ChapterPDF Available

Implementing a gender transformative research approach: Early lessons.

Research in development: Learning from
the CGIAR Research Program on Aquatic
Agricultural Systems
Douthwaite B,1 Apgar JM,1 Schwarz A,1 McDougall C,1 Attwood S,10 Senaratna Sellamuttu S11 and
Clayton T
Contributing authors
Aktar S,1 Apgar JM,1 Attwood S,10 Brown J, Chisonga N,12 Chea S,1 Choudhery A,1 Cole S,1 Clayton T,
Crissman C,
Douthwaite B,
Downing B,
Golam F,
Hak S,
Johnstone G,
Kabir K,
Kamp K,
Karim M,
Kato-Wallace J, Kimsan Lando LA,
Leem N,
Longley C,
Lunda Kalembwe J,
McDougall C,
Meng M,
Mulanda A,
Mulele S,
Manlosa A,
Muyaule C,
Mwasi D,
Nasilele F,
Teioli H,
Tolentino L,
Orirana G,
Paz-Ybarnegaray R,
Perez ML,
Rodericks A,
Saeni E,
Schwarz A,
Senaratna Sellamuttu S,
Siota F,
Songe M,
Rajaratnam S,
Sukulu M,
Sumaya MAD,
Wang J,
Ward A
and Waters-Bayer A
Authors’ Aliations
1 WorldFish
2 CARE International
3 Zambia Ministry of Agriculture and Livestock
4 Caritas-Mongu
5 Catholic Relief Services
6 Peoples’ Participation Service
7 International Livestock Research Institute
8 AVRDC – The World Vegetable Center
10 Bioversity International
11 International Water Management Institute
12 Water Aid, Zambia
This publication should be cited as: Douthwaite B, Apgar JM, Schwarz A, McDougall C, Attwood
S, Senaratna Sellamuttu S and Clayton T, eds. 2015. Research in development: Learning from the
CGIAR Research Program on Aquatic Agricultural Systems. Penang, Malaysia: CGIAR Research
Program on Aquatic Agricultural Systems. Working Paper: AAS-2015-16.
A work like this takes a great deal of time and eort and involves a great many people beyond the list
of lead and contributing authors. We would like to say “thank you to all the program sta who spent
hours compiling the data and information that provided the background for this analysis. Thank you
to all our partners who accompanied us on this journey. Most importantly, thank you to the men and
women in the communities where we work. Thank you for your hospitality and for your trust.
Learning from the implementation of the research-in-development approach 4
Developing community ownership in agricultural research 9
Collaborating for development impact: Learning from research partnership experiences 25
Implementing a gender-transformative research approach: Early lessons 41
More inclusive science for the poor: Linking farmers to researchers using the RinD approach 57
Synthesis: Learning about RinD 81
Notes 88
References 89
Chapter authors: Douthwaite B, Apgar JM,
Schwarz A, McDougall C, Attwood S, Senaratna
Sellamuttu S and Clayton T
Approximately 500 million people in Africa, Asia
and the Pacic depend on aquatic agricultural
systems for their livelihoods. Of these, an
estimated 137 million live in poverty1 (Béné
and Teoh 2014). They live in coastal zones and
along river oodplains and other wetlands,
where they are vulnerable to increasing
population pressure, natural resource depletion
and degradation, biodiversity loss, climate
change, sea level rise, and increasingly
frequent and severe extreme weather events.
The men and women who live in and depend
on these systems are an integral part of the
systems themselves. Socio-cultural systems
are inseparable from natural systems in that
livelihoods make use of both ecological
processes and the diversity of productive
options for growing and harvesting food and
other products that generate income and well-
being (Chiesura and de Groot 2003; AAS 2011).
Aquatic agricultural systems are vulnerable,
diverse, complex social-ecological systems2
to which people continue to apply traditional
management and productive practices in many
A central role for agricultural research in
complex social-ecological systems is to learn
how to use research processes and outputs
in ways that build the capacity of smallholder
farmers and shers to innovate faster, more
eectively and more equitably as a means to
poverty reduction. Men and women farmers
and shers living in aquatic agricultural systems
have always innovated to adapt to change
based on their indigenous and local knowledge.
Today, the increasing rate and scale of change
demands that smallholder farmers and artisanal
shers innovate better and faster than ever
before if they are to maintain a state of well-
being. Vulnerability varies by socioeconomic
group. Women and marginalized peoples tend
to be more vulnerable to sudden change and
often have less access to the range of resources
and factors needed to support innovation
(time, acceptability of risks in experimenting,
networks, etc.). Furthermore, inequities
in access to agricultural resources reduce
productivity and the ability to secure sucient
nutritious food throughout the year.
The CGIAR Research Program on Aquatic
Agricultural Systems (AAS) began operation
in 2011 with the aim of reducing poverty and
improving food security for small-scale shers
and farmers dependent on aquatic agricultural
systems (AAS 2011). As well as seeking to
generate outcomes that directly improve
the productivity and resilience of aquatic
agricultural systems through agricultural
research, the program set a goal of better
understanding how agricultural research
can itself innovate such that it meets the
challenge framed above—helping poor and
vulnerable people achieve more equitable and
more sustainable livelihoods from the social-
ecological system they are part of. To capture
the intent of this goal and to contrast the
program’s approach with “business as usual”
agricultural research, AAS coined the term
“research in development” (RinD).
In this document, business as usual refers to
the common problem-solving process used
in science where the researcher is understood
to stand objectively outside the system under
study and produce a research output, which is
then adopted and adapted by users to solve
a specic problem. In the business-as-usual
model, adoption or adaptation is usually not
the researcher’s concern. Typically, researchers
are neither recognized nor rewarded if users
adopt their output (Campbell et al. 2015).
The result is a disconnect between researcher
and user, resulting too often in research
technologies that do not meet local needs
and are abandoned. Consequently, much
technology development does not necessarily
have development impact. The optimistic
but common term used to describe these
technologies is “on the shelf.The business-as-
usual model has also been called the “pipeline”
approach (Sumberg 2005), the “central source of
innovation” model (Biggs 1990), and in industry,
the “delivery” mode or “over the wall” approach
(Leonard-Barton 1998).
In contrast, the term “research in development”
implies an approach where the research
is carried out within and as part of a more
complex social-ecological system. In this
approach, the distinction between “inside” and
outside” becomes less obvious and innovation
is seen as a process that links across them.
This does not mean that all research must be
implemented directly with farmers and shers;
indeed, there is a need for basic research to
support improved productivity in aquatic
agricultural systems (such as developing a
new variety of rice) that requires scientists
to work away from the farm. The emphasis
on linking and innovation, however, calls for
all agricultural research to be cognizant of
how its outputs support and engage with
local processes of innovation to achieve
development outcomes. This recognition
pushes agricultural researchers to think beyond
the specic problem they are aiming to address
and embrace a broader perspective on how
development is achieved.
Approaching systems through only their parts
means we run the risk of not appreciating the
whole. Poverty in social-ecological systems
is multifaceted, and the causes of inequality
are often hidden (Pelling 2010; Kabeer 2012).
Consequently, an approach to agricultural
research that aims to support poverty
alleviation and is particularly concerned about
marginalized peoples must look beyond the
easily identiable agricultural problems that
business-as-usual models are good at solving.
It must also understand underlying social
dynamics and the patterns of interactions
between stakeholders that may inhibit
equitable outcomes for all. This more complex
and socially aware approach to agricultural
research builds on and extends the experience
and learning from farming systems research
(e.g. Gilbert et al. 1980; Scoones et al. 2009) to
embrace underlying development processes
and appreciate patterns of interactions. It is
aligned with a growing eld of research and
practice in development that acknowledges
complexity (e.g. Jones 2011; Ramalingan 2013).
RinD intends, therefore, to take a more holistic
approach and look beneath the surface so that
agricultural research can equitably support
capacity to innovate and achieve sustainable
development outcomes.
The AAS program proposal (AAS 2011) dened
an RinD approach to agricultural research
as one that is cognizant of the multifaceted
nature of poverty and one that aims to address
challenges in complex social-ecological
systems. The RinD approach as it is now
understood by the program has evolved from
the initial intent to greater articulation of its
elements and requirements as the proposal has
been operationalized. As of October 2015, AAS
has been in operation and developing the RinD
approach for 3 ½ years in ve hubs.
Hubs are dened as “locations within key
aquatic agricultural systems where innovation
and learning can bring about development
outcomes” (AAS 2013, 5). As of May 2015,
AAS was working in ve hubs: the Barotse
oodplain in Zambia; the Southern Polder
Zone of Bangladesh; the Tonle Sap oodplain in
Cambodia; the Visayas-Mindanao region in the
Philippines; and Malaita and Western provinces
in Solomon Islands.
The biophysical and socio-cultural context
of each of the hubs is unique and requires
adaptation of the implementation approach
to each context to address relevant
development challenges. Consequently, the
RinD approach is being developed through
a case study approach to learning from
implementation, with each hub as a case of
RinD implementation. This forms a core part
of AAS research on the RinD approach, which
aims to generate lessons that are useful more
broadly in the eld of agricultural research and
development practice.
This working paper aims to synthesize and
share learning from the experience of adapting
and operationalizing the RinD approach to
agricultural research in the ve hubs. It seeks
to share learning about how the approach
is working in context and to explore the
outcomes it is achieving through initial
implementation over 3 ½ years. This learning
can inform continuation of agricultural research
in the second phase of the CGIAR research
programs and will be useful to others aiming
to implement research programs that seek to
equitably build capacity to innovate in complex
social-ecological systems. In the next section,
we describe what RinD was understood to be
in 2013 (Dugan et al. 2013), providing a starting
point for the chapters that follow, which
explore lessons about particular aspects of the
approach and their outcomes.
The RinD approach
The rst step under the RinD approach in the
hubs was to articulate a hub development
challenge collectively with stakeholders
through a participatory planning process.
Scoping and diagnosis of particular challenges,
both biophysical and socio-cultural, was
undertaken by multidisciplinary research teams.
The resulting hub development challenges
provide the guiding collective vision for how
agricultural research in each hub can contribute
to achieving development outcomes and set
up the program of work. Stakeholders then
agree to tackle the hub development challenge
and implement interventions. Planning the
interventions requires further articulation of
specic research agendas.
The approach used to implement these
interventions, described in Figure 1, utilizes four
elements: commitment to people and place,
participatory action research, using a gender-
transformative approach, and facilitating learning
and networking. The approach also requires two
enabling conditions: partnerships and capacity
development. The elements build on a range
of theories on and experience from agricultural
research-for-development experiences (e.g.
Hawkins et al. 2009; Hall et al. 2014).
Commitment to people and place is based on
the assumption that people have the potential
to innovate and bring about meaningful
change, and that a sustainable way to improve
livelihoods is to leverage this potential for deeper
and longer-lasting change (e.g. Chambers and
Ghildyal 1985; Hickey and Mohan 2004). AAS
aims to foster development within communities
through engaging the poor and marginalized
across scales to help improve their access to
and use of the process of agricultural research,
as well as the research outputs produced. This
takes time and commitment from researchers
working collaboratively with local stakeholders as
everyone learns together how to make the most
of the potential that lies within the system.
Participatory action research (PAR) is the core
engagement process that RinD uses to ensure
beneciaries are co-owners in the process of
nding solutions to their own problems and
in building their own capacity to reect and
innovate (Reason and Bradbury 2008), and is
described in Apgar and Douthwaite (2013).
A gender-transformative approach embodies
a commitment to and strategies for social
transformation that result in equity and equality
among diverse actors (Cole et al. 2014b). A
gender-transformative approach frames the
research process as one that combines technical
knowledge generation with equity-oriented
transformative learning. AAS seeks both to
address the visible aspects of gender and other
social gaps and to create opportunities for
actors to shift the underlying norms, attitudes,
practices or policies that shape these gaps.
Figure 1. Six elements that constitute the AAS RinD approach (adapted from Dugan et al. 2013).
Eective partnerships
Strengthened capacities
Impact on
poor and
to people and
Photo Credit: Sanjiv de Silva/IWMI
The participatory process of data collection.
Learning and networking stress the need for
adaptive management, learning and adapting as
hub programs of work are implemented, as well
as using monitoring and evaluation as another
set of tools to ensure this happens (Douthwaite
et al. 2014). RinD requires those involved to
be aware of their own mindset throughout
the implementation and to learn new skills,
such as facilitation and networking. Eective
partnerships acknowledge that intervening
meaningfully requires working with others, and
that building partnerships at all levels is the
pathway to greater development outcomes.
This paper is the result of program-level PAR.
Action researchers recognize that there are
multiple and overlapping levels of inquiry,
referred to as rst-, second- and third-person
action research (Reason and Torbert 2001).
First-person research refers to learning through
individual self-inquiry. Second-person research
is relational and includes reecting and learning
with peers in a community about a particular
area of theory or practice. Third-person
research refers to learning with stakeholders
about the broader issues that are the focus of
a specic inquiry. AAS engages all three levels
to surface and document learning that is used
to improve practice and enable others to learn
and to answer research questions about RinD,
contributing to the global discourse through
the production of international public goods
(Figure 2).
As the basis for identifying and measuring AAS
outcomes, AAS has implemented a learning
system that includes hub RinD implementation
teams who engage in third-person research
with hub stakeholders. These teams engage in
their own second-person research and annually
consolidate their learning around specic
research areas. Once a year, representatives
from hub teams come together for a cross-hub
review and engage in another level of second-
person research with their peers from other
hubs. Cross-hub learning is facilitated by a
global RinD team of researchers based outside
the hubs. This working paper is an output of
this cross-hub learning.
In January 2015, the AAS cross-hub review
brought together hub sta and the global team
at WorldFish headquarters in Penang, Malaysia
(AAS 2015). The review process enabled hub
teams to share what they had learned across
contexts and to articulate common themes. Prior
to the cross-hub review, each hub team had
carried out their own review with stakeholders
in which they reected on three topics: what
worked and what did not in implementing and
building capacity for RinD; early evidence of
outcomes; and the continued relevance of the
overall hub strategic framework (AAS 2014).
This review process identied the following six
areas of collective learning about RinD from
across hubs:
learning from community engagement
learning about partnerships
learning from the integration of the gender-
transformative approach
learning about how to make science more
learning about capacity development
generating a better articulation of RinD and
its value.
In the following chapters of this working paper,
the rst four areas of learning are investigated.
Figure 2. AAS learning system, including rst-, second- and third-person action research.
3rd person
2nd person
2nd person
1st person
RinD team
hub 2
RinD team
hub 1
Photo Credit: Silvia Sarapura/WorldFish
Community members in the Khulna hub attending an AAS meeting, Bangladesh.
Chapter authors: Apgar JM, Tolentino L, Golam
F, Aktar S, Orirana G, Saeni E, Chea S, Hak S,
Chisonga N and Lunda Kalembwe J
A core aspect of the AAS RinD approach is its
focus on engaging with communities through
a process known as
. This is a methodology
used in many practitioner-based elds to
support engagement of stakeholders in
the process of research in order to promote
empowerment and behavior change (e.g.
Reason and Bradbury 2008). AAS builds on
the long history of farmer participation in
agricultural research (e.g. Chambers and
Ghildyal 1985; Biggs 2008; Scoones et al. 2009)
and extends it through a purposeful approach
to community engagement.
The AAS RinD approach assumes that using
participatory engagement with stakeholders
in designing, planning, implementing and
learning from agricultural research will lead
to empowerment and ownership such that
more lasting outcomes can be achieved.
Examples from health (e.g. Tindana et al. 2007;
Nakibinge et al. 2009), education (e.g. Weerts
and Sandmann 2008; Butin 2010), business (e.g.
Bowen et al. 2010) and community development
(e.g. Tamarack 2007) illustrate that better results
can be achieved when communities are involved
in development processes that aect them. AAS
believes that engagement with a select number
of communities in a hub over the lifespan of the
program can inform and build a joint research
This chapter examines what we have learned
about community engagement and PAR.
Design of community engagement
within RinD
Commitment to people and place: PAR
across scales
Commitment to people and place and PAR are
both elements of the AAS RinD approach. In AAS,
PAR is composed of iterative, facilitated cycles of
planning, acting, observing and reecting with
stakeholders at the community and hub levels.
Through this process, stakeholders identify
and begin to address their own development
challenges through agricultural research. This
process allows communities to reect on how
change is happening, thus becoming an integral
part of the monitoring and evaluation system.
Figure 3 shows the two levels at which the
program engages with stakeholders.
The AAS PAR process starts with a
multidisciplinary research team scoping
the biophysical and social dimensions of
the aquatic agricultural system to identify
opportunities and development challenges.
This scoping leads to the selection of local sites
for community engagement and articulation
of a hub development challenge that guides
the program. Then, researchers engage with
communities and document community
visions, priorities and action plans, which are
owned by the communities. The nal step of
planning during the rst cycle is a workshop
that produces a program of work for the hub,
including research initiatives that address
community visions and support community
The next cycle of stakeholder engagement
starts with initiative planning. An initiative
includes research and development activities
with partners that directly support community
action plans and answer identied research
questions. Concurrently, communities
continue their cycle of planning, acting and
reecting on what they have learned through
implementation. An annual review workshop
provides opportunity to adjust initiatives and
community actions. The intent of continued
engagement at two levels is to build and
strengthen links between the local actions and
achievement of outcomes on the ground with
system-level processes of research and change.
Together, the two levels of engagement aim to
tackle the hub development challenge.
Figure 3. AAS program engagement cycle across scales in hubs. (AAR stands for after-action
Hub-level engagement
Review of
initial actions
and priorities
and initiatives
visions and
action plan
AAR and
Annual hub
review and
Annual hub
review and
Scoping and
AAR and
Box 1. Principles for PAR design and
implementation across AAS hubs
Ownership: The process is owned
by the participants, who
dene real-life problems to
address through PAR.
Equity: Facilitators recognize
multiple voices and power
relations and are mindful
of who is participating and
Shared analysis: The process emphasizes
jointly shared
responsibilities for data
collection and analysis
to support improved
understanding and action.
Feedback: Results of the process
are fed back to the
participants for ongoing
learning that supports
adaptation and
Principles of engagement
Engagement with hub and community
stakeholders through PAR is designed to
ensure that agricultural research helps achieve
community goals relating to production, food
security, nutrition, income, environmental
conditions, etc. The focus on practical solutions
to real-life concerns (e.g. reduced soil fertility,
water availability, reduced incomes, etc.) means
that the process is context specic. Many issues
and concerns, however, occur across dierent
farming communities, thus enabling sharing of
learning and scaling of impact. For consistent
implementation across AAS hubs, we identied
four principles to guide PAR implementation
(Box 1). These principles are consistent with
similar analyses (e.g. McTaggart 1991; Stringer
2007; Reason and Bradbury 2008).
Ownership. The rst principle assures that by
returning to their own community visions to
reect on what has been achieved, the men,
women and youth of the locality co-own the
process of research and the learning that
emerges throughout the implementation of
their action plans and the supporting initiatives.
Equity. This principle helps ensure that
facilitation teams (co-researchers) pay attention
to the multiple voices that inuence the
community vision and action plans and the
processes through which they are developed,
implemented and reected upon. This principle
is further strengthened through the program’s
transformative approach to gender (see the
gender chapter for further explanation of the
approach to gender).
Ensuring equity in the PAR process requires
strategies for the creation of “safe spaces” where
men, women and youth can freely express
themselves and safely question underlying
norms that contribute to inequity and
inequality. Specic research interventions to
support the achievement of community visions
need to be designed and implemented in ways
that are cognizant of social dierentiation and
implications for participation and benet. This
requires initial research to understand why
inequities and inequalities related to gender,
ethnicity and religion exist, as well as how they
aect choices and outcomes. That knowledge
can then be used to design activities that
facilitate change in underlying attitudes and
beliefs and manage any consequent tradeos.
Shared analysis. The third principle focuses
on an area of research practice: analysis.
Implementing this principle means that
researchers who are facilitating the PAR process
enable other stakeholders who are co-researchers
to take part in the analytical steps that lead to
greater understanding of a particular issue that
relates to the collective concern. Appropriate
data collection and analysis methods are used,
depending on the specic question being
addressed. Researchers have a responsibility to
proactively involve stakeholders in the process
such that the group as a whole can learn,
rendering the results of the research process
more useful and able to address real-life concerns.
Feedback. The fourth principle emphasizes the
commitment to support ongoing development
and enablement of joint learning. By
emphasizing feedback mechanisms, researchers
are required to think beyond production of
a research output and consider how to keep
the research connected to community visions,
particularly in relation to how outputs are used
and how they may contribute to achieving
desired development outcomes.
Box 2. Strength-based approach
When facilitators meet with people in
communities, they look for their strengths.
They do not start from their weaknesses. A
strength-based approach, or SALT, is a mode
of interaction with communities.
S : Stimulate, Support, Share
A : Appreciate
L : Listen, Learn, Link
T : Transfer, Team
Initiating community engagement: The
community life competence process
AAS recognized that implementing community
engagement was outside CGIAR’s area of
expertise when designing the program.
Consequently, the program developed
a partnership with Constellation,3 an
international nongovernmental organization
(NGO) with relevant experience who shared
similar goals of building local capacity to
respond to development challenges (see
the partnership chapter for more on shared
partnership learning). The community
life competence process developed by
Constellation was adapted to RinD and used
to initiate engagement with community-
level stakeholders in all ve hubs during
the rst cycle of engagement. Constellation
coaches worked closely with hub teams and
implementing partners to build their capacity
and guide implementation.
The community life competence process is a
strength-based approach in that it emphasizes
a particular mindset among facilitators (Box
2). The process comprises a number of steps
that lead to development of community action
plans. Community mobilization is initiated
through visits to selected communities to build
relationships, identify community strengths and
stimulate members of the community to think
critically about their situations. Mobilization
involves identifying local facilitators (both men
and women) who can act as a bridge between
the program and the community and who
become community researchers.
The next step is “dream building” to develop
a community vision of success. Men, women,
old and young are rst engaged separately
to create safe spaces for their own visioning
processes. Consolidation of the dierent visions
to develop a collective vision is facilitated
where desired and appropriate. From the
dream as articulated, community members (as
a collective or in separate groups, depending
on the context) then identify priority areas for
action. They conduct a self-assessment as a
critical reection on their situation in order to
identify constraints. This exercise is aimed at
identifying gaps between their present situation
and their desired state. Next, they prioritize
areas and identify actions that a group of people
in the community are motivated to undertake
to move towards achieving their vision of
the future (in many cases these are actions
for the whole community, such as building a
community-owned market). This stage is called
“prioritization and action planning. The result
is a set of community-owned action plans
(some collectively owned and some owned by
smaller groups or by a particular social group
such as women or youth) with commitment to
implement the plans using local resources.
Ongoing community engagement
through PAR
Communities then proceed to implementing
their action plans. The local community
facilitators work with program sta, who
support them in implementation. As
the research initiatives take shape and
implementation begins, areas of more specic
joint inquiry are identied. Examples include
productivity research in Bangladesh supporting
implementation of technologies for shaded
ponds; research on the impact of savings
and lending groups for income generation in
Zambia; research on access to markets in sh
value chains in Zambia; productivity research
to identify suitable sources of seed for SUPA
rice4 in Zambia; and piloting rice eld sheries
management practices in Cambodia.
Action plans and associated research lead
to the observation and reection step that
enables those involved to understand the
changes that may be occurring and to measure
their achievements. As Figure 4 illustrates, the
ongoing community engagement process is
the main vehicle for a village-level participatory
monitoring and evaluation system focused
on outcomes and learning. Community
action plans are revised on an annual basis,
building on what was learned the previous
year. Documentation of this process feeds into
program research to understand if and how the
RinD approach is working.
Community engagement
implementation models
In this section, we illustrate how the community
engagement design was implemented in the
AAS hubs. Table 1 describes the implementation
models used. The hubs vary in their biophysical
contexts, ranging from inland water systems
(Barotse and Tonle Sap oodplains) to coastal
marine systems (Visayas-Mindanao and Malaita)
and delta systems (Southern Bangladesh
Polder Zone). These systems face varying
degrees of ecosystem degradation, and a large
portion of the population in each are poor and
marginalized and depend heavily on the social-
ecological system for their livelihoods through
the provision of multiple ecosystem goods. The
issues associated with achieving community-
dened development aspirations and goals are
dierent in each hub, as they are driven by the
context and the program of work.
Contextual variations shaped how community
engagement was implemented:
Dierent biophysical systems and
varying agroecological zones. These
dierences inuence the degree of
livelihood dependence on capture sheries,
aquaculture, agriculture, livestock rearing
and wild biodiversity harvesting.
The hub development challenge and its
associated theory of change. While all hub
challenges focus on the potential of the
aquatic agricultural system, the specics of
the potential vary from the ood pulses in
the Barotse and Tonle Sap, to the salinity
gradient in the Southern Bangladesh Polder
Zone, to the rich natural resources in Malaita.
The cultural and social diversity found
within the hubs. The Barotse oodplain in
Zambia and the Malaita hub in Solomon
Islands are both territories of indigenous
peoples and use traditional governance
systems. In Cambodia, the Tonle Sap
oodplain is home to a majority of ethnic
Khmer and several other ethnic minorities,
but due to the area’s political history is
managed through a hierarchical government
Figure 4. Community engagement process designed as PAR and participatory
monitoring and evaluation for learning.
monitoring and
evaluation cycle
action research
of action plan
and learning
Visioning and
action planning
Local, district and national research
and development systems in place. The
presence of formal national or international
agricultural research systems in the hubs
varies, as does the level of development
History of WorldFish and CGIAR work and
the presence of ongoing bilateral projects.
The hubs sit on a continuum from areas with
a long history of CGIAR and WorldFish work
in-country and large bilateral programs also
implementing research (Solomon Islands and
Bangladesh), to areas with moderate bilateral
funding (Cambodia and the Philippines), to
one area with no prior WorldFish presence or
bilateral projects (Barotse oodplain).
The community engagement implementation
model in each hub has been signicantly
inuenced by previous partnerships with the
program in the hub and which relate, in part,
to the previous and ongoing CGIAR work
in the area and the existing local capacities
(see the partnership chapter for more detail
on partnerships). As is shown in Table 1, the
implementation model in all hubs consists
of a mix of local facilitation teams and
external support provided through NGOs or
other partners and AAS sta. The support
arrangements vary depending on who the
main program-implementing partners are. For
example, in the Visayas-Mindanao hub, the
primary supporting partners are government
organizations, while in Zambia and Cambodia
they are local NGOs. In Zambia, the Barotse
oodplain hub is the traditional territory of the
Lozi people, which requires the program to
work with the traditional governance system
(the Barotse Royal Establishment), and as a
consequence the village chiefs are members of
the community facilitation teams.
All hubs used the community life competence
process and were supported directly by
Constellation during the initial visioning and
action planning. This produced a similar yet
locally adapted process. The outputs included a
broad long-term vision for each community and
a number of community-owned action plans
that indicate where communities are motivated
and able to move towards achieving their
dream. In most cases, communities identied
the support they required from external agents
to implement their action plans, creating
opportunities for linking with support networks
and agricultural research. In all cases, these
outputs informed the development of the hub
strategic framework and the initiatives designed
to address the hub development challenge.
During initial community visioning and action
planning, there was some adaptation of the
community life competence process steps. In
most cases, separate groups of male, female
and youth participants rst developed their
own visions. A notable dierence among hubs
was the extent to which research was discussed
during the initial visioning and action planning.
In the Southern Bangladesh Polder Zone, for
example, the presence of many development
NGOs and projects coupled with researchers
playing a facilitation role in communities led
to a narrower focus of engagement on farmer-
, while in other hubs community action
plans were broader.
Building on the initial use of the community
life competence process, after-action reviews
became the main vehicle for implementing
the reection step (Figure 3). Dierent
strategies were used to create links between
the community-owned action plans and
implementation of interventions that form
the stakeholder-driven research initiatives in
each hub. For example, in Zambia, an early
opportunity to work with savings and internal
lending communities through partner support
created a unique way of implementing research
on use of PAR and the gender-transformative
approach while supporting community action
on increasing income.
Hub Hub context Implementation model
Remote coastal marine setting with little
infrastructure and few development
interventions. Mainly subsistence livelihoods.
Declining quality of marine and land resources
and increasing populations.
Engaged with three clusters of communities.
Total hub population: 137,596 (2009 census).
Community facilitators facilitate activities
and mobilize communities, while community
champions support them and provide a link
between the community and the program
team. Both roles are voluntary. AAS sta
(particularly a community coordinator)
supports facilitation and documentation of
the process. Partner NGOs and universities
provide technical support as required through
implementation of research activities.
Floodplain of the Zambezi River, traditional
territory of the Lozi people. Dual governance
system: the traditional authority referred to
as the Barotse Royal Establishment and the
central government. Lozi culture and livelihood
strategies intimately linked to the ood pulse;
seasonal movement of people and animals
from low to high lands. Poorest province
in Zambia. Fisheries important beyond the
oodplain and a recent decline in natural
resource management systems due to shifting
Engaged with 10 communities.
Total hub population: 522,298 (Central Statistics
Oce 2010).
Community facilitation teams include
community facilitators selected by
communities, traditional leaders and extension
ocers (Ministry of Agriculture and Livestock
sta). Community facilitators are paid a small
amount to cover their day-to-day expenses. The
local teams are supported by an NGO partner.
Specic interventions that support community
action plans are implemented with the support
of NGO partners and CGIAR scientists, who
provide technical expertise and implement
capacity-development activities.
Minimal presence of other NGOs in the 10
Tonle Sap
Seasonal ood pulse of the Tonle Sap Lake and
oodplain oers opportunities for improved
productivity, while water governance is a major
challenge for those who rely on it. Communities
looking to improve water governance for rice
productivity and sh farming. High incidence
of poverty despite the benet from the ood
pulse, particularly among oating villages
where livelihoods depend heavily on shing.
Engaged with 12 communities.
Total hub population: 1.5 million, of which
900,000 live in water-based or stilted villages.
Community facilitators from each of the
communities facilitate activities and mobilize
communities for the ongoing process.
Facilitators provide links to NGO partners
and the AAS team. Facilitators are volunteers
and receive a small per diem when they
attend events hosted by the AAS team and
NGO partners. NGO partners directly support
community facilitators with facilitation and
documentation. The AAS hub team, mainly
AAS sta, provides support for capacity
Polder Zone,
Coastal delta system with varying salinity
gradients in polders (oodplains enclosed by
embankments) aected by climate change.
Agriculture- and aquaculture-dependent
livelihoods. High population density and
disparity between land owners and landless.
Engaged with 16 communities in six districts.
Total hub population: 7.42 million.
Program ocers are AAS sta assigned to
each polder zone and who together with NGO
partner sta are responsible for facilitation
of all activities in communities. Scientists
from CGIAR and partners provide technical
support to AAS program ocers as they
implement PAR agendas. Hub sta provide
support on systematic documentation and the
participatory monitoring and evaluation system
in place.
Coastal marine areas of Visayas and Mindanao
provinces. High poverty rates and dependence
on shing and agriculture and highly vulnerable
to climate change. Development challenges on
governance of sheries and access to markets.
Engaged with eight communities.
Total hub population: 18.6 million.
Local community facilitators are employed
as eld research aids. Partner community
facilitators are sta of government partner
organizations who work closely with local
facilitators. AAS sta are organized as
community immersion teams and provide
direct support to partner and local community
facilitators, particularly with documentation
and monitoring and evaluation.
Table 1. Community engagement implementation models in each hub.
Learning about the community
engagement process
Through reection and analysis within and
across hubs, we have identied the following
lessons about how the community engagement
process is working and what its outcomes are.
Start with a community vision
In all hubs, the community life competence
process enabled a broad vision to be
articulated. At rst, there was concern that
such a broad vision would raise unrealistic
expectations and move outside the CGIAR
mandate. However, the process gives
communities the freedom to think about their
future on their own terms. By not limiting their
understanding of their livelihoods and lives
from an external perspective and not using
program language to frame issues on which to
focus, communities were able to identify their
strengths and take actions based upon those
strengths. In reection sessions during which
community members discuss their learning
and achievements, returning to the broad
community vision motivates them to continue
their journey. In this way, the role of research is
understood as supporting that journey from the
beginning, rather than leading the denition of
solutions to identied “problems” from the start.
For AAS, the visioning process has helped ground
an approach that looks at the whole system in
the local reality of the hub. Through building a
much broader understanding of the communities
and their aspirations, stakeholders can begin to
identify the relationships between various system
components. For example, in the Barotse hub,
sheries management was a major concern in 10
community action plans. This concern stimulated
action at higher levels, requiring strengthened
collaboration with important stakeholders such
as the Department of Fisheries, the Barotse Royal
Establishment and sh trader associations. In this
case, locally dened concerns resulted in a hub-
level response: the formation of village-based
sheries management committees as part of a
co-management approach that brings together
government, the private sector, traditional
leadership and the community. Understanding
local systems in the context of a broad vision
provides a big picture that helps inform the
research agenda, build on local strengths and
create links across scales.
Deepening engagement requires staging
and building trust
Across all hubs, the implementation phase
of initial community action plans was
accompanied by strategies to deepen
engagement. These strategies included
building a better understanding of the critical
and underlying issues that create opportunities
for research, and developing strategies that
include marginalized groups.
Varying strategies were used to build a better
understanding of the issues to be addressed
through interventions and that enabled
agricultural research to directly support the
actions of dierent groups within communities.
For example, in Bangladesh, homestead
agricultural systems are mainly managed by
women and are critical for household food
security and nutrition. Separate focus group
discussions with women led to a better
understanding of the challenges they faced,
such as not having access to quality seeds. As
a result, women farmers set up research trials
together with professional researchers from
local universities and government agencies and
have developed their skills for identifying the
best seed varieties for their household plots.
In the Philippines, focus group discussions
identied illegal shing gear as one of the
critical issues, leading to a multistakeholder
dialogue to bring illegal shers into sheries
management discussions. The trust built
through AAS researchers spending time in
the communities enabled the Balingasag
community to engage with a deeply rooted
issue. In Cambodia, the use of a coding
system enabled communities and facilitators
to collect and analyze qualitative data to
better understand the local situation and to
feed directly into the planning of the three
hub initiatives. One of them, the land and
water management initiative, now includes a
case study on Tram Pear Lake rehabilitation,
which was informed by the results of cross-
village analysis that illustrated the need to
improve water supply to increase rice and sh
No participatory process is perfect, and no
community is ever fully engaged from the
outset. Consequently, implementing the
principle of equity in community engagement
requires hub teams to be cognizant of who
was engaged in the process to begin with and
who was left out. Dierent strategies can be
used to build this understanding, and over time
strategies may be developed to reach out to
those who were not part of the initial broad
visioning process.
For example, in Malaita, the visioning process in
Alea and Kwai communities began by bringing
residents from surrounding villages together
at one location to develop community visions
and action plans. During implementation,
however, it became clear that even though
people were from the same tribe, families and
church, they were not accustomed to working
together. In contrast, in Fumatoo, where the
participants live closer together and consider
themselves one community, action plans were
collectively implemented with little diculty.
Ensuring broad participation, implementation
and ownership of a vision or an action plan
in the context of Malaita, therefore, required
adaption of the strategy to work with smaller
groups that are geographically close and
have the experience of working collectively. A
strategy was developed to deepen engagement
with a cohesive cluster and to plan to scale the
approach to neighboring hamlets to develop
dierent action plans.
In Cambodia, there are high levels of inequality
in villages. Here, the program works through
NGO partners who have been working in
the same villages for some time. This creates
a challenge for us to understand if the NGO
engagement process is broad enough and
reaches the marginalized. In order to better
understand who the program was engaging
with in relation to the whole community,
participatory well-being ranking was
conducted and used together with a coding
grid and a monitoring map. The coding system
allowed monitoring of who the program was
engaging with and provided a planning tool
to help identify the poor and marginalized
within communities. For example, in Tram
Pear village, use of the coding system enabled
the team to see that an action plan around
rehabilitation of the lake was the result of
participation by a group of people who all
lived close to and beneted from the lake. This
knowledge allowed the team to develop a more
household-focused strategy to reach out to the
most marginalized who lived far from the main
part of the village.
In Bangladesh, while the engagement process
remained open to all, farmers were selected
on the basis of their interest and motivation
to do research, which did not fully address
the gender or social dierence dimension.
Further engagement led to research on
small homestead shaded ponds intended
to help women overcome the challenge
of low productivity. Similarly, separate
discussions with men led to research on
eld crops. To understand the wealth status
of the participants, a participatory wealth
ranking method was developed. Community
members set up wealth ranking criteria for their
community and divided themselves according
to family income. This helped AAS sta
understand that the poorest sections of some
communities did not participate in PAR, as they
had neither land nor ponds.
In Zambia, 3 of the 10 communities in Senanga
District are home to two tribes: Mbunda and
Lozi. The Lozi are the original inhabitants,
and the Mbunda are immigrants who have
integrated over time. During the engagement
process, it was discovered that the Mbunda
coming into the communities found the
fertile land all taken by the Lozi. This pushed
the Mbunda into non-agriculture enterprises.
Discussions during the visioning exercise led
to action on canal clearing to free more land
for cultivation, which led to improved access to
land and more involvement by the Mbunda in
agricultural activities.
From across the cases, we see that starting
with a broad vision that can be implemented
through a relatively consistent methodology
across contexts must be accompanied by
contextualized strategies to dig deeper and
understand how to support communities and
groups within them to tackle their own issues
while avoiding elite capture. The deepening
process is also critical to identifying areas for
agricultural research. Treating community
members as co-researchers requires diagnostic
studies along with the engagement process
such that research interventions support
community motivation and change.
Digging deeper in the community engagement
process and fully embracing the equity principle
of PAR has not been easy. One challenge faced is
further discussed below regarding the research
teams’ mindsets and capacity to bring social
analysis and a critical lens into agricultural
research processes (see also the gender
chapter). Our experience suggests that staging
when to deepen engagement is important
to ensure the research process rst builds
trust. Working through a stronger relationship
built on trust means research is better able to
contribute to the development process and can
act upon hidden challenges that do not tend
to surface in initial action planning. With time,
the program can reach out to more, ensure the
marginalized are involved and begin to address
challenges that require deeper, transformative
Responding to broad issues requires
networking and partnerships
Not all concerns and opportunities identied
by communities as a starting point are
ones that a CGIAR research program could
engage with directly. Hub teams were initially
concerned about their inability to respond
to all the identied areas, and feared that
community residents might lose interest. This
created a tension between holding onto a
broad and holistic view of the development
process and staying true to the agricultural
research mandate of CGIAR. This tension
was managed by developing a strategy for
responding to broader community needs
through partnerships and playing a bridging
role so communities could connect to relevant
stakeholders with the capacity and mandate
to address concerns relating to infrastructural
development, health and sanitation, or delivery
of agricultural inputs.
For example, in the Philippines, the
community’s dreams and action plans were
presented to various stakeholders during a
consultation workshop that led to government
responses. In Pinamgo community, one of
the priority dreams was the repair of a solar
water system. As a result of sharing these
plans, the local government unit committed
to providing funds to buy a water pump.
In a case where the issue was education, in
Mancilang, the Department of Science and
Technology provided a number of scholarships
that enabled some of the youths to pursue a
college education. This has built trust in the
program and created an opening for the role of
agricultural research to contribute.
In Cambodia, NGO partners participating in
the engagement process have found ways to
link communities in order to address some
of their priority concerns. For example, as a
result of community visioning in one village,
people realized that to achieve their dream of
improving and diversifying livelihood activities
they needed to renovate a bridge that connects
them to the market and other communities.
Realizing that they needed further nancial
support, villagers shared their action plan with
various stakeholders, including a mung bean
association, a tour association and local sh
traders, to raise funds. Stakeholders provided
nancial support, and during a reection
workshop in March 2014, villagers reported
that with the bridge renovated they could
focus more on sh farming, as they had easier
access to the market. In this way, the program
can support responding to broader needs that
are linked to the ability to address agricultural
concerns through research.
In Bangladesh, two complementary strategies
were used. First, a research support team was
formed in Khulna in 2013 with scientists from
national research institutes and universities
to provide science and technical assistance to
farmers to carry out eld experiments. During
implementation, it transpired that the support
team was unable to deliver services at the
expected level to communities far from Khulna
city. In response, the hub team created a second
layer of research and technical support. The
new system consists of two layers of support
to communities: the original research support
team members (the rst layer) still provide
science and research support, but now include
more researchers from local research stations,
while the new second layer is made up of
members of local extension services (e.g. the
Department of Agricultural Extension and the
Department of Livestock) who can provide
support on an ongoing basis in more remote
areas. This system linked farmers to local
service providers and provided an enabling
environment to access technical advice. Second,
a process of building relationships between
development organizations, government
departments and farmers was facilitated, which
has resulted in farmers receiving poultry and
livestock vaccination and deworming services
from the Department of Livestock, thus
addressing one of their main concerns.
We have learned that the use of partnerships
as a strategy to address broader issues works
well, but is dependent on the presence of
stakeholders with the necessary expertise
working in the geographic area. For example,
in Malaita, some of the prioritized areas were
sanitation and health clinics. Although the
program does have links to relevant stakeholders
such as the Ministry of Health and NGOs, the
challenge remains that these agencies do not
have the funds nor the capacity to assist the
communities at this time. A similar challenge—
aligning the geographic focus of the program
with that of partners—emerged in the Barotse
oodplain. The local NGO that is anchoring the
community engagement process has had to
work outside its normal focal areas, which has
required more investment for the NGO to build
its presence in these locations and creates a
continued dependence on program funding to
support the broader development needs.
We are learning across cases that for an
agricultural research program to start with
a broad community vision and thus engage
with areas outside the agricultural eld,
partnerships and networking to support broad
development agendas should be in place or
must be developed alongside implementation
of research initiatives. This contributes to trust
and supports the holistic view of development
that communities pursue.
Systematic reection and documentation
enables adaptation
Community engagement as a process of PAR
includes a facilitated reection step during
which community participants take stock of
their learning and achievements and assess the
progress of their action plans. The reection step
allows the program to adapt the engagement
process. This is part of the system of monitoring
and evaluation for learning that functions across
scales of engagement—from community to
hub and across hubs. Implementation of this
part of the process was less scripted than the
initial community visioning and action planning
(which was informed by the community life
competence process), yet similar approaches
have been taken across hubs. All involved
systematic use of facilitated reection meetings
and documentation that captured outcomes
and lessons learned. The reection step is
proving to be a good vehicle for supporting
program adaptation.
For example, in the Malaita hub, AAS researchers,
along with partners and local resource people,
connect with the community through local
facilitators and community champions. After
every visit to communities, a trip report is
written and circulated to the whole program
team. This documentation process helped
identify that Alea was facing a challenge in
implementation of their action plans. Then,
during the community after-action review in
Alea, the hub team could better appreciate the
community dynamics that were challenging the
implementation of their plan. The joint reection
helped the community develop a strategy to
adapt the action planning and implementation
process. Community members conrmed that
they prefer coming together for learning. They
use a central demonstration site (for ongoing
work with development projects) for some of
their activities so that the surrounding villages
can join in but can also implement their own
actions in their own villages. Documentation
of the reection process provides input to the
strategy and is monitored on an ongoing basis.
In the Barotse oodplain, community
facilitators also reected on their performance
and challenges during the community after-
action reviews, which are implemented
every 6 months and include development of
new action plans informed by learning and
outcomes achieved. These reections have
helped the program and community facilitators
understand the challenge of an exponentially
increasing workload, as more activities are
underway relating to the research initiatives
and evolving requirements for documentation.
This understanding led to the development
of a strategy to invest in supporting emerging
champions and leaders who are already
facilitating and guiding various smaller PAR
groups. Reviewing documentation tools with
community facilitators in this way helps adapt
the system and builds capacity.
A similar strategy is used in Bangladesh to build
leadership capacity in emerging champions,
which helps to ground the engagement model
in communities. A system was designed that
emphasizes documenting the process during
the planning of each intervention and prior to
subsequent actions. This generates a running
record of what was done at each stage of
a development or change process, as well
as the outcomes associated with each step.
Program sta are responsible for running and
monitoring the process. This documentation
system has helped adapt the program design.
Also in Bangladesh, the documentation picked
up on an opportunity to support changes in
women’s access to land. At the beginning of the
program, some female farmers faced diculties,
as they did not have access to good land for
horticulture research. The model was adapted
to include household participation in the action
research, thus enabling women to use small
household plots.
In Cambodia, the facilitation teams used post-
session recording sheets and after-action
reviews to document their learning after every
facilitated event. Learning from the post-session
recording sheets has led to improvement in
how specic tools are used during sessions and
in the way the team facilitates the participation
of villagers overall. For instance, the facilitation
teams learned that community participation
was limited to a few groups in some of the
villages. A contributing factor to this narrow
participation was the village meeting style
of the rst sessions. In response, the team
adjusted the facilitation technique to start
with a SALT visit to households (Box 2) and
then implemented focus group discussions
across dierent wealth groups in the village.
Reections on how to engage both men and
women to improve the gender focus led to
a decision to have separate groups for most
In the Philippines, the initial strategy of
community immersion teams was implemented
in all communities. Through reection and
learning about what was or was not working,
the teams adjusted their strategy. For example,
in communities where the pressing issue was
enforcement of shery laws and the community
thought that it could easily be resolved by
employment of sh wardens, systematic
reection and documentation indicated that
a multistakeholder consultation workshop
was necessary before the employment of sh
wardens. In another case, abaca farmers reacted
to the recommendation of experts on how to
eradicate a virus that infested their plants (see
the inclusive science chapter for more details
on this case). The experts recommended the
removal of all potential host plants of the virus,
one of which is a crop that farmers presently
have on their farms and that provides good
income. Abaca farmers reacted strongly
against the recommendation, so the experts
conducted further research to conrm whether
the insect thriving in the replacment crop is the
host of the virus causing the disease. Another
strategy implemented was to conduct several
focus group discussions to fully understand
the perspectives of community members and
further enhance the credibility of experts to the
community. This approach put the community
members and the experts at ease with each
other and led to a satisfactory resolution and a
higher level of engagement.
Across the hubs, we have found that
implementation teams have embraced
reection processes and are systematically
facilitating reection on specic activities
and on the overall engagement process.
In all hubs, the engagement process has
evolved and is being rened as we learn
together with communities. Understanding
the nuances of how program implementation
takes shape in each location is only possible
if reection is happening across scales. The
cross-scale nature of the AAS monitoring
and evaluation for learning system builds
on these nested reection steps within the
PAR process. Challenges remain in designing,
testing, adapting and building capacity to use
documentation systems that support reection
and learning across scales. The documentation
challenge is not surprising given that use of
PAR requires co-ownership of processes and
learning and thus documentation should
support the learning of various stakeholders.
It therefore cannot be designed in advance
by researchers working alone, but requires
a meeting of researchers and stakeholders
to dene needs. Through our experience,
we are learning how to nd the right
balance of facilitation, ownership, trust and
documentation skills.
Shifting dependency mindsets and
managing expectations is an ongoing
The rates of poverty in all the hubs are high.
There are varying degrees of development
interventions across hubs (lowest in the Malaita
hub, highest in Bangladesh). In all the hubs
we have found a strong dependency mindset.
Taking a strength-based approach within
these contexts is challenging, as we cannot
immediately meet community expectations,
which are based on years of experience with
development projects that have provided aid
through a delivery approach and research
projects that have viewed them as passive
subjects. While communities appreciate the
use of visioning and action planning, it was
not in all cases a novel experience, and some
assumed that once the action plan was dened,
the project would deliver support and inputs.
Getting people to believe that they own their
action plan takes time.
Setting up research initiatives to build on and
support community action plans also takes
time, as these initiatives require stakeholder
engagement at hub level and at times further
scoping to identify researchable topics. During
the initial implementation phase, hub teams
felt pressure to deliver something tangible
to maintain momentum. How community
expectations and the resulting tensions have
been managed provide valuable lessons for
programs adopting a strength-based approach.
In the Malaita hub, the team developed and
signed community research agreements with
the communities. These played an important
role in managing people’s expectations, as
the agreement clearly spells out the areas of
involvement by both parties. This helped to
facilitate discussions on sensitive issues such
as payment of community members. The
ongoing management of expectations now
happens through the direct interactions of
the team with communities. Local resource
people, community facilitators and community
champions communicate important
information using local dialects and are honest
about what to expect.
In Bangladesh, an important strategy that
helped manage expectations was to ensure
that implementation of a few simple activities
started early. The PAR process started quickly
to address the issue of seed selection, with a
focus on homestead horticulture carried out
by women. With support from the research
team, farmers designed experiments to test the
productivity of ve okra varieties. The farmer-
researchers set up research plots, monitored,
recorded and analyzed the data, and shared their
results with the wider community. This research
was not cutting edge, but it focused on tangible
actions that helped build capacity and motivate
participants. The groups are now progressing
to more complex research that requires more
expert support, but they do so from a strong
base. The initial activity helped maintain
momentum while communities learned to
arrange for delivery of local expert support.
Dependency mindsets are most often
associated with the experience communities
have had with other programs that, in spite
of good intentions, treated them as subjects
and delivered solutions. Shifting this mindset
is an explicit objective of the RinD approach.
Emerging outcomes that have been evidenced
in hubs provide positive signals that some
communities have started on a journey towards
relying on their own strengths. In Bangladesh,
increased capacity to do research by farmers
participating in PAR activities is leading
to greater self-condence and increased
leadership by the poor. Similarly, in Malaita, the
program is documenting changes in attitudes
and behaviors of villagers who are now starting
to collectively address resource management
issues. In the Barotse hub, increased knowledge
is leading to more participation in collective
decision making. Similarly, in the Tonle Sap
hub, collective action has emerged that is in
part catalyzed by people engaging in processes
of learning and reection around their own
visions. These early outcomes suggest that
communities are moving along a pathway that
starts with shifts in their ability to organize and
tackle collective challenges, recognizing that
these changes are still fragile and limited to
those involved in the program.
Transforming ourselves is part of the process
From its inception, the program was cognizant
that engaging communities through PAR is not
a core skill of CGIAR. The program therefore
invested in capacity development. This started
with support from Constellation across all hubs
and has continued through varying support
strategies. The main modality for developing
the behaviors and skills required for PAR has
been through on-the-job training, coaching and
ongoing mentoring. As noted above, in all hubs
a culture of implementing after-action reviews
after every event has been institutionalized, and
teams have been using this process to reect
and learn—not just about the changes they see
occurring in the hubs, but also about their own
capacity to use a strength-based approach and
implement PAR.
All hub teams have embraced a new way of
working with communities and with each other.
The rst steps of community engagement
were taken with the specic guidance of the
community life competence process, which
required adaptation and contextualization.
As teams moved into the implementation
phase, they had less direct guidance and were
encouraged to design processes based on their
own experience with backstopping support. For
many, the lack of specic guidance and clearly
dened boundaries has been challenging.
Some have had little experience with a learning-
focused approach that starts with a broad
framework and requires contextualization.
Many were used to project implementation in
which the project has already decided what
it will focus on and comes in with a rigid plan
that is implemented according to a logical
framework and a schedule.
Teams also grappled with learning how to
let communities be in control of their own
development process. The Constellation SALT
mindset that was introduced to all teams (Box
2) is an example of how capacity development
has focused in part on shifting our own mindset
from a project-driven mentality to a strength-
based program approach. As implementation
teams, we appreciate that we also are in a
process of transformation.
We have identied facilitation skills as important
for ensuring quality in the community
engagement process, implementing PAR,
and understanding and using a gender-
transformative approach. Facilitation skills
include the capacity for active listening and
critical reection. To ensure a strength-based
approach, teams have found it important to
build team spirit to achieve a common vision
and understand our own role as bridges and
brokers rather than providers of solutions.
Identifying networking opportunities and
pursuing them is a skill that has enabled teams
to manage expectations and look across scales.
Rigorous and systematic documentation
and having a good plan for information
management and sharing are equally important.
Community ownership in
agricultural research
The program set out to implement an approach
to community engagement by embracing
community ownership as a principle of PAR
informed by practice and theory (e.g. McTaggart
1991; Stringer 2007). The use of PAR within RinD
has intentionally reached beyond a narrow
lens of farmers participating in research, or the
“research-driven farmer participation” model
as described by Okali et al. (1994). The program
took as a starting point a hypothesis that
achieving better and longer-lasting development
outcomes through agricultural research requires
that researchers engage with the development
process in earnest. Thus, the program is not
interested in participation simply as a means to
achieving eciency in the agricultural research
process (Sumberg and Okali 1997) but attempts
to understand how improved lives and livelihoods
of the poor and marginalized can be supported
through agricultural research. Researchers
use learning to reect back on the intent of
broadening the agricultural research agenda
through community engagement to see if there
is evidence that using PAR and starting from a
broad development-oriented agenda will enable
agricultural research to have greater impact.
From our learning, we can distinguish two ways
in which PAR builds community ownership. First,
at community level ownership means that the
research program is being led by community
members and is focused on achieving their
dreams, not those of the researchers and
facilitators. In practice, this happens through
the local facilitation teams and their capacity
to enable meaningful participation of
community members in planning activities to be
implemented, in the implementation itself and
in the learning that emerges.
Our experience indicates that starting with
a broad community vision of success is
instrumental in building community leadership
of the program, and we are nding that this
vision can begin to shift dependency mindsets.
Systematic facilitated reection that relates back
to the vision keeps this ownership alive and
the focus of the program on the community
development process, enabling research to
directly contribute to the change process. Local
anchoring of the learning process requires
ongoing reection by the program to adapt to
the evolving nature of community engagement.
Important steps include working in context
to identify appropriate partners, selecting
appropriate facilitators and building the support
systems required. Our experience suggests
that entering with a strength-based mindset is
critical to helping us adapt, learn and support
communities to believe in their own capacities.
These ndings provide evidence that the initial
intention to broaden agricultural research
through PAR is indeed supporting a journey of
shifting mindsets in communities and ourselves
and is paving the way for agricultural research to
unlock the potential that lies within.
Second, community ownership means
recognizing community members as co-
researchers in the joint research agendas that
emerge. Researchers must recognize that
community members co-own the identication
of research opportunities, design of
interventions, implementation and harvesting
of lessons. As we have illustrated with
examples throughout this chapter, community
members have been able to identify entry
points for research. In these cases, we are
identifying spaces for joint inquiry that build
on local motivation and specic development
challenges and feed into the overall research
interests of scientists working nationally and
globally. Early signs indicate some success
in supporting community voices to inform
research agendas, the most notable being
how the community PAR process has directly
informed the development of stakeholder-
owned interventions in Cambodia, such as the
development of pilots and case studies on land
and water management technologies.
The idea of communities co-owning joint
inquiry is not new to CGIAR, where farmers
have participated in the research process over
many years (see Becker 2000 for a review of
participatory research in CGIAR). A notable
experience is the work of the International
Center for Tropical Agriculture with local
agricultural research committees, known
by their Spanish acronym CIALs (Ashby et
al. 2000). Evaluations of years of work with
CIALs have shown that what started as a
methodology to build capacity of farmers to
implement formal research protocols together
with scientists led to an appreciation of the
processes of engagement. This led to adapting
the methodology to look beyond the eld
experiment and being cognizant of the broader
development process (Humphries et al. 2000),
as well as the value of the PAR process for
organizing and learning (Bentley et al. 2006) that
can lead to broader impacts. Our ndings build
on this appreciation for the broader process and
move towards a deeper sense of co-ownership
of the research process by communities.
Our reection on how we have been shifting
our own mindset to enable communities
to lead in research suggests that building
co-ownership is a process that needs to be
nurtured over time. The program has begun
to make headway in overcoming some of the
challenges through institutionalizing reection
processes and allowing for adaptive and exible
planning and implementation models. As
Becker (2000) suggests, historically some of
the challenges faced in using a participatory
approach within CGIAR have included a narrow
understanding of what science should do and
weak institutionalization. Our ndings provide
evidence that using PAR to guide community
engagement within the RinD approach is
enabling a broader view of how science can
contribute and has begun to institutionalize the
processes required to shift ourselves towards
a model of research that supports community
ownership. Thus the program is moving beyond
the historical challenges.
Perhaps the most signicant contribution of
these ndings is to illustrate that supporting
participation within a technical research agenda
alone is not sucient to build community
ownership. It is the link between the ongoing
PAR process of engagement to support a broad
community-owned and community-driven
development process and the participation
within specic interventions that agricultural
research can support that makes for a stronger
program that is more likely to build local
capacity to innovate and adapt.
We have explored and provided evidence from
the six areas of learning that help us appreciate
how community engagement can be successful
in building community ownership:
Starting with a broad community vision
is important for setting the tone of
engagement and helping the program
ground its systems approach.
Multiple contextualized strategies are
required to build trust and support a
deepening of engagement to achieve
authentic participation in the PAR process
and to nd researchable topics that become
the opening for specic technical agendas.
Given the broad starting point, programs
need to have contextualized strategies for
responding to critical development agendas
that fall outside their often limited scope of
expertise. Partnerships are critical for this
aspect of community engagement.
The iterative systematic reection of the PAR
process leads to learning and adaptation
of the implementation strategy to meet
local needs. Good documentation is a
major challenge in this process and one the
program is learning how to build capacity for.
An ongoing challenge in community
engagement is the expectations that come
with a dependency mindset. Creative
strategies have been developed for dierent
contexts to manage this tension.
The most important shift within the program
is the shift in our own mindset towards
a more strength-based approach. This,
we hope, is the beginning of a process of
transformation towards RinD.
If we reect upon the initial espoused theory
on how community engagement would be
implemented, we nd that the emphasis on
a strength-based model has indeed led to
a shift within the program team and within
communities. The transition from rollout to
implementation and from building a broad
agenda for community change to specic
agendas for agricultural research was not
as straightforward as the PAR cycles might
lead one to expect. The tension between
managing group consensus versus taking a
critical lens to understanding power dynamics
and building authentic participation across
all interest groups started to play out. This
tension does align, however, with a view of PAR
as a nonlinear process that grows organically
through relationships. While it was never
anticipated to be linear, to actually understand
this we had to build the skills of the program
teams around managing complexity. The
program initially underestimated what it takes
to build relationships and foster partnerships
to support broader research agendas and
community trust. On the other hand, the aim of
designing the program’s monitoring, learning
and adaptation mechanisms through PAR at
the community level seems to be working
as imagined. The institutionalization of after-
action reviews has been pivotal in building
learning into the program. The documentation
requirements of bringing a research lens to bear
on PAR may have been underestimated and
underplanned, and initial assumptions about
the capacity of development partners will need
to be revisited.
Chapter authors: Schwarz A,Cole SM,Downing
B,Perez ML,Kamp K,Crissman C,Johnstone
G,Paz-Ybarnegaray R,Aleem N,Apgar JM,Golam
F,Rodericks A,Longley C,Lunda Kalembwe J,
Mulele S,Muyaule C,Mulanda A,Mwasi D,Nasilele
F,Songe MM,Sumaya MAD,Wang J,Ward A
andWaters-Bayer A
Since AAS was rst designed, eective
partnerships have been central to the RinD
approach and essential to achieving the
program’s goals. The 2011 proposal outlined
a partnership strategy that drew on intensive
discussions with multiple partners during
proposal development and was built on three
core premises:
1. CGIAR is only one of many organizations
and networks engaged in working in
aquatic agricultural systems. Other research,
development and policy players together
spend many hundreds of millions of dollars
annually to improve the lives of people who
depend upon aquatic agricultural systems.
2. To add value in this complex institutional
environment, we need to identify where
and how the science insights we provide
can strengthen the focus and delivery of
other partners, and where the convening
and catalytic roles we play can foster more
eective coalitions of partners around our
3. Partners will devote the time and eort
required to work together only if the value
of doing so is clear. This requires that we
identify mutual needs and expectations.
It was anticipated that partners would
be engaged as core institutions, key
implementing partners and general
In 2012, the program began implementation
in three of its ve hubs (Figure 5): Bangladesh
(Southern Bangladesh Polder Zone), Zambia
(Barotse oodplain) and Solomon Islands
(Malaita Province). A rollout phase guided
by an AAS rollout handbook (CGIAR 2012)
rearmed the centrality of partnerships to
program success and required hub teams to
engage with new partners and consolidate
existing ones as early as possible. The handbook
identied opportunities and specic activities
for partner engagement and communication
throughout the rollout phase. In the planning
phase, methodological guidance focused on
communicating, using existing relationships,
lobbying and convincing; in the scoping phase,
the focus was on stakeholder consultation
workshops; and in the diagnosis phase,
the emphasis was on the engagement of
partners as members of design and diagnostic
teams and in the participatory workshop to
design hub strategic frameworks. In 2013,
program implementation was extended to the
Philippines (Visayas-Mindanao) and Cambodia
(Tonle Sap).
Rollout transitioned into community-
and stakeholder-led research design and
implementation in 2013 and 2014 (Figure 5).
This process, facilitated and supported by
AAS, resulted in the articulation of research
initiatives, theories of change and structured
research agendas to address hub development
challenges. At the same time, AAS sta in
focal hubs and across the programs science
themes gave high priority to identifying and
engaging more eectively with both science
and development partners.
A sharper focus on improving partner quality
and performance was supported by a draft
partnership framework, increased program
investment from the 2014 budget allocation
and additional funds from a results-based
management pilot program initiated by the
CGIAR Consortium (Downing et al. 2014).
The specic focus of this pilot was to increase
levels of support to a small number of existing
hub partnerships that had the potential
to accelerate current research design and
implementation, and to identify opportunities
to scale the learning from this work.
As described in the introductory chapter, in
January 2015, the annual cross-hub after-
action review brought together people from
across hubs and the global team at WorldFish
headquarters in Penang, Malaysia. Eective
partnerships was one of the common themes
identied from which useful learning was
emerging. Six case studies were identied during
the workshop, which have subsequently been
expanded upon by a group of partners and
AAS program sta. In this chapter we describe
the six case studies. For each, we reect on
the processes and the journey that led to the
emerging outcomes. We also synthesize lessons
from the case studies in context of current
literature around collective and knowledge
partnerships and oer guidance for nalizing the
AAS draft partnership framework.
Case studies
The AAS proposal notes that “global
partnerships are needed to leverage our
national and regional achievements and help
change development thinking and policy
globally” (AAS 2012a, 59). Global development
partners were expected to participate in
program implementation in the hubs, while
global research partners were expected to
develop collaborations on research themes.
The program outcomes were anticipated to
be achieved through three impact pathways.
One of these, pathway 3, seeks to use the
international public goods produced by the
program with the partners for “raising awareness
in the broader regional and global community”
(AAS 2012a, 27). This case study examines
the AAS experience with several research and
development partners who have the potential
to contribute to that global partnership
mandate: CARE, Prolinnova (a name based on
its mandate of “promoting local innovation in
ecologically oriented agriculture and natural
resource management”) and Constellation.
CARE’s interest in a partnership with AAS was
linked to the AAS program’s intent to have deep
impact in the lives of poor and marginalized
smallholder farmers and shers through
systems research and a commitment to gender-
transformative change within aquatic agricultural
systems. These goals aligned well with CARE’s
approach to gender equality and work
addressing the underlying causes of poverty and
marginalization. CARE is a key global partner
that participates in the program leadership team,
and in this role was able to contribute to and
learn from AAS conceptualization of gender-
transformative change and RinD in the early
stages of the program, as well as to contribute
lessons from various countries as AAS developed
its rollout process.
Later, attempts were made to deepen the
partnership within specic AAS hubs. In
Bangladesh, AAS sponsored the participation
of two sta members from CARE Bangladesh to
attend the Summer School on Gender organized
by the University of East Anglia in collaboration
with AAS. In Zambia and Cambodia, while
interest was high, CARE did not have the
capacity to expand to the same geographical
location as AAS. This means that within the
hubs, the relationship has not progressed to the
stage of including CARE activities in AAS core
communities but remains at the level of higher-
level inuencing and advocacy.
Figure 5. Timeline of development of documents and activities specically related to
partnerships during program rollout and implementation.
• Proposal
• Engagement
of core global
• Rollout
• Rollout in three
• Rollout in two
more hubs
• Implementation
in three hubs
• Results-based
pilot trial
• Draft revised
AAS partnership
• Reection on
New partnership
CARE continues to see value in this partnership,
including participating in meetings at the
CGIAR Consortium, explaining the value
proposition and potential of the partnership
between AAS and CARE, and collaborating on
signicant events such as International Rural
Womens Day and a scaling dialogue organized
by AAS in Rome in December 2014.
The partnership is not without its challenges.
In the absence of more visible work at the
hub level, it has been dicult to maintain
close collaboration and dialogue and to
convince some stakeholders of the value of
the partnership. Planning to scale impact is
beginning to address this challenge. Another
challenge relates to sta transitions. A lesson
learned here is the importance of investing time
and eort within both organizations to orient
new people to the partnership. CARE is hoping
to partner with AAS beyond research hubs and
to make links to other CARE programming.
As AAS reviewed the literature on what works in
RinD, it approached Prolinnova for their expertise
in community-led research and development.
Prolinnova has a long-established presence
in European discussions on the merits of the
approach and saw AAS as an ally. Their AAS-
contracted review of the impact of farmer-led
research supported by civil society organizations
revealed evidence of successes and highlighted
the opportunity for learning from existing
evidence and documentation. Together with
AAS, they are proposing follow-up research
to understand the mechanisms that lead to
enhanced capacity to innovate through farmer-
led research approaches. The results of the initial
research were presented in a European forum
and were published as an AAS working paper
(Wettasinha et al. 2014) and as a scholarly article
(Waters-Bayer et al. 2015), both of which have
been promoted through additional media and
other CGIAR research programs. A member
of Prolinnova joined the program’s strategic
leadership group (formerly the program
leadership team) and has become an active
ambassador for AAS and the RinD approach.
Prolinnova values the partnership with AAS
because it provides widely recognized evidence
to strengthen their case for promoting
farmer-led participatory research in ways that
strengthen capacity to innovate at the grassroots
level. The network, which spans Africa, Asia, Latin
America and Europe, hopes that through their
link with AAS, greater inuence can be exerted
at higher levels to create an enabling policy
and institutional environment for grassroots
innovation. Accordingly, the collaboration
between Prolinnova and AAS is now primarily
through impact pathway 3: to inuence the
global agricultural research and development
community. The partnership also allows the
Prolinnova network to gain deeper insights into
how community-driven agricultural research and
development can be supported more eectively.
Thus, the partnership provides a platform for
joint learning and advocacy for those within
the CGIAR system who are committed to a
transformative approach in research.
A core feature of the RinD approach is the use of
PAR to guide engagement with stakeholders at
hub and community level. This is our strength-
based engagement of women, men and youth
at the local level through a visioning and action-
planning process that identies opportunities for
agricultural research to support community goals.
From the outset, AAS and WorldFish recognized
that we did not have all the skills required to
implement this approach, so we approached the
Belgian-based international NGO Constellation.
Constellation shares the goal of supporting
community-driven change as a vehicle for
achieving development outcomes and has
over 10 years of experience in engaging with
communities. The partnership was established
through a memorandum of understanding
in 2012, with the main objective being to use
Constellation’s community life competence
process model for initiating and supporting
ongoing community engagement and ensuring
consistency in the approach across hubs. A core
principle of the partnership was the desire to
learn together about using a strength-based
approach to engaging communities in an
agricultural research program.
The partnership has been implemented over
3 years in all ve hubs (see the community
engagement chapter). It is based on
Constellation’s networked structure of
international and local coaches working
closely with the hub teams and hub partner
organizations responsible for implementing
community engagement. The Constellation
global point-person for the partnership has
engaged in joint planning with AAS program
leadership globally to adapt the community
life competence process model to the
implementation processes in hubs, as well as
participating in after-action reviews throughout
the rst 2 years. In 2014, a joint learning
paper was developed in which the following
partnership lessons emerged.
The partnership improved and was increasingly
successful over time due to systematic joint
planning activities across scales. This included
the involvement of Constellation coaches in
a number of hub activities during rollout that
helped them understand the complexity of
RinD and their role in supporting it.
The adaptation and use of the community life
competence process as a PAR process within
RinD required learning together how to support
a research process built on a strength-based
development approach.
After-action reviews enabled shared learning
and adaptation along the way. This is critical in
an emergent partnership that aims to address
collective goals within and across contexts. The
systematic use of after-action reviews enabled
the partnership to evolve and mature and even
inuence the memorandum of agreement
structure to ensure eective implementation
and joint learning.
A recent joint after-action review led to the
identication of challenges and tensions that
we have been able to manage through the
partnership, as well as the mutual understanding
that we needed to evolve the partnership
to a dierent modality. The hubs are now
implementing research initiatives, some of which
use PAR to work with communities and build on
the community life competence process work
led by Constellation. A more critical research
stance on community engagement as a process
within an RinD program is now required as the
overall research agenda around RinD evolves.
This leads us to shift the emphasis of the
community engagement process to evaluating
how the approach works in context. The current
work on community engagement therefore
moves beyond the expertise and interest of
Constellation. During 2014, writing up joint
learning was a challenge, which was recognized
as an indicator that Constellation was working
outside their area of core strength. In this new
phase of the program, the network of local
community facilitators that Constellation has
helped form becomes part of the AAS program
strategy for a dierent way of working with
Constellation recognizes that the joint work
with AAS has supported the growth of their
global movement of communities, facilitators
and coaches, stimulating community response
through the community life competence
process. They are pleased that this network
will grow and evolve beyond the extent of
the memorandum of agreement for program
implementation. While the relationship
between Constellation and AAS has
signicantly changed, the collective goal of
learning how to do things dierently and use
a strength-based approach to working with
communities is expected to enable us to nd
new ways of working in partnership.
Zambia: Barotse oodplain
Partner engagement activities started in the
Barotse hub in 2012. Organizations were
engaged primarily because of their expertise
in areas that would eectively contribute to
tackling the hub development challenge.
Each organization’s strengths were analyzed
to determine their main role during rollout,
and memorandums of understanding were
signed. This case study aims to highlight how
partnerships are transitioning from contractual
agreements to arrangements built on mutual
respect, joint planning, shared goals and honest
feedback. We argue that the transition is due
in part to critical reection processes that help
create spaces for partnerships to grow, develop
and gradually transform from contractual to
Case study trajectory (2012–2014)
This case study summarizes experiences of
partnerships with the Ministry of Agriculture
and Livestock, Catholic Relief Services (CRS),
Caritas-Mongu, and the Peoples’ Participation
Service (PPS). These organizations were all part
of the rollout and subsequent research and
development activities. Caritas-Mongu, the
Ministry of Agriculture and Livestock, and PPS
were involved in community engagement and
scoping studies in 2012. In 2013, the Ministry of
Agriculture and Livestock conducted mapping
and census exercises. The Department of
Fisheries, Caritas-Mongu and PPS all took part in
a sh value chain study. CRS and Caritas-Mongu
sta helped carry out an agro-biodiversity
assessment. Memorandums of agreement were
rst signed by partners before implementing
activities; however, it became clear that these
agreements could have been improved with
additional planning to ensure a common
understanding of objectives, an agreed-upon
process to develop integrated work plans, and
identication of the key people to be involved.
In early 2014, the planning process was adapted
to reect the learning from 2013.In April 2014,
an after-action review involving partners and
team members was convened by AAS. This was
a pivotal moment in the programs evolution.
Importantly, a collective accord was reached
that agreements and work plans would be
co-developed while acknowledging that this
joint planning would take time, eort and
coordination. This event was followed up with
another planning and reection meeting in
September 2014.
Activities and processes that inuenced the
case study trajectory
Certain activities and processes contributed
to shifts in the way AAS in Zambia is now
engaging in partnerships. Most notable are the
opportunities the program creates for sta and
partners to critically reect on and adjust ways
of working. For example, the April 2014 meeting
provided a safe space for partners to express
dissatisfaction with their lack of involvement
in developing agreements and work plans.
Discussions were held soon after the April
meeting between Caritas-Mongu, CRS and
WorldFish. All parties decided that a contractual
relationship was unsatisfactory, and that the
partnership arrangement should involve
greater interaction in planning, implementing
and reporting on AAS activities.
Another positive inuence on the quality
of partnerships was the additional support
provided by the results-based management
pilot. This catalyzed a deeper relationship with
Caritas-Mongu and the Ministry of Agriculture
and Livestock that resulted in the design of a
joint pilot project aimed at improving the ways
agriculture and nutrition extension services
are provided and knowledge is generated
and shared among partners, with the goal of
increasing productivity, improving nutrition
and contributing positively towards achieving
the Barotse hub development challenge.
Emerging outcomes related to partnerships
Over the course of 2014, the Barotse program
witnessed improved planning between sta
and partners, which in some cases led to
partners planning and using their own funds to
implement activities together based on shared
interests. This represented a shift from purely
contractual arrangements to partnerships that
were more collaborative. Joint planning has led
to better links between overall program goals
and research and development activities that
aim to address the needs of people in the 10
AAS focal communities. During the 2014 annual
stakeholder reection workshop, participants
highlighted that there is now greater
collaboration among partners. In addition, funds
to some partners were dispersed faster than in
years past. Nevertheless, not all memorandums
of agreement with partners were approved in
a timely manner by WorldFish. This was one
reason why some partners were unable to
complete all activities in their work plans.
A shift from contractual to collaborative
partnerships takes time to fully realize. For
example, during the annual reection workshop,
PPS expressed dissatisfaction about being
excluded from certain 2014 planning activities.
Providing honest feedback is a necessary rst
step towards better ways of communicating and
working together. The concern was revisited
in early 2015, and a strategy to improve the
partnership was agreed upon. Such examples of
partnership strengthening are becoming more
common in AAS in Barotse, enabling sta and
partners to improve relationships, build trust
and develop shared understandings through
learning by doing together as AAS evolves.
Bangladesh: Southern Bangladesh
Polder Zone
AAS recognizes that in Bangladesh a large
number of research and development
organizations are working to enhance
the well-being of people dependent on
aquatic agricultural systems. The program is
implementing RinD through partnerships in 16
communities and considers these partnerships
the key to successfully achieving outcomes and
impacts at scale. A number of the partners have
been involved in AAS since 2011 and have been
implementing the RinD approach to address
community-dened development goals. In this
case study, we describe some of the strategies
adopted for fostering multisectoral and cross-
disciplinary collaboration in the crowded
partner landscape of Bangladesh.
Case study trajectory (2011–2014)
Through the community life competence
visioning process conducted during rollout,
farmers, particularly women, expressed concern
about poor access to quality seeds for vegetable
production and subsequent household
consumption and sale. The challenge was
addressed through a hub productivity research
initiative as one part of a suite of activities to
tackle the hub development challenge. A cross-
disciplinary, multidimensional research support
team was convened. The research support team
was made up of scientists from research and
development organizations to provide science
support for community-led action research on
seed quality and productivity of homestead-
based horticulture crops.
Scientists from nine government and
nongovernment organizations made up
the research support team (Table 2), which
was led by a scientist from the Bangladesh
Agricultural Research Institute. A memorandum
of understanding outlined specic roles for
each member, and the program supported the
participation of individuals through payment
of a small honorarium. Subsequently, a second
research support team was formed under the
leadership of BRAC to support livestock fodder
research in 2013.
The primary responsibility of the research
support teams was to understand and analyze
the root cause of community-identied
problems (e.g. poor access to seed) and
design action research with people from the
communities to solve the challenge. Their
broader role was to build the AAS hub team
members’ research capacities and to help
identify the likely (best bet) technical options
for farmers and shers to base their research
eorts on. The research teams were to support
this through regular visits to farmer research
Year Partner Expertise Role
May 2013 Bangladesh Agricultural
Research Institute (BARI)*
Agricultural research
Research support
in community RinD
May 2013 Khulna University* Agricultural research
March 2013 Department of Agricultural
Agricultural extension
March 2013 Agricultural Training
Agricultural training
August 2013 BRAC* Research and development
August 2013 Department of Livestock
Livestock extension service
October 2014 International Maize and
Wheat Improvement Center
Agriculture eld crops
July 2014 Shushilan Action research program
community RinD
October 2014 Ashroy Foundation Gender Gender support
* Also involved in the subsequent research technical support system.
Table 2. Organizations involved in the research support teams and research technical support
system teams in Southern Bangladesh polder zone.
Within the rst 6 months of forming the rst
team, it became clear that members were not
able to deliver timely technical services to
communities located far from research stations
or universities where members were based.
During an after-action review attended by
research teams and AAS sta in 2013, strengths
and weaknesses of the model were identied
and potential solutions discussed. The model
was reimagined as a research and technical
support system, in which research support
team members continue to work in the same
way but an additional layer of supporting
partners is oered, including local extension
service provider oces in more remote districts.
The technical support system aims to link
farmers not only to new technologies and
scientic knowledge but also to local service
providers (public and private) who have a
presence in communities and are able to be
more responsive.
Activities and processes that inuenced the
case study trajectory
The development of the research technical
support system model was inuenced by a
number of converging factors. First, during
the community visioning process it became
clear that the interests of people in focal
communities were diverse and eorts to
address those interests needed to come from
multiple sources, not just the WorldFish sta
assigned to the communities. Second, to get
the science “right,” multiple research support
teams were needed, each with a specic area
of agriculture-related expertise. Third, the
theory of change developed by stakeholders
suggests that farmers and shers need to be
better connected to high-quality science, and
scientists need to engage more closely with
farmers and shers to ensure the science is
aligned with the challenges farmers and shers
face. Fourth, although the initial research
support team model began to create stronger
farmer-scientist relationships, research team
members had only a limited amount of time.
When the support team was expanded to a
system that included locally based research
extension and development actors with specic
expertise, a support mechanism was created
that was both science-based and locally
Through after-action reviews in 2013 and 2014,
partners identied some of the challenges,
emerging outcomes and ways forward. For
example, engagement between research
support team members was dicult given
their respective bureaucracies, and their overall
work requirements were demanding and often
conicted with those related to AAS. There
were few women researchers on the teams,
and capacity to conduct
was low, which
prevented some members from participating
in research activities. Also, contracts developed
with research support team members made
it dicult to engage with farmers and shers
outside a specic scope of work.
Emerging outcomes related to the case study
By 2014, individual research technical support
system members had come to appreciate the
value of RinD; however, it is less clear how much
that appreciation has become institutionalized.
Technical support system team members
now appreciate shers’ and farmers’ capacities
to innovate and are interested in using this
capacity to inuence their individual programs
of work. Capacities and condence levels of
farmers have been enhanced. For example,
farmers are regularly using science toolkits that
include simple measuring and weighing tools
to monitor the growth of their crops and are
better able to communicate with and access
expertise from scientists and other service
The technical support system has been
improved through an increased sense of
ownership and through formal agreements
with members’ respective organizations.
Improvements can still be made by being
more inclusive of multidisciplinary expertise
and private sector actors and by linking to
other platforms supported by AAS in the hub,
such as the knowledge sharing and learning
platform. The focus for the technical support
system moving forward is on strengthening
support systems that ensure farmers and shers
can access information, new technologies and
other services for continued adaptation and
Solomon Islands: Malaita hub
In 2011, WorldFish had been operating under
a memorandum of understanding with the
Solomon Islands government for more than
20 years and had collaborative relationships
with ministries responsible for sheries and
environment, as well as with most NGOs
working in the resource management sector.
Interactions outside this sector were incidental
(for example, if WorldFish was invited to
attend multisectoral workshops). Partnerships
within the sector also tended to be largely
transactional when funds were available
through WorldFish grants to contract locally
based NGOs or were limited to organizations
sharing information at partner workshops.
Joint planning was not a feature of these
partnerships except for specic donor projects
when explicit partners were named. When
AAS began to roll out in Malaita, a markedly
dierent approach to partnerships began to
Research capacity in agriculture and sheries is
generally low among organizations in Solomon
Islands, so there were two challenges. The
rst was to engage with partners outside the
traditional sheries sector, and the second was
to seek partners with the capacity to conduct
quality research to complement the sheries
research capacity of WorldFish in order to
address the hub development challenge. As
community priorities emerged during rollout,
and because implementing CGIAR Centers
(the International Water Management Institute
[IWMI] and Bioversity International)did not
work in Solomon Islands, it became clear that
research partnerships in the agricultural sector
would need to be identied and cultivated.
This case study reects on the evolution of
partnerships around one community priority
area of research, articulated as a research
initiative called “sustainable farming for and
nutrition and income, a cross-sectoral research
partnership with AVRDC – The World Vegetable
Center (AVRDC).
Figure 6. Trajectories (arrows), including aspirational trajectories (dotted lines), of core and key
implementing partners in Solomon Islands between 2011 and 2014.
Orange circles represent new key implementing partners not identied in 2011.
AVRDC = AVRDC – The World Vegetable Center
MPG = Malaita provincial government
KGA = Kastom Gaden Association
SPC = Secretariat for the Pacic Community
MAL = Ministry of Agriculture and Livestock
MFMR = Ministry of Fisheries and Marine Resources
MECDM = Ministry for Environment, Climate Change, Meteorology and Disaster Management
UQ = University of Queensland
* Alignment of purpose, trust and common results
Quality of partner*
Level of development
0 21 3 4 5
Case study trajectory from planning and
rollout to 2014
At a WorldFish science week in Penang in
July 2011, partnerships were identied to
implement AAS in Solomon Islands. At that
meeting, a version of Figure 6 was developed by
WorldFish sta to plot the status and trajectory
of some important relationships. Aspirations for
higher-quality relationships were identied for
most of the partners. These were particularly
ambitious for AVRDC and the Kastom Gaden
Association (KGA). The ministry responsible for
agriculture did not gure in our planning at
that time. Revisiting the diagram in 2014, the
KGA partnership had progressed somewhat,
but most notable is a markedly strengthened
(more collaborative and with more alignment
of purpose) relationship with AVRDC, plus the
inclusion of the Ministry of Agriculture and
Livestock and the University of Queensland.
Activities and processes that inuenced the
case study trajectory
In 2012, Solomon Islands AVRDC sta were
consulted during the scoping phase and
were participants in the rst stakeholder
consultation workshop, followed by the
design workshop where the hub development
challenge was validated by stakeholders. A
change in senior in-country sta at AVRDC
in 2013 stalled progress somewhat. Eorts
concentrated on building relationships at the
senior management level through one-on-
one meetings and communication between
Solomon Islands AAS program leadership and
the global theme leader responsible for AVRDC
projects in Solomon Islands, inuenced by other
one-on-one meetings with managers in Taiwan
and at WorldFish headquarters in Penang.
AVRDC has a common mission with the CGIAR
Centers. AVRDC has shown that vegetable
production is an integral part of livelihoods in
Solomon Islands. More than 90% of surveyed
households on Malaita and Guadalcanal engage
in vegetable production, which can contribute
on average more than 50% of total household
income. From the perspective of AVRDC, the
collaboration with WorldFish under AAS creates
opportunity for enhanced outcomes and
impacts for AVRDC’s projects funded by the
Australian Centre for International Agricultural
Research that, since 2007, have focused on
sustainable intensication of high-value
vegetables. The RinD and system concept is
new to AVRDC and so is considered a learning
In 2014, WorldFish invited local AVRDC sta to
contribute technical expertise to a publication
on food and nutrition among hub communities.
This joint publication (Jones et al. 2014) further
highlighted to both partners where our eorts
could be complementary. This led to AAS funds
being used to contract AVRDC national sta to
visit focal communities and scope opportunities
for supporting community action plans. Seed
funds were also provided to initiate some
farmer trials.
KGA and AVRDC local sta were involved in
developing the initial theory of change for the
sustainable farming and nutrition research
initiative for the Malaita hub. In late 2014,
as part of the results-based management
pilot, a participatory theory of change was
developed that included AVRDC, the University
of Queensland, the Ministry of Agriculture and
Livestock, and KGA. Developing joint theories
of change has been a powerful tool to develop
a coalition around the sustainable farming and
nutrition research initiative and to build capacity
in the hub team and partners to implement eld
trials as PAR.
Emerging outcomes related to partnerships
Through the relationship with AVRDC, AAS
has gained legitimacy with other agricultural
partners who play a larger role in extension and
networking than research organizations do.
Both the Ministry of Agriculture and Livestock
and KGA have committed to actions in a
participatory theory of change for the research
initiative. Hence, a collaborative coalition has
been built that acts as a bridge to a broader
network of agricultural partners. AVRDC is able
to take a leading role on the agricultural research
tasks with communities, while WorldFish
provides the bridge to the communities and
provides the opportunity for joint reection
and learning through after-action reviews at
the community, hub and national level. The
broader coalition has expressed commitment to
collective action around the hub development
challenge. In the coming years, AAS anticipates
supporting structured reection and learning to
strengthen the coalition’s capacity to address the
hub development challenge.
Philippines: Visayas-Mindanao hub
AAS implementation in the Visayas-Mindanao
hub started in February 2014 after a year of
regional and community-level consultations
with stakeholders in the ve regions that make
up the Visayas-Mindanao hub: Central and
Eastern Visayas, Zamboanga Peninsula, Northern
Mindanao, and Caraga. These culminated in
a design workshop and the development of
a hub-level strategic framework and theory
of change. To address limitations in resources
regarding the hub development challenge,
WorldFish pursued multilevel partnerships
to ensure successful implementation. The
regional consultations secured the buy-in of
collaborators and partners. This case study
focuses on the partnership approach used to
engage multiple partners to address the hub
development challenge.
Case study trajectory from planning to rollout
in 2014
As many local and international organizations
are working to improve the lives of people
living in aquatic agricultural systems in the
Philippines, the Visayas-Mindanao hub program
focused on where and how the program’s
science outputs could support the work of our
partners, and where the program’s convening
and catalyzing role could foster coalitions to
deliver more eective development outcomes.
In the early stages of program rollout, the
convening role was evident in partner network
analysis (Figure 7a), where WorldFish was
identied as being the central link for many of
the partners. After community engagement, the
network map took a very dierent form (Figure
7b). There were much stronger links across a
wider range of organizations, and WorldFish
was no longer the primary link between them.
We examine some of the processes that appear
to have inuenced this change.
Activities and processes that inuenced the
case study trajectory
Partnerships at the national level involved
engaging partners who have a mandate to
cover all regions in the country and whose
programs range from commodity-specic
to industry-based. These partners provide
funding for research that complements the
research initiatives of AAS. The Department of
Science and Technology’s PhilippineCouncil
for Agriculture, Aquatic and Natural Resources
Research and Development (PCAARRD) and
the Department of Agriculture’s Bureau of
Agricultural Research, in particular, have
been involved since program rollout in
activities, including regional and stakeholder
consultations and the design workshops. These
agencies are mandated to formulate policies,
plans and programs for science and technology
research and development in agriculture,
sheries, forestry and natural resource
management through a network of public and
private research institutions.
PCAARRD and Bureau of Agricultural Research
investments were instrumental in facilitating
scaling out and delivery of sustained support
for some main initiatives. A memorandum of
understanding was signed between WorldFish
and PCAARRD in 2012, followed by a specic
implementing agreement for AAS in 2013
that specied roles for each organization and
principles for collaboration. This resulted in
access to the industry science and technology
plans developed by PCAARRD and direct
investments in abaca rehabilitation and
aquaculture development in Southern Leyte.
The Bureau of Agricultural Research, on the
other hand, invested in AAS capacity-building
eorts (PAR, theories of change, scenario
building, etc.) for both the communities and
other local partners.
A number of partnerships are not only inuential
for implementation and research, but by being
embedded in local processes are anticipated
to improve sustainability. At the community
level, partnerships were pursued with local
state universities and colleges to provide more
sustained technical support. Through their
engagement in regional consultations and
design workshops, state universities and colleges
identied opportunities for linking with the
program through their graduate students and
ongoing research activities. These links proved
essential in securing funding for AAS initiatives
from PCAARRD in 2014, as the state universities
and colleges aligned research eorts with
community action plans.
Another important partnership for
sustainability was with local government units
in the communities where AAS works. Local
government units picked up the initiatives
Figure 7. Organizations engaged in AAS in the Visayas-Mindanao hub as mapped during
a partner analysis (a) during the early stages of the rollout and (b) after the
community engagement phase.
Icons represent organizations and stakeholders, while lines show who they are connected to through
information sharing, funding or activities. The larger the icon, the greater the number of direct connections.
In both diagrams WorldFish is represented by the large triangle.
Sogod_Baywide LGU_Sogod
PhilFIDA PhilFIDA_8 Radio_Tacloban
PARRFI LGU_Madridejos
for home gardening and made these a basis
for a city-wide program on organic vegetable
production and marketing. They were involved
especially at the barangay5 level in community
visioning and action planning and in identifying
the theory of change and hub development
Partnerships with other international NGOs
(CRS and Heifer International) also started at
rollout. For example, CRS recently contracted
WorldFish to assist in the rehabilitation of
typhoon-aected communities in Eastern
Visayas, allowing the scaling out of the program
in Eastern Samar and Leyte.
In 2014, the private sector (the Chamber of
Commerce and the Chamber of Handicrafts)
proved useful in helping partners consider
implications of technology development
from a business perspective. Their experience
of working within and shaping markets is
inuencing state universities and colleges,
government agencies, and other partners
to rethink their research agenda to ensure
adoption sustainability.
Emerging outcomes related to expanded
partner networks
Among the most signicant outcomes are the
Abaca Coalition. An alliance of agencies,
state universities and colleges, local
government units, and private sector
representatives have come together to
complement activities and co-invest in eorts
to rehabilitate abaca in Southern Leyte.
PCAARRD memorandum of understanding.
In addition to funding a study on sheries
demand and supply, PCAARRD invested
PHP 3.3 million to do comparative testing
of abaca hybrids and considers Southern
Leyte a testing area for abaca technologies
developed by the network.
Bureau of Agricultural Research
investments on capacity building. The
Bureau of Agricultural Research has so far
invested more than USD 300,000 on training
for PAR, theories of change and scenario
building, as well as community visioning
and action planning. Communities, along
with the local government units, are now
preparing projects for potential funding from
the Bureau of Agricultural Research.
Community Empowerment through
Science and Technology Program.
Regional oces (e.g. the Department of
Science and Technology) are now adopting
the RinD approach of community action
planning for the Community Empowerment
through Science and Technology Program
that will facilitate scaling out. With local
government unit investments, this will
ensure that community plans and visions for
development will be supported by science.
Cambodia: Tonle Sap hub
In this case study, we explore the experience
of engaging with partners during the initial 2
years of the program. We stress the importance
of capacity building in research methods and
design that can help develop collaborative
research capabilities and transform
development research from a contractual
arrangement led by scientists, external
institutions and programs into research that is
co-dened and co-managed by local partners,
stakeholders and communities to collectively
address complex development issues.
Engagement of partners with AAS began as
an informal sharing of information involving
partners who had a relationship or had worked
with WorldFish previously, and included
national government research institutions
and international development agencies. In
late 2012, as AAS was formally initiated in
Cambodia, a relevance assessment interview
instrument (McInnes and Johnstone 2012) was
developed to assess and evaluate the expertise
of local partners, their level of participation
and experience in the six AAS research themes
(productivity, markets, resilience, gender,
governance and knowledge sharing), and
their familiarity with and use of participatory
research methods. A total of 18 local partners
were selected and interviewed. Of these, eight
were identied as having relevant or potential
expertise and experience and were invited to
participate in AAS and the scoping of the Tonle
Sap hub in April 2013. The scoping provided an
initial mechanism for joint research discussions,
helped build mutual trust and resulted in the
rst draft of the hub development challenge.
Whose research agenda is it, anyway?
To undertake research in Cambodia, an
agreement with a government institute or
ministry is required. If the focus of the research
is people, then an arrangement is needed with a
government department to assist with approvals
at local government and community levels and
may also require their involvement to facilitate
various aspects of the research. WorldFish has
been formally recognized as a research institution
in Cambodia since 2003 and has a memorandum
of understanding with the Ministry of Foreign
Aairs and the Fisheries Administration.
The memorandum describes the principles,
conditions and timeframe of the partnership
and also provides WorldFish with the means to
carry out research with other organizations. This
is facilitated through separate agreements that
frame the research and expected outputs from a
partner. The use of memorandums of agreement
or understanding is a top-down process that
reinforces an externally dened and externally
led research agenda.
Under AAS, a more inclusive collaborative
research process in which researchers,
stakeholders and communities work in
partnership to exchange information, identify
and dene research issues, and frame these
as development challenges has been tested
in Cambodia. The process supports the
development of a partnership that co-manages
research that, with the appropriate levels of
investment in capacity building in research
methods and design, can decentralize the
research process and empower communities
to undertake collective action to address their
AAS Alliance development in 2014–2015
AAS has memorandums of agreement with 18
partners at local and national levels. Initially,
the main challenge was to bring the dierent
elements and expertise of the partners together
into a sustainable and focused arrangement
that could share and generate learning and
knowledge with the CGIAR Centers, other
partners and communities. The rollout process
in 2013 was structured to enable partners to
become engaged in the research development
process and included scoping, diagnosis,
community engagement and research design
phases, which collectively dened the AAS
research program.
In early 2014, 10 of the partners operating
in the hub formed the AAS Alliance (AASA),
whose goal was to generate research and
development knowledge, technologies,
institutional arrangements, methods and
insights and to share with partners and
communities across villages and the hub. The
stimulus for AASA was the various eorts made
by the program to build research capacity
in PAR as well as more traditional methods,
such as key informant interviews and focus
group discussions. The partners recognized
that sharing knowledge across the hub made
it more eective to operate collectively. The
process was documented through community
and stakeholder reections and reviews that
generated PAR guidelines (Nurick and Apgar
2014), case studies (Jore and de Silva 2015)
and outcome evidencing. Capacity building
was not limited to design and data collection
but included analysis and interpretation, much
of which was carried out in the community
with partners and community facilitators and
resulted in cross-village analysis.
By mid-2014, despite the existence of individual
partner contractual agreements with WorldFish,
the research agenda and dialogue was being
discussed collectively and directly with AASA.
The AASA included eight organizations
operating in the focal villages and two
national partners, Gender and Development
in Cambodia and a national research NGO,
the Analyzing Development Issues Centre
(ADIC). ADIC was contracted to coordinate
the documentation and communication of
research data between partners and also
operated as a facilitator between the CGIAR
Centers (WorldFish, IWMI and Bioversity) and
AASA partners. By the end of 2014, the idea of
formally recognizing AASA as a collaborative
research body was realized through additional
funding from the CGIAR results-based
management pilot that aimed to build upon the
existing AASA initiatives to develop a backbone
organization.6 AASA was provided with funds
and technical support to employ a partnership
ocer to coordinate partner engagement, as
well as funds to undertake research processes
and produce materials to share and generate
knowledge and evidence of impacts from the
By early 2015, AASA had developed a joint
proposal for funding submitted to the Water,
Land and Ecosystems Research Program, and
had developed a 3-year collective vision for the
organization with the goal “to work together
and share best practices at local, national
and international levels toward improving
livelihoods and welfare of the people, especially
the very poor, poor and vulnerable, and
improve natural resources management in the
Tonle Sap region. AASA has identied three
objectives to realize this goal:
Encourage men and women at all levels to
analyze and explore solutions in relation
to agricultural and sh production and
processing, and establish and maintain
market networks.
Build capacity for climate change resilience
to improve agricultural productivity and
manage water resources in an equitable way
and in collaboration with local authorities.
Improve the research and development
capacity of AASA and links to research and
development networks.
Building partner capacity is the key to
collaborative research
Building research capacity in AASA partners
has been instrumental in formulating and
transferring research that is not principally
dened externally, but is instead a program of
research that is co-managed by local partners
and communities (Pomeroy and Berkes 1997).
Figure 8 conceptualizes the role and potential
of researchers, partners and communities in
using collaborative research where the subjects
of the research are people and the natural
resources they rely on. One extreme of the
spectrum is where scientists, who are external
to the location and whose lives will not be
personally aected by its outcomes, have full
control over research, knowledge and learning.
The other extreme represents community-
controlled research, in which knowledge is
generated from within the community through
self-reection and used primarily to address
development issues faced by people in a
specic area. The AASA partnership, together
with researchers and communities, represents
a middle ground of collaborative research that
draws upon the strengths of both approaches,
where research institutions can co-manage
research and learning with local partners and
communities and provide technical inputs to
address development challenges that have
been co-dened with communities.
Figure 8. Spectrum of collaborative research. Source: Pomeroy and Berkes (1997).
Researchers inform
and consult
Researchers identify
and dene issues
Researchers, community
and partner cooperation,
communication and
information exchange
Community and partners
provide advisory role, joint
research action and partnership
Researcher under
community control
Cooperation between
communities to exchange
knowledge and learning
Spectrum of collaborative research
Co-managed research
Externally dened and
managed research
Community-dened and
-managed research
Community self-
reective research
Geographic location (Tonle Sap hub)
The case studies represent six partnership
journeys and emphasize dierent areas of
learning. The global case study illustrates the
experience of nurturing partners as part of
the RinD approach across the program, while
hub cases reect dierent starting places in
terms of capacity, relationships and partnership
contexts within the program sta and partner
organizations. Learning has been distilled into
three areas that were initially identied in the
draft partnership framework: learning about
conditions required to convene partners,
learning about how we sustain partners, and
learning about how we strengthen capacities
for leadership and to foster change (Table 3).
Conditions required to convene partners
It takes time and commitment to identify
the right partners who can understand and
connect with the hub development challenge
and create a shared vision for change through
agreed-upon action. This, combined with the
need for certain partners only at critical stages
of the program’s implementation, creates what
has been referred to elsewhere as a partner
continuum (Horton et al. 2009). In recognition
of this, AAS hub teams aim to create additional
spaces to convene new partners who bring
fresh insights, new understandings and capacity
to tackle the hub development challenge.
The hub cases, in particular, suggest that
more collaborative (rather than transactional)
partnerships are beginning to occur, in part
because processes have been set up and
supported by the program for sta and
partners to critically reect, share learning and
experiences, and use these to adapt their action
plans. The collaborative partnerships that have
evolved require trust and take time, eort and
coordination to mold and maintain. Without
sustained eort and commitment on the part of
all, there is the risk of lapsing back into old ways
of working. Eective collaborative partnerships
need to go beyond simple consultation.
Inclusive and participatory processes have
proven to be crucial.
Successful partnerships have been cultivated
through eorts to search for organizations
that have common goals and agendas,
enabling the development of a shared vision
for addressing hub development challenges.
Recognizing members of alliances as experts
on local conditions and priorities helps create a
better working partnership, as it considers the
interests of aected parties, fosters informed
debate, and exposes the costs, benets and
appropriateness of any planned programs.
Dialogue that includes the partners’ ideas and
priorities helps shape research projects and
leverage funding in support of community
action plans. Some partnerships move from
transactional to more collaborative ways of
working with little eort, while others require
more eort to ensure a successful transition.
Mutual understanding can take years to emerge
and require—at a minimum—the sharing of a
common purpose or goal (ADB 2011). Global
partner CARE participated in hub rollout
activities in Cambodia, but this did not result in
close collaboration in the hub despite ongoing
interest. A mutual understanding has been
reached that the most eective part for CARE
to play may result in a role that is not directly in
the hubs, but at a larger scale. Similarly, in the
Solomon Islands case, having implementing
partners involved from the scoping stage did
not initially seem to bear fruit, but through
ongoing participatory processes, alignment
of purpose was eventually established as a
foundation on which the AVRDC and AAS
partnership is now building. Program funds are
sometimes required to enable partners to move
outside the geographical range of their projects.
Learning Global Zambia Solomon
Bangladesh Cambodia Philippines
Conditions required to
convene partners X X X X X
How we sustain partners
How we strengthen
capacities for leadership
and foster change
Table 3. Where the three areas of learning were emphasized across case studies.
We now have enough experience to begin
to comment on the enabling conditions in a
partnership framework, which will be explored
further in subsequent papers. These conditions
include the following:
Establishing systems for shared
measurement and reection. Facilitating
data collection and measurement of results
consistently across all participants ensures that
eorts remain aligned, processes are equitable,
and participants hold each other accountable
through shared analysis and reection.
Mutually reinforcing activities. Partner
activities are dierentiated, yet are still
coordinated through a mutually reinforcing
plan of action. This approach will help
leverage the best capabilities of each partner.
Eective communication and learning.
Consistent and open communication occurs in
ways that build trust and mutually benecial
relations among all, supporting ongoing
learning that is potentially transformative.
Learning to sustain partnerships
It is the individuals within institutions who
undertake partnership activities (ADB 2011),
and the case studies have emphasized that
one-on-one relationship building is essential
for sustaining partnerships. Mutual trust in a
collaborative research partnership has to be
nurtured and developed over time and requires
a commitment of technical and nancial
resources by external research institutions and
programs. A further challenge to sustaining
partnerships relates to sta transitions, and this
highlights the importance of investing time and
eort to orient newcomers to the partnership.
Relationships should be institutionalized with
co-investments for shared action and advocacy.
Building on past gains while breaking new
ground, such as through joint publications, is
sometimes a useful way to build a common
vision for research. It may also be necessary to
modify the team makeup to include technical
service providers to local communities.
Learning to strengthen capacities for
leadership and foster change
As partnerships and coalitions began to mature,
lessons about deepening those partnerships
began to emerge. One lesson is the critical
importance of having strong leadership across
AAS: within communities, partners and the three
managing centers to model new behaviors,
embrace emergent thinking, and be successful
in convening and sustaining partnerships. This
reects learning (Kania and Kramer 2011) that a
backbone support organization that has the time
to perform functions such as facilitation, data
collection and reporting is one of the necessary
conditions for collective success. For global
partnerships, this needs to be multilayered (local
to global) to ensure co-ordination happens at
multiple levels (Patscheke et al. 2014).
In the case studies, AAS lead centers within
hubs either initially (e.g. Philippines) played
that supporting role or in some cases (Solomon
Islands) still do. With strong leadership there is
greater potential to delegate responsibility for
the research process to other partners. Capacity
building in science methods for AAS hub teams
and partner organizations was identied as
critical to the development of collaborative
research, and this capacity has to be incorporated
into the resourcing of the partnership from the
beginning of the research design.
These reections on what is being learned
across scales are part of the journey toward
sustained, equitable partnerships and coalitions
that deliver increased benets for the poor
and marginalized in program countries. A core
principle is the idea that interventions in a
complex system without a xed agenda can
be a powerful lever for change. The dialogue
and action space created in hubs was “safe”
precisely because our starting point was not
a xed intervention agenda. The “safe space
is one where a diversity of actors operating
in one geographical area can openly and
critically explore and eventually adjust their
interventions in the system. This also oers an
opportunity for evaluating the processes that
are emerging for generation and exchange of
knowledge (ADB 2011) within partnerships and
coalitions. Most importantly, neither partners
nor partnerships are static. The RinD approach
oers a practical and increasingly proven
methodology for engaging stakeholders in
dynamic complex systems.
Chapter authors: McDougall C, Cole SM,
Rajaratnam S, Brown J, Choudhury A, Kato-Wallace
J, Manlosa A, Meng K, Muyaule C, Schwarz A and
Teioli H
Agriculture research has made notable scientic
and productivity contributions over the past
decades (Alston 2010), yet the sustainability
and equity of its impacts have been questioned
in relation to its ability to benet women, the
poorest of the poor and socioeconomically
marginalized groups (Meinzen-Dick et al. 2003).
AAS has sought to address these limitations by
combining three streams of research-related
processes around an agreed-upon set of
development challenges:
Contextually relevant social and
biophysical agricultural research drawing
on participatory and other methods. This
is technical” AAS research and includes
research on aquaculture productivity, sh
value chains, oodplain management,
ecosystem services, and community-based
land and water governance.
Ongoing engagement of communities and
other actors in social learning processes
related to their jointly identied
development challenges, including
that seeks to nurture innovative capacity.
These are referred to as core RinD processes.
In relation to both of the above,
engagement of diverse local actors in
transformative reection and change
processes regarding underlying forces
and factors that shape equality and equity,
such as gender and social norms, attitudes,
practices and rules. This is the gender-
transformative approach.
The RinD approach aspires to develop scientic
insights and technologies, to combine
knowledge generation with enhancing the
innovative capacity of local actors, and to
increase the equity of the social, economic
and political structures that inuence
the livelihoods of poor and marginalized
households who depend on aquatic agricultural
systems (Kantor 2013). In going beyond more
common gender-mainstreaming aspirations
such as increasing women’s participation in and
access to technologies, the gender and socially
transformative aspect of RinD is expected to
contribute to a stronger foundation for more
equitable and lasting contributions of research
to development processes. We refer to this
socially transformative, equity-oriented element
of RinD as the gender-transformative approach
(see Kantor 2013; Kantor et al. 2015).
The AAS journey towards implementing a
gender-transformative approach has been a
learning-based process, combining conceptual
grounding, drawing on learning from others,
and experiential learning among research teams.
Given the newness of gender-transformative
research in the eld of agricultural research
and to the teams, the journey has involved
teams encountering and addressing multiple
challenges. With the understanding that other
programs or teams may face similar challenges
in the pursuit of gender-transformative research,
the goals of this chapter are to (i) highlight some
of the key challenges and learning regarding
how these can be eectively addressed and
(ii) share identied strategies for gender-
transformative research and examples of such
research that is in progress in AAS.
The reections and insights presented in this
chapter were generated through a two-stage
process: (i) identication of challenges and
related learning generated in a cross-hub
after-action review held in January 2015,
involving representatives from each of the ve
hubs, and (ii) drawing on and synthesizing
across new and existing written contributions
by hub team members regarding challenges,
learning and emergent examples of the
gender-transformative approach. The result is
a snapshot of the challenges faced in the start-
up phase of gender-transformative research,
highlights of learning about how to overcome
these, and a sketch of current gender-
transformative strategies and examples from
AAS research to date.
What is a gender-transformative
A gender-transformative approach to research
is an approach that “can be applied within
research to examine, question and, most
fundamentally, enable changes in inequitable
gender norms, attitudes, behaviors and
practices and the related imbalances of power
(IGWG 2010). Through encouraging critical
awareness among men and women of social
inequality and practices, [gender-transformative
approaches] help people challenge and
re-shape distribution of and control over
resources, allocation of duties between men
and women, and access to and inuence in
decision making (Caro 2009). They also enable
men and boys to question the eects of
harmful masculinity, not only on women, but
also on men themselves” (Meng 2015, 1). In
other words, a gender-transformative approach
seeks to generate understanding regarding
gender and the visible manifestations of gender
inequalities and inequities7 (such as gendered
roles and relations and their outcomes), and
to catalyze shifts in the norms, attitudes, and
formal and informal rules that underpin these
visible manifestations of inequality.
The need for a gender-transformative approach
emerges from the gap between research
(and development) practice and the eld of
gender’s conceptual development. In particular,
it emerges from the predominant focus of
gender eorts in research and development
on interventions that address individualized
demonstrations of gender inequality—gender
resource gaps—but ignores their wider social
causes” (AAS 2012b, 3). While this recognition
is relatively new in the eld of agricultural and
development research, it has been recognized,
and progress has been made in gender-
transformative approaches in other elds over
the past decades, most notably the eld of
A gender-transformative approach diers
from more commonly applied gender
mainstreaming approaches in agricultural and
development research in terms of the framing
of issues to be addressed (Cole et al. 2014a).
Gender mainstreaming focuses on addressing
visible manifestations of a gender gap, such as
women’s limited access to training or resources.
A gender-transformative approach adds a level
at which the central problem is framed. It does
so by adding a focus on the formal and informal
institutions underlying the visible gender or
social gaps — in particular, on (gendered) social
norms, attitudes, practices, processes, and rules
or policies. The reason for this focus is that it is
at this level that gender and social inequality is
produced and reproduced (Kabeer 1994).
This dierence in framing translates into a
dierence in goals as well. The goal of gender-
transformative research involves addressing
not only gender inclusion or more eective
technical innovations, but also catalyzing the
potential for shifts in any underlying informal
and formal institutions that inhibit equality.
The aim is to engage with and inuence these
institutions at multiple scales (from households
to communities to larger scales). As such, a
gender-transformative approach seeks to
engage women and men in research as a
social change process. Transformation towards
equity expands the range of aspirations,
options and opportunities available to
individuals, households and communities, as
well as increasing the agency of previously
marginalized actors, and thus their ability to
eectively act on their own potential (Cole et al.
A gender-transformative approach
operates by creating space for and sparking
increased critical questioning and awareness
(consciousness raising) of underlying attitudes
about rights, roles, capacities and values and
how these forces inuence individuals, families
and communities in relation to their livelihoods,
other aspirations and well-being. Bringing to
the surface the generally unquestioned norms
and practices and their inuence or costs for
individuals, families and communities can
spark cognitive shifts (McDougall and Ojha
forthcoming) towards more equity-enhancing
mindsets. These shifts in perceptions and
thinking can lead to more equitable roles,
relationships and practices between women
and men, and ultimately more equitable
development outcomes (Salazar 2014). Box 3
outlines the main characteristics of a gender-
transformative approach.
Looking back: Challenges in
developing and getting started
Conceptual challenges: Understanding
a gender-transformative approach and
moving from concept to application
All hub teams identied a fundamental
challenge from experience to date: the
conceptual complexity of a gender-
transformative approach, and diculty
in translating it from a broad conceptual
understanding to eective practice.
Compounding this challenge, the pioneering
nature of the concept within the eld of
agricultural research means that there were few
concrete examples on which the teams could
draw to ground their understanding.
Some team members suggested that the focus
on transformation is intimidating in that it
suggests new, profound or socially destabilizing
strategies. Similarly, some said that the fact
that the concept is almost always referred to
by its acronym (GTA) obfuscates the meaning
and reduces accessibility to the concept, even
for research team members who themselves
use the acronym. Taken together, these issues
reinforced diculties in translation into practice.
Reection on these challenges motivated
teams to identify an underlying and previously
unaddressed need for a more denitive, shared
and implementable understanding across the
program about what the gender-transformative
approach is in practice. Three questions and
their related challenges emerged:
What does the approach mean and what are
its goals in each hub? (the framing challenge)
How and through what strategies, when,
with whom, and by whom should teams and
partners apply this approach? (the capacity
How can it be integrated with the overall
hub research program involving core RinD
processes, such as
and community-level
visioning and reection, as well as technical
research initiatives around productivity,
governance, oodplain management and so
forth? (the organizational challenge)
The framing challenge: Gender-only focus
Hub teams identied as a challenge the
emphasis of the approach in practice around
gender as women and men (i.e. rather than
starting by engaging with broader issues of
social equity and equality early in RinD and
then extending this to gender in combination
with other socially constructed roles, relations,
values, and meanings and categories of social
dierence, such as wealth, ethnicity and caste,
class, and age). The Philippines hub team,
for example, found it challenging to foster a
collective sense of interest in and ownership
of a gender-transformative approach as an
element of RinD. Teams there were focused on
issues relating to men and women; subsequent
reection surfaced that gendered norms
were perceived by hub stakeholders to be
less generally signicant than power relations
and inequalities in opportunity structures
among other social groups in that context.
An overly narrow de facto framing of the
approach around gender (as a binary women-
men construct) may thus have operated as a
constraint on the approach’s development and
Box 3. Characteristics of a gender-
transformative approach
A gender-transformative approach
seeks to understand people within
their context, including in terms of how
culture, age and other aspects of social-
economic identity and other exogenous
factors and livelihood strategies (such as
remittances) aect and are aected by
makes explicit how social inequalities
intersect to aect their choices and
provides space for women and men to
engage in an iterative process of critical
learning, reection, questioning and
engages both women and men, as
transformative change stems from a
shared vision;
engages with dierent actors across
scales to redress the underlying norms
and power relations that enable social
Adapted from Kantor (2013).
The capacity challenge: From conceptual
understanding to practical application
All teams observed that shifting from a gender-
transformative approach as a concept to the
approach as an applied strategy required
building capacity among research sta and
partners, and this involved multiple challenging
factors. For example, the Zambia team said
that at the outset of the program the lack of
capacity to integrate the approach spanned
multiple levels and spheres: from research sta
designing activities, to partners integrating
gender-transformative approaches in activities,
to community facilitators facilitating PAR
processes. Specically, they noted that while
formal training appeared to be reasonably
eective in developing sta and partner
capacity, it was less eective at the community
facilitator level. For example, the team’s own
evaluation of the eectiveness of community
facilitator training indicated that after the
training of trainers, only 6 of 22 community
facilitators felt condent about their capacity to
facilitate gender-transformative sessions.
A second factor observed by all hub teams
relates to the point regarding conceptual
challenges. While building capacity in the sense
of knowledge and understanding is relatively
achievable through formal training, gender and
gender-transformative training may still leave a
signicant capacity gap in terms of teams being
able to make an eective shift from concept to
practice. Gender transformation training was
useful to lay a foundation of knowledge, but
only where the training linked directly to hub-
specic issues and research plans in progress
was the team able to make the shift to eective
practice in a relatively short time.
The organizational challenge: Working in
All hub teams noted organizational and
institutional challenges to operationalizing a
gender-transformative approach. One aspect of
this was that the gender-transformative work
was organized within the research program in
parallel to, rather than directly in connection
with, the PAR and community engagement
processes. For example, in Zambia, for most
of 2013–2014, the gender-transformative
initiatives and PAR activities were being
conceptualized, planned and implemented
separately. The lack of joint planning delayed
research team and partner understanding
of what the various initiatives and activities
were doing and slowed learning about how to
integrate gender transformation within these.
One hub researcher observed that it was as if
gender transformation and PAR were trying to
bypass one another.
These reections, along with the realization that
this separation was less than optimally eective,
laid the groundwork for more integrated
planning and strategies. This challenge was also
reected in organizational structuring in the
hubs and in the global program, which involved
generally separate gender and PAR sta and
partners. Teams said that while the PAR sta
worked in the hubs, the gender sta was in many
cases embodied, at least initially, in a single
gender research analyst. The gender analysts
being (generally) relatively junior and working
on their own reinforced the conceptual and
capacity challenges outlined above, including
overall diculties of communication, integration
of gender into hub programs of work, and
translation of the gender-transformative
approach into practical strategies.
Looking ahead: Meeting challenges,
making progress
While these challenges posed considerable
diculties, the teams persisted in seeking
ways forward with the gender-transformative
approach. These eorts, illustrated with
examples from Bangladesh, Zambia and
Solomon Islands, have led to a number of
Conceptual clarity and identifying principles
and strategies for action
The importance of demystifying the concept
within the research teams and among partners
has emerged as a fundamental lesson. Implicit
in this is encouraging and enabling the
understanding that a gender-transformative
approach need not be complex in terms of
strategies or separate from existing community
and multiscale engagement. In line with this
learning, teams are now aiming to translate
the big ideas of the gender-transformative
approach into action through a range of
practical strategies. The development of
strategies has come through a combination of
literature reviews, partnerships and capacity-
building processes. As outlined below, the
teams anticipate that taking a reective
learning-by-doing approach to these strategies,
treating them as learning opportunities, and
being ready to adjust as they progress will be
central to their success.
A second point of learning relates to
understanding gender-transformative research
as one type of gender research among several.
Central to this understanding is that not all
research activities in the program need to
be gender-transformative in nature. Rather,
Box 4. Rough typology of gender in agricultural research
Gender-integrated research (or descriptive gender and social analysis): Scientic quality relies
on research addressing gender and social dierence in terms of data being eectively and
accurately disaggregated, as appropriate to the context. Eectively assessing and analyzing
contexts and research results through a gender lens contributes to laying the foundation for
future gender-transformative work by increasing the collective understanding of the context
and the needs, opportunities and entry points for social change.
Strategic gender research distills widely applicable learning regarding gender, including
research in which women are the primary subject of the research. This could include, for
example, gendered dimensions of community access to decision making and benet sharing
in community-based sheries and natural resource management, and in particular how
governance can increase the ow of benets to women. Gender-strategic research helps enable
achievement of development outcomes at scale by understanding the gendered aspects of
technical, agricultural and governance learning available for use by a range of development
actors, including governments, bilateral agencies and civil society actors.8
Gender-transformative research is research that leverages the research process itself to directly
catalyze and contribute to gender-equitable shifts in the formal and informal rules, norms and
behaviors that underpin gender inequality in processes, practices and outcomes. Building
on the foundation built in AAS and the CGIAR Research Program on Livestock and Fish (such
as Cole et al. 2014a), this body of research aims to contribute to achieving the gender cross-
cutting sub-intermediate development outcomes of gender-equitable control of productive
assets and resources, reduced time burdens, and improved capacity of women and young
people to participate in decision making. For example, this could include research to catalyze
shifts in norms and rules addressing gender-equitable access to and control over key nancial
and productive assets. This process could involve development, application and assessment of
strategies within or in connection to the research process to spark critical questioning by men
and women regarding gendered rules, norms and behaviors. Questions could relate to how
gender-inequitable access to and control over key sh agri-food system assets and resources
(including aquaculture technologies and training, and nancial and other assets) inuences the
achievement of household and community aspirations; what factors shape access and control;
and how these factors can be addressed to create more equitable access and control—and in
connection to these—greater and more equitable achievement of local aspirations.
Source: Adapted from WorldFish et al. (2015).
gender in research can be seen as ranging
across a spectrum from all research involving
basic descriptive gender and social diversity
analysis (gender integrated), to strategic gender
research, to research involving a gender-
transformative approach (Box 4). Across all of
these, good practice for scientic quality and
ethics indicates that research activities need
to be designed and applied such that research
processes are accessible to and eectively and
equitably engage with or draw on a balance of
actors (i.e. be gender inclusive).
Re-framing gender in complex systems
Reection on experience, reinforced by the
literature (such as Resurreccion and Elmhirst
2008), underscores that the implementation
of a gender-transformative approach benets
from a broader, more nuanced and integrated
framing and implementation. This can be
seen as comprising a re-grounding of the
analysis and transformation in terms of both (i)
addressing multiple forms of socioeconomic
power and marginalization, which implies
a focus on poor or minority socioeconomic
groups (of both sexes), and (ii) engaging with
the multidimensional nature of gender.
In theory, the approach recognized the above
points from the program outset. Learning
from experience, combined with progress in
analysis of ndings, has further underscored
their signicance and the need to translate
these eectively into practice. Working with
multiple socioeconomic groups, enabling
reection, and undertaking analysis of the
gendered norms, practices, attitudes and
power relations that often disproportionately
impact women enables both more eective
and more inclusive social analysis for change
than does a narrow framing. Taking this forward
involves the research engaging with gender as
dynamic and context specic, and recognizing
that neither women nor men represent a
homogeneous category. Each gender category
is now being increasingly recognized as cross-
cut by, shaping, and in turn re-shaping multiple
other dimensions of social dierence, such
as wealth, ethnicity or caste, religion, and
relation to place. Additionally, this nuancing
is being connected with systems thinking and
political ecology perspectives (Resurreccion
and Elmhirst 2008; Locke et al. 2014) in terms of
understanding gender as involving a dynamic
and complex interplay of these categories
of dierence, running across multiple scales.
This understanding will avoid interpretation
of gendered roles or relations as static or pre-
determined, and instead re-emphasize the
mutual creation and re-creation at play among
socio-political, ecological and economic factors.
This more nuanced, integrated framing also
brings to the surface the political nature
of gender-transformative work. Given that
natural resource and development contexts
are inherently political and embody ongoing
contestation, conict, alliances and re-shaping
of power dynamics, both innovation and change
processes are likely to be unpredictable and
potentially conict-laden. As such, gender-
transformative research processes (as a form
of socio-technical change) should not be
anticipated to be predictable or smooth. Rather,
they will reect the complex and contested
nature of the context and the systems in which
they are embedded, and will need to address
conict management from an early stage.
Some external research has indicated that this
surfacing of latent tensions can ultimately be
constructive if the necessary supporting factors
are in place (McDougall and Banjade 2015).
This has already begun to be recognized in AAS
research (Kantor et al. 2015; Morgan et al. 2015),
and early strategies have begun to be developed
to work with this more complex perspective on
the system. For example, intrahousehold tensions
and conicts are being recognized and addressed
within the gender-transformative research in
relation to the microcredit initiative (see the
savings and internal lending communities,
known as SILC+GTA, illustration below).
Moving from concept to practice
Developing capacity through training,
partnerships, research and learning by doing
The program has taken a blended learning
approach to capacity building, combining
formal training with iterative learning in and
from the application of ideas (Sarapura Escobar
and Puskur 2014). A number of insights have
emerged from experience in relation to capacity
building through formal training:
Teams recognized that conceptual formal
training in gender-transformative approaches
is important, but insucient. To build
practical ways forward, training initiatives
need to directly connect with the specic
plans, activities and issues of the program.
Teams observed that focused capacity
development should start early, when
specic research initiatives are identied in
the program, rather than wait until other
aspects of the research are underway.
Teams identied that capacity development
around gender, along with the application of
gender research, is more eective when the
responsibility is shared across the team, rather
than being the sole responsibility of one or two
The connection between partnerships and
capacity building emerged as signicant. This
was both in terms of involving local partners in
capacity development around gender (to build
their understanding of its value in the research)
and for team members, including researchers
and community facilitators (to build relevant
skills, as well as in relation to scaling through
building networks and coalitions). Seeking
partners from outside the eld of agriculture
and development research who could lead
capacity development and engage in the
work as partners was extremely valuable. To
date, a combination of global partners played
this key role: The University of East Anglia in
developing capacities on gender theory and
analysis; Promundo in developing capacities to
implement gender-transformative approaches
in context; and Johns Hopkins University in
developing capacities in the specic area of
gender-transformative communication and
The knowledge generated through formal
research has played an important role in capacity
development for informing and operationalizing
the gender-transformative approach. For
example, in Zambia, drawing on the early
social and gender analysis ndings enabled
the team to base their research on empirical
data. This built credibility and thus condence
and momentum among engaged actors. It
also enabled the team to target their gender
research and gender-transformative approaches
more eectively than if they had been working
from general knowledge. Literature reviews
and expert dialogues have similarly helped to
ground and focus the research. For example, in
Bangladesh, targeted reviews helped to identify
priority areas for gender-transformative work in
relation to aquaculture.
Teams recognized the value of research
ndings in building a contextual understanding
of gender early in the program and using
these ndings as a foundation to identify
opportunities and entry points for gender-
transformative work. In four of the ve hubs,
delayed gathering and analyzing of social and
gender data and in-depth gendered context
studies led to a lack of useable information
regarding important local issues and entry
points for gender-transformative work in the
overall research. These teams agreed it would
have been better to undertake this analysis
earlier in the research and use it to inform
strategic planning around gender in various
RinD processes and initiatives. For example,
it could have been used to feed back into
community and hub PAR to spark gender and
social equity dialogue within those processes.
Just go ahead and “learn by doing emerged
as important rules of thumb for implementing
a gender-transformative approach. Given the
challenges outlined in this chapter, it is easy to
see how researchers are tempted to postpone
doing anything on gender issues. Teams said
in the reection sessions that it was important
to build on the capacity and information they
had and simply start with some small eort
in a learn-by-doing mode. As one researcher
commented, “We can’t wait for the perfect
time or perfect strategy: we need to just dive
in with GTA and learn as we go. For example,
participatory tools that were eventually used
in the social and gender analysis, such as
participatory wealth or well-being ranking,
could have been integrated usefully into the
community engagement processes early on.
Learning by doing expedites the learning
process. Some would say it is the learning
process. Moreover, from a systems perspective,
such an approach is appropriate to complex
systems. As such, formal training can be
complemented with space for and a culture of
team members regularly sharing, learning and
discussing the concept and its application and
then implementing another iteration of the
action-learning cycle. Moreover, team members
reported in the reection sessions that they had
begun to “develop our own habits or mindsets
as researchers of asking the ‘why questions
in relation to all aspects of the research and
context” (World Café session notes). Teams
also reported that it was also extremely
useful to have “outside eyes” on hub work to
help recognize when and where the gender-
transformative approach is evident (or not) and
how well it is or is not working.
Personal, relational and institutional shifts
and commitments
One factor that has enabled the teams to make
progress has been their commitment to the
gender-transformative approach at the level of
the individual researchers and the institution
involved. In several cases at the individual level,
this has been reected in the commitment of
the gender analysts and the unanticipated but
welcome commitment of other researchers
to gender issues. In Zambia (Barotse hub), the
relatively eective operationalization of the
gender-transformative approach relied rst
on establishing strong bonds and trust within
and between research and development
organizations. Once established, stakeholders
coalesced around salient social and gender
issues and began working together to achieve
better, more sustainable gender equality
development outcomes (Cole et al. 2014b).
At the institutional scale, WorldFish’s
commitment to gender is reected in the
resources invested, interest at the senior
scientist and management scales across
research themes and sites, and explicit
integration of gender into its overall aim. This
commitment has played a critical supportive
and enabling role in terms of institutional
willingness to support gender capacity
development and allow teams the freedom
to engage relatively uncharted territory in
agricultural research.
Emerging insights about a gender-
transformative approach
Overall insights
Several insights have emerged in relation
to implementing a gender-transformative
A gender-transformative approach is not
just about getting both women and men
together in the same room; it is about
bringing to the surface and initiating critical
reection and identifying options for change.
Its role is to engage diverse local women and
men in such critical reection and change
processes regarding underlying forces and
factors that shape equity, such as gender
and social equity-related norms, attitudes,
practices, processes and policies.
Building strong relationships among
scientists (and especially between social
and natural scientists), government and
development actors, and women, men and
youth in program communities is critical and
a prerequisite before change processes can
be initiated and realized.
Part of a researchers role is to facilitate
critical reection by asking questions
throughout the PAR process. These are
questions that help probe and bring to the
surface the underlying causes of imbalances
and the implications of the social and gender
status quo. These are the “why and the “so
what” questions.
Transformative gender work is a form of
social change, and as such it can only be
seeded, not forced or controlled.
Early understanding of the context, such
as through social and gender analysis and
gender benchmarking studies, can help
researchers understand the landscape
and inform core RinD processes from the
beginning. This type of early analysis can
also be fed back into these processes to help
researchers identify entry points for gender-
transformative work.
Gender-transformative research is a long-
term process that can be worked into the
research from the scoping stage, through
core RinD processes and throughout
technical initiatives.
There is no single strategy on which
the approach is based; rather, there are
numerous strategies relating to the principle
of facilitating critical reection (see below).
Examples of strategies for a gender-
transformative approach
There are multiple possible strategies for taking
a gender-transformative approach.What they
have in common is that they promote critical
reection and dialogue on gendered norms,
attitudes, behaviors and values and promote
the development of positive alternatives.
Moreover, the strategies also have in common
that they seek to empower individuals to take
up these gender-transformative practices by,
for example, promoting women’s agency to
participate actively in agricultural production
or enabling men to share household decision-
making power with their partners. Here we
present ve interconnected and overlapping
strategies or—more accurately—bundles of
strategies that AAS has been focusing on:
critical questioning, experiential learning, tools
for reection, communication for social change,
and networking.
Critical questioning for learning
Questioning deeply entrenched harmful
gender norms and practices is at the core of
the gender-transformative approach, and is
carried out via fostering group reection and
open dialogue within socially safe spaces.
While it is woven throughout all the strategies
presented, here we begin by presenting it as
a strategy within the community visioning
and reection cycles of RinD (i.e. community
engagement).Key points include the following:
Community and subcommunity processes
(such as visioning, planning and after-action
reviews) can prompt community members
to reect on harmful gender and social
norms and power relations. Questioning
harmful gender norms opens up spaces
for men and women to increase their
awareness of how unequal power dynamics
and harmful gender norms aect them as
individuals, their relationships, their families
and their communities, including in relation
to community goals and visions.
Critical questioning by community members
can be routinized through regular tracking
of gender and social equity in community-
based participatory monitoring.
There are no specic sets of questions, but
rather facilitators and researchers can help
to prompt regularly asking “why and “so
what” questions. In other words, researchers,
facilitators and community members
engaged in this critical dialogue seek to
go beyond reecting simply on roles and
responsibilities. Sex-disaggregated data
can help identify the root causes of gender
inequalities and their negative impact on
Strategic research on gender—such as
gender and social analysis ndings from the
site—can help to inform the researchers
and facilitators, as well as sparking critical
Experiential learning
This strategy applies critical questioning (above)
in combination with action. In other words, it
combines critical questioning and reection
with identifying and trying new ways of acting
or relating (such as new gender roles, or shifts
in gendered decision making). This action-
reection nature means this strategy ts with
or can be situated in PAR cycles of various kinds
or other action and learning-oriented activities.
This is illustrated in relation to microcredit and
aquaculture development in the Zambia and
Bangladesh illustrations described in the next
section. Key points include the following:
Experiential learning can take the form of
facilitated group sessions that integrate
critical questioning around gender
with a specic topic or activity (such as
microcredit) that is related to the overall
goal of a project or community action
plan. Participants in these group sessions
unpack how gender norms (as well as roles,
relations and behaviors) inuence the
activity and shape positive and negative
outcomes in relation to individuals,
households and the community’s identied
aspirations. Where the outcomes are
negative or inequitable, women and men
identify potential alternative norms, roles,
relations or behaviors (such as sharing
household work so that women can go
to savings meetings, or women and men
in households identifying joint goals to
reduce conict over spending choices,
or identifying ways in which women can
engage in local markets). Participants put
their identied solutions into action, testing
them and seeing to what extent they work.
They then return as a group to reect on
and learn from these experiences, and
iterate through further cycles of action and
Within these facilitated processes,
facilitators seek to create safe spaces
for women and men to reect on how
gendered norms, practices and rules
shape local realities and would impact on
desired (ideal) futures, in relation to specic
technical issues such as sh production,
markets and resource governance.
Experiential learning as a gender-
transformative strategy is potentially
potent because it can combine experiential
learning with new economic or technical
opportunities, as well as new capacities, and
potentially collective action.Together these
can contribute to building “power within”,
“power with and “power to”, as well as
shifting gender relations.
It is also a potentially powerful strategy
in that it can merge the social with
the technical by engaging in dialogue
around gender issues within technical
interventions. A main gender hypothesis
of AAS is that it is through implementing
a gender-transformative approach, hand-
in-hand with technology-focused and
livelihood-enhancing interventions, that
optimal results from both are achieved
(AAS 2012b). Achieving this marriage of
the social and technical requires rethinking
how technical interventions are delivered
(i.e. the process) and to whom, as well as
planning how purely social interventions
can be sequenced and layered with
technical ones. Examining whether and how
integrated packages of social and technical
interventions foster gender-transformative
change across contexts and social groups,
and how they aect technology adoption
and use, is a central research agenda for
gender-transformative approaches in the
agricultural sector.
Using participatory tools to spark dialogue
and questioning
Tools can be used as a strategy to support
critical questioning and dialogue, including
within a range of PAR processes. Examples
include the following:
Problem tree analysis that is applied to
unpack underlying roots of prioritized
technical issues can also eectively bring
to the surface underlying social and gender
Farming system analysis of roles and
relations of farming households can help
to illuminate contributions of women and
men to the household and bring to the
surface dialogue on working together within
households (see Solomon Islands mini-case).
Household-based visioning (Oxfam and
GADC 2014), or the gender road map, is a
powerful gender-transformative strategy
that has been tested in Cambodia. It is
designed to address unequal power relations
within the household. The model is targeted
specically to vulnerable couples facing
various issues such as poverty, domestic
violence, gambling and alcohol abuse.
Combining tools and experiential learning,
there are four steps to implementing the
gender road map:
о building capacity for beneciaries on
gender concepts;
о conducting monthly meetings with
beneciaries to identify gender issues
they are facing within the family;
о guiding couples to design their family
gender road map (a core step of the
model focused on household visioning—
what they want to be in the future
compared to their current situation);
о guiding couples to develop their action
plans and to hold monthly follow-up
sessions to monitor their progress.
There are a growing number of resources
that present specic gender-transformative
tools and how to use them, including in
the context of aquatic agricultural systems.
Examples include the following:
о The Bangladesh mini-case (this chapter)
drew on a range of social-consciousness-
raising exercisesthat included family
members, especially mothers-in-law
and spouses, based on the Helen
Keller International resource Nurturing
Connections (
о The Zambia mini-case (this chapter) drew
on tools presented in a forthcoming
manual by Promundo on savings
and loans groups with a gender-
transformative approach.
о A range of gender-transformative tools for
use on their own or as a series of sessions
in a workshop are presented in Promundo
and WorldFish’s forthcoming manual,
Engaging with Men and Boys on Gender.
Communication for social change, including
media and entertainment
Inspiring dialogue through media.
Media programs can take many forms,
including multimedia campaigns, radio
and TV programs, video productions, and
social media platforms. They contribute to
behavioral and social change by providing a
common language to address concerns, role
modeling positive choices, demonstrating
options for action, and above all, inspiring
people to talk about the issues raised
within their families, peer groups and
communities and throughout the country.
Dramatic stories and real-life testimonials
where people hear rsthand how someone
similar to them has been able to overcome
gender and economic challenges have
proven eective in motivating others to
take actions they may have previously felt
were too dicult to try. National dialogue
programs, where broadcasters across the
country discuss the same topic from various
angles for a set period of time (e.g. the role of
women and men in the sh value chain) can
be particularly useful for stimulating national
Photo Credit: Silvia Sarapura/WorldFish
Community members in the Barotse hub attending an AAS meeting with partners, Zambia.
dialogue and provoking conversations.
This dialogue encourages social normative
change, increases support for healthy
practices and actions, creates opportunities
for communities and groups to plan for
action,and, ultimately,enables improved
and sustainablehealth-enhancing action.
Participatory community theater for
development. This form of entertainment-
education methodology engages
community members to reect on their key
problems (e.g. gender and livelihoods) and
encourages them to voice their concerns,
plan together to overcome barriers, mobilize
the needed resources, and work in concert
with support from others when necessary. In
community theater for development, scripts
are based on investigations conducted
on the audiences’ lives to ensure they are
grounded in reality. Breaks are often taken
in the performance to get input from those
watching about how they would solve the
dilemma faced by the characters. Discussions
follow the performance to further stimulate
dialogue and reection about the topics
raised in the drama. These discussions often
feed into the script for the next performance.
Community radio programs. Radio is
still the most accessible medium in much
of the developing world. Community
radio is particularly important, as stations
are primarily established to be the “voice
of the people” within a set geographic
area. Programming focuses on topics the
community is concerned about and oers an
outlet for listeners to voice their concerns,
share their challenges and solutions, and
work together to solve community issues.
People who live within the community or
who go on a regular basis to collect rsthand
accounts of stories usually do the reporting.
Community radio provides much-needed
information, often in almost real time,
about events that are unfolding within
communities that are aecting people’s
daily lives. Programming can disseminate
information to increase knowledge,
motivate change, and inspire dialogue
and action around a variety of factors that
perpetuategender inequality.
Champions, coalitions and networks
This strategy involves partnering with well-
known and respected people, organizations
and networks at all levels to advocate for
gender transformation. At the community
scale, this can include local men and women
who display more equitable gender relations
and behaviors and can act as role models in
the community and in relation to the research
initiative. They can help to create safe spaces
to spark local discussions about gender norms
through feeding back information from
benchmarking. Beyond the community and
across scales, coalitions at the subnational and
national levels can advocate for systemic, legal
and regulatory changes that are necessary to
create an enabling environment for gender-
transformative action in communities,
institutions and households (e.g. Bangladesh
National Gender Working Group, Gender
Coalition in Southwest Bangladesh, and
knowledge sharing and learning platforms).
Examples of gender-transformative research
in action
Zambia: Savings and Internal Lending
Communities (SILC+GTA)
Women and men living in or along the Barotse
oodplain in Western Province are not only
some of the poorest in Zambia, but are also
vulnerable to demographic, socioeconomic
and climatic challenges (Cole et al. 2015).
The Lozi-speaking people who comprise the
majority population of this area depend on
the aquatic agricultural system for a variety of
livelihood opportunities. In an eort to enhance
the equity of the socioeconomic and political
structures that inuence the livelihoods of
the people dependent on the oodplain,
AAS operationalized a gender-transformative
approach in selected communities (Cole et
al. 2014a; 2014b). One of the rst actions was
the formation of the Savings and Internal
Lending Communities + Gender-Transformative
Approaches (SILC+GTA) pilot project.
SILC+GTA builds upon the CRS savings and
internal lending community (SILC) model—a
savings-led micronance initiative that helps
people in rural areas (where access to formal
nancial institutions is poor or nonexistent)
to create accessible, transparent, exible and
self-managed savings groups. The savings
accumulated are used to meet emergencies,
pay for anticipated expenses, capitalize on
business opportunities, and invest in productive
resources. The plus sign in SILC+GTA denotes
the integration of a gender-transformative
approach into this well-established micronance
methodology. SILC facilitators were trained to
implement gender-transformative sessions using
PAR processes that promote critical reection
and spark dialogue, action and learning with
women and men. Pre-pilot phase it was found
that SILC groups generally comprise women, yet
women’s domestic responsibilities, other socially
assigned roles and power struggles within
their homes make it dicult for them to attend
meetings and contribute larger sums of cash
to enable the pool of savings to grow. Spouses
of SILC group members were often unaware of
the purpose of their wives’ involvement in SILC
groups, felt jealous or insecure about their wives
participation and improved access to credit, and
provided little home support to their wives when
they were called for meetings. Additionally, men
did not tend to join SILC groups because they
believed that the nancial contributions would
benet other members more than themselves,
among other reasons (e.g. that such groups
are for women only). In some circumstances,
gender-based violence occurred when men
perceived their wives as economically more
empowered. On the positive side, SILC has been
shown over time to allow women to build their
business skills and use their capital to pay school
fees for their children or invest in businesses and
increase household incomes.
With support from Promundo-US, development
ocers from Caritas-Mongu and CRS, along
with researchers from WorldFish, designed
and began piloting the SILC+GTA model. The
rationale for the pilot was that by involving men
in SILC group formation and group activities
using PAR processes, it would be possible to
address harmful social and gender norms
and power relations that prevent SILC groups
from ourishing, as well as improving gender
relations within and outside the home, and as
a result, to achieve better and longer-lasting
development outcomes.
The SILC+GTA project is being implemented by
a group of multisectoral partners representing
local, provincial, national and global
organizations working in the Barotse oodplain
region. The project was informed by a social
and gender analysis conducted in 2013. Data
obtained from the social and gender analysis is
being used as one benchmark against gender-
transformative changes that are monitored
during SILC+GTA group meetings. As part of
the SILC+GTA project, a series of 12 focus group
discussion sessions were started. These SILC+GTA
sessions aim to stimulate discussion with women
and men on conceptual issues such as gender
and power, as well as on gender-based violence
and substance abuse, which were key issues
identied during the social and gender analysis.
PAR processes were embedded in the SILC+GTA
implementation methodology to promote and
foster spaces for reection, action planning and
knowledge sharing during the sessions.
Results and discussion
PAR processes have allowed women and men
to begin to realize that working together and
giving women the opportunity to be part of
economic activities can help enable them to
improve their livelihoods. At the same time, a
key outcome has been that communication
and engagement is contributing to building
social cohesion, trust and bonds among group
members. These features help individuals
support one another to improve their lives
through increased investments that lead to
increased savings and incomes. Women and
men have learned how to deal with conict, as
they demonstrated when they had to deal with
common issues within their community and
households. As a result of (usually) bi-weekly
encounters, both women and men have been
able to strengthen their leadership skills and self-
condence. One of the aims of the project is to
help facilitate women’s access to micronance to
improve their skills and capacities to participate
in household and community decision making.
Emerging results show that women are gradually
strengthening their decision-making abilities
and gaining respect from husbands, as well as
starting small businesses and increasing their
savings and incomes.
In sum, ndings have demonstrated that
eorts to achieve institutional changes in
the communities, as well as changes to
socioeconomic and political structures, are
needed. In addition, the SILC+GTA pilot project
is beginning to show that increases in income
and assets, as well as enhancement in resilience
and adaptive capacity, are necessary at the
individual level.
Bangladesh: Aquaculture technologies
Bangladesh is one of the world’s most densely
populated countries, with deep and widespread
poverty. Its water resources play an important
role in alleviating the country’s poverty and
ensuring food security through its sheries.
Approximately 20% of the rural population
(4.27 million households) own a household
pond. These ponds are usually very small in size
(50–150 decimals or 2023–6070 square meters)
and are owned by a single or several families
for various purposes, which may or may not
include aquaculture. As almost every household
in southern Bangladesh has small ponds
situated in the courtyards of households (called
homesteads), introduction of more intensied
sh farming to the women in the households
(who usually face mobility constraints) is
usually seen as the way forward to combating
nutritional and consumption needs, enabling
an extra income from an underutilized pond,
and further enabling women to take more
control over a household asset from which they
can contribute an income.
However, a recent CGIAR Research Program
on Climate Change, Agriculture and Food
Security and AAS-funded study that looked
into gender relations and technology adoption
in two projects funded by the United States
Agency for International Development (i.e.
the Cereal Systems Initiative for South Asia in
Bangladesh [CSISA-BD] and Aquaculture for
Income and Nutrition) has found that targeting
women for homestead sh technologies does
not necessarily mean that women are able to
use and adopt the technologies or receive the
benets that the technologies promise (Morgan
et al. 2016). The study ndings show that
women and men live in multidimensional layers
of relationships that need to be understood and
addressed when disseminating a technology.
The study ndings led to the piloting of a
revised extension approach in CSISA-BD’s
Faridpur hub (in southwest Bangladesh) that
tried to use the lessons learned from the
study. The revisions had to be made within the
Photo Credit: Md Zamal Uddin/WorldFish
Training session underway in Barisal, Bangladesh.
project’s time and budgetary constraints and
under pressure to deliver on numbers, which
many such donor-funded projects face.
The new approach involved merging the
technical sessions, with social-consciousness-
raising exercises, including exercises on
trust and teamwork. The strategies and
exercises were derived from the Helen Keller
International Nurturing Connections manual that
aims at challenging intrahousehold inequalities
and gender-discriminating practices that
hinder women’s successful adoption of and
benets from a technology. This merging of the
technical with the social aimed to help women
combat challenges they may face while trying
to apply the new knowledge that is gained.
The training was further modularized and
spread out across the entire production cycle,
enabling real-time application of the technical
knowledge. Other major changes in this pilot
included discarding the demonstration and
model farmer approach, forming smaller
preference-based learning subgroups, including
other family members in various sessions, and
using community theater groups in events to
create awareness on gender issues.
Results and discussion
Survey research methods and process
documentation are being used to monitor
the results of this gender-transformative pilot.
Based on the ndings and the ability of the
approach to foster gender-transformative
change while supporting technology adoption,
this pilot will be scaled out. Preliminary ndings
reveal women’s better scientic understanding
of pond management, which the women report
has led to a better status and respect within
the household. Other community members
were reported to seek out these household
members for advice on managing their own
ponds. Women’s self-condence and decision-
making ability increased and they gained trust
and respect from their husbands and other
family and community members. Specically,
the smaller preference-based learning
subgroups and exercises on trust and teamwork
helped to counter some of the group power
dynamics that the study helped to identify. The
inclusion of family members in the training
paved a way for women to attend the training
without hindrance. Two social and gender
analysis participants conrmed, “Because our
husbands, fathers-in-law and mothers-in-law
were included in some sessions, it was easier for
them to understand what we told them. They
don’t create any barriers to our participation.
Likewise, “Since they [other family members]
were included they heard it from the masters
themselves. They believe us now about the
benets of investing. Also, since input support
was uniform across all trainees as a result of
discarding the demonstration farmer model,
more harmony among the groups was created.
Finally, the technical livelihood incentive
made the attendance of family members and
participants more permissible in the social
messaging exercises, which involve games and
discussions around sensitive gender behaviors
and attitudes. For example, the women
reported behavior changes from household
members after an exercise on intrahousehold
food distribution. Accordingly, a consolidated
comment from the women participants was
as follows: There was an exercise with family
members on distributing food and on how
we usually make sure they eat better before
eating ourselves. So usually we don’t have
much on our plates. In the past, men didn’t
notice this. As long as they got a big piece or
the head, they were happy. Now, following this
exercise men check what we are eating. They
acknowledge that we work hard all day and
make sacrices and should eat equally.” Another
woman commented, “Our husbands ask us
before making purchases more than before. It is
because husbands are aware of the benets of
asking their wife’s opinion and since we women
were able to learn a lot from the training.
In sum, initial observation and feedback
revealed that understanding intrahousehold
gender dynamics and providing spaces for
women and men to reect on harmful practices
that prevent them from increasing their
household income, food supply and nutrition
through a gender-transformative approach in
technical initiatives increases participation for
Solomon Islands: Involving women and men
in aquaculture workshops, Malaita hub
In Solomon Islands, women and men are
involved in a diverse range of livelihood
activities. Terrestrial and freshwater resources,
inshore coastal areas, islands, and islets provide
opportunities for the people living in and
depending on these systems to capitalize on
their aquatic agricultural resources, of which
approximately 90% are held under customary
In rural communities, both men and women are
involved in community activities, in producing
food and generating income, and in preparing
food and taking care of their families, but their
roles vary by gender. Men may have more
opportunities to travel outside the community
to meetings and training sessions than do
women, who have the primary responsibility
for child care and work longer hours. These
dierent roles can aect whether and how
men and women are able to participate in
decisions about livelihoods and resource
management, as well as how they are impacted
by these decisions (Schwarz et al. 2014). When
opportunities arise for both men and women to
participate in meetings and workshops, there
may be social and cultural reasons that mean
women are less likely to speak up or contribute
toward decisions. It has been observed that
when selection of participants for training
opportunities relies on male community
leaders, most participants are men, even when
the leaders are explicitly requested to invite
women to events. When women do attend
community events with external organizations,
they often have a dual role of preparing food
for the participants, and as a result can spend
much of the meeting moving in and out, losing
the opportunity to participate fully.
To improve opportunities for both women and
men to benet from an emerging diversied
livelihood opportunity, gender-transformative
strategies were integrated into an aquaculture
project. Between 2012 and 2014, WorldFish
worked with more than 40 sh farmers in
the central region of the Malaita hub. Fish
farming is a new and emerging technology
in Solomon Islands. Farmer workshops and
training had been exclusively attended by men,
although researchers noticed that women
were participating in activities related to pond
To facilitate greater engagement of women
in accessing information and knowledge
about sh farming, in April 2014 the research
team made explicit eorts to engage married
couples in farmers’ workshops in which gender-
conscious facilitation was employed.
The workshops were organized and jointly
facilitated by AAS researchers who included
social and natural scientists specializing in
aquaculture. The facilitation process used a
tool that encouraged men and women to
draw a farming systems diagram in which they
mapped out their respective roles in their daily
livelihoods. They did this separately in groups
of women and men and then shared their
drawings with the full group. This demonstrated
that although men were the “face” of sh
farming, it was clear that women and children
were playing a signicant role.
Results and discussion
The farming systems diagrams highlighted
that the men were shing, farming, gathering
rewood and building houses. Women
highlighted reproductive roles such as food
preparation, water gathering, child and elder
care, house cleaning, and clothes washing, as
well as productive activities together with men
such as tending vegetable gardens, pig farms
and sh ponds. In aquaculture activities, men
were more involved in constructing the ponds,
stocking sh and harvesting, but depended
greatly on women’s and children’s support.
Women often fed the sh or took on all roles in
the absence of their husband.
When men and women shared these stories,
a couple that was already working together in
partnership and sharing roles around their sh
pond stood up and shared their experience,
encouraging other couples in the room that
they too could benet from working together
as a team. Women and men had the space to
identify the ways they could work together
and share the work that arises from operating a
homestead pond.
Since that time, women actively participated
in and contributed to a farmers’ meeting in
March 2015 and another farmers’ workshop
in June 2015. They have shown increased
condence to speak in front of men, and the
men have accepted the womens presence and
participation in recognition of the role they are
playing in this livelihood. As AAS plans future
gender-transformative approaches within PAR
with these sh farming families, reections
from these recent meetings show that women
felt that household-scale ponds were well
integrated into their daily livelihoods and did
not add a signicant burden to their daily work.
This integration of a gender-transformative
approach into PAR processes with pond farmers
has not only increased awareness among men
and women farmers but has also increased
the AAS team’s knowledge of the roles that
men and women are playing and the power of
employing gender-conscious facilitation in PAR
Poet Antonio Machado proposed that “paths
are made by walking. In this chapter, we have
highlighted some of the challenges involved
in developing and implementing a gender-
transformative approach within RinD and
shared team learning to date regarding how
these can be eectively addressed. This learning
has led us to multiple strategies and a solid
footing for our eorts. Moreover, in doing so,
our reections have suggested a revisiting of
Machado’s proposition. Our journey to date
has underscored the signicance not only
of learning by doing, but also of working to
bridge silos in this process, of being critically
reective together, and of learning from and
with research partners and diverse local people
over time. As such, in the context of a gender-
transformative approach, we might propose
that “paths are made by walking together.
Chapter authors: Douthwaite B, Kabir K, Karim M,
Lando LA, Longley C, Muyaule C, Perez M, Siota F
and Sukulu M
Other chapters in this working paper reect
on elements of the RinD approach. In this
chapter, we look at cases in hubs where the
implementation of RinD has led to more
inclusive science; that is, where farmers and
researchers are working together dierently
than might be expected from conventional
research-for-development approaches. Better
relationships between farmers and researchers
are an important outcome in the overall AAS
theory of change. This is because farmers in
aquatic agricultural systems are vulnerable
and becoming increasingly so as populations
increase, natural resources are depleted and
degraded, sea levels rise, and extreme weather
events become increasingly frequent and
severe. Farmers have always innovated to
adapt to change; however, their increasing
vulnerability requires that they innovate faster
and more eectively. Better links to researchers
have provided farmers with more connections
to sources of information and technology and
more opportunity to experiment and innovate.
In this chapter, after 3 ½ years of AAS
implementation, we reect on what four best
cases tell us about how RinD works and how
it is dierent from research-for-development
approaches that focus more on the generation
of new technology than on the relationships
between the people who generate and use it.
At the workshop in January 2015 described in
the introductory chapter, we identied four
best cases to illustrate where RinD has led to
more inclusive science. We agreed to adopt
a case study methodology after Yin (1989),
building our respective cases to structure
cross-hub learning and ensure the internal
validity of our conclusions. We choose to learn
from the best cases because in understanding
innovation, there is often less to learn where
things have not worked (Perrin 2002). The
cases were chosen as best examples of where
implementation guided by RinD has led to
researchers working with and responding to the
needs of farmers and shers in ways consistent
with overall program objectives. Hub case
study authors completed a rst draft of the
cases using a range of data sources, including
their own experience as participants, as well
as workshop and other process reports. Then
began an iterative process in which the lead
author queried the hub authors to produce nal
versions. As case study methodology suggests,
we paid particular attention to the sequence
of events and the plausibility of the causal
explanation linking them as a way of ensuring
the internal validity of each of the cases. Case
authors checked their narratives with other
members of the AAS hub teams to conrm
the sequence of events, explanations and
Case studies
Abaca rehabilitation in the Philippines
Lando LA and Perez M
In the Philippines, AAS works in the areas
shown on the map in Figure 9. The hub includes
areas of Visayas and Northern Mindanao.
AAS carried out community visioning and
action planning in eight barangays. In June
2013, two barangays in Sogod, Southern Leyte,
identied their main priority as rehabilitating
their abaca9 plantations from an infestation of
abaca bunchy top virus (ABTV).
Brgy. = barangay = smallest government administrative unit.
Figure 9. Visayas-Mindanao hub and the location of the AAS focal barangays.
Agusan del Norte
Zamboanga del Norte
Bohol Sea
Celebes Sea
Marine Region
Southern Philippine
Sea Marine Region
Northern Philippine
Sea Marine Region
Visayan Sea
Marine Region
Sulu Sea
Marine Region
Misamis Oriental
Brgys. Galas,
Dipolog City,
Zamboanga del
Brgys. Binitinan and
Waterfall, Balingasag,
Misamis Oriental
Brgy. Mancilang
Madridejos, Cebu
Brgy. Maac and
Mahayahay Sogod,
Southern Leyte
Brgy. Pinamgo,
Bien Unido, Bohol
The problem of abaca bunchy top disease in
the Visayas-Mindanao hub
In the early 2000s, an ABTV epidemic began
to seriously aect production in most of the
producing provinces (Raymundo et al. 2001).
Production fell by 15% from a high of 77,000
metric tons in 2000 to 65,000 metric tons in
2013 (Bureau of Agricultural Statistics 2014).
Government support for tackling ABTV largely
went to funding programs to eradicate infected
plants. These programs lacked grassroots
support because farmers wanted to continue
to grow abaca, not have it removed from their
farms. In Sogod and many other communities,
basic communication was hampered by a
misunderstanding resulting from the local word
for “medicine” being the same as “herbicide.
Farmers expected that plants would be
treated with medicine and recover. Instead,
technicians sprayed them with herbicide
and killed them. This led to a breakdown in
trust and poor implementation of replanting
programs. Technicians were afraid to go back to
communities because farmers were angry. As a
result, in 2013, farmers in Sogod and elsewhere
were not practicing eradication and eld
sanitation voluntarily or regularly. Despite the
failure, farmers still relied on the government to
do something.
Abaca rehabilitation programs only worked in
areas that had strong local government units
that had the power to mandate recommended
eradication and production practices.
Agencies working on abaca have tended to
be jealous of their mandates and treat each
other as competitors for funding. Researchers
have tended to see farmers as a source of
sample materials for disease management
and for breeding work and use their elds for
multilocation trials of varieties. Hybrids are
not yet available for general release. Some
institutions hold eld days for farmers to show
the progress of research work, not to acquire
feedback on whether the research is something
that farmers would use or even need.
Community engagement
The idea of rehabilitating abaca emerged
during the community life competenceprocess
visioning and action planning in June 2013.
Participants agreed that there would be no
more poor people in Sogod if abaca were
“given back to them. The AAS response was
to commission the National Abaca Research
Center (NARC), part of the Visayas State
University (VSU) based in Leyte and just 2 hours
away from Sogod, to conduct a rapid appraisal
of the feasibility of abaca rehabilitation. The
survey, completed in November 2013, found
that the two barangays were losing USD 2
million per year as a result of the drop in
abaca production (Tabada et al. 2013), a very
substantial fall in earnings given that about 6
in every 10 people in the two barangays are
living below the poverty threshold. The farmer
consensus appeared to be well founded.
A feasibility study was carried out by a
team of VSU-NARC researchers led by Drs.
Tabada, Abamo and Madayag. In setting
up the research, the AAS country program
leader, Maripaz Perez, used her professional
relationship with the VSU-NARC director, Dr.
Ruben Gapasin, and VSU president Dr. Jose
Bacusmo, which had developed when she
worked as undersecretary of the Department
of Science and Technology. In engaging
the researchers, AAS sta stressed the RinD
principle of putting farmer priorities rst and
so underlined the importance of involving
the farmers from the start to build their
understanding and ownership. The VSU-NARC
team began by visiting farmers’ homes to invite
them to come to a meeting to discuss survey
design and, more fundamentally, whether
it was still feasible to grow abaca in Sogod.
Through this process, they conrmed farmer
interest before beginning their usual rounds
of focus group discussions and key informant
interviews. Engaging farmers before the survey
work paid dividends when barangay ocials
coordinated with the local police to escort the
research team up the mountains to the abaca
farms, something that would not normally have
been expected to happen.
Their report concluded that abaca can be
restored in Sogod but only with the strict
implementation of eradication and production
protocols, including the use of resistant
varieties developed by the University of the
Philippines in Los Baños and VSU. AAS provided
the opportunity for VSU-NARC researchers to
share their results with the farmers, present
their recommendations for action and build
farmer buy-in for the proposed actions.
The farmers agreed to implement the
protocols. They initially asked for planting
materials and nancial support. AAS agreed
to provide planting materials in the form
of tissue-cultured hybrid seedlings, but not
money. A key principle behind the community
life competence process is that participants
should own and be responsible for their own
action plans motivated by a collective desire
to achieve a shared vision rather than receive
cash handouts. Constellation has learned that
programs that rely on achieving participation
through nancial inducements are unlikely to
lead to any sustained change in behavior. Other
strength-based approaches subscribe to the
same principle.
AAS monitoring in May 2014 revealed that none
of the farmers had acted on their action plans
because the strict eradication protocol dictated
that they had to eradicate karlang (a variety
of taro) from their farms, as it is an alternative
host to the aphids that carry ABTV. Also, farmers
wanted seedlings of their traditional varieties,
believing that the ber quality was better.
Karlang is their cash crop replacement for abaca
and so, not surprisingly, they refused to kill it.
AAS met with the VSU team to discuss the
impasse and what the next steps should be.
The researchers agreed with the farmers that
they would rst determine whether the aphids
found on alternate hosts, especially karlang,
were Pentalonia nigronervosa, the specic
vector for ABTV. If they were not, then the
karlang would not have to be eradicated. The
researchers also prepared a poster of frequently
asked questions written in the local language
and posted it in barangay halls.
During group discussions to negotiate the
karlang compromise, farmers suggested
including the neighboring barangays of
Javier and Maria Plana in the abaca work to
reduce the risk of reinfection. AAS sta saw
this as evidence of the farmers beginning to
understand the epidemiology of ABTV through
engaging with researchers. Farmers took on the
responsibility of talking to their peers in these
other barangays.
AAS facilitated a revisiting of community
dreams in all eight barangays in July 2014
as part of an annual PAR cycle. Despite the
onslaught of various typhoons (e.g. Haiyan)
and other natural disasters (e.g. landslides)
aecting the hub, AAS sta found that what
their Constellation coach had told them
was true—dreams don’t change until they
are fullled. Despite farmers not yet having
received planting materials, they conrmed
their dream to bring back abaca. They also re-
emphasized their preference for traditional but
ABTV-susceptible abaca because it produces
higher-quality ber.
Part of the delay in providing farmers with
planting materials was due to the fact that
seedlings are produced using tissue culture and
there were not enough for everyone. As a result,
the researchers had to design and negotiate
a seedling distribution system that would be
agreeable to all. After a series of conversations,
they agreed to start out with 70 farmers who
would receive 50 seedlings each. These farmers
would then repay the planting materials in
4–5 months when their seedlings produced
suckers. Each mother plant can produce 3–6
suckers in that period, and each farmer would
repay with 2 suckers from each mother plant
(thus 100 suckers). These would then be given
to two other farmers to plant, and so on until
all members of the abaca farmers committee
received 50 seedlings.
The interaction between farmers and
researchers over the provision of seedlings and
karlang proved a watershed. For the RinD team,
it marked the point where farmers collectively
started to believe they could help themselves.
Farmers started asking the researchers about
doing research on their own emerging
questions about abaca, and about whether
they could adjust the experimental protocols.
For example, one farmer suggested comparing
tissue culture materials against those growing
naturally that have been certied virus-free by
NARC. Another farmer requested that he do
his tissue culture trials on at land closer to his
house rather than in the hills where abaca is
usually grown. AAS sta facilitated agreements
that both farmers and researchers would
take actions based on each other’s opinions,
preferences and priorities.
In August 2014, farmers and researchers agreed
to have regular quarterly meetings. Farmers
were excited and agreed they would meet on a
monthly basis (even without AAS facilitation) to
compare their data and continue discussions on
eradication of the virus. However, many farmers
were still looking for nancial support and often
broached the subject at reection meetings. In
reply, AAS sta continued to argue for the RinD
strength-based focus of relying on their own
resources. Fellow farmers also urged them to
work on their farms instead of “complaining. The
hub RinD team has found that the farmers who
ask for nancial support continue participating
but return at the next meeting with the same
request for nancial support. The RinD team is
learning that changing this dependency mindset
will be a long process and will entail continuous
reinforcement of strength-based principles.
Engagement with hub and national-level
Two major events in national and hub-level
engagement were the stakeholder consultation
workshop and the design workshop, held in
September and November 2013, respectively.
The November design workshop produced
a strategic plan for AAS in the Philippines.
Abaca rehabilitation emerged as a priority of
an existing productivity initiative. An initial
theory of change workshop was then held
in January 2014 that further established
abaca rehabilitation as a main element of the
productivity initiative and improved partnership
for productivity enhancement as one of the
three abaca outcome pathways. The theory of
change workshop introduced participants to
articulating how an initiative will bring about
change and then detailing and testing that
theory during implementation as a way of
better appreciating how to leverage change.
AAS sta worked to engage relevant agencies
in support of this outcome pathway. There
had been a coalition called Abaca Disease
Management and Research Team (ADMART),
which was set up by a former vice governor.
This coalition stopped working in 2009 when
the vice governor was not re-elected and funds
given to the program could not be accounted
for. When AAS provided assistance to each hub
for strengthening partnerships, the AAS team
decided to use the funds to build on lessons
and old friendships from this coalition. AAS
facilitated a stakeholder meeting in July 2014
that included key research organizations (VSU-
NARC and Southern Luzon State University),
Sogod and Southern Leyte local government
units, and regional line agencies (Department of
Science and Technology Region 8, Department
of Agriculture Region 8, and Philippine Fiber
Industry Development Authority [PhilFIDA]
Region 8). Given the unfortunate history of
ADMART and competition between government
agencies for funding, AAS sta realized that
convening the group would be dicult. What
worked was the convening power of the AAS
country program leader through her previous
position in the Department of Science and
Technology, the reputation of Dr. Gaspasin,
NARC director, who has taught most of the key
people at some point, and the ability of AAS to
present itself as a neutral convener. Workshop
participants agreed that a coalition should be
re-convened to allow the agencies to work
AAS sta then worked to capitalize on this
agreement to organize an abaca stakeholder
consultation workshop in September 2014
(Box 5). The agencies presented their work on
abaca and engaged in an exercise to describe
future scenarios for the abaca industry in the
Philippines. During these conversations, the
agencies decided to formalize the coalition
discussed in July as a replacement for the
defunct ADMART. However, instead of sourcing
a common fund that the group would share,
they decided to begin working together
immediately in Sogod using their current
programs and budgets. For instance, PCAARRD
now plans to include Sogod in its target sites
for the abaca research and development
program, as well as setting up a community-
based science and technology farm. PhilFIDA,
in cooperation with Department of Science
and Technology 8, will channel the distribution
of tissue-culture-planting materials to Sogod
to support the PAR group’s research, and the
University of Philippines Institute of Plant
Breeding will include Maac and Mahayahay as
sites in the multilocation trials of abaca hybrids.
In keeping with local tradition and as a display
of the new partnership, the Abaca Coalition
was formally launched on 2 February 2015 in
Sogod with a motorcade. Members agreed
that a motorcade would be an inexpensive
means of letting people know who they are
and communicating their intent to mobilize
resources and work together. Each of the seven
local agencies brought one to two vehicles
that carried a banner showing the logo of the
member agency and the tagline: Kauban ta sa
Coalition Abaca (We are part of/We support the
Abaca Coalition). Community representatives
from Mahayahay hired a van, while the Maac
farmers brought their motorcycles. The
representatives from Javier and Maria Plana
rode in the agency vehicles. After the usual
opening ceremony, the vice mayor described
the progress of abaca rehabilitation so far, and a
small media conference followed that stressed
the coalition tagline “the future is bright with
abaca!” The motorcade nished at the Maac
barangay hall, where abaca seedlings were
distributed to farmers.
Emerging outcomes and learning
As of April 2015, the main outcome of the
work was that farmers, researchers, AAS sta
and Abaca Coalition members were engaging
and working in ways dierent from business
as usual. The organizations working on abaca
in the Philippines have their own mandates
and own ways of doing things, and as a rule
do not really talk to each other. Before the
AAS intervention, PCAARRD was not working
with PhilFIDA in Southern Leyte, which in turn
looked at NARC as a competitor. Except for
PhilFIDA, none of these organizations were
working in Sogod. PhilFIDA had reported the
completion of an eradication program in 2012.
The Department of Science and Technology
was working locally on its own priorities,
largely small livelihood projects based around
Box 5. Agencies attending the September
2014 consultation workshop
Philippine Fiber Industries Development
Authority (PhilFIDA)
Philippine Council for Agriculture, Aquatic
and Natural Resources Research and
Development (PCAARRD)
Institute of Plant Breeding, University of
the Philippines, Los Baños (IPB-UPLB)
Visayas State University National Abaca
Research Center (VSU-NARC)
Southern Leyte State University
Local government units
Department of Science and Technology,
Region 8
Private sector (chamber of commerce and
industry representatives)
food processing. Universities and research
organizations focused on their respective
research and development agendas. As a
neutral third party, AAS had the convening
power to bring these institutions together. As
long as none of them were cast as “leader or
“follower,” they could collaborate.
The AAS visioning and action planning process
was not enough in itself for farmers to become
proactive in rehabilitating abaca, despite
it emerging as their top priority. AAS team
members needed to continually remind farmers
that rehabilitating abaca was their program and
that they had the capacity to seek solutions
to their problems. Gradually, the farmers are
learning to appreciate the RinD strength-based
approach, in which the main input is convening
and facilitating spaces for farmers to engage
with each other and agencies who have a
mandate to help them.
Champions played an essential role in the early
successes of the abaca work. They provided
leadership for activities, rallied their networks to
the cause and provided initial resource support.
Champions were enabled rst by the realization
that they were part of the community and that
the community recognized them. Champions
saw how the inclusive and participatory
nature of the community visioning and action
planning resonated well with the community
and how the process could lead to more
sustainable initiatives that the community
could own.
The AAS team attributed part of the success
in changing mindsets to working within
existing governance and social structures. The
team always acknowledged the authority of
Box 6. Abaca Coalition champions
Barangay captains (Kapitan Fely and
Kapitan Raul)
Sogod mayor Imelda Tan
Vice mayor Rufo Olo
Department of Science and Technology
local facilitator Evelyn Tablante
Department of Science and Technology
regional director Engr. Ed Esperancilla
VSU-NARC researcher Dr. Ruben Gapasin
Abaca committee chairs Celso Ortiz and
Maximo Sotto
the barangay captains and local government
unit ocials, while stressing their mandate
to provide a service to their constituents.
They engaged partners according to their
mandate and provided an opportunity through
community engagement for them to deliver on
their mission. While the AAS team did organize
the Abaca Coalition, its members recognize that
AAS is only facilitating and does not seek to
replace their institutions nor their institutions’
Improving postharvest handling of sh in
the Barotse oodplain of Western Zambia
Longley C and Muyaule C
The AAS hub in Zambia is the Barotse oodplain
(Figure 10). Hub rollout began in 2011 with a
national consultation. AAS held a stakeholder
consultation workshop in June 2012, where
stakeholders agreed to work collectively to
address the hub development challenge, which
was “to make eective use of the seasonal
ooding and natural resources in the Barotse
oodplain system through more productive and
diversied aquatic agricultural management
practices and technologies that improve the
lives and livelihoods of the poor.
In the same workshop, stakeholders identied
access to markets and postharvest handling
of sh as priority areas. AAS then carried out
community visioning and action planning in 10
communities in August and September 2012.
Two of the seven priority areas identied were
also improved access to markets and improved
postharvest handling. The design workshop
in October 2012 to identify where hub
stakeholders might best support community
priorities established a value chain initiative that
would work on sh as part of the strategic plan.
Problems with postharvest sh handling
Postharvest sh losses are a major concern and
occur in most sh value chains throughout the
world (Partt et al. 2010). Nearly one-third of the
weight of sh harvested in Zambia is lost (Béné
2011). In the Barotse oodplain, postharvest sh
losses occur for a number of reasons, including
damage in nets, damage during transport,
Figure 10. AAS focal communities in the Barotse hub.
damage in processing and spoilage. Spoilage
reduces the price customers are prepared to
pay for the sh and is a particular problem in
the Barotse because of long travel times due to
poor roads and lack of refrigeration. Traders buy
fresh tilapia early in the morning and struggle
to sell the bulk during the day. Due to lack of
refrigeration, the quality and price drops over
the course of the day. Most customers wait
until evening when a sh seller is desperate for
buyers and will sell at a low price.
Processing sh is one way to reduce spoilage.
Current methods of processing sh in the
Barotse include sun-drying, smoking and, to
a lesser extent, salting. Salt sh is produced in
small amounts for markets in the Democratic
Republic of the Congo (DR Congo) and Angola.
The Department of Fisheries holds training
events for shers and traders on improved
methods of sh handling and processing
Traditional sun-drying and smoking tend to
produce brittle sh that are easily damaged
during packaging and transportation. Damage
is also caused by insects that lay their eggs in
the sh while it is drying in the sun, leading to
infestation by maggots. Insects and rodents
also eat sh in storage. To prevent these
problems, some processors resort to the use
of toxic substances to prolong the shelf life of
dried sh. It is highly likely that some of these
substances are harmful to humans. Where
rewood is in short supply, it tends to be
expensive, and dried cow dung is used as a fuel
for smoking. Salting of sh provides a good
technical option.
Community and hub-level stakeholder
Work on the value chain initiative began with
rice and sh value chain assessments carried
out from May to August 2013. AAS researchers
collected survey data from eight shing camps,
local and distant markets, harbors, and loading
points (WorldFish 2013). Most shing camps
selected were those where shers from the
AAS focal villages commonly go to sh. AAS set
up a sh value chain working group over the
same period to guide the analysis. The working
group included approximately 30 people from
the traditional authority (the Barotse Royal
Establishment), the Government of Zambia,
NGOs, market development organizations,
traders, and input and service providers.
AAS convened a sh value chain participatory
planning workshop in September 2013. The
aim of the workshop was to build ownership
of the initiative and agree to next steps based
on ndings from the assessment. This was
done through the participatory construction
of theories of change built on participants’
ideas about how interventions might bring
about desired changes. Participants included
members of the sh value chain working
group and people from the shing camps
surveyed, as well as from the 10 AAS focal
villages. During the workshop, participants
formed themselves into three interest groups
around the top three priority areas that
emerged during the workshop: (i) sheries
co-management; (ii) cooperatives, associations
and access to nance; and (iii) postharvest
processing. AAS then invited the three groups
to submit a proposal as to how they wished
to pursue their interest as part of a sh value
chain innovation platform. After the workshop,
the sh value chain working group met, and
on the suggestion of AAS, agreed to establish
themselves as an innovation platform. During
this meeting, the then Zambia country program
leader insisted that the platform had to engage
with shers from the focal communities. AAS
then hired a value chain coordinator, who set
up the platform and established regular joint
reection and planning meetings.