Articles
Water-Use Sustainability in
Socioecological Systems: A
Multiscale Integrated Approach
Cristina Madrid, Violeta Cabello, and Mario Giampietro
Human societies and ecosystems use water in different ways and at different scales, which complicates the study of water-use sustainability in
socioecological systems. We present a multiscale integrated assessment of societal and ecosystem metabolism, an innovative approach to the
quantitative analysis of water use that addresses the problem of multiple scales. It builds on the concept of metabolic pattern and the flow–fund
model of Georgescu-Roegen. We show how to define water resources and water use (expressed in hourly rates) for socioeconomic systems in
relation to the identities of relevant fund elements (relevant categories of human activity or land use) over a time span of 1 year. Similarly, we
define the limits on the human appropriation of water (aggregate withdrawal or damping per year) on the basis of the structural and functional
stability of ecological funds (defined over a much longer time scale) and the related land-use pattern.
Keywords: water metabolism, water-use sustainability, multiscale assessment, socioecosystems, MuSIASEM
W
ater scarcity is a socially driven phenomenon, fed by what Allan (2011) called a “deep social delusion”
about water that manifests itself as a mismatch between
physical water availability and societal water use. A clear
example of this is the location of intensive-irrigation agricultural land or human settlements in deserts. The physical
availability of water is determined by broadscale natural
dynamics, whereas its use is determined by societal dynamics operating at a much narrower scale. Governance of water
resources also has consequences at various scales (Laborte
et al. 2007). As a result, quantitative analyses based on the
consideration of only a single scale are unlikely to provide
an input for the management of water resources that is
adequate for both natural and social systems.
Several quantitative methods have been proposed to link
societal water use to ecosystem impact. However, none of
these methods explicitly addresses the issue of scale. The
problem is twofold. First, these quantitative methods are
deeply rooted in a reductionist approach to the definition
of water. In fact, water has different meanings for different
actors and scientists in different contexts and levels of analysis. Second, these methods do not bridge the results of the
analyses at different levels.
An integrated assessment of water use should be able
to address the shifting identity of a water mass perceived
differently in different contexts. For example, a typical classification of water may define it as either blue or green, depending on its origin (evaporated from surface or groundwater or
stored in the soil as moisture, respectively; Falkenmark 1995).
However, it is doubtful that this characterization (1 cubic
hectometer [hm3] of blue water or 1 hm3 of green water) will
always be useful when moving across different contexts and
levels of analysis. For example, if we need to check the number of jobs created in the paper sector per liter of water used,
the origin of the water is usually not relevant. Yet, the origin
of water is relevant for assessing the impact inflicted on the
ecosystem. Water extracted from ecosystems can certainly be
blue or green, but the potential usefulness of this classification will always depend on both the purpose of the analysis
and the final use of this water. Given that social systems can
control only (blue) freshwater, this has traditionally been the
main indicator used to characterize water use within social
systems, and there has been a systemic tendency to ignore
green water.
The theoretical concept of virtual water (Allan 1998),
beyond its operative characterization as the amount of water
needed to produce a good or a service, is a step toward the
adoption of a more flexible definition of water that is able to
connect different perceptions referring to different systems.
In fact, by adopting this concept, one can establish a water
BioScience 63: 14–24. ISSN 0006-3568, electronic ISSN 1525-3244. © 2013 by American Institute of Biological Sciences. All rights reserved. Request
permission to photocopy or reproduce article content at the University of California Press’s Rights and Permissions Web site at www.ucpressjournals.com/
reprintinfo.asp. doi:10.1525/bio.2013.63.1.6
14 BioScience • January 2013 / Vol. 63 No. 1
www.biosciencemag.org
Articles
interpreted as the possibility of reducing two distinct scales of analysis—the
distinct scales required to provide useful representations of ecosystem and
societal processes, for example—into a
Policy
single quantitative representation.
Building on the concepts of selfSelection of
goals, definition
organizing
dissipative systems (Prigogine
of priorities,
1978), autopoiesis (Maturana and Varela
evaluation of
results
1980), and evolutionary theory (Weber
Political ecology,
et al. 1988, Brooks et al. 1989, Depew
policy studies,
and Weber 1989), pioneers of theoscience for
governance
retical ecology (Margalef 1968, Odum
Social multicriteria
1971, Ulanowicz 1986) have developed
evaluation
various methodological approaches to
Conflict and
analyzing, in quantitative terms, the
institutional
patterns of the flow exchange of matanalysis
ter and energy within ecosystems. The
rationales of self-organization and
autopoiesis are useful for establishing a set of expected
relationships between the characteristics of the various
parts and subparts and those of the whole ecosystem across
hierarchical levels, as is exemplified by the classic ecological
representation of trophic structures.
In a similar line of thinking, the concept of societal
metabolism was introduced in the first half of the twentieth century as a means to study the link between energy
use and the expression of societal structures and functions
(Zipf 1941, White 1943, Cottrell 1955). Given the growing
concern with sustainability, this concept has reemerged in
various forms in the more recent literature (Martinez-Alier
and Schlüpmann 1987, Fischer-Kowalski 1998, Giampietro
et al. 2011). In this case, a set of expected relations between
the characteristics of the parts and those of the whole socioeconomic system can also be established when looking at the
metabolic patterns of societies (Giampietro et al. 2012).
That is, both society and ecosystems can be interpreted
as complex, self-organizing, dissipative systems capable
of stabilizing their own identity by reproducing a defined
metabolic pattern (Giampietro et al. 2011). However, the
mechanisms and processes involved in the expression of
their respective metabolic patterns are operating at different
scales. Therefore, it is impossible to perceive and represent
in quantitative terms the process of the self-organization of
SESs using a single scale and a single quantitative approach.
The MuSIASEM approach has been specifically developed
to deal with this impasse. MuSIASEM makes it possible to
analyze the metabolic pattern of energy and material flows
in society and ecosystems using different scales of analysis
and to establish a bridge across these scales in the related
sustainability assessments (Giampietro et al. 2011, 2012). In
the quantitative analysis of the metabolic pattern of societies,
we represent energy and material flows across structures and
functions (e.g., houses, crop fields) on a scale that is different
from the scale used for the analysis of flows across natural
structures and ecological functions (e.g., lakes, aquifers). For
Table 1. The different dimensions, issues, and disciplines that need to be
addressed in a holistic quantitative analysis of water-use sustainability and
the related role of a multiscale integrated analysis of societal and ecosystem
metabolism (MuSIASEM).
Ecological
Social
Socioeconomic
Issues
Water resources
(ecological funds)
Human
appropriation
of water
(societal flows)
Water withdrawal:
Water used in the
economic process
Disciplines
Ecology, Earth
sciences
Engineering,
agronomy,
industrial ecology
Economics,
sociology
multiscale
integrated analysis
Tools
MuSIASEM
MuSIASEM
MuSIASEM
Output
Ecological
constraints
analysis
Biophysical
constraints
analysis
Economic
constraints
analysis
nexus among social systems and between natural and social
systems (e.g., the water–food–trade nexus; Allan 2003). In
this way, assessments of virtual water are extremely effective
in helping people to understand the impacts that water consumption patterns do generate in ecosystems, be they near
or far from the place of consumption.
The problem of multiple scales is generated by the need
to transfer information across levels of representation.
The water flows that are relevant for the sustainability of
socioecological systems (SESs) can be perceived only at different spatiotemporal scales. To solve this conundrum, we
need a more complex approach to effectively integrating the
representation of the biophysical, ecological, social, and economic aspects of water use, analyzed at different hierarchical
levels and scales, into one coherent and holistic framework.
We propose as a solution, in this article, the multiscale
integrated analysis of societal and ecosystem metabolism
(MuSIASEM), a heuristic method for the integration of the
various dimensions, disciplines, and issues involved in such
a holistic quantitative analysis (table 1).
Linking the analyses of societal and ecosystem
metabolism
The recent gain in popularity of the notion of SESs underscores the need for a holistic approach to the analysis of the
interface between socioeconomic and ecological systems
(Berkes and Folke 1998, Young et al. 2006, De Aranzabal
et al. 2008). The basic features of SESs, common to the
many definitions and interpretations found in the literature,
are that human societies are embedded in the ecological
processes with which they have strong biophysical ties and
that the socioeconomic system and the ecological system in
which it is embedded should be considered as one complex
adaptive system, which is expected to express nonlinear
behavior and, therefore, is difficult to model.
Although effective in conveying the need for a holistic
approach, the concept of SESs may be misleading if it is
www.biosciencemag.org January 2013 / Vol. 63 No. 1 • BioScience 15
Articles
example, humanmade water purification systems can make
a volume of water drinkable in hours, whereas the natural
water cycle might take months or years to do the same.
Acknowledging that the study of societal and ecosystem
metabolism dictates the use of different scales does not
mean that the resulting analyses must be necessarily disconnected. Whenever the same pool of resources is being used
by both human societies and ecosystems, competition for its
use is practically inevitable. In MuSIASEM, we purposefully
establish a relationship between these two nonequivalent
quantitative analyses, and therefore, a thorough study of the
terms of this competition is warranted.
aspects of the SES and show how it works within the society. This analysis links the overall resource consumption of
society (level n) to the resource consumption by the various
functions (at level n – 1) and structural elements (level n – 2)
of society. Levels n, n + 1, and n + 2 describe the functional
aspects of the SES and show why the water metabolism in
society is possible in the first place. This analysis links society
(level n) to the characteristics of its boundary conditions
(level n + 1) and the processes guaranteeing the stability of
these boundary conditions (level n + 2).
Therefore, to check the sustainability of the metabolic
pattern of a society, in MuSIASEM, we use representations
belonging to two nonequivalent descriptive domains so as to
simultaneously assess the desirability of the resource flows
for the maintenance of the societal functions and structures
(internal triadic reading; levels n, n – 1, n – 2)—the view
from the inside—and the feasibility of the metabolic pattern in terms of the compatibility between the structural
and functional characteristics of society and the biophysical
constraints imposed by the embedding ecosystems (external
triadic reading; levels n, n + 1, n + 2)—the view from the
outside.
Using hierarchy theory to integrate the metabolic
pattern across multiple scales
Hierarchy theory is the branch of complexity theory dealing with the epistemological implications of multiple scales
(Grene 1969, Pattee 1973, Salthe 1985, Allen and Starr 1988,
O’Neill 1989). According to hierarchy theory, the same
system expresses different identities, depending on the
scale at which we observe it. The identity of an observed
system can be defined as each of the researcher’s perceptions of the investigated system as an entity distinct from
Two additional epistemological problems in the
its background and from other systems with which it is
assessment of water-use sustainability
interacting (Giampietro 2004). Therefore, the identity of
There are two additional epistemological predicaments
a system depends on the set of selected relevant qualities
plaguing the assessment of water-use sustainability and
(observable attributes) that are chosen
to reflect both the perception and the
n
n +2
n−2
n−1
n +1
representation of the system.
Each system’s identity has an associWater use
by society
Consumption
ated descriptive domain, the domain of
(life water)
Households
reality delimited by the interactions of
Surface
(blue water)
interest (Kampis 1991). This is where
Rain
we frame our representation of the
identity of the observed system. The
Agriculture
descriptive domain determines the corresponding scale, which is defined as
Effluents
the extent of the temporal or spatial
Soil
(green water)
boundaries of the description and its
Industry
resolution (or grain). Therefore, the
Percolation
coexistence of multiple scales implies
Taken from
Production
ecosystems
Services
the unavoidable coexistence of non(economy
Aquifers
Underground
water)
equivalent descriptive domains that
(blue water)
transfers
cannot be reduced to each other with a
formal system of inference (Mandelbrot
1967, Rosen 1985).
Triadic reading 1: The external view
Triadic reading 2: The internal view
Applying these basic concepts to
Figure 1. Hierarchical levels establishing a link between the external and
the analysis of an SES, we distinguish
internal views of societal water metabolism. The semantically open definition
five hierarchical levels of organization,
of water is shown by the changing role it acquires in the levels of each triadic
which are represented in figure 1 as two
reading. The labels blue and green water correspond to the water’s origin
contiguous triadic readings. The triadic
(i.e., evaporated from surface or ground water or stored in the soil as moisture,
reading on the right refers to society;
respectively). Economy water and life water correspond to the end use that
the one on the left refers to the embedwater is given (i.e., the water used for the economic sectors and the water used
ding ecosystems. Hierarchical levels n,
for drinking and personal care, respectively; Arrojo 2006).
n – 1, and n – 2 describe the structural
16 BioScience • January 2013 / Vol. 63 No. 1
www.biosciencemag.org
Articles
limiting the usefulness of conventional quantitative methods
of analysis: (1) how to define a volume of water (the quantity
to be measured) as a legitimate flow belonging to a given
category (e.g., what should be considered drinking water?)
and (2) how to include the external limit represented by
ecosystem integrity in the assessment of the maximum flow
of water that can be used by society.
MuSIASEM employs a semantically open definition of water as a
resource. In relation to the first point, the literature on water
studies abounds with definitions of the chemical compound
water. Depending on the underlying narrative, water is a
public good, a cultural element, a trade nexus, or an ecological niche. However, each of these definitions is semantically
closed—meaning that each term has a single fixed definition
that is related to the narrative and one that is restricted to the
scale of the corresponding descriptive domain. Therefore,
if we are to analyze water use across different hierarchical
levels of organization and scales, we must define in nonequivalent ways what water is and what water does in different descriptive domains.
Indeed, Zimmermann (1951) claimed that resources cannot be defined in substantive terms: “Resources are not, they
become” (p. 15). A certain amount of an element becomes a
resource only when it is able to provide a service to a specified end user. Therefore, a certain volume of water becomes
a resource only if its characteristics allow it to do so. For
example, for the end use of drinking water, only the volume
of water that possesses the set of attributes required for this
purpose qualifies as water resource. In line with this thinking, we employ a semantically open approach in our analysis
in order to give an identity to volumes of water while moving across hierarchical levels and scales. This means that for
different levels of analysis and different potential end users,
we must expect to find different definitions of water. This
approach allows for a cultural or context-dependent definition of water resource. For example, in Europe, the water of
the Ganges River would not meet the safety standards for
qualifying as drinking water, but in India, cultural practices
or a lack of alternatives create a different picture.
In the analytical framework of MuSIASEM, the attributes
characterizing a given volume of water as a water resource are
therefore defined by the identity of the structural and functional compartments of society using that water. For example,
the relatively high heat capacity of water (the attribute) and
the need for efficient cooling processes (the end use) define a
given waterflow as a resource for thermal power stations (the
functional or structural compartment). Within the multiscale
analysis, we discover that water resources are used in the process of reproduction by societal functional structures (at one
scale), but they are also ecosystem functional structures (at
a broader scale). Therefore, our conceptualization of water
use does not refer to the use of a given quantity of the chemical element water but to the services that a given volume of
water provides for the maintenance of the metabolic pattern
of observed systems (either a society or an ecosystem).
www.biosciencemag.org Including ecological integrity in the analysis. In relation to the
second epistemological problem, Aguilera Klink (1995)
proposed that water is an ecosocial asset, because it provides
services not only for socioeconomic systems but also for
ecosystems. Indeed, there is a wide range of ecohydrological
processes that rely on water’s presence (Brauman et al. 2007).
As part of these processes, water flows through ecosystems
guarantee the integrity of these systems by maintaining their
structural and functional relationships, from habitat provision to evapotranspiration. There is no ecosystem on Earth
capable of reproducing itself without using water. Indeed, it
is the pattern of water availability that delimits the boundary
conditions for the maintenance and reproduction of various
ecosystem types.
The water used by ecosystems is often qualitatively transformed (e.g., in terms of its quantity, quality, temperature, or
geographical and temporal reference) in a way that increases
its value for human end uses (e.g., drinking, irrigation,
regulation of flows, hydroelectric power), thus making it a
resource for social purposes. This process allows the provision of water’s ecosystem services (Willaarts et al. 2012).
Therefore, the interaction between the water cycle and
ecosystems contributes to the supply of water resources to
society and is crucial to water-use sustainability.
Georgescu-Roegen’s flow–fund model to frame the
quantitative analysis of water use in MuSIASEM
As was argued above, the simultaneous adoption of nonequivalent descriptive domains demands the use of semantically open definitions of useful water. For dealing with
these scale issues, Georgescu-Roegen (1971) proposed the
flow–fund model for the quantitative analysis of the biophysical processes underlying the functioning of a modern
economy. MuSIASEM uses this flow–fund model to provide both a semantic criterion to define the identities of the
elements of the system across scales and a formal criterion
to represent the relationships among them (Giampietro
et al. 2011).
A semantic criterion to define the identity of
elements across scales
In the MuSIASEM approach, fund elements are those elements of the observed system that remain the same across
the duration of the analysis (the extent of the chosen scale).
They represent what the system is and what the system is
made of. The idea of sustainability implies that these elements
are reproduced in the metabolic process. Flow elements, on
the other hand, are those elements that appear or disappear
over the duration of the analysis (outputs that are generated,
inputs that are consumed). They tell us what the system does
with regard to the interaction with the context (at the large
scale) and that among its internal components (at the local
scale) (Giampietro et al. 2011).
The question of whether a given element, in this case
water, should be classified as a fund or as a flow in the analytical representation of societal metabolism depends on
January 2013 / Vol. 63 No. 1 • BioScience 17
Articles
the chosen scale (i.e., the extent and grain) of the analysis
(Georgescu-Roegen 1971). When the required property of
a given volume of water is compromised during the time
scale of the analysis by its use, it must be considered a flow;
otherwise, it is a fund. Normally, social uses of water belong
to the category flow, because one or more water attributes
are lost in order to create economic value or to maintain
the social funds (e.g., water losing gravitational energy in a
hydroelectric plant, water becoming polluted because of its
use for cleaning, water evaporating or percolating away after
irrigation). However, there are rare cases in which water can
be considered a fund for social systems, such as ponds and
artificial lakes used for cultural or recreational purposes.
Given that the distinction between fund and flow is semantic, the same water volume can be perceived as a fund or as
a flow depending on the attributes and scale considered. For
example, a volume of water drawn from a river for irrigation is a flow that explains what the agricultural system of
a society does (narrow scale). On the contrary, the water of
that same river is a fund that expresses the identity of a lotic
ecosystem (broad scale). This semantic ambiguity in the definition of fund and flow elements makes it possible to study
the competition for the use of water between the two systems.
Indeed, in the analysis of the metabolic pattern of ecosystems,
the predictable flow regime of water in a river is a fund for
the ecosystem that needs that water to express the ecosystem’s
functions and preserve its biodiversity. A stable identity of
the ecological fund elements (e.g., rivers, lakes, aquifers) does
require a regulation of the input (e.g., affluents) in relation to
the output (e.g., effluents). At the global scale, this regulation
is enforced by the water cycle in its interaction with ecosystems. It determines the set of ecological water funds from
which humans can derive their water flows.
Formal criteria and quantitative indicators to represent expected
relationships between flows and funds. The relationship
between flows and funds has to be defined in qualitative and
quantitative terms. In relation to the qualitative aspects, the
flow–fund model indicates that in any metabolic system,
the identity of a flow depends on its end use, which is fund
specific. This expected relationship determines which water
flows are admissible in the accounting. For example, drinking
water (flow) for humans (fund) must satisfy certain criteria
(e.g., the absence of toxic substances and harmful micro­
organisms) and so must irrigation water (flow) for crop
plants (socioeconomic fund elements; e.g., salinity level).
In relation to the quantitative aspects, the nature of
metabolic systems allows us to define admissible values
for the flow:fund ratios that will guarantee the survival or
reproduction of the fund elements included in the analysis
(e.g., humans, plants, rivers). For example, humans must
consume, on average, about 2 liters of drinking water per
day per person—not much more and not much less; different types of crops require different quantities of irrigation
water per hectare; a freshwater ecosystem in a river needs a
minimum flow of water going through it.
18 BioScience • January 2013 / Vol. 63 No. 1
Therefore, a sound analysis of water-use sustainability
requires us to make informed preanalytical choices of relevant qualitative semantic attributes (e.g., what we mean by
economically “cheap” water) and related formal quantitative attributes (e.g., the cost of 1 cubic meter [m3] of water
expressed in a given currency) to be used as indicators that
will make it possible to properly track qualitative and quantitative changes in the flow and fund elements within a given
metabolic pattern. Indeed, within the analytical framework
of MuSIASEM, sustainability is closely related to the maintenance of the characteristics of the water flows within their
admissible ranges (qualitative check) and the flow:fund ratios
within the viable and feasible limits (quantitative check). A
flow extraction should not cause a permanent reduction
(i.e., a quantitative change) in or damage (i.e., a qualitative
change) to the fund elements that it feeds.
Therefore, the flow–fund model allows us to visualize,
quantify, and explain the nonequivalent roles that water
plays in the metabolic pattern of society and ecosystems
across scales and, therefore, to better explore the nexus
and competition between socioeconomic and ecological
processes.
Making the flow–fund model operational within MuSIASEM. The
preanalytical choice of defining a given volume of water as
either a flow or a fund determines the selection of the indicators of water use that will be included in our MuSIASEM.
These indicators consist of a system of measurable benchmarks that are able to characterize changes in the relevant
system structures and functions. For the indicators to be
defined, we must therefore properly identify the fund elements that describe the end uses of water and the semantic
qualitative criteria that delineate what kind of water should
be considered a water resource to those ends. The identification of these criteria allows us to assign relevant semantic
attributes to the flow elements. The formal quantitative
criteria that define how much water is needed for stabilizing
the societal and ecological metabolic patterns and that thus
allow us to establish the relative size of the fund elements
in the form of indicators, such as hours of human activity
devoted to maintain a certain social function or cultivated
area, should also be identified.
With all this said, the preanalytical choice of qualitative
criteria, attributes, and indicators obviously depend on the
purpose of the study and the specific characteristics of the
investigated SES.
In figure 2, we give an example of how to proceed with
these preanalytical choices using a multicriteria approach.
Here, an integrated set of criteria, semantic attributes, and
quantitative indicators of water-use sustainability is shown
in the form of a radar diagram organized around two
perpendicular axes. To the left of the vertical axis, we have
information about the metabolic pattern of the embedding ecosystems; to the right, we have information about
the metabolic pattern of the socioeconomic system. Above
the horizontal axis, we have the qualitative aspect of the
www.biosciencemag.org
Articles
Figure 2. Definition of a set of formal attributes and indicators in the preanalytical step of a multiscale integrated analysis
of water-use sustainability. The location in the radar indicates the desirability of the situation.
analysis (the choice of attributes and indicators); below it is
the quantitative aspect (the resulting assessment of quantities). The indicators shown have been assigned a value (the
black squares in the radar diagram) and a related meaning
(depending on its position in the viability domain: The further from the center it is, the better its performance is).
In the lower part of figure 2, we see that the preanalytical
definition of fund elements (either human activity on the
right or water bodies on the left) allows us to formalize the
considered situation in quantitative terms. For instance, in
this ­example, we see that water withdrawal from rivers is
excessive (the indicator shows a negative effect on this fund
element), water use by the residential sector satisfies the needs,
and the agricultural sector has plenty of water for irrigation.
In this particular example (figure 2), the value taken on by
the various indicators is characterized according to the flag
model (Gomiero and Giampietro 2005). In this model, for
each one of the chosen attributes, we select an indicator or
proxy variable (e.g., for the semantic attribute nutrient concentration, we select the formal attribute nitrogen concentration)
and define the related viability domain (the minimum or
maximum acceptable values) and a range of desirable values,
represented here by the different gray circles at the center.
As a result, we test the desirability of the actual supply of
water flows in relation to the various end uses in a society
www.biosciencemag.org (a check on the internal constraints) in qualitative (upper
part) and quantitative terms (lower part). The feasibility of
this metabolic pattern of water use in relation to the external
constraints imposed by the ecosystem metabolism can be
verified by assessing the impact of water withdrawal (defined
in quantity and quality) by society on the stability of the
ecological funds. This integrated characterization is useful
for studying whether human water withdrawal compromises
the integrity of the ecosystem.
For the water used in the economic process, the attributes
defining its role as an input flow are determined by preferences
and regulations set by social institutions (such as laws or market
transactions). Because water is used in different economic sectors and within these sectors for many different purposes and
functions, it is often impossible to define in substantive terms
its optimal use, even when considering only societal meta­
bolism. The consideration of water as a resource for ecosystems
hinders a consensus on it even more, because it is virtually
impossible to achieve full agreement on the quantification of
the biophysical limits of water appropriation by society, on
the damage that human water use causes, or on the desirability
of the preservation of the identity (i.e., the key characteristics)
of ecological funds (e.g., lakes, rivers, aquifers).
Therefore, a proper multicriteria frame, as is shown in
­figure 2, requires us to carefully go through the following
January 2013 / Vol. 63 No. 1 • BioScience 19
Articles
steps: selection of relevant criteria, identification of relevant
attributes, choice of indicators, gathering of data, definition of viability and desirability, evaluation of trade-offs,
and ­iteration of the whole process to increase its quality (Nijkamp et al. 1990, Vincke 1992, Allen et al. 2003).
Carrying out this kind of analysis in a technocratic way by
experts alone is not possible. The effort requires participatory
approaches to handle the unavoidable presence of legitimate
but contrasting perceptions and of large doses of uncertainty
in relation to the quantitative representation (Funtowicz and
Ravetz 1990, Checkland and Scholes 1999, Munda 2008).
A bridge between nonequivalent representations of
water metabolism on different scales
We show here the steps of constructing a multiscale integrated assessment of the water metabolism of SESs on the
basis of two nonequivalent representations of the metabolic
pattern of water at two different scales: societal metabolism
and ecosystem metabolism.
Adopting the small scale: The water metabolism of society seen
from the inside. In figure 3, we bring in focus the water
metabolism of society, adopting the internal view. The data
presented refer to the Catalonia region of Spain in 2007. This
view corresponds to a socioeconomic taxonomy of relevant
fund and flow categories for the accounting. The time span
of this representation is 1 year (extent), and the resolution
(grain) for the analysis of flows across fund elements is
1 hour.
The total water appropriated by society (at level n) is
extracted from water funds and amounts to 2956 hm3
per year. This indicator is useful for studying the compatibility between internal and external constraints because
it lies on the interface of the internal and external triadic
reading. This same volume of water can also be expressed,
within the internal triadic reading, as a rate of water use of
51 liters (L) per hour of human activity per year (equivalent to 1224 L per capita per day); this is the flow sustaining
the human population.
To establish a relationship between the overall flow of water
at level n and the water flows at lower levels n – i, we map the
water flows onto the fund element human activity. This fund
element provides the structural frame of our assessment of the
internal view. It serves this purpose well, because a structural
and functional subdivision of society on the basis of human
activity corresponds accurately to the various socioeconomic
Figure 3. The internal view: Water withdrawal and use in the Catalonia region of Spain in 2007 by economic sectors.
The data are from water-use surveys of the Spanish National Statistics Institute. Abbreviations: AGR, agriculture;
HH, households; hm3, cubic hectometers; IND, industry; L, liters; m3, cubic meters; PW, paid work; SG, services and
government. The economic productivity of labor is represented in black. The water-use rate is in gray.
20 BioScience • January 2013 / Vol. 63 No. 1
www.biosciencemag.org
Articles
end uses of water. In our example (figure 3), the total number
of hours of human activity per year of the whole society are
split between the household sector and the paid work sector
at level n – 1. The human activity in the paid work sector is
then further divided at level n – 2 among the agricultural,
industrial, and service and government sectors. The empirical
data of figure 3 show that the total flow of water use is largely
determined by the water use in the paid work sector and only
marginally affected by the household sector. Of the paid work
sector, the agricultural sector accounts for the lion’s share—
more than 70% of the total water use. This representation of
the societal water metabolism is focused on the direct uses
of water under human control and is therefore limited to the
withdrawal of water from superficial and subterranean water,
the reuse of water, and water from desalination.
As part of this internal view, we can also simultaneously
assess flows of water and money (defined over the same
fund elements) so as to provide information on the economic performance of water flows. This latter indicator is an
essential piece of information for a holistic assessment of the
water metabolism of SESs and for economic policymaking.
In figure 4, which shows the metabolic pattern of water use
in Catalonia for the period of 2000–2008, we illustrate how
to obtain this information. On the vertical axis, we have the
water-use rate (in m3 per hour of human activity), and on
the horizontal axis is the economic labor productivity (in
euros []) of gross value added per hour of human activity).
We see here again that water resources are used at markedly
different rates by the various economic sectors (be aware that
the graph has a logarithmic scale). We also see that the sector
with the highest hourly water-use rate (agriculture) has the
lowest economic productivity per hour. Indeed, the economic
Water-use rate (cubic meters per hour)
100
10
1
0.1
0.01
0.001
0
HH
5
10
15
20
25
30
Economic productivity of labor (gross value added, per hour)
CAT
AGR
PW
IND
SG
Figure 4. Metabolic performance of different societal
functions in the Catalonia region of Spain in 2007. The
data are from water-use surveys of the Spanish National
Statistics Institute. Abbreviations: AGR, agriculture;
CAT, the entire Catalonia region; HH, households; IND,
industry; PW, paid work; SG, services and government.
www.biosciencemag.org performance of water use in the agriculture sector is a mere
1 /m3, whereas the industrial and service and government
sectors reach values of, respectively, 224 and 916 /m3 (also
see figure 3). Note that the water resources used by the
various economic sectors shown are defined in different ways.
For instance, water used at the household level has different
attributes from water used in the agricultural sector.
Adopting the large scale: View of society interacting with the
embedding ecosystems. The price to pay for seeing the
dynamics of the water metabolism within the society (the
internal socioeconomic perception) is the adoption of a
scale that does not allow us to see the water dynamics of the
embedding ecosystems or the relationship between the local
ecosystems and the larger water cycle.
In figure 5, we focus on the water metabolism of the
ecosystems embedding society, adopting the external triadic
reading (also see figure 1). Ecological fund elements, such
as lakes, rivers, and aquifers, form the structural and functional compartments of the ecological processes operating
at levels n + i. Accordingly, in this representation, water
volumes are mapped and accounted for in spatial terms
(instead of in hours of human activity). The time extent
of this representation is necessarily much broader than
1 year—from decades to thousands of years, depending on
the type of ecosystem under analysis—and it is required to
define a stable identity of the ecological funds. Therefore,
the temporal resolution (grain) of the analysis of ecosystem
metabolism must also be much higher than the one used
for the analysis of societal metabolism. In order to establish
an interface between the two quantitative analyses and their
corresponding scales, we set the resolution for the analysis of ecosystem metabolism at 1 year so that it matches
the extent of the analysis of societal metabolism. In this
way, the total amount of water required or used by society
in 1 year (defined at level n, the interface; see figures 1, 3,
and 5) can be represented at the larger scale as an appropriation rate from ecological fund elements (figure 5).
On this larger scale, we prefer to define water use as the
amount of water appropriated by humans, rather than
simply as the amount of water withdrawn (figure 5). This
is to emphasize the analogy with the concept of human
appropriation of net primary productivity suggested by
Vitousek and colleagues (1986). Indeed, human appropriation of water from the fund elements of the ecosystem
may have serious consequences on the water cycle at the
larger scale if it destabilizes the integrity of local ecosystems (e.g., extensive deforestation, severe river decline).
The stabilization of favorable boundary conditions for
the expression of both societal and ecosystem metabolism
ultimately depends on the global water cycle. The global
water cycle is essential for the regulation of the temperature on Earth and for the stabilization of aquifers, rivers,
lakes, and seas. It is undoubtedly one of the most important metabolic patterns expressed by our planet. Indeed,
the maintenance of the water cycle requires a huge flow
January 2013 / Vol. 63 No. 1 • BioScience 21
Articles
Figure 5. The external view: The effects of the human appropriation of water in Catalonia in 2007 over the Ebro Basin and
the Internal Basins of Catalonia (CIC). The data are from water surveys of the Spanish National Statistics Institute and
the Catalan Water Agency. The white arrows represent the water use for refrigeration. Abbreviations: AGR, agriculture;
HH, households; hm3, cubic hectometers; IND, industry; popul., population; SG, services and government. A darker color
indicates the water’s loss of an ability to provide services. The water cycle renews the water resources that arrive to the
ecological funds.
of energy—44,000 terawatts (TW), or one-third of the
total solar energy reaching Earth (Taube 1985)—to keep in
motion and to thereby recover the water resources degraded
by the social system. This quantity is 2750 times the total
amount of energy controlled by humankind in 2010, which
was around 16 TW, or 504 exajoules per year (BP 2012). This
enormous difference between the amount of energy used
by nature for the water cycle and that used (i.e., controlled)
by societies for their own metabolism explains why the vast
majority of the processes determining the supply of freshwater for societal uses are beyond direct human control. In
fact, only desalination and water treatment for reuse take
place within the metabolic pattern of society. Nonetheless,
because of their huge energy and capital cost, these activities
account for only a negligible fraction of the water consumed
by society.
How bridging scales can make a difference. The gap between the
time horizons at which we are able to visualize the effects of
water use on the economy (the shorter horizon) and on the
22 BioScience • January 2013 / Vol. 63 No. 1
embedding ecosystems (the longer horizon) often makes it
difficult to conduct a comprehensive evaluation of policies
with a national or regional scope having influence on water
use. An analysis of the relationships between the water funds
providing the identity of the ecological systems (figure 5) and
the water flows used by society (figure 3) can be extremely
helpful in this regard.
We illustrate this with an example from “vegetable factory” greenhouses located close to the Tabernas Desert in
the province of Almeria, in southern Spain. In this area, the
high percentage of agricultural lands irrigated with groundwater (around 80%; Consejería de Agricultura y Pesca 2011)
is a consequence of the low average rainfall in the region
(around 190 millimeters per year; AEMET 2012). A shortterm, technical analysis of water use as a productive factor
has led local greenhouse producers and managers in Almeria
to achieve the highest efficiency in irrigation in Europe
(Contreras 2002). However, the long-term effect of this
technological improvement has been shown to be a classic
case of the Jevons paradox (Polimeni et al. 2008): In the case
www.biosciencemag.org
Articles
of Almeria, it has indeed led to a rapid expansion of greenhouses in the region and a concomitant increase in total
water use with disastrous effects for local water funds and
ecosystems. The enhanced economic performance of water
use in agriculture has marginalized its role in reproducing
ecosystem funds (i.e., ecological structures and functions)
with the result that, now, most of the rivers and aquifers in
that region are declared to be in a situation of risk under
the regulations of the European Water Framework Directive
(see supplemental material, available online at http://dx.doi.
org/10.1525/bio.2013.63.1.6).
The MuSIASEM approach can help avoid a situation
like the one in Almeria by bridging the two scales mentioned. This means that the flows of water use in society are
mapped against relevant categories of land uses in addition
to relevant categories of human activity. Given that water
availability has a quite fixed geographical reference, the
geographical link between the water funds and water flows
provided by the analysis of the land-use patterns becomes
crucial in avoiding such situations. Within the MuSIASEM
approach, we can distinguish between the natural landcover,
which uses water as a fund, and the human-driven land uses,
in which human colonization defines the role of water as a
flow. The study of the bridged triadic readings proposed here
would have shown the unsustainability of the increment on
irrigation land for greenhouse production and would have
led to policy advice against this option.
Conclusions
The analysis of the metabolic pattern of water allows us
to handle nonequivalent definitions of water as a resource
for the reproduction of socioeconomic fund elements and
ecosystem fund elements. The chemical element water
per se does not possess any attributes that permit us to
define its usefulness for either society or ecosystems. It is the
definition of water as “becoming a resource” in relation to
a specified end use that makes it possible to describe a set of
expected characteristics for water flows for different types
of autopoietic processes. By adopting this approach, the
qualitative and cultural aspects of societies can be included
in the definition of water resources and, for that reason, in
the quantitative assessment of water use.
An integrated quantitative characterization of the pros
and cons of water use should always address the societal
metabolic pattern and the corresponding desirability of
societal water use: How and why is a society using water (as
a flow element) inside the economic process? The ecosystem
metabolic pattern and the corresponding viability of societal
water use must be addressed: How and why do the embedding ecosystems (as fund elements) and other biophysical
conditions impose external constraints on the societal use
of water?
These two metabolic patterns must be studied at different scales. When water is considered as a resource for society (a flow element used to reproduce societal funds), the
attributes of performance defining its quality depend on
www.biosciencemag.org the functions or end uses that are essential or desirable for
society. The resulting quantitative analysis corresponds to
a descriptive domain with a time span of 1 year (because
of statistical constraints) and a grain of 1 hour. However,
when water is considered as a resource for ecosystems (a
fund needed for guaranteeing the stability of ecosystem
meta­bolism), the preservation of the identity (i.e., the key
characteristics) of the ecological funds (e.g., lakes, rivers,
aquifers) is the criterion that determines the biophysical limits to water appropriation by society. The resulting
quantitative analysis corresponds to a descriptive domain
with a much larger time span (e.g., decades, centuries) and
a grain of 1 year.
An integrated analysis of SESs necessarily employs indicators that are defined across nonequivalent descriptive
domains. As a consequence, we cannot rely on reductionist models and protocols (of the one-size-fits-all variety).
We have to adopt semantically open procedures and tailor
our quantitative analysis to our purpose and the specific
characteristics of both the social and the ecological systems
under study.
Acknowledgments
The authors are grateful to Sandra Bukkens for her excellent
editing job and to Jesús Ramos-Martín and three anonymous reviewers for their useful comments. This research has
been funded by the Agency for Management of University
and Research Grants of the Government of Catalonia
(grant nos. SGR2009-594 and SGR2009-600), the Catalan
Reference Network on Economics and Public Policies, and
the Spanish Ministry of Science and Innovation (Formación
de Profesorado Universitario program). Mario Giampietro
acknowledges the support of the Stellenbosch Institute for
Advanced Study to the program.
References cited
[AEMET] Agencia Estatal de Meteorología. 2012. Valores climatológicos
normales: Almería aeropuerto. AEMET. (1 October 2012; www.aemet.
es/es/serviciosclimaticos/datosclimatologicos/valoresclimatologicos?l=
6325O&k=and)
Aguilera Klink F. 1995. El agua como activo económico, social y ambiental.
El Campo 132: 15–27.
Allan JA. 1998. Virtual water: A strategic resource: Global solutions to
regional deficits. Ground Water 36: 545–546.
———. 2003. Virtual water—The water, food, and trade nexus: Useful concept or misleading metaphor? Water International 28: 106–113.
———. 2011. Virtual Water: Tackling the Threat to Our Planet’s Most
Precious Resource. I. B. Tauris.
Allen TFH, Starr TB. 1988. Hierarchy: Perspectives for Ecological Complexity.
University of Chicago Press.
Allen TFH, Tainter JA, Hoekstra TW. 2003. Supply-Side Sustainability.
Columbia University Press.
Arrojo P. 2006. El Reto Ético de la Nueva Cultura del Agua: Funciones,
Valores y Derechos en Juego. Paidós Estado y Sociedad no. 139.
Paidós.
Berkes F, Folke C. 1998. Linking Social and Ecological Systems: Management
Practices and Social Mechanisms for Building Resilience. Cambridge
University Press.
[BP] British Petroleum. 2012. 2011 in Review. BP. (2 October 2012; www.bp.
com/sectiongenericarticle800.do?categoryId=9037128&contentId=7068555)
January 2013 / Vol. 63 No. 1 • BioScience 23
Articles
Brauman KA, Daily GC, Duarte TK, Mooney HA. 2007. The nature and
value of ecosystem services: An overview highlighting hydrologic services. Annual Review of Environment and Resources 32: 67–98.
Brooks DR, Collier J, Maurer BA, Smith JDH, Wiley EO. 1989. Entropy and
information in evolving biological systems. Biology and Philosophy 4:
407–432.
Checkland P, Scholes J. 1999. Soft Systems Methodology in Action. Wiley.
Consejería de Agricultura y Pesca. 2011. Inventario de Regadíos 2008 y Su
Evolución en la Última Década. Junta de Andalucía, Spain.
Contreras S. 2002. Los regadíos intensivos del Campo de Dalías (Almería).
Pages 151–192 in Fernandez J, Selma MAE, eds. Agua, Regadío y
Sostenibilidad en el Sudeste Ibérico. Bakeaz.
Cottrell F. 1955. Energy and Society: The Relation Between Energy, Social
Change, and Economic Development. McGraw-Hill.
De Aranzabal I, Schmitz MF, Aguilera P, Pineda FD. 2008. Modelling of
landscape changes derived from the dynamics of socio-ecological systems: A case of study in a semiarid Mediterranean landscape. Ecological
Indicators 8: 672–685.
Depew DJ, Weber BH, eds. 1989. Evolution at a Crossroads: The New
Biology and the New Philosophy of Science. MIT Press.
Falkenmark M. 1995. Land-water linkages: A synopsis. Pages 15–16 in Food
and Agriculture Organization of the United Nations (FAO). Land and
Water Integration and River Basin Management. FAO.
Fischer-Kowalski M. 1998. Society’s metabolism. Journal of Industrial
Ecology 2: 61–78.
Funtowicz SO, Ravetz JR. 1990. Uncertainty and Quality in Science for
Policy. Springer.
Georgescu-Roegen N. 1971. The Entropy Law and the Economic Process.
Harvard University Press.
Giampietro M. 2004. Multi-scale Integrated Analysis of Agroecosystems.
CRC Press.
Giampietro M, Mayumi K, Sorman AH. 2011. The Metabolic Pattern of
Societies: Where Economists Fall Short. Routledge.
———. 2012. Energy Analysis for a Sustainable Future: The Multi-scale
Integrated Analysis of Societal and Ecosystem Metabolism. Routledge.
Gomiero T, Giampietro M. 2005. Graphic tools for data representation in
integrated analysis of farming systems. International Journal of Global
Environmental Issues 5: 264–301.
Grene M. 1969. Hierarchy: One word, how many concepts? Pages 56–58
in Whyte LL, Wilson AG, Wilson D, eds. Hierarchical Structures:
Proceedings. Elsevier.
Kampis G. 1991. Self-Modifying Systems in Biology and Cognitive Science: A
New Framework for Dynamics, Information, and Complexity. Pergamon.
Laborte AG, Van Ittersum MK, Van den Berg MM. 2007. Multi-scale analysis
of agricultural development: A modelling approach for Ilocos Norte,
Philippines. Agricultural Systems 94: 862–873.
Mandelbrot B. 1967. How long is the coast of Britain? Statistical self­similarity and fractional dimension. Science 156: 636–638.
Margalef R. 1968. Perspectives in Ecological Theory. University of Chicago
Press.
Martinez-Alier J, Schlüpmann K. 1987. Ecological Economics: Energy,
Environment, and Society. Basil Blackwell.
24 BioScience • January 2013 / Vol. 63 No. 1
Maturana HR, Varela FG. 1980. Autopoiesis and Cognition: The Realization
of the Living. Springer.
Munda G. 2008. Social Multi-criteria Evaluation for a Sustainable Economy.
Springer.
Nijkamp P, Rietveld P, Voogd H. 1990. Multicriteria Evaluation in Physical
Planning. North-Holland.
O’Neill RV. 1989. Perspectives in hierarchy and scale. Pages 140–156
in Roughgarden J, McCredie May R, Levin SA, eds. Perspectives in
Ecological Theory. Princeton University Press.
Odum EP. 1971. Fundamentals of Ecology, 3rd ed. Saunders.
Pattee HH. 1973. Hierarchy Theory: The Challenge of Complex Systems.
Braziller.
Polimeni JM, Mayumi K, Giampietro M, Alcott B. 2008. The Jevons Paradox
and the Myth of Resource Efficiency Improvements. Earthscan.
Prigogine I. 1978. Time, structure, and fluctuations. Science 201:
777–785.
Rosen R. 1985. Anticipatory Systems: Philosophical, Mathematical, and
Methodological Foundations. Pergamon.
Salthe SN. 1985. Evolving Hierarchical Systems: Their Structure and
Representation. Columbia University Press.
Taube M. 1985. Evolution of Matter and Energy on a Cosmic and Planetary
Scale. Springer.
Ulanowicz RE. 1986. Growth and Development: Ecosystems Phenomenology.
Springer.
Vincke P. 1992. Multicriteria Decision-Aid. Wiley.
Vitousek PM, Ehrlich PR, Ehrlich AH, Matson PA. 1986. Human appropriation of the products of photosynthesis. BioScience 36: 368–373.
Weber BH, DePew DJ, Smith JD, eds. 1988. Entropy, Information, and
Evolution: New Perspective on Physical and Biological Evolution. MIT
Press.
White LA. 1943. Energy and the evolution of culture. American
Anthropologist 45: 335–356.
Willaarts BA, Volk M, Aguilera PA. 2012. Assessing the ecosystem services supplied by freshwater flows in Mediterranean agroecosystems.
Agricultural Water Management 105: 21–31.
Young OR, Berkhout F, Gallopin GC, Janssen MA, Ostrom E, van der Leeuw
S. 2006. The globalization of socio-ecological systems: An agenda for
scientific research. Global Environmental Change 16: 304–316.
Zimmermann EW. 1951. World Resources and Industries: A Functional
Appraisal of the Availability of Agricultural and Industrial Materials.
Harper.
Zipf GK. 1941. National Unity and Disunity: The Nation as a Bio-Social
Organism. Principia.
Cristina Madrid ([email protected]) and Mario Giampietro
are affiliated with the Institute of Environmental Science and Technology
at the Autonomous University of Barcelona, Spain. Mario Giampietro’s
­primary affiliation is with the Catalan Institution for Research and Advanced
Studies (ICREA). Violeta Cabello is affiliated with the Department of Human
Geography at the University of Seville, Spain.
www.biosciencemag.org