Quantitative genetics study and development of a molecular marker for use in red raspberry (Rubus idaeus L.) breeding

Abstract

Introduction Red raspberry (Rubus idaeus L.) is an outcrossing (but self fertile) diploid fruit crop that is clonally propagated (homogeneous within cultivars) and highly heterozygous. Raspberries grown for freezing and process markets are generally harvested by machine which requires a specific combination of plant and fruit traits. Successful red raspberry cultivars need to have high machine-harvested yield as well as a number of fruit quality attributes such as firm fruit, high sugar and acid content, dark colour, good flavour and enhanced human health properties. Quantitative genetic studies and, more recently, molecular plant breeding methods using DNA markers and quantitative trait loci (QTL) analyses are tools for increasing the breeding efficiency through better understanding of genetic parameters and genes controlling complex traits. Aims The objectives of this study were (i) to investigate the inheritance and genetic parameters of yield and yield components of red raspberries, (ii) investigate the possibility of using key yield components to predict total machine-harvested yield of individual seedlings in a breeding programme, (iii) investigate the inheritance and genetic parameters of fruit quality traits and optimise tools for multi trait selection in a breeding programme, (iv) identify DNA markers associated with QTL for fruit quality traits. Materials and methods Two distinct populations were used. The first, used to study genetic parameters associated with machine-harvested raspberry yield and fruit quality was located in Washington State, United States and consisted of 45 parents giving rise to 85 families each with 12 seedlings from a double pairwise crossing design. Total yield (TYLD) and yield components; cane length (CLEN), cane number (NCAN), cane diameter (CDIA), number of buds/nodes (NBUD), percentage budbreak (PCBB), lateral length (LLEN), number of fruit per lateral (NFRT), berry weight (BWT) and fruit quality traits, fruit firmness (FIRM), soluble solid content (SS), titratable acid content (ACID), total anthocyanins (TACY) and total ellagitannins (TELG) were measured over three seasons and data analysed using mixed models and best linear unbiased predictors (BLUP). The second population, designed for genetic mapping from an interspecific F1 cross between black raspberry (R. occidentalis) and red raspberry (R. idaeus) was used to create a genetic framework map using candidate gene markers from a number of species. QTL analysis was conducted for fruit quality traits SS, ACID and TELG. Results Variance component and BLUP analysis of data revealed TYLD had low narrow sense heritability (h²=0.25) but some yield components had higher heritability (e.g. LLEN h²=0.44, BWT h²=0.69) and high genetic correlation with TYLD (e.g. BWT r=0.80, CLEN r=0.54). The yield components BWT, LLEN and NFRT had low genotype by year (G×Y) interaction and were more stable over all seasons than CLEN, CDIA, PCBB, NBUD and NCAN. BWT and LLEN measured in the first two seasons were the two yield components that provided the best estimate of TYLD on mature plants in the second and third seasons. These results were from breeding values derived from hand harvested plants, a harvest method that is time-consuming and expensive to measure in breeding plots. Using a bulk family-plot machine-harvest yield we developed a strategy for estimating machine-harvest yield breeding values for our individual seedlings and found higher genetic gain per generation using estimated individual machine-harvest breeding values (7.6%) than using hand-harvested breeding values (6.5%). Fruit quality parameters, FIRM (h²=0.54), SS (h²=0.73), ACID (h²=0.45), TACY (h²=0.67) and TELG (h²=0.46) content had moderate to high narrow sense heritability and all but ACID (r=-0.35) and TACY (r=-0.28) were positively correlated with TYLD. G×Y interaction was low for all fruit quality traits measured. We applied a selection index to enable multiple trait selection for high yield as well as fruit quality traits, FIRM, SS, TACY and TELG. For fruit quality traits, SS, glucose, fructose, ACID, citric acid and TELG we found QTL that explained significant amounts of variation for these traits on our parental maps. DNA markers that were derived from candidate genes were closely linked to QTL for these traits. Discussion and conclusions TYLD has low heritability and is difficult to breed for; however, this study has shown that key components of yield can be used to select high yielding types in raspberry breeding populations. Further, we have shown that it is an efficient breeding strategy to use a machine to bulk harvest full-sib family plots and use key yield component data to relate these data to individuals within families. This new and innovative technique has potential to revolutionise breeding for improved yield in machine harvested berry crops. We have shown that fruit quality traits have good heritability and have low G×Y interaction and that it is possible to select for high yield as well as multiple fruit quality traits (including those negatively correlated to total yield) early in the life of plants using a selection index. Genomic regions and DNA-based markers associated with fruit quality traits measured in this study offer potential for further development and future marker-assisted selection.