BREEDING POTENTIAL OF RICE GENOTYPES IN TWO AGROCLIMATIC CONDITIONS OF SULAIMANI - KURDISTAN REGION - IRAQ

This study was aimed to investigate the genetic variability of 26 rice genotypes and evaluation at two locations in Sulaimani governorate, Gaba and Chawtan which were completely different in their environmental condition during the season of 2019. The performances of the genotypes were analyzed at both locations as well as the average of both. Simple coefficients of correlation were used to assess the grain yield components and their relationships. Path analysis was used to determine the direct and indirect effects of such components on grain yield plant -1 . The genotypes were grouped based on the agro-morphological features using cluster analysis. Almost all of the traits at both locations and the average of both locations showed highly significant differences; Chawtan outperformed Gaba location for the traits no. of tillers plant -1 , no. of panicle plant -1 , 1000-grain weight, and plant -1 grain yield, whereas higher values were found for plant height and no. of grains panicle -1 for Gaba location. At both locations, there was a highly significant and positive association between the number of tillers and panicles in plant -1 and grain yield plant -1 . The number of panicle plants -1 had the most favorable direct effects on grain yield at both locations. At both locations, the highest positive indirect effect on grain output was provided by the number of tillers plant -1 via the number of panicle plant -1 . Based on the agro-morphological features, the rice genotypes were grouped into 5 clusters at both locations. At both locations, Cluster V gave better values regarding grain yield.


INTRODUCTION
The development of rice genotypes with greater production potential is the goal of Iraqi breeders of the Oryza sativa L. plant. Numerous varieties of rice have been developed and adapted because to the country's different agro-ecosystems (28). The average production per unit area in Iraq is too low comparing the other countries due to the decreasing in soil fertility as results of cultivating of wheat after the rice and also due to the high water consumption during the growing season which is not available in all of the cultivation area especially in summer (2,11,12). Increasing yield per unit area is one method for increasing rice production; another is to cultivate additional land. To increase grain production, the genetic improvement of rice must be strengthened. In order to create an effective selection method, it was essential to comprehend the relationship between yield and the characteristics that contribute to it. Multivariate analysis is one of the approaches that is most commonly used to assess genetic variability by breeders when determining improved rice varieties. For high production levels to be sustained, genetic diversity is essential (28). The most important decision a plant breeder makes is which parents to use in a hybridization strategy. Utilizing genetic divergence to choose parents for an effective hybridization and breeding program makes sense (29). According to (24), identifying the elements crucial for population-level genetic diversity may help in the selection of diverse parents for hybridization programs. In order to identify and measure diversity in a variety of rice using agro-morphological traits, several researchers have used multivariate analysis (16,21). A plant breeder must have a thorough awareness of the phenotypic and genotypic interactions of multiple economic aspects in order to choose and breed different varieties and lines of rice with increased yield potential (3,18). The strength and direction of the relationship between yield and its component traits are revealed through correlation analysis of yield and those traits. Twenty different varieties of rice were examined by Yong-xiang et al. (31) using 9 different traits. It was discovered that rice grain yield plant -1 and the number of grain panicles -1 were positively connected. Sabesan et al. (22) investigated 54 rice varieties and found that plant height and the number of plant -1 tillers were positively correlated with grain yield plant -1 . According to Hairmansis et al. (8), plant height had a negative effect on grain yield plant -1 , while 1000-grain weight had less of an effect (8). Florence et al. (7) examined 32 genotypes originating from various geographical areas. 32 genotypes were grouped into four groups via cluster analysis. 75 genotypes were examined by Pervaiz et al. (17). Using cluster analysis, 75 genotypes were divided into four main groups that distinguished between tall, late-maturing, thin aromatic types, and early non-aromatic genotypes. It is challenging to study rice genotypes in the Iraqi Kurdistan region due to a lack of available information and previous research on rice production in the area. Currently, rice yields at the Kurdistan region are low and not profitable due to a lack of improved varieties, poor agronomic practices, and the presence of pests and diseases. Improving rice productivity in Iraq, which is distinguished by extremely different agro-climates and various growing conditions, depends significantly on the presence of genetic diversity. The objectives of this study were to determine the degree of diversity and identify potential genotypes for rice hybridization based on agro-morphological traits and also to identify rice genotypes that have both high mean grain yield and stable yield performance across different environments for the region. Additionally, determine the degree of association of the yield and the other traits in each environment through the correlation analysis and partitioning of the correlation coefficient to determine direct and indirect effects through alternate pathways of various traits and grain yield through path analysis. Another goal of the current investigation is to group the genotypes into distinct clusters to use in further study.

MATERIALS AND METHOD Plant Materials
Twenty-six different rice genotypes were collected throughout Iraq for this study, these genotypes were named according to the region from which they were originally sourced as well as the local name (Table 1).  Table 2 and the water analyses are given in Table 3.

Statistical analysis
The data were statistically analyzed according to the methods of analysis of variance as a general test using Randomized Complete Block Design (RCBD) with three replicates; a combined analysis of variance across locations was also conducted for the studied characters. All possible comparisons among the means were carried out by using the Least Significant Difference (L.S.D) test at significant levels of 0.05 and 0.01. The correlation coefficients were calculated to determine the degree of association of characters with yield and also among the yield components themselves in each environment. Phenotypic correlations were computed by using the formula given by (24). The path coefficient analysis was carried out through the solution of the following equations as suggested by (24) through Analysis of Moment Structures (AMOS) Ver. 18 Software.
The hierarchical cluster analysis based on Euclidean Distance and Unweighted Pairgroup Linkage (UPGMA) was also performed to cluster the rice genotype's relatedness based on agro-morphological traits using the IBM SPSS program, Ver. 19 (26).

RESULTS AND DISCUSSION
Performance of the genotypes:Highly significant differences were detected among the genotypes for all of the studied traits at Gaba and Chawtan locations, according to the mean squares of the variance analysis which indicates the presence of huge differences among the studied genotypes (Table 4). This demonstrates the genotypes' significant potential for utilization as a genetic source for breeding reasons (14). The mean squares of the combined analysis of the variances were presented in Table 5, which indicated the presence of highly significant mean squares due to the locations for the traits plant height, no. of tillers plant -1 , no. of panicles plant -1 , no. of grains panicle -1 , 1000-grain weight, and grain yield plant -1 , while the mean squares due to the genotypes were highly significant for all of the studied traits. Concerning the (Genotypes×Environments) interaction they were indicated to be highly significant for all of the traits except for grain width which was significant only. These findings emphasize the importance of the genotypes, the environment, and the interaction effect of Genotypes×Environments. The interaction effects have significant impacts and cannot be ignored when studying rice growth in multilocations experiments.  The mean performance of the genotypes at Gaba location was presented in     The effect of locations on the studied traits were presented in Table 9

Correlation analysis
The correlation analysis among the studied traits at Gaba location was presented in Table  ( 10). No. of tillers plant -1 recorded highly significant and positive correlations with no. of panicles plant -1 and grain yield plant -1 with (0.977**) and (0.903**) respectively, whilst it correlated significant and negatively (-0.414*) with no. of grains panicle -1 . A highly significant and positive correlation was detected between no. of panicles plant -1 and grain yield plant -1 (0.897**), while it correlated significantly and negatively with no. of grains plant -1 (-0.458*), but correlated significantly and positively with grain length (0.432*). A significant and positive association was detected between grain length and grain yield plant -1 (0.486*) and the association between grain width with 1000grain weight was highly significant and positive (0.820**). Concerning the grain yield plant -1 , as noticed in the results selection genotypes with high values due to the traits no. of tillers plant -1 , no. of panicles plant -1 and grain length have a great impact on increasing the grain yield plant -1 in rice genotypes at Gaba location. Madhavilatha et al. (13) reported that no. of grains panicle -1 and 1000grain weights were positively associated with grain yield plant -1 . Khan et al. (10) found correlation of plant height and no. of tillers plant -1 was positive. Grain yield plant -1 had positive correlation with no. of grains panicle -1 . Sabesan et al. (22) observed that grain yield plant -1 was positively associated with plant height. Hairmansis et al. (8) reported that plant height had negative effect on grain yield plant -1 while 1000-grain weight had negligible effect on grain yield.   Table 11 shows the correlation analysis among the studied traits at Chawtan location. Plant height recorded a highly significant and negative correlation with grain length (-0.617**), whilst showing a significant and positive correlation with 1000-grain weight (0.467*). The association between no. of tillers plant -1 and no. of panicles plant -1 as well as grain yield plant -1 were positive and highly significant with (0.961**) and (0.844**) respectively. No. of panicles plant -1 recorded highly significant and positive correlation with grain yield plant -1 reached (0.828**). The correlation between no. of grains panicle -1 and grain length was highly significant and negative (-0.616**). Grain length recorded a significant and positive correlation with 1000grain weight (-0.429*). The correlation between grain width and 1000-grain weight was highly significant and positive (0.660**). Breeding rice genotypes have high values for the traits no. of tillers plant -1 , and no. of panicles plant -1 resulting in an increase in the weight of grain produced by the plant at Chawtan location. Rashid et al. (20) found that plant height showed significant positive correlation with panicle length, highly significant positive correlation with no. of tillers plant -1 , highly significant negatively correlation with no. of grains panicle -1 , significant and negative correlation with 1000grain weight and highly significant negative correlation with grain yield plant -1 , also they recorded that No. of tillers plant -1 showed nonsignificant positive correlation with no. of grains panicle -1 , non-significant negative correlation with 1000-seed weight and nonsignificant positive correlation with grain yield plant -1 . Iftekharuddaula et al. (9) were found that panicle length and 1000-grain weight were positively associated with rice grain yield plant -1 and plant height. Rasheed et al. (19) revealed by correlation that no. of tillers plant -1 was positively associated with grain yield plant -1 . Surek and Beser (27) reported grain yield plant -1 significantly correlated with no. grains panicle -1 . Borbora et al. (5) found that 1000-grain weight were highly associated with grain yield plant -1 .

Path coefficient analysis
The path coefficient analysis for grain yield plant -1 was presented in Table 12 at Gaba location, which indicated that, the maximum positive direct effect on grain yield recorded by no. of panicles plant -1 (0.537) followed by no. of tillers plant -1 and no. of grains panicle -1 with 0.497 and 0.392 respectively, while the highest positive indirect effect was 0.525 recorded by no. of tillers plant -1 via no. of panicles plant -1 and followed by 0.232 for grain length via no. of panicles plant -1 . The results of residual effect (R=0.019) revealed that 98% of the grain yield plant -1 was contributed by the traits studied in this experiment. The role of other independent variables which had not been included in this experiment was expected to influence grain yield only by 2%. This result indicates the adequacy of the traits that were included in this study.    Table 14 represents the Agglomeration schedule of the average linkage (Between Groups) at both locations, within cluster I, G 2 and G 4 had the greatest diversity with a distance coefficient of 90.028, which was likely due to the participation of different parents in their crosses or might have been the result of a mutation; whereas G 5 was most similar to G 6 (12.607) and the variation between them may be the result of their different origins. Within cluster II, G 7 and G 10 had most distance (95.504), while G 11 has a slight difference from G 15 (16.150). Concerning Gaba location, Table 15 shows that Cluster II includes genotypes recorded better values for panicle length, grain width, and 1000-grain weight. Cluster III included genotypes which were superior for plant height; Cluster V includes genotypes which were superior for no. of tillers plant -1 , no. of panicle plant -1 , no. of grain panicle -1 , and grain yield plant -1 . Cluster I and IV dose not record any superiority in any of the traits.   (14.720) and the variation between them may be the result of their different origins. Within cluster III, G 2 and G 17 had most distance (164.029), while G 16 has a slight difference from G 18 (19.181). At Chawtan location, Table  15 shows that Cluster I included genotypes recorded better grain width. Cluster II included genotypes which were superior for 1000-grain weight; Cluster IV included genotypes which were superior for grain width. Cluster V includes genotypes which were superior in Plant height, no. of tillers plant -1 , no. of panicle plant -1 , no. of grain panicle -1 , Panicle length, and grain yield plant -1 . Cluster III dose not record any superiority in any of the traits. Cluster analysis provided with a complete view of the variation present among the 26 rice genotypes and it might be use for the plant breeders for the genetic improvement of rice. According to many researchers results, there is genetic variation among rice genotypes, which could be grouped into a variety of unique clusters. Alamir (1) demonstrated the presence of genetic diversity among the investigated genotypes by grouping 36 low land rice genotypes with 12 morphological features into 7 clusters. Twenty irrigated lowland rice genotypes with 11 morphological characters were separated into 4 clusters by (4), who also demonstrated significant genetic variation within the studied genotypes. Twenty-four accessions of irrigated rice were divided into five clusters by (21). Worede et al. (30) divided Twenty-four upland rice varieties into 2 clusters based on 17 morpho-agronomic parameters. Thirty-two early maturing rice genotypes were divided into 3 clusters by (23). Thirty-nine genotypes of rice grown under irrigation were divided into 6 separate clusters by (6).