Virginia is the largest apple producing state in the Southeast region of the United States and ranks 10th in most cideries in the US. Natural, or un-inoculated, fermentation methods are of interest to cider producers due to the potential for generating unique and complex aromas and flavors via fermentation with naturally present microbiota. The objective of this study was to determine the effect of common pre-fermentation sulfite addition and pH adjustment on cider chemistry and sensory outcomes for naturally fermented high- and low-tannin apple cultivars. Four treatment conditions were applied to both the high- and the low- tannin cultivars: acidification only, sulfites only, acidification and sulfites, and a control with no pre-fermentation juice chemistry adjustment. The eight experimental ciders were fermented using the Pied de Cuve (PDC) method for natural fermentation. Cider chemistry and sensory parameters were determined, and the treatments imparted key differences in both. Key findings were analyzed for pH, titratable acidity, volatile acidity, malic acid, free/total SO2, yeast assimilable nitrogen (YAN), total polyphenols, residual sugars, and ethanol. For the acidified condition, the pH was lowered to 3.2 using malic acid. Cider pH ranged from 3.36 ± 0.04 to 3.72 ± 0.07, reflecting a general trend toward rising pH over the course of fermentation. Juice tannins were 0.244 ± 0.003 g/L for Harrison and 0.12 ± 0.01 g/L for GoldRush. Tannins decreased during fermentation; however, Harrison ciders maintained a higher range compared to GoldRush. Sensory characteristics were determined using a Descriptive Analysis (DA) with a trained panel which produced 28 descriptors. Results were examined via analysis of variance (ANOVA) and significant differences for apple cultivar, acid adjustment, and sulfite use were found for both chemistry and sensory parameters. The interaction between high- and low- tannin content and sulfite use had the most impact on the cider chemistry and sensory attributes. This study helps to shed light on the extent to which pre-fermentation pH adjustment and/or sulfite additions can influence the outcomes of natural cider fermentation in both high- and low-tannin cultivars. / Master of Science in Life Sciences / Cider, also known as "hard cider," is an alcoholic beverage fermented from apples. Virginia is the 6th largest apple producing state in United States and ranks 10th in number of cideries. Natural fermentation uses microorganisms that are present in the environment to ferment cider. This type of fermentation is of interest to the cider industry due to the unique aromas and flavors produced by this method. The objective of this study was to determine the effect of common fermentation management strategies: pre-fermentation sulfite addition and acid adjustment, on cider chemistry and sensory outcomes for naturally fermented high- and low-tannin apple cultivars. Eight experimental ciders were fermented using the Pied de Cuve (PDC) method, which is a type of natural fermentation. Cider chemistry and sensory outcomes were evaluated. The experimental treatments and their interactions imparted key differences in both chemical and sensory outcomes. Cider pH ranged from 3.36 ± 0.04 to 3.72 ± 0.07 reflecting a general trend toward rising pH over the course of fermentation. Juice tannins were 0.244 ± 0.003 g/L for Harrison and 0.12 ± 0.01 g/L for GoldRush. Tannin concentration generally decreased during fermentation; however, Harrison ciders maintained a higher range compared to GoldRush. Sensory characteristics of each cider were determined using a Descriptive Analysis (DA) study, a with a trained panel which produced 28 descriptors, 19 of which were significant. The interaction between high- and low- tannin content and sulfite use had the most impact on the cider chemistry and sensory attributes. This study helps to inform cider producers regarding the impacts of pre-fermentation acid or sulfite additions on natural fermentation, and how those impacts may vary among high- and low-tannin apple cultivars.

Philosophiae Doctor - PhD / Apple scab, powdery mildew and woolly apple aphid are a major concern for apple breeders and producers. Control of these diseases is a significant economic and marketing priority for the South African apple industry. Application of chemicals and orchard management practices are the main methods for controlling these diseases. These diseases require an average of 15 chemical sprays per season, which leads to increased production costs for the farmer. The increased cost of chemical based control programs and demand from consumers for ‘organic apples’ grown with very little to no chemical sprays makes it important to breed for commercial apple cultivars with endogenous disease resistance genes (R-genes). The use of genetic tools (apple genetic linkage maps and the apple genome sequence) to track and introgress endogenous R-genes in breeding and to confer durable disease resistance in commercial apple cultivars will lead to a more cost effective means of disease control for apple producers. Historically, most breeding programmes rely on recurrent conventional breeding systems. This involves the crossing of apple selections showing resistance to a given disease with a susceptible elite variety. This is followed by phenotyping the progeny to identify trees exhibiting segregating field resistance. Several crosses and backcrossing are required to produce resistant varieties and to fix the resistance trait using this breeding strategy. This breeding technique is time consuming, especially in perennial tree species such as apples, which have a long juvenile period. Molecular markers have enabled the building of genetic maps, which has allowed for tracking of the inheritance of genes contributing towards the observed resistances. This has given breeders the opportunity to start the implementation of marker-assisted-breeding (MAB) and marker-assistedselection (MAS). MAB and MAS greatly reduce the time required to select for favourable genotypes, given that MAB facilitates efficient selection for inherited traits at the seedling stage. With the publication of the apple genome sequence, the identification of the genes involved in disease resistances has been made possible and this will allow researchers to venture into cisgenics for apples, which will further reduce the time required for the introgression of desirable genes into commercial cultivars. The main thrust of this research was to generate dense genetic linkage maps for three mapping populations segregating for apple scab, woolly apple aphid and powdery mildew resistance. The three mapping populations are ‘Mildew Resistant’ x ‘Golden Delicious’, ‘Russian Seedling’ x ‘Golden Delicious’ and Malus platycarpa x ‘Mildew Resistant’ and are Malus full-sib outbreed mapping populations. The generation of the genetic maps was for use in the subsequent identification candidate disease resistance QTLs/genes that can be implemented in apple cisgenics. Integrated genetic maps using SSRs, DArTs and SNP marker data were generated for all the three crosses. The integrated map of ‘Mildew Resistance’ x ‘Golden Delicious’ consists of 1, 563 markers with a total map length of 1, 298.8 cM. The ‘Russian Seedling’ x ‘Golden Delicious’ genetic map is composed of 979 markers with a total map length of 1, 729.9 cM. The Malus platycarpa x ‘Mildew Resistant’ integrated map has 616 markers and a total map length of 1,324.3 cM. Due to the fragmentation of some of the linkage groups in the ‘Russian Seedling’ x ‘Golden Delicious’ and in the Malus platycarpa x ‘Mildew Resistant’ genetic maps, a phylogenetic analysis was performed to evaluate the genetic distances between the parents of the crosses in order to understand the cause of the fragmentation of these two integrated genetic maps. QTLs were detected through the statistical correlation of the phenotypic and map data using restricted Multiple QTL Mapping (rMQM) from MapQTL® 6.0. The genome-wide LOD score minimum QTL detection thresholds were determined using 10 000 permutations for each population. The minimum QTL detection threshold for accepting a putative QTL was then determined to be 4.5 for ‘Mildew Resistant’ x ‘Golden Delicious’ and 4.6 for both the ‘Malus platycarpa’ x ‘Mildew Resistant’ and ‘Russian Seedling’ x ‘Golden Delicious’ mapping populations. A total of 17 putative QTLs were detected for the ‘Mildew Resistant’ x ‘Golden Delicious’ population, 10 putative QTLs for the Malus platycarpa x ‘Mildew Resistant’ population and nine putative QTLs for the ‘Russian Seedling’ x ‘Golden Delicious’ population were detected for the three diseases under study. The two putative QTLs for apple scab resistance detected on LG 02 of the ‘Russian Seedling’ x ‘Golden Delicious’ map coincided with the loci previously identified as encoding two apple scab resistance genes Vh2 and Vh4 on ‘Russian apple’. The QTL for apple scab resistance identified on the proximal QTL of LG 02 co-localized with SNP marker R_8936738_Lg2 on the loci where Vh4 was previously identified. The distal QTL on LG 02 shown to encode the Vh2 R-gene was linked with the SNP marker R_32981524_Lg2. With ‘Russian apple’ being known to carry a natural pyramid of R-genes for apple scab on LG 02, therefore, the ‘Russian Seedling’ used in this study was screened by a set of 14 SSR markers to determine if it was related to ‘Russian apple. The 14 SSRs produced identical alleles to those amplified by ‘Russian apple’, which means “Russian Seedling’ and ‘Russian apple’ are closely related or identical. The LG 02 pseudo-chromosome sequence was extracted from the NCBI database housing the apple genome sequence and was then used to mine for the putative R-genes within the two QTL regions. The region corresponding to the Vh2 loci, which was roughly a 600 kb region, had two clusters of ABC (PDR) disease resistance related genes. These were predicted using a full Pfam domain search and were only detected on the negative strand. The 60 kb region corresponding to the Vh4 loci comprised a cluster of LRR domains that were also detected on the negative strand using a full Pfam domain search. This 60 kb region was further analysed using Phytozome and Genome Database for Rosaceae (GDR) leading to two candidate disease resistance genes being identified. Ten consensus gene sequences were present within the 60 kb region, with only two transcripts MDP0000657246 and MDP0000128458 identified as being disease resistance related genes. The MDP0000657246 was identified on the contig MDC000294 of the Malus x domestica reference genome as being a Leucine Rich Repeat protein kinase family, which is one of the most abundant disease resistance family mainly involved in the gene-for-gene resistance mechanism. The MDP0000128458 locus was identified on contig MDC015161 as being a Ser/Thr phosphatase 7. The Ser/Thr phosphatase genes have been associated with the regulation of MAP kinase cascades that have been shown to induce the hypersensitive response (HR) in tobacco. Therefore these two genes are likely to be the loci associated with the hypersensitive response associated with the infection of apples with race 4 of apple scab, carrying the Vh4 apple scab resistance gene. Recurrent putative QTLs were detected that still need to be validated in order to be used for MAB. The ‘Russian Seedling’ x ‘Golden Delicious’ cross produced a single powdery mildew resistance QTL located on LG08 and conferring a 1:1 resistance to susceptible phenotypic segregation ratio. These results indicate that the source of the resistance thus was a single dominant resistance gene. The ‘Mildew Resistant’ x ‘Golden Delicious’ mapping population also showed two stable QTLs one for powdery mildew on LG 03, which co-segregated with SNP GD_LG03snp00866 and in addition SNP R_13071892_Lg10 was also identified to be co-segregating with the QTL for apple scab resistance on LG10. However, none of these recurrent QTLs co-localized with known genes or QTLs. For the phylogenetic analysis, re-sequenced data using the Illumina® sequencing technologies and the apple SNP chip data for ‘Russian Seedling’, ‘Mildew Resistant’, Malus platycarpa, a Chinese accession of Malus sieversii and ‘Anna’ where used to infer relatedness of the five genotypes. The Chinese accession of Malus sieversii was included in the analysis since ‘Russian Seedling’ was thought to be relatively close genetically. Whilst ‘Anna’ is known to be a low chilling cultivar of Malus x domestica (Borkh) and therefore would add in the phylogenetic placement of ‘Mildew Resistant’ and Malus platycarpa. These were sequenced to coverage of approximately 60X for ‘Russian Seedling’ and 6X for the other four genotypes. The sequence data was aligned to the reference Malus x domestica cv Golden Delicious mitochondrial genome sequence. Phylogenetic analysis was then performed using both the data from the apple SNP-chip and the aligned mitochondrial genomes. The results from both sets of data supported the putative evolutionary distances between the five genotypes. ‘Russian Seedling’ and M. sieversii were closely related, while both were genetically divergent from the closely related ‘Anna’ and ‘Golden Delicious’ commercial cultivars. This analysis however indicated that ‘Mildew Resistant’ was relatively closely related to ‘Golden Delicious’ and hence the low number of markers showing segregation distortions for the ‘Mildew Resistant’ x ‘Golden Delicious’ population in the 17 LGs of the integrated map. However, the other two mapping population exhibited a high number of markers with segregation distortions. Markers which are closely associated with disease resistance to apple scab powdery mildew and woolly apple aphid resistance will play a major role in the identification of the genes responsible for the resistances being observed. The identification of the two candidate genes for the Vh4 gene associated with apple scab resistance will be the platform from which a cisgenic programme can be implemented in the South African apple breeding program.