Introduction

West African Dwarf (WAD) breed of sheep is the most numerous in the humid southwestern part of Nigeria, a large proportion of which is under extensive management system. The breed is primarily reared for its meat. This animal species is yet to undergo any specialized breeding for higher meat yield. Genetic improvement of its live weight is required to increase meat yield from this breed. Body measurements (BM) are simple and easily measured variables for estimating live weight although it is unlikely to be more accurate than direct measurement of live weight due to error in location of reference points and anatomical distortions produced when animal change position or posture or muscle tone. In most cases the derived equations can only be useful on-station where such information are required for the purpose of selection and breeding. Despite these limitations, body measurements have been used to evaluate breed performance and to characterize breed of animals. In addition, it has been used as a means of selecting replacement animals and evaluating breed in a controlled environment (Shrestha et al. 1984) and yearling weight in sheep (Aziz et al. 1981).

Among West African small ruminants, strong linear and geometric relationship between live weight and chest girth have been reported in Djallonke, Sahel x Djallonke, Red Sokoto goats and Yankasa sheep (Antobam 1983; Benyi 1997; Fasae et al. 2005). There is paucity of information on the relationship between live weight and body measurements of WAD sheep as in cattle (Davis et al. 1961; Biogolo and Meregali 1972; Young 1972; Spencer and Eckert 1988; Mgbere et al. 2005), WAD goats and crosses (Ozoje 1997; Benyi 1997), and other breeds of sheep (Weiner and Hayter 1974; Bhadula et al. 1979; Bhad et al. 1980; Shrestha et al. 1984; Ibiwoye et al. 1993). Furthermore, in most of the studies only chest girth was considered in prediction equations. In a breeding programme where improved live weight is the overall breeding objective other body measurements having strong correlation to live weight must be considered. Measurement of height at withers and body length may indicate crossbreeding (Hall 1991). This study was undertaken to obtain prediction equations for estimating live weight of WAD sheep from eight body measurements for the purpose of breed characterization and selection for genetic improvement.

Materials and methods

Location and climate of the study area

Animals used for live weight and body measurements were sourced among the smallholder farming community in Ajegunle Abeokuta, Ogun State, Nigeria. The location falls within latitudes 7° 5.5′–7° 8.0′ N and longitudes 3° 11.2′–3° 12.5′ E. The climate is humid and is located in the derived savanna zone of South Western Nigeria. It receives a mean annual precipitation of 1,455 mm, mean annual temperature of 34.7 °C and mean relative humidity of 82%.

Management of animals

A total of 210 WAD sheep, 97 males and 113 females, were selected for live weight and body measurements among the flocks of smallholder farmers. There was no birth record therefore determination of age was by dentition (FAO 1994). The estimated age of the animals was between 13 and 36 months. The sheep were intensively managed. Panicum maximum was cut and carried to their pens in the morning and supplementation with concentrates was done in the evening.

Live weight and body measurement procedures

Data was collected over a three-month period. Live weight (kg) of individual animal was determined prior to morning supply of feed to avoid error due to gut-fill using MeasuretecR hanging scale. Body measurements were obtained by the use of measuring tape calibrated in centimeters (cm) after restraining and holding the animals in an unforced position. Reference points for the body measurements were determined according to the procedure of Searle et al. (1989). Height at withers (HW) was obtained as the highest point over the scapulae vertical to the ground. Heart girth (HG) was obtained as the smallest circumference just behind the foreleg. Body length (BL) was obtained as the distance from the head of the humerii to the distal end of the pubic bone. Head width (HDW) was obtained as the distance between the outer ends of both eyes. Head length (HL) refers to the distance between the horn site and the lower lip. Loin girth (LG) was determined as the circumference round the animal just before the hind leg. The reference point for Length of hindquarter (LHQ) was the distance located between the 10th rib and the ventral tuberosity of the tuber iscshii while Width of hindquarter (WHQ) was the circumference the animal around the 10th rib.

Statistical analyses

Statistical analyses were carried out using SPSS Software version 10.0 (SPSS 1999) linear and non-linear regression procedures. Live weight was regressed on the body measurements separately for males and females (sex-specific), and for the pool data (not sex-specific). In the multiple regression equation, prediction equations were developed for live weight using a stepwise elimination procedure. The following models were used.

$${\text{W}} = {\text{a}} + {\text{bG}}\quad ...............................{\left( {{\text{linear}}} \right)}$$
$${\text{W}} = {\text{aG}}^{{\text{b}}} \quad ..............................{\left( {{\text{allometric}}} \right)}$$
$${\text{W}} = {\text{a}} + {\text{b}}_{1} {\text{G}}_{1} + {\text{b}}_{2} {\text{G}}_{2} + ... + {\text{b}}_{{\text{n}}} {\text{G}}_{{\text{n}}} .........{\left( {{\text{multiple}}\,{\text{linear}}} \right)}$$

where W = live weight; a = intercept; G = body measurement; b = regression coefficient of W on G; n = nth number of body measurement.

Results

Live weight and body measurements of WAD sheep

Table 1 shows the Least Square means of live weight and body measurements of WAD sheep. The males used in the experiment were of higher mean liveweight that the female although the highest individual weight was in the female. The male and female WAD sheep had a mean live weight of 16.99 kg and 15.20 kg, respectively. In all the body measurements, male WAD sheep had higher values compared to the females.

Table 1 Range and least square means of live weight (kg) and body measurements (cm) of WAD sheep
Full size table

Linear regression model

Parameter estimates of linear regression equations predicting live weight from body measurements of WAD sheep are presented in Table 2. In this study all the body measurements were good in predicting the liveweight of the WAD sheep except HDW, which had coefficient of determination (R2) below 50%. The highest R2 was observed in WHQ of rams and HG of ewes while WHQ depicts the best overall. The low values of the standard deviations showed the closeness of the predicted to the actual values.

Table 2 Linear regression parameters for estimating live weight from body measurements of WAD sheep
Full size table

Allometric regression model

Table 3 shows the parameter estimates of allometric equations for predicting live weight from body measurements of WAD sheep. The R2 values computed for the body measurements were generally higher (0.87–0.99) using allometric regression than linear regression (0.44–0.94). This yielded an average of 6% higher R2 value than linear regression. Heart girth (HG) and WHQ depicted the highest relationship to live weight in allometric equations compared to other body measurements. Heart girth (HG) accounted for 99% of live weight in both rams and ewes while WHQ accounted for 99% of live weight in rams and 98% in the ewes.

Table 3 Allometric regression parameters for estimating live weight from body measurements of WAD sheep
Full size table

Multiple linear regression model

The multiple linear equations for estimating live weight from body measurements of WAD sheep are presented in Table 4. All the eight body measurements were fitted into the model and through stepwise elimination procedure five of the body measurements were considered unfit in the model. The three body measurements that best fit the model are heart girth (HG), head length (HL) and width of hindquarters (WHQ) accounting for 95% of the live weight in the rams and 91% in the ewes.

Table 4 Multiple linear regression equations for estimating live weight from body measurements of WAD sheep
Full size table

Correlations of live weight and body measurements

Correlation coefficients obtained between the live weight and body measurements of WAD sheep are presented in Table 5. There was a high and positive correlation between the live weight and all body measurements. The highest correlation coefficient (r) was depicted by WHQ (r = 0.96) followed by HG (r = 0.94). The lowest correlation was depicted by HDW (r = 0.66). There was also a high and positive correlation among all the body measurements.

Table 5 Pearson correlation matrix of live weight and linear body measurements of WAD sheep
Full size table

Discussion

Differences in live weight and body measurements of rams and ewes show that these parameters are sex dependent. Ewes has slower rate of growth and reaches a lower mature size due to the effect of estrogen in restricting the growth of the long bones of the body. The values of HW, HG and BL are close to values of 50.5, 61.7, 44.5 cm and 50.2, 60.5, 43.2 cm reported for WAD rams and ewes, respectively (Hall 1991). The slightly lower value of these parameters could be attributed to few numbers of animals (23) involved in his work. There are no published references on HDW, HL, LG, LHQ and WHQ in WAD sheep suggesting that this breed has not been properly characterized.

Width of hindquarter (WHQ) and HG better described the relationship between live weight and body measurements of WAD sheep. Both WHQ and HG gave the best R2 in linear equation. Heart girth has been reported to be the most satisfactory single variable in estimating live weight (Ibiwoye et al. 1993) however, in this study, WHQ seems to compare favourably with HG in estimating live weight using linear equations. The lowest R2 value was exhibited by the HDW (R2 = 0.43) suggesting that HDW is not a reliable body measurement in predicting the liveweight of WAD sheep.

The results obtained with allometric equations supports the assertion that live weight tends to be underestimated by linear regression equations in animals with large chest girth (Brown et al. 1983). This also confirms earlier findings that the relationship between live weight and chest girth in sheep and goats as in cattle is curvilinear (Davis et al. 1961; Pander et al. 1989 and Benyi 1997). Heart girth explained 99% of variation in live weight of WAD sheep by allometric and 89% by linear regression. This result is in line with previous reports where HG has been reported to be the most satisfactory single variable in estimating live weight (Davis 1961; Biogolo and Meregali 1972; Spencer and Eckert 1988; Ibiwoye et al. 1993, Benyi 1997). In Iranian Mehraban sheep (Bathaei 1995), HW, BL and HG were the most significantly associated with variation in weight using allometric equations. No published work is available on the relationship of WHQ to live weight. Areas around the WHQ also include the rump, which is one of the six-hindquarter cut of choice in carcass evaluation and therefore would be an important parameter in a selection and breeding programme.

Parameter estimates in multiple linear regression model shows that more than one body measurement may be required for predicting live weight especially in breeding programme for WAD sheep in which overall live weight is the breeding objective. In this study, HG, HL and WHQ are the most important body measurements required for selection and breeding in WAD sheep. Bamiro (1991) reported that the use of combination of HG, length of brisket (LB) and HW enhanced the efficiency of predicting the live weight of Yankassa sheep. This implies that body conformation varies in different species of sheep.

The positive correlation coefficient in this study is in agreement with Mgbere et al. (2005) who reported similar relationship between live weight and body measurement in N’Dama cattle. This indicates that an increase in any one body measurement would result in a corresponding increase in live weight.

Conclusion

The pattern of relationship observed in this study indicates that for a breeder or stockman to have a fairly good knowledge of the live weight of WAD sheep, measurement of HG and WHQ will be useful. Selection and breeding based on these two body measurements could result in improved live weight in WAD sheep. Such measurements would also aid in recognizing and describing the complex relationships that contribute to adaptation of WAD sheep to humid tropical environment and its functional efficiency.