ECOTOXICOLOGY: A Comprehensive Treatment - Chapter 29 potx

Clements: “3357_c029” — 2007/11/9 — 18:37 — page 611 — #1
Part V
Ecosystem Ecotoxicology
In amnesiac revery it is also easy to overlook the services that ecosystems provide humanity. They enrich
the soil and create the very air we breathe. Without these amenities, the remaining tenure of the human
race would be nasty and brief.
(E.O. Wilson 1999)
Ecosystems represent the highest and final level of biological organization that we will consider in
our treatment of ecotoxicology. It is appropriate that we conclude with a discussion of ecosystems,
which have been considered by some ecologists to be the fundamental units of nature (Tansley
1935). The critical defining feature of ecosystems that is unique from other levels of biological
organization we have considered is the inclusion of abiotic variables. Ecosystem ecotoxicology is
necessarily a multidisciplinary science, and the ecosystem processes that respond to contaminants
go beyond those of populations and communities. Because these processes are often scale dependent
(Carpenter and Turner 1998), effects of contaminants on ecosystem function also vary across spati-
otemporal scales. Ecosystem ecologists have made tremendous progress developingbiogeochemical
models of nutrient dynamics, and these models can be readily adapted to predict contaminant move-
ment within and between ecosystems. Quantifying effects of contaminants on ecosystem processes
and demonstrating causal relationships between stressors and responses is challenging. As a con-
sequence, ecosystem responses are not routinely measured in ecological risk assessments. However,
characterization of ecological integrity based exclusively on structural measurements has provided
a somewhat incomplete picture of how ecosystems respond to anthropogenic perturbations (Gessner
and Chauvet 2002). Furthermore, the unprecedented rate of species extinction occurring at a global
scale (Wilson 1999) requires that ecologists and ecotoxicologists develop a better appreciation of
the relationship between community patterns and ecosystem processes. Finally, many ecosystem
processes are intimately connected to ecosystem goods and services that are essential for the welfare
of humanity. The goal of this section is to demonstrate how contaminants and other anthropogenic
stressors affect these critical ecosystem processes and related services.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 612 — #2
612 Ecotoxicology: A Comprehensive Treatment

REFERENCES
Carpenter, S.R. and Turner, M.G., At last a journal devoted to ecosystem science, Ecosystems, 1, 1–5, 1998.
Gessner, M.O. and Chauvet, E., A case for using litter breakdown to assess functional stream integrity, Ecol.
Appl., 12, 498–510, 2002.
Tansley, A.G., The use and abuse of vegetational concepts and terms, Ecology, 16, 284–307, 1935.
Wilson, E.O., The Diversity of Life, W.W. Norton & Company, New York, 1999.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 613 — #3
29
Introduction to
Ecosystem Ecology
and Ecotoxicology
29.1 BACKGROUND AND DEFINITIONS
Ecosystems behave in ways that are very different from the systems described by other sciences.
(Ulanowicz 1997)
Ecosystems can be seen more powerfully as sequences of events rather than as things in a place. These
events are transformations of matter and energy that occur as the ecosystem does its work. Ecosystems
are process-oriented and more easily seen as temporally rather than spatially ordered.
(Allen and Hoekstra 1992)
Although the term ecosystem is broadly recognized by the general public and appears frequently in
the nonscientific literature, ecologists and ecotoxicologists still struggle with a precise definition.
Recognition that groups of plants formpredictableassociationsacrossbroad geographic regions was a
significant breakthrough inthe history of ecology(Clements 1916), and early plant ecologistsdevoted
considerable effort to understanding the mechanisms responsible for these patterns. Perhaps because
of the tremendous influence of Frederic Clements on the field of ecology, the contentious debates
regarding holistic and reductionist interpretations of natural systems continued well into the 1930s.
These debates figured prominently in the establishment and maturation of the emerging field of
ecosystem ecology. Rejecting the Clementsian superorganism perspective that growth, development,
and senescence of a community was analogous to that of individual organisms, the term ecosystem
was first introduced by Arthur Tansley in 1935 when he appropriately recognized the difficulty of

studying biotic and abiotic components of natural systems in isolation.
Though the organisms may claim our primary interest, when we are trying to think fundamentally we
cannot separate them from their special environment, with which they form one physical system.
(Tansley 1935)
Thus, one distinguishing feature of ecosystem ecology, which was recognized early in its history,
was the necessity of considering integrated physical, chemical, and biological processes. Ecosystem
ecologists are not simply recognizing the influences of the physical environment but are considering
organisms and the abiotic environment as part of a single system. This holistic perspective is funda-
mentally different than how lower levels of organization have been treated in ecology. Likens (1992)
defined an ecosystem as a “spatially explicit unit of the earth that includes all of the organisms along
with all components of the abiotic environment within its boundaries.” One can see by this broad
definition that while the spatial extent of an ecosystem remains somewhat vague, the emphasis is on
including organisms and the environment. We will also see that because of the focus on movement
of energy and abiotic materials (e.g., C, N, P), ecosystem ecology integrates the fields of chemistry,
physics, and biology and is, therefore, necessarily a multidisciplinary science.
613
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 614 — #4
614 Ecotoxicology: A Comprehensive Treatment
29.1.1 THE SPATIAL BOUNDARIES OF ECOSYSTEMS
Because of the loosely defined spatial and temporal boundaries, some ecologists have argued that
ecosystems lack the logical interconnectedness typical of other levels of biological organization
(Reiners 1986). Clearly, the spatial boundaries of an ecosystem often extend beyond those of its
component populations and communities. These broad spatial and temporal boundaries of ecosys-
tems are necessary because they provide ecologists with the flexibility to match questions with
appropriate scales. For example, to quantify the mass balance of nitrogen or phosphorus in a lake
ecosystem, it is necessary to include materials contributed from the surrounding watershed. Sim-
ilarly, to quantify the transport of organochlorines or other persistent organic pollutants through
an aquatic food web, assessment of atmospheric sources may be required. Although flexibility in
defining the spatial and temporal scale of an ecosystem is necessary, the classic studies of ecosystem

dynamics have been conducted in systems with well-defined boundaries such as watersheds and
lakes. Thus, ecologists recognize the necessity of including inputs of materials from outside sources,
but in practice ecosystem boundaries are more precisely defined.
While Tansley considered ecosystems “the basic units of nature on the face of the earth,” there
remains some debate in the literature over whether ecosystems actually exist or are simply an artifact
of our inability to adequately describe nature (Goldstein 1999). Contemporary ecologists still ques-
tion whether the ecosystem is a physical construct, as defined by Tansley, or more like a theoretical
concept that serves to organize our thoughts and ideas. Early definitions attempted to place specific
boundaries on ecosystems, lakes being the most obvious example. However, we now recognize that
ecosystems are connected to and influenced by features outside these traditional borders. Allen and
Hoekstra (1992) note that it is unworkable to consider an ecosystem simply as a place on a land-
scape. Thus the question becomes, is ecosystem science simply the study of processes (as opposed
to patterns)? We can readily discuss properties of ecosystems (e.g., trophic structure), but recognize
that it may not be possible or prudent to enclose ecosystems in arbitrary boundaries.
Arelatively broad delineation of ecosystem boundaries will also influence the scope and coverage
of ecosystem ecotoxicology considered in the following sections. In our previous discussion of food
web ecotoxicology, we described the structure of food webs and how contaminants may influence
linkages among trophic levels. Analyses of connectance, trophic linkages, and food chain length
provide important insights into community organization and help explain variation in contaminant
levels among consumers. In the following sections, we will emphasize factors that affect contaminant
transport in ecosystems and the potential effects of contaminants on bioenergetics, nutrient cycling,
and other ecosystem processes.
29.1.2 CONTRAST OF ENERGY FLOW AND MATERIALS CYCLING
Although the flow of energy and the transport of materials through an ecosystem are generally
treated separately in most ecosystem assessments, these processes are so intimately linked that it is
often more practical to consider them simultaneously. For example, the flow of energy is closely
associated with the transfer of carbon through photosynthesis and respiration. One important dis-
tinction between the movement of energy and abiotic materials through ecosystems concerns the
second law of thermodynamics, which essentially states that some energy is dissipated as heat
with each energy transformation. It is well established that energy flow through biological systems

is a highly inefficient, one-way process, with approximately 10% of energy transferred from one
trophic level to the next (Slobodkin 1961). This inefficiency greatly limits the number of trophic
levels in an ecosystem and accounts for the rarity of large predators (Colinvaux 1978). In con-
trast, abiotic materials such as nutrients and carbon are cycled through ecosystems, and the amount
of these materials increases with trophic level (Figure 29.1). These differences between energy
flow and materials cycling are at least partially responsible for the process of biomagnification in
top predators observed for many organic chemicals. Although the amount of energy decreases,
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 615 — #5
Introduction to Ecosystem Ecology and Ecotoxicology 615
Energy
Chemicals
Zooplankton
Planktivores
Piscivores
Phytoplankton
FIGURE 29.1 Hypothetical changes in energy and chemicals in an aquatic food web. Because energy is
dissipated as heat as it is transferred through a food chain, it decreases with trophic level. In contrast, many
chemicals, including toxic and bioaccumulative substances, cycle through an ecosystem and may increase with
trophic level. (Modified from Stiling (1999).)
many abiotic materials, including contaminants, tend to increase in concentration with trophic
level.
29.1.3 COMMUNITY STRUCTURE,ECOSYSTEM FUNCTION
AND
STABILITY
The precise characterization of ecosystem properties has important implications for how we define
ecosystem resistance and resilience. Previous studies have reported that structural characteristics,
such as abundance or the number of species, are generally more sensitive than ecosystem processes,
such as energy flow or nutrient cycling (Schindler 1987). Consider the example of acidified lakes,
which have been studied extensively in ecosystem ecology. If we define resistance based on alter-

ations in primary productivity of an acidified lake, we may conclude the ecosystem was relatively
stable. However, if we assessed stability based on loss of species or changes in community com-
position, responses known to be considerably more sensitive, we may conclude that the system had
low stability. The important point is that populations and communities may appear to behave quite
differently when they are considered in isolation from ecosystems. The simplification of ecosystems
to component parts has also contributed to the controversy over the relationship between stability
and diversity described in previous sections. Attempts to define the stability of ecosystem processes
based on the diversity of its components (e.g., number of species) have met with mixed success.
29.2 ECOSYSTEM ECOLOGY AND ECOTOXICOLOGY:
A HISTORICAL CONTEXT
Compared to the study of population and community ecology, an ecosystem perspective is relatively
new in the history of ecology. There is also considerable variation among ecologists in their precise
descriptions of ecosystems, which have been compared to individual organisms and precisely engin-
eered (though relatively inefficient) machines. Ecosystems have been described as static or dynamic,
as open or closed, and as predictable or stochastic collections of unrelated, noninteracting species
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 616 — #6
616 Ecotoxicology: A Comprehensive Treatment
(Ulanowicz 1997). For some contemporary ecologists, the field of ecology is predominantly a study
of the movement of energy and materials through ecosystems. Others consider the movement of
materials to be an outcome of the interactions among organisms and with the abiotic environment.
These different characterizations reflect some uncertainty in the literature with respect to the ecosys-
tem as an object of study or simply a concept. Over the past 50 years, the predominant perspective of
an ecosystem has evolved from the idea of spatiotemporal constancy to coupled dynamics in space
in time. Despite this evolution, developing a comprehensive framework to address spatiotemporal
issues in ecosystem ecology remains a challenge (O’Neill et al. 1986), and how we describe an
ecosystem is often influenced by personal bias or point of reference.
29.2.1 EARLY DEVELOPMENT OF THE ECOSYSTEM CONCEPT
As noted above, Tansley (1871–1955) coined the term ecosystem and was the first to publish the
concept in a technical paper. In the History of the Ecosystem Concept in Ecology, Golley (1993)

argued that Tansley’s inclusion of biotic and abiotic processes in the definition of an ecosystem
was an attempt to resolve the conceptual disagreements among plant ecologists concerning the
hierarchical versus organismicnatureofacommunity. In 1942, theecosystem concept was formalized
by Raymond Lindeman into the “trophic dynamic aspect,” widely recognized as one of the most
significant contributions in the early history of ecology (Lindeman 1942). The most striking aspect of
this original work was Lindeman’s attempts to quantify seasonal dynamics of vegetation and animal
production in a small lake (Cedar Bog Lake, Minnesota) and to characterize an ecosystem based
on energy flow. He also organized different groups of species into categories based on their feeding
habits or trophic level (e.g., browsers, plankton predators, benthic predators). More importantly,
he highlighted the interactions between biotic and abiotic components of the ecosystem. Important
concepts such as the substitution of units of energy (calories) for biomass, estimates of production
based on turnover, and calculation of ecological efficiencies anticipated questions that would figure
prominently in contemporary ecosystem research. However, the most significant contribution of the
work was therecognition that energy, or morespecifically calories, was the mostappropriate currency
by which to characterize ecosystems. Ironically, Lindeman’s original manuscript was rejected by
Ecology, primarily because of its overly theoretical nature. The paper was accepted only after strong
appeal from Lindeman’s Ph.D. advisor, the famous Yale limnologist G.E. Hutchinson, and published
after Lindeman’s death in 1942.
While Lindeman’s classic paper introduced the trophic dynamic concept and formalized the study
of ecosystem ecology, it was the publication of Eugene P. Odum’s (1953) classic text Fundamentals
of Ecology a decade later that placed ecosystem studies in the mainstream of ecological research.
This textbook greatly influenced a generation of ecologists during a critical period of development
and allowed the ecosystem concept to finally emerge as a legitimate topic of ecological research.
Ecology was gradually attempting to move from a predominantly descriptive science concerned
primarily with natural history to a more mechanistic-based science that sought to achieve the status
of chemistry and physics. Interpretation of ecological processes using laws of thermodynamics
appealed to many ecologists. This work also initiated a series of disputes among ecologists regarding
the usefulness of mathematical models for quantifying ecosystem dynamics. Ecosystem ecologists
were criticized for reducing the complexity of ecosystems to fewer and fewer components, and
for simplifying interactions among these components using strictly deterministic models. Golley

(1993) notes that much of the ecosystem research conducted during this period was little more
than “machine theory applied to nature.” The ecosystem as a machine concept and the application
of large-scale ecosystem models, referred to as “brute force reductionism” (Allen and Starr 1982)
figured prominently in the earlyhistory of ecosystemresearch.Although there issome dispute that the
complex box-and-arrow models of system ecologists represent testable hypotheses (Golley 1993),
they at least provided ecologists with mechanistic explanations for patterns observed in nature.
Providing insight into mechanisms, which has long been considered the holy grail of ecological
research (Ulanowicz 1997), is likely to improve the ability of ecosystem ecologists to address
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 617 — #7
Introduction to Ecosystem Ecology and Ecotoxicology 617
applied issues. Although Odum’s textbook preceded the environmental movement by over a decade,
it appealed to a growing number of ecologists concerned with human impacts on natural systems.
At a time when humans were only beginning to understand the potential effects of their actions on
the environment, this book stands out as one of the first to emphasize the importance of including
anthropogenic activities in any assessment of ecosystem structure and function.
29.2.2 QUANTIFICATION OF ENERGY FLOW THROUGH
ECOSYSTEMS
The flow of energy described in the conceptual diagrams of Elton and Lindeman was quantified
in the early 1960s. These initial analyses confirmed theoretical predictions showing the relative
inefficiency of energy transfers from primary producers to herbivores and predators. Golley’s (1960)
classic study of energy dynamics conducted in an old field with a relatively simple food chain from
plants to herbivores (mice) and predators (weasels) (kcal/ha/year) showed that only a small fraction
of the energy in primary producers results in predator production (Figure 29.2). About 50% of the
sunlight striking the field is of the wavelength that can be used by plants, and only about 1% of
this is converted to Net Primary Production (NPP). Fisher and Likens (1973) quantified all organic
material input and output to develop an energy budget for Bear Brook, a small second-order stream
in the northeastern United States. Over 99% of the energy input to the stream was allochthonous,
indicating that Bear Brook was a strongly heterotrophic system.
Ecosystem-level studies by Golley (1960), Fisher and Likens (1973), and others demonstrated

that the movement of energy through an ecosystem could be quantified; however, the food chains in
these initial studies were relatively simple. Quantifying energetics of more complex systems proved
to be a daunting task. One significant event duringthis period facilitatedthe development of new tech-
niques to quantify energy and materials flow in ecosystems. Funding provided by theAtomic Energy
Commission (AEC) allowed researchers to study the distribution of radioactive materials in biotic
Incident sunlight
47.1 × 10
8
46.5 × 10
8
GPP = 58.3 × 10
6
NPP = 49.5 × 10
6
8.8 × 10
6
R
49.3 × 10
6
Available to mice
15.8 × 10
6
Consumption
25.0 × 10
4
17.0 × 10
4
R
7.4 × 10
4

Production
5.17 × 10
3
12.0 × 10
3
Consumption
5.82 × 10
3
5.43 × 10
3
260
Production
130
Import: 13.5 × 10
3
Population increase
1.57 × 10
3
R
20
Population increase
117
Plants
Mice
Weasels
Unused portions
FIGURE 29.2 Energy flow in an old field ecosystem showing the relative amounts of energy (kcal/ha/year)
from incident sunlight to top predators. (Modified from Golley (1960).)
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 618 — #8

618 Ecotoxicology: A Comprehensive Treatment
and abiotic compartments following intentional releases associated with tests of explosive devices.
It is no coincidence that several prominent centers for ecosystem studies in the United States, includ-
ing Oak Ridge National Laboratory and Savannah River Ecology Laboratory, were associated with
nuclear testing facilities and involved the emerging area of radiation ecology. Recognition that radio-
active materials moved between biotic and abiotic compartments and accumulated in food chains
was a significant discovery that linked basic and applied ecological research. A readily available
source of funding from the AEC certainly facilitated this association (Golley 1993). Experimental
techniques such as the addition of radioactive tracers improved the ability of ecologists to quantify
the movement of energy and materials through an ecosystem. By labeling primary producers with
a radioactive isotope, most commonly phosphorus-32 (
32
P), ecologists can trace the movement of
energy through a foodweb. Whittaker (1961) pioneered this technique in aquatic ecology and used
microcosms to measure movement of
32
P through an aquatic food web. Similar experiments were
conducted by Ball and Hooper (1963) in a Michigan trout stream. Tracer experiments became a
mainstay for the emerging field of radioecology that allowed ecosystem ecologists to estimate the
rate of movement of materials and energy through the system. The large number of studies that fol-
lowed reflected the growing perspective that energy is the universal currency in ecosystems and that
an understanding of energy flow was critical to the study of ecosystem ecology. This development
proved to be especially significant to the study of ecosystem ecotoxicology because many of the
transport and fate models used to quantify the movement of radioactive materials were eventually
adopted and modified for the study of contaminants.
Quantifying the movement of energy and materials through an ecosystem generally required a
mass budget approach in which inputs and outputs were measured. Thus, lakes and streams became
appropriate modelsfor thestudy ofecosystems because, unliketerrestrial ecosystems, theboundaries
were well defined. As we will see, recognizing the exchanges that take place between ecosystems,
such as the movement of materials from terrestrial to aquatic habitats, has become an important area

of ecosystem research (Ulanowicz 1997). In fact, despite precedence for the term ecosystem being
attributed to Tansley, earlier writings of Stephen Forbes, an American limnologist, also highlighted
the role of abiotic processes and interactions within communities and recognized the importance of
studying ecosystem function in addition to structure (Forbes 1887). Using extensive data collected
from Silver Springs, FL, H.T. Odum was the first to quantify the inputs and outputs of materials
through an ecosystem, thus calculating a mass budget and providing an estimate of metabolism
(Odum 1957). This approach proved especially insightful because estimates of mass budgets, either
of natural materials such as nutrients or of synthetic organic compounds such as pesticides, have
been the workhorse of ecosystem research.
29.2.3 THE INTERNATIONAL BIOLOGICAL PROGRAM AND
THE
MATURATION OF ECOSYSTEM SCIENCE
The early history of ecosystem science focused on three general areas of research: characterization
of the structure and function of whole ecosystems, quantification of energy flow, and estimation
of ecosystem productivity (Golley 1993). There was relatively little effort during this initial period
devoted to the study of nutrient cycling and the flow of abiotic materials through ecosystems. Per-
haps more importantly, relatively little funding was available to pursue what was considered to be a
somewhat intractable research topic. This changed in the early 1960s when the International Biolo-
gical Program (IBP) provided a unique focus on ecosystem research and, more importantly, funding
opportunities for large-scale and long-term ecosystem-level studies. The pioneering investigations
into biogeochemical cycling at Hubbard Brook Experimental Forest, New Hampshire demonstrated
that ecosystem-level questions were both manageable and could address critical applied issues (Bor-
mann and Likens 1967). By quantifying inputs and outputs of various nutrients, cations, and anions,
these researchers demonstrated that materials budgets for an entire ecosystem could be developed.
More importantly, they expanded the traditional boundaries of stream ecosystems to include the
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 619 — #9
Introduction to Ecosystem Ecology and Ecotoxicology 619
surrounding upland areas and pioneered the field of watershed research. From an applied ecotoxic-
ological perspective, the creation of a watershed budget for Hubbard Brook also provided some of

the first concrete evidence of the effects of acid rain on ecosystems in the United States (Likens et al.
1996).
29.3 CHALLENGES TO THE STUDY OF
WHOLE SYSTEMS
The answer for ecosystems lies neither in the elegant simplicity of classical physics nor in the fascination
for detail of natural history.
(Holling and Allen 2002)
Ecologists who believed whole ecosystems were the most appropriate scale of their investigations
soon realized they faced several significant challenges. An ecosystem perspective would require
estimates of biomass and production of all resident species—clearly an impossible task. If ecologists
were to study ecosystems in their entirety, a system-level approach was necessary. Various solutions
to this dilemma were offered, including limiting analyses to the few dominant species and assuming
that related species performed similar functions. The second alternative, the approach used by most
contemporary ecologists, was to assign species to functional groups and characterize energy and
materials flow through these groups. This approach required numerous simplifying assumptions,
and many ecologists were critical of the loss of information that occurred when aggregating feed-
ing habits of different species. In addition to minimizing species-specific differences, categorizing
organisms into functional feeding groups ignored seasonal and ontogenetic variation. Suter (1993)
lists several additional impediments to ecosystem-level assessments, including greater costs, lack of
standardization, lack of consensus over relevant endpoints, ecosystem complexity, high variation,
and relative insensitivity. Because much of ecosystem ecology remains purely descriptive, there has
been criticism that the hypothetico-deductive approach advocated by many philosophers of science
(Popper 1972) has been neglected. These practical and conceptual impediments partially explain
why ecotoxicologists have not pursued a more rigorous program of research in ecosystem assess-
ments. Clearly, the relevant question for many of these issues is how much detail can we ignore
and still have an adequate representation of overall ecosystem function. However, downplaying the
importance of species in favor of characterizing ecosystems based entirely on processes has received
harsh criticism, particularly in the field of conservation biology (Goldstein 1999).
Finally, our ability to understand effects of contaminants on ecosystems is both facilitated and
impeded by their self-organizing and cybernetic characteristics (O’Neill et al. 1986). The perspective

that ecosystems are controlled by stabilizing negative feedback relationships is relatively widespread
in ecology. Indeed, the ability of some ecosystems to quickly return to predisturbance conditions
following perturbation implies some degree of organization and homeostasis. This is encouraging
and suggests that, despite inherent complexity, ecosystems are legitimate objects of study and that
patterns and processes are tractable. However, the resilience and resistance of ecosystem processes
to disturbance may hamper our ability to quantify these responses.
29.3.1 TEMPORAL SCALE
In addition to questions regarding the appropriate boundaries and spatial scale, the relevant time scale
required to adequately quantify ecosystem-level processes requires careful consideration. According
to hierarchy theory, responses at higher levels of organization occur slower than those at lower
levels (Figure 29.3). For example, bioaccumulation of contaminants and resulting physiological and
behavioral alterations in organisms may occur rapidly following the discharge of toxic materials to an
ecosystem. However, it may require considerably longer time before we observe discernible effects
on ecosystem processes. The temporal scale of ecosystem responses is an important consideration
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 620 — #10
620 Ecotoxicology: A Comprehensive Treatment
Pollutant Input
Behavioral Response
Biochemical response
Response
Morphological
Response
Population Impact
Functional Changes
Time scale
(h)
10
1
10

2
10
5
0
10
3
10
4
Bioaccumulation
Structural Changes
Altered Performance
Pollutant input
Behavioral response
Physiological
response
Morphological
response
Population impact
Functional changes
0
Bioaccumulation
Structural changes
Altered performance
FIGURE 29.3 Time scale for responses to chemical pollutants at different levels of biological organization.
Effects of chemicals on physiological and biochemical endpoints are expected to occur within hours to days,
whereas community- and ecosystem-level responses may require months to years. (Modified from Sheehan
(1984).)
Response
Time (decades)
Response

Time (months)
FIGURE 29.4 Importance of temporal scale in ecosystem assessments. Response trajectories to ecosystem
perturbations measured over very short time scales (e.g., months), the typical duration of many ecological
investigations, will likely be quite different from those measured over longer time scales (e.g., decades).
in assessing impacts of anthropogenic perturbations. Ecosystem responses to global climate change
at one temporal scale (e.g., hundreds of years) may show very different responses over shorter time
periods (Figure 29.4). A sampling regime that is too short will not capture the patterns occurring
over longer time periods. It is therefore critical that the time scale of ecosystem responses and the
methodological approaches and sampling frequency designed to assess these responses match the
expected time scale of the perturbation. In a review of over 800 experimental and field studies,
Tilman (1989) reported that over 75% were limited to 1–2 years. Although a 2-year duration may be
adequate to characterize some ecosystem processes, these short-term ecosystem studies will likely
miss novel events, such as droughts or other natural disturbances, which are important features of
ecosystems.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 621 — #11
Introduction to Ecosystem Ecology and Ecotoxicology 621
29.4 THE ROLE OF ECOSYSTEM THEORY
29.4.1 S
UCCESSION THEORY AND THE STRATEGY OF ECOSYSTEM
DEVELOPMENT
Ecologists have long recognized that ecosystems change over time and that these changes are often
orderly and predictable. Characterizing long-term temporal changes in ecosystem structure and
function is problematic, and therefore researchers studying succession have relied heavily on space
for time substitutions. For example, by comparing vegetation patterns along newly formed sand
dunes of Lake Michigan, ecologists could visualize the transition from pioneer to climax species
along a defined spatial gradient (Cowles 1899). Because ecological succession is often initiated
following natural or anthropogenic disturbance, it is appropriate to consider temporal changes within
the context of ecosystem perturbation and recovery.
Odum’s (1969) classic paper “The strategy of ecosystem development” recognized the parallels

between developmental biology of organisms and succession of ecosystems. He stated that the
“strategy” of succession as a short-term process is basically the same as the “strategy” of long-term
evolutionary development of the biosphere. The overall strategy ofecosystem development according
to Odum was to achieve as large and diverse an organic structure as possible within the constraints of
available energy and materials. Due to the misgivings of many ecologists concerning the organismal
or superorganismal properties of ecosystems, Odum’s use of the word “strategy” is somewhat unfor-
tunate. However, this paper is especially significant because it was one of the first to relate ecosystem
processes such as succession and stability to anthropogenic disturbance. Ecosystem developmental
changes, such as shifts in the ratio of primary production (P) to respiration (R) and changes in
species diversity, are considered functional indices of ecosystem maturity. Because of the solid the-
oretical underpinnings and the generality of these responses, they may represent useful measures of
ecosystem responses to contaminants (Table 29.1).
Ecologists generally employ two very different methodologies to study ecosystem processes. The
first approach employs the traditional hypothetical-deductive method and relies on a combination of
induction, observation, and experimentation. This approach tends to be more site specific, and the
results often pertain toaspecific set of questionsina particular ecosystem. Much of applied ecosystem
ecology, in which researchers attempt to identify the causes of specific alterations in ecological
processes, uses this form of inquiry. In the second approach, ecologists develop theoretical principles
and mathematical models to draw inferences about processes that can be generalized across different
ecosystems. Sagoff (2003) discusses several conceptual obstacles faced by researchers using these
theoretical or “top-down” approaches. Because these obstacles pertain to how we define ecosystems
and how we isolate cause and effect relationships, a brief discussion is warranted here.
Perhaps the most serious challenge to theoretical ecosystem ecology is defining the class of
objects that constitute an ecosystem. Sagoff (2003) argues that most definitions are either over-
or under-inclusive. The broad definition cited above would include such diverse systems as the
bacterial assemblages living in a cow’s intestines as well as all of Lake Superior. Asecond challenge
is the remarkably diverse ways in which ecosystem ecologists have attempted to explain processes
occurring in nature. Ecologists have borrowed heavily from information systems, chaos theory,
statistical mechanics, thermodynamics, cybernetic systems, and hierarchy theory (to name a few)
to develop ecosystem models (Sagoff 2003). Finally, the use of mathematical models to address

applied issues in environmental biology may representthe greatest challenge to theoretical ecosystem
ecology. Although someresearchers are pessimistic aboutour ability tointegrate ecosystem models in
applied ecology (Sarkar 1996), it is likely that the following decades will see numerous opportunities
to test model predictions inalteredecosystems. Deviationinthe behavior of these systemsfrommodel
predictions may be used as a measure of the level of perturbation.
We recognize that it will not be possible or practical to study the response of all ecosystem com-
ponents to anthropogenic stressors. To measure effects of soil acidification on decomposition rate,
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 622 — #12
622 Ecotoxicology: A Comprehensive Treatment
TABLE 29.1
Examples of Population, Community, and Ecosystem
Attributes of Developing and Mature Systems
Attribute
Developing
Stages
Mature
Stages
Life history characteristics
Niche specialization Broad Narrow
Organism size Small Large
Life cycles Short; simple Long, complex
Community structure
Total organic matter pool Small Large
Inorganic nutrients Extrabiotic Intrabiotic
Species diversity and evenness Low High
Spatial heterogeneity Low High
Ecosystem attributes
Gross production:respiration (P/R) > or <1.0 ∼1.0
Gross production:biomass (P/B) High Low

Net community production High Low
Food chains Linear Web-like
Mineral cycles Open Closed
Nutrient exchange Rapid Slow
Role of detritus Less important Important
We suggest that many of these same attributes may be employed tocharacterize
the level of anthropogenic disturbance in ecosystems.
Source: Modified from Odum (1969).
ecosystem ecologists generally will not attempt to characterize structural composition of microbial
communities but simply rely on surrogate responses such as litter decay or respiration rates. There-
fore, in addition to deciding which groups of objects are appropriately classified as ecosystems, we
must also decide which processes or component parts are important and which can be omitted from
our characterizations.
In summary, because ecosystems are complex, dynamic, spatially variable, and often controlled
by interacting physical, chemical, biological, and socioeconomic factors (Gosz 1999), improving our
basic understanding of ecosystem processes and identifying solutions to applied problems will not
be easy. As noted in Chapter 20, mechanistic explanations for processes occurring at higher levels
of biological organization are often lacking or difficult to identify. While simplification of these
processes may be necessary to facilitate progress, we must acknowledge ecosystem complexity and
incorporate complexity into our investigations (Pace and Groffman 1998).
29.4.2 HIERARCHY THEORY AND THE HOLISTIC PERSPECTIVE OF
ECOSYSTEMS
The idea that ecosystems can be viewed as parts within parts has served as a focal point for the
debate between holistic and reductionist interpretations of ecological phenomena since Tansley first
introduced the ecosystem concept. Fundamentally, the question becomes, can we learn more by
studying ecosystem-level processes such as the flow of energy and the movement of materials or
should we devote more effort to studying the behavior of component parts (e.g., populations and
communities)? Central to this debate is the idea that emergent properties of an ecosystem cannot
be understood by studying the behavior of component parts. It is unlikely that we would make
© 2008 by Taylor & Francis Group, LLC

Clements: “3357_c029” — 2007/11/9 — 18:37 — page 623 — #13
Introduction to Ecosystem Ecology and Ecotoxicology 623
much headway in understanding how energy flows or materials are cycled through an ecosystem
by detailed analyses of populations and communities. The analysis of these processes requires an
ecosystem-level perspective. Similarly, although we can readily measure concentrations of persistent
organic chemicals in various compartments of an ecosystem, a complete assessment of fate and
transport requires a holisticecosystem perspective. We agree withCarney(1987) that there isacritical
need for better integration among population/community ecology and ecosystem-level ecology. This
integration will provide a much broader context for population and community ecologists to interpret
demographic patterns and species interactions. Integrated studies across these levels of organization
will provide ecosystem ecologists with a better understanding of mechanisms responsible for energy
flow, material fluxes, and nutrient cycling.
29.5 RECENT DEVELOPMENTS IN ECOSYSTEM
SCIENCE
In 1998, six decades after Tansley first introduced the term, ecosystem, a new journal dedicated
entirely to the study of ecosystem science was introduced. The debut of Ecosystems represented an
important turning point in the history of ecosystem science. Questions within the field regarding
the definition of an ecosystem and specific spatiotemporal boundaries may be resolved now that
ecosystem ecologists have a forum for discussion. Among the recent accomplishments within the
field of ecosystem science, Carpenter and Turner (1998) list several examples in applied ecosystem
ecology and ecotoxicology, including a better understanding of fate and transport of contaminants
(Table 29.2). Most notable among the contributions of ecosystem science to applied issues is the
improved understanding of the effects of acid deposition on ecosystem processes. Applied issues
will continue to play a prominent role in ecosystem science as ecologists struggle with the dual
challenges of basic research and environmental problem solving. The emphasis on applied issues
in ecosystem science is obvious from a casual review of the first few volumes of this new journal,
which reveals an impressive assortment of papers devoted to topics such as effects of contaminants
on ecosystem processes, effects of nutrient deposition and eutrophication, and responses to climate
change.
Another important development in ecosystem science is the recognition that humans play an

increasingly important role in controlling the flux of materials through aquatic and terrestrial eco-
systems. Although the inclusion of humans as components of ecosystems was initially proposed by
Odum (1953), ecologists have traditionally conducted research in areas presumed to be free from
human influences. While there remains some debate as to whether human-dominated landscapes can
be defined as ecosystems, most ecologists recognize that there are relatively few pristine ecosystems
and that any analysis of the structure and function of an ecosystem must account for anthropogenic
inputs. Even relatively remote areas located within protected habitats or wilderness areas receive
TABLE 29.2
Significant Accomplishments in Basic and Applied Ecosystem Science
• Understanding the movement of energy and materials through freshwater and marine ecosystems
• Assessment of feedbacks between plants and animals and the abiotic environment
• Understanding the causes and consequences of eutrophication
• Quantification of the fate and transport of contaminants through ecosystems
• Understanding abiotic controls of production and the relationship with climate change
• Understanding the importance of belowground processes
• Recognizing the scale dependence of ecosystem processes
Source: Modified from Carpenter and Turner (1998).
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 624 — #14
624 Ecotoxicology: A Comprehensive Treatment
inputs from atmospheric deposition. For example, Baron et al. (2000) reported increased nitro-
gen deposition in Rocky Mountain National Park, Colorado and associated changes in forest and
watershed processes. Finally, because humans also are an important component of the history of
ecosystems, processes measured in these systems may reflect previous disturbance events. Similar
to the community conditioning hypothesis described in Chapter 25, our assessments of ecosystems
must account for ecological legacies that may continue to influence processes (Carpenter and Turner
1998).
29.5.1 GENERAL METHODOLOGICAL APPROACHES
The general methods employed by ecosystem ecologists and ecotoxicologists to study ecosystem
processes are similar to those used by population and community ecologists (Figure 29.5). A com-

bination of modeling, comparative, long-term monitoring, and experimental approaches have been
used to address a variety of basic and applied research questions (Pace and Groffman 1998). Exper-
imental approaches often involve pulsing a system and measuring the associated signal as the pulse
passes through the system (Allen and Hoekstra 1992). Some of the classic experiments in ecosystem
ecology involved the addition of tracers or other materials that are then followed over time to differ-
ent compartments of the ecosystem. Measurements of energy and material transformations within a
system provide information on fate, whereas measurements of inputs and outputs provide estimates
of mass balance and ecosystem processing. Although some combination of modeling, comparat-
ive, long-term monitoring, and experimental approaches will be required to address many applied
questions, it is unlikely that these four approaches will contribute equally to the advancement of eco-
system science. Borrowing from a metaphor developed by Karr (1993) to describe issues in water
quality management, it seems reasonable to represent each approach as the adjustable legs of a four-
legged table placed on uneven ground. In this example, the relative length of each leg represents the
Experiments
Comparisons
Models and theory
Ecosystem
science
Experiments
Long-term
monitoring
Long-term
monitoring
Comparisons
Ecosystem
science
Models and theory
FIGURE 29.5 Descriptive, experimental, and theoretical approaches used to assess ecosystem processes to
perturbations. The relative size of the arrows indicates the importance of each approach. The lower panel is
an example common in many ecotoxicological investigations where long-term monitoring and comparative

studies were more important than experimental or theoretical approaches. (Modified from Pace and Groffman
(1998).)
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 625 — #15
Introduction to Ecosystem Ecology and Ecotoxicology 625
significance of the approach for addressing a specific question. For example, because experimental
introduction of persistent toxic chemicals into whole ecosystems is neither practical nor ethical,
many of the questions related to effects of toxic chemicals on ecosystem processes will be addressed
primarily using a combination of comparative and long-term studies. Experimental assessments of
the fate of persistent chemicals in ecosystems will likely be limited to small-scale microcosm or
mesocosm approaches. In contrast, modeling and long-term paleoecological approaches may be
more appropriate for predicting ecosystem responses to global climate change.
29.5.2 THE IMPORTANCE OF MULTIDISCIPLINARY RESEARCH IN
ECOSYSTEM ECOLOGY AND ECOTOXICOLOGY
It has become clear that basic and applied ecosystem research requires a multidisciplinary approach,
making it difficult to conduct ecosystem-level studies as an individual investigator. A review of
papers published in the new journal Ecosystems reveals relatively few single-investigator public-
ations. However, institutional barriers have impeded development of multidisciplinary approaches
and often stymied collaboration between researchers from different fields. Ecosystem-level research
was also hampered because of the difficulty in fitting abiotic aspects of ecosystem ecology into tra-
ditional biology departments where emphasis remained on individual research (Golley 1993). The
overwhelming complexity of ecosystems requires a multidisciplinary approach, but this same com-
plexity segregates investigators into more traditional and specialized areas of research (Maciorowski
1988). The existence of disciplines within a field of study is often necessary to understand this com-
plexity; however, some of the most innovative developments in science result from cooperation
among researchers in different disciplines. These interactions produce new research questions and
require participants to view complex issues from very different perspectives. Naiman (1999) provides
a unique perspective on the benefits and pitfalls of conducting interdisciplinary research as well as
several strategies for overcoming these obstacles. Opportunities for interdisciplinary research may
be found within the broad field of ecosystem ecology, with distinctly different disciplines in the

physical and chemical sciences, and across the range of spatiotemporal scales that engage ecologists
(Pickett et al. 1999).
29.5.3 STRONG INFERENCE VERSUS ADAPTIVE INFERENCE:
S
TRATEGIES FOR UNDERSTANDING ECOSYSTEM DYNAMICS
In the early 1970s, many ecologists were strongly influenced by the writings of philosophers of
science such as Popper, Kuhn, and Platt. Popper’s (1972) emphasis on falsification and Platt’s
(1964) process of using strong inference and multiple working hypotheses in scientific investig-
ations proved to be highly successful in physics and molecular biology and resonated well with
population and community ecologists. Strong aversion to committing type I errors (e.g., incorrectly
rejecting a null hypothesis) greatly influenced experimental designs and generated criticism of unrep-
licated ecosystem-level experiments(Carpenter 1989). The process ofsystematically culling unlikely
alternative hypotheses and deriving truth was perceived as the most rigorous application of Platt’s
“strong inference.” However, because of the complexity and dynamic nature of ecosystems, single
explanations are often insufficient, and unambiguous tests of hypotheses are unlikely to explain the
patterns observed in nature. Holling andAllen (2002) argue that strictly guarding against type I errors
may inhibit the creativity necessary to unravel these complexities. They propose a course of inquiry
called “adaptive inference” in which researchers shift concerns between type I and type II errors
(Figure 29.6). Attempts to minimize type II errors dominate the process initially, thereby allowing
researchers uninhibited speculation about cause and effect relationships. These initial propositions
(or “bold conjectures” as described by Popper) are themselves untestable, but capable of generating
testable hypotheses. The process gradually shifts from propositions to hypotheses and eventually to
models as researchers become more concerned about type I errors.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 626 — #16
626 Ecotoxicology: A Comprehensive Treatment
Theory Generation
Avoid type I error
Proposals
Hypotheses

Models
Observation of
pattern
Tests of
pattern
Formulation of
hypotheses
Consistency of
pattern
Formulation of
alternative
competing
hypotheses
Comparative tests
to risk hypotheses
Theory generation
Theory maturation
Avoid type II error
FIGURE 29.6 Adaptive inference in ecosystem research. Adaptive inference is a process in which concerns
over type II errors (i.e., incorrectly rejecting an alternative hypothesis) are less important early in the investiga-
tion, thereby allowing unrestrained speculation about patterns. The emphasis shifts over time to concerns over
type I errors (i.e., incorrectly rejecting a null hypothesis) to sort out false hypotheses. (Modified from Holling
and Allen (2002).)
29.6 ECOSYTEM ECOTOXICOLOGY
One primary focus of the field of ecology is to describe patterns and processes in the natural world.
Effects of contaminants on these patterns, such as differences in abundance or species richness
upstream and downstream from a toxic discharge, was the subject of community ecotoxicology
described in the previous chapters. Ecosystem ecologists are generally concerned with processes,
such as the rate of energy flow through a food web or the rate of nutrient cycling in a forest. Follow-
ing the recommendations of Ulanowicz (1997), these exchanges can be summarized in four simple

categories: (1) imports from the external environment; (2) internal transfers between compartments
(or species); (3) exports of usable materials of energy; and (4) dissipation of energy or conversion of
materials to its base form. Therefore, in its simplest form, ecosystem ecology attempts to quantify
energy and materials exchanges within these four categories. If the goal of ecosystem ecology is
to understand and quantify the flow of energy and materials through a system, our treatment of
ecosystem ecotoxicology will therefore be limited to the influence of contaminants and other anthro-
pogenic stressors on these processes. From a more practical perspective, a fundamental question is
whether alterations in the rate of energy and materials exchanges can serve as sensitive indicators of
anthropogenic perturbation.
Although ecosystems are widely regarded as fundamental units of ecology, effects of contamin-
ants on ecosystem processes have not received significant attention in the ecotoxicological literature
and are rarely considered within a regulatory framework. Despite the stated goal of maintaining both
structural and functional integrity of ecological systems, biological assessments in support of the
U.S. Clean Water Act emphasize population and community-level analyses. Other federal programs
within the United States, such as the Department of Interior Natural Resource Damage Assessment
program, rely almost exclusively on assessing responses at lower levels of organization (e.g., indi-
viduals and populations). Sheehan (1984) reviewed the effects of physical and chemical stressors on
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 627 — #17
Introduction to Ecosystem Ecology and Ecotoxicology 627
several ecosystem processes, including detritus processing, ecosystem metabolism, and food web
regulation. In light of the complexity of ecosystems and uncertainty in defining their spatiotemporal
boundaries, the focus on populations and communities in most ecotoxicological research is under-
standable. However, an ecosystem perspective is essential for predicting and understanding impacts
of chemicals and other stressors.
29.7 LINKS FROM COMMUNITY TO ECOSYSTEM
ECOTOXICOLOGY
Neither the process-functional approach nor the population–community approach can form a complete
theoretical foundation for ecosystem analysis.
(O’Neill et al. 1986)

29.7.1 ECOSYSTEMS WITHIN THE HIERARCHICAL CONTEXT
Ecology has many general theories but few have been fruitful in producing new insights.
(O’Neill et al. 1986)
Most ecologists readilyacknowledge the existence oflevels of organization within biological systems
and therefore accept the simple application of hierarchy theory to ecosystem ecology. The application
of hierarchy theory to the study of ecosystems may help resolve many of the issues inherent in the
study of complex systems. By decomposing complex systems into component parts, we can more
easily study mechanisms responsible for higher level, emergent processes. However, it is important
to recognize that, while this approach may be sufficient to study some ecological processes, such
as the influence of species interactions on energy flow, other processes will require a more holistic
analysis of emergent properties. Although hierarchy theory appears to be an appropriate model for
the study of ecosystems, significant progress has not been realized because of an overly simplistic
view of hierarchies and ecosystems (O’Neill et al. 1986). Hierarchy theory goes well beyond the
familiar level of organization concept described in many ecology textbooks. The levels within a
hierarchy are defined on the basis of differences in rate structures rather than an arbitrary assignment
to levels of organization.
Within ecological systems, hierarchies are often considered synonymous with levels of biological
organization. Within this hierarchy, ecosystems are located between communities and landscapes.
The critical distinction between communities and ecosystems is not simply a matter of size but
the relative emphasis on biotic and abiotic components. Population and community ecologists are
concerned with the interactions among individuals and species within an area. The abiotic envir-
onment is certainly recognized as being important and may influence these interactions, but it is
largely peripheral to the focus of the investigation. To the ecosystem ecologist, the organisms
and surrounding abiotic environment are integrated and constitute the unit of study. It is diffi-
cult to study the discrete components of an ecosystem in isolation. On the basis of hierarchy
theory, we would expect that ecosystems are composed of the next lower level (communities)
and are controlled by processes at the next higher level (landscapes). Because we have defined
ecosystems based on processing and flow of materials and energy, the hierarchical analysis is not
especially satisfactory because it assumes that these processes can be understood by studying the
populations and communities that comprise an ecosystem. O’Neill et al. (1986) warn against this

“naive reductionism” and note that, while ecosystems are indeed comprised of populations and
communities, these lower levels of organization are not necessarily the most appropriate for study-
ing ecosystem processes. Furthermore, most analyses of populations and communities either do
not consider abiotic factors, a critical component of ecosystems, or treat the abiotic environment
separately.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 628 — #18
628 Ecotoxicology: A Comprehensive Treatment
To place the ecosystem perspective in its proper context within a hierarchy of biological organ-
ization, consider the following example. Assume that a brown trout is feeding on aquatic insects in
a small mountain stream. A population ecologist would be especially interested in the consequences
of this predation event on demographic characteristics of either predators or prey. For example, does
the removal of prey species from a population contribute significantly to overall prey mortality?
Is survivorship or reproduction of the brown trout improved because of consumption of these prey
species? In contrast, a community ecologist would be more interested in the nature of the species
interactions and the implications for other species. For example, does the removal of individuals of
a particular prey species influence abundance and distribution of other species in the community?
Finally, the ecosystem ecologist is likely to consider this interaction strictly from a bioenergetics
perspective and focus on the transfer of calories (or other units of energy) between these two trophic
levels. Ecosystem ecologists are generally not concerned with the demographic consequences or the
strength of species interactions, but simply consider individual species as energy transformers.
O’Neill et al. (1986) contrast the population–community and process–functional approaches
to the study of ecosystems and discuss the consequences of viewing ecosystems from these very
different perspectives. The false dichotomy between structural and process-functional approaches to
the study of ecosystem science is similar to the historic controversy between the individualistic and
holistic viewpoints of Clements and Gleason on the nature of the community. Both perspectives were
narrowly focused on a set of observations and readily dismissed observations that were inconsistent
with these predisposed ideas. Although many ecologists believe otherwise, the relationship between
the structure and process of an ecosystem is not always obvious and in some situations may not
exist. This is at least partially owing to the fact that community and ecosystem ecologists use

different theories, vocabularies, and methodologies to study nature (Carney 1987). The redundancy
of functional processes andthefact that different assemblages of speciesarecapable of performing the
same function further complicates this relationship. The long list of ecosystem processes considered
critical to the assessment of ecosystem function further complicates the relationship between pattern
and process. Goldstein laments that
The list of possible descriptors of ecosystem processes and functions is limited neither by the myriad of
definitions of ecosystems nor by organismal considerations. Only the imagination of ecologists and the
shortcomings of language place a ceiling on the alleged number of ecosystem properties.
(Goldstein 1999)
Only recently, with the renewed interest in understanding how species diversity may influence
ecosystem function, have ecologists begun to clarify these relationships (Tilman et al. 1997). Despite
this progress, it is unlikely that we will ever be able to characterize the functioning of an ecosystem
based entirely on community structure and composition. Except in those rare instances where a
keystone species dominates a community and exerts strong control over ecosystem processes will
connections between pattern and process be clearly obvious. Splitting the traditional levels of organ-
ization and treating population–community and process-functional approaches as dual hierarchies
resolves many of the problems that ecologists have encountered when attempting to establish con-
nections between structural and functional components (O’Neill et al. 1986). According to this
model, the population–community hierarchy is regulated by organisms and species interactions
(e.g., competition, predation) whereas the process-functional hierarchy is regulated by mass balance
and principles of thermodynamics. More importantly, it is not necessary (or possible) to reduce one
dimension of the hierarchy to the other. As an example, Figure 29.7 contrasts the structural and
functional perspectives for the study of energy flow through a stream ecosystem. The structural per-
spective may consider questions related to energy flow, but most of the focus is on the importance of
the interactions among the groups. A functional perspective is less focused on effects of predators on
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 629 — #19
Introduction to Ecosystem Ecology and Ecotoxicology 629
Periphyton and
diatoms

Leaves and
detritus
Grazers
Shredders and
detritivores
Predators
Energy capture
Energy inputs
Structural emphasis Functional emphasis
Rate regulation
Consumers
Primary
producers
FIGURE 29.7 Contrastingstructuralandfunctional emphasis in anaquaticfoodchain.Thestructuralhierarchy
emphasizes population and community interactions whereas the functional hierarchy emphasizes ecosystem
processes and exchanges with the abiotic environment. (Modified from O’Neill et al. (1986).)
I
E
P
C
Ind
Eco
Pop
Comm
Individual
Ecosystem
Population
Community
Large scale
Intermediate

scale
Small
scale
FIGURE 29.8 An alternative depiction of individual, population, community, and ecosystem responses across
different spatial scales. Each level of organization is represented on each layer. Comparisons among levels of
the hierarchy can be made within a single spatial scale or between different spatial scales. This representation
allows investigators to consider how properties at one level of organization (e.g., ecosystem processes) and at
one spatial scale level are influenced by other levels of organization at different scales. (Modified from Allen
and Hoekstra (1992).)
prey populations and places more emphasis on the rate and regulation of energy inputs and capture
to the system.
Allen and Hoekstra (1992) provide a very different conceptual arrangement of the hierarchy of
biological organization. They suggest that separating an ecosystem to its component parts does not
reveal populations and communities, and use the analogy of a layer cake to illustrate connections
among levels of biological organization across differentspatialscales(Figure 29.8). Intheirdepiction,
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 630 — #20
630 Ecotoxicology: A Comprehensive Treatment
each level of organization is represented on each layer of the cake and comparisons among levels
of the hierarchy can be made within a single spatial scale or between different spatial scales. This
representation is intuitively appealing because it also allows us to consider how properties at one
level of organization (e.g., ecosystem processes) and at one spatial scale level are influenced by other
levels of organization at different scales. We should note that these modern conceptualizations of
ecosystems have been viewed skeptically by some ecologists. Ricklefs (1990) suggests that these
new approaches “may provide little more than a new set of jargon in the place of old conceptual
frustrations” and cautions against “legitimizing a mysticism that has no place in science.”
Regardless of which model we choose to illustrate where and how ecosystems fit within a hierarch-
ical framework, the fact remains that significant progresscan be achieved by integrated investigations
of ecosystem structure and function. From an applied perspective, the somewhat tenuous connection
between pattern and process may work to our advantage when designing monitoring systems to

assess ecological impairment. By including responses of both structural and functional endpoints,
we are clearly measuring different characteristics of an ecosystem. Arguments over which approach
is more fundamental to the ecosystem concept do not advance the discipline and may actually impede
progress. While recent developments in ecosystem science have significantly improved our ability
to assess effects of contaminants on mass flow through ecosystem compartments, the perception
remains that structural and functional approaches are clearly at odds (Cairns and Pratt 1986).
29.8 SUMMARY
In summary, measuring input and output of energy and materials in ecosystems with no consideration
of the species responsible for regulating these processes misses a fundamental component of the
system. Some understanding of the relative importance of different species to the movement of
energy and materials through an ecosystem will improve predictability and help understand variation
among ecosystems. Similarly, studying populations and communities in isolation from the abiotic
environment arbitrarily reducesthe importance ofabiotic factors. Furthermore, because of taxonomic
difficultieswithcertain groups of organisms (e.g., bacteria, nematodes), astructural perspective based
on abundance and species richness is not always possible. In the following chapters, we will adhere
to the conventional definition of an ecosystem that includes both biotic and abiotic components.
Although our focus will be on how contaminants affect ecosystem processes such as energy flow
and materials cycling, where our understanding of these processes is improved by information
on abundance and species richness we will also consider these structural elements. Although the
ecosystem ecologists may focus on the biogeochemical responses of an ecosystem to stressors, an
appreciation for the underlying physiological processes of individual organisms may enhance our
understanding of mechanisms.
In assessing the potential impacts of contaminants and other anthropogenic stressors on eco-
systems, we will distinguish between responses to localized disturbances, such as the discharge of
persistent chemicals to a lake, from the more pervasive effects of global stressors, such as increasing
atmospheric CO
2
levels. Although both disturbances may similarly influence ecosystem processes,
the consequences and interpretations of responses are quite different. In the former example, ecosys-
tems are reservoirs of these chemicals and the transformations within compartments are of primary

concern. In the second example, ecosystems are not only directly affected by increasing CO
2
levels,
but may also directly influence global atmospheric concentrations by sequestration in various com-
partments. Pace and Groffman (1998) characterize effects of the localized and widespread stressors
as footprints and fingerprints. Ecological footprints reflect the direct effects of localized stressors
and are a consequence of human appropriation of natural resources or discharge of toxic substances.
In contrast, the ecological fingerprints associated with global and atmospheric stressors are likely
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 631 — #21
Introduction to Ecosystem Ecology and Ecotoxicology 631
to be subtle and more difficult to detect. In the following chapters, we will first describe import-
ant ecosystem-level processes and then discuss how these processes are likely to be affected by
ecological footprints and fingerprints.
29.8.1 SUMMARY OF FOUNDATION CONCEPTS AND PARADIGMS
• The goal of ecosystem ecology is to understand and quantify the flow of energy and mater-
ials through a system. Ecosystem ecotoxicology will assess the influence of contaminants
and other anthropogenic stressors on these processes.
• One distinguishing featureof ecosystem ecology, whichwas recognized early inits history,
was the necessity of considering integrated physical, chemical, and biological processes.
• Although the flow of energy and the transport of materials through an ecosystem are gen-
erally treated separately in most ecosystem assessments, these processes are so intimately
linked it is often more practical to consider them simultaneously.
• The publication of Eugene P. Odum’s (1953)classic text, Fundamentals ofEcology, placed
ecosystem studies in the mainstream of ecological research and allowed the ecosystem
concept to emerge as a legitimate topic.
• Ecosystem-level studies conducted in systems with relatively simple food webs demon-
strated that the movement of energy through an ecosystem could be quantified; however,
quantifying energetics of more complex systems proved to be a daunting task.
• Experimental techniques such as the addition of radioactive tracers (e.g.,

32
P) improved
our ability to quantify the movement of energy and materials through ecosystems.
• Because quantifying the movement of energy and materials through an ecosystem gener-
ally required a mass budget approach in which inputs and outputs were measured, lakes
and streams became appropriate models because the boundaries were well defined.
• Our ability to understand effects of contaminants on ecosystems is both facilitated
and impeded by the fact that ecosystem processes are controlled by negative feedback
relationships.
• A better understanding of fate and transformation of contaminants is considered a major
accomplishment in the field of applied ecosystem science.
• Some combination of modeling, comparative, long-term monitoring and experimental
approaches will be required to address applied questions in ecosystem ecotoxicology.
• Despite the stated regulatory goal of maintaining both structural and functional integrity
of ecological systems, biological assessments in support of the Clean Water Act rarely
assess effects of contaminants on ecosystem processes.
• The critical distinction between communities and ecosystems is not simply a matter of
size but the relative emphasis on biotic and abiotic components. While population and
community ecologists are concerned primarily with the interactions among individuals
and species within an area, ecosystem ecologists consider these interactions within the
context of the surrounding abiotic environment.
REFERENCES
Allen, T.F.H. and Hoekstra, T.W., Toward a Unified Ecology, Columbia University Press, New York, 1992.
Allen, T.F.H. and Starr, T.B., Hierarchy: Perspectives for Ecological Complexity, University of Chicago Press,
Chicago, 1982.
Ball, R.C. and Hooper, F.F., Translocation of phosphorous in a trout stream ecosystem, In Radioecology,
Schultz, V. and Klement, W. (eds.), Reinhold, New York, 1963, pp. 217–228.
Baron, J.S., Rueth, H.M., Wolfe, A.M., Nydick, K.R., Allstott, E.J., Minear, J.T., and Moraska, B., Ecosystem
responses to nitrogen deposition in the Colorado Front Range, Ecosystems, 3, 352–368, 2000.
© 2008 by Taylor & Francis Group, LLC

Clements: “3357_c029” — 2007/11/9 — 18:37 — page 632 — #22
632 Ecotoxicology: A Comprehensive Treatment
Bormann, F.H. and Likens, G.E., Nutrient cycling, Science, 155, 424–429, 1967.
Cairns, J., Jr. and Pratt, J.R., On the relation between structural and functional analyses of ecosystems, Environ.
Toxicol. Chem., 5, 785–786, 1986.
Carney, H.J., On competition and the integration of population, community, and ecosystem studies, Funct.
Ecol., 3, 637–641, 1987.
Carpenter, S.R., Replication and treatment strength in whole-lake experiments, Ecology, 70, 453–463, 1989.
Carpenter, S.R. and Turner, M.G., At last a journal devoted to ecosystem science, Ecosystems, 1, 1–5, 1998.
Clements, F.E., Plant Succession: Analysis of the Development of Vegetation, Carnegie of Washington
Publication, No. 242, Washington, D.C., 1916.
Colinvaux, P., Why Big Fierce Animals are Rare. An Ecologists Perspective, University Press, Princeton, NJ,
1978.
Cowles, H.C., The ecological relations of the vegetation on the sand dunes of Lake Michigan, Bot. Gaz., 27,
95–117, 167–202, 281–308, 361–391, 1899.
Fisher, S.G. and Likens, G.E., Energy flow in Bear Brook, New Hampshire: An integrative approach to stream
ecosystem metabolism, Ecol. Monogr., 43, 421–439, 1973.
Forbes, S.A., The lake as a microcosm, III. Nat. Hist. Surv. Bull., 15, 537–550, 1887.
Goldstein, P.Z., Functional ecosystems and biodiversity buzzwords, Con. Bio., 13, 247–255, 1999.
Golley, F.B., Energy dynamics of a food chain of an old field community, Ecol. Monogr., 30, 187–206, 1960.
Golley, F.B., A History of the Ecosystem Concept in Ecology, Yale University Press, New Haven, 1993.
Gosz, J.R., Ecology challenged? Who? Why? Where is this headed?, Ecosystems, 2, 475–478, 1999.
Holling, C.S. andAllen, C.R., Adaptive inference for distinguishing credible from incredible patterns in nature,
Ecosystems, 5, 319–328, 2002.
Karr, J.R., Defining and assessing ecological integrity: Beyond water quality, Env. Toxicol. Chem., 12,
1521–1531, 1993.
Likens, G.E., The Ecosystem Approach: Its Use and Abuse, Oldendorf/Luhe, Germany, Ecology Institute, 1992.
Likens, G.E., Driscoll, C.T., and Buso, D.C., Long-term effects of acid rain: Response and recovery of a forest
ecosystem, Science, 272, 244–246, 1996.
Lindeman, R.L., The trophic-dynamic aspect of ecology, Ecology, 23, 399–418, 1942.

Maciorowski, A.F., Populations and communities: Linking toxicology and ecology in a new synthesis,
Env. Toxicol. Chem., 7, 677–678, 1988.
Naiman, R.J., A perspective on interdisciplinary science, Ecosystems, 2, 292–295, 1999.
O’Neill, R.V., DeAngelis, D.L.,Waide, J.B., andAllen, T.F.H., A Hierarchical Concept of Ecosystems, Princeton
University Press, Princeton, NJ, 1986.
Odum, E.P., Fundamentals of Ecology, Saunders, Philadelphia, 1953.
Odum, E.P., The strategy of ecosystem development, Science, 164, 262–270, 1969.
Odum, H.T., Trophic structure and productivity of Silver Springs, Florida, Ecol. Monogr., 27, 55–112, 1957.
Pace, M.L. and Groffman, P.M., Successes, limitations, and frontiers in ecosystem science: Reflections on the
seventh Cary Conference, Ecosystems, 1, 137–142, 1998.
Pickett, S.T.A., Burch, W.R., and Grove, J.M., Interdisciplinary research: Maintaining the constructive impulse
in a culture of criticism, Ecosystems, 2, 302–307, 1999.
Platt, J.R., Strong inference, Science, 146, 347–353, 1964.
Popper, K.R., The Logic of Scientific Discovery, 3rd ed., Hutchinson, London, England, 1972.
Reiners, W.A., Complementary models for ecosystems, Am. Nat., 127, 59–73, 1986.
Ricklefs, R.E., Ecology, 3rd ed., W.H. Freeman and Company, New York, 1990.
Sagoff, M., The plaza and the pendulum: Two concepts of ecological science, Biol. Phil., 18, 529–552, 2003.
Sarkar, S., Ecological theory and anuran declines, BioScience, 46, 199–207, 1996.
Schindler, D.W., Detecting ecosystem responses to anthropogenic stress, Can. J. Fish. Aquat. Sci. 44 (Suppl. 1),
6–25, 1987.
Sheehan, P.J., Functional changes in the ecosystem, In Effects of Pollutants at the Ecosystem Level, Shee-
han, P.J., Miller, D.R., Butler, G.C., and Bourdeau, P. (eds.), John Wiley & Sons, Chichester,
1984, pp. 101–145.
Slobodkin, L.B., Growth and Regulation of Animal, Holt, Rinehart and Winston, New York, 1961.
Stiling, P., Ecology: Theories and Application, 3rd ed., Prentice Hall, Upper Saddle River, NJ, 1999.
Suter, G.W., II, Ecological Risk Assessment, Lewis Publishers, Chelsea, MI, 1993.
Tansley, A.G., The use and abuse of vegetational concepts and terms, Ecology, 16, 284–307, 1935.
© 2008 by Taylor & Francis Group, LLC
Clements: “3357_c029” — 2007/11/9 — 18:37 — page 633 — #23
Introduction to Ecosystem Ecology and Ecotoxicology 633

Tilman, D., Ecological experimentation: Strengths and conceptual problems, In Long-Term Studies in Ecology,
Likens, G.E. (ed.), Springer-Verlag, New York, 1989, pp. 136–157.
Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M., and Siemann, E., The influence of functional diversity
and composition on ecosystem processes, Science, 277, 1300–1302, 1997.
Ulanowicz, R.E., Ecology, The Ascendant Perspective, Columbia University Press, New York, 1997.
Whittaker, R.H., Experiments with radiophosphorus tracer in aquarium microcosms, Ecol. Monogr., 31,
157–188, 1961.
© 2008 by Taylor & Francis Group, LLC