2011 Annual Report
1a.Objectives (from AD-416)
Refine the classification of cultivated tomato based on fruit morphology and genotype. Tomato fruit morphology categories will be studied using the program Tomato Analyzer (TA) and the effect of each fruit morphology gene will be estimated. First the tomato fruit classification scheme will be refined using 51 accessions that vary in longitudinal fruit shape. Data on transverse shape attributes such as pericarp and placenta thickness, and locule number which are features implemented in the latest version of TA will be added.
The Plant Genetic Resources Unit will collect data on nutritional traits for all three locations of the trials for the two years. Nutritional data will include lycopene, vitamin C, brix, and titratable acids. Data on nutritional traits and morphological traits will be analyzed for diversity and genetic relationships.
1b.Approach (from AD-416)
This collection will be grown at three different locations over two growing seasons: Wooster OH, Mills River NC, and Wellington Farm in Geneva, NY (Plant Genetic Resources Unit). This agreement covers the field grow out for these two years at Geneva, NY. From each accession, we will analyze 48 fruit (N=4992, i.e. 52 lines x 4 fruit/plant x 4 plants x 3 locations x 2 years) by scanning using the TA program.
Nutritional data will be collected from frozen fruit homogenates from the three locations. An indirect measure of citric acid, titratable acidity (TA), will be determined by titrating samples with NaOH. Ascorbic acid (vitamin C) will be estimated using a commercial kit (COSMO BIO CO, Japan) that measures both oxidized and reduced forms. Lycopene will be estimated using a Minolta Chroma Meter CR-300 that records L*a*b* color space. Each L*a*b* value represents the average of three measurements. Lycopene will be estimated using a regression model based on the transformed a*. Degrees Brix data will be collected using a Model DR103L digital refractometer (QA Supplies, Norfolk, VA). Juice from thawed homogenates will be place upon the refractometer. For each sample degrees Brix will be calculated as a mean of three readings for each replication for each accession.
Hierarchical cluster analyses will be performed interactively using different combinations of similarity/dissimilarity indices and linkage algorithms to reveal the most appropriate number of shape classes. This approach will also reveal the subset of TA-defined attributes that contribute most to the classification, and establish standardized tolerances for key attributes such that each shape class can be objectively defined regardless of source. The net result will be a uniquely semi-automatic and objective classification system useful for categorization based on fruit morphology. By growing the collection in three different environments over two years we will be able to establish the degree to which genotype x environment interaction affect fruit shape based on multi-way analyses of variance of each attribute and shape class. Importantly, these analyses will determine the variability in fruit morphology for certain genotypes and categories.
In this study, we hypothesized that environmental factors could impact both fruit shape and quality, and that such effects would vary by host genotype. To study the impact of the genotype and the environment on tomato fruit morphology, shape characteristics of 48 cultivars grown at 3 different locations (OH, NC, and NY) were obtained. These cultivars represent a mix of cultivars belonging to the eight shape categories previously identified by our laboratories. Individual samples consisted of approximately 20 fruits from each plant. Longitudinal and transverse fruit cutting images were analyzed using Tomato Analyzer software. Additionally, genotype information based on 20 different SSR markers as well as known shape genes FAS, OVATE, SUN, and LC were determined for each cultivar. In total, a matrix of over 58,305 data points was assembled and analyzed for variation by genotype, cultivar, environment, and interactions thereof. Additionally, data on fruit quality parameters (lycopene, total soluble solids, vitamin C, and total acidity) were collected from fruit of equivalent maturity harvested at all 3 locations. Analysis of State, Cultivar, Shape category, and Rep were made using the ANOVA General Linear Model.
Conclusions generated this year: Fruit morphology varied substantially, providing a unique opportunity to study important plant developmental pathways. Significant variation in several fruit shape attributes, fruit mass and locule number were observed for all of the predictor variables. Fruit shape index external II varied significantly by state indicating (P value from 0.000-0.061) for 9 of the 48 cultivars tested. Patterns for state effects were not the same, most followed the pattern of OH>NC>NY, but other patterns were observed as well. Among nine fruit shape categories, 3 (Ellipsoid, Obovoid and Heart) were affected significantly by state indicating and 5 of them showed significant (P<0.1) environment (state)* cultivar interaction. The environment-cultivar interaction indicated more complex patterns in different fruit shape classes response to state, probably due to the different phenotypic shape classes were not homogenous for genotype based on selected genes (i.e. FAS, OVATE, SUN, and LC) or neutral markers. Distribution of fruit shape morphology genes SUN, OVATE, LOCULE NUMBER (LC) and FASCIATED (FAS) was evaluated in previous work. Different alleles were considered as different genotypes and their sensitivity to the environment were analyzed. Our results showed significant state and cultivar*state interactions (P value from 0.000 to 0.035) within genotypes in terms of fruit shape index external II. Cultivar-specific variation was noted in all genotypes regardless of the alleles. Genotypes associated with 23 different neutral genetic markers were analyzed to test the sensitivity to environment as well as known fruit shape morphology genes. Our results showed that usually one, sometimes two groups within 20 genetic markers showed a significant state impact and state*cultivar interaction (P value from 0.000 to 0.076).
Monitoring activities for this project included site visits and regular email and phone contacts.