Rural Development in Madagascar - Aga Khan Development Network
Aga Khan Development Network

Rural Development in Madagascar

In Africa, one of the greatest challenges to development has been food security - the ability of farming populations to grow enough crops to feed their families through the "hungry" season. Typically, during this time, families eat only once a day.

Rural development in MadagascarIn the PLAR programme in Madagascar, farmers were encouraged to build on their traditional knowledge as they experimented with the PLAR toolkit on a 10 square meter test plot.In 2005, the Aga Khan Foundation (AKF) launched an integrated rural development project in the Sofia region of Madagascar with the objective of increasing rice yields by 100 percent. In keeping with the Foundation’s overall vision of development – that the programme be participatory and driven by the knowledge, experiences and choices of the beneficiaries themselves – the programme AKF did not dictate a single formula for higher yields.  Instead, it provided farmers with a “toolkit” that included a range of possible inputs, such as different seeds and fertilizers, and training, including mapping, planning, water management and field preparation.

The programme is called Participative Learning Action Research in Rice Management (PLAR).  It differs from the widely emulated System of Rice Intensification (SRI), which appears to offer a “top-down”, universal technological solution to agricultural development. In contrast, PLAR relies heavily on negotiation and agreement among multiple stakeholders, each of whom may have their own objectives, perspectives and interests. In PLAR, the end result is not the only point of focus; instead, what can be learned from the processes of planning, negotiating and finding solutions is also closely followed.  PLAR is therefore a holistic, bottom-up approach to natural resources management and rural change.

In the PLAR programme in Madagascar, farmers were encouraged to build on their traditional knowledge as they experimented with the PLAR toolkit on a 10 square meter test plot.  Then they shared the results and lessons of the tests with other members of the group. Farmers subsequently adopted certain techniques on their own fields. Experimentation combined with local knowledge led to innovations that were better adapted to local conditions and abilities -- the essence of the participative approach.

M A D A G A S C A R    R I C E    F I L M S
The series of videos below, produced in a dialect native to the programme area, was used as a training tool for scaling up from its initial pilot project to a larger number of farmers:
Film 1 (11mns 37secs)
Film 2 (15mins)
Film 3 (15mins 49secs)
Film 4 (10mins 38secs)
Film 5 (14mins 37secs)
Film 6 (19mins 39secs)
Successful innovation then led to replication. Farmers from the programme observed that their neighbours outside the group started “borrowing” certain techniques, such as planting in rows.

In 2009, average yields were more than 300 percent higher in fields where selected PLAR practices were adopted than in neighbouring fields where they were not.  Typically, yields grew from 1.5 tonnes per hectare to 5 tonnes per hectare – without the use of chemical fertilisers.  Farmers who had adopted the technique were even saying that the higher yields had enabled them to store enough food to carry their families through the “hungry” season.  The programme is now being scaled up to reach a majority of the rice cultivators in the country.

PLAR-IRM is a people-centered – as opposed to technology-centred – approach that aims to build the capacity of farmers and other actors in rice cultivation to discover and improve agricultural methods that are most suitable for their specific environment. The step-by-step process is based on collective adult learning that leverages the experiences of group members, each of whom experiments with improvements to rice cultivation and management.

While the approach can be applied to other crops, PLAR-IRM is specifically adapted to rain-fed lowland rice ecology, which currently represents about 35 to 40 percent of the rice growing area in sub-Saharan Africa. Expanding rice growing areas in the lowland ecology -- an estimated 130 million hectares (320 million acres) are available across the continent -- is very promising. This ecology lends itself to intensification and diversification in areas where some form of water control is ensured and where social and environmental issues related to development are thoroughly understood.

Rain-fed lowlands are characterized by low-precision farming (limited water control, low fertilizer input, sub-optimal crop management practices, narrow windows for sowing and harvesting, etc.). The underlying reasoning for the introduction of PLAR-IRM is that in such complex, diverse and low-precision rice ecologies, farmers are not always served well by “narrowly defined” technologies.  Whereas these technologies may be suitable for high-precision irrigated systems, farmers operating in lowlands need a wide range of “flexible” options. In these cases, one size rarely fits all.  Farmers try out new methods, fine-tune existing techniques and then integrate both traditional and new practices to create a “best fit” of crop management practices.

Coastal Rural Support Programme, Kenya BriefClick on the image to download brief The PLAR-IRM learning approach therefore does not recommend a “package of technologies” to farmers.  Instead, it exposes farmers to problems and opportunities, allows a process of participatory group learning and consolidates this learning through the identification of alternative rice management options. Famers discover and learn together what works, what needs to be adapted and why. Together with other stakeholders, farmers also identify major constraints, priority areas and how to take a step-by-step approach to improving management practices. Later in the PLAR cycle, farmers may go beyond identifying improvements to the production cycle to processing and marketing of the rice product.

The PLAR-IRM approach starts with groups of 25 to 35 rice farmers, which are either formed (normally on the basis of a common bond, such as use of the same water source for crop irrigation) or in some cases they are based on already existing entities. During the first years the groups are supported by a programme facilitator who animates the PLAR learning and innovation sessions. These modules intend to develop tools for a common understanding of the rice plant and the natural resource base as well as the development of planning tools for crop management such as land preparation, nursery establishment, transplanting, soil fertility and water, weeds and pest management. Also, topics related to harvest and post-harvest practices are captured in the PLAR-IRM modules, including storage and negotiation skills. The PLAR sessions aim to strengthen farmers’ capacity to observe, analyse, interpret, make decisions, innovate and share knowledge and experiences to improve rice farming.

PLAR is based on indigenous knowledge, practices and skills, and discussion around existing rice management practices, and their logic or justification is the starting point for all modules. PLAR is a learning-by-doing approach and farmers are therefore encouraged to try out any new ideas identified during PLAR sessions on well-marked “innovation spaces” (i.e., a section of their plot reserved for new techniques, the size of which is determined by the farmer). This allows them to assess the impact of such innovations on their rice yield, or (given the price of inputs, labour time in particular) on the profitability of the new techniques. They can then decide whether or not to integrate the innovation into their current management practices. These innovation spaces are regularly visited as part of PLAR learning sessions, which allows for farmers to learn from one another and develop essential skills in observation and analysis.