Effects of genetic control of subcutaneous fat deposition via using restricted selection indexes on live Performance and carcass characteristics of Pekin ducklings

Estimates of phenotypic and genetic parameters for live Performance and detaiied dissection traits of body components and carcass tissues in Pekin ducks were calculated and used to construct selection indexes. The expected reduction in potential gain in live weight, dressing percentage and carcass characteristics resulting from restricting change in subcutaneous fat level to zero were assessed. The aggregate genotype measurements were weight at slaughter, dressing percentage and percent subcutaneous fat. The index measurements were weight at hatching, weight at slaughter, weight gain, breast width, breast length and breast circumference. The restricted indexes as compared with the unrestricted would lead to minimum reduction in potential gain of 29.4% in total net merit, 96.4% in weight at slaughter, 49.0% in dressing percentage, 74.0% in dissected side weight and 75.0% in muscle to bone ratio. The potential reduction in percent abdominal fat and percent bone in side was decreased by at least 38.5% and 80.2%, respectively. Absolute genetic response results showed that selecting for breast width alone (rT, = 0.79) would be recommended for birds with substandard levels of SCF, whereas use of the restricted index containing weight at hatching and breast width (rTI = 0.54) would be advised for individuals with Standard fatness.


Introduction
The carcass fatness pattern that includes proportionately excessive subcutaneous fat generally represents wasted dietary energy (WEBSTER, 1977;LIN et al.,1980), reflects poor management and breeding program (FOWLER et al., 1976;HEATH et al., 1980, LECLERCQ, 1987;LILBURN, 1987) and could result in a wasted product, both quantitatively and qualitatively (KAUFFMAN, 1982).The importance of restrain-ing the subcutaneous fat to certain levels is still debatable and its impact on the live weight and carcass desirability remains to be demonstrated.The aim of the present paper was to test on Pekin ducks the hypothesis that reduction in potential gain in live weight, dressing percentage and other carcass compositional traits would be expected with use of selection indexes restricting changes in subcutaneous fat level to zero.

Material and Methods
Source of data.Sixty-one male and thirty-five female Pekin ducklings progeny of eight sires (twelve per sire) from the experimental poultry farm of the Faculty of Agriculture, Tanta University were used in the present study.The progeny were produced from one spring hatch.The ducklings were housed in temperature-controlled battery brooders for the first two weeks, after which they were reared on a litterfloored pens under uniform conditions.All ducklings reeeived feed and water ad.libitum.The diet contained approximately 22% protein and a metabolizable energy of 2900 Kcal/kg from hatching to 10 weeks (slaughter age).Traits considered.The weight of the birds were recorded at hatching (VI) and at slaughter (V2) permitting the calculation of weight gain (V3).Prior to killing, the following measurements were taken: breast width (V4), breast length (V5) and breast circumference (V6), as described by SWATLAND (1979).The birds were killed by severing the carotid artery and jugular veins.The head was removed at the atlantooccipital articulation.Heart, liver, and gizzard were separated and their sum of weights 'giblets' was taken (V7).The weight of carcass (without giblets) was recorded (V8).The abdominal fat was removed around the cloaca, gizzard and intestines and weighed (V9).Carcasses stored at -20 °C, were transferred to the Meat Laboratory of the Faculty of Agriculture, Ain Shams University for dissection.They were thawed for approximately 8 hrs (at 5 °C) while being in their bags.The dissected right side weight (V10) was calculated as sum of muscle (Vll), subcutaneous fat including skin (V12), intermuscular fat (VI3) and bone (VI4).This permitted the calculation of muscle: bone ratio (VI5) and muscle: fat ratio (V16).

Parameters estimation.
The data used for parameters estimation and selection index construetion included six live Performance traits representing the available sources of Information (VI through V6) and nine slaughter traits (V7 through VI6).The genetic parameters of the traits were estimated from sire components of variance and covariance by the least Squares and maximum likelihood program (HARVEY, 1990), aecording to the following mixed model: where: Y ijk = a trait measurement taken on the k ^ bird in the ij * subclass; \i = an effect common to all birds; Fi = a random effect of the i * sire; Sj = a fixed effect of the j * sex; eyic = a random error assumed N.I.D. (0, o e 2 ).
Definition of net income.The breeding objective was to maximize the value of net income (revenue -cost) for each bird.SOCHOCKA and WEZYK, 1971;0.28, PINGEL andJUNG, 1979, 0.35 PINGEL andHEIMPOLD, 1983).The h 2 value of 0.27 for weight gain is much lower than the estimate of 0.73 obtained in the Pekin ducks sample investigated by CLAYTON and POWELL (1979).The h -value for per cent abdominal fat in ducks seems to be moderate (0.41, Table 1).
Dressing percentage is less heritable in Pekin ducks (0.59, Table 1) than in Muscovy ducks (0.93, PILLA, 1974) and carcass weight is less heritable in the present sample of Pekin ducks than in that of CLAYTON and POWELL (1979) (0.74).Carcass compositional traits were moderately heritable (0.26 to 0.47) with the only exception of percent intermuscular fat in side (0.12).

(B) Correlations
The phenotypic and genetic correlations among traits are given in Table 2. Phenotypic association among index traits.Measures of breast conformation were positively correlated with one another (0.25 to 0.65), indicating that breast width could suitably replace breast length or breast circumference as index trait.The breast conformation was more strongly related to weight at slaughter (0.53 to 0.82) or to weight gain (0.53 to 0.83) than to weight at hatching (-0.09 to +0.02).The correlation between weight gain and weight at slaughter of 1.00 implies that the inclusion of both traits in one index is of no use.Phenotypic and genetic association among individual traits in aggregate breeding value.The basic traits (i.e.those having economic values) were strongly intercorrelated positively (r G = 0.82 to 0.89; r p = 0.52 to 0.74), implying that restricting change in per cent subcutaneous fat to zero would lead to considerable genetic reduction in weight at slaughter and dressing percentage.The basic traits appeared to be strongly correlated positively with dissected side weight, per cent intermuscular fat in side and muscle to bone ratio, but negatively correlated with per cent muscle in side and per cent abdominal fat in body.The two latter traits were practically independent (r G = -0.02;r p = -0.04).

Genetic association between individual traits in aggregate breeding value and those in index.
The dressing percentage and per cent subcutaneous fat were highly correlated positively with breast width (0.90 and 0.58, respectively) and weight at slaughter (0.89 and 0.82, resp.).The breast width and weight at slaughter were positively correlated with one another (0.54) and with weight gain (0.57 and 1.00, resp.), dissected side weight (0.68 and 0.99, resp.) and per cent intermuscular fat in side (0.14 and 0.38, resp.) and negatively correlated with per cent bone in side (-0.75 and -0.88, resp.).Whereas per cent muscle in side was practically independent of breast width (-0.01), it was moderately correlated negatively with weight at slaughter (-0.28).

(C) Economic values
The economic values were estimated at LE 7.0 x 10" 4 per gram weight at slaughter (WS), LE 1.1 x 10" per 0.01 dressing percentage (DP) and LE -8.7 x 10" 2 per cent subcutaneous fat (SCF).Therefore, the true breeding value (T) is-T = 7.0 x W 4 g ws + 1.1 x 10" 2 g DP -8.7 x 10" 2 g SCF where g is the additive genetic value for the considered trait, (a T = 21.60).(D) Indexes Table 3 gives the b-values, the relative value for each trait and the r T1 -values representing the multiple or simple correlation of index with genetic value for net merit.Breast width seems to be the most valuable source of Information in the indexes considered.This is due to its strong genetic relations with the basic traits (0.54, with weight at slaughter; 0.90, with dressing percentage; 0.58, with per cent subcutaneous fat).Its relative value rahged between 30.7 and 78.8% in the unrestricted indexes and attained 94.6 and 98.5% in the restricted indexes, ( 7 7(SCF) and 7 8(SCF) ), respectively.The relative values of breast length, breast circumference, weights at hatching and at slaughtering and daily gain in the most accurate indexes (7,, 7 2 ,1 5 ,1 6 , 7 7 and 7 8 ) did not exceed 1.1%.Indexes accuracy.The six unrestricted indexes involving breast width were the most accurate (r TI = 0.79 to 0.80).Selecting for breast width alone (7 6 ) would be as efficient as selecting for the füll index in its two combinations (7, and 7 2 ).Selection for weight traits alone would create a remarkable reduction in accuracy (r TI = 0.50 for 7 3 and 0.48 for 7 4 ) demonstrating that little can be gained by selection for these traits alone without information on breast conformation.Restricting change in per cent subcutaneous fat to zero through use of 7 7(SC F) and 7 8(SC F) would cause, respectively, 32.5 and 60.8% reduction in accuracy of selection since per cent subcutaneous fat is more related genetically to weight at slaughter (0.82) than to weight at hatching (-0.37).

Genetic response.
Results of the expected outcome in total merit and individual traits for 7 6 ,7 7 , 7 7( SCF) , 7g and 7 8(SC F) are given in Table 4. Selecting for the unrestricted indexes (7 6 , 7 7 and 7 8 ) should lead to increased net merit of LE 0.17 through the production of extra weight at slaughter (44.8 to 49.3 g) with higher dressing percentage (2.04 to 2.09%) to overcome the reduction from over fatness (0.98 to 1.09%).Increases in daily gain (0.67 to 0.73 g/day) and dissected side weight (27.3 to 29.0 g) could also be expected.The small change in per cent muscle in side is of little consequence since enough extra muscle to bone ratio of 0.12 unit is expected to result.Fat would be expected to decrease abdominally and increase intermuscularly with no great change in muscle to fat ratio.As compared to their unrestricted forms (7 7 and 7 8 ), the restricted indexes 7 7(SC F) and7 8 (SCF) would lead to obvious reduction in potential gain (Figure).Minimum reduction of 29.4% in total net merit, 96.4% in weight at slaughter, 89.6% in weight gain, 49.0% in dressing percentage, 74.2% in dissected side weight and 75.0% in muscle to bone ratio would be resulted.The potential reduction in per cent abdominal fat and per cent bone in side should be decreased by at least 38.5%) and 80.2%, respectively.Based on the absolute genetic response results (Table 4).It appears that if it is desired to optimize selection for the given aggregate genotype on condition that no genetic change would occur in percent SCF, individuals with Standard levels of SCF could be selected using the index as it would lead to no reduction in live body weight or carcass weight.In case of substandard levels of SCF, the single factor index I 6 = 15.05Breast width (rn = 0.79) would be recommended but with precaution as it is expected to result in a genetic increase in per cent subcutaneous fat of 1.09 unit after one round of selection.ERNST RITTER, Dummerstorf

Figure :
Figure: Expected reduction in potential gain in total merit, weight at slaughter, weight gain, dressing percentage, dissected side weight and muscle : bone ratio resulting from restricting change in subcutaneous fat (SCF) to zero.• l7(scro and D I^SCF) compared to their respective unrestricted forms, I 7 and lg as ((Restricted -unrestricted) / unrestricted, %)

The aggregate genotype, economic values and selection indexes. Net
The revenue was calculated as carcass weight (kg) * price (LE /kg) according to subcutaneous fat score based on a preliminary market study.For scores A, B and C where SCF level was respectively,

Table 2 .
Coefficients of phenotypic correlation (above diagonal) and genetic correlation (below diagonal) among live Performance and slaughter traits (Phänotypische Korre-

Table 4
Expected genetic changes per generation in total merit and individual traits (intensity of selection = 1.0) (Ausgewiesener genetischer Fort WH= weight at hatching; WS= weight at slaughter; SCF= per cent subcutaneous fat.