Livestock Research for Rural Development 28 (8) 2016 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Predicting live weight of grasscutters using their linear body measurements

B A Hagan, J K Nyameasem, A Asafu-Adjaye and K A Daffour-Oduro

CSIR-Animal Research Institute, P. O. Box AH 20 Achimota, Accra
bernardatohagan@yahoo.com

Abstract

A study was undertaken to predict the live weight (LW) of domestically kept grasscutters from body length (BL), tail length (TL), heart girth (HG) and head length (HL) using linear, quadratic and cubic functions. Data on LW, BL, TL, HG and HL were taken from 80 grasscutters between the ages of 12 and 18 months old. Linear, quadratic and cubic functions were fitted with LW as independent variable and BL, TL, HG or HL as predictor variables using the Generalized Linear Model (GLM) procedure of SAS. The correlations between the LW and linear body measurements were also determined for male, female and pooled data (both males and females).

The correlation coefficients between LW and body measurements ranged from 0.33 to 0.76 for pooled data. The coefficients of determination (R2) for the prediction equations ranged from 8.2% to 81.4% for pooled data. Using the linear function, BL (54.3%) and HG (58.2%) were better predictors of LW than TL (10.8%) and HL (8.2%). The use of both BL and HG in a given function explained better the variation in LW than the use of just one body measurement. In all three functions, prediction equations involving males gave the best R2  values compared with those from females and pooled data.

Keywords: body length, coefficient of determination, correlation, head length, heart girth, tail length


Introduction

The grasscutter is an important source of animal protein in Ghana (Opara 2010) and has one of the most expensive meats in the West African sub-region (Okeke and Mogbo 2013). The meat is a delicacy for both the rural and urban inhabitants. The successful domestication of the grasscutter has created a vibrant grasscutter farming business for the young and old in Ghana.

Knowledge of the live weight of livestock including grasscutter is important, as it is required in determining their feed requirement, breeding management, correct administration of drugs and marketing of animals (Baffour-Awuah et al 2000; Thiruvenkandan 2005; Sowande et al 2010). Live weight and changes in live weight could also be used in determining responses to genetic selection (Touchberry and Lush 1950). The accepted method of measuring live weight of animals is the use of calibrated electronic or mechanical scales. However, in most rural communities and villages in Ghana, where majority of the grasscutter farmers are found, poverty rates are very high. It is estimated that about 30% of the populace, mostly rural inhabitants, live on less than $1.00 a day. In these communities, weighing scales are either non available or unaffordable. Also in cases where cheap scales are present, farmers complain of frequent breakdowns of these scales. This results in frequent purchasing of these scales making them eventually expensive. Most grasscutter farmers have therefore resorted to the use of visual appraisal in the marketing of their animals. This has led to either underpricing or overpricing of animals during their sale.

Apart from morphological body measurements being used as indicators of sexual dimorphism in animals (Lindenfors et al 2007), they can also be used in predicting the live weight of livestock. This will enable farmers who cannot afford expensive weighing scales or frequently purchasing scales to still be in a position to estimate the approximate weight of their animals for whatever husbandry practices they need the weight for using simple measuring tapes.

Linear body measurements have been used to predict live weight of several livestock species including goats (Fajemilehin and Salako 2008; Slippers et al 2000; Tsegaye et al 2013), sheep (Thys and Hardouin 1991; Baffour-Awuah et al 2000; Afolayan et al 2006), cattle (Ahulu and Kpesese 1995; Francis et al 2002; Lukuyu et al 2016), rabbits (Chineke 2005; Egena et al 2012) and pigs (Daffour-Oduro and Naazie 2010; Adeola et al 2013). Information is however, scanty with respect to predicting live weight of grasscutters using morphological traits of the animal (Annor et al 2011; Udeh and Okonta 2013). Using a smaller data set (9 animals), Ikpeze and Ebenebe (2004) reported an almost unitary correlation coefficient between body weight and body length of grasscutter whilst Udeh and Okonta (2013) also reported that tail length and heart girth were best predictors of live weight of 5-month old female grasscutters.

The objective of this study was to predict the live weight of domestically kept grasscutters from different linear body measurements using linear, quadratic and cubic functions.


Materials and methods

Study area and management of animals

The study was carried out at the Grasscutter Unit, Pokuase Research Station of the Council for Scientific and Industrial Research-Animal Research Institute (CSIR-ARI), Accra. The area lies within the transitional agro-ecological zone on latitude 5° 41´N and longitude 0° 17´W. The mean monthly temperature and rainfall range from 23.3 to 24.7°C and 12.6 to 204.6 mm respectively (Ghana Meteorological Services Department 2009).

Animals were housed in 3-tier metallic wire mesh cages with each unit measuring 60×60×40 cm in dimension. Each animal was kept in a cage except during the breeding season where a female is joined to a male for mating. All animals were identified with metallic ear tags with unique codes. Cages were exposed to natural light, ventilation and ambient temperature.

Feed and feeding

Animal were fed freshly cut Panicum maximum daily. Fresh water was given in concrete made troughs. Both feed and water were given ad libitum. Left over feed and water were changed the next day.

Data collection

Measurements of live weight (LW), body length (BL), tail length (TL), heart girth (HG) and head length (HL) were taken from 35 male and 45 female grasscutters between the ages of 12 and 18 months old. All the measurements were taken in the morning before the animals were fed and watered. Live weights were taken with a digital weighing scale (in kilograms) after restraining the animals with nylon net. The linear body measurements were defined as described by Annor et al (2011) and were taken with a measuring tape in centimetres.

Body length is the distance from the tip of the nose to the tip of the tail.

Tail length is the distance from the base to the tip of the tail.

Heart girth is the circumference of the chest directly below the forelimb.

Head length is the distance from the tip of the nose to the level of the 7th cervical vertebrae.

Statistical analysis

Data were analysed using the generalized linear model (GLM) procedure of SAS (SAS 2008). First, data were subjected to least square analysis using GLM Type III procedure on the fixed effect of sex:

Yij = m + Si + eij

Where Yij = the trait under consideration (LW, BL, TL, HG, HL)

m = overall mean

Si = fixed effect of the ith sex of animal (male, female)

eij = random error term associated with each observation

Differences between means were separated by probability of difference (PDIFF) procedure of SAS (SAS 2008) at P<0.05.

Live weight of grasscutters were regressed on BL, TL, HG and HL separately for males and females and pooled data (both sexes) and analysed as follows:

i. Linear function

Y = a + bx

Multiple regression of LW on BL and HG for males, females and mixed sex were also carried out.

ii. Quadratic function

Y = a + bx + cx2

iii. Cubic function

Y = a + bx + cx2 + dx3

Where Y = Live weight

a = constant of the regression equation

b, c, d = regression coefficients

x = body length, tail length, heart girth or head length

The quadratic and cubic functions were fitted to only BL and HG because they were the only morphological traits that were significant (P<0.05) after stepwise regression of all four traits (BL, TL, HG and HL).

Correlations among all the study traits were also estimated for males, females and pooled data (regardless of sex). Coefficients of determination (R2) values were used to evaluate the functions.


Results

Table 1 shows the descriptive statistics of LW and linear body traits of grasscutter. Both LW and HL showed appreciable spread around their means. Males were higher than females in all the traits studied though the differences were not significant (P>0.05) for TL and HL (Table 2).

Table 1. Descriptive statistics of live weight (in kg) and linear body measurements (in centimetres) of grasscutter

Trait

N

Mean ± SE

SD

Minimum

Maximum

Live weight (LW)

80

2.60 ± 0.07

0.60

1.28

4.28

Body length (BL)

80

57.6 ± 0.49

4.42

47.50

68.70

Tail length (TL)

80

16.4 ± 0.18

1.61

11.00

20.20

Heart girth (HG)

80

29.3 ± 0.26

2.37

24.00

34.10

Head length (HL)

80

10.6 ± 0.17

1.56

8.00

15.50

N – Sample size; SE – Standard error; SD – Standard deviation



Table 2. Mean (± SE) of live weight and linear body measurements of grasscutter as influenced by their sex

Sex

N

LW (kg)

BL (cm)

TL (cm)

HG (cm)

HL (cm)

Male

35

2.90 ± 0.09a

58.9 ± 0.72a

16.5 ± 0.27

30.1 ± 0.38a

10.6 ± 0.26

Female

45

2.36 ± 0.08b

56.5 ± 0.64b

16.3 ± 0.24

28.7 ± 0.34b

10.6 ± 0.23

abMeans within columns with different superscripts are different at P<0.05

Table 3 presents the phenotypic correlation among all the traits studied. Live weight of grasscutter was positive and significantly correlated with all the linear body measurements in all three categories of sexes except for the correlation between LW and TL in females. Generally, the correlations between LW and BL or HG were stronger for all categories of sexes than those between LW and TL or HL. Body length is positive and significantly correlated with TL, HG or HL for all groups except for the correlation between BL and HG in females (P>0.05). The correlations between TL and HG, HL and HG were all low and not significant (P>0.05). Moderately positive correlations existed between TL and HL though that for females was not significant (P>0.05).

Table 3. Correlation between live weight (kg) and linear body measurements (cm) of male, female and mixed sex grasscutter

Sex

BL

TL

HG

HL

Male
Female
Pooled

LW

0.74***
0.68***
0.74***

0.45***
0.18NS
0.33***

0.83***
0.60***
0.76***

0.33*
0.33*
0.29**

 

Male
Female
Pooled

BL

0.67***
0.49***
0.58***

0.41**
0.23NS
0.39***

0.51**
0.66***
0.55***

 

Male
Female
Pooled

TL

0.18NS
0.01NS
0.12NS

0.43**
0.24NS
0.31**

 

Male
Female
Pooled

HG

0.10NS
-0.06NS
0.01NS

*P<0.05; **P<0.01; ***P<0.001; NSP>0.05

Table 4 shows LW prediction from BL, TL, HG or HL using linear functions including multiple regression function with both BL and HG at the same time. Males had prediction equations with higher coefficients of determination (R2) than those of the females and mixed sexes for all the linear functions. The use of both BL and HG in predicting LW using linear functions yielded the highest R2 values. With the exception of models involving TL and HL, all the models under the linear functions were highly significant (P<0.01).

Table 4. Linear regression equations for estimating live weight (LW) from body length (BL), tail length (TL), heart girth (HG) and head length (HL) for male, female and mixed sex grasscutters

Sex

Prediction equation

R2 (%)

Model (p)

Body length

Male

LW = -2.78 + 0.10BL

55.1

<0.0001

Female

LW = -2.34 + 0.08BL

46.0

<0.0001

Pooled

LW = -3.20 + 0.10BL

54.3

<0.0001

 

Tail length

Male

LW = 0.26 + 0.16TL

20.3

0.0067

Female

LW = 1.42 + 0.06TL

3.4

0.2250

Pooled

LW= 0.58 + 0.12TL

10.8

0.0030

 

Heart girth

Male

LW = -3.08 + 0.20HG

68.4

<0.0001

Female

LW = -1.75 + 0.14HG

36.6

<0.0001

Pooled

LW = -3.11 + 0.19HG

58.2

<0.0001

 

Head length

Male

LW = 1.07 + 0.17HL

11.3

0.0489

Female

LW = 1.48 + 0.08HL

10.6

0.0288

Pooled

LW = 1.42 + 0.11HL

8.2

0.0101

 

Body Length and Heart girth

Male

LW = -5.35 + 0.65BL + 0.15HG

87.9

<0.0001

Female

LW = -4.79 + 0.07BL + 0.11HG

67.2

<0.0001

Pooled

LW = -5.68 + 0.07BL + 0.14HG

81.4

<0.0001

Table 5 shows LW prediction from either BL or HG or both using quadratic function for the various categories of sexes. In all cases, the prediction equations for males had higher R2 values than those for females and mixed sexes. Also, quadratic equations involving both BL and HG at the same time yielded higher R2 values than those involving only one linear measurement, that is, BL or HG. All the models under the quadratic function were highly significant (P<0.01).

Table 5. Quadratic regression equations for estimating live weight (LW) from body length (BL), heart girth (HG), and both traits for male, female and mixed sex grasscutters

Sex

Prediction equation

R2 (%)

Model (p)

Body length

Male

LW = 1.64 – 0.06BL + 0.001BL2

55.4

<0.0001

Female

LW = -2.83 + 0.10BL – 0.0001BL2

46.0

<0.0001

Pooled

LW = 4.70 – 0.17BL +0.002BL2

55.3

<0.0001

 

Heart girth

Male

LW = -12.96 + 0.88HG – 0.011HG2

70.4

<0.0001

Female

LW = -15.90 + 1.13HG – 0.017 HG2

38.5

<0.0001

Pooled

LW = -4.65 + 0.30HG – 0.002HG2

58.3

<0.0001

 

Body length and Heart girth

Male

LW = -11.51 + 0.06BL + 0.58HG – 0.0072HG2

88.7

<0.0001

Female

LW = -10.67 + 0.07BL + 0.53HG – 0.0072HG2

67.5

<0.0001

Pooled

LW = -5.63 + 0.07BL + 0.14HG + 0.0001HG2

81.0

<0.0001

 

Body length and Heart girth

Male

LW = -7.19 + 0.15HG + 0.13BL – 0.0005BL2

88.0

<0.0001

Female

LW = -10.57 + 0.11HG + 0.27BL – 0.0018BL2

67.6

<0.0001

Pooled

LW = -6.15 + 0.14HG + 0.09BL – 0.0001BL2

81.0

<0.0001

Prediction of LW from either BL or HG using cubic function for the various sex categories is presented in Table 6. The R2 values for females were lower than those of the males and mixed sexes. All the models under the cubic function were also highly significant (P<0.01).

Table 6. Cubic regression equations for estimating live weight (LW) from body length (BL) and heart girth (HG) in male, female and mixed sex grasscutters

Sex

Prediction equation

R2 (%)

Model p

Body length

Male

LW = -88.89 + 4.59BL – 0.08BL2 + 0.0004BL3

56.1

<0.0001

Female

LW = -65.33 + 3.46BL – 0.06BL2 + 0.0004BL3

46.5

<0.0001

Pooled

LW = -60.30 + 3.22BL – 0.06BL2 + 0.0003BL3

55.9

<0.0001

 

Hearth girth

Male

LW = -87.05 + 8.55HG – 0.27HG2 + 0.0030HG3

71.2

<0.0001

Female

LW = 156.82 – 16.77HG + 0.60HG2 – 0.0071HG3

40.1

<0.0001

Pooled

LW = -58.36 + 5.84HG – 0.19HG2 + 0.0021HG3

58.6

<0.0001


Discussion

Males outperformed female grasscutters in LW and all the linear body measurements. Other workers have also reported similar findings in this same species (Adu et al 2002; Ikpeze and Ebenebe 2004; Jayeaola et al 2009; Annor et al 2011). This confirms the general notion that sexual dimorphism in body weight and body measurements indeed exist in animals including grasscutters. In grasscutters, where sex determination is a major challenge (Annor et al 2009), farmers could use differences in their linear body measurements to determine the sexes of animals of similar ages especially animals above 1-day old. At birth, there is no significant difference in body weight and linear body measurements of male and female grasscutters (Annor et al 2011).

With the exception of HL, all the mean linear body measurements for males and female grasscutters were slightly higher than those reported by Annor et al (2011) for 180-day old grasscutters. The differences in the trait means could possibly be due to the age differences of the animals. On the other hand, the BL of both male and female grasscutters in this study were within the range of 42-58 cm reported by Schrage and Yewadan (1999) for animals over 12 months old.

The generally medium to high positive correlations between LW and linear body measurements in this study agree with Annor et al (2011) and Jayeola et al (2009) in grasscutters, Daffour-Oduro and Naazie (2010) in pigs and Baffour-Awuah et al (2000) and Afolayan et al (2006) in sheep. This suggests that body measurements could be used as predictors of LW of matured grasscutters. Body length and HG will however, be better predictors of LW than TL and HL due to their relatively high correlation coefficients (r) with LW. The high correlation coefficients between LW and body measurements in males compared to females indicate that these body measurements can predict LW better in males than their female counterparts. Daffour-Oduro and Naazie (2010) reported similar findings in 3 to 6 months old indigenous pigs in Ghana. The medium to high positive correlations among BL, TL and HL suggest that these traits could be used in predicting each other. Likewise, selection for one trait could lead to the corresponding improvement in the other traits and males are more likely to respond to this selection. Heart girth had low and non-significant (P>0.05) correlations with TL and HL but not BL. This is contrary to report by Annor et al (2011) in relatively younger grasscutters (1 to 180-day old) where high correlations existed between HG and TL, HL or BL. The difference in the findings could be attributed to the ages of the animals involved. Since HG measures muscle tissues and TL, HL and BL measure skeletal tissues, it is possible that these tissues could be growing at a similar rate in young animals compared to older ones hence the lower correlations between HG and both HL and TL in older grasscutters (12 to 18 months old).

The coefficients of determination (R2) were used as measures of the goodness of fit for the models used in predicting LW from linear body measurements. In all three functions, there were differences in the way the same traits predicted LW in different sex categories of grasscutters. In all cases, the R2 values were higher for equations involving males than those of females and mixed sex categories. Indeed, a similar trend was observed for the correlations between body measurements and LW. These therefore suggest that body measurements predict LW better in adult male grasscutters than their female counterparts. This finding collaborates those of Ozluturk et al (2006), Daffour-Oduro and Naazie (2010) and Tsegaye et al (2013) in cattle, pigs and goats respectively.

Among the linear functions, modeling only one trait at a time, HG explained best the variation in LW followed by BL, TL and HL in that order. For the quadratic and cubic functions, where only HG and BL were used, the predictive ability of HG was also higher than BL. These findings agree with those of Birteeb and Ozoje (2012) in sheep, Tsegaye et al (2013) in goats, Francis et al (2002) in cattle and Annor et al (2011) and Udeh and Okota (2013) in grasscutters. The R2 values increased when both BL and HG were used to estimate LW in both linear and quadratic functions. This is an indication that LW could be predicted better using both BL and HG in the same prediction equation. Several workers have also reported similar results indicating that adding additional independent variables to the prediction equation increases the goodness of fit (Baffour-Awuah et al 2000; Yakubu 2010; Tsegaye et al 2013). For more reliable estimates of LW of matured grasscutters, farmers should use prediction equations involving both HG and BL.

Generally, the quadratic functions explained much of the variations in LW of matured grasscutters compared to the linear and cubic functions. However, where the prediction equation involved the use of only one body measurement, the cubic functions were slightly ahead of the other functions. This is in agreement with results of Ojedapo et al (2012) in layer chicken where cubic functions gave the highest goodness of fit compared with linear and quadratic functions.

In summary, the prediction equation with the highest R2 value was the quadratic equation involving both BL and HG to predict LW in male grasscutters (LW = -11.51 + 0.06BL + 0.58HG – 0.0072HG2). However, it is recommended that farmers use much simpler but reliable prediction equations like the linear equation involving both BL and HG (LW = -5.68 + 0.07BL + 0.14HG).


Conclusion


Acknowledgement

We are grateful to all the members of the Grasscutter Unit at the Pokuase station of Animal Research Institute, Ghana and particularly Messrs Yram Atipoe, Noble Defoe, Bright Ababio and Peter Hafudza for their assistance in collecting data for this study.


References

Adeola A C, Oseni S O and Omitogun O G 2013 Morphological characterization of Indigenous and Crossbred pigs in rural and Peri-urban areas of South-western Nigeria. Animal Science 7(1): 6-12.

Adu E K, Wallace P A and Ocloo T O 2002 Efficacy of sex determination in the greater cane rat, Thryonomys swinderianus, Temminck. Tropical Animal Health and Production 34(1): 27-33.

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.

Ahunu B K and Kpesese K W 1995 A note on linear body measurements in estimating liveweight in N’dama and West African shorthorn cattle. Legon Agricultural Research and Extension Journal 4: 97-99.

Annor S Y, Adu E K, Donkor J, Otsyina H and Ahiaba J 2009 In: Grasscutter Production: A Handbook. Qualitype Ltd. Accra, Ghana, 109 pp.

Annor S Y, Ahunu B K, Aboagye G S, Boa-Amponsem K, Djang-Fordjour K T and Cassady J P 2011 The genetics of morphological traits in the grasscutter. Livestock Research for Rural Development. Volume 23, Article #167. Retrieved April 2, 2016, from http://www.lrrd.org/lrrd23/8/Anno23167.htm

Baffour-Awuah O, Ampofo E and Dodoo R 2000 Predicting the liveweight of sheep by using linear body measurements. Ghana Journal of Agricultural Science 33: 207-212.

Birteeb P T and Ozoje M O 2012 Prediction of live body weight from linear body measurements of West African long-legged and West African dwarf sheep in northern Ghana. Online Journal of Animal and Feed Research 2(5): 427-434.

Chineke C A 2005 The relationships among body weights and linear dimensions in rabbit breeds and crosses. Journal of Animal and Veterinary Advances 4: 775-784.

Daffour-Oduro K A and Naazie A 2010 Predictive abilities of two linear body measurements of liveweight of the Ashanti Black Pig. Ghanaian Journal of Animal Science 5(1): 49-59.

Egena S S A, Akpa G N, Aremu A and Alemede I C 2012 Predicting body weight of Rabbit from linear body measurements at various ages by genetic group, parity and sex. 10th World Rabbit Congress, September 3-6, Egypt, 19-23.

Fajemilehin O K S and Salako A E 2008 Body measurements characteristics of the West African Dwarf (WAD) goats in deciduous forest zone of southwestern Nigeria. African Journal Biotechnology 7: 2521-2526.

Francis J, Sibanda S, Hermansen J E and Kristensen T 2002 Estimating body weight of cattle using linear body measurements. Zimbabwe Veterinary Journal 33: 15-21.

Ghana Meteorological Services Department 2009 www.meteo-gov.gh

Ikpeze O O and Ebenebe C I 2004 Relationships between physical body traits of the grasscutter (Rodentia: Thryonomyidae) in Akpaka Forest Reserve, Onitsha. Animal Research International 1(3): 160-163.

Jayeola O A, Onadeko S A, Ademolu K O and Awofeso O M 2009 Prediction of body weight from body measurements in cane rat: Thryonomys swinderianus (Temminck, 1827). The Zoologist 7: 168-175.

Lindenfors P, Gittleman J L and Jones K E 2007 Sex, size and gender roles: evolutionary studies of sexual size dimorphism. In: Sexual size dimorphism in mammals. D.J. Fairbairn, W.U. Blanckenhorn and T. Szekely (Eds.), Oxford University Press, Oxford, Pp. 19-26.

Lukuyu M N, Gibson J P, Savage D B, Duncan A J, Mujibi F D N and Okeyo A M 2016 Use of body linear measurements to estimate liveweight of crossbred dairy cattle in smallholder farms in Kenya. SpringerPlus 5: 63. http://doi.org/10.1186/s40064-016-1698-3

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 strains chickens. Asian Journal of Animal Science 6: 13-22.

Okeke J J and Mogbo T C 2013 Comparative study of growth performance of grasscutter fed on diverse foodstuff in captivity. International Journal of Advanced Biological Research 3(1): 85-89.

Opara M N 2010 The grasscutter I: A livestock of tomorrow. Research Journal of Forestry 4: 119-135.

Ozlururk A, Kopuzlu S, Guler O and Yanar M 2006 Determination of linear regression models for estimation of body weights of Eastern Anatolian Red cattle. Atatürk Üniv. Ziraat Fak. Derg. 37(2): 169-175.

Statistical Analysis System (SAS) 2008 User’s Guide. SAS/STAT® 9.2, Cary, NC: SAS Institute Inc.

Schrage R and Yéwadan L T 1999 In: Raising Grasscutters, Deutsche Gesellschaft fϋr Technische Zusammenarbeit (GTZ) GmbH, Eschborn, Germany, 99 pp.

Slippers S C, Letty B A and De Villiers J F 2000 Predicting the body weight of Nguni goats. South African Journal of Animal Science 30 (Suppl. 1): 127-128.

Sowande O S, Oyewale B F and Iyasere O S 2010 Age and sex dependent regression models for predicting the live weight of West African Dwarf goat from body measurements. Tropical Animal Health and Production 42: 969-975.

Thiruvenkandan A K 2005 Determination of best fitted regression model for estimation of body weight in Kanni Adu kids under farmer’s management systems. Livestock Research for Rural Development. Volume 17, Article #85 Retrieved April 4, 2016 from http://www.lrrd.org//lrrd17/7/thir17085.htm.

Thys E and Hardouin J 1991 Prediction of sheep body weight in markets in the Far North Cameroon. Livestock Research for Rural Development. Volume 3, Article #1. Retrieved April 11, 2016 from http://www.lrrd.org/lrrd3/1/hardouin.htm.

Touchberry R W and Lush J L 1950 The accuracy of linear body measurements of dairy cattle. Journal of Dairy Science 33: 72-80.

Tsegaye D, Belay B and Haile A 2013 Linear Body Measurements as Predictor of Body Weight in Hararghe Highland Goats under Farmers Environment Ethiopia. Global Veterinaria 11(5): 649-656.

Udeh I and Okonta B C 2013 Prediction of body weight from measurements in grasscutters. Tropical Animal Health and Production 45: 1331-1334.


Received 16 June 2016; Accepted 16 July 2016; Published 1 August 2016

Go to top