Androgenesis from a Lolium multiflorum × Festuca arundinacea hybrid to generate extreme variation for freezing-tolerance

New opportunities for plant breeding using androgenesis in Lolium × Festuca hybrids have been identified. Plants derived by anther culture from a Lolium multiflorum × Festuca arundinacea (5x) hybrid were screened for freezing-tolerance, and their post-freezing recovery compared. The androgenic population showed extreme diversity in freezing-tolerance. While the majority of androgenic plants had inferior freezing-tolerance compared with the freezing-sensitive L. multiflorum parent, 6% of the population were more freezing resistant than the freezing-tolerant Festuca parent. Novel Lolium and Festuca gene combinations resulting from rare meiotic events were recovered within the androgenic population. The two most freezing-tolerant androgenic plants carried virtually the entire F. pratensis subgenome of F. arundinacea. F. pratensis is known to carry genes for freezing-tolerance and would be expected to be the primary source of genes governing this trait within the F. arundinacea genome. The most freezing-tolerant androgenic plants were more freezing-tolerant than the hybrid plant from which they were derived. Consequently, androgenesis was also effective in removing factor(s) reducing the expression of freezing-tolerance within the L. multiflorum × F. arundinacea (5x) hybrid.     

The controlled introgression of Festuca arundinacea genes into Lolium multiflorum

Summary Using phosphoglucoisomerase (PGI/2) as a genetic marker, it has been shown to be possible to transfer genes from Festuca arundinacea into diploid Lolium multiflorum using the pentaploid hybrid L. multiflorum (4x) x F. arundinacea (6x). The pentaploid hybrid was sufficiently fertile to be used in reciprocal crosses with diploid. L. multiflorum. When used as the male parent, only two backcross generations were then required to reconstitute the diploid genotype. Intergeneric recombinants including a F. arundinacea PGI/2 allele were found among the diploid BC₂. Cytological data indicates that although the majority of chromosome associations involve only homologous Lolium chromosomes, associations involving Lolium and Festuca chromosomes also occur.Interpollinating the pentaploid hybrids prior to commencing a backcrossing programme increases the number of cycles of recombination and improves the chance of recovering intergeneric recombinants. The crossing programme described is proposed to be an effective method of introducing F. arundinacea genes into L. multiflorum.     

Designing grasses with a future – combining the attributes of Lolium and Festuca

Lolium species (considered the ideal grasses for European agriculture) are not sufficiently robust to meet many of the environmental challenges that face extensive agriculture in less favoured areas. Fortunately, adaptations to abiotic and biotic stresses exist amongst Festuca species related closely to Lolium. The complex of species has an enormous wealth of genetic variability and potentiality for genetic exchange, thus offering unique opportunities for the production of versatile hybrid varieties with new combinations of useful characters suited to modern grassland farming. The attributes of Lolium and Festuca can be combined into a single genotype by amphiploidy or alternatively, a limited number of characters can be selectively introgressed from Festucainto Lolium or vice versa. Androgenesis of the interspecific hybrids can generate genotypes combining characters that may not be recovered by sexual backcrossing. Genomic in situ hybridization(GISH) can differentially ‘paint’ the chromosomes of Lolium and Festuca and identify Lolium-Festuca recombinant chromosomes. GISH is valuable in the analysis of amphiploids, introgressions and androgenic genotypes and can be used to physically map introgressed traits. Introgression mapping is a powerful new approach to the mapping of traits and arises from a fusion of physical and genetic mapping. For example, in a diploidLolium introgression genotype with only one introgressed Festucasegment, the gene(s) for any Festucaderived trait expressed by the plant must be located within the segment. Using GISH and molecular markers, a dense but highly localised map of the Festuca segment is made in isolation of the Loliumgenome – this may simplify QTL analysis. This revised version was published online in July 2006 with corrections to the Cover Date.     

Androgenesis in a Lolium multiflorum × Festuca arundinacea hybrid to generate genotypic variation for drought resistance

The pentaploid hybrid of Lolium multiflorum and Festuca arundinacea (2n = 5x = 35) combines the high growth rate of L. multiforum with the drought resistance and freezing-tolerance of F. arundinacea. Unfortunately, it also displays the deleterious traits associated with Festuca, namely those associated with high leaf fibre content giving rise to poor palatability and digestibility. To access different combinations of these characters, anther cultures were initiated and regenerated into single embryo derived plants. The anther culture method was very productive since out of a total of 2349 androgenic plants derived from the same parent plant, 57% were green plantlets, although only 507 (22%) subsequently established into plants following transfer to soil. Chromosome counts of randomly selected lines showed that plants with euploid chromosome numbers (14, 21, and 28) would appear to have selective advantage during regeneration. There was wide variation between mature androgenic plants grown under field conditions in plant height, leaf length, leaf width, tiller number and herbage dry matter. The variation between genotypes in response to drought stress was assessed by placing replicate clones under rain-out shelters or under irrigated control conditions in the field. Herbage dry matter under drought was higher in a number of androgenic lines than either parents, but not higher than the pentaploid hybrid. Androgenesis was shown to be a highly effective procedure to expose diverse phenotypic variation all derived from the same Lolium × Festuca hybrid genotype. This revised version was published online in July 2006 with corrections to the Cover Date.     

Infection of Lolium and Festuca spp. by Drechslera siccans and D. Catenaria

Summary Drechslera siccans was the commonest of the species of this leaf spot pathogen found on single spaced plants of Lolium perenne at Aberystwyth. D. catenaria and D. dictyoides were also frequently isolated. A wide range of genotypes within the host species L. perenne, L. multiflorum, Festuca arundinacea and F. pratensis, and structural hybrids between F. arundinacea and L. multiflorum, were all equally susceptible to conidial infection by D. siccans in the glasshouse and showed similar development of leaf spot symptoms, even though Festuca spp. are not normally attacked in the field. However, when exposed to infection out of doors there was significant variation in resistance, both between species and often also between genotypes within the same species. As in so many other instances of fungal infection, tetraploidy appeared to confer a higher level of resistance. The degree of resistance of a F. arundinacea × L. multiflorum amphiploid and two succeeding backcross generations to L. multiflorum approximated to that of the more susceptible L. multiflorum parent. By contrast to D. siccans, when plants were inoculated with D. catenaria in the glasshouse, both intervarietal and intravarietal variation in leaf spot development was evident. Resistance to D. siccans appears to depend partly on a limitation of conidial production, and to D. catenaria on restriction of hyphal growth following infection, which may explain the relative discrepancies between field and glasshouse results. The implications for resistance breeding are discussed.     

Improved Drought Resistance in Introgression Lines Derived from Lolium multiflorum × Festuca arundinacea Hybrids

In a back-crossing programme involving Lolium multiflorum (the recurrent parent) and Festuca arundinacea, the diploid L. multiflorum phenotype was rapidly recovered with the inclusion, in some progeny, of a small number of genes from the fescue parent. In field drought trials derivatives of these backcross populations were on average less drought resistant than the L. multiflorum parental populations, but 3 % of individuals were as drought resistant as F. arundinacea. After only one cycle of selection and polycrossing these drought-resistant Lolium-like plants, the mean drought resistance of most progeny lines was significantly improved, in some cases to near that of F. arundinacea. Available evidence strongly indicates that this improved drought resistance was due to the transfer of genetic material from Festuca into Lolium. These populations will contribute to variety improvement and to our understanding of the genetics and physiology of drought resistance.     

Effects of introgressions from <Emphasis Type="Italic">Festuca pratensis</Emphasis> on winter hardiness of <Emphasis Type="Italic">Lolium perenne</Emphasis>

Previous studies reported that some genotypes with introgressed Festuca chromosome segment(s) in Lolium genome showed enhanced winter hardiness compared to Lolium. The aim of this study was to search comprehensively for the Festuca pratensis chromosome regions affecting winter hardiness-related traits when introgressed into the Lolium perenne genome. Association between F. pratensis introgression and winter hardiness-related traits (fall and winter hardiness indexes, early-spring dry matter yield, and freezing tolerance) were screened in the diploid introgression populations (n = 203) that had some F. pratensis chromosome segments introgressed. Eighty-four intron markers corresponding to unique rice genes randomly distributed across the genome were used for genotyping. Winter hardiness of almost all plants in the introgression populations was lower than that of the F. pratensis and triploid hybrid parents, but the average was higher than that of L. perenne. A significant positive effect of F. pratensis introgression on early-spring dry matter yield was detected on chromosome 7. This quantitative trait locus (QTL) was confirmed by linkage analysis using a backcross population with F. pratensis introgression in the target region of chromosome 7. However, the contribution of the newly identified QTL was rather small (6.7–9.6%), suggesting that superior winter hardiness of F. pratensis compared to L. perenne is conferred by multiple small-effect QTLs. We also detected a previously unreported negative effect of Festuca introgression on winter hardiness. Newly obtained QTL information in this study would contribute to the design of Festuca/Lolium hybrid breeding.     

Assessing success in gene transfer betweenLolium multiflorum andFestuca arundinacea

Summary AFestuca arundinacea (2n=6x-42) plant with three PGI/2 homoeoalleles marking three homoeologous chromosomes was crossed with aLolium multiflorum plant (2n=4x=28) with a different PGI/2 phenotype to give a pentaploid hybrid (2n=5x=35) with five chromosomes each marked by a different PGI/2 allele. This hybrid plant was backcrossed twice with diploidL. multiflorum (2n=2x=14) with a different PGI/2 phenotype. Numbers of interspecific recombinants involving chromosomes marked by PGI/2 were then determined in both backcross generations. In the BC1, recombinants involving only one PGI/2 allele were found but in the BC2, all three homoeologousF. arundinacea chromosomes carrying the PGI/2 locus recombined withLolium with one in greater frequency and the others in equal but lower frequency. The evidence supports claims made forF pratensis (2n=2x=14) andF. arundinacea var.glaucescens (2n=4x=28) being progenitors forF. arundinacea.     

Cytogenetic Studies of Hybrids between Festuca gigantea Vill. and Lolium multiflorum Lam.

The meiotic behaviour of the hybrid between Festuca gigantea (2n – 6x = 42) and Festuca gigantea (2n = 6x = 42) indicates distinct structural differences between the two species. These differences are also apparent in the degree of chromosome pairing observed in Lolium multiflorum×F. gigantea compared with previous reports on the L. multiflorum×F. arundinacea hybrids. Although the L. multiflorum×F. gigantea 8x amphiploid com Dines the complementary characters of the two species and is agronomically interesting, there are some irregularities in meiotic behaviour that could affect the stability of the amphiploid. Seed fertility in the amphiploid is high and preliminary studies show evidence of a high degree of self-fertility. The pentaploid hybrid between autotetraploid L. multiflorum×F. gigantea is sufficiently fertile to be used as the pollen plant in crosses with diploid L. multiflorum (BC₁). Backcrossing the BC₃ hybrid to L. multiflorum results in mainly diploid progeny. The possibilities, of using this crossing scheme to introgress F. gigantea characters into L. multiflorum is discussed as an alternative approach to amphiploid breeding as a means o: combining specific complementary characters of the two species.     

Utilization of flow cytometry for festulolium breeding (Lolium multiflorum (2x) × Festuca arundinacea (6x))

Festulolium is a hybrid between Festuca and Lolium species that has valuable agronomic traits from both grass species. The purpose of our breeding program is to produce hexaploid festulolium that introduces tolerance to summer depression into Italian ryegrass (Lolium multiflorum) by crossing it with tall fescue (Festuca arundinacea). However, we found the DNA ploidy of hexaploids was not stable and was reduced in successive generations. We aimed to find out how to obtain stable high-ploidy festulolium. F1 hybrids of L. multiflorum and F. arundinacea were produced. The F3 generation was produced from putative hexaploid F2 individuals by open pollination. The F4 to F6 generations were obtained by polycrossing. The DNA ploidy levels of F2 to F6 individuals were estimated by flow cytometry. Cytological characteristics of the F5 and F6 individuals were investigated by FISH and GISH. The DNA ploidy level of hexaploid festulolium was reduced and stabilized at almost the same level as a tetraploid. Seed fertility was inversely correlated with an increase in ploidy level. GISH revealed no preferential Lolium transmission. FISH with a telomere probe revealed that counting the exact number of chromosomes in festulolium was difficult. DNA ploidy level was strongly correlated with the number of chromosomes.     

Disturbed Mendelian segregations at isozyme marker loci in early backcrosses of Lolium multiflorum × Festuca pratensis hybrids to L. multiflorum

Summary The segregation of interspecific recombinant Festuca pratensis (Fp) alleles, introgressed into the germolasm of Lolium multiflorum (Lm), at four loci (PGI/2, AcP/2, GOT/3 and BAP) is described. Heterozygous (Lm/Fp) plants were backcrossed to L. multiflorum (2n=2×=14) and subsequent BC2 Lm/Fp sibling genotypes intercrossed.In crosses between BC1 heterozygous plants (Lm/Fp) used as males and L. multiflorum, there was a reduced transmission of the F. pratensis (Donor Parent) derived alleles in the populations with PGI/2 and AcP/2 marker loci compared to the reciprocal cross but the reduction was not significantly different in those with GOT/3 and BAP markers.Two PGI/2 marked BC2 half-sib families in particular exhibited a more extreme deficiency of Fp/Fp progeny plants than anticipated from the BC2 segregations indicating possible linkage to zygotic lethals. Deficiencies of F. pratensis alleles were, in most cases, less marked in BC2 half-sib families indicating that a further round of recombination had reduced the size of the introgressed chromosome segment or that deleterious linkages had been broken. A tendency towards heterozygote advantage was found in one BAP marked halfsib family.The significance for forage grass breeding of reduced transmission rates of Donor Parent alleles in early back-cross generations especially through the male gametes is discussed.     

Intergeneric symmetric and asymmetric somatic hybridization in Festuca and Lolium

Summary Intergeneric symmetric and asymmetric somatic hybrids have been obtained by fusion of metabolically inactivated protoplasts from embryogenic suspension cultures of tall fescue (Festuca arundinacea Schreb.) and unirradiated or 10–500 Gy-irradiated protoplasts from non-morphogenic cell suspensions of Italian ryegrass (Lolium multiflorum Lam.). Genotypically and phenotypically different somatic hybrid Festulolium mature flowering plants were regenerated.Species-specific sequences from F. arundinacea and L. multiflorum being dispersed and evenly-represented in the corresponding genomes were isolated and used for the molecular characterization of the nuclear make-up of the intergeneric, somatic Festulolium plants recovered. The irradiation of Italian ryegrass protoplasts with 250 Gy X-rays prior to fusogenic treatment favoured the unidirectional elimination of most or part of the donor chromosomes. Irradiation of L. multiflorum protoplasts with 500 Gy produced highly asymmetric (over 80% donor genome elimination) nuclear hybrids and clones showing a complete loss of donor chromosomes.The RFLP analysis of the organellar composition in symmetric and asymmetric tall fescue (+) Italian ryegrass regenerants confirmed their somatic hybrid character and revealed a bias towards recipient-type organelles when extensive donor nuclear genome elimination had occurred.Approaches aimed at improving persistence of ryegrasses based on asymmetric somatic hybridization with largely sexually-incompatible grass species (F. rubra and Alopecurus pratensis), and at transferring the cytoplasmic male sterility trait by intra- and inter-specific hybridization in L. multiflorum and L. perenne, have been undertaken.Abbreviations cpDNA chloroplast DNA - CMS cytoplasmic male sterility - 2,4-D 2,4-dichlorophenoxy-acetic acid - IOA iodoacetamide - mtDNA mitochondrial DNA - RFLP restriction fragment length polymorphism     

Breeding for hard red winter wheat cultivars adapted to conventional-till and no-till systems in northern latitudes

Summary Monosomic alien addition lines combining individual F. drymeja chromosomes and the L. multiflorum complement were isolated from the cross between the triploid hybrid L. multiflorum (4x) × F. drymeja (2x) and diploid L. multiflorum (2x). Chromosome pairing in the addition lines was studied at metaphase 1 of meiosis and the relationship between single F. drymeja chromosomes and the corresponding homologous pair in L. multiflorum is discussed. Trivalent frequency in the addition lines was higher than expected from observations of chromosome pairing in the triploid hybrid and there were differences between lines in the number of trivalent associations formed. There is some evidence to suggest that trivalent frequency is not entirely dependent on chromosome length and that transmission of the alien chromosome in the female is dependent on the size of the added chromosome. Morphological studies were made to assess the phenotypic effects of the addition of single F. drymeja chromosomes to the L. multiflorum complement. Two plants (2n=14) with recombination between a L. multiflorum and a F. drymeja chromosome were identified.     

Genomic constitution of Festulolium cultivars released in the Czech Republic

Genomic in situ hybridization (GISH) was used to characterize the chromosome constitutions of individual plants from a set of tetraploid and hexaploid cultivars of Festulolium developed and released in the Czech Republic from hybrids of Lolium multiflorum with Festuca pratensis and F. arundinacea. A simplified GISH protocol readily discriminated parental genomes in the hybrids and facilitated the screening of large numbers of plants per accession. The contribution of parental genomes in the cultivars tested ranged from predominance of chromatin from one of the parents to a more balanced contribution from both parents. However, in none of the cultivars were equal proportions of chromatin from both parents present. The parental contribution to the hybrids was both in the form of complete chromosomes or as chromosome translocations. In hexaploid cultivars from (L. multiflorum × F. arundinacea) × F. arundinacea hybrids the average numbers of complete L. multiflorum chromosomes ranged from 4.95 to 7.5 and the numbers of translocations from 6.33 to 10.21. Two tetraploid cultivars from (L. multiflorum × F. arundinacea) × L. multiflorum hybrids showed a strong prevalence of L. multiflorum chromatin and intergeneric translocations were rare. In the tetraploid cultivar ‘Perun’ of the L. multiflorum × F. pratensis hybrid there were 11.7 chromosomes of L. multiflorum and 14.7 recombined chromosomes on average. Reasons for the domination of one of the parental genomes in hybrid cultivars are not clear and are only partially explained by breeding history. Recombination rates of individual genomes in hybrids involving F. arundinacea were evaluated in double hybridization experiments. The results indicated a strong affinity of the L. multiflorum genome for the F. pratensis genome present in F. arundinacea and little affinity for the F. glaucescens genome. This suggests that introgressions from F. arundinacea into L. multiflorum are primarily limited to the F. pratensis genome which can be more readily accessed in L. multiflorum × F. pratensis hybrids.     

Temperature‐Dependent Growth Adaptation of Festuca pratensis from the Biophysical Viewpoint

Meadow fescue (Festuca pratensis), Italian ryegrass (Lolium multiflorum), their hybrid Festulolium braunii and perennial ryegrass (Lolium perenne) were grown hydroponically under vegetative (20 °C) and hardening (8 °C, 0 °C) regimes. The relative shoot/root growth ratio K, linear root growth rate and other parameters were estimated. When the temperature was lowered from 20 °C to 8 °C, the relative shoot/root growth ratio K of F. pratensis decreased to 0.6, i.e. much more than that of Lolium, suggesting that at 8 °C F. pratensis shoot growth blockage occurs. Further, by dropping the temperature from 8 °C down to 0 °C, the K ratio of F. pratensis increased significantly, while the linear root growth rate decreased much more markedly than in the other species – it suggests also a blockage of root growth. This growth slowdown of stress‐tolerant F. pratensis shoots and at a lower temperature also of its roots is inadequate to the direct impact of temperature and, thus, indicates a two‐step qualitative reorganization, i.e. transition into a qualitatively new state of stress. This phenomenon does not manifest itself in the less stress‐tolerant Lolium and Festulolium species but is expressed in stress‐tolerators, allowing them to achieve their strategic goal – to survive under extreme conditions.     

Culture of embryos and shoot tips for chromosome doubling in Lolium perenne and sterile hybrids between Lolium and Festuca

An improved method is reported for polyploid induction in Lolium (ryegrass), and in sterile F₁ hybrids between Lolium and Festuca (fescue). Two factors greatly increased the survival rate of colchicine‐treated embryos of Lolium perenne (perennial ryegrass) germinated and cultured in vitro (1) a high concentration of sucrose (100 g/1) in the germination medium and (1), maintenance at a low temperature of 10°C for 2 weeks after treatment. The maximum success rate for chromosome doubling among survivors of perennial ryegrass was 79.1%, and for L. perenne×Festuca arundinacea F₁ hybrid embryos it exceeded 40%. The same doubling treatment also works with shoot tip culture in non‐ flowering genotypes obtained by anther culture of L. multiflorum×F. arundinacea hybrids.     

Genetic analysis of forage grasses based on heterologous RFLP markers detected by rice cDNAs

Genetic polymorphism within and between three species of forage grasses, perennial ryegrass (Lolium perenne L), meadow fescue (Festuca pratensis Huds.) and tall fescue (Festuca arundinacea Schreb.), was analyzed using restriction fragment length polymorphism (RFLP) markers detected by rice cDNA probes developed at the Rice Genome Research Programme of Japan (RGP). One hundred and ninety‐seven rice cDNA clones were used for hybridization to genomic DNA of forage grasses. Many of the rice cDNA clones produced no visible band or only a smear with no discrete bands. Twenty‐three clones showed high efficiency cross‐hybridization to the genomic DNA of forage grasses. Genetic variation was evaluated for five varieties and one population of forage grasses using 12 polymorphic rice cDNA RFLP probes. Genetic variability within varieties as measured by Rogers’ genetic distance was considerably lower for the F. pratensis variety ‘Tomosakae’ than for the L. perenne and F. arundinacea varieties. To determine the genetic diversity between varieties of different species, cluster analysis was performed using data from the 12 RFLP probes. The two accessions of Lolium perenne were clustered more closely together than the three varieties of F. arundinacea. Two Japanese varieties of F. arundinacea were grouped in the same cluster. The variety‐specific RFLP markers were seen among six accessions of L. perenne, F. pratensis and F. arundinacea. Such variety‐specific RFLP markers would provide very useful tools for breeding programmes such as the intergeneric hybridization of Lolium and Festuca genera.     

Habitat and management affect genetic structure of Festuca pratensis but not Lolium multiflorum ecotype populations

Genetic diversity present in permanent grassland may be valuable for broadening gene pools in breeding programmes and for conservation of genetic resources. However, little is known about the amount of genetic diversity present at specific habitats and about site‐related factors affecting it. To identify valuable habitats, genetic diversity of 12 ecotype populations and four reference cultivars of both Festuca pratensis Huds. and Lolium multiflorum Lam. was analysed using Simple sequence repeat markers (SSR). Analysis of molecular variance revealed a larger within population variation for L. multiflorum (97.1%) than for F. pratensis (92.6%). F. pratensis ecotype populations were clearly separated from cultivars and formed three distinct subclusters according to the geographic regions they were sampled from. Differences between L. multiflorum ecotype populations and cultivars were small and no grouping of populations was observed. Thus, only F. pratensis ecotype populations were structured and habitat as well as management had a slight influence on genetic structure. This information may allow the design of individual strategies for targeted utilization of genetic resources in plant breeding programmes.     

Hybrids of ryegrasses and meadow fescue and their value for grass breeding

1. 1.
   In attempts to intercross Lolium perenne and L. multiflorum with Festuca pratensis in a glasshouse, 13,284 emasculated flowers of diploid, triploid and tetraploid Lolium plants were provided with pollen of diploid and tetraploid Festuca plants. Only the combinations 2n\sx4n and 3n\sx4n produced hybrids, viz. on average 5.3 and 1.0 mature hybrid plants per 100 pollinated flowers. In crosses performed in winter these averages were 12.0 and 4.9 respectively.     
   
   

Progenies of allotriploids of Oriental × Asiatic lilies (Lilium) examined by GISH analysis

With the aim of utilizing allotriploid (2n = 3x = 36) lily hybrids (Lilium) in introgression breeding, different types of crosses were made. First, using diploid Asiatic lilies (2n = 2x = 24), reciprocal crosses (3x − 2x and 2x − 3x) were made with allotriploid hybrids (AOA) obtained by backcrosses of F₁ Oriental × Asiatic hybrids (OA) to Asiatic cultivars (A). Secondly, the AOA allotriploids were crossed with allotetraploid (OAOA, 2n = 4x = 48), in 3x − 4x combination. Finally, the AOA allotriploids where crossed to 2n gamete producer F₁ OA hybrids (3x − 2x (2n)). Two types of triploids were used as parents in the different types of crosses, derived from: (a) mitotic polyploidization and (b) sexual polyploidization. Ploidy level of the progeny was determined by estimating the DNA values through flowcytometry as well as chromosome counting. The aneuploid progeny plants from 3x − 2x and reciprocal crosses had approximate diploid levels and in 3x − 4x crosses and 3x − 2x (2n) the progeny had approximate tetraploid levels. Balanced euploid gametes (x, 2x and 3x) were formed in the AOA genotypes. Recombinant chromosomes were found in the progenies of all crosses, except in the case of 2x − 3x crosses through genomic in situ hybridization (GISH) analyses. Recombinant chromosomes occurred in the F₁ OA hybrid when the triploid AOA hybrid was derived through sexual polyploidization, but not through mitotic polyploidization with two exceptions. Those recombinant chromosomes were transmitted to the progenies in variable frequencies.