Correlation and Path Coefficient Analysis for Agronomical Traits of Lowland Adapted Ethiopian Sorghum Genotypes [Sorghum bicolor (L.) Moench] Genotypes

TAFERE Mulualem¹, SENTAYEHU Alamrew², TAYE Tadesse³ and DAGNE Wegary⁴

Greener Journal of Agricultural Sciences, vol. 8, pp. 155-159, no. 8, 2018

^{1, 2}Jimma University, College of Agriculture and Veterinary Medicine, P.O.Box, 307, Jimma Ethiopia.

³Plant Breeder, Melkassa Agricultural Research Center, P.O.Box, 436 Adama, Ethiopia.

⁴Plant Breeder CIMMYT, ILRI Campus- P.O.Box, 5689 Addis Ababa, Ethiopia.

INTRODUCTION

Sorghum (Sorghum bicolor (L.)  Moench)  crop is genetically suited to hot and dry agro-ecologies with frequent drought, where it is difficult to grow other crops. In Ethiopian lowland areas sorghum is mainly grown for both food and feed (stover) purposes. Therefore, it can play a  vital  role  for  the  uplift  of  socio-economic status  of  the  farmers  in  the areas  through development  of  high  yielding  varieties  along  with reasonable  maturation date.

The study of associations among quantitative traits is important for assessing the feasibility of joint selection of two or more  traits and  hence for  evaluating  the effect  of  selection for secondary traits  on  genetic  gain  for  the  primary  trait  under consideration.  A positive genetic correlation between two desirable traits makes the job of the plant breeder easy for improving both traits simultaneously. The path coefficient analysis allows partitioning of correlation coefficient into direct and indirect contributions (effects) of  various  traits  towards  dependent  variable  and  thus helps  in  evaluating  the  cause-effect  relationship  as well as effective selection. Therefore, the present study is aimed to analyze and determine the traits having greater association with yield utilizing the correlation and path analysis for different traits in sorghum.

MATERIALS AND METHODS                                 

One-hundred ten early and medium maturing lowland adapted sorghum genotypes were used for the experiment. The trial was grown in alpha lattice design with two replications at two locations i.e. Meiso and Sheraro research sub-stations. Mieso is located 9°14′N, 40°45′E, and 1394 m.a.s.l. whereas Sheraro lies in 14.4N, 37.9 E, 1179 m.a.s.l. Both the areas are among the potential sorghum producing dry lowlands in the northern and eastern part of country.

Data to be collected

All agronomic data were collected on plot and plant bases using sorghum descriptors (IBPGR/ICRISAT, 1993). In each plot, randomly selected five plants were used to measure the following plant based characters.

Data analysis

Phenotypic (r_p) and genotypic (r_g) correlation coefficient

The correlation was estimated using the formula suggested by Miller et al. (1958):

Where,   Phenotypic correlation coefficient

           Phenotypic covariance between character x and y

Where, g_coxy= genotypic covariance between character x and y,

r_g = genotypic correlation coefficient,

= Genotypic variance of x character

= Genotypic variance of y character

= Phenotypic variance of x character

=Phenotypic variance of y character

Path coefficient analysis

The path coefficient analysis initially suggested by Wright (1921) and described by Dewey and Lu (1959) allows partitioning of correlation coefficient into direct and indirect contributions (effects) of various traits towards dependent variable and thus helps in assessing the cause-effect relationship as well as effective selection. Genetic correlations were further partitioned into direct and indirect effects using the path coefficient analyses following the method of Dewey and Lu (1959).

Where,mutual association between the independent character i (yield related trait) and dependent character, j (yield ) as measured by the genotypic correlation coefficient; is components of direct effect of the independent character (i) on the dependent character (j) as measured by the genotypic path coefficients; and  Σr_ikp_{kj =}summation of components of indirect effects of a given independent characters (i) on the given dependent character (j) via all other independent characters (k). Whereas the contribution of the remaining unknown characters are measured as the residual which is calculated as:

RESULTS AND DISCUSSION

Estimation of correlation coefficient of yield and yield related traits

The existence of variation alone in the population is not sufficient for improving desirable characters and hence, estimation of the extent and pattern of genetic variability existing in the available germplasm is essential to breeders. Breeders are also interested in the association that may exist between or among sorghum characters.

The  phenotypic  and  genotypic  correlation  coefficients  computed  among  yield  and  yield  related traits are presented in Table 1. Yield is a complex character which depends upon several component characters. Therefore, direct selection for yield is often not effective. Thus, it is essential to study the association of yield components with yield which is less influenced by environmental factors. Grain yield showed significant and positive phenotypic correlation with number of heads per plot (r=0.34**), panicle weight (r=0.19**) and hundred grain weight (r=0.21**). The strongest phenotypic association was observed between days to flowering and days to maturity (r= 0.53**) followed by hundred grain weight and plant height (r= 0.47**). On the other hand characters days to flowering, days to maturity and disease score showed negative and significant association with grain yield with the value (r= 0.18^**), (r= -0.17^**) and (r= -0.11^*) respectively.

At genotypic level grain yield showed positive and highly significant correlations with number of heads per plot (r=0.42**) and panicle weight (r=0.31**) (Table 1). The strongest genotypic association was observed between days to flowering and days to maturity (r= 0.71**) followed by plant height and hundred grain weight (r= 0.56**) and plant height and panicle length (r=0.42**). Generally, the values of genotypic correlation coefficients  were  higher  than  the  corresponding phenotypic correlation coefficients for most of the traits  suggesting  that  there  was  inherent relationship  between  these  traits.  This  is  in accordance  with  the  findings  of  Mahajan  et  al. (2011),  Ezeaku  and  Mohammed  (2006) and  (Alhassan  et  al.,  2008).

Table 1: Estimates of correlation coefficients at genotypic (above diagonal) and phenotypic (below diagonal) for yield and yield related traits among nine morpho-agronomic traits of sorghum genotypes evaluated at Meiso and Sheraro, 2016

 * P≤ 0.05; ** P≤ 0.01, ***P≤0.001, DTF= days to flowering, DTM=days to maturity, PHT=plant height, NPPP= number of panicles per plot, PAL=panicle length, PAW=panicle weight, HGW=hundred grain weight, DS=disease score, GYD=grain yield,

Path Coefficient Analysis 

Path  coefficient  analysis   provides  an  effective  means  of  finding  direct  and indirect  causes  of  association (Wright, 1921). In the present investigation, the path coefficient analysis was done with nine characters using estimates of direct and indirect effects of eight characters on grain yield based on phenotypic and genotypic correlation coefficients  (Table 2 and 3).

Phenotypic path coefficient analysis

The phenotypic direct and indirect effects of yield-related traits on grain yield of sorghum are presented in Table 2. Thus, days to maturity, plant height, number of heads per plot, panicle weight and hundred grain weight exerted positive phenotypic direct effect on grain yield. Whereas days to flowering, panicle length and disease score exerted negative direct effects on grain yield. However, days to flowering exhibited positive phenotypic correlation with grain yield due to their positive indirect effects through number of heads per plot, panicle weight and hundred grain weight. Similarly disease score exhibited positive phenotypic correlation with grain yield due to their positive indirect effects through days to flowering and days to maturity. The highest and positive phenotypic direct effects on grain yield were exhibited by number of panicles per plot (0.28), followed by hundred grain weight (0.17). Hence, these traits should be considered in further selection procedures for higher grain yield. The highest indirect effect belonged to number of heads per plot via panicle weight and hundred grain weight. Days to maturity had positive direct effect (0.08) on grain yield. However, it affects the yield negatively via plant height, number of heads per plot, panicle weight and hundred grain weight.  Although  the  number of panicles per plot,  panicle weight and hundred grain weight has statistically highly significant positive phenotypic direct effect on the yield, but it has negative indirect  effect  via  days to flowering  and  days to maturity. The positive phenotypic direct effect of plant  height  and  number of panicles per plot  on  grain  yield  is  in agreement  with  Mahajan  et  al.,  (2011).

Table 2: Estimates of phenotypic path analysis of the direct (bolded diagonal) and indirect (off-diagonal) effects of yield related traits on grain yield of sorghum genotypes evaluated at Meiso and Sheraro, 2016

DTF= days to flowering, DTM=days to maturity, PHT=plant height, NPPP= number of panicles per plot, PAL=panicle length, PAW=panicle weight, HGW=hundred grain weight, DS=disease score, GYD=grain yield

Genotypic path coefficient analysis  

The genotypic direct and indirect effects of yield-related traits on sorghum grain yield are presented in Table 3. Number  of  panicles  per  plot  exerted  positive  direct  effect  and  exhibited  positive significant  genotypic  correlation  with  grain  yield.  Days to maturity,  plant height, number of panicles  per plot, panicle length, panicle weight  and  hundred grain weight  showed  positive  genotypic  direct  effect  on  yield  and  also  had  positive correlation  with  grain  yield. These  traits  could  be  used  as  a  reliable  indicator  in  indirect selection for higher grain yield since their direct effect and association with grain yield were positive.

The  highest  genotypic  direct  effect  on  grain  yield  was  exerted  by  number of panicles  per plot (0.38) followed by days to maturity (0.18). The positive associations of these traits with grain yield were due to the positive indirect effects through other traits. Negative direct effects on grain yield were found for days to flowering (-0.21) and disease score    (-0.23), which is not agree with previously studied by Premlatha et al., (2006) also these traits exhibited negative correlation with grain yield. The  positive  genotypic  direct  effect  of  plant height  is  in  conformity  with  the  results  obtained  by Mahajan et al., (2011)  in their study on variability, correlation and path coefficient analysis in sorghum.

Table 3: Estimates of genotypic path analysis of the direct (bolded diagonal) and indirect (off-diagonal) effects of yield related traits on grain yield of sorghum genotypes evaluated at Meiso and Sheraro, 2016

DTF= days to flowering, DTM=days to maturity, PHT=plant height, NPPP= number of panicles per plot, PAL=panicle length, PAW=panicle weight, HGW=hundred grain weight, DS=disease score, GYD=grain yield

SUMMARY AND CONCLUSION

The relationship of different agronomic characters with each other and their relationship with yield is important. Grain yield showed significant and positive phenotypic correlation with number of heads per plot (r=0.34**), panicle weight (r=0.19**) and hundred grain weight (r=0.21**). the values of genotypic correlation coefficients  were  higher  than  the  corresponding phenotypic correlation coefficients for most of the traits  suggesting  that  there  was  inherent relationship  between  these  traits. Though, further evaluation of these and other genotypes of sorghum at more locations and over years is advisable to confirm the promising results observed in the present study.

In general, it may be concluded that the information from this study could be valuable for researchers and/or academicians who anticipate to know the direct and indirect effects of yield component traits in different varieties of sorghum.

ACKNOWLEDGMENTS

The author is thankful to the financial support from Ethiopian Institute of Agricultural Research and Jimma University. A special thanks to Dr. Taye Tadesse (national sorghum research coordinator), technical assistances and field workers at MARC, Meiso and Sheraro for the support rendered during field evaluation.

REFERENCES

Alhassan U, MY Yeye, DA Aba and SO Alabi (2008) Correlation and path  coefficient  analyses  for  agronomic  and  malting  quality traits  in  some  sorghum  [Sorghum  bicolor  (L.)  Moench] genotypes. J. Food Agric. Environ. 6(4): 285-288.

Dewey, D.R. and K.H. Lu, (1959). A correlation and path coefficient analysis of components of crested wheatgrass seed production. Agron. J., 51: 515-518.

Ezeaku IE and Mohammed SG (2006).  Character association and path analysis in grain sorghum. Afr. J. Biotechnol. 5(14): 1337-1340.

IBPGR/ICRISAT, (1993). Descriptors for Sorghum (Sorghum bicolor (L.) Moench). International Board for Plant genetic resources, Rome, Italy; International Crops Research Institute for Semi-Arid Tropics, Patancheru, India.

Mahajan, R.C., P.B. Wadikar, S.P. Pole and M.V. Dhuppe, (2011). Variability, correlation and path analysis studies in sorghum. Research Journal of Agricultural Sciences, 2(1): 101-103.

Miller, P.A., C. Williams, H.F. Robinson and R.E. Comstock. (1958). Estimates of genotypic and Environmental variances and co-variances in upland cotton and their implications in selection. Agric. J. 50:126 –131.

Premlatha, N., Kumaravadivel, N. and Veerabadhiran, P. (2006). Correlation and path analysis for yield and yield traits insorghum [Sorghum bicolor (L.) Moench]. Res.on Crops 7: 187-90.