Participatory Barley Breeding for Improving Production in Stress Environments

O. Kafawin, H. Saoub, S. Ceccarelli, Y. Shakhatreh, A. Yasin, A. S. Grando, A-R. Bwaliez, A. Khazaleh


Farmer participation is a new approach to barley breeding in Jordan, besides the conventional barley-breeding program. One cycle of decentralized participatory selection was conducted in eight locations in Jordan. These were: Khanasri, Ramtha East, Ramtha West, Ghweir (Station), Ghwair (Field), Rabba, Muaqure and Mohay. A set of 200 barley entries was planted during November 19-26/2002 in seven host farmers’ fields and in one Research station. These entries included fixed lines and segregating populations, in addition to two checks; one improved six-row released variety “Rum”, and a local check unreplicated in one research station and in the seven fields of host farmers. In each location, the plots were arranged in Incomplete Blocks Design of ten plots each, and the layout was ten rows and twenty columns. This allowed both an Incomplete Block Design analysis as well as a bi-dimensional spatial analysis. The local check yielded more than the improved variety Rum in 5 locations, but less in 2 locations. However, the differences were not significant. In four of six villages there were several lines outyielding the local check. There were large entries x locations interaction effects that were more than four times larger than entries effect. This confirms that the classification of entries varies greatly with the location. There were large differences between the correlation coefficients both between locations and within locations. Both breeders and the farmers generally selected for taller plants. However, in Ghwair there was no relation between the farmer’s or the breeder’s score and plant height, while in Rabba the breeders selected tall plants and the farmers selected short plants. The correlation coefficients between the breeder’s scores and grain yield and biological yield were not always significant, and were often lower than those relative to the farmer’s scores. Kernel weight was much more important as selection criterion for the farmer than for the breeder, and more important than spike length, which was strongly correlated with farmer’s score in Mohai and Ramtha. Eventually, the correlation coefficients between farmer’s and breeder’s scores varied from highly positive and significant, such as in Khanasri, Mohai, and Ramtha, to non significant such as in Gwheer and Rabba. During this first cycle of selection, farmers show a modest preference for two row types, which increased their frequency from 70.7% in original population to 74% among the selected entries, and a strong preference for white-seeded types which increased from about 66% to 83%, while the frequency of the black seeded types was reduced in half.

الكلمات المفتاحية

barley, Hordeum vulgare, genotype by environment interaction, farmer participation, participatory plant breeding

النص الكامل:

PDF (English)

المراجع العائدة

  • لا توجد روابط عائدة حالياً.