Varietal Evaluation and Genetic Variability in Rice (Oryza sativa L.) Genotypes of the Mid-Hill Region of Nepal

The objectives of this study were to identify promising rice genotypes and evaluate the genetic variance and effectiveness of selection of the rice varieties for several yield attributing traits. A varietal trial of fifteen rice genotypes was laid out in a randomized complete block design (RCBD) with three replications in a farmer’s field in Sundarbazar, Lamjung, Nepal during the rainy season of 2018. Analysis of variance indicated that all the genotypes showed significant variation for all the traits considered. The phenotypic coefficient of variation (PCV) was higher than the genotypic coefficient of variation (GCV) for all the characteristics being studied indicating the presence of environmental influence on the traits. High heritability coupled with high genetic advance as a percent of the mean was found for days to physiological maturity, number of tillers per m, plant height, leaf area, effective tillers per m, flag leaf area, test weight, grains per panicle, filled grains per panicle, harvest index, grain yield, and straw yield indicating that additive gene interaction is present in their inheritance. Direct selection can be effective for yield improvement in the populations through selection of these traits. Cluster analysis based on eighteen traits grouped the fifteen rice genotypes into four clusters. Cluster I was the largest and consisted of five genotypes. Radha 11, NR 119, and Sukhadhan-5 were the top performing genotypes having yield potentials of 5.78, 5.49, and 4.89 tons per ha, respectively.


Introduction
In Nepal, rice is the main staple crop grown in all three major ecological regions accounting for 73%, 24%, and 3% of the total rice cultivated in the Terai, Hill, and Mountain areas, respectively (MoAD, 2016). More than 70% of rice is grown under rainfed conditions, 9% under upland conditions, and 21% under partially or fully irrigated conditions (NARC, 2009). Rice fields occupy 1.36 million ha of land with an annual production of 4.23 million metric tons and productivity of 3.154 tons per ha (MoAD, 2016). The production and productivity of rice in Nepal are much less compared to other rice-growing countries. Agriculture in Nepal is a subsistence type where each farmer derives only a small part of their food requirements from fragmented plots of land cultivated under challenging conditions. Food insecurity is a major ongoing problem in Nepal, resulting in the rising problem of malnutrition in the country. The ever-increasing population is a major constraint in supplying enough food for each person's daily requirements. Climate change and natural calamities like drought, flooding, and fluctuations in rainfall patterns are also serious threats to food security.
The presence and magnitude of genetic variability for important traits in the gene pool are the basic guides for any breeding program (Akinwale et al., 2011). Such variability is either naturally present or created using various techniques (Pandey et al., 2009). The presence and magnitude of variability can be accessed by employing specific tools such as GCV, PCV, heritability, and genetic advance (GA), among others. Cluster analysis based on agromorphological and yield component traits has shown considerable genetic diversity among genotypes themselves and can group them into distinct clusters .
Previous rice researchers such as Yadav et al. (2019) conducted participatory evaluations of cold-tolerant rice in the mid-hill region of Nepal. Adhikari et al. (2018) also evaluated advanced lowland irrigated rice in Doti, Nepal. Bhandari et al. (2019) evaluated eleven rice genotypes in Lamjung and assessed their genetic variability. Therefore, this study was undertaken to assess the nature and magnitude of genetic variability, heritability, and genetic advance among these genotypes for further utilization in breeding programs.

Experimental site
The present study was carried out in a farmer's field in Sundarbazar, Lamjung, Nepal during the rainy season of 2018. The site is situated at 28.09°N latitude, 84.47°E longitude, and an altitude of 650 meters above mean sea level. Geographically the site is in the Gandaki province of Nepal.

Experimental design and genetic materials
The design of experiment was a RCBD with three replications and fifteen rice genotypes as the treatment as shown in Table 1. All the genetic materials were obtained from the Regional Agriculture Research Station (RARS), Khajura, Banke. Manabahu was used as the local check and Sukhadha-5 was used as the standard check. Twenty-five day old seedlings were transplanted in wells prepared in the puddled field. Seedlings were transplanted with a crop geometry of 20cm × 20cm and a plot area of 2m 2 , and then proper care was given with respect to water management, insect pest control, and nutrient management. Doses of 15 tons per ha of farmyard manure were applied at the time of land preparation and chemical fertilizer 60:20:20 NPK kg per ha in the forms of urea, diammonium phosphate (DAP), and potash, respectively. Full doses of phosphorus and potash and half doses of nitrogen were applied at basal doses while the remaining nitrogen was applied as split doses at the two stages of tillering and panicle initiation.

Data collection
Ten sample plants were observed for recording data. Observations were recorded for the phenological characteristics of days to booting, days to heading, and days to maturity; agro-morphological characteristics, namely plant height, flag leaf length, flag leaf area, panicle length, and soil plant analyzer development (SPAD) reading; yield and yield attributing characters, namely number of effective tillers per m 2 , number of tillers per m 2 , filled grains per panicle, grain yield, straw yield, harvest index, fertility percentage, and test weight (thousandgrain weight).

Calculation and statistical analysis
All taken data were tabulated and processed through MS Excel version 19. The coefficient of variation (CV), mean, and analysis of variance (ANOVA) were computed using R studio version 3.6.1. Least significant difference (LSD) was computed at the 5% level of significance. Clustering of genotypes was done by Minitab version 14. PCV and GCV were computed using the formulas suggested by Chaudhary & Prasad (1968) and Deshmukh et al. (1986), respectively. Broad sense heritability, GA, and genetic advance as a percentage of the mean (GAM) were computed by the formulas given by Robinson et al. (1949), Johnson et al. (1955), andFalconer (1996), respectively.

Analysis of variance
The analysis of variance suggested that there were significant differences for all the agromorphological and yield component traits of the studied rice genotypes as shown in Table 2. Similar results were reported by Bhandari et al. (2019). Adhikari et al. (2018) reported the existence of significant differences for days to flowering, maturity, plant height, panicle length, thousand-grain weight, and grain yield, which is accordance to our findings. Gyawali et al. (2018) evaluated seven rice varieties in Lamjung, Nepal and reported significant variation in the genotypes for the traits of days to booting, days to heading, days to anthesis, plant height, effective tillers per m 2 , thousand-grain weight, straw yield, and harvest index.

Mean performance of the rice genotypes on the phenological and agro-morphological characteristics
The performance of the fifteen rice genotypes based on the phenological and agromorphological traits are shown in Table 3. Brief explanations of each trait are given below.

Days to heading (DH)
Days to heading of the local landrace Manabahu (112 days) was found to be significantly higher and statistically on par with Sawa Mansuli sub-1, NR-601-1-9, and NR 1190. Hardinath-1 had the fewest days to heading (82 days) which was statistically on par with Chaite 4 and DRR 44. Hardinath-1 was one of the early genotypes and had a short vegetative phase and long reproductive phase.

Days to booting (DB)
The fewest days to booting was recorded for Hardinath-1 (75 days) which was statistically on par with Radha 11, Chaite 5, DRR 44, Bahuguni, Ghaiya 2, Chaite 4, and Cheherang sub-1. These genotypes were earlier in terms of initiation of the reproductive period. Manabahu showed the longest days to booting (102 days) and its reproductive phase started later.

Days to maturity (DM)
NR-601-1-9 showed the highest number of days to maturity (161 days) which was statistically on par with Manabahu, but Hardinath-1 (116 days) had the lowest number of days to maturity which was statistically on par with Chaite 5. This indicates that Hardinath-1 and Chaite 5 were early varieties, which can be selected for based on the variability analysis.

Leaf area (LA) (cm 2 )
Radha 13 showed significantly superior leaf area than all the other genotypes and was statistically on par with NR-601-1-9, whereas Sawa Mansuli sub-1 had the lowest leaf area.

Flag leaf area (FLA) (cm 2 )
NR-601-1-9 showed the highest flag leaf area which was statistically on par with Radha 13 and NR 1190, while Cheherang sub-1 had the lowest flag leaf area.

Mean performance of the rice genotypes in the yield and yield attributing traits
The performance of the fifteen rice genotypes based on the yield and yield components traits are shown in Table 4. Brief explanations of each trait are given below.

Number of tillers per m 2
Ghaiya 2 had a significantly higher number tillers per m 2 , which was statistically on par with NR 1190, Radha 11, Bahuguni, and Chaite 4, but the NR-601-1-9 variety had the lowest number of tillers per m 2 . All the varieties except NR-601-1-9, Radha 13, Hardinath-1, and Manabahu, showed a statistically similar number of tillers per m 2 with Sukhadhan-5.

Genetic variability
Estimation of the genetic variability was performed to determine the heritable and nonheritable portion of concerned traits, which is crucial in adopting breeding methods. The heritable portion of the overall observed variation can be determined by studying the components of variation such as the coefficients of genotypic and phenotypic variability, heritability, genetic advance, and genetic advance as a percentage of the mean, which is a prerequisite for the improvement of crop.

GCV and PCV
Estimated values of the GCV and PCV for all the traits evaluated in this study are shown in Table 5. The coefficients of variation of the studied traits showed that the values of the PCV were slightly more than the GCV for all the traits. This indicates the presence of environmental influence in the expression of these traits (Rasel et al., 2018). Earlier researcher on rice, such as Kumar et al. (2018) Kumar et al. (2018) identified high GCVs and PCVs for effective tillers per m 2 , filled grains per panicle, thousand-grain yield, harvesting index, and grain yield, which is accordance with our findings. Similarly, high GCVs and PCVs were recorded by Prasad et al. (2017) for grain yield, filled grains per panicle, thousand-grain weight, and effective tillers per m 2 , and by Srivastava et al. (2017) for filled grains per panicle. High GCVs and PCVs indicate the presence of high variance in the traits and less influence of the environment in the expression of these traits, hence, direct selection may be effective for improvement of these traits (Srivastava et al., 2017). Bhandari et al. (2019) reported high GCVs and PCVs for thousand-grain weight, straw yield, leaf area, grain yield, and effective tillers per m 2 .
Moderate values of the GCV and PCV were recorded for leaf area, effective tillers per m 2 (19.78 and 22.97), plant height (15.48 and 16.29), numbers of tillers per m 2 (15.18 and 18.49), and days to maturity (10.63 and 10.69). Similar results were also observed by Kumar et al. (2018) for days to maturity and plant height, as well as by Akhtar et al. (2011) and Babar et al. (2009). These values indicate the higher effect of the environment in the expression of these traits.
Low values of the GCV and PCV were recorded for days to booting (9.29 and 10.30), panicle length (8.89 and 9.52), fertility percentage (8.85 and 10.95), days to heading (9.96 and 10.67), and SPAD (4.75 and 6.26). Moderate GCVs but high of PCVs were recorded for leaf area (17.49 and 20.76) and effective tillers (19.78 and 22.97) indicating the high influence of the environment on the expression of these traits.
Low estimates of the GCV and PCV were exhibited by fertility percentage (8.85 and 10.95), panicle length (8.89 and 9.52), and chlorophyll content (4.75 and 6.26). Similarly, Prasad et al. (2017) and Srivastava et al. (2017) recorded low GCVs and PCVs for panicle length. These characteristics may be expressed by nonadditive genes and the possibility of improvement of yield through selection by employing recombination breeding (Srivastava et al., 2017). Panicle length is a valuable and important trait for yield improvement (Rasel et al., 2018).

Heritability (broad sense heritability)
Estimation of the genetic variability, which is crucial in determining breeding methods, was performed to determine the heritable and nonheritable portions of the concerned traits. Calculating the heritable portion of the overall observed variation is a prerequisite for the improvement of crop and can be determined by studying the components of variation such as the coefficients of genotypic and phenotypic variability, heritability, genetic advance, and genetic advance as a percentage of the mean. Broad sense heritability was calculated for all the traits evaluated in this study. Calculated values of heritability for these traits are presented in Table 6. The traits investigated in the present study expressed low to high broad sense heritability estimates ranging from 19.4% to 99.1%.
High values of broad sense heritability indicate that the traits being studied have the scope to improve through selection. These traits are less influenced by the environment and could be successfully transferred to the next generations if utilized in a hybridization program so that superior genotypic selection can be made through simple selection. Low heritability of traits indicates more influence of the environment in the phenotypic expression of that trait and hence, selection based on progeny or family testing should be done.

Genetic advance
Genetic advance as a percentage of the mean was found to be highest for straw yield (65.75%) followed by harvesting index (52.67%), filled grains per panicle (49.18%), grains per panicle (48.04%), thousand-grain weight (43.66%), flag leaf area (42.48%), grain yield (41.45%), effective tillers per m 2 (35.09%), flag leaf length (42.48%), leaf area (30.34%), plant height (30.31%), tillers per m 2 (25.69%), and days to maturity (21.76%). Similar high values of genetic advance as a percentage of the mean were previously recorded by Srivastava et al. (2017) for plant height, flag leaf length, number of grains per panicle, filled grains per panicle, test weight, and grain yield per plant. Medium genetic advance as a percentage of the mean values were found for days to booting (17.23%), panicle length (17.11%), days to heading (19.15%), and fertility percentage (14.75%). Low genetic advance as a percentage of the mean was found for SPAD reading (7.42%). High heritability coupled with high genetic advance as a percentage of the mean were recorded for days to maturity, number of tillers per m 2 , plant height, leaf area, flag leaf length, effective tillers per m 2 , grain yield, flag leaf area, test weight, panicle density, grains per panicle, filled grains per panicle, harvesting index, and straw yield.
High GCVs and PCVs, heritability, and genetic advance as a percentage of the mean were recorded for straw yield, harvesting index, filled grains per panicle, grains per panicle, thousandgrain weight, flag leaf area, grain yield, effective tillers per m 2 , and leaf area. These characteristics with high values of the GCV, PCV, and heritability accompanied by high genetic advance as a percentage of the mean might be transmitted to their progenies and therefore, phenotypic selection based on these characteristics would be effective. Panse (1957) stated that high values of heritability with low genetic advance indicated that the heritability was probably due to the effects of non-additive gene action. In general, the characteristics that show high heritability with high genetic advance are controlled by additive gene action (Patil & Lokesha, 2003) and can be improved through simple or progeny selection methods. Selection for the traits having high heritability coupled with high genetic advance is likely to accumulate more additive genes leading to further improvement of their performance (Johnson et al., 1955). The characteristics showing high heritability along with moderate or low genetic advance can be improved by intermating superior genotypes of a segregating population developed from combination breeding.

Cluster analysis
Cluster analysis of the fifteen rice genotypes in this study showed that the genotypes exhibited considerable genetic variability among themselves by occupying four clusters as shown in Table 6. These rice genotypes were grouped based on phenological, agro-morphological, and yield related traits, namely days to booting, number of effective tillers per m 2 , filled grains per panicle, flag leaf area, flag leaf length, fertility percentage, grains per panicle, grain yield, days to heading, harvesting index, leaf area, days to maturity, panicle length, plant height, SPAD reading, straw yield, test weight, and number of tillers per m 2 .
Cluster analysis showed that cluster I was comprised of five genotypes, cluster II consisted of four genotypes, cluster III was comprised of four genotypes, and cluster IV consisted of two genotypes (Figure 1). Cluster II genotypes were characterized as having the highest fertility, smallest palnicle lengths, and lowest thousandgrain weights. Cluster III genotypes had the lowest SPAD readings at the time of flowering, and the longest number of days for booting, anthesis, and harvesting. It was also characterized by having the longest panicle lengths, largest numbers of grains per panicle, largest numbers of filled grains per panicle, and highest straw yields but lowest harvesting indicies.
Cluster IV contained the genotypes with the highest SPAD readings at the time of flowering, largest leaf areas, flag leaf areas, and thousand-grain weights, in addition to the lowest flag leaf lengths, and lowest numbers of total and effective tillers per m 2 . Cluster I had the genotypes that were earliest in terms of booting, heading, and days to maturity, lowest leaf areas, lowest flag leaf areas, and lowest numbers of filled grains per panicle and grains per panicle. It was comprised of genotypes with the highest grain yields, harvesting indicies, and numbers of effective tillers per panicle.

Conclusions
The analysis of variance study showed the presence of adequate phenotypic variability among the tested genotypes for all the traits in this study. From the mean performance analysis, it was found that Radha 11 can be selected for higher grain yield. Radha 13 can be selected for grains per panicle, tallest plant height, and leaf area. Ghaiya 2 can be selected for the largest number of panicles per m 2 and effective tillers per m 2 . Hardinath-1 can be selected for maximum flag leaf area, chlorophyll content, and earlier heading and maturity, while the local landrace (Manabahu) also had a high values for panicle length, straw yield, flag leaf length, and maximum days to heading, booting, and maturity, and can be selected for these traits. The PCV was greater than the GCV for all the traits being studied, meaning that some degree of environmental influence was present in the expression of these traits. A small difference between the CGV and PCV suggests less influence of the environment on the expression of the traits. High GCVs and PCVs, higher broad sense heritability, and higher GAM of the traits 100.00 straw yield, filled grains per panicle, harvesting index, grains per panicle, flag leaf area, grain yield, thousand-grain weight, effective tillers per m 2 , and leaf area indicates that there is less influence of the environment in the expression of these traits. A higher proportion of variability is heritable and improvement can be made for these traits in the next generation through direct selection. High broad sense heritability and genetic advance for these traits indicate the presence of additive gene interaction in the expression of quantitative traits and showed more opportunity for selection of these traits for improvements in yield. Cluster analysis showed the presence of considerable genetic variability among the genotypes for further improvement of the genotypes.