Volume III
           Ecosystems/Modeling Workshop
                         OFFICE OF PESTICIDE PROGRAMS AND
                                 WASHINGTON, D.C.
    30 - AUGUST 1, 1975

                 Volume III

          Ecosystems/Modeling Workshop

                         FREDERICKSBURG, VIRGINIA
V^JULY 30 - AUGUST 1, 1975

FIRST YEAR OF PROGRESS are published in four volumes.  Volume I
contains speeches and discussion from the Plenary Session, the agenda,
and lists of participants and speakers. Volumes II,  III, and IV cover the
Toxicological Methods and Genetic Effects Workshop, the Ecosystems/
Modeling Workshop, and the Chemical Methods Workshop, respectively.

                 TABLE OF CONTENTS


Wednesday, July 30,  1975

    Dr. John L. Buckley	1
    Dr. Norman R. Glass	3

    Dr. James W. Gillett	5

    Dr. James R. Sanborn	29

    Dr. W. Peter Schoor	31
    Dr. W. Peter Schoor	39

    Dr. Allan R. Isensee	45

    Dr. Erik D. Goodman	53
    Dr. James W. Gillett	65
    John Bowser	73

    Dr. Riz Haque	79
    Dr. Fumihiko Hayashi	87

Thursday, July 31, 1975
    Carter Schuth	105
    Dr. James Hill	Ill


                            OPENING REMARKS
                            Dr. John L. Buckley*
  There are some detailed models which seem to be reasonably good for
what happens in local areas.  There are some conceptual models that seem
to me to be highly useful.  I guess I feel somewhat better about it now than
I did a year ago.  The tools available to us through the activities of many
of you in this room and others have really provided some additional data and
some additional methods that can be used for looking at this question  of
what happens to this particular set of chemicals in the environment.
  I'd make one other point and I make it because of the position that I was
in last night.  I was there to talk about the Substitute Chemical Program
and I did for about 15 minutes out of a two-hour program. The rest of the
session occasionally touched on the Substitute Cnemical Program, but more
than that,  it touched on why EPA had made some of the decisions that it had
  I was reminded that the quality of the decisions that get made isn't  going
to be any better than the quality of the information available at the time the
decisions were made.  I suppose that is an obvious statement. The other
half of the statement is that availability of good information doesn't assure
that good decisions will be made.  I don't  mean to put down any individual
decision but I note that the two aspects are separable  and at the same time
intricately related. It seems to me that our concern  here should be with
how we can improve the quality  of the information that can be used hi making
the decisions.
*Acting Deputy Assistant Administrator for Program Integration, EPA

                    OF FUTURE NERL PROGRAMS
                      Dr. Norman R. Glass
  Briefly, I'd like to make just a couple of very quick comments and then
move more directly into the program in some depth.  The first comment that
I'd like to make I think most of you are aware of but for those who aren't, let
me restate  it.  This Program was funded in May of 1974,  a scant year and a
half ago, and what that really means is that there are some portions of the
Program that are very new and some portions that are somewhat older.

  There are three primary components to  the  Program.  These three
objectives are to develop a microcosm methodology which can be used as a
pre-market screening device for proposed substitute chemicals; second, through
the development of the microcosm methodology we hope to be able to get at some
of the basic ecological processes which may be affected by certain of these sub-
stitute chemicals; and third, we hope to be able to integrate some of these data
and information by the use of modeling techniques.

  To carry out those three major program objectives there  are  four organi-
zational components that have been involved.  The four components correspond
roughly  to the four research laboratories within the  EPA's Office of Research
and Development that have expertise bearing on this problem.

  First  there's the Corvallis  Environmental Research Laboratory  of which
I am a part. We are responsible for coordination and overall project
management and direction and also for the terrestrial research part of the
Program and modeling.  The Gulf Breeze Environmental Research Laboratory
is the second major component of the Program and their responsibility  is
primarily in the marine and estuary portions.  Third, the Duluth Laboratory
* National Ecological Research Laboratory, Corvallis, Oregon

is responsible for fresh water toxicology and fresh water ecological research.
Fourth, the Athens Laboratory is responsible for some work in modeling as
well as work in fresh water ecology.

  In addition to the in-house Program, there is also an extramural portion
conducted through contract, grant,  and interagency agreements, and these
components supplement the in-house part of the Program.

  We have for example, Dr. Sanborn from the University of Illinois who will
be discussing part of the work that he and Dr. Metcalf have been doing there
under contract to the Corvallis Lab.  Dr. Isensee from USDA in Beltsvllle,
Maryland is going to discuss the  results of his and Dr. Kearney's work on
microcosms.  Dr.  Goodman from Michigan State is going to talk to us about
the progress of his grant which was funded one month ago, so this will be a
short story. Also Pete Schoor from Gulf Breeze may choose to cover some
portions of their extramural program in addition to his own in-house work.
Last, but not least,  we have three representatives of the Criteria and
Evaluation Division who will present the results of the contracts they've been
overseeing for the past year.

  In summary,  the Program is broken down into three components in terms
of medium into marine, fresh water,  and terrestrial.  It's broken down into
three objectives: the development of microcosms,  the use of microcosms to
improve our understanding of process, and the use of models for integration.
This Program is being discharged by the four laboratories which I have just
identified and by their corresponding extramural contractors.

  Finally, during the course of the past year to year and a half,  in addition to
discharging the research program per se,  we've also reviewed 25 or 26 chemi-
cals that have come from OPP under the Scientific and Mini-Economic Review

                     PROGRESS AND STATUS REPORT
                          Dr. James W. .Gillett**

   The Substitute Chemical Program (SCP) is charged by the Congress with
developing procedures for determining the safety of materials  registered for uses
for which other agents have been limited or deregistered. The National Ecological
Research  Laboratory is carrying out a program of research on the disposition and
effects of these pesticides in laboratory microcosms in terrestrial/atmospheric,
estuarine/marine,  and freshwater systems.  A conceptual model indicating the
interrrelationship within and between these media has been produced (2). The need
to know or project  accurately potential effects of chemicals deliberately or acciden-
tally added to the environment forms the basis for the development of laboratory
ecosystem simulators. Controlled environmental parameters (light, temperature,
humidity,  air and water movements) and defined biotic and abiotic elements are
combined to form a laboratory system representing processes typical of the real
world.  Parallel development of mathematical models and simulators permits inter-
pretation and extrapolation of effects.  Laboratory microcosms represent an inter-
mediate level of investigation between the "test-tube" or single animal toxicology
test and field studies of natural ecosystems.
Historical Background
   During the late  1960s Dr. Robert Metcalf  of the University  of Illinois led a
group of investigators (3) into developing a "farm pond" laboratory system
(Figure la) that has been used to assess the environmental fate and,  to a lesser
extent, effects  of over 100 pesticides and other hazardous chemicals. In spite of a
number of deficiencies inherent intrying to simulate anything as complex as a natural
* Co-author is Jay D. Gile
**National Ecological Research Laboratory, Corvallis Environmental Research
  Laboratory, Office of Research and Development, EPA, Corvallis, Oregon

ecosystem, these aquatic/terrestrial microcosms have demonstrated their
utility and basic validity by providing information on the disposition of chemicals
in the biota and in the abiotic elements (water,  "soil," and chamber.  The data
from experiments can be compressed into two indices of disposition:  Ecological
Magnification (E. M.), the  ratio of the concentration in any biological element
to the concentration in water or soil; and the Biodegradability Index (B. I.), the
ratio of polar metabolites (water-soluble breakdown products) to the remaining
chemical agent.  For persistent chemicals (for which broad experience in the
field provides a direct basis for comparison) the E. M. s are large (usually great-
er than 1000) and B. I. s are low (usually less than 0.5).  Conversely,  the E. M. s
are low (100 or less) and B. I. s high (greater than 1) for a number of agents,
including some which might be employed as replacements for the environmentally
difficult, persistent pesticides.
   As an element of the Substitute Chemical Program, NERL has continued EPA
support of the work of Dr.  Metcalf, who is now developing a terrestrial "mono-
culture" microcosm (Figure Ib)  that can  be  related to major crop usage of
such substitute chemicals.  Thus far, this  system (1) has been employed, with
aldrin,  DDT, dyfonate,  and methoxychlor.   Accountability of material has been
increased to about 65  to 75 percent and at least two specific research problems
of major interest have been generated.  In  the first instance,  Cole et al.  (1)
have quantified the phytotoxicity of aldrin/dieldrin to corn.   Second, they have
found a major and as yet unidentified metabolite of aldrin in corn.  It is highly
significant that microcosm research can develop new aspects for study even
with an agent that has been used for a quarter century.  Less spectacularly but
of greater  interest to  the program,  the tests to date with the "monoculture"
terrestrial  microcosm have shown the same patterns of residues and critical
indices (B.I.,  E.M.) observed in both the "farm pond" system and the real

                             Figure 1

     A.  Metcalf Aquatic/Terrestrial Microcosm
     B.  Metcalf Terrestrial "Monoculture" Microcosm
    SAND «3

                                              Mosquito  larvae

                                              Corn (Cotton,
                                              Prairie Vole
                 B    METCALF
                       "MONOCULTURE "-TERRESTRIAL
The NERL Terrestrial Microcosm

   At the NERL facility in Corvallis,  terrestrial microcosm research is being

extended an order of magnitude in size and scope in attempts to develop a system

capable of assessing fate and effects in greater detail and for longer periods of

time.  The basic system  is derived from considerations of the conceptual model

of pesticide movement (2) in air, soil, and biota and from the pioneering work of

Metcalf's group (1,  3). The Terrestrial  Microcosm Chamber (TMC) consists of

a box with associated  air and water systems and contains an artificial soil and a

variety of  biota.  General design objectives of the TMC have been to provide

capability  for:

   1. Complete mass  balance or accountability for an introduced pesticide;

   2. Repetitive samples so that the temporal course of disposition could be


                               Figure 2

     NERL Terrestrial Microcosm Chamber (TMC)
                      (AIR INLET)
                                                (TO AIR
                                           MONITORING PACKAGE)
                               •SPRING" RESERVOIR)
                                             (TO "GROUND-WATER*

                 Annual rye grass
                 Endemic flora
                  (algae, bacteria)
Endemic micro-fauna
 (protozoa, etc.)
Field cricket
Button quail
Gray-toiled vole
   3. A situation in which all or a major part of the life cycle of selected

organisms could be observed,  quantified, and evaluated in the face of chemical

insult;  and

   4. An apparatus and operational system of general availability and applica-

bility to disposition studies,  as a screen by industry or the Agency.

The  TMC

   The TMC (Figure 2) consists of a glass box  (1 x 0,75 x 0.61 m deep) with a

Plexiglas lid.  The glass joints are sealed with silastic aquarium cement covered

with aluminum tape. The lid has an access port (20 x 30 cm) in the center

covered with a 30 x 40-cm sliding Plexiglas door.  Around the upper edge of the

port hole and on both the lower and upper surfaces of the door are adhesive-

backed Teflon string. The door is held in place by a slide, the lower edge of
which is covered with the Teflon tape over a layer of 1/4-inch cork.
   The top and bottom exterior edges  of the box are bound with nylon-reinforced
tape; the exterior bottom 20 cm of the box are also covered with opaque duct tape.
The top edge of the box has a 6-mm cork trim covered with an epoxy resin, acting
as a seal between the lid and the box.  The lid, which is edged and crossed by
1/2 x  1-inch Plexiglas bars to act as reinforcing members, is compressed
against the box by 1/4-inch aluminum  bars (across the top along the outside edge)
connected by threaded rods to the  light bank support frame.
   The TMC rests on a 1-inch thick polyethylene foam cushion over a plywood
frame which is set up to provide a 2 percent grade (2 cm drop per 100 cm) longi-
tudinally.  A pair of TMCs are set under a light bank containing 16 96-inch, 100-W
blue white fluorescent tubes.  A double 100-W pink bulb incandescent fixture is
centered over each TMC.  The lighting is controlled by a cam timer such that  at
"dawn" the incandescent bulbs receive variable power (0 to 100 percent) over a
30-minute period followed by sets of four fluorescent tubes being lit at 15- to 20-
minute intervals. At "dusk" the procedure is reversed, with the four sets of
four tubes being turned off at 15- to 20-  minute intervals, followed by dimming
of the incandescents.  The light/dark period being used is 16 hours on/8 hours off.
   The temperature of the room is regulated by a high capacity air conditioner
which provides an ambient temperature of 21-24°C at "night" and 25-30°C during
the "day" when the lights are on.  The temperature at the soil surface ranges from
24° to 30°C, depending on time of day, degree of foliar coverage, and placement
of probe in the TMC, while air temperatures may average about 3° to 5°C higher
during the day.  The majority  of heat  input into the TMC  is from the light banks
located about 30 cm above the lid.
   The soil in the chamber has thermistor/psychrometer probes located at
various depths.  The leads are sealed into the bottom by  silastic covered with

                           Figure 3
             Schematic diagram of air and water systems of NERL TMC.
           a.  5n filter capsule (a' 47 mm 5v  Metricel filter); b. char-
           coal filter capsule; c.  25-1.  carboy; d. bypass valve; e.
           equivalent plumbing to each TMC; f. Micropump;  g. three-
           way valve for "rain" and/or "spring" supply;  h.  needle-valve
           controlled flowmeter;  i.  "spring" inlet tube; j. "rain" nozzles;
           k.  "groundwater" outlet tube;  1.  10-1. water sample reservoir;
           m.  air supply  inlet;  n.  air sampling system outlet;  o. flow-
           meter;  p.  400-mm Graham-type condenser  (water-cooled);  q.
           1-1. Pyrex condensate receiver;  r.  400-mm Allihn-type
           condenser (refrigerated to -15°C);  s.  gas washing bottles with
           coarse filter and 300-ml solvent; and t.  exhaust vacuum
           regulating valve.
                              SYSTEMS OF NERL TMC
                       WATER SYSTEMS
                   Olr -»
                                         f—"TMC"^  ^«r~l
                                         !    <    L#/M
                 (TYPE I)
epoxy adhesive.  The psychrometer function of all 20 leads ceased after the first

few weeks of use, but the thermistor function has remained operable.

The Air System (Figure 3)

   Air Supply;  The top of the TMC is fitted with a 10-cm nylon-screened (52 x

52 mesh) opening over which a Plexiglas disc has been attached by methylene

chloride.  The disc has a 7-mm I.D. glass tube cemented into it, which in turn

is attached externally to a polypropylene "T'^tube.  An excess of membrane- and
charcoal-filtered compressed air is delivered at 0.5 psig in the long leg of the
"T,"  replacing air removed from the chamber of the air sampling system  (v..i.).
The remainder flows out through the other limb of the "T."  Should the air supply
fail, room air would temporarily replace the filtered air.  Should the chamber
exhaust fail, the full flow would bypass the chamber.  Hence intricate and dif-
ficult control of internal TMC air pressure is obviated.  A filter or trap may be
attached to the external exit of the air system to assess losses in the event of
exhaust blockage.
  Air Sampling System:  The exhaust port in the lid is a 1/2-inch polypropylene
Swaglok bulkhead fitting capped on the proximal end with 52 x 52 mesh nylon screen-
ing and topped externally with a 47-mm  Millipore polypropylene filter holder en-
closing a  heavy cellulose prefilter and 5/*  Metricel filter.  Mounted on top of the
filter holder is a size 13 Gilmont rotometer,  which is used to adjust the air flow
rate from the TMC.
   A section of polypropylene tubing connects the flowmeter to a standard taper
joint into a 400-mm Graham-type water-jacketed condenser.  The bottom joint of
the condenser is  on the side arm of an adaptor mounted on a 1-1 Pyrex bottle and
below a refrigerated (-15 C)  Allihn-type condenser acting as a freeze-out  trap.
The refrigeration system is set to automatically defrost for 10-minute intervals
every two hours, but that function can be adjusted as needed to prevent blockage.
The bottle collects fluids from both condensers.  The exhaust exits the freeze-
out trap via a  standard taper glass joint to a 1/2-inch Pyrex tee attached to
nylon Swaglok fittings and passes through a double set of dual gas scrubbers
containing 300  ml each of an appropriate trapping solution. The scrubbers are
fitted with a 50-mm coarse porous disc located 8 to 10  mm above the bottom of
the vessel.  Tygon tubing is used to attach the last two scrubbers to an exhaust
manifold with a regulating valve.  All standard taper joints throughout are sealed

with Teflon or Fluochem tapered seals coated at the top with Apiezon lubricant.
The freeze-out condenser, scrubber traps, and the refrigeration lines are
cove red with 1/2-inch black polyethylene foam.   The  air sampling systems from
two TMCs  are mounted securely on a single frame which can be easily detached
(at the flow meter and exit) to provide better access to the TMC.
The Water System (Figure 3)
   The "Rain/Spring" System:  Similar to the air system, the water system of
the TMC has  two parts:  one for input or supply and  one for output or sampling.
Water is added to the TMC as needed from a common supply (reverse-osmosis
purified water in a 25-liter carboy) via either a "spring" or by "rain." A three-way
valve at each TMC permits both modes of application simultaneously and indepen-
dently. All components are polypropylene except as noted.
   Water is pumped (Micropump) from the reservoir through a 5^ Metricel
filter to the valve at the TMC.  For the "rain"  system the tubing connects to
two, parallel-mounted pairs of bulkhead fittings each sealed into the lid 25 cm
from the end  and 18 cm from the side. The "rain" nozzles  are ceramic discs
(Bete Fog Nozzle, Inc., Greenfield, Mass.) with 18 0.05-inch tapered and angled
(60°) holes.  Discs are supported internally by a Viton ring and externally by a
Teflon ring seal. Pores  in the disc are periodically cleaned by reversing the
disc for a short period.
   The "spring"  system supplies water to a tube extending 1 cm above the
surface of the soil.  A flowmeter (Gilmont No.  11) is used to regulate the delivery
rate of water, which is pumped by  air pressure (bypassing the pump at the
reservoir). The ascending arm of the "spring" system (located at the high end
of the TMC) is encased in a larger (0.5-inch) piece of polypropylene tubing to
permit flow directly to the sand-gravel under-stratum.  The entry tube is
sealed into the bottom of the TMC by epoxy cement over silastic.

   The "Groundwater"  System; At the low end of each TMC,  about 6 cm from
each side of one corner,  a groundwater sampling tube is sealed through the
bottom of the TMC with silastic and epoxy cement. The  tube projects  inside
about 1 cm and is covered by a loose-fitting, 10-cm long section of 1/2-inch
O. D.  tubing with a 1/4-inch hole drilled in the middle.  After exiting the TMC,
the tubing ascends to a 47-mm filter holder (containing both a coarse cellulose
pre-filter and a 5^ Metricel filter).  The line then descends to a shut-off valve
and a Micropump, which draws water from the TMC and passes it into a 10-liter
sample reservoir.
The Soil
   A 5-cm layer (about 40 kg) of washed and autoclaved river gravel passing
through a 1/2-inch mesh screen is applied carefully to the bottom of the terrarium.
Another 5-cm layer of 8-grit (coarse) washed river sand (about 40 kg) is then
overlaid, and both layers are washed with several changes (about 60 liters) of
water by pumping off and discarding the water until the  effluent is no longer
turbid.  About 15 liters of distilled water (enough to cover both layers to a depth
of about 6 cm) is added carefully with a sprinkling can.   The prepared soil or
potting mixture  is then added in three 40-kg portions  along with about 6 liters
of distilled water. After each addition of soil or potting mixture and water, a
1/2 -inch plywood board (the size of the surface  of the TMC and with a small
cutout for the "spring" pipe) is laid on the soil and topped with two lead bricks
(about 15 kg each).  After standing overnight, the bricks and board are removed
and the next portion of soil and water is added as before. The soil should be
slightly (about 3 cm) below the top edge of the duct tape shielding the sides of
the soil from light. A final addition of 2 cm of mixture occurs at planting of the
   Soil thermistor probes are located specifically as  to coordinates and depth
during the soil addition.  Leads must be long enough so that the unburied sensors
are out of the way during compaction by the board and so that shielding is

provided during operation.  Some soil invertebrates are added before the final
addition of soil and water.
    The soil or potting mixtures utilized thus far have been arbitrary and a matter
of experimentation.  Washed 20-grit sea sand produced little foliage even when
used with a complete hydroponic supplement (1).  Invertebrates therein suffered
severe abrasion.  Standard potting mixtures, such as Jiffy-Mix Plus,  have an
inordinate amount of organic matter (50 percent peatmoss) and would not be
relevant to activity in normal soils.  Excellent productivity and compatibility
with quail, voles,  and various invertebrates were found for this material, how-
ever.  The currently used mixture is composed (on a weight basis) of 10  percent
hydrated Jiffy-Mix Plus, 45 percent 20-grit washed sea sand, and 45 percent
Turface (Wyandotte Chemical Co,), a heat-expanded illite clay soil amendment.
The Plants
   With approximately 2 cm more soil to be added to the terrarium, the seeds
are added by (a) planting according to a specific pattern (coordinates, depth) or
(b) scattering a given weight of seeds as evenly as possible over the surface.
Species planted according to (a) include corn (Zea maize. "Golden Jubilee" or
Midget Golden Bantam), peas (Pisum satrum).  sorghum (Sorghum vulgaris).
and a variety of kole crops (Brassicasidael Species planted according to (b)
have  included alfalfa (Meticago sativa). Johnsongrass (Sorghum halpense).
perennial ryegrass (Lolium perenne). and annual bluegrass (Psoa annua).  A
wide variety of indigenous algae have been observed,  including greens and blue-
greens. Their appearance has been discouraged by addition of a fine layer of
sand over the top of the final 2 cm of soil  added after addition of seeds.   How-
ever, as the surface is disturbed (e.g., by digging of the vole), the algae become
   In any given terrarium,  a botanical array has been sought which will provide

   1.  Differential herbicidal or phytotoxic response (broad-leaf vs.  grasses);
   2.  Economic crop (corn, soybeans) vs. weed (ryegrass, Johnsongrass);
   3.  Sound food basis for vole (ryegrass, alfalfa, corn, soybean),  and
   4.  A conformation supportive of anoles (corn, soybean, Sorghum spp.).
   Corn and soybean are two of the crops most intensely impacted by pesticide
usage, and a number of the substitute chemicals will find heavy application to
these  very large acreages.  In the Pacific Northwest grass seed crops are of
substantial economic importance.   Perennial ryegrass has the greatest acreage,
but little pesticide usage is required.  On the other hand, this plant is one of
the most prevalent and economically significant weeds in other cereals in the
region.   Its growth habit and ease of handling make it an ideal component of the
   Other grasses are of less importance, either as crops or weeds,  but John-
songrass is a major weed problem  in corn and cotton areas of the Midwest and
Southeast.  It has proven difficult to handle (highly variable germination,  propen-
sity for alleopathic effects on other plants) and appears to be less than ideal.
Additionally the voles regard it as a less desirable foodstuff.
   Alfalfa and other legumes in addition to soybeans do well in supporting a
variety of economic insects and are widely grown throughout the country. Their
growth habits are excellent for the  terraria, since they can form dense plantings
of substantial biomass.   The vole of choice (v. i.) is native to  alfalfa and pastures
in the Willamette Valley  and prefers these crops to all others offered.
The Animals
   In  this first phase of microcosm development the following species have been
included for the reasons  noted.  In general, the criterion of availability from open
sources has been followed.

   Nematodes: No endemic species of nematodes have been isolated from
unsterilized soil mixture.  Two easily identified bacteriophagic nematodes have
been added in water to the soil system prior to planting:  Pristionchus iheriteiri
and Cephalobus perseghis.  The latter becomes dominant after about 30 days.
These species were selected primarily for their ease of culture by the Soil Ecol-
ogy Unit at NERL.  Nematodes occupy a significant point in the food web as
consumers of bacteria elicited by plant root  exudates, animal wastes, etc., and
as food for other soil  invertebrates.   Phytopathogenic nematodes, while of
interest both in theory and practice,  could introduce  a problem of infestation
interrupting the basic energy converting system upon which the principal orga-
nism (e.g., vole) is dependent.
   Earthworms:  Commercially available Lumbricus spp. provide another
important link between soil-bound pesticides and other biota. The worms are
conveniently and effectively added just prior to the final compaction of the soil.
Their activity  in redistributing soil residues is very  evident at early stages of
terrarium growth.  Effects of agents on these species are judged significant,
since loss of earthworm populations may alter soil character and productivity
so markedly.  In addition to Lumbricus spp., laboratory-reared Enchaetriad
worms are being studied, since the latter have a higher soil temperature toler-
ance limit.
   Snails:  The common garden snail, Helix pomata. is readily reared in large
numbers.   The adults are not preyed upon by other microcosm inhabitants,  but
the young which emerge from eggs laid deep (15  cm)  in the soil are prey for
voles and possibly other vertebrates. Being strict herbivores, the snails pro-
vide a constant competition for the vole  and  serve  as terminal repository of
plant-borne residues.  These snails represent the order Mollusca and with slugs
constitute a major pest of agricultural and ornamental crops.

   Meal Worms;  Third and fourth instar larvae of Tenebrio molitar are readily
reared or obtained commercially.  They are added primarily as a food source
for the vertebrates, but their subterranean and scavenging habits also mean that
they  are exposed to soil-bound pesticides and other residues and hence can effect
bio-transfer of these agents or their metabolites.  These larvae are added prior
to additions of vertebrates and at several times during the life of the terrarium.
Excess adults from the laboratory colony have also been added in some tests
and fare well.
   Crickets:  Adult brown house  crickets (Achetus domesticus  are obtained in
large numbers  inexpensively, either from commercial producers or by rearing
in the laboratory.  As omnivorous/herbivorous orthopters, they serve several
roles in the microcosm ecosystem, but act primarily as a food source for
vertebrates.  Their food and activity patterns  present opportunity for transfer
of residues from soil and plant material to their predators.  Reproduction in
the soil provides a sensitive index of toxicity,  while the young continue to occupy
a role in the food web somewhat  similar to that of the adults.  A substantial
portion of the life cycle (10-12 weeks) can be completed in the TMC.
   Houseflies;  Musca domestica L. are easily reared in quantity and provide
an excellent food species for insectivores.  The toxicological background  know-
ledge on this species is superior to that of any other included among the TMC
biota, since houseflies have been a leading pest and are so easily manipulated.
The particular  strain employed herein is the maximally susceptible (e.g., lack-
ing genes for dieldrin resistance on Chromosomes II,  in, and V),  blind (white
eye), vestigial winged,  mutant.   Since these can neither fly nor avoid predators
by sight, they are an ideal prey.  This strain has been thoroughly investigated
in the course of a number of studies on insecticide resistance, degradation,
and microsomal induction as detailed in the works of our source,  Dr. Terriere's
group at Oregon State University (c.f., Schonbrod et al.,  1968; Yu  and Ter-
riere, 1973).

    Pillbugs: Armadillarium and Porcella ssp. gathered in the field and reared
 in the laboratory are included as representative of the arthropod detritovores.
 Adults are generally not subject to predation, although cannibalism can be a
 problem when food is in short supply.  Young pillbugs appear to be vulnerable
 to their own and other species,  thus providing a food link between decaying
 organic matter  and higher organisms.
    The greatest concern about dispersive chemical usage relates to effects on
 higher animals, including man.  Selection of appropriate models was based on
 the following criteria:
    1.  The principal orders affected should be represented, especially Mam-
 malia and Aves:
   2.  The species should have an ecological niche not grossly distorted by the
 physical size and contents of the terrarium;
   3.  The species should show relatively "normal"  physiological processes
 and behavior under the conditions  of the  terrarium;
   4.  They should be generally available or easily reared; and
   5.  Some toxicological background should be available.
   Thus far the species selected qualify  according to all criteria but the last.
   Gray-Tailed Vole: Microtus canicaudus is a locally native vole (40-60 g),
now a resident in the alfalfa and ryegrass fields of the Willamette Valley.
Dr. Larry Forslund of Oregon State University has been carrying out extensive
reproductive behavior studies on this and other Microtus spp. for the past
several years. These voles, which mate well in captivity,  have an 18-day
gestation period with post-partum estrus; hence gravid females, (about 55-60 g)
are readily available.  Normal litter size is five  to seven pups born hairless and
blind.  Weaning occurs at 15 to 20 days of age.
                                    18              .

   The vole is sufficiently omnivorous to permit it'to consume the entire bfotic
contents of the terraria (except adult snails and birds).  They have been observed
eating all plant species tested (except mustard); all invertebrates added except
adult  snails; quail eggs; and even a two-week old Japanese quail chick (about 30 g)
that was found dead in the burrow. The anoles are believed to be eaten by the
vole since in only a few instances anoles added to the terraria have been recov-
ered  at termination (e.g.,  at the loss of all surface vegetation).  Only two dead
anoles have been found in an operational TMC. The intense appetite  of this
rodent and its ability to adapt well to terrarium life commend it for use in bio-
magnification studies.
   Button quail; Although it was originally intended that the toxicologically well-
explored Japanese quail (Coturmx coturnix japonica) should be the avian species
in the terrarium, that bird was unsuccessful. On the recommendation of James
Keith, Denver Wildlife Center, U. S. Fish and Wildlife Service,  the button (or
Chinese painted) quail (Excalfactoria chinensis) has been employed in the TMC.
Although basically a seed-eater, this species  is smaller (40-60 g adult) and more
omnivorous than other galliforms which might be used. Males and females are
distinctively marked and are handled fairly easily.  They are available from most
exotic bird distributors and breeders, but have not received systematic attention
nor any toxicological study.  They lack a cloacal gland and hence should be excel-
lent subjects for artificial insemination (pers. comm. Dr.  George Arscott), a
matter providing for side experiments using individual males exposed to agents
in the TMC.
   "Chameleons"  (Anoleus sp.): Very little study of the effects or even of
residues of pesticides  in reptiles has taken place, even though these species
can be especially important in a number of ecosystems. Since Anoleus sp.  are
readily obtained and maintained in captivity and since these are  active insectivores,
they seem a reasonable component in the terraria.  They are  arboreal animals,

 however, and require (or at least prefer) habitats 10 to 30 cm off the ground in
 the TMC.  Heavy stemmed grasses (e.g., corn, sorghum,  or Johnsongrass)
 provide an adequate habitat. They feed only on live insects, adults or larvae,
 and in turn appear to be preyed upon by the vole.
 Results and Discussion
    Pesticides have not yet been tested in the TMC system,  as efforts to date
 have concentrated on developing the operational parameters and requisite under-
 standing of the combinations of biotic and abiotic elements. Tests in preparation
 include studies of sampling procedures and appropriate analyses to ascertain the
 functionability of the design of both the TMC and experiments. The current test
 protocol is shown in Table 1 and typical  results for such an experiment are
 illustrated in Figure 4. The average terrarium appears to provide 5-7 kg of
 plant material in 25 to 30 days.  Button quail require supplemental feeding of
 seeds (10 g of mixed alfalfa and ryegrass) every second or third day, as assessed
 by body weight changes. Added crickets are consumed quickly (about 10-15
 crickets/bird/day).  Once the vole is added,  however, the competition between
 the vole and quail requires higher food inputs to prevent starvation of the quail.
 Then 10 g of seeds and 25 crickets were added daily.  Oyster shell (10 g) is
 added weekly for laying female quail and snails.
    The pregnant female vole literally destroys the terrarium. Initial impacts
 involve clipping of alfalfa and soybean leaves, tunneling and nest building, and
 in two instances chopping down of  the corn plants.  The growth of the corn
 shades the terrarium and substantively alters the soil temperature.  Why two
 voles elected to raze the corn without eating it is unclear,  but the degree of un-
predictability of such behavior is not tolerable in attempts to generate a consist-
 ent microclimate.  For 13 other voles,  the corn was left until last, and then the
leaves were attacked first.  Hence, inclusion of corn as a test crop risks con-
 sistency of operation which should be taken into account in design of experiments
for pesticides not used on corn.

               Table 1:     Protocol for Typical Experiment

                         TYPICAL EXPERIMENT


                -3         Add soil mixture
                           Start light/temperature cycle

                 0          Cfreat with pesticide]
                           Plant seeds
                           Add earthworms, Tenebrio larvae,  pillbugs.
                           Bring water up to 1/3 level by rain

                1-3        Bring water up to level by rain

                 7          Add 150 crickets

                 14         Add anoles

                 21         Add male and female quail
                              alfalfa and  rye seed
                              Tenebrio larvae

                 28         Remove quail
                           Add more seed, crickets
                           Add gravid vole

               42-56      Remove vole, terminate terrarium
   The voles survive an average of 21.4 days after introduction and produce

an average of 4.6 pups per litter brought to weaning.  However,  some voles

have not whelped or the pups have not survived at all.  Possibilities for abor-

tion or cannibalism of young must be kept in mind in evaluating effects of any

treatment.  Where pups have been weaned,  the terrarium is characteristically

well-tunneled, with a nest about 10 cm in diameter located just above the  sand

layer deep in the soil toward the upper end of the TMC.  Such nests will have

several (2-5) entrances and are filled with rye or alfalfa as nesting material.

Frequently, but  not always,  storage areas for food will be found in other por-

tions of the tunnel system.

                                   Figure 4

                         Results of Typical Experiment

                    TERRARIUM IA
                Vole (Microtus canicaudus)
            ,—-r--—• v PUPS
                Anoles  (Anoleus spJ
                Quail  (Excalfactoria

* 20
                                     EGGS  • •
                                    (rimovtd) >
                                       2 broktn, 7i*t, 4 hatched
                                                        > -o
                                                        3 £
^W ••• trf

for 16 to 18 days at 38°C.  The quail make a rough nest of grass, almost cave-
like,  which simplifies locating the eggs,  apparently laid about the same time
each day.  When the eggs were not collected in early experiments, they were eaten
by the vole.  No direct interaction of the quail and vole were observed.
   The anoles fared less well, partly because the quail and vole were more
aggressive in acquiring the insects (especially crickets and Tenebrio).  As the
supporting foliage and stems were eaten, the anoles began disappearing.  The
anoles would display their vermillion throat patch whenever the vole approached
them, but no vole has been observed attacking the anoles.  The average anole
weighs about 2.5 g and would not be able to defend itself except by flight.  Even
the addition of an artificial "tree" (a cross-tee  of 5-mm bamboo set up 10 to 15
cm above the soil surface) did not provide sufficient protection.
    The insects survive  to varying degrees.  The commercially obtained crickets
showed 50 percent survival times of about two weeks in the absence of predators.
Tenebrio larvae matured to adults, but the latter did not reproduce.  The crickets
produced large numbers of young which appeared about 15 to 20 days after adults
were  added.  No cricket matured beyond the third instar stage during the course
of these experiments (45 to 60 days).  Houseflies lasted only one to two days post-
emergence due to predation, and the pupae had to be suspended in a container from
the  lid in order to obtain emergence.  Otherwise the vole or quail would have
eaten the pupae.
    Earthworms did not  survive well at the operating temperatures used in these
studies,  although other tests in the soil mixture at 20 to 22°C gave good (60 per-
cent or better) survival. The vast majority of earthworms recovered have been
located at the edge deep in the soil.
   Snails survive without incident and almost without exception.   Young are
not always produced, but egg masses or young snails occurred in aobut 70 per-
cent of the TMCs with 4 or more snails.  Snails create a problem in that their

 feces and slime trails on the box obscure visibility and are difficult to clean off
 (e.g., for recovery of pesticides).  As condensate forms on the top and sides,
 algae blooms also occur, but these are not suppressed by the snails.   The early
 stages of young snails are eaten by the vole.
   Nematodes are under more intensive study because the observed distributions
 thus far seem to be highly indicative of the extent of the edge effect in the TMC.
 Concentrations of nematodes in the 5-cm  wide zone at the edge, where vole and
 quail paths provide trampled and decaying vegetation and animal feces, may be
 up to 100 times greater than in the center top (upper 5-cm) zone.  Populations
 differ by an order of magnitude less at 10 cm depth.
   Pillbugs also seem to fare well in the terraria,  although survival is limited
to about 20 to 30 percent recovered after 45 days.  They are usually found on the
surface and at the edge,  but are active throughout the TMC.  Both Tenebrio
larvae and pillbugs are found feeding on dead adult crickets and Tenebrio.
   The plants also have had varying degrees  of success.  Moisture must be
carefully regulated to low levels to prevent necrotic lesions  on the soybean
leaves.   Soybeans will develop inflorescences but not fully flower under the light
conditions; corn will develop spikes, but not tassels.   Ryegrass goes  to seed
after 50 to 60 days.  Alfalfa never blooms and growth is retarded relative to field-
or greenhouse-grown plants. Bluegrass does moderately well under soybeans,
but none  of the grasses did well in the presence of Johnsongrass.  Three to 30
plants of the latter would  sprout  from 10 g of seed.  The sorghums (S. vulgaris
and S. halpense) had the same growth habit as corn and were about as poorly
accepted by the vole, i.e., only toward the end of vegetative growth of the TMC.
Among the  other plants with characteristics suitable to the TMC are included
several beans (dwarf lima, cowpea,  chickpea, sweetpea, garden pea)  which
are readily accepted by the vole.  The  brassicacious plants (several lettuces,
cabbages, brussel sprouts, kale, mustard, and spinach) grew slowly and were

not well accepted.  Carrots and peppers grew very poorly.  All plants tested
in the currently used potting mixture (Jiffy-Mix/sand/Turface) had heavy root
development, in some cases down into the sand/gravel base.  Root development
was heaviest in sand alone, but foliar parts were minimal.  Root development was
less in Jiffy-Mix alone, although foliar development was adequate.
   Thus far the 15 TMCs have operated in such a way as to provide convincing
evidence that the future study of pesticides and other chemicals is feasible in
more complex systems than attempted heretofore.  Yet to be tested are those
portions and operations of the system that involve analysis and treatment with
pesticide for determination of the functionability of the TMC in that regard.
Still further ahead in testing and design is the  mathematical  model of the TMC
that is being assembled (6).  At the end of the  first year of effort  in the Substitute
Chemical Program, the NERL Terrestrial Microcosm project then has developed
a conceptual model of pesticide movement, employed that  model in design and
construction of a terrestrial microcosm test system, and  is ready to begin tests
with pesticides.  The first series of tests will employ 14C- dieldrin and 14C-
methyl parathion to investigate aspects of operation.  The second series of plan-
ned tests will compare the fate of   C-labeled p_-nitrophenol from parathion,
methyl parathion, and pj-nitrophenol itself.  The third series of tests will com-
pare pentachlorophenol,  hexachlorobenzene, and pentachloronitrobenzene.
This latter group, an  analogous series of substituted benzene rings differing
only at one group, provide comparisons derived from the consequent physico-
chemical differences and also are related in pesticidal characteristics. HCB
is a major contaminant of PCNB; POP,  a major herbicide in its own right,  is also
a metabolite of HCB.  The fungicide captan (14C-trichloromethyl-labeled) may
also be compared in this series since the three benzenoid chemicals all have
fungicidal or fungistatic activity.


1.  Cole, L.K., Metcalf, B.L., and Sanborn, J.  R.:  Environmental fate of
    insecticides in terrestrial model ecosystems.  Presented at Ecological
    Society of America, Corvallis, Oregon, August 20, 1975

2.  Gillett, J. W.,  Hill, IV,  J.R., Jarvenin, A. W., and Schoor, W.P.:
    A conceptual model for the movement of pesticides through the environment.
    U.S. Environmental Protection Agency Ecological Research Series EPA
    660/3-74-024 (1974)

3.  Metcalf,  R. L., Sangha, G.K., and Kapoor, I. P.: Model ecosystem for the
    evaluation of pesticide biodegradability and ecological magnification.
    Environ.  Sci. Technol. £:  709-713 (1971)

4.  Schonbrod, R.D.,  Khan,  M.A.Q., Terriere,  L. C., and Plapp, Jr., F.W.:
    Microsomal oxidases in the housefly:  a survey of fourteen strains.  Life
    Sci.  1 (Part D ;  681-685  (1968)

5.  Yu, S.J., and Terriere, L. C.:  Phenobarbital induction of detoxifying
    enzymes  in resistant and  susceptible houseflies.  Pestic.  Biochem. Physlol.
    3.:  96-101 (1973)

6.  Haefner,  J.A., and Gillett, J.W.:  Mathematical model of terrestrial labor-
    atory microcosm.  Presented at Substitute Chemical Program meeting,
    Environmental Protection Agency, Fredericksburg, Virginia, August 1,1975
   QUESTION:  You made a statement that you were going to look at the nutrient
cycles.  What did you have in mind?

   DR. GILLETT: Particularly nitrogen, phosphorus, sulfur, and carbon.  We

have a soil ecology group working on the same processes in litter decomposi-

tion in association with coal-fired power plant siting.  If they can develop

adequate methodologies that are conveniently used, we will probably use them.

   QUESTION:  Had you considered the effect of these chemicals on this micro-
flora? Were you thinking of that route or are there so many ways to tackle it?
   DR. GILLETT:  There are so many ways and one of the problems, which I
may discuss later about the mathematical simulation, is trying to discriminate
between the significant processes by sensitivity analysis of the mathematical
model.  This requires that several laboratories do a lot more work on eco-
system simulation in the laboratory and in mathematical models.
   COMMENT:  In answer to your question, we're trying to develop something
that will work to look at changes at ATP concentration, changes in direct count
to get handles on the bacterial populations and also looking at CO2 production
and oxygen consumption to try to get a handle on the process and process rate.
   QUESTION:  I can't tell whether you had ultra-violet light there or not.
   DR. GILLETT:  No, we do not have and that is a problem in terms  of photo-
   QUESTION:  What is the  effect on your controls of possible starvation of your
organisms during the tests? For example, a quail might run out of food or the
vole might eat most of the food toward the end. Mice sometimes will have re-
productive effects if they are starved.
   DR. GILLETT:  Regarding our voles, we could have a real problem.  For
instance, once the last bit of food disappears,  the pups are  extremely
vulnerable.  She has a litter of six  or seven and they are about five to  seven
grams.  The  demands on the mother and the demands on the  system are extra-
ordinarily high.
   The pups can die quickly because they do not have enough energy input.  We
have to find means  of identifying where she is  and what her condition might be.
We are considering  a thermistor telemetric device which will measure her body
temperature, because if that starts dropping,  then her energy consumption is

 abnormal.  The device will tell us not only where she is but also something
 about her condition. That is getting very sophisticated in terms of what we
 need to know about disposition and may not be worthwhile  in terms of the over-
 all cost of the experiment.
   You have to understand the processes within this chamber just as you  must
 understand the processes in the environment. You can intuitively and deduc-
 tively reach some understanding of what is happening.  This system does not
provide answers; in the absolute sense it provides guidance to where to ask the
best questions  and what to look for.  Whether biomagnification or a multiplicity
of metabolic pathways are present,  this will reveal it. It  will not tell you the
routes of metabolism in cows.  It will not tell you what will happen in mothers'
milk, but  it  will give you a profile around which you can build a good research


                   Dr. James R. Sanborn*
                         NOT AVAILABLE


*Illinois Natural History Survey, Urbana, Illinois

                         Dr. W. Peter Schoor*

  This report has not been published and its contents must therefore be considered
  The reason for using mirex is to establish an exposed control type system.  That
is the point at which we now are.
  A simple model ecosystem was developed to study the fate and effects of pollu-
tants in the estuary.  The approach used is intermediate in complexity and represents
a refinement beyond single-species bioassays in that a food chain permits the study
of the ability of an animal to escape predation after exposure to nonlethal concentra-
tions of such compounds.  We have thus made use of one of the most fundamental,
yet most sensitive, biological indicators of behavior.
  This criterion of effect has been largely disregarded in recent years,  especially
since the advent of molecular characterization.  The present state of fundamental
biochemical research, however, permits the use of enzymatic reactions  as unequivo-
cally sensitive tools only in rare cases,  especially outside of mammalian pharma-
cology and toxicology.
  Metcalf et jid. (1971) described an experimental fresh water ecosystem which has
produced valuable data on the biodegradation and bioconcentration of many compounds.
Our experimental ecosystem is similar to theirs and will be used as a screening
device for toxic materials to identify the most toxic and/or most highly concentrated
compounds that would be potentially hazardous to the estuarine environment.  This
ecosystem can serve until a scientifically more rigorous and elaborate system has
been developed to assess more accurately detrimental effects of pollutants on the
estuary.   We are presently engaged in this task, but find that nature steadfastly
refuses to be simulated  on a short-term basis.
*Gulf Breeze Environmental Research Laboratory, Estuarine Systems Branch


  The components of our ecosystem represent a simple food chain that includes
 a primary producer (turtle grass - Thalassia testudinum), a primary consumer
 (grass shrimp - Palaemonetes vulgaris), and a secondary consumer (pinfish - Lagodoj
 rhomboides).  This choice represents an important part of the food web found in
 turtle grass communities  (Hansen, 1969; Adams and Angelovic, 1970)  which, ranging
 in depth from 0.5 to 4.5 m, cover much of the bottoms in estuaries of the northern
 Gulf of Mexico (Humm, 1956).  It also assures a year-round, continuous and reliable
 source of organisms which adjust readily to laboratory conditions.
  Mirex was  chosen as toxicant because it is a chemically fairly stable compound
 not readily hydrolyzed, oxidized, reduced, or otherwise chemically altered. This
 stability permits long-term exposures with reasonable assurance that  the parent
 compound will be the one actually investigated.  Few other compounds interfere
 directly with its quantitative analysis by gas  chromatography.  Moreover, mirex
 can also be introduced into the water without the aid of a solubilizer, since it is
 used in bait form adsorbed to corn cob grit.  Any induced leaching thus provides
 amounts of mirex in the water as they would  occur in the environment.
  The physical size of the system was chosen to allow water samples of 1 to 2
 liters to be taken without undue  disturbance of the "steady-state." It is also the
 minimum  size in which to keep a community  consisting of the following components:
 *  160 liters of artificial seawater (distilled water and marine mix) at a salinity of
 20 parts per thousand.
 * 4 cm of sand.   Stock of sand had been dredged from Santa Rosa Sound, Florida,
 and consisted of 25 percent coarse particles  (#35 standard sieve) and 75 percent
medium particles (#120 sieve).
 • 75 turtle grass plants, Thalassia testudinum, (175-210 g total) planted over 1500
cm of the bottom.
 • 75 adult grass shrimp, Palaemonetes vulgaris, (30-35 mm rostrum-telson length,
0.20-0.25 g each).

 * 2 juvenile pinfish, Lagodon rhomboides, (90-95 mm total length,  9-12 g each).
  The tanks were covered and water temperatures maintained at 20±1 C.
Fluorescent lights were used to provide 12-hour periods of light and dark.  The
dissolved oxygen content was kept relatively high by artificial aeration.  The tem-
perature, pH, dissolved oxygen, and turbidity were measured at regular intervals.
  After equilibration of the system for 4-6 days, mirex was introduced Into the
water by means of an air-lift column, in which a continuous flow of air causes water
to percolate through a column containing mirex in bait form {Borthwick, 1975).
The amount of mirex in the columns represented five  times  field-rate application
and was chosen because earlier studies (Tagatz et al., 1975; Borthwick, 1975) showed
that concentrations thus produced in the water did not cause significant mortalities
in grass shrimp exposed for a period of 3 weeks.
  Water, sand, turtle grass, and shrimp were sampled at intervals (Table 1), and
mirex residues were determined using a Hewlett-Packard Model 5700 gas chromato-
graph with  a  Ni electron-capture detector.  An OV  101 column (2 percent OV 101
on gas chrom Q, 100-120 mesh) was operated at 195°C with the detector at 300°C
and the argon/methane (10:1) carrier gas at a flowrate of 60 ml/min.

                               Table 1

           Exposure                         Amount Mlrex (/ig)
            (Days)                      Shrimp2            Water
               1                          0.4                15
              4                          1.1                16
               7                          2.0                12
              10                          2.6                  9.2
              13                          2.5                  5.0
1  Simple average of two replicate tanks.
2  Total residue in component remaining at corresponding length of exposure.

  After an exposure period to mirex of 16 days, chi-square analysis of the surviving
 shrimp showed nonsignificant mortalities in all tanks (Table 5).  Predators (pinfish)
 were added at that point and mortalities of shrimp observed at intervals of 1, 2 and 3
 days.  The results  are depicted in Table 5.
  The water quality data indicate that our experimental ecosystem was not
 significantly stressed,  if at all.  The pH values are normal for those of seawater
 (7. 5-8.4) and the dissolved oxygen was close to saturation,  The water was not turbid
 Appearance of new blades of turtle grass indicated growth under test conditions in
 both control and exposed tanks.
  The amount of mirex in the water was deliberately chosen to reflect  nonlethal
 levels for the time of exposure, which was substantiated by chi-square analysis.
 This permitted predation on grass shrimp which were not dying, but obviously affecte
 by mirex to such an extent as  not to be able to escape predation to the  same degree
 as control  animals.  Chi-square analysis showed significance at the 5 percent and
 1 percent levels.

                               Table 2
                                 Amount Mirex
1 Simple average of two replicate tanks.
2 Total residue in component remaining at corresponding length of exposure.

                                   Table 3:  Averages (/*g/kg) of Mirex
No Plants
No Plants
No Plants
No Plants
Water Sand
Plants Plants
No Plants
0.12 0.30
0.057 0.65
0.031 0.50
0.030 1.10
0.018 1.50
No Plants

                                      Table 4:  Adjusted (Water) Averages (ng/kg) of Mirex
Exposure Hepato-
(Days) pancreas
No Plants
1 280
4 370
7 360.
10 500
13 650
No Plants

No Plants

No Plants


Water Sand
Plants Plants
No Plants*
0.12 0.30
0.057 0.65

0.031 0.50
0.030 1.10
0.018 1.50
No Plants

* Adjusted

                                                    Table 5
Tanks Mirex
Control/ (^g/l)
« 1/1 ».061
(0. 13-0. 019)
1/1 °-°51
' (0.11-0.017)
3/3 0. 025
% Survival Chi-
(Numbers) Square
Control Exposed
(52 of 63)
(49 of 63)
(177 of 189)
68 N.S.
(43 of 63)
81 N.S.
(51 of 63)
88 N.S.
(167 of 189)
Predation % Survival
(Days) (Numbers)
Control Exposed
1 44
(23 of 52)
2 16
(8 of 49)
3 24
(42 of 177)
(10 of 43)
(0 of 51)
(5 of 115)
Chi- 2


19. 4°

1  Average concentration of mirex throughout exposure period.
2  N.S.  = nonsignificant; a «= significant at J
3  Omitted 1 tank because one pinfish die/1.
2 N.S.  = nonsignificant; a <= significant at 5% level (x2, 1 d.f.  = 3.84); b = significant at 1% level (x2,  1 d.f. = 6.63).

  We find,  thus, that while the criterion of death may be useful in single-species
bioassay, in an ecosystem bioassay an effect on the predator-prey interactions may
have far greater implications regarding the total welfare of the system.  In addition,
concentrations at which this behavioral effect occurs  are quite low and come closer
to the environmental levels.

                           Dr. W. Peter Schoor*
   This report has not been published and its contents must,  therefore, be
considered provisional.
   The reason for using mirex is to establish an exposed control type system.
That is the point at which we are now.
   It has been common practice in the past to classify the substratum of the
aquatic environment as a "sink" for pesticides and heavy metals.  Yet benthic
organisms abound and it is recognized that many of these organisms are instru-
mental in the physical turnover of  this substratum,  tt was thusly of interest to
   1. Whether or not the substratum represents a "sink" to benthic organisms,
   2. What behavioral changes may occur In these organisms in a pesticide-laden
substrate, and
   3. If pesticides  can be channeled back into the pelagic system.
   We  chose the lugworm (Arenicola cristata) for our studies because of its rela-
tive abundance in the shallow estuaries of Northwest Florida. According to D'Asaro
(1974), the following are characteristics of the lugworm:
   1. The estimated density is six to ten adults per square meter;
   2. Except for ease of damage to external gills,  they are a hardy species;
   3. They adjust well to laboratory conditions, including reproduction; and
   4. They are  easily maintained on a diet of dried, ground turtle grass (Thalassia
*Gulf Breeze Environmental Research Laboratory, Estuarine Systems Branch

   Two 180-liter covered glass tanks (120 x 30 x 50 cm) were filled to a height of
 30 cm with alternating layers of sand (75 percent coarse particles - #35 standard
 sieve; 75 percent medium particles - #120 standard sieve) and organic silt (from
 pristine Spartina marsh). A photo-period of 10 hours light and 14 hours dark was
 established using four 40 watt fluorescent bulbs mounted 30 cm above each tank.
 Filtered seawater,  average  salinity 26 ppt,  was added to fill the tanks.  The
 temperature of the water was maintained at 20 + 2° C.  Algae of the genus Ecto-
 carpus were blended and introduced into the tanks to form a growing mat on the
 substratum.  Four adult worms were then introduced in each aquarium and allowed
 to adjust for 3 days, after which time one tank was exposed to  mirex by means of
 an air-lift column (Borthwick, 1974). One-liter water samples (replaced water)
 were taken 5 days per week  and analyzed for mirex by gas chromatography.
   In order to make observations of the surface activity of the  worms, vertical
 lines, 10 cm apart, were drawn on the outside of the tanks ending precisely at the
 surface of the substratum.   The tanks were in this fashion divided into 12 equal
transects perpendicular to the tanks' longitudinal axes.  Daily  sightings were made
 along each transect, and the following surface features recorded:
   1. Active head shafts (feeding funnels),
   2. Inactive head  shafts,
   3. Tail shafts (respiratory holes), and
   4. Egg masses.
   Each feature was recorded together with the time of appearance and,  given a
significance of one, recorded in toto. The exposure to mirex was discontinued at
day 30, and fresh algae were reintroduced in both tanks at days 37 and 55.  The
experiment was terminated at day 75.
Results and Discussion
   Table 1 shows the concentrations of mirex in the water of the exposed tank.
The variation seen is most likely due to the adsorption of mirex to particulate

                 Table 1:  Mirex Concentrations  in Water
                 Days Elapsed                 Mirex Cone.
0. 036
matter (some ultracentrifugal experiments suggest as much as 80 percent), pos-
sibly the algae.  Our limit of detection for 1 liter of water samples was
   Table 2 represents the total of the surface features appearing per day through-
out the duration of the exposure to mirex.  As pointed out previously, each feature
was assigned a relative significance of one in order to quantitate a behavioral
effect.  This "relative" significance most likely can be better defined after more

                  Table 2
      29  ,

























































experimentation; however, used as defined here, statistically significant behavioral
changes occur after day 23.  The exposure to mirex was terminated at day 30, and
both tanks observed for another 45 days.  On day 37 more algae were introduced

   It has thus been shown that:
   1. The burrowing activity of Arenicola can affect the distribution of mirex in
the substrate;
   2. Low concentrations of mirex in water can affect burrowing behavior as
measured by surface activity; and
   3. The free-swimming juvenile stage can act as a source of mirex to any
      DR. GLASS:   If I could just address a couple of questions in general that
  were raised earlier, I think most of the answers to your questions about
  stability and the reasons for using microcosms are contained in the theory and
  lore that go along with microcosm development, which started many decades
  ago, and partly some of the problems we've had experimentally are the result
  of working with microcosms.  We don't plan them to be the end all of experi-
  mental apparatuses but they are Inexpensive and fairly rapid and useful In
  many respects.

                        Dr. Allan R. Isensee*

  The contract  that our laboratory has with the EPA on the Substitute Chemical
Program was finalized in late May 1975.  Progress thus far has involved methodology
development and the search for a biologist to work on the project.  One of the primary
contract objectives is the development of a system that incorporates the useful aspects
of both the static and flowing ecosystems.  The static, as presently employed in this
laboratory, subjects a pesticide to some typical environmental conditions in order
to determine their effect on the behavior and accumulation of the compound by eco-
system components.  The flowing systems have usually been designed to study the
effect of the compound on a  component of the environment.  If some of the major
elements of these systems can be "married," the result should be a more compre-
hensive model ecosystem.  This report will describe the ecosystems that we have
used and how these systems will be used and modified in relation to goals of the
Substitute Chemical Program.
Ecosystem Design
  The model ecosystems used in this laboratory (1,2) are entirely aquatic and con-
tain organisms representing parts of two food chains (algae to snails, and daphnids to
mosquito fish or fingerling bluegill).  The  C-labeled pesticide is adsorbed to soil
(100 grams),  introduced into the  system,  and flooded with 4 liters of water.  This
mode of entry simulates the major route of  unintentional entry of pesticides into the
aquatic environment, namely, erosion of soil from agricultural land that has been
treated with pesticides.  The surface soil, which usually contains the highest concen-
tration of pesticide, is also the most susceptible to erosion.  This rationale  of pesti-
cide introduction is also useful in determining what soil concentration to use.
Based on the behavior of a pesticide in soil, e.g., mobility, adsorption,  and rate
*Beltsville Agricultural Research Center, Agricultural Environmental Quality
 Institute,  USDA
**Contract No. IA6-D5-F811, Fresh-Water Micro Ecosystem Development and
  Testing of Substitute Chemicals

 of degradation, and its method of application, an estimate of the concentration in
 the surface few millimeters or centimeters can be made.  We normally use this
 estimated field concentration, a 10 x rate, and a 10   x rate.  The actual water
 concentration of pesticide depends not only on the amount of pesticides  added, but
 also on the rate and extent of desorption from soil.
  The tanks are placed in a shallow water bath (22+2 C) in the greenhouse and
 receive only natural light.  One day after flooding, daphnids,  snails, and algae are
 added.  Water samples are taken three times a week and analyzed by standard liquid
 scintillation methods.  The system is allowed to operate for 30 days.  At 30 days
 samples of daphnids are taken for analysis and two mosquito fish or fingerling bluegill
 are added. The system is terminated 3 days later when the fish have consumed all
 daphnids and  all organisms are harvested.  In some studies, two fingerling channel
 catfish are added and exposed for an additional 6 days after the other organisms are
  The fish and snails are homogenized in methanol and 1-ml samples are analyzed
by standard liquid scintillation methods.  If a high concentration of pesticide is
indicated, the extract is further analyzed by thin-layer chromatography to determine
if and how much metabolism has occurred. The algae and daphnids are assayed for
total radioactivity by determining   CO0 from standard  combustion and scintillation
methods.  The results are expressed as ppm (of the parent compound based on   C
content) in the fresh tissue and the bioaccumulation ratio (cone,  in tissue/oonc. in
water) calculated.
  Larger systems have also been used (4), usually containing 10 kg of soil, 80 liters
of water, and fingerling channel catfish and crayfish in addition to the organisms used
in the smaller systems.  This larger system is useful in obtaining more biomass
(for metabolism work) and can support larger organisms.  The  exposure time and
methods of analysis are similar to those used in the smaller systems.
  Two modified systems are presently under development. One is a recirculating
static model ecosystem (Figure 1).  Its major advantage over the above systems is
that fish are exposed for the entire 30-day period  along with the daphnids and other
organisms.  A 3-day exposure may not be  sufficient for equilibrium to occur.

In addition, fish placed in a system after a 30-day period may be exposed only to
pesticide metabolites; thus, uptake of compounds by the fish after 30 days may differ
qualitatively and quantitatively from that occurring during a continuous 30-day
exposure.  Some daphnids pass through the screen to provide food for the fish and
thereby complete the daphnids to fish food chain; enough daphnids remain to maintain
a reproducing population. The number of daphnids that move through the screen is
inadequate to provide their entire food supply, thus requiring supplemental feeding.
The pump is designed to  maintain the same concentration of pesticide  on both sides
of the glass partition and to transport fecal matter to the larger chamber where
daphnids may use it as food.  This system has a total water volume of 16 liters but
could easily be designed  for any convenient size.

                              Figure 1
      Air In
                                   'P*rco'Qf«"'  Wot., Pump
                     Aluminum Screen
                   'Water Depth  1
                  Glass Partition
Water Level

  Initial testing (minus pesticides) has shown it to be satisfactory, except that the
daphnid population was unstable.  One pesticide, hexachlorobenzene, has been inves-
tigated in both the static and recirculating systems (Table 1).  Statistically identical
results were not obtained, but treatment conditions (soil aged 1 year vs.  freshly
treated)  were different and may be partly responsible for the different results.
However, all results were well within an order of magnitude, which is often cited
as the normal variation of model ecosystem-derived data.
Table 1: Comparison of the Static and Recirculating Static Model Ecosystem
         for the Accumulation of l^C-HCB by Several Organisms
Algae Snails
	 RtnHr. — 	
— ppm-

10a             1.72      1.561*0.190    2.805±0.252       1.612*0.359   3.509*0.550-
	Recirculating Static	
9.2             7.9       4.51*0.46      0.58*0.17       0.95*0.10      3.20*0.69
a  Concentration of  C-HCB in 100 grams soil treated 2 days before use (static system),
   and concentration of 14C-HCB in 500 grams soil aged 1 year (recirculating static
   system).  The static and recirculating static systems contain 4 and 16 liters of water,
b  Average of three replicates plus or minus the standard error of the mean.
  A further modification, shown in Figure 2,  is designed to maintain a stable daphnid
 population and be used as either a static or a flowing ecosystem.  The system as shown
 has a total volume of 60 liters.  The major difference between the two systems is the
 screen and gate combination and the slanting glass barrier.  The 32-mesh screen
 restricts most daphnids while the 9-mesh screen allows all daphnids to pass. The gate
 can be  opened and closed for such time as to release enough daphnids to contribute


to the food chain,  yet not deplete their population.  Fecal matter collects on the
lower screen where it disintegrates, passes through the screen, and provides
food for daphnids.  The system shown  in Figure 2 is connected to a serial diluter
and has been operated for about  30 days with favorable results.  All organisms,
including daphnids, thrived in the system.
                               Figure 2
                          Top view
                         9 m*ih icrttn
                               Glass gat*
                                            32 mtsh m*«n
                                  Glais barritr
                                                      To drain or
                                                        r*cireiilarlng pump
                        Sid* vi*w
  The serial dilution apparatus that has been constructed for this project is
similar to that described by Mount and Brungs (3) and is designed to deliver the
test chemical at concentrations of 0. 5, 0.1, 0.05,  and 0.01 of the highest
concentration plus an untreated control.

   The next phase of the study will be to start the actual testing and comparing of
these static and flowing systems using simazine and trifluralin.  These initial studies
will be designed to determine how smoothly the systems operate and whether there
are any major differences in the behavior and accumulation of chemicals between
organisms exposed in the static and flowing systems.

1. Isensee, A. R. and Jones, G. E.: Distribution of 2,3,7,8-tetrachloro-dibenzo-
   jo-dioxin (TCDD) in aquatic model ecosystem.  Environ. Sci. Tech. _9:668-672
2, Isensee, A. R., Kearney,  P. C., Woolson,  E. A., Jones, G. E., and Williams,
   V. P.: Distribution of alkyl arsenicals in model ecosystems.  Environ. Sci. Tech*
   7_:841-845 (1973)
3. Mount, D.  I. and Brungs, W. A.:  A simplified dosing apparatus for fish toxicologj
   studies.  Water Res. 1.^1-29 (1967)
4. Schuth, C. K., Isensee, A. R., Woolson, E. A., and Kearney,  P. C.:  Distri-
   bution of 14C and arsenic derived from (14C) cacodylic acid in an aquatic eco-
   system.  J. Agr. Food Chem. 22j999-1003 (1974)
    DR.  MACEK:  Are you going to make any effort to try to discriminate between
  the residue that you detect in an atropic level that comes directly from uptake of
  water as opposed to that portion of it that may be a function of transfer by the
  food chain?  And in the experimental design, are you going to discriminate
  between those two sources ?
    DR.  ISENSEE:  No,  I don't intend to with this system. The  concentrations we
  will work with in the flowing system will probably be based on the concentrations
  we get from soil in the  recirculating static. The flowing system will be tested
  with and without soil, allowing the soil to have its own interaction, act perhaps
  as a sink,  perhaps to stimulate microbial breakdown. At least  four organisms,
  representing two trophic levels, will be used to measure accumulation and the
  results compared to the data obtained from the recirculating systems.

  MR. JOHNSON:  What's the function of all of the rather large number of
trophic levels?
  DR. ISENSEE:   These are the organisms we have worked with in the past
and we feel it is better to obtain accumulation data on several organisms rather
than one.

  QUESTION:  What do you do if you use a pesticide more toxic to daphnia
than it is to fish?

  DR. ISENSEE:   In those cases you will operate without some component of
the system.  However,  the whole idea is to work with concentrations which will
be realistic in terms of toxicitles.  For example, methyl parathion has an 11)50
for daphnids of 4.8 parts per billion, and about 8 parts per million for fish.
Therefore, in order to have daphnids surviving, the operating concentration
should be in the 1 or 2 parts per billion range.

   QUESTION:  When you're separating your fish and other components,  how
do you handle the nutrition of the fish during that period?

   DR. ISENSEE:  We'll have to go to supplemental feeding because the system
obviously does not generate enough daphnia to feed the fish.
   QUESTION:  What type of supplemental feeding?
   DR. ISENSEE:  Commercial fish chow.
   QUESTION:  How stable are the populations of organisms?

   DR. ISENSEE:  We've run the flowing system for 30  days and gotten blooms
of daphnids.  Apparently there is enough fecal material  from fish plus the
effect of soil to get a good daphnid population.

   QUESTION:  Are the experiments carried out in the greenhouse?

   DR. ISENSEE:  Yes.  Due to space limitations, all experiments are per-
formed in the greenhouse.  This results in some algae problems, but has the
advantage of allowing photodegradation to occur.  Some compounds have been
photodegraded in our systems.

   MR. JOHNSON:  If your daphnids are able to reproduce fairly well and develop
a large population, that could result in a really interesting shunt. Then for com-
pounds that are not very stable, the daphnids could be taking compounds out of
the water and doing other things to them.  I was just curious if you could repro-
duce that aspect of it.

   DR0 ISENSEE:  With a flowing system you never have to worry about concen-
trations because you have the toxins coming in all the time.  To a certain extent
the soil also works well, acting as  a reservoir from which the compounds desorb
and supply the toxin to the water.

   QUESTION: In 30 to 60 days, are unstable populations a problem?

   DR. ISENSEE: We have normally had stable populations in the small
four-liter tanks.

                     Dr. Erik D. Goodman*

  Society is becoming acutely aware of the necessity to develop a means of
anticipating environmental problems.  Post facto environmental management risks
an increasing array of undesirable and perhaps irreversible consequences. The
impact review process initiated by the National Environmental Protection Act clearly
emphasizes anticipatory decision-making.
  Despite the clear need for such an anticipatory decision paradigm based on sound
guidelines, in most cases the scientific structures required to predict responses of
biological systems have not yet been developed.  Predicting these responses requires
understanding of the basic mechanisms of component processes and how these are con-
strained by environmental conditions.  Most  environmental impact assessments to
date have relied on correlation or black box simulation models that fail to reveal the
structure-function relationships.  The mechanism-oriented approach to be utilized in
this research will attempt to represent each  process as realistically as is practical,
given constraints on data and complexity, and without resorting to a single mathe-
matical form, e.g., linear relationship,  to represent all processes.
  Over the short  term,  the research will build upon existing developments in lake
modeling at Michigan State University (MSU); current lake models will be revised
and expanded to include  the effects of pesticides introduced directly into the lake for
aquatic pest control and pesticides applied terrestrially that  reach lakes in sufficient
quantities to represent potential  problems.   Over the longer term, the research focus
will be extended to include the terrestrial impacts of selected pesticides and the
dynamics of the pesticides' transport and breakdown.  Due to the relatively limited
state of knowledge of quantitative ecosystem  dynamics in a terrestrial setting, it
*Department of Electrical Engineering and Systems Science, Michigan State University

will take considerably longer to develop a process-oriented predictive model to
assess this type of pesticide impact; however, even early explorations should point
out critical areas for investigation in the field and laboratory.
  A thorough literature search is  under way as the first step of the program, and
the relevant data obtained will be employed to develop preliminary computer models.
The relative importance of the various parameters will be assessed,  and the sensi-
tivity of the models to parameter magnitudes will be tested by mathematical modeling
and computer-simulation techniques used in the current eutrophication studies at MSU-
Computer models will provide the framework for organizing and integrating the
information gathered, insuring that missing or incompatible data are  recognized as
requiring further study.
  The literature review will include as "principal"pesticides,  malathion, parathion,
fenthion, azinphosmethyl, simazine, and atrazine, and as "subsidiary" pesticides,
methyl parathion,  chlorpyrifos, diazinon, trifluralin, PCNB,  and captan.  Effects of
all  six "principal" pesticides and of some or all of the six "subsidiary" pesticides
will be modeled by focusing on:
    1.  The amounts of the selected pesticides which enter the system;
    2.  The chemistry of the pesticides in their environment;
    3.  The distribution of the pesticides within the system, as mediated by physico-
chemical processes;
    4.  The direct  effects  of the selected pesticides on specific target and nontarget
organisms  in the system;
    5.  The interrelationships among biotic components of the system (for example,
among phytoplankton, zooplankton, bacteria, aquatic insects,  and fish);
    6.  The indirect pesticide effects (e.g., food-chain-mediated) on the structure
and function of the community; and
    7.  The degree to which pesticide impacts are modified by the turnover rates of
the biotic components (e.g., in a eutrophic vs. oligotrophic lake  or fertilized vs.
unfertilized field) or by other stresses on components of the ecosystem.

  In all cases,  causative factors will be sought and characterized.  Special attention
will be paid to both direct and indirect responses of the various elements of the food
web to the pesticide.
  The approach to be followed is Indicated by the structure being developed for
physical and biological processes in the lake ecosystem (Figure  1).  Variables not
homogeneously distributed in the lake are maintained separately for each structural
element, i.e.,  ring or layer.  For example, temperature, light, pesticide residue
level, population densities of several fish,  algae, zooplankton, insects, and macro-
phytes, and several chemical concentrations in each element are calculated at each
time step.  Movement between adjacent elements occurs by passive mixing as well
as by locomotion where appropriate.  Spatial heterogeneity must be explicitly repre-
sented, since many feedbacks between biological and physical processes are often
expressed via changes in the spatial distribution of materials and energy.   The level
of structural detail (i.e., number  of layers), however, can be chosen to fit the
particular problem under study.
  Figure 2 shows a flow diagram for a typical nutrient through the biota within each
littoral zone ring.  Flow rates depend on population levels and on additional variables
such as pH, temperature, and light.  Pesticides may directly influence the dynamics
of these populations or their nutrient fluxes; conversely the organisms can alter the
amounts and distributions of pesticide residues.
  Figure 3 Illustrates flow through the biota In a limnetic zone layer.  A sample
of the output from  a preliminary run of the limnetic zone model  for a hypothetical
lake is depicted in Figure 4. Phosphorus concentration changes with depth and time
in a pattern reflecting the relative complexity of component dynamics built into the
  Parameterization Is clearly a critical step In making the model  useful.  While
parameter estimates are based largely on work reported in the  literature, some
parameters must be evaluated empirically during this  project.  Laboratory and field
studies already underway include measurements of algal sinking (or flotation) rates and

                                  Figure 1

Hypothetical lake in cross-section and plan views.  In this example the limnetic zone
is represented by a stack of nine horizontal layers of equal thickness; uppermost
layers are bounded laterally by the innermost littoral ring, and lower layers are
bounded laterally by the sediments.  The littoral zone is here composed of three con-
centric rings numbered consecutively from the shoreline to the limit of vascular
A 1 Iltt0r^
ring 1 x ^>>>-' 9
ring. —

ent of
limnetic zone
.A J
layer 1
layer 2



4* Ov
X fc, ^^_^ , »
• 0 surface
2 o,
3 rr
5 t"
6 ra
9 bottom
               LAKE CONTOUR HAP
           above	V
                                  • shoreline

                                   Figure 2
Trophic structure and the flow of organic carbon, phosphorus,  and other nutrients
(ignoring possible gas phases)  in a littoral zone.  Exchanges of each functional group
with the dissolved pool are depicted by dotted arrows.


Allochthonous 	 J


>n...> . ....j

, 	 limnetic
=*scds" *
r •




— ^


limnetic .
_ 	 — 	 — 	 *
             allochthonous	>

                                   Figure 3
 Cycling of phosphorus (or other nutrient with no gas phase) in the limnetic zone.
 Zooplankton size categories are based on susceptibility to fish predation; phytoplankton
 size categories are based on susceptibility to grazing.
to Ipwer layers and sedf*~
         to littoral  **"

from drainage basin
from atmosphere
diffusion/mixing from below
diffusion/mixing from littoral
                                                                                -\  egeitlon
 experiments on zooplankton population dynamics; studies to begin within the next
 month include 1) pH and inorganic carbon effects on plankton seasonal dynamics,
 2) pesticide effects on periphyton in artificial stream channels, and 3) growth rates
 of periphyton and phytoplankton as functions of light, temperature,  pesticide, and


                                  Figure 4
Simulation depth profiles of ambient phosphorus concentration (ppb) from April 8
through August 6 in a hypothetical cylindrical lake with low phosphorus loading.
                *.ee -
                 6. 68 -
                               ftMBlEHT PHOSPHORUS
                                                        KEY	PP8
                                                        A -  .zoee-ci
                                                        0 "  13066*68
                  '   .W  |M-8  I  "6-8    72.8   95.8  I  123.
                        12.8   36.0   60.0    84.0    168.
                                  TIMS IN DAYS
   While the aquatic impact modeling builds upon a fairly extensive literature and
 an earlier lake modeling program, the terrestrial impact work must begin by develop-
 ing the  same type of structured representation for a terrestrial ecosystem that is
 shown for the aquatic system in Figures 1 through 3.  The initial focus will be on
 orchard and forest ecosystems, because they have relatively undisturbed flora and
 fauna (compared to tilled systems), because they are exposed to significant inputs
 of pesticides and have shown undesirable responses (1),  and because of the experience
 of our research team in studying impacts of persistent pesticides  on a forest ecosystem.
 At a later stage, attention will be given to tilled crop ecosystems  such as those of corn
 and cotton.  Concurrent studies also at Michigan State University  will be particularly
 valuable in evaluating pesticide runoff to streams and lakes.  Further stream pollution
 modeling studies are being proposed at Michigan State,  which would help in the con-
 ceptual linking of the pesticide impacts on terrestrial and aquatic  ecosystems.

  Validation of the models developed will be undertaken when the parameterization
activities are completed.  These studies will include both observational and experi-
mental (manipulative) comparisons.  The residues resulting from routine pesticide
applications in normal practice will be monitored at specific sites, and the effects
on species numbers (or biomass) and their residue levels will be compared with the
predictions of the model.  In addition, it is expected that a limited number of small-
scale direct pesticide applications to aquatic and terrestrial sites will be made in
order to quantify the responses of the ecosystem for comparison with model outputs.
  In summary, the research program is in the process of shifting its focus from
strictly aquatic ecosystem studies toward studies of terrestrial ecosystems, with
the goal of producing realistic and reasonably general models to help predict alterna-
tive pesticide impacts.

  1.   Croft, B. A., and Brown, A. W.  A.:  Responses of arthropod  natural
      enemies to insecticides.  Ann. Rev. Entomol. 30:285-335 (1975)
    QUESTION:   I'd like to know whether your mathematical models will be
 deterministic or stochastic, suggestive or predictive?

    DR. GOODMAN:  They are deterministic and predictive.  We're not trying
 to build into our functions any probability distributions and look at them from
 that point of view.  We will be accomplishing a similar thing by doing high-low
 level sensitivity analysis—that is, putting in high and low values for variables
 and determining what happens.   We'll be trying to bound the responses that we
 might get.  That doesn't give us exactly the same type of information that you'd
 get with a stochastic model; however, it reduces the model one level of

  QUESTION:  What mathematical form are you going to be using to your
model -- differential equations, difference equations, or what?

  DR. GOODMAN:  We'll be using difference equations, and at this point for
our aquatic models we're using a time step of a tenth of a day in determining
rates.  Obviously this requires higher level representations for some rates than-
would otherwise be used, and that requires our performing some calculations
relatively often. We're trying to economize wherever we can   though, by choosing
an appropriate time interval for various processes.
  QUESTION:  The rates of accumulation, concentration and dissipation—
since these are continuous constants, how are you going to describe them in
mathematical terms?

  DR.  GOODMAN:  We have to integrate over short time periods in order to
translate from a continuous process into a discrete process; in digital computer
simulation we're limited to dealing with discrete processes.  But the methodology
for translating from continuous time processes to discrete time processes is
fairly well developed and we're using that methodology. We'll be parameterizing
the system,  however, as difference equations rather than strictly differential
  QUESTION:  Are you using "systems analysis" in the sense of contemporary
engineering, or input-output analysis, or what methodology?
  DR. GOODMAN:  We're not doing input-output analysis.  We're using discrete
time state-space systems theory such as is employed generally in engineering
studies of control theory problems.

  QUESTION:  Is your model a distributed lag model?
   DR. GOODMAN:   It is not explicitly written as a  distributed lag model. There
are, however,  for many populations, multiple life~stages which essentially yield


 lags in the model.  Thus, you could analyze it as distributed lag model, but that's
 not the particular methodology from which we're trying to build it.
   QUESTION:  You have a framework there that you showed us to describe
 nutrient dynamics in the lake, and I wondered how you're going to go about lying
 the pesticide in there with that?

  DR. GOODMAN:   For each organism that we represent and each of the major
processes that may be involved in transfer of biomass among organisms, we will
be trying to determine for particular individual pesticides how they accumulate
in the organisms, if they do; what mortality they exact, if they  do; and then
how they influence rates in other fashions.  We expect to see some more or less
direct  effects which will depend, of course, very heavily on the concentrations
that we're looking at,  the range we're looking at.

  We also expect to see some indirect effects; that is, effects which are not
directly caused by the pesticide but are mediated by the perturbations you've
caused in populations.  If you wipe out all of one species,  you're going to change
the trophic dynamics of the system and we have to try to put some bounds — this
is really the goal — on what those effects may be and how they may proliferate
through the system.  Can you set up unstable oscillations?  Can you cause effects
of this sort on the ecosystem?
   QUESTION:  Sir, you've described a full system here.  If I  understand it
right,  based on the diagram that you showed, it would be a fairly detailed, almost
microscopic view of many aspects of the system, in order to include every aspect
to consider and relevant to a whole system,  rather than a macroscopic view of
the system. Is that correct?
  DR. GOODMAN:   I think our notion is that we don't want to get any more
microscopic than we have to, but where we see a process that Is influenced by
the dynamics of the pesticide itself, we do have to break it up.  That's why, for


example, we're hung up on the algal dynamics -- because of their tremendous
potential for cleaning things up or passing them up to higher trophic levels.
We plan to do much less detailed models of higher elements, even though they
are integrative in nature.

   QUESTION:  Have you perchance done any rough calculations on the number
of parameters you will want to do sensitivity analyses for?
   DR. GOODMAN: We've sat down and looked at that kind of thing and obviously
we're not going to be able to do total, global sensitivity analyses on all of these
parameters.  Many of these will be done on chunks of the model — on components
and then larger behavioral characteristics will be assembled to get the total
system behavior.

   DR. GLASS:   One thing Erik didn't mention in his talk is they've been
developing this model — what?  Two, three years under NSF funding?

   DR. GOODMAN: Well, one year under funding and two years with no funding.

                        Dr. James W. Gillett**

  Those of you who have had an opportunity to meet Jim Haefner and know him will
appreciate the humility with which I present his work.  Unfortunately, he was with us
only seven months, and I wish it could have been much longer.  The difficulty of pre-
sent whole ecosystem mathematical models has just been illustrated by Dr. Goodman's
•work.  Our problem is in some ways simpler and in some ways more difficult, to
try to represent mathematically the disposition of a pesticide within a terrarium.
Fortunately, we have  some good help on this, especially those people at Oregon State
University, who are deeply involved in similar work, and those at Athens Research
Lab:  Jim Hill, Jim Falco, Ray Lassiter,  and George Bailey. Discussions on the
pesticide  Runoff and Transport Model (Bailey et_al., 1973) have been very useful.
Unfortunately, their system is not applicable to our model because the intensity of
interactions within a microcosm changes the scope of work.
  In most systems —  I think Dr. Kenaga would support me on this — the  mass depo-
sition of a pesticide is not greatly involved with the biota.  The disposition within
the biota may become critical in relation to the total biotic structure.  Where the
mass of the pesticide  goes is not affected as much by the biota,  as by the media.
Within a microcosm,  within the confines of an intense,  closed system, the biota
become very critical.  Thus we have two sets of processes on which we are working
as parts of the model.  The first set is the so-called pesticide transport process
/pTP)» the second set is the biological control processes (BCP).  These  latter pro-
cesses with non-pesticidal inputs and outputs regulate functions which cross over
to pesticide transport processes and which receive inputs from pesticide transport
processes (Figure 1).
 *Co-author is James Haefner
 **National Ecological Research Laboratory, Corvallis Environmental Research
   Laboratory, Office of Research and Development,  EPA, Corvallis, Oregon


                                Figure 1
                                 _^__j	^
             Pesticide               TI    I              Outputs  of
              Inputs  	n   PTP   I	*- Pesticide  and
          /MlnPHts             |           |                Outputs
          (Nutrients,	^   BCP    	*- (Wastes, Growth,
            Energy)            U	_J          (Reproduction, etc.)
  m the real world it is likely that the PTP box is very large and the BCP box is
very small. In our microcosm we suspect that they are much more equal in size
and, hence, much more consequence can be attributed to the interactions between
the two.  Those arrows between the boxes constitute the system control process (SCP)»
We view this system control as a means of predicting the disposition of a pesticide
within the terrarium.  The purpose of doing the mathematical model is not to provide
a means of predicting the disposition of the pesticide in the real environment,  in a
total ecosystem or any realistic segment of an ecosystem, such as a crop land.  It
is to provide us with a means of analyzing the structure  of our terrarium and the
results from it.  We hope to understand the processes in the terrarium  and in the
model using this tool.
  A number of mathematical models have provided input into our concepts. The
Stanford hydrology model and the subsequent Wisconsin hydrology  models, which
led to the PRT, are an order of magnitude or greater in scale, so that they are not
directly applicable.  Also, the work done on soil columns, the very elegant work of
Lindstrom, Boersma and Hall, the work of Lichtenstein's group, and the work done
at NERL in Athens also are not directly applicable.  These are based on very hyper-
fine processes, highly resolved and relatively well understood, but they have no
direct relationship to the larger world involving biological processes and movement.

  The part of the system which is least understood of all in which we are involved is
the below-ground ecosystem.  It is very difficult to work in, partly because we can-
not observe it in the sense that we can observe fish in an aquarium.  We have diffi-
culty sampling it and difficulty constructing it.  That is, we cannot make a soil the
-way we can make artificial seawater, brackish water, or fresh water.
  m addition to the differences in the kinds of matter that are in soil, we do know
that the differences between the types of dissolved organic carbon  and the types of
particulate matter have great consequences on the function of that  soil system.
These functions  are affected by such potent environmental forces as temperature,
moisture, and cation availability. The controllers of these processes may be micro-
organisms or macroorganisms  (mites — the shredders; earthworms — the vertical
movers and distributors) which change the spatial, geometric distribution of these
materials and thus their biological availability within the system.
  Using McBrayer's model in general,  Dr. Haefner has constructed a similar model
 /Figure 2) which reflects most of the processes with which we are concerned, not

                                 Figure 2
i             	r._y _ _,. — ,

                               Figure 3
just in our terrestrial microcosm for pesticide studies, but also in microcosms for
litter decomposition involved in evaluation of coal-fired power plant siting.  Parti-
culate organic carbon divided into arbitrary size classes can be handled either sto-
chastically or strictly on a class basis. The movement of animal and plant organic
carbon, feces, carcasses, exoskeletons on one hand,  and leaves, stems, and root
exfoliates on the other into dissolved organic carbon is extremely critical.   Similarly,
we need to look at the fungal breakdown, particularly of the woodier materials (e. g.,
stems) (Figure 3).  The relationships between the fungi and fungivores and between
the bacteria and bacteriovores are very complex and not well understood.  However,
there may be means of discriminating between these different functions, using certain
optical techniques that Dr. Bruce lightheart has been working on in NERL that will
allow us to discriminate between live and dead bacteria and fungi, which in turn will
allow us to make  some estimates of active biomass.  All of the optical techniques and
chemical techniques in this area suffer in some regard, however.

                                     Figure 4
                 FOOD (From PLANT
                 FAUNA, SOILS, and
                 SOIL ORGANISMS)
                                          To  other. FAUNA
                                          	^_	f*	
   . j   Blood   (Hemolymphr£
                    To Above-Ground
               To Edge Subsystem
                Subsystem 4
                                       Figure 5
                                       From Above-Ground
             To Edge-S-
                                     To Edge Dissolved Surface
  Figure 5 illustrates the consequences of the biological control and pesticide trans-
port processes interaction,  the control process.  The terrarium has an "edge" and
a "center. "  Nematode concentrations at the edge may be two orders of magnitude
greater because  of greater moisture content and particulate organic matter,  providing
for greater bacterial growth.  In addition to the lateral zones, the model has vertical
zones:  "center above-ground, " "surface,  " "soil intermediate" and "groundwater. "

A new concept being explored is that of "isolated" and "non-isolated" soil.  "Non-
isolated" soil is interacting directly with the atmosphere; "isolated" soil is not inter-
acting with the atmosphere. The proportion of isolated soil may be temporal related
to the changes in moisture and heat content and to biologic activity. The concept of the
non-isolated and isolated soil may be a useful approach to assessing volatilization and
movement of materials from the soil.
   These compartments of the model represent concentrations for the parent chemi-
cal. For each metabolite or breakdown product,  a similar, parallel model is assumed.
The difficulty that has  not been surmounted by any modeler to date involves modeling
multiple inputs of the metabolite from  a given chemical into a parallel model of the
metabolite. It would appear to be  mathematically feasible, although elaborate and
frought with unknowns. In a number of instances, however, it is the metabolite in
which we are interested and not the parent compound.  While one has hope for deriv-
ing a mathematical model of the disposition of a pesticide, the problem  of integrating
the metabolites with the parent compound would seem to be a serious difficulty in the
future development of  terrestrial microcosm models.

                              John Bowser*
  This research program will be conducted under an interagency agreement
between EPA and USDA.  Funding for the agreement will come from the
Substitute Chemical Program and USDA.  Since the work is just getting started
at this time,  there are no experimental results to present.
  I will discuss the need for the research and why it's being funded under the
Substitute Chemical Program.  I will also outline what we hope to accomplish
through this work, how we plan to conduct the experiments, and why the work
is being done in cooperation with USDA.
  The work to be performed under this agreement deals with two separate
but interrelated problems.   The first is the potential for alteration in rates of
incidence of plant disease in crops which have been treated with pesticides..
The second problem involves possible alterations  in pesticide degradation
patterns  which can result from the use of various  pesticide combinations.
  Ill first discuss the work concerning the effect  of pesticides upon the
incidence of plant disease.   A number of  reports have appeared in the literature
which indicate that in some instances use of certain pesticides may result in
change in the incidence of certain plant diseases.  These reports include
descriptions  of instances in which severity of plant disease was increased
(1, 4, 5,  6),  as well as Instances in which the rate of incidence of plant disease
decreased concurrent to the use of certain pesticides (2, 3, 5, 6).
  These findings  are of interest to us since, in order to carry out the
Substitute Chemical Program properly, we must be able to anticipate what will
* Office of Pesticide Programs, Criteria and Evaluation Division, EPA

happen if one pesticide is substituted for another. We must have a way to pre-
dict whether the adoption of a new pesticide use pattern will result in a change
in plant disease patterns faced by the pesticide user.  That is, will the user
suffer greater or lesser losses due to plant disease,  or will there simply be no
change observed in the incidence of plant disease?
   We found the current literature lacking in regard to this problem.  First of
all, it is sparse.  For many of the compounds to be  evaluated under the Substitute
Chemical Program, there is no information of this type.  Secondly, the studies
which are available are generally laboratory or greenhouse studies which have
not been confirmed by field  observations; therefore, their value in determining
the results  which will be experienced by the user is unclear.  Thirdly, the
possible effects of pesticide combinations in this area have not been examined;
that is, the studies which have been conducted have been on single compounds,
not on combinations of compounds.
   In view of the problems which I have outlined above, we have designed a
research program which is  intended to meet the following goals.  First, we
would like to fill some of the data gaps In regard to the chemicals we're
interested in under the Substitute Chemical Program. Secondly, we'd like to
conduct experiments under both greenhouse and field conditions so that we'll be
able to correlate results observed in the greenhouse with results observed in
the field. Finally,  we'd like to demonstrate a relatively simple and inexpensive
procedure which can be followed by other researchers in the future.
   The initial work will be laboratory or greenhouse  studies of 20 to 25 pesticides
or combinations of pesticides.  These studies will include Investigations of the
incidence of plant disease, the effects of the pesticides on antagonists to the par-
ticular pathogens under study, and the effects of the pesticides on mlcroblal
activities which are normally associated with maintenance of soil fertility.  The
lab and greenhouse studies will be followed by studies of four or five pesticides
or pesticide combinations under field conditions.


   The crop-pathogen combinations, as well as some of the pesticides, which
 have been selected for these studies are listed in Tables 1 through 3.
                                 Table 1
Rhlzoctonia solani
Herbicides:    Dinitroaniline (trifluralin)
               Thtocarbamate (EPTC)
                  Table 2
 Phytophthora megasperma
 Phenolic dinoseb
Dinitroaniline (trifluralin)

 Thielaviopsts basicola
 Phthalate deriv.
 Aniline deriv.
                                   Plus phenolic dinoseb
                                   Plus thiocarbamate (EPTC)
                                   Plus phenolic dinoseb
      Plus aniline deriv. (alachlor)
      Plus benzole acid deriv.  (chloramben]
      Plus dinitroaniline (trifluralin)
      Plus triazine (metribuzin)
(trifluralin) plus phenolic DNBP

  As you can see, we have decided to concentrate on pre-emergence herbicides
 in this study.  The herbicides will be applied at rates consistent with normal
 agricultural practice.  Furthermore, we've made an attempt'in the selection of
 these pesticides to include representatives from the major chemical families.
   Dr. George Papavizas and Dr. Jack Lewis of the USDA's Soilborne Diseases
 Laboratory will be the USDA project officers for  this portion of the agreement.
  Dr. Donald Kaufman of the USDA's Pesticide Degradation Laboratory will
 be the project officer responsible for studies to be conducted under the second
 half of this agreement.  The work will be designed to detect changes in the
 patterns of degradation of various types of pesticides which occur when these
 pesticides are used hi combination with other pesticides.
  Dr. Kaufman's work will involve  initial laboratory screening of 20 pesticide
 combinations.  This initial work will then be followed by field tests on four or
 five pesticide combinations.  The compounds used for this work will Include
 representatives of major pesticide chemical families.
  The USDA was  selected to carry out this research primarily because of the
 excellent combination of expertise and facilities available at the Agricultural
 Research Center  at Beltsville,  Maryland. A contributing factor in our decision
 was the USDA's willingness to assume a portion of the financial costs of the

1.   Chandler, J.  M., and Santelmann,  P. W.:  Interactions of four
    herbicides with Rhizoctonia solani on seedling cotton.   Weed Set. 16; 453-456
2.   Cole, A. W., and Batson,  W. E.:  Effects of diphenamid on Rhizoctonia
    solani, Pythium aphanidermatum and damping-off of tomato. Phytopathology
     65_: 431-434 (1975)
3.  Harvey, R. G., Hagedorn,  D. J., and DeLoughery, R.  L.:  Influence of
     herbicides on root-rot in processing peas.  Crop Sci.  15:67-71 (1975)

4.  Perich, J. A., and Lockwood, J. L.:  Influence of atrazine on the severity
    of fusarlum root rot in pea and corn.  Phytopathology 65; 154-159 (1975)

5.  Richardson, L. T.:  Effect of insecticides and herbicides applied to soil
    on the development of plant diseases.  I. The seedling disease of barley
    caused by Helminthosporium savitum P.K. and B.  Can. J. Plant Sci.  37:
    196-204 (1957)

6.  Richardson,  L. T.: Effect of insecticides and herbicides applied to soil
    on the development of plant diseases,  n. Early blight and fusarium wilt
    of tomato.  Can J.  Plant Sci. 39:30-38 (1959)
  QUESTION: Which factors are you going to examine in your soil fertility

studies ?
  MR. BOWSER:  Well be concentrating on microbial activities such as
ammoniflcatlon and nitrification.  We feel that these indicators are more
valuable than simple  estimates of microbial populations  in the soil.

                               Dr. Riz Haque*

   I would like to point out that this program has just been started, so I don't have
any scientific results to report,  but I will describe the rationale and conceptual scheme
behind this program. I shall begin with Regulation Three,  the Federal Insecticide,
Fungicide, and Rodenticide Act, as amended, which states that a pesticide,  in order
to be registered, must perform its intended function without unreasonable adverse
effects on man or the environment.
   This is one of the most important regulations as far as the registration of pesti-
cides is concerned.  During 1974,  under appropriation PL 93-135, the Substitute
Chemical Program was started.  The objective of this Program is to find substitutes
for problematic pesticides.  The registration of pesticides or the finding of substi-
tutes for undesirable or problematic pesticides requires a critical evaluation of data
from various disciplines and the establishment of the effects of pesticides on man and
environment.  Some of the factors used in the evaluation of pesticides are 1) physical
and chemical properties,  such as water solubility, vapor pressure,  melting point,
boiling point, and structure; 2) toxicology — we have to know about acute and chronic
toxicity; 3) metabolism — what the metabolites of the particular pesticides in soil,
plants, or animals are; and 4} fate and behavior — the transporting soil, bicaccumu-
lation, residues in water, air,  and soil.
   However, the problem of defining an unreasonable adverse effect on the environ-
ment is extremely difficult.  Furthermore, other factors,  such as the mere number
of pesticides and metabolites and the existence of various formulations for any parti-
cular pesticide, make this task  very tedious and cumbersome.  The introduction of
simulated modeling techniques has given a new approach to the study of behavior of
pesticides in the environment; however,  we still do not have any practical model.
*Office of Pesticide Programs,  Criteria and Evaluation Division, EPA

   One of the most important questions which we are faced with now is whether we can

somehow simplify the evaluation process for pesticides.  The purpose of this bench-

mark chemistry program is to qualitatively predict the behavior of a pesticide from a

knowledge of the chemodynamic parameters  of selected benchmark pesticides.

   The words benchmark and chemodynamic are relatively new, so I would like to

define these terms.  Benchmark is the name given to a group of pesticides selected

from  each important class, based on the structure and use pattern.  The idea behind

benchmarks is that if we know the behavior or properties of one or two pesticides from
each important class, then we can predict the behavior of any chemical on a quali-

tative basis.  I would like to point out that this list (Table 1) is just selected randomly.

on the basis of structure, and there is no special reason for any particular compound

to be included.  We hope that if you know two or three chemicals from each class,

then you can predict the behavior of others from that class.

                Table 1: Suggested Benchmark Pesticides

  Chemical Class                      Pesticide
  Aliphatic Acids
  Benzole Acids
  Phenoxyalkanoic Acids
  Quaternary Salt and Picolinic Acid

  Natural Pesticides
  Aromatic Nitro Compounds
Chloropropham, Alachlor
Dicamba, Chloramben
Carbaryl, Carbofuran,  Methomyl
DBCP; Methyl Bromide
Lindane, Dleldrin, Methoxychlor,  Mirex
MSMA,  Methyl Mercury, Tributyltin
Malathion, Diazinon, Methyl Parathion,
Phorate, Fensulfothion Chlorpyrifos,
DDVP, Leptophos
Dinoseb, PCP
Picloram, Paraquat
Atrazine, Simazine
Diuron,  Monuron
Dichlobenil, Benomyl, Altosid,  Bifenox,

   As you see, this represents aliphatic acids,  carbamates, dinitroanilines, organo-
phosphates, organometallics, phenols, triazines, ureas, and so on.  If we take into
account the structure and use pattern, these may represent 80 percent of the pesti-
cides used  in the United States.
   Chemodynamics is a field which describes the use of physical chemical properties
and the processes associated with them in predicting the behavior of a chemical in
various parts of the environment. For example, if we want to predict the behavior
of a chemical in the air, then we should know the vapor pressure and  vapor loss.  If
we want to  know the behavior in water, we should know solubility, heat of solution,
and half-life for hydrolysis. Similarly, for soil we have to know adsorption/desorp-
tion, leaching,  and degradation. For biota we  measure the octanol/water partition
coefficient  or fish bioaccumulation.   Then hi order to account for the  total degradation
we study photodegradation.  These parameters can give us a rough estimate of the
physical as well as biological behavior of pesticides.
   Once we have the parameters, we put them in a matrix form and try to build up
an environmental profile.  From this profile we can predict the behavior of each
class of pesticides.  Figure 1 is a schematic diagram of how we screen a pesticide
through this benchmark program. We have a large number of pesticides from which
we select the benchmarks.  On the benchmarks we carry out the chemodynamic studies
and then build up environmental profiles.
                               Figure 1
                         Environmental Profile

                                    81 .

   In Figure 2 we have the various chemodynamic parameters,  such as water, air,
 soil, biota, soil degradation, and photodegradation, from which we build up the
 profile.  The idea behind the benchmark program is to obtain the benchmarks, measure
 the chemodynamic parameters, put them in the matrix form, and then, from the
 principle of pattern recognition, build up a stereotype profile for each class of com-
                             Figure 2
                       Benchmark Pesticide
                          Soil Degradation
   This program is divided into six parts, and it is expected that the work on each
part will be carried out in different labs.  It's very unlikely that we could find one
lab which is fully equipped and has expertise in all these areas.  For the air work,
we will measure the vapor pressure, vapor loss from soils, and effect of temperature
and moisture on vapor loss.  For the water part, we will measure solubility, dis-
sociation constant, hydrolysis,  effect of temperature, and pH.  For soils, we will
measure the adsorption, desorption, leaching, effects of temperature and moisture,
and pH.  For biota, we will measure the  octanol/water partition coefficient and fish

   It has been shown by Dow scientists that the octanol/water partition coefficient
could be used as a guide for accumulation of pesticides. We propose that, if you
measure the octanol/water partition coefficient,  you can tell the degree of accumu-
lation of the pesticide in a living system.
   For soil degradation we  measure the rate of degradation in soils and identification
of metabolites.  In photodegradation we study the effect of sunlight and also the identi-
fication of products.  These are the various factors which are involved in the bench-
mark concept, and we expect that in each part we'll be studying the same pesticides
under similar environmental conditions,  so that  results obtained from one part will
be comparable to results from the others.  We also hope to study the same number
of pesticides and similar pesticides, so that we won't have a gap when we try to put
all the parts of the jigsaw puzzle together to get  the profile.
   Secondly, we  should use the same grade of pesticides.  This is necessary in order
to observe any surprising effect of the impurities.  Third, similar analytical methods
should be used as far as possible.  These may include spectroscopic techniques, GLC,
and radioactive tracer methods.  Fourth, quality assurance — we should make sure
that all the results have similar analytical precision.  Fifth, the studies  should be
carried out with active ingredients as well as technical formulations.   In the litera-
ture, most of the work on pesticides has been done with active ingredients; whereas
we use pesticides in field as formulations, either as wettable powder  or dust.  There-
fore, we should make sure that by changing from a pure chemical to a formulation
these parameters are not affected significantly.  Finally,  we need good communica-
tion — among all these parts we expect that there'll be open exchange of information.
   The benchmark chemistry program is going to be divided into two phases.   Phase
one, which involves the evaluation of literature, is being carried out in cooperation
with Ms. Pulliam, and this work has been contracted to George Washington University.
We expect the work to be completed shortly.

   With George Washington University we have identified several experts in this
country who are especially suited to carry out the work.  As consultants, they are
helping George Washington University evaluate the literature work.  Dr. Spencer is
supplying information on air work; Dr. Freed of Oregon State University, on water;
Dr. Farmer of University of California, Riverside,  on soil adsorption; Dr. Metcalf
of the University of Illinois, on bioaccumulation and octanol/water partitioning;
Dr. Kaufman of USDA, Beltsville, Maryland, on soil degradation; and Dr.  Crosby
of University of California, Davis, on photodegradation.
   Phase two is the actual  lab work, which is barely underway.  Only two parts, the
air part and soil degradation part, have been signed under interagency agreement
with USDA.  Dr. Spencer is carrying out the vapor pressure determination and vapor
loss  studies,  while Dr. Kaufman is carrying out soil degradation studies, including
the identification of metabolites.  The other parts  — water, soil, biota, photo-
degradation, and modeling — will be starting shortly,  and we expect to have some
input from the Office of Research and Development and the Office of Toxic  Substances.
   Finally I'd like to summarize the scope and advantages of this benchmark program.-
Number one,  it would be used in the Substitute Chemical Program.  For example,
if you have to find a substitute for chemical X, then from  the use pattern you may
find that there are a thousand or maybe several hundred chemicals available as sub-
stitutes.  By comparing the profile of this compound X with all these chemicals on
the benchmarks, you may be able to eliminate many of them and so you can concentrate
on only a few chemicals.  In this way, use of the benchmarks can simplify the Substitute
Chemical Program.
   Number two, benchmarks may be utilized in the registration process. For example,
if you have a new chemical for registration,  you are provided various data by the
registrant. You can compare the profile of this chemical with the benchmarks and
find out where any potential problems may exist.  I must point out that this benchmark
concept would not replace the data requirements but simply assist the evaluation

   Number three, from the benchmark program we'll find out where these chemicals
will concentrate in the environment.  This should assist in planning long-term moni-
toring and toxicology studies.
   Number four, this benchmark program can be applied to other toxic  chemicals
and, in this respect, I would like to point out that we are helping the National Science
Foundation. The Foundation plans to employ this benchmark concept for other toxic
   Number five, we are going to carry out these studies on many compounds, and
I'm sure we'll find some interesting research problems.
   Number six,  most of the experiments described in the benchmark program are lab
work and are of relatively short duration,  so we should get this information quickly.
   And number seven is, for my own bias, simplicity.
   COMMENT: After listening to this dialogue on benchmark chemicals, I don't
know why we have to submit any data for registration of a compound.  You plan to
do it all over again,  and these data have already been taken care of quite adequately.
   DR. HAQUE:  You have a good point there, and we have given consideration to
this point.  If you look in the literature, you will see that this work has been carried
out in different labs, under different environmental conditions.  It is very difficult
to compare the results.
   COMMENT:  If someone wants to redo all the degradation, all the literature
work,  and all the experimentation we've done in the field, it's not going to be done
simply.  It takes a long time, in excess of six,  seven years and about $8 million
to get a pesticide in the marketplace,  and they're not all. the same.
   DR. HAQUE:  I would like to point out that if you have followed my talk, you
will note there are two phases of  this program. Phase I is the literature work.

 From the literature work,  if we are convinced that the data are available, then we
 are not going to do experimental work.
    COMMENT:  As far as the compound that we have that's  in review by the
 Substitute Chemical Program, everything I saw on that slide is available.
    MS.  SCHUTH:  There are a lot of areas where data are not available,  like
 water solubility and vapor pressure.  When available, there  is standardization of
 measurement techniques. We're not talking about going  out and repeating all your
 studies. We're just talking about defining five or six parameters and having them
 all measured in the same way,  so that these values will be useful in our models
 and will be more valuable for comparative purposes.
    QUESTION:  I question one other thing. I can agree with your point on water
 solubility.  It's 20, 25,  30,  35 degrees centigrade, and volatility under a variety
 of conditions,  I can understand this, but I can't say that the  benchmark chemical
 is going to satisfy  and be comparable in the total series of compounds that belong
 in the same class. I think the literature points that  out very clearly. They metab-
olize  somewhat differently.  They have different products.
   DR.  HAQUE: Therefore, we have more than one compound in each class, and
we are saying that, if you look at, for example, chlorinated hydrocarbons,  then
you have the problem of bioaccumulation. We are trying to  identify the problems
in each  class,  or various characteristics of types, and we are not  really going  to
do all the studies.  There are just a few parameters we are going to be studying
under exactly similar conditions.  Some of the data in the literature are really not
very good.
   QUESTION:  I wondered  if you'd considered how you would approach this with
reference to benchmarks when you get a compound and have several different
functional groups related between classes ?
   DR.  HAQUE:  Then we may have to compare several benchmarks from the
several  groups,  for example, one compound  may fall in two or three classes.
We may have to look at different classes.

                        FOR HERBICIDE PHYTOTOXICITY
                               Dr. Fumihiko Hayashi*
  An Important part of the Substitute Chemical Program is to select,  evaluate,
and test pesticides that may be used as alternates to certain registered products.
Phytotoxicity of pesticides is currently identified and measured by observable symp-
toms which may or may not be the result of biochemical disruptions at cellular
levels.  Such physiological or physical Injury can be easily discerned and Is not the
subject of this research project.  Rather I would like to address the subject of how
to measure invisible phytotoxicity resulting from pesticide usage.
  Symptoms displayed by plants from biochemical disruptions are results of bio-
chemical stimuli at cellular levels and may result in alterations of plant growth
and development.  Effects of pesticides on plants have not been measured in both
susceptible and tolerant plant species at cellular and molecular levels,  therefore
we have little or no biochemical Information as to why susceptible and resistant
plant species exist.
  Basically a herbicide Is a chemical which may be used to kill plant species grow-
ing In places where they are not wanted.   Therefore most herbicides are phytotoxlc
substances.  In view of the Importance of pesticides for quality and quantity of food
and fiber productions and for safety of use, it is apparent that a better understanding
of modes of action of pesticides  In both susceptible and tolerant plants Is needed.
Such an understanding could lead to a rapid and reliable screening procedure for
substitute herbicides and plant growth regulators.
  With this background, I shall  review with you our project #68-01-2482, which has
for Its study objective the rapid  Identification and measurement of biochemical
""Criteria and Evaluation Division, Office of Pesticide Programs,  EPA

 changes in plant cells due to exposure to herbicides.  The Department of Botany,
 University of Michigan, was awarded this study contract.  Three professors,
 three technicians,  and five  graduate assistants have been engaged since July 1974
 on this project.  We in the  Criteria and Evaluation Division of the Office of Pesti-
 cide  Programs have set the following criteria for the rapid screen procedure.
   The scope of work is to positively  Identify three or more modes  of action
 simultaneously,  including inhibition of mitochondrial respiration or electron
 transport, inhibition of protein synthesis controlled by DNA-dependent RNA,
 and inhibition of photosynthesis and Hill reaction.  The acceptable  biochemical
 method must be rapid,  reliable,  and  accommodate the following: 1) detect pos-
 sible simultaneous presence of the above modes of action in a single susceptible
 plant species; 2) provide reliable data within 48 hours; 3) be reliable at the 95
 percent level; 4) allow multiple,  simultaneous analysis.
  DR. SIKKA:  Dr. Hayashi, what other plant species could be used?
  DR. HAYASHI: So far, we are using one plant species, soybeans.
   Table 1 shows common names  of herbicides which are used In this research
 project.  As you see, this is a chemical group classification of herbicides:
phyenoxys, phenol, arsenical, blpyrldyllum, carbamates,  thlocarbamate,  dlnitro-
 aniline, triazines,  triazole, ureas, and  uracil.
   Figure 1 shows a schematic presentation of electron transfer and energy
transfer systems in plant mitochondria.  Plant mitochondria are believed to con-
tain two electron transport chains.  The main phosphorylatlve chain Is Inhibited
by cyanide terminally and by antimycin A (abbreviated AA in this figure) in the
middle, and a non-phosphorylation chain withdraws electrons from the flavo
protein (abbreviated as FP in the figure) level to oxygen via a flavo protein and
a second oxidase.  The second chain is cyanide- and antimycin A-resistant.
The exact nature of the cyanide-resistant chain has not been clarified yet,  but
it  is  estimated to consist of about 25 percent of total respiratory activity of mung

                                       Table 1:  Herbicides Studied
Chemical Group
Arsenical s
§ Thiocarbamate
Dinitro aniline
Common Name
Diuron (DCMU)
Chemical Name
(2, 4-dichlorophenoxy) acetic acid
(2,4, 5-trichlorophenoxy)acetic acid
Disodium methylarsonate
Hydroxydimethylarsine oxide
6, 7-dihydrodipyrido (l,2-a:2',l'-c) pyrazinedium ion
Isopropyl-m_-chl orocarbanil ate
£-(2 , 3-dichloroallyl)-diisopropylthiocarbamate
o- , a , a -trifluoro-2, 4-dinitro-N, N-dipropyl-p-toluidine
4, 6-dinitro-£-cresol
3-amino-i, 2, 4-triazole
3-(p-Chlorophenyl)-l, 1-dimethylurea
3~(2f, 4-Dichlorophenyl)-l, 1-dimethylurea
tert-butylcarbamic acid ester with
K & K
K & K
K & K
                                           3-(m -hydroxyphenyl)-!, 1-dimethylurea

 bean mitochondria.  The main chain is coupled to the energy conservation pathway
 of ATP formation at the three sites.  Although the exact mechanism by which ATP
 is formed is not elucidated yet, it is believed to go through a so-called "high-energy"
 intermediate state.
                                 Figure 1
                                                              02  (CN~ sensitive)
(CN" insensitive)
   The current assay techniques for studying possible inhibitions by herbicides
or mitochondrial activity depend heavily on measurement of oxygen consumption.
Oxygen uptake is started by an addition of a substrate, and it is further stimu-
lated by addition of ADP to the reaction medium which contains inorganic
   Tables 2 and 3 show the results of 15 herbicides tested on respiratory
activity.  Seven herbicides gave no effect at the starting minimal concentrations
on mitochondrial activity.  Bromacil, chlorpropham,  dluron,  DNOC, and
propham only showed a slight effect at mM concentrations two orders of magnitude

              Table 2: Effects of Weakly Active Herbicides
                        on Mitochondrial Activity*
Ki (mM)
ca. 5
— — -
ca. 0.7
ca. 0.6
Max. inhibition 37% at 10 mM
80% inhbition at 10 mM
Max. inhibition 40% at 10 mM
80% inhibition at 10 mM
Max. inhibition 40% at 10 mM
85% inhibition at 10 mM
Slight uncoupling at 0.3-1 mM
90% inhibition at 3 mM
Max. uncoupling at 0. 3-1 mM
Uncoupling at 1 mM
Max. inhibition 50% at 3 mM
Max. inhibition 75% at 3 mM
Max. inhibition 95% at 3 mM
Max. inhibition 95% at 3 mM
* Effects of weakly active herbicides on mitochondrial activity.  Various
  concentrations of a herbicide are given at the second state 4, and the
  effects on that state or on the subsequent state 3 rate (induced with about
  150 f*M ADP) were examined.  Substrate concentration used:  10 mM
  succinate or 30 mM malate.  The Ki value refers to the concentration of
  a herbicide which inhibits 50 percent of the second state 3 rate.

          Table 3:  Effects of Herbicides on Respiratory

               Activity of Isolated Corn Root Tips*


Effect on
<>2 Uptake
55% down
45% up
C0£ Evolution
0 •
157% up
59% up
85% up
21% down
203% up
75% up
* Effects of herbicides on respiratory activity of isolated corn root tips.
 0 denotes no significant effect of herbicide; up_ signifies percent increase
 above control; down means percent decrease.  Control values are 570
 nmoles/min.'g fresh weight for oxygen uptake and 605 nmoles/min.'g
 fresh weight for carbon dioxide evolution.

                Table 4:  Effects of Herbicides on Mitochondrial Activities,  Using an Oxygen Electrode
Concentrations (mM) Tested Which Showed
Electron Energy
No Effect Uncoupling . Transfer Transfer
Inhibition Inhibition
3.0 — — 	
	 	 > 0.05*
0.001-1 >1
> 3. 0 1-10
	 > 0.02 > 0.01
Substrate Used
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
Succinate and Malate
       * Diallate titration curves are different for succinate oxidation and malate oxidation.

  As exemplified by DNOC which acted to uncouple at low concentrations and
caused incomplete inhibition of electron transport at high concentrations,  the
herbicides which affect mitochondrial activity at the  mM level may affect  other
metabolic activities more potently.  It is of interest  to note that the inhibitors
of the Hill  reaction, such as atrazine and simazine,  do not affect mitochondrial
activity at  a mM concentration, 'while monuron,  an inhibitor of photosystem II,
induces particle uncoupling at a high concentration.   Another frequently used
inhibitor of photosystem n, diuron,  inhibits respiratory electron transfer at the
high concentration.
  These results, together with insensitivity of the chloroplast system to anti-
mycin A, oligomycin and DNP, can  be considered to indicate that the nature of
electron transport and coupled phosphorylation is different in mitochondria and
  Of the 15 herbicides tested,  three gave no effect in the oxygen evaluation
assay at a  mM concentration range.   All the rest inhibited electron transfer at
concentrations less than one mM. Most potent inhibitors of electron transfer
are bromacil, followed by atrazine,  karbutilate, monuron, simazine, and tri-
fluralin. Chlorpropham and propham are relatively  ineffective, requiring con-
centrations one order of magnitude  higher than the above group.
   The results from the herbicidal effects on the oxygen consumption assay
show that most herbicides do not affect the photosystem I activity at mM  con-
centrations.  It can be concluded that most herbicides inhibit the photosystem
H activity.
   Of the 15 herbicides tested, four  herbicides have  an inhibitory effect on
DNA synthesis and five herbicides inhibited protein synthesis when we tested
using   Oleucine plus 3H-thymidine or 14C-uridine incorporated into DNA  and
RNA fractions of corn root tip cells.

    Table 5:  Effect of Five Herbicides on Incorporation of Precursors

        into DNA,  RNA, and Protein Fractions of Corn Root Tips*
Herbicide Used






% of '

1, 300
Pro tain

*Twenty-flve root tips were Incubated with *4 C-leuclne + 3H-thymtdlne or *4C-urldlne for
2 hours and fractionated Into DNA, RNA, and protein with the enzymatic method described
In the last quarterly report. Incorporation In CFM Is an average of duplicate runs ex-
pressed on the basis of 25 root tips.

       Table 6:  Effect of Some Herbicides on Incorporation of Precursors

            into DNA,  RNA,  and Protein Fractions of Corn Root Tip Cells*


Herbicide Used





2x 1(T5
2 x 10"5

2 x 10*4
2 x NT6
2 x 10"5











*Ten root tips were Incubated with 14C-leucine +• 3H-thymldlne or l^c-urldlne for 2 hours
and fractionated into DNA,  RNA, and protein as described in the text. Incorporation In CPM
is an average of duplicate runs expressed on the basis of 10 root tips*

  We also examined morphological changes of plant material with or without
herbicide  application.  An electron microscopic study is important for evaluating
physiological and chemical modes  of action.
  Figures 2 and 3 show some results of our morphological experimental work.
When we treat with 10 micro trifluralin,  the number of microtubules along the
cell wall was reduced.  Figure 2 is a water control for the spindle fiber.
                              Figure 2
             Untreated Cell Wall (Water Control) 17,000 X


              Figure 3
 Treated with Trifluralin 10 mM, 17, 000 X
             Figure 4
Untreated Spindle Fiber (Water Control) 17,000 X


  After treatment with trifluralin the spindle fibers disappeared completely
(See Figures 4 and 5).  Figure 6 shows a water control of bromacil.  When we
apply  10~6M of bromacil, you can see the enlarged mitochondria and the donut
shaped mitochondria (Figure 7).  With 10~6M treatment we sometimes cannot
see the biochemical response, but we may be able to detect some mitochondrial
or morphological changes using an electron microscope.
                             Figure 5
                 Treated with Trifluralin 10 mM, 17,000 X

              Figure 6
Untreated Mitochondria (Water Control) 25,000 X
             Figure 7
   Treated with Bromacil 10~6M, 25, 000 X

  It is possible to evaluate the mode of action of the herbicide by the morpho-?
logical changes.  From the various previously described experimental work,
we concluded that the following screening methods will be reasonable and reliable
for phytotoxicity.
  Once more I'd like to explain the flow of electrons from water to carbon
dioxide in plant cells.  Figure 8 is a schematic chart of how to metabolize from
water molecule to carbohydrate,  or some other carbon compounds with carbon
dioxide. This is also called the Hill reaction.  And as far as we studied,  many
herbicides are effective to photosystem II.  We can evaluate mode of action for
those herbicides by the changes of electron transport systems.
  Our proposed evaluation method of phytotoxicity has four steps. For post-
emergence herbicides we put the whole intact plant into the chamber.  We then
apply the herbicide either on the top or the root.  We'll measure carbon dioxide
using the IE gas analyzer.  The root growth inhibition will be measured by ER gas
analyzers and oxygen electron.  In the case of pre-emergence herbicides,  we put
the seed on a petri dish,  keeping It In the chamber, and measuring carbon dioxide
evolution using an IR gas analyzer (Figures 9 and 10).
  I'll discuss the second step later.  The third step is to measure respiratory
changes in mitochondrial suspension solution with or without herbicide application.
Step four is to check for morphological changes using the electron microscope.
  How we are currently  constructing the Instrument Is diagrammed In Figure 11.
This is the second step.   Our preliminary experiments in which a leaf was held
between two color filters indicate that the changes of leaf fluorescence can be
very easily detected visually.  The construction of the box which houses a leaf
sample and two  color filters must be completed before obtaining qualitatively
comparative data for screening various herbicides.
  The instrument for the instant evaluation method has two filters.  The primary
filter has 450 mM absorption, and the secondary filter is 600 mM. We put a  sample
leaf between the two filters and measure fluorescence intensity.

                                    Figure 8

                            Electron Flow System In Chloroplast
                                                       (FROM MEDIUM)
                        »02+4H" (CYCLIC)

              PAST EMERGENCE
               PRE EMERGENCE
                                       Figure 9
                 AND RESPIRATION)

                                             O.ECO,  ELECTRODES (ANALYSIS
                                                          OF RESPIRATION)

                                             K+ ELECTRODE (ANALYSIS OF
                                                        ION TRANSPORT)



                               Figure 10







                               Figure 11

               LP —LIGHT POWER SUPPLY
                L --LAMP
               hf —HEAT FILTER
               pf —PRIMARY FILTER (CA.  «0mu)
               sf —SECONDARY FILTER (CA. 600mu)
                R —RECORDER

  I hope someday the basic concept of our screening technique using the electron
transfer system or respiratory system can be used in establishment of an evalua-
tion method for all biologically active chemicals, including pesticides.  Toxicity
to mammalian and plant cells would be evaluated by changes of electron transport
at the molecular level using some unit of evaluation instead of LDso as at present.
I believe that the electron transfer system will be a useful screening technique for
evaluation of substitute chemicals among the pesticides.


                            Carter Schuth*

   I'd like to thank the Office of Research and Development and the Office of Pesticide
 Programs for this opportunity to advertise the environmental fate and effects testing
 program of the Office of Toxic Substances.  As you may know, the several versions
 of the Toxic Substances Control Act currently pending in the Congress all have pro-
 visions for EPA to require certain types of test data from industry on chemicals
 and chemical classes in order to determine if a chemical substance presents an
 unreasonable risk to health or the environment.  The task for us is thus not the testing
 of chemicals but the description and interpretation of adequate test methodologies,
 the results of which will serve as bases for decision-making as to the need for regu-
 latory action by EPA.  I would like to discuss with you this morning the part of our
 program which is directed toward environmental fate and effects testing, the objec-
 tive of which is the "evaluation and demonstration of methods for environmental
 testing of chemicals."
  As  an aside,  let me elaborate  a bit on the significance of environmental fate
 and/or effects.  For reasons which are probably best described as bureaucratic or
 for the sake of simplicity and organization, we tend to deal with fate separately from
 effects, especially in the discussion which is  to follow.  More emphasis seems to be
 placed on fate in our program at this time.  This is due to our feeling that an under-
 standing of environmental fate is essential in  arriving at some  estimate of the prob-
 ability of observing (and  causing) environmental and human health effects, i.e.,
 exposure.  A need crying to be filled is the development and demonstration of reliable
 and simple screening techniques for these effects.  Bear in mind that Bob Metcalf's
 original idea was to develop in his model ecosystem a "white rat" for the environ-
 mental toxicologist as an appropriate laboratory model.  It is not sufficient to estimate
*Offlce of Toxic Substances, EPA

fate without some attention to possible effects.  I would strongly encourage the behav-
ioral work of the Gulf Breeze Lab and others as a promising enterprise. In addition,
I'd like to pass on a comment brought to my attention by John Buckley.  We can all
collect reams of data from any given model, but as regulators and the regulated,
what is the information we need and, very importantly, what is the information we
don't need?
   Now, the return to my topic — I wish to discuss with you the environmental testing
activities of the Office  of Toxic Substances.  In support of this objective is a contract
effort involving three separate but related activities—methodology assessments;
demonstration projects; and chemical- or problem-specific information services.
Whereas I want to concentrate my comments on the second topic, I will briefly mention
the other two.
   A number of methodology assessments have been completed or are in the final
stages of completion.   These are state-of-the-art reviews involving a compre-
hensive literature search and review, interviews  involving a comprehensive literature
search  and review,  interviews with specialists in the various fields, and evaluations
by the contractor as to the usefulness and convenience  of the test methodologies.
The following topics comprise a listing of the reports that have been undertaken by
this program:
  •    Survey of industrial test practices
  •    Persistence  and routes of degradation of chemicals in the environment
  •    Transport of chemicals in the environment
  •    Methods to assess the effects of chemicals on terrestrial animal species
  •    Methods to assess the effects of chemicals on plants
Most have been completed and are available through the National Technical Information
Service in Springfield,  Virginia.  I have prepared a handout with the exact title,  an
abstract, and information on ordering the reports.  If there are insufficient copies
of this for those who are interested, I will be glad to mail this information to anyone
who gives me his or her name and mailing address at the end of this session.

  In addition to information on the methodologies for determining the environmental
fate of chemicals,  we are sometimes interested in knowing what information is avail-
able on the fate or effects of specific chemicals.  We are also interested in keeping
abreast of current developments in the various methodologies.  For these services,
we turn to task order contracts which can produce information  and reports on a very
short schedule.
  Several reports  under the category of chemical-specific information services are:
   •  Environmental fate of selected carcinogens
   •  Benzene: environmental sources of contamination, ambient levels, and fate
   •  Environmental fate of selected PNAH's
   *  Environmental fate of selected aza-arenes
This information is gathered from the published literature, scientists working the
field, and educated guesses on the part of the contractors.  To date we have investi-
gated the environmental fate of six carcinogens (as defined by the Occupational Health
and Safety Administration) and benzene.  Ongoing are studies of the environmental
fate of polynuclear aromatic hydrogens including some nitrogen heterocycles.
  I'd like to turn now to the part of our program that involves model systems
directly.  In the following list of demonstration projects, the studies listed under
parts 1 and 2 have been started under FY-75 funding.
       1.  Reproducibility and variability
           Modular food chain
           Model ecosystem
       2.  Validation studies
           Model stream
       3.  Physical-chemical measurements
           Octanol/water partitioning coefficients
           Water solubility
           Vapor pressure
The third category is proposed for this year.

  Among the many criticisms of laboratory models for determining environmental
fate and/or effects is the recurring charge that we know very little or nothing about
the reproducibility and variability of the various systems.  In response to this, we
have undertaken the two studies shown in part 1.  Dr. Thomas Johnson of the Fish
Pesticide Laboratory in Columbia, Missouri has been working for several years
on what he has called a "modular food  chain." This model system appeals to us as
being relatively uncomplicated,  inexpensive, and indicative of actual environmental
conditions.  I need not mention that concentration of toxicants in food chains is of
primary concern to EPA.  In order to  assess the modular food chain's utility in a
regulatory mode, we have signed an interagency agreement with the Department of
the Interior for the Fish Pesticide Laboratory to investigate the  reproducibility and
variability of the modular food chain using four chemicals and various concentrations
of these chemicals.  By analogy, the model terrestrial-aquatic ecosystem first
described by Metcalf et al. has been met with less than universal acceptance. We
have signed a contract with the Illinois Natural History Survey to undertake a study
of the reproducibility of that model using the same four compounds. , In this  way, we
hope to get good comparisons of each system.
  The intent of this project is to better understand  the two systems rather than
generate new information about potentially hazardous chemicals.  In fact, it can be
viewed as the study of two simple and inexpensive systems using "benchmark"
chemicals.  In addition to studying the reproducibility of each system with respect to
the  same chemicals, we are interested in assessing the utility of each system with
respect to certain  types of chemicals.  So, we have set up  criteria for the selection
of chemicals in order to achieve this comparison.  We wish to look at compounds
which display a range of C-14 isotopes as well as being well-characterized as to
possible metabolites.  We have selected decachlorobiphenyl, hexachlorobenzene,
di-2-ethylhexyl phthalate, and atrazine.
  In addition to wanting information on the reliability of the simpler models, we ask
ourselves what can they tell us about the "real world."  Dr. Howard Johnson of the
Michigan State University has signed a contract to study the fate and effects  of two


of the previously mentioned chemicals in model stream communities.  These model
streams have been set up in a coverted fish hatchery in Paris, Michigan.  They are
supplied with communities of organisms and pristine water from a stream which
has been diverted to flow through the concrete channels in the hatchery.  Dr. Johnson
and his staff will determine the fate of the chemicals in the various trophic levels
of the model streams  as well as observe effects of the chemicals on the organisms.
Our rationale in selecting the model streams involved the belief that field studies
are both very difficult and expensive.  Analogies of the model stream results and
monitoring data will be possible and will be useful in validation of our models and the
environment we strive to protect.
  It is our intention as regulators to require of industry only the necessary infor-
mation on its products that is needed for responsible decision-making.  It has been
suggested to us that these model systems are too complicated and that the measure-
ment of simple physical and chemical parameters gives the  same information.
This is an area that we are now investigating and feel sure will fit into the program
that I've described to you this morning.
  The last category encompasses the utility of physical and  chemical parameters
to indicate the environmental fate and transport of chemicals. As we discussed
yesterday, these numbers are useful in the utilization of models being developed as
well as for comparisons  with "known" compounds.


                           Dr. James Hill*
                          NOT AVAILABLE


* Environmental Research Laboratory, Athens, Georgia


   DR. BUCKLEY: We may have a fairly lively discussion over the next hour or two,
and I don't really have any ground rules as to how vigorous or rough or contrary wise
we ought to be.  I have a number of observations to make,  and some of them are per-
haps designed to arouse a certain amount of controversy.  Some of them  are intended
to see if we can figure out some ways in which we can do business a little bit better.
   Some of them may be controversial in a sense because they rather run counter to
 the traditional ways in which science is done, of trying to understand all about every-
thing ,  and I'd like to turn some of that around.  It's true that I'm in the  research part
of EPA and not in the regulatory part, but I am constantly and continuously made
aware  of the fact that my only reason for existence is to serve that part  of the Agency
which is concerned with regulation. So it seems to me there are some general
questions that we need to think about.
   Carter Schuth mentioned this morning that we need to somehow or another concen-
trate on what it is that we need to know. To turn that the other way around,  what are
the things that we can get along without knowing? At the San Antonio conference on
 evaluation of chemicals In the environment — I dont have a copy with me, and my
quote may not be precise — I pointed out, and I would emphasize again now, that the
 game we're involved in really is trying to acquire the least amount of  Information
necessary as a basis for decisions. It's not how can we find out as much as possible
 about a particular substance or what happens.  I realize that not all of us have the
 same crystal balls, and that if we do look in the crystal ball, we may  not all see the
 same Image.  I think we  really ought to be thinking as deeply as we can  about what
 things about chemicals and their interactions with organisms are-the important ones
 that are necessary so that sensible judgments can be made.

   There's kind of an irreverent question that I keep asking people when they propose
 a different kind of research.  It is,  "What can I do with that piece of information if
 you generate it that I can't do without it?" or "What can I do better if you tell me
 that? "  And then sometimes after they come with a piece of information which has
 been painstakingly gathered, it seems to me very reasonable to ask, "So what?" I
 don't mean it in a flippant sense.  But I mean, if you tell me that, say, the nitrogen
 cycle is interfered with in a particular circumstance,  that doesn't help me much if
 I'm an administrator.  What I  really want to know is,  'What difference does it make
 in terms of human well-being? " And it's just possible it doesn't  make a bit of
 difference in human terms.  We know that some of the fungicides that are used do
 interfere with the nitrification process,  and so long as we're using these on agri-
 cultural lands where we apply fertilizers with a liberal hand, a couple of odd pounds
 of nitrogen plus or minus in the cost of production don't make that much difference.
   On the other hand, if you do this in a forest system, where you don't add nitrogen,
 maybe the answer  to the question is very different.  So I think there are some things
 to be thought about in different terms. Not all ecosystems are equally important in
 terms of man.  Not all  ecosystems are  equally vulnerable'. Within the elements of
 vulnerability,  not all processes are equally vulnerable in all systems.
   Somehow or another you have to think about these things in terms of finally de-
 ciding whether you're going to  allow a particular chemical to be used in a particular
way.  Back of all this hangs the question of "What happens to the  chemical in the
 environment?" Now we've got some real problems. We dont know what's happened
to a lot  of chemicals we've already put in the environment. And now we're talking
under the Toxic Substances Act of determining what would happen to a chemical if we
did put it in the environment.  I submit that that's worthy of a considerable amount
 of thought.
   After we have our minds all made up about where it would go and how long it would
be there, then we're faced with a question of "What does it do while it's there?"

Finally, I suppose, the payoff question is "What difference does it make if it does
cause some particular effect?" Those are not scientifically phrased questions, but
basically they are what we're about here.  What we are trying to do here, I think,
is to understand which things are important to learn about, how little we can get by
with in our knowledge of these, and to give some reasonable projection of what things
will be like.
   I guess that's the end of my lecture as such, but I had two or three other things
that came to mind as I listened.  I think when Carter Schuth was talking this morning
about not being interested just in fate and transport, I'm not interested solely in that
either. I think the body of information we have in general and our ability to make
some kinds of estimates on effects is so much greater than our ability to make
reasonable estimates on what happens to the chemical,  that I'm anxious at the moment
to push studies of fate and transport.  It's not because I think the other things are un-
important, but I have some gut feeling that I can cope with that part of it better.
   Another thing, I didn't hear much comment after Pete Schoor's presentation
yesterday when he was talking about two elements of behavior that were being looked
at, and I guess I think behavior is  a subtle sort of thing to get hold of, and I think it's
something that's been grossly underdone.  I think that's true in terms of what happens
in the environment.  I know it's true in terms of what happens to people.  And this is
a general purpose comment, as far as behavioral research, which I expect to make to
the toxicologists worried about human  well-being as well as here.
   I think there are some very sensitive indicators that may be applicable to the "real
world" that maybe we ought to be exploiting somewhat more.  I hope that those of
you who are working actively with these model systems and systems models will put
your heads together on how we can get at the important things, what the important
things are, where there are some promising leads that haven't been looked at, and
how far we can extrapolate from what we get.

   In these terms, what is it that these systems represent?  When we feed a mouse a
 compound to see whether a tumor is produced, it isn't really that we're worried about
 the mouse population.  What we're really worried about is man. We've made a judg-
 ment that says mice and rats and some other animals In some sense or another repre-
 sent man, and we do extrapolate.  We worry about how good the extrapolations are.
   The problem I see here is that we really don't worry enough about what our tests
 are meant to represent and how far we can legitimately extrapolate  from them and
 with what levels of confidence.  1 hope in your discussions that you'll think about
 that. I think the work of standardization or comparability that  Carter Schuth was
 talking about this morning is important. If Joe Smith does it, and Harry Jones, can
 they both start from sort of the same recipe on how you do it?  Do they come out the
 same way or are their results quite different?
   I don't think we've done enough of such comparisons. I think this undertaking to
 look at comparability among systems that we're inclined to use is probably very use-
 ful.  To what degree do they represent the real world or portions of the real world?
 The  final thing that I would keep stressing is to think about what questions are
   I would hope you can start at the most integrated level of the system that you can
 afford  to deal with and undertake to see whether this net end result shows anything.
 If it doesn't, for the moment at least,  I would be inclined to turn to  something else
 and go  on.  Now I may have to come back to this later on,  but as a first cut, it seems
to me we ought to do that.
   That's not only what I meant to say,  but it's probably more than I meant to say.
   DR. EATON:  I'm curious to know what the other behavioral  responses you referred
to were.
   DR.  BUCKLEY: Well, I guess I was thinking in Pete Schoor's talk,  what he was
 using in  that particular case was  predation and  ability to escape predation as  a

sensitive indicator of effect on the prey.  Provided you don't end up with levels that
are harmful to the predator, more easily available prey may be a good thing on a
temporary basis for the predator. Or you can turn it the other way around, and the
inability to escape the predator may  In fact be distinctly harmful to the prey species.
  Obviously you have to examine it from  both ends in terms of the net biological effect.
What I was driving at was that if you go back to whatwe used to do in looking atan effect
of a chemical on the estuarlne environment, we tried static tests in jars, and we dis-
covered that didn't work very well.  So then we decided we had to have flow through
systems.  So we did that.  Then we looked at shrimp — pink shrimp, brown shrimp,
white shrimp, even grass shrimp, but mostly not grass shrimp, the others — to come
up with some 96-hour LCgos. We discovered that shrimp were difficult to work with
that way, so we threw out the LC_n,  and we began talking about an ECPft, which was
                              ou                                50
an effect concentration, and  the reason we did that was because the shrimp would
become immobilized, but it just wouldn't die within 96  hours.
  And, you know, it was reasonable to assume that,  were it out on the open ocean
and immobile, it probably wouldn't survive very long.  Then you begin to wonder if
there weren't levels lower than this, where ability to move would be impaired, but
not in a way that you could observe,  and  I think what people at Gulf Breeze did was
to put two species together and let the predator be the indicator of the effect on the
prey species, a behavioral characteristic in its inability to escape,  to perform in
its usual way.
  I suspect there may be many clues of this sort, if we look at what we've been doing,
that we haven't yet exploited. It seems to me almost impossible that this is the sole
example of that kind of thing. Does that  answer,  John?
  DR. EATON:  I was wondering if you had more examples.
  DR. BUCKLEY: Well, yes. There is another one.  In work on the lugworm, and
again it's behavior in that sense, it's looking at the configuration of the surface and

whether treated and control worms make the same patterns.  It Is a lot different
from an LC   or an LD   — perhaps more sensitive — I don't know enough about
           50         50
the lugworm to answer the question as to whether that's important.
   DRe  JOHNSON:  John, you made statements that I thought were very interesting,
and I thought,  since I'm working with ecosystems, perhaps I should give a little bit
of my philosophy.  I think we have a  mixed group of representatives here, and I'm
very concerned about how our data are interpreted. First of all, I'm with the Fish
and Wildlife Service.  By training I'm a microbiologist who suddenly became inter-
ested in metabolites and wondered what happened to them, and so I was drug into
food chains, and I was never clever enough to figure out how to get out of it.
   But John mentioned one  term here — phrasing that struck me  —  and that is
that we should consider everything  in human terms.   I think  this is rather
a dangerous philosophy.  I really think, when we begin to look at ecosystems, we
cannot isolate man from the rest of his environment, and frankly I look at man,  with
the small m, in  relationship to other species.  And, I think, if we tend to think of
man only in terms of the ultimate species and ignore the ecosystems and the other
species that are involved in ecosystems, we're headed for trouble.
  There was a  second statement, that we  should develop  an ecological hierarchy
of different ecosystems.  Certain systems are not as important as others. I really
don't feel that we're at that level of sophistication that we can really decide that this
particular indicator or that particular indicator is of no importance, and I think that
at this very early stage if we begin to develop these hierarchies, we're in very
serious trouble.  Therefore, when I look at the particular studies that we are in-
volving or developing, I think that they are merely indicators of possible effects on
aquatic life — I kind of like the term "red flags" — and they may help industry or
may help us as far as a particular type of compound — red light, stop — but no
farther than that.

  But I think,  as we deal with the public even on this,  it's the Judeo-Chrlstian con-
cept of really why we're here, and where we have this major clash right now is how
do you deal with the environment?  Is that a stopping over place? I'm not trying to
preach this, but it's something with which we have to legally deal.  Or are we con-
sidered .ajDOsteri?  Are we considering  2250 or the  year  3000, and how do we
handle these resources?
  When I see the words "human terms" underlined,  I think that is a very dangerous
thing.  I think that's all I had to say on that particular point. I have other things when
we get into other data, but just philosophically I think this is a very  important point.
   DR. BUCKLEY:  In my view, man is the center of the universe; he is because he
is the thinking organism who is doing the worrying about it.  Now in no sense does
that mean to me that I'm not concerned about transgenerational effects.  I worry about
my offspring and my offspring's offspring, _ad Inflnitum.  And I'm a firm opponent of
the use of discount rates in the process of making decisions as they apply to bio-
logical things.  So in no way am I meaning to put this down.  I have  a nagging feeling
that all species in one way or another are important, if I'm just smart enough to under-
stand how.
   So when I talk about "human terms, " I'm really trying to say that you people who
are biologists, who are scientists of other sorts, who are worrying about this, owe
me as pseudo-administrator some understanding of  why some of  those things are
important.  And it doesn't mean that they aren't.  Furthermore, we can talk about
all the rest of these living things as, if you will, a life-support system for  man;
they are Important to us.
   I'm not meaning for a moment to put down the significance or importance of
esthetic considerations in here either. You know, I want the air to be clear so the
crest of the ridge is crisp when I look at it.   That's just as important to me as many
other things are.  Other people are hung up on different kinds of things, but you

can't measure them all.  They don't all have a physical or measurable biological
response, and that doesn't make them unimportant.
   So I just want to state my position on what I mean by in human terms, because I
don't mean it in a narrow, immediate sense.  I don't mean if you can't fish for it,
if you can't hunt for it, or if you can't eat it, or if you can't take pleasure from look-
ing at it, that it's of no importance. What I do mean is that all of these things have
a significance to man, and to the degree that we can, we ought to be explicit.
   A second point I wanted to make and perhaps should have made before is the whole
business of decisions. One of the things that a lot of us fail to recognize is that we in
EPA make these on kind of a time schedule.  The guy who ends up having to make the
decision often has to make the decision on the basis of what is known. Now if he
decides not to make the decision, he has just made a decision.  If he says 1 don't
                         *            . -
know enough to decide, in fact what he has  decided Ln positive terms  is that the
status quo is all right, and that is just as positive a decision as the other.  There's
no way you can avoid making decisions.
   Our job is to try to get at least all of what is now known to the guy making the deci-
sions.  If we've got some evidence that says, "Look, the world's not likely to fall apart
tomorrow, and by tomorrow I'll know this much more, so why don't you let it rock.
along until tomorrow, and I'll tell you some more, and then you decide. "  That's
reasonable, but you need to give him some basis for holding off. You have to remem-
ber this character of the decisions, and often, it seems to me, we do forget them.
   DR. VAN HOOK: I've got a comment on doing business better. It would be bene-
ficial to those of us not in EPA but under either  a contract from EPA, ERDA or RANN,
If there could be a little better Information dissemination on just what projects are
In existence.  I just became aware of the Substitute Chemical Program a month ago,
and that's my fault, but I'm here to find out what's happening in that Program rela-
tive to microcosms.  There are other sectors of EPA that are doing  model ecosystem
work, other sectors of the Government (NSF), and it's difficult to find out what's
going on In a timely manner.

  If you wait on the Smithsonian scientific information exchange, the projects are
over,  buried, and dead.
  DR. BUCKLEY:  Yes,  and you might not put the good ones in there anyway.
  DR. VAN HOOK: That's true.  I just make the comment that it would be nice to
come up with some mechanism to let the people around the country know what's
going on.  Let me give you an example, and I don't want to flag RANN as doing it
right, but we put together a trace contaminants abstracts and trace contaminants
directory for RANN which tries to keep the people around the country informed as to
what projects are ongoing relative to trace elements, and it's been fairly successful.
There is now movement to get one going on organlcs, but It would be good if we
could all find out what's happening in a timely manner.
   DR. BUCKLEY:  The point being addressed is that there are a number of people
In other agencies,  universities, and Industry who, for different reasons, are funding
and actively pursuing these kinds of model systems.  Somehow or  another, It hasn't
coalesced yet into one of these invisible colleges where all  the practitioners know
the others.  Maybe it's not the kind of thing which can do that.  And I guess I would
certainly tend to agree with Dr. Van Hook that some kind of a newsletter or reporting
service or other Informal thing would be useful.
   All of us have a certain secret!veness on parts of this, where we're reluctant to
give away new Ideas, at least until we've got something underway.  On the other hand,
there are beginning to be so many of us who have some of the same forces driving us,
I really think that we ought to think about the possibility of an Informal newsletter,  an
abstract bulletin, Intermittent reports.
   I remember, for example, there used to be a PCB  newsletter.  It never had any
 status at all.  I guess it was the Duluth lab that sort of assembled all this stuff and
 mailed It around to anybody who wanted It.   It wasn't  a publication, but It did, In
 fact, give current results of what was under way.  It began as a result  of consensus,

 a recognition that it would be useful, and it stopped,  not by directive or anything else,
 but it just didn't seem to be useful anymore.  It might even rear its head again.
   But the point I'm making is I think there are a lot of ways and I would hope that,
 if we agree that's a useful thing, we ought to try and structure some way for it to
   DR. EATON: I don't think that all the people involved in the Substitute Chemical
 Program have a good idea at this point of exactly where they are in regard to the
state-of-the-art of micro-ecosystem testing and applicability.  There might be  a
 need for something one  step beyond a newsletter,  such as a summary of the state-
 of-the-art of this area,  that would attempt to Interpret the significance of what has
 been done.
   It seems to me that after one year we should have progressed to the point where
 we are well informed about what various people are- doing in this area and that  this
 should be the basis for making some decisions about where we're going, and I think
 it might be very helpful to have the information summarized.
   DR. BUCKLEY:  I think  it might too. The only reason I'm reluctant to speak'to
 it is that I'm not sure who to point the finger at or how to get it done.  Let me make
another observation, and that is that, because  of the interest of the Office of Toxic
Substances in this set of questions, we went out of our way to make sure that those
persons as well as those who were going to be  working with them were present at
this  symposium.  I suspect that we have about  as large a group of people who are
working on or associated with these model systems assembled in this room as  we're
likely to  catch under one roof in the near future.
  Maybe one of the things that we all might resolve to do before we leave is to sit
down and write  a paragraph about who we are and what we're doing and which systems
we're using.  And I don't feel that those things that I've had to say necessarily  ought
to clutter up the record for all time. I'll willingly give up my remarks for a kind of

informal, partial, incomplete directory of "who's doing what, " if you'll cooperate
by providing a very short statement on your work by the end of the session this
   I think it would be helpful from industry, to the extent that they feel free to dis-
close this, and certainly if there are aspects of it which you don't feel free to disclose,
but there are others which you can, I'd certainly be anxious to see those.  May I also
point out that one of the things that you can do in this way, and its hard to bring your-
self to it, but it's very helpful to know who's tried something and failed.  It really
saves the next guy time. We in Government laboratories,  in EPA In particular,
practically never get around to doing that.
   In those laboratories that report regularly to their colleagues and the scientific
literature through standard journals and so on, that tends to turn up.  Where it
doesn't turn up is in internal reports and so on, because what you're looking for there
are  positive results.   I suspect there's a lot of information on how not to go about
this business which people know, and were they willing to share, would be very help-
ful to others somewhat newer to the game.
   MS. SCHUTH:  I wonder would there be any value in  having people submit
information to someone  (perhaps me) and I'll copy it  and distribute  it  to
anyone who's interested.
   DR. BUCKLEY:  Yes, I think that's a very good  point.
   COMMENT:  Perhaps you might have the  participants stand and give their names,
addresses, and a couple of  sentences regarding their work in the field. This would
be a quick and .easy method of recording this information now and  avoiding further

   DR. BUCKLEY: Probably not in the transcription but in the reporting end of it.
 Can I invite each of you who is practicing in this area in its broadest sense to at
 least  put  your name and address,  institutional affiliation,  where  you are,
 something about the system or systems that you're working with and something about
 the characteristics of the substances you're working with,  the kinds of questions
 you're trying to investigate.  If you could do that this afternoon, before the 2:30
 adjournment, it would give us something to put together as a starting place.
   MS. SCHUTH:  What I'll do is have them  typed,  each  on a single page,
stapled together and  mailed  to  anyone  who  lets  me  know he  or she is
   DR. BUCKLEY: Okay, we have done something positive during this meeting.  There
 may be some other positive things too, but now we have a measurable accomplishment.
 Could we now leave the mechanics and turn to the substance of how you go about some
 of these things, and since I don't know anything about substance, only mechanics, I'd
 like to turn the microphone back to Dr. Glass and perhaps to Ms.  Schuth, who may
 have some further things to say.
   DR. GLASS:  I'm not sure I can help much on the  substantive matters either, but
 I think now might be a good time to enter into the discussions concerning how useful
 these Indices are that we've all talked about, how applicable are some of the numbers
 that we get to the general problem of the  effects of these chemicals in the environ-
 ment, and  one question that came to my mind this morning while Dr. Johnson was
 making his comments concerning microbiology.  Is it possible that the microbio-
 logical system is somewhat more diverse perhaps than other types of biological sys-
 tems and that,  because of this, maybe there is enough redundancy in that system that
 you don't need to be concerned about whether or not  particular species of micro-
 organisms  die as a result of your toxic chemical ?
   I don't know that that's true.  I am just asking if maybe there's  some difference
 between the kind of diversity,  for example, that you observe in the micro-community


relative to the sort of diversity or other community parameter that you might measure
on larger communities. I wonder if there is any merit to the thought that possibly
diversity Indices are different, depending on the type of community It is.
   COMMENT: In determining about microcosm — we really do not have a micro-
cosm as such, but a modular system — and we use microorganisms merely as a
type of indicator of the accumulation of particles.  Matter of fact,  we ended up using
yeast because we found out it didn't make any difference in the bacteria or algae.  We
used yeast simply because it would be very readily available to people, and they didn't
have to have a lot of background on how to culture organisms or then you get into  a
raging debate on what organism to use.  And so we discovered that yeast were quite
   As a matter of fact, it appears that even autoclaved killed yeast may form the same
basis for studying the accumulation of particulate  matter, biological particulate matter
for this particular value.  But responding to the type question, it just so happened
that we're quite interested in this area for another reason.  One of the major problems
that has always plagued administrators, I think, is the interpretation of the degrada-  -
tion of a particular compound.  Joe Blow says his compound is just no problem what-
soever, it's readily degradable. And then, lo and behold, we look at our national
monitoring program or something of this sort, and we find out that the compound is
present at horrendous rates.  Something is wrong somewhere.
   We find out that perhaps this individual has done a pure culture study or something
of this type,  which is really relatively nonsense for environmental work, or he may
have done a study in sludge, where this particular compound is never going to be, even
if you get into this type of ecosystem.  So it becomes a question of how do you define
this type of a system  so that you can do degradation studies?
   We have to zero In Immediately and not look at the whole world but just  look at
heterotrophs.  This is what  we are doing and we have begun work along these lines.

 And we do see changes in the organisms, and these do influence degradation.  How
 important they are, we don't know, but there  is something that does occur, and that
 is the microflora during degradation periods do change.  What we're interested in
 is have they changed to such an  extent as to change what's going to happen in the
 environment, so that you now are looking at an organism that in the environment
 never would be greater than ten to one or two per some unit, and now we find out
 that, due to the conditions,  that this organism is at very,  very high levels.
    This is a major consideration, and we are going to do extensive work along these
 lines,  just to determine this, but strictly in terms of heterotrophs.
    DR. GLASS:  Another way to look at it is to ask if the material degrades and not
 is there a change in the composition of the organisms that degrade it. What you
 really care about is whether or not the compound disappears over some reasonable
 length of time.
    COMMENT: Yes, but there is another flip side to this, which is another project
 that has concerned us a great deal,  and that is that we have been so concerned with
 the parent compounds.  A good example of this,  I think, is methoxychlor,  on which
 Dr. Metcalf did some work,  ft was a very interesting study, and this has been in
 the literature repeatedly as  a biodegradable compound.  It is biodegradable in the
 sense that it has lost its initial toxicity, and it had been hydrolyzed, but we still see
 the ring structure, and we begin to ask ourselves what's happening to these com-
 pounds, these persistent things,  such as pesticides, that have sat around and no
longer have  the apparent biological activity.  We are beginning to look at these in
terms of their effect on geochemical cycles, because we feel that perhaps we're over-
looking these and that we may be having marginal effects of one or two percent, which
have an effect on ultimate  productivity as far as our resources are concerned.
   COMMENT:  The whole issue of whether the metabolites  are too from the breakdown

  COMMENT:  Perhaps this shift in materials, this shift in certain colonies,  certain
types of organisms, may influence the geochemical cycle, particularly nitrogen,
sulfur, etc.,  that may have a very important effect on productivity.
  This  particular topic you're discussing relates to bound and conjugated
pesticide residues,  which the Vail conference was principally about, and the bio-
availability, if you'd like, of conjugates or bound or unavailable type materials to
the environment.  This not only includes plants, earthworms, the whole mix,  what-
ever was available for testing, and I think the crux to the whole problem is  it was
established, I think, that almost everything that we put there is going to still be
there In some form or other. It's not just going to go to CO0 and water very quickly.
  The whole thing is the significance of something in the environment which we may
call "unnatural." If it's not bioavallable to a number of different test species that
we  determine, how significant are those results?  I think that's the crux to the whole
problem,  that we have to set some criteria and test species here, and if it's not
bioavailable as we define it, then how else can we go until the state-of-the-art
continually progresses?
  But if there is  something there at some level, as we determine by radiolabellng
techniques,  and It's there 100 percent, but In some other supposedly unavailable,
nontoxic substance, yes, we should be concerned about It, but the significance that
we place upon it may be a much lower priority than something which is more  avail-
    COMMENT: I suppose It would matter what compartment or component you found
this material residing in.  If you found it bound to a soil particle,  for example, you
might be less concerned about it than If you found It In some protein compound In an
   COMMENT: For Instance, in relation to the soil thing,  Cleve Goring and John
Hammecker of Dow had gone through an elaborate mathematical model as well as

 through practical applications of a large number of chemicals being bound to soil,
 and discussed the relevance and significance of things that are occluded, bound,
 that you can't chemically seemingly get out of the soil.  Now it's there, and it might
 be there for thousands of years, millions of years,  but is it really going to affect
 our environment, either us or microcosms,  the whole works? Philosophical
   QUESTION:  I'm wondering if we can get back to diversity just for a minute.  It's
 my opinion that some methods for determining diversity have progressed a little bit
 further than analysis of the significance of them.  This is opinion  based on prac-
 tically'no experience. It's also my opinion that it is an obvious tool in cases where
 there is gross contamination and that changes results that are easy to see and
 interpret under those circumstances.  I'm wondering specifically If anybody would
 care to enlighten me a little bit regarding the sensitivity of diversity as a measure
 of effect of substances on the environment.  It's strictly out of ignorance that I'm
 asking that question.
   COMMENT: There have been several review articles In the last five or ten years
 that look at different types  of Indices and test their sensitivity, mathematically
speaking,  and which  apply different  indices to the same  data  set and  different
 data sets to the same index, etc. I'm not sure that there is a clear pattern that's
emerged from those reviews,  other than to say that you don't always get the same
conclusion from two different  Indices applied to the same data set, for example.  If
you had one set of data from which you could calculate a diversity index, you calcu-
late two indices, and you might discover that you come to conflicting conclusions or
at least different conclusions.
  QUESTION:  Is that the same thing regarding sensitivity?  It seems to me one of
the big drawbacks at this point in time regards our  knowledge about the sensitivity
of diversity in terms of their being  able to  make evaluations regarding effect.

   COMMENT:  I think I understand your question.  I just don't know the answer.
   DR. KENAGA: We talked yesterday a bit about the benchmark concept summarized
by Dr. Rlz Haque of EPA.  That Includes benchmark chemicals and benchmark test
methods, the way I look at It.  Dow has done quite a bit In this area for a good many
years.  You've heard our researchers' names mentioned on several occasions.  We
feel that this has been a very useful concept for us in research, If nothing else. In
research we found, for example, that octanol/water partitioning coefficients and
soil organic partitioning coefficients and fat residues In fish have a very good cor-
relation in coefficient numbers.  Bloconcentratlon factors are calculated on the basis
of those techniques, and so they have a lot of things In common.
   Now there are several ways to go about studying environmental pollution.  One  is
by studying a minute detail for your entire life.  Another way is to determine the
important ways in which a chemical dissipates in the environment.  The study of the
physical and chemical properties and the proper method of comparing them with
other chemicals gives you a great lead to that in research.
   I have been encouraging this  latter" concept for a long time through my publications.
I feel dedicated to the fact that this is a useful tooU However, government regu-
lations are what we all have to face In industry.  For example, in the  appendix to
the new Guidelines there Is a test method on octanol/water partitioning coefficient
and there are numbers that are suggested — I think it's 100 — as to whether some-
thing might be an Indicator of bloconcentratlon from the octanol/water partitioning
   However, this Is not one of the test methods suggested In the Guidelines itself.
So there isn't a matching up of  those numbers and test methods In the Guidelines
themselves. We have tried to make use of partitioning coefficients in trying to
register products and maintaining that these coefficients are correctable and useful
as indicators as to what might happen in different segments of the ecosystem.

 However, we don't get very far with EPA in using those partitioning coefficients when
 there's nothing in an EPA checkbox that says that you have to do this test method,
 because they don't seem to know how to interpret them. And we're hoping that, if
 we do establish within EPA that those benchmark characteristics and others men-
 tioned by Riz Haque are valid, then we can use them in our arguments as to which
 test methods need to be done to answer the questions as to distribution of pesticides
 in the environment.
   I realize that it may sound to some industry people like we're being stuck with
 test methods which are not the same as our own, but I would submit to you that
 nearly all of the Guideline test methods are becoming that way, and our only hope
 is that we encourage the use of test methods that are valid and are really meaning-
 ful. In the present test methodology on model ecosystems there's only one that even
 has an EPA test method guideline to It,  and that's the flowing water bioconcentratlon
 test.  Another one that's Inferred but not written out Is the test method where you
 treat  soil and put the soil hi water and then have the catfish exposed to It about a
 month after the soil has Incubated the compound.  But that's not really In the method.
 We've been talking a lot here about many different model ecosystems. It would be
 frightening for Industry to be faced with four or five different model ecosystems to
 have to use In the registration process for each compound.
   That's not to say the research Isn't needed for model ecosystems.  And I certainly
would be the last to discourage that. 1 can see that terrestrial model ecosystems .
 require different organisms than aquatic ecosystems.  The partitioning of these
compounds in environment are all subject to water,  which Is the major liquid In
all ecosystems, and to fat, which occurs commonly hi plants and animals.  The
general partitioning trend of the chemical In the environment may depend upon the
 ratio of fat and water to each other.  Use of the data that we have on residues In fat
 tissues  In pesticide-treated cattle or other animals that we usually have had to sub-
mit with the registration of the pesticide supplements partitioning coefficients that
occur In fish and other ecosystems.


   There are many ways in which the total input of the information gained by use of
 test methods named in the Guidelines lead you to the same direction and the same
 conclusion.  Industry certainly hopes that with  the Toxic Substances Act coming up
 for passage we don't have two regulatory groups with two different bureaucratic
 sets of test methods for each evaluation (i. e.,  fish and wildlife, physical and
 chemical properties, etc.).
   Now I've been assured that this will not happen, but I will wait and see.  I laud
 the people in the lower echelons who, I'm sure, believe that there should not be, but
 I'm fearful of the upper echelons.
   COMMENT: I have two special comments.  One was in relation to Ms. Schuth's
 work about the octanol/water partition coefficient.  This could be a reliable indi-
 cator of bioaccumulation, but I think that using this coefficient in the fate studies
 may lead to some  misleading conclusions,  because other factors like specificity,
 localization at certain sites, degradation, excretion, play a very Important role in
 determining the toxicity of a compound.  So I think we should be somewhat cautious
 there in applying partition coefficients to the fate studies.
   And my second  comment was in relation to Dr.  Hayashi's work which he presented
 yesterday in using the La vitro system  as a screening method for detecting the toxi-
 city or the mechanism action of a compound.  I think using systems such as chloro-
plast, isolated chloroplast or mitochondrium, or cytochrome P450 and'exposing
 them to a certain chemical  in test tubes could be very misleading.
   I  think if we want to use this in vitro system, it should be in vivo exposure,  fol-
lowed by In vitro testing,  and I'd like to hear comments from some  other people
 along these lines,  because it has been pretty well  demonstrated in certain pesticides,
 for example,  a very nontoxlc compound like PCNB.  If you take a cell-free prepar-
 ation of cell membranes and feed the _ln vitro system, this compound will show you
 all kinds of membrane alterations hi the cell because,  being a sulf Idal compound,  it

reacts with the proteins, amino groups.  On the other hand,  if you feed this thing
to animals, apparently you're not going to see anything there.  This has been already
demonstrated with other pesticides like Atrazine. All we know is that the corn is
very resistant and soybeans are very sensitive.
   If you expose the corn plant and soybean plants to the herbicide and then measure
the photosynthetic rates in the leaf discs, you will find that corn  is not inhibited and
soybeans are very sensitive.  On the other hand, if you take  the chloroplast of corn,
which is a  resistant species, and expose this to the chemical Atrazine, you will see
almost 100 percent inhibition.
   The point I'm trying to make  is that this work entirely dealt with the in vitro ex-
posure,  and I think we should try to make the interpretation  very carefully, and this
is a very hazardous approach.
   COMMENT: I have to explain something  about our research project.  First of all,
when we evaluate any herbicide mode of action,  we have to think  about physiological
result, physiological  phenomenon, either one as the result of biochemical reaction.
First,  we are taking the various kinds of biochemical reactions,  both in vitro and
in vivo systems.  As I reported yesterday,  the time was so limited,  so I just picked
up the important data, and right now we just concluded what  kind of method for
evaluation  of a herbicide.  We are concerned about the best way to use whole,  intact
plants and  also some  in vivo systems. If you have seen my report,  right now we are
using some susceptible plants, Intact susceptible plants, actually soybean plants,  to
evaluate herbicide, using oxygen uptake.
  As you have seen in the  slide of the 15 herbicides, 12 herbicides are effective to
the oxygen evolution.   In order to evaluate the mode of action of a herbicide, the
best way is to use the whole intact plants.  When we use the  corporal system,  we
have a lot of troubles. Once we were thinking about evaluating the mode of action
of a herbicide using a protein pattern on a corporal system,  but it was not success-
ful.  Every plant species has a different response to the various  herbicides or

  Right now we are using the whole,  intact plant.  Of course, before we made the
conclusion, we used in vivo and in vitro systems of chloroplast and metachondrial
  COMMENT: Yes, I think we should try to distinguish if the mode of action is the
mechanism by which a chemical produces its toxic effect, Irrespective of whether it
reaches the site of action or not.  But what I think that you're trying to do is to find
a sensitive screening method.
  This is an inherent mechanism.  If a plant is inhibited by certain biochemical
reactions,  if we go by the whole plant, like a resistant plant, we will never be able
to see this mechanism of action because by the time it reaches the chloroplast,
reaches the site of action, it will get degraded into a nontoxic compound.  The only
way you can see the mechanism of action is by working a system in which we have
eliminated the factors of degradation, eliminated the uptake, transportation.
  This is a screening method for determining toxiclty of the compound, not a
screening method for determining the mechanism of action of a compound. That's
my whole point.
  QUESTION: Do you have any idea of what kind of methodology we actually have
to use?
  COMMENT: No, what I was saying was that if you are going to use a screening
method using chloroplast, industry is going to use chloroplast or metachondria of
P450 as sort of enzyme system, isolate the enzyme system.  I don't think this is
going to give us the meaningful Information,  because under those circumstances you're
exposing the  system to the highest concentration of the chemical, which apparently
will never happen in the whole animal or the whole plant.
  If you establish a system In your research and then on that basis you recommend
It to Industry or to universities saying,  "Look,  we have this chloroplast,  this Is our
system,  you go and screen all your new chemicals In that, " I'm sure you're going
to come up with very misleading information based on these studies because there

are many,  many compounds which apparently don't produce that.  If that was the
case, we won't have species sensitive to herbicide and species resistant to herbi-
cides.  In the literature, most of the compounds are in plants which have the differ-
ential sensitivity or resistance,  a lot of times it's due to the differential uptake or
differential breakdown of the compound.
   So this is the thing that we have to keep in mind,  that we should try to keep away
from isolated chloroplasts and isolated metachondria.  If you want to do it, you should
expose the  whole plant to the chemical and then try to isolate its chloroplasm,
metachondria or whatever other organisms you want to work with.
   COMMENT:  I think probably you misunderstood that first we use the whole,  in-
tact plant.  You remember that we put the whole, intact plant in a chamber and de-
tected carbon dioxide, using the aniode gas analyzer. Why are we evaluating carbon
dioxide  stimulation and the respiratory effect?  Because first we have to evaluate
the mode of action of the herbicide using the whole,  intact plant.
   COMMENT:  That's the point I'm trying to make, that you cannot study the mech-
anism of action in  the whole plant, because you don't know what you have done  to the
pesticide.  You don't know what is producing the toxic effect.
   COMMENT:  We talked about a practical test, and you're talking about a mode of
   COMMENT:  I still maintain that it's difficult because you are adding other factors
because one doesn't know what has happened to chemicals before they produce an
effect — maybe the chemical has been converted Into compound A or has been con-
verted into compound B, and that's the one  that produces the effect.
   COMMENT:  I still am not clear about how to evaluate the mode of action.
   COMMENT:  You should Isolate your system.
   COMMENT:  You still disagree, because we have done some experimental work
using Isolated chloroplast.  This Is not of practical use and I think I'd like to discuss


this problem with you.  Too many factors are involved; for example, the surface of
the chloroplast is directly touched to the herbicide solution.  If we evaluate the herbi-
cidal mode of action — those data are not of practical use.
   QUESTION: If EPA is charged with regulating the occurrence of these chemicals
in the environment, the distribution, the fate, and the ultimate effects, and if indus-
try is also charged to produce these various materials, free enterprise,  and all
that sort of thing, how do we, the contractors like myself and others who work for
EPA, get the information from industry that helps us scientifically devise a system
that supposedly we're going to build to give to industry to help them evaluate chemi-
   Now there are a number of grants that have been let where the contractor is told
help devise a "sensible screening tool" that industry can use.  How do we know in-
dustry's going to accept it?  EPA obviously could force it, but that's not the best of
all situations. And my question is, how do we get EPA and industry together to
find out what both needs, so we could get the happy compromise?
   COMMENT: You don't want to reinvent the wheel.
   COMMENT: Exactly. They've done a lot of things that we're unaware of.
   DR. BUCKLEY: I guess there's no clear-cut answer, obviously, to what you've
just asked.  One of the things that EPA tries to do in its regulatory action is to be
as open as it can along the way.  Now I understand that we're criticized for both
extremes, being too open and, on the other hand, never talking to anybody.  I submit
that  this meeting,  for example,  is one opportunity to exchange views or  to begin to
understand what other parts of the community are doing.  It seems to me there is a
second stage. Let's assume that the object of this whole exercise were  to erect a
system which could be imposed by the Government,  followed by Industry, with  the
object of net gain by society.
   One way of doing this Is for EPA through contractors,  by whatever mechanism they
can, to assemble Information and to put this out in the form of a proposal for comments.

We're now in a position where we don't even have a law (the Toxic Substances Act).
We can still shape that.  But there will come a time when there's some kind of
   One of the things we might consider is an "advanced notice of proposed rule
making" in which we set forth a proposal, follow it with a series of public meetings
in different parts of the country to bring together people from all sorts of interests
to discuss  what's impractical about it and why you can't do  it that way or, conversely,
why you must do it that way because there's no other way of getting at the information.
   The usual practice in EPA is to issue an advance notice of proposed rule making
to try to pick the biggest of the bugs out of the system that way.  We then go forward
with proposed rule making in which people realize that we are being serious about
it, and sometimes elephants instead of bugs come walking out of the woodwork.  And
finally, taking into account the comments, we publish a set of regulations which have
effectively the force of law and which have been looked at by everybody. That doesn't
mean you all agree on it.
   COMMENT:  May I suggest that perhaps there's another  way which  has been uti-
lized recently very effectively, I think, and that is through  the American Society of
Testing and Materials, whereby all segments of those people who are  in a position
both to use and regulate these things have the opportunity to agree before it is put
into the rule making process  and which gives a lot more ample time for discussion
of the pros and cons of the various methods, and theoretically requires consensus
agreement on the methods prior to the time that they are  published and made avail-
able as being another alternative means of accomplishing some of these.
  DR. BUCKLEY:  I think that's a good point. I think to  the degree that it can be
used, it's a very useful process.  I submit that what's going to happen in practice
with the Toxic Substances Bill,  however,  is one day it will  pass, and all of a sudden
there will be an urgent drive  to get guidelines to industry, on the street, whatever

they're called.  They may have the force of law; they may simply be guidelines; or
they may be something in between, but somehow or another that's going to happen.
  We've got a few months ahead of us yet in which we can still do some sensible
things jointly, but the closer we get to needing to put out something that people must
do,  the tighter the deadlines,  the greater the difficulty in doing that.  And I would
just urge that we talk as freely among ourselves as we can.  As Gene Kenaga has
pointed out, industry has got a wealth of information that they have worked with
over the years with the deliberate intent in the pesticide area,  for example, of try-
ing  to find what are those things that will be good to kill pests and won't also do a
lot of bad things at the same time.
  These are, in fact, screens because there are innumerable  potential chemicals,
some of which are more promising than others.  As a practical business matter,
they've been trying  to decide which ones to develop.  And it's obvious that they
don't want to get to  the end of the scheme and discover that this Is fine and every-
thing, but unfortunately,  it makes people drop dead or, for some other reason, is
just totally unacceptable.  Somehow or another those of us in academia or  in re-
search institutes or hi regulatory groups or whatever, do really need to extract
whatever we  can from these systems.
  DR. KENAGA:  I think industry screening techniques can be used for evaluation
of many compounds.  Screening is a word we've used here a little loosely.  I think
it means different things to different people.  In Industry screening Is when you take
a compound and,  if you've had it for the first time,  and you submit it to a lot of
test methods to find out whether it's going to filter biologically out and become a
viable pesticide candidate. And In that sense Industry has many, many screening
techniques that we've gotten from anybody that looked like they're practical.  We
don't care where they come from, as long as they give us an answer that is practical.
  We've had a lot of experience in evaluating techniques or test methods that are
misleading.  So we're a little allergic to those.  And one thing that a company really

hates is to have the first compound tested on a new test method, which nobody knows
how  to interpret or use,  and then somebody  publishes data and  says what  a
terrible compound it is.
   QUESTION:  The subject matter came up before about test indices which are ulti-
mately the objective of looking at model ecosystems and, whether or not you look at
one type or another, at least have some common ground of comparison.  I don't per-
sonally do a catfish study.  I have it done for me. I was involved in assisting on some
things for registration of compounds.  I get back numbers from which you can take
and calculate the bioavailability, rate of degradation indices.  How do we judge in
industry the meaningfulness of these indices,  and what importance do we place upon
   Are there guidelines that define some numbers which are acceptable or unacceptable?
Obviously, it has to depend on what kind of compound.  The compound itself may have
different toxicology and other things like that that make it completely safe, so you'd
have a higher number.  But are there guidelines or numbers so you can judge your-
self a little bit?
   I go back to the data,  I start to assemble it. But we may continue on another six
months or a year's program before we submit the whole package.  Maybe that's an
alarming thing right when I get back and I look at it and I say,  "Oh, oh, we'd better
stop right now."  Do we have to wait to get to  the Government so they can have them
review it so they can come back to us a year,  a year and a half or two later, and say
forget it fella, you're out of luck now?  Is there something that we can assess the
situation on?
   COMMENT:  If you take the magnification values that have been derived in the
terrestrial model ecosystem, and you go to Illinois Farm Ponds,  you find that the
values fall right in that order of magnitude.
   I don't think the absolute values are what you're interested in, it's the relationship,
as you point out, between compound X and compound Y which is known to persist and
has already been shown to be persistent.

   DR. BUCKLEY:  I wouldn't even want to draw the analogy. Although the
Guidelines suggest what you might do with another chemical, a given chemical X
which is in the same family as another chemical,  many factors are going to be
significantly different.  There can be a fair amount of difference,  the distinctions
of metabolism,  other factors.
   COMMENT:  Use patterns.
   COMMENT:  The whole works.  So that you can't draw the parallel.  You
might have a biological accumulation, and something of 200 which you normally
define as being  bad, whereas another number may be 75 or 50, and that may be
just as bad.  It  depends upon the chemical.
   COMMENT:  One must take many, many fragments of information and put
them together before one can make a decision.  I know a compound in particular
that will show a very high accumulation in fish, but it is readily degradable in
hydrosoil.  If you were to take fish alone or take  the octanol coefficient and
look at those two, you would say that it's a no-go compound.  But if you were to
look at the degradation element,  you'd find out that it's rapidly degraded.  This
changes the story.  You must take these and put them all together, and I think
we all agree on that.
   DR. BUCKLEY:  Let me take what you said and push it one step further.  If
you took that octanol number and found it was high,  what that says to me is I'd
better look at fish if it's going to get in the water.  On the other hand, if it's low,
I have a pretty good feeling that it's not likely to accumulate in the fat.  That
doesn't keep it out of the bones,  it doesn't do a number of other things.  It
doesn't answer all questions,  but it does give you a clue that you need to know
 more. And you,  as an industry guy making judgments,  are just as competent
 as those who will make the decisions in Government in thinking about what the
 next question is and whether you need to pursue it or not.

   Another thing that has been emphasized here two or three times and probably
 needs to be continuously reiterated is that if a substance is not particularly toxic
 and you don't care if it's around at a part per 1000, then accumulation may not
 mean anything.  On the other hand, if accumulation will result in levels within
 an order of magnitude  on the range where you think there may be a problem, in
 that case you're going  to be a lot more careful.  I think it comes back to this word
   Again, I don't think any one of these is definitive,  but I think they point you to
 other questions you need to ask and,  hopefully,  if the screen is worth anything,
 it will save you answering some of those questions.
   QUESTION:  There's one more  aspect to that question posing that nobody has
 approached yet.  As you're running the test with a given compound through
 everything that you can stop at, at  that point you've determined that you have
 enough information to decide that it shouldn't be developed any further, is there
 some way to keep in contact during the development process to allow you to save
 money ?
   MS. SCHUTH: That's a business decision.
   COMMENT:  Yes, I don't think  it's appropriate for Government.
   QUESTION: If that is not the question, I will retract it,  but is that  part of the
question? Did I  interpret you correctly as saying that?
   DR. BUCKLEY:  It seems to me that that is a business decision, and just like
Government decisions, there are two sides to it. If the promise in terms of bucks
in the pocket from that compound is extremely high, you may  be willing to accept
some  results that you wish hadn't happened and then pursue it to a considerable
extent beyond there.  On the other  hand, if this is an exploratory thing and you
don't see the potential for a big market with it, and it's got a couple of strikes

against it in terms of characteristics that will be hard to live with, you'd
probably drop it.  I don't think we ought to get into it.
   COMMENT:  There you are at the business end of it, in terms of the econom-
ics of can you make enough out of it  in that way to make it worth pursuing and
digging up the information needed.
   DR. BUCKLEY:  I've tried twice  and I'll try again.  It seems to me one of
the things I'd like to know is,  are these any good? I don't mean indices, I mean
model systems.  If they are, what are they good for, and more to the point than
that, if they have some promise, what can we do about them to make them better?
Are there some questions that we've overlooked?  It seems to  me we've been all
around the question and we've tiptoed up to it a half dozen times, and nobody has
been crude enough to tell Jim Gillett that his system was too complicated. No-
body was crude enough to suggest that the terrestrial one might be too simple.
   What I'm interested in from you people who are working in  this area accumu-
lating data is some feel for things that would be better,  or things that can't be
done, or things that are worth looking at that you haven't gotten to yet.
   I saw an aquatic system yesterday and I was curious about atmospheric inputs
and outputs.  I wondered if you couldn't use that same system  and extract an ad-
ditional piece of information by controlling the atmosphere  over it.  I'm reminded
of some work that EPA had done in our earlier days — not  all that much earlier
either — when we were obligated apparently to work on things that were in water.
If it left the water into the atmosphere, then it wasn't considered  our problem.
It was somebody else's problem.  Somehow or another, it's all one environment.
   Doesn't somebody have something to say on both the shortcomings and the
possibilities?  Are there some imaginative ways we can tie things together that
we haven't yet?  Are there some questions about the way things are now being

   COMMENT: Jiin Haggar's model which is under development was a tool of
the design to model the microcosm itself in the hopes that we could do something
like a sensitivity analysis on the model and determine if there were aspects of
the microcosm that could be deleted and not lose too much information.  That's
in a state of development at the moment; it's not completed, but I think that's
one possible approach.
   COMMENT: It seems to me that Dr. Johnson in his food chain studies has
been dealing with some of those questions, but not necessarily always in the same
box.  Sometimes it's several boxes, and you get to a particular point and you
move the experiment from that box to another.   Maybe you don't literally have to
move them,  but now that you know how  far it gets there you  can find other ways
to look at it.
   COMMENT: For several years when we were thinking about how we can put
together some system to test compounds, we asked would it be degraded in an
aquatic  system, would it be accumulative, would it be transported, and could it
be done for under $5,000. 00 or under $500.00,  or some value that seems real-
istic, particularly if a company had 18  or 20 different compounds, or if the
Agency  had a particular herbicide or insecticide that it felt was very important
in managing certain types of resources  .
   We came up with mathematical models — I'm using the word model probably
in the broadest sense.  I apologize for that,  but we felt our particular system
could not be put together in one unit because we  could not control.  We had a
problem of controlling biomass  ratios in relationship to nutrients, and so we
came up with what we called modular,  or used it in this modular fashion.  We
can look at each segment by itself,  independently of the other, and then inter-
marry these at our own convenience.  We felt that we could very accurately
study uptake, we could very accurately  study degradation, and we feel that this
sort of thing works reasonably well.

   One other thing we felt was a weakness of microcosms systems is that it's
very difficult to follow the nutrients. When you're feeding a particular organism
fish or something of this sort, are you feeding him at a realistic level?  Are you
starving him or are you feeding him at such a rate that this is unrealistic?  With
our system we can calculate these or we  can adjust these.  Then you also know
what is happening as far as the nutrients  are concerned and so you have many
handles. Our model was taken apart.
   To go back to your question of how good they are,  we feel that they are only as
good as the people who operate them, and maybe in two more years with all this
testing Carter will find out if they are reproducible.  We do feel that they offer,
in conjunction with other data, red flags.  They offer indications of probable dif-
ficulties .
   We feel that an index is useful because it does help those persons in the
hierarchy who have to make the decision when all the data are gathered, and
that's usually either a vice-president somewhere along the line or perhaps an
administrator who is a gifted scientist but who may not be really up on all of the
idiosyncracies and the difficulties.
   Thus, if we give them some of these values with the proper guidelines along
the way, that this means so arid so, we feel that these numbers could be of some
use. We have based all of our accumulation data on  DDT.  We know a great deal
about DDT,  its ubiquity in the environment.  It is quite easy to work with in the
   Therefore, if I say a compound has an index that's,  say, based  on DDT in
unity, an index of 110 — we'll say DDT has an index of 100 — this gives you
some idea that this is a problem.  If it has an index such as some of the triazines
these have 0.001 or something like this — it indicates  to you that triazine as far
as accumulation in these aquatic  systems is essentially no problem.  This is the

 type of information you can come to, but it does need to be used in conjunction
 with other data.  The index, we feel, has value, and I like it.  If you can come
 up with better reasons for approaching it from some other way,  fine.  We are
 constantly asked what does 75X mean? Is  this bad, good, or indifferent?
   The answer is this is a value.  It isn't good or bad, it has no value.  Scien-
 tific data are not  good or bad.   It  is a number,  and 75 or  1, OOOX  could
 mean under that circumstance,  it has  this  value.  Under environmental  circum-
 stances,  in use in conjunction with other data,  it may really mean that nothing
 has no value.
   QUESTION: You use the word index as  sort of a single parameter type index.
   COMMENT: Yes, but it could be multi.  We're using it as single right now.
   QUESTION: Bioaccumulation.
   COMMENT: Yes, right.
   QUESTION: It seems to me that for an  index to be usable you would almost
have to have it be a multidimensional index.
   COMMENT: Yes, and this is the next thing I would like to bring up.   The great
 problem is how to weigh these.   The Metcalf and Sanborn methods use an index
 based on degradation.  You could use one based on LC   or not index, but all these
 fragments of data could be put together and you could end up with five or six or
 seven of these, and come up with a particular numerical value which would tell
you something about that compound environment.
   I think of the analogy of the old heart.  Have you ever seen these  at the
doctor's office? As I was  looking at it, I thought this is marvelous because
it was indicating whether you'd have a heart attack, whether you're in bad shape,
such things as your age, the number of cigarettes you smoke, whether you have
a high lipid diet, or how much cholesterol you have in your blood.  You have all


these values.  These are arbitrarily given values,  and you add these up.  There
is a little index which says if you're within such and such values, you're going to
die before you leave the office.
  This sort of thing might be useful in some of our compound problems,  but I
don't know how you'd weigh them.  This is where the argument is.  How would
you weigh an LC   of a part per billion which is a range for the organic chlor-
ines such as DDT, versus a compound of the organophosphates which perhaps
would be 100 parts per billion,  and then go into things as effect on reproduction?
How would you weigh that? This would be the ultimate.  No one at our laboratory
agrees with this; they think it's great but then we get into the arguments over
values.   But that's the ultimate.
  DR.  BUCKLEY:  I worry a little bit about questions of simplification and
oversimplification.  It's clear that what we want out of this whole situation are
simple ways of evaluating complex events.  What I hope we don't end up with are
simplistic ways of evaluating these which really have no particular merit
associated with them.
  I don't mean that as a put-down to indices, because I submit,  for example,
that one that we all  use in one sense or another in our daily lives is an appre-
ciation of what the GNP,  the gross national product, is.  We may not know with
precision but you have kind of a gut feel that when it goes up,  certain things are
happening,  and when it goes the other way, there are certain things associated
with it.   It's a very arbitrary kind of thing, put together out of those things
which are conveniently measureable.  It has no intrinsic merit; the  ingredients
are not necessarily all causally related, or relatable in ways that we can even
specify, but they do tell us in a single number that things are getting "better" or
"worse" in terms of our national economy.
   I submit that we may, with the passage of time,  find  that it doesn't tell us at
all what we want because at the point in time GNP was developed, the presumption

 was that continuing growth was the mode that this nation and the civilized world
 would be in. If we change our goals and decide that stability of some sort or
 another or some lessened growth rate is the desired mode, then the way we use
the measure may be very different.  We may both want to change the index and
 perhaps change our interpretation of it.
   I think that this same kind of thing would apply to the indices here.  I'd like
 to see if we can turn now from this  to the physical experimental systems.  Let
 me ask a question of Dr. Johnson.  What questions can you answer with your
 food chain model?  What things that are important can you not measure with it?
   DR. JOHNSON:  We feel that we can reasonably and accurately reproduce
 some values as far as uptake from the aquatics in microorganisms. In our
 total food chain we attempted to look at three trophic levels: a microorganisms
 trophic level,  if there is such a thing; a filter feeder; and a fish. .
   These represent distinctive niches,  and we hoped to interpret the uptake and
determine what the plateau levels were over a given period of time. We're
hoping to determine what happens to the compounds — were they degradable, is
the nature of the compound being changed now, or do you get into somewhat of
the Pandora-box syndrome type of thing ? You are able to know what happens at
a trophic level  that you're interested in, and for us it was the fish and so we can
look at those things.
   We  have about five areas that we can look at independently or simultaneously,
and we can get  both qualitative and quantitative information.  One other thing
which we haven't emphasized, but it's quite important,  is the elimination of the
compound, that is,  if it is picked up, how long does it stay in the particular
organisms ?  We are interested in the half-life of the compound.

   Those are the bits of information that come to mind readily, and which we
looked for.  You can come up,  if you like,  with an ultimate index if you want to
compare it or just if you have straight values under these things.  You can also
quite readily in these systems do all sorts of manipulating with biomass changes.
You can stress your animals, you can do temperature studies.  These are all
the most obvious, but they're relatively easy to do in these systems.
   DR. BUCKLEY:  This is a more general question, but it's triggered by what
you said.  What consideration do you give to the way in which this  substance
might reach the environment?  For example, the organophosphorus pesticides
are ordinarily used intermittently.
   DR. JOHNSON:  This is a very important consideration.  I would have had to
go through every aspect and probably given you a complete seminar of what I
tried to do. This goes into what is called toxic or dosage or concentration rel-
evancy.  How is a compound going to get there?  Is it going to be an effluent?
What is its concentration? Is its projected level one that you may  find in water?
   The one that we feel is the most important, at least that we use a great deal,
is based on acute toxicity to a component of our model.  We either used  inver-
tebrates or fish, depending upon the compounds.  This is a very important
consideration because it's quite easy to use a concentration, water concentra-
tion far and exceeding.
   One of the problems is consideration of such things as the introduction of the
chemical,  whether  it's going to be intermittent or  is it going to be continuous.
This is a very basic thing which some people call static versus continuous.
These are considerations with which you have to work.  We lean toward  being
continuous for most of our food chains, but we also have set-ups for  statics.
The chemical is introduced once, but as you deal with the organochlorines,
these get into all sorts of difficulties.

   In regard to the other question about things we can't do, I think the difficulty
 is an integrated thing where everything is all together at once.  We cannot do
   DR.  BUCKLEY:  Even if your chemical may be introduced only once, the
 metabolites or some of the other breakdown products might be continuously
 available to the next trophic level, or the next along the chain.
   DR.  JOHNSON: Yes, exactly.
   DR.  BUCKLEY:  We haven't heard very much from our industry counterparts
 in this microcosm business.  I'd really be interested in finding out to what extent
 microcosms and apparatus of that type are in any kind of use in industry. Are
 there any tests that are either routinely performed on a number of chemical
 types?  Can anybody shed any light on the use of these types of apparatus, out-
 side of the  Government or academic community?
   DR.  GELLETT: There's nobody here from Ciba-Geigy  which is the only
 company I've heard  that is doing it.
   MR.  NEWBY:  I'm Loy Newby, Ciba-Geigy.  We are not conducting eco-
 system  research. We are in the  process  of developing a modular system, that
 is, our  Basel laboratories are so engaged.  My concern here  is that although
 the ecosystem work is personally very interesting, when it comes down to a
 company, we meet with the Registration people (who are conspicuous by their
absence here) and essentially  the buck stops there. We do a dynamic study, or
we do some other aspect of a fish residue study.
   DR. BUCKLEY:  So you don't have any sort of special protocol or procedures
that you send prospective chemicals through and hope to come  to some under-
standing of  their possible ecological effects before you entered the registration
business ?

   MR. NEWBY:  Certainly the Guidelines and the Appendices list a wide range
of studies that we have to do.  We go beyond this to some degree, but not in the
microcosm or ecosystem area.  As Jim pointed out, he thought we were engaged
in such research because one of our colleagues from Basel recently visited
here,  trying to come up with ideas for a system which probably will be a modu-
lar type.
   You must take into consideration that in a modular system you may get a high
degree of accumulation which I'm not  always sure what it means, but I will agree
with Tom Johnson and Jim Sanborn, it throws up a flag.  We have some concern,
so let's next look at it from a more realistic standpoint. Let's apply it to the
soil, age the soil, then add water and fish, a bottom feeder,  determine how much
accumulates in the fish,  and you may  get a completely different story.
   DR, BUCKLEY:  The point is, you go through those sorts of steps.  How
elaborate are those steps,  or is that something you don't feel free to really go
   MR. NEWBY:  No, I can't go into all of it certainly.  We feel when we present
our data  in an environmental submittal, we have a pretty good idea of what that
compound is going to do  in the environment.
   DR. AMUNDSON: We are not involved in developing microcosms; however,
we have used them. We have sent compounds  to Dr.  Sanborn, Illinois Natural
History Survey; Dr. Kearney and his  associates, USDA; and to Bionomics for
studies of various kinds.
   In spite  of the emphasis on the development of ecosystem models for the
prediction of bioaccumulation,  one can get a good approximation of the poten-
tial of a compound to bioaccumulate by looking at water solubility, partition
coefficient, soil degradation, and animal metabolism.  Some compounds which

are quite persistent in the soil environment are readily metabolized by animals
and fish, so one has to evaluate several parameters.
   We feel that the current Guidelines for Registering Pesticides in USA are
fairly reasonable, and I feel industry can live with these Guidelines.  The im-
portant thing is interpretation by the Agency of the data which are submitted for
registration purposes.
   Dr. D. G. Crosby,  University of California, Davis, has developed another
index of environmental concern which he has applied to several pesticides.  This
work will be published by him in a paper entitled "The Toxicant-Wildlife
Complex." The formula for calculating the index includes solubility, P values,
and toxicity to selected aquatic species.  Perhaps it is another way of evaluating
compounds as they relate to the environment.
   QUESTION:  So it is a multidimensional thing?
   DR. AMUNDSON: That's correct, and I think the same thing is true with
the Bionomics fish accumulation,  with Jim's tests in Illinois, or with
Kearney's tests at Beltsville.  They are indicators, and we routinely send all
candidate  pesticides through one or more of these tests,  just to get some
idea  of what their problems might be later on.
   COMMENT:  From your comments I gathered that the new Guidelines are
really not that great a change from your current practices, that maybe they're
all spelled out in one place.
   DR. AMUNDSON: They've been used for some period of time. It's kind of
a continuum.
   COMMENT:  They appeared in the June 25th issue of the Federal Register.
There are 60 days through August 26th  within which to make comment regarding
the content of the Guidelines and Appendices.  It's worth everybody's time to

look at them and comment if you have an opinion regarding their completeness
or content.
   QUESTION:  I would like more guidance not on just what to do or how to do
it, possibly what the environmentalists feel would be appropriate, but also how
I can judge and determine the significance of every experiment I do.
   Obviously with my limited exposure to things that I'm just doing, or seeing
something in the literature that comes out several years later after it's been
done,  I can only judge from a limited point of view.  I'm not asking EPA to
make  a business decision for me which really is my decision in upper manage-
ment.  I'm not just asking for guidelines that suggest screening procedures or
tests, I would like to be aware of the significance of those tests with some handle,
qualitative or quantitative, so that when I see something, I can make some judg-
ment about this, whether it is something that I should be concerned about or do.
On a couple of occasions I have called Ron Ney and said, "Okay, Ron, what does
this mean?"
   The answer is submit the thing and when it comes through and we review it,
then we'll tell you what it means.  That takes time.  Time is a lot of money,  not
only from the standpoint of the work we'd be continuing to do on that project,  but
because there may be other priorities to which you'd like to switch.  Time is
money being spent to continue a given project for another year or year and a
half before you get a readout from the reviewers at EPA as to what it means.
I would like to be in a position to be able to  make some assessments of my own,
more  than just gut feelings,
   COMMENT: I think in Cleve Goring's book there is an environmental profile
that lists water solubility, volatility, toxicity to quail, and a number of other
factors.  There are little bars, and it's all listed in terms of either a concen-
tration or an LD.   It's always a concentration, and you can draw a little

stream going down through there and if you're in the stream, you're pretty
safe through all of these factors.  If you have one or more of these that fall
outside of that little directional line, then you should start querying the thing.
These relatively simple factors  which you do early in the game are less ex-
pensive than a two-year carcinogenic study.  But you really would like to know
what it means,  like in simple octanol/water.
   MS. SCHUTH:  What you're asking is what are the criteria on which de-
cisions are  based? That's obviously the job of the Criteria  and Evaluation
Division and they have written some of the criteria and are continually involved
in this process. For example,  they've written criteria which involve soil
persistence. That's what you're asking for and that's what you're getting.  It
takes a lot of time and it's a  big job.
   Bioaccumulation is one of those concerns which is a current item for criteria
development at the present time. I think as an  industrial representative it is
certainly within your  province to push on the Registration Division people and
ask, "What  are the criteria that you're using?" They can in turn push on the
Criteria and Evaluation people to define philosophy within which your results
will be interpreted.  Even if  they wanted to give you a number, I don't think you
want a number because  you have to evaluate each chemical according to its use
as well as other things.
   Those criteria are forthcoming.  I agree it would be nice to have them all
now.  The Criteria and  Evaluation Division has had a lot of other jobs to do too
and these criteria maybe haven't received as much attention as they deserve.
But push for it,  just say to the Agency, "It's registration criteria that we need.
These  delays are keeping our products off the market."
   COMMENT:   All I'm  saying is there should be a mechanism and hopefully it
will be forthcoming,  partly through the benchmark chemistry program, partly

through other things, where you can draw profiles of things, something where
we can at an early stage in the game assess what we're going to do with the
chemical, not just from the process standpoint but from all these things.
   COMMENT:  That's one of the reasons for the microcosm approach.  The
idea is that it may not tell you when you've got a completely safe compound.
There may be things that you would overlook.  You would tend to err in that
direction.  The theory is that you're going to detect any significant problems
early, using some kind of integrative mechanism like a vole or whatever the
organism might be that's in your microcosm.  That tells you that you better
do a little more extensive testing on that material before you go  too much
further with it.
   QUESTION:  Since I don't really work in the system, I'm not really  aware of
the disadvantages of the system,  or as Dr. Buckley had said, what is it telling
you and what is it not telling you? Is  it possible that by trying to integrate all
these species all at once, and then treating the whole system with the pesticide,
you've 'totally distorted the thing? Are we distorting the system because you're
exposing everything all at once to it and that doesn't really happen always in the
   COMMENT:  The answer is very definitely yes, we're distorting it, for the
same reason that we do high chronic dosage studies  and 2-year feeding trials
to try to accelerate toxic advantage, or in drug testing do accelerated  90-day
feeding studies, logarithmic increases of doses and  so forth.  We  have to
reach some sort of conclusion quickly. Like you say, time is money.  We're
accentuating the biotic elements in terms of population densities, in terms of
interactions.  The processes which we're trying to represent in the  environ-
ment are yet there.  Predation, for example, is a process. Soil binding is a
process. Leaching is a process. Volatilization is a process.  We're including

 these processes in the systems.  This is the reason that Allan has a very
 complex, physical structuring because he doesn't want predation at certain
 stages of development, but he does want predation of the adults.
   He has to construct his flow-through system to get a certain kind of
 biological process simulated within the laboratory system. We have to have
 intensity of this process at a rate that we  can measure. If we had absolutely
 perfect simulation counters and no quenching and good carryover  of all of our
 material, if we could get the technical material labeled in the same way that we
 can get the pure quail to put in  our chambers,  we would improve the sensitivity
 of our analysis in terms of the  one percent or  the 10th of a percent impurity.
   I would like to see us be able to do that.  I would say from a microcosmic
 experiment,  one could predict very accurately whether or not one should run
 extensive leaching studies or a very simple, short-term leaching study.  If
 you have 65 percent loss  of dieldrin in 30  days or 20 days, I would suggest
 that you do a lot more photochemistry and a lot more  soil loss and a lot more
 loss from glass plates studies than you might want to  do for a compound for
 which you had 10 percent loss in the total  system or 10 percent unaccountability
 in your  total system in 90 days  or 60 days.
   If we can increase the intensity of the interactions,  particularly the biological
 interactions which is where the action is in terms of ecosystems effects, then
 we can come up with better rate constants.  Maybe these are not equivalent to
 the rate constants of nitrification or ammonification in the environment, or of
 predator-prey  relationship, but they are representative on a comparative
 basis.   We use dieldrin as a standard, they use DDT as a standard,  Pete can
 use mirex.  We can standardize it against a reference chemical,  and as long as
 we understand what the between system meaning is, then we can make some
 judgment as to what sort of questions we want to investigate in a much more
sophisticated, detailed way, using a specialist in that  area.

   I'm not a photochemist and I haven't worked in photolysis for 13 years, and
yet I know that if we look at the TLC profile of a microsomooxydase-treated
preparation,  or of a glass plate-photolyzed preparation, we're going to see
similar qualitative, metabolic patterns, and we know where to go from there if
we  really want to know what those chemicals are.  If you have a chemical that
looks good in the field and in other applications,  you're going to follow it out
that far. I've seen the work that you do and you follow it down the line until
you find no toxicity in your products.
   COMMENT:  I'd follow that up quickly.  My point was that the whole system —
all  of the vertebrates,  all of the invertebrates,  the whole works — is  totally
exposed to the maximum level of the pesticide all at once, whereas that may
not really be the situation.  You said, yes, we have to stress the situation in
order to get an answer out  of it. Are the numbers, whatever is generated from
each individual species, weight averaged?  In other words, I've seen the  results
from Metcalf's study on our one compound, and I see the numbers  for all of the
species. They are submitted, and the problem is what weight do you place upon
each species now?  It may  be so much farther down the food chain, although it
has a high number, that it would have never gotten to that point.
   My point about an artificial situation is that maybe it's in the mechanics of
dealing with the  numbers that you get from there to get the final, index again,
but that it has to be weight averaged in order to really look, because something
that you  say is bad for that one small microorganism may never get there.
    COMMENT:  In any kind of modeling context, whether it's mathematical or
an econometric model, for example, on locating an industry, there are all
kinds of simplifications built into that sort of thing, and I don't see how you can
avoid that.   Models are to give you general ideas or general indications of
whether you've got a problem or not.

   QUESTION:  Is this one aspect of half a dozen different systems that exist
when comparisons are made between different systems?  Will it ever get to a
point where one system is considered to be the best and everybody follows that
as a standard, or will there be several systems that get to be best?
   COMMENT: That will happen.
   QUESTION:  Weight averaging might be an inclusion in the interpretation of
the data.  That will  be one point.  We talked about interrelations this morning
of different factors and mathematical model.   These factors were not spelled
out exactly, but I got the impression they were exclusively scientific points of
data. However, there are other social and economic factors that are actually
related to those scientific points.  In other words, how many millions of acres
of corn are planted or destroyed? How many people does that wind up feeding?
What is the starvation rate? How many jobs  does that affect? How many kids
are born?  It can go on and on.  Those are points which maybe we as scientists
say, "That's not for me to judge," but they have to be -included somewhere.
   COMMENT:  That concept is really the general Qase of how you approach a
model from a topological point of view,  and the model could include socioeco-
nomic characteristics.
   COMMENT:  That I didn't know.
   COMMENT:  Production.
   COMMENT:  That's what I wanted  to know.
   COMMENT:  Any number of features.  The method itself can be generalized
to a model  of any degree of complexity.