Covariance between Line per se and Testcross Performance
- O. S. Smith
The relationship between line per se and testcross performance was evaluated theoretically to examine possible reasons for the low values, reported in the literature, for this correlation. An expression for the correlation between line per se and testcross performance was developed. This expression was evaluated using restrictive assumptions for allellic frequencies and effects. Computer simulation was used to estimate the correlations in the more general case. These results indicated that, for traits conditioned by a large number of loci showing complete dominance, the correlations between line perse and testcross performance are expected to be less than 0.5. This is due to the masking effects of favorable dominant alleles in the tester. The computer simulation results agree with previous empirical results which showed that use of a high performance (good) tester will reduce the genetic variance among testcrosses. Additionally, the use of a good tester reduced the correlation between line per se and testcross performance. Correlations between line per se and testcross performance to a related (tester from the same population) and unrelated (tester from a different heterotic group) tester also were examined in the simulation study. Correlations were generally low, although higher for testcrosses to a good unrelated tester than for testcrosses to a good related tester.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © .