الفهرس 
ACKNOWLEDGEMENTSOnly 14 pages are availabe for public view
 |  | from | 72 |  |  |
| | | from | 72 | | |
Abstract
Two consecutive cycles of selection were imposed on five F2 populations derived from crosses four of which had the elite-very long spike (>25 cm) line 6 as a common parent. The first cycle was a divergent selection for spike length conducted in the favorable environment of the optimal sowing date and the response was measured under favorable and heat stress conditions of a late sowing date. Positive responses to selection for longer spikes were obtained which were greater under favorable conditions (averaged 13.43%) than under heat stress (8.66%) whereas the responses for shorter spikes were less in magnitude being 2.24% and 5.02% in the two environments, respectively.
The longer spike F3 selections were more sensitive to heat stress (8.14 average reduction) than the shorter spike selection (6.94% reduction) and the least reduction was that of the unselected bulks (4.03%). The realized heritability of spike length was greater under favorable conditions (0.22-0.56) than under heat stress (0.18-0.41).
Concurrent positive responses to selection for longer spikes were obtained in grain yield per spike which averaged 25.35% in the favorable environment and 18.03% under heat stress indicating remarkable improvement in heat tolerance. The correlation between spike length and grain yield per spike was stronger (r=0.91, P<0.01) under favorable than under heat stress conditions (r=0.65, P<0.01). selection for greater number of grains per spike imposed on the F3 plants selected for spike length under heat stress resulted in significant responses which averaged 14.65%. The realized heritability was very low (ranging from 0.10 to 0.28)
Selection for greater number of grains per spike resulted in correlated responses in grain yield per spike which averaged 17.64%. The correlation between number of grains per spike and grain yield per spike was almost complete (r=0.98, P<0.01) under heat stress. The concurrent positive responses produced in spike length in F4 with selection for number of grains per spike (averaged 9.20%) was almost equal to that produced by the direct selection in the F3 (8.66%) indicating that the selection advance effected in the F3 has been maintained in the F4 .
The F3/F4regression for spike length under heat stress was high and significant (b=0.85±0.07) indicating the high heritability of this trait.
In conclusion, phenotypic selection for longer spikes under heat stress followed by a cycle of selection for number of grains per spike under heat stress was capable of improving heat tolerance in wheat.