| dc.description.abstract | Roses (genus Rosa) are among the most popular ornamental plants. Traditional rose breeding is a slow and tedious process, but breeding efficiency can be improved by marker-assisted selection. Marker-assisted selection, however, requires a thorough understanding of the genetic control of the traits of interest. To characterize the genetic control of certain traits of interest, eight segregating diploid rose families were developed. Families were phenotyped for black spot and cercospora resistance, defoliation, flower intensity, and plant architecture (number of primary shoots, height, length, width, longest dimension, volume, apical dominance, and growth habit). Families were genotyped for single nucleotide polymorphisms (SNPs) using genotyping by sequencing. Seventy-three rose cultivars were genotyped and phenotyped in the same manner.
Heritability was estimated for both datasets. Broad-sense heritability was high for black spot, defoliation, and flower intensity, but low for cercospora. All four traits also had low narrow-sense heritability, indicating a high degree of non-additive effects. Architecture traits generally had low to moderate broad-sense heritability and low narrow-sense heritability, again indicating non-additive effects. Genotype by environment interactions were generally high within a year, reflecting the growth of the plants over the course of the year, but relatively low over years. Narrow-sense heritability estimates for length, width, longest dimension, and apical dominance were slightly higher in once-flowering genotypes than continuous flowering genotypes, suggesting that some germplasm has stronger additive effects for these traits.
Association mapping was performed for both datasets. Three clusters of associations were identified for black spot and cercospora on chromosomes 2, 3, and 6. When flowering type was controlled for, five clusters associated with flower intensity were identified on chromosomes 2, 4, and 5. Ten clusters associated with plant vigor (height, length, width, longest dimension, and volume) were identified. Vigor clusters on chromosomes 1, 2, 4, and 7 may coincide with previously identified quantitative trait loci (QTLs), but the six other clusters appear to be novel.
In conclusion, disease resistance, defoliation, flower intensity, and architecture traits had a range of heritability estimates with mostly non-additive heritability. Potential genomic regions controlling these traits were identified but require validation. To facilitate this, a high-density integrated consensus linkage map was developed from the three largest families in preparation for a QTL analysis. Future work will use this map in QTL analyses, enabling marker-assisted selection for these traits of interest. | en |
