Using 11 microsatellite markers, we investigated the allelic variation and genetic structure of Cryptomeria japonica, across most of its natural distribution. The markers displayed high levels of polymorphism (average gene diversity=0.77, average number of alleles=24.0), in sharp contrast to the lower levels of polymorphism found in allozyme and cleaved amplified polymorphic sequence markers in previous studies. Little genetic differentiation was found among populations (F(ST)=0.028, P<0.001), probably because the species is wind-pollinated and long-lived. No clear relationship between Nei's genetic distances and geographical locations of the populations were found using the principal coordinate and unweighted pair-group method with arithmetic averaging analyses. The lack of such trends might be due partly to microsatellite homoplasy arising from mutation blurring the genealogical record. However, there was a trend towards high allelic diversity in five populations (Ashitaka, Ashiu, Oki-Island, Yakushima-Island-1 and -2), which are very close to, or in, refugial areas of the last glacial period as defined by Tsukada based on pollen analysis data and current climatic divisions. We postulate that these refugial populations might have been less affected by genetic drift than the other populations due to their relatively large size.

We investigated gene flow and pollen contamination in a seed orchard of Cryptomeria japonica D. Don using six microsatellite markers. The quality of a seed crop is determined by many factors, including the rate of pollen contamination from outside sources, degree of self-fertilization, and equality of the clones as pollen donors. The surveyed seed orchard consisted of 62 clones and a total of 360 ramets. The seeds from 12 mother trees in the seed orchard were investigated using six highly polymorphic loci with high multipaternity exclusion probability (0.999). The estimated average pollen contamination rate was high at 47.78% (±4.12%), and the clonal self-fertilization rate was 2.22% (±0.94%) when we assumed that null alleles were present. The high pollen contamination rate could dramatically reduce the quality of the seed crop. The contamination rate differed from tree to tree, but the differences were not related to the location of each individual. The levels of contribution as pollen donors differed significantly amongst clones in the seed orchard. Differences in flowering phenology and relative pollen fecundity may be responsible for the differences in paternal contributions. The effective distance of pollen dispersal in this C. japonica seed orchard seemed to be greater than previously estimated.

BackgroundHigh-density linkage maps facilitate the mapping of target genes and the construction of partial linkage maps around target loci to develop markers for marker-assisted selection (MAS). MAS is quite challenging in conifers because of their large, complex, and poorly-characterized genomes. Our goal was to construct a high-density linkage map to facilitate the identification of markers that are tightly linked to a major recessive male-sterile gene (ms1) for MAS in C. japonica, a species that is important in Japanese afforestation but which causes serious social pollinosis problems.ResultsWe constructed a high-density saturated genetic linkage map for C. japonica using expressed sequence-derived co-dominant single nucleotide polymorphism (SNP) markers, most of which were genotyped using the GoldenGate genotyping assay. A total of 1261 markers were assigned to 11 linkage groups with an observed map length of 1405.2 cM and a mean distance between two adjacent markers of 1.1 cM; the number of linkage groups matched the basic chromosome number in C. japonica. Using this map, we located ms1 on the 9th linkage group and constructed a partial linkage map around the ms1 locus. This enabled us to identify a marker (hrmSNP970_sf) that is closely linked to the ms1 gene, being separated from it by only 0.5 cM.ConclusionsUsing the high-density map, we located the ms1 gene on the 9th linkage group and constructed a partial linkage map around the ms1 locus. The map distance between the ms1 gene and the tightly linked marker was only 0.5 cM. The identification of markers that are tightly linked to the ms1 gene will facilitate the early selection of male-sterile trees, which should expedite C. japonica breeding programs aimed at alleviating pollinosis problems without harming productivity.

Thirty four microsatellite markers for Cryptomeria japonica D. Don were developed by searching three types of library: a database of C. japonica cDNA sequences, a standard non-enriched genomic DNA library and a microsatellite-enriched library using magnetic particles. The enrichment of microsatellite sequences using magnetic particles is very efficient compared to the other two methods both in terms of the numbers of markers generated, and in the polymorphism they detect. The microsatellites developed from the genomic DNA library generally have longer repeat sequences and are more polymorphic than those from cDNA. All the developed microsatellite markers in this study showed polymorphism among 28 plus trees selected from locations scattered throughout Japan. The mean number of alleles per locus (MNA) detected in the 28 plus trees ranged from 2 to 21 with an average of 7.5. The Polymorphism Information Content (PIC) ranged from 0.160 to 0.936 with an average of 0.666. Co-dominant segregation of alleles in a three-generation pedigree of C. japonica was demonstrated at 34 microsatellite loci, and the segregation was not distorted from Mendelian expectation for all loci. In 12 out of 34 loci, a null allele was detected. Key relationships between polymorphic parameters, such as MNA and PIC, and the characteristics of microsatellite sequences, such as the longest repeat number, total repeat number and total number of nucleotides, were investigated using rank correlation coefficients, Kendall's tau. A positive correlation was found between repeat lengths and polymorphisms. The markers provide sufficient resolution for investigating gene flow within forests and seed orchards, and for genome mapping.

We investigated gene flow in five Cryptomeria japonica D. Don seed orchards of two different types (common and miniature) at widely spaced locations using microsatellite markers. The quality of a seed crop is determined by many factors, including pollen contamination from outside sources, self-fertilization, and the proportion of contributions from constituent clones. Contamination rates were found to vary among ramets both within seed orchards (10.0-76.7% in the most variable seed orchard) and among seed orchards (35.0-65.8% on average). Among ramets, there were significant negative correlations between pollen contamination rate and their distance from the orchard edge; among seed orchards, there were significant positive correlations between the pollen contamination rate and the C. japonica forest area nearby. Some proportion of the pollen (10.7% of total contamination) also migrated from parts of the orchards that had not been treated with gibberellin to induce flowering. Self-fertilization rates varied among seed orchards (1.4-4.4% on average), and there were significant positive correlations between self-fertilization rate and the number of ramets per clone in both types of seed orchard. Contributions as pollen donors differed significantly among clones in all seed orchards. The distance between planted ramets, flowering phenology, and relative pollen fecundity may also have contributed to observed differences in paternal contribution. The influence of these factors on genetic potential did not differ greatly between the two types of orchards.