A population 'consensus', partial linkage map of Picea abies Karst. based on RAPD markers

We built a ''consensus'' partial linkage map based on RAPD markers using 48 sibships of eight megagametophytes each from a natural population of Norway spruce. A RAPD linkage map for a single individual from the same population had previously been constructed. Using 30 random decamers that had yielded 83 RAPD markers in the single-tree map, eight megagametophytes for each of the 48 sibships were screened. The linkage relationship among markers was estimated considering each family of eight megagametophytes as a progeny of a phase-unknown backcross mating between a heterozygous mother and a fictitious 'recessive' father. Markers were assigned to windows using LOD = 2.0 and theta = 0.4 as thresholds, and ordered using a criterion of interval support greater than or equal to 2.0. For eight ''windows'' of recombination selected on the single-tree map, we investigated the consistency of marker order in the two maps. We adopted restrictive criteria for rejecting co-linearity between the two locus orders. For each window we imposed the most likely locus order obtained from one data set to the other (and vice versa), obtaining two symmetrical log-likelihood differences. We considered the hypothesis of colinearity rejected when both symmetrical differences were significant (Delta LOD > 3.0). By bootstrapping a subset of markers for each window (highly informative, 'framework' loci chosen on the previous single-tree map using a matrix correlation method) the sampling variability of the single-tree and population maps was estimated. As expected the population map was affected by a larger variability than the single-tree map, Heterogeneity in pairwise recombination fractions among groups of sibship revealed a (possible) alternative genomic arrangement detected within a single recombination window.