A backeross between the inbred parents, Lycopersicon esculentum (LA 490, recurrent parent) and Solanum pennellii (LA 716) constitutes ideal material for an investigation of linkage between enzyme marker genes and loci determining quantitative traits (QTL): the chromosomes of the parents are the same in number and are homosequential; the parents differ for well-defined alleles at 12 enzymatic loci, which represent at least 8 of the 12 chromosomes, and large differences exist between the parents for measurable metric traits.The four metric traits investigated in this study were inherited in typical quantitative fashion; the marker segregations fit monogenic ratios for all except four loci, for which there were deviations of three favoured alleles of the recurrent parent, in agreement with frequent observations in interspecific hybrids. Association between QTL and mapped enzyme markers were detected statistically from calculations based on the normal distribution. In this manner a minimum of 21 0Th were mapped.Two pairs of linked enzyme markers permitted application of a three-point mapping procedure by comparing the data with models expected for various positions of QTL. Data for three Oil fitted predictions specified for a locus closer to one marker than the other, and data for one OTL met all criteria save one for a locus between the two markers.Certain QTL coding for stigma exsertion and leaf ratio have effects opposite to those expected from parental values; i.e., for these traits both parents contribute alleles coding for positive and negative effects, resulting in the observed transgressive variation. When values observed for Oil linked with pairs of independent markers are compared with sums of separate values for each, significant epistatic interactions are detected. Although the extent of epistasis could not be determined precisely, it is probably not large. The paired interactions identified additional OTL whose individual effects were not otherwise detectable. Enzymatic markers offer a number of advantages for the analysis of quantitative genetic variation which include: (a) lack of detectable effects on the morphology and physiology of the soma, (b) codominant expression, permitting exact identification of genotypes, and (c) lack of epistasis, allowing classification of any number of such markers segregating simultaneously. A tester stock with linked enzyme markers on each chromosome therefore could conceivably be utilized to analyze the total quantitative differences between two parents in a single segregating progeny. Such a tester stock might be especially useful for probing exotic germplastn for genes controlling quantitative characters of economic importance. The methods of analysis outlined here could easily be adapted for analysis of F2 populations or generations from other mating designs where a fuller array of genetic variation would be revealed.

For several years it has been recognized that introduction of plant cells into culture results in genetic changes. These genetic alterations have been recovered in the plants regenerated from cell cultures. More recently it has been recognized that this method of introducing genetic changes into crop plants could be used to develop new breeding lines. The technology of introducing genetic variation by using cell culture has been termed somaclonal and gametoclonal variation. This paper reviews the history of this technology and offers genetic documentation of somaclonal variation in tomato. As this variation represents a new tool for the plant breeder, breeding strategies for the use of this variation are presented and discussed. Somaclonal and gametoclonal variation are new tools for the geneticist and plant breeder that permit reduction in the time period for new variety development and that permit access to new classes of genetic variation.

No abstract

The extent of correlation was estimated between isozyme genotypes and the four widely segregating characters - leaf segment W/L ratio, stigma exsertion, fruit weight, and seed weight - in the first backcross of F1 Lycopersicon esculentum x Solanum pennellii to the former parent. The inbred parents differ in their alleles at the 12 tested isozymic loci, which are known to mark a minimum of eight of the twelve tomato chromosomes. Based on the isozyme data, a mean heterozygosity value, ¯H, was calculated which estimates the proportion of pennillii alleles in each individual. Correlations between mean heterozygosity and observed levels of each quantitative trait were highly significant and positive or negative as expected from the relative parental values. Plants with the lowest mean heterozygosity - i.e., closest to the esculentum zymotype also had mean values closest to those of this parent amongst the whole backcross population for each of the quantitative traits.Bivariate and multiple regression analysis was used to evaluate the ability of isozymes vs diagnostic morphological characters to estimate the portion of recurrent parent genes carried in each backcross individual. The results suggest that isozyme data gives better estimates than single diagnostic morphological characters and approach the level obtained by combinations of three morphological traits. Since electrophoretic determinations are made on small seedlings, selection at that stage can effect great savings of space and effort by greatly deminishing the size of the population needed at maturity. As such, isozyme selection would precede morphological selection but not replace it, thus the predictive value of these biochemical markers as well as diagnostic morphological characters could be obtained.