The aim of the present study was to assess whether near infrared reflectance spectroscopy (NIRS) and re-gression equations are the practical and accurate approach of nutritional assessment of common feedstuffs. Therefore two experiments were conducted to study the effect of amino acid determination methods on broiler performance. In experiment I, two hundred thirty four male Ross broiler chicks were fed a commercial starter diet from d 1 to 16 of age followed by the test diets from d 17 to 21 to determine the digestible amino acids content of wheat, corn and soybean meal. The 3 test diets consisted of corn, wheat or soybean meal was used as the sole sources of protein. On d 21, birds were killed and ileal digesta were collected for determining standardized ileal amino acid digestibility (SID). The in vivo values of SID were compared with values predicted by NIRS and regression methods. Experiment II was conducted to compare the effect of feed formulation based on SID determined by in vivo with regression and NIRS prediction methods on broiler performance. The results showed that the SID contents of wheat and corn predicted by NIRS and regression method were close to in vivo values. Values predicted by NIRS and regression for soybean meal varied when compared with those of in vivo assay. Although, these differences were not appeared when comparing performance of broiler fed diets formulated based on SID values estimated through in vivo trial, NIRS and regression methods. This study reveals that feed formulation based on digestible amino acids estimated by NIRS and regression methods are reliable for feed formulation in commercial poultry production. 

Keywords