Management of genetic variability in rice (Oryza sativa L. and O. glaberrima Steud.) by breeders and farmers in Sierra Leone

M.S. Jusu

Research output: Thesisinternal PhD, WU


<p>The new kind of relationship between crop scientists and farmers in difficult environment and the idea of participatory crop improvement is documented in a sizeable body of literature. The quantitative data on how these two groups work for the mutual understanding and co-operation is, however, limited. The main issues considered in this thesis include how close or far apart are breeders' and farmers, how do different groups of farmers select, or otherwise manage their rice planting materials, what is the morphological evidence for "intermediate" rices, and what are the opportunities for geneflow under farmer management of rice varieties? The focus on farmer selections in the <em>O. glaberrima</em> species, or selections morphologically "intermediate" between <em>O. glaberrima</em> and <em>O. sativa,</em> is a common thread running through the various chapters.</p><p>The first part of this study introduces small-scale rice farming in North-western Sierra Leone in three Chiefdoms (Kambia, Tonko Limba and Bramaia) and considers the history of rice farming and rice research in the area, discusses ethnicity and the history of slavery as factors in shaping the attitudes of different ethnic groups to their rice varieties, and outlines the main production ecologies for rice farming in the region.</p><p>The second part looks at how these farmers manage seed. It discusses the organization of farm labor, farmer seed selection and development activities, the role of gender in seed processing, and considers all the stages in seed processing through harvesting into storage. It concludes with comments on farmers' knowledge of varieties and soil types, and information about the original sources of farmer varieties. What farmers consider important in selecting rice varieties was investigated. A detailed study of the genetic diversity in farmers varieties including varieties of <em>O. glaberrima</em> and <em>O. sativa</em> , together with "intermediate" types using univariate and multivariate methods. Farmers varietal management was also investigate to understand why some Susu (one of the ethnic groups studies) farmers mix varieties before seeding. The last part focuses on flowering in rice, as the crucial "window of opportunity" for natural outcrossing be institutionalized, so that breeders and farmers work in partnership. Potential biotechnology developments may make farmer selection skills more, not less, important.</p><p>The results revealed that farmers in the three case study chiefdoms mainly produce annual crops. Farmers in Magbema Chiefdom have significant holdings of cash crops. And those in Bramaia farmers earn cash from pepper and groundnuts. Rice remains by far the most important food crop, and this is reflected in level of varietal diversity. Traditional grains (sorghum and Digitaria) continue to supplement rice, but there has been a significant spread of introduced roots and tubers (cassava and sweet potato) varieties in recent years. Science has had more impact on roots and tubers than on a deep knowledge of a crop grown for millennia. Farmers in the research mainly grow rice on the upland-swamp continuum and have more lowlands than upland varieties. Farmers report dependence on the wider regional system for rice germplasm. Magbema farmers tend to get new local varieties from Tonko Limba and Bramaia. Tonko Limba and Bramaia farmers tend to get new wetland varieties from Magbema, an area with a long history of wetland (cultivation in mangrove blackish water) and associated (fresh water) swamps.</p><p>Rices from species <em>O. glaberrima</em> remain important in the region. The proposed 'intermediate' (farmers selected inter-specific hybrid) rice, <em>Pa Three Month</em> is A main variety. Farmers varietal development and seed management was variable according to gender and ethnicity. There was also a variation in the ways farmers saved seed are planted. There was a clear interaction of genotype, environmental and socio-cultural interaction in the management of rice varieties by farmers in the study area.</font>There was a lack of correlation between farmers and scientists selection criteria indicating divergences in the way farmers and researchers view selection decisions. There was considerable diversity in local rices, suggesting farmer selection is not yet "bottlenecked". <em>Pa three month</em> clustered independently of the main <em>O. glaberrima</em> and <em>O. sativa</em> cultivars, suggesting there is something distinctive about this farmer selection. A dendrogram from the multivariate analysis clearly separated <em>O. sativa</em> and <em>O. glaberrima</em> materials, and differentiated those <em>O. glaberrima</em> varieties farmers consider to be "traditional" and "new" (including <em>Pa three month</em> ).</p><p>The results confirmed that farmer indigenous knowledge as apparent in rice selection is well grounded. Marker analysis should now be used to explore these materials further. The results also confirm that the Susu farmers who mix rice varieties before seeding varieties before seeding on the uplands is advantageous. The most important finding was that RYT is larger than unity when <em>O. sativa</em> and <em>O</em> . <em>glaberrima</em> varieties are grown in mixtures under upland condition, without fertilizer application. There are also benefits in terms of out-crossing and subsequent farmer selection. The result on flowering in rice, as the "window of opportunity" for natural outcrossing reported that flowering intervals in rice vary from c. 5 to 15 days, and that this variation is to some extent under genetic control. <em>Pa three month</em> is a variety with one of the longest flowering periods. Plants with long flowering period presumably have some advantage in terms of chances of out-crossing.</p><p>Conclusion from this study of farmer management of rice genetic resources show that farmers have a wide range of planting material but need outside input in terms of materials and knowledge. The central experiment in the research reported above was to get farmers to select across a range of their own and researchers' materials. Despite the participating farmers being among the world's most impoverished people working in highly insecure war-time conditions it was clear that they knew what they were looking for. Likewise, they could always explain the selection choices they made. There was considerable agreement among the farmers about what they chose and why, even though these choices would not necessarily agree with researchers' choices. But there was also considerable variation by ethnic group and environment, despite the three groups having lived together in one rather small region for perhaps many hundreds of years.</p><p>Farmers have their own indicators of performance and quality, not well anticipated by breeders' criteria, but despite this international high-yielding ( <em>sativa)</em> material proved to be the single most frequently chosen class of material. Results make clear that low-resource farmers like variety (a range of options) and not just a few high performing varieties. By exploring the context of farmers' choices this thesis has made clear that breeders cannot afford to ignore local cultural, historical and environmental circumstances. Factors as various as exposure to the slave trade and inter-regional commerce, desiccation, soil impoverishment, G x E interaction, and variations in modes of labor mobilization and farming style (e.g. when, how and by which genders harvesting and seed purification are managed) all tend to account for why a sample of male, upland-rice-farming heads of households in three contiguous and small chiefdoms tends to diverge (but systematically) in their selection responses to a single body of trial material. In relation to low-resource farming, the devil, as the saying goes, is in the detail.</p><p>In general, morphological analysis supports the idea that there may be considerable variation in the local rices, both within and between species groups, perhaps as the result of the meeting of different streams of material where coastal and interior trade routes have converged, and that the potential for selecting upon, or breeding from, this local material is still not exhausted. One of the vehicles for mixing and merging among these convergent local gene pools may be the habit of some Susu farmers to interplant distinct varieties in the same field in the hope of spreading risks and ensuring supplies of nutritionally favored varieties. This may also provide the incidental basis for considerable natural out-crossing upon which farmers subsequently select. More generally, however, the thesis confirm considerable potential for gene-flow among local rices, and between local rices and improved cultivars in upland rice farming in Sierra Leone, even if more detailed experiments are now needed.</p><p>Farmers have adopted, and continue to show interest in RRS varieties, but mainly the pure-line selections from local varieties introduced during the 1970s. Much could still be done to extend access to these varieties beyond the immediate station vicinity (farmers no more than 75 km from the station were still not familiar with these releases even 25 years after their first introduction). But much more worrying, farmers hardly showed any interest in the current advanced lines from RRS. International (including WARDA hybrids) and local materials both attracted more interest in the selection trials.</p><p>Comparing "ideotypes" suggested quite a gap between farmers and national rice professionals. An under-funded national program has been almost totally dependent for funds on the international programs led by IRRI. Loyalty to one's patron is a marked characteristic of social and political systems in Sierra Leone (Richards, 1986). But perhaps the distantly formulated IRRI Standard Evaluation criteria are too rigidly applied?</p><p>What seems now to be needed to revitalize the program for low-resource upland farmers is the further extension of the kind of methodology explored in this thesis. Breeders need to learn that their varieties sometimes do good by stealth - that releases many times "disappear", but perhaps not without first enriching the local gene pool, and that this boosts farmers' confidence in their own selection practices. Breeders, therefore, need both to select and cross, but also to monitor farmers' selection processes, and adjust according to the lessons they can derive from such monitoring. They also need to learn the lessons from farmers' experiments with mixtures. In low-input conditions choosing complementary mixtures of <em>O. sativa</em> and <em>O. glaberrima</em> (the approach of some farmers) may be more fruitful than an inter-specific hybridization program.</p><p>But to pick up on such lessons some institutional innovation is needed to make farmers' selection experiments a regular part of an interactive process binding together farmers and researchers in a necessary and complementary relationship.</p><p>Donors will have to adjust their research support and assessment processes accordingly. They will need to channel more funding through decentralized initiatives such as the Community Biodiversity Development & Conservation program. But they will also have to consider new approaches to monitoring and evaluation. Sending an economist to track innovations may be less important than it once was. The new approach may be better based on marker probes to track where the useful genes have gone, and who makes use of them, in farmers' practice as well as in station releases.</p><p>Whatever the case, this thesis makes plain that farmer agency is a resource not to be lightly disregarded in plant improvement. As we move into a new world shaped by radical bio-technologies such as apomixis it may transpire that there is much more scope for alternatives to the top-down Green Revolution programs planned and regulated by far-sighted but remote international scientists. There may be much more scope for the tailoring of plants to local social and environmental need. If so, we may come to regard the farmer's eye for selection and experimentation as crucial skills enabling this to happen. Perhaps one day farmer knowledge itself will become a candidate for a Nobel Prize.</p>
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Richards, P., Promotor, External person
  • Stam, P., Promotor, External person
  • Elings, A., Promotor, External person
Award date21 Dec 1999
Place of PublicationS.l.
Print ISBNs9789058081827
Publication statusPublished - 1999


  • rice
  • Oryza sativa
  • genetic variation
  • plant genetic resources
  • farmers' attitudes
  • plant breeding
  • Sierra Leone


Dive into the research topics of 'Management of genetic variability in rice (Oryza sativa L. and O. glaberrima Steud.) by breeders and farmers in Sierra Leone'. Together they form a unique fingerprint.

Cite this