Theor Appl Genet (1989) 78:65-75 9Springer-Verlag 1989 RFLP analysis and linkage mapping in Solanum tuberosum C. Gebhardt, E. Ritter, T. Debener, U. Schachtschabel, B. Walkemeier, H. Uhrig and F. Salamini Max-Planck-Institut fiir Ziichtungsforschung, D-5000 K61n 30, FRG Received December 23, 1988; Accepted December 30, 1988 Communicated by G. Wenzel Summary. A morphologically and agronomically hetero- geneous collection of 38 diploid potato lines was ana- lysed for restriction fragment length polymorphisms (RFLPs) with 168 potato probes, including random genomic and cDNA sequences as well as characterized potato genes of known function. The use of four cutter restriction enzymes and a fragment separation range from 250 to 2,000 bases on denaturing polyacrylamide gels allowed the detection of RFLPs of a few nucleotides. With this system, 90% of all probes tested showed useful polymorphism, and 95% of those were polymorphic with two or all three enzymes used. On the average, 80% of the probes were informative in all pairwise comparisons of the 38 lines with a minimum of 49% and a maximum of 95%. The percentage of heterozygosity was deter- mined relative to each other for each line and indicated that direct segregation analysis in F1 populations should be feasible for most combinations. From a backcross involving one pair of the 38 lines, a RFLP linkage map with 141 loci was constructed, covering 690 cMorgan of the Solanum tuberosum genome. Key words: Solanum tuberosum - RFLP - Linkage map Introduction Besides direct sequence comparisons, restriction frag- ment length polymorphisms (RFLPs) are the most sensi- tive tool for the detection of DNA differences within or between species. The molecular basis of RFLPs is the loss or gain of a restriction site due to a point mutation within the enzymes recognition sequence, or a molecular event leading to insertion, deletion or inversion. Both situa- tions result in a length difference of genomic restriction fragments detectable on Southern blots. RFLPs are genetic markers with several advantages compared to conventional markers. They describe direct- ly the genotype instead of the phenotype and, therefore, are not influenced by the environment. The number of RFLP markers which can be mapped is only limited by the molecular differences existing between available genotypes; it is also proportional to the effort applied to detect them. RFLP studies and detailed RFLP linkage maps are recognized as important contributions of mo- lecular genetics to plant breeding (Beckmann and Soller 1986; Burr et al. 1983; Tanksley 1983) as well as to the fields of plant taxonomy and evolution (Song et al. 1988; Hosaka and Hanneman 1988 a, b). In recent years RFLP linkage maps have been devel- oped in several diploid plants, the most advanced being maize (Helentjaris et al. 1986; Helentjaris 1987), tomato (Bernatzky and Tanksley 1986a; Helentjaris et al. 1986; Zamir and Tanksley 1988), lettuce (Landry et al. 1987 b), rice (McCouch et al. 1988), pepper (Tanksley et al. 1988) and Arabidopsis (Chang et al. 1988). The cultivated po- tato (Solanum tuberosum ssp. tuberosum) is a tetraploid with poorly developed cytogenetics. In fact, no linkage map is available for the species, although a few linkages among isoenzymes have been obtained by Douches and Quiros (1987; see also Desborough 1983). A comparative RFLP map between tomato and po- tato has been recently obtained from an interspecific cross involving the wild species S. phureja and S. cha- coense and a diploid S. tuberosum line (Bonierbale et al. 1988). However, it seemed desirable to us to produce a RFLP linkage map within a gene pool of diploid S. tuberosum, where many crosses of agronomic interest can be analysed with the aim of mapping disease resis- tance genes or quantitative trait loci. Therefore, we con- ducted a survey of RFLPs present in a collection of 38 diploid potato breeding lines, which were selected for