In order to construct a mapping population with high level of polymorphism and normal fertility as an intention to reduce segregation distortion, we screened a total of 74 accessions of cultivated rice varieties (0. sativa) and the Asian common wild rice (0. rufipogon) for DNA polymorphisms with 76 RFLP probes. A total of 144 crosses resulting from intermating these accessions were evaluated for F₂ fertility. A cross between Aijiao Nante, an indica cultivar widely used in southern and central China in the 1960 and 1970s', and an accession of the common wild rice (P16) from Dongxiang County, Jiangxi Province was selected for developing the mapping population. An F₂ population of 172 individuals was raised in 1994 and has been maintained vegetatively ever since both in Wuhan and Hainan (South China Sea) Island.

The parents were screened with a total of 391 probes from the Cornell group (prefixed RG and RZ) and 380 probes from RGP (prefixed G, C, or R). We also included a number of probes from various sources including cDNAs of oat (CDO), genomic clones of barley (BGL) and wheat (WG). The F₂ population was assayed individually with a total of 390 probes including 180 from the Cornell group, 200 from RGP and 10 from other sources. Over all, 412 polymorphic loci were detected with the 390 probes, of which 369 probes each resolved a single polymorphic locus and the remaining 21 probes each detected two or more polymorphic loci.

A framework map, consisting of 359 loci, was constructed using Mapmaker 3.0 (Lincoln et al. 1992) with LOD threshold 3.0. The remaining 53 loci were placed on their most probable intervals with LOD scores less than or equal to 2.0. The average distance

in the map was 4.1 cM between adjacent marker loci. The map well integrated the Cornell and RGP maps (Fig. 1). The total length of the resulting map, 1705 cM, is 14.4% longer than the Cornell map and 8.5% longer than the RGP map. Eleven of the 12 chromosomes in our map are longer than their respective counterparts in the Cornell and the RGP maps: chromosome 7 was 8% shorter in our map than in the Cornell map and chromosome 4 was 1.1 % shorter than that in the RGP map. There were four gaps, with intervals larger than 25 cM, located on chromosomes 3 (27.2cM), 5 (33.6cM), 7 (23.1 cM) and 8 (22.6cM), respectively. These gaps are larger than the existing gaps in the Cornell and RGP maps. It should also be noted, however, that the markers in this map have bridged some of the major gaps in the Cornell and RGP maps due to the integration of the markers from both maps. Examples are the largest gap of 21.2cM on chromosome 4 of the RGP map, which was reduced to an interval of 10.6cM, and the largest gap of 23.0cM on chromosome 7 of the Cornell map, which was filled with a number of markers from the RGP map.

The orders for majority of the marker loci are very well conserved between the maps. Together, 322 loci resolved by 312 probes appeared in the same order as they did in the Cornell and RGP maps. The remaining loci fell into three different categories: (1) 34 loci, detected by 31 probes (23 from Cornell and 8 from RGP), were mapped to the same chromosomes as they were in the other two maps, but their relative positions were somewhat exchanged with their neighboring loci; (2) 12 loci, detected by 12 probes respectively (11 from Cornell and 1 from RGP), were mapped to chromosomes different from where they appeared in the Cornell and RGP maps, and; (3) 34 loci, resolved by 25 probes ( 13 from Cornell and 12 from RGP), appeared in this map but not in the Cornell or RGP maps: most of these loci were resolved by probes detecting multiple loci (11 of the 12 RGP probes were in this category), indicating the possibility that different copies were mapped in different studies.

Although the cross was selected on the basis of normal fertility of the F₁ hybrid with the intention to reduce segregation distortion, it is evident that segregation distortion was still common involving 124 of the 412 loci (Fig. 1). Segregation distortion was significant at the 0.01 probability level for 96 of the 124 loci and at the 0.05 probability level for the remaining 28 loci. Although skewed segregations were observed in 15 contiguous genomic regions involving 10 of the 12 chromosomes, they occurred mainly in three chromosomal segments: the entire short arm of chromosome 3, almost entire chromosome 6 and the major portion of chromosome 12. Most of the distorted segregations were in favor of the alleles from the cultivar Aijiao Nante; homozygotes for the Aijiao Nante alleles were in excess at 100 of the 124 loci located on chromosomes 3,5,6, 11 and 12, respectively. In contrast, homozygotes for the alleles from the wild parent were favored at only 20 loci located on chromosomes 1, 2 and 4. In addition, segregations at the four contiguous loci in the centromeric region of chromosome 7 were skewed in favor of heterozy-gotes. Many of the genomic regions showing distorted segregations were consistent with previously published results (Causse et al. 1994; Harushima et al. 1996; Xu et al. 1997). An extreme case is the segregation distortion occurring on the short arm of chromosome 3 which was detected in almost all the published data. In connection with many previous

114 Rice Genetics Newsletter Vol. 14

findings including high proportion of segregation distortion observed even in an intrasubspecific cross (e.g. Yu et al. 1997), we concluded that segregation distortion is a common phenomenon in rice.

This map may be useful in several ways. Firstly, it will provide a cross-reference to the two widely used high density linkage maps, thus facilitating information transfer from one map to the other or vice versa. Secondly, the wild parent has a number of desirable characteristics for rice improvement, such as tolerance to biotic and abiotic stress. The development of this map will greatly facilitate the analysis and exploitation of these characteristics. Thirdly, inclusion of a wild rice, the putative progenitor of the cultivated rice, as one of the parents of the mapping population will provide the opportunity to investigate the genetic basis of evolutionary changes associated with the domestication processes.