EPA-600/3-84-013
January 1984
STRUCTURE AND ORGANIZATION
OF PERSISTENT AQUATIC LABORATORY
COMMUNITIES EXPOSED TO THE
INSECTICIDE DIELDRIN
by
William J. Liss, Becky L. McClurken-Lilley and
Douglas S. Lee
Oak Creek Laboratory of Biology
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon 97331
Cooperative Agreement CR80745702
Final Report
Project Officer
Steven F. Hedtke
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-84-013
3. RECIPIENT'S ACCESSION NO.
PHP & 141.183
4. TITLE AND SUBTITLE
Structure and Organization of Persistent Aquatic
Laboratory Communities Exposed to the Insecticide
Dieldrin
5. REPORT DATE
January 1984
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W. J. Liss, B. L. McClurken-Lilley, and D. S. Lee
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
Oak Creek Laboratory of Biology
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon 97331
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR8.07457
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-600/03'
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Sixteen aquatic communities composed of persistent populations of guppies (Poecllia
reticulata), amphipods (Gammarus fasciatus), snails (family Planorbidae), planaria
(Dugesia sp.), and various microinvertebrates were established under laboratory
conditions. Eight of these communities received low energy input with low habitat
availability while the remaining eight communities had high habitat availability and
received high energy input. At each level of input, guppy populations in two systems
were exploited at either 0, 10, 20, or 40 percent of the population biomass each
month. Macroinvertebrates were also sampled monthly for population counts and
biomass measurements. After each system reached near steady-state conditions, 1 pg/1
of dieldrin was introduced into one system of each treatment. These systems were
allowed to reach new steady-state conditions.
The response of the systems was dependent upon both the energy input/habitat and
exploitation levels. The low energy input/habitat systems were more sensitive to
dieldrin particularly at higher levels of exploitation. The influence of
organization and environment on population persistence and system structure are
explored theoretically with isocline models and the implications for aquatic ecology
and environmental management strategies are discussed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Croup
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
INTRODUCTION
Management of toxic substances must be based upon good understanding
of their effects not only on individual organisms but also on populations
and communities. Accomplishing this presupposes some theoretical perspective
for understanding populations and communities. In this report we present
our perspective, apply it in understanding the structure and organization
of laboratory communities exposed to a toxicant in relation to the
environments of the communities, and explore some of the broader implications
of this perspective for understanding and control of toxic substances. Much
of the report deals with research conducted as a part of a 5 year project
supported by USEPA. The goal of the project was to advance understanding
of the capacities and performances of individual organisms, populations,
and simple communities exposed to toxicants. The first three years of
this work, supported by a research grant, was reported by Liss et al.
(1980). The work reported here deals with research conducted over the last
two years of the project.
The objectives of that research were:
1. Determine and explain, under different sets of environmental
conditions, the effect of the insecticide dieldrin on the
persistence, structure and organization of laboratory
communities, including the ability of the systems to recover
from toxicant perturbation. Environmental conditions include
different levels of habitat availability and rates of energy
and material resource input and different rates of population
exploitation.
2. In ancillary experiments, determine and explain, under different
sets of environmental conditions, the effects of dieldrin on the
-------
persistence, structure and organization of systems that are of
simpler organization and shorter duration than the laboratory
communities, yet still capable of representing population and
simple community responses. The response to toxicants of these
simpler systems will be compared to toxicant response in the more
complex systems to determine the extent to which simple models
can provide some understanding of toxicant effects in complex
systems.
Some of the results from the first three years work must necessarily
be included here to provide continuity and background for interpretation of
the last two years of research.
We view the perspectives for understanding populations and communities
and any possible utility they may provide as a way of thinking about toxic
substance effects in relation to system structure and organization, system
dynamics, and system environment as being more important than the particular
empirical results reported here. In essence, the laboratory community
research was designed, conducted, and interpreted according to our perspective.
The empirical research is useful for evaluating the utility of the perspective
for understanding systems and its conformity with observational experience.
The more complex laboratory communities discussed here are not necessarily
intended as tools for testing toxic substance behavior and effects.
The performance of a system is anything the system does as a whole.
Performances include structure or state, change in structure, yield, and
persistence. The performance of a system at a particular time can be under-
stood as being jointly determined by the capacity of the system and conditions
in its environment at that time (Warren et al. 1979). A system with a given
-------
capacity will exhibit different performances (or different magnitudes of a
given kind of performance) under different sets of environmental conditions.
In this sense, the capacity of a system can be understood to be all of its
possible performances in all possible environments. The performance of a
system will change if either its capacity or its environment changes.
The way that a system is organized determines its capacity to perform.
Thus any performance of a system, including its response to a toxicant, can
be understood to be jointly determined by the organization (capacity) of the
system and conditions in the environment of the system. As individual
organisms, populations, communites or any other natural system develop or
evolve, their organization and thus their capacity to perform changes.
Exposure to toxicants, exploitation, and other affects of man on natural
systems may substantially alter the capacity of the systems. These are the
most profound effects that man's activities may have on natural systems
for if capacity is significantly altered, man's effects on natural systems
may be largely irreversible (Warren et al. 1983). Thus, a system whose
capacity has been fundamentally changed may not be able to recover or return
to its original or previous state(s) even if its environment were returned
to its previous state(s).
System organization we take to be a theoretical concept entailing
incorporation, concordance, and interpenetration of the capacities and
performances of the subsystems and their environments (Warren et al. 1983).
Structure—the apparent form of the system as a whole—we take to be a more
empirical, observable performance than organization. For relatively simple
systems, such as we will deal with here, structure can entail the kinds of
species composing the system and their distribution and abundance in space
-------
and time. We will deal theoretically and empirically with both steady-state
and dynamic system structure. Organization of such simple systems can
entail the ways in which the species are incorporated or unified into a
system, with emphasis on their capacities in relation to the capacity of the
system as a whole, the concordant or harmonious, rule-like relations between
the species that integrate or link them together to form a system, and the
interpenetration, permeation, or interspersion of species populations,
emphasizing that interactions between different species consist of inter-
actions between individuals or groups of individuals of the different species,
Population interactions such as predation and competition play an important
role in organizing the laboratory communities.
Mathematical models can be used to symbolize, partially articulate,
and provide a perspective on system structure and organization. The
following generalizations pertaining to system structure and dynamics in
relation to system environment will be illustrated with isocline models
and demonstrated empirically in the laboratory communities:
1. Under different states of the environment of a natural system,
the system will come to have different steady-state structures
and organizations and thus can be understood to be a multisteady-
state system.
2. Dynamics or changes in structure of an n—dimensional system can
be understood as an n-dimensional trajectory in phase space in
continuous pursuit of an n-dimensional steady-state point whose
location in phase space is continually changing as a result of
changes in the state of the environment of the system.
-------
Isocline models of the kind to be discussed here provide one of the
principle bases for both illustrations of these generalizations and inter-
pretation of the laboratory community results. More detailed information
concerning model derivation can be found in Booty (1976), Liss (1977), and
Thompson (1981).
Isoclines and phase planes representing a simple predation system is
shown in Figure 1. Systems with somewhat more complex organization than that
shown in Figure 1 can also be represented with systems of isoclines on phase
planes, but the predation system will suffice to illustrate the generalizations.
The intersecting isoclines define systems of steady-state relationships
between the populations composing the system. The position and form of the
isoclines is determined by systems of graphs or equations representing the
rates of change in biomass of the populations composing the system (Booty
1976, Liss 1977, Thompson 1981). A complete set of isoclines on all phase
planes provides at least a partial view of the structure and organization
of a system in relation to environmental conditions.
The structures of systems are in continuous developmental and evolu-
tionary change. Even so, these systems can be understood ideally as being
multisteady-state systems. For a system to be understood as being a
multisteady-state system it is necessary only to be able to conceive of
the environment of the system as having different states. In the example
shown here (Fig. 1), for each set of environmental conditions I and E, that
is, for each state of the environment of the system, there exists a single
steady-state point on each phase plane, each of these points being a two-
dimensional projection of a single, multidimensional system steady-state
point in phase space. The set of these two-dimensional points defines the
-------
Figure 1. Phase planes and isocline systems representing the inter-
relationships between populations in a system represented as
where C,H,P, and R comprise the system and E, units of harvest-
ing effort, and I, rate of input of plant resources, are factors
in the environment of the system. Predator biomass is plotted
on the y-axis of each phase plane and prey biomass is plotted on
the x-axis. On each phase plane, the descending lines identified
by different rates of plant resource input, I, are prey isoclines.
Each prey isocline is defined as a set of biomasses of predator
and prey where the rate of change of prey biomass with time is
zero. The ascending lines on each phase plane are predator
isoclines. Each predator isocline is defined as a set of bio-
masses of predator and prey where the rate of change of predator
biomass with time is zero. Each intersection of a predator and
prey isocline is a steady-state point where the rate of change
of both predator and prey biomass with time is zero. The
positions and forms of the isoclines can be deduced from equations
or sets of graphs representing the rates of gain and loss of each
of the populations (Booty, 1976; Liss, 1977). At a particular
level of I and E, a single steady-state point exists on each
phase plane, the set of these points defining the steady-state
biomasses of C,H,P, and R. The points that define the steady-state
biomasses of C,H,P,R at Med I and OE (circles), 30E (squares),
90E (triangles) and 150E (hexagons) are shown. In this simple
system, at each input rate, increased E brings about reduction
in steady-state biomass of C, increases in H, reductions in P,
and increases in R. Increases in I shift the steady-state
relationships between predator and prey to the right on each
phase plane, essentially increasing the biomasses of C,H,P, and
R. After Liss and Warren (1980).
-------
LOW MED HIGH
5r
cs>
V)
O
m
UJ
tc. '
ui
1234
HERBIVORE BIOMASS (H)
LOW MED HIGH
I I 1
1234
PLANT BIOMASS (P)
OH
I 23 4
PLANT RESOURCE (R)
-------
steady-;state structure of the system for a given set of environmental con-
ditions. Changing the state of the system's environment (for example,
from Med I, OE to Med I, 30E or Med I, 150E) brings about a change in the
steady-state structure of the system. A system may have an infinite
number of possible steady-state structures.
In the predation system shown here, a single system steady-state point
exists for each environmental state. Systems represented with different
sets of equations or with different values for the constants in the equations
may have more than one potential system steady-state point at some or all
environmental states. That is, isoclines need not have the uniform,
monotonically increasing or decreasing shapes shown in Figure 1. They may
be bowed or looped (e.g. Rozenzweig, 1968), thus creating the possibilities
for multiple intersections for a given set of environmental conditions.
For the most part, this will not affect the general conclusions we wish to
draw and these conclusions can be most clearly illustrated with systems
having single system steady-states for a given set of environmental conditions,
On each phase plane, trajectories represent the changes through time
in biomasses of the populations composing the system. If environmental
conditions are fixed, trajectories on each phase plane will converge upon
the steady-state points locating the steady-state structure of the system
under those conditions (Fig. 2). In natural systems, environmental con-
ditions are rarely constant for long enough periods of time to permit
systems to reach steady-states. Thus, trajectories are in constant pursuit
of a steady-state point whose location in phase space is continuously
being shifted as a result of changes in environmental conditions. For
an n-dimensional system, the trajectory on each phase plane is a two-
dimensional projection of an n-dimensional trajectory in phase space.
8
-------
Figure 2. Phase planes and isocline systems illustrating a possible effect
of different concentrations of a toxicant (T) on the structure
of a simple community. In this example the toxicant directly
affects only the carnivore population. The presence of the
toxicant lowers the predator isocline at each E, the extent
to which it is lowered depending upon the effect of the partic-
ular toxicant concentration on carnivore growth, reproduction,
and survival. Steady-state structure at HIGH I, OE, OT (circles);
HIGH I, OE, 2T (squares); LOW I, 90E, 2T (triangles) is shown.
Trajectories of biomasses of carnivore (C), herbivore (H),
plant (P), and plant resource (R) originating at point 0 are
shown to converge on each of these steady-states under each
particular set of environmental conditions. Introduction of
toxicant at a concentration of 2T reduced carnivore biomass,
increased herbivore biomass, reduced plant biomass and increased
plant resource at each combination of levels of I and E (for
example, compare circles and squares at HIGH I, OE on all
phase planes). After Warren and LLss (1977).
-------
5-1
LOW MED HIGH I
LOW MED HIGH I
LOW MED HIGH I
2.5 H
3.5 H
5H
-------
Introduction of toxic substances can alter the structure and organization
of systems (Fig. 2). The response of systems to toxic substances is affected
by their organization and conditions in their environments such as I and E
(Warren and Liss, 1977). For example, in Figure 2, the carnivore population
is able to persist at LOW I when it is heavily exploited (90E) if a toxic
substance (T) is not present (the predator isocline identified by 90E, OT
intersects the prey isocline identified by LOW I). But, under these same
environmental conditions at a toxicant concentration of 2T, the carnivore
population is driven to extinction (the prey isocline identified by LOW I
does not intersect the predator isocline identified by 90E, 2T; C trajectory).
Under different sets of environmental conditions the carnivore is able to
persist, although at reduced biomass, at a toxicant concentration of 2T.
Persistence at 2T is possible at LOW I when the carnivore is unexploited
(OE), or at MED I and HIGH I when C is heavily exploited (90E). The
laboratory communities provide an empirical demonstration of the importance
of system organization and system environment in determining response to
toxicants.
Both predation and competition are important in organizing the labora-
tory communities and determining their response to a toxicant. If an
exploited competitor C2 utilizing prey species H were added to a system
such as that shown in Figure 1, the prey isoclines at each I wpuld, in
effect, "explode" into an infinite family of prey isoclines (Booty 1976,
Liss 1977, Thompson 1981), each parameterized by a particular level of
harvesting effort on the competitor (Fig. 3). In this particular example,
addition of a competitor shifted to the left the prey isoclines defining
the steady-state relationships between Cl and H at each I, so altering
steady-state Cl and H densities and consequently the densities of other
11
-------
Figure 3. Phase plane and isocline systems showing the steady-state relation-
ships between a carnivore Cl and its prey H in the system
C2 is a competitor of Cl for H and is exploited by E2. El and
E2 are different harvesting systems. Prey isoclines become
parameterized not only by I but also by E2, the number of units
of effort harvesting C2. When C2 is not present (OC2), steady-
state densities of Cl and H are indicated at LOW 1, 2E1 (open
square), LOW I, 3E1 (open triangle), HIGH I, 2E1 (solid square,
and HIGH I, 3E1 (solid triangle). Addition of C2 shifts the
prey isoclines to the left at each I lowering the steady-state
biomasses of Cl and H at each El (for example, compare biomasses
of Cl and H at LOW I, 2E1, OC2 — the open square — with the biomasses
of these populations when C2 is present and unharvested at LOW I,
2E1, OE2 — the vertically-barred square). The greater the intensity
of harvesting of C2 (the lower its biomass is maintained) the less
the prey isocline is shifted to the left at each I. This example
illustrates the particular effect of introducing a competitor
into a system such as that shown in Figure 1. The impact of
introduction of a competitor may vary considerably depending upon
the organization of the system into which the competitor is
introduced. Isoclines on this phase plane were generated through
computer iteration using the simulation command control language
SIMCON (Worden 1976, Thompson 1981). (We appreciate the assis-
tance of Grant Thompson, Oak Creek Laboratory of Biology, in
generating this phase plane.) After Lee (1983).
12
-------
HIGH I
OE2 OC2
2E1
3E1
13
-------
species in the system. Note that the prey isocline identified by LOW I,
OE2 does not intersect the predator isocline identified by 3E1. Under
these conditions Cl and C2 cannot coexist, Cl being driven to extinction
by the presence of an unharvested competitor (vertically-barred triangle).
Cl and C2 can coexist if environmental conditions change in such a way
that predator and prey isoclines intersect in positive phase space. This
can occur if 1) El is reduced from 3E1 to 2E1 (vertically-barred triangle
to vertically-barred square), 2) E2 is increased from OE2 to 1E2 (vertically-
barred triangle to horizontally-barred triangle), and 3) I is increased
from LOW I to HIGH I (vertically-barred triangle to stippled triangle).
Thus competitive coexistence is dependent upon the levels of I and E.
A generalized isocline model summarizing some possible effects of
changes in system environment and organization on steady-state relationships
between a predator and its prey is shown in Figure 4. Changes in the
levels of any of the factors identifying the predator and prey isoclines
can alter the position and form of the isoclines and so shift the location
of the steady-state point in phase space, thus altering system steady-state
s tructure.
Not only the presence of a competitor of the predator, as is shown in
Figure 3, but also the presence of competitors of the prey or trophic levels
leading to the prey can alter the position of the prey isoclines at each I
(Booty, 1976). Physico-chemical factors (temperature, oxygen, etc.) affecting
the prey or trophic levels leading to the prey can also bring about shifts
in the prey isoclines at each I. Physico-chemical factors directly affecting
the recruitment, growth, or survival of the predator can alter the position
and form of the predator isoclines. Species that are prey of the predator
but do not compete with other prey can also bring about shifts in the predator
14
-------
Figure 4. A. Generalized phase plane and isocline systems representing
the interaction between predator and prey, and some possible
effects of energy and material input rate (I), competition,
physico-chemical factors, toxic substances, exploitation (E) ,
and alternative prey. Steady-state points are indicated at
the intersections of predator and prey isoclines. At each I,
presence of a competitor, toxic substances, or physico-chemical
factors affecting the prey can shift the prey isocline on
the phase plane. For example, the presence of a competitor
or a toxic substance affecting the prey may shift the prey
isocline to the left at each I (solid prey isocline to dashed
prey isocline). Physico-chemical conditions more favorable
to the prey may shift the prey isocline to the right at each
I (solid prey isocline to dotted prey isocline). At each E,
presence of alternative prey and toxic substances directly
affecting the predator can shift the predator isocline. For
example, the presence of a toxic substance directly affecting
the predator may shift the predator isocline downward at each
E (solid predator isocline to dashed predator isocline at
20E and 40E). This is the response shown in Figure 2.
Physico-chemical conditions more favorable to the predator
may shift the predator isocline upward at each E.
B. Predator steady-state yield curves and how the magnitudes
of these curves can be affected by changes in I, competition,
toxicants, physico-chemical factors, and alternative prey.
Each curve is derived from a prey isocline on the predator-
prey phase plane. After Liss (1977).
15
-------
LOW I
HIGH I
g
oo
cr
o
Q
UJ
a:
a.
'\ Competition
Competition Physico-Chem Factors
Physico-Chem Factors Toxicant
Toxicant - ., „
40 E^ Alter Prey
1 |_ Physico-Chem Factors
-\--.-40E,TOX
PREY BIOMASS
Q
_l
LJ
o:
Q
UJ
rr
a.
HIGH I
PREDATOR BIOMASS
16
-------
isoclines. A toxic substance affecting the prey or lower trophic levels
leading to the prey can alter the prey isoclines at each I. A toxicant
directly affecting predator survival, growth, or reproduction can bring about
shifts in the position and form of the predator isoclines at each E, as
s hown in Figure 2.
From each prey isocline a steady-state predator yield curve can be
derived as shown in Figure 4 (Liss 1977, Thompson 1981). Changes in the
magnitude of factors parameterizing prey isoclines (I, competition, etc.)
shift the location of the prey isoclines and so lead to changes in the
magnitude of the yield curves. Changes in the magnitude of factors para-
meterizing predator isoclines can also lead to changes in the magnitude
of the yield curves if, at each E, they alter the location of the steady-
state point on a particular prey isocline and thus alter the steady—state
biomass and yield the population is able to maintain at that particular E.
Toxic substances directly affecting a predator and/or affecting its prey
or trophic levels leading to the prey can bring about changes in the
magnitude of predator yield curves.
The phase planes and isocline system in Figure 4 depict some of the
kinds of factors that bring about shifts in the location of predator and
prey isoclines and so affect system structure. However, with this model
or any other, it is not possible to generalize about the particular effect
of any factor on system structure. For example, we cannot conclude that
the addition of a competitor will, in all kinds of systems, shift prey
isoclines to the left and reduce steady-state predator and prey densities,
although this may be an intuitive result. The particular effect on system
structure of the addition of a competitor or other parameterizing factor,
17
-------
as indicated by the magnitude and direction of shift of predator and prey
isoclines, depends upon the organization of the system, that is, upon the
other kinds of species composing the system and the nature of the inter-
actions or interrelationships between them (Thompson 1981).
18
-------
MATERIALS AND METHODS
Laboratory Community Experiments
Sixteen aquatic communities were established, each composed of persistent
populations of guppies (Poecilia reticulata), amphipods (Gammarus fasciatus),
snails (family Planorbidae), planaria (Dugesia sp«), and benthic microinverte-
brates including flagellates, rotifers, nematodes, gastrotrichs, and protozoans
(Liss et al. 1980). Green and blue-green algae and diatoms were present.
Habitat and escape cover for invertebrates was provided by a substrate of
1.5 cm quartzite gravel four cm. deep. A gelatinous mixture of 60 percent
alfalfa and 40 percent Oregon Test Diet (Sinnhuber et al. 1977) served as
primary energy and material input in the laboratory communities. The
nitrogen content of the alfalfa ration was 3.0 percent.
Each laboratory system was maintained in a fiberglass tank measuring
1.2m x l.lm x 0.4m and holding 560 liters of water. This was continuously
exchanged by a 600 milliliter per minute flow of heated well water. Water
temperature (21 ± 1°C), dissolved oxygen (8.2 ± 0.5 ppm), and pH (7.8 ± 0.1)
were maintained at nearly constant levels. Light was provided by fluorescent
lights placed above each tank. Intensities ranged from 15 to 23 foot-candles
at the surface of the water. Light was maintained at this low level to prevent
blue-green algae blooms. Photoperiod was controlled by a timer set for 14 hours
light and 10 hours darkness.
Initially all 16 laboratory systems were established with three circular
nests of quartzite gravel covering 20 percent of the bottom area of each tank
and each received a 0.6 gram per tank daily alfalfa-OTD ration. This treat-
ment is identified as low energy and material input and low habitat availabi-
lity, or LOW I. In April 1975, 200 amphipods were stocked in each system, with
19
-------
the size distribution of amphipods introduced into each system being similar.
Sediment accumulation and development of a benthic microflora and microfauna
ensued. Snails and planaria inadvertently colonized all the systems and
became major components of some of the communities. By November 1976, groups
of 37 guppies (4.5 grams), each with a similar size distribution and sex
ratio, had been stocked in all systems. Four guppy populations were exploited
at one of four rates, 0, 10, 20, or 40 percent of the biomass of the popula-
tion present at the time of sampling (OE, 10E, 20E and 40E, respectively).
The systems were sampled every 28 days.
In March 1978, eight of the laboratory systems (two at each guppy
exploitation rate) were modified to establish a higher level of energy and
material input and habitat availability. The gravel habitat and escape cover
was increased to cover 95 percent of each tank bottom. Energy and material
input was increased to 4.0 grams of alfalfa-OTD ration daily. This treatment
will be identified as HIGK I. Planaria are extremely effective predators on
amphipods and were capable of driving amphipod populations to extinction if
left uncontrolled. At HIGH I, planaria became abundant and in four of these
systems, one at each guppy exploitation rate, planaria control was instituted
by manually removing planaria during sampling. This proved to be an effective
means of planaria control. Planaria were either absent or maintained at low
densities in planaria-controlled systems (Table 2).
In April 1978, when four of the systems at LOW I (one at each guppy
exploitation rate) established near steady-states, continuous introduction of
one ppb of the organochlorine insecticide dieldrin was begun (Liss et al.
1980). Near steady-state (NSS) structure for the laboratory systems under
each set of environmental conditions (I and E) was assumed when the trajec-
tories of biomasses of the interacting populations in the system fluctuated
20
-------
in a very restricted region of phase space relative to previous fluctuations.
When the systems established NSS's, 10 to 18 monthly measurements of structure
were necessary to adequately define the domain of NSS behavior. Dieldrin
introduction into these systems was terminated in October 1979 and recovery
of the systems from dieldrin perturbation was observed through September 1981.
Continuous introduction of one ppb of dieldrin into the four systems
at HIGH I in which planaria were controlled began when these systems estab-
lished NSS's. Dieldrin introduction into the systems whose guppy populations
were exploited at OE and 10E began in September 1981. Dieldrin was intro-
duced into the remaining two systems (20E and 40E) in January 1982. Dieldrin
was continuously introduced into all of these systems through September 1982,
when the experiment was terminated.
Leeches and the amphipod Hyallela azteca inadvertently colonized several
of the systems. Leeches prey on young snails and if they reach high densities
they can reduce snail biomass. They increased in abundance in several systems
in late 1980 and, in early 1981, leech control was instituted. Control was
very effective in most systems, eliminating leeches entirely or severely
reducing their abundance (Table 2). Hyallela began to colonize the systems
in late 1979 and early 1980. This species became abundant in several systems
and appears to have had an impact on system recovery at LOW I and on NSS
structure at HIGH I.
Organisms in each laboratory system were censused and its guppy
population exploited every 28 days. Guppies were netted from the tanks.
Individual length and weight measurements were taken for immature guppies
(standard length of 10 to 14 millimeters) and for mature females (standard
length greater than 14 millimeters). Total number and weight were recorded
for mature males and for newborns (standard length less than 10 millimeters).
21
-------
The method of exploitation was similar to that employed by Liss (1974).
A systematic exploitation schedule was developed for each exploitation rate.
The schedule provided for the removal of a proportion of the population
corresponding to the exploitation rate. At each exploitation rate, all sizes
of fish were exploited with the same intensity.
Monthly exploitation of the guppy populations simulated the impact of
harvesting by man or natural mortality. Heavy exploitation resulted in size
distributions occasionally having up to 80 percent of a population's biomass
residing in one large female. More or less than the intended percentage of
biomass was thus often exploited in any one month. This led to some fluctu-
ations in population biomass; over many months, however, mean exploitation
rates were near the intended percentages.
Population number, biomass and yield (i.e. the biomass of the catch
at a given sampling date) was determined for each guppy population. Size-
specific fecundity of exploited females was also determined.
Sampling procedures also included the temporary removal of all benthic
invertebrate populations, sediment, and gravel substrate. Individuals of
all macroinvertebrate populations (snails, amphipods, planaria) were sized
and the number and biomass of each size group was determined. From this,
population number, biomass, and size structure were determined. All amphipods,
snails, sediments, and unexploited fish were returned to the tanks after
s ampling.
The insecticide dilution and delivery system was similar to the continuous
flow dilution apparatus described by Chadwick et al. (1972). A solution having
a constant toxicant concentration was produced by passing water through
a column of 1.5 cm quartzite gravel coated with technical grade dieldrin.
22
-------
Concentrations of dieldrin in the water were determined weekly. Following
standard extraction procedures, dieldrin analysis was done using a Varianaero-
graph 2000 gas chromatograph equipped with an electron capture detector.
The work to which the objectives stated in the Introduction pertained
was originally proposed to span a three year period. During the second year
funding was reduced. At the end of the second year, the Cooperative Agreement
was terminated due to lack of funds. Due to reduction in funding and early
termination, analysis of the concentrations of dieldrin in the tissues of
organisms was not possible.
Ancillary Experiments with Laboratory Communities of Simple Design
The laboratory communities used in these experiments were intended to
serve, in some sense, as "microcosms of microcosms", that is, as systems that
are simpler in design than the larger, more complex laboratory systems, but
still capable of representing population and simple community behavior and
response to toxicants. Toxicant behavior and effects in these simpler systems
were then to be compared to the behavior and effects of the toxicant in the
more complex laboratory systems. The comparison was to be made in order to
determine the extent to which simpler models of complex systems could provide
some understanding and prediction of toxicant behavior and effects in the
complex systems.
Systems composed of only guppy populations, only snail populations, and
guppy and snail populations together were established (Table 1; Lee 1983).
Guppies and snails were the most abundant populations in the more complex
laboratory systems and were competitors for the alfalfa ration. Energy and
material input into these systems was in the form of an alfalfa-OTD ration,
which was introduced at rates that were the same as the rates of introduction
23
-------
Table 1. Laboratory ecosystems designed for ancillary experiments. After
Lee 1983.
System
snail-alfalfa
guppy-alfalfa
guppy-s nail-alfalfa
Number of Tanks
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Input Rate
4.0
0.6
0.6
0.6
0.6
4.0
4.0
4.0
0.6
0.6
0.6
4.0
4.0
4.0
2
Exploitation Rate
_
—
0
25
40
0
25
40
0
25
40
0
25
40
1) gms alfalfa day
2) % biomass removed month"
24
-------
into the more complex laboratory systems. Guppy populations were exploited.
Water temperature, light intensity, and light-dark cycle in the ancillary
experiments were the same as in the more complex laboratory systems.
Each system resided in a 40 liter glass aquarium that had been adapted
for flow-through usage. Each aquarium received 200 ml min"1 of well water
at 21°C ± 0.5°C and was exposed to a 14/10 hour light-dark cycle. The
systems received light intensities of 25 foot-candles. Each tank had approx-
imately 3,450 cm^ of cover available to newborn guppies and snails in the form
of floating plastic plants.
Fourteen of the systems received a low rate of energy and material
input in the form of an alfalfa ration (60% alfalfa and 40% Oregon Test Diet)
and the other fourteen systems received a high rate of energy and material
input (Table 1). These treatments will be identified as LOW I and HIGH I,
respectively.
Twelve of the systems, six at LOW I and six at HIGH I, had only guppy
and algae populations, with guppy populations in two systems in each subgroup
of six being exploited at 0 percent, 25 percent, or 40 percent of the popula-
tion biomass present at the time of sampling (OE, 25E, and 40E). Twelve
systems were composed of guppy, snail, and algae populations with six systems
at LOW I and six at HIGH I. Again, guppy populations in two systems in each
subgroup were exploited at OE, 25E, and 40E. The remaining four systems,
two at LOW I and two at HIGH I, were maintained with only snail and algal
populations.
Each system was sampled every 28 days. The length, weight, and numbers
of all individuals in the guppy and snail populations were recorded. Guppy
populations in these systems were exploited in the same way as the populations
25
-------
in the more complex laboratory systems. In order to keep micro-invertebrates
to a minimal level, each tank had accumulated sediments siphoned out every
four days. Twice a month, for each system, the sediment samples were saved
and dry weights were determined in order to keep track of changes in relative
sediment densities through time.
The procedure for conducting these ancillary experiments was intended
to be similar to that employed in conduct of the more complex laboratory
community experiments; to allow the systems to establish NSS's and to
adequately define the regions of NSS behavior prior to toxicant introduction.
It was hoped that the simpler laboratory systems would establish NSS's more
rapidly than the more complex systems and so the experiments would be of
shorter duration while still being capable of representing population and
simple community performances. Such systems would be useful microcosms.
The time required for the simpler systems to establish NSS's and for
their NSS domains to be defined was nearly as long as that required in the
more complex laboratory systems. The simpler systems required 9 to 15
months to establish NSS's, depending upon the levels of E and I. Several
months were then required to adequately define the NSS domain of system
behavior. Thus, because of the time needed for establishment and documen-
tation of NSS's and the shortened funding period of the Cooperative Agreement,
toxicant had not been introduced prior to preparation of this report. The
experiment is being continued with funding from other sources and toxicant
eventually will be introduced.
26
-------
RESULTS AND INTERPRETATION
Trophic Organization of the More Complex Laboratory Communities
Trophic organization entails that aspect of community organization
based on interactions between species populations for food resources.
Figure 5 represents the inferred major trophic relations in the laboratory
systems.
Organic sediments including the alfalfa-OTD ration and microorganisms
were the common prey for three of the major populations: guppies, amphipods,
and snails. The microorganism component included nematodes, flagellates,
rotifers, gastrotrichs, and protozoans (Finger, 1980). Differences in
alfalfa input, 0.6 or 4.0 grams per day, were the major source of differences
in sediment densities between systems at LOW I and HIGH I.
Guppies are omnivorous, live-bearing, cannibalistic fish. Adult
females (up to 42 mm and 2.0 grams) and mature males (up to 20 mm and 0.1
grams) were observed consuming the alfalfa-OTD ration, sediments, and
amphipods. Stomach samples showed the presence of materials and micoorganisms
identified in the sediments, plus amphipod parts. Newborn guppies were
observed eating alfalfa and picking through the sediments.
The amphipods in these laboratory systems were herbivorous crusta-
ceans that ranged in size from 0.5 to 15.0 millimeters (0.04-4.0 mg).
They were observed feeding in the sediments and on the alfalfa-OTD ration.
Amphipods moved freely throughout the tank and were found among the rocks
and on the sides of the tanks when guppies were absent or at low densities.
Their movement was usually limited to among the rocks and the sediment
when guppies were present. Amphipods were prey of the guppy populations,
although probably not as preferred a prey as the alfalfa ration. Amphipods
were also competitors of the guppies for the alfalfa-OTD ration.
27
-------
EXPLOITATION
PLANARIA
DIELDRIN
GUPPIES
AMPHIPODS
-f
+
SNAILS
-f
ORGANIC SEDIMENTS ATTACHED
and ASSOCIATED MICROORGANISMS ALGAE
ALFALFA-OTD
RATION
H-
LIGHT
ENERGY
Figure 5. Kinetic diagram representing inferred trophic
interrelations in the laboratory communities.
28
-------
Snails were introduced as eggs attached to aquatic plants. Snails
were observed eating attached algae on the sides and bottom of the tanks
as well as feeding in the sediments. They were competitors of both guppy
and amphipod populations for the alfalfa-OTD ration. Increased I brought
about dramatic increases in snail biomasses. Censusing of snail populations
in both HIGH I and LOW I systems was initiated about this time.
Planaria were inadvertently introduced and were more effective
predators on amphipods than were guppies, the amphipods being unable to
escape from planaria in the rock substrate. Planaria were observed in
all systems at low densities (i.e., 1 to 30 individuals) as early as
October 1977. Following the increase in energy and material input and
substrate cover in eight systems, amphipod populations increased up to
15 times in biomass. Planaria populations increased following the
increase in amphipods, and this led to drastic declines in amphipod
populations. Intensive planaria control in four HIGH I systems allowed
the amphipods to again increase. Planaria were not controlled in the
remaining four systems at HIGH I and they eventually eliminated all the
amphipods.
Attached algae included a mixture of blue-green and green algae and
diatoms. These algae and associated microorganisms were a food resource
for snails and probably other species in the laboratory systems.
Multisteady-state Structure and Organization Prior to Dieldrin Introduction
Habitat availability and energy and material input rate (I) and
exploitation (E) were defined as components of the environment of the
laboratory communities. Different environmental states were obtained by
fixing I and E at different levels. We did not expect any system to establish
29
-------
a perfect steady-state, for this is a theoretical concept. Rather we expected
that system behavior would be localized to a region in phase space (a near
steady-state) and, further, that the region or localized domain of behavior
would be different under different sets of environmental conditions.
NSS structures of the laboratory communities prior to dieldrin intro-
duction are represented on guppy-amphipod and guppy-snail phase planes shown
in Figures 6A, 7, 8, and 9. NSS relations between amphipods and snails can be
inferred from these relations. Mean NSS biomasses of the species composing
the systems are given in Table 2. In the four systems at LOW I that were
exposed to dieldrin, snail biomasses prior to dieldrin introduction were not
determined.
At LOW I, increased E brought about reductions in NSS guppy biomasses
and increases in NSS amphipod biomasses (Fig. 6). Increased amphipod biomasses
apparently resulted from reduction in intensity of predation and possibly
competition by guppies owing to reduction in guppy biomass.
Prior to introduction of guppies into the laboratory systems, amphipod
populations reached maximum biomasses of 7.5 grams, the range at the time of
guppy introduction being 0.5 to 5.0 grams. Introduction of guppies resulted
in establishment of NSS amphipod biomasses that were much lower than the
biomasses that existed prior to guppy introduction.
At HIGH I, NSS relationships between populations were shifted to the
right on each phase plane, resulting in increased NSS biomasses of guppies,
amphipods, and snails at each E (Figs. 8, 9, Table 2). This broadly
conforms to the responses to increased I shown in Figures 1 and 3.
At HIGH I in systems in which planaria were controlled (open symbols,
Fig. 9), the relationship between guppy and snail biomasses was an inverse
30
-------
Figure 6. Guppy-amphipod phase plane at LOW I. A. Near steady-state
behavior prior to toxicant introduction. B. Behavior of the
systems during toxicant introduction. C. Behavior during
recovery from toxicant introduction. Fish were restocked twice
at 40E during recovery. Areas encircling open symbols in B
and C represent near steady-state behavior prior to toxicant
introduction. (After Woltering et al., in prep.)
31
-------
O OE
A 10 E
O 20 E
D 40 E
0.2 0.4 as 1.0 1.2 1.4
B
1.6
2.0
AMPHIPOD BIOMASS(g)
32
-------
Figure 7. Guppy-snail phase planes at LOW I. A. Near steady-state
behavior during toxicant introduction. B. Behavior of the
systems during recovery from toxicant introduction. Fish were
restocked twice at 40E during recovery. (After Weltering et al.,
in prep.)
33
-------
5-
in
<
s
o
Q.
a
D
O
5 -
15
SNAIL BiOMASS (g)
34
-------
Figure 8. Guppy-amphipod phase plane at HIGH I. Open symbols indicate near
steady-state behavior at HIGH I prior to toxicant introduction.
Closed symbols indicate behavior at HIGH I during toxicant
introduction. The trajectories show actual system behavior
prior to establishing near steady-states. The hatched section
represents the domain of behavior of the systems at LOW I.
(After Woltering et al., in prep.)
35
-------
10 20 30
AMPHIPOD BIOMASS (g)
36
-------
Figure 9. Guppy-snail phase plane at HIGH I. Outlined symbols indicate near
steady-state behavior at HIGH I for guppy-snail-planaria systems.
Open symbols indicate near steady state behavior at HIGH I for
guppy-snail-amphipod systems prior to toxicant introduction.
Closed symbols indicate behavior at HIGH I for guppy-snail-
amphipod systems during toxicant introduction. The trajectories
show actual system behavior prior to establishing near steady-
states. The hatched section represents the domain of behavior
of the systems at LOW I. (After Weltering et al., in prep.)
37
-------
O
20 40 80
SNAIL BIOMASS (g)
100
38
-------
Table 2. Guppy, amphipod, snail, planaria, and leech near steady-state behavior mean blomasses (B), number of sample periods
(n), and biomass range during n in laboratory ecosystems.
10
vo
SYSTEM
LOW I
Prior to
Toxicant
LOW I
Toxicant
Intro.
LOW I
Recovery
HIGH I
Planaria
HIGH I
Prior to
Toxicant
HIGH I
Toxicant
Intro.
OE
IDE
20E
40E
OE
IDE
20E
40E*
OE
10R
20E
40E
OE
10E
20E
40E
OE
10E
20E
40E
OE
10E
20E
40E
n
16
13
13
18
8
29
9
3
26
IS
26
26
11
7
8
9
15
13
17
17
10
14
10
10
GUPPIES
lj(g) Range(g)
7.08
5.55
3.72
1.29
5.05
3.74
1.99
0.00
9.06
4.80
3.41
0.94
39.40
15.39
10.04
5.38
35.30
21.55
13.79
7.31
31.18
18.04
10.72
6.27
5.87- 8.03
4.68- 6.10
2.78- 4.11
0.71- 1.96
4.71- 5.39
2.35- 5.00
1.31- 2.55
7.25-10.73
3.17- 5.91
2.19- 4.27
0.00- 2.09
33.81-42.69
14.47-16.37
9.53-10.41
4.96- 5.94
32.06-39.79
20.30-23.14
11.27-17.02
5.04- 9.28
28.38-34.19
16.63-19.71
8.88-12.49
4.84- 8.44
n
16
13
13
18
8
29
9
3
26
15
26
26
11
7
8
9
15
13
17
17
10
13
10
10
AMPHIPODS
¥(g) Range(g)
0.09
0.30
0.16
1.00
0.13
1.31
0.41
1.52
0.48
1.81
1.30
2.55
0.00
1.08
0.74
0.00
13.28
19.38
14.78
16.32
17.54
18.48
19.86
14.15
0.01- 0.33
0.22- 0.45
0.03- 0.24
0.81- 1.54
0.09- 0.20
0.23- 2.45
0.25- 0.59
1.03- 2.53
0.04- 1.02
1.13- 2.60
0.84- 2.73
0.86- 4.67
0.00- 0.00
0.09- 2.80
0.28- 1.37
0.00- 0.00
9.56-15.42
16.02-21.59
12.24-17.75
10.33-23.11
12.76-20.85
15.79-23.35
16.30-22.29
10.12-17.13
n
4
24
8
3
26
15
26
26
11
7
8
9
15
13
17
17
10
14
10
10
SNAILS
~B(g) Range(g)
N.S.
N.S.
N.S.
N.S.
6.2
5.2
11.2
9.0
4.7
4.1
7.4
4.4
27.06
42.0
79.6
94.14
11.22
10.14
14.98
21.14
6.40
8.34
10.01
25.62
4.8- 8.0
2.7- 8.5
8.2-14.2
8.8- 9.2
1.8- 9.8
3.1- 6.4
0.1-15.8
3.3- 8.0
18.60-35.56
35.1 -48.3
68.4 -89.7
90.54-104.56
5.12-20.66
5.22-14.64
5.36-25.03
11.83-32.16
4.46- 9.16
5.24-12.45
5.26-16.02
19.47-30.82
n
8
29
8
3
26
15
26
25
11
7
6
9
15
13
14
17
10
14
10
10
PLANARIA
¥(g) Range(g)
F
f
M
M
<
<
0.02
0.05
<
0.01
<
0.05
0.48
1.31
0.49
1.59
<
<
0.24
0.01
0.00
0.00
0.17
0.00
0.00- <
0.00-0.01
0.01-0.05
0.03-0.08
0.00-0.01
0.00-0.04
0.00- <
0.01-0.13
0.24-1.02
0.02-2.88
0.07-1.72
0.82-2.50
0.00-0.01
0.00- <
0.03-0.59
0.00-0.06
0.00-0.00
0.00-0.00
0.10-0.29
0.00-0.00
LEECHES
Density
N
N
N
N
N
F
N
N
N
F
F
N
M-N
N
N
M-N
M
M
F
H
N
N
N
N
N.S. « not sampled. 0.00 - no individuals found. < •= less than 0.01 g. N - none, F - few, H « moderate, H <• high.
* This represents system structure after extinction.
-------
one, with reductions in NSS guppy biomass owing to increased E being
accompanied by increases in NSS biomass of the snail competitor. Amphipods
were abundant in these systems (Table 2). Interestingly, leeches maintained
a relatively high density in the system at 40E but did not severely reduce
snail biomass. Snails may have been able to sustain relatively higher
predation intensities by leeches at HIGH I, 40E because the low density of
the guppy competitor and the high rate of energy and material input may
have made more food available for snails, leading to higher rates of
reproduction and production.
At HIGH I in systems where planaria were not controlled (outlined
symbols, Fig. 9), amphipod populations were maintained at very low levels
or driven to extinction by planaria predation (Table 2). That planaria
were capable of preying effectively on amphipods was verified by placing
planaria and amphipods together in small dishes and observing predation
take place. The form of the relationship between guppies and snails in
the systems in which planaria were not controlled was an inverse one, similar
to that observed in the HIGH I systems in which planaria were controlled.
The relationship, however, was shifted to the right on the phase plane. At
each E, snail populations maintained higher NSS biomasses when planaria were
abundant and amphipod densities severely reduced, than when planaria were
controlled and amphipods were abundant (Table 2). Reduction in intensity
of competition with amphipods owing to reduced amphipod densities may have
allowed snails to maintain these higher densities. NSS guppy densities
did not appear to be greatly affected by amphipod elimination.
When the amphipod competitor was absent, changes in NSS guppy biomass
had a much greater effect on NSS snail biomass. In these systems there was
40
-------
over a three-fold difference in snail biomass between OE and 40E. In the
systems in which amphipods were present the difference in snail biomass
between OE and 40E was less than two-fold.
NSS amphipod biomasses in systems where planaria were controlled appeared
to be unaffected by changes in guppy and snail biomasses resulting from
guppy exploitation (Fig. 8). Because of the large amount of rock substrate
available as an amphipod refugium at HIGH I and a greater availability of
their preferred food, the alfalfa ration, guppies may not have preyed as
effectively on amphipods and thus had little direct predatory effect on their
biomasses. The nature of the competitive interactions of amphipods with
guppies and snails is less clear. Competitive interactions between amphipods
and snails appear to be relatively intense, as evidenced in Figure 9 by the
large decreases in snail biomasses that occurred when planaria were controlled
and amphipods were abundant. And yet, in systems where planaria were control-
led, changes in snail biomass brought about by changes in the biomass of
guppies had little apparent effect on amphipods (Fig. 8). The situation is
further confused by the presence of Hyallela (densities unknown) in the
systems at OE and 20E, with the systems at OE being composed mostly of
Hyallela.
Structure and Organization of the Laboratory Communities
After Introduction of Dieldrin
In general, the responses of the laboratory communities to continuous
exposure to one ppb of dieldrin broadly conformed to the responses outlined
in the Introduction (Fig. 2). Responses entailed alteration in system
structure and organization, characterized by shifts in location or position
of the system NSS points in phase space. All systems, with the exception
of the system at LOW I, 10E, established NSS's during exposure to dieldrin
41
-------
(Figs. 6B, 7A, 8, 9). Although dieldrin altered system structure and
organization, in the sense of shifting the location of NSS points, about
the same forms of relationships between populations were maintained. Thus,
at LOW I, after dieldrin introduction, an inverse relationship was maintained
between NSS guppy and amphipod biomasses (with the exception of 10E) and
between NSS guppy and snail biomasses (Figs. 6B, 7A). At each E, during
dieldrin introduction, the guppy population maintained lower biomasses and
the amphipod population higher biomasses than the biomasses these populations
maintained prior to dieldrin introduction (Table 2). Since censusing of
snail populations at LOW I did not begin until about six months after dieldrin
introduction, effects on snail biomass in the four systems exposed to dieldrin
cannot be determined. At HIGH I, after dieldrin introduction, the forms of
the NSS relations between populations were similar to the forms of these
relations prior to dieldrin exposure (Figs. 8, 9).
Experiments were conducted at our laboratory on the effects of
dieldrin on individual organisms of the species present in the laboratory
systems to enable us to better explain the effects observed in the systems.
The 96-hour LC50 of newborn guppies was about 5 ppb and that of adult females
about 20 ppb. Dieldrin concentrations of 1 ppb and 2 ppb affected age-
specific growth, survival, and reproduction of guppies (Kulbicki 1980).
The 96-hour LC50 for amphipods was about 50 ppb. Sublethal effects on
amphipods could not be reliably determined, because of difficulties in
maintaining individuals and populations in aquarium studies outside the
laboratory communities. Thus, a dieldrin concentration of one ppb in the
laboratory systems probably directly affected guppy survival, growth, and
reproduction (Weltering 1981, Liss et al. 1980) and only indirectly affected
other populations through its effects on guppies.
42
-------
At LOW I, OE, after Introduction of dieldrin, there was an immediate
decrease in guppy biomass (Fig. 6B). Amphipod populations increased
slightly as a result of the decrease in biomass of their guppy predator and
competitor. Numerous male and adult female guppies died in the first few
months after exposure. Such mortality was not apparent in the other three
systems receiving dieldrin. At OE prior to dieldrin introduction, the
guppy population maintained a relatively high NSS density and apparently
there was less food available per individual and slower individual growth
than at higher exploitation rates (Liss et al. 1980). Thus, the fish were
in relatively poorer condition and apparently were highly susceptible to
dieldrin intoxication.
After the introduction of dieldrin, the system appeared to establish a
new NSS at a guppy biomass of about five grams and an amphipod biomass of
about 0.4 grams. The system maintained this structure for about eight
months. Although dieldrin was still being introduced, the system began to
recover from toxicant perturbation, with guppy biomass gradually increasing
and amphipod biomass decreasing. The structure of the recovering system
eventually overlapped the NSS structure that existed prior to dieldrin
introduction.
At LOW I, 10E, amphipod biomass became highly variable after dieldrin
introduction, exhibiting as much as a four-fold difference in density (Fig. 6B),
The reasons for this are not known, however, it was not due to colonization
of this system by other species. These large fluctuations in amphipod den-
sity and to a lesser extent guppy density (about a two-fold fluctuation)
preclude identification of a well-defined NSS region on the guppy-amphipod
phase plane. However, a NSS region is much better defined on the guppy-
snail phase plane (Fig. 7B). The relatively large fluctuations in amphipod
43
-------
biomass did not appear to greatly influence guppy and snail dynamics.
At LOW I, 20E, the system established a new NSS structure in the pre-
sence of dieldrin with NSS biomass of guppies lower and amphipod biomass
higher than their NSS biomasses prior to dieldrin introduction (Fig. 6B).
At LOW I, 40E, guppy populations went extinct after 15 months of con-
tinuous exposure to dieldrin (Figs. 6B, 7A, also see Fig. 2). For the
first eight months after introduction of dieldrin, there were no obvious
changes in the structure of the system. Guppy populations exploited at 40E •
maintained very low densities and often had size distributions with over
70 percent of the biomass in one large female. This sometimes resulted
in significantly more or less than the prescribed 40 percent of the biomass
being exploited in any given month. Over many months, however, mean exploita-
tion rate was near 40 percent.
This system had been exploited by 77 and 82 percent in two of the 18
months prior to dieldrin introduction. The population recovered from these
incidences of "overexploitation". After ten months of continuous exposure
to dieldrin, the system was again overexploited at 75 percent (a one gram
female). The population did not recover in the following five months and
went extinct with the removal of a 0.2 and 0.5 gram female. The two fish
carried a total of 42 eggs and embryos. For seven months prior to extinction,
the number of newborn fish was at least 70 percent lower than the number of
newborns present prior to dieldrin introduction. There had been no newborn
fish present in the tank for three months prior to extinction.
Severe reduction in density, simplification of age structure, and
possible reduction in genetic variation and adaptive potential sometimes
associated with low population densities (Franklin, 1980, Soule 1980,
44
-------
Frankel and Soule 1981) must surely make populations more susceptible to
chance extinction. Thus, we can never be sure that extinction was causally
related to dieldrin exposure, but the available evidence makes this a
plausible explanation.
Severe reduction and elimination of offspring recruitment occurred for
several months prior to extinction. These low levels of recruitment were
not apparent prior to dieldrin introduction at similar adult densities.
Furthermore, the few females that were present prior to extinction carried
sufficient eggs, if they would have survived and grown, to replenish the
adult stock. This suggests that reduction in offspring survival and
recruitment to mature size classes may have been the proximate cause of
extinction. Offspring survival may have been reduced due to accumulation
of dieldrin in eggs.
In the laboratory community at LOW I, 40E into which dieldrin was not
introduced, the guppy population persisted for 65 months, until the termina-
tion of the experiment, occasionally being subjected to the same kind of
"overharvest" as the population exposed to dieldrin. This further supports
the conclusion that guppy population extinction was related to exposure to
dieldrin.
Finally, nine months after extinction, when dieldrin introduction had
been terminated, the system was restocked with guppies. Nine months
afterward this population essentially went extinct, the only remaining fish
being an adult male. The system was again restocked. Possibly the fish
were able to accumulate in their eggs enough dieldrin from that remaining
in the organic sediments and their food organisms to again bring about
reductions in offspring survival.
45
-------
After extinction of guppies, amphipod biomass increased, reaching
levels that the population had not attained since guppies were introduced.
In addition, behavioral changes occurred, the amphipods moving more freely
throughout the tank rather than restricting their movement primarily to
the rock nests.
During introduction of dieldrin, NSS guppy biomass was inversely
related to NSS snail biomass. Increased E resulted in a reduction in NSS
guppy biomass and an increase in NSS biomass of the snail competitor. This
is the same form of relationship observed at HIGH I (Fig. 9).
At LOW I, system response to dieldrin varied with exploitation rate,
responses ranging from perturbation and recovery at OE to guppy extinction
at 40E. The response of the laboratory communities to dieldrin appeared
to be dependent upon the level of I as well as E. At HIGH I, all systems
established new NSS's after introduction of dieldrin (Fig. 8, 9). The
alteration of system structure and organization at each exploitation rate
was much less than at LOW I. At HIGH I, during dieldrin introduction, the
guppy population at each E maintained somewhat lower NSS biomasses than the
biomasses it maintained prior to dieldrin introduction. However, at HIGH I
the changes in mean NSS biomasses of guppies as well as amphipods induced by
exposure to dieldrin were not nearly as great as the changes that occurred
at LOW I (Table 3).
At HIGH I mean NSS biomasses of snails at OE, 10E, and 20E appear to
have been reduced somewhat by exposure to dieldrin (Table 2). However, the
domain of NSS behavior of snail biomass prior to dieldrin introduction is
rather large (Fig. 9, Table 2). Although mean NSS snail biomasses prior to
and after dieldrin introduction are different, the domain of behavior of snail
biomass after dieldrin introduction is within the domain of behavior prior
46
-------
Table 3. Percent change in mean NSS biomasses following toxicant
introduction.
SYSTEM GUPPIES
LOW I HIGH I
OE
10E
20E
40E
-28.7 -11.7
-32.6 -16.3
-46.5 -22.3
-100.0 -14.2
AMPHIPODS
LOW I HIGH I
+44.4
+366.7
+156.3
+41.0
+32.1
-4.6
+34.4
-13.3
SNAILS
LOW I* HIGH I
-43.0
-17.8
-33.2
+21.1
* No snail data were taken in systems prior to toxicant introduction.
47
-------
to dieldrin introduction. Thus, for snails, the mean values may not be
indicative of the actual magnitude of effect on snails.
At LOW I, the guppy population exploited at 40E was driven to extinction
apparently by exposure to dieldrin. At HIGH I, the guppy population at
40E persisted for nine months after dieldrin introduction, until the
experiment was terminated. This population was not exposed to dieldrin for
as long a period of time as the population at LOW I. However, the population
maintained a relatively high density and there was no indication from
observations of offspring survival and recruitment that the population was
in danger of extinction. Prior to introduction of dieldrin into the system
at HIGH I, 40E, the mean NSS number of newborn fish (fish less than 10 mm)
was 60. For the four successive months prior to termination of the experiment,
the number of newborns observed in the system was 64, 99, 93, and 63. At
LOW I, 40E, for many months prior to extinction the number of newborns was
considerably lower than the number maintained prior to dieldrin introduction.
Population persistence may be determined by the levels of I as well as E, as
is shown in Figures 2 and 3.
Recovery of Systems at LOW I From Dieldrin Perturbation
Changes in community structure and organization after termination of
dieldrin introduction were examined in the four systems at LOW I to determine
if the systems would recover to structures they maintained prior to dieldrin
introduction. Over the time period in which recovery was evaluated the
systems maintained structures that were different from the structures they
maintained during dieldrin introduction (Figs. 6B, 6C, 7A, 7B), however they
did not return to the structures they maintained prior to toxicant intro-
duction (Fig. 6A). Recovery of these systems was evaluated for nearly
48
-------
two years. Thus ample time was available for the systems to return to their
original structures.
Differences in system structures prior to recovery and structures
during recovery were related to changes in system organization. During the
period of recovery inadvertent colonization of the systems by the amphipod
Hyallela azteca and leeches occurred. This colonization was unrelated to
dieldrin exposure. Furthermore, unavoidable long-term changes in the system
such as accumulation of organic sediments also may have occurred. These
kinds of changes in system organization brought about changes in system
capacity. A system whose capacity has changed cannot recover or return to
Its original structure(s) even if the environment can be restored to its
original state(s). Changes in system capacity brought about by colonization
make it difficult to ascertain whether capacity had been altered by exposure
to dieldrin.
In the system at OE, amphipod and guppy biomasses were higher than the
biomasses these populations maintained prior to dieldrin introduction or
during dieldrin introduction (Fig. 6C, Table 2). Guppy biomass had begun to
increase even before termination of dieldrin introduction at this exploitation
rate. Snails maintained about the same biomass during recovery as they
maintained during dieldrin introduction (Fig. 7A, B). Hyallela became the
dominant amphipod species in this system, nearly excluding Gammarus, which
persisted only at very low densities when Hyallela was not present (e.g.
Fig. 6A).
At 10E, during recovery, the guppy population returned to about the same
density it maintained prior to dieldrin introduction, although guppy biomass
was somewhat more variable during recovery (Fig. 6C). Amphipod biomass was
49
-------
considerably higher during recovery than prior to or during dieldrin intro-
duction. The reasons for this are not known. Interestingly, few Hyallela
were present in this system. Snail biomass during recovery was about the
same as the biomass the snails maintained during dieldrin introduction
(Figs. 5A, B). Both amphipod and snail biomasses were less variable during
recovery than during dieldrin introduction.
At 20E, during recovery, guppies maintained about the same biomass they
maintained prior to dieldrin introduction (Fig. 6C). Amphipod biomass was
considerably higher during the recovery phase than prior to recovery.
Hyallela colonized this system and was fairly abundant. Snail biomass was
lower and far more variable during recovery than during dieldrin introduction
(Fig. 5A, B). Leeches colonized the system and these together with an
increase in the abundance of the guppy competitor may have accounted for the
reduction in snail biomass.
The guppy population at 40E had suffered extinction during dieldrin
introduction, after which amphipod biomass increased considerably (Fig. 6B, C),
Nine months after extinction occurred and three months after dieldrin intro-
duction had been terminated, guppies were restocked at a density and size
structure similar to that which existed prior to dieldrin introduction. After
restocking, amphipods were reduced to the biomass they maintained prior to
dieldrin introduction (Fig. 6C). Nine months after restocking the guppy
population effectively went extinct, being composed at that time of only
one male. Guppies were again restocked. Some Hyallela were present in
this system. Snails maintained lower biomasses during recovery than during
dieldrin introduction (Fig. 7A, B). The reasons for this are not clear.
At OE, the guppy population maintained a higher biomass during recovery
than it maintained prior to and during dieldrin introduction. It is possible
50
-------
that the capacity of this population was altered evolutonarily as a conse-
quence of exposure to dieldrin. This will be considered in the discussion.
At 10E and 20E, the guppy populations maintained about the same biomasses
during recovery as they maintained prior to dieldrin introduction.
Amphipod population biomasses at all E were higher during recovery
than at any time prior to the recovery phase. At OE, 20E, and possibly
40E, this may have been related to colonization of the systems by Hyallela.
Hyallela may make more efficient use of the resources than Gammarus and/or
be less susceptible to biomass reduction due to predation and possibly
competition from guppies and snails.
At 20E and 40E, snail populations maintained lower biomasses during
recovery than during exposure to dieldrin. This may be related to the
presence of leeches in the systems and increases in densities of the guppy
competitor.
At 20E, 40E, and, at least in part, at OE changes in community organ-
ization brought about through colonization of the systems by Hyallela and
leeches can account for differences in structure of the communities
during the recovery phase and prior to recovery. Colonization, by bringing
about changes in community organization and consequent changes in system
capacity, shifted the location in phase space of the system NSS points at
each level of I and E.
Guppy Life History Patterns and Guppy Population Production and Yield Prior
to and During Dieldrin Introduction
The life histories of individual organisms include the patterns of growth
and reproduction they manifest when exposed to different sets of environmental
51
-------
conditions. Guppy size-specific relative growth and reproduction rates near
steady-state were affected by exploitation rate, exposure to dieldrin, and
level of habitat availability and energy and material input. These effects
were reported elsewhere (Liss et al. 1980) and will only be summarized here.
Size-specific guppy growth was able to be determined only at LOW I.
Size-specific relative growth rates were density-dependent, increasing with
decreases in population biomass brought about by increased E. At LOW I,
the highest growth rates occurred at 40E where population biomass was
lowest, and the lowest growth rates occurred at OE where population biomass
was greatest. In general, at both HIGH I and LOW I, size-specific fecundity
also appeared to be density-dependent. At each E, size-specific fecundity
and probably growth were greater in fish at HIGH I than in fish at LOW I,
this apparently reflecting the greater availability of food.
At LOW I, dieldrin appeared to reduce the relative growth rates of
juvenile fish and small mature females up to 24 mm in length and possibly
of larger fish. Dieldrin also reduced size-specific fecundity of mature
female guppies. At LOW I, 40E, reduced growth and fecundity but especially
poor survival of newborn fish may have been responsible for bringing
about population extinction.
NSS curves of guppy production, or total tissue elaboration, and
yield are shown in Figure 10. At each I, increased E resulted in reduction
in NSS guppy biomass, essentially shifting production and yield values from
right to left along each dome-shaped curve. Increased I increased guppy
food resources, this resulting in increased guppy biomass, production, and
yield at each E. Thus the magnitude of the production and yield curves
increased as a result of an increase in I. Dieldrin altered guppy life
history patterns by reducing size-specific growth and reproduction, as well
52
-------
Figure 10. Mean NSS guppy production and yield as a function of 'mean NSS
population biomass for LOW I (smaller symbols) and HIGH I
(larger symbols). Open symbols indicate NSS production and
yield prior to toxicant introduction and closed symbols indicate
production and yield during toxicant introduction. Production
at HIGH I was not able to be determined, but the relative
position that such a curve would occupy is shown by the dashed
line. (After Weltering et al., in prep.)
53
-------
z
o
o
§1
Q- c
4.0
3.2
2.4
1.6
0.8
/
\
i i i i i i i i i i i i
8
MEAN GUPPY BIOMASS (grams)
\
j i
12 16 20 24 28 32 36
54
-------
as survival, this accounting for decreased guppy biomass and reduction in
the magnitude of production and yield curves at each I. These kinds of
changes in the magnitude of yield curves conform to those derived in
Figure 4.
Structure and Organization of Simple Laboratory Communities
in Ancillary Experiments
The laboratory communities in the ancillary experiments were intended
to be persistent systems having simpler organization and being of shorter
duration than the more complex laboratory communities. NSS structure and
organization of the systems is shown in Figures 11 and 12 (Lee 1983). NSS
domains of behavior are reasonably well-defined for most systems.
The alfalfa ration, food resource of both guppies and snails, is intro-
duced daily and becomes part of the organic sediment. Thus organic sediment
biomass may be an index of food biomass and this should be related to the
biomass of guppies and snails. At both LOW I and HIGH I, when the snail
competitor is not present in the system, guppy biomass is inversely related
to sediment biomass (Fig. 11), with increased E resulting in a reduction in
guppy biomass and an increase in sediment biomass. The relationship between
guppy and sediment biomass was shifted to the right on the phase plane when
I was increased from LOW I to HIGH I (Figure 11B). At HIGH I, the biomasses
of both guppies and sediments are greater than the biomasses they maintained
at corresponding E at LOW I. These responses are similar to those derived
in Figures 1 and 3.
At both HIGH I and LOW I, the presence of a snail competitor shifted the
relationship between guppies and organic sediments to the left on the phase
planes (Fig. 11). When snails were present, the biomasses of both guppies
and sediments were less than the biomasses they maintained when snails were
55
-------
Figure 11. Phase plane representing NSS guppy and sediment biomasses at LOW
I (A) and HIGH I (B) when the snail competitor is absent (G-A
systems, solid symbols) and when snails are present (G-S-A
systems, open symbols). NSS behavior of systems at OE (circles),
25 E (triangles), and 40E (squares) is shown. Actual trajec-
tories of G-A systems at HIGH I, OE (a) and HIGH I, 40E (b),
and G-S-A systems at HIGH I, OE (c) and HIGH I, 40E (d) are
shown. RCO is the region of common origin of all trajectories,
that is, the biomasses at which populations were first intro-
duced into the systems. After Lee (1983).
56
-------
A
0.2 0.3 0.4 0.5
ORGANIC SEDIMENTS (gm)
60
50
40
30
co
CO
o
CD 20
I
CO
u_
10
LOW I
SYSTEM
DOMAIN
B
1.0 1.5 2.0 2.5
ORGANIC SEDIMENTS (gm)
3.0
3.5
57
-------
Figure 12. Phase plane representing NSS guppy and snail biomasses at LOW I
(A), and HIGH I (B). NSS behavior for systems at OE (circles),
25 E (triangles), and 40E (squares) is shown. Actual trajec-
tories of G-S-A systems at HIGH I, OE (c) and HIGH I, 40E (d)
are shown. RCO is the region of common origin of all trajec-
tories that is, the biomasses at which populations were first
introduced into the systems. After Lee (1983).
58
-------
o>
« 6
CO
1 5
o
03
4
CO
A
IRCO*
x..'
12 16 20
SNAIL BIOMASS (gm)
24
28
32
B
60
90 120 150
SNAIL BIOMASS (gm)
180
210
59
-------
absent. Presumably the presence of the snail competitor made less food
available for guppies, which was reflected in reduction in sediment biomass.
Consequently guppies were not able to maintain the NSS densities they
maintained when snails were not present. Furthermore, at both LOW I and
HIGH I, guppy populations at 40E were driven to extinction or were about to
be driven to extinction in the systems in which the snail competitor was
present.
At each I, in systems where snails were present, the phase plane
relationship between guppy and snail biomasses was an inverse one, with
a decrease in guppy biomass brought about by increased E being accompanied
by an increase in the biomass of the snail competitor (Fig. 12). Increased
I shifted the inverse relationship between these populations to the right
on the phase plane, increasing the biomasses of both guppies and snails
at each E. Thus the relationships between NSS guppy and snail biomasses
observed in the ancillary, simpler laboratory communities were similar to
the relationships between these populations observed in the more complex
laboratory systems (compare Figs. 7, 9, and Fig. 12).
60
-------
DISCUSSION
In the Introduction to this report, we advanced some premises or
generalizations that partially define a perspective for understanding
natural systems, their dynamics, and their response to toxicants. The
perspective primarily provides a way of thinking about natural systems
and was used to design, conduct, and interpret the research reported here.
It may be helpful to repeat the premises or generalizations here.
Our view entails the following:
1. Under different states of the environment of a natural system,
the system will come to have different steady-state structures
and organizations and thus can be understood to be a multisteady-
state system.
2. Dynamics or changes in structure of an n-dimensional system can
be understood as an n-dimensional trajectory in continuous pursuit
of an n-dimensional steady-state point whose location in phase
space is continually changing as a result of changes in the state
of the environment of the system.
We can add an additional generalization:
3. Any performance of a system, including its response to a toxicant,
is jointly determined by its organization (capacity) and conditions
in its environment.
Energy and material resource input and habitat availability, I, and
exploitation, E, were defined as part of the environment of both the more
complex and the simple laboratory systems. Different environmental states
were generated by fixing I and E at different levels. At each environmental
state, the systems established NSS's, or localized domains of behavior in
phase space (Figs. 6, 7, 8, 9, 11, 12). Under different sets of environmental
conditions, or different environmental states, the systems established
different NSS structures. Thus, the laboratory systems can be understood
as being multisteady-state systems in the sense that we have used that
concept here.
61
-------
Even in systems that are as simple and intensively studied as these
laboratory communities, the roles of the dominant species in organizing the
communities are difficult to ascertain. NSS relationships between populations
in the communities were similar to those derived in Figures 1, 2, and 3. At
both LOW I and HIGH I, increased E reduced NSS biomass of guppies, leading
to an increase in NSS biomass of the snail competitor (Figs. 7, 9, 12). At
HIGH I, the NSS relationship between guppies and snails was shifted to the
right on the phase plane, with both guppies and snails maintaining higher
biomasses than they maintained at LOW I (Fig. 9, 12B). The role in organizing
the communities of guppy-snail competition for food was made more clear
through experiments with the simpler laboratory systems. In these systems,
an inverse relationship existed between NSS guppy biomass and the NSS biomass
of organic sediments, an index of food level (Fig. 11). At HIGH I, both
guppies and organic sediments maintained higher biomasses at each E than they
maintained at LOW I (Fig. 1JB). These relationships are similar to the ones
derived in Figure 1. The addition of the snail competitor shifted to the
left the relationship between guppy and sediment biomass at each I (Fig. 11).
At both HIGH I and LOW I, the biomasses of both guppies and organic sediments
were lower when the snail competitor was present than when it was absent.
This is similar to the results theoretically derived in Figure 3 when an
unharvested competitor was added to the system.
In the systems in which the snail competitor was absent, guppy popula-
tions were able to persist at 40E at both LOW I and HIGH I (Fig. 11, solid
squares; Fig. 3, open triangle). However, at 40E at both level of I, guppy
populations were driven to extinction in the systems in which the snail
competitor was present (Fig. 11, open squares; Fig. 3, vertically-barred
62
-------
triangle). Guppy and snail populations were able to coexist at both levels
of I when guppies were less heavily exploited (Fig. 11, open circles and
triangles; Fig. 3, vertically-barred square). Further, as suggested in
Figure 3, the populations may have been able to coexist if the rate of
input of the alfalfa-OTD ration, I, were increased to a level greater than
4.0 grams per day and/or if the snails were harvested.
The role of amphipods in organizing the communities is less well under-
stood than the roles of guppies and snails. In the systems at HIGH I in
which planaria were controlled, amphipods were very abundant, maintaining
densities that were equal to or a little greater than the densities of snails.
In these systems snails maintained NSS biomasses at each E that were lower
than the biomasses they maintained in the systems in which amphipods were
absent, the NSS relationship between guppies and snails being shifted to
the left on the phase plane (Fig. 9). This is evidence that amphipods
affected snail populations, perhaps through competition for food. However,
at HIGH I, changes in E which brought about changes in guppy and snail
biomass had no affect on the biomass of amphipods (Fig. 8). It seems as
though the competitive relationship between amphipods and snails, at least
at HIGH I, was unidirectional; amphipods, through changes in their density,
can alter snail density, but changes in snail density have no apparent affect
on amphipods.
If this is so, guppies have the capacity to affect snail populations
through competition with them for food and by altering amphipod densities
through predation. At HIGH I, apparently, amphipods may not have been preyed
upon very heavily by guppies perhaps because they were protected from predation
by the greater availability of rock substrate and/or because guppies preferred
63
-------
to feed upon the alfalfa - OTD ration, which was much more attainable than
amphipods and was much more abundant at HIGH I than at LOW I.
At LOW I, NSS guppy biomass was inversely related to NSS amphipod biomass
(Fig. 6). Increased E resulted in reduction in the biomass of guppies and
brought about an increase in amphipod biomass. Prior to stocking of guppies
in these systems, amphipods maintained biomasses of up to 7.5 grams. After
guppies were stocked, the highest mean NSS biomass amphipods attained was
1.08 grams at 40E. Further, amphipod biomass at this E increased rapidly
to about 4.0 grams after guppy extinction (Fig. 6C).
These changes in amphipod biomass were probably brought about' chiefly
through predation by guppies on amphipods, although competition for the
alfalfa-OTD ration may also have been involved. At LOW I, rock substrate
which served as a refugium from predation was much less abundant than at
HIGH I. At LOW I rocks covered only 20 percent of the bottom of each
tank. Predation on amphipods by guppies was probably much greater at LOW I
than at HIGH I.
Amphipods were maintained at such low densities at LOW I that they could
not have been the major food source for guppies and were probably ineffective
competitors of both guppies and snails. Amphipod biomasses were kept at low
levels through predation by guppies, whose densities were maintained by
feeding on the more accessible alfalfa-OTD ration. Thus, the dynamics of
amphipod as well as guppy populations would have been quite different had
an alternative food for guppies not been available. At LOW I, guppy and snail
populations were inversely related and probably competed for food. At the
same time, however, an indirect facilitative interaction between guppies and
snails may have existed, for the guppy kept amphipods, a potentially effective
64
-------
competitor of snails, at low densities. The structure and dynamics of
communities are an outcome of very complex direct and indirect interactions
between species. This makes it difficult to anticipate or predict a priori
what the effect will be on community structure of the addition or loss of a
species or a change in density of an existing species.
Colonization of a community by new species changes its organization and
thus its capacity. Communities with different capacities will exhibit
different performances, including structure, yield, and response to toxicants,
even under the same set of environmental conditions. In terms of isocline
models, changes in system structure due to changes in system capacity are
reflected in alteration in the location in phase space of the isocline systems
and thus of the steady-state points under each set of environmental conditions
(Thompson 1981). Thus, steady-state points may be shifted in phase space not
only due to changes in system environment (e.g. I and E) but also due to
changes in system organization and thus capacity. Change in organization
through colonization by new species and extinction of old is the process of
community development.
Systems at HIGH I in which amphipods were present and those in which
they were absent had different organizations and capacities and so had
different structures even at the same levels of I and E (Fig. 9). At each
level of I and E, the location in phase space of the NSS domains of these
systems was different. Similarly, in the simpler systems, communities in
which snails were present and those in which they were absent had different
organizations and exhibited different performances at given levels of I and
E (Fig. 11). And, after introduction of dieldrin had been terminated at
LOW I, the communities were colonized by Hyallela and leeches altering their
organization. Consequently they did not return to the structures they
65
-------
maintained prior to dieldrin introduction (Fig. 6C). Colonization by new
species and disappearance of old is not the only way in which system
capacity may be altered. It may also be changed through evolution of the
populations composing the system. This will be discussed below.
Toxic substances may alter the structure and organization of ecological
systems, such a change bringing about a shift in location in phase space of
the n-dimensional system steady-state point existing at each set of environ-
mental conditions (Fig. 2). In the laboratory communities, exposure to
dieldrin brought about these kinds of shifts in NSS community structure
(Figs. 6B, 8, 9).
The response of a system to a toxicant is determined jointly by system
organization and system environment (generalization 3). The response of the
laboratory communities to dieldrin was related to the levels of environmental
factors I and E. At LOW I, response was dependent upon E, ranging from
perturbation and recovery at OE to population extinction at 40E (Fig. 6B).
At each E, the structure of the system at HIGH I was not altered as much by
exposure to dieldrin as system structure at LOW I (Figs. 8, 9).
Using a theoretical example, we can further explicate the interplay
of system environment and system organization in determining system response
to a toxicant. Let us consider as an example predator extinction, such as
occurred for the guppy population in the more complex laboratory communities,
for population extinction is one of the more serious effects a toxicant may
have on a system. Referring back to Figure 4, assume the presence of a
competitor of the predator shifts the prey isocline at each I to the left.
At LOW I, 40E when a competitor is present but toxicant has not been introduced,
the predator population is able to persist (Fig. 4, solid square). This is
analogous to the situation in the laboratory community at LOW I, 40E prior
66
-------
to dieldrin introduction (Fig. 6A). Introduction of a toxicant at a suffi-
ciently high concentration may so lower the predator isocline parameterized
by 40E that it no longer intersects the prey isocline parameterized by LOW I
and the predator population is driven to extinction (Fig. 4, vertically-barred
square). At LOW 1, the guppy population exploited at 40E was driven to
extinction apparently as a result of exposure to dieldrin (Fig. 6B). Heavily
exploited predators in systems of low productivity may be particularly vulner-
able to extinction as a result of exposure to a toxicant. The life history
and evolutionary reasons that may underlie this will be discussed below.
Any change in system environment that leads to a positive intersection
of the predator and prey isocline will facilitate the persistence of a
predator exposed to toxicant. This includes changes that shift the prey
Isocline to the right (e.g. increased I, horizontally—barred square; more
favorable physico-chemical conditions for the prey; etc.) and/or shift
the predator isocline upward (e.g. decreased E, vertically-barred circle;
more favorable physico-chemical conditions for the predator; etc.). Thus,
in the laboratory communities, the guppy population exploited at 40E was
able to persist when exposed to dieldrin when I was increased to HIGH I
(Figs. 8, 9). Further, the less heavily harvested populations—those
exploited at OE, 10E, and 20E—were able to persist at LOW I when exposed
to the pesticide (Fig. 6B). However, it is quite likely that at HIGH I had
E been increased to some level greater than 40E the guppy population would
have been driven extinct by exposure to dieldrin. And, at levels of E less
than 40E, guppy populations would not be able to persist in the presence of
dieldrin if I were reduced to a level lower than LOW I.
Changes in system organization may also have profound effects on predator
persistence. Competitors of the predator or competition on lower trophic
67
-------
levels may shift the prey isocline to the left. Thus, in the absence of
competitors, the prey isocline at a given I would lie further to the right
on the phase plane (see Fig. 3), possibly bringing about a positive inter-
section of the predator isocline parameterized by 40E, TOX and the prey
isocline parameterized by LOW I (Fig. 4, stippled square) and enabling a
predator exposed to toxicant to persist. This suggests that, in the laboratory
communities, had competitors of the guppy not been present, the population
at LOW I, 40E may have had a better chance of persisting when exposed to
dieldrin. Ancillary experiments confirm that guppy populations can maintain
much higher densities, especially at high E, when their snail competitors
are absent from the system (Fig. 11). Thus the presence of species that are
competitors of a predator but are not affected by the toxicant as much as
the predator population may render the predator more sensitive to toxicant
perturbation and more vulnerable to extinction. Since competition may be a
ubiquitous process in natural systems, it may be important in determining
population and community response to a toxicant. In general, any change in
system organization or system environment that shifts prey isoclines to the
left and/or lowers predator isoclines renders a (predator) population more
vulnerable to extinction by a toxicant by lowering its steady-state density
or creating conditions in which isoclines do not intersect in positive phase
space.
Changes in environmental conditions not only bring about changes in
community structure and organization, but also concomitant changes in
individual organism life histories. Individual organisms have life history
capacities to alter their life history patterns in response to changes in
their environments in ways favoring their survival and reproduction and the
68
-------
persistence of their populations (Warren and Liss, 1980). It is through such
developmental alteration in life history patterns, as well as evolutionary
alterations in the kinds of individual organism life history capacities
composing populations, that populations maintain concordance with and so are
adapted to their environmental systems.
The community provides the environmental context in which individuals
develop and populations must be adapted to persist. Different community
structures—different kinds and densities of prey, predators, and competitors,
etc.—constitute different developmental environments for individuals and
different evolutionary environments for populations. Changes in community
structure bring about developmental changes in individual organism life
history patterns as well as evolutionary changes in the kinds of individual
organism life history capacities composing populations. Thus, different
life history patterns and capacities can be associated with different
community structures.
In the laboratory communities, changes in conditions in the environment
of the'community such as I, E and exposure to dieldrin altered NSS community
structure. Entailed in this is alteration of life history patterns and life
history capacities. Referring again to the model shown in Figure 1, at each
I, increases in E result in reduction in steady-state predator biomass and
increases in steady-state prey biomass. Steady-state predator life history
patterns that may exist at lower and higher exploitation rates are summarized
in Table 4.
As exploitation rate and, thus, mortality due to fishing increases, we
expect reduction in length of life and number of reproductions per lifetime.
But due to reduction in predator biomass and increase in the biomass of the
food resource, we might also expect faster juvenile and adult predator growth,
69
-------
Table 4. Some possible predator life history patterns at low and high
exploitation rates (After Kulbicki, 1980).
Life History Traits
Low Exploitation Rate
High Exploitation Rate
Longevity
long lifespan
low adult mortality
shorter lifespan due to
very high mortality
resulting from
exploitation
Number of
clutches in a
lifetime
many due to
long lifespan
fewer due to the high
mortality and shorter
lifespan
Growth Rate
low due to low food
availability and
high population
density
high due to high food
availability and low
population density
Age at first
reproduction
later due to low
food availability
and low adult
mortality
earlier due to high food
availability and/or
higher mortality
Size at first
reproduction
smaller due to
slower growth
larger size at first
reproduction
Clutch size
low food availability
will result in small
clutches
larger clutches because
of high food availa-
bility and better growth
70
-------
increased size at first reproduction (or perhaps decreased age at first
reproduction), and increased fecundity. These kinds of changes in guppy life
histories were observed in the laboratory communities (Weltering, 1980, Liss
et al. 1980) as well as in other exploited guppy populations (Liss, 1974),
in clupeids (Burd and Gushing, 1962; Beverton, 1963), in trout and perch
(Aim, 1959), and in whitefish (Miller, 1956). Thus, when some populations
are exploited at high rates, they may develop life history patterns normally
associated with higher values of the intrinsic rate of increase (r) - more
rapid growth, earlier maturity or larger size at maturity, and higher fecundity
at first and at subsequent reproductions. Development of such life history
characteristics can be understood as adaptations to environments that bring
about increased mortality, shortening of lifespan and reduction in number of
reproductions per lifetime. This is simply to say that density-dependent
changes in life histories are adaptive and enable populations to persist
where mortality rate is high. Populations whose organisms cannot manifest
these kinds of characteristics may not be able to persist at high levels of
exploitation (or mortality).
Toxic substances may so alter life history patterns and the adaptive
capacities of populations that populations are no longer able to persist in
their environments or may persist at reduced densities under given sets of
environmental conditions. In the laboratory communities, dieldrin apparently
altered guppy life history patterns by reducing growth and fecundity and,
at LOW I, 40E, by increasing mortality of offspring (Weltering 1980, Liss
et al. 1980). At LOW I, 40E, dieldrin may have caused extinction of the
population by effectively preventing the individuals from exhibiting the life
history patterns—more rapid growth, higher fecundity, increased offspring
71
-------
survival—that adapted the population to persist at this very high exploitation
rate.
Thus heavily exploited populations, where rapid growth, high fecundity,
and good juvenile survival are essential for persistence, may be more
"sensitive" to reductions in growth, fecundity, and survival caused by
toxic substances. At lower exploitation rates, alterations in life history
pattern caused by dieldrin resulted in reductions in guppy population density,
but the populations were able to persist. But surely higher concentrations
of toxicant, which would bring about more severe alterations in life history
patterns—greater reductions in growth, fecundity, and survival—would cause
extinction of populations exploited at even these lower rates.
Now, at higher I, because of the greater availability of food, a
population may be able to persist when exposed to a toxicant and exploited
at 40E, as shown in Figure 4. The range of exploitation rates over which
populations can persist when exposed to a given level of toxicant may be
greater at HIGH I than at LOW I. But, as we discussed before, there will
surely be some exploitation rate higher than 40E at which a population even
at HIGH I will not be able to persist.
At any time, the capacity of a population to adapt evolutionarily to
changes in its environment, including exploitation and toxic substances, is
determined by its genetic organization. But evolutionary changes entail
changes in genetic organization and thus alteration in the capacity of a
population to adapt to future changes in its environment. Evolution can
be understood as a continual change in a population's adaptive capacity
(Warren and Liss 1980). Changes in genetic organization are accompanied by
changes in the kinds of individual organism life history capacities composing
the population. Thus changes in community capacity may be brought about
72
-------
through evolutionary changes In the capacities of populations composing the
community as well as by changes in organization resulting from species
colonization and extinction.
Evolutionary changes in populations can be brought about by exploitation
(Schaeffer and Elson 1975, Silliman 1975, Moav et al. 1978, Ricker 1981) and
exposure to toxic substances (Culley and Ferguson 1969). At LOW I, the
guppy population exploited at OE was able to adapt developmentally and
perhaps evolutionarily to the presence of dieldrin and so recover from
d ieldrin perturbation. Natural selection could have favored those individuals
that had the life history capacity to survive, grow, and reproduce most
efficiently in the presence of dieldrin—the so-called "resistant" Individuals,
These individuals exhibit life history patterns that are well adapted to the
presence of dieldrin, thus enabling the population to increase in blomass
while still being exposed to the pesticide.
/
The population had begun to recover from toxicant perturbation long
before toxicant introduction into that system was terminated. Mean NSS
biomass of the population after termination of dieldrin introduction was
about 30 percent greater than the mean NSS biomass the population maintained
prior to dieldrin introduction (Table 2). This provides some evidence
that the capacity of the population had been altered. The difference between
mean NSS biomass after termination of dieldrin and mean biomass prior to
dieldrin introduction is not nearly this great for any of the other popula-
tions. Although Hyallela had colonized this system during recovery, amphipod
biomass was still far too low to have been a significant food resource for
the guppies. Thus the changes in amphipod species composition and biomass
during recovery cannot account for the increase in guppy density.
73
-------
However, in adapting to the toxicant the population at OE may have lost
some of its capacity to adapt to other environmental conditions. Thus the
population may not adapt and perform as well under different or continually
changing environmental conditions as it would have adapted and performed if
it had not been exposed to toxicant.
In the laboratory communities at LOW 1, populations exploited at 10E
and 20E did not recover from toxicant perturbation throughout the time that
dieldrin was being introduced. The unexploited population maintained a
greater density than the populations exploited at 10E and 20E. Perhaps
greater population density increased the probability that "resistant" genes
or sets of genes were present in the population and so speeded evolutionary
adaptation and recovery from perturbation. And, of course, the population
exploited at 40E, which maintained a very low density was apparently partic-
ularly sensitive to dieldrin and was driven to extinction.
Several workers have argued that, for genetic reasons alone, reduction
in density renders natural populations more vulnerable to extinction
(Franklin 1980, Soule 1980, Frankel and Soule 1981, Kapuscinski 1983). They
argue that reduction in population size may be accompanied by inbreeding,
leading to reduction in genetic variation (reduction in heterozygosity)
and fixation of deleterious alleles. The life history traits most severely
affected by inbreeding are the so-called "fitness traits"—those associated
with survival and reproduction. The problem is thought to be especially
severe in isolated populations that are not subjected to infusion of new
genes by gene flow. Inbreeding and its negative effects on life history
traits—growth, reproduction, and survival—would act in opposition to any
adaptive density-dependent increases in these performances that would tend
to be brought about as a result of reduction in the density of a population
74
-------
and increase in the density of its food resources. Toxicants that themselves
reduce growth, survival, and reproduction would, in effect, reinforce the
negative effects of inbreeding at low population densities. Moreover,
inbred individuals may be far more susceptible to toxicants; their growth,
survival, and reproduction may be affected more by a toxicant than non-inbred
individuals.
Loss of genetic variation would amount to reduction in the capacity of
a population to adapt evolutionarily to changes in environmental conditions,
including toxicant introduction. Thus intensive exploitation may reduce the
adaptive capacities of population by reducing their density and consequently
genetic variation (Kapuscinski 1983). Populations subjected to high mortality
rates and maintained at low densities may be especially sensitive to toxic
substances for life history reasons—the individuals need to maintain high
growth and reproduction rates for the population to persist but both toxicants
and inbreeding effects reduce these—and for evolutionary reasons—the popula-
tions have a severely reduced capacity to adapt evolutionarily to the toxicant.
To relate this to guppy population extinction brought about by dieldrin
is purely speculative, but certainly many of the conditions conducive to
intensive inbreeding and reduction in genetic variation were present in the
population at LOW I, 40E. Prior to dieldrin introduction, the population had
maintained very low densities for many generations, being composed at most
of only a few mature fish at any one time. Further, the population was
completely isolated from gene flow.
At HIGH I, 40E, individual guppies were able to survive, grow, and
reproduce better than fish at LOW I due to greater food availability. The
density of the population at HIGH I, 40E was much greater than the density
at LOW I and, thus, perhaps genetic diversity was also greater. All of these
75
-------
factors may have enabled the population at HIGH I, 40E to better withstand
toxicant exposure.
For the most part, to this point, we have been concerned with systems
at steady-state for two related reasons. First, we wished to illustrate
theoretically and empirically some of the system generalizations posed in
the Introduction, including the multisteady-state nature of systems and the
relation between system steady-state structure and system environment.
Second, system steady-states have explanatory utilty, for they are the
"targets" of system trajectories; their location in phase space, and thus
where trajectories want to go, changes with changes in environmental
conditions. Both the complex and the simple laboratory communities estab-
lished (near)steady-states. Their structure and organization was given
meaning with multisteady-state models. We evaluated effects of a toxicant
under different sets of environmental conditions in terms of the extent to
which the NSS location of the system in phase space (NSS system structure)
was altered by exposure to the toxicant.
In natural systems and in many kinds of laboratory systems environmental
conditions may be continually changing and systems may never establish steady-
states or even NSS'.s. The second generalization of the set posed in the
Introduction states that system dynamics, or changes in system structure
and organization, can be understood as an n-dimensional system trajectory
in continuous pursuit of an n-dimensional system steady-state point whose
location in phase space is continually changing with changes in the state of
the system's environment. What, then, are some implications of changing
environmental conditions in understanding the effects of toxic substances
on ecological systems and how can these effects be evaluated?
76
-------
Figure 13 . Phase plane representation of the relationship between Cl and HI in a
system. El, units of harvesting effort on Cl, and I constitute the
environment of the system. Stable trajectories are generated when I
and E repeatedly cycle in the sequence (1) LOW I, 2E1, (2) MED I, 6E1,
(3) HIGH I, 4E1, and (4) MED I, 2E1. Trajectory A represents changes
in Cl and H when toxicant is not present. The steady-state points the
trajectory tracks toward under the different sets of environmental
conditions are indicated by open circles. Under the same environmental
cycle, trajectories B and C represent changes in Cl and H when toxi-
cant causes reductions in Cl growth, reproduction, and survival of
10 percent and 25 percent respectively. Predator isoclines and
steady-state points (solid circles) at each E and I for the 25 percent
reduction are shown. Only the trajectory is shown for the 10 percent
reduction. The particular sequence of environmental conditions used
here has no special significance. It is intended as an example of
how interacting populations respond numerically to changes in environ-
mental conditions. The isoclines and trajectories on this phase plane
were generated through computer iteration using the command control
language SIMCON. (We appreciate the assistance of Grant Thompson,
Oak Creek Laboratory of Biology in performing these simulations.)
77
-------
LOW I MED j HIGH I
2E 1
C1
2E 1 25% TOX
4E 1
6E 1
25% TOX
6E1 25%TOX
H
78
-------
Shown in Figure 13 is a stable "control" trajectory (A) generated when
environmental conditions change in a regularly repeated sequence (1) LOW I,
2E1,; (2) MEDI, 6E1; (3) HIGH I, 4E1; (4) MED I, 2E1. The steady-state points
that the trajectory tracks toward under the different sets of conditions are
indicated by the open circles. Before the trajectory converges on a particular
steady-state, environmental conditions change, altering the location of the
steady-state point in phase space and consequently diverting the trajectory
toward a new steady-state. A "stable" trajectory pattern is eventually
established, repeated with every cycle of change in environmental conditions.
A toxic substance alters the location of the steady-state point in phase
space at each I and E, the extent of these changes being dependent upon
system organization and the particular levels of I and E. This alters the
form and location in phase space of the stable trajectory generated by a
particular sequence of change in I and E. Shown in Figure 13 are the stable
trajectories generated by the same sequence of I and E used to generate the
"control" trajectory, but with toxicant-induced reduction in growth, repro-
duction, and survival of Cl of 10 percent (B) and 25 percent(C). The
trajectories of systems exposed to the toxicant occupy different regions of
phase space than the control trajectory. There is considerable overlap of
the trajectory generated by a 10 percent reduction with the control trajectory.
Such a difference in system behavior may be difficult to detect under field
conditions. The trajectory generated by a 25 percent reduction occupies a
different region of phase space than the control trajectory. It is interesting
that at MED I, 6E1, 25% TOX the intersection of the predator and prey isoclines
is along the HI axis. This means that had conditions remained fixed at MED I,
6E1, 25% TOX, Cl would have been driven to extinction. But before the
trajectory could converge on this point, conditions changed; El was reduced
79
-------
from 6E1 to 4E1 and I increased from MED I to HIGH I, effectively shifting
the predator isocline upward and the prey isocline to the right and generating
a positive intersection in phase space. Thus, the extent of translocation of
the stable toxicant trajectory in phase space is no indication of the potential
severity of toxicant effects. Had environmental conditions not become more
favorable a 25 percent reduction in growth, reproduction, and survival would
have led to the extinction of Cl.
Fundamentally, when viewed within our perspective, the form of a system
trajectory is determined by the location of the system in phase space relative
to the steady-state point it is going toward and how that point is shifted
in phase space as environmental conditions change (Liss et al., in prep.).
This is so, we believe, for any performance of any system whatsoever, including
community structure and organization, population density, production, and
yield, and toxic substance uptake and concentration. If the environment
of a system had no regular cycle of change, a trajectory would wander through-
out phase space, pursuing an ever-shifting steady-state point and having no
particular unique form. It may be that under these conditions toxic substances
may be more difficult to detect under field conditions, unless they are severe,
because there is bound to be considerable overlap in domains of behavior of
systems that have been exposed to a toxicant and systems that have not.
But more than this, if we insist on confining ourselves to short-term
experiments on natural or laboratory systems and not worrying about envir-
onmental conditions, we must begin to wonder about the meaning of our
observations and their relevance to possible effects a toxicant may have on
natural systems. Any single measurement or series of measurements defines
only a single point on a trajectory or a segment of a trajectory. This may
have little significance if environmental conditions are continually changing
80
-------
and the trajectory is moving throughout phase space; it may tell us little
about toxicant effects and thus may be misleading in attempting to manage
toxic substances.
Frequently, in laboratory microcosm research, experiments are conducted
under a single set of environmental conditions. Unless the system establishes
a steady-state, the measurements represent only points on a trajectory going
toward a steady-state point, but the location of the point in phase space
is unknown. And, even if the system were to establish a steady-state, this
would represent only one in an infinite family of possible steady-states.
Toxicant effects measured under this particular set of conditions may have
little significance, for system response may vary with conditions in the
environment of the system. If the laboratory community experiment reported
here was conducted under only one set of environmental conditions, one may
get quite different impressions of the effect of dieldrin depending upon
whether conditions were fixed at, say, LOW I, OE, LOW I, 40E, or HIGH I, 40E.
Conducting experiments on any ecological system over ranges of important
environmental conditions provides better definition of the domain of toxicant
effects on the system (Warren and Liss 1977). This simply amounts to better
understanding of toxicant effects and this understanding must be the basis
of toxic substance control.
With laboratory systems we are attempting to model developing, adapting,
persistent natural systems of great organizational complexity. It is not
clear that any laboratory system can model natural systems very well. An
important consideration in developing laboratory models is the kind or class
f
of system to be modeled, for different kinds or classes of systems have
different capacities (Warren et al. 1983). The model should reflect some-
thing of the organization, capacity, and performances of that kind or class
81
-------
of natural system if there is to be any hope of extrapolating laboratory
results to natural systems.
In many kinds of microcosms, resources are inadequate and environmental
conditions insufficient to maintain persistence of the systems for any length
of time. In these systems what is it that we are measuring when we investigate
toxic substance behavior and effects? Essentially the origin (0,0) is the
steady-state point on a phase plane upon which such a system will converge.
The toxicant effect we measure, at best, is simply how much faster an already
dying system will meet its demise. We believe these sorts of systems should
not be favorably viewed as toxicant testing tools simply for reasons of
expedience, that is, simply because they can produce data over a relatively
short time period. These systems may fulfill the criteria of being short-term,
replicable systems but they may fall far short of other important criteria—
they possess few of the properties of developing, adapting, persistent
ecosystems. Microcosms of this kind should not be passed off as methods for
evaluating toxicant behavior and effects in ecosystems. The information on
transport, accumulation, and metabolism of toxicants generally garnered from
these microcosms may be better obtained with individual organism feeding
experiments through incorporation in the experiment of different food consump-
tion rates and other kinds of differences in environmental conditions
(Warren and Lass, 1977).
But by suggesting that well-conducted individual organism experiments
can provide the same kind of information gained from short-term, nonpersistent
microcosms, we do not wish to imply that the individual organism test will
suffice for determining impacts on higher levels of organization. We do not
believe this is so. Individual organism experiments when conducted under
82
-------
different sets of environmental conditions and with a life history perspective
(Sterns 1976, Warren and Liss, 1980, Kulbicki 1980), can be useful in under-
standing the impacts of a toxicant on life history patterns and capacities
and probably provide a lot of the same kind of information on uptake rates
and metabolism as is obtained in many kinds of microcosm studies. But, based
solely upon information of toxic effects on individual organism growth,
survival, and reproduction in the laboratory, we have almost no way of
knowing how a toxicant may affect the densities, yields, genetic structure,
adaptation, persistence, etc. of populations or the structure, organization,
development, and persistence of communities. There is no theoretical basis
for supposing that effects on population and community capacities and per- .
formances can be simply predicted from effects on individual organisms. The
kinds of results obtained in our laboratory community experiments could not
have been predicted from the results of any individual organism experiments
with dieldrin (Liss et al. 1980). And, the organization of natural communities
is far more complex than the organization of our laboratory communities.
Laboratory models of ecosystems are probably needed to illustrate or
demonstrate theoretical generalizations pertaining to toxic substance
behavior and effects and provide "tests" of the effects toxicants may have
on natural systems. Their utility depends upon the frameworks, theories,
and models available for their design, conduct, and interpretation. We are
concerned, however, that proper evaluation of community or ecosystem level
effects may be fundamentally incompatible with the criteria that toxicity
tests be essentially short-term and replicable.
83
-------
REFERENCES
Aim, G. 1959. Connection between maturity, size, and age in fishes.
Experiments carried out at the Kalarne Fishery Research Station.
Ann. Rep. Inst. Freshwater Res. Drotinghalm. 40:5-145.
Beverton, R.J.H. 1963. Maturation, growth, and mortality of Clupeid and
Engraulid stocks in relation to fishing. Rapp. Proces Verbaux Reun.
Cons. Perm. Intern. Explor. Mer. 154:44-67.
Booty, W.M. 1976. A theory of resource utilization. MS Thesis, Oregon
State University, Corvallis.
Burd, A.C. and D.H. Cushing. 1962. Studies on the Dunmore herring stock.
I. A population assessment. J. Cons. Intern. Explor. Merc. 29:277-301.
Chadwick, G.G., J.R. Palensky, and D.L. Shumway. 1972. Continuous-flow
dilution apparatus for toxicity studies. Proc. of 39th Pacific NW
Industrial Waste Management Conference, p. 101-105. Portland, Oregon.
Culley, D.D., and D.F. Ferguson. 1969. Patterns of insecticide resistance
in mosquito fish Gambusia affinis. J. Fish. Res. Bd. Can. 26:2395-2401.
Finger, S.E. 1980. Effects of perturbation on community structure and
organization of aquatic microcosms. MS Thesis, Oregon State University,
Corvallis.
Frankel, O.K. and M.E. Soule. 1981. Conservation and Evolution. Cambridge
University Press, New York.
Franklin, I.R. 1980. Evolutionary change in small populations. In
Conservation Biology; An Evolutionary-Ecological Perspective, p. 135-
149. M.E. Soule and B.A. Wilcox (eds). Sinauer, Massachusetts.
Kapuscinski, A. 1983. A fitness model for fisheries management. Ph.D.
Thesis in preparation. Marine Science Center, Oregon State University,
Newport.
Kulbricki, M.L. 1980. The effects of dieldrin and different food levels
on life history tactics of the guppy (Poecilia reticulata Peters).
MS Thesis, Oregon State University, Corvallis.
Lee, D.S. 1983. Competition and the multisteady-state structure and
organization of systems of populations. MS Thesis, Oregon State
University, Corvallis.
Liss, W.J. 1974. Dynamics of exploited guppy populations exposed to dieldrin.
MS Thesis, Oregon State University, Corvallis.
Liss, W.J. 1977. Toward a general theory of exploitation of fish popula-
tions. Ph.D. Thesis, Oregon State University, Corvallis.
84
-------
Liss, W.J. and C.E. Warren. 1980. Ecology of Aquatic Systems. In R.L.
Lackey and L. Nielsen (eds.), Fisheries Management, p. 41-80. Blackwell
Scientific Publications, Oxford.
Liss, W.J., D.M. Weltering, S.E. Finger, M.L. Kulbicki, and B. McClurken.
1980. Organization and adaptation of aquatic ecosystems exposed to
the insecticide dieldrin. NTIS, USEPA.
Liss, W.J., G.G. Thompson, and C.E. Warren. In preparation. Structure,
organization, and productivity of multisteady-state systems: a
perspective and some implications to fisheries theory.
Miller, R.B. 1956. The collapse and recovery of a small whitefish fishery.
J. Fish. Res. Bd. Can. 13:135-146.
Moav, R., T. Brody, and G. Hulata. 1978. Genetic improvement of wild fish
populations. Science 201:1090-1094.
Ricker, W.E. 1981. Changes in the average size and average age of Pacific
salmon. Can. J. Fish. Aquat. Sci. 38:1636-1656.
Sinnhuber, R.O., J.D. Hendricks, J.H. Wales, and G.B. Putnam. 1977. Neoplasm
in rainbow trout, a sensitive animal model for environmental carcin-
ogenesis. Ann. N.Y. Acad. Sci. 298-389.
Schaeffer, W.M., and P.F. Elson. 1975. The adaptive significance of
variations in life history among local populations of Atlantic salmon
in North America. Ecol. 56:577-590.
Soule, M.E. 1980. Thresholds for survival: maintaining fitness and
evolutionary potential. In Conservation Biology; An Evolutionary-
Ecological Perspective, p. 151-169. M.E. Soule and B.A. Wilcox (eds.)
Sinauer, Massachusetts.
Stearns, S.C. 1976. Life history tactics: A review of the ideas.
Quart. Rev. Biol. 51:3-47.
Thompson, G.G. 1981. Theoretical framework for economic fishery management.
M.S. Thesis, Oregon State University.
Warren, C.E. and W.J. Liss. 1977. Design and evaluation of laboratory
ecological system studies. Ecological Research Series, USEPA, EPA-
600/3-77-022.
Warren, C.E., M.W. Allen, and J.W. Haefner. 1979. Conceptual frameworks
and the philosophical foundations of general living systems theory.
Behavior. Sci. 24:296-310.
Warren, C.E. and W.J. Liss. 1980. Adaptation to aquatic environments.
In R.L. Lackey and L. Nielsen (eds.), Fisheries Management, p. 15-40.
Blackwell Scientific Publications, Oxford.
85
-------
Warren, C.E., W.J. Llss, G.L. Beach, C.A. Frissel, M.E. Gregor, M.D. Hurley,
D. McCullough, W.K. Seim, G.G. Thompson, and M.J. Wevers. 1983.
Systems classification and modeling of watersheds and streams. Draft
report, Dept. of Fisheries and Wildlife, Oregon State Univ., Corvallis.
Worden, D. 1976. SIMCON: a simulation control command language. Oregon
State Univ. Sea Grant College Program, Corvallis.
Weltering, D.M. 1981. Organization and adaptation of aquatic laboratory
ecosystems to the resource availability, exploitation, and a toxicant.
Ph.D. Thesis, Oregon State University, Corvallis.
Weltering, D.M., S.E. Finger, B. McClurken, and W.J. Liss. In preparation.
Structure and organization of laboratory ecosystems exposed to the
insecticide dieldrin.
86
------- |