Triploid tulips have agronomically desirable traits such as vigorous growth and large flower size, but only a portion of all cultivated tulips is triploid. To apply 2n pollen to polyploid breeding of tulips, the polyploidizing agent, nitrous oxide gas (N 2 O), was applied to bulbs. In tulips, meiosis in anthers occurs inside the bulbs from mid-to late-October. When meiosis in anthers (excised from bulbs) reached metaphase I, we treated other bulbs of the same clones with N 2 O for 24-48 h. Most of the treated plants produced pollen grains with a wide-ranging or bimodal size distribution, indicating a mixture of n, 2n and aneuploid pollen grains. The use of pollen containing a relatively high proportion of giant pollen grains tended to yield larger numbers of triploids in the progeny. The number of giant pollen grains could be increased when N 2 O-treated pollen grains were suspended in 10% sucrose and then sieved through a nylon mesh. Very few polyploids were observed in some cross combinations, even those involving pollen with a relatively high proportion of giant grains. Even so, this low polyploid yield most likely is due to a triploid block, because the capsules obtained in the crosses of the diploid × N 2 O-treated plants contained some abnormal seeds, which were mostly triploid. Embryo culture was useful in rescuing abnormal embryos. The present study reveals that 2n pollen can be produced at high frequency using N 2 O during tulip breeding.

Chromosome morphology was studied in diploid cultivars of Tulipa fosteriana and T. gesneriana (2n = 2x = 24) and triploid Darwin hybrids (2n = 3x = 36) developed from interspecific crosses of T. gesneriana and T. fosteriana. Chromosomes were arranged in the karyotype according to decreasing total length. Based on our karyotypic analysis, we propose that median chromosomes may serve as markers for diploid genotypes. Discriminant analysis with respect to total chromosome length and short arm length showed a significant difference between the size of the larger median chromosomes of T. gesneriana and T. fosteriana. Comparison of median chromosome length in Darwin hybrid tulips showed that two larger chromosomes and one smaller chromosome were derived from T. gesneriana and T. fosteriana, respectively. This finding was clearly and unambiguously confirmed by simultaneous hybridization of differentially labeled genomic probes of T. fosteriana and T. gesneriana to metaphase chromosomes of the triploid The first and second author have contributed equally to this paper

In this study, Tulipa fosteriana was found to contain 45S rDNA repeat units of 9.7 and 9.5 kb, in which at least 7 types of 45S rDNAs were identified by restriction site analysis. For 5S rDNA, repeat units ranging from 364 bp to 396 bp were identified. The diploid cultivars (2n = 2x = 24) 'Christmas Dream' and 'Queen of Night,' representing the horticultural group T. gesneriana, and 'Red Emperor', belonging to T. fosteriana, were compared cytogenetically using cloned 5S and 45S rDNAs. Fluorescence in situ hybridization (FISH) analysis identified many rDNA sites located on each chromosome in the diploid genomes. For example, we identified 71 sites of 5S rDNA and 10 sites of 45S rDNA in 'Red Emperor'. Additionally, FISH analyses enabled construction of karyotypes for these cultivars. Karyotype comparison of T. gesneriana cultivars showed conservation of repetitive rDNA unit positioning. A clear difference in chromosome size and signal pattern was observed between T. gesneriana and T. fosteriana cultivars.Here we demonstrate the unique nature of the highly repeated 5S rDNA units in these Tulipa species and the usefulness of FISH karyotyping with cloned 5S and 45S rDNAs to clearly distinguish between chromosomes from T. gesneriana and T. fosteriana.

No abstract

Most Liliaceae plants have the tetrasporic Fritillaria-type embryo sac and normally form diploid embryos and pentaploid endosperms derived from a 4:1 maternal-to-paternal genome ratio (4m:1p) after double fertilization. Here we characterize embryo sac and endosperm formation in Tulipa spp. of Liliaceae. Chromosome analysis using seeds derived from 2x x 2x crosses of Tulipa gesneriana (2n = 2x = 24) identified diploid chromosome number in the endosperm. Similarly, flow cytometric analysis confirmed diploid endosperm formation in T. gesneriana, T. fosteriana (2n = 2x = 24) and T. greigii (2n = 2x = 24). To further study the possible mechanism of diploid endosperm formation, we made interploidy crosses of triploid (2n = 3x = 36) x diploid in which aneuploid seeds with various chromosome numbers (2n = 25-36) were produced. Again, flow cytometric analysis confirmed the same ploidy level in both embryos and endosperms at all aneuploidy levels, suggesting that only a single haploid polar nucleus contributes to endosperm formation at fertilization. Histological observation further confirmed the physical separation of two polar nuclei by a large vacuole in the Fritillaria-type embryo sac of T. gesneriana that appeared to prevent the fusion of the two polar nuclei that originated at the micropylar and chalazal ends before fertilization. Taken together, these results indicate that diploid endosperms (1m:1p) are normally formed in Tulipa spp. by fusion of the micropylar polar nucleus (n) and a spermatid (n) but not by normal triple fusion. We also show that tulip endosperm partially overcomes the triploid block mechanism that occurs in interploidy crosses. Based on these observations, the possible role of triple nuclear fusion in double fertilization is discussed.