Knowledge co-production and co-dissemination for climate change adaptation: insights based on Chokwe district, Mozambique

Climate change demands proactive discussions among stakeholders to jointly produce and disseminate knowledge that addresses local needs and informs policies. Research on how these stakeholders’ knowledge co-creation platforms operate considering the ecological and socioeconomic dynamics remains limited. This research attempts to fill this gap by studying knowledge co-production and co-dissemination for climate change adaptation in the Chokwe district, Mozambique. It uses qualitative data drawn from 9 focus group discussions and 70 semi-structured interviews. Research findings show that everyday conversations, community meetings with externals, field visits and training programs are platforms of knowledge co-creation. Throughout these platforms, horizontal co-dissemination occurs with one-to-one conversations and neighbourhood meetings. Vertical co-dissemination is observed in disaster management bureaucratic settings, followed, for example, by extension workers in their interactions with other formal entities and farmers. Transversal co-dissemination involves cross-sectorial stakeholders collaborating trough community meetings and training programs. Then, television and radio were indicated as key additional tools informing people about climate change and disasters, albeit primarily through unilateral arrangements. Furthermore, knowledge co-creation differs across spatial and temporal contexts. In disaster prone areas, it focuses on climate adaptation. Meanwhile, in non vulnerable areas, development based knowledge co-creation dominates. The decision making process depend on how local population see and trust certain collaborative outcome. Overall, co-creation platforms for climate change adaptation involve both top-down and bottom-up perspectives since the knowledge from outside stakeholders reaches local communities and vice versa. Nevertheless, this is not empty of power struggles, exclusions, institutional learning, and other dynamics. Therefore, co-creation initiatives should forge structural transformations, generating solid partnerships, strategic, sustainable, longer, inclusive, and accountable adaptation programs.

Climate change is a current challenge referring to the long-term alteration in local or global climate patterns, this may result from natural variability and/ or human activities (Chukwuemeka et al., 2024). With these alterations, temperatures have been rising at an alarming rate since the 1970s, and intense debates on these changes began in the late 1980s, besides with climate events unfolding rapidly, the frequency and intensity of disasters are increasing, impacting human survival, and generating more challenges for adaptation (Beniston, 2010; Ashby and Pachico, 2012). Since 1970 the average annual disasters events are increasing with a peak in 2000, then a slight drop between 2012 and 2021, yet over the last six decades, a disaster overlapping situation is occurring, with more countries experiencing one disaster before the other one ends, for example storms and floods, heat wave and wildfire (IRFC, 2023).

Generally, we are moving into unknown domain with adaptation solutions, including the knowledge adaptation systems being strained (Schleussner et al., 2024). With these challenges, it becomes more evident that climate change impacts are increasingly requiring collaborative efforts to unfold power relations and foster joint learning for collective benefits. One of the entry points is the shift from traditional centralised governance to inclusive governance, in which actors recognise various interests and collectively discuss policies. Thus, we grasp inclusive governance as a crucial framework promoting the active participation of several stakeholders, including marginalised groups, to facilitate community-driven strategies (Soriano et al., 2024). How does this occur? What are the outcomes and potential challenges emerging from these processes, especially in local contexts? These questions remain under researched, and responses can take different angles. Researchers have dealt a distance with co-creation methodologies and frameworks (Wainman et al., 2026; Şentürk et al., 2026) as well as regarding the barriers and strengths of stabilising and running co-creation platforms (Aliska et al., 2025; Abderhalden et al., 2025). In this paper, we assess inclusive governance in the context of climate knowledge co-creation for adaptation in Chokwe District, Mozambique.

Addressing climate adaptation and inclusive governance involves navigating through ontological perspectives as a form to explore what climate change is, or its existence and the problem itself, as well as the epistemic approach to focus on the strategies to deal with climate change (Goldman et al., 2015). Different forms of knowing, for example scientific and local knowledge may apply, either independently or collaboratively to respond climate change impacts. But in any approach, the application remains with challenges. Yet, what is knowledge, what is science, and how different knowledge systems are seen in climate adaptation? Knowledge is societal and dynamic product detained, organised, and created through social interactions, as well by accumulating, organising, and evaluating facts, data, and information (Weichselgartner and Pigeon, 2015). Science is an enabler to obtain knowledge that is rigorous, empirically tested and justified based on the application of vital formulated methods, emphasising strict rules (Müürsepp, 2013). Moreover, while scientific knowledge remains as a categorical knowledge produced trough formal methods as well as measurable and with worldwide claims (Raymond et al., 2010; Nyadzi et al., 2021), local knowledge is associated to perspectives based on experimental, localised, and cultural aspects transmitted throughout generations (Li et al., 2016; Uddin, 2022). Although local knowledge is not purely local, since local populations interact with outsiders and experience other ways of living which demand adjustment of the available local knowledge (Artur, 2011) and scientific knowledge remain dominant position, while the knowledge from local communities is often overlooked (Kita, 2026). It is crucial to emphasise that scientific knowledge struggles to shift from universal solutions to provide contextualised and robust explanations of the uncertainties and the local adaptation (Knutti, 2019) and local knowledge is advocated as distinctive, and powerfully located font of wisdom, meaning that its integration with other knowledges enriches local climate adaptation (Makondo and Thomas, 2018; Mustonen et al., 2022). Then, this paper focus on the perspective that scientific and local knowledge are expected to be complementary. As Metha et al. (2019) also argue, uncertainties due to climate change demand the link between scientific knowledge and local knowledge, because these knowledge systems are exposed to limitations, and the interaction among them may influence reliable adaptation strategies. Along this, the interest and demand for inclusive governance in climate change adaptation are expanding. This spans from the recognition that inclusive governance enlarges the understanding of participatory learning, development pathways, and how distinctive knowledge systems operate (Utter et al., 2021). Thus, to promote such an approach, multi-stakeholder platforms emerge, allowing different knowledge systems to be reflected in climate change adaptation and a co-creation process to unfold (Offat and Miriam, 2015).

Fagiewicz et al. (2021) structure co-creation as a process relying on active participation of several groups: academia, business, Government, and society. Mauser et al. (2013) postulate knowledge co-creation as comprising three stages: co-design, denoting the outlining of the scenario; co-production, referring to the collaborations on a specific issue; and co-dissemination as the publication and open discussion. Likewise, for Brandsen and Honingh (2018), co-creation is wide-ranging, covering the definition of the problem, its unfolding, and other processes before a decision. Within co-creation discussions, Ziervogel et al. (2021) posit co-production as a capacity building approach supporting the understanding of local problems and adoption of context specific adaptation responses. Then, for other scholars, dissemination stands as “an active approach of spreading evidence-based interventions to the target audience via predetermined channels using planned strategies” (Rabin et al., 2008, p.118). It also implies an “interactive process of communicating knowledge to target audiences so that it may be used to lead to change” (Ordoñez and Olivier, 2017, p.871). Furthermore, the interaction occurring among stakeholders at the same level is defined as horizontal dissemination, while when it is done in a hierarchical form or top-down approach is referred to as vertical, and when involving several stakeholders with different knowledge and power lines, the dissemination is called transversal (World Bank, 2012; Acunzo et al., 2014; Lingga et al., 2023). Henceforth, in this paper, we use the term co-creation to refer to the full collaborative learning process to produce and share knowledge. By co-production, we refer to a concrete stage of collaboration to generate a solution. Co-dissemination is used to mean a joint responsibility to spread knowledge and/or share co-production outcomes, considering different lines of stakeholders and interactions.

Knowledge co-creation does not occur in an organisational vacuum. All knowledge is co-produced through socio-political relations as it arises within certain social backgrounds, identities, and power interactions (Jasanoff, 2004; Nightingale, 2016). Concurrently, stakeholders have different means to define the main concerns and motivations, create and validate knowledge, then responsibilities and time management also differ (Vincent et al., 2020; Stier and Smit, 2021). Moreover, it is worth realising that unequal scales of knowledge and different interests influence conflicts (Ahlborg and Nightingale, 2012), and due to elitism, some local populations embrace the interests of leaders, rather than their own expectations (Nightingale and Ojha, 2013). In Nehring et al. (2025) view, a “pure” co-production as part of co-creation can also be challenging and does not remove the inequalities among participants. Hence, how ideas, power and resources are conceptualised, negotiated, and executed by different groups at distinct scales becomes pertinent to climate inclusive governance (Tanner and Allouche, 2011). Accordingly, a power analysis before and during any co-creation intervention is necessary to foster genuine empowerment and inclusion of marginalised groups given the fact that co-creation platforms may challenge narratives and power relations (Helga et al., 2022; Metha et al., 2019).

Chokwe District provides a captivating case for analysing co-creation discussions, because historically the District has been between development efforts and disaster management. Chokwe hosts the largest irrigation scheme of the country (Ismael et al., 2021), and due to that, over the years it has witnessed an inflow of private and Government actors who have engaged with local farmers for development programs (Struijk, 2013). Additionally, as a region vulnerable to floods and droughts (Jonas et al., 2024), it has also endorsed an inflow of humanitarian organisations, donors, and other disaster governance structures. The Government has its institutions and frameworks to manage disasters as well as partners ranging from United Nations Agencies, International Non-Government Organizations and National Non-Government Organizations engaging in activities covering several sectors, such as education, health, shelter, food security, nutrition, water, sanitation, hygiene, and protection (UNOCHA, 2019). For illustration, when Chokwe was flooded in 2000, Oxfam had interventions in water, sanitation, and public health (Oxfam, 2000). In 2013 floods, Red Cross was engaged to evacuate people, support with first aid, operationalise health campaigns and other coordinated activities (IFRC, 2013). Save the Children was another entity that in 2000 focused in supporting families with food, non-food products and other health support also giving strong attention to children, then in 2013 among other assistance, the organisation supported some families with shelter supplies and construction tools kits to (re) building their houses (Save the Children, 2000; Bovens, 2013). Furthermore, LIVANINGO, Malhalhale, Kulima and other organisations with National representation, and origin based on humanitarian activities effects to the civil war which ended in 1992, later expanded their interventions behind humanitarian assistance to donor funded projects focusing on disaster recovery, sustainable and/or rural development, social justice, environment and so on.1

Alongside these climatic events and humanitarian interventions, migration to South Africa and Maputo City has been increasing and is (re)structuring how communities live in Chokwe (Midgley et al., 2012). For example, a survey indicated the prevalence of female headed households, associated to high HIV/AIDS index and male labor migrants who returned home sick and shortly passed away. The same survey concluded that migration affected demographic figures, labour division and local knowledge patterns, with women holding more experience, for example in local knowledge about seeds (Waterhouse et al., 2004). As Pellizzoli (2010) also argue, male migration affected many family’s relations and women responsibilities because their work and presence in agriculture increased and, in some cases, remittances (from male migrant workers) were fundamental to cover land and farming costs. Hence, the convergence of several stakeholders’ involvement in disasters, migration and gender patterns provide a fertile environment for multiple and challenging knowledge systems, and so a dynamic field for studying the interactions and outcomes of these knowledge systems.

Generally, knowledge is crucial in climate change adaptation, yet climate change is challenging these knowledge systems, which require co-creation for adaptation. While there are clear positions that knowledge co-creation involves power dynamics among stakeholders and strong governance structures, a gap remains about how current co-creation processes involve with rapid ecological and socioeconomic dynamics, and stakeholders with limited resources, mostly dependent on donors funding. Considering these intersections, the study contributes to the ongoing debates on climate knowledge co-creation providing empirical evidence on what platforms and circumstances co-creation fosters inclusive governance. So, focusing on collaborative learning, adaptive governance, and institutional learning as part of adaptive co-management framework, this paper discusses the process of co-production, co-dissemination, and use of knowledge for climate change adaptation in Chokwe District. For that, the research investigated the following questions: What informal and formal platforms exist for discussing local knowledge (LK) and scientific knowledge (SK) for climate change adaptation in Chokwe? Who participates in these discussions? How do these platforms operate? Is local knowledge incorporated into framing adaptation, or is it merely acknowledged within scientific settings? What are the features of climate knowledge co-creation at local/ informal and institutional levels?

Theoretical framework: adaptive co-management

The two key features of the study; co-production and co-dissemination of knowledge in climate change adaptation, are broadly framed as part of inclusive governance and adaptive co-management to bring collective well-being.

Adaptive co-management results from the bridging of adaptive management and collaborative management fields. Adaptive management emerged in the 1970s through resilience and socio-ecological systems studies, with a focus on learning and decision-making processes (Hurlbert, 2018). During the 1990s, the adaptive management framework was questioned due to its limitation in dealing with complex socio-ecological systems involving multiple stakeholders and interests; then co-management approaches were advocated (Hasselman, 2016; Islam et al., 2018). Co-management is viewed as a problem-solving knowledge co-production process (Berkes, 2009) and an inclusive decision-making process based on trust (Ahmad and Esposito, 2025). Hence, adaptive co-management refers to “an ongoing process that allows stakeholders to share responsibility within a system where they can explore their objectives, find common ground, learn from their institutions and practices, and adapt and modify them for subsequent cycles” (Fabricius and Currie, 2015, p. 148). Importantly, through adaptive co-management, horizontal and vertical relationships are created, which then foster collective learning (Hurlbert, 2018). Furthermore, four key foundation’s structure adaptive co-management: (1) Governance, which is about empowering settings based on institutional, policies and legislative engagements; (2) Institutional learning based on continuous evaluations of co-creation initiatives and outcomes (3) Collaborative learning built across stakeholders with distinctive interests, ranks and scales and (4) Stakeholders engagement trough consistent and repeated collaborations (Fabricius and Currie, 2015).

Adaptive co-management is challenging in settings with scarce stakeholder capacity, weak governance, reduced trust, and a need for urgent solutions (Fabricius and Currie, 2015). Nevertheless, this is an appropriate framing for this research because it permits focusing on collaborations within the uncertainties imposed by climate change impacts. Thus, we concur with the perspective that adaptive co-management is a prospective tool amongst governance alternatives to adjust unsustainable socio-ecological systems, since “Ecological and social uncertainty is acknowledged as inherent to governance and is best addressed with collaborative processes and recognition that multiple sources and types of knowledge are relevant to problem solving” (Armitage et al., 2009, p.96). Then, throughout data analysis and discussion this paper focus on the following foundations of this theory: collaborative learning, adaptive governance, and institutional learning.

Data collection

The research design for empirical data collection is based in three phases. The first phase occurred in May–June 2023 as an exploratory visit conducted in Chokwe District to gather relevant information for the main fieldwork. During this first visit, conversations occurred with District Service for Economic Activities (Serviço Distrital de Actividade Económicas - SDAE), some farmers’ associations in the Chokwe irrigation scheme, technicians at District Service for Planning and Infrastructure (Serviço Distrital de Planeamento e Infraestrutura - SDPI), professors and researchers at Instituto Superior Politécnico de Gaza and a representative of Lower Limpopo Irrigation Scheme (Regadio do Baixo Limpopo-RBL)2.” During this visit, some interviews were done, and the data are used in this research as presented along this section, and based on the geographical settings livelihoods, disaster experience and other socioeconomic dynamics, the five communities were identified to be visited during the main fieldwork. The second phase was conducted in September 2023 as the main fieldwork. This fieldwork initiated with some meeting with district authorities to inform about the stay and the activity in the area. In this process, Non-Government Organisations (NGOs) were interviewed based on their active involvement in co-creation with local populations, farmers associations were also interviewed in groups considering the activities in Chokwe. Then, in local communities sampling included local leaders, household head who participate in co-creation platforms, as well as other household members who were able to share their experiences in climate adaptation strategies. The third phase was a validation and follow up research conducted in June 2025. During this event participants included extension workers, farmers, and local leaders, the conversation occurred in the same place trough separated group discussions, then when consensus was demanded we turned into multistakeholder discussions. So, the preliminary findings were validated with the participants, and supplementary data about knowledge co-creation process features were gathered also confirming the data saturation. Though it was initially scheduled for late 2023 or 2024, this activity was delayed due the socio-political constrains that imposed safety risk and travel limitations from Maputo to Chokwe, combined to delays in securing additional fieldwork funding.

In any cases, the focus group discussions or individual interviews, the facilitation made all possible efforts to make sure that voices were heard either individually or in group. Focus group discussions were relevant to provide relevant data from individual, collective, and interactive perspectives regarding concerns affecting a group (Clark et al., 2021). So, to apply the method, the research question focused on how/when/where farmers, extension workers and local leaders interact with peers as well as with other stakeholders regarding farming activities and other community issues under climate change. For the individual interviews, participants were selected through snowballing, meaning that each interviewed participant indicated other potential participants, until the saturation point was reached (Naderifar et al., 2017; Gierczyk et al., 2023). So, in this process, each interviewee was free to help in the identification of other neighbours who have been living in Chokwe for considerable time and had disaster experiences. The saturation point occurred when after several interviews in Chiaquelane, Chokwe town and Conhane, additional conversations did not bring different patterns or dimensions to the study. These interviews were crucial to capture the existing interactions among community members, as well as between them and formal entities. The conversations also explored aspects related to the co-creation participants profile/influence, topics of discussion, learning processes, dissemination and use in neighbourhoods and farming.

In all conversations, informed consent was done verbally due to high illiteracy rate, so there was time for introductions (names, origin, institution etc). Then an explanation about the research and why decided to conduct it in Chokwe was provided. Subsequently, the expectation was explained, emphasising the interest to hearing and learning from the participant/s about their experiences with climate change etc. To move on, other conversations were related to recording and personal information; clarified and agreed on confidentiality, right stop the conversation, to ask questions along the interview and the use of the information which was to write articles and give presentations at the university and other places, involving fellow students, researchers and other academic/related stakeholders. To confirm the participation, a final question was to know if he/she wanted to move with the conversation and answer the questions and to be recorded or not. Then, if decided to move his/her choice was followed and again highlighted about confidentiality in the whole research. Moreover, with exception of conversation with formal entities, extension workers and three household representatives, all other conversations were conducted in changana, the local language. Not all interviews were recorded, since some participants had concerns about being recorded. The few records had simultaneous translation and transcription done manually after the field activities that day, when there was enough time, or later after the field work.

Research participants

This research draws on 9 focus group discussions and 70 semi-structured interviews. Focus group discussions were conducted with extension workers (n = 1), local leaders (n = 2) and farmers (n = 6). The individual interviews were conducted with community members (n = 63), NGOs representatives (n = 4), Government technicians from SDPI (n = 2) and RBL representative (n = 1) (See Table 1).

Researcher positionally and reflexivity

The principal researcher had no prior fieldwork in Chokwe, which provided a clear or objective foundation for the data collection, analysis, and report. The researcher had continuous reflexivity being aware on how personal experiences, previous reading and believes could shape field work interactions, data collection, analysis and so on. Therefore, for research integrity, notes were systematically taken to register additional aspects related to participants behaviour while answering specific questions, field dynamics, and other observations. The researcher embraced a learner perspective putting the participants at the centre of the research, and being dedicated to hear and respect their opinions, traditions, practices and accurately reflect them in the whole research process.

Data analysis

The study uses thematic analysis, which involves familiarisation with the data, coding, theme identification, theme review, construction of the narrative in relation to themes and sub-themes, and evidence to support the analysis (Clark et al., 2021; Ahmed et al., 2025). Hence, to permit an immersive contact with the data, through deductive (manual) coding/ codes predefined based on research questions and/or the literature review, rather than from the data (Bihu, 2023), the process considered the definition of codes such as discussions, interactions, meetings, conversations, sharing, learning, and teaching. These terms were created based on reflections from the literature review, research question and theoretical framework since they mostly imply to a process involving more than one person/ stakeholder, which in turn developed an assumption of possible co-creation process, presented in the next figure as the key or broader theme.

Once this was defined, a reflective exercise drawn on the theoretical framework, research questions, previous studies, and reports (see for example, Dale and Armitage, 2011; Singh et al., 2021; Sitoe, 2023) permitted to create a list of key stakeholders, which were then linked as the first set of subthemes for analysis. The next step consisted in creating a list of possible collaborative platforms, which included several aspects, such as Government/sectorial meetings, field visits, training, workshops, school farming, farmers fairs, media debates, community meeting, informal conversations and so on. Then, after repeatedly reading the printed transcripts, using different markers and sticky notes a gradual exercise of colour marking/highlighting was manually done in the following order: (1) Identify and highlight the excerpts matching the initial the codes and (2) Use the highlighted excerpts to separately mark information about stakeholders and (3) Categorise the excerpts according to the co-creation platforms. In this process, the information from the latest fieldwork events, was integrated after the event, following the previously defined codes, theme, and subthemes, since this data set also reinforced data saturation claims.

Despite the active and reflective involvement in developing the codes, the theme, and sub-themes, at this stage, with significant familiarisation with data, some of the previously listed stakeholders and collaborative platforms were excluded, remaining with the one that were supported by the empirical information. It was in this stage that knowledge co-dissemination especially through media, was introduced as separate subtheme, supported by the data showing little about media as collaborative and actively inclusive tool, because local populations have access to such platforms, but show concerns on how the information is produced and shared to convince them about some adaptation actions. So, here a form of inductive approach was used to complement the definition of subthemes. Then, the marked materials and other notes were repeatedly organised, read and used as a guide in the refined writing process of the final narrative.

To proceed, the chart (See Figure 1 Analytic structure) provides the overview of the process, which indicates co-creation as a process expected to flow from both top-down and bottom-up perspectives. However, taking the bottom-up approach as the most significant to the discussion regarding co-creation and inclusive governance, the presentation and discussion prioritise a bottom-up structure, emphasising: (1) Local communities as stakeholders with local knowledge and categorised in level three, leading to everyday conversations as an informal co-creation platform; (2) Government agencies in the first level and part of the second level (NGOs, extension workers, and other technicians) form multisectoral interactions leading to community meetings. Then, field visits and training programs integrated into activities developed by NGOs and the Government are also discussed and (3) Acknowledging that all the interactions also enable the co-dissemination of knowledge, the study looks to the media as a tool to be transformed into an additional inclusive dissemination channel and a co-creation platform. Throughout the paper, some quotes are used to provide relevant examples from the interviews and group discussions.

In this section, we first examine co-creation platforms that target all community members through everyday conversations and community meetings involving formal entities. Second, we look to co-creation processes focused on farmers throughout field visits and training programs with external actors.

Co-creation targeting all community members

Everyday conversations

Everyday conversations in the neighbourhoods take place through scheduled meetings, casual talks with neighbours, and peer interactions in farming areas. They arise from how local populations perceive and experience changes in farms and houses, as well as from the limitations in developing suitable solutions based on what they have been learning and doing. In general, everyone participates in these conversations, and successful experiences are replicated within the community. These conversations are not limited to knowledge exchange but also to the exchange of goods, favours, opportunities, alliances, money/saving groups, and power disputes. For example, people exchange seeds that are more tolerant to drought or share market prices and talk about the best and worst buyers/sellers in the neighbourhood or outside the research area, as well as other advice to cope with specific events:

The local leaders invite us to neighbourhood meetings. The meetings about climate related issues occur when the rainy season, storms, or droughts are approaching or disturbing our community. We exchange ideas about organising our belongings or household documents [such as identification cards, medical and school files, etc.], reinforcing the roofs, looking for wild fruits during the dry season, and so on. [Community member].

These informal spaces are not power neutral. For example, during disasters, for some households, family relocation to resettlement areas and adoption of new adaptation practices emerging from the co-creation process might require male authorisation, even when he was outside the district.3 Alongside this, male migration to look for livelihood alternatives away from Chokwe is changing gender relations. Then, knowledge co-creation processes are massively involving women who are developing autonomy, spaces, and knowledge to adapt to the context. As some participants mentioned, women are now, more than ever, actively engaged in conversations that enlighten and shape specific knowledge to deal with climate change impacts. In the research area, women are part of different development and disaster risk reduction groups. Besides, their active involvement and leadership in community, farming, or other social groups is notable.

Generally, over time, women have taken more responsibilities in agricultural and household activities. Then, they become more empowered in deciding the options for household well-being. Hence, living and dealing directly with climate hazards, their expertise is shared in everyday conversations with both males and females, as highlighted in some interviews:

We always communicate about abnormal things that happen on the farms; we look for solutions among ourselves; even without the extension worker or other project technician who works with us, everyone [male and female] speaks [Local farmer association members].

Key informant interviews reveal that community members who are more resilient and foster wellbeing, despite all the challenges, are more likely to be visited by their peers and asked for everyday conversations. These conversations are somehow connected and perpetuate a cyclical opportunity of knowledge co-production and co-dissemination. For instance, when a household representative does not join the neighbourhood meetings, later, he/she approaches the leader or other neighbours to talk about what happened in the meetings. In the case of farmers, when applicable knowledge is discussed in the neighbourhood, it is also taken to the farm talks and vice versa. Moreover, due to their occupations, lack of local knowledge or other factors, some community members are exposed to other sources of knowledge. Others, to contribute to local disaster management, volunteer and join formal training facilitated, for example, by the Disaster Risk Management Institute and other partners. Then, since they continue to join everyday conversations, this also diversifies the knowledge discussed in the informal co-creation process.

Community meetings: an exercise of inclusive governance in climate knowledge co-production?

Multisectoral collaborations leading to community meetings. Multisectoral collaborations bring together several competencies to discuss and collaborate on climate related issues. For example, the National Institute of Meteorology (INAM) monitor meteorological events, and the National Directorate of Hydraulic Resources Management (DNGRH) manages the hydraulic infrastructures and resources. The National Institute for Disaster Risk Management and Reduction (INGD) takes responsibility for managing, preventing, and responding to disaster risks. The SDPI works on planning, territorial organisation and infrastructure management at the district level, and the District Services for Economic Activities (SDAE) oversees the local development, covering agriculture, commerce, livestock, industry, and tourism. Then, at several levels, the listed and other government agencies have meetings, workshops, or other events to cooperatively design solutions on climate related issues.

In Chokwe, the meetings involving several sectors also have the district administrator’s participation to share updates and coordinate the interventions in the district. Hence, local government interventions in disaster management have a multisectoral approach to ensure that all perspectives are considered and are based on factual data from each competent Government agency.4 Moreover, these kinds of collaborations at a formal level are also expanded through the work with private sector stakeholders. In an interview with NGOs technicians, it was mentioned that the work with other NGOs and Government agencies was strong. For instance, they have coordinated strategies to prioritise training and capacity building programs in communities, as well as the work of local disaster committees trained by the INGD. 5 Since our main interest is in knowledge co-creation involving the LK, we explore multisectoral coordination as an entry point for a more participatory approach in knowledge co-creation among local populations, the Government, and partners. Thus, in the following subsection, we argue that community meetings with external entities can be an approach to inclusive governance for climate change adaptation.

Community meetings with external partners

They are part of the partnership among communities, Government agencies and private actors engaging in reflections about climate related events affecting local populations. Similar to what happens in the informal co-creation platform, women actively engage in public discussions. Government and NGOs technicians mentioned the contact with community members to address different issues in Disaster Risk Reduction (DRR). The interactions occur through formal and scheduled meetings in the community. The local Government may initiate the meetings and invite private partners, or the NGOs may be the ones inviting local Government representatives. Then, the meetings are coordinated with local authorities as formal and frequent focal points in communities: village leader, head of 10 houses - chefe de 10 casas, head of block - chefe de quarteirão, secretary of the neighbourhood -secretário do bairro, who inform and mobilise community members to participate.

According to the participants, the group of technicians attending the meetings usually forms a multisectoral team6 aligned to the events expected or occurring in the visited community: droughts, floods, plagues, or other issues. In the context of disaster preparedness, most of the meeting agendas are prepared by the formal entities. However, local leaders have an assured prestige toward them. They are crucial to endorse the officials’ presence in the community, influence the topic of discussion, express their viewpoints, and connect the externals to the local participants, a kind of customary protocol when outsiders visit local communities. As discussions unfold, other community members are heard, and all stakeholders structure responses to the matter under discussion. Nonetheless, the local perception of the Government’s or other formal entities’ presence to co-create climate adaptation knowledge differs in two groups: (1) those who continue to live in vulnerable areas (Chokwe town and Conhane) and (2) the relocated to safe zones (Chate and Chiaquelane).

Meetings with local populations in vulnerable areas

This group acknowledged the presence of formal entities to discuss climate related issues, highlighting mapping routes for evacuation; identifying sensitive areas or vulnerable groups in the neighbourhood; discussing cultivation practices and learning about early warning systems and the actions needed in each situation. For example, local populations and farmers helped technicians identify routes for water to reduce the damage along the irrigation scheme7, and farmers also coordinate watering dates and other strategies based on the interactions with the RBL managers in the irrigation scheme8. Other practices were also learned in the meetings with externals, as exemplified in the following quotes:

I heard about meetings in the neighbourhood. Community members and technicians exchange knowledge and design collective strategies to deal with the rainy season. Although I do not participate so much because I must work, I know from family members and neighbours that they talk about reinforcement of roofs, relocations, and other instructions [Community member].

The Government come here to teach us about rainfall forecasting and early warning systems. They also ask us questions, and we share our experiences; then we find solutions to specific discussions [Community member].

Meetings with local populations in safe zones

Except for a few households participating in farming projects, many who are relocated in Chate and Chiaquelane understand that the presence of the Government and partners is scarce. Formal entities’ visits to talk about droughts, floods, and alternative livelihoods were frequent in the first years after flood relocations. Currently, the few visits focus on the discussion of basic services: electricity, water supply, education, infrastructure, and so on:

When the Government visit us, we talk about the lack of water, electricity, and roads. Compared to the first years after relocation, now there are no frequent meetings. I do not see the government or others talking about climate change here. We just hear the local leader motivating us to cultivate in our yards or in the lowland, for those who have access to it [Community member].

These days, I do not see the Government or other organisations. Last year, we discussed with them about excavating wells to overcome water scarcity. They asked about our opinions, and they mentioned little about climate change impacts [Community member].

Co-creation focused on farmers

The discussion about everyday conversations and community meetings with externals has shown that knowledge co-creation covers several issues affecting local populations. However, since agriculture is the main economic activity in Chokwe and climate change affects the farmers’ practices, the next section focuses on two platforms: field visits and farmers’ training programs.

Field visits

Field visits are frequent when there are resources from projects or other Government programs targeting farmers, mainly organised in associations. So, the platform permits the participation of farmers who remain active in their association or farming in certain areas covered by the interventions. And it is mostly coordinated by extension workers and/or NGO technicians. The gender patterns in field visits are like other platforms discussed previously; women are actively engaged in knowledge co-creation to improve agricultural practices.

NGOs technicians and/or extension workers9 share SK produced in their formal settings, hear, and discuss with farmers about soil composts, barns, irrigation techniques, crop varieties and other strategies to increase productivity under climate change. In Chiaquelane, the resettlement area of 2013, participants of the project implemented by an NGO in the lower land highlighted that they discuss farming issues with the Government, NGO technicians, and occasionally with researchers10. Extension workers acknowledged that in the construction of improved barns using local materials, farmers indicate the best clay and covering material, then the technicians collaborate with them on how to build the storage. Other results from knowledge co-creation included the adoption of multiple use crops such as sweet potatoes, cassava, and cowpeas, and the production of bio-pesticides.11 With plagues and financial constraints to buy pesticides, the knowledge co-production of bio-pesticides is developing, as shown in some conversations:

We have a co-production of bio-pesticides. Local populations identify the plants to combat pests, and together with the extension workers and project technicians, we discuss the preparation and testing on farms [NGO technician].

I just finished preparing my bio-pesticide to spray on my farm. I used chilli and soap. I learned this in the workshop with other farmers and technicians working with us in a project [Farmer].

In co-production processes, not all results are expected to be consensual. For instance, to treat livestock pests, farmers prefer local knowledge, which involves the use of local plants, salt, and other techniques. This happens for two main reasons: farmers want to avoid contamination of cattle in public bathing areas, and modern products are expensive and mostly available in South Africa.12 In agricultural practices, some farmers avoid preparing the land in advance because it is very hot, the animals are weak, and the soil is harder. So, they prepare the land immediately with sowing during the first rains, as they consider that there is already grass for the animals, the soil is looser and cooler. Mulching is rejected by certain farmers, who informed extension workers that it hides rats, termites, and its execution (fetching and using grass) is labour-intensive.13 Meanwhile, farmers targeted by an NGO project in Chiaquelane had good results using sugar cane mulch to retain the moisture, protect the soil, as well as jack bean plants due to their biomass capacity and nitrogen fixation14.

In field visits, extension workers’ participation is seen in two ways: 1) Disseminate the knowledge co-created in a formal setting, which in turn is subject to co-creation with farmers, and 2) Use the results of co-creation with farmers to feed the formal co-creation knowledge process. However, the first one turns out to be problematic, particularly in the case of disasters such as floods. When events do not occur as predicted by the competent agencies and so informed by extension or disaster management workers, and/or due to the protocols to be followed, there are delays in informing farmers; limitations to co-creating knowledge with farmers occur. Subsequently, extension workers are seen as unprofessional and liars.15

Training programs

Training programs are another form of knowledge co-creation in the agriculture sector. They also require funds and commitment to contribute to local capacity building. So, they are usually linked to projects promoting school farms, workshops, farmers’ fairs, and other related activities. Regarding the participants, the interviews indicate that apart from formal entities, training programs mostly include leading farmers and farmers’ association representatives nominated by their peers. Although leading farmers come from local communities, they are trusted based on their prestigious position in farming practices; they are seen as role models with the power to stimulate other farmers’ practices. Leading farmers and farmers’ association representatives keep in contact with other farmers and ensure the replication of the experiences and co-produced knowledge during the training programs.

A leading farmer coaches circa thirty-five farmers in his/her community16, while a farmers’ association representative may work with over fifty active members17. Out of the five farmers’ associations we spoke with, three were led by women. We also found that women are working as leading farmers, as is the case with one of the research participants. However, when training programs require travel, some female leading farmers may stay behind. As we noticed in the dialogue conducted in June 2025, only male leading farmers were available and joined the event. In farmers’ associations, when the main representative cannot participate in training programs, the secretary, a member with writing skills and responsibility to document some association discussions, or other eligible members are likely to represent the group18.

To improve farming practices, several activities are developed in the training programs, some of them involving the Government. The projects foster interactive methods, allowing the collective assessment of agricultural practices before their replication. So, demonstration fields are part of the strategies to test the co-produced knowledge.19 Highlighting the co-production of bio-pesticides, the project implementer brings techniques based on other experiences and asks the participants to share their knowledge as well as to interact with peers in their communities to support the identification of native plants to be used in the co-production of bio-pesticides20. Then all the techniques (from the project implementor and farmers) are collectively evaluated, as extracted, for example, from an interview:

The project implementors gave us beneficial knowledge. For example, to produce bio-pesticides, they arrived with tobacco and other products just to show us their techniques in the production of bio-pesticides. Then, they asked us to come up with new and local ideas to produce bio-pesticides. In the workshops, we showed aloe vera, [Aloe barbadenis miller], “beijo da mulata” [ Catharanthus roseus - rose periwinkle], chilli, and other plants. We tested several local plants, around 30, and we moved on with those that showed efficiency. Now, as trained farmers, we have the responsibility to share with our fellow farmers [Famer].

Knowledge co-dissemination: rethinking the role of mass media

The co-creation platforms discussed in the previous section integrate different arrangements of co-dissemination, because when stakeholders gather to co-produce knowledge, they simultaneously co-disseminate it. So, co-creation fosters co-dissemination in horizontal, vertical, and transversal forms. The horizontal way is based on daily interactions among local populations. The vertical form involves hierarchies and formal, bureaucratic structures in which, for example, extension workers depend on information from other Government authorities to communicate it to farmers. The transversal co-dissemination focuses on various stakeholders through community meetings, field visits, and training programs, shaping broader discussions of climate change adaptation interventions. Yet, involved in structural and relational settings, knowledge co-dissemination can also be explored using other creative and outreach channels. Henceforward, we focus on how the media can be capitalised in co-dissemination as well as in the whole process of co-creation. A step toward such transformation is the appropriate use of audiovisual materials that integrate local, inclusive, and accessible language21. As an example, and an initial point regarding the use of media, local populations in Chokwe district acknowledged that in addition to knowledge shared through peers, technicians, phone conversations and other local arrangements, they heard Government agencies sharing disaster information on radio and television.22 However, this unilateral and vertical dissemination (from government agencies to communities) was not taken seriously. Local populations said that they doubted it and preferred to trust their experiences and local social networks. Some stayed in the risk area until they saw the threat, or until when through phone calls with people in the neighbouring areas, they heard about the water levels and direction. This is a somewhat recurrent local approach to deal with floods (see Christie and Hanlon, 2001; Domingos and Nagamatsu, 2024). Concurring with Atkinson (2023), people end up relying on social networks because the trust in the Government to respond timely and efficaciously to disasters is decreasing.

Collaborative learning and decision-making

Everyday conversations are seen as informal collaborative learning efforts expanding capacity building, social networks, and the sense of belonging to a certain group willing to construct social ties allowing them to have collective discussions, learning and actions for climate adaptation. Through these informal interactions, local populations led adaptation processes mostly through their local knowledge and lived experience. Meanwhile, local communities with opportunity to work with other stakeholders holding different knowledge engage in more diversified processes to collectively produce new knowledge. Yet, in all cases the decisions regarding the learnt issue are influenced by the evidence and trust that specific knowledge is applicable and grant good results. This approach posits decision making as an evidence based learning process, which more that trust issue is about reducing the risk and vulnerability that can emerge from such climate adaptation strategy. Other scholars argue that informal networks in knowledge co-creation are rooted in trust, collective memory, and action (Vergotine et al., 2025). Besides, factors such societal justice, creativity and collective encouragement support adaptation process which also rely on the “art of noticing,” where local communities become more attentive to their surroundings, related changes/challenges and contribute to adaptation strategies (Korsbrekke et al., 2025). This means that co-creation needs to increase stakeholders’ enthusiasm and awareness about the matter of discussion as well as to balance the mutual responsibilities’, also providing the feeling of ownership without an overstretching specific stakeholder (Abderhalden et al., 2025). Thus, to allow ownership and follow up actions coming from collaborative learning processes, we need to ensure that different stakeholders are fairly involved and their views and practices are observed and discussed for expected policies and actions (Matti et al., 2022).

Collaborative learning shaping institutional frameworks

Agenda setting, spatial and time based disparities

When formal entities have more space to define the agenda for community meetings, regardless of the relevance given by community members, this reveals power dynamics denoting their dominance in the issues to be discussed. Government and partners are somehow present in local communities, but with different agendas. These perceptions might create a divergence in co-creation processes involving Government agencies. On the one hand, the expansion of climate knowledge co-production and co-dissemination in risk areas helps people sustain their lives in these areas, while they benefit from the available services, infrastructures, and irrigation schemes. However, the frequency and intensity of disasters continue to expose them to risk, and some of them have already experienced disasters and have been offered land in safe zones. On the other hand, in safe zones, there is a shift in the agenda to focus on the provision of basic services. So, a failure to co-produce and co-disseminate knowledge for climate change adaptation in safe, but “bare” and hard zones for agriculture and alternative livelihoods is perceived and creates vulnerabilities. Consistent with Eriksen et al. (2021), arguing that adaptation interventions may intensify, rearrange, or produce new forms of vulnerability. Consequently, an open question emerges: Does not this approach (occurring for decades) foster cyclical vulnerability of both groups, while postponing and/or increasing the costs for adaptation interventions in Chokwe district?

Communication, trust, and institutional effects

Trust issues regarding the role of extension workers in disaster contexts push some farmers to doubt or slowdown in following the procedures (e.g., remove the equipment, leave the area) unilaterally defined and recommended by formal entities. This concurs with Sacyaten (2025) work, asserting that in a disaster context, delays due to inconsistent or minimised information, absence of preparation and time consuming responses reduce community trust, engender hesitancy, and weaken community obedience. Additionally, delays and the accuracy of information during a disaster are not new in the context of Mozambique. As an illustration, assessing the case of Licungo in 2015, a lack of accurate information about floods, failure to inform the magnitude of the event, and impacts were observed (GoM et al., 2015), and the technical problems with community radio, lack of electricity to charge phones and receive information from the local disaster committees aggravated the loss due to disasters (Domingos and Nagamatsu, 2024). Thus, in climate knowledge co-creation, we look at time management as a matter of protection, where a failure can be interpreted as a mistake or unforeseen situation. But also, as a setting for information vacuum, increasing vulnerability, and attempting to expand control over local populations through imposition of “last-minute” actions to be taken out of a co-creation process involving them or not discerning authentic needs. Thomas and Warner (2019) discuss related issues, introducing the idea of “weaponising” vulnerability. They argue that some settings aggravate uncertainty by disallowing people indispensable and suitable support, while resources are directed to groups in conditions to respond to climate hazards. Importantly, these situations in Chokwe may also indicate some limitations in the current governance frameworks, where possible reform may foster adequate resources mobilisation, responsibility and actions of key institutions expected to be involved and coordinate activities with extension workers. Then, clarifying institutional processes and related responsibilities is crucial because comprehensive framework of the context, publicly inclusive and robust technical approaches are key in climate governance (Hermelin and Gustafsson, 2024; Sanz et al., 2025). Besides, it is relevant to highlight that when stakeholders or institutions perceive that their work require minor engagement, or their role turnover is rapid, there may be obstacles for institutional learning, which later affect the engagement and acceptance of local communities (Pickard et al., 2026).

Translating disagreements into extensive collaborative learning opportunities

Farmers’ claims regarding some disagreements on practices debated in co-creation processes are genuine. But the lack of consensus should not impede co-creation; rather, it is a prospect for further knowledge co-creation. For instance, farmers arguing that mulching techniques are not working can have co-creation events with farmers holding different perspectives and results in neighbouring communities. Hence, the interactions involving Government agencies and NGOs also need to maximise collaborations serving as a follow-up on the challenges and lessons from different interventions. Regarding other claims on the sowing period, the local knowledge cannot be ignored, since farmers have their own valuable experience. However, with the current climate uncertainties, altering farming strategies, the decision on utilising certain farming practices or get weather and seasons information should also be taken considering the triangulation of local knowledge with scientific experiences, such triangulation is likely to produce more reliable solutions. Overall, co-creation based on field visits provides the opportunity for collective learning, living the reality in the field. This makes the platform crucial to well inform formal entities and create a room for stronger institutionalisation, clear policies, and a long-term approach in knowledge co-creation. For instance, Kehinde et al. (2025) use the term “bridging ties” to demonstrate that other stakeholders such as Governments and NGOs are crucial in knowledge sharing and discussion of collective actions. Furthermore, scholars indicate that farmers learn a lot through extension workers and peers, while other prominent actors such as researchers and media can also be involved in co-creation (Oladele and Gbayan, 2025). Thus, the most important thing in these participatory processes is the develop of collaborative learning opportunities that fits farmers needs and are acknowledged by all participants (Nataliningsih, 2016).

The time frame shaping instrumental and integrated collaborative learning

Knowledge co-creation involving local populations, Government, and NGOs fosters inclusive governance, but it still aims to mostly inform actions. Meetings are expected to be followed by specific actions to respond to floods, droughts, or other disasters. So, our criticism emerges from the perspective that such an approach may be effective in the short term since it creates pressure to actively respond to climate disturbance. Yet, in the long term, questions arise about structural issues in strategic, sustainable, and accountable adaptation programs. This may reveal a form of collaborative learning known as instrumental learning, which parenthetically emerge from stakeholders’ interactions, without a consistent plan to follow up (Sommerauer et al., 2026). Meanwhile, under specific project, knowledge co-creation open opportunities to register the good practices usually disseminated horizontally among farmers, vertically or transversely through other activities involving Government agencies, NGOs, and local communities. So, projects engaging public-private stakeholders can inform not only immediate actions, but also medium or long-term practices, as well as policies, capacity building initiatives and development of future adaptation projects. Besides, having most of the targeted farmers organised in specific groups, their assessment becomes more accessible and relevant in stakeholders’ agendas. For this form of collaborative learning process ensuring a considerable timeframe and follow up, Sommerauer et al. (2026) use the term integrated learning process, which has its concept and progressive monitor attached to possible impactful results.

The findings about project related activities such as training and field activities (which focus mostly on farmers groups) support the arguments that integrated learning interventions generate positive outcomes. For instance, Covane et al. (2026) argue that training about climate change and farmers’ association membership play a significant role for climate change adaptation in regions of Mozambique. Nonetheless, these outcomes also involve limitations and entry points for additional reflections regarding the local adaptation initiatives. Many stakeholders (Government and NGOs) frequently depend on donors funding, for example the Non-Government entities interviewed for this work, spoke considering the experiences from specific projects. Likewise, farmers associations and leading farmers who participated in the study were mostly supported by technicians or extension workers contracted through projects. Then, although local communities gain and produce knowledge with these stakeholders, once the project is closed there may be limitations for continuous, flexible, documented, and inclusive process of co-creation. One can say that the projects may provide tools/ capacity building to endure its impacts in the communities, and clearly local communities may also be engaged to innovate and foster co-creation trough the informal platforms, as done in everyday conversations. However, we also need to acknowledge that the withdrawal of these formal entities clearly creates a void and subsequent challenges specially, under rapid climate change impacts influencing the use of the existing knowledge systems. Along this, the acquisition of resources to implement some initiatives or projects may also take time and face limitations based on local, national, and international agendas. As Wainman et al. (2026) claim, more inclusive collaborative initiatives demand stronger plans, fundings for long term programmes which also enable trustful, shared, and productive relationships.

Overall, centralisation, governance fragmentation, resource limitation, knowledge gaps and policy disharmony throughout sectors and governance levels, as well short term focus and limited evaluation mechanism remain as challenge for successful collaborative learning, but multilevel governance systems with local, national and international efforts are crucial to foster resilience, long term frameworks as well as to ensure the measurement of social, ecological and economic impacts of co-creation processes (Yulianto et al., 2025; Jaisridhar et al., 2025). So, more public-private alliances and programs are expected to foster institutionalisation of collaborative platforms, resource mobilisation and monitoring of local solutions (Daza and Coq, 2021).

Inclusive adaptive governance

Gender dynamics require more efforts to challenge some social barriers

The unveiled gender dynamics indicate how exposure to risks fosters women to take proactive actions to respond to climate change impacts. With climate change, women have more responsibilities; mostly in household and agricultural/ other local livelihood activities. While everyday conversation and initiatives which do not demand for example travels, ensure, and expands women sphere of influence and engagement in civic discussions, some rigid social norms may still impose limits in their participation in other settings. For instance, when some women cannot travel to attend training programs, the restriction to directly interact with other peers and formal entities can to some extend influence how women understand and apply certain knowledge. Along this, if some female leading farmers end up delegating, for example, a male peer to participate in training, it is still appreciable in their leadership spectrum. But if the local setting is not strong enough to promote gender equality, female leading farmers may also lose their legitimacy and influential role in the area, because somehow the first and direct contact with formal entities in training programs, reveal certain privilege and legitimacy to talk to other farmers. Thus, prior training programs or other related initiatives, adaptive governance frameworks (including public and private sector programs) should engage in understanding the social dynamics of the local populations and address all the barriers for inclusive co-creation processes. Such barriers could be for example, multiple and burnout roles, household dynamics requiring male decision regarding travels and so on. By addressing these complexities, it is expected that collaborative learning process will promote equitable initiatives which empower women, promote gender equality, and actively hear the voices of women who are more vulnerable to climate impacts and own large experiences and insights that contribute to local climate change understanding and adaptation. For instance, in agri-food systems women participation has been crucial due to their knowledge, skills and role within their households and communities (Bryan et al., 2024).

Moreover, understanding and following these dynamics is especially relevant in the context of Chokwe where, historically, migration and HIV/AIDS have influenced the absence of man, and many households head have been women (O’Laughlin, 2009). Disasters and external agencies presence also induced the changes in gender relations, with the understanding that continuous educational and capacity building opportunities would support the effective and active participation of women (Wiles et al., 2005). Furthermore, previous studies demonstrate that the limitation to include women in issues behind household related issues is likely to exclude them and their children from harnessing the climate adaptation initiatives (Lestari, 2017). Capacity building, inclusive policies, participatory governance, gender transformative tools and sectorial collaborations are necessary for fair resilience, and this require the remotion of structural barriers and evidence that policies acknowledge the needs of the diverse stakeholders and gender aspects for inclusive and equitable solutions (Adhikari and Ghimire, 2025). Thus, ensuring that gender related limitations and vulnerabilities are understood and addressed in climate adaptation initiatives influence sustainable and resilient responses (Balabantaray, 2023), and foster inclusive governance frameworks, recognising women as leaders with effective and justifiable responses influenced by their experiences (Biswas and Barua, 2025).

Inclusive governance trough media as a co-creation

What happens in Chokwe reveals how pertinent it is to use media (mostly television and radio) not just as a channel to vertically disseminate specific knowledge, but rather to transform it into a trustworthy participatory outreach channel. This research acknowledges that different knowledges demand appropriate assessment and use of specific tools for dissemination. Then, collaborative mechanisms, socially inclusive and interactive debates also showing the local realities (based on specific projects, initiatives, or other situations) and discussing both scientific and local knowledge, can spur climate knowledge co-dissemination as well as a cyclical co-creation platform with large participants influencing adaptation policies. So, without abandoning the traditional media, which still relevant mostly for rural communities, this study proposes the combination of the platforms and promotion of more interactive or participatory approach in substitution to the unilateral communication. For instance, in Mozambique some of these traditional platforms are already combining their transmissions with the use of their own social media, either for lives or just sharing some video/audio excerpts. Then from the perspective of collaborative learning and adaptive governance this combination can enlarge the involvement of local populations and valuing their knowledge in media debates with other key stakeholders, and it is more likely to strengthen the relationship between formal entities and local populations, as well as the designed actions to be taken in a disaster context.

Furthermore, it is relevant to explore these combinations in collaborative learning because, local populations are already demonstrating frustration to the usual information sharing process done unilaterally, and the use of social media is becoming more visible and challenging the traditional media. Other studies argue that the top-down communication and governance structure generate barriers for inclusion, and a co-creation approach is crucial for risk communication, collaborative learning, and trust building (Aliska et al., 2025). Additionally, when several channels are strategically used, considering the context and diversity of the audiences, the available information is comprehensive, and when communities are involved in dissemination there is a strong stimulus for community consciousness (Palttala et al., 2012; Abunyewah et al., 2020; Sacyaten, 2025). Social media is putting pressure to traditional platforms and is now playing a relevant role in several climate communication dimensions such as sharing information, mobilising people, or promoting dialogue (Schäfer, 2025). Hence, leveraging the use of technologies it must foster capacity building, flexibility, interactive learning, and decentralised decision making (Yulianto et al., 2025).

This research was set to understand the processes of knowledge co-production and co-dissemination referred to as co-creation. The analysis shows an amalgam of stakeholders and interests from different levels. At the local level, in everyday conversations, local populations engage in casual conversations and neighbourhood meetings to learn from each other about forms of adapting to climate change. In these conversations, nearly everyone participates, although people with vast experience and proven successful techniques are more likely to be approached by their neighbours. So, a horizontal knowledge co-dissemination process also occurs in this informal co-creation platform. From the central level, formal multisectoral meetings lead to community meetings. All community members are invited to the meetings with formal entities, aiming to discuss climate events. Additionally, field visits and training programs targeting farmers permit the engagement of farmers and technicians, either from Government agencies or Non-Government Organisations to co-produce and co-disseminate farming solutions. Vertical knowledge co-dissemination is observed in settings where, for example, extension workers depend on specific Government authorities’ information to talk to farmers. Then, transversal co-dissemination occurs in platforms involving stakeholders from different institutions and sectors: community meetings, training programs and field visits. The study also argues for the transformation of media (mostly television and radio) into interactive and mixed communication combined to social media as a form to embrace more participatory, expansive and outreach to foster collaborative learning in climate change.

The analysis indicate that power structures are dynamic and influence the outcome and uptake of the co-created knowledge. Women are now actively engaging in climate change discussions, which alters the previous social structures where men dominated civic participation. Yet, in some circumstances, social norms prevails and women need male authorisation regarding some decisions, meaning that the stakeholders involved in co-creation initiatives still need to understand and address the remaining social barriers. In community meetings with external entities, usually the formal (outside) entities bring the agenda, which shows a certain dominance in the outline of the matter of discussion. Yet, these entities also depend on customary deliberations from the local leaders to be connected to the local populations. Moreover, the participants in the training are leading farmers and association representatives, people with certain prestige among their peers or within communities. This reveals a certain power pattern defining who has direct contact with the formal stakeholders and leads the interpretation, testing, and dissemination of the co-produced knowledge, so who join these training programs have more chances to be respected and heard by the peers. While extension workers play a great role in knowledge co-creation, which yields appreciable results in agricultural practices. When disasters strike, the bureaucratic steps expose power dynamics. The extension worker’s role becomes almost invisible as the disaster management authorities and other stakeholders, without frequent and direct engagement with farmers, take control. Concomitantly, delays in sharing information with farmers and other local actors become a constraint. So, from farmers/local populations, trust issues emerge and interfere with extension workers’ efforts. Extension workers experience reveals how disaster institutions need to reflect on the current governance frameworks and ensure more engagement with local communities. Furthermore, the disagreements coming from some collaborative process are pointed out as crucial factors to expand co-creation process involving more stakeholders with different expiries or results regarding the matter of discussion.

Overall, co-creation platforms involve an adaptive co-management approach where different knowledge systems are discussed to define adaptation solutions. Either in formal or informal collaborations, the decisions taken by local populations, mostly depend on their trust to the outcome and evidence that it works. Here, both top-down and bottom-up perspectives are involved, since knowledge from outside/formal actors reaches local communities and vice versa. Nevertheless, this is not empty of power struggles, exclusions, and other social dynamics. Therefore, with the discussion regarding the institutional learning, adaptive governance, and collaborative learning process, we conclude that co-creation initiatives should not be limited to design specific and immediate actions, but rather need to forge structural transformations, engendering solid partnerships, strategic, sustainable, longer, inclusive, and accountable adaptation programs.