摘要

Two experiments were conducted to evaluate the nutritive value and maximum safe level (MSL) of pennycress meal (PM) for broiler chicks. In experiment 1, a total of 480 chicks was fed either mash or crumbled diets containing zero, 5, 10, or 15% PM for 18 d (8 diets; 6 replications per diet). In experiment 2, a total of 660 chicks was fed mash diets containing zero, 3, 6, 9, 12, or 15% of either PM or canola meal (CM; a comparative reference) for 14 d (11 diets; 6 replications per diet). Analytical results show that PM is a good source of protein (similar to 31% CP) and it is very comparable to CM (similar to 36% CP). However, it contains higher erucic acid (similar to 1.68 vs. < 0.021%), glucosinolates (sinigrin) (similar to 63.5 vs. < 0.163 mu mol /g), and crude fiber (18.60 vs. 9.27%) compared to CM. In experiment 1, increasing PM from zero to 15% resulted in linear reductions (P < 0.05) in FI, BWG, and FCR at 10 days. Above 10%, performance responses were affected for FI and BWG at 18 d, respectively. An estimated MSL of 10% PM based on orthogonal contrast was optimal for satisfactory FI and BWG. The MSL as estimated by brokenline linear (BLL) and broken-line quadratic (BLQ) models was 9.12 +/- 0.50 and 7.0 +/- 1.27%, respectively. In experiment 2, growth performance at 14 d was reduced above 9% due to PM inclusion. CM inclusion did not affect growth performance at 14 d, suggesting 15% to be safe. The MSL for maximum growth performance varied depending on the statistical analysis as follows: 12% by orthogonal contrast and LSD, 15% by the Scheffe test, 10.84 +/- 0.57 by BLL, and 8.61 +/- 1.29 by BLQ. In conclusion, PM can be included in broiler starter diets as a protein source but its inclusion should be limited to no more than 8.5%. Different statistical procedures give different MSL and this should be considered when interpreting the data.