Methods

Leaves of Vitis vinifera L cultivars were collected from the ampelographic collection of the Laboratory of Plant Physiology and Biotechnology, University of Crete and the Institute of Viticulture, Floriculture and Vegetable Crops of Heraklion (National Agronony Research Foundation; outdoors). Leaves were kept frozen at ' 80'C until extraction.

Amplification primers sequences for nuclear microsatellite loci from Vitis riparia (Steinkellner et al., 1999), ssrVrZAG 21, ssrVrZAG 47, ssrVrZAG 62, ssrVrZAG 64, ssrVrZAG 79 ssrVrZAG 83 and from Vitis vinifera (Thomas and Scot, 1993; Bowers et al., 1996, Lefort et al., unpublished) VVS2, VMD5, VMD7, UCH2, UCH3, UCH11, UCH12, UCH17, UCH 18, UCH19, UCH29, UCH35, UCH40, UCH47 were used for DNA amplification. PCR amplifications were carried out in 96-wells propylene plates in 20 'l final volume reaction mixtures in a PTC-100 thermal cycler (M.J. Research Inc., Watertown, Ma.,USA). PCR reactions were as follows: 1 'M of each primer, 100 'M of each dNTPs (Biofinex, Praroman, Switzerland), 1.5 mM MgCl2 in the following buffer 75 mm Tris HCl (pH 9.0), 50 mM KCl, 20 mM (NH4)2SO4, 0.5 units Taq polymerase(Biotools, Madrid, Spain) and 50 ng DNA template. The forward primer was in each case labelled with the Licor IR800 fluorochrome. The following thermal cycling protocol was applied : 95 'C for 5 min, 10 cycles of 15 s at 51 'C, 15 s at 94 'C, followed by 23 cycles of 15 s at 51 'C, 15 s at 89 'C and terminated immediately at 4 'C. Amplification success was checked by electrophoresis in 2% agarose gel in 1xTBE buffer (Tris, boric acid, EDTA pH 8.0) and ethidium bromide staining. PCR products analysis was carried on in 2% polyacrylamide 7M urea sequencing gels in a Licor 2400 DNA Sequencer (Licor, Lincoln, Nebraska, USA) and alleles were sized with the software Gene Profiler v3.54 (Scanalytics, Fairfax, Va., USA).

According to Echt et al. (1998), the chloroplast genome, which does not exist in a heterozygous state and does not genetically recombine, 'may be viewed as a single locus and all sequence variation interpreted as giving rise to different haplotypes of the genome'. An alternative approach may consider the chloroplast genome as 'a circular haploid chromosome wherein sequence variation generates different alleles within individual, non recombinant loci'. We used the proposed terminology where the term 'locus' refers to 'a cpSSR site, defined by the termini of a PCR primer pair', and 'alleles' to 'length variants at a cpSSR site'. We used six primer pairs designed for dicotyledonous angiosperms by Weising and Gardner (1999): ccmp2, ccmp3, ccmp4, ccmp6, ccmp7 and ccmp10.
PCR amplification was performed using a Perkin Elmer l 9600 thermal Cycler with the following profile : 5 min denaturation at 95 'C , 5 min at 80 'C during which the enzyme is added, 25 cycles [1 min at 94'C , 1 min at 50 'C, 1 min at 72'C], followed by a final step at 72'C for 8 min. Reactions were carried out in 96 wells polypropylene plates in a 25 'l final volume containing dNTPs (0.2mM each), 2.5 mM MgCl2, 0.2 'M of each primer, 10x reaction buffer (Pharmacia), 10 ng DNA and 1 u of Taq polymerase (Pharmacia). PCR amplifications were triplicated at all loci for all DNA samples.

Identity 1.0 (Sefc and Wagner, 2000) was used for formatting the microsatellite data into the Microsat format. A distance matrix was constructed from microsatellite data with Microsat (Minch et al., 1997). The distance used was [-Log (proportion of shared alleles)]. A phenogramme was then drawn by using Kitsch of the Phylip package (Feselstein, 1989) and Treeview (Page, 1996).