Interspecific hybridization in the genus Tulipa L.

The genus Tulipa L. comprises about 55 species. The tulip species are classified in two subgenera, Tulipa and Eriostemones, which are subdivided into five and three sections respectively. Commercial tulips are mainly cultivars of T. gesneriana L . and of Darwin hybrids, the latter of which are obtained by interspecific hybridization between T. gesneriana and T. fostefiana Hoog ex . Irving. The present-day tulip assortment could be improved considerably by introducing traits from other tulip species, such as resistance for diseases and short forcing period. With conventional breeding methods, hybrids can only be obtained from a limited number of interspecific crosses. Most interspecific crosses are hampered by crossing barriers. The main goal of the research on crossing barriers in tulip was to bypass crossing barriers between cultivars of the present-day assortment and the other tulip species.This main goal was divided into two subgoals: (1) the identification of crossing barriers in crosses between T. gesneriana and a number of representative tulip species and (2) the development of techniques to bypass these barriers.Pre-fertilization development and post-fertilization development have both been studied in a number of crosses to identify crossing barriers. Pre-fertilization development was analyzed in crosses between cultivars of T.gesneriana and 13 tulip species from all eight sections of the genus Tulipa (Chapter 2). Pollen tube growth in the pistil and pollen tube penetration in the ovules showed much variation between the crosses in progamic development. Depending on the cross, pollen tubes grew as far as the stigma or the style or continued growing down into the ovary. This resulted in percentages of ovules with pollen tube penetration varying between 0%-79%. The percentages ovules with pollen tube penetration differed between the flowers of a specific cross, between the different years and between the maternal genotypes used.The progamic phase and embryo and endosperm. development were studied in the incongruent cross T.gesneriana x T. agenensis DC. (former name T. oculus-solis St. Amans) (Chapter 3) and in a compatible T. gesneriana cross. Less pollen tubes penetrated the ovules in the interspecific cross in comparison with the compatible cross. In the interspecific cross, the embryo development was more often retarded and higher percentages of ovules with aberrant embryogenesis (from 32 days after pollination) were found than in the compatible cross. In most ovules with aberrations in embryo development, also the endosperm showed abnormalities. Endosperm degeneration seems to be the major cause of starvation of the embryo before seed maturity in the cross T. gesneriana x T. agenensis.Several techniques have been studied for bypassing pre-fertilization barriers in interspecific tulip crosses: the cut-style method, the grafted-ovary method and placental pollination (Chapter 4). The application of the cut-style method or the grafted-ovary method in the interspecific crosses did not improve the pollen tube growth and pollen tube penetration in the ovules as compared to ordinary stigmatic pollination. The percentage of ovules with pollen tube penetration was also not increased after placental pollination. However, the pollination procedure applied for placental pollination might still be optimized to improve the percentage of ovules penetrated by pollen tubes.The effect of hormone treatments and embryo rescue techniques on the ability of bypassing post-fertilization barriers was studied. Ovaries were treated with the cytokinin BAP (0. 1 % or I %) or the auxin NAA (1 %) at 12 days after pollination (Chapter 4). Seeds of the cross T . gesneriana x T. agenensis were obtained on the plant after treating ovaries with 0. 1 % BAR Seed production on the plant from this cross has not been reported previously. Additional research is needed before definite conclusions can be drawn about the effect of hormone treatments for bypassing crossing barriers in a wider range of interspecific tulip crosses.Two embryo rescue techniques were studied: ovary-slice culture and ovule culture. Compatible T. gesneriana crosses were used as model system. The efficiency of direct ovule culture and ovary-slice culture followed by ovule culture was compared for cultures started at 2-9 weeks after pollination (Chapter 5). The influence of media composition on the percentage of ovules showing germination was also studied (Chapter 6). In most cases, the percentage of embryos that germinated increased significantly with a more advanced developmental stage of the embryos at the start of the culture. The lower efficiency at early culture dates is due to embryo abortion and retarded embryo development. The germination percentages for ovary- slice culture followed by ovule culture, started at various dates, were for some culture dates comparable to direct ovule culture, but for other culture dates significantly higher. The length of the period of ovary-slice culture prior to ovule culture mostly did not affect the germination percentage.Media composition influenced the percentage of ovules showing germination after embryo rescue. The germination percentage was influenced by the sucrose concentration used for ovary-slice culture. 9% Sucrose resulted in higher germination percentages as compared to 5 % sucrose. The sucrose concentrations in media used for subsequent ovule culture (3 %, 5%) or in media used for direct ovule culture (3%, 6%, 9%) did not influence the germination percentages. Analysis of the carbohydrates concentrations revealed that the total concentration of carbohydrates decreased with 22%-50% in media for ovary-slice culture (9% sucrose), whereas the total concentration of carbohydrates remained rather constant in ovule culture (3% sucrose) applied after ovary-slice culture. Comparable germination percentages were obtained by using media with the full or half of the concentrations micronutrients and macronutrients of the MS-medium during ovary-slice culture and subsequent ovule culture. For direct ovule culture, started at 4, 6 and 8 weeks after pollination, the germination was not improved by the addition of the cytokinin BAP (0.01 and 0. 1 mg/1), nor by the use of liquid shaken culture.An in vitro pollination procedure has been developed in order to perform an integrated system of pollination, fertilization and embryo rescue techniques under optimal controlled environmental conditions. Once an in vitro pollination procedure has been developed, it can also be used for the post-fertilization culture of whole ovaries. By using compatible intraspecific T. gesneriana crosses as model, the effect of the following media components on seed set and seed germination could be studied (Chapter 7): concentration of macronutrients and micronutrients of MS-medium, concentration sucrose, fructose and glucose, the auxin NAA, the cytokinin BAP, the gibberellin GA 3 , spermine, casein hydrolysate, the buffer MES and the use and type of agar. The application of MS-medium at full strength, with 5%-7% sucrose and 1 mg/l NAA proved to be suitable. Analysis of carbohydrate uptake revealed that on liquid media significantly more carbohydrates were absorbed by the ovaries than on agar solidified media. However, the numbers of seeds produced were higher on media with agar than on liquid media. The A of Daichin agar improved the seed set as compared to bacteriological agar. The addition of the buffer MES to liquid media, to control the pH, affected the seed set positively. The effect of the use of casein hydrolysate needs additional research. Other components tested had no or negative effects on the seed set and/or seed germination.Unique hybrids have been obtained from the crosses T. gesneriana x T. praestans Hoog and T.gesneriana x T. agenensis by using hormone treatments and/or ovary-slice culture and ovule culture. These techniques proved to be suitable for rescuing embryos of incongruent interspecific tulip crosses. Improvement of culture conditions and culture media can still increase the efficiency, especially at early culture dates. Bulblet formation in vitro after embryo germination and the transfer of the bulblets grown in vitro into the soil were not the subject of our research, but proved to be problematic. For hybrid plant production, it is therefore of great importance to improve the methods of hybrid plant recovery after embryo germination in vitro. For the introduction of genes for resistance in the cultivar assortment of tulip by means of the in this study developed methods, it is of importance to locate possible resistance genitors in the tulip species. Bypassing F1-sterility due to the formation of triploids and sterile diploids in interspecific tulip hybridization, also needs additional research. Traits from a number of tulip species such as forcing time, flower colours, flower shapes and other morphological characteristics can already be introduced in the present-day assortment with the aid of the described embryo rescue techniques. This can give a new impulse to e breeding of tulips