Content uploaded by Richard J.T. Klein
Author content
All content in this area was uploaded by Richard J.T. Klein on Dec 10, 2015
Content may be subject to copyright.
NATURE CLIMATE CHANGE | VOL 3 | APRIL 2013 | www.nature.com/natureclimatechange 305
opinion & comment
COMMENTARY:
Limits to adaptation
Kirstin Dow, Frans Berkhout, Benjamin L. Preston, Richard J.T. Klein, Guy Midgley
and M. Rebecca Shaw
An actor-centered, risk-based approach to defining limits to social adaptation provides a useful analytic
framing for identifying and anticipating these limits and informing debates over society’s responses to
climate change.
An inadequate multilateral response
on reductions in global greenhouse-
gas emissions has resulted in
much greater emphasis on adaptation
to address the growing risks of climate
change. Natural and social systems oen
have signicant capacities to adapt,
including the potential in social systems
for transformative adaptation in response
to climate-related risks1. Global-scale
assessments of vulnerability to climate
change have identied ‘key vulnerabilities’2
and ‘tipping points’3 suggesting thresholds
in biophysical systems that, if exceeded,
would pose major threats to sustainability
and human welfare, and are thus ‘reasons
for concern’4. However, it is dicult to
establish how the rate and magnitude of
climate change and biophysical thresholds
might relate to the capacity to adapt in
social systems. Some adaptation limits
have been clearly identied, primarily for
ecological systems, exemplied by species
extinctions. But little is known about limits
in social systems — whether there are social
limits to adaptation5, what inuences their
likelihood, where these might lie, who they
would aect and what the consequences of
reaching such limits might be.
e existence of adaptation limits
has broad implications. If the capacity
to adapt is unlimited, a key rationale for
reducing emissions of greenhouse gases is
weakened and replaced by considerations of
adaptation costs and benets, and of equity
concerns. However, research suggests that
opportunities and resources to adapt may
be nite for many social actors, whether
these are individual households, businesses
or governments6. Breaching adaptation
limits will result in escalating losses or
require transformational change. Hence,
there is an urgent need to identify and
predict where limits are likely to occur in
order to assess and prepare for the potential
consequences. Here we propose a risk-based
approach to dening adaptation limits,
provide two examples, and raise some
implications and research needs highlighted
by this perspective.
e utility of the current literature on
adaptation limits is weakened by ambiguity.
Terms such as thresholds, limits, barriers
and constraints are used interchangeably,
yet their meanings dier. Although an
adaptation barrier or constraint represents
a stressor or an impediment to adaptation
that can in principle be overcome6, an
adaptation limit implies a level of adaptive
capacity, broadly dened, that cannot be
surpassed. Providing a sound conceptual
approach is a necessary rst step to enable
progress in building knowledge about limits.
We believe an actor-centred approach to
dening social adaptation limits will bring
clarity, and can inform practical action.
Adaptation is primarily intended to
reduce climate-related risks to things we
value7. e concept of risk includes extra
elements that are useful in developing
a clear denition of adaptation limits;
notably the consequences, likelihood and
uncertainty of climate-related hazards8.
Risk perceptions inuence a homeowner’s
desire to live in a particular place, a forestry
company’s management strategy in the face
of climate-related hazards or an insurance
company’s unwillingness to provide disaster
coverage. Adaptation processes can be
viewed as attempts to keep risks to valued
objectives — such as a home by the beach or
a protable forestry business — at a tolerable
level in the face of climate-related threats.
We recognize that actors, from individuals
to corporations and governments, may
dier in their perceptions, experiences and
evaluations of risks and in their willingness
to take actions to abate risks. For instance,
farmers in the same region may dier in their
perceptions of the magnitude, consequences
or uncertainty of climate-related risks, in their
tolerances of potential crop and related losses
and in their willingness to shi to new crops
or management practices. Governments may
dier in their approach to crop insurance.
To simplify this complexity, Klinke and
Renn have argued that actors will implicitly
or explicitly place risks to their valued
objectives into one of three categories
involving dierent types of response8,9:
acceptable risks are risks deemed so low that
further eorts in risk reduction (adaptation)
are not justied; tolerable risks relate to
situations where adaptive, risk-reduction
eorts are required for risks to be kept
within reasonable levels10; and intolerable
risks are those which fundamentally
threaten a private or social norm —
threatening, for instance, public safety,
continuity of traditions, a legal standard
or a social contract11 — despite adaptive
action having been taken. On reaching an
intolerable risk level, we normally expect a
discontinuity of behaviour in order to avoid
the risk, whether this is a homeowner’s
decision to move, or a forester selling o
land, as the alternative is increasing losses.
e question of what is acceptable, tolerable
or intolerable remains with the individual
actors, as they shape collective responses.
Figure1 depicts these risk types,
focusing on the relationship between
the frequency and intensity of adverse
impacts. According to this representation,
adaptation occurs within the zone dened
Many communities in highly-
vulnerable regions like the
Arctic are already facing limits
in their capacity to adapt.
A risk-based approach to
limits is scalable, broadly
applicable and readily intuitive
by a broad array of actors.
© 2013 Macmillan Publishers Limited. All rights reserved
306 NATURE CLIMATE CHANGE | VOL 3 | APRIL 2013 | www.nature.com/natureclimatechange
opinion & comment
as representing a tolerable risk. For instance,
a farmer seeking to cultivate a specic crop
under increasingly stressed water resources
will invest in available adaptation options to
raise the eciency of water use, increasing
adaptive eort as access to water resources
becomes more constrained. At some point,
adaptation eort under the existing regime
will become disproportionate to the benets
and a new adaptation action, such as
irrigation, is needed to maintain a farming
livelihood. is new adaptation would
allow farming or other valued objectives
to continue. But, at some point, no new
adaptation options are available to respond
to growing risks, or the level of adaptive
eort required to maintain valued objectives
becomes infeasible. At this point the farmer
may, for example, choose to abandon
farming altogether.
e blurred boundaries between the
zones of risk seen as acceptable, tolerable
and intolerable signify the diversity among
actors and the potential for debate over the
level and distribution of risk. In the absence
of new adaptation options or resources, the
threshold for intolerable risks represents a
point at which an actor must either live with
the risk of escalating loss and damage12,
or transform behaviour to avoid the risk1.
Such a discontinuity in risk or behaviour
is symptomatic of an adaptation limit
being reached.
We therefore propose a denition of an
adaptation limit as a point at which an actor
can no longer secure valued objectives from
intolerable risk through adaptive action.
Two rather dierent examples illustrate the
denition. First, we take rice cultivation in
South Asia. Rice pollination and owering
has a threshold temperature of 26°C
(at night), with a 10% decline in yield for
every 1°C increase in temperature above
that13. In this example, the adaptation limit
is established by the inability to breed rice
varieties that pollinate with night-time
temperatures above the 32–35°C range14.
e valued objective is to produce rice as a
staple crop and for export. e intolerable
risk is a level of loss in rice production,
farmer livelihoods, income from exports
and food security. Rising temperatures
increase the future probability that rice
harvests may fail.
Such failures would probably impose
economic losses on farmers, as well as
generate broader economic and political
impacts. If no aordable alternative supplies
of rice can be found, it could entail excessive
costs to consumers and/or changing
dietary practices. e increasing threat
of these impacts could lead farmers and
policymakers to change long-established
practices of rice cultivation and pursue
security though alternative crops. For many
reasons, change in response to limits can be
a complex process. e tolerable degree of
risk diers among actors and debate could
slow collective adaptive action. e level
of disruption caused by escalating losses
and discontinuities at the limit boundary
could potentially be mediated by designing
eective adaptation processes for managing
change associated with the limit. Better
recognition of the limit could reduce losses
during such a transition, whereas failing
to address the risk as a limit is reached
could result in catastrophic economic and
social costs.
Our second example is at the societal
level, and concerns a risk to cultural
continuity7. In the mid- to late eenth
century, aer about 400years of settlement,
the complex and vibrant Norse Greenland
society came to an end. is is oen seen
as a failure to adapt to climatic changes in
the ‘Little Ice Age’. In fact, the story of this
collapse represents an example of limits to
adaptation. Norse Greenlanders adapted
in a variety of ways by shiing to new
ways of exploiting marine mammals as
harsh climatic conditions forced declines
in agriculture and domestic livestock
production15. But faced with growing
competition from Inuit hunters, declining
trade in ivory and fur with Norway, and a
worsening climate, these adaptations were
insucient to maintain risks to community
continuity at tolerable levels.
We believe an approach to dening
adaptation limits linked to the (in)
tolerability of risks is useful because it
engages with the social, institutional and
cultural contexts shaping adaptation and
risk. It also incorporates the role of social,
economic and cultural values in dening
adaptation limits5,7. By starting from the
perspective of social actors, this approach
recognizes that adaptation limits always
need to be dened from the perspective of
a specic actor facing the loss, be that an
individual, community or region. is also
applies to global-scale biophysical changes
articulated in the key vulnerabilities2
described by the Intergovernmental Panel
on Climate Change . Identifying ‘dangerous
climate change’ relies on understanding
which actors will be aected and whether
they are facing adaptation limits. As the rice
and Norse cases show, a risk-based approach
to limits is scalable, broadly applicable and
readily intuitive by a broad array of actors.
Focusing on the potential intersection
of intolerable risks and valued objectives
directs our attention beyond the biophysical
impacts to identifying the broader social
dimensions of potential losses. is will be
part of accounting for, debating over and
allocating resources at adaptation limits.
Given the diculties in determining
limits to adaptation, there is an urgent need
for research in key domains — including
agriculture, water resources management
and disease control — to determine
where limits may exist so that actors may
anticipate and plan to mediate the hardships
that cannot be avoided. e capacities
to provide early warnings and to operate
Very frequentVery rare
Frequency of adverse impact
Intolerable risks
Intensity of adverse impact Catastrophic Negligible
Acceptable risks
Tolerable risks
A
d
a
p
t
a
t
i
o
n
l
i
m
i
t
L
i
m
i
t
o
f
a
c
c
e
p
t
a
b
l
e
r
i
s
k
Figure 1 | Acceptable, tolerable and intolerable risks in relation to adaptation limits. Figure drawn by
Yuk a Es t ra d a, IP CC , ba s ed on re f . 8 .
© 2013 Macmillan Publishers Limited. All rights reserved
NATURE CLIMATE CHANGE | VOL 3 | APRIL 2013 | www.nature.com/natureclimatechange 307
opinion & comment
across scales are two important features of
such eorts. Beyond these, a concern for
adaptation limits draws attention to the
design, capabilities and trust in institutions
needed to implement risk management
in extremis. Many communities in highly
vulnerable regions — such as the Arctic —
are already facing limits in their capacity
to adapt, and losses that are dicult
to compensate for. As climate change
accelerates, increasingly more communities,
regions and sectors will begin to approach
these limits. We need to be aware of this
gathering storm of crises, the potential
for changes in risk and behaviour at limits
and the likelihood that those changes will
generate challenging debates. Researchers
need to begin making progress in predicting
and anticipating adaptation limits, and
policymakers need to start making plans for
managing the consequences of exceeding
adaptation limits.❐
Kirsti n Dow 1*, Frans Berkhout2, Benjamin Preston3,
Richard Klein4,5, Guy Midgley6 and M. Rebecca Shaw 7,8
are in the 1Department of Geography, University of
South Carolina, Callcott Building 709 Bull Street,
Columbia, South Carolina 29208, USA, 2Institute
for Environmental Studies, VU University,
De Boelelaan 1085, Amsterdam, 1081HV,
Nether lands, 3Oak Ridge National Laboratory,
One Bethel Valley Road, PO Box2800, MS-6301,
Oak Ridge, Tennessee 37831-6301, USA, 4Stockholm
Environment Institute, Kräriket 2B, 106 91
Stockholm, Sweden, 5Centre for Climate Science
and Policy Research and Department of ematic
Studies, Linköping University, Sweden, 6South African
National Biodiversity Institute, Kirstenbosch, P/Bag x7,
Claremont 7735, Cape Town, South Africa, 7School
of Agricultural, Earth, and Environmental Sciences,
University of Kwazulu-Natal, Pietermaritzburg,
P/Bag x101, Scottsville 3209, South Africa,
8Environmental Defense Fund, 123 Mission St,
28th Floor, San Francisco, California 94105, USA.
*e-mail: kdow@sc.edu
References
1. Kates, R.W., Travis, W.R. & Wilbanks, T.J.
Proc. Natl Acad.Sci. USA 109, 7156–7161 (2012).
2. Schneider, S.H. et al. in IPCC Climate Change 2007: Impacts,
Adaptation and Vulnerability (eds Parry, M.L., Canziani, O.F.,
Palutikof J.P., van der Linden, P.J. & Hanson, C.E.) 779–810
(Cambridge Univ. Press, 2007).
3. Lenton, T. et al. Proc. Natl Acad. Sci. USA 105, 1786–1793 (2008).
4. Smith, J.B. et al. Proc. Natl Acad. Sci. USA
106, 4133–4137 (2009).
5. Adger, W.N. et al. Climatic Change 93, 335–354 (2009).
6. Moser, S.C. & Ekstrom, J.A. Proc. Natl Acad. Sci. USA
107, 22026–22031 (2010).
7. Adger, W.N., Barnett, J., Brown, K., Marshall, N. & O’Brien, K.
Nature Clim. Change 3, 112–117 (2012).
8. Klinke, A. & Renn, O. Risk Anal. 22, 1071–1094 (2002).
9. Renn, O. Risk Governance: Coping with Uncertainty in a Complex
Wor ld (Earthscan, 2008).
10. Hall, J.W., Brown, S., Nicholls, R.J., Pidgeon, N.F.
& Watson, R.T. Nature Clim. Change 2, 833–834 (2012).
11. Adger, W.N., Quinn, T., Lorenzoni, I., Murphy, C. & Sweeney, J.
Nature Clim. Change 3, 330–333 (2013).
12. Verheyen, R. & Roderick, P. Beyond Adaptation: e Legal
Duty to Pay Compensation for Climate Change Damage
(WWF-UK, 2008).
13. Peng, S. etal. Proc. Natl Acad. Sci. USA 101, 9971–9975 (2004).
14. IPCC Climate Change 2001: Synthesis Report
(eds Watson, R.T. etal.) (Cambridge Univ. Press, 2001).
15. Dugmore, A.J. etal. in Adapting to Climate
Change: resholds, Values, Governance
(eds Adger, W.N., Lorenzoni,I. & O’Brien, K.L.)
96–113 (Cambridge Univ. Press, 2009).
Acknowledgements
e authors would like to thank Mozaharul Alam,
Habiba Gitay, Susanne Moser, James urlow, and
Koko Warner for their constructive comments.
© 2013 Macmillan Publishers Limited. All rights reserved
