Resilience of Ecological Systems
Chris J Kettle
Lecture 1: 04.03.2016
Stability, Resilience and Complexity, Ecosystem management in the Anthropocene.
We are living in a period of rapid and enormous global change. At no other time in the Earth’s history
has a single species had so much impact on the earth. Scientists refer to the geological time period
as the “anthropocene”(see Box 1). Population growth and technological advances have enabled us to
impact broadly and deeply across all the earths “natural” processes. For example, the emissions of
hazardous and toxic chemicals (CFC’s), Green House Gases (GHG’s), global-scale deforestation and
land use change, over fishing, urbanisation and mass migration and global travel. All of these factors
exert enormous pressure on the Earth system.
Box 1: In the last three centuries human population growth has increased 10 fold to 6 billion. With on
average one cow per family the global Cattle stock is estimated at 1.4 billion. Urbanisation has
increased 13 fold in the last 100 years alone, more than 50% of accessible fresh water is used for
humans consumption ,between 25% and 35% of primary production from the oceans is removed by
fisheries. Landcover under crops has doubled in the last 100 years corresponding to more than 20%
global decline in forests. More nitrogen is fixed artifically for fertilizer production than naturally but
terrestrial ecosystems. See ref 7 and other papers by P J Crutzen form more information.
The past century has largely focused on technological advances to ‘maximise’ yields, profits and
productivity, with relatively little consideration of the environmental consequences. This has led to
collapse in many natural ecosystems, examples include, fisheries, soil erosion and water shortages
due to deforestation, desertification and eutrophication of lakes. Despite these system collapses, the
Earth system persists. As ecologists and environmental scientists, our challenge is to inform better
management systems, which meet human needs while minimising the negative consequences to the
earth systems through ‘optimisation’ rather than ‘maximization’. Technological advances have
contributed to rapid global degradation but can also provide many opportunities for understanding
and resolve environmental problems and complexities.
For the early part of the 20th Century the dogma in natural resource management was largely one of
‘ecological equalibria’, where the objective was to manage in ‘states’. Managers and ecologists
believed that ecological systems have a tendency to return to a ‘steady stable state’ through selfregulatory processes or homeostasis. There is now an increasing body of evidence that this earlier
view of complex ecological systems does not hold - in fact many ecosystems appear to have
thresholds which lead them to flip from one configuration to an alternative state.
Thresholds
One outcome of the complex interactions of fast and slow variables is that systems may suddenly,
and unpredictably, undergo dramatic changes. It seems that there are thresholds in many systems
that separate (in time and space) two or more different system states. Once we cross these
thresholds, dramatic changes in the economic, ecological or social systems unfold, and these may
lead to huge ecological degradation, economic loss, or social conflicts. Examples include the collapse
of fisheries, the sudden transformation of grasslands to deserts, the transformation of coral reef
systems to algae-dominated communities3, and the outbreak of pests in forests and agriculture
following years of pesticide applications or fire suppression. Under such circumstances traditional
management agencies and processes become paralyzed – they are unable to respond effectively to
these new conditions because all models, concepts, and knowledge is based on a single ecosystem
state, which may vary, but which does not undergo wholesale change to another totally different
state. As traditional agencies fail to cope with the changes, the public lose trust in their ability to
manage, and in the governments’ ability to implement effective policies.
The Pathology of Resource Management
The lack of flexibility of management implementing institutions has been called the pathology of
resource management4. This implies an overemphasis on the control of natural systems and process
by humans so-called ‘command and control’, here the focus was on the prevention of disturbances
and the maintenance of equilibrium (constant) states. This type of management may initially be very
successful in delivering increased productivity by reducing detrimental impacts, but eventually it may
lead to environmental systems that are increasingly vulnerable to large scale changes because the
normal checks and balances that are imposed by natural disturbances are not occurring. Thus, the
inherent variability and uncertainty in natural systems is replaced with human control, but this, over
time, leads to slow changes in ecological, social and economic components, and ultimately to
dramatic changes as certain thresholds are crossed.
Additionally, management agencies become increasingly less flexible and adaptable to natural
changes because their whole philosophy is aimed at avoiding or preventing natural variability.
Economic sectors also become increasingly dependent on management practice that aims to deliver
a constant supply of product, when in fact this can only be achieved through the control of the
natural variability expressed by ecosystems. When a collapse does occur, these economic systems
are therefore much more vulnerable.
1.
Briske DD, Washington-Allen RA, Johnson CR, Lockwood JA, Lockwood DR, et al. (2010) Catastrophic Thresholds: A
2
Synthesis of Concepts, Perspectives, and Applications. Ecology and Society 15. Interesting general reading on these topics
include Something New Under the Sun: An Environmental History of the Twentieth-Century World by John Robert McNeill
3
and Paul Kennedy, and The World Is Flat: A Brief History of the Twenty-first Century by Thomas L. Friedman For coral
reefs, see Mumby, P.J. et al. (2006) Fishing, trophic cascades, and the process of grazing on coral reefs. Science 311, 98-101.
4
An excellent paper that introduced this concept is that of Holling, C.S. and Meffe, G.K. (1996) Command and control and
the pathology of natural resource management. Conservation Biology, 10: 328-337.
What makes ecosystems stable?
If we wish to avoid collapses, or at least incorporate sudden ecosystem changes into current
management practice, then we need to understand both why shifts in ecosystem states occur, and
what maintains ecosystem stability.
Ecosystems have a certain amount of resilience to change. In other words, if an ecosystem is
disturbed (by human action or as a result of natural events) it is usually able to recover after a period
of time to its original (or near original) state.
This robustness is derived from the fact that ecosystems are the products of processes that operate
at a variety of scales. A disturbance may occur at one scale and temporarily impact the processes
operating at that scale, but other processes operating at other spatial and temporal scales will serve
to bring the system back to the steady state. Thus an extremely dry summer may impact the
vegetation of a grassland community, perhaps even causing changes in the structure of a community,
but the large scale and long-term processes of seed dispersal and soil hydrology remain largely
unaffected allowing for re-colonisation and maintenance of soil moisture.
A second reason is that natural ecosystems have functional diversity: when conditions change, some
species will suffer, but others will benefit, the overall result at the landscape or ecosystem scale
being no large change in biomass, productivity or system processes. Functional diversity is effectively
a natural insurance against perturbation. Biodiversity is thought to be a primary determinant of
functional diversity, and hence biodiversity underlies the ability of an ecosystem to recover from
disturbance. Consequently, ecologists argue that loss of biodiversity degrades the stability of
ecosystems which could lead to dramatic changes in ecosystem states.
The spatial heterogeneity or patchiness of a landscape also offers protection from disturbances,
partly because a patchy landscape limits the extent over which a disturbance is expressed, but also
because following disturbance, patches that are little affected by the disturbance act as source
populations for recolonisation.
Myths of Nature
There are a number of different ways in which ecosystems have been perceived. Such viewpoints
have been called “myths of nature”, that is partial truths that describe some aspects of the way
nature behaves.
These perspectives are illustrated in Figure 1. In these images, the ball indicates the state of the
environment, and the surface on which it sits represents the forces acting on the ecosystem should
the environment be perturbed. Pushing the ball in any direction represents a disturbance or
management intervention.
In the first diagram, Nature Flat indicates that there are no natural forces that affect the direction in
which an ecosystem moves, and therefore humans can manipulate nature in any direction they want.
In this scenario nature is amenable to human control and largely predictable—we simply manage the
environment as we wish. Issues of resource use and development are considered to be entirely
within human control, and movement of the ecosystem in one direction can be, with appropriate
management intervention, reversed. This is the promethean viewpoint: humans can overcome any
environmental problem.
Figure 1. Four myths of nature.
Nature Balanced refers to nature existing at some equilibrium condition: there is always a tendency
to return to the original state following disturbance. You push the ball a little and it will return to the
equilibrium state represented by the bottom of the depression. This is the viewpoint that is
associated with traditional resource management systems prevalent in forestry and fisheries.
Harvesting of trees or fish is expected to be followed naturally by a gradual recovery of the harvested
population to the carrying capacity. The carrying capacity is the equilibrium point to which
populations are expected to return to. The equivalent concept at the community level is the “climax
vegetation type”. This scenario forms the basis of much modeling work that underpins management
agency policies, and includes at its core the concepts of the logistic model of growth.
Nature Anarchic is equivalent to viewpoints that emphasise the fragility of nature. A small
perturbation will cause the ecosystem to change dramatically with no possibility or recovery. This
perspective is held by many environmentalists who refer to the fragility of, for example, rainforests,
coral reefs, species rich grasslands etc. All of these views are partly correct, but each on its own is an
incomplete description of reality.
More realistic is the Nature Resilient view, where there are some ecosystem states that are fragile,
and some that are relatively stable with a local equilibrium, but there is also the possibility that the
ecosystem can flip from one state to another (when the ball is pushed hard enough that it escapes
the stability domain and falls into another). This view includes the concept of thresholds and
alternative stable states, and therefore it implies that the resiliency of an ecosystem is limited.
Equilibrium and Dynamic Ecosystem Perspectives
If we ignore the Nature Flat and Nature Anarchic views as being overly simplistic, we are left with the
traditional resource management perspective (Nature Balanced) and the more complex, dynamic,
and unpredictable, Nature Resilient view. Nature balanced emphasizes the equilibrium of nature,
while Nature resilient recognizes local equilibria, but emphasises the dynamic nature of ecosystem
change. An equilibrium perspective focuses on the carrying capacity, and its central management
idea is that of achieving a maximum sustained yield. The dynamic view focuses on stability domain
boundaries, thresholds, and the possibility of switching to alternative states, and management
therefore has to be adaptable. We now recognize that key features of ecosystems include: • Periodic
and unpredictable disturbance and change • Considerable patchiness and several alternative
recovery profiles • Multiple equilibria representing different states • Thresholds that are uncertain
Given these features, we may question what we exactly mean by ‘stability’.
What is Stability?
Stability can be defined in terms of the resilience an ecosystem has in response to change. However,
there are at least two ways of defining resilience. The first is the speed of recovery to the original
state after a perturbation, and assumes an equilibrium state in the direction of which the recovery
will proceed. This implies a predictable recovery profile (we can predict, for example, that following
selective harvest of trees, a forest will return to its pre-harvest state in 30 years). Managers will try to
determine the resiliency (rate of recovery following impact) of an ecosystem and manage resource
use to ensure optimal recovery for maximum efficiency. This perspective falls within the traditional
resource use management paradigms and has been called “engineering resilience” to reflect the
desire of human control. The second definition of resilience is the amount of stress, or size of
disturbance, that can be absorbed before the ecosystem switches to an alternative stable state. This
view recognizes the possibility that a system might not recover to its original state and instead
changes to something very different (for example the collapse and subsequent failure of fish stocks
to recover). Concepts that are implied by this definition of resilience are ecosystem persistence,
adaptiveness, variability and unpredictability, and the term “Ecosystem resilience” is often applied to
reflect this complexity. Management recognizes multiple stable ecosystem states, and therefore
proceeds with caution. The precautionary principle becomes important. Such management is
appropriate when there is uncertainty about the dynamics of the ecosystem.
Spruce-fir Forests and Insect Outbreaks An example (see Figure 2) will illustrate how ecosystems
may have resilience but can also be subject to sudden and dramatic change. Spruce-fir forests in
North America are subject to herbivory by insects, notably the spruce budworm. When the forest
stand is young the structure is relatively open and there is little foliage. Consequently, birds that eat
the budworm are able to control budworm populations and maintain them at low population
densities. As the tree population slowly matures it becomes increasingly difficult for the birds to find
their prey in the denser foliage. After several decades the formation of a dense mature forest stand
and the corresponding loss of predator efficiency allow the budworm to be released from predator
pressure leading to a budworm outbreak. The rapid population expansion of the budworm is
maintained for a few years owing to the extensive forest areas, but also causes widespread
defoliation and forest death. Following a few years of decomposition and seedling establishment, the
forest returns to the young stage where birds can once more control budworm populations. Here we
see local equilibria (budworm populations maintained at low population densities throughout most
of the life of the forest trees), and also thresholds (the point at which birds are no longer able to
control budworm populations, not because of increasing budworms, but because of a reduction in
bird foraging efficiency), collapses (widespread forest mortality), and alternative ecosystem states
(low budworm populations and high budworm populations).
Figure 2. The dynamics of the Spruce-fir forests of North America.
It is very possible that similar dynamics operate within the Alpine spruce-fir forests in Switzerland.
The Adaptive Cycle This example can be generalized into a conceptual model of system dynamics
that has been termed the Adaptive Cycle5(Figure 3). This adaptive cycle, which incorporates the
concepts of equilibria, but also thresholds, collapse, fast and slow variables and alternative states,
can be equally applied to the dynamics of economic and social systems as well as ecological systems
(see references on web site for more information on this).
Gunderson and Holling (2002) Panarchy. Island Press.
Hysteresis
An important element of alternative stable states is that once an ecosystem has switched from one
state to another it may be very difficult, and very expensive, to return it to the original state. This is
the concept of hysteresis6, and is illustrated in Figure 4. +A clear lake (blue line) is maintained by
vegetation that soaks up the additional nutrients derived from fertilizer runoff, with only a small
increase in turbidity due to increasing phytoplankton. Once a critical turbidity is reached, however,
the light reaching the ground vegetation is not sufficient to maintain them, and the vegetation dies.
At this point turbidity increases dramatically, because without vegetation the phytoplankton-grazing
fish have no refuge from predators and therefore also decline, allowing a rapid explosion in
phytoplankton. We are now in the alternative state of a turbid lake that lacks vegetation (red line).
To return to the original state nutrient input must be reduced to very low levels (far below the point
which triggered the change from clear to turbid) because there are no longer any phytoplanktonfeeding fish to reduce plankton biomass (figure from Scheffer et al. 2001).
Complexity from Simplicity
We now realize that ecosystems dynamics is exceedingly complex, but also that this complexity
arises from the interaction of a few very simple interactions. The idea that complexity arises from
very few critical variables and processes that operate over different scales in space and time is
receiving considerable research interest6-7. It also represents a challenge for the management of
ecosystems over the long term. Implications for Management In conclusion, there are a number of
issues that ecosystem managers should recognize: • Ecosystem shifts cause large losses of economic
and ecological resources • Restoration may require extensive action (hysteresis) • Gradual reduction
in resilience is rarely noticed • Management should focus on: – maintaining resilience, and NOT
preventing disturbance – sustaining a large stability domain – slowly-changing variables: land use,
nutrient stocks and biodiversity.
6
See Scheffer et al. (2001) Catastrophic shifts in ecosystems. Nature 413: 591-596. 6 See, for example, Fragile Dominion
7
(1999) by Simon Levin, or Deep Simplicity (2005) by John Gribbin, both very readable books. Steffen W, Crutzen PJ,
McNeill JR (2007) The Anthropocene: Are humans now overwhelming the great forces of nature. Ambio 36: 614-621.