Knowledge of genetic diversity among prospective parental lines is important for the success of a hybrid rice breeding program as it determines the magnitude of heterosis in F1 hybrids to a large extent. Genetic diversity can be measured by different methods such as pedigree analysis, morphological data and using molecular markers. Inter simple sequence repeat (ISSR) markers are cost effective and informative for genetic diversity analysis.

The objective of this work was to compare the genetic diversity among 275 parental lines using morphological and molecular data. The 35 maintainer lines (B lines) and 240 restorer lines (R lines) included those developed indigenously and at IRRl. Thirteen morphological characters were studied in five plants of each line. Three highly informative primers (UBC 834, 841 and 842) were used singly in ISSR-PCR. The amplified products were resolved using 6% denatured polyacrylamide gel electrophoresis and silver staining. All data was analyzed using NTSYS pc 2.0 ( Rohlf 1994 ).

The similarity coefficient based on morphological data ranged from 0.02 to 1.00. All nodes were between 0.02 to 0.35 indicating a wide diversity. The dendrogram showed that 2 R lines (EPLT-104, HKR 93-1) were most distinct followed by 3 R lines (Pusa Basmati, NLR-33358 and NSN-5). These 5 lines were distinct from the other 270 lines, which grouped in about 22 small clusters. The B lines were interspersed with R lines with at least one B line in 21 clusters. One pair of B and R lines (PMS 10B, IR61614-38) could not be distinguished from each other.

In molecular analysis, 17034 ISSR markers were scored at 81 band positions. The percentage polymorphism revealed and resolving power of the primers 834, 841 and 842 were 100, 91.3, 95.5 and 4.1, 10.1, 14 respectively. The similarity coefficient based on molecular data ranged from 64 to 99. The nodes were between 0.64 to 0.99, indicating more ability to differentiate and more similarity than based on morphological data. The dendrogram could help distinguish all the 275 lines except 2 pairs of R lines (IR62037-93, IR62917-97 and 205-14, 203-3-P1). Lines derived from the same crosses grouped together indicating the robustness of the clustering. There were 7 such pairs of lines.

Two R lines (NSN5, 1189-1) were distinct and separated from the 273 accessions at similarity coefficient of 64. The next to separate were two B lines NP16 and NP45 at 68 followed by two more lines, a B line (MRST-16) and one R line (IR44675-101). The remaining 269 lines grouped in 6 large clusters. It was significant that B lines were clustered together unlike in the dendrogram based on morphological data. As many as 22 B lines grouped in cluster 1 of 38 lines. These 22 lines included IR 58025 B, the B line of IR 58025 A a popular cms line which has been used in 11 of the 18 released hybrids in India. Two B lines (PHS 75, NRI 17-4-2) were in the next closest cluster. Five indica x japonica derived B lines including three New Plant Type lines (NPT30, NPT34, NPT38) and two B lines (NRI-52-P3, SC5-12-04-8) formed the next closest cluster. NPT30 and NPT34 were 97% similar. Of the 3 remaining B lines CRMS 32 B, IR 62829 B grouped in one and IR69628 B in another cluster. Thus most of the B lines are very closely related. Clusters 4 and 6 consisted largely of the IRRI lines. There

were 12 pairs of R lines the members of which were more than 95% similar. The R lines of popular widely grown hybrids KRH2, CNRH3, CORH2, NDRH 2, ADTRH 1 and APHR 2 were in one cluster and those of three hybrids Sahyadri, Pant Sankar Dhan-1and DRRH-1 in another cluster which was distant from that of their B line IR 58025B (Table 1). Lines derived from indica x japonica crosses were frequent in clusters 1, 3 and 6 each with 7 or 8 such lines and absent in clusters 2 and 4 the clusters containing the restorers of popular hybrids. This confirms earlier observations that potential maintainers have more japonica alleles and potential restorers have more indica alleles (Devanand et al 1999). Six restorers derived from indica x indica crosses grouped together. There were two groups each of 4 isocytoplasmic restorers. Pusa Basmati was in Cluster 2 and C20R, IR36, IR64, IR66 in cluster 4. Similar clustering of these 5 accessions was obtained using SSR markers (Ravi et al 2003).

The two similarity matrices (morphology data and molecular data) for 39675 pair wise comparisons were not significantly correlated indicating that the diversity based on morphological data was different from that based on molecular data. The study based on molecular marker data points to the need for widening the genetic base of both the parental lines. This would help in increasing the magnitude of heterosis further.

References

Devanand, P.S., J. Wan, M. Rangaswamy and H. Ikehashi, 1999. Isozyme divergence between maintainers and restorers in hybrid rice breeding program in India. Crop Sci. 39: 831-835.

Ravi, M., S. Geethanjali, F. Sameeyafarheen and M. Maheswaran, 2003. Molecular marker based genetic diversity analysis in rice (Oryza sativa L) using RAPD and SSR markers. Euphytica 133: 243-252.

Rohlf, F.J., 1994. NTSYS-Pc: Numerical taxonomy and multivariable analysis of system version 2.2. State University of New York, stony brooks NY.