Content uploaded by Oula Seitsonen - Sakarin-Pentin Ilarin Oula
Author content
All content in this area was uploaded by Oula Seitsonen - Sakarin-Pentin Ilarin Oula on Jan 20, 2023
Content may be subject to copyright.
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Toward a holistic understanding of pastoralism
Pablo Manzano, Daniel Burgas, Luis Cadahıa, Jussi T. Eronen, Alvaro Fernandez-Llamazares,
Slimane Bencherif, Øystein Holand, Oula Seitsonen, Bayarmaa Byambaa, Mikael Fortelius, Marıa
E. Fernandez-Gimenez, Kathleen A. Galvin, Mar Cabeza, and Nils Chr. Stenseth
SUMMARY
Pastoralism is globally significant in social, environmental, and economic terms. However, it
experiences crises rooted in misconceptions and poor interdisciplinary understanding, while being
largely overlooked in international sustainability forums and agendas. Here, we propose a
transdisciplinary research approach to understand pastoralist transitions using (1) social, economic,
and environmental dimensions, (2) diverse geographic contexts and scales to capture emerging
properties, allowing for cross-system comparisons, and (3) timescales from the distant past to the
present. We provide specific guidelines to develop indicators for this approach, within a social-
ecological resilience analytical framework to understand change. Distinct systems undergo similar
transitions over time, crossing critical thresholds and then either collapsing or recovering. Such an
integrated view of multidimensional interactions improves understanding of possible tipping points,
thereby supporting better-informed decision making. The need for a paradigm shift in pastoralism
science and policy is pressing. This research approach, including participatory methods, can provide
the solutions urgently needed.
GRAPHICAL ABSTRACT
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Introduction
Why pastoralism and why now?
Pastoralism is the most widespread land use worldwide.1, 2, 3 Present in over 100 countries
through all inhabited continents (Figure 1), its extent may be up to 60% of the world's terrestrial
area—but see4,5 for gaps and uncertainties in estimates. With between 50 and 500 million people
living from this practice,6, 7, 8, 9 it represents an important contribution to the agricultural GDP in
many nations—e.g., 88% in Mongolia10 and 50% in Kenya.11 Beyond its economic value, extensive
pastoralism also provides significant environmental, social, and cultural contributions.12
Pastoralist systems are often present in harsh and highly variable regions (Figure 1). These social-
ecological systems (SES) have risen and fallen since their origins millennia ago, but the last decades
have witnessed an increasing frequency and magnitude of sudden livestock production losses.15.
This global pattern threatens the future viability of pastoral livelihoods and poses great challenges
for achieving many of the UN Sustainable Development Goals16 in many nations. In contrast,
forecasted climatic changes threaten agricultural production in many limiting environments where,
in turn, pastoralism may remain the most resilient and adaptive livelihood.17 Such characteristics,
coupled with a growing human population and a predicted increase in meat demands,18 make
investing in pastoralism research—and supporting policies—a timely endeavor.
The pressing need to advance the science and policy around pastoralism is justified by a number of
reasons. First, crises in pastoralism have been linked to policies that undermine the adaptability of
pastoralist systems, such as those promoting sedentarization and land privatization.1,4,19, 20, 21
These policy trends have been associated with the persistent demeaning of pastoralist livelihoods,
often perceived as an obsolete or inferior alternative compared with other livelihoods.22,23 Second,
this marginalization is reflected in the weak presence of pastoralism at international environmental
and economic policy forums and global sustainability agendas. UNFCCC COP24 showcased this
problem, with panels specifically dedicated to mountains,24 oceans,25 farmers or indigenous
peoples,26 but none on rangelands or pastoralists, and no organized presence of pastoral interest
groups. Third, a recent UN report highlights widespread knowledge gaps as reasons behind
detrimental policies5, emphasizing (1) the challenge of studying the diverse systems—as depicted in
Figure 1—under a single umbrella, and (2) interdisciplinary integration being essential if we want
the most updated research to be incorporated into practice. It is important to point out that most
research on pastoralism so far has been isolated and fragmented across disciplines and geographic
regions, with few efforts cutting across them. To the best of our knowledge, there is no global
appraisal that maps out the relevance of pastoralism in the context of planetary sustainability, and
also the rangelands upon which they depend.
Analytical framework and theoretical considerations
This paper uses a social-ecological resilience framework to develop a theory of pastoralism that
incorporates multiple slow, controlling, and fast-moving social, ecological, and institutional
variables operating at multiple scales from the local to the global.27,28 Maintaining resilience thus
entails the management for a diverse set of parameters and their feedbacks. Social, ecological, and
political processes are path dependent,29, 30, 31 such that legacies of past events affect the
dynamics of the current SES.
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Figure 1. Pastoralism globally
The map shows an estimate of the extent of pastoralism, after the report by Nori and Davies,13 and
the pictures illustrate different pastoralist societies that herd
diverse livestock (sheep, goats, cattle, camels, llamas, reindeer). We understand pastoralism as a
production system based on extensive livestock production
that has a high reliance on common-pool natural forage resources. The often high spatiotemporal
variation of resources is managed through livestock mobility,
although this can take place at very different scales. They are low-input, low-output systems.14
Pastoralism offers comparative advantages for sustainable food
production and livelihoods in areas where cultivation is risky or unsuitable: tundra, boreal forests,
mountains, and drylands.
(A) USA sheep herding (Wikimedia Commons: ARS_sheep_herding.jpg [public domain]).
(B) Spanish transhumance (Barcex/Wikimedia Commons: Madrid_-
_XX_fiesta_de_la_trashumancia_-_131006_105048.jpg [cc-by SA 3.0 unported]).
(C) Quechua girl and alpaca (Donkeet/Wikimedia Commons: Gwalpaca.jpg [cc-by SA 3.0
unported]).
(D) Huaso (LBM1948/Wikimedia Commons: Laguna_Blanca_06.jpg [cc-by SA 4.0]).
(E) Yaake (Dan Lundberg/Flickr: 1997_276-18A_Yaake_demonstration.jpg [cc-by SA 2.0]).
(F) Beja people (Nikswieweg/German Wikipedia: Bedscha.jpg [cc-by SA 3.0]).
(G) Maasai (Andreas Lederer/Flickr: Young_Masai_herder.jpg [cc-by 2.0]).
(H) Daasanach (Rod Waddington/Flickr: Older_Woman,
_Dassanech_Tribe,_Ethiopia_(21884095049).jpg [cc-by SA 2.0 Generic]).
(I) Sami reindeer herder (Mats Andersson/Flickr: Reindeer_herding.jpg [cc-by 2.0]).
(J) Gujjar lady (Nishit Dey/Pexels: https://www.pexels.com/photo/gujjar-lady-from-pushkar-
1150846/[public domain]).
(K) Australian cattle rancher (Wikimedia Commons: VRDcattle.jpg [public domain]);
(L) Bakkerwals (Laportechicago/English Wikipedia: Bakkerwals.jpg [cc-by 2.5]).
(M) Mongolian goat herder (Taylor Weidman/The Vanishing Cultures Project/Wikimedia
Commons: Mongolia_Herding_Life4.JPG [cc-by SA 3.0 unported]).
(N) Yakut reindeer herder (Somogotto/Wikimedia Commons: Yakut_Sakha_herder.jpg [cc-by SA
4.0]).
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
This social-ecological resilience framework includes both adaptive co-management and adaptive
governance approaches (see Box 1 for definitions). These approaches, operationalized through
institutional arrangements, have the potential to empower local decision making and link local
communities to vertical and horizontal organizations (e.g., for funding, policy), and emphasize the
role of collaboration and social learning.32, 33, 34, 35 This framework provides the ability to
describe the broad-scale historical processes of pastoralism, the nested institutional arrangements
across scales, the social and ecological outcomes, and the political, economic, and social drivers
that constrain or enable pastoralism. Strategies to enhance social-ecological resilience include
maintaining biological, economic, and cultural diversity, sustaining a mix of stabilizing feedbacks
and creative renewal that build natural and social capital (the productive base), fostering social
learning and innovation through adaptive co-management, facilitating adaptive governance
arrangements that are flexible, redundant, and diverse to adapt to changing conditions, and taking
advantage of windows of opportunity to transform the system to sustainable futures.36 The ultimate
goal of the resilience-based framework is aligned both with an understanding of pastoralism and the
sustainable development goals37 in its effort to provide the capacity for ecosystems to maintain the
ecosystem services upon which people depend, while at the same time enhancing human livelihoods
and well-being.38
Box 1
Definitions of relevant terms for the resilience theoretical framework
Adaptive co-management: process by which institutional arrangements and ecological knowledge
are tested and revised in a dynamic, ongoing, self-organized process of learning-by-doing (after
Folke et al.39).
Adaptive governance: flexible, polycentric institutional arrangements that are nested across
horizontal and vertical linkages; diverse, redundant, and overlapping institutional arrangements
(after Folke et al.,36 Berkes,40 and Kofinas41).
Governance: formal and informal institutions, policies, rules, and practices that shape human-
environment interactions (after Folke et al.36 and Kofinas41).
Social-ecological resilience: capacity of groups and individuals to adapt or transform in the face of
change in social-ecological systems, particularly unexpected change, in ways that continue to
support human well-being (after Folke et al.42).
Transdisciplinary research: research process where academics from different disciplines, as well as
stakeholders from the civil society and other sectors, co-produce knowledge (after Guimarães
et al.43).
A primary concern in resilience and resilience-based ecosystem stewardship is to understand the
factors associated with exogenous controls and maintenance of critical slow and fast variables.44
Exogenous controls include regional climate and national-level policies, economics, and
governance (Box 1). These are not managed for per se but should be recognized as regulating
factors that condition the state of slow and fast variables.
Slow variables are controlling factors that regulate ecosystem structure and function, and are
buffered by stabilizing effects so that they remain relatively constant over time. Critical slow
variables include functional types of plants, disturbance regimes, or cultural ties to the land that
regulate stability, maintain historical legacies, and provide ecological and social memory. Fast
variables (e.g., annual growth, wildlife density, agent-specific preferences) respond to daily,
seasonal, and inter-annual variation. Rapid changes in fast variables may serve to overwhelm slow
variables leading to a regime shift.45,46 This is characteristic of systems in panarchy, hierarchical
SES that are interlinked in continual adaptive cycles of growth, accumulation, release, and
renewal.47,48 During the first two phases of such cycles, the system undergoes a steadily
accumulation of potential that is not clearly visible and gives an impression of stability, apparently
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
conserving the balance among its elements. Such apparent balance is, however, broken when the
accumulated potential approaches critical tipping points and is then released, giving way to a phase
of crisis but also of renewal. Classic examples of these conservation-release patterns would be the
accumulation of biomass that finally releases fires and germination of sun-loving seeds previously
suppressed by shadow; or the accumulation of social tensions in the usual functioning of a society
that finally escalate to the point of releasing social unrest and giving way to the proposal and
construction of new, even revolutionary social structures. The founder conditions (plant community,
governance scheme) established at the beginning of the new cycle are going to define and select the
structure of the next conservation phase, which may not be equal to the old one because of the loss,
addition, or rearrangement of elements (species, ideas) in respect to the former cycle. Tempo over
the cycle phases is not equal, for example, the conservation phase may be hundreds of years and the
release may be in days. It is difficult to sometimes see the release phase, only that the system has
changed. Therefore, maintaining resilience in SES requires simultaneous monitoring of and
management for slow and fast variables and their interactions to avoid undesirable state
transitions.44 In addition, it is important to recognize the role of historical specificity in explaining
dynamics of the current SES (SES path dependency). Such a social-ecological resilience (Box 1)
prism is largely missing in the study of pastoralism, despite pastoral systems having all the relevant
characteristics. Better recognition, characterization, and monitoring of slow and fast variables in
these systems, as well as their interactions and dependencies is paramount, as they condition the
resilience and sustainability of pastoralism. With this framework in mind, the following sections
address the elements needed to progress in this direction.
Need to move toward transdisciplinarity
Pastoralist traditional knowledge is inherently holistic,49, 50, 51, 52, 53, 54 evolving from
sustainable use of resources with millennia of adaptive management systems in their environment.
In contrast, pastoralist research tends to be compartmentalized, failing to inform decision making in
external interventions and national policies. Both scholars and practitioners have repeatedly pointed
to the lack of comprehensive SES understanding as an underlying cause5,55 of current pastoralist
crises.
Advances in both sustainability science and practice will require transcending disciplinary
boundaries.56 This need is particularly important for strongly coupled SES—not because such
systems would be particularly vulnerable or less resilient, but rather because connections between
their elements and unexpected reactions due to nonlinearities may make them more difficult to
manage or understand. In these systems small changes in one factor, such as a change in livestock
prices, land tenure, access to markets, or a drought event (fast variables), can cause abrupt,
cascading effects at different nested system scales resulting in system-wide reorganizations,36
displaying consequences in apparently unrelated factors. Changes in slower variables, such as land
degradation57 or shrub encroachment,58 can also determine systemic changes. As the inter-relations
between people and ecosystems change in scope and intensity, with globalization increasing the
scales at which interactions occur, understanding the dynamics of SES through integrated
approaches is needed to foster sustainable development.59
Although growing interest in SESs has increased collaboration across disciplines, and between
science and society in general,60 this progress is not yet well reflected in the science and policy on
pastoralism.5 Admittedly, the degree of integration of the economic, social, and environmental
dimensions has been growing in pastoralism research.61, 62, 63 Yet, the focus remains, at most, on
a reduced number of interactions (“environmental × economic,” “economic × social,” etc., e.g., see
Figure 2). Such limited integration has led to economic and environmental misconceptions that still
permeate policy-making.1,4,19, 20, 21, 22, 23 A more integrated study of pastoralism will greatly
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
improve our understanding of past transitions as well as ease the path toward achieving
sustainability goals.
In this perspective we show how moving toward a more holistic, transdisciplinary (Box 1), and
global understanding of pastoralism will support the 2030 Agenda for Sustainable Development.16
We refer to similar transitions brought about by Farming System Research to agricultural
development.72,73 Arising as a response to reductionist views and unexpected negative trade-offs
from modern interventions, the Farming Systems field recognized the need to become more
integrative, systemic, and comprehensive, as well as needing to account for multiple spatial and
temporal scales, including historical developments and geographic traits. This led to a new
paradigm in late 1970s characterized by interdisciplinarity, and even transdisciplinarity involving
farmers in the research process.72,74 To advance in this direction we endorse an approach that uses
(1) evidence across the social, economic, and environmental dimensions of sustainability, (2)
diverse geographic contexts and scales to capture emerging properties and allow for cross-system
comparisons, and (3) timescales from the distant past to the present, allowing for the analysis of
trajectories of pastoralist societies. We believe that such a diachronic, multi-scale analysis will not
only add to the science of pastoralism but will also reveal the shortcomings of different
development and governance interventions.75,76 Such analysis will help uncover novel and
transformative approaches to sustainable futures, thus laying out potential pathways to help solve
some of the problems pastoralist systems face today. We discuss the use of suitable sustainability
indicators, adapted to pastoralist systems, that support the identification of metrics that are locally
based, but globally relevant. Finally, we elaborate on why the elements listed above are critical and
how a common approach can integrate them across different spatiotemporal scales.
Call for comparative studies across time and space
Unsurprisingly, in pastoralism research the more comprehensive the interdisciplinary effort has
been, the narrower the geographical extent at which pastoralism has been studied. Despite some
integrative works,63,77, 78, 79, 80 uncoordinated efforts have resulted in isolated, local, or single-
system studies that limit synthesis and generalizations across pastoralist systems. Such gaps have
hampered transferability of policies, practices, and lessons between systems and cultural contexts.
Examining geographical gradients and contexts
The geographic context of pastoralism is often characterized by the intersection of multiple
environmental and social gradients. These typically include productivity gradients within a given
pastoralist system, with limiting climatic factors that are optimally and dynamically exploited
across seasons.17 Examples include the optimal use of pastures in Spanish81 or Sahelian82,83
transhumance corridors extending over 1,000 km, the strategies to exploit patchy and ephemeral
dryland resources among Fulani herders17,83 or the management of reindeer herd movement to avoid
seasonally parasite-infested areas.84 Examples are also available for aridity-humidity gradients
overlaying with distance to markets or population density, shaping center-periphery gradients in
Eastern African pastoralist systems.85 Such heterogeneity comes with differing development
challenges and calls for adaptation of policies to local contexts.
Similarly, different political, economic, and cultural contexts may reveal differences in pastoralist
adaptations under otherwise similar ecological conditions and explain large-scale land-use changes
driving pastoral ecosystems to collapse. An example of this is the differing influence in the
Serengeti-Mara ecosystem surroundings exerted by the contrasting open versus state-controlled
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
economies of Kenya and Tanzania, respectively, causing much wider land-use change and
fragmentation in Kenya.86 Conversely, a less-open economic and political system in China's Inner
Mongolia Autonomous Region, where less space is given to traditional land management than in
Mongolia, exemplifies drivers of land degradation.87, 88, 89 In the same line, cultural contrasts and
modernization events with differential diffusion across communities have uneven consequences,
well-illustrated by the penetration of firearms from conflict areas in Sudan that put at a
disadvantage those accessing guns later.90
Figure 2. Examples of shortcomings at integrating social, economic, and environmental
dimensions needed to understand pastoralism
(A) Social patterns explained through environmental-economic perspectives.
The cattle complex is often understood as a group of traits of socio-cultural origin and oftentimes
dysfunctional or irrational. However, an environmental and/or economic perspective brings
understanding to the functional value of these choices. For example, long-horn cattle, first thought
of as of low productivity, have a necessary trait to facilitate heat dispersal in tropical
environments.64 Similarly, the accumulation of cattle before a drought can be better understood as a
strategy to maximize the chances to rebuild the herd after the stress time.65
Image credit: Sarah McCans/Wikimedia Commons: Ankole_Cattle.jpg (cc-by 2.0).
(B) Environmental patterns explained through economic perspectives.
Overgrazing is an overarching concept routinely attributed to pastoralist practices, implying that
grazing practices lead to degradation. Suspected factors are the use of livestock species that are ill-
suited to local conditions, or climatic changes that force shifts in plant communities and increase the
sensitivity to grazing. Evidence from Northern Africa,66 the Middle East,67 and reindeer-grazed
lichen rangelands in Scandinavia68 indicate, however, that fodder provision policies are probably
behind many land degradation events by supporting herd sizes beyond sustainable numbers.
Image credit: Helge Busch-Paulick (Grand-Duc @ Wikipedia)/Wikimedia Commons:
Reindeer_in_finnish_fell-2.jpg (cc-by-sa 3.0).
(C) Economic patterns complemented by environmental perspectives.
Life-cycle analysis of intensive versus extensive livestock keeping indicates high greenhouse gas
emission impact per product unit of the latter, but such economic evaluation does not consider many
valuable ecosystem services that pastoralists are known to provide to society. However, greenhouse
gas impacts are greatly balanced by the wide array of environmental benefits provided by the
extensive system.69 Additional biological effects such as the lower albedo from forests also nuance
life-cycle analysis considerations.70,71 Image credit: Dguendel/Wikimedia Commons: Cap_Gris-
Nez,_flock_of_sheep.jpg (cc-by 4.0).
Thus, to synthesize relevant patterns and processes across systems, pastoralism research should start
embracing such multidimensional variability, along with its geographical variation (gradients).
Research designed to cover variability in multiple factors will allow us to (1) better separate or
identify drivers of change and (2) better understand system dynamics, nonlinearities, and the
potential for successful interventions. The former is particularly relevant for central and debated
topics in pastoralism research, such as rangeland degradation. Traditional grazing management, i.e.,
management by indigenous or traditional pastoralists, has often been seen as the driver of rangeland
degradation, even if governance/management factors, such as regulation of use or botanical
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
knowledge and non-equilibrium ecological dynamics, show that traditional management has a
neutral or beneficial effect on rangeland conditions.1,91,92 This misunderstanding of traditional
practices has resulted in policies that encourage or mandate sedentarization and land privatization
under the pretext that mobile pastoralists have mismanaged and overstocked rangelands.93,94 Only if
we measure covariates (social, environmental, and economic dimensions) over gradients or
contexts, will we be able to identify the slow and fast variables at different scales that affect
resilience. Such gradients, covering variability for the relevant domains, will often require
comparing apparently disparate systems from across the world (see Figure 1), to further our
understanding of roles of governance, development, and market access, among other factors, on a
number of processes in pastoralist systems. Such analyses can encompass such diverse systems as
cattle herders in Switzerland, horse pastoralists in Mongolia, and sheep herders in Algeria. In other
cases, or for particular questions, sufficient variability will be found at smaller spatial scales,
regionally or nationally.
Understanding the importance of scale
Noteworthy is also the integration across scales. Pastoralism has been described as a complex
system characterized by fractal properties: properties or behaviors that are self-similar at different
spatial or temporal scales. This applies both to territorial81 and socio-economic domains.95 The
concept of “nutag” (homeland) in Mongolia by which herders interchangeably refer to their grazing
territory, province, or country depending on where they are,96 or clan-related proverbs in the Horn
of Africa and Middle East, e.g., I and Somalia against the world; I and my clan against Somalia; I
and my family against the clan; I and my brother against the family; I against my brother,97 provide
simple examples of social scalability. Such properties allow systems to scale up or scale down
quickly in response to perturbations. Recognizing the properties of such systems is important for
understanding escalating responses (e.g., violence, herd, or people movements). Management
options or interventions should therefore take advantage of such self-organized nestedness as a
source of resilience, because disrupting it, either unintentionally or deliberately by, e.g., bringing
alternative, top-down imposed schemes, may also impair adaptation strategies and lead to undesired
states as, for example, poverty traps.98,99 This is because the components of the nested social
organization are flexible and contingent on the issues or perturbations that affect the functioning of
the SES. This flexibility is necessary because many pastoral systems are in harsh and highly
variable regions (see definition of pastoral systems in Figure 1). When systems of management
become rigid in any single scale (as in sedentarization), pastoralism can become vulnerable and
therefore fall into poverty unless inputs into the system increase (e.g., alternative livelihoods). Yet,
these scaling properties are rarely recognized, let alone studied formally. To advance understanding
of these dynamics, we suggest developing comparative studies across time, geographic, and socio-
political contexts and over multiple scales in time and space. There are examples where common
factors examined in separate studies can be integrated. This is the case for geographically explicit
factors that condition the practice of pastoralism in environmental and socio-economic terms, both
at the local Southern Ethiopian scale100 and at the regional East African one.85
Temporal depth is key to advance understanding
A more comprehensive understanding of the past may help identify and prepare pastoralists for
ongoing crises and critical tipping points, even though environmental drivers may have become less
determinant in the present than they were in the past, and socio-economic drivers have potentially
gained relative importance.101,102 Accessing data from the past is challenging, however, and more so
the farther back in time we look. Consequently, most of the integration has concentrated on the
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
recent past.81,103, 104, 105, 106 Conversely, the long-written history of pastoralism and the
abundance of archaeological records allow for integration along longer time periods. Since its
emergence in Western Asia, at around 10,500 years before present for goats (ca. 500 years later for
cattle), the development of pastoralism has been a process that involved both social and
environmental drivers and varied from region to region.107, 108, 109, 110, 111, 112, 113, 114 In
Africa and Eurasia, the development of pastoralism was connected to increased variability in
rainfall114, 115, 116 but also to socio-economic dynamics. The repeated emergence of pastoralist
societies and their shared attributes today is arguably associated with pastoralism being an optimal
strategy adapted to harsh and uncertain environments.14 Yet, pastoral societies have undergone
several important transitions, with small societies growing in number and complexity up to
empires117, 118, 119 followed by crises or collapses.120 Comprehensive understanding of past
major transitions is still meager at best, with many crises attributed to environmental drivers due to
a lack of consideration of social, political, and economic perspectives.105 Modeling of complex
systems based on archaeological evidence that study pastoralist settings offer very promising
results.121 Along these lines, an important remaining archaeological question is whether pastoralism
developed as local adaptation within extensive agropastoral communities, or separately as mobile
pastoralism.108, 109, 110,122,123 The past is therefore interesting by itself, with relevant
archaeological questions, such as the origin of pastoralism. However, the past is also relevant to
understand the present, since it can be connected to current processes through comparative
transitions, whose study is possible thanks to improvements in technology, data, and
interdisciplinary analysis.
Integrative knowledge management methods toward a solution
Exploring pastoralism transitions
We propose a comparative research approach based on aligning the trajectories of diverse
pastoralist systems through history, which allows the inclusion of all the key elements described
above. This approach is motivated by the recurrent pattern of decline and recovery observed for
most pastoralist systems, whether measured as economic or environmental decays. For instance,
land degradation pulses followed by recovery have been observed in Australia124 and Mediterranean
Europe125; income reduction caused by high human population growth reverts after countries reach
a demographic transition, when pastoralists achieve secure income and specialization in high-value
niche markets.126, 127, 128 Some downturns, however, lead to system collapse. This has been the
case in Southern Finland, for example, where intensive meat production completely displaced
traditional extensive practices, and associated High Nature Value habitats have consequently
become the most threatened in the country.129 We build on the qualitative similarities of these
trajectories and theorize that if thresholds or tipping points could be identified through systematic
empirical data collection and analysis, interventions could be undertaken to prevent collapses or
accelerate recoveries across all domains.
We find that the development of Kuznets curves and related theories130 serves as a useful analogy in
this regard. These curves were born as a graphic description of the process where nations
undergoing industrialization first see their economic inequality increase, and later decrease. The use
of the curve has been extended to diverse economic indicators against time—suitable for exploring
trajectories of single systems/nations131,132—or against development metrics or indicators of
economic growth—used either to explore the evolution of a system or to compare across systems.133
Perhaps more influential have been extensions that focus on environmental degradation,130
suggesting that environmental quality first decreases and then starts recovering at certain levels of
development, industrialization, and/or environmental awareness of the population. While some
empirical evidence gives support to the shape of the curve, its generalization has been debated,134
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
working better for some indicators than for others. We do not claim that pastoralism transitions
follow a particular U-shaped curve, but propose compiling empirical data and exploring patterns in
a similar way along an x axis that indicates time or development. We believe the concept is useful
for comparing disparate pastoralist systems that are undergoing transitions and are at different
development stages. Indeed, we find some examples where pastoralism has been used in the
representation of such curves.135
If we take overall pastoralism in the Sahara as a specific case, where efficiency of rangeland use is
the metric of focus, we would observe a decline, followed by a long hiatus, and a more recent
recovery. The decline can be explained by the climatic trends leading to the Saharan desiccation.136
Yet, the recovery has not happened because of the climatic trend reverting, but rather because
technological developments (e.g., truck transportation) have allowed pastoralists to access and graze
short-lived pastures again.137 We are able to observe the decline-recovery dynamic described by
Kuznets U-shaped curves once we integrate elements from archaeology and modern agronomy.
Few are the cases for which we have such an understanding, and Figure 2 offers examples of the
shortcomings derived from trying to understand pastoral systems without integrating information of
several dimensions.
In Box 2 we illustrate the application of this approach to three different pastoralist systems and for
longer time frames, based on qualitative interpretations. In the example, we draw on how Goal 2,
Zero Hunger, from the Sustainable Development Goals16 evolves along time to measure one aspect
of sustainability, using food security as a specific metric/indicator. Comparison of food security
aspects in the distant past and in the present is challenging, as it is for other variables, because of
limited data availability or quality for past indicators (Figure 3; see also section on indicators
below). Osteoarcheological studies138 and new developments around palaeodemography139 promise
feasible comparisons. This conceptual exercise exemplified in Box 2 should be understood as a
proposition for future exploratory research, where the y axis should preferably include diverse
quantifiable indicators across sustainability domains (see next section). We also show in Box 2 how
one can similarly explore trajectories for single pastoralist systems through time (e.g., Pastoralism
in Mongolia; Box 2 Figure A), or combine disparate pastoralist societies (East Africa, Central Asia,
Arctic; Box 2 Figure B) in a single curve and at different points in time using an appropriate x axis.
The metrics we plot here on the y axis indicate that a system may move from a situation of high
sustainability (e.g., high food security) and down to levels that approach sustainability limits when
crossing critical domains. From those critical domains, the system may either collapse or recover as
in the panarchy adaptive cycle.48 Such critical domains could be determined empirically through the
comparisons of multiple systems, and considering sustainability indicators pertaining to different
dimensions. Note that, when mapping different pastoralist systems along the same curve, we may
observe responses at slower or faster paces. The pace of change is likely related to the slow and fast
variables that affect where they are on the adaptive cycle,48 such as the placement of pastoralist
societies along political or economic gradients of centrality-periphery. Pace of change can also
correlate with the degree of cultural exchange, the innovations incorporated and also with how they
enter into critical pitfalls or exit from them, with differing investments to conserve environmental,
social, or cultural elements. Such representations allow us to add temporal depth and compare
trajectories looking for common patterns in transitions across systems. The drivers of major
inflection points may be particular and revealed only when simultaneously studying connections
with variables in multiple domains. Transferability of policies and management measures from the
right-hand side of the curve to the left-hand side should be encouraged, allowing tunneling through
critical domains while avoiding points of no return.140
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Box 2
An exploratory exercise of three pastoralist trajectories
In a qualitative, exploratory exercise, we compare the trajectories of food security among three
pastoralist societies in Eastern Africa (Daasanach), Central Asia (Mongolian), and the Arctic
(Sami). The Daasanach's past in the area has experienced no major tipping point and their food
security is increasingly threatened by a growing population,141 evolving from biodiverse,
sustainably managed landscapes that the European explorers encountered in the 19th century142—it
could be therefore described as having a “slow pace” of transition framed in a periphery situation.
Mongolian pastoralists show a tipping point just before the expansion of the Mongolian
Empire117,143: restrictions by the neighboring Chinese Empire were overcome through extraordinary
good production years that would have triggered the dispersal of the population across Eurasia,
initiating a phase of improvement in food security later prolonged by technological
improvements—a “moderate pace” of transition. Sami reindeer herders would have evolved from a
relatively rapidly deteriorating situation in the few last few centuries, coinciding with colonization
and political marginalization,144,145 to a very rapid transition that led them to adopt well-being
standards of developed economies. This shows a “rapid pace” of transition that is a consequence of
their placement close to the global centrality area of Northern Europe. Such differential paces and
their outcomes point to the potential of differential transition paces to avoid collapse and shift out of
the critical zone, or not. Moreover, the probability of a whole livelihood to collapse would be
related to the livelihood's sector in crisis and its capacity to affect all other sectors.
Box 2 figure. Dynamics and comparability of different pastoralist systems.
(A) Historical trajectories of food security domain for three different pastoralist systems plotted
against the log of years before present (BP); Daasanach from northern Kenya/southern Ethiopia,
Sami reindeer herders from Fennoscandia, and herders from Mongolia. Crashes, such as the one
observed for Pleuropneumonia and Rinderpest in the Daasanach curve would be present in all
pastoralist systems but are diluted in the general trend of the curve.
(B) Properties from different pastoralist systems at various times can be mapped along a
standardized measure of degree transformation of the livelihood (see text). Pastoralist systems
entering into the critical domain are prone to collapse.
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Figure 3. Model for developing pastoralism indicators that span across different time scales
The graph shows potential indicators (color bars) for social, economic, and environmental
dimensions, with some indicators spanning across all periods of interest, while others are only
available or relevant for more recent time periods. A few example indicators are suggested on some
bars (cursive). While long-term indicators will favor large-scale understanding of system dynamics,
finer scale and recent times will be more relevant when monitoring management actions. The graph
also depicts the involvement of decision makers and civil society in the development of these
metrics.
We insist that the approach illustrated by Box 2 should not be limited to one indicator in a single
domain (“food security” in the figure) but instead explore transformations in all domains, and their
interactions. But to move forward, from the qualitative conceptualization presented to empirically
based theories of pastoralism, requires first the development of a relevant set of sustainability
indicators. We also need a better characterization of such measurable indicators as descriptors of
slow and fast variables in different domains.
Sustainability indicators for pastoralism research and policy development
The research approach just discussed is not only useful to explore past transitions. Some of the key
metrics used to identify accumulating slow variables, critical fast variables, important transitions,
and tipping points of past pastoralist trajectories can, similarly, be used to monitor current
pastoralism sustainability and suggest paths for improving it in the future. Thus, sustainability goals
can also be used to guide the selection of these metrics within the framework of social-ecological
system resilience. In this context, a number of general guidelines and indicator frameworks have
been developed—e.g., Bellagio Principles,146 Sustainable Development Goals,147,148 or the IPBES
Conceptual Framework.149 Some review works extensively discuss the strengths and weaknesses of
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
diverse sustainability indicators,150, 151, 152, 153, 154 providing advice on how to choose the
right metrics, particularly for decision making and policy implementation.
Yet, all around the world, pastoralism is often subject to pre-determined monitoring schemes and
reporting indicators that are often externally codified, and that may not support the visions, goals,
and needs of pastoral communities. Moreover, indicators developed without community
participation often overlook, misrepresent or discount the knowledge systems and values that
underpin the connections between pastoral communities and their ecosystems.155,156 For instance,
development of these indicators have overlooked the particular characteristics of pastoralist systems
(see below); hence, their suitability for capturing the vulnerabilities of these systems requires
examination. As a first step we (1) summarize as a starting point, important characteristics of
pastoralism that could be reflected in sustainability indicators, (2) elaborate on the desired dual role
of the indicators (i.e., to gain long-term system understanding and to advance policy
implementation and evaluation) and their validity across temporal and spatial scales, (3) identify
available data sources or transferability of indicators across disciplines, and (4) present a set of
methods to put pastoralists at the center of the process of indicator development.
Considerable indicator work has been conducted for livestock production systems, but either for
intensive livestock farms or for particular regions.152,157,158 The shift in focus we call for here
highlights the need to consider pastoralism characteristics such as mobility dependence, common-
pool resources, low-input/low-output processes, and coexistence with natural or semi-natural values
(see definition of pastoralism in Figure 1). We stress that relevant indicators should thus capture
changes in these characteristics in addition to other indicators identified as relevant for strongly
linked SES. The effort made by the Sustainable Rangeland Roundtable provides a good starting
point with 64 potential indicators selected for rangeland systems.157 Some of the listed indicators
map well the pastoralism characteristics mentioned above (e.g., indicators related to common
access: land tenure, land use, and ownership patterns by land size classes; or indicators related to
coexistence with nature: population status and geographic range of rangeland-dependent species),
yet key indicators of mobility, and relevant indicators of resilient social structures are largely
missing. Overall, environmental indicators identified to date are viewed as comprehensive or
promising, whereas socio-cultural and economic indicators present more challenges, and are
recognized as being in an earlier stage of development.157,158 In fact, despite all the promising
indicators identified, researchers call for (1) further development of particular indicator themes, (2)
a more manageable list of indicators that can better evaluate impacts and ease of data collection,
and (3) to empirically test interactions between indicators, within and between dimensions.158 The
approach to explore trajectories presented here may offer this opportunity.
We have argued that a lack of a holistic system understanding is partially responsible for the
pastoralism crises observed, and we have advocated for analyses at broader temporal scales,
through pre-history to present times. We contend that some sustainability indicators may also be
suitable to explore past crises and find equivalents to recent or present ones (Figure 3). We note that
different proxies may be needed for different time periods. For example, indicators of food security
that may reflect resilience of pastoral communities could be quantified through different measures
of diet diversity, sourced from osteological composition from fossil bones for the past, from
questionnaires, or participant observation in the present—all of these would be relevant to the
example we present in Box 2. This will require careful consideration of paces of change (the slow
and fast variables mentioned above) and choice of indicator units as well as adequate quantification
of uncertainties. When possible, indicators should be measured with consistent methodologies
through time. Indeed, some indicators often used in paleontological or archaeological studies can be
extended to infer and also compare present states—e.g., ecometrics.159, 160, 161 In particular,
some relevant social and economic indicators will often be scarce for pre-history periods (such as
indicators of erosion of traditional institutions, market access, or market value) while others may
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
only be important at present (e.g., incorporation of new technologies, or impact of international and
national policies).
Importantly, indicators should be comparable for different geographical contexts and should not be
susceptible to small differences in methods or measures across settings (e.g., follow developments
of livestock grazing schemes such as the livestock units valid to compare across pastoralist
systems162 through space and time). We echo the calls for locally relevant yet globally applicable
indicators, and, when possible, consider interactions that link local to global sustainabilities in an
increasingly interconnected or telecoupled world, noted as particularly important in the global
livestock production system.163, 164, 165 Opportunely, valuable data are already available from
multiple sources, yet with patchy distribution—thus requiring substantial efforts to complement
such existing data.
To fill data gaps and identify indicators that are relevant for local community decision making, it is
paramount to work with pastoral communities within transdisciplinary, collaborative research
settings that are co-designed and co-produced.166, 167, 168 Universalism in indicator development
has often fallen short in recognizing culturally grounded perspectives of sustainability,156 leading to
calls for more active engagement of pastoral communities in research, in an effort to develop
indicators that better reflect their needs, views, and knowledge systems.155,169 Here, we propose
taking advantage of participatory methods to support the identification of indicators that are
culturally relevant and that provide legitimate knowledge-in-use for community decision making.
Such methods can take many different forms, including biocultural-oriented monitoring,156 models
of continual engagement,155 knowledge co-production,170,171 participatory action research,172 or a
Multiple Evidence Base approach for connecting different knowledge systems,173 among many
others. Endeavors in this vein can support the identification of metrics that are culturally
appropriate and attuned to both local needs and global priorities. Finally, collaborative indicator
development can serve a purpose in re-positioning pastoralists at the center of the research process
itself,171 recognizing them not just as stakeholders, but also as knowledge and rights holders.174
Given that many aspirationally participatory approaches perpetuate extractive dynamics with local
communities, we urge researchers to carefully consider recommended guidelines for genuine
engagement with indigenous and pastoralist communities.175
Indicator choice should not be carved in stone. Instead, paraphrasing Ostrom,176,177 we need a
multilevel, nested system of indicators, with both a set of variables relevant for studying a single
SES, as well as a common set of variables that allows comparison among pastoral SES. The
analytical framework presented here, the need to compare scales, and the need to integrate
pastoralists' interests should steer the development of an indicator classification, and stimulate a
global movement for collecting them. The IUBS-funded Global Integrative Pastoralism Program178
and other initiatives from the League for Pastoral Peoples179 or the International Livestock Research
Institute180 are following such steps and working to bridge these gaps, establishing databases that
are likely to be reinforced with the possible declaration of an International Year of Rangelands and
Pastoralists in 2026.181
Way forward
The understanding of trajectories in pastoralist societies, and the characterization of the factors that
drive them through globally accepted indicators, should be used to find solutions by informing
decision making at the macro level (e.g., national policies). When applying our research approach,
we propose it is possible to identify how external interventions interact with responses at the micro
level (pastoral undertakings; day-to-day local decisions), affecting this way the sustainability of the
practice. To achieve this, we call for the integration of pastoral knowledge, interests, and values182
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
in the indicator set through a co-production process183,184 as many solutions will actually arise from
pastoralist communities themselves.76
We acknowledge that our call for locally relevant metrics is sometimes in tension with the urgent
need to develop globally comparable indicators. On the one hand, metrics capturing information in
the same way regardless of place-based contexts often lack specificity.156 On the other hand,
indicators developed at the local level are often difficult to upscale, given their place-based
specificity. However, there are several mechanisms that allow for bridging the gap between local
and global indicators (e.g., Hicks et al.185). In particular, there are different processes in which
locally appropriate, yet globally applicable, indices have been developed, based on pre-defined
broad domains that are comparable cross-culturally (i.e., indicator groupings), but based on locally
adapted and culturally specific metrics and indicators.156,186 Alternatively, indicator frameworks177
also allow for comparisons across indicator classes. Some of these approaches draw inspiration
from middle range theories (i.e., generalizations that describe causal mechanisms within certain
contextual bonds, see Meyfroidt et al.187).
The identification of solutions for pastoralism is inherently complex, not only because they are
challenging to trace but also because the trajectories of the different sustainability domains are
tightly interdependent. While progress can be observed, e.g., in the trajectories of economic and
environmental domains, deep shortcomings in the social domain can drive the whole system to
collapse. An example of this would be the differential emigration of women.81 The loss of women's
knowledge and capacities, as in, e.g., dairy production,188 renders generational continuity
impossible (Figure 4A). Conversely, while well-directed early action in one single domain could
avert total collapse, approaching tipping points that are dangerously close to collapse would erode
existing knowledge. For instance, women's emigration may not reach the level of system collapse,
but may be high enough to affect women-associated dairy-processing knowledge, such as the
elaboration of cheese varieties whose sale can boost the pastoralist economy. Such unsatisfactory
social outcomes would negatively impact the future economic and environmental outcomes of the
livelihood, even if total collapse is avoided (Figure 4B). Other losses caused by approaching
critically low sustainability levels include species extinctions in the environmental domain, or loss
of culturally adapted product consumption patterns in the economic domain. Only by understanding
the need to act simultaneously across domains, early enough, and through well-informed action,
loss of knowledge, adaptation options, and wasted economic development potential would be
minimized (Figure 4C).
To enable such identification of solutions, it is essential to establish global databases (preferably
open access) that collect and avail data from a wide array of different pastoralist systems, and that
turn such data into comparable, transferable indicators. The transdisciplinary and holistic nature of
SES research should be extended to the solutions and implementation phase, where interventions
nested in the different sustainability domains must be not only simultaneous but also coordinated.
These phases should also include pastoralists from the systems targeted by the interventions. This
includes adaptive co-management where users and resource managers, operating at different scales,
can have the means to experiment, monitor, deliberate, and reactively manage resources at the local
level.189 It also includes adaptive governance, where the social dimensions associated with
management and decision making can be addressed.36
In summary, we encourage an approach where: (1) sets of relevant indicators are compiled across
scholars, practitioners, and pastoralists globally; (2) the transferability of indicators across systems,
temporal dimensions, and scales is evaluated, and suitable standardizations applied; (3) their
suitability to identify system transitions, especially tipping points, is assessed; and (4) the subset of
indicators found useful is collaboratively appraised and their use further promoted to monitor
pastoralism sustainability both globally and locally, aiming at enhancing favorable policies across
scales. We believe that the indicators developed through the proposed research approach may
FINAL DRAFT, published in One Earth 4(5): 651–665 https://doi.org/10.1016/j.oneear.2021.04.012
Figure 4. Evolution of indicators of sustainability domains under different policy/intervention
scenarios
Graphs represent a selected social (red line), economic (blue line), and environmental (green line)
factor as an example, related to the explanation in the main text. Stars represent interventions.
(A) System collapse dynamics driven by a fatal deterioration of social factors after a partial
recovery of economic and environmental factors.
(B) A well-guided single domain intervention (e.g., on economics) can avoid the collapse vortex,
yet it is not enough to successfully escape from low sustainability levels that can bring to future
collapses or avoid the loss of some elements in each domain, such as species extinction or loss of
cultural traits.
(C) Early interventions that are simultaneous across all sustainability domains can avoid critical
levels and facilitate the achievement of pastoralism's full potential.
become better integrated into global science-policy interfaces and assessments, contributing to
better recognize pastoralism in global environmental and economic forums or sustainability
agendas. This will pave the way for producing better-informed and more effective policies and
interventions, which will ensure pastoralism resilience and sustainability in the long run.
