
Unidentified
468
RAETACF
f Roeta .cpp.
Mollusca
Bivalvia
Mactridac
469
PISIDIUM
‘isidium .cpp
Mollusca
Bivalvia
Pisidiidae
470
ANODONTA
nodonra spp.
Mollusca
Bivalvia
Unionidac
471
FERRISSI
rernssia spp
Mollusca
Gastropoda
Acroloxidae
472
COLUMBEL
olumbella .cpp
Mollusca
Gastropoda
Columbellidac
473
MELANELL
lelanella .cpp
Mollusca
Gastropoda
Ehmidae
474
LYOGYRUS
vogrus .cpp
Mollusca
Gastropoda
Hydrobiidae
475
BUSYCON
usvcon .cpp.
Mollusca
Gastropoda
Me longenidae
476
PHYSELLA
1 hysella spp.
Mollusca
Gastropoda
Physidae
477
PROMENET
‘romenetus .cpp
Mollusca
Gastropoda
Planorbidae
478
PLEUROCE
‘Ieurocera .cpp
Mollusca
Gastrupoda
Plcuroceridae
479
NEMERTIN
lemertinea
Nemertinea
Nemertinea
Unidentified
480
PHORONIS
‘horosiis .cpp.
Phoronida
Phoronida
Phoronidae
481
SIPUNCUL
:puncula
Supuncula
Sipuncula
Unidentified
-------
GALLAGHER & G sSLE - 81
Nol CODE I G us SPP. I PHYLYUM I CLASS I FAMILY
— FRESHWATER TAXA
482
ENCHYTRA
nchy:raeldae
Annelida
Oligochacta
Enchytraeidae
483
LUMBRICU
wnbriculidae
Annelida
O ligocliacta
LumbricuLidac
484
ARCTLOMO
rcreonws lomondi
Annelida
Oligochaeta
Naididae
485
BRATUNID
ratislawaunideniada
Annelida
Oligochaeta
Naididac
486
CHAETOGA
haetogaster spp.
Annelida
Oligochacta
Naididac
487
DERODIGI
ero diguata
Annehda
Oligochacta
Naididae
488
NAISPARD
Pals pardalis
Annelida
Ohgochacta
Naididae
489
NAISPSEU
Pals pseudob:usa
Annelida
Oligochaeta
Naididae
490
PIGUPIfiCH
iRuendla michiganens:
Annelida
Ohgochaeta
Naididae
491
SLAVAPPE
iavu*a appendiculoJa
Annelida
Oligochaeia
Naididae
492
SPECJOSI
pecariajosinae
Annelida
Oligochaeta
Naididae
493
siarr AND
ephensonzana gandvi
Annelida
Oligochacta
Naididae
494
STEPTRIV
Annelida
- Oligochaeta
Naididae
495
496
STYLLACU
v1ana lacustris
Annelida
Oligochacta
Naididae
AULOLIMN
ulodrilus limnobius
Annelida
O ligochaeta
Tubificidae
497
AULOPAUC
ulodrilus paucichaeta
Annelida
Oligochaeta
Tubificidae
498
AULOPIGU
ulodrilus pigueti
Annelida
Oligochaeta
Tubificidae
499
AULOPLUR
ulodrilus pluriseta
Annelida
Oligochaeta
Tubificidae
500
BRANSOWE
ranchiura sowerbvi
Annelida
Oligochaeta
Tubifjcidae
501
HABESPEC
faber c i. spec,osus
Annelida
Oligochaeta
Tubificidae
502
ILYOTEMP
Ivodrilus templetoni
Annelida
Oligochaeta
Tubificidae
503
ISOCFREY
sochaeridesfreyi
Annelida
O ligochaeta
Tubificidac
LIMNCERV
unnodrilus cerv&x
Annelida
Oligochaeta
Tubificidae
505
506
507
LIMNCLAP
LIMNIIOFF
1° 2
.imnodrilus hoffineisteri
Annelida
Annelida
Oligochaeta
Oligochaeta
Tubificidae
Tubificidae
LIMNUDEK
.unnodr:lus udekenuanu
Annelida
Oligochaeta
Tubificidae
508
QUISMULT
)ugs:adrilus multisetosus
Annelida
Oligochacta
Tubificidac
509
TELMVEJD
e1matodrslus vejdovskvi
Annelida
Oligochaeta
Tubificidae
510
TUBIPIWI
c ae
Annelida
Oligochaeca
Tubificidae
511
TLJBIFIWO
ub
Annehda
Oligochacta
Tubificidac
512
TUBIBROW
ubificoides brownae
Annelida
O ligochaeta
Tubificidae
513
TUBUIETE
Annelida
Oligochaeta
Tubificidac
514
515
AXARUS
xarus spp.
Arthropoda
Chironomidae
Chironomini
CHIRONOM
7uronomus spp
Arthropoda
Chironomidae
Chironomini
516
CLADOPLE
1adop1ema spp
Arthropoda
Chironomidac
Chironomini
517
CRYPFULV
ryptochironomus fulvus
Arthrc oda
Chironomidae
Chironomini
518
CRYPTOTE
‘rvp:otendapes spp
Arthropoda
Chironomidae
Chironomini
519
520
521
DEMICRYP
DICRNERV
) mIcrW,toChIi,omus
)icrotendipes ner.’osus
Arthropoda
Arthropoda
Chironomidac
Chirononudac
Chironomini
Chironomini
DICROTEN
)icrotend:pes spp
Arthropoda
Chironomidae
Chironomini
522
ENDOCHIR
r.ndochsronomus spp
Axthropoda
Chirononudae
Chironomirn
523
GLYPTOTE
;Ivp:otendspes spp
Arthropoda
Chironomidac
Chironomini
HARNISCH
Iarntschia sr’n
Arthroooda
Chironomidac
Chirononuni
-------
82 EMAP-EVPCOMMUNTrYSTRUCFURE
No
525
CODE
GENUS SPP.
PHYLYUM
CLiss
FAMILY
MICROCHI
icrochironomus spp.
Azthropoda
Chironomidae
Chironomini
526
PAEA LAD
aracLadopdma wp.
Azthropoda
Qnronomidae
hironomini
arai werbomiella .
Arthropoda
Chironomidac
Chironoinini
527
PABAIAUT
‘olypedilwn t? podura
Arth ,oda
Chironomidae
Chironomini
528
POLYTRIP
‘:ev4ochironomus s .
Aiihropoda
Chironomidae
hironomini
529
PSEUDO H
::c:ochirononws spp.
Azthropoda
Chironomidae
Chironomini
530
STICTO H
531
NANOCLAD
lwwcladius sn,.
Axthropoda
Chironomidae
Oithocladiinae
532
COELOTAN
oe1o:an us spp.
Arthropoda
Chironomidae
Tanypodinac
533
534
535
PROCHOLO
TANYPUS
rocladuis (Hoiotanypus
‘anypus spp.
Aithropoda
Aithropoda
Cbimnoinidae
Chironomidae
Tanypodinae
Tanypodinac
CLADOTAN
dotanytarsu5 spp.
Arthropoda
Chironomidae
Tanytarsini
536
RHEOTANY
Uzeotany:arsus .cpp.
Azthropoda
Chironomidac
Tanytarsini
537
TANYTARS
anytarsus . cpp.
Anhropoda
Cluronomidae
Tanytarsini
538
DUBIRAPH
)ubiraphia .cpp.
Arthropoda
Coleopiera
Elmidae
539
STENELMI
tenelniis spp.
Arthropoda
Coleoptera
Elmidae
540
BF 72iA
ezzia .cpp.
Arthropoda
Diptera
Ceratopogonidae
541
PALPOMYI
‘alpomvsa .cpp.
Arthropoda
Diptera
Ceratopogonidae
542
PROBEZZI
robezva .cpp.
Arthropoda
Diptera
Ceratopogonidae
543
SPHAEROM
,haeromias .cpp.
Arthropoda
Diptera
Ceratopogonidac
544
CHAOPUNC
haoborus pwzctipenn:s
Arthropoda
Diptera
Chaobondae
545
DOLICHOP
olichopodidae
Arthropoda
Diptera
Dolichopodidae
546
BRACHYCE
rachvcercus .cpp.
Anhropoda
Ephemeroptera
Caenidae
547
CAENIS
aenis .cpp.
Arthropoda
Ephemeroptera
Caenidae
548
HEXALIMB
exagenla IL,nboia
Aiihropoda
Ephemeroptera
Epheinendac
549
HEXAGENI
fexaRenia .cpp.
Arthropoda
Ephemeroptera
Epheinendae
550
HYDROPTI
ydroptila .cpp.
Arth 1 u oda
Trichoptera
Hydropu lidae
551
OECETIS
)eceus .cpp.
Arthrgpoda
Tnchoptera
Leptocendae
EMAP TAXA POOLED WIT
H OTHER TAXA
552
ASYCHIS
553
AMPEABDI
mpehsca abdita
Arthropoda
Amphipoda
Ampeliscidac
554
AMPEVADO
ntpelisca vadorum
Anhropoda
Amphipoda
Ampeliscidae
555
ORCHOMEN
rchomenella .cpp.
Atthropoda
Amphipoda
Lysianassidae
556
YOLDIA
‘oldia .cpp.
Mollusca
Bivalvia
Nuculanidae
557
PANDORA
pandora .cpp.
Mollusca
Bivalvia
Pandor idae
558
PANDORID
andoridae
Mollusca
Bivalvia
Pandondac
559
SOLEMYA
olemva .cpp
Mollusca
Bivalvia
Solemyidac
560
SOLEMYID
olemyidae
Mollusca
Bivalvia
Solemyidac
561
MUSCULIU
dusculium .cpp.
Mollusca
Bivalvia
Sphaerudae
562
TELLINA
ellina .cpp.
Mollusca
Bivalvia
Teilinidae
563
EUDORELL
udorella .cpp.
Arthropoda
Cumacea
Leuconidae
564
MICRATRA
4crophiopholis atra
Echinodermata
Ophiuroidea
Amphiundae
565
MELINNA
4elinna spp.
Annelida
Polychacta
Ampharetidae
566
APHESPEA
phelochaera sp. A Blake
Annelida
Polychacla
Cirratulidae
567
MONTBAPT
Ionncelluia baptisteae
Annehda
Polychaeta
Cirratulidae
568
MONTDORS
Ionticelluia •
orsobranch,alzs
Annclida
Po lychae ta
Cirratulidae
569
PHERUSA
iierusa .cpp.
Annelida
Polychaeta
F labelhgendae
570
LUMBHEBE
coletoma hebes
Annelida
Polychneta
Lumbrineridae
-------
GALLAGHER&GRASSLE ________ _____________ 83
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
571
ASYCELON
ábaco elongwus
Annelida
Polychaeta
Ma ldanidae
572
LEITOSCO
eitoscoloplos spp.
Annelida
Polychacta
Orbirnidac
573
OWENIA
iwenia spp.
Annelida
Polychacta
Oweniidae
574
PECTINAR
ec:inana spp.
Annelida
Po lychaeta
Pecdnanidae
575
BRANCLAV
rania clavata
Anneida
Po lychaeta
Syllidae
576
BRANS WED
Iransa swedmarki
Annelida
Polychaeta
Syllidae
577
TYPOALTE
‘posv1lis alternata
Annelida
Polychacca
Syllidae
578
TYPOSPE1
\ipasvilis sp. I NMFS
Annelida
Polychacta
Sy llidae
579
AMPH1TRI
mphitrit,nae
Annelida
Po lychaeca
Terebe l lidae
Dropped EMAP-E Taxa (See Text)
580
NEVEDUPL
757
ERPODFAM
rpodellidae
Annelida
Hirudinea
Erpode lhdae
758
HIRUDINE
lirudinea
Annelida
Hirudinea
Unidentified -
779
DERO
)ero spp.
Annelida
Oligochaeta
Naididae
780
NAIDIDAE
laididac
Annelida
Oligochaeca
Naididae
781
STEPHENS
tephensoniana spp.
Annelida
Oligochaela
Naididac
782
TUBIFICO
ubiftco,des spp.
Annelida
Oligochacta
Tubificidac
784
AMPHARTD
inpharetidae
Annelida
Polychaeta
Amphare cidae
785
ARABELLA
rabella spp
Annelida
Polychacta
Arabel lidae
786
ARABELLI
rabellidae
Annelida
Polychaeta
Arabellidac
787
CAPITELD
apitellidae
Annelida
Polychacta
Capite llidae
788
NOTOMAST
folotnastus spp
Annelida
Polychacta
Capitellidac
789
CIRRATUL
:irratu lidae
Annelida
Polychaeca
Cirraculidac
790
EUNICIDA
unicidae
Annehda
Polychacta
Eunicidac
791
FLABELLI
labelligendae
Annelida
Po lychaeta
Flabel ligendae
792
GLYCERA
lvcera spp
Annelida
Po lychaeta
G lyceridae
793
GLYCERID
lvcendae
Annelida
Polychaeta
Glyceridae
794
GONIADID
oniadidae
Annelida
Po lychaeta
Goniadidae
795
GYPTIS
vpt is spp
Annelida
Polychaeta
Hesionidae
796
HESIONID
esionidae
Annelida
Po lychaeta
Hesionidae
797
MICROPHT
dicroph:halinus spp
Annelida
Polychaeca
Hesionidac
798
LUMBRIND
umbnnendae
Annelida
Polychae ca
Lumbnneridae
799
LUMBRINE
co1ero,na spp
Annelida
Polychaeta
Lumbrineridae
800
MALDANID
1aldanidae
Annelida
Polychaeta
Ma ldanidae
801
NEPHTYID
ephtyidae
Annelida
Po lychaeta
- Nephtyidae
802
NEPHTYS
Jephrvs spp
Annelida
Polychaeta
Nephtyidae
803
NEREIDAE
Iereididae
Arnielida
Polychacta
Nereididac
804
ONUPHIDA
nuphidae
Annelida
Po lychaeta
Onuphidac
805
OPHELIID
pheliidae
Annelida
Po lychaeta
Ophe liidae
806
IRA VISIA
ravisia spp
Anneluda
Po lychae ca
Ophe liidae
807
ORBINIA
)rbinia spp.
Annelida
Polychaeta
Orbiniidae
808
ORBINIU)
)rbiniidae
Annelida
Po lychaeta
Orbinuidae
809
SCOLOPLO
coloplos spp
Annelida
Po lychaeta
Orbinndae
810
OWENIIDA
)wcniidae
Annelida
Po lychaeta
Oweniidac
811
ARICIDEA
rsc:dea spp
Annelida
Polychaeta
Paraonidae
812
PARAONID
araonidac
Annelida
Polychacta
Paraonidae
813
ETEONE
Ivpereteone spp
Annelida
Po lychaeta
Phyllodocidae
814
PHYLLODO
hvllodoce spp
Annelida
Polychacta
Phy l lodocidae
PHYLDCDE
hyllodocidac
Annelida
Po lychaeta
Phyl lodocidae
-------
EMAP-E VP COMMUNITY STRUCTURE
No
CODE
GENus Sn ’.
PHYLYUM
CLASS
FAMILY
816
PJLARGID
ilargidac
Annelida
Po lychaeta
Pi largidae
817
SIGAMBRA
iRa,nbra spp.
Annelida
Polychacta
Pi largidae
818
HARMOTHO
!armorhoe spp.
Anneida
Polychaeta
Polynoidac
819
LEPIDONO
epidonotus spp.
Anneida
Po lychaeia
Polynoidac
820
POLYNOID
olynoidae
Annelida
Polychacta
Polynoidac
821
SABELLAR
abellariidae
Annelida
Po lychaeta
Sabel lanidae
822
EUCHONE
uchone spp.
Annelida
Polychacta
Sabellidae
823
FABRICIN
ábricinac
Annehda
Polychacta
Sabel lidae
824
SABELLID
abellidae
calibregmatidae
ilograninae sp. A Moms
Annelida
Polychae ta
Sabel lidae
Scalibregmatidae
Se ipu lidae
Annelida
Polychacta
825
SCALIBRE
Annelida
Polychacta
826
FILOGRAN
827
HYDRDIAN
Fydroides dianthus
Annelida
Polychaeta
Serpuhdae
828
HYDRPROT
Tydroides protulicola
Annelida
Polychaeta
Serpuhdae
829
HYDROIDE
rydroides spp.
Annelida
Pojychacta
Serpulidac
830
SERPULID
erpulidac
Annelida
Polychaeta
Serpuhdae
831
SIGALION
igahonidae
Annelida
Po lychaeta
Sigalionidac
832
STENELAI
thenelais spp.
Annelida
Po lychaeta
Sigalionidac
833
POLYDORA
olydora spp.
Annelida
Polychacta
Spionidae
834
PRIONOSP
rsonospio spp.
Annelida
Po lychaeta
Spionidae
835
SCOLELEP
colelepis spp.
Annelida
Po lychaeta
Spionidac
836
SF10
gno spp.
Annelida
Polychacta
Spionidae
837
SPIONIDA
pionidac
Annelida
Polychacta
Spionidae
838
SPIRORBD
pirorbidae
Annelida
Polychaeta
Spirorbidae
839
SPIRORBI
,irorbj.c spp.
Annelida
Po lychaeta
Spirorbidae
840
AUTOLYNI
wtolyninae
Annelida
Polychaeta
Syl lidae
841
AUTOLYTU
wolyws spp.
Annelida
Po lychaeta
Syllidac
842
BRANIA
franw spp.
Annehda
Polychacta
Syllidac
843
EXOGONE
.xogone spp.
Annchda
Polychacta
Syllidac
844
PIONOSYL
‘ionosyllis spp.
Annelida
Polychacta
Sylhdae
845
SPIIAEROS
phaerosyll:s spp.
Annelida
Po lychaeta
Syllidae
846
STREPTOS
:rep:osyllis spp.
Annelida
Polychaeta
Syllidac
847
SYLLIDAE
yllidae
Annehda
Polychaeta
Syllidac
848
SYLLIDES
yllides spp.
Annelida
Po ychaeta
Syl lidae
849
TYPOSYLL
‘ypoiyWs spp.
Annelida
Polychacta
Syl lidae
850
PISTA
1 ista spp.
Annelida
Polychae ta
Terebellidae
851
POLYCIRN
olycimnae Unidentified
Annelida
Polychaeta
Terebellidae
852
POLYCIRR
o1yc:rrus spp.
Annelida
Po lychaeta
Terebe l lidae
853
TEREBELL
erebellidae
Annehda
Polychaeta
Terebellidae
854
POLYCSUB
- DCCP
Annelida
Polychaeta
Unidentified
855
POLYCCAR
olyChaeta: Other
Annelida
Polychaeta
Unidentified
856
POLYCSUR
-
Annelida
Polychacta
Unidentified
857
POLYCHAE
Annelida
Polychacta
Unidentified
581
AMPELISC
mpelisca spp
Arthropoda
Amphipoda
Ampeliscidac
582
AMPITHOE
mpuhoe spp.
Arthrcpoda
Amphipoda
Ampithoidac
583
AMPITHOI
mpithoidac
Arihropoda
Amphipoda
Ampithoidac
AORIDAE
ondae
Arthroi ,oda
Amohinoda
Aoridae
-------
f A! t af u Z 85
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
585
LEMBOS
.embos spp.
Aithro oda
Amphipoda
Aoridae
586
i p’rocnE
£ptocheIrus spp.
Arthrci ,oda
Amphipoda
Aoridae
587
MICRODEU
4icrodeutopus spp.
Aithropoda
Amphipoda
Aoridae
588
UNCIOLA
Jnciola spp.
Azthrupoda
Amphipoda
Aondac
589
AEGILONG
eginina iongicornLc
Ai11 oi,oda
Amphipoda
Caprellidae
590
CAPRANDR
aprelia andreae
Axthropoda
Amphipoda
Caprellidae
591
CAPRPENA
aprelia penantLc
Arthropoda
Amphipoda
Caprel lidae
592
CAPRELLA
aprella spp.
Azthropoda
Amphipoda
Capre llidae
593
CAPRELLI
aprellidae
Anhroçoda
Amphipoda
Caprellidae
594
LUCOINCE
uconacsa uicerta
Aithropoda
Ai phipoda
Caprelhdae
595
PARATENIJ
‘aracaprelia zenws
Anhropoda
Amphipoda
Caprdllidae
596
COROPHIU
orophiwn spp.
Atthropoda
Amphipoda
Corophiidae
597
GAMMARID
ammandae
Aithropoda
Amphipoda
Gammaridae
598
GAMMARUS
ammarus spp.
Aithwpoda
Amphipoda
Gammandae
599
ACANTHOR
ca,uhohaustoriu.c spp.
Ai*opoda
Amphipoda
Haustoriidae
600
RAUSTIDA
austoriidae
Azthropoda
Amphipoda
Haustoriidae
601
PARAHAUS
arahaustonus spp.
Artluopoda
Amphipoda
Haustoriidae
602
PROTOHAU
mtohaustonus spp.
Anhropoda
Amphipoda
Haustonidae
603
PHOTIS
hotis spp.
Anhropoda
Amphipoda
Isaeidae
604
ERICTHON
ricthornus spp.
Azthropoda
Amphipoda
lschymcendae
605
LILJEBOR
iljeborgiidae
Arthropoda
Amphipoda
Lilieborgiidae
606
LISTRIEL
.istnella spp.
Artbropoda
Arnphipoda
Liljeborgiidae
607
LYSIADAE
.ysianassidae
Arthropoda
Amphipoda
Lysianassidae
608
MELITIDA
4eliudae
Arthropoda
Amphipoda
Me litidae
609
MONOCULO
4onoculodes spp.
Arthropoda
Amphipoda
Oedicen tidae
610
PHOXOCEP
‘hoxoc ha1idae
Arthropoda
Amphipoda
Phoxocephalidae
611
RHEPOXYN
Uicpoxynius spp.
Arthropoda
Amphipoda
Phoxocephalidae
612
PODOCERI
odoccndae
Arthropoda
Amphipoda
Podoceridac
613
STENOTHO
tenothoc spp.
Arthropoda
Amphipoda
Stenothoidae
614
AMPHIPOD
mphipoda: Other
Anhropoda
Amphipoda
Unidentified
625
HOMAAMER
Fomarus americanus
Arthropoda
Astacidea
Nephropsidae
674
CHRNMDAE
iironomidae
Arthropoda
Chimnomidae
chironomidac
675
CHIRONIM
hironomini
Arthropoda
Chironomidae
Chirononuni
676
CRYFTOCH
ryptochironomus spp.
Aiihropoda
Chironomidae
Chironomini
677
POLYPEDI
blypedilum spp.
Arthropoda
Chironomidae
Chironomini
PROCLADI
ocIadius spp.
Arthropoda
Chimnomidae
Tanypodinae
679
TANYTTRB
anytarsini
Arthmpoda
Ch,ronomidae
Tanytarsini
680
BALABALA
Ialanus balanoides
Arthiopoda
Cimpedia
Balanidac
681
BALACREN
lalanus crenatus
Aiihropoda
Cirripedia
Balanidae
682
BALAIMPR
lalanus improvisus
Arthropoda
Cimpedia
Ba lanidae
683
BALANUS
lalanus spp.
Anhropoda
Cirnpedia
Balanidae
684
BALAVENU
lalanus venustus
Arthropoda
Cirnpedia
Ba lanidae
685
CLADOCER
ladocera
Arthropoda
Cladocera
Unidentified
686
COLLEMBO
ollembola
Arthropoda
Collembola
Unidentified
687
CALANOID
:alanoida
Arthropoda
Copepoda
Calanoida
688
CALIGOID
ahgoida
Arthropoda
Copepoda
Caligoida
689
HARPACTI
arpacticoida
Anhropoda
Copepoda
Harpacticoida
690
BODOTRUN
lodotnidac
Arthropoda
Cumacea
Bodotnidac
691
CUMACEA
umacea
Arthrovoda
Cumacea
Unidentified
-------
86 EMAP-E VP COMMUNITY STRUCTURE
No
692
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
BRACHYUR
rachyura
Arthropoda
Decapoda
Brachyura
693
CANcIRRO
ancer irioratus
Arthropoda
Decapoda
Canmdae
2!
CANCER
ncor sp
A opoda
Decapoda
Cancndae
695
CARIDEA
andea
Azthropoda
Decapoda
Catidea
696
HIPPOLYT
lippolytidac
Arthropoda
Decapoda
Hippolytidac
697
LIBINIA
ibima SPP.
Arthmpoda
Decapoda
Majidae
698
MAJIDAE
laiidae
Arthrcpoda
Decapoda
Majidae
699
PAGURIDA
aguridae
Aithropoda
Decapoda
Pagugidae
700
PAGURUS
agurus spp.
Arthropoda
Decapoda
Paguridac
865
PALAPUGI
alaemonetes puglo
Arthropoda
Decapoda
Paiaemonidae
701
PENAEIDA
enacidac
Arthropoda
D poda
Penacidae
702
TRACCONS
Arthropoda
Decapoda
Penaeidae
703
704
DISSMELL
issodactylus mellitac
Arthropoda
Decapoda
Ptnnotheridae
PINNCHAE
innixa chaetopterana
Arthropoda
Decapoda
Pinnotheridae
705
PINNRETI
innixa retmens
Arthropoda
Decapoda
P,nnotheridae
706
PINNSAYA
innixa sayana
Arthropoda
Decapoda
Pinnotheridac
707
PINNIXA
innixa spp.
Arthropoda
Decapoda
Pinnotheridae
708
PINNOTHR
innotheres spp
Arthropoda
Decapoda
Pinnotheridae
709
PINNOTHE
innothendae
Azthropoda
Decapoda
Pinnotheiidae
710
CALLSAPI
:allinectes sapidus
Arthropoda
Decapoda
Poriunidae
711
CALLINEC
:auinectes spp.
Anhropoda
Decapoda
Portunidac
712
CARCMAEN
:arcinus macnas
Arthropoda
Decapoda
Portunidae
713
OVALIPES
)valipes spp.
Arthropoda
Decapoda
Portunidac
714
PORTUNID
ortunidae
Arthropoda
Decapoda
Porturndae
715
DECAPODA
)ecapoda
Arthropoda
Decapoda
Unidentified
716
XANTHIDA
anthidae
Arthropoda
Decapoda
Xanthidae
717
CERATFAM
:eratopo gpnidae
Arthropoda
Diptera
Ceratopogonidac
718
DIPTERA
)iptera
Arthropoda
Diptera
Unidentified
724
EPHEMFAM
phemendac
Arthropoda
Ephemeroptera
Ephemendae
760
HYDRACAR
Iydracanna
Arthropoda
Hydracanna
Unidentified
761
INSECTA
nsecta
Arthropoda
Insecta
Unidentified
762
ANTHURID
nihundae
Arthropoda
Isopoda
Anthundac
763
CYATHURA
:yathura spp.
Arthropoda
Isopoda
Anthundae
764
CHIRIDOT
:hindotea spp.
Arthropoda
Isopoda
Idoteidae
765
ERICHSON
nchsonella spp
Arthropoda
Isapoda
Idoteidac
766
IDOTEA
dotca sop.
Arthropoda
Isopoda
Idoteidae
767
IDOTEIDA
dotcidae
Arthropoda
Isopoda
Idotetdae
768
ISOPODA
sopoda: Other
Arthropoda
Isopoda
Unidentified
769
LIMUPOLY
.imulus polyphemus
Arthropoda
Merostomata
Limulidae
772
HETEFORM
leteromysis formosa
Anhropoda
Mysidacea
Mysidae
773
MYSIDAE
lysidac
Arthropoda
Mysidacca
Mysidac
774
MYSIALMY
lysidopsis almyra
Arthropoda
Mysidacea
Mysidae
775
MYSIBIGE
4ysidopsis bigelowi
Arthropoda
Mysidacea
Mysidac
776
MYSIDOPS
4ysidopsis spp.
Arthropoda
Mysidacea
Mysidae
777
NEOMAMER
Jeomysis americana
Arthropoda
Mysidacea
Mysidae
783
OSTRACOD
)stracoda
Arthropoda
Ostmcoda
Unidentified
861
PYCNOGON
ycnogonida
Arthropoda
Pycnogonida
Unidentified
862
TANAIDAC
anaidacea
Arthropoda
Tanaidacea
Unidentified
863
HYDROFAM
lydroptilidac
Arthroooda
Tnchootera
Hvdroytihdae
-------
(ALLACHER & GRASSLE 87
No
CODE
GENUS SPP.
PIIYLYUM
CLASS
FAMILY
665
ALCYONID
Icyomdium spp.
Biyozoa
Bryozoa
A lcyorndiidae
666
CALLCRAT
allopora craticula
Bryozoa
Bryozoa
Cal loporidae
667
TURBDI H
urbice11opora dichocom
Bryozoa
Sryozoa
Celleponnidac
668
MEMBTENU
dembranipora cenuis
Bryozoa
Bryozoa
Membraniporidae
669
ANGUPALM
thguineUa palmata
Bryozoa
Bryozoa
No lellidae
670
S HIUNIC
chizoporella unicornis
Bryozoa
Bryozoa
Schizoporellidne
671
AMATVIDO
imathia vidovici
Bryozoa
Bryozoa
Vesicu lanidae
617
BOSTPILU
Chordata
Ascidiacea
Molgulidae
618
619
MOLGAREN
4oIgula arenata
Chordata
Ascidiacea
Molgulidae
MOLGMANB
1oIgula manhattensis
Chordata
Ascidiacea
Mo lgu(idae
620
PEROVIRI
erophora vindis
Chordata
Ascidiacea
Perophoridae
62
AMARSTEL
.maroucium stellatum
Chordata
Ascidiacea
Polyclinidae
62
BOTRSCHL
3ot iIlus schiosseri
Chordata
Ascidiacea
Styelidae
623
CNEMMOLL
nemidocarpa mollis
Chordata
Ascidiacea
Stvelidae
624
ASCIDIAC
.scidiacea
Chordata
Ascidiacea
Iinidenufied
Branchiostomidne
Branchiostomidae
673
BRANCARI
Iranchiostoma caribaeun
Chordata
Cephalochordata
672
BRANVIRG
ranchiostoma canbaeun
Chordata
Cephalochordala
615
PARARAPI
aranthus rapiformis
Cnidaria
Anthozoa
Actinostolidae
616
ANTHOZOA
nthozoa
Cnidaria
Anthozoa
Unidentified
626
ASTERIAS
stenas spp.
Echinodermata
Astaroidea
Asteriidae
627
ASTKKOID
stemidea
Echinodermata
Asteroidea
Unidentified
719
ECHINODE
chinodermata
Echinodermata
Echinodermata
Unidentified
720
ARBAPUNC
i.rbacia punctulata
Echinodermata
Echinoidea
Arbaciidae
721
ECHIPARM
chinarachnius parma
Echinodermata
Echinoidea
Echinarachnidae
722
MELLQUIN
esperforaia
Echinoidea
Me l liiidae
723
ECHINOID
chinoidea
Echinodermata
Echinoidea
Unidentified
759
HOLOTHUR
1olothuroidea
Echinodermata
Holothuroidea
Unidentified
756
HEMICHOR
lemichordata
Hemichordata
Hemichordata
Unidentified
770
MISCELLA
liscellanea
Miscellanea
Miscellanea
Unidentified
771
NOORGPRS
Jo Organisms Present
Miscellanea
Miscellanea
Unidentified
628
ANOMSIMP
nomia simplex
Mollusca
Bivalvia
Anomiidac
629
ANOMIA
nomia spp
Mollusca
Bivalvia
Anomiidae
630
ANOMSQUA
nomia squamula
Mollusca
Bivalvia
Anomtidae
631
ARCIDFAM
icidae
Mollusca
Bivalvia
Arcidae
632
ASTARTE
siartc spp
Mollusca
Buvalvia
Astarudae
633
ASTARTID
starudae
Mollusca
Bivalvia
Astartidae
634
MYTILEUC
4ytilopsis leucophaeta
Mollusca
Bivalvia
Dreissenidae
635
GALEOMMA
aleommatacea
Mollusca
Bivalvia
Galeommatacea
63
LYONSIA
.yons la spp
Mollusca
Bivalvia
Lyonsiidae
637
MACTRFAM
lactndae
Mollusca
Bivalvia
Mactridae
638
MYSEPLAN
lysella planulata
Mollusca
Bivalvia
Montacuiidae
639
MYSELLA
lysefla spp.
Mollusca
Bivalvia
Montacutidae
6 Q
CRENDECU
:renella decussata
Mollusca
Bivalvia
Myuuidac
641
CRENGLAN
renella glandula
Mollusca
Bivalvia
Mytilidae
CRENELLA
:rcnella SPP.
Mollusca
Bivalvia
Mytilidac
643
GEUKDEMI
leukensia demissa
Mollusca
Bivalvia
Mytilidae
ISCHRECU
schadium recurvum
Mollusca
Bivalvia
Mytilidac
645
MODIOLUS
4odiolus soy
Mollusca
Bivalvia
Mvtilidae
-------
88 EMAP.E VP COMMUNITY STRUCTURE
No
646
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
MUSCNIGE
lusculus niger
Mollusca
Bivalvia
Mynuidac
647
MUSCULUS
lusculus spp.
Mollusca
Bivalvia
Myu lidac
648
MYTILIDA
lynuidae
Mollusca
Bivalvia
Mytilidac
649
MYTIEDUL
Aynlus edulis
Mollusca
Bivalvia
Myu lidae
650
NUCULANI
Iuculanidae
Mollusca
Bivalvia
Nuculanidac
651
NUCULA
lucula spp.
Mollusca
Bivalvia
Nuculidac
652
PECTINID
ectinidae
Mollusca
Bivalvia
Pecnnidae
653
BARNTRUN
larnea truncaca
Mollusca
Bivalvia
Pholadidae
654
PHOLADID
holadidae
Moliusca
Bivalvia
Pholadidae
655
SOLECFAM
olecurtidae
Mollusca
Bivalvia
Solecurndae
656
TAGELUS
aielus spa.
Mollusca
Bivalvia
So lecumdae
657
SOLENIDA
olenidae
Molluscs
Bivalvia
Solenidae
658
TELLINID
eIIimdae
Mollusca
Bivalvia
Tel linjdae
659
THRACIID
iiraciidae
Mollusca
Bivalvia
Thraciidae
660
THYASIRI
hyasiridae
Mollusca
Bivalvia
Thyasindae
661
BIVALDEP
bvalvia. Other - DC I)os lt
Mollusca
Bivalvia
Unidentified
662
663
BIVALSUS
i o? ers
Molluscs
Bivalvia
Unidentified
BIVALVIA
hvalvia: Other -
Jnidcntified
Molluscs
Bivalvia
Unidentified
664
UNIONIDA
Jnionidae
Mollusca
Bivalvia
Unionidae
725
BUCCINID
luccinidae
Mollusca
Gastropoda
Buccinidae
726
CAECIDAE
aecidae
Mollusca
Gastropoda
Caecidae
727
CAECUM
aecum spp.
Mollusca
Gastropoda
Caecidae
728
CREPCONV
: piduIa convexa
Mollusca
Gasiropoda
Calyptraeidae
729
CREPCOFO
repiduIa
onvexa-forn icwa
omplex
Mollusca
Gastropoda
Ca lyptraeidae
730
CREPFORN
repidula fornicata
Molluscs
Gasuopoda
Ca lypiraeidae
731
CREPMACU
repidula maculosa
Mollusca
Gastropoda
Ca lyptraeidae
732
CREPPLAN
repidula plana
Molluscs
Gastropoda
Calyptraeidae
733
CREPIDUL
:repidula spp.
Mollusca
Gasuopoda
Calyptraeidae
734
ANACHIS
nachis spp.
Mollusca
Gastropoda
Columbcl lidae
735
COLUMBLD
:olumbeliidae
Molluscs
Gastropoda
Co lumbellidae
736
CYLICHNE
:ylichndua SPP
Mollusca
Gasiropoda
Cy lichrndac
737
EPITONIU
pitonium spp.
Mollusca
Gastropoda
Epitoniidae
738
CRATENA
:ratena Spp.
Mollusca
Gastropoda
Face linidae
739
HYDROBIA
lydrobia spp.
Mollusca
Gasiropoda
Hydrobtidac
740
HYDROBU
Iydrobiidae
Molluscs
Gastropoda
Hydrobndae
741
LYMNAFAM
.ymnacidae
Mollusca
Gastropoda
Lymnaeidae
742
NASSARIU
(assarius spp.
Molluscs
Gastropoda
Nassarndae
743
NATICA
lazica spp.
Mollusca
Gastropoda
Naticidac
744
NATIcmA
laticidac
Mollusca
Gastropoda
Naticidae
745
NUDIBRAN
udibranchia
Mollusca
Gastropoda
Nudibranchia
746
PLANORBI
‘lanorbidac
Mollusca
Gastropoda
Planorbidae
747
FARGOA
argoa spp.
Mollusca
Gastropoda
Pyramidellidae
748
ODOSTOMI
)dostomia spp.
Mollusca
Gastropoda
Pyramidel lidae
749
PYRAMIDE
yramidellidae
Molluscs
Gastropoda
Pyramidellidac
750
TURBONIL
urbonilla son.
Motlusca
Gastrocoda
Pvramidellidae
-------
GALLAGHER & GRASSLE - 89
No
CODE
GENUS SPP.
PHYLYUM
CLASS
FAMILY
751
SCAPHFAM
caphandndae
Mollusca
Gastropoda
Scaphandiidae
752
TURRIFAM -
ruriidae
Mollusca
Gastropoda
Tumdae
753
GASTROPO
3asuopoda: Other
Mollusca
Gastropoda
Unidentified
754
V1TRINEL
iitnnellidac
Mollusca
Gasuopoda
Vitrinellidae
755
VIVIPARI
T iviparidae
Mollusca
Gasiropoda
Viviparidae
858
CHAEAPIC
:haetopleura apiculata
Mollusca
Polyplacophora
Chaetopleundae
859
POLYPLAC
olyplacophora
MoUusca
Polyplacophora
Urndenufied
778
NEMATODA
lematoda
Nematoda
Nematoda
Unidentified
864
TURBELLA
urbellana
Platyhelminthe
Turbellaria
Unidentified
860
PORIFERA
onfera
Porifera
Porifera
Unidentified
— Taxa in the EMA
868 NAISCOMM Fiats communss
866 AMPRABDI tmphiop1us abdita
867 HIATARCT ‘-fiateI1a arcr:ca
P-E Species List but not p
Annelida Oligochaeta
Echinodermata Ophiuroidea
Mollusca Bivalvia
resent
Naididac
Amphiundae
Hiatellidae
-------
90 EMAP-E VP COMMUNITY STRUCTURE
APPENDIX IV SAMPLE CLUSTER ANALYSIS
A cluster analysis of the 1918-sample EMAP-E VP benthic data at CNESS (CNESS, m=25, UPGMA
Sorting). All samples with fewer than 25 individuals were dropped (a requirement with a random sample
size of 25). The pared data set consisted of 1736 samples and 466 species. New Bedford Harbor STA
099, a degraded estuarine station (Schimmel et al. 1993, Table B-2) is bolded. This degraded EMAP-E
VP sampling site exhibits considerable variation among months and years.
CNESS Distance (NESSm = 25)
0.04 0.30 0.40 0.10 0.80 1.00 1.20
NAME
VA90—001 29AUG90 1 1
VA90-OOi 29AUG90 2 2
VA9O—001 29AUG90 3 3
VA9O-018 03AUG90 3 55
VA90—173 30JUL90 2 456
VA9O—173 30JUL90 3 457
VA9O-018 03AUG90 1 53
VA9O-018 03AUG90 2 54
VA91—173 06AUG91 1 661
VA91-173 06AUG91 2 662
VA91—173 06AUG91 3 663
VA90-158 10AUG90 1 420
VA90-158 10AUG90 2 421
VA9O—158 10AUG90 3 422
VA92-527 30JUL92 1 1284
VA92—527 30JUL92 2 1285
VA92-527 30JUL92 3 1286
VA92-512 09AUG92 1242
VA92-512 09AUG92 1243
VA92-512 09AUG92 1244
VA92-173 27JUL92 1069
VA92-173 24AUG92 2 1073
VA93-173 07AUG93 1 1432
VA93—173 07AUG93 3 1434
VA93—173 02SEP93 1 1435
VA93-173 02SEP93 3 1437
VA92-538 20AUG92 1 1303
VA92—538 20AUG92 3 1305
VA92-538 20AUG92 2 1304
VA92—173 24AUG92 1 1072
VA92-173 24AUG92 3 1074
VA93-173 07AUG93 2 1433
VA92-515 08AUG92 1 1251
VA92-5i5 08AUG92 3 1253
VA93-173 02SEP93 2 1436
VAS3-702 26AUG93 1 1656
VA93-102 26AUG93 2 1657
VA93-702 26AUG93 3 1658
VA92-173 27JUL92 2 1070
VA92-173 27JUL92 3 1071
VA93-669 22AUG93 1 1585
VA93-669 22AUG93 2 1586
VA93-669 22AUG93 3 1587
VA9O-002 29AUG90 1 4
VA90-002 29AUG90 2 5
VA90-002 29AUG90 3 6
VA90—099 04SEP90 3 267
VA9O-123 21JUL90 2 338
VA90-123 21JUL90 3 339
VA90—122 21JUL90 1 331
VA90—123 21JUL90 1 337
VA91—373 03AUG91 1 863
VA91-373 03AUG91 2 864
VA91—373 03AUG91 3 865
VA9O-122 21JUL90 2 332
VA90-174 20JUL90 1 458
VA9O-174 20JUL90 459
VA9O-174 20JUL90 460
VA9O-174 30JUL90 1 461
VA9O-1 7 4 30JUL90 3 463
VA90-174 30JUL90 2 462
VA91-369 04AUG91 1 854
VA91-369 04AUG91 2 855
VA9L—369 04AUG91 3 856
VA90—161 20JUL90 1 432
VA90—161 20JUL90 2 433
VA9O—028 31JUL90 1 89
VA90-094 30JUL90 1 253
VA90—094 30JUL90 3 255
VA90—094 30JUL90 2 254
VA92—528 29JUL92 1 1287
VA92-528 29JUL92 3 1289
Vk92-528 29JUL92 2 1288
VA90-086 01AUG90 1 226
VA90-150 19JUL90 1 394
VA90-150 19JUL90 2 395
VA9O—150 19JUL90 3 396
VA90—155 30JUL90 2 412
VA9O-155 30JUL90 3 413
VA90—155 30JUL90 1 411
VA9O—100 21SEP90 2 272
VA90—100 21SEP90 3 273
‘JA9O—122 21JUL90 3 333
XD
-------
( AI.IAruPR & GRASRLE
91
VA92—509 09AUG92 1
VA92-509 09AUG92 2
Vk92-509 09AUG92 3
VA93—150 13AUG93 1
VA93-150 13AUG93 3
VA93—150 13AUG93 2
VA93—711 19AUG93 1
VA9O-122 21SEP90 1
VA90-122 21SEP90 3
VA9O—122 21SEP90 2
¶1?31—404 24JUL91 1
VA91—404 24JUL91 2
VA91—404 24JUL91 3
VA92—045 02AUG92 1
VA92-045 02AUG92 2
VA92-045 02AUG92 3
Vk92-045 27AUG92 1
VA92—045 27AUG92 3
VA92-045 27AUG92 2
VA93—659 21AUG93 I
VA93—659 21AUG93 3
VA93—659 21AUG93 2
VA92—150 17AUG92 1
VA92-150 17AUG92 2
VA92-150 17AUG92 3
VA92-150 27AUG92 2
VA92-150 27AUG92 1
VA92-150 27AUG92 3
VA9I—310 04AUG91 1
VA91—370 04AUG91 2
VA91—371 03AUG91 3
VA91-371 03AUG91 1
VA91—371 03AUG91 2
VA91-370 04AUG91 3
VA90—099 04SEP90 1
vAIO—099 04SEP90 2
VA90-162 20JUL90 2
VAO0-13S 15AUG90 1
VA9O—135 13AUG90 2
VA90-135 15AUG90 3
VA92—489 30JUL92 2
VA93-685 10AUG93 1
VA9)—685 10AUG93 2
VA93—683 10AUG93 3
VA92—534 02AUG92 1
VA90-168 10AUG90 1
VA90—168 10AUG90 2
VA9O—163 10AUG90 3
VA92-489 30JUL92 1
VA92-489 30JUL92 3
VA93—682 08AUG93 1
VA93-684 07AUG93 1
VA93—684 07AUG93 3
VA91-405 04SEP91 1
VA9 I-405 04SEP91 3
VA91-403 04SEP91 2
VA91-136 01AUG91 1
VA9 I-136 01AUG91 2
VA93—136 30AUG93 2
VA9O—162 20JUL90 1
VA91-363 03AUG91 1
VA91—363 05AUG91 3
VA91-363 05AUG91 2
VA90-008 31AUG90 1
VA9O-008 311.XX90 3
VA90-008 31AUG90 2
VA90-009 31AUG90 1
VA90-152 30JUL90 1
VA9O-132 30JUL90 2
VA90-152 30JUL90 3
VA92-451 10AUG92 1
VA92-431 10AUG92 2
VA92—431 10AUG92 3
VA9O-045 20JUL90 1
VA9O-045 20JUL90 2
VA90-045 20JUL90 3
VA90—251 30AUG90 1
VA9O-251 30AUG90 2
VA9O-251 30AUG90 3
VA92-517 26AUG92 1
VA93-602 13AUG93 2
VA90—009 31AUG90 2
VA90-009 31AUG90 3
VA90-260 20SEP90 1
VA90-260 20SEP90 3
VA9O-260 20SEP90 2
VA91-368 04AUG91 1
VA91—368 04AUG91 2
VA91.368 04AUG91 3
VA93—602 08AUG93 2
VA91—342 23JUL91 2
VA91-342 23JUL91 3
VA90-106 14AUG90 1
VA9O-106 15SEP90 1
VA90-106 15SEP90 3
VA90—106 1SSEP9O 2
VA91—370 06AUG91
VA91—410 09AUG91
VA90—106 14AUG90
VA90-106 14AUG90
VA93-722 21AUG93
VA93—722 21AUG93
VA93—722 21AUG93
VA9O-108 24AUG90 1
VA90—108 24AUG90 2
VA90-108 24AUG90 3
VA91-410 09AUG91 2
-------
92
EMAP.E VP COMMUNITY STRUCrUR
VA91—410 09AUG91 3 957
VA91—376 06AUG91 3 867
VA92—530 28JUL92 1 1290
VA92—530 28JUL92 2 1291
vA92—530 28JUL92 3 1292
VA93—735 07AUG93 1 1732
VA93—735 07AUG93 2 1733
VA93—735 07AUG93 3 1734
VA93—658 26AUG93 3 1569
VA9O-095 30AUG90 1 256
VA9O—095 30AUG90 2 257
VA9O—095 30AUG90 3 258
VA93—671 21AUG93 1 1591
VA93—671 21AUG93 3 1593
VA93—671 21AUG93 2 1592
VA93—686 01SEP93 1 1620
VA9O—050 20JUL90 1 146
VA93—627 09AUG93 2 1.511
VA93-627 09AUG93 3 1512
VA9O—050 20JUL90 3 147
VA9O—190 15AUG90 2 495
vA90-063 25AUG90 1 1.79
VA9O-063 25AUG90 2 180
VA9O-063 25AUG90 3 181
VA90-060 03AUG90 1 171
VA90-060 03AUG90 2 1.72
VA91-286 11AUG91 1 715
VA93—653 27AUG93 1560
VA91-280 09AUG91 703
VA9O-1.90 15AUG90 494
VA90—190 15AUG90 3 496
VA92—452 09AUG92 1 1096
VA92—452 09AUG92 2 1097
VAO3—617 22AUG93 3 1491
VA90—084 14AUG90 1 220
VA92-486 28AUG92 1 1177
VA92-486 28AUG92 3 1178
VA90-084 14AUG90 2 221
VA9O-086 13SEP90 1 229
VA90-086 13SEP90 2 230
VA90-086 13SEP90 231
VA9O—084 14AUG90 222
VA9O—081 27AUG90 215
VA9O-081 27AUG90 3 216
VA90—183 06AUG90 472
VA93-617 22AUG93 1 1490
VA90-132 26JUL90 1 355
VA9O-132 26JUL90 2 356
VA9O-132 26JUL90 3 357
VA91-362 05AUG91 1 842
VA91-362 05AUG91 2 843
VA91—362 05AUG91 3 844
VA92-452 09AUG92 3 1098
VA9O-183 06AUG90 1 470
VA90—183 06AUG90 2 471
VA93-653 27AUG93 2 1561
VA93-653 27AUG93 3 1562
VA93—050 29JUL93 2 1404
VA91-330 17AUG91 1 783
VA9O—044 21JUL90 3 133
VA90-100 21SEP90 271
VA93-732 17AUG93 1728
VA93-1 .50 02SEP93 1 1429
VA93-1.50 02SEP93 1430
VA93-1 .50 02SEP93 3 1431
VA93-732 17AUG93 1726
VA93—732 17AUG93 2 1727
VA91-417 11AUG91 1 972
VA91—417 11AUG91 3 974
VA91—421 10AUG91 2 982
VA91—421 10AUG91 1 981
VA91-421 10AUG91 3 983
VA91—417 11AUG91 2 973
VA91—419 10AUG91 1 978
VA91-419 10AUG91 3 980
VA91-419 10AUG91 2 979
VA91-418 11AUG91 1 975
VA91-418 11AUG91. 2 976
VA91—418 11AUG91 3 977
VA9O-118 11AUG90 1 322
VA9O—119 11AUG90 3 324
VA90-L18 11AUG90 2 323
VA9O-1 .20 11AUG90 1 328
VA9O—L20 11AUG90 3 330
VA9O-120 11AUG90 2 329
VA9O-L19 11AUG90 1 325
VA9O-119 11AUG90 2 326
VA90-119 11AUG90 3 327
VA92-532 27AUG92 1 1293
VA92-532 27AUG92 3 1295
VA92-532 27AUG92 2 1294
VA92-515 08AUG92 2 1252
VA91-348 24JUL91. 1 821
VA91-349 24JUL91 1 824
VA91-349 24JUL91 2 825
VA91-349 24JUL91 3 826
VA91-348 24JUL91 2 822
VA91-348 24JUL91 3 823
VA9O-L04 08AUG90 1 286
VA9O-J.04 08AUG90 2 287
VA90-104 08AUG90 3 288
VA90—105 08AUG90 1 292
VA9O-105 08AUG90 2 293
VA9O-105 08AUG90 3 294
VA90-256 10AUG90 1 624
VA90-2S6 10AUG90 2 625
VA90-256 10AUG90 3 626
-------
GALLAGHER & GRASSLE
VA93—666 14AUG93 1
VA93—666 14AUG93 3
VA93-666 14AUG93 2
VA90—256 23SEP90 1
VA90-256 23SEP90 3
VA93—714 19AUG93 1
VA93—714 19AUG93 2
VAS3—714 29AUG93 3
VA91-400 12SEP91
VA92-559 14AUG92
‘ 11 .92-559 14AUG92
VA92-559 14AUG92
VA93—711 19AUG93
VA93—711 19AUG93 2
VA92-565 30JUL92 1
‘11.92-565 30JUL92 3
VA92—565 30JUL92 2
VA9O-003 29AUG90 1
VA90-003 29AUG90 3
‘11.90-033 30AUG90 1
VA90—033 30AUG90 3
VA90-003 29AUG90 2
VA90-019 29AUG90 1
VA90—019 29AUG90 2
VA90—019 29AUG90 3
VA90-033 12SEP90 I
VA9O-033 12SEP90 3
‘11.90-033 12SEP90 2
VA90—033 30AUG90 2
VA91—340 07SEP91 1
VA91—340 07SEP91 2
‘11.91-340 07SEP91 3
‘11.92—466 21AUG92 1
VA92—466 21AUG92 3
VA92—491 14AUG92 1
VA92—491 . 14AUG92 2
VA92-491 14AUG92 3
VA92—485 28AUG92 1
VA92—485 28AUG92 2
v 1 .92—4e5 28AUG92 3
‘11.92—465 21AUG92 1.
VA92-465 21AUG92 2
VA92—46 5 21AUG92 3
VA92-466 21AUG92 2
‘11.90-164 01AUG90 1
VA90-164 01AUG90 2
‘11.90-164 01AUG90 3
VA90-202 02AUG90 1.
‘11.90-202 02AUG90 2
‘11.90-202 02AUG90 3
VA9O—203 01AUG90 2
VA9O-203 01AUG90 3
VA9O -040 19AUG90 1
‘11.90—040 19AUG90 2
VA90-040 19AUG90 3
VA90-087 20AUG90 1
‘11.90—087 20AUG90 2
VA9O—039 11AUG90 1
VA90-039 11AUG90 2
VA9O-039 11AUG90 3
VA90-061 25JUL90 1
‘11.90-061 25JUL90 2
VA90-061 25JUL90 3
VA90-061 06SEP90 1
VA9O-061 06SEP90 2
VA90-061 06SEP90 3
VA9 I—266 17AUG91 1
VA91-266 17AUG91 2
VAS2-453 10AUG92 1.
‘11.92-453 10AUG92 3
VA92-453 10AUG92 2
VA93-60 7 27AUG93 1
VA93-60 7 27AUG93 2
VA93-607 27AUG93 3
‘11.92-050 27AUG92 2
VA92-474 22AUG92 1
VA92-474 22AUG92 2
VA92-474 22AUG92 3
VA93—615 04SEP93 1
VA93—615 04SEP93 2
VA93-615 04SEP93 3
VA93-618 03SEP93 1
VA93-618 03SEP93 2
‘11.93-618 03SEP93 3
VA9O-047 29JUL90 1
‘11.90-047 29JUL.90 2
VA90-047 29JUL90 3
VA9X—276 16AUG91 3
VA9O-145 21JUl.90 1
VA9O—145 22JUL90 2
VA90—145 21JUL90 3
VA92-050 03AUG92 1
‘11.92—483 15AUG92 1
‘11.92—483 15AUG92 2
VAS2—483 15AUG92 3
‘11.90—086 01AUG90 2
VA90-150 14SEP90 1
V 1 .93—045 07AUG93 1
VA93—045 02SEP93 1
VA9O—086 01AUG90 3
VA92— 510 10AUG92 I
‘11 .92—510 10AUG92 3
VA92—510 10AUG92 2
VA91—263 03AUG91 1
VA91—296 21AUG91 3
‘11.91—263 03AUG91 2
VA91—26] 03AUG91 3
‘11.91—267 17AUG91 2
1579
1581
1580
627
629
1685
1686
1687
925
1354
1355
1356
1678
1671
1373
7
9
101.
103
8
56
57
58
104
106
105
102
802
803
004
1126
1128
1191
11.92
1193
1174
1175
1176
1123
1124
1125
1127
436
437
438
533
534
535
537
538
122
123
124
232
233
119
120
121
173
174
175
116
177
178
680
681.
1099
1101.
1100
1463
1464
1465
1037
1150
1151
1152
1404
1485
1486
1492
1493
1494
140
141
142
699
388
389
390
1034
1168
1169
1170
22’
39’
1398
1401
228
1236
1238
1237
674
733
675
676
684
93
-------
EMAP-E VP COMMUNITY STRUCTURE
¶11.91—267 17AUG91 3 685
VA91—045 15JUL91 1 642
VA93—045 07AUG93 2 1399
VA91-432 16JUL91 1 1004
VA91 —432 16JUL91 3 1006
¶11.93-045 01SEP93 3 1403
V1.91—432 16JUL91 2 1005
VA91—433 16JUL91 3 1009
¶11.91—433 16JUL91 2 1008
VA91—433 16JUL91 1 1007
VAU1—434 16JUL.91 1 1010
V 1 .93-633 01SEP93 1 1519
¶11.93—633 01SEP93 3 1521
¶11.93—633 01SEP93 2 1520
VA9Z-434 16JUL91 2 1011
VA91-434 16JUL91 3 1012
VA91-308 29AUG91 2 746
VA91-344 26JUL91 1 814
VA91—344 26JUL.91 815
VA9 I—344 26JUL91 3 816
¶11.91—308 29AUG91 747
¶11.91—311 27AUG91 752
VA91—311 27AUG91 1 750
VA91—311 27AUG91 751
VA91-3i6 12SEP91 1 759
¶11.91-316 12SEP91 760
¶11.91-316 12SEP91 761
¶11.91-045 15JUL91 — 643
¶11.91—045 15JUL91 3 644
¶11.91-308 29AUG91 1 745
VA92-457 15AUG92 1 1108
¶11.93—045 01SEP93 2 1402
VA93-631 09AUG93 2 1517
VA93—631 09AUG93 3 1318
¶11.93—631 09AUG93 1 1516
¶11.91—307 27AUG91 1 742
¶11.91-307 27AUG91 2 743
VA9 I—307 27AUG91 3 744
VA91—322 15AUG91 1 771
VA91-322 15AUG91 3 773
VA91-322 15AUG91 2 772
¶11.93-045 07AUG93 3 1400
VA91-317 29AUG91 1 762
VA92—488 15AUG92 1 1182
VA91-317 29AUG91 2 763
¶11.92—488 15AUG92 2 1183
VA92—488 15AUG92 3 1184
¶11.91-317 29AUG91 3 764
VA92-517 26AUG92 2 1258
¶1A90—041 20JUL90 1 125
¶11.90—041 20JUL90 2 126
VA9O—041 20JUL90 3 127
¶11.91-292 22AUG91 1 722
VA9O—057 25JUL90 3 1.64
¶11.90-057 25JUL90 1 162
VA9O—057 25JUL.90 2 1.63
VA91-292 22AUG91 723
¶11.91-292 22AUG91 724
VA9O—042 03AUG90 1.28
¶11.90—042 03AUG90 1.29
¶11.90-042 03AUG90 3 130
¶11.91-060 09JUL91 1 648
VA9 I—060 09JUL91 650
¶11.91-060 09JUL91 649
VA9 I-427 10JUL91 1 989
VA91-428 10JUL91 1 992
¶11.91-428 10JUL91 993
VA91—428 10JUL91 3 994
¶11.91—427 10JUL91 990
¶11.91-427 10JUL91 991
VA91-291 11AUG91 719
VA9I—291 11AUG91 720
VA9 I—291 11AUG91 3 721
VA91—284 12AUG91 1 709
VA91—284 12AUG91 2 710
VA91—284 12AUG91 3 711
VA9I-285 22AUG91 1 712
¶11.91-285 22AUG91 2 713
VA9I-285 22AUG91 3 714
VA92-472 23AUG92 2 1145
¶11.92-475 24AUG92 2 1154
VA92-475 24AUG92 3 1155
VA92-OSO 27AUG92 1 1036
¶11.92-487 02AUG92 2 1180
VA92-487 02AUG92 3 1181
¶11.92-060 05AUG92 2 1046
VA92-060 30AUG92 1 1048
¶11.92-060 30AUG92 3 1050
VA92-060 05AUG92 3 1047
¶11.92—060 30AUG92 2 1049
¶11.92-060 05AUG92 1 1045
¶11.92-472 23AUG92 3 1146
VA92-478 24AUG92 1 1162
¶11.92-478 24AUG92 2 1163
¶11.92—478 24AUG92 3 1164
¶11.92-475 24AUG92 1 1153
VA92-472 23AUG92 1 1144
VA92-487 02AUG92 1 1179
VA93-624 08AUG93 3 1507
¶11.93-624 08AUG93 2 1506
VA91—261 03AUG91 1 668
¶11.91—261 03AUG91 2 669
VA91-261 03AUG91. 3 670
VA9O-087 20AUG90 3 234
VA93-624 08AUG93 1 1505
VA92-050 27AUG92 3 1038
VA91-295 21AUG91 1 728
-------
vA91-295 21AUG91 3 730
VA9 I—295 21AUG91 2 729
VA90-055 24AUG90 1 159
VA9O-055 24AUG90 2 160
VA9O-055 24AUG90 3 161
VA93—060 07AUG93 1 1411
VA93-060 07AUG93 3 1413
VA93—060 07AUG93 2 1412
VA93—060 26AUG93 1 1414
VA93—060 26AUG93 2 1415
VA93-060 26AUG93 3 1416
VA93—622 07AUG93 1 1500
VA93—626 03SEP93 1 1508
VA93—626 03SeP93 2 1509
VA93—626 03SEP93 3 1510
VA9O—065 16AUG90 1 182
VA9O-065 16AUG90 2 183
VA9O-065 16AUG90 3 184
VA9O-065 07SEP90 1 185
VA9O—065 07SEP90 2 186
VA90-065 07SEP90 3 187
VA93-638 30JUL93 1 1528
VA93-638 30JUL93 2 1529
VA93—638 30JUL93 3 1530
VA91-296 21AUG91 1 731
VA93—611 28AUG93 1 1475
VA93—611 28AUG93 2 1476
VA93—611 28AUG93 3 1477
VA91-429 11JUL91 1 995
VA91-429 11JUL91 2 996
VA91-429 11JUL91 3 997
VA91-430 11JUL91 1 998
VA9I-430 11JUL91 3 1000
VA91—430 11JUL91 2 999
VA91—431 11JUL91 1 1001
VA91-431 11JUL91 2 1002
‘11 .91—431 11JUL91 3 1003
VA92-050 03AUG92 3 1035
‘11.90—028 31JUL90 2 90
VA90-028 31JUL90 3 91
VA92-540 28AUG92 1 1309
‘11.92-540 28AUG92 3 1310
VA90-150 14SEP90 2 398
VA91-150 22AUG91 1 658
‘11.91—150 22AUG91 3 660
VA91 -150 22AUG91 2 659
VA90—053 03AUG90 1 153
VA9O-053 03AUG90 3 155
VA90—053 03AUG90 2 154
VA9 I-282 12AUG91 1 704
VA9 I-282 12AUG91 3 705
VA9O-007 24JUL90 19
‘11.90-007 24JUL90 20
‘11.90-001 24JUL90 21
¶1A90—059 22JUL90 1 168
VA90-059 22JUL90 170
VA92-513 26M 92 2 1246
‘11.92-513 26AUG92 3 1247
VA90-044 21JUL90 1 131
VA90—044 21JUL90 2 132
VA90—256 23SEP90 2 628
VA9O-059 22JUL90 2 169
VA91-26 7 17AUG91 1 683
VA91—266 17AUG91 3 682
VA91—296 21AUG91 2 732
VA90-03 4 18AUG90 1 101
VA9O-034 18AUG90 109
VA90—034 18AUG90 — 108
VA91-305 23AUG91 1 739
VA91—305 23AUG91 3 741
VA91-305 23AUG91 2 740
VA9L—318 23AUG91 1 765
VA91-318 23AUG91 3 167
VA91-318 23AUG91 2 766
VA93—634 17AUG93 1522
VA93—634 17AUG93 3 1524
VA93-634 17AUG93 1523
VA9O-151. 19JUL90 399
VA9O- ISI 19JUL90 400
‘11.90-151 19JUL90 401
VA9O—153 30JUL90 - 405
VA9O-153 30JUL90 2 406
VA90-153 30JUL90 3 407
VA92-495 16AUG92 1 1203
VA92-495 16AUG92 2 1204
VA92-495 16AUG92 3 1205
VA92-496 16AUG92 1 1206
‘11.92-496 16AUG92 2 1201
VA92-496 16AUG92 3 1208
VA91-327 24AUG91 1 777
VA91—32 7 24AUG91 2 778
‘11.91-327 24AUG91 3 7 9
VA91-328 24AUG91 1 780
VA9 I—328 24AUG91 781
VA91-328 24AUG91 782
VA90—046 04AUG90 - 137
VA9O—046 04AUG90 3 139
‘11.90—046 04AUG90 2 138
VA9O—051 31JUL90 1 148
VA9O—051 31JUL90 2 149
VA90—051 31JUL90 3 150
VA90—054 31JUL90 1 156
VA90—054 31JUL90 2 157
VA9O—05 4 31JUL90 3 158
VA93—601 18SEP93 1 1454
VA93—601 18SEP93 2 1455
VA93—601 18SEP93 3 1456
‘11.91—265 18AUG91 1 677
GALlAGHER & GiuSsLE
95
T 1
-------
96
EMAP-E VP COMMUNITY STRUCTURE
VA91-265 IaAUG9L 2 678
VA9I-265 1.8AU091 3 679
VA91—276 16AUG91 1 697
VA9L—276 16AUG91 2 698
VA91—279 23AUG91 1 700
VA91-279 23AUG91 3 702
VA91-279 23AUG91 2 701
VAS2—462 11AUG92 1 1117
VA92—462 11AUG92 3 1119
VAS2-462 11AUG92 2 1118
VA92-470 23AUG92 1 1138
VA92—470 23AUG92 2 1139
VA92-470 23AUG92 3 1140
VA92-473 22AUG92 1 1147
VA92—473 22AUG92 2 1148
VA92-473 22AUG92 3 1149
VA92-454 08AUG92 1 1102
VA92-454 08AUG92 2 1103
VA92—454 08AUG92 3 1104
VA92-460 13AUG92 1 1111
VA92-460 11AUG92 3 1113
VA92-460 11AUG92 2 1112
VA93-608 04SEP93 1 1466
VA93-608 04SEP93 3 1468
VA93-608 04SEP93 2 1467
‘ 11.93-612 04SEP93 2 1479
‘11.93-612 04SEP93 3 1480
VA93-612 04SEP93 1 1478
VA91-283 23AUG91 1 • 706
VA91-283 23AUG91 2 707
VA91—283 23AUG91 3 708
VA93-622 07AUG93 2 1501
VA91-271 16AUG91 1 692
VA91-271 16AUG91 3 694
VA91—271 16AUG91 2 693
‘11.93-616 27JUL93 1 1487
VA93—616 27JUL93 3 1489
‘11.93—616 27JUL93 2 1488
VA90-005 24JUL90 1 13
‘11.90-005 24JUL90 3 15
VA9O-005 24JUL90 2 14
VA90-006 24JUL90 3 18
VA9O-006 24JUL90 1 16
VA9O-006 24JUL90 2 17
‘11.92-513 26AUG92 1 1245
VA9O-031 29AUG90 1 95
VA9O-031 29AUG90 2 96
VA90-031 29AUG90 3 97
VA90-032 04AUG90 1 98
‘11.90-032 04AUG90 2 99
VA90—032 04AUG90 3 100
VA90-035 03AUG90 1 110
VA90—035 03AUG90 2 111
‘17.90-035 03AUG90 3 112
‘11.92-503 18AUG92 1 1215
VA92-503 18AUG92 3 1217
VA92-503 18AUG92 2 1216
VA9O-015 28AUG90 1 44
VA9O-015 28AUG90 2 45
VA90-015 28AUG90 3 46
VA90-023 19AUG90 1 71
VA9O-023 19AUG90 2 72
‘17.90-023 19AUG90 3 73
VA9O-014 28AUG90 1 41
VA9O-014 28AUG90 2 42
VA9O-014 28AUG90 3 43
VA9O-016 29AUG90 1 47
VA93—663 19AUG93 1 1576
VA91-338 26JUL91 1 796
VA91—338 26JUL91 2 •79 7
VA91—338 26JUL91 3 798
VA9O-016 29AUG90 2 48
VA9O-020 03AUG90 1 59
VA9O-020 03AUG90 3 61
VA9O-020 03AUG90 2 60
VA9O-016 29AUG90 3 49
VA9O-258 10AUG90 1 631
VA91-335 06SEP91 1 787
‘11.91-335 06SEP91 2 788
VA91-335 06SEP91 3 789
VA91—337 06SEP91 1 793
VA91—337 06SEP91 2 794
‘11.91-337 06SEP91 3 795
VA90-048 22JUL90 1 143
VA90—048 22JUL90 3 145
VA9O—048 22JUL90 2 144
VA9O-085 13AUG90 1 223
VA90-085 13AUG90 2 224
VA9O-085 13AUG90 3 225
VA92—546 29AUG92 1 1323
VA92-546 29AUG92 2 1324
VA92-546 29AUG92 3 1325
VA93—688 31AUG93 1 1626
VA93-688 31AUG93 3 1628
‘11.93—688 31AUG93 2 1627
VA92-53S 03AUG92 1 1297
VA92—535 03AUG92 2 1298
VA92—535 03AUG92 3 1299
VA90-012 04AUG90 1 33
VA90-012 04AUG90 2 34
‘11.90—012 26AUG90 1 35
VA9O-0i2 28AUG90 3 37
¶1A90-012 28AUG90 2 36
‘11.90-017 28AUG90 1 50
VA90-017 28AUG90 2 51
VA9O-017 28AUG90 3 52
VA92—508 12AUG92 1. 1230
-------
( AtjACI4P & GRAS LE
97
VA92-508 12AUG92 2 1231
VA92-508 12AUG92 3 1232
VA92—5O5 10AUG92 2 1222
VA9O—144 21JUL90 1 385
VA93—663 19AUG93 2 1577
VA93—663 19AUG93 3 1578
VA92—505 10AUG92 1 1221
VA92—505 10AUG92 3 1223
VA9O—154 18AUG90 1 408
VA90—154 18AUG90 3 410
VA90-154 18AUG90 2 409
VA93—723 22AUG93 1 1706
VA93—723 22AUG93 2 1707
VA93—723 22AUG93 3 1708
VA93—641 16AUG93 1 1537
VA93—641 16AUG93 3 1539
VA93-641 16AUG93 2 1530
vA93-724 22AUG93 1 1709
VA93—724 22AUG93 2 1710
VA93—724 22AUG93 3 1711
VA93—604 26JUl.93 1 1458
VA93—604 26JUL93 2 1459
VA93—604 26JUL93 3 1460
VA9O—258 10AUG90 2 632
VA90—258 10AUG90 3 633
VA90—259 10AUG90 1 634
UA9O-259 10AUG90 635
VA9O-010 30AUG90 29
VA92—517 26AUG92 3 1259
VA9O—204 21AUG90 539
VA9O—204 21AUG90 540
VA9O—204 21AUG90 541
VA90—191 15AUG90 1 497
VA90-191 15AUG90 499
VA90—191 15AUG90 498
VA91—342 23JUL91 1 808
VA91—375 07AUG91 1 866
VA91—37 7 07AUG91 1 868
VA91—377 07AUG91 2 869
VA91—377 07AUG91 3 870
VA9O-021 21JUL90 1 62
VA90-021 21JUL90 2 63
VA90-024 21JUL90 1 74
VA9O-024 21JUL90 3 76
VA9O-024 21JUL90 75
VA92-544 08AUG92 1319
¶JA93-695 03AUG93 1642
VA93—695 03AUG93 1643
VA93-696 02AUG93 1644
VA93-696 02AUG93 1646
VA93-695 03AUG93 — 164].
VA90-079 13AUG90 1 212
VA92-079 10AUG92 1 1051.
VA92—079 10AUG92 2 1052
VA9O—079 13AUG90 2 213
VA9O-079 13AUG90 3 214
VA92-079 10AUG92 3 1.053
VA92-079 26AUG92 2 1055
VA92—079 26AUG92 3 1056
VA92-544 08AUG92 1 1317
VA92—544 08AUG92 2 1318
VA92-549 08AUG92 1 1332
VA92-549 08AUG92 2 1333
VA92-549 08AUG92 3 1334
VA93—696 02AUG93 2 1645
VA92-019 26AUG92 1 1054
VA93-079 03AUG93 1 1417
VA93-079 03AUG93 3 1419
VA93—079 03AUG93 2 1418
VA93-079 01SEP93 1 1420
VA93-079 01SEP93 2 1421
VA93-079 01SEP93 3 1422
VA9O—159 19JUL90 1 423
VA90—159 19JUL90 2 424
VA90-159 19JUL90 3 425
VA90-160 19JUL90 1 429
VA90-160 19JUL90 3 431
VA9O—160 19JUl.90 2 430
VA90-159 11SEP90 2 427
VA90-159 11SEP90 3 428
VA9O-159 11SEP90 1 426
VA91-079 23AUG91 2 652
VA91-079 23AUG91 3 653
VA91-387 25AUG91 1 893
VA91-387 25AUG91 2 894
VA91—387 25AUG91 3 895
VA92-541 06AUG92 1 1311
VA92-541 06AUG92 2 1312
VA92-541 06AUG92 3 1313
VA91—0 7 9 23AUG91 1 651
VA91—379 05AUG91 873
VA91-379 05AUG91 874
VA92-025 26AUG92 1025
VA92-025 26AUG92 2 1026
VA92 -025 26AUG92 3 1027
VA91-385 22AUG91 1 887
VA91-385 22AUG91 2 888
VA91-385 22AUG91 3 889
VA93-718 14AUG93 2 1695
VA9O-021 21JUL90 3 64
VA90-070 28JUL90 1 188
VA9O-070 28JUL90 2 189
VA90-070 28JUL90 3 190
VA93-718 14AUG93 1 1694
VA93—710 14AUG93 3 1696
VA90-022 23JUL90 1 65
VA9O-022 23JUL90 3 67
VA90-022 23JUL90 2 66
-------
98 EMAP-E VP COMMUNITY STRUCTURE
¶7A93—693 30JUL93 2 1636
VA92-025 04AUG92 1 1022
VA92-025 04AUG92 2 1023
VA92-025 04AUG92 3 1024
VA93—693 30JUL93 1 1635
VA93—693 30JUL93 3 1637
V1.30-025 22JUL90 1 17
V590-025 22JUL90 2 78
VA9O—025 22JUL90 3 79
VA93—025 01AUG93 1 1392
VA93—025 01AUG93 3 1394
VA93—025 01AUG93 2 1393
VA92—539 04NJG92 1 1306
VA92-539 04AUG92 2 1307
¶FA92-539 04AUG92 3 1308
VA93—713 13AUG93 2 1683
VA93-713 13513093 3 1684
VA90-026 22JUL90 1 80
VA90-026 22JUL90 2 81
VA90-026 22JUL90 3 82
VA9O—022 14SEP90 1 68
Vk90-022 14SEP90 3 70
VA9O—022 14SEP90 2 69
VA92-547 11AUG92 1 1326
VA92-547 11AUG92 3 1328
VA92—547 1151 )092 2 1327
VA90—096 31JUL90 1 259
VA90-096 31JUL90 2 260
VA90-096 31JUL90 3 261
VA91-379 05AUG91 1 872
VA9L-025 17AUG91 1 639
VA9L-025 17AUG91 2 640
VA91-025 17AUG91 3 641
VA93—713 13AUG93 1 1682
VA93—025 31AUG93 1 1395
VA93—68 7 31AUG93 1 1623
VA93—687 31AUG93 1625
VA93-687 31AUG93 1624
VA93—025 31AUG93 1396
VA93-025 31AUG93 1397
VA90-070 08SEP90 191
VA90-070 08SEP90 192
VA90-070 08SEP90 193
VA90-156 30JUL90 • 414
VA90-156 30JUL90 3 416
VA90-156 30JUL90 2 415
VA9 I-381 15AUG91 1 878
VA91-381 15AUG91 2 879
VA91—381 15AUG91 3 880
VA9 I-382 15AUG91 1 881
VA91—382 15AUG91 3 883
VA9L-382 15AUG91 2 882
VA9O-104 09SEP90 1 289
VA90-104 09SEP90 2 290
VA90-104 09SEP90 3 291
VA9O—157 30JUL90 2 418
VA9O-157 30JUL90 3 419
VA9O-157 30JUL90 1 417
VA9O-026 22SEP90 1 83
VA90—026 22SEP90 2 84
VA90-026 22SEP90 3 85
VA91-392 23AUG91 1 908
VA9L—392 23AUG91 2 909
VA91-392 23AUG91 3 910
VA93—686 01SEP93 2 1621
VA93-686 01SEP93 3 1622
VA93-703 25AUG93 1 1659
VA93-703 25AUG93 2 1660
VA93-703 25513093 3 1661
VA91-380 17AUG91 1 875
VA9 I-380 17AUG91 3 877
VA9 I-380 17AUG91 2 876
VA92-566 15AUG92 1 1315
VA92-566 15AUG92 2 1376
VA92—566 15AUG92 3 1377
VA93-721 2051)093 1 1700
VA93-121 20AUG93 2 1701
VA93-721 20AUG93 3 1702
VA92-561 16AUG92 1 1360
VA92-561 16AUG92 2 1361
VA92-561 16AUG92 3 1362
VA91—414 13SEP91 1 963
VA9 I—414 13SEP91 2 964
V591-414 13SEP91 3 965
VA90-004 29AUG90 1 10
VA9O-004 29AUG90 2 11
VA9O-004 29AUG90 3 12
VA9O-144 21JUL90 2 386
VA90-144 21JUL90 3 387
VA9 I-341 22JUL91 1 805
VA91—341 22JUL91 2 806
VA91-341 22JUL91 3 807
VA90-203 01AUG90 1 536
V591—262 15AUG91 1 671
VA91-262 15AUG91 2 672
VA91—262 15AUG91 3 673
VA91-230 17AUG91 1 689
VA91-270 17AUG91 3 691
VA91-270 17AUG91 2 690
VA90-029 05AUG90 1 92
VA9O—029 0551)090 3 94
VA90—029 05AUG90 2 93
VA9O—166 27JUL90 1 442
VA90-166 27JUL90 2 443
VA9O-167 27JUL90 1 444
VA9O-167 27JUL90 2 445
VA90-167 27JUL90 3 446
-------
GALLAGHER & GRASSLE 99
VA93—720 16A2 93 1 1697
VA93—720 16AUG93 2 1698
VA93—720 16M 93 3 1699
9O—O99 1U5090 1. 262
VA90—099 15AUG90 2 263
VASO—099 15AUG90 3 264
VA92—567 31JUL92 1 1378
VA92—S67 31JUL92 3 1.380
VA92—567 31JUL92 2 1379
VA90-071 24SEP90 2. 194
VA9O—071 24SEP90 3 196
VA9O—071 24SEP90 2 195
VA90—075 03AUG90 1 203
VA90—07S 03AUG90 2 204
VA9O—075 03AUG90 3 205
VA91—394 07SEP91 1 911
VA91—394 07SEP91 2 912
VA9 I—394 07SEP91 3 913
VA90—037 16AUG90 1 116
VA90—037 16AUG90 3 118
VA90—037 16AUG90 2 117
VA91—408 30JUL91 3 951
‘11 .92—562 29JUL92 1. 1363
‘11.92—562 29JUL92 2 2.364
‘11.92—562 29JUL92 3 1365
VA92— 563 31JUL92 2 1367
VA92—563 31JUL92 3 1368
VA92-563 31JUL92 1. 1366
VA91—402 12SEP91 1. 931
‘11.91-402 12SEP91 3 933
‘11.91—402 12SEP91 2 932
VA91—407 29JUL91 1 946
vA91—407 29JUL91 3 948
VA91—407 29JUL91 2 947
VA91—415 30JUL91 2. 966
VA91-415 30JUL91 2 967
VA91—415 30JUL91 3 968
VA91—408 30JUL91 1 949
VA91-408 30JUL91 2 950
VA93—681 09MJG93 1 1609
VA93—681 09AUG93 2 1610
VA93—681 09AUG93 3 1611
‘11.91—416 26JUL91 1 969
‘11.91—416 26JUL91 2 970
VA91—416 26JUL91 3 971
VA90-076 29JUL90 1 206
VA90—076 29JUL90 2 207
‘11.90—076 29JUL90 3 208
‘11.93—698 27AUG93 1 1650
V 1 .93—698 27AUG93 2 1651
VA93—698 27AUG93 3 1652
‘11.93-699 17SEP93 1 1653
VA93—699 17SEP93 3 1655
VA93-699 17SEP93 2 1654
VA90—078 18AUG90 2. 209
VA9O—078 181.21090 2 210
¶11.90-078 18AUG90 3 211
¶11.91—388 161.21091 1 896
VA91—388 161.21091 2 897
VA91—388 16AUG91 3 898
‘11.92-548 10AUG92 1 1329
¶11.92-548 10AUG92 2 1330
¶11.92-548 10AUG92 3 1331
¶11.91-401 04SEP91 2. 928
‘11.91—401 04SEP91 3 930
VA91-401 04SEP91 2 929
VA93-704 25AUG93 1 1662
VA93-704 25AUG93 3 1664
VA93-704 25AUG93 2 1663
VA91—406 29JUL91 1 943
¶11.91-406 29JUL91 3 945
VA91-406 29JUL91 2 944
VA93-709 13AUG93 1. 1671
VA93—709 13AUG93 2 1672
¶11.93-709 13AUG93 3 1673
VA92-568 16AUG92 1 1381
‘11.93-568 16AUG92 3 1393
VA92-568 16AUG92 2 1)82
VA93-690 27AUG93 2. 1632
VA93-690 27AUG93 3 1634
VA93-690 27AUG93 2 1633
VA91-386 22AUG91 1 090
VA91-386 22AUG91 2 891
VA9 I-386 22AUG91 3 892
‘11.92—542 06AUG92 2. 1314
VA92—542 06AUG92 2 1315
VA92—542 06AUG92 3 1316
‘11.90-010 30AUG90 2. 28
‘11.90-205 21AUG90 2. 542
¶11.90-205 21AUG90 2 543
VA9O-205 21AUG90 3 544
VA92-457 15AUG92 2 1109
VA92—457 15AUG92 3 1110
VA91—314 24JUL91 1 753
VA91—314 24JUL91 2 754
VA91-314 24JUL91 3 755
‘11.91-437 18JUL91 1 1019
‘11.91-437 18JUL91 3 1021
¶11.91-437 18JUL91 2 1020
‘11.92-501 28AUG92 1 1209
‘11.92-501 28AUG92 2 1210
¶11.92-501 28AUG92 3 1211
VA90-165 02AUG90 1 439
¶11.90-165 02AUG90 3 441
‘11.90-165 02AUG90 2 440
¶11.93-652 28AUG93 1 1558
VA93-652 28AUG93 3 1559
-------
EMAP-E VP COMMUNITY STRUCTURE
100
VA93—676 20AUG93
VAS3—676 20AUG93
VA93—676 20AUG93
VA90—177 12AUG90 1
VA9O—177 12AUG90
VA9O-177 12AUG90
VA91-384 27AUG91 1
V 91—384 27AUG91. 3
VA91-384 27AUG91 2
VA92—537 21AUG92 1
VA92—537 21AUG92 2
VA92—537 21AUG92 3
VA90—198 12AUG90 1
VA9O—198 12AUG90 2
vA9O-198 12AUG90 3
VA93—729 29AUG93 3.
VA93—729 28AUG93 2
VA93—729 28AUG93 3
VA9O-013 30AUG90 3
VA9O—083 24AUG90 3
VA92—518 11AUG92 3
VA9O—114 06AUG90 7 .
VA90—114 06AUG90 2
VA90—114 06AUG90 3
VA9O-115 06AUG90 1’
VA9O—115 06AUG90 3
VA9O—115 06AUG90 2
VA9O-083 24AUG90 1
VA9O-112 27JUL90 1
VA9O—112 27JUL90 3
VA9O-112 27JUL90 2
VA91—339 28JUL91 1
VA91-339 28JUL91 2
VA91—339 28JUL91 3
VA91-288 10AUG91 1
VA91-288 10AUG91 3
VA92-506 06AUG92 1
VA92-506 06AUG92 2
VA92-506 06AUG92 3
VA92-507 02AUG92 1
VA92-507 02AUG92 3
VA92-507 02AUG92 2
VA91-288 10AUG91 2
VA93—649 27AUG93 1
VA93-649 27AUG93 3
VA93-649 27AUG93 2
VA9O-113 27JUL90 1
VA90—113 27JUL90 3
VA9O—113 27JUL90 2
VA92—477 04AUG92 2
VA92-476 04AUG92 3
VA9Z-477 04AUG92 1
VA92-477 04AUG92 3
VA92—504 06AUG92 1
VA92-504 06AUG92 3
VA92-504 06AUG92 2
VA92-476 04AUG92 1
VA92-476 04AUG92 2
VA92-511 03AUG92 1
VA92-511 03AUG92 3
VA92-511 03AUG92 2
VA91-336 16AUG91 1
VA91—336 16AUG91 2
VA91—336 16AUG91 3
VA92—492 31JUL92 1
VA92-492 31JUL92 2
VA92-492 31JUL92 3
VA92-490 30JUL92 1
VA92-490 30JUL92 3
VA92-490 30JUL92 2
VA93-640 10AUG93 1
VA93-640 10AUG93 2
VA93-640 10AUG93 3
VA93-642 10AUG93 3
V7.93-642 10AUG93 2
VA9O-083 24AUG90 2
VA9O—184 27JUL90 1
VA90-184 27JUL90 2
VA9O—184 27JUL90 3
VA9O-185 27JUL90 1
VA9O—185 27JUL90 3
VA9O—185 27JUL90 2
VA9O-129 27JUL90 1
VA90-129 27JUL90 3
VA9O-184 14SEP90 ‘
VA93-642 10AUG93
VA91-343 05SEP91
VA91—343 05SEP91
VA91—343 05SEP91
VA90-184 14SEP90
VA91-269 04AUG91 3
VA91-397 27AUG91
VA91-397 27AUG91
VA90-133 11AUG90
VA9O-133 11AUG90
VA90-133 11AUG90 3
VA9O-142 24JUL90
VA90-142 24JUL90
VA9O—142 24JUL90 3
VA9O—129 27JUL90 2
VA90—184 14SEP90 2
VA9O-058 17AUG90 1
VA90-058 17AUG90 3
VA90-058 17AUG90 2
VA9 I-058 17JUL91 1
VA91-0 58 17JUL91 2
VA91—058 17JUL91 3
VA91—436 17JUL91 3
-------
VA93—058 02SEP93 1 1408
VA93-058 02SEP93 3 1410
VA92—058 30AUG92 2 1043
VA93—058 04AUG93 1 1405
VA93-058 02SEP93 2 1409
vA93—058 04AUG93 2 1406
VA93—058 04AUG93 3 1407
VA91-315 24JUL91 1 756
VA91—315 24JUL91 3 758
VA93—643 09AUG93 2 1544
VA93-643 09AUG93 1 1543
VA93—643 09AUG93 3 1545
VA91—436 17JUL91 1 1016
VA91—31 5 24JUL91 2 757
VA91—435 18JUL91 2 1014
VA91—435 18JUL91 3 1035
VA91—435 18JUL91 1 1013
VA91—436 17JUL91 2 1017
VA92—058 03AUG92 1 1039
VA92-058 03AUG92 2 1040
VA92—058 037.UG92 3 1041
VA92—058 30AUG92 1. 1042
VA92—058 30AUG92 3 1044
VA90—102 01AUG90 3 277
VA9O-102 01AUG90 3 279
VA9O—147 03AUG90 3 393
VA90—147 01AUG90 2 392
VA90-102 01AUG90 2 278
VA90—147 01AUG90 3 391
VA9O-103 31JUL90 3 283
VA90—103 31JUL90 2 284
VA90—103 31JUL90 3 285
VA90-221 26AUG90 2 582
VA93—684 07AUG93 2 1635
VA9O-1.02 19SEP90 1 280
VA90-102 19SEP90 3 282
VA90—102 19SEP90 2 281
VA9 I—397 27AUG91 1 916
VA9O-130 11AUG90 1 352
VA9O-130 11AUG90 2 353
VA90-231 19JUL90 1. 598
VA90-219 28JUL90 1 576
VA90-219 28JUL90 3 578
VA9O—252 26AUG90 2 615
VA90—219 28JUL90 2 577
VA9O-252 26AUG90 3 636
VA91-389 16AUG91 3 901
VA9O-128 27JUL90 1 346
VA90—128 27JUL90 3 348
VA9O-128 27JUL90 2 347
VA9O—2 53 26AUG90 2 618
VA9O-253 26AUG90 3 619
VA93—621 20AUG93 2 1499
VA9O-193 26JUL90 1 506
VA90-193 26JUL90 2 507
VA9O-193 26JUL90 3 508
VA90—192 26JUL90 1 500
VA90—192 26JUL90 2 501
VA90-192 26JUL90 3 502
VA90—192 07SEP90 1 503
VA90-192 07SEP90 2 504
VA90—192 07SEP90 3 505
VA91—269 04AI. 91 1 686
VA91-269 04AUG91 2 687
VA93—621 20AUG93 1 1498
VA90-136 05AUG90 1 364
VA90-136 05AUG90 3 366
VA90-13 7 05AUG90 2 368
VA9O-13 7 05i UG90 3 369
VA9O-136 05AUG90 2 365
VA9O-137 05AUG90 1 367
VA92-552 21AUG92 1 1337
VA92-552 21AUG92 2 1338
VA92—552 23AUG92 3 1339
VA92-481 06AUG92 1 1165
VA92—481 06AUG92 3 1167
VA92-481 06AUG92 2 1166
VA92—555 22AUG52 1 1346
VA92—555 22AUG92 2 1347
VA92—555 22AUG92 3 1348
VA92-136 04AUG92 1 1057
VA92-136 04AUG92 2 1058
VA92-136 04AUG92 3 1059
VA92-514 04AUG92 1 1248
VA92—514 04AUG92 2 1249
VA92—514 04AUG92 3 1250
VA90-143 24JUl.90 2 383
VA92-136 29AUG92 1 1060
VA92-136 29AUG92 3 1062
VA92-136 29AUG92 2 1061
VA93-136 03AUG93 2 1423
VA93—136 03AUG93 3 1424
VA91—319 23JUL91 1 768
VA92—522 20AUG92 1 1270
VA92—522 20AUG92 2 1271
VA92—522 20AUG92 3 1272
VA91—294 30JUL91 1 725
VA91-294 30JUL91 2 726
VA91—294 30JUL91 3 727
VA91—34 7 27AUG91 1 819
VA90-199 11AUG90 1 524
VA9O—199 11AUG90 3 526
VA90—199 11AUG90 2 525
VA93—694 26JUL93 1 1638
VA93—694 26JUL.93 2 1639
VA93—694 26JUL93 3 1640
VA90-212 11AUG90 1 560
I A I1 R & (RASSLE
101
-------
102 ____________ _________- EMAP-E VP COMMUNITY STRUCTURE
VA9O-212 11AUG90 2 561
VA9O-212 11AUG90 3 562
VA93-710 26JUL93 2 1674
VA93—710 26JUL93 3 1675
VA91-31.9 23JUL91 2 769
VA91—31.9 23JUL91. 3 770
VA93-667 02AUG93 1 1582
VA93-667 02AUG93 3 1584
VA93—667 02AUG93 2 1583
VA93-606 14AUG93 2 1461
VA93—606 14AUG93 3 1462
VA9O-011 30AUG90 1 30
VA92-520 20AUG92 3 1266
VA92-520 20AUG92 1 1264
VA9O-011 30AUG90 3 32
VA9O-011 30AUG90 2 31
VA92-520 20AUG92 2 1265
VA9O-124 27AUG90 3 342
VA9O—124 27AUG90 1 340
VA9O-125 27AUG90 1 343
VA9O—125 27AUG90 2 344
VA9O-125 27AUG90 3 345
VA90—124 27AUG90 2 341
VA9O-013 30AUG90 1 38
VA9O—013 30AUG90 2 39
VA90—223. 26AUG90 1 581
VA90—250 30AUG90 1 608
VA90—250 30AUG90 3 610
VA9O-250 30AUG90 2 609
YA9O-221 26AUG90 3 583
VA90—036 31AUG90 1 113
VA90—036 31AUG90 2 114
VA9O-036 31AUG90 3 115
VA92-51.6 08AUG92 1 1254
VA92—51.6 08AUG92 3 1256
VA92-516 08AUG92 2 1255
VA93-650 29AUG93 1 1552
VA93-650 29AUG93 2 1553
VA93—650 29AUG93 3 1554
VA90—100 09AUG90 1 268
VA90—100 09AUG90 2 269
VA90-100 09AUG90 3 270
VA90-U1 09AUG90 1 307
A90-111 09AUG90 2 308
VA9O—111 09AUG90 3 309
VA93—619 28JUL93 1 1495
VA93—619 28JUL93 2 1496
VA93—619 28JUL93 3 1497
VA93-623 08AUG93 1 1502
VA93—623 08AUG93 3 1.504
VA93—623 08AUG93 2 1.503
VA9O-224 18AUG90 1 585
VA9O-252 26AUG90 1 614
VA92-493 29JUL92 1 1197
VA92-493 29JUL92 2 1190
VA92-493 29JUL92 3 1199
VA9 I-395 21AUG91 2 914
VA91—395 21AUG91. 3 915
VA90-027 05AUG90 1 86
VA9O-027 05AUG90 2 87
VA9O-027 05AUG90 3 88
VA9 I—409 26JUL91 1 952
VA91—409 26JUL91 2 953
VA91-409 26JUL91 3 954
VA90—052 26AUG90 1 151.
VA90-052 26AUG90 2 152
VA92—551 10AUG92 1 1335
VA92—551 10AUG92 3 1336
VA92-569 27JUL92 7. 1384
VA92-S69 27JUL92 2 1385
VA93—716 14SEP93 1 1688
VA93—716 14SEP93 3 1.690
VA93-73 .6 14SEP93 2 1.689
VA93—697 04AUG93 1 1.647
VA93-697 04AUG93 2 1.648
VA93—697 04AUG93 3 1.649
VA91-391 03SEP91. 1 905
VA91—391. 03SEP91. 3 907
VA91—391 03SEP91. 2 906
VA92—455 08AUG92 1 1105
VA92—455 08AUG92 2 1106
VA92-455 08AUG92 3 1107
VA9O—257 22JUL90 1 630
VA92—557 28JUL92 2 1349
‘ 11 .92-557 28JUL92 3 1.350
VA92—564 30JUL92 1. 1369
VA92—564 30JUL92 2 1370
VA92—564 30JUL92 3 1371
VA91—389 16AUG91. 1. 899
VA91—389 16AUG91. 2 900
VA91-390 05SEP91 902
VA91-390 05SEP91 903
VA91-390 05SEP91 904
‘11.93—717 15SEP93 1691
VA93—717 15SEP93 1.692
‘11.93-717 15SEP93 3 1693
VA92—558 28JUL92 • 1351.
VA92-558 28JUL92 2 1352
VA92-5S8 28JUL92 3 1.353
VA91-413 09AUG91. 1 960
VA91-413 09AUG91 3 962
‘11.91-413 09AUG91. 2 961
‘11.93—660 22AUG93 1 1.573
‘11.93-660 22AUG93 3 1.575
VA93—660 22AUG93 2 1574
VA93—654 15AUG93 1 1563
VA93—654 15AUG93 3 1.565
‘11.93-654 15AUG93 2 1.564
-------
GALLAGHER & GRAsSLE 103
VA9O-10 7 14AUG90 1 301
VA90—10 7 14AUG90 3 303
VASO-10 7 14AUG90 2 302
VA92—545 28AUG92 1. 1320
VA92-5 4 5 28AUG92 2 1321
VA92—545 28AUG92 3 1322
VA93— 7 12 14AUG93 1. 1679
VA93—712 14AUG93 3 1681
vA93-712 14AUG93 2 1680
VA93—689 28AUG93 1 . 1629
VA93—689 28AUG93 2 1630
VA93—689 28AUG93 3 1631
VA9O—073 04AUG90 1 197
VA90-0 7 3 04AUG90 2 198
VA90—013 04AUG90 3 199
VA9O—0 7 4 03AUG90 1 200
VA90—0 74 03AUG90 3 202
VA90—074 03AUG90 2 201
vA91—399 06SEP91 1. 922
VA91399 06SEP91 2 923
VA91—399 06SEP91 3 924
VA92-553 18AUG92 1 1.340
VA92—553 18AUG92 2 1341
VA92-553 18AUG92 3 1342
VA93- 7 0 7 17SEP93 1 . 1668
VA93-70 7 17SEP93 3 1670
VA93-70’ 17SEP93 2 1669
VAP2—S54 18AUG92 1 1343
VAS2-554 18AUG92 2 1344
VA92—554 18AUG92 3 1349
VA93-706 18SEP93 1 1665
VA93—706 16SEP93 2 1666
VA93-706 16SEP93 3 1667
VA91—398 06SEP91 1 919
VA91-398 06SEP91 2 920
VA91-398 06SEP91 3 921
VA90—088 26AUG90 1 235
VA90—088 26AUG90 2 236
VA90-088 26AUG90 3 237
VA92—1 7 8 23AUG92 2 1079
VA92-1 7 8 23AUG92 3 1080
VA9O—088 21SEP90 1 238
VA9O—088 21SEP90 2 239
VA90-088 21SEP90 3 240
VA9O-1 72 26AUG90 1 453
VA9O-1 72 26AUG90 2 454
VA9O-172 26AUG90 3 455
VA9O—21 6 10AUG90 1. 569
VA9O-216 10AUG90 3 571
VA9O—216 10AUG90 2 970
VA91—422 29AUG91 3 986
VA92—S 7 0 23AUG92 1 1386
VA91—422 29AUG91 2 985
VA91-42 2 29AUG91 1 984
¶FA92-570 23AUG92 2 1387
VA92—5 7 0 23AUG92 3 1388
VA93—215 03SEP93 1 1453
VA91—300 29JUL91 2 737
VA92—1 7 8 23AUG92 1 1078
VA90—090 26JUL90 1 244
VA90—090 09SEP90 3 249
VA90—140 15AUG90 1 373
VA90—140 15AUG90 3 375
VA9O—195 26JUL90 2 513
VA93-6 7 3 29JUL93 1 1597
VA93—6 7 3 29JUL93 3 1599
VA93—673 29JUL93 2 1598
VA9O—090 26JUL90 3 246
VA9O—195 26JUL90 1 912
VA9O—101 11AUG90 3 276
VA9O—101 11AUG90 2 275
VA92—525 22AUG92 1 1279
VA9O—228 20AUG90 3 594
VA93—610 16AUG93 3 1474
VA90—214 11AUG90 1 963
VA90—214 11AUG90 2 564
VA90—214 11AUG90 3 565
VA91—396 09AUG91 3 834
VA93-610 16AUG93 1 1472
VA93—610 16AUG93 2 1473
VA9O-101 11AUG90 1 274
VA90-090 05SEP90 1 247
VA91-215 29AUG91 2 667
VA90-091 14AUG90 3 251
VA93—628 19AUG93 1 1513
VA9O-215 10AUG90 1 566
VA90-140 15AUG90 2 374
VA90-209 22AUG90 1 551.
VA91—]S1 30JUL91 1. 828
VA92-215 12AUG92 1 1087
VA92-21.5 12AUG92 3 1099
VA90-21 7 09AUG90 1 572
VA92-571 24AUG92 2 1390
VA93-219 27JUL93 3 1452
VA92-5 7 1 24AUG92 3 1391
VA9O-21 7 09AUG90 2 573
VA90-090 26JUL90 2 245
VA90-195 26JUL90 3 514
VA92-525 22AUG92 3 1280
VA90-091 14AUG90 1 249
VA90—19 7 05AUG90 1 518
VA90-19’ 05AUG90 2 519
VA9O-19 7 05AUG90 3 520
VA9O-215 10AUG90 2 567
VA93-635 23AUG93 1 1525
VA9O-215 10AUG90 3 568
VA93—635 23AUG93 2 1526
-------
104 EMAP.E VP COMMUNITY STRUCTURE
VA90—091 14AUG90 2 250
VA93—635 23AUG93 3 1527
VA91—300 29JUL91 3 738
‘11.90—209 22AUG90 2 552
VA93—609 15AUG93 1 1469
‘11.93-728 15AUG93 ‘ 1718
VA9O-209 22AUG90 553
‘11.93-609 15AUG93 1471
‘11.93—609 15AUG93 1470
‘11.93—728 15AUG93 1719
VA91—298 30JUL91 735
VA93-628 19AUG93 1514
VA93-628 19AUG93 1515
VA93—728 15AUG93 1717
VA9O—196 05AUG90 515
VA9O—196 05AUG90 — 517
VA9O—196 05AUG90 2 516
VA91—351 30JUL91 2 829
VA92—468 14AUG92 2 1133
VA90—218 09AUG90 1 574
VA90—218 09AUG90 2 575
‘11.93—726 28JUL93 1 1712
VA93—726 28JUL93 2 1713
VA93-726 28JUL93 3 1714
‘11.90-130 11AUG90 3 354
VA9O-232 20AUG90 1 602
V&90—232 20AUG90 2 603
VA9O-232 20AUG90 3 604
VA92—178 31JUL92 1 1075
VA92—178 31JUL92 2 1076
‘11.92—178 31JUL92 3 1077
VA90—169 21AUG90 1 450
‘11.90-169 21AUG90 3 452
VA90-169 21AUG90 2 451
VA9O-178 19AUG90 2 468
VA91-365 11AUG91 1 848
‘11.91-365 11AUG91 3 850
‘11.91-365 11AUG91 849
‘11.90—228 20AUG90 1 592
‘11.90—228 20AUG90 593
VA90-233 19JUL90 605
‘11.90-233 19JUL90 607
‘11.90-233 19JUL90 — 606
VA90-188 26AUG90 1 488
‘11.90—188 26AUG90 2 489
VA90—188 26AUG90 3 490
VA90—200 05AUG90 1 527
VA90—200 05AUG90 2 528
‘11.90-201 05AUG90 3 532
VA9O-201 05AUG90 2 531
‘11.90-201 05AUG90 1 530
VA91-300 29JUL91 1 736
VA9O-200 05AUG90 3 529
VA9O-210 23JUL90 3 556
VA92-464 17AUG92 3 1122
VA92-464 17AUG92 1 1120
‘11.92—464 17AUG92 2 1121
‘11.92—469 16AUG92 1 1135
¶11.92—469 16AUG92 3 1137
VA92-469 16AUG92 2 1136
‘11.93-178 27JUL93 1 1438
VA93-178 27JUL93 2 1439
VA93-178 27JUL93 3 1440
VA93—178 31AUG93 1 1441
¶11.93—178 31AUG93 2 1442
¶11.93-734 26JUL93 1 1729
VA93-734 26JUL93 1731
VA93-734 26JUL93 1730
‘11.90-210 09SEP90 1 557
VA9O-210 09SEP90 2 558
VA9O-210 09SEP90 3 559
VA9O-178 19AUG90 1 467
VA9O-178 19AUG90 3 469
VA93-188 25AUG93 3 1449
‘11.93-1.88 25AUG93 1 1447
VA9O-208 22AUG90 1 548
VA9O-208 22AUG90 2 549
VA9O-208 22AUG90 3 550
‘11.93—178 31AUG93 3 1443
VA92-188 26AUG92 1 1084
VA92-188 26AUG92 2 1085
‘ 11.92-1.88 26AUG92 3 1086
VA91—309 29JUL91 1 748
VA93-613 17AUG93 1 1481
‘11.93—613 17AUG93 3 1483
VA9I-309 29JUL91 2 749
VA93—188 25AUG93 2 1448
VA93—639 21AUG93 2 1532
VA9O-210 23JUL90 1 554
‘11.90-210 23JUL90 2 555
‘11.90-1.89 26AUG90 1 491
VA9O-189 26AUG90 2 492
VA9O-189 26AUG90 3 493
VA9 I—1.88 22JUL91. 1 664
VA92—467 16AUG92 1 1129
‘11.92—467 16AUG92 3 1131
VA92-467 16AUG92 2 1130
VA93-655 03AUG93 1 1566
VA93-655 03AUG93 3 1568
VA93-655 03AUG93 2 1567
VA91-353 30JUL91 1 830
VA91—353 30JUL91. 2 831
VA91—353 30JUL91 3 832
VA92-521 28AUG92 1 1267
VA92-521 28AUG92 2 1268
VA92-521 20AUG92 3 1269
VA93—639 21AUG93 1 1531
VA93-639 21AUG93 3 1533
-------
GALLAGHER & GRASSLE - - ________________________________ tos
vA92-188 27JUL92 1
Vk92-188 21JUL92 3
VA92—188 27JUL92 2
‘Th92—461 14AUG92 1
VA92—461 14AUG92 2
VA92—461 14AUG92 3
VA93—613 17AUG93 2
VA93—651 01AUG93 1
VA93—672 01AUG93 1
V1.93—672 01AUG93 2
VA93-672 01AUG93 3
VA9X—275 05AUG91 2
VA91—275 05AUG91 3
VA93—215 27JUL93 1
VA93—670 01AUG93 1
VA93—670 01AUG93 2
VA93—670 01AUG93 3
VA9O-089 01AUG90 1.
VA90-089 07AUG90 2
VA9O-254 15SEP90 3
VA90—089 07AUG90 3
VA9O—186 25AUG90 2
VA90—187 25AUG90 2
vA90-187 25AUG90 3
VA9O—186 25AUG90 3
VA9O—187 25AUG90 1
VA9O—186 25AUG90 1
VA90-254 25JUL90 1.
VA9O-254 25JUL90 2
VA90-254 25JUL90 3
VA91-411 28AUG91 1.
VA92—484 06AUG92 1.
VA92—484 06AUG92 2
VP.92—484 06AUG92 3
VA93—645 10AUG93 I
VA93-64S 10AUG93 2
VA93—645 10AUG93 3
VA93—731 19AUG93 1
VA92—468 14AUG92 1
VA92-468 14AUG92 3
VA92—560 22AUG92 1
VA92-560 22AUG92 2
VA93-215 27JUL93 2
VA93-736 27JUL93 2
VA9O-143 24JUL90 3
VA92—494 29JUL92 I
VA92-494 29JUL92 2
VA92-494 29JUL92 3
VA93—731 19AUG93 2
VA93—731 19AUG93 3
VA90—207 22AUG90 1
VA9O-207 22AUG90 2
VA90-207 22AUG90 3
VA9O-253 26AUG90 1
VA91-347 27AUG91 3
VA90-220 18AUG90 2
VA92—471 18AUG92 3
VA92—471 18AUG92 1
VA92—471 18AUG92 2
VA90-194 26JUL90 1
VA90-194 26JUL90 2
VA9O-194 26JUL90 3
VA91-298 30JUL91 1
VA90—139 26JUL90 1
VA9O-139 26JUL90 3
VA9O—139 26JUL90 2
VA92-519 05AUG92 I
VA92-519 05AUG92 2
VA92-519 05AUG92 3
VA91-346 27AUG91 3
VA9L-356 09AUG91 I
VA91-188 22JUL91 2
VA91-188 22JUL91 3
VA91-333 22JUL.91
VA91—326 23JUL91
VA91—326 23JUL91 1
VA91—326 23JUL91
VA92-502 27JUL92 —
VA92-502 27JUL92 3
VA92-502 27JUL92 2
VA93-188 02AUG93 1
VA93—188 02AUG93 3
VA93-188 02AUG93 2
VA93-651 01AUG93 2
VA93-651 01AUG93 3
VA91—333 22JUL91 2
VA91—333 22JUL91 3
VA90-093 23AUG90 2
VA91-403 21AUG91 1
VA91—403 21AUG91 3
VA91-403 21AUG91 2
VA9O-143 24JUL90
VA90-223 11SEP90 3
VA90-225 29JUL90
VA90-225 29JUL90
VA92-560 22AUG92
VA91—350 28AUG91
VA90-227 08AUG90 1
VA9O-227 08AUG90 2
VA9O-227 08AUG90 3
VA90-229 08AUG90 1
VA90-229 08AUG90 3
VA92-526 22AUG92 1
VA92-526 22AUG92 2
VA92-526 22AUG92 3
VA93—6 7 8 27JUL.93 3
VA91—424 29AUG91 1
-------
VAS2-524 21AUG92 1
V7.92-524 21AUG92 3
VA92—21 5 12N3G92 2
VA92—21 5 23AUG92 1
VA92—215 23AUG92 2
VA92—215 23AUG92 3
VA92—571 24AUG92 1
VA92—523 21AUG92 3
VA92—524 21AUG92 2
VA93—127 29JUL.93 3
VA93—736 27JUL93 1
VA93—736 27JUL.93 3
VA90—229 08AUG90 2
vk93-678 27Jul.93 1
VA90-231 10SEP90
VA9O—231 10SEP90
VA90-231 10SEP90
VA93—678 27JUL93
VA9O—141 24JUL90 1
VA9O—141 24JUL90
VA90—141 24JUL90
VA9O-220 18AUG90 1
VA90-225 29JUL90 1
VA92-523 21AUG92 1
VA92-523 21AUG92 2
VA93—727 29JUL93 1
VA91-357 09AUG91 1
VA91—360 10AUG91 1
VA91—357 09AUG91 3
Vh91-357 09AUG91 2
VA91-360 10AUG91 2
VA91-358 10AUG91 1
VA91-3 58 10AUG91 3
VA91-424 29AUG91 3
VA91-411 28AUG91 3
VA93—6 7 5 28JUL93 1
VA93-6 7 5 28JUL93 2
VA93-675 28JUL93 3
VA91-090 05SEP91 1
VA91-090 05SEP91 2
VA9 I-346 27AUG91 1
VA91-400 12SEP91 2
V&91-400 12SEP91 3
106 - EMAP.E VP COMMUNiTY STRUCTURE
1276
1278 ____________
1088
1090 _______
1091 ___________
1092 ___________
1389
1275
1277
1716
1735
1737
596
1606
599
600
601
1607
376
378
377
579
586
1273
1274
1715
835
840
837
836
841
838
839
988
959
1600
1601
1602
654
655
817
926
927
IDNO I I
CNESS Distance (NESSm = 25)
LEVEL 0.04 0.20
0.40
0 • 60
0 • 80
1.00 1.20
I ___
-------
GALLAGHER & GI ssLE 107
APPENDIX V SPECIES CLUSTERS FOR ALL 551 EMAP-E VP TAXA
All EMAP-E VP taxa were clustered using the species-clustering methods described in Trueblood etal. (1994).
Species are clustered using single linkage clustering of cos e, where 9 is the angle between pairs of species
vectors in the Gabriel covariance biplot (Figure 15).
Pearson’s R
LZVZL 0.97
NPJ ID U0
ABLAPARA 1
CRYPTOrE 135
LUMBRICU 275
pA2AL1 ur 375
ALMYPROX 12
M7IOMSPA 57
B000TRII 70
COuR EL 118
DOLICHOP 154
OORVRUDO 157
HEXBLD 226
NAISPMD 316
PAGUACA 360
PRO7 ET 439
SLAVAPPE 480
TELMVE3D 524
8 1?1ALTE 63
PLE LAB 408
PSEUBOR! 445
POLYI DU 424
VITRFLOR 550
?.STACR 48
NAN0 RAY 318
TUBIBROW 333
GNOUIIINU 195
POLIIIERO 414
SCOLQUAD 470
SPHAERØD 484
POLIPOLI 415
BUSYCON 81
PSEUDOcH 447
STEPTAND 501
CRYPFULV 134
MICROCHI 304
ASTACAST 47
PROTODRI 443
PISIR 60 403
MICRFRAG 302
OPHRYOTR 353
PROTCHI..E 440
ARICCERU 42
P.STA.SPEA 49
KESI .ON 222
M IR0SE 221
PROrxEFE 442
CAECJO)E4 84
TANAPSAN 518
SCOLSOUR 472
STERUNIS 505
TRAVSPEB 530
BAmPARK 61
PARALNGI 376
B000SPEA 69
DONAVARI 155
I4ANCSTEL 299
CALLLAEV 90
CIIXRCOEC 107
PAP.AATrE 368
OPHEBICO 350
PMAKOLM 374
ACANSIXI 3
PARAFULG 373
POLYSPEB 428
srREAREB 508
PSEUMINO 448
DISPUNCI 152
POLYSPE?. 427
PROI IGL 444
ANCIDEPA 27
PROTDEXC 441
PSEUCARO 446
DORIOBSC 156
STLLCONV 513
ORBYRISE 354
PARADSPB 372
ANPHARCT 19
PAR?.SPEC 381
ANCIJQNE 29
OW PUSI 358
SQUIEBPU 494
STEBBOA 495
TANYORSI 520
ACMMILL 2
rRAVSPEA 529
APANMAGN 34
PARACAUD 369
CXRROSPB 112
ONUPEREB 348
NEPHBUCE 320
LEPTTERU 260
SCOLTV.XP. 473
SPMAOUAD 486
ACROCPAN 4
0.84 0.68 0.52 0.37 0.21 0.05
-------
ios EMAP.E VP COMMUNITY STRUCTURE
PAGUMINU 361 _______
ASTYLUN . 52
S.r (D U 499
AEQUtRR& 8
POLYWESS 430 _________________________
aocLxAxh 68
CARAROBS 94
PAGUPOLL 363
FARGBUSH 192
MAG .ON& 285
3PIO3C 488 ______
NICOZOST 333
PAPOCARC 382
APHSLOCN 35
BRAZ4WZU. 78
STREPETT 510
BUS7 .ZO 80
SIGAAJ Z 475
SPXSSOLX 493
CALLSETI 91
COPBCONT 119
PSETJOBLZ 449
iiAR 0(ACG 216
ARICWASS 44
RHEPHUDS 459 _________________
DRIL.SPEB 160
LEVISPEA 263
STREVARI 511
AUTOSPEA 58
SABEVULG 463
CAI3LSPEB 98
DORVSPM 158
NEPHPICT 331
ORBISWAN 355
PARAP UO 380
RHEPEPIS 458
178
BATECATH 60
ELASLAEV 165
ANOPPETI 33
HAMISOLI 212
TAGEDIVX 516
SABRELON 462
AMPIV R 18 ___________________
Z IP2CCRTU 41
PYGOELZO 453
RUDXNAGL. 461
UNCXDISS 542
APOPPYGN 36
NERQ4ERC 299
T .LASXL 523 I ____________
DIOPCUPR 151
EPITHUMP 173
sr vALI 500
ETEO?OLI 179 ________________________
TURB’AEO 537
TURBEPEB 539
AMPIVALI 22
SYLLVERR 514
CERATUBU 101
KURTATRO 241
OXYUPHIT
CAECSPEB 87
NATXPUSZ 322
SXL.ICOST 478
SYNCANER 515
B E&3VA 16
PODAOBSC 413
SPIOCOST 489
TAW.SPEA 519
SRANCLSW 76
EXCGDISP 187
SPHATAYL 487
CALLBREV 89 _________________________
POLY CIM 419
CYATSURB 136
CYMACOMP 141
ZOOTBALT 233 _____________
JAERI4AAX 239 ____________
LACUVINC 242
NICRABER 300
ERxcA rE 175
MICRGRYL 303
NOTOLURZ 336
OLIGOCEA 347
AIIPHORNA 20
EOBRSPIN 171
HARMINUR 215 ______
LYSZALBA 280 1 1
PRIOHETE 431
COROACUT 122
DE3CATHEC 146
GARMOCEA 198
MICRANOM 301
MICRRANE 305
D MXCR 144
PISTPALI4 405
COROINSI 125
IDOTPHOS 234 ___________
ODONPULG 341 ____________________________
SEILADAI( 474
LEPISQUA 254
JMSMARM 240
AMPILONG 21
COROACHE 121 ______________________________________
LEPTDUBI 257
-------
GALLAGHER & GRASSLE 109
f vAsP!A 66
000SSPEA 343
Ufl T 1X 276
SOL.EVELU 481
ARABSPEA 38
L 4DSMIT 250
iaCRSiNX 307
pARALO7 377
NEANAR 323
NEOPSATI 327
CAULBXOC 97
000ECACE 153
EUNISANG 185
UROSCIN! 547
R 416
OPSA1 169
NEANVIRE 325
: — -:--—i-—-—————---——-—-——-—
KARPBCLL 291
ANACLAPR 23
PE RPX0L 387
PHYLAREU 395
GIXCAN!R 203
NASSTRXV 320
MUCRMUCR 310
PAGULONG 362
CTi.13102 140
SCOLBOUS 467
PROCVXCI 438
BPITRUPX 174
N COT T 283
OPHXIJROI 352 I ___________________
KALMSPEB 287
P1CT000L 384
BOONBISU 71
ILTAGBSO 235
ErzoxETE 180
STBLBEUZ 509
LEXTROBU 249
MICRSCZE 306 ______
NYAAR 4 313
SPIOSETO 492
BOONII4PR 72
MELD4ITX 298
POLICOBS 418 ________________________________________________________________________
PARAAEST 367
CRASVIRC 132 — ‘
(OGSPEA 189
UNCXSERE 545
COR0rUBE 129 _______________________
!RXCBRAS 176
PIONSPEB 401
CP.ANSEPT 131 _______
PARACYPR 371
PROCCORN 435
EUPLCAUD 186 ______________
LEPXSUBL. 255
LTONIWAL 279 _____________ - I I—.——
SCOLRUBR 471
ACTEORTZ 6
BRADVILL 75
HUTOU.CR 229 _____________________
PAN000UL 365
PITANORE 406 r
SIPUNCUL 479
NINONIGR 334
NUCUDELP 340
NOTOSPIN 338
PARGBART 191
KELIXACU 297
POLThA 423
SCOLNEBE 469
TYPOAL._J 541
AGZ.ACXRC 9
BYBLSERR 82 ____________ _______________
GONIGRAC 208 _________________
COROCRAS 124 ______
CERAPINE 100
ASTAUNDA 50
CTCLBORE 138
IIIPPSERR 221
CALTPSPA 92
SCOLCAPE 468
SPIOPILI 490
CXRRSPEA 113 _________________
PHOTPOLL 392 ________________________________________________ I
UNCXINER 543
CHONINPU 109
OPHEACUII 349
ORC7O(XMJ 356 __________
PHYLMUCO 397
ElcoGvERU 190
EUCHELEG 182 ___________________________
!SCHANGU 237
L.AONICROT 247
A ZAGRS 17
UNCIIRRO 544
COROBONN 123
PHOTDEBT 391
STERCAND 503
PHYLHACL7 396
EXOGHEBE 188
-------
EMAP-E VP COMMUNITY STRUCTURf
110
£RXCPASC 177
OPHIGIGA 351
XEI0SPE 295
PTXLT 1U 452
PROJTIZ 433
HARx rE 214
PO1.YCAUL 417
rrBPDS 309
PHOIQiOLa 394
sT IGRk 497
ANOEGRAC 30
ARC?ISLA 39
HAVESCAB 220
ARIGMAMA 45
DTOPXONA 163
AS710i P 51
wARAc rr 527
NAR21600R 219
PLNOSPDJ 411
S171 .E(X 506
LEFrPING 258
PROLI4INU 389
PLEUXNER 409
POLYQUAD 425
SPIOLINI 491
TRoaluur 532
POLYSOCE 426
CLYZftORO 116
EUDOPUSI 184
DXASSCUL 148
XYXXXNFU 315
STERINER 498
PSEURERE 451
EUCHINCO 183
HYPELONG 232
MYRIOCUI. 314
STERSCUT 504
TERESTRO 525
DRILL.ONG 159
ANONLZL4 32
DIASQUAD 147
CAUDAR 4 96
GLYCROBU 205
LADNICE 246
HARPPROP 218
000SSULC 344
AGLAVERR 10
NACRZCNA 284
EPXTGREE 172
ANCIMART 28
POLYGORD 422
LUNEACIC 274
CABXINCE 83
SCALINPL 466
N TZN 326
LXSTSMIT 271
TRALASSI 526
ANACOBES 24
CAECREGU 85
MELANELL 296
SPXAACIC 483
CAECSPEA 86
T 3RRSPEA 540
GASTSPEA 200
PERTPULC 385
PIONSPEA 400
BOONSESZ 73
I4ARPSMG 292
COROSEST 127
L WE3S 251
POLYGIAR 420
HEXAANGU 224
CAPITELL 93
OVALOCEL. 357
LAEVMORT 245
UPQGAFFI 546
LEPICONE 252
THARSPA 528
CERIAMER 102
PHERAFFI 388
LIVISRAC 262
NZPEXRCX 330
SIGAT 1T 477
NUCUA U 339
YOZ.DLIRA 551
FARGGIBB 193
ACTICANE 5
LISTCLYM 270
LOI *IEDU 273
SACCXOWA 464
PARAPX 379
PSEUPAUC 450
ODOS O 342
PARVMULT 383
PBOROEXS 390
ALIGELEV 11
NASSVXBE 321
PODALEVI 412
Az.p a ETE 13
GITANOPS 202
PARAWrE 378
CRATPXLP. 133
HARGRAPA 213
ARABIRMU 37
LEPXVMZ 256
ARXCFRAG 43
OGYPALPM 346
AMASCAPE 14
SIDIRE 170
GLYCOIBR 204
-------
1
GALLAGHER & GRASSLE 111
MIADTP.AN 26
EUCEPRAE 181
GAI@4SU11I 199
BHMIHETE 63
NEPXCRYP 329
PRIOPERK 432
XRLHSPk 286
NEREGR&Y 332
CXRRGR?.N 111
PflOTP X1N 393
P I14ARG 386
c3thEVARX 104
PISTCRXS 404
GYPTVITT 210
SXGABA S 476
pO1 GXBB 421
CERAIRRI 99
•NJR&INTR 538
LISTBARN 269
601104PA 337
ASABOCUL 46
D Z 293
,TTP.11. 164
COSSSOYE 130
MEDICALI 294
I10LxLa Zz 311
NOTOLOBA 335
LTC50LI 206
SAYEDIES 465
zJUCM 261
AMNILIXO 15
QUISMULT 454
DXCRNRRV 149
TANYThRS 522
FERRISSI 194
aR P.Z3 196
ERCHYTRA 161
L EOFY 261 _________________
103171111 534
Lx)o UDEK 268
IIANOCLAD 319
VALVSINC 548
RHE0TANY 457
IRRLNT 106
POLYTRIP 429
ANODONTA 31
OLYPTOTE 207
BP.ACXYCE 74
1DOCHIR 168
HYDROPTI 230
PHYSELLA 398
DICROTER 150 —
GONIVIP.G 209
CARHIS 88
OECETIS 345
STICTOCH 507
ARCTLOXO 40
BITHTSMT 64
CORBFLUM 120
TRICORYT 531 —
D 4ICRYP 143
SPIU 2ROM 485
LA&VFUSC 244
HEXP.GERI 225
LI)0(CLAP 266
BRAIVNID 79
ELLICOMP 166
PIGUMICH 399
SP!CJOSI 482
AXARUS 59
CHASTOGA 103
GAMMFASC 191
CYRNYRAT 142
PLEUROCE 410
AULOLX) 7 53
AULOPIOD 55
PROCXOL.O 436
DUBIRAPH 161 -
BRANSCWE 77 -
DERODIGI 145
CRIRONOM 108
CINCWXNIC 110
MUSCTPJ.N 312
PISIDIUM 402
STERELMI 496
PARACLAD 370
STYLLA U 512
AULOPMJC 54
C8A070110 105
PROBEZZX 434
cov o!3S 117 -
TANYPUS 521 -
HAZSPSEU 317
AULOPLUR 56
ISOCFREY 238
HARNISCH 217
ZLYOTD(P 236
LI76 CERV 265
CLADOPLE 114
LYOGYRUS 277
VALYTRIC 549
103 17110 535
PROCSUBL 437
STEPrRIV 502
-------
112 EMAP-E VP COMMUNiTY STRUCTURE
DEZZXk 62
CLADOI?.N 115
PALPOIW 364
HABE$PEC 211
azv z 290
IO .ZGZ 67
C TPOLI 137
PEfll(MR 460
OROLACU 126
MOBSFLOR 228
MCNOSPE1 308
LZPTP&ON 259
Lfl? IU 212
MANMEST 288
LA2OCULV 243
LEITPRAG 248
Z.EPXDYTI 253
CASSOVAL 95
0BALT 281
RA2I0C 456
TUBIHITE 536
CYCLVARX 139
T Z!PLEB 517
‘ — 201
C0 TTC 282
UTEVILI 223
H1CDRTRU5 231
10 90
0.97 0.84 0.66
0.52 0.37 0.21 0.05
Pearson’s r
------- |