Inheritance of root glucosinolate content in canola

Recent research indicates that high levels of specific glucosinolates (GSL) in canola roots reduce nematode hosting, and suppress nematode and fungal pathogens of subsequent crops. The potential to increase levels of root GSL s is therefore of interest, yet little is known of the genetics and subsequ ent opportunities for breeding root GSL in Brassica spp. A population was d eveloped from crosses between canola cvv. Karoo (high root GSL) and Monty ( low GSL). Parents, and F-1 and F-2 progeny, were grown in the CSIRO Phytotr on in Canberra and total root GSL, 2-phenylethyl (2PE), %2PE in roots, and indolyl GSL contents measured. Among-entry variation for flowering time was related (r = 0.15-0.42, P < 0.05) to changes in total root and component G SL contents. When corrected for variation in flowering time, significant (P < 0.05) variation was observed for root GSLs among parents, between F-1 an d F-2 generations, and among F-2 progeny. Repeatability was moderate to hig h for most GSLs measured in the glasshouse. Maternal genetic effects were u nimportant, while genetic variation for 2PE, %2PE, and 3- indolylmethyl GSL (3IM) contents were largely under additive genetic control, although there was some nonadditive gene action for %2PE. When parents and 100 F-2:3 prog eny were evaluated in replicated field plots in autumn and spring 1999, bro ad- and narrow-sense heritabilities were moderate to high (0.43-0.93) for t otal root, 2PE, and %2PE GSL contents. Differential family sensitivity to s owing date reduced broad- and narrow-sense heritabilities for all indolyl r oot GSLs. Parent-offspring and realised heritability estimates were smalles t for all root GSLs measured in the autumn sowing. This suggested that care should be taken when glasshouse screening for field root GSL content. Gene tic correlations with total root GSL content were largest for 2PE (r(g) = 0 .99), %2PE (0.88), and 3IM (-0.82) contents. Phenotypic correlations for to tal root GSL and 2PE contents were also strong (r(P) = 0.98) but environmen t decreased this correlation for %2PE (0.66) and particularly 3IM (-0.24). Strong genetic and phenotypic relationships for 2PE and total root GSL cont ent suggested 2PE should provide a robust and reliable predictor in screeni ng for high total root GSL content in canola. High heritability and large a dditive gene action should enable selection for high 2PE content in early g enerations of a canola breeding program.