An assessment of the relationship between body weight and body measurements of indigenous Nigeria chickens (Gallus gallus domesticus) using path coefficient analysis

S S A Egena, A T Ijaiya and R Kolawole

Department of Animal Production, Federal University of Technology, P.M.B 65, Minna, Niger State, Nigeria.
acheneje.egena@futminna.edu.ng

Abstract

Path analysis was used to carry out an assessment of the relationship between body weight and body measurements of indigenous Nigeria chickens. The chickens used for the experiment were raised under extensive system of management. The objectives of the study were to deduce the components that best describe the relationship between body weight and body measurements and also come up with optimized regression models that could be used in predicting body weight in the chickens. The parameters measured were body weight (BW), body length (BL), heart girth (HG), wing length (WL), shank length (SL) and shank thickness (ST). Pair wise correlation between body weight and the linear body measurements were positive and highly significant (r= 0.166 - 0.697 in males and 0.411 - 0.714 in females; p<0.01). Path analysis showed that body length had the greatest direct effect on body weight (path coefficient = 0.417 in males; 0.428 in females, respectively) while the least direct effect was observed for SL in males (path coefficient = 0.100), and BG in females (path coefficient = 0.033). The highest percent direct contribution in males was observed to be 17.39% (BL) followed by 4.75% (WL) while the least was observed to be 1.00% (SL). Equally, the highest percent direct contribution in females was observed to be 18.32% (BL) followed by 5.06% (WL) while the least was observed to be 0.11% (BG). The highest percent combined contribution in males and females was 0.07 and 0.06% (both by BL via WL). The optimized linear regression model with a coefficient of determination (R²) value of 0.58 (males) and 0.66 (females) included forecast indices such as BL, BG, WL and ST (males) and BL, WL, SL and ST (females). The regression models generated could be used in predicting the body weight of indigenous Nigerian chickens and also for selection purposes.

Key words: correlation, direct and indirect effects, path analysis, percent contribution, regression

Introduction

The Nigerian indigenous chicken breeds like other indigenous chickens of Africa represent a great repository of genetic variation whose potential has unfortunately being left untapped and unexploited. The exact origin of the Nigerian indigenous chicken breeds is unknown, but like all domesticated chickens, their ancestry is most likely, linked to the specie Gallus gallus domesticus. The indigenous chicken breeds of Africa play a very vital role in the socio-economic life of their keepers, providing them with readily available meat, egg (Kitalyi 1998; Zaman et al 2004) and manure. They also provide a source of income to rural and semi-urban resource poor individuals (Ssewanyana et al 2003). The meat of the indigenous chickens are said to be more palatable than the meat obtained from the exotic broilers breeds (Kolawole 2010). This might be because of exposure of the birds to natural sources of nourishment composed basically of grasses, insects, worms and the like. The birds have become well adapted to the different ecology where they are found due largely to natural selection. One drawback of the indigenous chickens is their compact and slow rate of growth. Their growth performance however could be improved through careful selection for body weight and or associated traits such as body weight gain, feed: gain ratio and linear body measurements. 

Growth involves increase in size and changes in functional capabilities of the various tissues and organs of animals (Ojedapo et al 2012). Growth in farm animals is however very complex and not a straightforward affair. This is because there are combinations of factors which may favour or disfavour it. One of such factors is the interrelationship between body weight and its associated traits like body dimensions. While this interrelationship has been found useful in predicting body weight in poultry (Monsi 1992; Gueye 1998), it has not been able to point out the actual causal of growth. There is the need therefore to make use of other methods of evaluating this relationship. One method that easily comes to mind is path coefficient analysis. This is due to its ability to ascertain the direct impact of one variable or factor on another variable or factor, and its capacity to split the correlation coefficient into direct effect (path coefficient), and indirect effects (effects exerted via other independent variables). Path analysis therefore measures the relative importance of causal factors which provides information leading to effective selection during improvement programmes (Gelalcha and Hanchinal 2013). 

The objective of this study is to assess the relationship between body weight and body measurements in indigenous Nigerian chickens using path coefficient analysis. This will help in determining the nature of character association in male and female indigenous Nigerian chickens.

Materials and Methods

Experimental animals and their management 

Seven hundred and fifty indigenous Nigerian chickens of both sexes (543 males and 207 females) were randomly selected in certain smallholders farms in villages located within the three administrative zones of Niger State in the north central part of Nigeria (Figure 1). Niger state is located in the sub-humid savannah area of Nigeria around 30˚ 2′ North and 11˚ 3′ East. The state has a land area of 80,000 square kilometres with maximum altitude at its highest point of 1475 m above sea level. The state experiences distinct dry and wet seasons with annual rainfall varying from 1100 mm in the north to 1600 mm in the south. The dry season lasts for 6 to 7 months (October to April) in the northern part of the state, and 4 to 5 months (November to March) in the southern part. The maximum temperature which does not exceed 39˚C is experienced between March and June, while the minimal temperature (as low as 21˚C) is usually experienced between December and January. The animals were extensively managed. Hence they mostly subsist on available household wastes and crop residues.

Figure 1: Map of Nigeria showing the study area.

Traits measured 

The traits measured include body weight and five linear body measurements. The body parts measured were body length (BL), measured as the distance from the tip of the beak, through the body trunk to the tail; body girth (BG), measured as the circumference of the breast region; wing length (WL), measured as the length of the wing from the scapula joints to the last digit of the wing and shank length (SL), measured as the length of the tarso-metatarsus from the hock joint to the metatarsal pad. Shank thickness (ST) was measured as the diameter of the tarso-metatarsus just below the spur. Body weight of individual birds was measured using a mechanical hanging balance of 2.5 kg. The body parts were measures using tape rule (cm) while shank thickness was determined (mm) using a pair of vernier calliper. The metric measurements were as described by Fayeye et al (2006). The measurements were carried out by the same person in order to avoid between individual variations. 

Statistical analysis 

Means, standard deviation and coefficients of variation of body weight and body measurements of animals adjusted for sex effects were computed using Microsoft Excel. The initial values of the parameters measured were transformed to generate the standardized version from the unstandardized variables using the means and standard deviations as described by Akintunde (2012). The standardized data was then subjected to regression and bivariate correlation analysis using SPSS (2001). The standardized partial regression coefficients called direct path coefficients were calculated thus: 

σX₁/σY = ‘P₁’, the path coefficient from X₁ to Y.

σX₂/σY =  ‘P₂’, the path coefficient from X₂ to Y.

σX₃/σY =  ‘P₃’, the path coefficient from X₃ to Y.

σX₄/σY =  ‘P₄’, the path coefficient from X₄ to Y.

σX₅/σY =  ‘P₅’, the path coefficient from X₅ to Y. 

Where Y is the effect and X_1, X₂, X₃. X₄ and X₅ are the causes. The indirect contributions of X_1,  X₂, X_3, X₄ and X₅ to Y were worked out as follows: 

Y₁ = P₁ + P₂RX₁X₂ + P₃RX₁X₃ + P₄RX₁X₄ + P₅RX₁X₅

Y₂ = P₁RX₁X₂ + P₂ + P₃RX₂X₃ + P₄RX₂X₄ + P₅RX₂X₅

Y₃ = P₁RX₁X₃ + P₂RX₂X₃ + P₃ + P₄RX₃X₄ + P₅RX₃X₅

Y₄ = P₁RX₁X₄ + P₂RX₂X₄ + P₃RX₃X₄ + P₄ + P₅RX₄X₅

Y₅ = P₁RX₁X₅ + P₂RX₂X₅ + P₃RX₃X₅ + P₄RX₄X₅ + P₅                                                    

Where R = correlation coefficient between the variables. The equations illustrate the splitting process for a 5 factor variables with one effect variable Y. The multiple linear regression model adopted for the studies was: 

Y = a + b₁X₁ +b₂X₂ + b₃X₃ + ---------------- + bpXp 

Where Y = the dependent or endogenous variable (body weight), a = the intercept, b = the regression coefficients and Xs are the independent or exogenous variables (BL, BG, WL, SL, ST).

Results

Morphometric traits 

The result of the descriptive statistics of body weight and body measurement traits of indigenous Nigerian chickens is presented in Table 1. Sexual dimorphism was in favour of male indigenous Nigerian chickens for BW, BL, BG, WL, SL and ST. Shank length varied the most in male indigenous chickens, followed by BW and ST. In the female indigenous chickens, BW showed the most variation, followed by ST and SL. The least variation was observed for BG in both the male and female indigenous Nigerian chickens.  

Pair-wise correlation 

The coefficients of correlation between BW and body measurements of indigenous Nigerian chickens are presented in Table 2. Correlation between BW and the body measurements were all positive and highly significant. The highest correlation was between BL and WL in the males (r = 0.719) and, BW and BL in the females (r = 0.714).

Direct and indirect effects 

The direct and indirect effect of morphological measurements on BW in male and female indigenous Nigerian chickens is presented in Table 3a and 3b. Body length made the greatest direct contribution to body weight of male indigenous Nigerian chickens (0.417) while the lowest direct effect was made by SL (0.100). In the female indigenous Nigerian chickens, BL made the greatest contribution to BW while BG made the least contribution. When combined, the indirect effects influencing BW were observed to be larger than the direct effects.

Percentage contribution of parameters 

Body length in males and females (17.39; 18.32%) made the greatest percent contribution to the body weight of indigenous Nigerian chickens (Table 4). The lowest percent contributions were made by SL (male; 1.00%) and BG (female; 0.11%). The greatest combined percent contribution in males was 0.07% made by BL via WL, while in the female, the greatest was 0.06% made by BL via WL.

Establishment of preliminary regression and optimized equations 

The following equations with their coefficients of determination (R²) were obtained from simple regression between BW and body measurements: 

Y = -0.000004 + 0.417BL + 0.117BG + 0.218WL + 0.010SL + 0.192ST …. i (male indigenous Nigerian chicken; R² = 0.58). 

Y = -0.0000008 + 0.428BL + 0.033BG + 0.225WL + 0.208SL + 0.131ST…. ii (female indigenous Nigerian chicken; R² = 0.66). 

After the removal of the redundant variables from the initial regression equations, the optimized but much simplified equation models with their coefficient of determination were: 

Y = 0.000004 + 0.418BL + 0.118BG + 0.217WL + 0.192ST …… i (male indigenous Nigerian chicken; R² = 0.58). 

Y = 0.0000009 + 0.424BL + 0.238WL + 0.216SL + 0.135ST………. ii (female indigenous Nigerian chicken; R² = 0.66).

Discussion

The sexual differences observed in the chickens is consistent with earlier reports (Deeb and Cahaner 2001; Zaky and Amin 2007; Yakubu and Salako 2009). Godonou (2002), Dossou (2005) and Youssao et al. (2010) also reported sexual differences in indigenous chicken breeds of Benin republic. The high variation observed for SL (male) and BW (female) means that these traits could be selected for subsequent genetic improvement. This is because genetic variation between and within breeds of animals are potential raw materials for genetic improvement. The high variability in these traits might not be unconnected with environmental factors acting on the animal. The average BW, BL and SL observed in the present study is higher than those reported by Sonaiya (2003) and Yakubu and Salako (2009). The higher values suggest that the birds were mostly of the Fulani ecotype indigenous Nigerian chickens known to have higher body weight and associated traits. The average body weight however falls within the range of 1.35-2.5 Kg reported for Wareng chicken of Indonesia (Hasnelly and Armayanti 2006).  

The positive and strong nature of the correlation between BW and body measurement traits means that BW could be estimated from body measurements. This is because growth in animals could be evaluated from the component parts of the animal (Wolanski et al 2006). This means that an improvement in the body measurements will invariable lead to a corresponding improvement in the BW of the indigenous Nigerian chickens especially if the correlation is positive as was observed in the present study. Similar high correlation coefficients between BW and body measurements have been reported in indigenous chickens of Senegal (Gueye et al 1998), Jinghai yellow chicken (Yang et al 2006) and in Gaga chicken of Indonesia (Sri Rachma et al 2013).  

The direct effects (path coefficient) of BL, BG, WL and ST of male indigenous Nigerian chickens were significant while SL was not significant (Table 3a). The total value of the indirect effect of SL was however large. The insignificant nature of SL inferred that the correlation between BW and the trait was largely due to indirect effect. This indirect effect was via BL. Body weight could be estimated therefore in male indigenous Nigerian chickens using BL, BG, WL and ST. Body girth was observed to have insignificant direct effect while the other traits all had significant direct effect in female indigenous Nigerian chickens. The total value of indirect effect for BG was however large. The large indirect effect was obtained via BL indicating that the significant correlation observed between BW and BG was principally via BL. Since BL, WL, SL and ST have significant direct effects on BW, it means they could be used in the estimation of BW in female indigenous Nigerian chickens. The higher direct effect of BL (male and female) observed in the present study however contradicts the findings of Yakubu and Salako (2009) who reported thigh circumference (male) and comb height (female) as having the highest direct effect in a population of indigenous chickens sampled in Nasarawa state, Nigeria. Path analysis provides a comprehensive means of determining factors affecting BW in indigenous Nigerian chickens. The results obtained from the analysis will provide information which could be used for selection purposes and ultimately, during animal improvement programmes. This is because correlation alone cannot provide the exact contributions made by the growth attributes to the overall body weight of the indigenous chickens. The use of path analysis to compute the degree of linkage between BW and body measurements has been reported in American Bronze turkeys (Mendes et al 2005). Body length will play a major role in any genetic improvement programme targeted at the indigenous Nigerian chicken by virtue of its great contribution (directly or indirectly) to the overall BW of the chickens. Selecting and improving BL will impact positively on the BW of the indigenous Nigerian chicken. 

The direct effects of SL (in males) and BG (in females) were not significant and hence were removed from the initial equations. This is in order to obtain an optimized equation that could be utilized in estimating the live weight of indigenous Nigerian chickens. The removal of variables that are redundant or not significant in order to obtain optimized equations has been reported by Yakubu and Salako (2009). The removal of the redundant variables did not lead to any change in the R² value of both equations. The optimized equations indicated that combined the non-redundant variables represent 58 and 66% of the variation of body weight of indigenous Nigerian chickens. Hence, they could be used in estimating the body weight of indigenous Nigerian chickens.

Conclusion

• Phenotypic correlation showed that BW correlated positively and significantly with all of the traits studied in both sexes of indigenous Nigerian chickens.
  
  
• The results of path analysis however revealed that SL and BG did not contributed directly to BW in male and female indigenous Nigerian chickens respectively.
  
  
• The implication is that body weight of indigenous Nigerian chickens could be estimated accurately using body measurements such as BL, BG, WL, and ST in males, and BL, WL, SL and ST in females.
  
  
• Selecting and improving these traits will impact positively on the BW of indigenous Nigerian chickens.

References

Akintunde A S 2012: Path analysis step by step using Excel. Journal of Technical Science and Technologies 1(1): 9-15.

Deeb N and Cahaner A 2001: Genotype-by-environment interaction with broiler genotypes differing in growth rate. 1. The effects of high ambient temperature and naked-neck genotype on lines differing in genetic background. Poultry Science 80: 541-548.     

Dossou F 2005: Caracterisation phenotypique des types genetiques de populations locales de volailles de l’espece Gallus gallus dans la Commune d’Abomey-Calavi. Memoire de fin de formation a l’EPAC, UAC, Abomey-Calavi. Pp: 49. 

Fayeye T R, Ayorinde K L, Ojo V and Adesina O M 2006: Frequency and influence of some major genes on body weight and body size parameters of Nigerian local chickens. Livestock research for rural development, 18, Article #3 http://www.lrrd.org/lrrd18/3/faye18037.htm  

Gelalcha S and Hanchinal R R 2013: Correlation and path analysis in yield and yield  components in spring bread wheat (Triticum aestivum L.) genotypes under irrigated condition in southern India. African Journal of Agricultural Research 8(24): 3186-3192. DOI: 10.5897/AJAR2012.6965   

Godonou Y D 2002: Le systeme de production du poulet local des elevages suivis par le Programme d’Appui au Developpement de l’Aviculture Villageoise (PADAV) dans la region de Ouake: production, commercialisation et possibilite d’amelioration. Theses Ing. Agro. FSA. UAC. Benin. Pp: 87. 

Gueye E F, Ndiaye A and Branckaert R D S 1998: Prediction of body weight on the basis of body measurements in mature indigenous chickens in Senegal. Livestock Research for Rural Development, 10, Article #3  http://www.lrrd.org/lrrd10/3/sene103.htm

Hasnelly A and Armayanti R 2006: Performans ayam Merawang betina dewasa berdasarkan karakter kualitatif dan ukuran ukuran tubuh sebagai bibit. In: Lokakarya Nasional Inovasi Teknologi dalam Mendukung Usaha Ternak Unggas Berdaya saing, held in 2006. Puslitbangnak, Bogor. Indonesia. Pp: 95-74. 

Kitalyi A J 1998: Village production systems in rural Africa. Household food security and gender issue, FAO. http://www.fao.org/docrep/003/w8989e/w8989E00.HTM 

Kolawole  R 2010: Frequency and effect of spur gene on metric parameters in the Nigerian local chicken in a southern guinea savannah area of Nigeria. B.Tech. Thesis, Department of Animal Production, Federal University of Technology Minna, Niger state, Nigeria. 

Mendes M, Karabayir A and Pala A 2005: Path analysis of the relationship between various body measures and live weight of American Bronze turkeys under three different lighting programs. Tarim Bilimleri Dergisi 11: 184-188.

Monsi A 1992: Appraisal of interrelationships among live measurements at different ages in meat type chickens. Nigerian Journal of Animal Production 19(1&2): 15-24

Ojedapo L O, Amao S R, Ameen S A, Adedeji T.A, Ogundipe R.I and Ige A.O 2012: Prediction of body weight and other linear body measurement of two commercial layer strain chickens. Asian Journal of Animal Sciences, 6: 13-22. DOI: 10.3923/ajas.2012.13.22

Sri Rachma A B, Hiroshi H, Muh. Ihsan A.D, Lellah R and Kusumandari I P 2013: Study of body dimension of Gaga’ chicken, germ plasm of local chicken from south sulawesi-Indonesia. International Journal of Plant, Animal and Environmental Studies 3(4): 204-209.

Ssewanyana E, Onyait A O, Ogwal J, Mukasa B, Nsamba P and Masaba J 2003: Characteristics of rural chicken in apac and kumi districts of Uganda. Uganda Journal of Agricultural Sciences 8: 159-164.

Sonaiya E B 2003: Producing local livestock-improving rural livelihoods. In: Nigerian Livestock: A goldmine for economic growth and food security. Taiwo, A.A., Raji, A.M., Ogbonna, J.U. and Adebowale, E.A. (eds). Proceedings of 28th Annual Conference of the Nigerian Society for Animal Production (NSAP). Ibadan (Nigeria), 16-20 March 2003. Pp: 458- 465. 

SPSS 2001: Statistical Package for Social Sciences. SPSS Inc., 444 Michigan Avenue, Chicago, IL60611. 

Wolanski N J, Renema R A, Robinson F E, Carney V L and Fanchert B I 2006: Relationship between chick conformation and quality measures with early growth traits in males of eight selected pure or commercial broiler breeder strains. Poultry Science 85: 1490-1497.

Yakubu A and Salako A E 2009: Path coefficient analysis of body weight and morphological traits of Nigerian indigenous chickens. Egyptian Poultry Science 29(3): 837-850. 

Yang Y, Mekki D M, Lu S J, Yu J H, Wang L Y, Wang J Y, Xie K Z and Dai G J 2006: Canonical correlation analysis of body weight, body measurements and carcass characteristics of Jinghai Yellow chicken. Journal of Animal and Veterinary Advances 5: 980-984. http://medwelljournals.com/abstract/?doi=javaa.2006.980.984 

Youssao I A K, Tobada P C, Koutinhouin B G, Dahouda M, Idrissou N D, Bonou G A, Tougan U P, Ahounou S, Yapi-Gnaore V, Kayang B, Rognon X and Tixier-Boichard M 2010: Phenotypic characterisation and molecular polymorphism of indigenous poultry populations of the species Gallus gallus of savannah and forest ecotypes of Benin. African Journal of Biotechnology 9 (3): 369-381.  DOI: 10.5897/AJB09.1220

Zaky H I and Amin E M 2007: Estimates of genetic parameters for body weight and body measurements in Bronze turkeys (Baladi) by using animal model. Egyptian Poultry Science 27: 151-164.

Zaman M A, Sorensen P and Howlider M A R 2004: Egg production performances of a breed and three crossbreeds under scavenging system of management.  Livestock Research for Rural Development, 12, Article # 1 http://www.lrrd.org/lrrd16/8/zama16060.htm

Received 30 January 2014; Accepted 4 February 2014; Published 1 March 2014

Go to top