No abstract

SummaryIn 1975, tests with UK populations of Plasmodiophora brassicae not only revealed a lack of effective clubroot resistance in swedes (Brassica napus), but also the outstanding resistance of the European Clubroot Differential (ECD)04 (B. rapa). It was, therefore, decided to transfer the resistance genes from ECD04 to swedes, using the most pathogenic UK population of clubroot (C56) available for screening purposes. An autotetraploid form of ECD04 was crossed with tetraploid kale (B. oleracea) using the latter as female parent. One of the euploid, 2n = 38, hybrids secured by embryo rescue in 1976 was crossed to the swede cultivars Marian and Ruta Øtofte. Three further backcrosses of clubroot resistant plants to lines derived from modern swede cultivars were made over the period 1980 to 1982. Selfing commenced in 1983 to produce F2 populations. From F3 to F5 there was family selection for yield and agronomic characters, as well as single plant selection for clubroot resistance. In 1991, the six most promising F5 families were multiplied for subsequent evaluation in replicated yield trials in Dundee. The most promising family entered official trials at the beginning of 1993 and, 2 years later, was added to the National List as cv. Invitation and granted Plant Breeders' Rights. The first certified seed was sold in 1996, 20 years after the original synthetic B. napus was produced.The breeding programme provided evidence for only one of the three postulated dominant genes in ECD04 being required for resistance to C56 and also good evidence of differential resistance from tests with other clubroot populations. Hence, whilst the differential resistance in cv. Invitation should prove useful in the UK in the immediate future, it may not be durable in the longer term. It is, therefore, argued that the next and more difficult goal to achieve should be to introduce high levels of non‐differential resistance from B. oleracea.

In 1989 seven F, hybrid swedes were produced in polythene tunnels. The female parent was an inbred line derived from cv. Ruta Otofte, and made selfincompatible with an S-allele introduced by backcrossing. Three more hybrids were produced in 1990 in a similar way with an inbred line from cv. Criffel as female parent. In trials in Dundee over the period 1990 to 1992, better parent heterosis for dry matter yield in the hybrids ranged from 1.9% to 19.2%. There was no such heterosis for dry matter percentage nor for mildew resistance.The highest yielding hybrid was the one between the Criffel line and an inbred line from cv. Magres. It averaged 9.9% heterosis over the two years 1991 and 1992, outyielded seven commercial cultivars, but had a lower yield (12.3 t/ha) than an F6 line SS5 (13.3 t/ha) derived from the cross between the Criffel and Magres lines by pedigree inbreeding with selection.It is concluded that high yielding heterotic F,'s should be the starting point rather than the finishing point of swede breeding programmes, and that research into breeding methods should concentrate on improving the efficiency of pedigree inbreeding following hybridisations, rather than on ways of producing F, hybrids.

Marrow-stem kale (Brassica oleracea var. acephala L.) is grown in the United Kingdom, New Zealand and similar climates, for feeding cattle and sheep during autumn and winter. Population improvement by half-sib family selection for higher digestibility and digestible organic matter yield, lower levels of antimetabolites, and clubroot resistance is described and analysed as a selection experiment. The foundation population consisted of 120 clubroot resistant plants from the 16 most resistant marrow-stem cultivars in an assessment of 96 cultivars of Brassica oleracea. At the end of the first generation of selection adequate resistance had been achieved. After five generations, dry-matter content and digestible organic matter yield had been increased, the latter without an undesirable increase in height. Indolyl glucosinolate (source of goitrogenic thiocyanate ion) content had been reduced, but not S-methylcysteine sulphoxide (the haemolytic factor) content. Selection differentials each generation were small and heritabilities of family means were moderate to low, making improvement per generation slow. It was demonstrated that faster progress could be made by switching from half-sib to full-sib family selection, as expected from theory. Multiplication of the half-sib family with the highest digestibility (DOMD%) achieved an improvement in this trait, without a reduction in digestible organic matter yield, and was marketed as cultivar Grampian.

The results are presented of two single seed descent (SSD) breeding programmes for swedes (Brassica napus L. var.