Ploidy level was estimated in Hydrangea macrophylla (Thunb) Ser. using flow cytometry. For H. macrophylla ssp. macrophylla, 42 diploid and 19 triploid cultivars were identified. All 14 H. macrophylla ssp. serrata (Thunb.) Makino cultivars tested were diploids. Somatic chromosome counts confirmed the ploidy of three diploid (2n = 2x = 36) and three triploid (2n = 3x = 54) cultivars. Stomatal guard cell length and pollen diameter of H. macrophylla ssp. macrophylla diploid cultivars were smaller than those of triploid cultivars. However, because the range of measurements for the diploids overlapped that of the triploids, neither guard cell nor pollen measurements are recommended for determining ploidy of H. macrophylla cultivars. Fertility was estimated using pollen staining and controlled pollinations. Stainable pollen for triploid cultivars averaged 63% and ranged from 25% in ‘Masja’ to 85% in ‘Marechal Foch’. Viable seed was obtained when four triploid cultivars were used as pistillate or staminate parents in controlled pollinations to diploid H. macrophylla ssp. macrophylla cultivars. A bimodal distribution of pollen sizes, which is suggestive of unreduced gamete production, was observed in one cultivar; however, more detailed genetic and cytologic studies are needed to elucidate the mechanism behind triploid formation in H. macrophylla taxa.

The potential of producing an intergeneric hybrid between Dichroa febrifuga Lour. and Hydrangea macrophylla (Thunb.) Ser. was investigated. Reciprocal hybridizations were made between a D. febrifuga selection (GUIZ 48) and diploid (‘Veitchii’) and triploid (‘Kardinal’ and ‘Taube’) cultivars of H. macrophylla. Embryo rescue was employed for about one-third of the crosses that produced fruit, and the rest were allowed to mature on the plant and seed-collected and germinated. Reciprocal hybrids, which were verified with simple sequence repeat markers, were produced from both embryo rescue and seed germination and with both diploid and triploid H. macrophylla cultivars. Hybrids were intermediate in appearance between parents, but variability in leaf, inflorescence, and flower size and flower color existed among the hybrids. A somatic chromosome number of 2n = 6x = 108 was tentatively proposed for D. febrifuga GUIZ 48. Chromosome counts and flow-cytometric measurements of nuclear DNA content indicated that some of the hybrids may be aneuploids, but neither analysis was definitive. Although hybrids with H. macrophylla as the pistillate parent did not form pollen-producing anthers, D. febrifuga × H. macrophylla hybrids had normal-appearing anthers that produced abundant pollen. F2 and BC1 progeny were obtained using D. febrifuga × ‘Veitchii’ hybrids. This work documents the first step in an effort to combine desirable horticultural features from D. febrifuga and H. macrophylla.

No abstract

Previous attempts to use interspecific hybridization to combine flower color and cold hardiness in Hydrangea have not produced the desired results, with confirmed hybrids being weak, sterile or aneuploid. In all cases, H. macrophylla (Thumb.) Ser. was used as the source of flower color. This work investigates the use of H. involucrata Sieb. as an alternative source of flower color in Hydrangea interspecific hybridizations. Controlled reciprocal pollinations of H. involucrata with two cultivars of H. arborescens L. and three cultivars of H. paniculata Sieb. were made. Hybridity of progeny was verified using RAPD markers and confirmed with chromosome counts and morphological comparisons of hybrids and parents. Plants were obtained only when H. involucrata was used as the pollen parent. No hybrids between H. paniculata or H. arborescens `Annabelle' and H. involucrata were produced. Seven H. arborescens `Dardom' × H. involucrata progeny showed either a sum of the RAPD bands of both parents or banding patterns that matched those of H. involucrata. Leaf blade length and length/width ratio of the hybrid were intermediate to its parents. Chromosome number in the hybrid (2n = 34) was also intermediate between H. arborecens (2n = 38) and H. involucrata (2n = 30). One `Dardom' × H. involucrata plant flowered in 2005. While pollen staining indicated a very low level of fertility, we will continue to evaluate the possibility of using the hybrid for producing advanced filial or backcross progeny.

No abstract