Agronomic characters of some traditional rice (Oryza sativa L.) cultivars in Sri Lanka

Grain yield in rice is a complex trait and highly dependent on the other agronomic characters. Agronomic characters and grain yield of twenty traditional rice cultivars were evaluated at Faculty of Agriculture, University of Ruhuna in Maha -2010/2011 and in Yala - 2011. The experiment was conducted according to the randomized complete block design with three replicates. Twenty plants from each cultivar were evaluated for the selected characters, Plant height, Leaf blade length, Leaf blade width, Number of tillers, Number of reproductive tillers, Panicle length, Number of spikelets per panicle, Number of fertile spikelets per panicle, Number of infertile spikelets per panicle, Seed length, Seed width, 100 seed weight and Yield per plant. Recommended modern rice cultivar Bg 379/2 was used as the reference. Deviations of each character in different traditional rice cultivars from those of recommended rice cultivar Bg 379/2 were recorded. The average plant height of the tallest rice cultivar, Podihatatha was 198 cm and that of the shortest rice cultivar, Rathranwee was 68 cm. The longest leafblade and longest culm length were also belonged to (9.6 tillers per plant) and the highest average number of reproductive tillers (9 per plant), were recorded by rice cultivar Mahasudu wee . The highest value of 100 seed weight was recorded by rice cultivar Galpa wee . However, the most important economical character, the highest yield per plant(28.52g/plant) was recorded by Thanthiribalan which was significantly higher than that of in the recommended rice cultivar Bg 379/2 (26.5 g/plant). Significant correlations were found between the yield per plant and the characters such as number of reproductive tillers (r = 0.692), panicle length (r =0.565) and number of spikelets per panicle (r = 0.761). Agronomic data collected in the present study would be important to understand the suitability of an individual rice cultivar for the farmer field. DOI: http://dx.doi.org/10.4038/jur.v1i1.6150 Journal of the University of Ruhuna 2013 1(1): 3-9


Introduction
Improvement of rice grain yield is the main target of breeding program to develop rice varieties. Grain yield is a complex trait, controlled by many genes and highly affected by environment. In addition, grain yield is also related to other characters such as plant type, growth duration, and yield components (Yoshida 1981). Sri Lanka has a rich treasure of traditional rice cultivars and there are about 2000 conserved traditional rice varieties (Priyangani et al., 2008). Many are very high in nutritional value and have medicinal properties and most are resistant to extreme climatic conditions, soil conditions, diseases and pests. Conservation of genetic materials has been gaining importance as natural resources diminish (Yano & Sasaki, 1997;He et al., 1999;Rabiei et al., 2004). Reintroduction and improvement of traditional rice cultivars selected by mass selection method have been reported in different countries (Gravois & Mc New, 1993;Jannink et al., 2000;Almekinders & Elings, 2001;Gyawali et al., 2007). Some Sri Lankan traditional rice cultivars have been screened for abiotic stress tolerances such as submergence and salinity (Ranawake et al., 2010a, some Ranawake et al., 2010bRanawake et al., 2011a, Ranawake et al., 2011b. Tolerant rice cultivars for those stresses have been identified. Information on agronomic characteristics of these rice cultivars are needed for the further utilization of them in rice improvement programmes. Correlations between yield and other agronomic characters are Table 1. PGRC Accession numbers and names of traditional rice cultivars (PGRC, 1999).
important for the production of high yielding rice cultivars.
The objective of the present study was to investigate general phenotypic characters of selected traditional rice cultivars and to understand the correlations between agronomic characters and grain yield of these cultivars. Genetic variability of important traits and their association with grain yield in rice has also been studied previously (Silva, 2009).

Materials and Methods
Twenty Sri Lankan traditional rice cultivars were used for the experiment. Seeds were obtained from Plant Genetic Resource Center (PGRC) Gannoruwa. The accession numbers and names of the traditional rice cultivars used for the study are in given in Table  1. Seeds were kept at 50 o C for 5 days to break dormancy and then they were dipped in 70% alcohol for 2 minutes and washed properly with distilled water. Seeds were then dipped in 2% Clorox for 30 minutes and were washed properly with distilled water. Seeds were kept in an incubator at 35 o C for 7 days under dark condition and germinated seeds were planted in trays for two weeks. Two week old seedlings were then transplanted in the paddy field according to the randomized complete block design with three replicates and 20 plants per replicate with 15 cm X 20 cm spacing. Fertilizers were applied at rates of Urea 50 kg/ha, TSP 62.5kg/ha, MOP 50kg/ha at sowing time. Urea at a rate of 37.5 kg/ha was also applied as top dressing after 2 weeks of planting and 7 weeks of planting.
The rice yield, yield components and other characteristics were determined according to the method of Standard Evaluation System for Rice (IRRI, 1988). The average results of the above measured parameters are given in table 3. Data were analyzed using statistical analysis software; SAS institute inc., (2000). Bg 379/2 was used as the reference recommended rice cultivar.     (Table 3). In the present research, plant height was positively correlated with leaf blade length (r = 0.793), leaf blade width (r = 0.505), seed width (r = 0.18) and yield per plant (r = 0.347) whereas it was negatively correlated with 100 seeds weight (r = -0.316) ( Table  4). However, it has been reported that plant height was affected by many factors like plantation method, plant density and fertilizer application (Beser & Genctan, 1999;Aide & Beighly, 2006;Gozubenli, 1992).
There were strong positive correlations of panicle length with number of spikelets per panicle (r = 0.631), 100 seeds weight (r = 0.457) and yield per plant (r = 0.565) (

Conclusion
Broad variations in agronomic characters have been found in tested traditional rice cultivars. Characters such as number of reproductive tillers, panicle length and number of spikelet per panicle are directly responsible for the final yield whereas characters such as plant height, leaf blade width and numbers of total tillers are not very important as yield determinants. However, all these characters are taken in to consideration when the rice cultivars are selected for the commercial cultivation. The present study gives average values for each parameter. These primary data can be utilized when traditional rice cultivars are screened for future breeding programs.

Acknowledgement
Authors acknowledge PGRC Gannoruwa for providing plant materials.