Selection Procedures for Developing Fertility Restorer Lines in Wheat

The effectiveness of selection for anther extrusion in fertility restorer lines of wheat (Triticum aestivum L. em Thell.) to improve cross-pollination potential and the relationship between the self-fertility of the restorer lines and the fertility of their testcrosses were examined in six R/B (restorer/conventional line) crosses and one R/R (restorer/restorer) cross. A common restorer parent, R109, was used in all crosses. One hundred and fifty-two F₆ R-lines, selected from 722 F₅ lines, were evaluated in yield trials at three locations in North Dakota in 1979. All F₆ lines were testcrossed to male steriles with Triticum timopheevi Zhuk. cytoplasm, and fertility of testcross progeny was used as the criterion for fertility restoration potential. The effect of restorer genes on several agronomic and quality traits also was investigated. Regressions of the F₆ on F₅ generations indicated that selection for anther extrusion in the F₅ was effective although somewhat less effective than selection for heading date and plant height. Forty percent of the variation in the F₆ for anther extrusion was determined by variation in the F₅ compared to 64 % for heading date and 72 % for plant height. R-line anther extrusion was only slightly indicative of fertility restoration potential; the R² for testcross F₁ seed set vs. F₆ R-llne anther extrusion was 10%. Among the 152 R-lines, there was a low negative association between anther extrusion and 200-kernel weight and a low positive association between R-line test weight and testcross seed set, When lines were grouped by testcross fertility, no differences among testcross fertile, testcross partially fertile, and testcross male-sterile lines were found for any trait except R-line anther extrusion and testcross anther extrusion. The R109-derived restorer genes and their associated linkage blocks had no deleterious effects on agronomic traits.