Impact of variable corn nutrient content, AME prediction, and xylanase inclusion on growth performance

An experiment was conducted to investigate the effect of corn apparent metabolizable energy (AME) value, on broiler performance associated with geographical location (source), xylanase inclusion, and formulation. A 3 (corn source) × 2 (formulation; adjusted or non-adjusted value) × 2 (xylanase inclusion, 0 vs 16,000 BXU/kg) factorial randomized complete block desig study was conducted which included 10 replicates of 20 male broilers per replicate for a 42 d experiment. In the non-adjusted corn diets, each dietary treatment contained the same percentage of each corn source while the adjusted corn diets formulated based on a corn near-infrared (NIR) nutrient analysis and AME prediction equation (AB Vista, Marlborough, UK). In all cases, the predicted AME of the corns was lower than the NRC (1994) AME value, resulting in additional fat inclusion. The dietary program consisted of 3 dietary phases; starter (d 1 to 18), grower (d 19 to 33), and finisher (d 34 to 42) with growth performance monitored at feed changes. Xylanase increased (p < 0.05) starter feed consumption and d 18 body weight (BW) (2.1%), and an increase in BW (4.6%) was observed using the predicted AME values. Corn source continued to impact broiler performance throughout the remainder of the trial as corn source A consistently outperformed (p < 0.05) corn source C with increased BW and reduced mortality corrected FCR. Dietary adjustment with AME predicted values for each corn source consistently improved (p < 0.05) growth performance. Corn source A had the highest average BW with a 2.9% increase (p < 0.05) over corn source C. Early feed conversion rate (FCR) was improved (p < 0.05) with the predicted AME value diet-fed broilers, decreasing FCR by 3%. Multiple significant interactions were observed between the corn source and the adjustment method. For example, with d 42 FCR, using the predicted AME value adjustment had an impact of 9 and 10 points for corn source B and C, respectively, but only 2 for corn source A. This experiment demonstrates the impact of variable corn nutrient content and the potential improvement when using NIR technology for ingredient nutrient specifications.