PARTICIPATORY EVALUATION AND SELECTION OF TRITICALE (X-Triticosecale Wittmack) VARIETIES IN MAO-KOMO DISTRICT OF BENISHANGUL-GUMUZ REGIONAL STATE, WESTERN ETHIOPIA

Abstract

Triticale varieties were recommended for production in Ethiopia following research undertaken for many decades. However, these varieties were not evaluated and introduced in districts of Benshangul Gumuz regional state having potential for the production of the crop. Therefore, the present research was undertaken to evaluate eight triticale varieties and assess the effect of locations and genotype x location interaction on yield and yield related traits, and to identify farmers’ preferred high yielding varieties. The field experiment was laid out in randomized complete block design with three replications at four locations in Mao-Komo district during the 2020/21 main season. Agronomic data were subjected to analysis of variance (ANOVA), mean comparison and genotype x environment interaction (GEI) analysis, while matrix ranking was used for data analysis collected about farmers’ preferences. Results from (ANOVA) for each location revealed the presence of significant differences among the triticale varieties for all agronomic traits except for days to maturity at Tongo, above- ground biomass yield at Tongo and Shoshor locations, and for harvest index at research station. The results of the combined ANOVA over locations showed significant effects of location on all traits except biomass yield. On the other hand, the interaction of variety x location was significant only on days to heading, days to maturity, number of fertile tillers per meter square and grain hectoliter weight. The combined analysis of variance over locations for grain yield showed that the mean square for interaction of genotype × location was not significant and GEI contributed 45.9%, 15.8% and 10.7% to the total sum of squares. The results indicated the observed variation among varieties for grain yield was due to the genetic differences and the effect of locations. The varieties had 3008.2 kg ha-1 mean grain yield over four locations in the range between 2604.2 kg ha-1 at Tongo and 3420.8 kg ha-1 at research station. Abdisa variety had significantly highest mean grain yield of 4110.3 kg ha-1 while Motti (2385.3 kg ha-1 ) and Dersolign (2544.7 kg ha-1 ) had significantly lower mean grain yield over locations than the other varieties. Farmers used disease resistance, grain yield, tillering capacity, spike length and plant height for selection, and the pair-wise ranking showed that grain yield was the top most selection criterion, followed by disease resistance and tillering capacity. In the participatory varietal selection, Abdisa and Dilfekar varieties were the most preferred varieties by farmers at all locations. Dilfekar variety had higher mean grain yield of 3322.9 kg ha-1 over location, and only exceeded by Abdisa variety. Therefore, these two varieties were selected based on agronomic data and farmers’ preference and could be recommended for production in the study areas.