Published in Futures Vol 31 #7 pp.

721-742

AN ECOSYSTEM APPROACH FOR SUSTAINABILITY: ADDRESSING THE

CHALLENGE OF COMPLEXITY.

James J. Kay Henry A. Regier

Environment & Resource Studies Institute for Environmental Studies
University of Waterloo University of Toronto

Michelle Boyle George Francis

Network for Ecosystem Environment & Resource Studies
Sustainability and Health University of Waterloo

The dynamics of ecosystems and human systems need to be addressed in the context
of Post-Normal Science grounded in complex systems thinking. We portray these
systems as Self-Organizing Holarchic Open (SOHO) systems and interpret their
behaviours and structures with reference to non-equilibrium thermodynamics; holons,
propensities, and canons; and information and attractors. Given the phenomena
exhibited by SOHO systems, conventional science approaches to modelling and
forecasting are inappropriate, as are prevailing explanations in terms of linear causality
and stochastic properties. Instead, narratives in the form of scenarios to depict
morphogenetic causal loops, autocatalysis, and multiple possible pathways for
development need to be considered. Short examples are given. We also link SOHO
system descriptions to issues of human preferences and choices concerning the
preferred attributes of particular SOHO systems, and to the implications for achieving
them through adaptive management, monitoring and appropriate structures for
governance. A heuristic framework to guide reasoning for this is presented, and
reiterative steps for applying it are identified. In this way we provide a coherent
conceptual basis, in the workings of both natural systems and decision systems, for the
practice of Post-Normal Science.

1. Introduction
Much discussion about ecosystem management, or taking an ecosystem approach
emphasizes the need to work across all manner of human boundaries at different
geographic scales. However the growing understanding of the dynamics of ecosystems
entails much more [1]. In this paper we suggest an approach to understanding these
systems in the context of “Post-Normal Science” grounded in complex systems thinking.
At its heart is the portrayal of ecological and human systems as Self-Organizing
Holarchic Open (SOHO) systems whose dynamics are predominated by both positive
and negative feedback processes operating over a range of spatial and temporal
scales. These systems exhibit loose hierarchical structures, various emergent
phenomena, and relatively sudden reconfigurations from one state of system
organization to another. Some changes in these systems are inherently unpredictable.
 Futures Vol 31, #7, Sept. 99, pp.721-742 2

The understanding of SOHO systems requires a major change in some of the ways in
which science and decision making are conducted. Traditional reductionistic
disciplinary science and expert predictions, the basis for much of the advice given to
decision makers, have limited applicability. Narratives about possible futures for given
SOHO systems are better able to capture the richness of possibilities. Other
epistemological “mindsets” or causal metatypes [2] must be brought to bear, notably
explanations based on morphogenetic causal loops that involve both positive and
negative feedback processes and autocatalysis. Expectations that decision makers can
carefully control or manage changes in societal or ecological systems have also to be
challenged. Adaptive learning and adjustment, guided by a much wider range of human
experience and understanding than disciplinary science, are also necessary.
This paper sketches a theoretical approach for understanding the dynamics of SOHO
systems. It also situates this approach within an extensive heuristic framework for
relating human choices and preferences concerning the attributes of SOHO systems to
the implications for achieving them through adaptive management, monitoring, and
governance. The main focus of this paper is on the features of SOHO systems that
require different modes of human responses, and also constitute the initial steps for
using the heuristic framework.

2. Self-Organizing Holarchic Open (SOHO) Systems

Complex systems thinking follows in the tradition of von Bertalanffy's general systems
theory, and draws upon other concepts from the “new science” emerging over the past
three decades, for example, catastrophe theory, chaos and complexity theory, non-
equilibrium thermodynamics and self-organization, and Jaynesian information theory.
The phenomenon of special interest is self-organization.[3]

The Emergence of SOHO systems

Spontaneous coherent behaviour and organization occurs in open systems (such as
ecosystems and human systems). The key to understanding such phenomena is to
realize that open systems are processing an enduring flow of high quality energy
(exergy). In these circumstances, coherent behaviour appears in systems for varying
periods of time but can change suddenly whenever the system reaches a catastrophe
threshold, [4] and “flips” into a new coherent behavioural state. [5] (A simple example is
the vortex which spontaneously appears in water from draining a bathtub, or more
dramatically, the appearance of tornadoes “from nowhere”).
Kay and Schneider [6] examined the energetics of open systems and have taken
Prigogine's work one step further. An open system with exergy (high quality energy)
pumped into it is moved away from equilibrium, but nature resists movement away from
equilibrium. This is the second law of thermodynamics restated for non-equilibrium
situations. When the input of high quality energy and material pushes the system
beyond a critical distance from equilibrium, the open system responds with the
spontaneous emergence of new, reconfigured organized behaviour that uses the high
quality energy to build, organize and maintain its new structure. This reduces the ability
of the high quality energy to move the system further away from equilibrium. As more

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 3

high quality energy is pumped into a system, more organization emerges, in a step-wise
way, to dissipate the exergy. Furthermore, these systems tend to get better and better
at "grabbing" resources and utilizing them to build more structure, thus enhancing their
dissipating capability. There is however, in principle, an upper limit to this
organizational response. Beyond a critical distance from equilibrium, the organizational
capacity of the system is overwhelmed and the system's behaviour leaves the domain
of self-organization and becomes chaotic. As noted by Ulanowicz [7] there is a
"window of vitality", that is a minimum and maximum level in between which self-
organization can occur.
The theory of non-equilibrium thermodynamics suggests that the self-organization
process in SOHO systems proceeds in a way that captures increasing resources
(exergy and material); makes ever more effective use of the resources; builds more
structure; and enhances survivability. [8] These seem to be the kernel of the
propensities of self-organization. This conception of self-organization, as a dissipative
system, is presented in Figure 1.
How these propensities manifest themselves as morphogenetic causal loops and
dissipative processes is a function of the given environment (context) in which the
system is imbedded, as well as the available materials, exergy and “information”, the
latter defined as factors embedded internally within the system that constrain and guide
the self-organization. The interplay of these factors defines the context and associated
constraints on the set of processes which may emerge. Generally speaking, which
specific processes emerge from the potential set are uncertain.
Self-organizing dissipative processes emerge whenever sufficient exergy is available to
support them. Once a dissipative process emerges and becomes established it
manifests itself as a structure.[9] These structures provide a new context, nested within
which new processes can emerge, which in turn beget new structures, nested within
which... Thus emerges a SOHO system, a nested constellation of self-organizing
dissipative process/structures organized about a particular set of sources of exergy,
materials, and information, embedded in a physical environment, that give rise to
coherent self-perpetuating behaviours. [10]

Holons, Propensities and Canons

Koestler [11] perceived the unit/element features of systems, and posed the term
“holon” (a term derived etymologically from whole/part) as an entity that exists
contextually in a nested network of holons referred to as a “holarchy”. A holarchy is a
generalized version of a traditional hierarchy (not to be confused with Allen's notion of
hierarchy [12] ) with reciprocal power relationships between levels rather than a
preponderance of power exerted from the top downwards. A holon of particular interest
for an observer occurs in some holarchic relationships with mutual causality guiding
reciprocal interactions between a holon and proximate contiguous holons of different
scales –- inside, outside and lateral to the holon of interest.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 4

Figure 1: A conceptual model for self-organizing systems as dissipative structures. Self-organizing

dissipative processes emerge whenever sufficient exergy is available to support them. Dissipative
processes restructure the available raw materials in order to dissipate the exergy. Through catalyse,
the information present enables and promotes some processes to the disadvantage of others . The
physical environment will favour certain processes. The interplay of these factors defines the context
for (i.e. constrains) the set of processes which may emerge. Once a dissipative process emerges
and becomes established it manifests itself as a structure. These structures provide a new context,
nested within which new processes can emerge, which in turn beget new structures, nested within
which... Thus emerges a SOHO system, a nested constellation of self-organizing dissipative
process/structures organized about a particular set of sources of exergy, materials, and information,
embedded in a physical environment. The canon of the SOHO system is the complex nested
interplay and relationships of the processes and structures, and their propensities, that give rise to
coherent self-perpetuating behaviours, that define the attractor

SELF-ORGANIZING SYSTEM

DISSIPATIVE
CONTEXT

∞
PROCESS
Physical Environment

Flows:
Exergy (energy)
Material
Information DISSIPATIVE
STRUCTURE

Ulanowicz [13] developed further some proposals by K. R. Popper [14] to extend a

perception of indeterminacy in the quantum realm to other scales of phenomena by
generalizing the usual Newtonian concept of force to obtain a notion of systemic
dynamical cohesion which Popper called a “propensity”. A propensity is always
contextual (as are Koestler’s holons) rather than universalistic as in the Newtonian
sense of “force”. Ulanowicz proposed that a mutual-causal kind of autocatalysis plays a
self-organizing role in Popper’s propensity, perhaps in generating dynamical cohesion
through forces that act asymmetrically, and not symmetrically as in the Newtonian
sense. Ulanowicz also implied that “dynamical cohesion” tends to be attenuated in a
step-wise manner at the interfaces of interacting holons, both with respect to nested and

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 5

non-nested kinds of relationships and that these attenuations may be perceived as

boundary-like.
For our purposes, a central question to be addressed by the narrative description of a
SOHO system is an elaboration of its propensities. The elaboration delineates the
mutual causality of the feedback loops and autocatalytic process which give the system
its coherence as an entity. We refer to this set of propensities, which characterize a
holon, as its "canon".

Attractors
A SOHO system exhibits a set of behaviours which are coherent and organized, within
limits. The nexus of this organization at any given time is referred to as an attractor.
The term "attractor" comes from the state space description of the behaviour. The
system has a propensity to remain in a limited domain of state space (for example a
gravity well). It behaves as if it were "attracted" toward this domain and hence the term
"attractor". As SOHO systems evolve they shift between attractors within the SOHO
system's overall state space. The re-organization that these shifts entail is not smooth
and continuous but rather is step-wise. The system “flips” its organizational state in
often dramatic ways.
Ecosystems have multiple possible operating states or attractors, and may shift or
diverge suddenly from any one of them. The notion of alternate stable states in
ecosystems is not well known in the ecological, and particularly the resource
management communities, but it is also not new [15]. Yet the importance of this notion
for explaining ecosystem phenomena remains largely unexplored.
Human systems exhibit similar phenomena of self-organization, such as economic
markets, communication networks, and urban expansion or rural contraction. Self-
organization does not deny human agency, but suggests only that the collective striving
of many individuals and organizations can lead to larger scale structures which are both
unplanned and unexpected. The striving itself may be directed towards realizing
aspirations (an “attractor”?) and reducing the gradients between perceived conditions
and desired goals. Human systems also exhibit the dynamics of SOHO systems.[16]

3. Science and Decision Making in the Context of Complexity

Some facets of complexity have been highlighted in this discussion. The first is that
self-organization inherently involves internal causality. A self-organizing system has the
ability to maintain itself at an attractor despite changes in its environment. It is possible
for a system’s environment to change substantially, without the system exhibiting major
change. Self-organizing systems can respond in a synergistic way to multiple
environmental factors such that changes in the system cannot be tied categorically to
specific environmental factors. Furthermore, such systems have the capability to
generate new behaviour which may emerge independently of changes in the
environment.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 6

Table 1: Properties of complex systems to bear in mind when thinking about SOHO
systems.

•NON-LINEAR: Behave as a whole, a system. Cannot be understood by simply

decomposing into pieces which are added or multiplied together.

•HIERARCHICAL: Are holarchically nested. The system is nested within a system

and is made up of systems. The "control" exercised by a holon of a specific level
always involves a balance of internal or self-control and external, shared,
reciprocating controls involving other holons in a mutual causal way that transcends
the old selfish-altruistic polarizing designations. Such nestings cannot be
understood by focusing on one hierarchical level (holon) alone. Understanding
comes from multiple perspectives of different type and scale.

•INTERNAL CAUSALITY: non-Newtonian, not a mechanism, but rather is self-

organizing. Characterized by: goals, positive and negative feedback, autocatalysis,
emergent properties and surprise.

•WINDOW OF VITALITY: Must have enough complexity but not too much. There is a
range within which self-organization can occur. Complex systems strive for
optimum, not minimum or maximum.

•DYNAMICALLY STABLE?: There may not exist equilibrium points for the system.

•MULTIPLE STEADY STATES: There is not necessarily a unique preferred system

state in a given situation. Multiple attractors can be possible in a given situation and
the current system state may be as much a function of historical accidents as
anything else.

•CATASTROPHIC BEHAVIOUR: The norm

Bifurcations: moments of unpredictable behaviour
Flips: sudden discontinuities, rapid change
Holling four box cycle Shifting steady state mosaic

•CHAOTIC BEHAVIOUR: our ability to forecast and predict is always limited, for
example to between five and ten days for weather forecasts, regardless of how
sophisticated our computers are and how much information we have.

This capacity to organize and maintain itself about an attractor is the hallmark of a self-
organizing system. Such behaviour requires causal explanation in terms of synergistic
positive and negative feedback loops. The explanation must also take into account
different scales and types of interactions. An adequate description of self-organizing
behaviour must be in terms of holarchical, multi-observer/disciplinary perspectives and
morphogenetic mutual causality. Linear cause and effect explanations, in principle,

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 7

cannot provide the kind of causal explanation required to describe self-organizing

behaviour. In addition to the complexity associated with the issue of causality and self-
organization, there is the possibility of more than one appropriate self-organizing
response (i.e. multiple attractors). There is not necessarily a unique preferred state to
be deduced from scientific arguments. Furthermore, system responses are a function of
their particular histories or circumstances of the moment. Thus there is an element of
irreducible uncertainty about self-organizing behaviour, uncertainty about what may
come to pass as well as uncertainty about what ought to come to pass.
Table 1 summarizes the properties of SOHO systems and notes features which have to
be addressed for understanding them. The nature of SOHO systems requires that they
be studied from different types of perspectives, each emphasizing certain processes
and structures such as the geophysical (abiotic), ecological (biotic) and human cultural
components, and at different scales. There is no unique correct perspective. Rather, a
diversity of perspectives is required for understanding, and the relative emphasis placed
on some perspectives over others depends on the purpose of the inquiry.
In situations dominated by self-organizing behaviour, the properties of inherent
uncertainty and emergence limit the capacity to predict how the situation will unfold. In
principle, it will not be possible in many situations to construct accurate quantitative
models which forecast the future to the degree required for anticipatory management.
Anticipatory management is based on the premise that it is possible to predict and
anticipate the consequences of decisions and hence to make a proper decision once all
the necessary information is gathered to make a scientific forecast. But anticipatory
management is a central doctrine of the "normal" orthodox science for decision making
and is, in principle, not sufficient to deal with the complexity associated with the
dynamics of SOHO systems.

Other epistemological traditions are required for understanding SOHO systems. For
example, Maruyama [17] identified four causal metatypes that, as related to scientific
substance and process are:
H: explanation in terms of cause and effect models, where the cause and effect
relationship can be deterministic or probabilistic and is often linear in nature. No
reciprocal relationship between effect and cause is considered and hence there
are no causal loops in this mode of explanation. For example- sequential cause
and effect analysis in the old engineering science mode used safety factors in the
design of works to cover features of reality that do not conform to this mindscape.
Similarly, it was assumed that incremental decreases in anthropogenic impacts
on environments will result in incremental improvements in environmental quality.
I: explanation in terms of independent event models in which the most probable
states of an isolated system are random distributions of independent events, and
non-random features tend to decay. From this perspective any systemic linkages
within an ecosystem are only opportunistic and temporary, hence any synergism
among the effects of different human uses is unlikely and the term “ecosystem”
has little meaning;

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 8

S: explanation in terms of homeostatic causal loop models, that is, negative

feedback that maintains steady-state. The cause is controlled by the effect.
Causality may be probabilistic or deterministic. Homeostatic causal loops can
maintain organized heterogeneity in a system. This form of explanation
underlies the old "Clementsian" view of ecological systems in which succession
to a climax community led early preservation naturalists to hope that an
ecosystem could remain in perpetuity in a particular preferred state; and
G: explanation in terms of morphogenetic causal models, that is explanations that
involve both positive and negative feedback loops and autocatalysis. These
probabilistic or deterministic loops of mutual causality can increase the pattern of
heterogeneity towards higher levels of organized complexity. For example,
consider the task of ecosystem regeneration from an unacceptably altered
ecological system (say a eutrophic state in a lake) into a more desirable (for
example an oligotrophic state in a lake) but not fully predictable state with strong
self-organizing and self-sustaining capabilities, with constrained autocatalysis.
This would be accomplished by changing a number of contextual elements (for
example phosphorous loading) beyond critical thresholds so as to precipitate a
flip in ecosystem state.
The understanding of SOHO systems draws upon all four mindscapes, especially G.

4. Preparing the Narratives

In the post-normal paradigm, a scientist's role in decision making shifts from inferring
what will happen, that is, making predictions which are the basis of decisions, to
providing decisions makers and the community with an appreciation, through narrative
descriptions, of how the future might unfold. As noted earlier, these narratives consist
of several scenarios of how the SOHO systems in question might evolve. These
narratives focus on a qualitative/quantitative understanding that describes:
• the human context for the narrative
• the hierarchical nature of the system;
• the attractors which may be accessible to the system;
• how the system behaves in the neighbourhood of each attractor, potentially in terms
of a quantitative simulation model;
• the positive and negative feedbacks and autocatalytic loops and associated
gradients which organize the system about an attractor;
• what might enable and disable these loops and hence might promote or discourage
the system from being in the neighbourhood of an attractor; and
• what might be likely to precipitate flips between attractors.
These narratives are in the service of informing decision makers and the community
about:
• possible future states of organization of the system;
• understanding of conditions under which these states might occur;
• understanding of the tradeoffs which the different states represent;
• appropriate schemes for ensuring the ability to adapt to different situations;

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 9

• and perhaps most importantly the appropriate level of confidence that the narrative
deserves, that is our degree of uncertainty.
Having sketched a picture of the possibilities in the future, it remains for scientists to
suggest ways of mitigating and adapting to the inevitable surprises, both surprises in the
form of unexpected flips to known attractors and those that involve flips to new
attractors which correspond to heretofore unknown manifestations of system
organization. Table 2 summarizes the analysis of ecosystems as SOHO systems.

Table 2: Analysis of ecosystems as Self-organizing Holarchic Open Systems : The

term ecosystem here is used in its broadest sense and refers to natural as well as
human constructs, such as economic systems. Note that while the activities of
ecosystem analysis are written out below in a linear way, the analysis itself is not linear.
Thus the activities will be accomplished in different sequences depending on the case
at hand.

A. Define the ecosystem

Type:
What perspectives will be used to look at the system? (abiotic, biotic, human
cultural...)

Scale and Extent (the horizontal perspective, where do things begin and end?)
What are the boundaries of observation?
What are the processes which define the whole?
What are the boundaries of the ecosystem, the holon of focus?

Holarchy: (The vertical perspective, what is a part of what?)

Define the Nested Holons (nested systems); this defines the contextual
relationships.

Holarchic Structure
Delineate the vertical and horizontal relationships between holons

B. Describe the ecosystem as a self-organizing entity

The attractors (organizational states) and their domains:
What are the attractors?
In what direction will the ecosystem tend to develop? What are its propensities?
(Self-organization theory of dissipative structures helps answer this.)
What is the behaviour of the ecosystem about the attractors?
(Homeostatic, Stable, Figure ∞, Unstable but persists, chaotic?)
Are there bifurcation points?
What are the potential flips between attractors?
What triggers the flips?
How can we monitor for them?
The context:

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 10

What is the interplay of energy, exergy, available materials, information and

environmental conditions (in space and time) which shapes the ecosystem?
Morphogenetic casual models
The synergistic relationships, the cycles, the feedback loops, virtual worlds.
The canon:
Think carefully about the dissipative processes and structures, their interplay (e.g.
Figure ∞), their scale and extent, the nested holons and their interactions and
connections, the information available to the ecosystem, and the environmental
conditions it is set in.
(Ecological history and non-equilibrium thermodynamics help answer this)

C. How do we evaluate Integrity for this ecosystem?

(What states of ecosystem organization are acceptable to us?)
What are the ecological, economic and other processes (at each of the nested levels)
we value and/or need?
How do we identify these?
How do we measure the status of these processes?
(Notice that this takes us back to step A above)

Which attractors represent unacceptable ecosystem conditions?

D. Is this integrity threatened?

What are the external influences which could effect the organizational status of the
system?

What are the thresholds of flips to the unacceptable attractors? (states of ecosystem
organization)

E. How do we maintain integrity in this system?

How do we mitigate known threats (that is contextual changes which promote undesired
attractors)?

How do we promote positive influences? (For example; fire in a prairie, subsidies for
clean technologies)

How do we monitor the ecosystem so as to detect changes due to previously

unidentified external influences?

F. How to deal with Emergent Complexity..........

When all is said and done, our ability to predict is severely limited. Unexpected events
and trends will occur. Surprise will happen, complexity will emerge. We must therefore
rely on anticipatory and adaptive management.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 11

Always remember "The system imbedded in another system imbedded in another

system imbedded in another system ........." and the challenge of sustaining a dynamic,
changing, evolving, self-organizing ecosystem.

Narratives I. Systems and Issues of Interest

The narrative to describe a SOHO system must begin with identification of the relevant
set of constituent holons, the self-organizing entities of interest. Consideration must be
given to the types of perspectives required and the appropriate scales of investigation
[18]. The SOHO systems of interest are critically related to the human issues to be
addressed. Care must be taken to identify the issues at hand and the appropriate
perspectives and scales of investigation necessary to deal with these issues in the
SOHO systems context. This identification process can only occur in the context of
human values and requires bringing a diversity of views to bear on the question at hand.
This in turn, often depends upon who is invited to the consultation or negotiation
sessions to design the collaborative inquiry, and the values and concerns each
stakeholder brings to the table. Issues of inclusiveness, relative power among
stakeholders, and extent of mutual trust in the process are all relevant .
One example of regional development in a rather isolated region in northern Peru
illustrates this. When the issues of sustainability of food production and nutrition for rural
residents were investigated by external researchers, the focus of attention was on cattle
ranching and range management. When a more inclusive consultation was held locally,
it became apparent that fish, not cattle, are the important food source. This shifted the
research focus from cattle ranching and range management to the condition of riverine
fisheries. [19]
Interest in maintaining “natural” ecosystem conditions in a small site at the edge of
Kitchener, Ontario, for environmental education purposes, posed questions about the
conditions and desired compositional changes in forest stands and in a stream-pond
system flowing through the site. However, both are driven by subsurface
hydrogeological systems which could not be managed directly. The hydrogeological
systems were affected, however, by land use changes, through urbanization, occurring
away from the site of interest, and involving landowners not involved with concerns
about the impacts on the site of interest for educational purposes . Thus a critical
system type or perspective, that of land ownership patterns and land uses in the region
of interest, needed to be considered in addition to the conventional biological analysis of
the site.[20]

Narratives II: Exergy, canons, and attractors

The task of characterizing SOHO systems is to describe how these systems unfold over
time. The description is a narrative, literally a story, that is qualitative with multiple
threads of explanation, portraying a number of possible pathways for development (or
storylines). The narratives characterize the attractors and canon of SOHO systems in

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 12

terms of how the local context of exergy, materials and information and biophysical
environment, and the global propensities of capturing more resources (exergy and
material), making more effective use of the resources; building more structure, and
enhancing survivability, give rise to the emergence of the nested structures and
processes which constitute a self-organizing holarchic open system. Two examples
summarized below demonstrate the role of the dissipative system model and
thermodynamics in generating a narrative description of ecological systems as SOHO
systems.
Regier and Kay [21] provide an example from Lake Erie which proposed a two-attractor
catastrophe cusp model (Figure 2) as a way of integrating much empirical information of
how aquatic systems might transform under powerful, careless human interventions.
Two different attractors for shallow lakes have been identified [22]. In the
oligotrophic/benthic state, a high water clarity bottom vegetation ecosystem exists. As
nutrient loading results in increasing density of planktonic turbidity in the water, the
internal state of the adapting ecosystem eventually hits a catastrophe threshold and the
ecosystem then flips into an eutrophic/pelagic state. Lake ecosystems have been found
which may be perceived to flip between these attractors irregularly. (Lake Erie appears
to be currently in the midst of such a flip, from pelagic to benthic.) At least three quite
different descriptions of such a lake will be needed, one for the pelagic state, one for the
benthic, and one for the intermediate stage as the system flips between attractors. An
ecosystem adapts through a shift in phase to major fluctuations in factors at a larger
scale in the overall regional holarchy.
The essence of the canon of the benthic system is that it depends on solar energy
reaching the bottom, for the exergy necessary to energize the system. The solar exergy
is captured by the green matter on the bottom and is transformed into forms appropriate
to power the benthic processes. These include predation and grazing of the pelagic
system, thus suppressing it. Various means emerge to maintain the ecosystem at the
benthic attractor. Notable among these are means for keeping the water clear so solar
energy will reach the bottom and means for keeping the water column free of sufficient
exergy which would empower the pelagic attractor.
The pelagic system, on the other hand, depends on exergy in the water column to
energize it. Solar energy may be in the water column. However, unless the materials
necessary for the existence of dissipative processes, which can utilize the solar energy,
are present in the water column, nothing can be done with the solar energy, so it has no
exergy. For example, in many lakes, available phosphorus in the water column limits
the level of photosynthesis by phytoplankton. Beyond a critical level of available
phosphorus in the water column, there is enough availability of solar energy (i.e.
sunlight exergy) to support the phytoplankton bloom necessary for the activation of the
pelagic attractor. Once this occurs, the solar energy capture happens nearer the
water's surface instead of at the bottom and means emerge for promoting and
maintaining the pelagic attractor. Of course by its very presence the pelagic system
shades the benthic from irradiation by the sun, thus decreasing the exergy at the
bottom.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 13

Figure 2: Benthic and Pelagic Attractors in shallow lakes

Two different attractors for shallow lakes have been identified. In the benthic state, a high water clarity
bottom vegetation ecosystem exists. As nutrient loading increases the turbidity in the water, the
ecosystem hits a catastrophe threshold and flips into a hypertrophic, turbid, phytoplankton pelagic
ecosystem. The relationship of these two attractors, from a thermodynamic perspective, is as follows:
Let us assume that the benthic attractor is dominant and that the rate at which phosphorus is being
added to the water is increasing. The benthic system has means of deactivating phosphorus.
However the amount of active phosphorus will increase, albeit slowly, effectively increasing the
exergy in the water column. As this exergy increases a critical threshold is passed which allows
the pelagic system to self-organize to coherence. Once this occurs the exergy at bottom
decreases rapidly due to shading (turbidity) thus catastrophically de-energizing the benthic system.
This results in the eventual re-activation of the phosphorus in the bottom muds which the benthic
system had previously deactivated, thus strengthening the pelagic attractor even more.
Assuming the pelagic attractor is dominant and if the level of active phosphorus in the water
column decreases, a critical threshold is again reached below which it is no longer possible to
capture enough solar energy to energize the pelagic system. In effect, the exergy in the water
column decreases below the minimum level for the window of vitality of the pelagic system. As
this occurs the exergy at the bottom increases thus re-energizing the benthic system. And so the
aquatic system flips back and forth between the pelagic and the benthic regime depending on
where in the water column the sunlight's exergy is available to energize the system.
Increased solar exergy at bottom

Pelagic Attractor
Turbidity

Benthic Attractor

Nutrients

Increased Exergy in the Water Column

Elsewhere, Kay and Regier [23] sketched a more detailed partial narrative of Lake Erie
as a SOHO system. This narrative weaves together the themes of organism, species,
ecosystem, landscape and biome in the context of physical environment, climate and
human habitation and the changes therein. Some of the crucial morphogenetic causal
loops, particularly those involving phosphorus, and their relationships to the canon of

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 14

the pelagic and benthic attractors are outlined. The narrative takes the form of a
multilayered account of the ecosystem's operation from different perspectives and
scales. While some individual elements of the narrative consist of traditional scientific
models and descriptions, the synthesis of these elements together into a narrative
transcends normal scientific descriptions.
In this narrative of Lake Erie, the feedback loops, which buffer the system from changes
in external influences, are of particular importance. The benthic attractor has elaborate
feedback schemes, operating at different spatial and temporal scales, for limiting the
phosphorous in the water column. The pelagic attractor has elaborate schemes to
accomplish just the opposite. The way in which changes in context enable and disable
these feedback loops, and their associated canons, thus re-enforcing attractors or
triggering flips between them, has received little attention from the scientific community.
Yet our work would suggest that it is precisely these questions concerning the "flip" from
one attractor to another, through accounting for how environmental influences (context),
acting at different spatial and temporal scales, disable one feedback system while
enabling another, that we must understand. This is essential to comprehending the
relationships between human activities and changes in the ecology of the lakes.
Another example of the notion of canon and attractor, and the ability to characterize
them in terms of the form of exergy utilized, is Holling's four-box cyclical model of
terrestrial ecosystems.[24] (See figure 3 at end of paper.) The first trajectory is the
"exploitation" to "conservation" thermodynamic branch which culminates in the "climax"
community. The biological attractor is the autotrophic system (i.e. a forest). The canon
is expressed, for example, as the growth of a forest to maturity and this is energized by
solar energy. However in the process of increasing the utilization of solar energy and
hence building more structure, much exergy is stored in the biomass. This has the
effect of moving the system further and further from thermodynamic equilibrium as it
develops.
When, as Holling puts it, the inevitable accident (fire, windstorm, or pest outbreaks)
happens, suddenly much exergy is available in the form of dead biomass. This exergy
energizes a new biological attractor, the heterotrophic or decomposer system. This is
the thermodynamic branch which runs from "release" to "re-organization". As the
system progresses along this path it releases the stored nutrients while using the stored
exergy. Eventually the stored exergy runs out and the heterotrophic system collapses.
However in the process it has released the nutrients necessary for the re-emergence of
the solar energy-powered system. This interplay between two biological attractors,
which are organized around different forms of exergy, materials, and information is
played out giving rise to the landscape we see.

Narratives III: Morphogenetic Causal Loops

The internal causal schemes which maintain the attractor and the canon of SOHO
systems can be described in terms of morphogenetic causal loops made up of positive
and negative feedbacks, some of which generate autocatalysis. Ulanowicz [25]
discusses the importance of these morphogenetic causal loops to the understanding of
ecosystems. Two simple examples, taken from DeAngelis [26] are presented below.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 15

Consider forests in dry mountainous areas of the world. Often, as moisture laden
clouds pass over bare mountains, they will not drop rain because of the heat reflected
from the bare rocks. However as a forest develops on a mountain the re-radiated heat
decreases [27]. As the re-radiated heat decreases more rain falls, which promotes more
forest growth, which promotes more rain fall....
In southeastern Australia the dominant trees are sclerophyllous eucalypts, but the
undergrowth consists of lush mesophytic vegetation. Normally these circumstances
would give rise to a temperate rain forest. However these systems are subject to
frequent fire, which would not occur if the mesophytic vegetation dominated. Fire
increases soil leaching and sclerophylls are better adapted to poorer soils than
mesophylls. Thus the dominance by sclerophyllous forest depends on fire and the
occurrence of fire depends on the dominance by sclerophyllous forest. The
morphogenetic causal loop of sclerophyllous dominant forest, fire, and soil infertility
obstructs the development of temperate rain forests.
Another important aspect of SOHO systems is the role of morphogenetic causal loops in
maintaining the canon of a system in spite of a changing context [28].Consider for
example the acidification of lakes. The acidity in the precipitation changed substantially,
but rather incrementally, over a number of years. However, the pH of the receiving lake
waters did not change substantially, relatively speaking, over the same period [29]. In
our terms, the pH of precipitation is part of the context of the SOHO system. While this
changed substantially over time, the lakes maintained their canon through a series of
feedback loops that largely buffered them from the environmental change. Eventually
the inflows or runoff reached a level of acidity which exceeded the compensatory
capacity of these loops. Once this happened, the effectiveness of the SOHO system to
maintain its existing state decreased, which in turn decreased the capacity of the loops
to compensate, which decreased the effectiveness of the SOHO system.... and then
quickly the canon unraveled and the SOHO system flipped to another attractor, in this
case a "dead" lake. The narrative description of a SOHO system must not only
delineate the morphogenetic causal loops, but also the contextual circumstances in
which they can and cannot operate. Doing this in effect defines the domains of the
attractors, the resiliency of the canon, and its window of vitality.

5. Decision Making in the Post Normal Mode

Funtowicz and Ravetz [30] distinguish problem-solving strategies for different
circumstances defined by the inherent uncertainties in the situation and the severity of
consequences arising from the decision to be made. “Normal science”, either as applied
science or mission-oriented research, succeeds where the relative uncertainties are low
(and most of it can be handled by standardized procedures), and the stakes or
outcomes associated with decisions to be made are modest. In contrast, SOHO
systems are situations for which there may be little useable science, high levels of
inherent uncertainty, and severe potential consequences from decisions that have to be
made. These are “Post-Normal Science” situations. Some similarities between this
depiction of decision making circumstances and “mindsets” to understand SOHO
systems are noted in Table 3.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 16

Table 3: A comparison of the normal applied, professional consultancy and Post-normal

science approaches to environmental concerns. (Based on discussions with
S.Funtowicz and J. Ravetz)

Normal Conventional Professional Post-normal Science and Inquiry

Applied Science Consultancy
Essentials
Certainty Uncertainty (reducible in Uncertainty (irreducible in principle)
principle, we lack knowledge)
Low stakes Intermediate stakes High stakes
Facts: Truth found Solution: client happy, society is Resolution: a course of action chosen
satisfied
Results
Hard Try to be hard Soft
Predictable Error reduced to an acceptable Unpredictability a fact of life
level
Quantitative Quantitative ± Quantitative + Qualitative
In the service of
Truth Client in a societal institutional Decision makers, policy, public
framework
Judgment of results
Truth accepted No mistakes (i.e. surprises) Quality of process, integrity
Peer review Holds up in court, client happy Holds up to public scrutiny, move
forward
Mode of Inquiry
Hypothesis testing Problem solving Ecosystem approach
Pursuit of TRUTH Mission and product oriented Pursuit of understanding
Reductionism Holarchic
Analysis Analysis + Design Analysis + Design + Synthesis
Explanations
Linear cause and effect Non-linear, negative feedback Negative + Positive feedback,
autocatalysis, morphogenetic
causal loops
Mechanistic Mechanistic + Cybernetic Synergistic, emergence,
Stability Control, homeostasis Change, evolution, ∞ cycles
Efficiency Efficiency + Adaptation
Extremum principles Local optimum, tradeoffs
Laws Propensities and constraints
Forensics
Fact Interpretation Testimony
Characteristics
Objective, one correct view Subjective, client-consultant view Subjective, plural
Value free Limited values Ethical, Integrity
Predictive Management Control Management Anticipatory + Adaptive Management
Physics Engineering Ecological Economics

In the post-normal situation, the shared understanding of decision making about issues
takes on a different complexion. Decision making becomes what it has always been
about, finding our way through partially undiscovered country rather than charting a
scientifically determined course to a known end point. But to what end? Decisions must
be made about which of the systemic possibilities (i.e. attractors) to promote and which
to discourage. Tradeoffs must be made. Decisions must also be made about how to

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 17

deal with the inherent uncertainties, what risks to take, what contingencies to plan for,
what backups to have in place. These decisions must be informed by science, but in
the end they are an expression of human ethics and preferences, and of the socio-
political context in which they are made. This of course raises the question, who
decides? At the very least, those who might be affected by the outcomes should have
some role in making the decisions.
Given that the ability to forecast is limited, management and decision-making strategies
must focus on maintaining a capacity to adapt to changing environmental conditions.
Adaptive management involves a very different agenda than anticipatory management
[31]. For example, the issues of requisite redundancy, contingency planning and
designing human systems which can evolve, come to the forefront. In adaptive
management, differences between how the future actually unfolds and how it was
anticipated that the future would unfold, are seen as opportunities for learning. This is
in sharp contrast to anticipatory management which sees such deviations as "errors" to
be avoided. Much of the agenda of adaptive management is learning through
experimentation rather than focusing on error avoidance. Adaptive management is not
meant to displace anticipatory management but rather to complement it. The program
of post-normal science is to provide a basis for the understanding necessary to unravel
complexity (emergence, irreducible uncertainty, internal causality), so that we may
successfully anticipate, when possible, and adapt, when appropriate or necessary, to
changes in the self-organizing systems of which we are an integrated and dependent
part.

Choosing a path through sets of SOHO systems not only benefits from bringing a G-
type understanding of mutual causality. It also needs to involve those who can identify a
range of human preferences and issues to be addressed, i.e. a variety of actors and
stakeholders whose desires, perceptions, and knowledge must be integrated with
scientific understanding.
A framework for this integration is presented in Figure 4 (see end of document). It
serves as a heuristic for the adaptive ecosystem approach, where it is accepted that
human systems and biophysical systems are mutually interrelated in complex ways.
This framework presumes that decisions about environmental issues involve mapping
out a vision of how the landscape of human and natural ecosystems should co-evolve
as a self-organizing entity to meet human preferences. The left hand box in Figure 4 is
about developing a SOHO narrative description as discussed in this paper. This
delineates the system’s possibilities and constraints, and the contextual realities of the
situation. Essential to the narrative is the holarchical description of the system. Its
relevance depends on choosing the appropriate system type perspectives and scales
for observation, the appropriate processes and structures for study. The
appropriateness of this choice depends on the issues which are of human concern. The
business of establishing an issues framework is the undertaking represented by the
right hand box. It involves establishing who the actors and stakeholders are, what their
values and concerns are, what their visions for the future are. Because people's
perceptions, concerns and visions will be altered during the exercise of developing the
narrative description, the undertakings of developing an issues framework and a SOHO

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 18

narrative depend on each other and thus are recursive, as is appropriate in the context
of morphogenetic causal loops.
The inherent potentialities of the SOHO systems combined with human vision and
preferences gives rise to scenarios of possible and desirable futures. Each of these
scenarios will represent different sets of tradeoffs and require planning and negotiations
among stakeholders to reconcile these tradeoffs and to develop a plan, or pathway for
the future. The nature of this plan is that it encourages and discourages, as is
appropriate, human activities based on a vision of how the integrated human systems
and ecosystems should co-evolve as a self-organizing entity. It also identifies the
institutional arrangements for governance necessary to act and adapt to the way in
which the self-organization unfolds.
In this mode, governance is an activity that focuses on the SOHO systems and adjusts
the vision based on how the self-organization process is unfolding. Management seeks
to translate the vision into reality by maintaining the context for self-organizing systems,
rather than intervening in the system in a mechanical way as is done under the H-type
mindset. Generally speaking management concentrates on the relationship between
human systems and ecosystems, and on guiding the human side of the relationship.
Monitoring is the activity of observing the human and natural self-organizing systems
and synthesising the observations into a narrative of how the situation is actually
unfolding. This narrative is then used as the basis for governance and management,
that is, for revisioning, and adapting human activities, as is necessary.
In summary, in this post-normal mode of decision making, people provide an image of
how they would like to see their landscapes and social systems evolve. Science,
drawing upon several epistemological traditions, helps identify known constraints and
possibilities of the SOHO systems. A dialogue must ensue which explores the desired
and the feasible, and reconciles these in a vision of how to proceed. Scientists inform
this dialogue by providing the narratives through a process in which they participate as
equals with others in the task of articulating the vision, and identifying pathways for the
future. To us, this constitutes an ecosystem approach for sustainability as it has to be
interpreted in the context of Post-Normal Science grounded in complex systems
thinking.

Acknowledgements
Many, many conversations with David Waltner-Toews and Tamsyn Murray helped
refine these ideas. Other members of the informal “Dirk Gently gang” have also
contributed much (Silvio Funtowicz, Jerry Ravetz, Mario Giampietro, Martin O'Connor
and Gilberto Gallopin). Our students, Nina-Marie Lister, Kate Oxley, Richard Martell,
and Beth Dempster (University of Waterloo) and Charlotte Sunde (Massey University,
New Zealand) and other members of our post-normal science discussion group at
Waterloo have also made invaluable contributions. Finally J. Kay would like to
acknowledge the constant encouragement and support of T.F.H. Allen and R.E.
Ulanowicz.

© James Kay et al. 1999

 Available nutrients

CONSERVATION

REORGANIZATON
•Autotrophic Climax
•Heterotrophic Climax

To another
•Pest
attractor
•Fire
•Storm
Exergy consumption

© James Kay et al. 1999

Futures Vol 31, #7, Sept. 99, pp.721-742

EXPLOITATION

RELEASE
Figure 3: Holling's four box model as a dual thermodynamic branch system.
19

Connectedness
Stored exergy
 Futures Vol 31, #7, Sept. 99, pp.721-742 20

For the four box model, there are two axises

Horizontal axis: Stored exergy. The amount of exergy stored in biomass. This is
related to the amount of nutrients bound in the biomass.

Vertical axis: Exergy consumed. The rate at which exergy is utilized by the system.

From a non-equilibrium/self-organization theory perspective, the four box proceeds as

follows

Starting at exploitation: If there is sufficient materials and biological information

available, then dissipative processes will emerge which utilize the exergy in the solar
energy. In other words, some organisms will take advantage of the available resources.

The thermodynamic direction of all self-organizing processes is to increase its utilization

(consumption) rate of exergy (Kay). In this case the developmental path taken also
involves increasing biomass and hence stored exergy (Jørgensen). The more exergy
stored, the bigger the structure, the better able it is to utilize exergy, the bigger it gets,
etc. This is the direction of the first thermodynamic branch. The exergy source is
solar energy.

Problem: The more exergy that is stored in the system, the more likely (according to
the restated second law of thermodynamics) that some dissipative process will emerge
to take advantage of it. So fire, pest outbreaks etc., occur that take advantage of all the
exergy stored in the biomass. The paradox is that the more effective the system is at
consuming exergy, that is the more organized it is, the more exergy it contains and
hence the more likely it is to be consumed by another self-organizing process (i.e. fire,
pest outbreaks etc.)

So conservation represents a point of maximum thermodynamic organization in the

sense that the system is utilizing the available exergy as fully as possible. But it also
represents a point of maximum thermodynamic risk as it as far out of equilibrium as is
possible. (Distance from thermodynamic equilibrium is measured by exergy content.)

In the language of attractors, there are two attractors, the attractor of maximum exergy
consumption and the attractor of local thermodynamic equilibrium. For this particular
thermodynamic branch the attractor of maximum exergy consumption is moving in
opposition to the local equilibrium attractor. The conservation point is the place where
the two attractors are in balance. For some systems this balance is precarious for
others less so, but in the end the local equilibrium attractor is always dominant

Once the inescapable happens, that is release, a new source of exergy is available for
use, that is the exergy in the stored biomass. Again it inevitable that this new exergy
source will be utilized. As always, the self-organizing process unfolds in a direction of
increasing exergy utilization, except that the processes involved are fundamentally

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 21

different and instead of storing biomass and hence exergy, they release the exergy in
the stored biomass and at the same time release the stored nutrients. This is the
direction of the second thermodynamic branch. The exergy source is stored energy.

Eventually the reorganization point is reached, that is point where the stored exergy
runs out. But now the raw materials are available to start along the first type of
thermodynamic branch again. Which specific branch is followed is a function of the
biological information, nutrients and current environmental conditions. (And this is the
point where biodiversity is so crucial, as this is the point where resiliency matters.)

To summarize: There are two thermodynamic branches, that is self-organizing

pathways that are followed. One (from exploitation to conservation) is driven by the
exergy in solar energy and involves increasing biomass and hence stored exergy. The
other (from release to reorganization) is driven by stored exergy and involves the
release of the exergy and hence biomass. The direction of both is increased exergy
utilization. The ecosystem alternates between these two sources of exergy and hence
follows two qualitatively different pathways of self-organization. The specifics are
determined by the environmental conditions, available resources and biological
information, the latter usually being the determining factor.

The first branch has been traditionally referred to as succession, or growth and
development. Biologically it is the attractor for the autotrophic system. The second
branch is about creative destruction, that is decomposition. Biologically it is the attractor
for the heterotrophic system.

So the exploitation point is one of minimum exergy use and storage.

The conservation point is one of maximum exergy use and storage.
The release point is one of minimum exergy use and maximum storage
The reorganization point is one of maximum exergy use and minimum storage.

Note:
Exergy: the quality of energy. It measures how much work can be extracted from an
energy source. In essence it tells us how good a fuel, an energy source is.

The restated second law of thermodynamics (Kay and Schneider): The more exergy
there is, the more likely it is that a self-organizing dissipative system will emerge to take
advantage of it. In biology, the more exergy available, the more likely some organism
will make use of the opportunity.

The further a system is moved from thermodynamic equilibrium (which is measured by

exergy content), the stronger the tendency to return to thermodynamic equilibrium.

Thermodynamic branch: The developmental path taken by a self-organizing system as

it develops.

© James Kay et al. 1999

Figure 4: An adaptive Self Organizing Holarchic Open (SOHO) System approach to ecosystem sustainability
and health. [32] (A better quality graphic can be found at www.fes.uwaterloo.ca/u/jjkay/about/diamond.html)

DEVELOP A SOHO ECOSYSTEM DEVELOP AN ISSUES

DESCRIPTION FRAMEWORK

· Identify stakeholders and other

Science · Identify ecosystem of interest Culture
participants (actors)
and other (extent) and
· Describe current governance and
Knowledge · Hierarchical description (scale Values
institutional relationships and their
and type, nesting)
evolution
· System description (structure &
· Issue identification and analysis
processes, external influences)
by actors (i.e. Checkland CATWOE)
· Attractor analysis (including
· Define integrity & sustainability
feedbacks and propensities)
in terms of a vision for the future
· Evaluation of current integrity General
Complex · Hierarchical mapping of human
Systems · Synthesis (narrative description Vision
concerns and preferences onto the
Theory of self-organizing behaviour) SOHO ecosystem description.

ecological vision and

SCENARIOS
possibilities preferences

SYSTEMS COLLABORATIVE
APPROACHES Which self-organizing entities PROCESSES
(SOHO and attractors) do we want
to encourage and when?

DESIGN THE CONTEXT

Design the human & ecological
infrastructure and circumstances required
to encourage the desired SOHO
© James J. Kay, M. Boyle, 1999

The continuing process of learning, revisioning, ONGOING ADAPTIVE MANAGEMENT

resolving tradeoffs, and planning by the parties
to adapt to the unfolding situation. This will
entail the ongoing evolution of governance
arrangements.
Development and implementation of
GOVERNANCE strategies to promote/ discourage
self-organization in the context of
the communal vision and plan. This
involves the identification of
external contextual changes, flows
into and from the system, and
MONITORING MANAGEMENT feedback loops which are to be
The continuing collection of and encouraged and discouraged.
synthesis of information into a Generally this entails managing
narrative of the present and human activities rather than direct
anticipated evolution of the system intervention in the system.
 Futures Vol 31, #7, Sept. 99, pp.721-742 23

1 Holling, C.S. , "The Resilience of Terrestrial Ecosystems: Local Surprise and Global Change,
Sustainable Development in the Biosphere, W.M.. Clark and R.E.. Munn, eds. (Cambridge: Cambridge
University Press, 1986), pp. 292-320;
Holling, C.S. "Cross-scale Morphology, Geometry, and Dynamics of Ecosystems,” Ecological
Monographs, 62:4 , pp. 447-502.(1992)
Kay, J.J. 1984. Self-organization in living systems. Ph.D. thesis. Systems Design Engineering,
University of Waterloo, Waterloo, Ontario. 458 pp.
2 Maruyama, M. 1980. Mindscapes and science theories. Current Anthropology 21: 589-599. Reprinted
in excerpted form as: Most frequently found mindscape types, pp. 1-30 in M.T. Caley and D. Sawada
(eds.). 1994. Mindscapes: the Epistemology of Magoroh Maruyama. Gordon and Breach Science
Publishers, Langhorne, Pennsylvania. xxiii + 206 pp.
3 For more information see:
di Castri, F. The Evolution of Terrestrial Ecosystems. in Ravera, O., Ed. Ecological Assessment of
Environmental Degradation, Pollution and Recovery. : Elsevier Science; 1987: pp.1-30
Casti, J. L. Complexification: Explaining a Paradoxical World Through the Science of Surprise. NY:
Harper Collins; 1994.
Jantsch, Erich. The self-organizing universe : scientific and human implications of the emerging paradigm
of evolution. Pergamon Press, 1980.
Kay, 184, Op cit
Nicolis, G., Prigogine, I.; "Self-Organization in Non-Equilibrium Systems"; Wiley-Interscience, 1977.
Nicolis, G., Prigogine, I.; "Exploring Complexity"; Freeman, 1989.
Peacocke, A. R. The Physical Chemistry of Biological Processes. : Oxford University Press; 1983;
Wicken, J. S. Evolution, Thermodynamics, and Information: Extending the Dawinian Program. : Oxford
University Press; 1987.
4 A “catastrophic threshold” is a point of discontinuity at which continuous change of some variables
generate sudden discontinuous responses. See Appendix 1 in:
Kay, J.J. 1991. A nonequilibrium thermodynamic framework for discussing ecosystem integrity.
Environmental Management 15(4): 483-495.
5 Nicolis and Prigogine, 1977, 1989 op cit.
6 Schneider, E.D, Kay, J.J., 1994a "Complexity and Thermodynamics:Towards a New Ecology", Futures
24 (6) pp.626-647, August 1994
Schneider, E.D, Kay, J.J., 1994b, "Life as a Manifestation of the Second Law of Thermodynamics",
Mathematical and Computer Modelling, Vol 19, No. 6-8, pp.25-48.
7 Ulanowicz, R. E. 1997. Limitations on the Connectivity of Ecosystem Flow Networks. IN: A. Rinaldo, &
A. A. Marani (eds), Biological Models: Proceedings of the 1992 Summer School on Environmental
Dynamics., pp.125- 143. Venice, Istituto Veneto di Scienze, Lettere ed Arti, Venice.
8 Kay, 1984,, Schneider and Kay, 1994a,b op cit and Kay, J.J., Schneider, E.D., 1992. "Thermodynamics
and Measures of Ecosystem Integrity" in Ecological Indicators, Volume 1, D.H. McKenzie, D.E. Hyatt, V.J.
Mc Donald (eds.), Proceedings of the International Symposium on Ecological Indicators, Fort Lauderdale,
Florida, Elsevier, pp.159-182.

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 24

9 In the case of a vortex in the bathtub water, the exergy is the potential energy of the water, the raw
material is the water and there is no information, the dissipative process is water draining, the dissipative
structure is the vortex. The vortex will not form until enough height of water is in the bathtub, and if too
much height of water is present, laminar flow occurs instead of a vortex.
10 In an ecological setting, examples of the structures are the individuals of species, breeding
populations, forests etc. The processes are reproduction, metabolism, evapotranspiration etc. The
context is the available set of nutrients and energy sources in a physical environment. The information
includes the biodiversity.
11 Koestler, A. 1978. Janus: a Summing Up. Hutchinson, London, U.K. vii + 354 pp.
12 Allen, T.F.H.; Bandurski, B.L.; King, A.W. The Ecosystem Approach: theory and ecosystem integrity:
International Joint Commission; 1993 (Report to the Great Lakes Science Advisory Board).
Allen, T. F. H.; Hoekstra, T.W. Toward a Unified Ecology. New York: Columbia University Press; 1992.
Allen, T. F. H.; Starr, T. B. Hierarchy: Perspectives for Ecological Complexity: University of Chicago
Press; 1982.

13 Ulanowicz, R.E. 1996. The propensities of evolving systems. Pp. 217- 233 In: E.L. Khalil and K.E.
Boulding (Eds.) Evolution, Order and Complexity. Routledge, London. 276p.

14 Popper, K.R. 1990. A World of Propensities. Thoemmes, Brussels.

15 The notion of alternate stable states in ecosystems and their implications are examined in Kay (1984),
Holling (1986), Kay (1991), Kay and Schneider (1994), op cit and
Kay, J.J.. 1997. Some notes on: The Ecosystem Approach, Ecosystems as Complex Systems. In: T.
Murray, & G. Gallopin (eds.) Integrated Conceptual Framework for Tropical Agroecosystem Research
Based on Complex Systems Theories;, pp. 69-98. Cali, Colombia. Centro Internacional de Agricultura
Tropical, Working Document No. 167.
Ludwig, D., B. Walker, B & Holling, C. S., 1997. Sustainability, stability, and resilience. Conservation
Ecology. 1(1): Article 7.
16 Described for example by Tainter, J.A. The Collapse of Complex Societies , Cambridge University
Press, Cambridge MA, 1989, and Chase-Dunn, C. and T.D. Hall, Rise and Demise: Comparng World-
Systems, Westview Press, Boulder CO, 1997.
17 Maruyama op cit
18 As discussed for example by: Allen and Starr, 1982, Allen and Hoekstra, 1992, Allen, Bandurski, and
King 1993, op cit and King, A.W. Considerations of Scale and Hierarchy. Woodley, S.; Kay, J. ; Francis,
G., (eds). Ecological Integrity and the Management of Ecosystems: St. Lucie Press; 1993: 19-46.
19 Murray, T.P., Kay, J.J., Waltner-Toews, D., Raez-Luna, E. "Adaptive Methodology for ecosystem
Sustainability and Health: Development and Application in the Western Amazon Lowlands" (in press)
20 This site and the decision process surrounding its management are described through a detailed case
study using an adaptive ecosystem approach in Kay, 1997, op. cit. and
Lister N-M and Kay, JJ 1999 "Celebrating Diversity: Towards Adaptive Planning for Biodiversity
Conservation." In: S. Bocking (ed.) Canadian Perspectives in Biodiversity. Toronto: Broadview Press. in
press.
21 Regier, H.A. and J.J. Kay. 1996. An heuristic model of transformations of the aquatic ecosystems of
the Great Lakes - St. Lawrence River Basin. Journal of Aquatic Ecosystem Health 5: 3-21. Kluwer
Academic Publishers.
22 For examples, see:

© James Kay et al. 1999

 Futures Vol 31, #7, Sept. 99, pp.721-742 25

Blindow, I.; Andersson, G.; Hargeby, A., and Johansson, S. Long-term pattern of alternative stable states
in two shallow eutrophic lakes. Freshwater Biology. 1993; 30:159-167.
Carpenter, S. R., & K. L. Cottingham. 1997. Resilience and restoration of lakes. Conservation Ecology 1
(1): Article 2.
Scheffer, M. 1990. Multiplicity Of Stable States In Freshwater Systems. Hydrobiologia 200/201: 475-486.
Scheffer, M. Ecology of Shallow Lakes. London: Chapman and Hall; 1998.
23 Kay. J., Regier, H.., 1999. "An Ecosystem Approach to Erie’s Ecology" in M. Munawar, T.Edsall,
S.Nepszy, G. Sprules & B. Shute (eds), International Symposium. The State of Lake Erie (SOLE) - Past,
Present and Future. A tribute to Drs. Joe Leach & HenryRegier, Backhuys Academic Publishers,
Netherlands, (in press)
24 Holling op cit
25 Ulanowicz, R.E. 1997. Ecology: The Ascendent Perspective. Columbia University Press, New York.

26 DeAngelis, D. L.; Post, W. M.; Travis, C. C. Positive Feedback in Natural Systems. Berlin: Springer-
Verlag; 1986
27 Schneider and Kay, 1994 op cit
28 Rapport, D.J., H.A. Regier and T.C. Hutchinson. 1985. Ecosystem behavior under stress. American
Naturalist 125: 617-640.
Rapport, D.J. and H.A. Regier. 1995. Disturbance and stress effects on ecological systems. Pp. 397-414
in S.E. Jorgensen and B.C. Patten (eds.) Complex Ecology: the Organization, Feedback and Stability.
Prentice-Hall, New York.

29 Stigliani, W.M. Changes in valued "capacities" of soils and sediments as indicators of non-linear and
time-delayed environmental effects. Int. J. Env. Mon. and Assess.; 1988; 10: 95-103
30 Funtowicz, S.O. and J. Ravetz. 1993. Science for a post-normal age. Futures 25: 735-755.
31 Holling, C.S. , (ed), "Adaptive environmental assessment and management" Wiley, 1978.
Gunderson, L. H.; Holling, C. S.; Light, S. S. Barriers and Bridges to the Renewal of Ecosystems and
Institutions. USA: Columbia University Press; 1995.
32 This framework has gone through a number of incarnations:
Kay, J.J., 1994, "The Ecosystem Approach, Ecosystems as Complex Systems and State of the
Environment Reporting", prepared for North American Commission for Environmental Cooperation, State
of the North American Ecosystem meeting, Montreal, Canada. 8-10 December, 1994, 42 pages
K. Oxley, Education in Support of the Ecosystem Approach at the Huron Natural Area. M.E.S. thesis.
Environment and Resource Studies, University of Waterloo, Waterloo, Ontario. (1998)
M. Boyle: An Adaptive Ecosystem Approach to Monitoring. M.E.S. thesis. Environment and Resource
Studies, University of Waterloo, Waterloo, Ontario. (1998)
Murray et al, op. cit.
and is discussed in fuller detail in a planning context in
Lister N-M, 1999"Celebrating Diversity: An Adaptive Planning Framework for the Conservation of
Biological Diversity in Urbanising Landscapes. Ph.D. thesis. School of Planning , University of Waterloo,
Waterloo, Ontario.

© James Kay et al. 1999