1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Predicting Body Weight Based on Body Measurements at Different Ages and sex in 2Saburai Goat AKHMAD DAKHLAN*, ARIF QISTHON, MUHAMMAD DIMA IQBAL HAMDANI1Department of Animal Husbandry, Faculty of Agriculture, Universitas Lampung Jl. Prof. Sumantri Brojonegoro No.1 Gedung Meneng Bandar Lampung, Lampung, Indonesia 35145 Correspondence: akhmad.dakhlan@fp.unila.ac.id Abstract | This 2study aimed to predict live body weight (BW) of Saburai goat based on body measurements (body length = BL, shoulder height = SH, chest girth = CG) in Saburai goats at different ages and sex. A total of six hundred fifty-four data generated from 214 Saburai goats (108 female goats and 96 male goats) aged 0-12 months were used in this study. We used analysis of correlation and regression model between body measurements and BW using R program. Determination coefficient (R2), 17root mean square error (RMSE), Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used as a reference in finding the best model in addition to using stepwise regression to find a simpler and more efficient model. The result showed that BL, SH and CG positively correlated to BW with the highest correlation value generally was CG followed by BL and SH. 2Combination of BL and CG was the best predictor for BW and showed the more parsimonious regression model for predicting BW at all ages and sex. Age and sex of goat affected the regression model suggesting that different sex and age had own regression model to predict BW of Saburai goat. Keywords | Body measurements, Body weight, Correlation, Regression model analysis, Saburai goat 30 INTRODUCTION 31 Saburai goat is a new composite breed goat and a superior goat in Lampung Province and 32 was established 22by the Ministry of Agriculture of the Republic of Indonesia in 2015 as a 33 Lampung local goat (Kementan, 2015). The Saburai goat is a small ruminant livestock as 34 a result of 2grading-up between females Ettawa Grade (EG) goat (Ettawa x Kacang goat) 35 2and male Boer goats with genetic composition of 25% EG goats and 75% Boer goats. 36 Saburai goat is much more preferred by the village communities because of its rapid 37 growth exceeding other local goats in Indonesia such as Kacang goat, Rambon goat or 1 38 EG goat, although their body weight is lower than Boer goats. In 2018 Saburai goat 39 population reached to 3,293 heads (Sulastri et al., 2019). 40 The average adult body weights of male and female Saburai goats were 37.27 ± 41 7.38 kg and 33.81 ± 6.12 kg, respectively, with body length of 58.10 ± 14.94 cm and 42 51.34 ± 10, respectively, shoulder height of 61.79 ± 8.95 cm and 55.67 ± 6.81 cm, 43 respectively, and chest girth of 63.78 ± 8.06 cm and 55.95 ± 9.05 cm, respectively 44 (Kementan, 2015). According to Sulastri et al. (2014), the average of birth weight, 45 weaning weight and yearling 52weight of mixed male and female Saburai goats were 3.02 46 ± 0.66 kg, 19.67 ± 1.54 kg, and 42.27 ± 2,12 kg, respectively. 47 In general, Saburai goats were raised by farmers for breeding and as savings that 48 can be cashed at any time needed for the needs of their daily life. When they sell their 49 goat, the price was based on prediction of goat body weight visually by the middleman 50 (buyer). This prediction is not accurate because the prediction is subjective and this is 51 detrimental to farmers. To overcome this problem, we should find a method in predicting 52 body weight which is practical in the field conditions where facilities are limited in rural 53 area, one of them is by using body measurements as predictors for live body weight 54 without weighing livestock first (Trisnawanto 1et al., 2012; Musa et al., 2012; Paputungan 55 et al., 2013). In addition, knowledge about livestock 13body weight is important for 56 maintenance management, for example, to determine the amount of feed and drug dosage 57 to be given, and for selection purposes. 58 Several 29studies have shown that body measurements of goat can be used to estimate 59 their body weight (Khan et al., 2006; Tekle, 2014; Natsir 1et al., 2010; Basbeth et al., 2015; 60 Hazza et al., 2017; Berhe, 2017; Abd-Allah et al., 2019; Dakhlan et al., 2020; Dakhlan et 61 al., 2021; Waheed et al., 2020). If we see the cross section of the goats, the body resembles 2 62 a geometric shape in the form of a tube. According to Isroli (2001), the volume of the 63 tube itself is actually body weight. The width of the base on the tube is the circumference 64 of the chest while the height is the length of the body. This 1combination of chest girth and 65 body length results in a tube volume which is called as body weight. That is why in many 66 studies chest girth and 19body length were the best predictor for live body weight (Topal & 67 Macit, 2004; Afolayan 16et al., 2006; Chitra et al., 2012; Shirzeyli et al., 2013; Iqbal et al., 68 2013; Waheed et al., 2020; Dakhlan et al., 2020; Dakhlan et al., 2021). 69 Studies on 53using body measurements to predict live body weight43of goat have been 70 done in many region of the world. In previous studies, 6the relationship between body 71 weight and body measurements of goats was based on a certain age and on a certain 72 gender. There has not been much research on how age and gender influence the regression 73 model. The present study was 1aimed to analyse the correlation and regression model 74 between live 3body weight and some body measurements (body length, shoulder height, 75 chest girth) of Saburai goat at different ages and sex, and to find the fittest regression 76 model in estimating 34body weight based on body measurements. 78 MATERIAL AND METHODS 79 The data used in this study were 654 kids aged 0-12 months generated from 108 80 Saburai does and 96 Saburai bucks recorded from 2010-2016. For the purposes of 81 analysis, the data 30were divided into 3 age groups, namely 0, 3-6 and 9-12 months of age. 82 Live body weight was taken using scales. 44Body length (BL) was measured as the 83 distance9from the anterior edge of shoulder to the posterior edge of ischium using a meter 84 tape; shoulder height (SH) is the distance from the ground surface where the goat is 85 trampling to 1the highest part of the shoulder using a measuring stick with the goat in an 77 3 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 upright position; 23chest girth (CG) was measured as the circumference of the chest just behind forelegs using meter tape. Data tabulation was done with the Excel program. 21Descriptive analysis was performed using mean, standard deviation (SD), minimum, maximum and coefficient of variation (CV%) of the data. 31Correlation and regression model analysis between body measurements (body length, shoulder height and chest girth) and live body weight were performed using R program. Linear and multiple linear regression model used 1were as follows: 1. Y = a + b1*BL 2. Y = a + b2*SH 3. Y = a + b3*CG 4. Y = a + b1*BL + b2*SH 5. Y = a + b1*BL + b3*CG 6. Y = a + b2*SH + b3*CG 7. Y = a + b1*BL + b2*SH + b3*CG where Y is live body weight (kg), BL is51body length (cm), SH is shoulder height (cm), CG is chest girth (cm), a is intercept or regression constant, 49b1, b2, and b3 are regression coefficient of BL, SH and CG, respectively. Shapiro-Wilk normality test and QQ-Plot were used to check normality of residual of each regression model. Pearson’s correlation coefficient was used 14to evaluate the relationship between body weight and body measurements. Determination coefficient (R2), 20RMSE (root mean square error), AIC (Akaike Information Criterion) and BIC (Bayesian Information Criterion) were used as a reference in the search for the best model in addition to using stepwise regression to find a simpler and more efficient model (simplicity of models). The higher R2 and the smaller RMSE, AIC and BIC of a model is the better. 12Student t-test was used to investigate the effect of sex and age on predicted of two regression models compared. ANOVA test was also performed 38to evaluate the effect of sex and age on the slope of two regression models compared (Dakhlan, 2019; Astuti, 4 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 2007) by comparing the two regression models with and without interaction with sex or age. In addition, visualization of the two regression models can be used to determine whether the two regression models were different or not. To evaluate multicollinearity among predictor variables (body measurements) we calculated the VIF (Variance Inflation Factor). If the VIF > 10 means that there were multicollinearity of predictor variables in the data set. RESULTS Description 2of Body Weight and Body Measurements of Saburai Goat 3Data on body weight and body measurements of Saburai goats at different groups of age and sex are presented in Table 1. Normality test revealed that all data used in this study were normally distributed. 1In general, body weight and body measurements of male Saburai goat were5higher (P<0.05) than those of female goat. In the early age (at birth) the 42body weight and body measurements for both males and females were observed to be almost similar, but with age advancement males performed higher (more weight, longer 48body length, higher shoulder height, and bigger chest girth) than their counterpart females. The results showed that overall means along with standard deviation for body weight at 0, 3-6, and 9-12 month of ages were 3.41 ± 0.49, 17.29 ± 4.79, and 31.77 ± 7.58 kg, respectively for females and 3.30 ± 0.33, 23.25 ± 6.61, and 35.93 ± 7.16 kg, respectively for males. The body length of female in the three age groups was recorded to be 31.18 ± 2.49, 46.62 ± 5.21, 59.48 ± 7.26 cm, 3respectively, while that of the male in these age groups were found to be 34.21 ± 3.17, 50.53 ± 6.93, 63.51 ± 5.99 cm, respectively. The average of shoulder height in female in three age groups was noted to be 31.81 ± 2.69, 49.29 ± 5.31, 60.63 ± 6.97 cm, respectively, 3while that of the male in 5 138 these age groups were recorded to be 33.71 ± 2.15, 46.66 ± 6.31, 60.81 ± 5.49 cm, 139 15respectively. The mean chest girth in the female of three age groups was found to be 35.03 140 ± 3.65, 54.53 ± 6.87, 68.30 ± 7.95 cm, respectively, while that of the male in the same 141 age groups were investigated to be 34.33 ± 2.13, 60.97 ± 7.57, 76.08 ± 5.72 cm, 142 respectively. 143 2TABLE 1: Statistics of body weight (kg) and body measurements (cm) of Saburai 144 goats aged 0, 3-6, and 9-12 Months Age Sex Parameters BW BL SH CG 0 month (at birth) Female (n = 108) Average 3-6 months SD Maximum Minimum CV% Male (n = 96) Average SD Maximum Minimum CV% Female (n = 138) Average SD Maximum Minimum CV% Male (n = 157) Average SD Maximum Minimum CV% 3.41 31.18 31.81 0.49 2.49 2.69 4.4 38 38 2.4 25 27 14.38 7.99 8.47 3.3 34.21 33.71 0.33 3.17 2.15 4.5 41 38 2.8 24 28 9.95 9.28 6.37 17.29 46.62 49.29 4.79 5.21 5.31 30 58 58 10 36 38 27.69 11.17 10.77 23.25 50.53 46.66 6.61 6.93 6.31 39 67 61 9.5 37 35 28.43 13.71 13.53 35.03 3.65 48 24 10.43 34.33 2.13 42 28 6.21 54.53 6.87 69 40 12.61 60.97 7.57 77 39 12.42 9-12 months Female (n = 102) Average SD Maximum Minimum CV% Male (n = 119) Average SD Maximum Minimum CV% 31.77 59.48 60.63 7.58 7.26 6.97 55 77 77 21 47 49 25.85 12.2 11.49 35.93 63.51 60.81 7.16 5.99 5.49 60 76 75 24 49 54 19.93 9.43 9.03 68.3 7.95 86 51 11.64 76.08 5.72 93 66 7.51 145 1Note: BW = body weight; BL = Body length; SH = Shoulder height; CG = Chest girth; 40SD = standard 146 deviation; CV = Coefficient of variation 147 6 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 The birth weights of current study (3.40 ± 0.50 kg for females and 3.30 ± 0.34 kg for males of Saburai goats) were higher than those of other local goats such as Boerawa goats (50% Boer goat; 50% EG goat), EG goats, Boerka goats (50% Boer goat; 50% Kacang goat) and Kacang goats with birth weights of 112.91 kg, 2.36 kg, 2.43 kg and 1.87 kg, respectively (Harris et al., 2009). In comparison to birth weight of female Saburai goat from previous study by Adhianto et al. (2017) (3.3 ± 0.4 kg) and Adhianto et al. (2019) (3.1 ± 0.3 kg), this current study was relatively the same. The similarity result with previous study with the same breed of goat may be because of the same regency with relatively the same environment. Weaning weights (3 months of age) of Saburai goats (not presented here) in this study (14.75 ± 2.89 33kg for females and 18.57 ± 4.12 kg for males) were higher than those of other local goats such as Boerawa goats and EG goats with weaning weights of 14.1128 ± 0.71 kg and 12.93 ± 0.56 kg, respectively (Sulastri and Dakhlan, 2006). However, this current results were relatively the same as Dakhlan et al. (2009) results who reported that weaning weights of Boerawa goats and EG goats that were traditionally reared were 18.40 kg and 16.81 kg, respectively, whereas for rational rearing with the addition of concentrates were 22.95 kg and 18.06 kg, respectively. This result 1confirmed with the result reported by Adhianto et al. (2017 and 2019) in 39female Saburai goat in Gisting and Sumberrejo district, Tanggamus regency, which were 16.10 ± 3.40 kg and 14.90 ± 3.70 kg, respectively. The current study finding (at 3-6 and 9-12 month of age) are in partially agreement with those of Thiruvenkadan (2005), with body weight were higher in the present study. Khan et al. (2006) reported similar result with current study, those were 14.50-18.60 (kg) live body weight, 58.70-59.60 (cm) body length, 56.50-60.30 (cm) height at withers and 7 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 57.60-59.10 (cm) hearth girth, in Beetal goats at 4-12 month of age. Whereas Raja et al. (2013) reported mean body weight of 10.14 ± 0.28 and 9.82 ± 0.26 (kg); body length of 46.38 ± 0.43 and 46.46 ± 0.41 (cm); withers height of 51.05 ± 0.46 and 50.62 ± 0.43 (cm) and heart girth of 49.99 ± 0.44 and 49.47 ± 0.42 (cm), respectively for male and female of Beetal goats at 3-6 month, 7which are not in agreement with the findings of the present study. This differences may be because of different breed and environmental and rearing management. Correlation and Regression Model 5between Body Weights and Body Measurements of Saburai Goats Correlation coefficient between body weight and body measurements of female and male Saburai goats at different sex and age 1are presented in Table 2. Results of this study26showed that body length (BL), shoulder height (SH) and chest girth (CG) positively correlated to body weight (BW) with the highest correlation value generally was CG followed by BL and SH. Furthermore, correlation among predictor variables were under 0.90 meaning that different predictor variables have no similar predictive relationship with the BW. In addition, multicollinearity test showed that among predictor variables (body measurements) had VIF (25variance inflation factors) value lower than five indicating that 25multiple regression model can be applied as predictor variables for BW in this study (Dakhlan, 2019). The present study showed that the correlation coefficient 47between body weight and body measurements increased with increasing ages. The results 1of current study confirmed with the result reported by Khan et al. (2006) that generally 18correlation coefficient between body weight and body measurements increased with increasing ages. However, current result differed from those reported by Basbeth et al. (2015) which the 8 197 18correlation coefficient between body weight and body measurements decreased with the 198 increasing ages. 46This different finding might be caused by the different breed and 199 environment of the goats. 200 201 TABLE 2: Phenotypic correlations between live body measurement traits BW BL SH CG Female 0 month (at birth) BW 1.00 BL 0.72 1.00 SH 0.34 0.43 1.00 CG 0.62 0.77 0.56 1.00 Female 3-6 month 202 203 204 BW 1.00 BL 0.73 1.00 SH 0.54 0.78 CG 0.82 0.71 1.00 0.71 1.00 Female 9-12 month BW 1.00 BL 0.82 1.00 SH 0.72 0.84 CG 0.90 0.76 1.00 0.71 1.00 Male 0 month (at birth) BW 1.00 BL 0.48 1.00 SH 0.50 0.36 CG 0.66 0.45 1.00 0.62 1.00 Male 3-6 month BW 1.00 BL 0.67 1.00 SH 0.47 0.80 1.00 CG 0.89 0.65 0.48 Male 9-12 month BW 1.00 BL 0.83 1.00 SH 0.67 0.47 1.00 CG 0.95 0.84 0.60 1Note: BW = body weight; BL = Body length; SH = Shoulder height; CG = Chest girth 1.00 1.00 The results of this study differed from those by Khan 36et al. (2006) and Abd-Allah et al. (2019). Khan et al. (2006) reported that the height at withers or shoulder height of 9 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 goats aged 4-12 months had the highest correlation (r=0.75) with body weight followed by heart girth or 35chest girth (r=0.62) and body length (r=0.49) in the Livestock Research and Development Station Surezai, Peshawar, Pakistan. Furthermore, Abd-Allah 19et al. (2019) reported that body length had the highest correlation (r=0.95) with body weight in female Shami goats in subtropical areas in Egypt, while in male goats the highest correlated body measurements to 24body weight was heart girth (r=0.98). The current study corroborate the findings of a study by 24Adeyinka and Mohammed (2006), Raja et al. (2013), Berhe (2017) and Waheed et al. (2020) that 37chest girth had the highest correlation with body weight in Nigerian Red Sokoto Goats, in Attappady Black goats in Kerala, India, in Maefur goat in Tigray, Northern 41Ethiopia, and in Beetal goats in Punjab, Pakistan, respectively. Abd-Allah et al. (2019) reported that 4chest girth had the highest correlation (r=0.98) with body weight in male Shami goats in subtropical areas in Egypt, while in female goats the highest correlated body measurements to body weight was heart girth (r=0.95). While 45Iqbal et al. (2013) reported body length correlated with body weight higher compared to 50height at withers, heart girth, rump and forehead with correlation coefficient of 0.805, 0.766, 0.767, 0.088 and 0.229, respectively. Regression models (equation 1-7), along with their criteria of model selection, between 4body weight and body measurements of female and male Saburai goats at different age are presented in Table 3 and Table 4. 224 TABLE 3: Regression models 2of body weight on body measurements of female Saburai 225 goats aged 0, 3-6 and 9-12 months along with criteria of model selection. Regression model Intercept (a) Coefficient of regression (slope) b1(BL) b2(SH) b3(CG) RMSE r R2 AIC BIC Goats aged 0 month (at birth) BW=a+b1BL BW=a+b2SH BW=a+b3CG -0.99 0.14(0.00) - - 0.34 0.72** 0.52 74.2 82.04 1.44 - 0.06(0.00) - 0.46 0.34** 0.12 134.93 142.78 0.5 - - 0.08(0.00) 0.38 0.62** 0.38 98.54 106.38 10 226 227 228 229 230 231 232 233 BW=a+b1BL+b2SH BW=a+b1BL+b3CG BW=a+b2SH+b3CG BW=a+b1BL+b2SH+b3CG -1.1 0.13(0.00) -0.99 0.12(0.00) 0.53 - -0.95 0.12(0.00) 0.01(0.65) - 0.002(0.92) 0.003(0.87) - 0.02(0.12) 0.08(0.00) 0.02(0.14) 0.34 0.62** 0.33 0.70** 0.38 0.55** 0.33 0.65** 0.52 0.53 0.38 0.33 75.99 73.71 100.53 75.68 86.45 84.17 110.99 88.76 Goats aged 3-6 months BW=a+b1BL BW=a+b2SH BW=a+b3CG BW=a+b1BL+b2SH BW=a+b1BL+b3CG BW=a+b2SH+b3CG BW=a+b1BL+b2SH+b3CG -14.06 0.67(0.00) - - 3.25 0.73** 0.54 723.25 732.03 -6.69 - 0.49(0.00) - 4.02 0.54** 0.29 781.48 790.26 -13.77 - - 0.57(0.00) 2.75 0.82** 0.67 676.44 685.22 -13.26 0.72(0.00) -0.06(0.45) - 3.25 0.67** 0.54 724.65 736.36 -18.56 0.28(0.00) - 0.42(0.00) 2.55 0.84** 0.71 658.2 669.9 -12.22 - -0.08(0.21) 0.61(0.00) 2.73 0.75** 0.67 676.79 688.5 -15.39 0.46(0.00) -0.33(0.00) 0.50(0.00) 2.34 0.78** 0.76 636.39 651.02 Goats aged 9-12 months BW=a+b1BL BW=a+b2SH BW=a+b3CG BW=a+b1BL+b2SH BW=a+b1BL+b3CG BW=a+b2SH+b3CG BW=a+b1BL+b2SH+b3CG -18.94 0.85(0.00) -15.47 - -25.55 - -20.24 0.77(0.00) -30.49 0.33(0.00) -29.72 - -29.88 0.37(0.00) - 0.78(0.00) - 0.10(0.46) - 0.17(0.03) -0.06(0.53) - - 0.85(0.00) - 0.62(0.00) 0.75(0.00) 0.63(0.00) 4.35 0.82** 5.25 0.72** 3.34 0.90** 4.33 0.80** 2.96 0.91** 3.23 0.88** 2.95 0.88** 0.67 0.51 0.8 0.67 0.85 0.82 0.85 416.09 442.94 378.65 417.52 363.63 375.87 365.21 422.88 449.73 385.44 426.57 372.68 384.92 376.52 Note: number in parenthesis indicated the significance of the regression coefficient (b1, b2 and b3); r = Pearson’s correlation coefficient, R2 = 10coefficient of determination; RMSE = root mean square error, AIC = Akaike information criterion, BIC = Bayesian information criterion, 4BW = body weight; BL = Body length; SH = Shoulder height; CG = Chest girth. **significant at P<0.01; bold typed indicated the best regression model. TABLE 4: Regression models 2of body weight on body measurements of male Saburai goats aged 0, 3-6 and 9-12 months along with criteria of model selection. Coefficient of regression (slope) Regression model Intercep t (a) b1(BL) b2(SH) b3(CG) MSE r R2 AIC BIC Goats aged 0 month BW=a+b1BL BW=a+b2SH BW=a+b3CG BW=a+b1BL+b2SH BW=a+b1BL+b3CG BW=a+b2SH+b3CG BW=a+b1BL+b2SH+b3CG 1.62 0.05(0.00) 0.29 0.48** 0.23 0.73 - 0.08(0.00) 0.28 0.50** 0.25 -0.18 - - 0.10(0.00) 0.25 0.66** 0.43 0.14 0.04(0.00) 0.06(0.00) - 0.26 0.59** 0.35 -0.44 0.02(0.01) - 0.09(0.00) 0.24 0.64** 0.47 -0.45 - 0.02(0.15) 0.09(0.00) 0.24 0.65** 0.45 -0.65 0.02(0.02) 0.02(0.23) 0.07(0.00) 0.24 0.67** 0.48 Goats aged 3-6 months BW=a+b1BL BW=a+b2SH BW=a+b3CG BW=a+b1BL+b2SH -9.27 0.64(0.00) 0.33 - -24.03 - -7.31 0.79(0.00) - 0.49(0.00) - -0.20(0.06) - - 0.78(0.00) - 4.86 0.67** 5.82 0.47** 3.03 0.89* 4.81 0.61** 0.46 0.22 0.79 0.47 36.89 34.21 9 23.01 4.41 8.88 4.87 948.22 1004.5 799.41 946.47 44.39 41.71 16.5 33.01 14.41 18.88 17.37 957.39 1013.6 7 808.58 958.7 11 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 BW=a+b1BL+b3CG BW=a+b2SH+b3CG BW=a+b1BL+b2SH+b3CG -26.33 0.16(0.00) -25.2 - -24.95 0.25(0.00) - 0.05(0.23) -0.12(0.05) 0.68(0.00) 0.75(0.00) 0.67(0.00) 2.91 3.01 2.87 0.87** 0.81** 0.79** 0.81 0.79 0.81 Goats aged 9-12 months BW=a+b1BL BW=a+b2SH BW=a+b3CG BW=a+b1BL+b2SH BW=a+b1BL+b3CG BW=a+b2SH+b3CG -26.98 0.99(0.00) -17.03 - -54.35 - -42.71 0.79(0.00) -53.89 0.12(0.12) -57.79 - - 0.87(0.00) - 0.47(0.00) - 0.20(0.00) - - 1.19(0.00) - 1.08(0.00) 1.07(0.00) 3.99 5.3 2.29 3.28 2.25 2.11 0.83** 0.67** 0.95** 0.88** 0.92** 0.91** 0.69 0.45 0.9 0.79 0.9 0.91 788.71 799.95 786.63 476.74 524.54 383.32 446.2 382.84 371.65 BW=a+b1BL+b2SH+b3CG -57.4 0.15(0.04) 0.21(0.00) 0.93(0.00) 2.05 0.94** 0.92 369.32 Note: number in parenthesis indicated the significance of the regression coefficient (b1, b2 and b3); r = Pearson’s correlation coefficient, R2 = 10coefficient of determination; RMSE = root mean square error, AIC = Akaike information criterion, BIC = Bayesian information criterion, 4BW = body weight; BL = Body length; SH = Shoulder height; CG = Chest girth. **significant at P<0.01; bold typed indicated the best regression model. 800.93 812.18 801.91 484.03 531.84 390.62 455.92 392.56 381.37 381.47 Based on R2, RMSE, AIC and BIC criteria, 1the best regression model for predicting body weight of female Saburai goats at 0, 3-6 and 9-12 months (Table 3) was using combination of BL and CG, 2combination of BL, SH and CG, and combination of BL and CG predictors, respectively. On the other hand for male (Table 4), 2combination of BL and CG was the best predictor for BW at 0 month of age, while 1combination of BL, SH and CG was the best predictors for predicting BW both at 3-6 and 9-12 month of age. In all the best regression model, CG was the highest regression coefficient indicating that CG determine the highest variation for BW, except in female Saburai goat at age 0 month which BL (0.12) was higher in regression coefficient than CG (0.02). Among the criteria, AIC and BIC confirmed with the result of stepwise regression analysis method with the highest R2 except in regression model for female at 9-12 month of age in which equation 5 and 7 had the same R2 (0.85) with the regression coefficient of SH in equation 7 was not significant (0.53) so that the best equation noted to be equation 5. Two regression models for male goats at birth (0 month) showed similar R2 (0.47 in equation 5 and 0.48 in equation 7), but the best regression model was equation 5 because in equation 7 12 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 regression coefficient of SH was not significant (0.23). However, two regression models for male at 3-6 month of age had the same R2 (0.81) for equation 5 and 7 but the best regression model was equation 7 with all coefficient of regression were significant (0.00 for BL, 0.05 for SH and 0.00 for CG). Even though all 8body measurements included in the current study could provide together a better prediction of Saburai goat BW because of the higher R2, 2combination of BL and CG showed the more parsimonious regression model with BW and remain high value of correlation. Furthermore, if we wanted to use single body measurement, CG is 1the best predictor to BW compared to BL and SH. It was recorded that 5the highest and strongly positive (P< 0.01) correlation of single body measurement to BW was CG both for the female and male Saburai goats compared to BL or to SH. 1The result of this study was in accordance with the result reported by Raja et al. (2013), 1Chitra et al. (2012), Adeyinka and Mohammed (2006), Berhe (2017), Habib 2et al. (2019), Abd-Allah et al. (2019), Dakhlan et al. (2020), and Dakhlan et al. (2021) 2that CG was the best predictor to BW of goat. In addition, CG 27is the easiest way to measure for live body weight prediction in the field conditions. Effect of Sex and Age on Regression Model Using Single Body Measurements of Saburai Goats Result of t-test between predicted body weight of two regression models with different group of age showed that different age affected on the regression model. This means that different group of age has own regression model. Furthermore, ANOVA test result indicated that different group of age has different slope of regression line (regression coefficient). We can see from Figure 1 that the regression line of different age group are not parallel for each predictor indicating that the three different ages had 13 279 280 281 282 283 284 285 286 287 288 289 290 291 different slope and different predicted 13body weight using the three body measurements. In other words, estimating 13body weight based on each of body measurement may use different regression models for female Saburai goat at different group of age. The same situation happened in male Saburai goats (Figure 2) where different age group had own regression model. FIGURE 1: 2Scatter plot and regression line between body weight and body measurements of Female Saburai goats at three different age group FIGURE 2: Scatter plot and regression line 6between body weight and body measurements of male Saburai goats at three different age group 14 292 293 294 295 296 297 298 299 300 Figure 3-5 show the 1scatter plot and regression line between body weight and body measurements at three different age group. FIGURE 3: Scatter plot and regression line 6between body weight and body measurements of female and male Saburai goats at age 0 month (at birth) 2FIGURE 4: Scatter plot and regression line between body weight and body measurements of female and male Saburai goats at age 3-6 month 15 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 FIGURE 5: 2Scatter plot and regression line between body weight and body measurements of female and male Saburai goats at age 9-12 month The result of t-test indicated that different sex affected on regression model (P<0.05) for each predictor at 0 month of age. ANOVA test result indicated that only body length had different slope (P<0.05) of the regression model. Based on Figure 3 we can see regression line between female and male were significantly different (not parallel and interact each other) for body length predictor, while for shoulder height and chest girth the two regression line between female and male are almost parallel. Overall, different sex had different regression model. At age 3-6 and 9-12 months only chest girth had different slope of the regression line between female and male Saburai goat, while for body length and shoulder height had parallel slope between female and male. The result of t-test and ANOVA test indicated that different sex at age 3-6 and 9-12 month of age had different regression model. CONCLUSION 16 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 1Chest girth had the highest correlation and was the best predictor for body weight of Saburai goat at all ages and sex 1if using single body measurement. The combined chest girth and body length had a higher, significant correlation and more parsimonious model in predicting body weight of the Saburai goat compared to other combination of body measurements at all ages and sex. Sex and age influenced the regression model of body weight to all body measurements at all ages and sex suggesting that different age or sex should use different regression model. Results of this study suggest as well that 32chest girth and body length could be used as selection criteria to improve body weight of Saburai goats. ACKNOWLEDGEMENTS The authors acknowledge to Dean of Faculty of Agriculture, University of Lampung for funding this research and publication cost. Farmer in Sumberrejo district Tanggamus regency Lampung province deserve great thanks for providing data for this study. 12CONFLICT OF INTEREST All authors declared that there is no conflict of interest. AUTHOR’S CONTRIBUTION AD arranged and designed the research, 28analyzed data, and wrote the manuscript. AQ critically reviewed the manuscript. MDIH collected and tabulated the data and studied the literture. REFERENCES Abd-Allah S, Abd-El Rahman HH, Shoukry MM, Mohamed MI, Salman FM, Abedo AA (2019). Some body measurements as a management tool for Shami goats raised in subtropical areas in Egypt. Bulletin of the National Research Centre, 43(1): 0–5. 17 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 5https://doi.org/10.1186/s42269-019-0042-9 Adeyinka IA, Mohammed ID. (2006). Accuracy of Body Weight Prediction in Nigerian Red Sokoto Goats Raised in North Eastern Nigeria Using Linear Body Measurement. Pak. J. Biol. Sci. 9(15): 2828-2830. https://doi.org/10.3923/pjbs.2006.2828.2830 Adhianto K, Anjani AR, Sulastri, Dakhlan A (2019). Qualitative characteristics of Saburai goats at weaning in Tanggamus District Lampung Province Indonesia. In The 8th International Seminar on Tropical Animal Production (pp. 56–60). Yogyakarta. Adhianto K, Sulastri S, Hamdani MDI, Novriani D, Yuliani L (2017). Performans Kambing Saburai Betina Di Wilayah Sumber Bibit Kabupaten Tanggamus Provinsi Lampung. Jurnal Ilmiah Ilmu-Ilmu Peternakan, 20(1): 9–16. Afolayan RA, Adeyinka IA, Lakpini CAM (2006). The estimation of live weight from body measurements in Yankasa sheep. Czech J. Anim. Sci., 51(8): 343–348. Astuti M (2007). Pengantar Ilmu Statistik untuk Peternakan dan Kesehatan Hewan (1st ed.). Bogor: Binasti Publisher. Basbeth A, Dilaga W, Purnomoadi A (2015). Hubungan Antara Ukuran-ukuran Tubuh Terhadap Bobot Badan Kambing Jawarandu Jantan Umur Muda Di Kabupaten Kendal Jawa Tengah. Anim. Agric. J., 4(1): 35–40. Berhe WG (2017). Relationship and Prediction of Body Weight from Morphometric Traits in Maefur Goat Population in Tigray, Northern Ethiopia. J. of Biometrics & Biostatistics, 8(5): 1–6. https://doi.org/10.4172/2155-6180.1000370 Chitra R, Rajendran S, Prasanna D, Kirubakaran A (2012). Prediction of body weight using appropriate regression model in adult female Malabari goat. Veterinary World, 5(7): 409–411. https://doi.org/10.5455/vetworld.2012.409-411 Dakhlan A, Sulastri, Damayanti I, Budiyah, Kristianti K. (2009). Does productivity index of Boerawa does and Etawa grade does fed by traditional and rational foodstuff. In: The 1st International Seminar on Animal Industry (pp. 248–252). Bogor: Faculty of Animal Science, Bogor Agricultural University. Dakhlan A. (2019). Experimental Design and Data Analysis Using R (1st ed.). Sleman, Yogyakarta: Graha Ilmu (Matematika). Retrieved from https://grahailmu.id/product/experimental-design-and-data-analysis-using-r/ Dakhlan A, Hamdani MDI, Putri DR, Sulastri, Qisthon A (2021). Short communication: Prediction of body weight based on body measurements in female saburai goat. Biodiversitas, 22(3): 1391–1396. https://doi.org/10.13057/biodiv/d220341 Dakhlan A, Saputra A, Hamdani MDI, Sulastri (2020). Regression Models and Correlation Analysis for Predicting Body Weight of Female Ettawa Grade Goat using its Body Measurements. Adv. in Anim. Vet. Sci., 8(11): 1142–1146. https://doi.org/10.17582/journal.aavs/2020/8.11.1142.1146 Habib MA, Akhtar A, Bhuiyan AKFH, Choudhury MP, Afroz MF. (2019). Biometrical Relationship between Body Weight and Body Measurements of Black Bengal Goat (BBG). Current J. App. Sci. Tech. 35(2): 1-7. https://doi.org/10.9734/CJAST/2019/v35i230172 Harris I, Dakhlan A, Suharyati S (2009). Performance of Grade-1 Kids as A Result of Grading-Up Between Local Goats and Boer Goats. In: The 1st International Seminar on Animal Industry (pp. 48–50). Bogor: Faculty of Animal Science, Bogor Agricultural University. Hazza ANH, Lestari CMS, Sutaryo (2017). Hubungan antara Ukuran-Ukuran Tubuh 18 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 dengan Bobot Badan Kambing Peranakan Etawah Betina dewasa di Kabupaten Klaten. Agromedia, 35(1): 14–20. Iqbal M, Javed K, Ahmad N (2013). Prediction of body weight through body measurements in Beetal goats. Pakistan J. Sci., 65(4): 458–461. Isroli (2001). Evaluasi terhadap Pendugaan Bobot Badan Domba Priangan Berdasarkan Ukuran Tubuh. Jurnal Ilmiah SAINTKES, 8(1): 90–94. Kementan (2015). Kambing saburai. Surat Keputusan Nomor 359/Kpts/PK.040/6/2015. Tentang Penetapan Rumpun Kambing Saburai. Khan H, Muhammad F, Ahmad R, Nawaz G, Rahimullah, Zubair M (2006). Relationship of body weight with linear body measurements in goats. J. Agric. Biolog. Sci., 1(3): 51–54. Musa AM, Idam NZ, Elamin KM (2012). Regression Analysis of Linear Body Measurements on Live Weight in Sudanese Shugor Sheep. Online J. Anim. Feed Res., 2(1): 27–29. Natsir A, Asja MA, Nasrullah, Yusmasari, Nurhayu A, Murray P, Murray-Prior R (2010). The relationship between heart-chest girth, body length and shoulder height, and live weight in Indonesian goats. In: The 5th International Seminar on Tropical Animal Production (pp. 441–445). Yogyakarta, Indonesia. Paputungan U, Hakim L, Ciptadi G, Lapian HFN (2013). The estimation accuracy of live weight from metric body measurements in Ongole grade cows. J. the Indonesian Trop. Anim. Agric., 38(3): 149–155. https://doi.org/10.14710/jitaa.38.3.149-155 Raja TV, Venkatachalapathy RT, Kannan A, Bindu KA. (2013). Determination of best- fitted regression model for prediction of body weight in Attappady black goats. Global Journal of Animal Breeding and Genetics 1(1): 20-25. Shirzeyli FH, Lavvaf A, Asadi A (2013). Estimation of body weight from body measurements in four breeds of Iranian sheep. Songklanakarin J. Sci. Tech., 35(5): 507–511. Sulastri, Adhianto K, Dakhlan A, Hamdani MDI, Siswanto S (2019). Population Performance of Saburai Goat at Saburai Goat Breeding Area, Tanggamus regency, Lampung Province. In: IOP Conference Series: Earth and Environmental Science (Vol. 372, pp. 1–7). Purwokerto: IOP Publishing Ltd. https://doi.org/10.1088/1755- 1315/372/1/012022 Sulastri, Sumadi, Hartatik T, Ngadiyono N (2014). Performans Pertumbuhan Kambing Boerawa di Village Breeding Centre, Desa Dadapan, Kecamatan Sumberejo, Kabupaten Tanggamus, Provinsi Lampung. Sains Peternakan, 12(1): 1–9. https://doi.org/10.20961/sainspet.v12i1.4758 Sulastri S, Dakhlan A (2006). Comparation on does productivity index between Boerawa and Ettawa grade goat at Campang village, Tanggamus, Lampung. In: The 4th ISTAP "Animal Production and Sustainable Agriculture in The Tropic (pp. 315–319). Yogyakarta: Faculty of Animal Science, Gadjah Mada University, November 8-9, 2006. Tekle T (2014). Predicting live weight using body measurements in Afar goats in north eastern Ethiopia. Momona Ethiopian J. Sci., 6(2): 18–32. https://doi.org/10.4314/mejs.v6i2.109619 Thiruvenkadan AK (2005). Determination of best-fitted regression model for estimation of body weight in Kanni Adu kids under farmer's management system. Livestock Res. Rural Develop. 17(7): Article #85. Topal M, Macit M (2004). Prediction of body weight from body measurements in 19 438 439 440 441 442 443 444 445 446 Morkaraman sheep. J. Appl. Anim. Res., 25(2): 97–100. https://doi.org/10.1080/09712119.2004.9706484 Trisnawanto, Adiwinarti R, Dilaga WS (2012). Hubungan antara ukuran-ukuran tubuh dengan bobot badan dombos jantan. Anim. Agric. J., 1(1): 653–668. Waheed HM, Moaeen-Ud-din M, Khan MS, Saif-Ur-rehman M, Nawaz-Ul-rehman MS (2020). Prediction of monthly body weight from body measurements in beetal goats reared under field and farm conditions. J. Anim. Plant Sci., 30(1): 1–8. https://doi.org/10.36899/japs.2020.1.0003 20