Fewer crop species, fewer cultivated varieties per species, less diversity within each variety are three "symptoms" of the erosion of biodiversity in cultivated plots. It is commonly assumed that the massive use of improved varieties instead of local varieties and anthropic pressure are the prime culprits behind this impoverishment. Is this a cliché or reality? The only way to find out is to monitor the changes in crop genetic resources. This has now been done for local rice varieties in Guinea, a reserve country for the genetic diversity within the two cultivated rice species: Oryza glaberrima from Africa and O. sativa from Asia.

Are local rice varieties disappearing? What strategies are required to conserve them? CIRAD and its African partners have been working in Guinea since 2000 to find answers to those questions. Their research has been conducted on several levels.

Stable or even slightly greater diversity

On a national level, the researchers inventoried the common names of the varieties used by farmers between 1996 and 2001. This meant surveying almost 1700 farms in 79 villages. Furthermore, in 2003, samples were collected from six villages in Maritime Guinea and compared, using molecular markers, with samples taken by a survey mission to the same village in 1980 and kept in cold storage in Montpellier ever since. The results obtained ran counter to the alarmist vision of genetic erosion. The number of rice varieties and genetic diversity were stable, or had even increased slightly. Since 1996, when improved varieties were introduced, the number of varieties, which varied from 4 to 40 depending on the village and the region, had increased by 10%. There had thus not been any loss of local varieties in Guinea.

The substantial varietal diversity observed is typical of subsistence agriculture: more than 80% of the varieties grown were local. Each village could thus allow for the range of prevailing agroecological conditions and different uses of rice. However, almost 90% of the varieties inventoried were only grown by a small number of farmers, and despite the observed diversity, these "minor" varieties are now under strong threat of extinction. Moreover, there was not only diversity in terms of the number of varieties, but also within each of those varieties. Each variety was the sum of a large number of pure lines, and the proportion of those lines varied from one farm to another. This "multi-line" structure can be put down to how the farmers manage their rice varieties, ie frequent exchanges and replacement of varieties and seeds, and cropping and seed production practices that favour genetic mixes and recombination.

50% of the genetic wealth of varieties in a village held on a single farm

As regards preserving the diversity within each local variety, in situ conservation on farms, which is compatible with agricultural development, looks like the only feasible option. In fact, it would be impossible to sample all the lines that make up a variety and keep them ex situ, for instance in a cryobank. The researchers working on the study thus characterized the varieties grown in Maritime Guinea, in two villages with contrasting production systems. To this end, they used descriptors combining common names and molecular markers (short DNA sequences). The results showed that a single village may hold the equivalent of 70% of regional diversity. On a more detailed analysis level, a large farm may hold 50% of the genetic wealth of a village. As a result, a small number of villages and farms is therefore sufficient to cover the genetic diversity of a whole region such as Maritime Guinea. This type of structure could eventually be extended to cover the whole of the country.