Cultivated pineapple is self-incompatible but sets seeds readily when cultivars from different groups are crossed. Pineapple cultivars are heterozygous, and hybridization is a valuable method in generating widely variable genotypes through gene recombination. With clear objectives and judicious selection of the parents, the hybrid population would present a spectrum of variation within which the desired improved genotypes may be found. It was believed that many important characteristics, such as high carotene content, greater translucency and lower acidity, that could not be achieved by clonal selection could be realized by hybridization (Wortman and Kerns, 1959).
Considerations in hybridization fertility, compatibility and direction of cross. Fertility and compatibility are important considerations in deciding the parents to be used for hybridization and also the direction of the cross, i.e. deciding which parent should be paternal or maternal. Chan (1986) found that, for crosses between 'Smooth Cayenne', 'Queen', 'Singapore Spanish' and 'Johor', a hybrid between 'Smooth Cayenne' and 'Singapore Spanish' was compatible, but there was differential seed set between the crosses and their reciprocals. 'Smooth Cayenne', for instance, has low fertility and was a poor female parent because only a few seeds were formed. However, reciprocal crosses using 'Smooth Cayenne' as the pollen donor provided satisfactory seed set in all crosses. In the extreme case, the 'Smooth Cayenne' female crossed with 'Johor' was incompatible but the reciprocal produced about 1000 seeds in a well-pollinated fruit (Chan, 1986). 'Smooth Cayenne' was also shown to be a poor female parent, bearing only about 200 seeds per fruit. When a reciprocal cross was made with the much more fertile 'Perolera' as the female parent, there were more than 2000 seeds per fruit (Cabral et al, 1993).
The direction of the cross was reported by Loison-Cabot and Lacoeuilhe (1990) to have an influence on the performance of the hybrids. In a cross between 'Smooth Cayenne' and 'Manzana', the latter used as the female parent improved the yield of the hybrid population compared with the reciprocal. In the case of total soluble solids (TSS), 'Smooth Cayenne' as the female parent generated progenies with higher TSS content.
synchrony in flowering. Synchrony in flowering of parents is obviously an important consideration in hybridization. The time taken for appearance of red-heart from time of hormoning is variable between various pineapple cultivars. Inflorescence appearance in 'Smooth Cayenne' is 10-12 days later than in 'Singapore Spanish' and 'Queen', respectively. This means that, while 'Singapore Spanish' and 'Queen' may be induced to flower simultaneously, 'Smooth Cayenne' should be induced about 2 weeks earlier in order that inflorescences of all varieties appear simultaneously for hybridization. Where the time to anthesis is unknown for each parent cultivar, a percentage of plants of each cultivar should be induced at, say, 5-day intervals.
time of day for hybridization. Dawn to 9 a.m. is probably the best time for pollination (Cabral et al., 1993; Sanewski, 1998). However, pollination may be done during the entire length of the day, although seed set may vary depending on the time of day and appears to be temperature-related (Chan, 1986). Higher seed set was obtained when crossing was done during the early morning or cooler evening and a significantly lower seed set when it was done during the hotter midday or early afternoon.
segregation in hybrid populations. Evaluation of hybrid populations from biparental hybridization indicated that there is wide segregation of progenies for most quantitative characters (Chan, 1989, 1991; Loison-Cabot, 1990; Sanewski, 1998). For TSS, for example, the range of values showed trans-gressive segregation, i.e. the minimum and maximum values of the hybrid progenies exceeded the lowest and highest values of both parents. Sanewski (1998) reported that the range of TSS from hybrids of the 'Queen' x 'Smooth Cayenne' cross was 9-24%, while the parents had a narrower range of 14-19%. The coefficients of variance in the hybrid population for many quantitative traits quite often exceeded twice that of the parents (Chan, 1989, 1991). As a matter of fact, the variation generated through hybridization of even closely related cultivars may be as large as that obtained for some pineapple collections (Chan, 1989). This wide variation will provide good opportunities for selection of improved recombinants.
In hybridization for improvement of pineapple, the choice of parents has usually been based on the strength of one in complementing the weakness of the other. For example, the ascorbic acid content of 'Smooth Cayenne' is improved using 'Perolera' ('Manzana'), which has higher ascorbic acid (Loison-Cabot, 1987), and 'Queen' for increasing TSS of 'Smooth Cayenne' winter fruits (Winks et al., 1985). Where biparental crosses are focused narrowly on the improvement of a single character, there is danger that the donor parent might inadvertently introduce other adverse characters.
In practice, breeders are always faced with selection of a set of economic traits and not just single characters. In this case, selection of suitable parents may be complicated. Chan (1991) computed breeding values for parents based on consideration of eight agronomic characters. The values were computed by obtaining the product of the percentage of progenies that qualified for selection in each of the eight characters. 'Queen' and 'Smooth Cayenne' were better all-round parents for hybridization because 1.1-1.2 % of the progenies qualified for selection when eight characters were considered. This was twice as much compared with the 'Singapore Spanish' parent. Sanewski (1998) used a similar computation and found that advanced hybrids, such as '73-50' developed by PRI, had a high breeding value. Sanewski (1998) concluded that these hybrids are valuable as parents because they have accumulated desirable genes during earlier multiple crossings and subsequent selection. Breeding values are helpful for guidance in the choice of parents and in ascertaining the probability of obtaining useful progenies. They are also useful for determining the population size of the hybrid progenies. Quite evidently, higher breeding values of parents will yield higher percentages of promising hybrids, thereby reducing the need for larger populations.
The high level of heterozygosity of pineapple cultivars reduces the chances of quickly obtaining progenies with a large complement of specific characteristics and it is this heterozygosity that is largely responsible for the slow rate of progress in most breeding programmes. Strategies such as selfing (Cabral et al., 2000), backcrossing and sib-crossing need to be investigated more thoroughly, especially considering the demonstrated inefficiencies of combining heterozygous genotypes. Other strategies to circumvent the heterozygosity of potential parents such as the development of homozy-gous haploid parents by anther and ovule culture have now been reported (Benega et al, 1998a; Cabral et al, 2000).
Hybridization produces recombinants, but potential selections that meet all the selection criteria of the breeder will appear very rarely. Therefore, very large populations are usually required to increase the chances in the draw of this genetic lottery. With the breeding values of parents known, the population size that may be required can be estimated (Chan, 1991). The number of selections in the second cycle that the breeder would want to evaluate must first be ascertained. At MARDI, it was put at 300 selections and hybrid populations of 23,000-45,000 progenies, depending on the parents' breeding values, will be needed to have a realistic chance that the desired number of potential recombinants would be generated. Loison-Cabot (1987), using 'Smooth
Cayenne' and 'Perolera' parents, recommended 30,000 progenies for each annual selection cycle, while Cabral et al. (1993) and Benega et al. (1998b) maintained 23,000 and 34,000 progenies, respectively, to obtain 120-169 first-cycle selections. The PRI breeding programme has been the most extensive to date and, at its peak, was producing over 1 million seedlings per year. Up to 100,000 seedlings were developed from some of the better parent combinations (Williams and Fleisch, 1993). Sanewski (1998), using some of these PRI selections, obtained 300 selections from populations of 50,000 seedlings because the parents have very high breeding values.
Some interest has been shown in the use of other species for imparting characteristics such as disease resistance. Collins (1960) indicated that, where crosses with wild germplasm were performed, only 1000 progenies were necessary in the first generation but 15,000-25,000 per backcross were required for each of four backcross generations.
Pollen is collected from flowers at anthesis early in the morning. The flowers are removed from the inflorescence by making three deep triangular incisions into the base of the flower and then gouging them out. This method of collecting pollen from excised flowers is better than removing only the stamens. Pollen in excised flowers should stay fresh longer and dehydrates less rapidly. There are no conclusive data on the storage of pollen, but Kerns (1931) indicated that 'Smooth Cayenne' pollen could be stored in cool, dry air for 15 days with minimal loss of viability.
During hand-pollination, the petals of the flower are removed to expose the freshly dehisced anthers with an abundance of pollen. The anther is picked up with a fine pair of forceps and gently brushed on to the stigma of the flower to be pollinated. In most cases, there is no necessity to emasculate the flowers before pollination because most cul-tivars are self-incompatible. In cases where fidelity of the pollen source is vital, as in genetic studies, or when natural pollinators are present, the contamination of foreign pollen can be prevented by using a fine mosquito netting or paper bag placed over the inflorescence (Leal and Coppens d'Eeckenbrugge, 1996). Depending on varieties, the number of flowers in an inflorescence ranges between 75 and 150. About three to seven flowers open daily and 2-3 weeks may be required to complete pollination of the whole inflorescence (Chan, 1986).
Was this article helpful?