Premium
Doubled Haploid Testcross Design for Predicting Corn Hybrid Performance Using Molecular Markers
Author(s) -
Mowers Ronald P.,
Nyhus Kris A.,
Tragesser Scott L.,
Foster David J.
Publication year - 2018
Publication title -
crop science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.76
H-Index - 147
eISSN - 1435-0653
pISSN - 0011-183X
DOI - 10.2135/cropsci2017.02.0086
Subject(s) - hybrid , biology , doubled haploidy , molecular marker , trait , population , quantitative trait locus , genetics , genetic marker , microbiology and biotechnology , agronomy , gene , computer science , demography , sociology , programming language
This is an historical account of experiments done from 1995 to 1999 at Garst Seed Company using doubled haploid (DH) technology, considered company secret at the time but now routine in many companies. We introduce a crossing design, the Doubled Haploid Testcross Design (DHTD), to combine DH and molecular marker technologies to rapidly develop and test new hybrids for replacing elite commercials and to predict future hybrid corn ( Zea mays L.) field performance. A DHTD experiment in 1997 allowed estimation of marker‐trait relationships and phenotypic general combining effects. The statistical analysis gave predictions of performance for all possible (untested) hybrid combinations between 77 female and 100 male DH lines, either based solely on molecular markers or based on hybrid phenotypic means. A second experiment in 1998 tested the validity of these predictions, contrasting molecular marker‐predicted best with phenotypic‐predicted best hybrids. Among the 1998 validation set of hybrids, the subset predicted to be best using molecular markers had statistically significant 6.9 and 20.4% advantages in performance index compared with subsets predicted to be in the middle and bottom thirds, respectively. In validation tests for hybrids predicted from only phenotypic general combining abilities (GCA), the 48 highest predicted hybrids had significantly better average performance than the four population‐parent hybrid checks and the best DH × parent hybrid checks. The 48 GCA testcross predicted best hybrids were significantly better in performance index than the marker‐best‐predicted hybrids. However, marker‐based prediction for specific traits such as grain moisture and lodging resulted in significant improvement over phenotypic‐selected hybrids.