Allopolyploidy plays an important role in plant evolution and confers obvious advantages on crop growth and breeding compared to low ploidy levels. The present investigation was aimed at synthesising the first known chromosomally stable hexaploid Brassica with the genome constitution AABBCC. More than 2,000 putative hexaploid plants were obtained through large-scale hybridisation from various combinations of crosses between different cultivars of Brassica carinata (BBCC) and B. rapa (AA). The majority of plants after two generations of selfing within selected hexaploid plants (H(2)) were aneuploid, and only 80 plants (4.6%) had the expected hexaploid chromosome number (2n = 54). The hexaploid ratio increased to an average of 23.0 and 26.3% in the H(3) and H(4) generations, respectively, and was accompanied by an increase in pollen fertility. The appearance of aneuploid plants in each generation could be detected having various chromosomal abnormalities at meiosis. The frequency of hexaploid plants varied significantly among different cultivar combinations, from 0 to 56% in the H(4) generation, and it showed a positive correlation with pollen fertility. The frequency of SSR allelic fragments lost or novel alleles gained was significantly lower in H(4) than in H(2) and H(3), which reflects increasing genome stability in H(4). The A and C genomes were significantly less stable than the B genome, which may mainly result from frequent homoeologous pairing and rearrangements between the A and C genomes. Methods to establish a stable hexaploid Brassica crop by intercrossing these lines followed by intensive selection are also discussed.

Accumulated evidence has shown that each of the three basic Brassica genomes (A, B and C) has undergone profound changes in different species, and has led to the concept of the "subgenome". Significant intersubgenomic heterosis was observed in hybrids between traditional Brassica napus and first generation lines of new type B. napus. The latter were produced by the partial introgression of subgenomic components from different species into B. napus. To increase the proportion of exotic subgenomic components and thus achieve stronger heterosis, lines of first generation new type B. napus were intercrossed with each other, and subjected to intensive marker-assisted selection to develop the second generation of new type B. napus. The second generation showed better agronomic traits and a higher proportion of introgression of subgenomic components than did the first generation. Compared with the commercial hybrid and the hybrids produced with the first generation new type B. napus, the novel hybrids showed stronger heterosis for seed yield during the 2 years of field trials. The extent of heterosis showed a significant positive correlation with the introgressed subgenomic components in the parental new type B. napus. To increase the content of the exotic subgenomic components further and to allow sustainable breeding of novel lines of new type B. napus, we initiated the development of a gene pool for new type B. napus that contained a substantial amount of genetic variation in the A(r) and C(c) genome. We discuss new approaches to broaden the avenue of intersubgenomic heterosis in oilseed Brassica.

Intersubgenomic heterosis in rapeseed has been revealed in previous studies by using traditional Brassica napus (A(n)A(n)C(n)C(n)) to cross partial new type B. napus with A(r)/C(c) introgression from the genomes of B. rapa and B. carinata, respectively. To further enlarge the genetic basis of B. napus and to facilitate a sustained heterosis breeding in rapeseed, it is crucial to create a population for substantial new type B. napus diversified at both A/C genomes. In this experiment, hundreds of artificial hexaploid plants (A(r)A(r)B(c)B(c)C(c)C(c)) involving hundreds of B. carinata/B. rapa combinations were first crossed with elite lines of partial new type B. napus. The pentaploid plants (AABCC) were open-pollinated in isolated conditions, and their offspring were successively self-pollinated and intensively selected for two generations. Thereafter, a population of substantial new type B. napus mainly with a genomic composition of A(r)A(r)C(c)C(c) harbouring genetic diversity from 25 original cultivars of B. rapa and 72 accessions of B. carinata was constructed. The population was cytologically verified to have the correct chromosome constitution of AACC and differed genetically from traditional B. napus, in terms of the genome components of A(r)/C(c) and B(c) as well as the novel genetic variations induced by the interspecific hybridisation process. Synchronously, rich phenotypic variation with plenty of novel valuable traits was observed in the population. The origin of the novel variations and the value of the population are discussed.

Brassica carinata (BBCC, 2n ¼ 34) possesses many good agronomic characters and, to provide more genetic information about this species and utilize it further, 110 accessions of B. carinata were tested for genetic variation by using 233 amplified fragment length polymorphism markers, so that a dendrogram could be constructed. To combine the good traits of B. carinata and B. rapa (AA, 2n ¼ 20) with those of B. napus (AACC, 2n ¼ 38) which is the major rapeseed variety in China, interspecific crosses between these two species have been made which have resulted in 276 doubled hybrids (AABBCC). These hexaploids combine positive characters and can be used for crossing with B. napus. Brassica carinata as the female parents exhibited extensive differences in interspecific crossability and 20 high crossability accessions were identified. The Poisson Regression Model analysis result (P < 0.0255) indicated that such differences were due to the genotype of the accessions. Accessions with high crossability could promote gene flow within the genus Brassica.

No abstract