The research, led by Clara Léonie Diebold of the University of Bern together with Peter Messerli, Paul Clément Harimalala, Sarah-Lan Mathez-Stiefel, and Julie Gwendolin Zaehringer, examines four conservation and development initiatives in Madagascar. These are the kinds of hybrid projects that have become ubiquitous across the Global South: interventions that promise to safeguard biological diversity while simultaneously improving local livelihoods, often funded by international organizations and implemented by NGOs working with communities. On paper, such projects almost universally pledge to be participatory and inclusive. What actually happens inside their decision-making machinery, however, has remained surprisingly opaque.

To pierce that opacity, the team constructed a novel analytical lens that combines three complementary frameworks. The first concerns participation itself, distinguishing between forms of engagement that genuinely share decision-making power and those that merely consult. The second distinguishes knowledge sources, separating local knowledge held by community members, scientific knowledge produced by researchers, and practitioner knowledge accumulated by project staff. The third distinguishes knowledge types: systems knowledge, which describes how a socio-ecological system works; target knowledge, which defines what goals and values the intervention should pursue; and transformation knowledge, which specifies how change can actually be achieved on the ground. By applying this triple framework across the full trajectories of the four initiatives, the researchers could trace how participation and knowledge inclusion evolved over time, rather than snapping a single static portrait.

The first major finding is a reason for cautious optimism. Across the four initiatives, local participation and the inclusion of diverse knowledge sources generally increased as the projects unfolded. Early phases, often dominated by external experts drafting plans, gave way to later stages in which community members were consulted, workshops were held, and local ecological insights were folded into project design. This trajectory suggests that the conservation sector’s decades-long critique of top-down fortress conservation has genuinely filtered into practice, at least in the Malagasy initiatives studied. Projects are, on the whole, opening their doors wider than they once did.

But the second finding cuts against any easy celebration. Even as participation broadened, it frequently remained consultative in character. In practical terms, this means local actors were asked for their views, their observations, and their traditional ecological knowledge, yet the consequential decisions about budgets, zoning, rules, and priorities continued to be made elsewhere. The researchers describe this as a risk of knowledge extraction: projects harvest valuable local information without granting communities genuine influence over what is done with it. A farmer may know precisely where the forest regenerates fastest and why households clear particular plots, but if that knowledge simply informs a plan drafted in an office in the capital or abroad, participation becomes a form of data collection rather than shared governance.

Underlying this pattern, the study finds, are deliberate choices made by project managers and funders. The degree to which local actors were involved and multiple knowledge sources were combined did not happen by accident; it depended on decisions taken by those holding the purse strings and the implementation mandate. This insight highlights two intertwined realities. On one side sit power asymmetries, the structural inequalities between internationally funded organizations and rural communities that no amount of goodwill automatically dissolves. On the other side sits agency: project leaders are not helpless cogs in a system. They can choose to design genuinely co-productive processes, and the study suggests many have room to do so. The same tools that can extract knowledge can, in different hands, distribute power.

The analysis of knowledge types produced perhaps the most technically striking result. When initiatives failed to integrate systems, target, and transformation knowledge, the consequences were not abstract; they materialized as significant setbacks in the projects’ trajectories. An initiative might possess excellent systems knowledge, a rigorous understanding of deforestation dynamics, for example, yet stumble because target knowledge was thin: the goals of conservation actors and local livelihood aspirations were never explicitly reconciled. Another might share clear objectives but lack transformation knowledge, the practical know-how about governance, incentives, and implementation that turns intention into durable change. The framework thus functions as a diagnostic instrument, allowing practitioners to identify precisely which kind of knowledge is missing before that gap derails an intervention.

Taken together, these findings elevate knowledge co-production from a fashionable buzzword to an operational necessity. Co-production, in the sense used here, means iterative processes in which scientists, practitioners, and community members jointly generate understanding, define goals, and design pathways of change, rather than exchanging information across a one-way pipeline. The authors argue that such processes must explicitly address power relations, engage plural worldviews, and deliberately weave together all three knowledge types. That is a demanding prescription. It requires funders to budget time and money for sustained engagement, managers to cede some control, and researchers to treat local knowledge as a partner rather than a resource. But the Malagasy evidence suggests the alternative, consultation without influence and integration without intention, carries real costs in project effectiveness and legitimacy.

The broader resonance of the study extends well beyond Madagascar’s forests and hillsides. Conservation and development initiatives now span the planet, from community forestry in the Himalayas to marine protected areas in the Pacific, and nearly all face the same governance dilemma: how to reconcile global biodiversity priorities with local rights, needs, and knowledge systems. International frameworks, including recent global biodiversity targets, increasingly recognize the importance of indigenous and local knowledge, yet recognition on paper does not automatically translate into influence in practice. By providing a concrete, replicable analytical lens, the study gives both researchers and practitioners a way to audit their own projects: Who participates, in what capacity? Which knowledge sources enter the room? Are systems, target, and transformation knowledge all present and connected?

For a field often polarized between those who prioritize nature protection and those who prioritize poverty alleviation, the study’s message is refreshingly constructive. The question, it suggests, is not which knowledge matters in the abstract, but how different knowledges can be brought together deliberately and equitably in each specific place and project. Where that happens, conservation and development initiatives stand a far better chance of achieving their twin promises. Where it does not, even well-funded, scientifically sophisticated interventions risk repeating an old failure mode: protecting landscapes while marginalizing the people who know them best, and who must live with the consequences of every decision made on their behalf.

Subject of Research: Participation and knowledge integration in decision-making within conservation and development initiatives in Madagascar

Article Title: Which knowledge matters? Decision-making in conservation and development initiatives

Article References: Diebold, C. L., Messerli, P., Harimalala, P. C., Mathez-Stiefel, S.-L., & Zaehringer, J. G. (2026). Which knowledge matters? Decision-making in conservation and development initiatives. PLOS Sustainability and Transformation, 5(5), e0000248. https://doi.org/10.1371/journal.pstr.0000248

Image Credits: AI Generated

DOI: 10.1371/journal.pstr.0000248

Keywords: conservation, development, Madagascar, knowledge co-production, participation, governance, local knowledge, scientific knowledge, power asymmetries, biodiversity, sustainability, decision-making