Plant breeders would welcome new tools to improve selection efficiency for complex traits such as improved flavor, especially since this is only one of many complex traits that a breeder has to integrate into improved cultivars. Using tomato (Lycopersicon esculentum Mill.) flavor as an example, a major obstacle to improvement is the lack of measurable traits to select for. It has been suggested that improved flavor can be achieved by increasing the soluble solids and acidity. Both of these traits are not simply inherited, but if fruit sampling is adequate, they can be measured and selected. Studies have located several molecular markers linked to high soluble solids, but some are also linked to undesirable traits such as small fruit size or low yield. Thus, the molecular markers are not being used in breeding programs at this point. Moreover, other studies have shown that flavor is also influenced by an array of aromatic volatiles. The importance of some of the volatiles has been reported, but the volatile profile that consistently results in superior tomato flavor is still not known. Molecular manipulation of a biochemical pathway has been done to increase the concentration of one volatile with positive results. However, this manipulation does not solve the overall flavor improvement problem. Furthermore, environment plays a profound role in tomato flavor, and this aspect needs to be dealt with if a branded high-quality product is to be successfully marketed. There are also flavor issues related to fruit firmness, pedicel type, and plant habit. In summary, molecular techniques may be useful in providing some incremental improvements for complex traits like tomato flavor, but more knowledge about targets to manipulate is required. There does not appear to be any cheap or easy solutions. If molecular approaches are to be commercially successful, they will have to be tied closely to a breeding program dedicated to the same goal.
J. W. Scott and W. L. George Jr.
Flowers of 3 parthenocarpic tomato (Lycopersicon esculentum Mill.) genotypes, ‘Severianin’ (pat-2), PSET-1 (pat-2) and RP 75/79, and one nonparthenocarpic genotype, ‘Tiny Tim’ (pat-2+), were given various pollination treatments in a greenhouse study: a) natural pollination, b) pollination with a hand vibrator, c) emasculation followed by pollination with mentor (heat-killed) pollen, d) emasculation only, e) removal of stigmas, and f) removal of the distil half of the anthers and styles. RP 75/59 had greater total fruit set than did the pat-2 genotypes. Treatments preventing pollination inhibited seed formation of all genotypes and inhibited fruit set of ‘Tiny Tim’. Mentor pollen did not increase parthenocarpic fruit set for any genotype. Vibration of the flowers completely inhibited parthenocarpic fruit set for all genotypes except PSET-1, which had 15.7% parthenocarpic fruit. No difference in fruit size or days to ripening for seeded vs. seedless fruit were evident for the pat-2 genotypes. Seeded fruit of RP 75/59 were larger than seedless fruit (34.2 g vs. 20.3 g), and also took longer to ripen (2.9 days).
J.W. Scott and J.P. Jones
Forty-two Lycopersicon pennellii Corr. D'Arcy accessions, from the Tomato Genetics Stock Center, were inoculated for resistance to Fusarium wilt race 3 at the 3-leaf and cotyledon stage. All were over 90% healthy when inoculated at the 3-leaf stage but had greater disease incidence at the cotyledon stage. Crosses were made between healthy plants within each accession. Using this seed, 39 accessions were 100% healthy and 3 were over 96% healthy when inoculated at either stage. Seventeen F1's with susceptible parents were tested for race 3 and all had over 80% healthy plants. Twenty-two accessions were tested for Fusarium wilt race 1 and race 2. For race 1, 21 were 100% healthy and 1 was 91% healthy, For race 2, 20 were 100% healthy, 1 was 96% healthy, and 1 was 75% healthy. Forty accessions were screened for Fusarium crown rot and Verticillium wilt. For crown rot, LA 1277, LA 1367, and LA 1657 were over 95% healthy, 6 other accessions were over 68% healthy and several others had over 50% healthy plants, All 40 were susceptible to Verticillium wilt race 1. L. pennellii appears to be a good source of resistance to Fusarium sp. but not to Verticillium wilt.
J.W. Scott and J.P. Jones
Lycopersicon pennellii accession LA 1277 was crossed to tomato (L. esculentum) and the F1 was backcrossed to tomato. Self-pollinated seed was saved from backcross plants and seedlings derived were inoculated with Fusarium oxysporum Schlecht f.sp. radicus-lycopersici Jarvis and Shoemaker, the causal agent of Fusarium crown and root rot (FCRR). Seed was saved from resistant plants that were self-pollinated and screened until homozygous resistance was verified five generations after the backcross. Three homozygous lines were crossed to Fla. 7547, a tomato breeding line susceptible to FCRR but resistant to Fusarium wilt races 1, 2, and 3. Subsequently, backcrosses were made to each parent and F2 seed were obtained. The three homozygous FCRR-resistant lines were also crossed to Ohio 89-1, which has a dominant gene for FCRR resistance presently being used in breeding programs. F2 seed were obtained from these crosses. These generations were inoculated with the FCRR pathogen. The resistant parents, F1, and backcross to the resistant parents were all healthy. The backcross to the susceptible parent and the F2 segregated healthy to susceptible plants in 1:1 and 3:1 ratios, respectively. Thus, the resistance from LA 1277 was inherited as a single dominant gene. This gene was different than the gene from Ohio 89-1 because susceptible segregants were detected in the F2 generation derived from the two resistant sources.
J. W. Scott and J. B. Jones
Two hundred eight-four Lycopersicon spp. genotypes reported to have some resistance to bacterial pathogens of tomato (L. esculentum Mill.) were inoculated in the field with Xanthomonas campestris pv. vesicatoria (XCV), the incitant of bacterial spot, and rated for disease severity in summer 1982 and/or summer 1983. One line tested in 1983, Hawaii 7998, had no definite XCV lesions and later was determined to be resistant to XCV in the laboratory. Genotypes with the highest levels of resistance during 2 years of testing were: Ohio 4013-3, Ohio 4014-4, Heinz 1568-F3, [(Subarctic Delite × MH1) × H603] F5, L556, ‘Campbell-28’, PI 127813, Heinz 603-F11, PI 224573, ‘Monense’, ‘Heinz 2990’, and PI 324708. Genotypes with highest levels of resistance in one year of testing were PI 379032 and ‘Burgess Crack Proof. In 1982, PI 270248- ‘Sugar’ had a high level of resistance to XCV on fruit, but foliage was susceptible.
J. W. Scott and C. L. Emmons
Ten tomato cultigens were crossed with L. peruvianum accessions PI 126443 and PI 129152. Fruit (536 total) were harvested between 15 and 65 days after anthesis (DAA). Culturable embryos were obtained from 13% of the fruit. There were 140 embryos plated, from which 36 plants were obtained (7% of fruit, 26% of embryos plated). 'Campbell 28', Fla. 7217, and Fla. 7182 were the most efficient tomato lines for producing F1 plants, there was no difference between the L. peruvianum accessions. No embryos were obtained beyond 57 DAA. No trend in embryo viability was detected between 15 and 56 DAA. Of 248 backcross fruit, 94 embryos were plated (38% of fruit) and 15 plants were obtained (6% of fruit, 16% of embryos plated). Female parents with the best percentage of plants per fruit crossed were Fla. 7217, Fla. 7215, and 'Campbell 28' with 15, 8, and 7%, respectively. No plants were obtained from 45 crosses on Fla. 7182.