Читать книгу Cucurbits - James R. Myers - Страница 32

Augustin Sagaret and Charles Naudin tried to cross melon with cucumber in the mid-19th century, but without success. All later investigators trying to cross these distantly related species have also failed, but interspecific hybrids have been produced for other species of Cucumis (Fig. 3.2), as well as for Cucurbita (Fig. 3.3), Citrullus, Luffa, Momordica, Trichosanthes and other cucurbit genera. Cucurbita is an example of where interspecific gene transfer has been utilized successfully for crop improvement. However, Cucumis sativus has been improved using genes from C. sativus var. hardwickii. Also, C. sativus was crossed with wild Cucumis hystrix and the chromosomes doubled to create a new allotetraploid, C. hytivus. Citrullus lanatus has been improved using Citrullus mucosospermus, C. amarus and C. colocynthis.

Fig. 3.2. Crossability polygon of Cucumis species. Arrows point in the direction of the female parent. Moderately to strongly self-fertile and cross-fertile hybrids (thick solid line); sparingly self-fertile and moderately cross-fertile hybrids (thin solid line); self-sterile, usually not cross-fertile hybrids (dash-and-dot line); inviable seeds or seedlings (dashed line). Absence of a line indicates that seeded fruit were not obtained. (Nijs and Visser, 1985. Reprinted courtesy of Kluwer Academic Publishers.)

Fig. 3.3. Crossability polygon of Cucurbita species. ‘Digitata Complex’ includes C. digitata, C. palmata, C. cylindrata and C. cordata, which are considered subspecies of C. digitata by some scientists. All crossing combinations have been tried in at least one direction, except for C. pedatifolia with C. maxima, and C. radicans with C. pedatifolia, C. ficifolia, C. ecuadorensis, C. okeechobeensis, and C. digitata sensu lato. Early works describing hybridization attempts with C. radicans were in error as described in Merrick (1990); the material was misidentified. Other published sources, as well as the unpublished work of Tom Andres (New York, 1996, personal communication), were used to create this diagram. Solid lines indicate an F₁ hybrid that is at least partially fertile; dashed lines indicate a viable but sterile F₁ plant.

Although the production of interspecific hybrids is only the first step in a rather long process, F₁ hybrids of Cucurbita maxima × C. moschata are used directly to produce elite cultivars. Both parental species are monoecious, having many more male than female flowers, but the interspecific hybrid is gynoecious or predominantly gynoecious in sex expression. The interspecific hybrid is usually productive if conditions for pollination are favourable (e.g. a monoecious cultivar is grown nearby to provide pollen, and bees are working the field). The unusual case of a gynoecious hybrid being produced by crossing two monoecious species also occurs in the cross of C. pepo × C. ecuadorensis.

The cross C. maxima × C. moschata may be difficult or easy to make depending on parental combinations (Castetter, 1930; Yongan et al., 2002; Karaağaç and Balkaya, 2013). For difficult crosses, many pollinations may be needed to set each fruit, and only a few seeds per fruit are produced. Breeders in Japan have been able to make this cross so successfully that they market interspecific F₁ seed commercially. They have found C. maxima and C. moschata parents that cross well and are more compatible than most members of these species. ‘Tetsakabuto’, the first popular interspecific hybrid squash cultivar, is a cross of ‘Delicious’ (C. maxima) × ‘Kurokawa No. 2’ (C. moschata). Seed production for the interspecific hybrid is most prolific when C. maxima is the maternal parent.

The cross C. maxima × C. pepo is difficult but not impossible to make; the F₁ is highly sterile. C. pepo and C. moschata can be crossed, but compatibility is influenced greatly by the choice of parents, with some cultivars crossing more readily than others. Wall and York (1960) reported that the cross was more easily made when one of the parents was an F₁ hybrid, which increased gametic diversity. The bush gene of C. pepo has been introgressed into C. moschata to incorporate compact plant habit into that species. Zucchini yellow mosaic virus (ZYMV) resistance, derived from ‘Nigerian Local’ (C. moschata), has been moved into C. pepo to produce resistant cultivars such as ‘Tigress’ and ‘Jaguar’.

Of all the Cucurbita species, C. moschata and C. argyrosperma are the most closely related and hybridization may occur in nature, especially where these two species are sympatric in Central America (Wessel-Beaver, 2000). The interspecific cross is easy to make, although only with C. moschata as the male parent (Wessel-Beaver et al., 2004). F₁ hybrids with commercial potential are under development in Mexico (Ortiz-Alamillo et al., 2007).

Wild Cucurbita species are being used to develop disease-resistant squash. In disease studies, C. ecuadorensis and C. foetidissima (buffalo gourd) were found to be resistant to a greater number of viruses than other species of Cucurbita tested (Provvidenti, 1990). Buffalo gourd is difficult to use in squash breeding because of its distant relationship and incompatibility with the cultivated species (Fig. 3.3). However, virus resistance alleles have been successfully introgressed from C. ecuadorensis to C. maxima. Multiple virus-resistant germplasm derived from C. maxima × C. ecuadorensis was developed at Cornell University, USA, and provided to breeders in 1985. ‘Redlands Trailblazer’, a winter squash resistant to ZYMV, watermelon mosaic virus (WMV) and papaya ringspot virus (PRSV), was bred in Australia from the same interspecific cross. Cultivars of C. pepo vary considerably in their compatibility with C. ecuadorensis (Robinson and Shail, 1987). Distant interspecific crosses of Cucurbita can be made more successful if embryo culture is used.

Resistances to cucumber mosaic virus (CMV) and powdery mildew have been transferred from Cucurbita okeechobeensis to C. pepo and C. moschata. The cross C. okeechobeensis × C. pepo is difficult to make, even with embryo culture, and sterility is a serious problem with the interspecific hybrid. These difficulties are overcome by the use of a bridging species, for example crossing C. pepo to the F₁ of C. okeechobeensis × C. moschata (Whitaker and Robinson, 1986).

Because undesirable traits (e.g. bitterness) of a wild parent are often dominant, unacceptable horticultural types predominate in the F₂ generation of these interspecific crosses. However, only a single backcross to ‘Butternut’ (C. moschata) of the F₁ of ‘Butternut’ × C. okeechobeensis ssp. martinezii was needed to produce disease-resistant (powdery mildew and CMV) plants with fairly good fruit characteristics. In general, the number of backcrosses to the domesticated parent that are needed depends on the number of genes that are different between the two parents. Once acceptable progeny are produced, self-pollination is performed to obtain uniformity.

Wild species of Cucurbita could probably be used further in squash breeding programmes to provide resistance to other pathogens and insects. Other desirable traits, such as the drought tolerance and gynoecious sex expression of C. foetidissima, could be transferred as well.