Abstract
Thirty indica rice (Oryza sativa L.) genotypes comprising different ecotypes of Bangladesh were studied to ascertain genetic divergence and relationship among and between them. The research was carried out during the period from July 2001 to the end of 2003.Twenty seven cultivated (Traditional cultivars) and three wild rice genotypes were studied for nine quantitative characters related to divergence of the genetic traits. Days to heading, maturity, plant height, panicle length, filled spikelet per panicle, total spikelet per panicle ands panicle weight contributed largely toward varietal divergence. Clusters showing maximum divergence among them were also genetically distant from each other for the relevant traits. Proximity of the cultivars with wild genotypes indicated that many of the cultivated genotypes were very close to the wild ones, which gave a positive indication of the presence of close phylogenetic relationship among them. Five genotypes (2 wild and 3 cultivated) were selected one from each of Aus, Aman and Boro ecotypes; and two wild genotypes on the basis of proximity for morpho-physiological characters. Wild 4855, Jhora 4325(wild), Dular, Habigonj Aman 2 and Jagliboro were evaluated and characterized for selected important qualitative and quantitative characters, viz. pigmentation in plant parts at different developmental stages, presence or absence of awn, period to reach at reproductive stages, panicle and yield contributing characters. Low Euclidian distance was found among Jhora 4325, Jagliboro, Dular and Habigonj Aman 2, respectively, indicated the close relationship among these three with each other for the quantitative characters. With respect to isozyme banding pattern, the zymograms of the selected five rice genotypes fall into two categories: Jhora 4325 and Habigonj Aman 2 showed one pattern and Wild 4855, Jagliboro and Dular had a different pattern. Overall low biochemical distance was present among Jhora 4325, Jagliboro, Habigonj Aman 2 and Dular respectively than Wild 4855 for leaf peroxidase. Only Habigonj Aman 2 was found to be weekly photoperiod sensitive under natural day length whereas other 4 genotypes Wild 4855, Jhora 4325, Jagliboro and Dular were photoperiod insensitive. Smaller variation was observed in days to flowering in wild genotypes than cultivated ones. Studies on embryo germination and regeneration ability of rice in tissue culture media revealed that only Jhora 4325 showed seed dormancy even in the culture media, in field condition seed dormancy was found in Jhora 4325 and Wild 4855. Ability for callus induction, organogenesis and plantlet establishment was found more in the two wild genotypes than the three cultivated ones. For Seed set percentage in diallel progenies (without reciprocal) of wild and cultivated rice genotypes ranged from 4.91 to 65.77%. Crossability of Jhora 4325 was found highest with Jagliboro followed by HBAman2, Dular and Wild 4855. A high degree of Genetic and sexual affinity was observed for spikelet fertility. The observations indicated the strong relationship among the five rice genotypes regarding the reproductive system. Inheritance of pigmentation and appearance of awn in F, and F2 generation derived from crosses showed that gene(s) controlling pigmentation and awn formation were dominant. Seedling base, internode and apiculus pigmentation was found monogenically controlled (3:1) in all cross combinations. Stigma color was found to be digenically controlled (9:7) in case of the single Habigonj Aman 2 x Jagliboro cross and monogenecally controlled in all other cross combinations, similar situation was found in respect to awn formation. The monogenic and digenic ratios in the F2 generation indicated pleiotropic and complementary gene actions for pigmentation and awn formation. Genetic component analysis showed overdominance gene action for seed germination rate, primary branch per panicle, secondary branch per panicle, filled spikelet per panicle, total spikelet per panicle, and partial dominance for days to 50% flowering, days to maturity, plant height, culm length, last internode length, no. of internode, leaf length, leaf width, tiller no. per plant, panicle length, unfilled spikelet per panicle and panicle weight. Narrow sense heritability was high for most of the characters. Both additive and dominant gene effect was observed for all of the characters except for number of internode, leaf width, panicle length, primary branch per panicle and panicle weight where non- genetic factor was in operation. Dominant alleles were found to be more frequent than recessive alleles in the parents for most of the characters. Proportions of dominant and recessive alleles were found in both directions. Equal distribution of dominant and recessive alleles was noticed in the parents for maximum number of characters. Preponderance of recessive alleles had positive effects to the desirable direction in parents for the characters- like: seed germination rate, days to maturity, last internode length, leaf length and unfilled spikelet panicle1· Expression of midparent heterosis was found to be low for most of the characters and estimated inbreeding depression was highest for plant height, tiller number per plant, panicle length, primary branch per panicle, secondary branch per panicle, total spikelet per panicle in the hybrid generation studied. Low midparent heterosis and high inbreeding depression also represent genetic relationship among the selected five rice genotypes. Based on the findings in respect of proximity in performance of various characters, Euclidian distance, peroxidase enzyme activity, crossability, F, pollen and spikelet fertility, GCA SCA effect patterns and heterosis - inbreeding depression magnitude - the possible evolutionary differentiation mode in the selected rice genotypes could be as follows: Jhora 4325- Jagliboro → Dular ➡>> Habigonj Aman 2. The other possible evolutionary links with Wild 4855 as putative progenitor could be as: Wild 4855 Dular→ Hab. Aman 2; or, Wild4855 → Hab. Aman 2 → Dular.