EPA-600/3-78-030
March 1978
Ecological Research Series
TECHNIQUES FOR SAMPLING AND ANALYZING
THE MARINE MACROBENTHOS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Corvallis, Oregon 97330
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency. have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. ‘Special” Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-73-030
March 1978
TECHNIQUES FOR SAMPLING AND ANALYZING
THE MARINE MACROBENTHOS
by
Richard C. Swartz
Marine and Freshwater Ecology Branch
Con/all is Environmental Research Laboratory
Marine Science Center
Newport, Oregon 97365
CORVALLIS ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97330
-------
DISCLAIMER
This report has been reviewed by the Corvallis Environmental
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recomendation
for use.
11
-------
FOREWORD
Effective regulatory and enforcement actions by the
Environmental Protection Agency would be virtually impossible
without sound scientific data on pollutants and their impact
on environmental stability and human health. Responsibility
for building this data base has been assigned to EPA ’s Office
of Research and Development and its 15 major field installations,
one of which is the Corvallis Environmental Research Laboratory.
The primary mission of the Corvallis laboratory is research
on the effects of environmental pollutants on terrestrial,
freshwater, and marine ecosystems; the behavior, effects and
control of pollutants in lake systems; and the development of
predictive models on the movement of pollutants in the biosphere.
This report describes sampling designs, collection methods,
laboratory techniques, and data analysis procedures for in-
vestigation of the response of marine macrofaunal benthic com-
munities to pollutional stress. It is part of a series of
reports on sampling guidelines for benthic, zooplankton, phyto-
plankton, demersal, and intertidal marine assemblages.
A. F. Bartsch, Director
Corvallis Environmental
Research Laboratory
11 1
-------
ABSTRACT
This report presents guidelines for the quantitative assess-
ment of the effects of marine pollution on benthic comunity
structure and population dynamics. The sampling design addresses
the number and location of stations, survey frequency, sampling
gear, replication of samples, screening and preservation of
biological samples, and the collection of abiotic data. Recom-
mendations are given for the sorting, identification, enumeration,
and weighing of benthic specimens. The section on data analysis
suggests indices for detecting changes in species composition,
density, dispersion, diversity, richness, dominance, and spatial-
ten oral faunal homogeneity.
iv
-------
FOREWORD.
ABSTRACT
LIST OF TABLES
ACKNOWLEDGMENTS
SECTIONS
I. CONCLUSIONS
II. INTRODUCTION
III. SAMPLING
111
LOCATION AND NUMBER OF STATIONS
SAMPLING GEAR
REPLICATION
SURVEY FREQUENCY . .
COLLECTION OF GRAB SAMPLES
SCREENING AND PRESERVATION
ABIOTIC DATA
IV. SAMPLE PROCESSING
SORTING AND SPECIES IDENTIFICATION
ENUMERATION
BIOMASS
OTHER OBSERVATIONS
V. DATA ANALYSIS
PRESENTATION OF DATA . .
POPULATION CHARACTERISTICS
COMMUNITY CHARACTERISTICS
VI. REFERENCES
• . iv
• . vi
• . vii
10
•11
11
11
14
• 14
• 19
25
CONTENT S
1
3
4
4
6
6
7
7
8
8
V
-------
TABLES
No. Page
Recommended Format for Presenting Numeric
Density Data 15
2 Example of the Calculation of the
Dissimilarity Index, Matrix, and
Dendro gram 23
vi
-------
ACKNOWLEDGMENTS
I thank Drs. Donald Baumgartner, Donald Boesch, and
Allan Michael for reviewing drafts of this report.
vii
-------
SECTION I
CONCLUSIONS
Alteration of the structure of benthic macrofaunal communities can
indicate the effects of pollution and natural stress on marine ecosystems.
The benthic survey techniques recommended here are not meant to be
standard methods, but rather guidelines to the kind of investigation
necessary to obtain meaningful results.
Survey designs must realistically reflect the time and resources
required for accurate field and laboratory processing of the samples. A
design which specifies as few as ten stations will necessitate a major
sampling effort. A fixed station grid or transect is recommended for
surveys of ocean disposal sites and point sources of pollutants. The
null hypothesis to be tested assumes no significant differences in
biotic conditions between control and presumably stressed sites. Sub-
stantial environmental heterogeneity would require stratified sampling.
Five 0.1 m 2 Smith-McIntyre grabs should be taken at each station
and cruises should be conducted at least once every three months. The
minimum screen size must not exceed 1 .0 mm. Bottom water temperature,
salinity, and dissolved oxygen, and pollutant concentrations in the
sediments, water and biota should be monitored at each station. A core
from all grabs must be preserved for analysis of sediment size distribution.
All specimens belonging to the major macrofaunal taxa (amphipods,
polychaetes, molluscs, echinoderms, etc.) must be identified to the
species level, enumerated, and weighed. Wet biomass estimates should be
converted to ashfree dry weights. Accurate species identifications are
a vita] part of the benthic survey. They should be conducted by exper-
ienced biologists with the aid of the best available keys and reference
collections prepared by or in consultation with expert taxonomists. It
1
-------
is often appropriate to make special observations on the biology of
selected species, e.g., size distribution, disease frequency, or repro-
ductive success.
No single aspect of population or community structure can serve as
an unequivocal index of biotic response to stress. Possible changes in
species composition, density, dispersion, diversity, richness, dominance,
and spatial-temporal faunal homogeneity should always be examined.
Patterns of these structural parameters should be compared with each
other and with other aspects of the biology of the berithos.
Biotic response to human perturbation is difficult to predict. For
example, diversity and density do not always decrease in altered marine
environments. Also, slight modificatons of benthic community structure
are not necessarily deleterious. Beyond statistical significance,
interpretation of the ecological importance of biotic change must ultimately
lie in the judgment of experienced ecologists.
2
-------
SECTION II
INTRODUCTION
This report presents quantitative methods for assessing the condition
of benthic communities. The species composition, density, and structure
of the macrofaunal benthos are good indicators of the effects of stress
on marine ecosystems. This is especially true for the impact of materials
which accumulate on the bottom and affect the biota directly through
toxic action or indirectly through altered sediment characteristics.
Since dredge spoils,sewage sludges, and some other materials dumped at sea
fall into this category, benthic surveys may often be required to provide
guidance for EPAs Ocean Disposal Permit Program. The methods described
here are also appropriate for research on undisturbed benthic habitats.
There are no standard techniques for benthic sampling and data
analysis. Ecologists use different collecting gears, sieve sizes,
mathematical indices, etc., according to the unique requirements of
individual investigations. The techniques recommended here will generally
provide the data necessary for a comparative, quantitative analysis of
the subtidal benthos of estuarine and coastal waters. Although these
methods are not absolute requirements, alteration of the survey design
should result in at least an equally rigorous investigation. The thorough
review of benthic survey methods edited by Holme and McIntyre (1971)
provides additional information on sampling designs.
3
-------
SECTION III
SAMPL ING
LOCATION AND NUMBER OF STATIONS
The location and number of stations must be determined for each
survey according to:
1. Objectives of the survey
2. Size and configuration of the survey area.
3. Environmental conditions at the site, especially spatial
changes in sediment characteristics, depth, salinity, biotic
assembla9eS, and pollutant concentrations.
4. Physical and chemical characteristics of the material to be
dumped or dredged, and predictions of dispersion patterns.
5. Need for “control” samples taken at comparable habitats which
will not be exposed to the same degree or kind of stress.
6. Human and material resources of the survey.
Sufficient information will seldom be initially available for all of
these factors. A preliminary examination of the biological, chemical,
and physical characteristics of the survey area may prevent wasted
effort. The presurvey need only consist of qualitative sampling or
direct observations of the bottom.
The number of stations is limited by the time required for post-
cruise processing of biological samples. Following the sampling design
4
-------
presented below, a scientist familiar with the species at the site will
require at least two weeks to sort, identify, enumerate, and weigh
specimens taken at a single station. It is desirable to collect mare
samples than may be needed, but stations which are critical to the
survey should be occupied first. It is not necessary to make benthos
collections at every water quality station. Survey designs which specify
more than ten biological stations are very expensive both in field costs
(especially at offshore sites) and laboratory effort.
Locate stations at offshore dumping grounds along at least two
transects crossing one another at the center of the site and extending
beyond its perimeter. One transect should be parallel to the predicted
direction of net motion of dumped materials on the seabed. Stations must
be located at the center, margin, and beyond the margin of the site. The
latter serve as “controls” but a reference station(s) further away may
also be necessary. It probably will not be possible to find a control
habitat which differs from the survey site only in the absence of dredging
or dumping activities. However, the reference stations can serve as
controls in the sense that the benthos may not be affected to the same
extent as within the dump or dredge site.
Location of stations along transects lends itself to a gradient
analysis of the effects of stress. Random location of stations would
permit a description of the biota of the entire site, but the number of
stations necessary for such a design is usually prohibitive.
If substantial natural changes in environmental conditions occur in
the survey area, it is necessary to stratify sampling by placing stations
in each habitat type. For example, if both coarse sand and mud bottoms
are found at a dump site, stations (including controls) must be placed
in areas of each sediment type.
5
-------
Benthic surveys of dredged channels must include two or more transects
perpendicular to the channel. Locate at least three stations along each
transect at the center, margin and outside of the channel. In estuaries
and bays the location of transects should reflect salinity gradients.
Relatively minor changes in depth can be signficant if the survey area
includes a portion of the intertidal zone.
SAMPLING GEAR
A variety of benthic grabs have been used for quantitative sampling
(Holme, 1971). The ubiteh area is essentially constant, but the depth
of penetration varies with sediment characteristics and must be recorded
for each sample. The Smith-McIntyre (1954) grab is recommended for use
in estuaries and on the continential shelf. It is mechanically reliable
and samples a reasonably large area (0.1 m 2 ). Screens on the back of
the jaws can be removed to permit direct observation of the surface of
the sediments. Cores for meiofaunal, chemical, and geological analyses
can be removed from the undisturbed sample. Additional weights can be
added to the Smith-Mcintyre grab to sample sediments that are difficult
to penetrate. Other grabs, especially the van Veen and box corer, can
also be used for quantitative surveys. Anchor dredges such as the one
used by Sanders, Hessler, and Hampson (1965) can provide large benthic
samples from the deep sea.
REPL ICAT ION
The number of replicate grabs per station depends on the community
and population characteristics of interest. Accurate estimates of the
density of a rare species with a patchy distribution might require
hundreds of samples. The design of benthic surveys at dredging and
dumping sites should provide the data necessary for an analysis of the
spatial-temporal distribution of the more cornon species and the biotic
6
-------
assemblage as a whole. A single sample per station will not permit a
statistical evaluations of variations. Mclntyre 1 s (1971) recommendation
of five grabs per station should be followed when possible. The number
of replicates should never be less than three. If necessary reduce the
number of stations so that an adequate number of grabs can be taken at
each.
Stations should be sampling areas (perhaps 0.25 km square) rather
than discrete points and random sampling should be attempted within each
I!stationhl This minimizes the possibility that due to patchiness or
other causes the collections might not be representative of true variations
in biological conditions.
SURVEY FREQUENCY
Because major seasonal changes occur in the structure and function
of benthic assemblages, long-terni monitoring programs should include
surveys at no more than three-month intervals. Quarterly surveys can be
designed so that samples are processed and analyzed before the next
cruise. Baseline studies of biological conditions should continue for
at least one annual cycle. An analysis of annual variations will require
two or more years of sampling.
COLLECTION OF GRAB SAMPLES
A power winch (preferably hydraulic) and a davit, boom, or A-frame
arrangement are essential for lifting a weighted Smith-McIntyre grab
onto the deck. After the grab is placed on a waist-high stand, remove
the screen on one side to see if an adequate sample has been taken.
Reject samples which are very shallow or which lost sediments during
retrieval . Record the maximum depth of the bite and take cores (approxi-
mately 4 cm in diameter) from each sample for sediment size distribution
and chemical analyses. A core may be taken for the study of meiofauna.
7
-------
Record the color, texture and obvious vertical stratification of the
sediments and section the cores if they cannot be preserved intact.
Note the presence of tubes, burrows, and epibenthic organisms. If the
stand is equipped with a large funnel, the remainder of the sample can
easily be dropped into a bucket for storage until the sediments can be
screened. Record the volume of sediments in the bucket as a second
estimate of the size of the sample.
SCREENING AND PRESERVATION
Sieving sediments and removing organisms from the screens are
critical parts of the benthic survey and should be done with great care.
The sieve mesh size must not exceed 1.0 mm. A good sieve can be constructed
by nailing a 40x40 cm screen to the bottom of a 10 cm deep wooden frame
equipped with sturdy handles. The screen should be sealed to the wood
with silicone rubber to prevent animals from crawling into the crevices
at the edges. Filtered seawater can be used to wash the sediments
through the screen. Samples can be washed more efficiently and less
destructively by dipping and shaking the bottom of the sieve in a tub of
seawater. Remove the larger organisms with a pair of forceps and then
wash the remaining sediments (sometimes more than a liter) into one
corner of the screen and spoon them into a separate jar. Fix the organisms
and sediment residue in a 10% formalin-seawater solution and transfer
them after a week to 70% ETOH-5% glycerine for permanent storage. A
label giving appropriate station and sample data written in indelible
ink on high quality rag paper must be placed in each jar. Specimens to
be used for chemical analyses must be frozen if they cannot be analyzed
immediately.
ABIOTIC DATA
The water depth andbottom water temperature, salinity, and dissolved
8
-------
oxygen must be recorded at each station. The grain size distribution of
the sediment samples is determined for sand by sieving through a Wentworth
scale screen series and for the silt-clay fraction by the pipette method
(Buchanan, 1971). Other chemical and physical analyses of the water,
sediments, and biota will usually be necessary to establish correlations
between biotic and abiotic conditions.
9
-------
SECTION IV
SAMPLE PROCESSSING
SORTING AND SPECIES IDENTIFICATION
Specimens should first be sorted to higher phylogenetic levels
(order or above). It may not be practical to identify specimens from all
phylogenetic groups to the species level. For example, meiobenthic
nematodes and copepods are occasionally retained on 1.0 mm screens and
their complete identification may not justify the necessary taxonomic
effort. However, it is essential to identify the species of the most
important taxa, i.e. , those higher groups (amphipods, polychaetes,
pelecypods, gastropods, echinoderms, etc.) which dominate the macrobenthos
in numbers of species and individuals, biomass, or functional significance.
Accurate species identifications are a vital part of a macrobenthos
survey. The paucity of trained taxonomists will force most projects to
develop their own classification capability. This is a major task and
will require weeks, perhaps months before a novice can confidently
identify an amphipod or polychaete to the species level. Throughout the
project the same person should identify all specimens from each phylo-
genetic group.
The best available keys must be acquired before species identifi-
cations are attempted. A reference collection of the major taxa in the
samples should be obtained from or constructed with the aid of expert
taxonomists. The reference specimens must be keyed out to confirm the
classifiers familiarity with the appropriate taxonomic characters.
Comprehensive keys do not exist for all taxa from all areas. If a
clearly distinct species cannot be identified beyond the familial or
generic level, it can simply be given a number, e.g . Nereid 1 or Ampelisca
2. If the species is coimion, it should be sent to a taxonomist for
identification.
10
-------
ENUMERATION
Specimens of all except colonial species should be counted individually.
Record colonial bryozoans, hydroids, etc. as present or absent. Count
fragments only when they clearly represent a single organism. Place
specimens of each species in separate vials labeled with the species
name and sample number. If possible store the collections permanently
and never discard them before final action is taken on a permit application.
BIOMASS
Determine the wet weight after blotting for each species in each
sample to the nearest 0.1 mg. The wet weight should include hard parts
of the body, but not tubes and protective coverings not attached to the
body. Report biomass as ash-free dry weight, estimated from the wet
weight data by appropriate species-specific conversion factors. These
factors can sometimes be found in the literature or can be determined by
the relation of wet weight to the difference between the weight after
drying at 105 C for several hours and the weight after incineration at
500 C. The conversion factors may change seasonally due to growth and
reproductive cycles.
OTHER OBSERVATIONS
Information beyond the biomass, number of individuals and identity
of each species should be recorded when preliminary results indicate a
need for more detailed study.
Biological Tissue Analysis
The concentration of chemical pollutants in biological tissues
should be determined when there is evidence of direct toxic action or
11
-------
bioaccumulation and magnification through the food chain. Special
collecting methods may be necessary because grab samples do not consis-
tently provide the quantity of the same species required for some chemical
analyses. It is best to determine the concentrations in specific organs
of individual specimens, but it may be necessary to pool whole bodies of
all specimens of the same phylogenetic group.
Size Distribution
The length, width or diameter of each specimen of a particular
species can be measured to document differences in size frequency distri-
bution. The precision of measurement should not be less than 1/20 of
the size range of the collection. An ocular micrometer or set of dividers
and rule are the best measurement tools for most benthic organisms.
Reproductive Condition
Reproductive success and seasonal cycles can be described by the
presence of external egg masses, ripe gonads, or other criteria of
reproductive condition. The size, sex, and if possible, stage of sexual
maturity must also be recorded for each specimen.
Disease
The type and incidence of all diseases and abnormalities must be
noted. Affected specimens should be sent to an invertebrate pathologist
for examination.
Mei obenthos
The 1 .0 mm screen was recommended for the study of the macrobenthos
because it will usually retain a sufficient diversity and quantity of
12
-------
organisms to make spatial-temporal comparisons of biological conditions.
It this is not true, a finer screen (0.25 or 0.50 mm) must be used to
collect smaller organisms. The taxonomy of the meiobenthos is less well
known than that of the macrobenthos; most of the benthos passes through
l.Ommscreen and enumeration is consequently more difficult; and the
finer screensretain a much larger fraction of the sediments. Except for
these problems, study of the meiobenthos would be required routinely.
The meiobenthos cores from the grab samples can be screened independently
or, for a larger sample, a fine screen can be stacked below the 1 .0 mm
screen during the initial sieving. The meiobenthos should be processed
in the same manner as the macrobenthos. The data should be analyzed
separately and perhaps later pooled with information on the macrobenthos.
13
-------
SECTION V
DATA ANALYSIS
Macrobenthos analysis can determine if ecologically important
differences exist between the biota of control and dump sites. Causal
relationships can be indicated through correlations between biotic and
abiotic factors. Bio-mathematical indices and statistical tests are
part of the analysis, but the interpretation of benthic data must include
the judgement of experienced ecologists.
Every aspect of the biology of a species can be affected by stress.
The emphasis of this survey design is on species composition, abundance,
and community structure. The scientific literature must be reviewed for
additional information that will permit a more complete understanding of
observed or predicted biotic responses to dredging and dumping activities.
PRESENTATION OF DATA
Numeric and biomass densities should be given in separate tables
for the collections at each station (Table 1). These tables give infor-
mation at the population (single species) and community (multi—species)
levels. Discussion should begin with the characteristics of the ecolo-
gically and economically important populations. Equations for bio-
mathematical indices (species diversity, faunal affinity, etc.) are
given below, and general statistical formulae (standard deviation,
analysis of variance, etc.) can be found in most statistics texts.
POPULATION CHARACTERISTICS
Species Composition
The presence or absence of species is the most basic result of a
14
-------
Table 1 . Recommended format for presenting numeric density data from the five grabs taken at each station. A similar table must also he given
for the hiomass data
95 Percent
Confidence
Limits of Mean
37.5-169.3
42.8-86.4
5.1 -16.1
0-2.7
0-2.5
98.4-262.8
3.1-4.9
0.339-0.439
Species
Grab Number Total Mean Standard
1 2 3 4 5 Deviation
Coefficient
of Dispersion
Pseudunciolao lbiquua
J T0C9 PY
I2cJi i9?C9T’ Pst951 kP
Acanthobaustorious milisi
t ]_t
______ 160 145 33 112 67 517 103.4 53. 1 27.27
82 74 56 73 38 323 64.6 17.5 4.80
11 4 15 0 4 53 10.6 t4 .23
____ 0 3 0 7 0 5 1.0 1.4 1.96
3 0 0 1 1 5 1.0 .2 1.44
Numeric Percent Cumulative
Rank of Percent
Total of Total
Total 256 226 104 197 120 903 180.6 56.2 24.27
Number cf Species 4 4 3 5 4 5 4.0 0.7
Species Diversity (H) 0.367 0.338 0.424 1.392 1.426 0.395 0.389 0.04
Dominanc (Complement of 1.507 (1.403 0.594 1)545 0.579 0.541 0.542 0. 1)5
Simpson Index)
1
57.3
57.3
2
35.8
93.1
3
5.9
99.0
4
0.6
99.6
-------
biotic survey and can be very informative, especially for organisms
known to be stress tolerant or intolerant. The literature on marine bio-
assays and the results of previous benthic investigations can be
helpful in the identification of indicator species. The distribution of
the benthos at existing dump sites is sometimes a consistent pattern of
presence of many species at the periphery and absence of most at the
center of the site.
Abundance
Dominant species are often ubiquitous and differences in their
densities (both biomass and nUmber of individuals) must be examined
quantitatively. The mean density and its standard deviation and confidence
limits (Table 1) can be used to describe spatial-temporal changes in
abundance. The confidence limits of the density of individual species
are usually very large because of patchiness and sampling error. Increasing
the number of replicate samples from 5 to 20 does not greatly improve
the estimates of mean density (McIntyre, 1971).
The Coefficient of Dispersion (variance: mean ratio) indicates
whether the distribution of a species on the bottom is random (CD=l),
clumped (CD>l), or even (CD
-------
Biological Tissue Analysis
Spatial-temporal changes in the concentration of pollutants in
biological tissues should be correlated through regression analysis with
concentrations in the water, sediment, and dumped materials. Some data
is available in the literature on threshold tissue concentrations for
deleterious effects of heavy metals, pesticides, etc. Knowledge of
feeding mechanisms and trophic position is important in analyzing the
accumulation and magnification of pollutants. Through tissue analysis
it may be possible to relate ecological alterations to a specific component
of dumped materials. In turn, this information would permit modification
rather thancessation of dumping practices.
Size Frequency Distribution
Size frequency distribution can document the absence of a particular
size class (due perhaps to the failure of a year class in a stressed
environment) or establish seasonal growth cycles and rates. Histograms
provide convincing evidence of major difference in size distributions
and chi-square tests can assess statistical significance in less obvious
instances.
Reproductive Condition
Chronic exposure of the benthos to sublethal pollutant concentrations
may block reproductive activity. Spatial differences in the frequency
of mature females with egg masses (or other criteria of reproductive ’
success) must be reported. Reproductive seasons can be described by
plotting percent ovigerous females against time. Information on growth
and reproductive cycles is helpful in selecting the best time of the
year for dredging or dumping.
17
-------
Disease
The statistical significances of changes in disease incidence can
be tested by chi-square. Diseased specimens should be examined and
described by an invertebrate pathologist. The size frequency distribution
of afflicted specimens may indicate differences in exposure or suscep-
tibility to stress.
F i she r i e S
The appearance of any life stage of commercial species in macro-
benthic collections must be emphasized. The abundance, spatial distri-
bution, size frequency, local fishery catches, and other pertinent data
should be used to assess the existing or potential significance of the
area as a rìursey or fishing ground. The impact of dredging or dumping
on transient fishery stocks and those which cannot be sampled effectively
by grabs or other benthic gear must be considered.
Other Population Characteristics
Other information from field observations and the literature about
the life history of the niost abundant species should be summarized.
Data on normal environmental requirements (temperature, salinity, sediment
preference, etc.) are important. If they are narrow, will a slight
perturbation be disruptive? Could the species be introduced to a new
habitat via dredge spoils and compete with the endemic fauna? At dump
sites the mode of associatio with the substrate (tube dwelling, burrower,
attached or motile epifauna, etc.) may determine the consequences of
burial by dredge spoils. The iterature may indicate which species
dominate the flow of energy in benthic systems. If nothing is known
about the susceptibility of kley species to stress, appropriate bioassays
should be conducted for the materials to be dumped (Environmental
Protection Agency/Corps of Engineers, 1977).
18
-------
COMMUNITY CHARACTERISTICS
Changes in community structure must be analyzed on the basis of
both number of individuals and biomass. Abundance, species diversUy,
and fauna] homogeneity indices should be calculated for all taxa cor bine
and separately for the major phylogenetic groups (Amphipoda, Polychaeta,
etc.) in each survey.
Abundance
Abundance at multispecies levels should be reported as the nuiher
of individuals and biomass in each sample and in the pooled samples at
each station (Table 1). The density of taxonomically and crophLal1 .’
related groups of species (e.g., haustoriid aiuphipods) may he more
informative than that of all species in the collection. Tear formatiur
and statistical comparisons between stations are the same as given in
section IV for population density.
Species Diversity
Species diversity is a function of the number of species (richnes)
and the distribution of individuals among the species (Lloyd and ;hari
1964). This is a broad ecological concept and no single di\rsityiri .
can be accepted as a unequivocal mathematical definition. The three
indices recommended here are sensitive to changes in differeni spect.
of community structure. Their interpretation as indicators of multi--
species response to stress requires very careful consideration of the
theoretical significance of the indices.
Richness can be expressed as the number of species (S) collected
per unit effort or area. Obviously this is not an estimate of the total
number of species in the community and it is valid only for comparative
19
-------
study. Constant sampling effort can usually be incorporated into survey
designs and S’s for different samples can be directly compared. If
effort varies, Hurlbert’s (1971) equation can be used to estimate the
number of species that would have been present if effort had been constant
(SCE):
so
SCE=SO —
NcE NO(EcE/EO)
where: SCE and NCE are the number of species and individuals, respec-
tively, for constant effort.
S 0 and N 0 are the number of species and individuals, respectively,
in the original sample.
n 1 is the number of individuals of the .th species.
ECE and E 0 are the constant and original sampling efforts.
For all samples, ECE E 0 .
The most important feature of the pattern of distribution of indi-
viduals among species that determines “effective” diversity is the
extent to which the assemblage is dominated by the abundant species.
Fewer species will appear per unit number of individuals in samples from
corriliunities in which dominance concentration is relatively high. Dominance
concentration can be determined by Simpson’s (.1949) index (S.I.). S.I.
is not greatly dependent on sample size and it is not necessary to
adjust to constant effort before calculating the index.
20
-------
0 n.(n.-l)
S.I. = N(N-l)
For index values to be postively related to diversity they can be
expressed as the complement of S.I. (McIntosh, 1964).
To many ecologists both richness and the species frequency distribution
are integrated in a single concept of diversity. The Shannon-Weaver
information—theoretical measure of mean species diversity per individual
(H’) is sensitive to both components of diversity and is a very popular
index of “overall” diversity (Pielou, 1970).
SO
H’ 1/n (N 0 logN 0 - ri log n )
i=l
The tables of Lloyd, Zar, and Karr (1968) facilitate the calculation
of this index if a computer program is not available.
Diversity patterns have sometimes been misrepresented as comprehen-
sive indicators of the “health” of aquatic ecosystems. Obviously no
single index value could summarize all aspects of community structure.
A basic limitation to diversity indices is that they are not sensitive
to changes in species composition. However, it would be equally erroneous
to dismiss species diversity as having no application in pollution
biology. Such fundamental ecological concepts as dominance and the
number of niches (richness) are certainly pertinent.
Faunal Homoyeneity
A simple percent dissimilarity index can document spatial-temporal
faunal homogeneity between stations. In Table 1 the Percent of Total
column gives the percent each species contributes to the total number of
individuals or biomass collected in all samples at each station [ lOO(n 1 /N)1.
A percent dissimilarity index is calculated for all pairs of stations as
21
-------
the complement of the sum of the minimum value of n 1 /N for all species
common to the two stations. Index value range from 1.0 when the collec-
tions have no species in common, to 0.0 when all species are common and
the distribution of individuals is the same on a percentage basis.
An example of the calculation of the dissimilarity index, matrix,
and dendrogram is given in Table 2. The dissimilarity index values for
all possible pairs of the five hypothetical collections (A-E) are calculated
and presented in the original dissimilarity matrix. The collection pair
with the lowest dissimilarity index (B and 0) is then combined and a
second matrix constructed treatingBD as a single collection. The
dissimilarity between BO and other collections, e.g., A, is the mean
dissimilarity in the original matrix between the collections in the
combined group and A, i.e. , B-A .80, D-A = .91, BD-A = (.80 + .91)/2 =
.86. This procedure is repeated until all collections are combined in a
single group. The results are presented in a dendogram which shows the
hierarchial relationships between collection groups (Table 2). In this
example, two collection clusters with high intra-group faunal homogeneity
are evident (B-D and A-C-E). Identification of clusters in dendrograms
is subjective and may not always be as straightforward as in Table 2.
If all collections are made in areas of similar biotic conditions,
attempts to discriminate more than one cluster may be misleading. If
collections are made along a continuous gradient, well-defined clusters
may not be apparent. However, this site clustering technique may often be
useful in describing quantitative differences in faunal homogeneity
between natural and disrupted benthic assemblages. Thorough reviews of
nun rical classification have been made by Clifford and Stephenson (1975)
and Boesch (1977).
Other Community Characteristics
The degree of similarity in the distribution of species pairs can
be analyzed by the same index given above for comparing station pairs.
Clustering techniques can then be used to identify groups of species
22
-------
Table 2. Example of the calculation of the dissimilarity index, matrix,
and dendrogram.
E
Collection
A
n 1 ni/N n. n ./N n n./N n. n./N n. n./N
1 110 .69 10 .11 46 .51 5 .04
2 42 .26 4 .04 30 .33 0 .00
3 0 .00 30 .33 5 .05 57 .50
4 8 .05 48 .52 10 .11 52 .46
Total 160 LOU 92 1.00 91 1.00 114 1.00
Dissimilarity Index (A-B) = l-(.ll +.04 +.00 +.05) = 0.80
Index values for all other collection pairs are given in the
similarity matrix below.
I. Original
A B C D
.80
.18
.69
.91
.17
.80
.35
.75
.23
III. Group AC Formed
AC BD
BD .80
E .29 .80
a)
- ,
C
4- )
-V.-
0.5
1
. 7 -
E
. 7-
C l )
U,
O.O
1.0 —
Dissimilarity Matrices
.86
Dendrogram
II. BD Combined
A BD
60
80
14
2
156
.38
.51
.09
.01
.99
original dis-
C
BD
C
E
IV. Group ACE Formed
ACE
BD .80
V. Group ABCDE Formed
r—1 I
1
1
1
D A
C
E
B
C
0
Species
B
C
D
E
.86
.18
.35
.75
.80 .23
23
-------
which have similar distribution patterns. The appearance or disappearance
of species assemblages in response to ecological alterations could be
investigated through such techniques. Boesch (1973) gave a good example
of the application of this type of classifactory analysis to macrobenthic
communities.
If sufficient information exists on the feeding mechanisms, trophic
levels or habitat preferences of individual species, it would be important
to describe any changes that occur in the dominant patterns at the
community level. Species diversity and other indices might be calculated
for the infauna, detritus feeders, or some other assemblage defined by
ecological similarity and interaction.
Abnormal seasonal fluctations and even longer-term trends towards
deterioration can be detected through community structure analysis
(Warriner and Brehmer, 1966; McErlean etal., 1973). Data on temporal
changes are especially important in baseline surveys because of the need
to discriminate between natural temporal variations and those caused by
chronic exposure to stress.
24
-------
SECTION VI
REFERENCES
Boesch, D. F. 1973. Classification and community structure of macro-
benthos in the Hampton Roads area, Virginia. Mar. Biol . 21: 226-244.
Boesch, D. F. 1977. Application of numerical classification in ecological
investigations of water pollution. U.S. Envir. Prot. Ag. Ecol. Res.
Ser. EPA-600/3--77-033. 115 p.
Buchanan, J. B. 1971. Sediments. In: International Biological Program
Handbook No. 16. Blackwell Scientific Pubi., Oxford. pp. 30-52.
Clifford, H. T. and W. Stephenson. 1975. An Introduction to Numerical
Classification. Academic Press, New York. 229 p.
Environmental Protection Agency/Corps of Engineers Tech. Comm. on Criteria
for Dredged and Fill Material. 1977. Ecological evaluation of pro-
posed discharge of dredged materials into ocean waters. Implementation
Manual for Sec. 103 o— P.L. 92—532. U.S. Army Engineer Waterways
Experiment Station, Vicksburg, Mississippi.
Greig-Smith, P. 1964. Quantitative Plant Ecology. Buttersworth,
London. 256 p.
Holme, N. A. 1971. Macrofaunal sampling. In: International Biological
Program Handbook No. 16. Blackwell Scientific Publ., Oxford.
pp. 80-1 30.
Holme, N. A. and A. D. McIntyre (eds.). 1971. Methods for Study of Marine
Benthos. International Biological Program Handbook No. 16. Blackwell
Scientific Pubi ., Oxford. 334 p.
Huribert, S. H. 1971. The nonconcept of species diversity: A critique
and alternate parameters. Ecology 52: 577-586.
Lloyd, M. and R. 3. Ghelardi. 1964. A table for calculating the equit-
ability component of species diversity. 3. Anim. Ecol. 33: 217-225.
Lloyd M., 3. H. Zar, and 3. R. Karr. 1968. On the calculation of
information - theoretical measures of diversity. Amer. Midl. Natur.
79: 257—272.
25
-------
McIntosh, R. P. 1967. An index of diversity and the relation of certain
concepts to diversity. Ecology 48:392-404.
McIntyre, A. D. 1971. Introduction: Design of sampling programmes.
International Biological Program Handbook No. 16:1-il.
Pielou, E. C. 1970. An introduction to mathematical ecology. Wiley-
Interscience New York. 286 p.
Sanders, H. L., R. R. Hessler, and G. R. Hanipson. 1965. An introduction
to the study of deep—sea benthic faunal assemblages along the Gay
Head-Bermuda transect. Deep-Sea Res. 12:845-867.
Smith, W. and A. D. McIntyre. 1954. A spring-loaded bottom sampler.
J. mar. biol. Ass. U.K. 33:257-264.
Simpson, E. H. 1949. Measurement of diversity. Nature 163:688.
Warririer, J. E. and M. L. Brehmer. 1966. The effects of thermal effluents
on marine organisms. Air Wat. Poll. mt. J. 10:277-289.
26
-------
TECHNICAL REPORT DATA
(Please read Instructions on the rel’erse before completing)
2. 3. RECIPIENT’S ACCESSION NO,
and Analyzing the Marine
5 REPORT DATE
8. PERFORMING ORGANtZAT cN REPORT N V.
NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
Agency 1BA6fl8
Research Laboratory 11.Cb T t iACT .GRANTNO.
Ecology Branch
Newport, Oregon 97365
AND ADDRESS 13. TYPE OF REPORT AND PERLOD COVERED
Laboratory-Corvallis inhouse
Development 14. SPONSORING AGENCY CODE
Arjencv
EPA/600/02
guidelines for the quantitative assessment of the effects
on benthic community structure and population dynamics. The
addresses the number and location of stations, survey frequency,
replication of samples, screening and preservation of biological
collection of abiotic data. Recommendations are given for the
enumeration, and weighing of benthic specimens. The
analysis sugg—sts indices for detecting changes in species com-
dispersion, diversity, richness, dominance, and spatial-
homogeneity.
KEY WORDS AND DOCUMENT ANALYSIS
bJDENTIFIERSIOPEN ENDED TERMS C. COSATI icld.Uroup
Field B
Grou
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
34
20. SECURITY CLASS (This page)
Uncl assified
22. PRICE
EPA Form 2220—1 (Rev. 4—77) PREVIOUS EDITION IS OBSOLETE
27
* U. S. GOV5RNM NT PSINTING OPrICS 1978—7cb .516/122 SEGLON 10
------- |