Selection indices for improving body weights in Saudi Aradi and Damascus goats

Data on body weight records of Saudi Aradi goat (A) and Damascus goats (D) were used to construct different selection indices. Records of 314 kids produced by (18 Sir and 67 Dam) for Saudi Aradi goat (A) and of 175 kids produced by (18 Sir and 145 Dam) for Damascus goat (D) were used for the study.  The variance component estimated from sire-dam mixed models was utilized to calculate heritability estimates. Seven selection indices constructed using phenotypic and genetic variances  and  co  variances  for  Saudi  Aradi  (A)  and  Damascus  goats  (D).  The original index (supposed to be 100% efficient in the genetic sense) was constructed to include all the 3 traits while the other three reduce indices include only two traits, in consequence one trait of the three traits (included in the original index) was dropped, while the last three reduce indices include only one trait in consequence two traits of the three traits (included in the original index) was dropped. High values of partial regressions coefficients (b´s) were obtained in weight at six month (6M) For Aradi goats (A) and weaning weight (WW) for Damascus goats (D) in all indices constructed. On the other hand, Comparing the three indexes (I₅ = 2.122 BW, I₆ = 2.352 WW and I₇  = 2.232 6M), For Aradi goats (A) the b value of birth weight (BW) was lower than that of weaning weight and weight at six month in the single indices, but, Comparing the three indexes for Damascus goats (D) (I₅  = 3.217 BW, I₆ = 1.699 WW and I₇ = 1.075 MW), the b value of birth weight (BW) was greater than that of weaning weight and weight at six month in the single indices. Estimates of (V%) indicate that including weight at six month in selection programs offer more economic-genetic contribution than those of other growth traits studied (birth and weaning weights). But, for Damascus goats (D) estimates of (V%) indicate that including weaning weight in selection programs offer more economic-genetic contribution than those of other growth traits studied (birth weight and weight at six month). The percentage of total economic-genetic gain (H%) accounted for by gain in each trait for most indices constructed indicated that weight at six month would contribute up to 74.99% of total economic-genetic gain for Aradi goat and up to 57.39% of total economic- genetic gain for Damascus goat which is higher than birth and weaning weights. Selection on I₁ and I₂ for Aradi goat gave the greatest improvement in 6M (+2.157 kg) and in WW (+1.180 kg). While, selection on I₁ for Damascus goat gave the greatest improvement in WW (+1.400 kg). Correlation between the indices constructed and each individual trait in the aggregate genotype (r_ix) for Aradi goat ranged from 0.109 to 0.391 for birth weight and from 0.084 to 0.669 for weaning weight and from 0.089 to 0.669 for weight at six month. Furthermore, the high correlation for weaning weight and weight at six month indicated that selection per generation based on any index constructed would actually lead to a high genetic gain in weaning weight and weight at six month. Also, these correlations for Damascus goat ranged from 0.020 to 0.641 for birth weight and from 0.343 to 0.599 for weaning weight and from 0.030 to 0.479 for weight at six month. Furthermore, the high correlation for weaning weight and weight at six month indicated that selection per generation based on any index constructed would actually lead to a high genetic gain in weaning weight and weight at six month. While the high correlation for birth weight in (I₅) only for two breeds indicated that the expected genetic change in this trait based on (I₅) would be larger than other indices. The maximum accuracy of selection In Aradi goat (RIH = 0.699) was obtained using the index (I₁). But the accuracy was decreased (RIH = 0.507) when 6M was ignored from the full index (I₄). Selection for 6M alone would be more efficient (RIH = 0.668) than selection for WW (RIH = 0.507) alone or BW (RIH = 0.094) alone. Also, the maximum accuracy of selection Damascus goat (RIH = 0.56) was obtained using the index (I₁). But the accuracy was more decreased (RIH = 0.482) when WW was ignored from the full index (I₃) than (RIH = 0.511) when 6M was ignored from the full index (I₄) Selection for 6M alone would be more efficient (RIH = 0.48) than selection for WW (RIH = 0.466) alone or BW (RIH = 0.235) alone. The efficiencies (RE) of different indices constructed relative to original index (I₁). For Aradi goat, three trait index,  (I₂) based on weaning weight and  weight at six month, index (I₃) based on birth weight and  weight at six month and index (I₇) based on weaning weight and  weight at six month were 99% as effective as the index including the three traits (I₁), while the effectiveness of (I₅) was only 14.00%. But, for Damascus goat one trait index,  (I₂) based on weaning weight and weight at six month were 99% as effective as the index including the three traits (I₁), while the effectiveness of (I₅) was only 42.00%.