Acta Hortic. 1283. ISHS 2020. DOI 10.17660/ActaHortic.2020.1283.18 Proc. XXVI Int. Eucarpia Symposium Section Ornamentals: Editing Novelty Eds.: P. Franken, C. Tränkner and U. Drüge 137 Improving selection efficiency in ornamental breeding H. Molenaar 1,a , R. Boehm 2 and H.-P. Piepho 1 1 University of Hohenheim, Institute of Crop Science, Biostatistics Unit, Stuttgart, Germany; 2 Klemm + Sohn GmbH & Ko KG, 70378, Stuttgart, Germany. Abstract The ornamental breeding scheme is often clone breeding. The greatest genetic variability exists in the seedling generation and each genotype is represented by a single plant. During the primary selection of seedlings, the population size is drastically reduced from thousands to hundreds of genotypes. But, as the population size is reduced, the genetic variability is reduced, too, from the seedling to the first clonal generation. Therefore, differences between individuals become more difficult to detect. A way to increase the precision of estimating treatment effects from augmented designs or to detect genotypic differences in spite of reduced genetic variation is to exploit the information of relatives by a family-index (Lush, 1947), which is estimated based on the intra-class correlation coefficient. An intra-class correlation coefficient of one as the one limiting case implies that the individual performance can be fully explained based on the family background, whereas an intra-class correlation coefficient of zero, the other limiting case, indicates that the performance of individuals is independent of the family background. When the intra-class correlation coefficient varies between zero and one, family-index selection is considered, otherwise individual selection is considered. Therefore, the main difference between individual and family-index selection lies in the adjustment of an individual’s effect by the intra-class correlation coefficient. Both selection methods were illustrated in Dianthus caryophyllus L. pot for flower longevity and in Dianthus caryophyllus L. cut for root quality. In almost all cases, family-index selection outperformed individual selection in terms of heritability and can therefore increase the selection efficiency in ornamental breeding programs. Keywords: BLUP, BLUE, two-phase design, phenotypic selection, family-index selection, individual selection, ornamental breeding INTRODUCTION Clone breeding is often the breeding scheme of ornamentals. In the seedling, generation the greatest genetic variability exists and each genotype is represented by a single plant. In later clonal generations, replicated testing is possible. Experimental designs that are suitable for testing unreplicated genotypes are augmented designs. Thereby, block effects are estimated solely by the use of checks and hence allow to control the error of an experiment. The population size is drastically reduced from thousands to hundreds of genotypes during the primary selection of seedlings. In the clonal generations, individual genotypes may be tested in replicated designs, for example in resolvable incomplete block designs, as applied in Dianthus caryophyllus L. or Pelargonium zonale (Boxriker et al. 2017; Molenaar et al., 2018) or in randomized complete block designs in later breeding stages as the number of individuals is more and more reduced (Molenaar et al., 2017). Higher precision of estimated treatment effects can be expected, because the block effects are estimated from the individuals included in all replicates. But, as the population size is reduced, the genetic variability is reduced, too, from the seedling to the first clonal generation and therefore, differences between individuals become more difficult to detect. A way to increase the precision of genotypic effect estimates from augmented designs or to detect genotypic differences in spite of reduced genetic variation is to exploit the information of relatives done by a family-index (Lush, 1947), which is estimated based on the a E-mail: Heike.Molenaar@uni-hohenheim.de