大豆不同季節交互選拔效能之探討

The Efficiency of Disruptive Seasonal Selection in Soybean

林順福(S. F. Lin) ； 魏趨開(C. K. Wey) ； 詹國連(K. L. Chan) ； 周政宏(J. H. Jou)

《中華農業研究》 37卷3期 (1988/09) Pp. 291-298

https://doi.org/10.29951/JARC.198809.0007

摘要

以74、75及76年度，各春、秋兩季之品系比較試驗及區域試驗中共408個參試品系爲材料，以各試驗產量等第爲基準，探討春、秋兩作交互連續選拔之效果，結果如下：（一）三層次之分析結果均顯示春、秋兩作交互選拔結果弊多於利：(1)春、秋兩作間產量等第相關分析：除76年中級品系比較試驗春、秋兩作相關值顯著外，其他各品系比較試驗及區域試驗之相關值均不顯著。(2)優良品系損失百分率分析：在各季無論以CK爲選拔標準，或選拔比例爲50%，均造成60%以上優良品系之損失，而選拔強度提高（選拔比例爲25%）時，優良品系損失率增至80%以上。(3)各季較優之五品系於另一季之變動分析：絕大多數各期作較高產之前五品系，在另一期作之表現均不若前一期作，甚至低於對照，而遭致淘汰。（二）由不同年度相同季節間產量等第相關程度大於不同季節間，顯示單季選拔效果大於春、秋兩季交互連續選拔。（三）建議大豆高產育種，在高級世代（F6以後）之品系比較試驗或區域試驗，宜採用兩年四季爲一循環之綜合評估法。

並列摘要

For studies on the effectiveness of disruptive seasonal selection in soybeans, 10 sets of yield trials containing 408 lines were conducted in the spring and fall seasons during 1985-1987, The results were summarized as follows:
1. Three evidences indicating that disruptive seasonal selection had less efficiency in soybean breeding programs:
(1) Most trials showed no significant rank correlation between spring and fall crop seasons except the intermediate yield test in 1987.
(2) Disruptive selection according to check variety or 50% proportion for high yield lines in each season of spring and fall also gave more than 65% preferable lines lost, and more than 80% lost when selection proportion decreased to 25%.
(3) Due to the variation of yield rank, most preferable 5 lines in one season were lost in the other season in the same year.
2. In view of the significantly and highly positive rank correlation within springs and within falls in various year, we concluded that selection in successive spring or fall was much more effective than disruptive seasonal selcteion.
3. The recommend from this experiment was by using two couple seasonal selection per cycle in yield test, thus would be more effective in soybean breeding.