Date of Graduation

5-2023

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Plant Science (PhD)

Degree Level

Graduate

Department

Agricultural, Food and Life Sciences

Advisor/Mentor

Margaret Worthington

Committee Member

Renee Threlfall

Second Committee Member

Ya-Jane Wang

Third Committee Member

Jacquelyn Lee

Fourth Committee Member

Larry Purcell

Keywords

descriptive sensory analysis;gene expression;GWAS;high throughput phenotyping;polygalacturonase;red drupelet reversion

Abstract

For most fresh-market fruit crops, texture is an important trait that strongly affects both shipping potential and consumer opinion. Efficient, scalable phenotyping methods are required by breeding programs to effectively select for improvements to fruit texture quality. In muscadine, we have developed a recommendation for characterizing complex muscadine grape texture profiles by comparing the results of breeders’ ratings, descriptive sensory panel results, and an array of instrumental protocols. Regression models were constructed to predict awareness of skins, crispness, hardness, and visual separation explaining 85%, 91%, 82%, and 83% of variance respectively. Genotypes that scored most highly in breeders’ ratings of overall texture had soft skins and firm flesh, suggesting that both qualities are important targets for texture improvement in muscadine. We have also developed an R shiny based web-application, called ShinyFruit, for image-based analysis of fruit morphology and color quality. ShinyFruit was tested against manual methods of size and red drupelet reversion (RDR) estimation in a diverse population of blackberry cultivars and breeding selections. ShinyFruit results shared a strong positive correlation with manual measurements for blackberry length (r = 0.96) and significant, albeit weaker, correlations with manual RDR estimation methods (r = 0.62 - 0.70). Further validation of ShinyFruit’s potential was provided when it was used to generate phenotypic datasets across a genome wide association study (GWAS) panel of 300 diverse blackberry genotypes. This GWAS panel is the first reported in autotetraploid blackberry, and numerous quantitative trait loci (QTL) for blackberry texture and RDR were identified, spanning chromosomes Ra01, Ra02, Ra03, and Ra06. All QTL associated with RDR were located on Ra02 and most of these 212 single nucleotide polymorphisms (SNPs) were in high linkage disequilibrium. Three variants on homologs of polygalcturonase (PG), pectin methylesterase (PME), and β-glucosidase explained 27% of variance in fruit firmness and were located on chromosomes Ra06, Ra01, and Ra02 respectively. Both fruit firmness and RDR appear to be complex, moderately heritable traits, which may be most effectively incorporated into a genomic selection model. Expression-level evidence suggests that an inhibitor of PME may be associated with the fruit firmness QTL identified on Ra01. The expression of this PME inhibitor was negatively correlated with PME activity through fruit development in the ‘crispy’ fruited A-2453T. Expansin-like proteins were also expressed more highly in A-2453T compared to the soft-fruited ‘Black GemTM’, suggesting that this protein family could play a unique role in the ‘crispy’ texture phenotype. By combining newly developed phenotyping methods with informative genomic and transcriptomic datasets, we provide a strong foundation for continued research. Future improvement of texture in blackberry should prioritize training genomic selection models which could be trained on our accumulated datasets and supported by ShinyFruit phenotyping.

Download

Included in

Horticulture Commons

Share

COinS