Abstract: The research was conducted to determine the Standardized Ileal Digestible Lysine (SID Lys) requirement of male broiler chicks from 0-14 days of age and to study the carry-over effects of early Lys deficiency in growing chickens. In the starter, graded levels of SID Lys in the experimental diets were achieved by addition of Lysine-HCl at the expense of cornstarch obtaining dietary SID Lys ranging from 0.88-1.28% of diet. Linear regression analysis was significant (p<0.05) for Body Weight Gain (BWG), Feed Intake (FI), Breast Meat (BM) and Thigh Yield (THG) to dietary Lys. Both of Linear Broken-Line (LBL) and Quadratic Broken-Line (QBL) models were fitted to determine SID Lys requirement. Regarding response criteria and regression models, SID Lys requirements of starting broiler chicks were from 1.06-1.29% of diet. Assuming 83% lysine digestibility of basal diet, the requirements were 1.28-1.55% of diet on total amino acids basis which were much higher than NRC recommendations. From 14-28 days of age, chickens were fed common grower diet according to NRC. Linear regression analysis was significant for BWG, Feed:Gain (FG), BM and TGH yield (p<0.05). Early lysine deficiency significantly reduced BWG, feed efficiency, BM and TGH yield (p<0.05). This study clearly showed that the very lysine-deficient fed chickens could not mount a compensatory growth.

INTRODUCTION

Although, all Amino Acids (AA) have regulating effects on growth and protein metabolism (Tesseraud et al., 2008), it has been documented that Lys effects on protein accretion is critical especially on breast meat yield (Tesseraud et al., 1996a, b, 2008) and Lys deficiency significantly increased protein degradation in pectoralis major muscle due to heightening the expression of m-calpain and cathepsin B in breast muscle (Tesseraud et al., 2008). Leclercq (1998) showed that higher levels of Lys in the diet would affect body composition by increasing breast meat yield and decreasing abdominal fat percentage. In comparison, other AAs such as Threonine (Thr) and Valine (Val) had no such pronounced effects (Leclercq, 1998). Dietary Lys is the most critical amino acid to breast muscle formation early in development (Tesseraud et al., 1996b, 2008). Labadan et al. (2001) clearly showed that Lys requirement for breast meat yield is higher than for weight gain and feed efficiency. Breast muscle contains a high concentration of Lys so it would be expected that dietary Lys concentrations can have a large influence on breast meat development (Kerr et al., 1999). In NRC (1994) starter period for broiler chickens is defined as a first 3 week of age but in practical situation, starter period for the modern broiler strains with high growth rate has been considered as first 2 week of age. The aims of this study are determination of SID Lys requirements of male broiler chicks from 0-14 days of age and evaluation of the carry-over effects of early Lys deficiency in the grower period.

MATERIALS AND METHODS

Animal and diets: Regarding amino acid requirements of starting broiler chicks, most of the research addressing the requirements during Days 7-21 of age and there is little information on the first 2 week of age. A corn-soybean-cottonseed meal basal diet was formulated to meet or exceed the NRC (1994) requirements for all nutrients except for Lys. Digestibility of amino acids in feedstuffs were obtained from AminoDat (2005) (Version 3.0, Degussa AG, Germany) and calculated metabolizable energy, crude protein, calcium and available phosphorus were taken from NRC (1994). Three hundred sixty 1 day old male Ross 308 chicks were used in a complete randomized design consisted of six treatments with four replications each having 15 birds. The experimental diets were isocaloric and isonitrogenous in which CSM substituted for SBM at 50% by weight (Table 1) and then graded levels of digestible Lys were obtained with addition of Lysine-HCl at expense of cornstarch (Table 2). The birds were offered feed and water ad libitum.

Table 1:	Lysine content of experimental diets
Increment: 0.08%

Table 2:	Composition of basal and grower diets
1Vitamin premix provides per kg of diet: vitamin A, 4,500 IU; vitamin D3, 2,250 ICU; vitamin E, 50 IU; thiamin HCl, 15 mg; riboflavin, 15 mg; nicotinic acid, 50 mg; folic acid, 6 mg; pyridoxine, 6 mg; biotin, 0.6 mg; vitamin B12, 0.02 mg; choline Cl, 2,500 mg; d-calcium pantothenate, 20 mg; menadione sodium bisulfite, 1.5 mg; butylated hydroxytoluene, 100 mg; glucose to make 12 g; 2Mineral premix provides per kilogram of diet: CaCO3, 25.6 g; CaHPO4.2H2O, 5.6 g; KH2PO4, 14 g; NaCl, 5.1 g; MnSO4.H2O, 0.33 g; FeSO4.7H2O, 0.33 g; KI, 0.003 g; CuSO4.5H2O, 0.05 g; ZnO, 0.1 g; CoCl2.6H2O, 0.0017 g; NaMoO4.2H2O, 0.0083 g; Na2SeO3, 0.0004 g

In the growing period, all birds were fed grower diet according to the NRC (1994) from 14-28 days of age and compensatory growth was measured for performance traits.

Feed intake was measured weekly and chicks were weighted at 0, 7, 14, 21 and 28 days of age. Feed:gain was determined as feed consumption divided by the live weight adjusted for mortality. Two birds of each pen were used for breast meat and thigh dissection. The breast muscle was carefully dissected and immediately weighted.

Statistical analysis: The calculated SID Lys values were used for statistical analysis with treatment levels ranging from 0.88-1.28% in increments of 0.08%. Five analyses were conducted: PROC REG (SAS, 2002) using linear response to evaluating potential effects of digestible lysine; PROC REG (SAS, 2002) using quadratic response to evaluating potential effects of digestible lysine (SAS, 2002); PROC NLIN (SAS, 2002) conducting linear broken-line model based on Robbins et al. (2006); PROC NLIN (SAS, 2002) conducting quadratic broken-line model based on Robbins et al. (2006) and PROC GLM (SAS, 2002) to separate differences among means using LSD option of SAS (Table 3). Digestible Lys requirements were estimated using both linear broken-line and quadratic broken-line models when a significant response occurred (p<0.05) (Fig. 1a-c). R² and sum of square residual of each model were used to compare nonlinear models (Table 4). Regression models for estimation SID Lys requirements were as follow:

Table 3:	Growth performance of broiler chicks fed varying levels of digestible lysine from days 0-14 of age
a-cValues within rows without a common superscript are significantly different (p≤0.05)

Fig. 1 (a-c):	Estimation lysine requirement for breast meat yield using Quadratic broken-line model

Table 4:	Lysine requirements (% of diet) of starting broiler chicks using different models and response criteria
1Linear Broken-Line (LBL) and Quadratic Broken-Line (QBL) are y = L+U (R-x) and y = L+U (R-x) (R-x), respectively if R<x. L = Ordinate value; U = Random component of the slope; R = Abscissa of the break point; 2SE = Standard Error; 3SSR = Sum of Squared Residual; 4CI = Confidence Interval

RESULTS AND DISCUSSION

In a dose-response experiment, SID Lys requirements were obtained at 1.29 and 1.15% for breast meat yield using QBL and LBL models respectively (Table 4). Sum of Squared Residuals (SSR) and R² were used to compare two models (Hernandez-Llamas, 2009; Pesti et al., 2009). Sum of squared residuals of linear and quadratic broken-line models for BM yield were 455.030 and 474.050, respectively. It represents clearly that QBL model fitted data better than LBL model. Also, R² of the QBL model for BM yield is higher than for LBL model regression. Quadratic broken-line model resulted in a wide confidence interval and in comparison with LBL model, QBL model covered higher range of the SID Lys requirements (Table 4). Depending on response criterion, R² and SSR for each model is variable. For example, regarding BWG, the value of SSR of LBL model is lower than value for QBL model, showing better fitting data with LBL model. Pesti et al. (2009) stated that model efficacy would be affected by response criterion. Using QBL model, SID Lys requirements for BM, BWG and THG yield were 1.29, 1.15 and 1.14%, respectively. Choosing different responses resulted in a variety of requirement values.

The extent of deficiency significantly affected the performance of broiler chicks during 14-28 days of age (Table 5). Significant effects of linear regression model revealed that the low lysine-basal diet was deficient from 0-14 days of age. Lysine deficiency significantly reduced the performance in starter and grower periods (p<0.05). The compensatory response was different among lowest and highest Lys-diets. Chickens fed 0.88% digestible lysine could not compensate growth in the following growth period at all. This effect was pronounced on BWG, FG and BM yield (Table 4). Linear regression analysis revealed that increasing dietary SID Lys in starter period resulted in better response in terms of BWG, FG, TGH and BM yield (p<0.05) during the growth period.

Table 5:	Growth performance of broiler chicks fed varying levels of digestible lysine feom days 14-28 of age
a-cValues within rows without a common superscript are significantly different (p≤0.05)

To maximizing profit feed formulation, the importance of choosing the correct model which relates animal response to feed composition has been documented (Lerman and Bie, 1975). The exact model will be functions of feed quality, feed intake, genetics and environmental conditions (Pesti et al., 2009). Among the nonlinear models, LBL has been used extensively for determination of amino acid requirements in the most species (Hernandez-Llamas, 2009). The most advantage of LBL model is the clearly definition of requirement for the performance variables (Pesti et al., 2009). Since, it assumes an unrealistic, abrupt shift at the plateau when a smooth transition actually occurs. So, the requirement would be underestimated. Alternatively, Robbins et al. (2006) used curve-and-plateau approach to estimate amino acid requirement. They applied quadratic equation in combination with straight-line equation at the plateau. Quadratic broken-line regression resolve the problem of underestimation resulted from LBL regression to some extent and ascending quadratic segment models diminishing marginal productivity until the level of the requirement is reached (Pesti et al., 2009). The results of this experiment are in agreement with Pesti et al. (2009) and Hernandez-Llamas (2009). To eliminating this underestimation, Booth modified LBL model so as to simulate a smooth curvilinear transition from ascending portion to the plateau and a breakpoint in the transition curve used for estimating requirement. They showed that diphasic model represents an improvement over the LBL model. As shown in Table 4, estimated SID Lys requirement using QBL model is 11% higher than value obtained from LBL model. Pesti et al. (2009) found 15% higher value using QBL model. Different dietary crude protein of basal diets and protein contributing ingredients of the experimental diets and also response criterion may be attributed to the difference between two experiments (Kerr et al., 1999; Leclercq, 1998; Pesti et al., 2009).

The SID Lys requirement for BM yield was higher than for BWG and THG yield (Table 4). Pesti and Miller (1997) suggested that the optimal levels of amino acids were not the same for the maximization of weight gain, feed conversion or breast muscle development. On other hand, the effect of dietary Lys on body composition has been demonstrated by several authors (Kidd et al., 1998; Kerr et al., 1999; Halevy et al., 2000; Moran and Bilgili, 1990; Leclercq, 1998). This effect is pronounced on the BM especially. Kerr et al. (1999) suggested that feeding dietary lysine above required levels for BWG improves BM deposition. Additionally, Acar et al. (1993) demonstrated that Pectoralis major is more sensitive to Lys intake other than muscles. This response also has been documented in the current study. Regardless of used model, lysine requirement for thigh yield is lower than for breast meat muscle synthesis (Table 4). It has been postulated that higher lysine requirement for BM yield and a large influence of dietary lysine on Pectoralis major may be related to the high lysine content of that muscle (Kerr et al., 1999; Acar et al., 1993). Labadan et al. (2001) estimated lysine requirement for BM yield at 1.32% of total amino acid in the diet which is lower than value estimated from this research. This difference may be related to the protein quality of the basal diet (Kidd and Fancher, 2001). Adverse effects of anti-nutritional factors in CSM and lower Lys digestibility of CSM would increase the needed Lys for normal growth and performance. Researchers used CSM in the basal diet and higher estimation of Lys requirement may be attributed to this event.

Halevy et al. (2000) found that reduction in muscle growth and low ability to mount a compensatory growth of chicken-fed lysine deficient diet may relate to the rapid maturation of fiber muscle in broilers and irreversible effect on fiber development. Kidd et al. (1998) also showed such effects of early malnutrition of lysine in broiler chicks and its detrimental influence on subsequent growth.

CONCLUSION

Estimation the SID Lys needs of broiler chicks using QBL model released more accurate results than LBL model and underestimation resulted from LBL model would be eliminated, since quadratic ascending portion of QBL regression models diminishing marginal productivity. The results of this research clearly showed that inadequate Lys nutrition of starting broiler chicks lead to decrease performance in the subsequent growth period and chickens fed very low-Lys diets could not mount a compensatory response in terms of BWG, FG and BM yield.