The difficulty in finding a superior recombinant among the widely varying progenies obtained through direct hybridization of highly heterozygous pineapple cultivars is the main cause of the failure of many breeding programmes (Collins, 1960; Williams and Fleisch, 1993; Coppens d'Eeckenbrugge and Duval, 1995). This led some researchers to evaluate the potential of inbreeding in pineapple breeding. Thus, Collins (1960) selfed a mutant of 'Smooth Cayenne' and observed in the resulting progenies a loss of vigour and a wide range of variation for qualitative and quantitative traits. In the second generation, inbreeding depression was so severe that only a few plants produced fruits and none of them set seed. Backcrosses of first-generation inbreds to 'Smooth Cayenne' restored vigour. Collins (1960) suggested the possibility of using this method to synthesize new cultivars very similar to 'Smooth Cayenne'. Following a similar idea, Coppens d'Eeckenbrugge and Duval (1995) proposed the introduction of one cycle of selfing before the final hybridization to produce inbred genotypes from the parental cul-tivars. Crossing genotypes with 50% homozygosity would considerably lower the variability of the hybrid progeny. In addition, studies on segregation of selfed progenies would provide specific information on the genetic nature of the parental cultivars, allowing identification and selection of recessive favourable alleles and elimination of codominant and recessive unfavourable alle-les. The final hybridization between the best inbreds from two different cultivars would restore vigour.
In pseudo-self-compatible cultivars, a few inflorescences would be sufficient to obtain the small number of self-seeds required. In addition, selfing is much less labour-intensive than crossing since these few seeds can be obtained by simply bagging the inflorescence before anthesis (Coppens d'Eeckenbrugge et al., 1993). However, for strongly self-incompatible cultivars, a high number of selfings may be necessary. Attempts have been made to overcome self-incompatibility with X-rays (Marr, 1964) or gamma rays (Subramanian et al., 1981). Bhowmik and Bhagabati (1975) tested bud pollination, style-grafting and naphthalene acetic acid (NAA) application, but only the last technique produced a few self-seeds in 'Queen'. Others have proposed obtaining homozygous plants through the production of doubled haploids. However, Collins's results throw some doubt on the viability of this approach. Highly depressed and sterile homozygotes obtained through this method would be of little use for breeding.
Coppens d'Eeckenbrugge et al. (1993) observed similar seed abortion rates after self- and open pollination of 71 cultivated and wild clones, which indicates no depression effect at the zygotic or embryonic level. No difference was observed in germination of self- and cross-seeds or in the early development of the resulting seedlings from 'Española Roja', 'Primavera', two clones of A. comosus var. ananassoides and two clones of A. comosus var. bracteatus. When transplanted to the field, these seedlings showed about 50% size reduction compared with progenies from open pollination. In contrast, a strong growth depression was observed in the early development of seedlings of self-progenies from 'Manzana' and from a clone of A. comosus var. ananassoides, causing their death in the nursery.
Was this article helpful?