The present study was carried out at Gemmeiza Agricultural Research
Station during 200012001 and 200112002 seasons. Diallel cross excluding reciprocats
among nine parents of wheat namely: Peg"s"/I HD22061 Hork's· (P1). CeTIIA (P2).
KVZ I BJ'rs" (P3). Sakha 61 (P4). Giza 168 (P5). Sids 6 (PS), CAR 422 f ANA II
URES (P7). Gemmeiza 5 (pa) and Gemmelza 3 (P9) were used to estimate hybrid
vigour. general and specific combining ability. phenotypic and genotypic correlation
coefficients and cluster analysis for yield and its variables viz.: plant height. spike
length. number of spikeletslspike. number of spikes/plant. 1000 -kernel weight. grain
weight/spike. number of grains/spike and grain yield/plant. Highly significant
differences among genotypes, parents and crosses were recorded for an studied
traits. GCNSCA ratio exceeded the unity in all traits except grain weighVspike and
grain yield/plant. Thls indicated the importance of additive and addItive X additive
genetic effects controlling the majority of the studied traits. While the non-additive
gene effects had the highly importance for grain weighVspike and grain yield/pi ant.
The estimates of heterosis for grain yield/plant indicated that thirty crosses out of 36
F1 hybrid$ significanUy surpassed their better parent wilh percentage ranged from
6.94 % for P4 X P6 to 98.84 % for PS X P8. These relatively high heterotic
percentages along with the variability existed among all diallel set Increase the
chance of good recombinations that can be isolated in the following generations
particularly. when selfing In the follOwing generations gives an essentially
homozygous stale and enhances the role of selected plants in reducing the masking
effect of dominance. Results revealed that P1. P7 and P9 were the best combiners
for yielding ability and three or four ot its attrbutes. also two crosses P1 x P5 and P1
x P9 were the best specific combining ability effects for grain yieldJplant and its
attributes. All correlation coefficients between grain yleld/ptant and its components
were significant with positive expression except spike length. Clusters were formed by sequentially dividing groups of genotypes using
un-weighted pair grouped method using arithmetic average (UPGMA). Clusler
analysis produced four main groups. These groups are split into many subgroups
based on similarity and diSSimilarity of genotypes. The results Indicated that
genotypes 1 (P1). 5 (P1 X P5). 7 (P1 X P7). 18 (P3). 20 (P3 X P5). 28 (P4 X P7). 30
(P4 X P9). 32 (P5 X P6). 38 (P6 X P8) and 42 (P7 X P9) have a high distance level
between each other and will produce good newly genetic combination if they are used
in a crossing ptogram.