Progress, prospects and challenges in a recurrent selection programme in apple

This recurrent selection programme in apple was set up to 1) broaden genetic base, 2) maintain genetic variability in the course of population improvement, 3) study the genetics of important horticultural traits, 4) select parents based on breeding value, and 5) maintain long term selection response. The base populations consisted of 520 (mostly) OP families from North America, Europe, UK, Australasia and Kazakhstan. Populations were divided into 4 sublines and planted in randomised incomplete blocks at three HortResearch sites in Havelock North (39 degrees 40' S 176 degrees 53' E), Nelson (41 degrees 04' S 173 degrees 00' E) and Clyde (45 degrees 12' S 169 degrees 18' E), from 1993 to 1996. Individual trees were evaluated for six tree, four pest and disease resistance and twenty eight fruit characters, from 1995 to 2001. Parents for the next cycle were chosen using a combination of backward (family) selection and forward (individual phenotype) selection. Most traits were considered neutral in the selection process; that is, both high and low values of a trait were of potential interest. For pest and disease resistance and fruit disorder incidence, non-susceptible individuals only were chosen. Individual trees were selected from 80% of the families, and frequency distributions of representative traits suggested that diversity was maintained and the populations slightly improved. The genetic parameter estimates for the populations appeared promising but they may need to be re-assessed in the next generation using controlled crosses (to increase their accuracy and also test for reproducibility), before they can be utilised in the programme. The future direction and challenges of the programme are discussed in the light of the above results.