| Budget Amount *help |
¥3,500,000 (Direct Cost: ¥3,500,000)
Fiscal Year 2001: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 2000: ¥1,100,000 (Direct Cost: ¥1,100,000)
Fiscal Year 1999: ¥1,600,000 (Direct Cost: ¥1,600,000)
|
|---|
| Research Abstract |
It is proposed that the efficiency of selection methods in plant breeding should be measured not by the expected genetic advance E(R), but by S/C (S=the probability that the desired genetic advance is achieved, and C=the cost that is spent for that). A selection method that maximizes S/C gives the largest number of success under the same long-term resource investment (Yonezawa et al. 1999), although it does not necessarily maximize E(R) in individual selection trials. In the research of this year, the efficiencies of five systems of selection for outbreeding crop plants, i.e., MAS (uniparental mass selection), FAM1 and FAM2 (selection between plus within half-sib falmilies; 1 with random crossing with all plants and families, and 2 with random crossing within each family), HSP (selection with half-sib progeny test), and FSP (selection with full-sib progeny test), were compared based on the Monte Carlo calculations of S/C (Plants per selection cycle are 5000 in MAS, FAM1 and FAM2 and 10000 in HSP and FSP; selection intensity is 5% in each selection cycle and system; 10000 repeated runs). As far as evaluated by the traditional criterion, i.e., E(R) under infinite population size, a more sophisticated selection system should be more efficient. However, Monte Carlo calculations showed that, because random drift occurs, the genetic advance expected under infinite population size is not always realized when a practically possible finite population size is used When evaluated by S/C, the simplest selection system of the five, MAS, is not substantialy inferior to any of the other systems. The most sophisticated systems HSP and FSP are not efficient because the probability of success S in these systems does not increase sufficiently to compensate for the extra investment for progeny test.
|
