Complexity of Coupled Human and
Natural Systems
Jianguo Liu,1
* Thomas Dietz,2
Stephen R. Carpenter,3
Marina Alberti,4
Carl Folke,5,6
Emilio Moran,7
Alice N. Pell,8
Peter Deadman,9
Timothy Kratz,10
Jane Lubchenco,11
Elinor Ostrom,12
Zhiyun Ouyang,13
William Provencher,14
Charles L. Redman,15
Stephen H. Schneider,16
William W. Taylor1
Integrated studies of coupled human and natural systems reveal new and complex patterns and
processes not evident when studied by social or natural scientists separately. Synthesis of six case
studies from around the world shows that couplings between human and natural systems vary
across space, time, and organizational units. They also exhibit nonlinear dynamics with thresholds,
reciprocal feedback loops, time lags, resilience, heterogeneity, and surprises. Furthermore, past
couplings have legacy effects on present conditions and future possibilities.
C
oupled human and natural systems are integrated
systems in which people interact
with natural components. Although many
studies have examined human-nature interactions
(1–5), the complexity of coupled systems has not
been well understood (6, 7). The lack of progress is
largely due to the traditional separation of ecological
and social sciences (8). Although some scholars
have studied coupled systems as complex adaptive
systems (9, 10), most of the previous work has
been theoretical rather than empirical.
An increasing number of interdisciplinary
programs have been integrating ecological and
social sciences to study coupled human and natural
systems (e.g., social-ecological systems and
human-environment systems). Here, we synthesize
six case studies to demonstrate the approaches
used and results found (Fig. 1 and table S1).
These studies are on five continents: the Kenyan
Highlands in Africa (Kenya); the Wolong Nature
Reserve for giant pandas in China (Wolong);
Central Puget Sound of Washington (Puget
Sound) and Northern Highland Lake District
of Wisconsin (Wisconsin) in the United States;
an area near Altamira, State of Pará, Brazil
(Altamira); and Kristianstads Vattenrike of Sweden
(Vattenriket) (Fig. 1). They include urban
(Puget Sound), semi-urban (Vattenriket), and
rural areas (Altamira, Kenya, Wisconsin, and
Wolong), and they are in developed countries
(Puget Sound, Wisconsin, and Vattenriket) and
developing countries (Altamira, Kenya, and
Wolong). These studies are in different ecological,
socioeconomic, political, demographic, and
cultural settings, and they encompass a variety of
ecosystem services and environmental problems
(table S1).
These studies share four major features. First,
they explicitly address complex interactions and
feedback between human and natural systems.
Unlike traditional ecological research that often
excluded human impacts or social research that
generally ignored ecological effects, these studies
consider both ecological and human components
as well as their connections. Thus, they measure
not only ecological variables (e.g., landscape
patterns, wildlife habitat, and biodiversity) and
human variables (e.g., socioeconomic processes,
social networks, agents, and structures of multilevel
governance) (11), but also variables that
link natural and human components (e.g., fuelwood
collection and use of ecosystem services).
Second, each study team is interdisciplinary, engaging
both ecological and social scientists
around common questions. Third, these studies
integrate various tools and techniques from ecological
and social sciences as well as other disciplines
such as remote sensing and geographic
information sciences for data collection, management,
analysis, modeling, and integration (11–15)
(table S1). Fourth, they are simultaneously contextspecific
and longitudinal over periods of time
long enough to elucidate temporal dynamics. As
such, these studies have offered unique interdisciplinary
insights into complexities that cannot
be gained from ecological or social research alone.
Reciprocal Effects and Feedback Loops
In coupled human and natural systems, people
and nature interact reciprocally and form complex
feedback loops. For example, local residents
in Wolong use forests as fuelwood for cooking
and heating. As forests near households were depleted
due to fuelwood collection (16), local residents
had to collect fuelwood from areas far
away (17). Because these forests are bamboo
forests (habitat for the endangered giant panda)
and the bamboo in the forests is the staple food
for the panda, fuelwood collection has led to
substantial deterioration in forests and panda
habitat (16). To prevent further degradation and
restore panda habitat, the Chinese government
began to implement three major conservation
policies several years ago, which help both local
residents and panda habitat. In Kenya, local
residents convert forests into cropland and intensively
cultivate land without supplying additional
nutrients, in some cases for more than 100 years.
Soil degradation with the resulting decreases in
crop yields and greater food insecurity hastens
conversion of remaining forests to agriculture.
Similarly, in Altamira, 255,739 hectares (ha) of
forests had been converted into pasture and
cropland as of 2003. As soil quality declines,
fertilizers must be applied, crops are shifted to
those with lower nutrient requirements, or more
forests are converted into cropland (there were
still 136,913 ha of forested area in 2003).
Feedback between human and natural systems
in the agricultural and tourism sectors of
developed countries is in many ways similar to
feedback in developing countries. For example,
local people (76,000 in 2005) in Vattenriket benefit
from ecosystem services that are the result of
long-term human management of the agricultural
landscape. In Wisconsin, ecosystem conditions
affect tourism, which is the mainstay of the economy,
but economic development and ecosystem
exploitation from tourism often degrade the qualities
that attract tourists.
The ecological and socioeconomic patterns
and processes in urban coupled systems are different
from those in rural areas. They are mediated
by factors such as the urban form, built
infrastructure, and location and consumption preferences
of heterogeneous households and businesses.
For example, in Puget Sound, a distinctive
spatial heterogeneity can be observed across an
urban to rural gradient in relation to diverse
development patterns (18). Land-cover changes
influence biophysical processes (e.g., water purification)
and stream biotic integrity (15). Furthermore,
changes in land cover due to development
in turn affect land value and real estate markets, as
evidenced by values of real estate having up to a
6.5% premium associated with forest cover (19).
REVIEW
1
Center for Systems Integration and Sustainability, Department
of Fisheries and Wildlife, Michigan State University,
East Lansing, MI 48824, USA. 2
Environmental Science and
Policy Program, Michigan State University, East Lansing, MI
48824, USA. 3
Center for Limnology, University of Wisconsin,
Madison, WI 53706, USA. 4
Department of Urban Design and
Planning, University of Washington, Seattle, WA 98195,
USA. 5
The Beijer International Institute of Ecological Economics,
The Royal Swedish Academy of Sciences, Post Office
Box 50005, SE 104 05 Stockholm, Sweden. 6
Stockholm
Resilience Centre, Stockholm University, SE 106 91 Stockholm,
Sweden. 7
Department of Anthropology, Indiana University,
Bloomington, IN 47405, USA. 8
Cornell International
Institute for Food, Agriculture, and Development, Cornell
University, Ithaca, NY 14853, USA. 9
Department of Geography,
University of Waterloo, Waterloo, Ontario N2L 3G1,
Canada. 10
Trout Lake Station, University of Wisconsin,
Boulder Junction, WI 54512, USA. 11
Department of Zoology,
Oregon State University, Corvallis, OR 97331, USA.
12
Workshop in Political Theory and Policy Analysis, Indiana
University, Bloomington, IN 47408–3895, USA. 13
Research
Center for Eco-Environmental Sciences, Chinese Academy of
Sciences, Beijing, 100085, China. 14
Department of Agricultural
and Applied Economics, University of Wisconsin,
Madison, WI 53706, USA. 15
Global Institute for Sustainability,
Arizona State University, Tempe, AZ 85287–3211,
USA. 16
Department of Biological Sciences, Stanford University,
Stanford, CA 94305–5020, USA.
*To whom correspondence should be addressed. E-mail:
jliu@panda.msu.edu
www.sciencemag.org SCIENCE VOL 317 14 SEPTEMBER 2007 1513
onSeptember18,2019http://science.sciencemag.org/Downloadedfrom
Dynamics of human-nature systems are influenced
by many factors, including government
policies and contextual factors in which local
processes are shaped by larger-scale and ultimately
global-scale processes (20). Both markets
and governance can cause decisions made in one
place to affect people and ecosystems far away.
For instance, economic opportunities in cities
attracted many local residents from Wolong to
work in cities in the past several years, thus reducing
fuelwood collection and consumption.
Compared with the migrant workers in cities
from Wolong, however, more than a thousand
times more tourists from around the world increased
the demand for fuelwood through consuming
local products, whose production may
require fuelwood and electricity.
Nonlinearity and Thresholds
Numerous relationships in coupled systems are
nonlinear. In Wisconsin, for instance, fallen trees
that provide critical fish habitat in lakes and
streams drastically decrease when housing density
exceeds about seven houses per kilometer of
shoreline (Fig. 2). Bird richness in the Puget
Sound landscape with single-family housing and
fragments of native forest increases nonlinearly
with forest cover and peaks when 50 to 60% of
the land is forested (Fig. 3) (21).
Thresholds [transition points between alternate
states (22)] are common forms of nonlinearity.
In Vattenriket, an intentional participatory
process mobilized stakeholders laying the groundwork
for a shift from conventional management
to adaptive co-management (23). Cultural values
and environmental concerns prompted local stakeholders
to build new knowledge, develop new
visions and goals, and create new social networks.
The result of these community activities was a
new and more suitable governance system of
adaptive co-management of the landscape.
System behaviors shift from one state to another
over time (temporal thresholds) and across
space (spatial thresholds). Altamira depicts a temporal
threshold, whereas Wolong demonstrates a
spatial threshold. Deforestation rates in Altamira
are high during the first 5 to 7 years of settlement
and then decrease rapidly. In Wolong, as the distance
between locations of households and fuelwood
collection sites increases, panda habitat
decreases, reaching a minimum at a distance of
approximately 1800 m (17). When the distance
between households and fuelwood supplies is
small, the total area for fuelwood collection is
small and thus panda habitat is better protected.
When the distance is large (>1800 m), fuelwood
collection is scattered throughout a large region
and affected areas can recover relatively quickly.
When the distance is approximately 1800 m, local
residents’ fuelwood demand is met by cutting
most available trees and causes more habitat loss
(17).
Surprises
When complexity is not understood, people
may be surprised at the outcomes of humannature
couplings. For example, smelt (Osmerus
mordax) was initially introduced to Wisconsin
as a prey species for game fish such as walleyes
(Stizostedion vitreum), but smelt ate juvenile
walleyes leading to loss of walleye populations.
In Puget Sound, growth management policy
has caused urban density to intensify inside the
urban growth boundary while unintentionally
facilitating sprawl outside the urban growth
boundary.
Conservation policies can also generate unintended
perverse results. In Wolong, for instance,
high-quality panda habitat degraded faster
after the area was established as a reserve than
before the reserve’s creation (24). To prevent
Fig. 1. Map highlighting major attributes of the six coupled human and natural systems (location; spatial extent; population size; and ecological, economic, and
administrative attributes). To save space, short names within the parentheses represent the coupled systems. See table S1 for more detailed descriptions.
14 SEPTEMBER 2007 VOL 317 SCIENCE www.sciencemag.org1514
REVIEW
onSeptember18,2019http://science.sciencemag.org/Downloadedfrom
further degradation, a natural forest conservation
program was introduced in 2001 for local residents
to monitor illegal harvesting. Unexpectedly,
a large number of new households formed in
2001 because many households decided to split
into smaller ones to more effectively capture
subsidies (20 to 25% of the average household
income) given to households as part of the program.
The household proliferation and reduction
in household size (number of people in a household)
increased demand for fuelwood and land
for house construction (25).
Some ecosystems can only be sustained
through human management practices, whereas
many conservation efforts preclude such human
interference. For example, the wetland site under
the Ramsar Convention (an international treaty
for the conservation and sustainable use of wetlands)
in Vattenriket was set aside for conservation
purposes, but the wetland became overgrown
when grazing was halted. This unintended consequence
led to an understanding of grazing as
essential to maintaining this wetland system (23).
Legacy Effects and Time Lags
Legacy effects are impacts of prior human-nature
couplings on later conditions. Among the six
sites, legacies vary in duration from decades to
centuries. The shortest legacy is in Altamira, a
frontier area where the land tenure system imposed
by the government in 1970 still shapes the
present spatial pattern of land-cover change,
human population distribution, and human
activities. The longest legacy is in Vattenriket,
where the landscape has been affected by human
actions such as using wet grasslands over
hundreds of years.
Legacy durations in the other sites fall
somewhere in between. In Wolong, current forest
types in areas at lower elevations
(1200 to 3000 m) are
shaped by forest harvesting
three to nine decades ago. Introduction
of a keystone species
can restructure a fish population
for decades or longer, as has
been demonstrated in the Wisconsin
study area which started
fish stocking in the 1930s.
Long-term (up to 100 years)
continuous cultivation in Kenya
has decreased crop yields, with
most of the degradation occurring
during the first 15 to 20
years after conversion from forest
to agriculture. In Puget
Sound, landscape patterns are
influenced by infrastructure built
decades or even a century ago.
The ecological and socioeconomic
impacts of humannature
couplings may not be
immediately observable or predictable
because of time lags
between the human-nature
interactions and the appearance of ecological
and socioeconomic consequences. In Kenya,
there is a time delay between investment in soil
improvement and increases in income. In
Vattenriket, the city of Kristianstad stopped
taking its drinking water from the Helgeå River
in the 1940s because untreated industrial and
household sewage had accumulated
several decades earlier.
Disturbances to groundwater
quality can take a long time to
appear “downstream” because
groundwater movement between
adjacent lakes can take
centuries. In Puget Sound, ecological
effects of the Growth
Management Act adopted in
the State of Washington in
1990 could not have been observed
in less than 8 years (26).
The length of lags attributable
to a single cause may vary
for different indicators; conversely,
different causes may
become apparent over different
time periods for the same indicator.
The former can be seen in
Altamira, where changes in crop
prices quickly affect planting of
annuals but effects on planting
(or abandonment) of perennials (such as cocoa
and black pepper) often are delayed. As to the
latter situation, changes in the price of electricity
quickly affect panda habitat in Wolong because
of sharp changes in fuelwood demand, but
spacing of births within households has a much
slower effect (27). Energy for cooking is needed
daily and fluctuations in the price for electricity
may quickly force local residents to use more
fuelwood (thus destroying forests and panda
habitat), whereas it takes a longer time for children
to establish new households that increase
demand for energy.
Resilience
Coupled systems have different degrees of
resilience—the capability to retain similar structures
and functioning after disturbances for
continuous development (10, 28, 29). Resilience
can be affected by many factors. In Wolong, for
example, larger areas with fast-growing tree
species are more resilient to fuelwood collection
than are smaller areas with slower-growing trees.
In Kenya, remittances from relatives employed in
urban areas minimize food insecurity due to crop
failures caused by droughts and poor soil fertility.
Human intervention also plays a key role in
maintaining resilience. For instance, in Vattenriket,
sustaining the resilience of the wetland landscape
requires grazing by cattle and incentives to
make grazing economically viable. Partially because
of the actions of environmentalists, Puget
Sound is still home to one of the last intact oldgrowth
forests in the United States despite rapid
urbanization. In Wisconsin, social-ecological
resilience comes from the good condition of
many ecosystems; the intention of Native Americans
to manage their lands and lakes sustainably;
the mosaic of tribal, private, and state ownership;
and innovations in ecosystem management by
various stakeholders (tribal governments of
Native Americans, lake associations, formal
research organizations, and nongovernment
organizations) (12).
Heterogeneity
Human-nature couplings vary across space, time,
and organizational units. The socioeconomic
differences among people in Wisconsin lead to
different choices and behaviors, which in turn
result in very different ecological outcomes than
Fig. 3. Change in avian richness with progressively more forest
(less human settlement) in the human-influenced landscape in
Central Puget Sound region of Washington, United States. [modified
from (21), with permission]
Fig. 2. The relationship between fish habitat (logs per kilometer) and
house density in the Northern Highland Lake District of Wisconsin,
United States. [modified from (38), with permission]
www.sciencemag.org SCIENCE VOL 317 14 SEPTEMBER 2007 1515
REVIEW
onSeptember18,2019http://science.sciencemag.org/Downloadedfrom
one would find were everyone to have the same
preferences for ecosystem services. In Altamira,
different settlement cohorts follow similar trajectories
of land use, but the magnitude of changes
in important variables like rates of deforestation
varies as a product of exogenous and endogenous
factors (e.g., local, regional, and global political
economy) (30). For the Kenyan highlands, it is
common to find families with soils of different
quality and as a result, different crop yields.
Coupled human-natural systems are not
static; they change over time. Although the human
population sizes have increased in all six
study sites over the past several decades, the
resultant ecological impacts have differed. In the
Kenya study area, human population size has
doubled over the past 30 years, causing a marked
reduction in farm size. Smaller farm size has led
to growing maize during both rainy seasons to
meet family demand, but this practice has
accelerated the rate of soil degradation and increased
poverty. An increase in recreational land
use in Wisconsin led to a 4.6-fold increase in
housing density (from 3.7 to 17.2 units/km2
)
from 1940 to 2000. For Puget Sound, between
1991 and 1999 land area covered by development
increased by 620 km2
(31.5% increase),
while forest cover declined by 714 km2
[10.3%
decline (26)]. Temporal changes take place not
only inside a coupled system, but also across its
boundaries. In Wolong, a rapidly increasing number
of domestic and foreign tourists have made
the system much more tightly coupled to the
national and global economy.
Spatial variations exist in all coupled systems.
For example, more fuelwood is collected in areas
of Wolong with easy access and little enforcement
than in forested areas with more
challenging topography or strict enforcement. In
Vattenriket, habitats and management practices,
local stewardship associations, social networks,
and multilevel institutions vary across the landscape
(11). Landscape heterogeneity of Puget
Sound increases with the degree of urbanization,
but differs substantially within the region
depending on urban land-use patterns, infrastructure,
and spatial distribution of activities
(18). In Wisconsin, people have preferentially
settled around lower-elevation lakes, which tend
to have riverine inputs, low to moderate dissolved
organic carbon, low to moderate nutrients,
and relatively diverse sport fish communities (31).
In Altamira, fertile soils permit cultivation of
cocoa and sugar cane, whereas on poorer soils,
pasture and manioc cultivation are more common.
Conclusion and Outlook
Results such as those reviewed here benefit from
and help advance the integration of ecological
and social sciences. The approaches used and the
results from these studies can be applied to many
other coupled systems at local, national, and
global levels. For instance, the finding that the
number of households increased faster than the
human population size in Wolong over the past
three decades has led to the discovery that this
trend is global and is particularly profound in
the 76 countries with biodiversity hotspots (25).
The Lake Futures Project (32) in Wisconsin was
a prototype used to develop approaches for the
Millennium Ecosystem Assessment scenarios
(33).
Comparison of these studies provides important
insights into diverse complex characteristics
that cannot be observed in a single study. The
types of surprises found in the case studies differ,
although all of them originated from the interactions
between human and natural systems. All
six studies have demonstrated legacy effects, but
legacy durations varied from decades to centuries.
Because of the independent nature of these
studies, information from one study is not
necessarily available in or transferable to other
studies. To increase the extent of generalizing
from case studies, future research on coupled
systems must include not only separate sitespecific
studies but also coordinated, long-term
comparative projects across multiple sites to capture
a full spectrum of variations (14, 34, 35).
Furthermore, all the studies in this review focus
on interactions within the system, rather than
interactions among different coupled systems. As
globalization intensifies, there are more interactions
among even geographically distant systems
and across scales (36, 37). Thus, it is critical to
move beyond the existing approaches for studying
coupled systems, to develop more comprehensive
portfolios, and to build an international
network for interdisciplinary research spanning
local, regional, national, and global levels.
References and Notes
1. G. P. Marsh, Man and Nature (Belknap Press of Harvard
Univ. Press, Cambridge, MA, 1864).
2. P. M. Vitousek, H. A. Mooney, J. Lubchenco, J. M. Melillo,
Science 277, 494 (1997).
3. National Research Council, Our Common Journey
(National Academy Press, Washington, DC, 1999).
4. W. L. Thomas Jr., Man’s Role in Changing the Face of the
Earth (Univ. of Chicago Press, Chicago, IL, 1956).
5. B. L. Turner II et al., The Earth as Transformed by Human
Action: Global and Regional Changes in the Biosphere
over the Past 300 Years (Cambridge Univ. Press,
Cambridge, UK, 1990).
6. S. H. Schneider, R. Londer, The Coevolution of Climate
and Life (Sierra Club Books, San Francisco, CA, 1984).
7. F. Berkes, J. Colding, C. Folke, Eds., Navigating
Social-Ecological Systems: Building Resilience for
Complexity and Change (Cambridge Univ. Press,
Cambridge, UK, 2003).
8. E. A. Rosa, T. Dietz, Int. Soc. 13, 421 (1998).
9. S. A. Levin, Fragile Dominion: Complexity and the
Commons (Perseus Publishing, Cambridge, MA,
1999).
10. L. H. Gunderson, C. S. Holling, Eds., Panarchy:
Understanding Transformations in Human and Natural
Systems (Island Press, Washington, DC, 2002).
11. L. Schultz, C. Folke, P. Olsson, Environ. Conserv. 34, 140
(2007).
12. S. R. Carpenter et al., Bioscience 57, 323 (2007).
13. J. Liu et al., in People and the Environment: Approaches
for Linking Household and Community Surveys to Remote
Sensing and GIS, J. Fox, V. Mishra, R. Rindfuss, S. Walsh,
Eds. (Kluwer Academic Publishers, Boston, MA, 2003),
pp. 241–263.
14. E. F. Moran, E. Ostrom, Seeing the Forest and the Trees:
Human Environment Interactions in Forest Ecosystems
(MIT Press, Cambridge, MA, 2005).
15. M. Alberti et al., Landsc. Urban Plan. 80, 345 (2007).
16. J. Liu et al., Conserv. Biol. 13, 1360 (1999).
17. L. An, M. Linderman, A. Shortridge, J. Qi, J. Liu, Ann.
Assoc. Am. Geogr. 95, 54 (2005).
18. M. A. Alberti, Int. Reg. Sci. Rev. 28, 168 (2005).
19. M. D. Oleyar, A. L. Greve, J. C. Withey, A. M. Bjorn, (www.
urbaneco.washington.edu/publications/oleyar_etal.pdf).
20. F. Berkes et al., Science 311, 1557 (2006).
21. J. Marzluff, Urban Ecosyst. 8, 157 (2005).
22. W. A. Brock, in Punctuated Equilibrium and the Dynamics
of U.S. Environmental Policy, R. Repetto, Ed. (Yale Univ.
Press, New Haven, CT, 2006), pp. 47–77.
23. P. Olsson, C. Folke, T. Hahn, Ecol. Soc. 9(4), 2 (2004).
24. J. Liu et al., Science 292, 98 (2001).
25. J. Liu, G. Daily, P. Ehrlich, G. Luck, Nature 421, 530 (2003).
26. M. Alberti et al., 2005 Landscape Benchmarks
(Washington State CTED, Olympia, WA, 2005).
27. L. An, G. M. He, Z. Liang, J. Liu, Biodivers. Conserv. 15,
2343 (2006).
28. B. Walker, A. J. Meyers, Ecol. Soc. 9(2), 3 (2004).
29. C. Folke, Glob. Environ. Change 16, 253 (2006).
30. S. D. McCracken et al., Photogrammetric Eng. Remote
Sens. 65, 1311 (1999).
31. J. J. Magnuson, T. K. Kratz, B. J. Benson, Eds., Long-Term
Dynamics of Lakes in the Landscape (Oxford Univ. Press,
Oxford, UK, 2006).
32. The Lake Futures Project (http://lakefutures.wisc.edu).
33. Millennium Ecosystem Assessment (www.MAweb.org).
34. P. Olsson et al., Ecol. Soc. 11(1), 18 (2006).
35. S. R. Carpenter, C. Folke, Trends Ecol. Evol. 21, 309
(2006).
36. J. Liu, J. Diamond, Nature 435, 1179 (2005).
37. O. Young et al., Glob. Environ. Change 16, 304 (2006).
38. G. G. Sass et al., Fisheries 31, 321 (2006).
39. We thank T. Baerwald for his support and inspiration for
this paper. We also thank R. May and three anonymous
reviewers for constructive comments and suggestions;
E. Brondizio, N. Reo, and L. VanWey for their helpful
input; and S. Li for her able editorial assistance. We
acknowledge J. Marzluff for providing data for Fig. 3 as
well as numerous collaborators and students for their
contributions to data collection and analysis in the
projects described here. Financial support was provided
by NSF (Dynamics of Coupled Natural and Human
Systems, and North Temperate Lakes Long-Term
Ecological Research site), NIH, National Aeronautics and
Space Administration, National Natural Science
Foundation of China, Michigan State University (Michigan
Agricultural Experimental Station, Rachel Carson Chair in
Sustainability, University Distinguished Professorship,
and Environmental Research Initiative), Swedish Research
Council for the Environment, Agricultural Sciences and
Spatial Planning, and Swedish Foundation for Strategic
Environmental Research.
Supporting Online Material
www.sciencemag.org/cgi/content/full/317/5844/1513/DC1
Table S1
10.1126/science.1144004
14 SEPTEMBER 2007 VOL 317 SCIENCE www.sciencemag.org1516
REVIEW
onSeptember18,2019http://science.sciencemag.org/Downloadedfrom
Complexity of Coupled Human and Natural Systems
Schneider and William W. Taylor
Timothy Kratz, Jane Lubchenco, Elinor Ostrom, Zhiyun Ouyang, William Provencher, Charles L. Redman, Stephen H.
Jianguo Liu , Thomas Dietz, Stephen R. Carpenter, Marina Alberti, Carl Folke, Emilio Moran, Alice N. Pell, Peter Deadman,
DOI: 10.1126/science.1144004
(5844), 1513-1516.317Science
ARTICLE TOOLS http://science.sciencemag.org/content/317/5844/1513
MATERIALS
SUPPLEMENTARY http://science.sciencemag.org/content/suppl/2007/09/11/317.5844.1513.DC1
REFERENCES
http://science.sciencemag.org/content/317/5844/1513#BIBL
This article cites 22 articles, 3 of which you can access for free
PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions
Terms of ServiceUse of this article is subject to the
is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title
(print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience
American Association for the Advancement of Science
onSeptember18,2019http://science.sciencemag.org/Downloadedfrom