Livestock Research for Rural Development 27 (9) 2015 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Phenotypic characterization and weight estimation from linear body traits of West African Dwarf goats reared in the transitional zone of Ghana

P T Birteeb and R Lomo

Department of Animal Science, Faculty of Agriculture, University for Development Studies, P. O. Box TL 1882, Tamale, Ghana.
bpetert2000@yahoo.com

Abstract

The knowledge of weight estimation and phenotypic characterization is paramount in goat production and marketing practices hence this research was undertaken to characterize West African Dwarf (WAD) goats phenotypically and develop a model for predicting the weight of WAD goats in the Transitional zone of Ghana. Body weight (BW), linear body measurements (heart girth-HG, body length-BL, height at withers-WH, rump height-RH and tail length-TL), and coat colour pattern were measured on 325 goats from the National Goat Breeding station at Kintampo and subjected to chi-square, general linear model, correlation, principal component factor and regression analyses.

 

The results revealed that goats having two colours dominated (61.2%), and brown colour was very common as a solid or in combination with other colours. The overall averages of 60.60.6, 51.70.4, 46.50.4, 48.60.4, 9.30.2cm, and 27.60.5 kg were obtained for HG, BL, WH, RH, TL and BW respectively. Male and female goats were generally similar in most of the traits except WH (P = 0.03). Coat colour did not have any significant influence (P > 0.05) on any trait. Phenotypic correlation between pairs of body traits ranged from low to moderate (0.09 – 0.71), with highest correlation being between BW and BL. Two principal component (PC) factors were extracted to explain body composition:  the first PC consisting of HG, BL, HW, and RH explained 58.2% of total variation; the second PC (TL) accounted for 20.3%. The first PC indicated the general body size of the goats. The combination of HG and BL yielded the best accuracy (64.5%) in weight estimation, while prediction from the two PC’s was only moderate (62.3%). The explained body variations can be exploited by the goat breeding station and individual researchers in selection programs to improve goat breeding in the transitional zone in the country.

Key words: body weight, characteristics, principal component


Introduction

The characterization of local genetic resources depends on the knowledge of the variation of morphological traits, which have played a very fundamental role in classification of livestock based on size and shape (Yakubu et al 2010a).   Size and conformation are important characteristics in meat animals especially ruminants. Traditionally, animals are usually assessed visually, which is a subjective method of judgment (Abanikannda et al 2002). Body size and shape measured objectively could improve selection for growth by enabling the breeder to recognize early maturing and late maturing animals of different sizes. Measurement of various body conformations are of value in judging quantitative characteristics of meat animals and are also helpful in developing suitable selection criteria. Body measurements and live weights taken on live animals have been used extensively for a variety of reasons both in experimental work and in selection practices (Lawrence and Fowler 2002). Body measurements have been used to evaluate breed performance and to characterize animals. In addition, they have been used as a means of selecting replacement animals (Sowande and Sobola 2008).

 

The knowledge of morphological body measurements of goats could be exploited to aid adequate management and production of goats. Afolayan et al (2006) stated that the accuracy of functions used to predict live weight or growth characteristics from live animal measurements is of immense financial contribution to livestock production enterprise. Several researchers have shown that body measurements provide great convenience for the prediction of body weight without weighbridges or scales (Birteeb and Ozoje 2012; Okpeku et al 2011; Yakubu 2009; Pesmen and Yardimci 2008; Adeyinka and Mohammed 2006; Afolayan et al 2006).

 

The West African Dwarf (WAD) goat is a common and popular breed in Ghana. WAD is a major component of the indigenous livestock genetic resources especially in the rural communities of Ghana. This breed is well adapted to, and produces and reproduces under the local environmental conditions although its productivity is generally low compared to its counterparts in other parts of the world. Uncontrolled breeding coupled with attempts by breeders and farmers to improve the performance of the indigenous African breeds through the introduction of exotic animals and crossbreeding practices are gradually leading to the erosion and complete masking of important survival traits, such as disease resistance associated with indigenous livestock as well as the extinction of certain breeds (Gizaw et al 2011). The characterization of African small ruminant populations will play a major role in the maintenance of these autochthonous genetic resources as the basis for future improvement at both the production and the genetic levels (Birteeb et al 2012).

 

Knowing the morphological measurements of WAD goats will be very useful for good animal management, including understanding medication doses, adjusting feed supply, monitoring growth and choosing replacement males and females (Slippers et al 2000). Knowledge on method of weight estimation will also be very useful in goat production since most farmers do not have weighing scales for measuring liveweight. Therefore, this study was conducted to characterize the WAD goat phenotypically and to determine suitable model(s) for liveweight prediction from body measurements. 


Materials and Methods

Study area and climate

 

The study was conducted at the National Goat Breeding Station at Kintampo in Ghana. Kintampo is located on latitude 803’ N and longitude 0143’W (Ghana Districts 2014). The area falls within the transitional or derived savannah ecological zone, having a mixture of tall trees, shrubs and grasses as the vegetation type.  The mean annual rainfall is about 1300 mm.  About 69% and 31% of the rains occur in the wet (April-October) and dry (November-March) seasons, respectively.  The mean monthly temperature is about 27oC while relative humidity values are 73.8% at 09.00 h and 56.7% at 15.00 h (Karikari and Blasu 2009).

 

Management of experimental animals

 

The animals were kept on-station where they grazed on cultivated pastures made up of Cynodon plectostachyus, Cajanus cajan, Stylosanthes spp, Panicum maximum, Panicum minimum and Ficus spp established by the station. They were supplemented with feed made up of wheat bran, maize chaff and cotton seed cake. The animals were regularly vaccinated and dewormed against diseases and pests. The animals were housed at night in roofed pens with concrete floors. Clean water together with mineral supplement/salt licks were provided ad-libitum after grazing.

 

Data collection

 

A total of 325 (29 males and 296 females) matured WAD goats were randomly sampled from the breeding station. The parameters measured were; Rump height (RH), Height at withers (HW), Body length (BL), Wither height (WH), Tail length (TL), Heart girth (HG), live body weight (BW), Coat colour, animal’s age and sex.

 

The live body weight (Kg) of the animals were measured using Griffith electronic weighing scale and linear body measurements (LBM’s) taken using a tailor’s measuring tape graduated in centimeters. The procedures for measuring the (LGM’s) are outlined below as described by Birteeb et al (2012).

 

Height at Withers (WH): The distance from the surface of a platform on which the animal stands to the withers.

Heart Girth (HG) or Chest circumference: It is a circumferential measure taken around the chest just behind the front legs and withers.

Body Length (BL): It is the distance from scapula (at shoulder) to the pin bone.

Rump Height (RH): The distance from the surface of a platform to the rump.

Tailed Length (TL): This is a measurement taken from the base of the tail to the tip.

Visual assessments were made in identifying the qualitative traits like coat colour and sex.

The ages of animals were traced using the date of birth as individual animals were identified by tags.

Photo 1. The measured linear body traits
Statistical analysis

 

The data set was analyzed using SPSS version 17. The chi-square goodness of fit test was done for distribution of sex, age and colour. The fixed effects of sex, age and colour on linear body measurements were tested using general linear model (GLM), the model given as:

Where Yijk = individual observation of each body traits;

μ = overall mean;

Si = fixed effect of ith sex (i = male, female);

Aj = fixed effect of jth age (j = 1 year old, 2 years old, and ≥ 3 years old);

Ck = fixed effect of kth colour pattern (k = unicolour, bicolour, tricolour);

(SA)ij = interaction effect of ith sex and jth age;

(SC)ik = interaction effect of ith sex and kth colour pattern;

(AC)jk = interaction effect of jth age and kth colour pattern;

(SAC)ijk = interaction effect of sex, age and colour pattern;

εijk = random error associated with each record (ε ~ N(0, σ2))

 

Bivariate phenotypic correlations among the morphological measurements were determined using Pearson’s correlation coefficient. The body composition of the goats was then analyzed by employing Principal Component Factor Analysis (PCA). The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s Test of Sphericity were employed to determine the validity and accuracy of the data set for factor analysis. The appropriateness of the factor analysis was further tested by examining the communalities. Body weight was then regressed on linear body measurements and the extracted principal components (PC) to test their use in predicting body weight of goats.


Results and discussion

Distribution of the qualitative traits of WAD goat

 

There was unequal representation of the different colours (Table 1). The animals had about ten coat colour patterns with the dominant being pied brown and white coat colour combination. The tricolour pattern comprised the combination of any three of brown, black, white and grey. Clearly there is a large variability in coat colour pattern. The dominant occurrence of brown colour (as solid or in combination with other colours) may suggest the association of this colour with a trait that is selected for at the station. The study agreed with that of Hassen et al (2012)  in which coat colour patterns observed were brown, white, black with  the most frequent coat colour being  white with spots (20.7%), followed by brown with patches (17.5%) and brown (15.4%). In Ethiopia, there were varied coloration patterns amongst the goat populations sampled with spotted pattern (36.1%) followed by patchy (32.4%) and plain (30.4%) of various colors. Black colored animals, however, are believed to have superior adaptation to seasonal cold weather or cold nights as the dark pigment helps them to warm up earlier than goats with other coat colors (Robertshaw 2006).

Table 1. Distribution of the qualitative traits of WAD goats

Traits

Class levels

Frequency

Percentage %

Chi-square

p

Sex

Male

29

8.9

219

<0.001

Female

296

91.1

Age

One year

59

18.2

226

<0.001

Two years

235

72.3

Three years

31

9.2

Coat colour

Unicolour

Brown

50

15.4

358

<0.001

Black

15

4.6

Grey

4

1.2

Bicolour*

Brown/White

78

24

Brown/Black

70

21.5

Black/White

26

8.0

Grey/White

12

3.7

Grey/Black

9

2.8

Grey/Brown

4

1.2

Tricolour

57

17.5

* Bicolours: i.e. occurrence of two colours in either patchy or spotted form.

Quantitative traits of West African Dwarf goat

 

Goats in the present study were similar in linear body dimensions to the indigenous goats of Benin (Dossa et al 2007), but weighed heavier than all goats in Okpeku et al (2011) although the Red Sokoto had higher wither height and chest girth. However, the goats in the present study were quite small and very light when compared to Saanen goats (Pesmen and Yardimci 2008).  The fixed effects of sex, coat colour and age on the body traits are presented in Table 2. Among the traits only height at withers was influenced by sex, where the nanny goats were longer than billy goat counterparts. All other traits were similar for both sexes (Table 2). This finding disagrees with the report in goats (Okpeku et al 2011) and sheep (Birteeb et al 2012) in which males were superior to females in all the body measurements.  Age affected all measured traits except tail length (Table 2). [T3] The present finding is similar to Yakubu (2009) where only HW was significantly influenced by sex. The ten colour patterns were categorized into uni-colour (solid coat colour), bi-colour (two coat colours) and tri-colour (three combined colours) to determine colour effect on the traits. Coat colour had no influence on any body measurements in the flock under study.

Table 2. Least square means (±S.E) of live body weight (kg) and linear body traits (cm) of WAD goat

Body
trait

Overall
mean

Sex

Age (years)

Coat colour

Male

Female

p.

One

Two

Three

p.

Unicolour

Bicolour

Tricolour

p.

HG

60.6±0.6

60.5±0.5

60.7±1.2

0.573

55.3±0.9c

61.3±1.1b

66.7±1.1a

<0.001

61.6±1.0

59.5±0.6

60.4±1.5

0.863

BL

51.7±0.4

51.7±0.4

51.6±0.9

0.597

47.1±0.7c

53.1±0.7b

55.1±0.8a

<0.001

51.0±0.7

51.4±0.4

52.9±1.1

0.579

WH

46.5±0.4

45.8±0.4b

47.5±0.9a

0.030

44.5±0.6c

46.9±0.7b

48.5±0.8a

<0.001

47.1±0.7

46.2±0.4

46.0±0.0

0.821

RH

48.6±0.4

48.8±0.3

48.4±0.9

0.511

46.3±0.6c

48.3±0.6b

52.2±0.8a

<0.001

48.7±0.7

48.4±0.4

48.8±1.0

0.414

TL

9.3±0.2

9.1±0.2

9.7±0.4

0.209

9.0±0.3

9.6±0.3

9.2±0.3

0.549

9.3±0.3

9.4±0.2

9.4±0.4

0.966

BW

27.6±0.5

28.5±0.4

26.8±1.0

0.802

20.1±0.7c

28.6±0.7b

36.3±0.9a

<0.001

28.2±0.8

26.8±0.4

28.5±1.2

0.944

a, b, c Means with different superscripts in a row differ significantly for sex and age

Phenotypic correlations of body traits

 

Most correlations among traits were only moderate (0.19 – 0.58) with the highest between BW and HG and BL. Pesmen and Yardimci (2008) also recorded the highest correlation between liveweight and HG. The correlations in this study were generally lower than those reported in earlier studies (Okpeku et al 2011; Pesmen and Yardimci 2008; Khan et al 2006).

Table 3. Phenotypic correlations among body measurements of WAD goat

HG

BL

WH

RH

TL

BW

HG

-

BL

0.459**

-

WH

0.347**

0.485**

-

RH

0.472**

0.506**

0.579**

-

TL

0.191**

0.147**

0.237**

0.226**

-

BW

0.666**

0.706**

0.462**

0.542**

0.094ns

-

** p< 0.01 ; ns correlation not significant

Body composition of WAD goat

 

The KMO test recorded a significantly high value (0.694) which confirmed the appropriateness of the procedure and accuracy of the results. It implied that true factors existed and the data were factorable. However, this KMO value was lower than 0.91 for Ghanaian sheep (Birteeb et al 2012), 0.92 for sheep in northern Nigeria (Yakubu et al 2011) and 0.85 for Uda sheep (Salako 2006). The wider differences indicates that the present goat data was not as appropriate as the sheep data (Birteeb et al 2012; Yakubu et al 2011; Salako 2006). The Bartlett’s Test was also highly significant (P <0.001), indicating enough support for the validity of the factor analysis of the data set. After a Varimax rotation of the component matrix, two underlying principal component (PC) factors were extracted to explain a total of 78.6% of the generalized variance (Table 4).

Table 4. Communalities and variance explained by PC factors

Traits

PC Factors

Communalities

1

2

Heart girth

0.82

-0.12

0.69

Body length

0.75

0.40

0.73

Height at withers

0.81

-0.44

0.85

Rump height

0.88

-0.26

0.84

Tail length

0.49

0.76

0.83

Eigenvalue

2.91

1.02

Explained variance (%)

58.2

20.3

Cumulative variance (%)

58.2

78.6

The two PC’s implied that variation in the phenotypic characteristics of the West African dwarf goat can by explained in two directions. The first PC comprised of heart girth, body length, height at withers and rump height which seemed to indicate the general body size of the goat, and explain a variation of 58.2% (Table 4). The second PC represented the tail and accounted for 20.3% of the total variation. The generally low correlation of the tail length with other traits, coupled with the high loading of the tail length as the main trait on the second PC may mean that the tail length has little or no coordination with the other traits in WAD goats. Much higher variations were earlier reported in WAD and Red Sokoto goats (Okpeku et al 2011). The communalities indicate the performance of each trait in the presence of other traits in explaining the generalized variation of the body morphology. The communalities where generally high (Table 4) though lower than those in WAD and Red Sokoto goats (Okpeku et al 2011). Birteeb et al (2012) and Okpeku et at (2011) reported that the first factor explained the maximum variation. The association of heart girth, body length, height at wither and rump height with the first PC is in accordance with Okpeku et al (2011). This may suggest that animals that are tall will equally have long body lengths and wide heart girths, hence will generally be large. In Ghanaian sheep, it was observed from the first PC that an animal large for one trait was generally large for all traits (Birteeb et al 2012). The PC information could be used to select animals based on a group of variables rather than a single trait. Yakubu (2010) stated that factor analysis could be exploited in breeding and selection programmes to acquire highly coordinated animal bodies using few body components.

 

Live weight estimation

Table 5. Parameter estimates, prediction accuracies and probabilities of regression models

Variable (s)

Parameter estimates

Adj. R2 (%)

p

β0

β1

β2

β3

HG

-15.1

0.71

-

-

44.2

<0.001

BL

-23.7

1.00

-

-

49.8

<0.001

HW

-7.19

0.76

-

-

21.1

<0.001

RH

-15.5

0.89

-

-

29.2

<0.001

TL

24.4

0.39

-

-

00.6

0.091

HG+BL

-37.1

0.46

0.72

-

64.5

<0.001

HG+HW

-29.2

0.61

0.43

-

50.1

<0.001

HG+RH

-29.7

0.56

0.48

-

50.7

<0.001

BL+HW

-30.0

0.89

0.26

-

51.5

<0.001

BL+RH

-34.4

0.82

0.41

-

54.2

<0.001

HW+RH

-22.1

0.37

0.68

-

32.2

<0.001

HG+BL+RH

-41.2

0.42

0.66

0.20

65.4

<0.001

PC1+PC2

27.9

4.73

-1.54

-

62.3

<0.001

Adj. = Adjusted; PC = Principal component

Body weight was regressed on each of the body dimensions in simple and multiple linear regressions. All the regression models were highly significant (p<0.001) except only when tail length was used as a predictor (Table 5). The resultant coefficients of determination (Adj. R2) were generally low as the highest accuracy was obtained from three predictors. The use of the two extracted PC factors from Table 4 yielded moderate prediction accuracy (Table 5). The low prediction accuracies could have been the consequence of the moderate correlations between BW and each of the traits (Table 3). The prediction accuracies in this study were far lower than 89% reported by Okpeku et al (2011) in the indigenous goats in southern Nigeria. In Ethiopia, the use of HG yielded 80.7% and 9.46% while the use of BL yielded 53.2% and 29.0% for two goat populations respectively (Hassen et al 2012). Again, the prediction accuracies of simple linear regressions herein were comparable to the range of 25.3% to 63.4% for older WAD sheep but lower than most values obtained for WALL sheep (Birteeb and Ozoje 2012). The prediction accuracies of the multiple models in this study were far lower than the range of 71% to 95% obtained in Saanen goats (Pesmen and Yardimci 2008). The moderate R2 values of multiple regression models herein supports the suggestion by Adeyinka and Mohammed (2006) that the use of other linear measurements along with heart girth could improve the predictability of the resultant equations. However, further additions of traits to a model need to be evaluated in light of the marginal gain in prediction accuracy. The moderate accuracy from the use of PC’s suggested that PC factor analysis is useful not only in describing animal composition but also in weight prediction. It also means that the use of PC’s describes animals (WAD goats) better than individual body measurements.


Conclusion


Acknowledgement

The authors wish to thank and appreciate the management and staff of National Goat Breeding station at Kintampo for availing the animals for this study, and also being instrumental in the data collection. God bless you.


References

Abanikannda O T F, Leigh A O and Olutogun O 2002 Linear measurements based discriminant classification of Zebu cattle in Lagos State. In: Fanimo AO and Olanite JA  (eds). Contributory role of animal production in national development. Proceedings of the 7th Annual Conference of Animal Science Association of Nigeria, held at the University of Agriculture, Abeokuta, September 16-19, 2002. pp 355-356.

Adeyinka I A and Mohammed I D 2006 Accuracy of Body Weight Prediction in Nigerian Red Sokoto Goats raised in North Eastern Nigeria using Linear Body Measurement. Pakistan Journal of Biological Sciences. 9: 2828-2830.

Afolayan R A, Adeyinka I A and Lakpini C A M 2006 The estimation of live weight from body measurements in Yankasa sheep. Czech Journal of Animal Science. 51: 343-348.

Birteeb P T and Ozoje M O 2012 Prediction of live body weight from linear body measurement of West African long-legged and West African dwarf sheep in Northern Ghana. Online Journal of animal and Feed Research. 2: 425-434.

Birteeb P T, Olusola S P, Yakubu A, Adeke M A and Ozoje M O 2012 Multivariate characteristics of the phenotypic traits of Djallonke and Sahel sheep in Northern Ghana. Tropical Animal Health and Production. Do1 10 1007/s11250-012-4.

Dossa L H, Wollny C and Gauly M 2007 Spatial variation in goat populations from Benin as revealed by multivariate analysis of morphological traits. Small Ruminant Research.73: 150-159.

Ghana districts 2014 Kintampo Municipal Assemble, Ghanadistricts.com. http://www.ghanadistricts.com/districts/?news&r=10&_=37, Retrieved on 27th July, 2014.

Gizaw S, Komen H, Hanote O, van Arendonk J A M, Kemp S, Haile A, Mwai O and Dessie T 2011 Characterization and conservation of indigenous sheep genetic resources: A practical framework for developing countries. ILRI Research Report No. 27. Nairobi, Kenya, ILRI

Hassen H, Baum M, Rischkowsky B and Tibbo M 2012 Phenotypic characteristics of Ethiopia indigenous goat populations. African Journal of Biotechnology. 11: 13838-13846.

Karikari P K and Blasu E Y 2009 Influence of nutritional flushing prior mating on the performance of West African dwarf goats mated in rainy season. Livestock Research for Rural Development. 21 (7). http://www.lrrd.org/lrrd21/7/kari21103.htm

Khan H,  Muhammed F, Ahmed R, Nawaz, Rahimullah and Zubair M 2006 Relationship of body measurement of body weight with linear body measurements in goats. Journal of Agricultural and Biological Science. 1: 51-54.

Lawrence T L J and Fowler V R 2002 Growth of farm animals. 2ndEdn. CABI Publshing, Oxon, UK.  374p.

Okpeku M, Yakubu A, Olusolas P, Ozoje O M, Ikeobi C O, Adebambo O A and Imumorin I G 2011 Application of multivariate principal component analysis to morphological characterization of indigenous goats in Southern Nigeria. Acta Agriculturea Slovenica. 98:101-109.

Opong-Anane K 2006 Country pasture forage research profile, pp. 3-20.

Otoikhian C S O, Otoikhian A M, Akporhuarho O P and Isidahomen C 2008 Body measurement parameters as a function of assessing body weight in goats under on-farm research environment. African Journal of Agricultural. 4: 135-140.

Pesmen G and Yardimci M 2008 Estimating the live weight using some body measurements in Saanen goats.AchivaZootechnica. 11: 30-40.

Robertshaw D 2006 Mechanism for the control of respiratory evaporative heat loss in panting animal. Journal of Applied Physiology. 101: 664-668.

Salako A E 2006 Principal component factor analysis of the morphostructure of immature Uda sheep. International Journal of Morphology. 24: 571–574.

Sikosana J L N (2008) Goat production and management. In: Rooyen and Homann (eds.). Enhancing incomes and livelihoods through improved farmers’ practices on goat production and marketing: Proceedings of a workshop organized by the Goat Forum, 2–3 October 2007, Bulawayo, Zimbabwe. PO Box 776, Bulawayo, Zimbabwe: International Crops Research Institute for the Semi-Arid Tropics. 84 pp.

Slippers S, Lett B A and De Villiers J K 2000 Prediction of the body weight of Nguni goats. South African Journal of Animal Science. 30: 127-128.

Sowande O S and Sabola O S 2008 Body measurements of West African dwarf sheep as parameters for estimation of live weight. Tropical Animal Health and Production. 40: 433-439.

Yakubu A 2009 Fixing collinearity instability in the estimation of body weight from morpho-biometrical traits of West African Dwarf goats. Trakia Journal of Science. 7: 61-66.

Yakubu A 2010 Path coefficient and path analysis of body weight and biometric traits in Yankasa lambs. Slovakia Journal of Animal Science. 43: 17-25.

Yakubu A, Idahor K O, Haruna H S, Wheto M and Amusan S 2010a Multiple analysis of phenotypic differentiation in Bunaji and Sokoto Gudali cattle. Acta Agricuturae Slovenica. 96: 78-80.

Yakubu A, Salako A E and Abdullah A-R 2011 Varimax rotated principal component factor analysis of the zoometrical traits of Uda sheep. Archivos de Zootecnia. 60: 813–816.

Yakubu A, Salako A E, Imumorin I G, Ige A O and Akinyemi M O 2010b Discriminant analysis of morphometric difference in West African Dwarf and Red Sokoto goats. South African Journal of Animal Science. 40: 381-387. 


Received 21 June 2015; Accepted 8 August 2015; Published 1 September 2015

Go to top