ORNL
v>EPA
Oak Ridge
National
Laboratory
United States
Environmental Protection
Agency
Research and Development
Operated by
Union Carbide Corporation
for the Department of Energy
Oak Ridge, Tennessee 37830
Office of Environmental
Processes and Effects Research
Washington, DC 20460
ORNL/EIS-160
EPA-600/9-80-019
April 1980
Interlaboratory
Evaluation of
Microcosm Research
Proceedings of
the Workshop rr
ode
Library
12CH3 POP ;.-iylvania Av -iue, NW
W'as,:?nn'-on DC '^04oO
Ly-r^
Ii
U.S. ElTvT
KDI31.-3),
-------
RESEARCH REPORTING SERIES
Research reports of the Off ice 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 MISCELLANEOUS REPORTS series.
This document is available to the public through theNationalTechnical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/9-80-019
ORNL/EIS-160
April 1980
INTERLABORATORY EVALUATION OF MICROCOSM RESEARCH:
PROCEEDINGS OF THE WORKSHOP
Athens, Georgia
September 18-19, 1979
Edited by
Rizwanul Haque, Workshop Chairman
Office of Environmental Processes and Effects Research
Office of Research and Development
Washington, D.C. 20460
J. Vincent Nabholz, Project Officer
Environmental Review Division
Office of Pesticides and Toxic Substances
Washington, D.C. 20460
Michael G. Ryon, Compiler
Health and Environmental Studies Program
Information Center Complex/Information Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
Date Published: April 1980
OAK RIDGE NATIONAL LABORATORY
OAK RIDGE, TENNESSEE 37830
operated by
UNION CARBIDE CORPORATION
for the
DEPARTMENT OF ENERGY
U. S.
EDI50S, fi.J. 0§817
-------
DISCLAIMER
This report has been reviewed by the Office of Environmental
Processes and Effects Research, U.S. Environmental Protection Agency,
Washington, B.C., 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 recommendation
for use.
ii
-------
CONTENTS
Foreword v
Abstract vii
1. Introduction 1
2. Microcosm Design 1
3. Regulatory Opportunities 5
3.1 Regulatory Needs 5
3.2 Potential Regulatory Applications 6
4. Research Recommendations 7
4.1 Priority Research 8
4.1.1 Protocol Development 9
4.1.2 Interlaboratory Testing 10
4.1.3 Protocol Validation 10
4.2 Long-Term Research 12
4.2.1 Data Bases for Comparative Ecosystem Toxicology . . 12
4.2.2 Test Methods for Interspecific Interactions .... 14
4.2.3 Identification of Holistic Ecosystem Attributes . . 14
4.2.4 Development of Criteria for Test System
Performance 15
4.2.5 Stochasticity and Non—Steady-State Dynamics .... 16
5. Integration of OEPER Microcosm Program 17
5.1 Introduction 17
5.2 Committee Structure 18
5.3 Committee Functions 19
5.4 Committee Operations 20
Appendix 21
iii
-------
-------
FOREWORD
Microcosm research is one of the major components of the OEPER
research program and is being carried out in six of the seven laboratories
of this office. It is supported through various appropriations including
TSCA and FIFRA. Microcosm research is also being actively pursued in many
non-EPA laboratories. Although microcosms represent a potentially power-
ful tool for predicting transport, fate, and effects of toxic chemicals in
the environment, this research tends to be somewhat controversial. There
are disagreements among our own laboratories as to the use, limitations,
and cost effectiveness of microcosms in the regulation of chemicals.
I believe that the six OEPER laboratories actively involved in micro-
cosm research represent a unique capability for developing this tool for
predicting the fate and effects of chemical substances in aquatic and
terrestrial ecosystems and groundwater systems. However, in view of the
complexities involved and huge resource requirements, it is desirable
that the microcosm research in OEPER laboratories be efficiently coordi-
nated and integrated.
This report will aid OEPER in evaluating and integrating microcosm
research in the OEPER environmental research laboratories and stimulate
communication between this office, active researchers, and the user pro-
gram offices.
Allan Hirsch
Deputy Assistant Administrator
Office of Environmental Processes
and Effects Research
-------
-------
ABSTRACT
This workshop, held at Athens, Georgia, September 18-19, 1979, was
to aid the Office of Environmental Processes and Effects Research (DEFER)
in evaluating and integrating microcosm research in the OEPER environ-
mental research laboratories. Participants discussed the design, advan-
tages, and limitations of microcosms; the value and potential role of
microcosms in the regulatory decision-making process; the identification
of priority and long-term research needs to refine microcosm methodology
as a data source for hazard and risk assessments; and the creation of a
microcosm technology committee to assist OEPER in coordinating and inte-
grating microcosm research. These proceedings consist of recommendations
made by special task groups on each of the major topics discussed.
vii
-------
-------
1. INTRODUCTION
On September 18-19, 1979, a workshop was held at the Sheraton Inn
and Conference Center In Athens, Georgia, to discuss the current research
in microcosm technology and its potential- value as a tool in regulatory
assessments of chemical fate and effects. The workshop was sponsored by
the Office of Research and Development (ORD) and the Office of Environ-
mental Processes and Effects Research (DEFER) of the U.S. Environmental
Protection Agency (EPA). Specific topics discussed at the workshop in-
cluded: the design and advantages of microcosms; the value and poten-
tial role of microcosms in the regulatory decision-making process; the
identification of priority and long-term research needs to refine micro-
cosm technology as a data source for hazard and risk assessments; and
the creation of a microcosm technology committee to assist ORD and OEPER
in coordinating and planning microcosm research. The workshop partici-
pants represented all major areas of microcosm research, including ter-
restrial, aquatic, marine, and air. Most of the U.S. research centers
working with microcosms were represented. A complete list of participants
and their affiliations is given in the Appendix. This report is a summary
of the findings and recommendations of the workshop.
2. MICROCOSM DESIGN
The design of microcosms is dependent upon their intended uses.
Interest in microcosms generally revolves around their potential to pre-
dict the behavior of selected biological and physical processes In natural
ecological systems. Scalar reduction of natural ecosystems or portions
of ecosystems enables investigators to include many kinds of interacting
organisms and/or processes at costs consistent with the size of the
microcosm. A quasi-quantitative relationship between microcosm size,
complexity, and time is presented in Figure 1. In general, it is assumed
that as the size of the microcosm increases, the ability to appropriately
include larger fauna and flora also increases, as well as the ability to
measure not only their functional but also their numerical responses.
However, increases in the dimensions of experimental systems are followed
-------
ORNL-DWG 80-9797
10-20 yr
1-2 yr
MONTHS
WEEKS
HOURS
TOTAL ECOSYSTEMS WITH LARGE VERTEBRATE POPULATIONS
SMALL ECOSYSTEMS WITH SMALL VERTEBRATE POPULATIONS OR
COHORTS OF LARGE VERTEBRATES- EXPERIMENTAL ECOLOGY
(EXPERIMENTAL WATERSHEDS)
LARGE PLANTS, MACROINVERTE8RATE PLANTS, INVERTEBRATE
GRAZERS, AND INVERTEBRATE PREDATORS
MEDIUM-SIZED PLANTS, GRAZER POPULATIONS,
MORE VERTEBRATE INVOLVEMENT
MACROSCOPIC PLANTS, INVERTEBRATE
GRAZERS, INVERTEBRATE PREDATORS
MICROSCOPIC PLANTS AND
ANIMALS (PRIMARY
PRODUCTIVITY. GRAZING)
BACTERIAL
POPULATION
(DEGRADA-
TION)
10-1
TEST FLASK
TUBE
10°
AQUARIA,
TERRARIA
101
LABORATORY
CHAMBERS
10Z
103
GREENHOUSES,
LARGE BAGS,
FIELD PLOTS,
PONDS
3H04
LAKES,
FORESTS,
FIORDS
DISTANCE (CROSS SECTION, METERS)
Fig. 1. Relationships of time, scale, and biological complexity
for laboratory and field microcosms and natural ecosystems.
-------
by concomitant increases in cost. For example, a 250-ml glass container
costs approximately $2.50 while a large plastic sphere capable of holding
1.9 x 10* ml costs $250,000.
Small-scale microcosms are convenient not only from the perspective
of representing complex biological assemblages but also because replica-
tion is possible. Appropriate control systems can be compared using
standard statistical techniques to other sets of treatment microcosms
insulted by various perturbants. In addition, small-scale but complex
microcosms are inexpensive to acquire and maintain. Another implicit
advantage of microcosms as opposed to simulation models is the lack of
any assumptions regarding a quantitative knowledge of interactions and
relationships between the various physical and biological elements.
As a result of the above characteristics, microcosms can be used as
a cost effective and viable tool in the management of large-scale ecologi-
cal systems which are or have the potential to be perturbed by various
anthropogenic substances or processes. However, there are a number of
assumptions associated with extrapolating results derived from microcosms
to the natural ecological system or processes being simulated. Specifi-
cally, it is assumed that the biological events observed in microcosms
are either equal to or closely related to those expressed in the natural
system. Rarely has this assumption been tested. As a result, develop-
ment of various management strategies or regulations for natural systems
based upon data derived from microcosms will always be subject to ques-
tion until such assumptions are proved.
To summarize, microcosms do possess some characteristics which
appear from a descriptive, qualitative point of view to be similar to
those occurring in natural systems. Applications of perturbants to such
systems and the ensuing results may be correlated, within the time frame
of the experiments and geographic location, to those realized within
natural systems which have been similarly perturbed. However, the
researcher has the responsibility of demonstrating that such a correla-
tion can, in fact, be made. Finally, it was recognized that certain
fundamental characteristics of microcosms such as scale, complexity, and
replicability are directly related to fabrication, operation, and main-
tenance costs.
-------
The application of microcosms to the regulatory decision-making
process is limited by their design and attainment of certain characteris-
tics which allow them to produce suitable results. Some of the more
important of these qualities include:
1. Microcosms used either for screening or for detailed simulation of
a particular ecosystem must exhibit demonstrable and measurable
ecosystem-level and community-level processes. This is necessary
in order for them to be used with any assurance in predictions of
ecosystem-level partitioning and biotransformation of chemicals and
their bioactive residues, or in predictions of effects on other
ecosystem-level attributes such as nutrient cycling, community pro-
duction and respiration, predator-prey interactions, and shifts in
species dominance or successional patterns.
2. The results of microcosm tests must be reproducible. Industry must
be assured impartial consideration of each chemical tested. Regula-
tory agencies must be able to compare results of one test with
another, much as the "benchmark" physical and chemical studies are
used in pesticide registration.
3. Microcosms must be realistic. However, the degree of realism, or
the ability of microcosm results to be extrapolated to the environ-
ment, varies depending on whether the microcosm is 'used for screen-
ing or for detailed simulation of a particular ecosystem. Screening
tools need only display ecosystem-level interactions and need not
necessarily have a physical world analog, since additional testing
will follow if adverse effects are observed. Consequently, screen-
ing tools must be sensitive enough to perturbations that they will
trigger additional testing, when appropriate, while still being
robust enough that controls will be reproducible.
A microcosm which is designed to simulate a specific ecosystem and used
to quantatively estimate the effects of potentially harmful chemicals
should have real world analogs. This is required to facilitate system
validation and extrapolation of results in detailed environmental
assessments.
-------
3. REGULATORY OPPORTUNITIES
3.1 REGULATORY NEEDS
There are currently immediate needs for microcosm protocols that can
couple a particular ecosystem with concentration and worst-case-situation
data to assist regulatory agencies in assessing the environmental fate
and effects of chemicals. Environmental assessments of chemicals are
presently being conducted under mandates of the Toxic Substances Control
Act, the Federal Insecticide, Fungicide, and Rodenticide Act, the Federal
Water Pollution Control Act, and the National Environmental Protection
Act. There are currently several uses for microcosms in filling regula-
tory needs:
1. Microcosms are needed as screening tools to aid industry and govern-
ment in the identification of those compartments of the environment
with the greatest risk from the marketing and use of chemicals,
thereby directing further research efforts —i.e., a "go — no go"
decision in the development of a promising chemical and the selec-
tion and initiation of appropriate environmental studies.
2. Microcosms may be used to validate environmental assessment decision-
making schemes that do not presently require the submission of micro-
cosm data for every chemical. The utility of microcosm data could be
measured against predictions by mathematical models of environmental
fate and effects, which depend largely on physical and chemical param-
eters and structure-activity relationships. Thus, the flexibility of
the decision maker to substitute one type of data for another, to re-
quire certain data, and to evaluate the relative degree of certainty
for decisions relying on each data base could be determined. It is
also probable that microcosm data could serve to fine tune mathemat-
ical models to assure better agreement among predictions from differ-
ent data bases at the initial environmental assessment level.
3. Microcosms may be used to provide quality control and to supplement
data submitted by industry. Again in this case screening microcosms
are viewed as the most feasible technique since they can evaluate
the larger number of chemicals requiring additional environmental
data with lower associated manpower, time, and cost requirements.
-------
4. Larger microcosms that simulate specific ecosystems, such as aquatic
models (Perez-type at Narragansett) and terrestrial models (Corvallis-
type), will be useful for later detailed environmental assessments of
those chemicals identified as having potentially adverse environmental
properties but which have equally great social and economic benefits.
These larger microcosm systems, the results of which are more easily
extrapolated to physical environmental analogs, may be used to devise
mitigation procedures or to serve as the backbone either for regula-
tory controls on the use of the chemicals or for their withdrawal from
the marketplace.
3.2 POTENTIAL REGULATORY APPLICATIONS
Most environmental regulatory actions require in some form a deter-
mination of risk or impact. Tools available for making such determina-
tions include laboratory testing, mathematical modeling, field monitoring,
and knowledge derived from previous case experience. In the case of EPA's
chemical registration and premanufacture notification programs (Federal
Insecticide, Fungicide, and Rodenticide Act and Toxic Substances Control
Act, respectively) field data are generally unavailable, so the burden
must be borne by laboratory tests (physical analog models) and predictive
(mathematical) models. Most available laboratory test methods today are
highly simplistic in purpose and utility. This is advantageous in that
it maximizes the feasibility and unequivocal characterization of physical
and biological components and attributes of ecosystems, but it is dis-
advantageous in that it ignores or presumes adequate understanding of
the interactions between these components and attributes.
Microcosms (multicompartmental laboratory test systems) are needed
in regulatory programs as part of the array of available testing proto-
cols and as support for the development and validation of mathematical
models. The range of possible uses for microcosms in a regulatory pro-
gram context is extremely broad for two reasons. First, presently avail-
able technology for chemical testing and assessment demonstrates many gaps
in scientific logic and empirical tools. Second, the state of microcosm
technology is still embryonic and therefore holds promise of yet undeter-
mined benefits.
-------
The potential for such benefits arises largely from the fact that
multicompartmental laboratory test systems have attributes that simply
do not exist in single compartment systems. For example, the opportunity
to observe chemical behavior at the water-sediment interface or in inter-
stitial waters is provided only in a multicompartmental system. Similar-
ly, the effects of chemicals on phenomena such as competition or other
interspecific interactions, and therefore on community succession and
resultant community structure, are unobservable in presently available
single-species test systems.
It would be a mistake to judge the regulatory usefulness of micro-
cosms on the basis of presently available systems, which are largely not
yet ready for application and have been developed to address only some
of the conceivable objectives of such systems. It would also be a mis-
take to set unattainable criteria for applying microcosms in a regulatory
program, such as, for example, requiring quantitative proof that micro-
cosms produce data that fully reflect the behavior of natural systems
through mechanisms that are totally understood; no other test systems
available today meet such criteria.
4. RESEARCH RECOMMENDATIONS
Recommended research was separated into two sections: priority
research and long-term research. The priority research section was
initially characterized as short-term objectives of an integrated micro-
cosm program, but it was realized that some high priority items could
not be accomplished in the short term. The distinction between short-
term and long-term research is somewhat artificial since all the recom-
mendations taken together actually form a continuum, but such a separa-
tion is necessary to identify the research most needed to assist the
regulatory agencies in assessing the environmental fate and effects of
chemicals. All suggested research was considered contributory to the
advance of the state of the art of microcosm technology and the regula-
tory decision-making process.
Current priorities in regulatory program needs regarding the develop-
ment of microcosms are perceived as follows:
-------
1. Screening microcosms that, along with other tests and information
(e.g., exposure and structure-activity considerations), can help
determine the need and direction of further testing of the toxic
hazard and/or environmental fate of a chemical.
2. Microcosms designed to simulate a specific ecosystem or subsystem
that are considered as advanced predictive and confirmative tests.
These systems can be undertaken when more detailed characterization
of the fate and/or ecological effects of a chemical is considered to
be warranted.
3. Microcosms or multicompartmental systems that can be used in a
research context as tools to improve understanding of the ecological
mechanisms that mediate chemical impact. These systems may or may
not be used to routinely test chemicals for their environmental fate
and effects.
All of these microcosms can be experimentally manipulated to aid in
the development or validation of mathematical models by identifying key
environmental variables and the sensitivity of model behavior to them.
4.1 PRIORITY RESEARCH
Microcosms provide unique opportunities to pursue ecological research
on the environmental fate and effects of chemical substances. However,
standardized uses of microcosms for routinely testing chemicals for the
regulatory decision-making process are not available at this time. There-
fore, current research should emphasize the refinement of microcosm meth-
odology so that it may be used as a standard technique in the regulatory
decision-making process. This refinement involves three related tasks:
(1) development of streamlined testing protocols, (2) interlaboratory
testing of protocols, and (3) protocol validation. Protocol development
must precede interlaboratory testing, while validation could occur con-
currently with protocol development and interlaboratory testing.
-------
4.1.1 Protocol Development
Testing protocols using microcosms are needed for the implementation
of the Toxic Substances Control Act. Testing protocols will be the
foundation of test standards, which will become the regulations required
under Section 4(a) of the Toxic Substances Control Act. They will have
the force of the law and therefore will be subject to vigorous require-
ments for format, content, and language. Thus, testing protocols are
essential to the development of consistent, defensible standards. Not
only must a protocol include the conditions and procedures which need to
be followed exactly for successful and replicable test performance, but
a protocol must also state scientific importance, scope, and limitations
of the test.
Microcosm protocols are needed for a wide spectrum of ecosystems
and subsystems, especially those systems with the most social relevance.
There is a particular need for screening microcosms. These microcosms
may contain soil or sediment, water, microbes, invertebrates, and/or
plants. They may be synthetic systems of artifically assembled compo-
nents not designed to simulate a specific ecosystem, although they could
consist of cores excised from a particular system. An early qualitative
determination of a chemical's probable fate and effects could be made.
Screening microcosms appear to be of greatest value immediately follow-
ing analyses with fate models. In this position, such microcosms could
confirm predictions of the fate model, and perhaps become a substitute
for several single-component tests. Most important, screening microcosms
could reduce both the time and cost of subsequent testing by limiting and
directing the type and number of followup tests. Thus, screening micro-
cosms could expedite the risk assessment process.
Prior to protocol documentation, microcosms need to be evaluated
for their potential use in testing protocols. Hopefully, this will per-
mit protocols to be developed more rapidly. A set of criteria to aid in
this evaluation is suggested, stated in the form of a series of questions.
These include, but should not be limited to, the following:
1. What specific hypothesis(es) is the microcosm being developed to test?
2. What specific attribute(s) will be used to measure the fate and envi-
ronmental effects of a chemical substance?
-------
3. How amenable is the microcosm to incorporation into a laboratory
testing protocol that could be widely used by various laboratories?
Has the microcosm been optimally scaled down for incorporation into
a laboratory test system? Has the microcosm retained a significant
amount of its ecological realism?
4. How representative are the system components compared to other pos-
sible selections of components?
5. Are the system components readily available? Can they be easily
cultured?
6. How well does the test system conform to the criteria for test sys-
tems outlined by the Toxic Substances Control Act?
7. What is the basis for extrapolation of results from the system to
the environment?
The order of presenting these questions is not important. The goal of
this process should be to produce a definitive description of and a
rationale for the test procedure.
4.1.2 Interlaboratory Testing
Interlaboratory testing, i.e., round robin testing, must follow the
development of all protocols. Properly designed round robin tests permit
the accuracy and precision of a testing procedure to be determined. Know-
ing the reproducibility of results for a promulgated test standard gives
the regulatory agencies more confidence in using test results in the risk
assessment and decision-making processes. Interlaboratory testing can
help identify problems not observed during protocol development and allows
improvements to be made in the procedures. In addition, the cost (or
range of costs) of a protocol can be determined. This will allow regula-
tory agencies to evaluate the cost effectiveness of their testing
procedures.
4.1.3 Protocol Validation
Microcosms are currently used as research models of natural systems,
as aids in determining chemical fate and effects in the environment, and
as methods to validate the applicability of mathematical models to natural
10
-------
systems. They have been recommended for testing chemicals routinely for
the regulatory decision-making process. However, if these systems are
to be used to their fullest capacity, their validity as predictive tools,
both separately and together, needs to be evaluated. Therefore, priority
research to test the validity of these systems is recommended.
A systematic research effort should be initiated to evaluate the
veracity with which microcosms exhibit those effects observed in field
situations. An effort should be made to identify those perturbed ecosys-
tems for which a sufficient amount of background information is available
to permit characterization of the changes produced by specific pollutants
or environmental factors. Studies should be conducted to determine the
extent to which microcosms of varying levels of complexity exhibit the
.effects observed in the field when subjected to the same perturbing
factors. Accordingly, emphasis should be placed on the following research
areas:
1. Scale considerations: As research questions are formulated for which
a microcosm approach is necessitated, the design of the microcosm
will be greatly facilitated if background information on scale is
available. The conditions should be established, including the size,
that will permit the persistence of a particular trophic level, tro-
phic interaction, or system component in the microcosm. For example,
under most appropriate environmental conditions fish populations can-
not be maintained in closed microcosms on natural, system-derived
food unless the microcosm size approaches that available in large
outdoor models. Research is needed to develop information on the
effect of scale, particularly as it applies to the design of micro-
cosm systems for both fate and toxicity studies.
2. Functional and structural parameters: Research is needed to develop
and assess structural and functional parameters for comparing a micro-
cosm to similar real world environments and to other microcosms.
Functional parameters such as dissolved oxygen patterns, nutrient
cycling, production-to-respiration ratios, biomass, and heterotrophic
activity should be evaluated for their potential as indicators for
comparing the activities of a microcosm to activities found in various
natural settings. Structural parameters such as species composition,
11
-------
trophic structure, and population dynamics should be investigated in
a similar manner.
4.2 LONG-TERM RESEARCH
Many of the research initiatives believed to be needed are currently
under way but still in early stages of development. Ongoing work in
microcosm research has shown that various techniques are sensitive means
of predicting the environmental fate of chemicals, but only limited prog-
ress has been made in applying existing techniques to the prediction of
chemical effects. The use of relatively complex microcosms as toxicolog-
ical tools is conceptually appealing because of the inherent integration
of direct and indirect effects of toxic chemicals.
Preliminary efforts by various groups seem to fall into three cate-
gories or philosophies of microcosm development: (1) artificial assem-
blages of biota that represent functional groups; (2) natural biotic
communities that are studied in situ; and (3) natural communities that
are removed to the laboratory. As yet, these approaches have not been
tested sufficiently to judge their respective potentials as toxicological
tools.
Many kinds of information needed as a basis for further research
initiatives will be obtained from this ongoing work during the next sev-
eral years. However, there is a continuing need for long-term attention
to the items addressed in this section.
4.2.1 Data Bases for Comparative Ecosystem Toxicology
Different kinds of ecosystems, like different kinds of organisms,
will vary in their sensitivity to a particular stress. To relate the
dose-response curve for one test species to a broader range of organisms
requires a knowledge of the relative sensitivities of organisms. Like-
wise, the response of one ecosystem (natural or laboratory) will be of
general predictive value only if we know the relative sensitivites of a
broad range of ecosystems. There are a number of substances (e.g., cer-
tain trace elements, certain pesticides, and No. 2 fuel oil) for which
12
-------
I
dose-response observations have been made in a variety of natural eco-
systems. Such data should be compiled for several substances to provide
frequency distribution curves of ecosystem sensitivity, against which
the sensitivity of particular microcosm tests could be compared. The
construction of such data bases represents an empirical approach to
"calibrating" microcosm results for general predictions of safe exposure
levels in nature.
An additional area of potential need is the identification and analy-
sis of long-term monitoring data for ecological parameters. These data
bases would aid in identifying historical trends and in determining the
amount of natural variability in ecological attributes. A background
would be provided against which more informed risk assessments can be
made using microcosm results.
The development of these data bases will not only be useful from a
comparative viewpoint, but will also aid in the validation process. These
data bases can be constructed by using information from the literature and
from specifically designed experiments. For example, the data bases could
include information generated by the following comparisons:
1. Relate data from uncontrolled perturbations of natural ecosystems
(e.g., Kepone in the James River) to microcosms of varying scale and
complexity designed to simulate either a process or subsection of the
perturbed natural system. The investigator would attempt to establish
qualitative and quantitative relationships between the two systems for
long and short time intervals.
2. Perform controlled perturbation experiments in the field where pos-
sible and in laboratory microcosms of varying scale and complexity
to establish relationships of state variable dynamics between both
systems.
3. Perform 1 and 2 above in different geographic locations for purposes
of identifying the ecosystem independent and dependent variables
and/or processes.
4. Identify the variables and/or processes of interest with respect to
sensitivity and relevance to definition of system state.
13
-------
4.2.2 Test Methods for Interspecific Interactions
To adequately fulfill the mandate of the Toxic Substances Control
Act, laboratory test systems must be developed for predicting the effects
of chemicals on those ecological phenomena that are indicative of inter-
specific interactions and community structure (that is, attributes which
can be measured only at levels of biological organization above the
population and below the ecosystem). Recent work has shown that indices
of community structure and possibly of function are at least as sensitive
as the "most sensitive species" approach. Attributes must be identified
which represent competitive, plant-herbivore, predator-prey, host-parasite,
and mutualistic relationships among species. In addition, community-level
phenomena must include competition within functional groups, structural
parameters, successional relationships, food chains, and food webs. There
is also a need for better microcosm testing and data interpretation tech-
niques to predict biomagnification of substances under relatively realistic
steady-state conditions.
Inclusion of such laboratory test systems in the risk assessment pro-
cess would impart more realism and a more integrated or holistic character
to this process. The integration of multiple direct effects and indirect
effects on interspecific relationships by the microcosm approach moves us
many steps closer to projecting impacts on resources that have real or
perceived value (e.g., fisheries). The formulation of a standard proto-
col for each ecological phenomenon identified is perceived as too narrow
in scope. For each attribute, development of a set of protocols, each
designed for a particular ecosystem type, is envisioned. For example,
at the very minimum, test systems for the terrestrial, freshwater, and
marine environments are needed.
4.2.3 Identification of Holistic Ecosystem Attributes
Easily measured indicators of ecosystem-level effects in microcosms
must be found. Holistic or integrative attributes of ecosystems (such
as energy flow, nutrient cycling, and homeostasis) are being investigated
theoretically, and these studies have begun to appear in the open litera-
ture. One fact that emerges is the dependence of many of these indices
14
-------
upon the same data base, namely, the flow network among the constituent
compartments of the ecosystem. For example, Finn's cycling index,
Mulholland's conditional entropy, May's connectivity, and Ulanowicz's
ascendancy are all parameters calculated from the network of flows between
system compartments. Currently, there is little agreement as to whether
such attributes are clearly identifiable or useful as parameters to be
evaluated in the assessment of chemical hazard. The development of
practical techniques for measuring these theoretically derived attributes
is recommended. In addition, ecological principles and theory need to be
applied and tested in microcosms to identify the holistic attributes that
are the most sensitive indicators of change in structure or function of
various ecosystems. The effort to search for and validate the use of
holistic attributes should receive increased attention. At the least,
it will help to refine our theoretical understanding of ecology.
4.2.4 Development of Criteria for Test System Performance^
Since the intended application of microcosm test results is to be
an integral part of the overall risk assessment, the performance of a
developed test must be quantified in terms of the goals for which it was
designed, its internal variance, and its simulation of natural variance.
Following such evaluation of test performance, appropriate statistical
design and criteria of performance can be established prior to routine
use of the test system. Statistical inferences about differences between
stressed and unstressed (treated vs control) conditions are founded in
probability theory, and the validity of test results, in a regulatory
sense, depends on the attention paid to this aspect.
No general criteria exist for assessing what constitutes a success-
ful microcosm. Such evaluations, of course, depend upon the purpose for
which a test study is developed. In cases where microcosms are designed
to closely simulate a natural system, the importance of variability in
that natural system cannot be ignored. For example, if a microcosm is
designed to study the fate and effects of an organic chemical, realistic
results might require that the microcosm be validated by comparing the
rates of major pathways, identifying major degradation products, assessing
15
-------
steady-state bioaccumulation ratios, and/or examining one or more exo-
toxicological effects in the microcosm and the natural system.
On the other hand, microcosms designed as screening tests might in
principle be less rigorously referenced to a given natural system, pro-
vided the necessary complexity of biological interaction is included. A
screening microcosm could supply guidance to further testing of a chemical.
For example, screening data could indicate which ecosystem components act
as reservoirs and which effects require more detailed testing. In addi-
tion, predictions of media partitioning models that are based on simple
physical and chemical properties can be evaluated and confirmed. Here
the validation might be of a more general nature and might include the
assurance that results seen are similar to what would be found in natural
systems. The information base is not available to establish criteria for
the degree of similarity that should be expected for either simulation
microcosms or screening microcosms. Experimentation and attempts at such
validation should be continued on a long-term basis so that appropriate
criteria can be formulated.
4.2.5 Stochasticity and Non—Steady-State Dynamics
Data taken from laboratory microcosms usually are in the form of a
time series for each compartment of interest. Often the series are com-
pared in the original form, a procedure which on occasion could result
in misleading conclusions. For example, the history of nitrate concen-
trations in two identical microcosms may exhibit similar form, but be
shifted in phase. When compared at any one instance, the values are
markedly different. Analysis of the variance applied to the primitive
series would likely infer that the two systems do not replicate. Cal-
culation of the autocorrelation function or power spectrum function, how-
ever, eliminates the effects of phasing. Appropriate statistical tests
carried out upon these transformations might reveal that the time dynamics
are essentially the same.
Similarly, time series analysis presents one of the few methodolo-
gies applicable to non—steady-state dynamics in microcosms. Here one is
reminded of the classical analysis of electrical engineering. Single
16
-------
frequency input to a system results in a spectrum of outputs which serves
to characterize the response of the system. In like manner the investi-
gator could sinusoidally vary the inputs and stresses to his system one
by one. The resultant spectra of the compartments should characterize
each response to the input being varied. Furthermore, cross-correlation
of compartment series can be used to identify the interactions among the
compartments of the system. Comparison of cross-correlation spectra be-
tween control and perturbed microcosms should allow one to focus upon
those interactions most sensitive to the chemical substance under study.
Thus, while present microcosm studies are usually executed under
constant laboratory environmental conditions, much more information
could be obtained from model systems by temporally varying the inputs
and environment in a controlled manner. Elucidation of the non—steady-
state dynamics of microcosms and of their responses to chemical perturba-
tions should, therefore, be pursued by systematically varying laboratory
conditions. The judicious use of time series analysis appears to be
necessary in extracting the maximum possible information from non—steady-
state laboratory ecosystems.
5. INTEGRATION OF OEPER MICROCOSM PROGRAM
5.1 INTRODUCTION
It may be desirable to have some manner of formal integration of
the OEPER microcosm program to carry out the following functions:
1. reduce or avoid redundance in systems development, while attempting
to establish appropriate commensuration (especially between media
and among media) at the same level of testing function (e.g., screen-
ing, system validation, and monitoring support);
2. establish a sense of continuity and appropriate common goals, so that
necessary conjunction of research is achieved;
3. establish a mechanism for communication between researchers and
program office staff (e.g., Water Quality and the Office of Pesticides
and Toxic Substances), extending that communication to the scientific
community through a newsletter; and
17
-------
4. establish an internal peer review mechanism regarding, for example,
interlaboratory testing of protocols or innovative technology.
A substantial part of these functions can be met by semiannual meet-
ings of the involved researchers and peers in the appropriate user offices.
Indeed, it would be counterproductive to add additional program reviews,
extra meetings, and more paperwork. If agreement can be reached among
researchers on the chemicals to be used in protocol development and round
robin testing, then a single, cost effective purchase of radiolabeled
material can be shared. If criteria for operations, structure, and func-
tion are agreed upon, then the products and outputs are more likely to be
of utility to the program offices.
Thus, the thrust of this integrative program is to facilitate (not
constrain) microcosm technology developments within limited resources
(particularly of money and staff time). It is important to keep in mind
that this element of the Public Health Initiative, substantially expand-
ing ORD resources in microcosm technology, was sold to Congress as an
integrated package for exposure assessment; it is not a blank check.
Hence, OKD management has a responsibility to see that laboratory prod-
ucts fit within the mandate of the legislation. Therefore, it is pro-
posed that a committee be established to carry out the aforementioned
functions within that management framework. This group would be called
the Microcosm Technology Committee (MTC).
5.2 COMMITTEE STRUCTURE
The MTC would consist of all active OEPER researchers (at least one
representative from each laboratory carrying on this work) and representa-
tives of the user program offices. This group would meet annually or, at
most, semiannually to review progress, improve communication, and establish
goals or criteria as desirable. The group would gather under the manage-
ment and auspices of the Deputy Administrative Assistant (DAA) of OEPER.
18
-------
5.3 COMMITTEE FUNCTIONS
The MTC would fulfill several functions, including providing for
the following:
1. Research efficacy. An annual review, held early enough to provide
ample time for input to the research committees, would provide a
forum for working out the most efficient and effective means of
organizing laboratory and contract research. This would provide an
opportunity for workers in the same media to get together (which
might otherwise be unlikely) and for those working toward the same
ends (e.g., screening tests) to adjust their techniques to fill
voids.
2. Continuity. There is no guarantee that a given researcher or labora-
tory program will necessarily continue, but the MTC would allow new
researchers to develop the close contacts vital to effective research.
Furthermore, it is expected that some common goals and directions
will evolve from these meetings that will assist each researcher in
doing the job, provide the program offices with better understanding
of the limitations and potential of microcosm technology, and allow
researchers to discuss the needs of each program office.
3. Communication. Among the most important functions of the committee
will be the direct communication links that it establishes between the
researchers, program offices, and ORE) management. It is proposed that
a newsletter, published by contract, be undertaken immediately.
4. Internal peer review. Since interlaboratory testing and validations
of microcosm technology are likely to be expensive and more time-
consuming than simple toxicity assays or chemical fate tests, it is
proposed that the MTC serve as a clearinghouse or peer review group
to coordinate this function according to accepted criteria. Such a
function would assist management and the research committees in find-
ing appropriate resources to bring products of microcosm technology
to the stage of application by user offices.
19
-------
5.4 COMMITTEE OPERATIONS
The group could elect a planning committee, or steering group, to
organize the annual or semiannual meetings to assure that the program is
likely to achieve its necessary functions. This steering committee would
work with ORD management in initiating and supporting committee activities,
such as the newsletter. Otherwise, the functions of the group will be
maintained simply by the OEPER staff between meetings. It seems desirable
for the MTC to meet at laboratories (preferably ORD) where microcosm work
is ongoing and where in-depth discussions are more likely to arise re-
garding a particular technology. The meeting should be held annually
(in spring or early summer) or, at most, semiannually at a time conducive
to the greatest representation by researchers and users. Sponsorship of
the meeting by the DAA/OEPER would facilitate travel and attendance.
20
-------
APPENDIX
LIST OF PARTICIPANTS
Beverly Ausmus
Battelle Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
William Cooper
Department of Zoology
Michigan State University
East Lansing, Michigan 48640
Sidney Draggan
Division of Policy Research
and Analysis
National Science Foundation
1800 G Street NW
Washington, D.C. 20550
William Dunlap
Robert S. Kerr Environmental
Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, Oklahoma 74820
Alfred Galli
Office of Research and Development
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Jeffrey Giddings
Environmental Sciences Division
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
James Gillette
Environmental Research Laboratory
U.S. Environmental Protection Agency
200 S.W. 35th Street
Corvallis, Oregon 97330
Anna Hammons
Environmental Sciences Division
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
Rizwanul Haque
Office of Environmental Processes
and Effects Research
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Pat Bonney Hartman
Information Division
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
Steve Hedtke
Newtovm Fish Toxicology Station
U.S. Environmental Protection Agency
3411 Church Street
Cincinnati, Ohio 45244
Michael Heeb
Office of Pesticides and Toxic
Substances (TS-792)
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Ray Lassiter
Environmental Research Laboratory
U.S. Environmental Protection Agency
College Station Road
Athens, Georgia 30601
Po-Yung Lu
Information Division
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
21
-------
John Matheson III
Environmental Impact Staff
Bureau of Veterinary Medicine
Food and Drug Administration
Rockville, Maryland 20857
Perry McCarty
Civil Engineering Department
Stanford University
Stanford, California 94305
Vincent Nabholz
Office of Pesticides and Toxic
Substances (TS-792)
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Brock Neeley
Environmental Sciences Research
Dow Chemical Company
Midland, Michigan 48640
Jon Parker
Radiological and Environmental
Research Division
Argonne National Laboratory
Argonne, Illinois 60439
Timothy Parsons
Department of Oceanography
University of British Columbia
Vancouver, British Columbia
Canada
Kenneth Perez
Environmental Research Laboratory
U.S. Environmental Protection Agency
South Ferry Road
Narragansett, Rhode Island 02882
Michael Pilson
Graduate School of Oceanography
University of Rhode Island
Kingston, Rhode Island 02881
Hap Pritchard
Environmental Research Laboratory
U.S. Environmental Protection Agency
Sabine Island
Gulf Breeze, Florida 32561
James Reisa
Office of Pesticides and Toxic
Substances (TS-792)
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Courtney Riordan
Office of Environmental Processes
and Effects Research
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
Robert Ross
Information Division
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
David Specht
Environmental Research Laboratory
U.S. Environmental Protection Agency
200 SW 35th Street
Corvallis, Oregon 97330
Steven Spigarelli
Radiological and Environmental
Research Division
Argonne National Laboratory
Argonne, Illinois 60439
Frieda Taub
University of Washington
Seattle, Washington 98105
Robert Ulanowicz
Chesapeake Biological Laboratory
Center for Environmental and
Estuarine Studies
University of Maryland
P.O. Box 38
Solomons, Maryland 20688
Gunter Zweig
Office of Pesticides Programs
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460
22
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/9-80-019
2.
3. RECIPIENT'S ACCESSION-NO.
4, TITLE AND SUBTITLE
Intel-laboratory Evaluation of Microcosm Research:
Proceedings of the Workshop
5. REPORT DATE
April 1980 date of issue
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R. Haque, OEPER: J. V. Nabholz, OPTS;
M. G. Ryon, ORNL
8. PERFORMING ORGANIZATION REPORT NO.
ORNL/EIS-160
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Information Center Complex
Information Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
10. PROGRAM ELEMENT NO.
AIVLIA/AIZLID
11. CONTRACT/GRANT NO.
IAG-78-D-X0453
12. SPONSORINC
Office of
AGENCY NAME AND ADDRESS ,„,.,. ,
Environmental Processes and Effects Research
Office of Research and Development
U.S. Environmental Protection Agency
Washington, B.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/16
15. SUPPLEMENTARY NOTES , . ... ^/r. c
Prepared in cooperation with the Environmental Review Division, Office of
Pesticides and Toxic Substances
16. ABSTRACT
This workshop, held at Athens, Georgia, on 18-19 September 1979, was to aid the
Office of Environmental Processes and Effects Research (OEPER) in evaluating and
integrating microcosm research in the OEPER environmental research laboratories.
Participants discussed the design, advantages, and limitations of microcosms; the
value and potential role of microcosms in the regulatory decision-making process;
the identification of priority and long-term research needs to refine microcosm
methodology as a data source for hazard and risk assessments; and the creation of a
microcosm technology committee to assist OEPER in coordinating and integrating
microcosm research. These proceedings consist of recommendations made by special
task groups on each of the major topics discussed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Aquatic ecosystems
Terrestrial ecosystems
Aquatic microcosms
Terrestrial microcosms
Testing protocols
Microcosm design
06/F
IS. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS fTTiisReport)
Unclassified
21. NO. OF PAGES
33
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
6PA Form 2220-1 O-73)
23
-------
-------
EP 600/9 EPA
180-019 Env. Proc. & Effects Res
AUTHOR
Interlaboratory evaluation of
T1TLEmicrocosm research; proceed-
ings of the workshop
DATE DOE
GAYUDRD 45
BORROWER'S NAME
-------
-------
DATE DUE
-------
> ra C
B.
i
Hi
i
S I
i I
"O
>
t = I!
I
I
O
------- |