To increase the speed and reduce the cost of constructing a genetic map of Actinidia species (kiwifruit), for use in both breeding and functional genomics programmes, we sampled microsatellites from expressed sequence tags (ESTs) to evaluate their frequency of occurrence and level of polymorphism. Perfect dinucleotide repeats were the microsatellites selected, and these were found to be numerous in both the 5' and 3' ends of the genes represented. The microsatellites were of various lengths, the majority being repeats with the pattern (CT)(n)/(GA)(n). One hundred and fifty microsatellites, each with more than 10 dinucleotide repeat units, were chosen as possible markers, and when these were amplified, 93.5% were found to be polymorphic and segregating in a mapping population, with 22.6% amplifying more than one locus. Four marker categories were identified. Fully informative markers made up 27% of the total, 36.2% were female informative, 25.8% were male informative and 10% partly informative. The mapping population was an intraspecific cross in the diploid species Actinidia chinensis, with parents chosen for their diversity in fruit and plant characteristics, and for their geographical separation. Linkage was tested using the software 'Joinmap' and a LOD value of 3. The distribution of the EST-based markers over the linkage groups obtained appeared to be random, taking into consideration the small sample size, that the number of linkage groups (31) exceeded the chromosome number of n=29, and that a number of markers were not assigned to any group. Some microsatellite markers which amplified more than one locus mapped to separate linkage groups. According to our study in A. chinensis, EST-derived microsatellites give large numbers of possible markers very quickly and at reasonable cost. The markers are highly polymorphic, segregate in the mapping population, and increase the value of the genomic map by providing some functional information.

No abstract

Microsatellite marker transfer across species in the dioecious genus Actinidia (kiwifruit) could offer an efficient and time-effective technique for use during trait transfer for vine and fruit improvement in breeding programmes. We evaluated the cross-species amplification of 20 EST-derived microsatellite markers that were fully informative in an Actinidia chinensis mapping family. We tested all 20 markers on 120 genotypes belonging to 21 species, 5 with varieties and/or chromosome races. These 26 taxa included 16 diploids, 7 tetraploids, 2 hexaploids and 1 octaploid, and represented all four taxonomic sections in the genus. All 20 markers showed some level of cross-species amplification. The most successful marker amplified in all genotypes from all species from all sections of the genus, the least successful amplified fragments only in A. chinensis and A. deliciosa. One species, A. glaucophylla, failed to amplify with all but 2 markers. PIC (Polymorphism information content) values were high, with 14 of 17 markers recording values of 0.90 and above. Sequence data demonstrated the presence of the microsatellite in all the amplified products. Sequence homology was less 5' of the microsatellite and increased toward the start codon of the translated region of the EST from which the marker was derived. The data confirm that EST-derived microsatellite markers from Actinidia species show cross-species amplification with high levels of polymorphism which could make them useful markers in breeding programmes.

Variation in the time of flowering within vines of two Actinidia chinensis (Planch.) var. chinensis cultivars, 'Hortl6A' and '37-3-18A' ('18A'), is described. The flowering capacity of the two cultivars was very different, with 106 terminal flowers and 50 lateral flowers/m 2 of canopy on '18A' vines compared to only 39 terminal flowers/m 2 on 'Hortl6A' vines. However the variation within vines was similar. The most consistent systematic trend in the time of flowering was within canes, where time of flowering varied by up to 7 days. Shoots near the apex of the cane produced more flowers and they opened earlier than shoots near the cordon. Vines of both cultivars produced a large number of canes near the trunk and few at the end of cordons. However, time of flowering was not consistently affected by position on the cordon, cane size, or type of fruiting wood. On ' 18A' vines, lateral flowers formed a separate population, opening 4-5 days after terminal flowers, and should be monitored separately. Any sampling scheme developed from these data should H00018Received 10 July 2000; accepted 22 February 2001-take account of variation within canes. Vines should be managed to reduce variation in flowering date, which has been linked to unwanted variation in fruit maturity at harvest.