Methods for selecting flower quality based on consumer evaluation

Summary The quality ranking of a sample of Gerbera cut-flowers varies greatly from one person to another. Further, there are significant differences between the rankings assigned by a panel of trained horticulturalists and a panel sampled from the general public. Therefore, the selection of a panel to evaluate the flower quality component of a breeding program is very important to the outcome. If the purpose of a program is to breed (those) flowers preferred by the general public then the selection of flowers should be done by a panel sampled from the general public. The prediction of consumer preferences by multiple linear regression techniques using eleven flower characters was not successful. Much greater predictive value was obtained by measuring the perceived information content of flowers. This indicates that flower characteristics which increase flower complexity may be preferred. Implications to the practical breeding program are discussed.