Resilience:
Exploring Stability and Change in
the Dynamics of Systems
Jan Sendzimir
International Institute of
Applied Systems Analysis
Laxenburg, Austria
sendzim@iiasa.ac.at
Ecological Succession
South-eastern North America
(After E.P. Odum 1971 Fundamentals of Ecology)
2
The result of a
century of fire
suppression in
North America?
Vulnerability 
More than 180
million hectares
extremely
vulnerable
to fire.
3
Sudden Collapse of the Oldest,
Richest Fishery on Earth
1900 tons
Annual
Catch
Of
Cod
(1000 tons)
More than 400 years
25 years
90 tons
Northwest Atlantic Cod Harvest (1895 – 1993)
2003 – after 10 years, no sign of recovery
4
Catastrophic Examples of
Sudden Shifts and Flips
Coral Reefs
coral vs. algae
Arid Landscapes
shrubland vs. grassland
Shallow Lakes
eutrophic vs. clear
North Florida Forest
– longleaf pine savanna & fire vs.
hardwood forest without fire
5
Outline
 Shallow Lake dynamics
 Resilience Theory
– Stability Landscapes
• Visual descriptions of systems dynamics
– Boreal Forest Case Study
– Factors Influencing Resilience
• Control of disturbance
• Regulation of Renewal
 Australian Rangeland Case Study
– Managing for resilience
 Summary
6
Lake Eutrophication
The flip from clear to turbid water
Some lakes remain clear for decades until one
summer storm churns up the sediments, and it
remains turbid for decades, despite all “cures.”
7
Eutrophication
can proceed
slowly for decades
as a natural enrichment that fills lakes.
Human waste can
accelerate the process
to a few years.
8
Possible ways in which
ecosystem equilibrium
states can vary with
conditions such as
nutrient loading,
exploitation or
temperature rise.
Clean
Turbid
Poor
Enriched
Conventional
Models of
Relations between
Ecosystem States
And Conditions
In a and b, only one
equilibrium exists for
each condition.
9
Percent
Of Lake
Covered
By MacroPhytes
Phosphorus in Water
Response of charophyte vegetation in the shallow Lake
Veluwe to increase of the phosphorus concentration in
the 1960s.
10
Hysteresis
Percent
Of Lake
Covered
By MacroPhytes
Response of charophyte vegetation in the shallow
Lake Veluwe to increase and subsequent decrease of
the phosphorus concentration. Red dots represent
years of the forward switch in the late 1960s and early
1970s. Black dots show the effect of gradual reduction
of the nutrient loading leading eventually to the
backward switch in the 1990s.
11
If the equilibrium curve is folded
backwards (c), three equilibria can
exist for a given condition. Equilibria
on the dashed middle section are
unstable and represent the border
between the basins of attraction of
the two alternative stable states on
the upper and lower branches.
12
Outline
 Shallow Lake dynamics
 Resilience Theory
– Stability Landscapes
• Visual descriptions of systems dynamics
– Boreal Forest Case Study
– Factors Influencing Resilience
• Control of disturbance
• Regulation of Renewal
 Australian Rangeland Case Study
– Managing for resilience
 Summary
13
Resilience Theory
 Invert the normal pessimism
– “If the world really is collapsing, why do so many
ecosystems persist?”
 Develop common tools to study the decline,
collapse or persistence of ecological,
economic and social systems.
 You are resilient if your identity persists:
– In the face of shock or disturbance the same set
of organizing processes remain to control the
behavior and appearance of a resilient system.
14
15
Lynx-Hare
Phase
Space
revealing a simple symmetry within
complex fluctuations
16
State
Space
Views of
Ecosystem
Dynamics
17
Landscape and State Space Views
of “Industrial Optimism”
18
Stability Landscape View
of “Ecological Pessimism”
19
Stability Landscape View of
Multiple Stable States
20
Stability Landscape View
of Evolution
Shift from one domain to the next
as the rules change
As it changes, a system
modifies its own possible states.
Here a smaller and smaller
perturbation can shift the
equilibrium from one stability
domain to another.
Finally the stability domain
disappears and the system
spontaneously changes state.
21
If the equilibrium curve is folded
backwards (c), three equilibria can
exist for a given condition. Equilibria
on the dashed middle section are
unstable and represent the border
between the basins of attraction of
the two alternative stable states on
the upper and lower branches.
22
Shallow lake dynamics from clear to turbid and back
Figure 3 External conditions affect
the resilience of multi-stable
ecosystems to perturbation. The
bottom plane shows the equilibrium
curve as in Fig. 2. The stability
landscapes depict the equilibria and
their basins of attraction at five
different conditions. Stable
equilibria correspond to valleys; the
unstable middle section of the folded
equilibrium curve corresponds to a
hill. If the size of the attraction basin
is small, resilience is small and even
a moderate perturbation may bring
the system into the alternative basin
of attraction.
23
Resilience
 “…the amount of change or disruption
that will cause an ecosystem to switch
from being maintained by one set of
mutually reinforcing processes and
structures to an alternative set of
processes and structures.”
– G. Peterson 2000
24
Ecosystem Resilience
Dynamic Exchanges between
Stability and Disturbance
 Stability is recognized for its contributions to
productivity and bio-geochemical cycles.
 Like ‘invigorating’ gymnastics, disturbances
contribute to diversity, structure and
resilience.
 The engine of evolution and resilience.
– Not disturbance alone
– Nor stability alone
– But the cycling between them
25
Collapse of Resilience
 Surprise from Cross-scale Interactions
– Occasionally Natural systems develop to a
stage of “over-maturity” where elements are
over-connected.
– They become accidents waiting to happen.
– Then collective activities of small scale
processes can “cascade upward” and cause
the system to flip to another system type.
26
Outline
 Shallow Lake dynamics
 Resilience Theory
– Stability Landscapes
• Visual descriptions of systems dynamics
– Boreal Forest Case Study
– Factors Influencing Resilience
• Control of disturbance
• Regulation of Renewal
 Australian Rangeland Case Study
– Managing for resilience
 Summary
27
Global Geographical
Distribution of Taiga Forest
28
Spruce Budworm in Boreal Forest
in New Brunswick, Canada
29
Spruce
Budworm
Adults and
Larvae
30
Simulated Dynamics of Boreal Forest
Budworm and Foliage Surface Area
31
Forest Growth Changes the Rules:
Feedback becomes more likely
32
The Dynamics of Change:
Paradoxical Twins
Unpredictable Change - Surprises
 Smooth, continuous change suddenly
interrupted by reversal or collapse.
Predictable Change - Return Times
 Fires, Floods, Pest Outbreaks
How do we reconcile these
contradictions?
33
System Dynamics
focus attention on destruction and re-organization
as well as growth and conservation
“The adaptive cycle
…a useful metaphor not a testable hypothesis.”
(Carpenter et al. 2001)
34
Outline
 Shallow Lake dynamics
 Resilience Theory
– Stability Landscapes
• Visual descriptions of systems dynamics
– Boreal Forest Case Study
– Factors Influencing Resilience
• Control of disturbance
• Regulation of Renewal
 Australian Rangeland Case Study
– Managing for resilience
 Summary
35
Ecological Resilience
Measures system integrity as the
capacity to absorb disruption and remain
the same kind of ecosystem.
Emerges from cross-scale
interactions
Depends upon:
Control of Disturbance
Regulation of Renewal
36
What Promotes Resilience?
Control of Disturbance
– Disturbance Frequency and Intensity
– Technical Restrictions
– Chesapeake Shellfish Fishery
– Herbivore grazing/browsing
– Fire or logging in forests
– Development in floodplain
– Local rain cycle in river basins
37
What Promotes Resilience?
Control of Disturbance
– Capacity to Absorb Disturbance
– Landscape morphometry
– Room for the River Program - Rhine river
– Habitat availability
– Ability to migrate (connectivity of landscape)
– Spatial Heterogeneity (mangroves, eel grass)
– Processing and Cycling of Resources
– Cross-scale functional reinforcement
– Within-scale functional diversity
38
What Promotes Resilience?
Regulation of Renewal (or
Regenerative potential)
–Stored Resources
–Soil depth, organic content,
seed bank
–Water (aquifer, lake, river)
–Nutrients in biomass
39
What Promotes Resilience?
Regulation of Renewal
–Facility of Response
–Recolonization distance
–Proximity of Youth (Kobe Earthquake)
–Biodiversity
–Cross-scale functional diversity
–Capacity to adapt, to generate novelty, to
innovate
40
What Promotes Resilience?
Regulation of Renewal (Regenerative potential)
–Availability of Information
–Viability of cultural information transfer Cultural Capital
–Language (Norway surrenders to English)
–Customs (education, discourse)
–Politics and institutions
–Human Memory & Population Age Structure
– Cree People and Caribou (Birkes)
41
Outline
 Shallow Lake dynamics
 Resilience Theory
– Stability Landscapes
• Visual descriptions of systems dynamics
– Boreal Forest Case Study
– Factors Influencing Resilience
• Control of disturbance
• Regulation of Renewal
 Australian Rangeland Case Study
– Managing for resilience
 Summary
42
Australian Rangelands
43
Geographical Distribution of
Australian Rangelands
44
For Ecosystems with threshold
effects and multiple stable states
 Supply of ecosystem services
– Can arise from many combinations of state
variables.
– Depend more on the stability domain the system is
in than any particular combination of state
variables.
 Management strategy
– Sustain or enlarge the stability domain (system
configuration) rather than emphasize a particular
state or variable to maximize production.
45
Ecosystem Dynamics and
Services in Australian Rangelands
 Ecosystem services (livestock products)
– Depend on amount of grass which
depends on amount of shrubs
– When shrub cover (area) exceeds a
threshold, there is not enough grass to
sustain a fire that will control the shrubs.
– The system moves on an undesirable
trajectory toward a Shrub stability domain.
46
Management for Resilience in
Australian Rangelands
Possible trajectories of a 2-variable system through time.
The positions of the dashed lines on the axes represent
critical threshold levels. (Walker, B. et al. 2002)
Ratio
of
Debt
Income
More grass
Ratio Woody vegetation / Grass
More shrubs
47
Not maximization of one goal: profit or biodiversity.
Expanding the Domain of
Desirable Options
 Biophysical axis
– increase the proportion of perennial
species in the grass sward, control grazing
pressure
 Socio-economic axis
– increasing access to alternative (external)
sources of income
• (game farming, tourism).
48
Summary
 Phase Space and Stability Landscapes
visually describe system dynamics.
 Resilience theory allows us
– to describe how different variables
influence system dynamics at different
scales.
– to manage for how a system moves and
retains its integrity as opposed to single
goals.
49
Sources of Uncertainty
Complexity
 Feedbacks, delays in system
interactions
– Completely confuse conventional models
• Causation: Life is a series of events
• Change: proceeds smoothly and monotonically
 Changing structure - shifting
relationships
50
Old Standards of toxicology in Jeopardy
1. EDC’s have many counter-intuitive properties:
• threshold assumption and non-monotonic effects
• synergy
• exposure during early ontogeny (sensitive periods)
• transgenerational effects
2. Integrating behavioral and evolutionary ecology
• limitations of using inbred strains
• limitations of tests in artificial conditions
e.g., increase susceptibility to predation and infection
Sources of Uncertainty:
Complexity
 Changing structure - shifting
relationships
– Ecosystems have more than one
equilibrium, each determined by a different
set of key processes
– Shifts in dominance of key processes
cause shifts from one equilibrium to
another.
52
P Dir ect I nput
to L ake
Sediment ation
F r action
-
P I nput t o Soil P hosphor us P diffusion Soil
in Soil
to L ake
P Neutr a lization
P Neutr a lization
F r action
P ation
Sediment
L ake Dep th &
M or phom etr y
P hosphor us
in Water
Wind Stor m
Water M ixing
P flushed out
P R esus pens ion
& Recy cling
P hy topla nktiv or es
B ottom
feeder F is h
A lgal Bloom
& T ur bidity
B iom ass in
A lgae
Turbid Algal
state processes
P hosphor us
in Sedime nt
M ax P
R ecy cle
R ate
P F lush F r action
Shallow Lake Model
Different sets of
Processes dominate
To maintain each
Stability domain.
-
B iom ass &
Nutr ients
B iom ass in
M acr ophy te
s
P B ur ial
P B ur ial fr action
Clear water Macrophyte
state processes
53
Ways of explaining reality
Events
Patterns, Trends
Systemic Structures
Mental Models
What just happened?
What’s been happening?
Have we been here or some place similar before?
What are the forces at play contributing to
these patterns?
What about our thinking allows this situation
to persist?
54
New challenges for toxicology
From Endocrine Disruptor Chemicals
1. EDC’s have many counter-intuitive properties:
• threshold assumption and non-monotonic effects
Toxic
Non-Toxic
effect
dosage
Conventional
Assumption
Observed
Dose-Effect Relations
• Synergy
At a certain dosage, Chemical A is safe by itself but toxic in
combination with Chemical B
Why Panarchy Theory?
 Rationalize the interplay between:
– Predictable and unpredictable
– Evolutionary change and persistence
 Explore the world where different
variables are nested inside of one
another and change at different scales
in space and time.
56
Panarchy
A Cross-scale Nested Set of Adaptive Cycles
57
Why Panarchy Theory?
To Account for Dynamics
 Within a level
– Adaptive cycle describes the engine of
novelty, Creative Destruction, and renewal
or reorganization.
 Between levels
– Revolt – the cascade upward of tiny events
– Remember – the context of the next larger
level at climax constrains the next smaller
level in times of renewal
58
Cross-Scale Interatcions:
Revolt and Remember
59
Surprise in Florida
Bay
A
Sea grass
Clear Water
B
Muddy Water
Algae Blooms
Florida Bay
60