Reconciling conservation and development requires enhanced integration and broader aims: A cross-continental assessment of landscape approaches

Full text

Article
Reconciling conservation and development requires
enhanced integration and broader aims: A cross-
continental assessment of landscape approaches
Graphical abstract
Highlights
dWe learn from 380 practitioners using landscape approaches
in the subtropics and tropics
dGlobally, three distinct forms of applying landscape
approaches exist in practice
dPerformance is driven by duration and investments
across goals
dInternational stakeholders are key partners but often have
misaligned aims
Authors
Natalia Estrada-Carmona,
Rachel Carmenta, James Reed, ...,
Louise Willemen, Camilla Zanzanaini,
Wei Zhang
Correspondence
n.e.carmona@cgiar.org
In brief
Reconciling conservation and
development is urgent, particularly in
tropical and subtropical regions where
forest-agricultural landscapes are under
pressure. To this end, cross-sectoral and
integrated approaches to managing
agricultural landscapes are considered
necessary, potentially able to balance
divergent interests and enable
coordinated action. In practice, they are
diverse. We analyzed the features of
landscape approaches across three
continents and identify three types,
assess their performance, and provide
practical lessons to improve their design
and implementation to benefit people and
nature.
Estrada-Carmona et al., 2024, One Earth 7, 1–16
October 18, 2024 ª2024 The Author(s). Published by Elsevier Inc.
https://doi.org/10.1016/j.oneear.2024.08.014 ll

Article
Reconciling conservation and development requires
enhanced integration and broader aims: A cross-
continental assessment of landscape approaches
Natalia Estrada-Carmona,
1,20,
*Rachel Carmenta,
2
James Reed,
3,4
Ermias Betemariam,
5
Fabrice DeClerck,
1
Thomas Falk,
6
Abigail K. Hart,
7
Sarah K. Jones,
1
Fritz Kleinschroth,
8,9
Matthew McCartney,
10
Ruth Meinzen-Dick,
11
(Author list continued on next page)
SUMMARY
Expectations for agricultural landscapes in subtropical and tropical regions are high, aiming for conservation
and development amid climate change, unfair trade, poverty, and environmental degradation. Landscape ap-
proaches (LAs) are gaining momentum as means to reconcile expectations, although they face multiple chal-
lenges, including unclear distinctions among LAs and stakeholder involvement. We studied 380 LAs from
three continents via questionnaires with landscape managers (2012–2015 and 2021) and identified three LA
types through cluster analysis: an ‘‘integrated’’ type with longer-term, multisectoral goals involving various
stakeholders early in the design and two shorter-term types focused on sectoral priorities of preservation
or production. Better-performing LAs are associated with longevity, inclusivity, and diversified investments
across goals, notably those enabling social justice. International stakeholder analysis shows broad support
for LAs but identifies gaps between support and LAs’ needs. The growing interest in LAs is promising. Yet,
underpinning effective and lasting LAs that reconcile multiple expectations requires better support.
INTRODUCTION
Multifunctional landscapes are crucial to achieving develop-
ment, conservation, land degradation, and climate change
goals. However, efforts to achieve these goals individually
through site-level and siloed interventions such as land sparing,
1
green revolution innovations,
2
or ‘‘fortress’’ conservation
3
often
exacerbate environmental challenges and social injustices,
1
Multifunctional Landscapes, Bioversity International, Montpellier, France
2
School of Global Development and the Tyndall Centre for Climate Change Research, Norwich Research Park, University of East Anglia,
Norwich, UK
3
Governance, Equity, and Wellbeing, CIFOR-ICRAF, Bogor, Indonesia
4
School of Global Development, University of East Anglia, Norwich Research Park, Norwich, UK
5
Soil and Land Health, CIFOR-ICRAF, Nairobi, Kenya
(Affiliations continued on next page)
SCIENCE FOR SOCIETY In subtropical and tropical regions, agricultural landscapes supply essential food,
feed, and fiber, among various ecosystem services to both rural communities and growing cities worldwide.
These landscapes are biodiversity hotspots and cultural centers but face pressures from climate change,
commodity trade, poverty, and environmental degradation. Holistic efforts like landscape approaches
(LAs) are increasingly used to balance trade-offs at the landscape level and address these interconnected
challenges to achieve multiple goals, such as conservation, sustainable production, livelihoods, and gover-
nance. Our results from three continents highlight distinct ways to apply LAs, despite shared challenges hin-
dering their performance. All LAs focus on capacity building and coordination, while long-term support, in-
clusion, and a diversified portfolio of interventions improve performance. Our findings offer actionable
insights, notably for international organizations, to enhance support for LAs.
One Earth 7, 1–16, October 18, 2024 ª2024 The Author(s). Published by Elsevier Inc. 1
This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
ll
OPEN ACCESS
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

particularly in tropical and subtropical regions.
4
Experts in devel-
opment and conservation argue that landscape approaches
hold promise to pursue multiple goals simultaneously, achieving
effective and equitable multifunctionality.
5–9
Landscape ap-
proaches can help to bring together stakeholders from different
sectors for collaborative and integrative planning to develop
landscape-level interventions that enhance socio-ecological
systems’ long-term resilience and sustainability.
10–12
In practice,
landscape approaches lack a standardized definition and spatial
delimitation but typically subscribe to a set of widely accepted
principles.
13,14
While these principles and aims are generally
acknowledged, landscape approaches take multiple forms and
go by various names that involve multiple concepts, such as
ecosystem approach, jurisdictional approach, biosphere
reserve, integrated watershed management, biological corri-
dors, foodscapes, and foodsheds, among others.
12,15
Thus,
the term ‘‘landscape approaches’’ (LAs) encompasses a diverse
set of efforts to tackle complex challenges at the land-
scape level.
The evidence confirms the wide application of LAs, which
cover different contexts and multiple continents.
12,13,16–20
How-
ever, in this suite of apparently diverse initiatives that identify as
LAs, there may be discernible distinctions in their efforts on the
ground. Regional evidence from Latin America and the Carib-
bean (LAC),
21
a global literature review,
7
and a group of case
studies from the Satoyama Initiative
12
point to distinct types of
LAs. Yet, whether those types exist or persist in other regions,
such as Asia or Africa, is unclear. Similarly, it’s unclear whether
the distinct types of LAs face the same range of barriers to oper-
ationalization identified in the scientific literature (e.g., Vermunt
et al.,
22
Pedroza-arceo and Weber,
23
and Meli et al.
22–24
) and,
more importantly, whether the factors associated with LAs’ per-
formance are shared or vary across types. While substantial
knowledge exists on various aspects of LAs as a general
concept,
6,9,10,16–18,20,25,26
detailed analyses of types and their
performance are available only for LAC.
21
Therefore, a cross-
continental analysis of the comparable regional assessments
from LAC,
16
Asia,
17
and Africa
18
is essential for understanding
what LAs are, how they differ, and how performance is associ-
ated with contexts and types. More broadly, identifying LA-
type-dependent enablers will be central to guiding, with
evidence, the growing endorsement and support of LAs by
international organizations, practitioners, funders, and private
enterprises.
6–9,26
Multiple intervention strategies that enable achieving sustain-
able development and conservation are put forward and imple-
mented in LAs. For example, intervention strategies to improve
agriculture sustainability cover a wide range of practices,
including conventional and, more recently, agroecological prac-
tices, notably in (sub)tropical regions.
27–29
Similarly, intervention
strategies may be used to nudge behavioral change through
standards or certification, land use zoning, or coordination.
29
Also, intervention strategies can focus on enabling certain ideas
of social justice through income generation, training, food secu-
rity, or health.
30–32
In practice, evidence of how those interven-
tion strategies or bundles of strategies are implemented across
LAs and how these influence LAs’ performance remains limited.
Carmenta et al.
21
analyzed LAs in Latin America and proposed
that three central intervention strategies (aiming to foster behav-
ioral change, enable social justice, or improve agriculture sus-
tainability) were operating within and across their typology of
LAs, and training and coordination were the most common
sets of interventions.
21
However, they found no consistent
pattern between intervention strategies and performance across
LA types.
21
This is an important knowledge gap, and empirical
evidence of how intervention strategies used by LA practitioners
are associated with performance is essential to better support
LAs and understand their effectiveness.
6,14,26
Interest in LAs is high, despite the substantial and sustained
support these complex efforts need. Some LA proponents sug-
gest three catalysts for achieving desired change in LAs: gover-
nance, markets, and finance.
9
The scientific literature has
focused on the governance aspects of LAs,
22,33,34
with the other
two catalysts of markets and finance receiving less attention.
Markets and finance can greatly contribute to delivering multiple
benefits at scale and innovating synergistic solutions to complex
challenges.
3,9,19
Yet, the involvement of international stake-
holders, including organizations directly linked to markets and
finance opportunities, such as the private sector, remains con-
tested. Some authors indicate a high involvement,
35
whereas
Jeff Milder,
12,13
Marcela Quintero,
14
Roseline Remans,
1,15
Diego Valbuena,
16
Louise Willemen,
17
Camilla Zanzanaini,
18
and Wei Zhang
19
6
Innovation Systems for the Drylands Program, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Hyderabad, India
7
Water Program, The Nature Conservancy (TNC), Sacramento, CA, USA
8
Ecosystem Management Group, Department of Environmental Systems Science, ETH Z€
urich, Z€
urich, Switzerland
9
Institute of Environmental Planning, Leibniz Universit€
at Hannover, Hannover, Germany
10
Sustainable Water Infrastructure and Ecosystems, International Water Management Institute (IWMI), Colombo, Sri Lanka
11
Transformation Types, International Food Policy Research Institute (IFPRI), Washington, DC, USA
12
Rainforest Alliance, New York, NY, USA
13
Department of Natural Resources, Cornell University, Ithaca, NY, USA
14
Research Strategy and Innovation, International Center for Tropical Agriculture (CIAT), Cali, Colombia
15
Glocolearning, Brussels, Belgium
16
Landscape Architecture and Spatial Planning, Wageningen University and Research, Wageningen, the Netherlands
17
Faculty of Geo-Information Science and Earth Observation, University of Twente, Enschede, the Netherlands
18
Nature Makers Lab, Hong Kong, Hong Kong SAR, China
19
Natural Resources and Resilience Unit, International Food Policy Research Institute (IFPRI), Washington, DC, USA
20
Lead contact
*Correspondence: n.e.carmona@cgiar.org
https://doi.org/10.1016/j.oneear.2024.08.014
ll
OPEN ACCESS Article
2One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

others flag an absence or low and conflicting engagement.
36
The
engagement of other international stakeholders indirectly sup-
porting market and finance access in LAs by generating evi-
dence and leveraging funds and capabilities, such as ‘‘research
for development’’ (R4D) organizations (R4D), remains heavily
understudied. Yet, certain R4D agencies, such as CGIAR, deliver
public goods that influence the type and management of agricul-
tural lands and impact the potential for multifunctionality in these
areas. Increasingly, the CGIAR recognizes the landscape as a
unit of intervention and analysis for systems transformation,
particularly for achieving multiple interconnected human,
ecological, and agricultural development goals.
37
Despite their
interest, action and influence in this space, there has been little
empirical work assessing how R4D advances the concept and
implementation of LAs.
Here, we fill these multiple knowledge gaps by analyzing the
full set of 357 LA projects, programs, or research initiatives,
collected in 2012–2015 and deriving from three regional assess-
ments (LAC, Asia, and Africa)
16–18
(hereafter referred to as the
‘‘cross-continental dataset’’). We excluded the European LA
assessment data due to differences in protocols precluding
comparative analyses. Additionally, we supplemented the
cross-continental dataset with additional data collected in
2021 from a set of 23 LAs with CGIAR researchers leading or
engaging and covering the same three regions (hereafter
referred to as the ‘‘extended dataset’’). The extended dataset
enabled a deeper evaluation of a key R4D consortium (CGIAR)
contributing to sustainable development via LAs in the studied
regions. We analyzed the LAs through a cluster analysis by using
multiple-factor analysis (MFA) and the hierarchical clustering of
principal components (HCPC). The cluster analysis enabled us
to identify central distinctions and similarities between LAs’ op-
erationalization across continents. Additionally, we conducted
a non-parametric analysis to identify variables associated with
LAs’ performance in the cross-continental dataset and each
cluster of LAs. Last, we analyzed all listed partners and specif-
ically analyzed international organizations or stakeholders
engaged to assess those present or missing across LAs. Our
findings suggest distinct ways to operationalize LAs, resulting
in different performance levels. We also found common chal-
lenges shared across LA types, and our results support the
notion that ‘‘how’’ these complex efforts are supported could
be more relevant than the ‘‘what’’ is supported. Our research,
grounded on LAs from three continents, contributes to
advancing the theoretical knowledge of LAs and pinpointing
key opportunities to improve their practical application. We pro-
vide actionable guidance for international organizations and
stakeholders to more effectively support LAs in achieving multi-
ple goals that benefit people, nature, and future generations.
RESULTS AND DISCUSSION
Distinct LA types implemented on the ground
It is important to first note that the data used in this analysis is
based on questionnaire data reported by individuals leading or
managing the LAs. This may introduce bias, as respondents
could conceivably choose to submit information that they
perceive would make a positive impression or, conversely, with-
hold information they consider to be negative. However, respon-
dents were made aware that donors would not view the data and
that the questionnaire was to learn from current strategies,
including their difficulties and barriers, to advance research
and practice, which we hope helped to reduce bias. Nonethe-
less, the potential for introduced bias exists and should, there-
fore, not be overlooked when interpreting the results (see Study
limitations and caveats for more details).
The cluster analysis on the cross-continental dataset (357 LAs)
confirms that central distinctions between LAs exist
7,12,21
(Fig-
ure 1). This finding remains consistent when assessing
the extended dataset involving the CGIAR-associated LAs
(Table S4). A wide range of categorical and quantitative vari-
ables, strongly and significantly associated with each cluster
(or ‘‘type’’ of LA). These distinctions cover multiple features,
including duration since inception (LA longevity), dominant land
uses, management options employed, motivations driving LA
inception, sectors and stakeholders engaged, and actions or in-
vestments led by LAs (Figures 2 and 3;Table S3). Validation of
the clusters is strong, with <10% (8 of 100) being misclassified
through a different test (Table 1;experimental procedures;
Table S1).
We labeled the LA types evident within the clusters as ‘‘pres-
ervation,’’ ‘‘production,’’ and ‘‘integrated’’ and thereby highlight
the salient differences. For example, the most common actions
in production LA types relate to agriculture, while four of the
seven actions in the conservation domain were pursued only
rarely (<20% production LAs) (Figure 2;Tables S2 and S3). Pres-
ervation LA types focus on the conservation of water, soil, or for-
est resources. We selected the term ‘‘preservation’’ to avoid
confusion with the conservation domain in the original question-
naire. In preservation LA types, actions such as ‘‘other commu-
nity-based natural resource management activities’’ and
‘‘training or capacity building programs to support natural
resource management’’ were dominant in these landscapes
(>80% preservation LAs) (Figure 2;Tables S2 and S3). Mean-
while, integrated LA types invested in actions across domains
(Figure 2;Tables S2 and S3). We found that production- and
preservation-oriented LAs tend to list significantly narrower mo-
tivations, fewer stakeholders and sectors involved, and fewer
and less diverse actions across domains (e.g., sustainable agri-
culture, conservation, coordination, and livelihoods), which
aligns with previous LA-type assessments that also identified
more sectoral-like LAs.
7,21
Therefore, even though landscape
managers identified their initiatives as integrated LAs, in fact,
they appear only marginally integrated and, instead, the preser-
vation and production types tend to have a sectoral lens focusing
on specific goals. In contrast, those clustering as integrated
types did have broader aims, actions, and stakeholder engage-
ment and can, therefore, be considered LAs reflecting more inte-
grated efforts (Figure 2;Table S3).
Although commonly and collectively called ‘‘integrated LAs’’ in
the scientific literature, only a third of LAs in our dataset appeared
as integratedLAs (Figures 1 and 2). Compared withthe production
and preservation LA types, the integrated LAs focus on more
diversified landscapes with a greater diversity of prevailing major
land uses (Figure 2B). These integrated LAs last longer, involve
more sectors (e.g., agriculture and natural resources; Table S3)
and actively invest across domains (Figures 2A, 2E, and 2G).
They also tend to engage with more stakeholders, such as local
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 3
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

government leaders or staff, local farmers’ associations, local
non-governmental organizations, local universities or research
centers, and women’s associations (Figure 2FandFigure 3).
Stakeholders—notably local stakeholders—in the integrated LAs
exhibit greater participation in early design stages (Figure 2). Inte-
grated LAs also mobilize more intervention strategies commonly
used in conservation and development (Figure 2H).
Therefore, the integrated LAs align most closely with seminal
work describing good practices for LAs,
10,35
previous descrip-
tions of integrated LAs from literature reviews,
7
and regional ty-
pologies of LAs.
21
Carmenta et al.
21
found two integrated types
when conducting the LAC regional assessment: one more
focused on ‘‘participation and legislation’’ and the other on ‘‘cer-
tification, institutions, and participation.’’ Yet, when analyzing
the cross-continental data, this differentiation disappears, re-
sulting in only one cluster of integrated LAs. Overall, the charac-
teristics of integrated LAs suggest that these efforts tend to
incorporate inclusive landscape governance at the design stage
and potentially incorporate local knowledge, social relations,
stakeholder needs, and grounded aspirations. Encouragingly,
although the integrated LAs in the cross-continental dataset
only represent about one-third of all approaches, they are
distributed evenly across the three continents (Figure 1).
Strong differences distinguish the sectoral production and
preservation LAs. Production LAs were developed in response
to more motivations across domains but aimed to resolve issues
within shorter time frames and involving fewer sectors, stake-
holders, and actions (Figures 2A, 2D–2H, and 3;; Table S3). Pres-
ervation LAs exhibit the most comprehensive management
options emphasizing monitoring and evaluation, adaptive man-
agement, and baseline and geospatial assessments (Figure 2C;
Table S3) but had narrower motivations for LA inception (Fig-
ure 2D). Yet, preservation LAs mobilized more actions across
all domains and intervention strategies, except those related to
agriculture (Figures 2G and 2H; Table S3).
Multiple factors, such as local conditions, priorities, and the
overall landscape starting point, may determine the LA type
applied.
7,12,19
For example, LAs in landscapes predominantly
comprised of conserved and protected areas with high cultural
values vs. predominantly degraded lands with fragile human-na-
ture interdependence face different challenges and demand
contrasting efforts to operationalize LAs.
13
However, the pre-
dominance of LA types with sectoral lenses raises the question
of whether this dominance is due to the landscape’s initial start-
ing point or to a large proportion of LAs having their origins in
more sectorial approaches. The latter may now be evolving
and adapting toward integrated LAs or simply just adopting
new nomenclature around their activities to maintain access to
funding, given the rising interest and investment in the concept.
All landscape managers agreed with our LA inclusion criteria
when completing the survey and believe their interventions to
classify as an LA (experimental procedures). This potentially
means that, within LAs with a sectoral focus, their efforts are
potentially a step further toward integration than purely sectoral
projects; however, we did not collect attribute data on the latter.
It may also mean that there are divergent understandings of what
constitutes an LA, which has been cited as a potential problem
arising from the current lack of definition, or, in the worst case,
that the term is being co-opted and applied to essentially secto-
ral approaches. However, all initiatives are likely attempting
some form of integration owing to the finding that, on average,
none of the LA types presented null or extremely low investments
across domains or investment strategies, except for increased
inputs (Figure 2). This evolution from sectoral to integrated ap-
proaches has been reported in the scientific literature from a
small set of case studies.
19
Figure 1. Distribution and distinct types of LAs across three continents (extra 23 LAs identified in 2021 and cross-continental database iden-
tified in 2012–2015)
LAs are color coded based on the cluster or type.
ll
OPEN ACCESS Article
4One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

Figure 2. Characterization of LA types (clusters) based on continuous variables
Values are standardized for comparability and represent the standardized mean, standard errors, and variab les strongly differentiating each cluster according to
the MFA/HCPC analysis. Values diverging significantly from the mean according to the v test (pvalue %0.05) are indicated as – or + .
(legend continued on next page)
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 5
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

The specific case of R4D offers further learning related to sec-
toral LA types. The cross-continental dataset has 38 LAs with
CGIAR engagement, of which 29 clustered in sectoral types
(17 production LAs and 12 preservation LAs). Similarly, of the
additional 23 LAs identified only through the CGIAR networks,
18 of them clustered in sectoral types (12 production and 6 pres-
ervation; Table S4C). Multiple reasons may explain the coherent
dominance of sectoral LAs in R4D organizations: 76% LAs in
2012–2015 and 78% LAs in 2021. For example, efforts to
move solely from work centered on technology adoption to con-
ducting systems-oriented research in and for development are
taking place but are relatively recent.
39
Besides, researchers
have flagged some fundamental discrepancies between the
support provided by international stakeholders and what LAs
need, including discrepancies in timelines (e.g., short-term
versus long-term support), investment priorities that encompass
global versus local needs, top-down versus bottom-up gover-
nance, narrow versus holistic monitoring systems, and funding
targeted to increase technology adoption versus strengthening
social empowerment, among others.
39–43
Therefore, the inertia
of technology-centered culture and projects, mixed with short-
term funding and engagement, constrains LAs maturation (e.g.,
Douthwaite et al.
39
). Addressing and understanding these dis-
crepancies is needed to support, rather than hinder, the locally
driven long-term and process-oriented efforts central to inte-
grated LAs.
(A–G) Questionnaire themes.
(H) Actions post classified into intervention strategies commonly used in conservation and development are presented but excluded from the MFA/HCPC
analysis as well as total stakeholders involved. See Table S3 for the categorical variables strongly differentiating each type (cluster). Intervention strategies
are as follows: AS aims to improve agriculture sustainability, BC aims to foster behavioral change; and SJ aims to contribute to social justice by fostering
empowerment, autonomy, health, and equity.
Figure 3. Proportion of stakeholders by level involved in each type and at different stages (design, implementation, and design and imple-
mentation)
* and indicate variables strongly and significantly associated with each typ e or cluster of LAs according to the c2 test (pvalue < 0.05), and variables can be either
rare (<20% LAs in each type) or common (20–80% LAs in each type), respective ly. See Tables S2 and S3 and the data
38
for more details. NGO, non-governmental
organizations.
ll
OPEN ACCESS Article
6One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

The level of outcomes reported across domains, a surrogate
measure of LAs’ performance, is strongly influenced by the
type of LAs (Figure 4). Hence, LAs with a sectoral lens are
more likely to fall short in achieving multiple outcomes, since
these reported significantly fewer outcomes, on average, for
each domain (Figure 4). In contrast, integrated LAs report
achieving significantly more diversified outcomes across sus-
tainable agriculture, conservation, coordination, and livelihood
domains, suggesting that these are multifunctional landscapes
(Figure 4). Therefore, our data reinforces previous claims that in-
tegrated LAs make the most progress toward achieving multiple
sustainable development goals.
7,21
Our data did not allow us to
assess whether sectoral LAs naturally transition toward inte-
grated LAs when time and resources are sufficient. Therefore,
new tools like the Mixing Board Tool can be useful to assess
LA trajectories by characterizing LAs according to selectable,
scalable, and measurable attributes.
44
Similarities and differences across LA types
We found evidence of shared tendencies across LA types.
Training was the most widely used implementation mecha-
nism across LAs, regardless of type (Figure 2H). This is a
coherent finding when looking only at LAs from the LAC
regional assessment.
21
Hence, our empirical data show that
practitioners across regions built capacity through training
potentially as a means to overcome deficits in capacity for
LA leadership, systems thinking, LA implementation, stake-
holder engagement, negotiation, conflict resolution, moni-
toring, and reporting—all central aspects of integrated
LAs.
13,22,45–47
LAs need substantial and continuous invest-
ment in capacity building.
10
Gaining a deeper understanding
of the role of capacity building in tandem with enabling factors
on behavior and perception change across stakeholders in
LAs should not be underestimated.
47
We found that LAs invested heavily in coordination (Figure 2G),
a result also found elsewhere,
21
likely owing to the multiple chal-
lenges faced in operationalizing LAs (Figure 2B).
26
For example,
coordination needs are manifold and include continuously iden-
tifying and engaging with diverse stakeholders who may have
divergent motivations and values.
20,22
Hence, LA leaders must
negotiate competing interests, facilitate consensus on the vision
and priorities, mediate power imbalances, secure sustainable
funding for coordination and implementation, monitor large
areas and processes, strengthen local governance capacity,
and deal with incoherent, incompatible, or unsupportive policies
or structures.
22
Thus, our results obtained from LA practitioners
reinforce the notion that, regardless of the type, LAs require
strong coordination to manage multiple and often unpredictable
or ‘‘wicked’’ challenges and to address complex, transformative,
and dynamic processes.
22,23
Table 1. Top variables explaining total outcome variance in four performance analyses
Performance analyses
Cross-continental
dataset
Production LAs Preservation LAs Integrated LAs
n357 113 118 126
Explained var (%) 37.38 8.59 24.4 12.2
Variables (units) Avg. rank No. metrics Avg. rank No. metrics Avg. rank No. metrics Avg. rank No. metrics
Total intervention
strategies used (%)
15 15 25 15
Longevity 3 5 5 4 1 5 3 5
Intervention strategies
enabling social justice
(SJ) (%)
35 35 45 55
Stakeholders involved
during implementation
(no.)
223935574
Total motivations (no.) 5 5 26 4 7 4 36 3
Intervention strategies
enabling agriculture
sustainability (AS) (%)
55 84 400* 25
AS: agro-ecological
practices (%)
74 55 480* 45
SJ: tenure, equity,
culture (%)
113 102* 9 4 353
Intervention strategies
enabling behavioral
change (BC) (%)
114 192* 7 5 290*
Variable values show the average variable ranking across five importance metrics. For example, the number of intervention strategies commonly used
in conservation and development deployed was consistently ranked as one of the most important variables associated with more outcomes achieved
across analyses. * indicates variables not selected consistently by at least three of the five variables’ importance metrics (see experimental procedures
and the full ranking in the dataset
38
). LA, landscape approach.
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 7
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

The early inclusion and engagement of stakeholders enable
transformative processes by identifying and reconciling
competing values and collectively defining the shared future
vision of the landscape and its inhabitants.
11,42,48
Scholars
have argued that ‘‘sharing concerns and a sense of urgency’’
is a basic component of integrated approaches.
19
Despite the
theoretical importance and need to engage sectors and stake-
holders early on, we find little evidence that early engagement
is widely implemented, particularly in the case of sectoral LAs
(Figures 2E and 2F). This gap suggests that the time, resources,
skills, and trust needed for consensus-building are potentially
lacking across many LAs, common bottlenecks in LAs,
22
notably
marked in sectoral LAs. For example, researchers from CGIAR
and other R4D organizations report struggling to adopt or sustain
multistakeholder platforms beyond project cycles in their at-
tempts to transition from work centered on technology adoption
to systems-oriented research.
39–41
This may be due to short-
term funding, which rarely values the critical initial investment
to engage with landscape actors as central drivers of landscape
change and to sustain well-functioning multistakeholder plat-
forms as key spaces to negotiate landscape futures in an inclu-
sive, representative, and equitable way.
19,40,49,50
LA longevity, inclusivity, multisectoral investments, and
interventions to enable social justice associated with
performance
The idea of ‘‘silver bullet’’ solutions to complex challenges in
development and conservation agendas persists.
51–55
However,
our data suggest that ‘‘how’’ these complex efforts are sup-
ported could be more relevant than the ‘‘what’’ is supported.
For example, LA practitioners use diverse intervention strategies
(Figure 2H). Indeed, the diversity of strategies used across LA
types is highly associated with LAs’ performance (Table 1). Our
data also indicate that not all intervention strategies have the
same impact. Across LA types, the more strategies used to
enable social justice by fostering empowerment, autonomy, eq-
uity, and health, the better LAs’ performance. We also found that
the longer duration of the LA from its inception (LAs’ longevity)
and a higher total number of stakeholders involved during imple-
mentation are consistently associated with LA performance
across types (Table 1). Time and inclusivity most likely contribute
to less tangible to measure but central aspects of well-func-
tioning and place-based landscape governance systems, such
as improved leadership, increased justice, mediated power im-
balances, established trust, and strengthened cooperation.
49,56
We acknowledge that LAs’ performance can impact their
longevity in that successful LAs might be able to obtain more
funding and, therefore, last longer. However, our data limits
testing this feedback loop (see Study limitations and caveats).
LAs reporting more motivations, either because these occur in
more complex contexts or because they were better character-
ized in the LA design, resulted in more outcomes across ana-
lyses (Table 1). Therefore, we confirm that long-term, better
framed, more inclusive efforts and a diversified portfolio of ac-
tions—notably those enabling social justice—are central to
LAs’ performance. Perhaps, unsurprisingly, integrated LAs
demonstrate this best.
Other variables associated with achieving more outcomes
were relevant in certain performance analyses. For example,
larger investments in intervention strategies enabling agriculture
sustainability in general, and specifically through ‘‘agroecologi-
cal practices,’’ were associated with more outcomes in all ana-
lyses except in preservation LAs (Table 1). Social justice inter-
vention strategies enabling, specifically, tenure, equity, and
culture, were associated with better performance in all but pro-
duction LAs (Table 1). Finally, intervention strategies aiming to
support behavioral change were notably relevant in the cross-
continental and preservation LAs. The remaining variables
were important for one type, suggesting that their contribution
to LA outcomes is context or case dependent (Table 1).
Therefore, across LA types, LAs performing better are those
that last longer, are more inclusive, are better framed, and deploy
a diversified portfolio of interventions. Furthermore, investing in
sustainable agriculture, notably through agroecological prac-
tices, is associated with achieving more outcomes. Among the
agricultural activity intervention strategies, agroecological prac-
tices, such as intensification through agroforestry, agrobiodiver-
sity, soil conservation, and home gardens, were the most pur-
sued to across LA types. Conversely, conventional agricultural
practices, such as intensification with mechanization, fertilizers,
Figure 4. Self-reported outcomes used to suggest LA performance by cluster or type and across domains
The boxplot displays performance score distribution (minimum 25%, median 75%, and maximum values 100%; black squares show the mean). Each point
represents an LA initiative or project, and the non-overlapping notch boxes suggest that medians differ among groups at a 95% confidence level. See Table S2 for
a description of the variables.
ll
OPEN ACCESS Article
8One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

and pesticides, were rarely pursued by LAs across types (Fig-
ure 2H; Table S3).
Our analyses of LAs’ performance measure it as the overall or
total of self-reported outcomes resulting from the LAs’ efforts.
The variables analyzed explained 37% of the overall outcome
variability across LAs (Tables 1 and 2). Although this is a rela-
tively high value for a cross-continental dataset that compiles
information from open, complex, and diverse contexts, it
also suggests that many factors beyond those captured by
our questionnaire influence LA performance. For example,
although 380 landscape practitioners confirm an alignment be-
tween our inclusion criteria and their landscape, we found
distinct LA types, highlighting the various perceptions of what
constitutes an LA. Likewise, the regional assessment question-
naire was designed to characterize LAs rather than to capture
broader social and political-economic factors influencing LAs’
performance. Finally, the explained variability also shows that
our questionnaire design for characterizing LAs better repre-
sents preservation types, since the analyzed variables from
our questionnaire explain approximately a quarter (24%) of
their outcome variability (Table 1). By contrast, our question-
naire and the variables analyzed explained only 12% and 8%
of the outcome variability for integrated and production LA
types, respectively, indicating that a better understanding of
the factors influencing the performance of these types is
needed (Table 1).
Which international stakeholders are sitting at the
LA table?
Our data on LA stakeholders reflect international organizations’
wide interest and support of LAs across regions (268 organiza-
tions; Table S5). However, despite multistakeholder engage-
ment being a core principle of LAs, we found limited or low levels
of international stakeholder engagement (Figures 3 and 5). Of the
different organization types, development agencies were the
most commonly engaged across LA types and were represented
across all regions (Figure 5A; Table S5). Conservation organiza-
tions were more commonly involved in preservation and inte-
grated LAs (Figure 5A), mostly in Africa and Asia (Figure S2).
United Nations organizations were also common across types
and regions, whereas R4D organizations were more commonly
involved in production LAs (Figure 5A; Table S4). When looking
at individual organizations, their percentage of engagement in
LAs remains low (Figure 5B). Only the World Wildlife Fund
(WWF) (preservation LAs), Global Environment Facility (GEF)
(preservation and integrated LAs), and the United Nations Devel-
opment Program (UNDP) (in three clusters) were involved in
slightly more than 10% of the LAs (Figure 5B). We found that
Table 2. Summary of the experimental procedures, variables, and tests used to respond to each research question
Research question Method Variables Significance or strength tests
(1) What distinct LA
types exist across
continents and what
are their similarities
or differences?
defining clusters of LAs
through multiple factor
analysis (MFA) and hierarchical
clustering of principal
components (HCPC):
(sample size for the
cross-continental dataset = 357,
main document and extended
dataset = 380 LAs supplemental
information)
assessing cluster accuracy
through random
forest/classification
(sample size = 357 main
document and 380 LAs
supplemental information)
124 variables from the
questionnaire (question
numbers 1–83 and 85–125;
Table S2), excluding outcomes, total
stakeholders not engaged,
intervention strategies, and
partner-related variables
clusters 124 variables
from the questionnaire
(question numbers 1–83
and 85–125; Table S2),
excluding outcomes, total
stakeholders not engaged,
intervention strategies, and
partner-related variables
agglomeration through Ward’s
method; minimizes the within-cluster
variance when identifying the optimal
number of clusters (Figures S1A–S1E).
v test (continuous variables) and
c2 test (categorical variables) to
identify variables strongly and
significantly associated with each
cluster (Table S3; Figures 2 and 3).
classification error matrix showing
predicted vs. actual classification
(Table S1).
(2) What variables are
associated with more
outcomes?
identifying variables explaining
LA total outcomes variance;
total outcome is used as a
surrogate for overall performance;
analysis through random
forest/regression and
randomForestExplainer
(four analyses conducted:
one with the full cross-continental
dataset (357 LAs) and one for
each cluster of LAs: integrated
(126 LAs), preservation (118 LAs),
and production (113 LAs)
total outcomes (variable
number 152, Table S2)103
variables from the questionnaire
(question numbers 1–3, 85–88,
and 157–172; Table S2),
excluding outcomes
per domain, total stakeholders
not engaged, total
actions per domain, and
international partner-related
questions
Consensus among five distinct metrics
of variable importance: (a) number
of trees, (b) minimal depth, (c) times
a root, (d) node purity increase,
(e) mean-squared error increase
(3) What international
organizations are present
or missing across LAs?
frequency of international
organizations’ involvement
(sample size cross-continental
dataset = 357 LAs)
all international organizations
named in partner’s
theme/questionnaire
(variables 153–156; Table S2)
N/A; see the full list of organizations,
classification, and engagement
across clusters in Table S5
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 9
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

Figure 5. Top international groups and individual organizations partnering with different LA types
Shown are the top 10 international organization types (A) and top 20 unique international organizations (B) partnering on the implementation of LAs. The data
source is color differentiated, and CGIAR centers have a dark outer line in (B). AsDB, Asian Development Bank; AWF, African Wildlife Foundation; CARE,
Cooperative for Assistance and Relief Everywhere; CATIE, Tropical Agricultural Research and Higher Education Center; CEPF, Critical Ecosystem Partnership
Fund; CI, Conservation International; CIFOR-ICRAF, Center for International Forestry Research and World Agroforestry, CGIAR; DFID, UK Department for In-
ternational Development; EU, European Union; FAO, Food and Agriculture Organization of the United Nations; FFI, Fauna & Flora International, GEF, Global
Environment Facility; GIZ, Gesellschaft f€
ur Internationale Zusammenarbeit; IFAD, International Fund for Agricultural Development; IUCN, International Union for
Conservation of Nature; IMFN, Ibero-American Model Forest Network; IRRI, International Rice Research Institute, CGIAR; IWMI, International Water Management
Institute, CGIAR; KfN, Kreditanstalt f€
ur Wiederaufbau; No int, no international organizations are involved; not detailed, specific names or acronyms were not listed,
only general categories (e.g., international non-governmental organizations [NGOs]); SDC, Swiss Agency for Development and Cooperation; SNV, Stichting
Nederlandse Vrijwilligers; TNC, The Nature Conservancy; UNDP, United Nations Development Program; UNEP, United Nations Environment Program; UNESCO,
United Nations Educational, Scientific and Cultural Organization; USAID, United States Agency for International Development; USFWS, United States Fish and
Wildlife Service; WB, World Bank; WCS, Wildlife Conservation Society; WWF, World Wildlife Fund. See Table S5 for the full list of international organizations.
ll
OPEN ACCESS Article
10 One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

33 international organizations were involved across the three LA
types, confirming that the same organization can operationalize
LAs differently (Figure 5; Table S5).
Despite the potential of LAs to contribute to human health and
nutrition and the wide range of partners listed, we did not find in-
ternational health organizations listed as partners across LA
types (Table S5). Aspects contributing to human health, like
food security—one intervention strategy enabling social jus-
tice—represented one of the lowest investments across LA
types and, most notably, within sectoral LAs (Figure 2H;
Table S3). Our findings show, for instance, that partnerships
with health organizations and professionals are missing in LAs
and represent an untapped potential for linking LAs to the plan-
etary health agenda and for incorporating nutrition-sensitive (ac-
cess to healthy foods), health-sensitive (pollution reduction), or
zoonotic disease risk (OneHealth) interventions in LAs.
57–61
Besides human health, LAs can also greatly contribute to the
overall well-being of landscape inhabitants, particularly of the
farmers, herders, fisherfolk, and forest communities who are
landscape stewards (e.g., Carmenta et al.61 and Hanspach
et al.
62,63
). Implementing holistic, systems-driven, and locally
driven LAs has a higher positive impact on local constituents’
well-being compared to single-sector style LAs, which tend to
report fewer positive and more negative outcomes.
62
Therefore,
given that farmers lead suicide and depression rates in countries
where statistics for this issue exist, and where conventional agri-
cultural models dominate,
64–69
assessing the comparative
advantage of integrated LAs on farmers’ well-being beyond sub-
sistence or income-related aspects is both a critical necessity
and an opportunity for tackling this often overlooked yet globally
prevalent phenomenon.
Another stakeholder group in our dataset with low or limited
engagement is the private sector, despite its considerable influ-
ence on landscape dynamics, social equity outcomes, biodiver-
sity, and revenues.
12,70
For example, foreign agribusiness;
extractive industries such as mining, oil, or gas; or forestry indus-
tries were largely absent across LA types and engagement
stages (>80% LAs; Figure 3). Both international and in-country
agribusinesses rarely participated in LAs, specifically in sector-
focused production and preservation LAs. Only integrated LAs
worked more frequently with this sector (30% of integrated
LAs) (Figure 3). The low number of international private-sector or-
ganizations listed as LA partners confirms the private sector’s
poor direct engagement. For example, less than 10% of the
LAs in each type reported engaging with companies or corpora-
tions. However, their indirect engagement through their chari-
table foundations or funds was slightly higher (<13%) (Figure 5A;
Table S5). Some other LA reviews indicate a common direct
involvement of international companies and corporations.
35
More recently, business coalition partners (e.g., One Planet
Business for Biodiversity [OP2B] or Forest Positive) are adopting
LAs,
71,72
and efforts to connect landscape practitioners with in-
vestors have recently begun (e.g., Landscape Finance Lab). Still,
multiple sources found null or low involvement of the private
sector in LAs, which aligns with our findings.
11,22,36,45,73
Regardless of current private sector engagement in LAs, this
stakeholder could play a key role and contribute to LAs’ success
by connecting them with markets and finance opportunities for
sustained funding and innovation.
9,19,20,74
Private-sector parties
could benefit from engagement in LAs by enhancing the stability
of key geographies for their value chains, reducing reputational
and operational risks, and improving compliance with standards,
consumers’ demands, and global agreements (e.g., zero
deforestation).
20,35,36,74
While engaging the private sector in LAs can present benefits,
engagement can also reinforce conflicts of interest and perpet-
uate entrenched positions in contradiction with local values or
views. As such, private-sector engagement can increase or rein-
force power imbalances and jeopardize communities’ sover-
eignty, autonomy, and overall social and environmental justice
in the landscapes in which they engage.
20,75
LA practitioners
have consistently highlighted challenges in reversing power im-
balances and addressing external pressures or threats to their
landscapes.
10,17,20,47
Past experiences show that the private
sector’s motivations beyond sectoral or short-term economic
objectives are often insufficient to engage in LAs.
36,74
Shifting
private sector actors toward long-term engagement and equi-
table access/benefit-sharing goals is challenging.
45
Although
meaningful partnerships and collaborations can exist when the
private sector’s interests align and parties commit to ongoing
engagement,
74,76
a fruitful and just engagement in LAs will also
require a stronger demand, accountability, and legal compliance
tailored to different types of private sectors.
36,74
Conclusion
LAs can be and ougth to be more central in achieving multiple
global agendas, including the Sustainable Development Goals
and Kunming-Montreal Global Biodiversity Framework.
77
To do
so, the increasing support toward these complex and dynamic
processes should be tailored to enhance how the most inte-
grated LAs are being operationalized, and to support those
more sectoral approaches to continue developing their cross-
sectoral connections, local relevance, stakeholder engagement,
and landscape-level outcomes. Despite confirming the exis-
tence of distinct LA types, we also found shared challenges hin-
dering their performance and persistence. For example, across
types, the investments in and role of capacity building and coor-
dination in LAs should not be underestimated. Likewise, early
stakeholder engagement is theoretically acknowledged but
lacking in practice; hence, understanding how to pass from the-
ory to practice is critical for LAs’ performance. Our results point
out time and inclusion as key aspects of better-performing LAs.
Hence, more attention is needed to measure intangible and diffi-
cult-to-measure core components of LAs enabled through time
and inclusion, such as trust, leadership, justice, power dy-
namics, and cooperation. Finally, LAs seem to benefit more
from diversified portfolios of intervention strategies, especially
those that enable social justice. Regarding gaps from the LAs
in the cross-continental dataset, we confirm that certain key
stakeholders can engage more in LAs (e.g., private sector and
R4D). However, how to ensure that their short-term engagement
and sectoral interests propel rather than undermine locally driven
stakeholder efforts pursuing justice and sustainability in multi-
functional landscapes remains an open question and requires
collectively learning from past errors.
78–82
Overall, evidence
shows the importance and need of putting agency and jus-
tice
83–87
at the center of international support, so it helps LAs
thrive. Finally, LAs’ impact and contributions to human health
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 11
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

and overall well-being are missing or poorly documented, open-
ing new and promising lines of research and interventions. The
growing interest and support for LAs is promising. We hope
our findings will enhance current efforts and research aiming to
support well-functioning and lasting LAs.
EXPERIMENTAL PROCEDURES
Characterization and selection of LAs in the cross-continental
dataset
The cross-continental dataset includes 357 LAs identified through regional as-
sessments focused on and implemented in LAC, Africa, and Asia between
2012 and 2015.
16–18
LAs implemented in Europe were also characterized
25
;
however, the European questionnaire was different, placing a larger emphasis
on cultural heritage and values, which hindered integrating data collected for
other regions. The authors of the LAC, Africa, and Asia regional assessments
pre-identified potential LAs through practitioner networks and through an on-
line search (see original articles for more details). The inclusion criteria for pre-
identifying LAs were ‘‘projects, programs, platforms, initiatives or sets of activ-
ities that (1) explicitly seek to improve food production, biodiversity or
ecosystem conservation and rural livelihoods; (2) work at a landscape scale
and include deliberate planning, policy, management, or support activities at
this scale; (3) involve inter-sectoral coordination or alignment of activities, pol-
icies, or investments at the level of ministries, local government entities, farmer
and community organizations, NGOs, donors, and/or the private sector; and
(4) are highly participatory, supporting adaptive, collaborative management
within a social learning framework.’’
After the LAs were pre-identified, local landscape leaders or managers were
invited to complete an online questionnaire that started by listing the inclusion
criteria mentioned above. The respondent was asked to respond to the ques-
tionnaire if their project, program, or activities aligned with the criteria listed.
The questionnaire was designed to characterize LAs in terms of longevity
(years since inception), land uses where initiatives take place, landscape man-
agement options used, motivations that drove their creation, stakeholders and
sectors engaged and at what stage (i.e., design, intervention, or both), actions
carried out by the initiative, and attributable outcomes achieved by the LA.
Characterization and selection of LAs in the extended dataset
In 2021, we used the same method, questionnaire, and LA inclusion criteria as
for the regional assessments to capture CGIAR’s and its partners’ wider
engagement in LAs. We identified LAs with CGIAR engagement across the
13 CGIAR centers in two steps. First, we identified researchers working within
LAs before identifying and characterizing CGIAR LAs. We contacted 986 of the
3,169 CGIAR researchers listed in the institutional directory in 2021. Contacted
researchers were selected as follows: 406 were tagged in the CGIAR expert
finder with the keyword ‘‘landscape,’’ and 580 were randomly selected across
the 13 CGIAR centers to guarantee contacting >30% of the research staff in
each center. We contacted researchers to (1) learn about their work on LAs,
(2) identify landscapes where CGIAR researchers are involved, and (3) refer
us to more CGIAR researchers working on LAs. Through this effort, we map-
ped 55 CGIAR researchers working in or with LAs. We asked them to describe
their LAs through an online questionnaire intentionally similar to the regional
assessments deployed in the ArcGIS123 platform (see Table S2 for the themes
and questions in the questionnaire). CGIAR researchers described an addi-
tional set of 23 landscapes in 2021. Overall, the extended dataset (cross-
continental + 23 CGIAR) includes 380 LAs being implemented in 69 countries
across three continents (Figure 1;Table S2).
38
Results obtained using the
extended dataset are presented in Tables S3 and S4.
The questionnaire was structured in three levels. First, the structure con-
cerned themes such as motivations, actions/investments, stakeholder groups
involved, sectors involved, land uses, management types created to support
LAs, and others. Second, some of the questionnaire themes included ques-
tions linked to different domains. For example, motivations that led to the cre-
ation of the LAs can be linked to five domains: sustainable agriculture, conser-
vation, livelihoods, climate change, and coordination. Third, individual
responses providing information for each question are also referred to here
as variables and are the most granular level of analysis. Using the same inclu-
sion criteria and questionnaire for characterizing LAs across continents allows
cross-continental comparability. However, it required some minor alignment
before further analyses. This process required removing certain variables
asked inconsistently across sites. One example of this inconsistency is the var-
iable ‘‘rocky terrain, desert,’’ which was an option for characterizing the land-
scape’s major land uses only collected in Asia but not in Afria or in LAC. These
rare inconsistencies represented 12 questions or 9% of the questionnaire. The
core questions, equivalent to 122 non-calculated variables, remained the
same across data sources (Table S2).
Closed-ended individual responses were converted to binary or nominal
variables, with their frequencies used to create theme- or domain-level indices
(Table S2). Index values were weighted by the number of questions in each
domain or theme and scaled to 0–100 points to facilitate comparison
(Table S2). For example, total outcome is measured as the number of out-
comes reported across domains weighted equally (Table S2). Hence, higher
outcome values imply more outcomes reported across domains and, there-
fore, higher landscape performance. Overall, the cross-continental dataset in-
cludes 104 LAs in LAC, 87 LAs in Africa, and 166 LAs in Asia. Of the 357 LAs
listed, only 38 (11%) mentioned at least one of the 13 CGIAR centers in the re-
spondent’s affiliation or as a partner. This seems to be a relatively low number,
given that each assessment included at least one author from the Center for
International Forestry Research and World Agroforestry (CIFOR-ICRAF) or
the Alliance of Bioversity and International Center for Tropical Agriculture
CGIAR centers and used CGIAR research networks to identify the LAs.
In both the cross-continental and extended datasets, LAs are referred to by
different names, such as watershed/basin management, biological corridor,
model forest, protected, conservation, buffer area, and others. This diversity
in names for LAs aligns with previous findings.
15
We used information
describing LAs, not the name of the project or landscape. For example, for
the clustering, we did not differentiate LAs among those called biological cor-
ridors or basins (Tables 2 and S2).
Cluster analysis to identify distinct LA types
In 2020, Carmenta et al.
21
conducted a cluster analysis on the LAs from LAC
and found four distinct clusters. Two clusters tend to use a sectoral lens either
focused on agriculture or conservation, and two integrated clusters either
focused on ‘‘participation and legislation’’ or ‘‘certification and institutions.’’
21
Additionally, Carmenta et al.
21
proposed grouping LA actions into intervention
strategies commonly used in conservation and development. Then, the au-
thors assessed the incidence of these intervention strategies across contexts
and LA performance. Intervention strategies commonly used in conservation
and development are promising strategies discussed in the scientific lite rature.
For example, agricultural activity intervention strategy refers to the types of
agriculture-related activities such as agroecological practices or intensifica-
tion (adapted from Rasmussen et al.
59
); Technical coupling intervention strat-
egy refers to the strategies for fostering behavioral change, such coordination,
legislation, and certification (adapted from Phalan et al.
29
). A people-based
intervention strategy aims to foster empowerment, autonomy, health, and eq-
uity (adapted from Phelps et al.,
30
Duchelle et al.,
32
and Wilebore et al.
31
).
We followed and adjusted the methodology proposed by Carmenta et al.
21
to classify LA actions into intervention strategies and identify LA types clus-
tering similar initiatives. Our methods are an exploratory analysis (see Table 2
for a summary). LA type identification through cluster analysis includes infor-
mation from the questionnaire related to the initiative’s duration from inception
(longevity), major land uses, management, motivations driving inception, sec-
tors and stakeholders engaged, and actions led by and attributable to the land-
scape efforts (Tables 2 and S2). We excluded outcomes-related variables from
the cluster analysis. Also, we excluded highly correlated variables (R0.8), such
as the post-classified actions into intervention strategies to avoid redundancy
(Figure S3). Similarly, the total number of stakeholders engaged at the national
level was also highly correlated with the total number of stakeholders engaged
overall; hence, we removed the latter variable (Table 2). Finally, we only
included the variables shared across regions and data collection efforts,
including core and calculated variables for a total of 124 variables (Tables 2
and S2).
The cluster analysis involved two steps. The first step was to remove statis-
tical noise from potential collinearity issues in the dataset through an MFA.
88,89
An MFA handles categorical and continuous variables and balances the
ll
OPEN ACCESS Article
12 One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

influence of questionnaire themes with different numbers of variables.
88,89
The
second step was to conduct an HCPC analysis on the MFA results to measure
the multidimensional distance among individual LAs through Ward’s method,
which measures the error sum of squares after merging two clusters in a bot-
tom-up approach.
88
HCPC describes the significant effect of each variable in
each cluster through the v test for continuous variables and c
2
test for categor-
ical variables.
88
Therefore, continuous variables with a v test value above or
below 1.96 (equivalent to p%0.05)
88
are considered significant in this anal-
ysis. Similarly, categorical variables significantly associated with each cluster
are those with a c
2
test and p degrees of freedom (variable levels) resulting in
p%0.05.
88
Strongly and significantly associated categorical variables can be
dominant (i.e., present in R80 of the LAs in the cluster) or rare (i.e., present
in <20 of the LAs in the cluster). We used strongly and significantly associated
continuous and categorical variables to describe each cluster (Figures 2 and 3;
Table S3). We labeled each cluster based on its most salient characteristics.
Finally, we tested the accuracy of the cluster groupings through an indepen-
dent method—the random forest classification
90
method (clusters 124 vari-
ables; Table 2). This test estimates a cluster classification error by comparing
defined clusters with MFA and HCPC against predicted clusters with random
forest. Random forest classification uses a bootstrapping process to split the
data randomly and repeatedly into training (30% of the data) and testing
samples (70% of the data) to calculate the misclassification error between
both samples.
90
This process is repeated 10,000 times (trees grown) to calcu-
late the error rate or number of LAs classified in the same cluster or other clus-
ters (Table S1).
90
Identifying variables and intervention strategies contributing to LA
performance
We used the random forest regression method (overall outcome variables) to
assess non-linear relations among variables associated with LAs’overall perfor-
mance (Table2). This method accounts for the largenumber of explanatory vari-
ables in relation to the sample size.
91,92
We ran this analysis twice using the 124
variables used in the cluster analysis plus the 11 variables from the post-classi-
fied investment strategies (162 variables in total) and replacing actions with the
post-classified investment strategies to explore the incidence of thesevariables
related to commonlyused strategies in developmentand conservation. Both an-
alyses rank variables similarly (Table 1). The main difference is that, when
running actions and investment strategies, both highly correlated variables.
For example, total intervention strategies (%) and total number of actions in-
vested appear ranked as the first and second most important. Therefore, we
ran the analysis only with investment strategies to avoid redundant variables
in the ranking. We used randomForestExplainer and considered five indepen-
dent importance metrics measuring the contribution to each variable to explain
the total self-reported outcome variance.
93
The importance metrics evaluate
variable importance differently, such as the number of nodes wherethe variable
was selected to splitthe node, mean minimal depth where the variableis used to
split the tree, times a variable is used to split the root node, node purity increase
(reduction in the sum of the squared error after splitting a node with a variable),
and mean-squarederror increase after a variable is permuted.
93
All metrics cap-
ture different variable contributions or importance in explaining outcome vari-
ance.
93
We shortlisted variables based on two criteria: first, variables with the
lowest averageranking across importancemetrics where the lowest valuesindi-
cate higher importance and second, variables shortlisted as the top 10 most
important in at least three of the five independent importance metrics.
Identifying international organizations present or missing
across LAs
We screened for partner organizations in three sections of the questionnaire:
respondents’ institution, partners inside the landscape, and partners outside
of the landscape (Tables 2 and S2). We searched and verified each organiza-
tion’s name and recorded international partners. When respondents provided
only an acronym, we searched the internet for the acronym in tandem with the
name of the landscape and country. We assigned broad categories for orga-
nizations based on the description given on the organization’s official web-
page. Because we could not find an official page for five organizations, we
used information from other sources such as Wikipedia or LinkedIn
(Table S5). In certain cases, where we could not find the organization or cate-
gory, we tagged projects as ‘‘unclear’’ (Table S5). We used descriptive statis-
tics to describe the frequency of individual or groups of international organiza-
tions engaged across LAs types. All the analyses were conducted in R and
included multiple packages.
94
Study limitations and caveats
Our research contains the same limitations highlighted in the regional assess-
ments and previous attempts to identify LA types through cluster analysis in
LAC.
3,16–18
First, the search for projects employing an LA was based on an
internet search and practitioners’ international networks (e.g., CGIAR centers ),
which may exclude grassroots-led initiatives without an internet presence or
access. The questionnaire was made available to respondents online or via
email. Second, responses were given by the LA project leader or manager
as only one person, self-reporting. However, it was considered that this person
may be biased toward reporting more, in an attempt to convey overperform-
ance, or may indeed report fewer investments and outcomes due to incom-
plete knowledge of the initiative if the practitioner was new to the position.
Third, investments and outcomes are self-reported, present/absent data
types. The data does not include the level of effort, time, resources, or
outcome reached and/or other information validating outcomes from a third
party. Fourth, the questionnaire was designed to take a snapshot of the LA
projects rather than to gather a detailed description; thus, it may exclude
certain activities, such as access to credit; certain outcomes, such as non-ma-
terial subjective and relational ones; or impacts, such as, for instance,
discerning how these outcomes affected the different stakeholders. Fifth,
and last, the questionnaire is a one-time snapshot, which does not include
key information on trajectories and historical legacies in each landscape.
Despite these caveats, this is the largest dataset across continents systemat-
ically describing LAs. It offers a rich starting point to better understand and
analyze LAs as described and operationalized by practitioners.
RESOURCE AVAILABILITY
Lead contact
Requests for further information and resources should be directed to and will
be fulfilled by the lead contact, Natalia Estrada-Carmona (n.e.carmona@
cgiar.org).
Materials availability
This study did not generate new unique materials.
Data and code availability
The extended dataset with LAs can be downloaded from Dataverse.
38
Scripts
can be made available upon request.
ACKNOWLEDGMENTS
We are grateful to all survey respondents for taking the time to characterize
their landscape through a lengthy questionnaire. We thank Claudia Ringler,
Leigh Ann Winowiecki, Pay Drechsel, and Yodit Kebede for insights during
the inception of the project. We thank EcoAgriculture Partners and the Land-
scapes for People, Food, and Nature Initiative, who funded the regional as-
sessments and permitted the combination of the LA dataset. This research
was conducted as part of the project ‘‘Flipping the Coin: Managing Agricultural
Landscapes to Tackle Multiple Global Challenges,’’ funded by the CGIAR Wa-
ter, Land, and Ecosystems (WLE) Research Program. This research received
support from the One CGIAR NEXUS Gains Initiative – Realizing Multiple Ben-
efits Across Water, Energy, Food, and Ecosystems (Forests, Biodiversity). This
research also received support from the Agroecological Transitions for Build-
ing Resilient and Inclusive Agricultural and Food Systems (TRANSITIONS) pro-
gram, which is funded by the European Union through its DeSIRA initiative and
managed by the International Fund for Agricultural Development, specifically
the project ‘‘‘Private Sector Incentives and Investments (PSii) for climate
change, resilience and environmental sustainability,’’ under European Com-
mission grant agreement 2000003771. J.R. acknowledges funding support
from the International Climate Initiative (IKI) of the German Federal Ministry
for the Environment, Nature Conservation, Buildings, and Nuclear Safety
(BMUB) under grant 18_IV_084. We thank Silvia Araujo de Lima for help with
identifying and documenting international organizations. We also thank Glenn
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 13
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

Hyman and Olga Spellman (Alliance of Bioversity International and CIAT Sci-
ence Writing Service) for editorial review and copyediting of this manuscript.
AUTHOR CONTRIBUTIONS
N.E.-C. developed the methodology, conducted the formal analysis and data
curation, and wrote the original draft. R.C. and J.R. contributed to the develop-
ment of the methodology. J.M., A.K.H., N.E.-C., F.D., and C.Z. contributed to
the original study design and data collection in LAC, Africa, and Asia. All au-
thors contributed to the conceptualization of the research and reviewed and
edited the original draft.
DECLARATION OF INTERESTS
The authors declare no competing interests.
SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.
oneear.2024.08.014.
Received: March 26, 2023
Revised: April 18, 2024
Accepted: August 27, 2024
Published: September 20, 2024
REFERENCES
1. Perfecto, I., and Vandermeer, J. (2008). Biodiversity conservation in trop-
ical agroecosystems: A new conservation paradigm. Ann. N. Y. Acad. Sci.
1134, 173–200. https://doi.org/10.1196/annals.1439.011.
2. Van Etten, J. (2022). Revisiting the adequacy of the economic policy narra-
tive underpinning the Green Revolution. Agric Hum Values 39, 1357–1372.
https://doi.org/10.1007/s10460-022-10325-2.
3. Carmenta, R., Barlow, J., Bastos Lima, M.G., Berenguer, E., Choiruzzad,
S., Estrada-Carmona, N., Franc¸ a, F., Kallis, G., Killick, E., Lees, A., et al.
(2023). Connected Conservation: Rethinking conservation for a tele-
coupled world. Biol. Conserv. 282, 110047. https://doi.org/10.1016/j.bio-
con.2023.110047.
4. Obura, D.O., Declerck, F., Verburg, P.H., Gupta, J., Abrams, J.F., Bai, X.,
Bunn, S., Ebi, K.L., Gifford, L., Gordon, C., et al. (2023). Achieving a nature-
and people-positive future. One Earth 6, 105–117. https://doi.org/10.
1016/j.oneear.2022.11.013.
5. Laurance, W.F., Sayer, J., and Cassman, K.G. (2014). Agricultural expan-
sion and its impacts on tropical nature. Trends Ecol. Evol. 29, 107–116.
https://doi.org/10.1016/j.tree.2013.12.001.
6. Sayer, J.A., Margules, C., Boedhihartono, A.K., Sunderland, T., Langston,
J.D., Reed, J., Riggs, R., Buck, L.E., Campbell, B.M., Kusters, K., et al.
(2017). Measuring the effectiveness of landscape approaches to conser-
vation and development. Sustain. Sci. 12, 465–476. https://doi.org/10.
1007/s11625-016-0415-z.
7. Freeman, O.E., Duguma, L.A., and Minang, P.A. (2015). Operationalizing
the integrated landscape approach in practice. Ecol. Soc. 20, art24.
https://doi.org/10.5751/ES-07175-200124.
8. B€
urgi, M., Ali, P., Chowdhury, A., Heinimann, A., Hett, C., Kienast, F.,
Mondal, M.K., Upreti, B.R., and Verburg, P.H. (2017). Integrated land-
scape approach: Closing the gap between theory and application.
Sustainability 9, 1371. https://doi.org/10.3390/su9081371.
9. Deneir, L., Scherr, S., Chatterton, P., Hovani, L., and Stam, N. (2015). The
Little Sustinable Landscapes Book (Global Canopy Programm (Oxford).
10. Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E.,
Venter, M., Boedhihartono, A.K., Day, M., Garcia, C., et al. (2013). Ten
principles for a landscape approach to reconciling agriculture, conserva-
tion, and other competing land uses. Proc. Natl. Acad. Sci. USA 110,
8349–8356. https://doi.org/10.1073/pnas.1210595110.
11. Reed, J., Chervier, C., Rumi, J., Davison, B., Kaala, G., Teddy, B.M.,
Connor, A.O., Siangulube, F., Yanou, M., Sunderland, T., et al. (2022).
Co - producing theory of change to operationalize integrated landscape
approaches. Sustain. Sci. https://doi.org/10.1007/s11625-022-01190-3.
12. Torralba, M., Nishi, M., Cebria
´n, M.A., Quintas, C., Marı
´a, S., and Martı
´n,
G. (2023). Disentangling the practice of landscape approaches : a Q -
method analysis on experiences in socio - ecological production land-
scapes and seascapes. Sustain. Sci. 18, 1893–1906. https://doi.org/10.
1007/s11625-023-01307-2.
13. Reed, J., Van Vianen, J., Deakin, E.L., Barlow, J., and Sunderland, T.
(2016). Integrated landscape approaches to managing social and environ-
mental issues in the tropics: learning from the past to guide the future.
Global Change Biol. 22, 2540–2554. https://doi.org/10.1111/gcb.13284.
14. Erbaugh, J., and Agrawal, A. (2017). Clarifying the landscape approach: A
Letter to the Editor on ‘‘Integrated landscape approaches to managing so-
cial and environmental issues in the tropics.’’. Global Change Biol. 23,
4453–4454. https://doi.org/10.1111/gcb.13788.
15. Scherr, S.J., Shames, S., and Friedman, R. (2023). Defining integrated
landscape management for policy makers.
16. Estrada-Carmona, N., Hart, A.K., DeClerck, F.A., Harvey, C.A., and Milder,
J.C. (2014). Integrated landscape management for agriculture, rural liveli-
hoods, and ecosystem conservation: An assessment of experience from
Latin America and the Caribbean. Landsc. Urban Plann. 129, 1–11.
17. Zanzanaini, C., Trần, B.T., Singh, C., Hart, A., Milder, J., and DeCle rck, F.
(2017). Integrated landscape initiatives for agriculture , livelihoods and
ecosystem conservation : An assessment of experiences from South
and Southeast Asia. Landsc. Urban Plann. 165, 11–21. https://doi.org/
10.1016/j.landurbplan.2017.03.010.
18. Milder, J.C., Hart, A.K., Dobie, P., Minai, J., and Zaleski, C. (2014).
Integrated landscape initiatives for african agriculture, development, and
conservation: a region-wide assessment. World Dev. 54, 68–80. https://
doi.org/10.1016/j.worlddev.2013.07.006.
19. Ros-Tonen, M.A.F., Reed, J., and Sunderland, T. (2018). From Synergy to
Complexity: The Trend Toward Integrated Value Chain and Landscape
Governance. Environ. Manag. 62, 1–14. https://doi.org/10.1007/s00267-
018-1055-0.
20. Sayer, J., Margules, C., Boedhihartono, A.K., Dale, A., Sunderland, T.,
Supriatna, J., and Saryanthi, R. (2014). Landscape approaches; what
are the pre-conditions for success? Sustain. Sci. 10, 345–355. https://
doi.org/10.1007/s11625-014-0281-5.
21. Carmenta, R., Coomes, D.A., DeClerck, F.A., Hart, A.K., Harvey, C.A.,
Milder, J., Reed, J., Vira, B., and Estrada-Carmona, N. (2020).
Characterizing and Evaluating Integrated Landscape Initiatives. One
Earth 2, 174–187. https://doi.org/10.1016/j.oneear.2020.01.009.
22. Vermunt, D.A., Verweij, P.A., and Verburg, R.W. (2020). What hampers im-
plementation of integrated landscape approaches in rural landscapes?
Curr. Landscape Ecol. Rep. 5, 99–115. https://doi.org/10.1007/s40823-
020-00057-6.
23. Pedroza-arceo, N.M., Weber, N., and Ortega-Argueta, A. (2022). A knowl-
edge review on integrated landscape approaches. Forest@ 13, 312–324.
24. Meli, P., Schweizer, D., Winowiecki, L.A., Chomba, S., Aynekulu, E., and
Guariguata, M.R. (2022). Mapping the information landscape of the united
nations decade on ecosystem restoration strategy. Restor. Ecol. 31, 1–14.
https://doi.org/10.1111/rec.13810.
25. Garcı
´a-Martı
´n, M., Bieling, C., Hart, A., and Plieninger, T. (2016).
Integrated landscape initiatives in Europe: Multi-sector collaboration in
multi-functional landscapes. Land Use Pol. 58, 43–53. https://doi.org/
10.1016/j.landusepol.2016.07.001.
26. Reed, J., Ros-Tonen, M., and Sunderland, T.C.H. (2020). Operationalizing
integrated landscape approaches in the tropics. (CIFOR), 1–233. https://
doi.org/10.17528/cifor/007800.
27. Rasmussen, L.V., Fagan, M.E., Ickowitz, A., Wood, S.L., Kennedy, G.,
Powell, B., Baudron, F., Gergel, S., Jung, S., Smithwick, E.A., et al.
(2020). Forest pattern, not just amount, influences dietary quality in five
ll
OPEN ACCESS Article
14 One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

African countries. Global Food Secur. 25, 100331. https://doi.org/10.
1016/j.gfs.2019.100331.
28. Tamburini, G., Bommarco, R., Wanger, T.C., Kremen, C., van der Heijden,
M.G.A., Liebman, M., and Hallin, S. (2020). Agricultural diversification pro-
motes multiple ecosystem services without compromising yield. Sci. Adv.
6, eaba1715. https://doi.org/10.1126/sciadv.aba1715.
29. Phalan, B., Green, R.E., Dicks, L.V., Dotta, G., Feniuk, C., Lamb, A.,
Strassburg, B.B.N., Williams, D.R., zu Ermgassen, E.K.H.J., and
Balmford, A. (2016). How can higher-yield farming help to spare nature?
Science 351, 450–451. https://doi.org/10.1126/science.aad0055.
30. Phelps, J., Carrasco, L.R., Webb, E.L., Koh, L.P., and Pascual, U. (2013).
Agricultural intensification escalates future conservation costs. Proc.
Natl. Acad. Sci. USA 110, 7601–7606. https://doi.org/10.1073/pnas.
1220070110.
31. Wilebore, B., Voors, M., Bulte, E.H., Coomes, D., and Kontoleon, A. (2019).
Unconditional transfers and tropical forest conservation: Evidence from a
randomized control trial in sierra leone. Am. J. Agric. Econ. 101, 894–918.
https://doi.org/10.1093/ajae/aay105.
32. Duchelle, A.E., Simonet, G., Sunderlin, W.D., and Wunder, S. (2018). What
is REDD+ achieving on the ground? Current. Curr. Opin. Environ. Sustain.
32, 134–140. https://doi.org/10.1016/j.cosust.2018.07.001.
33. Van Oosten, C., Runhaar, H., and Arts, B. (2021). Capable to govern land-
scape restoration? Exploring landscape governance capabilities, based
on literature and stakeholder perceptions. Land Use Pol. 104, 104020.
https://doi.org/10.1016/j.landusepol.2019.05.039.
34. Reed, J., Kusters, K., Barlow, J., Balinga, M., Borah, J.R., Carmenta, R.,
Chervier, C., Djoudi, H., Gumbo, D., Laumonier, Y., et al. (2021). Re-inte-
grating ecology into integrated landscape approaches. Landsc. Ecol. 36,
2395–2407. https://doi.org/10.1007/s10980-021-01268-w.
35. Arts, B., Buizer, M., Horlings, L., Ingram, V., Oosten, C.V., and Opdam, P.
(2017). Landscape approaches : a state-of-the-art review. Annu. Rev.
Environ. Resour. 42, 439–463.
36. Upla, P., Reed, J., Moombe, K.B., Kazule, B.J., Mulenga, B.P., Ros-tonen,
M., and Sunderland, T. (2022). Assessing the potential for private sector
engagement in integrated landscape approaches : Insights from value-
chain analyses in southern zambia. Land 11, 1549. https://doi.org/10.
3390/land11091549.
37. CGIAR (2021). CGIAR 2030 Reasearch and innovation strategy:
Transforming food, land, and water systems in a climate crisis.
38. Estrada-Carmona, N., Hart, A.K., and Zanzanaini, C. (2024). Data for:
Reconciling conservation and development requires enhanced integration
and broader aims: A cross-continental assessment of landscape ap-
proaches. Harvard Dataverse). https://doi.org/10.7910/DVN/M6ARB7.
39. Douthwaite, B., Apgar, J.M., Schwarz, A.-M., Attwood, S., Senaratna
Sellamuttu, S., and Clayton, T. (2017). A new professionalism for agricul-
tural research for development. Int. J. Agric. Sustain. 15, 238–252.
https://doi.org/10.1080/14735903.2017.1314754.
40. Schut, M., Cadilhon, J.J., Misiko, M., and Dror, I. (2018). Do mature inno-
vation platforms make a difference in agricultural research for develop-
ment? a meta-analysis of case studies. Exp. Agric. 54, 96–119. https://
doi.org/10.1017/S0014479716000752.
41. Schut, M., Kamanda, J., Gramzow, A., Dubois, T., Stoian, D., Andersson ,
J.A., Dror, I., Sartas, M., Mur, R., Kassam, S., et al. (2019). Innovation plat-
forms in agricultural research for development. Exp. Agric. 55, 575–596.
https://doi.org/10.1017/S0014479718000200.
42. Hellin, J., Amarnath, G., Challinor, A., Fisher, E., Girvetz, E., Guo, Z.,
Hodur, J., Loboguerrero, A.M., Pacillo, G., Rose, S., et al. (2022).
Transformative adaptation and implications for transdisciplinary climate
change research. Environ. Res, Climate 1, 23001. https://doi.org/10.
1088/2752-5295/ac8b9d.
43. Chervier, C., Piketty, M.-G., and Reed, J. (2020). Theories of change and
monitoring and evaluation types for landscape approaches. In
Operationalizing integrated landscape approaches in the tropics (CIFOR
Bogor), pp. 78–88.
44. Waeber, P.O., Carmenta, R., Carmona, N.E., Garcia, C.A., Falk, T., Fellay,
A., Ghazoul, J., Reed, J., Willemen, L., Zhang, W., and Kleinschroth, F.
(2023). Structuring the complexity of integrated landscape approaches
into selectable, scalable, and measurable attributes. Environ. Sci. Pol.
147, 67–77. https://doi.org/10.1016/j.envsci.2023.06.003.
45. Reed, J., Ickowitz, A., Chervier, C., Djoudi, H., Moombe, K., Ros-Tonen,
M., Yanou, M., Yuliani, L., and Sunderland, T. (2020). Integrated landscape
approaches in the tropics: A brief stock-take. Land Use Pol. 99, 104822 .
https://doi.org/10.1016/j.landusepol.2020.104822.
46. Reed, J., van Vianen, J., Barlow, J., and Sunderland, T. (2017). Have inte-
grated landscape approaches reconciled societal and environmental is-
sues in the tropics? Land Use Pol. 63, 481–492. https://doi.org/10.1016/
j.landusepol.2017.02.021.
47. Riggs, R.A., Achdiawan, R., Adiwinata, A., Boedhihartono, A.K., Kastanya,
A., Langston, J.D., Priyadi, H., Ruiz-Pe
´rez, M., Sayer, J., and Tjiu, A.
(2021). Governing the landscape: potential and challenges of integrated
approaches to landscape sustainability in Indonesia. Landsc. Ecol. 36,
2409–2426. https://doi.org/10.1007/s10980-021-01255-1.
48. Schmidt, L., Falk, T., Siegmund-Schultze, M., and Spangenberg, J.H.
(2020). The objectives of stakeholder involvement in transdisciplinary
research. A conceptual framework for a reflective and reflexive practise.
Ecol. Econ. 176, 106751. https://doi.org/10.1016/j.ecolecon.2020.106751.
49. Brock, S., Baker, L., Jekums, A., Ahmed, F., Fernandez, M., Montenegro
De Wit, M., Rosado-May, F.J., Me
´ndez, V.E., Anderson, C.R., DeClerck,
F., et al. (2024). Knowledge democratization approaches for food systems
transformation. Nat. Food 5, 342–345. https://doi.org/10.1038/s43016-
024-00966-3.
50. Ros-Tonen, M.A.F., and Willemen, L. (2021). Editorial: Spatial tools for in-
tegrated and inclusive landscape governance. Environ. Manag. 68,
605–610. https://doi.org/10.1007/s00267-021-01548-w.
51. Biggs, R., Schl€
uter, M., Biggs, D., Bohensky, E.L., Burnsilver, S., Cundill,
G., Dakos, V., Daw, T.M., Evans, L.S., Kotschy, K., et al. (2012). Toward
principles for enhancing the resilience of ecosystem services. Annu.
Rev. Environ. Resour. 37, 421–448. https://doi.org/10.1146/annurev-envi-
ron-051211-123836.
52. Ostrom, E. (2007). A diagnostic approach for going beyond panaceas.
Proc. Natl. Acad. Sci. USA 104, 15181–15187. https://doi.org/10.1073/
pnas.0702288104.
53. Ostrom, E., Janssen, M.A., and Anderies, J.M. (2007). Going beyond pan-
aceas. Proc. Natl. Acad. Sci. USA 104, 15176–15178. https://doi.org/10.
1073/pnas.0701886104.
54. Meyfroidt, P., De Bremond, A., Ryan, C.M., Archer, E., Aspinall, R.,
Chhabra, A., Camara, G., Corbera, E., DeFries, R., Dı
´az, S., et al. (2022).
Ten facts about land systems for sustainability. Proc. Natl. Acad. Sci.
USA 119, e2109217118. https://doi.org/10.1073/pnas.2109217118.
55. DeFries, R., and Rosenzweig, C. (2010). Toward a whole-landscape
approach for sustainable land use in the tropics. Proceedings of the
National Academy of Sciences 107, 19627–32. https://doi.org/10.1073/
PNAS.1011163107.
56. Meinzen-Dick, R., Zhang, W., Eldidi, H., and Priyadarshini, P. (2022).
Landscape governance engaging stakeholders to confront climate
change. In 2022 global food policy report: Climate change and food sys-
tems (International Food Policy Research Institute), pp. 64–71. https://
doi.org/10.2499/9780896294257.
57. Baudron, F., Duriaux Chavarrı
´a, J.Y., Remans, R., Yang, K., and
Sunderland, T. (2017). Indirect contributions of forests to dietary diversity
in Southern Ethiopia. Ecol. Soc. 22, 1–23. https://doi.org/10.5751/ES-
09267-220228.
58. Baudron, F., Tomscha , S.A., Powell, B., Groot, J.C.J., Gergel, S.E., and
Sunderland, T. (2019). Testing the various pathways linking forest cover
to dietary diversity in tropical landscapes. Front. Sustain. Food Syst. 3,
474255. https://doi.org/10.3389/fsufs.2019.00097.
59. Rasmussen, L.V., Fagan, M.E., Ickowitz, A., Wood, S.L.R., Kennedy, G.,
Powell, B., Baudron, F., Gergel, S., Jung, S., Smithwick, E.A.H., et al.
(2019). Forest pattern, not just amount, influences dietary quality in five
ll
OPEN ACCESS
Article
One Earth 7, 1–16, October 18, 2024 15
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014

African countries. Global Food Secur. 25, 100331. https://doi.org/10.
1016/j.gfs.2019.100331.
60. Gergel, S.E., Powell, B., Baudron, F., Wood, S.L.R., Rhemtulla, J.M.,
Kennedy, G., Rasmussen, L.V., Ickowitz, A., Fagan, M.E., Smithwick,
E.A.H., et al. (2020). Conceptual links between landscape diversity and
diet diversity: A roadmap for transdisciplinary research. Biosci ence 70,
563–575. https://doi.org/10.1093/biosci/biaa048.
61. Perfecto, I., Chaves, L.F., Fitch, G.M., Hajian-Forooshani, Z., Iuliano, B., Li,
K., Medina, N., Morris, J., Jime
´nez, B.O., Rivera-Salinas, I.S., et al. (2023).
Looking beyond land-use and land-cover change: Zoonoses emerge in
the agricultural matrix. One Earth 6, 1131–1142. https://doi.org/10.1016/
j.oneear.2023.08.010.
62. Carmenta, R., Steward, A., Albuquerque, A., Carneiro, R., Vira, B., and
Estrada Carmona, N. (2022). The comparative performance of land
sharing, land sparing type interventions on place-based well-being.
People and Nature 5, 1804–1821. https://doi.org/10.1002/pan3.10384.
63. Hanspach, J., Jamila Haider, L., Oteros-Rozas, E., Stahl Olafsson, A.,
Gulsrud, N.M., Raymond, C.M., Torralba, M., Martı
´n-Lo
´pez, B., Bieling,
C., Garcı
´a-Martı
´n, M., et al. (2020). Biocultural approaches to sustainabil-
ity: A systematic review of the scientific literature. People and Nature 2,
643–659. https://doi.org/10.1002/pan3.10120.
64. Kanamori, M., and Kondo, N. (2020). Suicide and Types of Agriculture: A
Time-Series Analysis in Japan. Suicide Life-Threatening Behav. 50,
122–137. https://doi.org/10.1111/sltb.12559.
65. Guseva Canu, I., Bovio, N., Wild, P., and Bopp, M.; Swiss National Cohort
SNC (2021). Identification of socio-demographic, occupational, and soci-
etal factors for guiding suicide prevention: A cohort study of Swiss male
workers (2000–2014). Suicide Life-Threatening Behav. 51, 540–553.
https://doi.org/10.1111/sltb.12746.
66. Ringgenberg, W., Peek-Asa, C., Donham, K., and Ramirez, M. (2018).
Trends and Characteristics of Occupational Suicide and Homicide in
Farmers and Agriculture Workers, 1992–2010. J. Rural Health 34,
246–253. https://doi.org/10.1111/jrh.12245.
67. Mohanty, B. (2005). ‘We are Like the Living Dead’: Farmer Suicides in
Maharashtra, Western India. J. Peasant Stud. 32, 243–276. https://doi.
org/10.1080/03066150500094485.
68. Deffontaines, N. (2014). La souffrance sociale chez les agriculteurs:
Quelques jalons pour une compre
´hension du suicide. E
´tudes Rural. 193,
13–24. https://doi.org/10.4000/etudesrurales.9988.
69. Milner, A., Spittal, M.J., Pirkis, J., and LaMontagne, A.D. (2013). Suicide by
occupation: Systematic review and meta-analysis. Br. J. Psychiatry 203,
409–416. https://doi.org/10.1192/bjp.bp.113.128405.
70. Scheidel, A., Ferna
´ndez-Llamazares, A
´., Bara, A.H., Del Bene, D., David-
Chavez, D.M., Fanari, E., Garba, I., Hana
cek, K., Liu, J., Martı
´nez-Alier, J.,
et al. (2023). Global impacts of extractive and industrial development pro-
jects on Indigenous Peoples’ lifeways, lands, and rights. Sci. Adv. 9,
33–35. https://doi.org/10.1126/sciadv.ade9557.
71. OP2B (2022). Working towards regenerative and restorative agricultural
value chains: Key learnings from OP2B member initiatives
(OP2B-WBCSD).
72. FPC (2022). Driving transformational change throughout the value chain
(The COnsumer Goods Forum’s Forest Positive Coalition of Action).
73. Hermans, F., Sartas, M., Van Schagen, B., Van Asten, P., and Schut, M.
(2017). Social network analysis of multi-stakeholder platforms in agricul-
tural research for development: Opportunities and constraints for innova-
tion and scaling. PLoS One 12, e0169634. https://doi.org/10.1371/journal.
pone.0169634.
74. Minderhoud, K. (2021). Landscape approach. In Leassons learnt and look-
ing back on five years of the learning agenda on landscape innovation
(Solidaridad).
75. Ros-Tonen, M.A.F., Van Leynseele, Y.P.B., Laven, A., and Sunderland, T.
(2015). Landscapes of social inclusion: Inclusive value-chain collaboration
through the lenses of food sovereignty and landscape governance. Eur. J.
Dev. Res. 27, 523–540. https://doi.org/10.1057/ejdr.2015.50.
76. Dudley, N., Baker, C., Chatterton, P., Ferwerda, W.H., Gutierrez, V., and
Madgwick, J. (2021). The 4 Returns Framework for landscape restoration
(UN Decade on Ecosystem Restoration Report. Commonland, Wetlands
International Landscape Finance Lab and IUCN Commission on
Ecosystem Management).
77. Li, Q., Ge, Y., and Sayer, J.A. (2023). Challenges to Implementing the
Kunming-Montreal Global Biodiversity Framework. Land 12, 2166.
https://doi.org/10.3390/land12122166.
78. Gadsden, G.I., Golden, N., and Harris, N.C. (2022). Place-based bias in
environmental scholarship derived from social – ecological landscapes
of fear. Bioscience 73, 1–13. https://doi.org/10.1093/biosci/biac095.
79. Kashwan, P., V Duffy, R., Masse
´, F., Asiyanbi, A.P., and Marijnen, E.
(2021). From Racialized Neocolonial Global Conservation to an Inclusive
and Regenerative Conservation. Environment 63, 4–19. https://doi.org/
10.1080/00139157.2021.1924574.
80. Chambers, J.M., Massarella, K., and Fletcher, R. (2022). The right to fail?
Problematizing failure discourse in international conservation. World Dev.
150, 105723. https://doi.org/10.1016/j.worlddev.2021.105723.
81. Kholongo, W., Lambert, E., Kaunda, E., Katengeza, S., and Malunga, A.
(2020). Key factors that influence sustainability of community based advo-
cacy groups after phaseout of donor support. International Journal of
Sustainable Development Research 6, 73. https://doi.org/10.11648/j.
ijsdr.20200604.13.
82. Adeyanju, S., Addoah, T., Bayala, E., Djoudi, H., Moombe, K., Reed, J.,
Siangulube, F., Sikanwe, A., and Sunderland, T. (2021). Learning from
community-based natural resource management ( CBNRM ) in ghana
and zambia : Lessons for integrated landscape approaches23, pp. 1–25.
83. Martin, A., Armijos, M.T., Coolsaet, B., Dawson, N., AS Edwards, G., Few,
R., Gross-Camp, N., Rodriguez, I., Schroeder, H., GL Tebboth, M., and
White, C.S. (2020). Environmental Justice and Transformations to
Sustainability. Environment 62, 19–30. https://doi.org/10.1080/
00139157.2020.1820294.
84. D’Ignazio, C. (2022). Creative data literacy. Inf. Des. J. 23, 6–18. https://
doi.org/10.1075/idj.23.1.03dig.
85. Reyes-Garcı
´a, V., Tofighi-Niaki, A., Austin, B.J., Benyei, P., Danielsen, F.,
Ferna
´ndez-Llamazares, A
´., Sharma, A., Soleymani-Fard, R., and Tengo
¨,
M. (2022). Data sovereignty in community-based environmental moni-
toring: Toward equitable environmental data governance. Bioscience 72,
714–717. https://doi.org/10.1093/biosci/biac048.
86. Dencik, L., Hintz, A., Redden, J., and Trere
´, E. (2019). Exploring data jus-
tice: Conceptions, applications and directions. Inf. Commun. Soc. 22,
873–881. https://doi.org/10.1080/1369118X.2019.1606268.
87. Heeks, R., and Renken, J. (2018). Data justice for development: What
would it mean? Inf. Dev. 34, 90–102. https://doi.org/10.1177/
0266666916678282.
88. Husson, F., Josse, J., and Page
`s, J. (2010). Principal component methods
- hierarchical clustering - partitional clustering: why would we need to
choose for visualizing data? Technical Report, 1–17. http://factominer.
free.fr/docs/HCPC_husson_josse.pdf.
89. Page
`s, J. (2015). Multiple Factor Analysis by Example Using R (Chapman
and Hall/CRC), pp. 1–272. https://doi.org/10.1201/b17700.
90. Breiman, T., Cutler, A., and Classification, D. (2011). Package ‘
randomForest.
91. Paluszyska, A. (2017). Structure Mining and Knowledge Extraction from
Random Forest with Applications to the Cancer Genome Atlas Project
(University of Warsaw Master Thesis), pp. 1–76.
92. Liaw, A., and Wiener, M. (2002). Classification and regression by
randomForest. R. News 2, 18–22.
93. Paluszynska, A., Biecek, P., and Jiang, Y. (2020). randomForestExplainer:
Explaining and visualizing random forests in terms of variable importance.
94. R Core Team (2023). R: A language and environment for statistical
computing.
ll
OPEN ACCESS Article
16 One Earth 7, 1–16, October 18, 2024
Please cite this article in press as: Estrada-Carmona et al., Reconciling conservation and development requires enhanced integration and broader
aims: A cross-continental assessment of landscape approaches, One Earth (2024), https://doi.org/10.1016/j.oneear.2024.08.014