López-Sesé, A.I., U.S. Department of Agriculture, Vegetable Crops Unit, University of Wisconsin Madison, 1575 Linden Dr., WI 53706, United States
Staub, J., U.S. Department of Agriculture, Vegetable Crops Unit, University of Wisconsin Madison, 1575 Linden Dr., WI 53706, United States
Katzir, N., Department of Vegetable Crops, Agricultural Research Organization, Newe ya'Ar Research Center, P.O.B. 1021, Ramat Yishay 30095, Israel
Gómez-Guillamón, M.L., Experimental Station la Mayora, Consejo Superior de Investigaciones Científicas (CSIC), Algarrobo (Málaga) 29750, Spain

The population structure of 15 Spanish melon (C. melo L.) accessions, mostly of Group Inodorus, was assessed by the analysis of 16 individuals of each accession using 100 random amplified polymorphic DNA (RAPD) bands produced by 36 primers, and allelic variation at 12 microsatellite (SSR) loci (23 alleles). A relatively high level of polymorphism (25.6%) was detected using RAPD markers, and eight SSR loci (66.7%) were useful in discriminating accessions. Cluster analysis using RAPD- and SSR-based genetic distance estimates resulted in similar and consistent groupings of most of the accessions studied. The mean genetic distance and standard error among accessions estimated by RAPD variation was 0.421 ± 0.099, and mean SSR-based genetic distance estimate was 0.285 ± 0.141. Albeit many dominant markers examined were fitted to a 3:1 test ratio, deviation from this ratio and from Hardy-Weinberg expectations for many SSR loci suggests that some populations were in genotypic disequilibrium. Moreover, a higher level of genetic variation was observed between Cassaba market classes than within accessions, suggesting that, depending upon the accession, allelic fixation has occurred in these populations. The relatively high level of heterogeneity observed (different band morphotypes and cluster grouping within a particular market class), however, indicates that the Spanish melons examined possess a relatively broad genetic background. An appraisal of accession population structure such as the one reported herein indicates that bulk sampling techniques coupled with molecular analysis techniques that employ a unique array of discriminating markers can provide information leading to effective strategies for diversity analyses of large collections.