

Barley, mainly two local landraces, Arabic Abiad and Arabic Aswad, covers about 1.5 million hectares of cultivated land in Syria. However, the average yield from these landraces is less than 1 t/ha. Barley breeders have developed several promising new varieties that can increase yields by an average of 20% over local landraces, without the need for additional inputs. For these new varieties to be successfully adopted, the seed should be easily and widely available to farmers.
Scientists from ICARDA assessed the role of farmer-to-farmer seed exchange
in the distribution of new varieties by tracing the flow of seeds from 52
farmers in 24 villages in Syria. Barley breeders supplied each farmer with
100-200 kg of seed of the new varieties in the 1994/95 season, and traced
the ensuing distribution for five years. The researchers also examined the
reasons for farmers acceptance of promising new barley varieties in
different agroecological zones, and the extent to which farmer-to-farmer distribution
of the seed was an autonomous process.
Most of the farmers involved in the study had already collaborated with the ICARDA barley improvement program, either through on-farm trials or by attending field days. Five promising new barley varieties (Arta, Rihan, Zanbaka, Tadmor, and WI 2291) were distributed to them in the first year. Some farmers selected more than one variety and others chose only the one they considered most suitable for their environment. Farmers grew the new varieties in the same way as their local barley and without any supervision from ICARDA or national extension agents.

On average, farmers sold
60% of grain as feed, kept 13% to feed their own sheep, and retained 7% as
seed for the next season. They sold only 3% as seed to other farmers. Barley
is therefore regarded principally as a cash crop in this area. In the markets,
grain from the new varieties fetched similar prices to that from landraces
when sold as feed, but farmers were able to obtain higher prices when selling
it as seed.
The high adoption rates achieved in this study illustrate the importance of farmer participation in evaluating new varieties and distributing seed. Farmer-to-farmer seed transfer appears to be a viable option for the dissemination of new varieties of cereals such as barley, especially when seed companies have been unable to meet local demand. Community-level seed technology needs to be developed to guarantee quality and to establish trusted local seed experts who can ensure a constant flow of new germplasm into communities. Such experts could be key partners in participatory germplasm improvement programs.
After five years, seeds of the new varieties had spread to 60 villages (Fig. 24). The process of technology transfer is summarized in Table 18. The total number of farmers monitored over the course of the study continued rising to reach 206 in the last year, but the number receiving new seeds each year declined after reaching a peak in the 1996/97 season.
Growers were divided into two types: new growers, planting the varieties for the first time, and adopters who planted the new varieties again after evaluating them. The adoption rate peaked at 75% in the second year and then declined slowly, with no

