Comparison of the observed and predicted frequencies of transgressive segregates for yield and related traits in two bread wheat populations.

摘要

The genetic parameters estimated in seven basic sets of generation mean analysis and F2 triple test-cross progenies were used to predict the frequency of recombinant inbred lines derived from F2 bread wheat populations of Harrier X S308 and RedpollX HD2204 crosses grown over two environments. The validity of predictions was tested by comparing the expected frequency with the observed frequency of transgressive segregates scored on family means of 183 F6 lines. These lines were randomly derived through single seed descent in each of the F2 populations. In majority of cases, a close goodness of fit was achieved between the predicted and observed frequency of recombinant inbred lines. Recombinant inbred lines were obtained exceeding the parental limits for plant height, peduncle length, 100-grain weight and days to heading. The skewed distribution observed for predicted frequency of recombinant inbreds was due to the presence of additive X additive type epistatic interaction. The results of the experiments clearly demonstrated that genetic analysis of a character in early generation of a cross of bread wheat can provide the necessary information to predict accurately the likely outcome of inbreeding in the F2. Breeding strategies for recovery of recombinant inbred lines having the desired level of performance are suggested.