The instrumentation required for objective evaluation of flavor has been available for only a decade and a half. Most information on the chemistry of flavor has accumulated within the past 5 years. There have been few studies on intercultivar variation of compounds important to flavor. Because an appreciation of the physiology of flavor and an understanding of flavor chemistry are basic to the genetics of flavor, it seems appropriate to discuss briefly these topics.
Investigation of the polyene-carotene and volatile compound content of tomatoes showed that there is a high correlation between the following relationships: Geranylacetone and farnesal with phytofluene plus 2 ζ-carotene plus neurosporene: 6-methylhept-5-en-2-one and citral with neurosporene plus 2 lycopene plus 2 prolycopene plus γ-carotene plus δ-carotene: β-ionone with γ-carotene plus 2 β-carotene plus α-carotene; α-ionone with δ-carotene plus α-carotene. Also, there appears to be a relationship between farnesylacetone and 2 phytoene plus phytofluene. The basis for establishing these relationships was oxidation of the polyene-carotenes at the first conjugated diene bonds. These relationships may exist because the volatile compounds result from oxidation of the polyene-carotenes.
The volatile compounds of tomato cultivars ‘Campbell 146’ and ‘Campbell 1327’ were studied by gas-liquid chromatography. Heritable concentration differences were found for 3 compounds identified by retention time and infrared spectroscopy as 2-isobutylthiazole, methyl salicylate and eugenol. Investigation of concentrations and flavor thresholds indicated that they probably contribute to tomato flavor.
A study of the inheritance of the 3 compounds indicated that 2-isobutylthiazole concentration was determined by a single gene with additive effects. Concentrations of methyl salicylate and eugenol were determined by single genes closely linked in the coupling phase, with dominance for low concentration.
Traditionally, plant breeders have been concerned with haracteristics such as yield and disease resistance which can be evaluated subjectively or by straightforward objective methods where differences are qualitative and maximum manifestation of the trait is desired. Quality has not been a principal objective in most plant breeding programs; frequently, it has been an afterthought. Once the other desired characteristics have been achieved, there is an attempt to select for adequate quality. The attention that quality characters have received varies greatly among the numerous quality traits. Color, for example, has received considerable attention because of its importance to appearance and consequently salability. In contrast, nutritional value and flavor have been mainly neglected by plant breeders.
There is continual change in the acidity of tomato fruits during development and maturation. The concn increases during development and reaches a maximum near incipient color, then decreases until well beyond maturity. The inheritance of citrate and malate concn in ‘Tondo Liscio’ (TL) and PI 263713 (PI) is controlled by a single gene for each compound. The dominant alleles, which were linked in the coupling phase in PI, condition a high concn of citrate and a low concn of malate. Recombination was about 18%. Study of divergent tomato accessions indicated that there are factors which condition higher and lower concn of citrate than the range delimited by TL and PI. There appears to be more than one malate factor controlling higher concn, but many tomato accessions are similar to the dominant low parent (PI). Current evidence indicates that there is no practical reason to breed for a specific citrate/malate ratio.
Flavor evaluation studies with ‘Campbell 146’ and ‘Campbell 1327’ indicated that 2-isobutylthiazole concn and solids/acid ratio are important to the flavor difference between these 2 cultivars.
Fifty-five tomato lines from divergent sources were analyzed to study relationships among components contributing to quality variation. There were highly significant correlations among major components contributing to solids. As total solids increased, the proportion of reducing sugars, the predominant organic compounds, increased. Despite a highly significant negative correlation between pH and titratable acidity, there was a wide range in the [H+]/titratable acidity ratio. The data indicate that differential buffering is primarily responsible for this variation and that phosphate content of the fruit is a prime factor in differential buffering. Citrate, but not malate, concn was correlated with titratable acidity.
Inheritance of the viscosity potential of tomato (Lycopersicon esculentum Mill.) fruits was studied using parents and F1, BCP1, BCP2, and F2 progeny from the crosses VF145-7879 × VF109 and VF145-7879 × 9039-M. Differences in viscosity from both crosses were controlled by 3 or fewer genes. In the VF145-7879 × VF109 cross there was bimodal distribution of BCP1 and F2 progeny which supports relatively simple inheritance. Partition analysis of the VF145-7879 × 9039-M cross supported the view that 3 or fewer genes were involved in viscosity difference. The heritability estimates were high for both crosses (0.75 and 0.68) and analysis of genetic variance showed that additive genic effects were most important. In the VF145-7879 × VF109 cross there was small, but significant, additive × dominance interaction whereas in the VF145-7879 × 9039-M there was no interaction but there were dominance effects. Study of the heritabilities of 5 components of the alcohol-insoluble solids did not clarify the genetic analysis of viscosity potential. Possible effects of competition for photosynthate on the genetic analysis are discussed.
This very complex topic must be dealt with only superficially here. Volumes have been written about sensory evaluation, and the best that this short paper can do is to generalize about sensory evaluation of horticultural commodities from our experience with tomatoes.