本试验旨在建立单胃动物仿生消化系统模拟猪饲料原料消化后还原糖释放量的测定方法，为饲料养分生物学效价的评估提供参考。还原糖释放量与上样量的线性关系研究以玉米-豆粕混合物（75%玉米+25%豆粕）为研究对象，上样量设0.2、0.4、0.6、0.8和1.0 g 5个处理，每个处理4个重复，每个重复1根消化管；方法的重复性检验以玉米-豆粕混合物、大麦、花生粕和米糠为研究对象，每个样品设3个批次，每个批次4个重复；方法的可加性检验设19个处理，其中处理1~7分别为玉米、大麦、高粱、大豆粕、花生粕、棉籽粕和米糠，处理8~19为2种或2种以上饲料原料按不同比例组合制备的12种混合饲料，每个处理4个重复，每个重复1根消化管，在仿生消化系统中模拟猪消化后测定各处理的还原糖释放量。结果表明：上样量在0.2~0.8 g时，还原糖总释放量与上样量呈显著的线性关系（R²=0.999 2），还原糖相对释放量在559.56~582.70 mg/g DM变化，变异系数为1.66%，而上样量为1.0 g时，还原糖相对释放量比上样量为0.2~0.8 g时的平均值下降5.37%；3个批次的大麦、花生粕、米糠和玉米-豆粕混合物的还原糖释放量的批内变异系数、批间变异系数和总变异系数均不大于1.68%，批间最大相对偏差分别为0.68%、1.50%、1.39%和0.29%；12种混合饲料还原糖释放量的实测值显著高于计算值（P<0.05），而还原糖释放量的计算值与实测值的线性回归模型与y=x相重合（截距P=0.480 5；斜率P=0.514 1）。由此得出，当上样量在0.2~0.8 g时，上样量与还原糖释放量呈显著线性关系；仿生消化法测定猪饲料原料还原糖释放量的重复性和可加性满足定量分析的基本要求。

This study was conducted to establish the method for determining the released amount of reducing sugar of feed ingredients digested with simulated digestion system to mimic the digestion of pigs, which provided a reference for the evaluation of feed nutrient bioavailability. A corn-soybean meal mixture (75% corn+25% soybean meal) was formulated to test the linear relationship between the released amount of reducing sugar and sample volume. Five treatments of sample volume were consisted of 0.2, 0.4, 0.6, 0.8 and 1.0 g. Each treatment contained 4 replicates with 1 digestion tube per replicate. Corn-soybean meal mixture, barley, peanut meal and rice bran were used to test the repeatability of the method. Each sample was determined for 3 batches with 4 replicates in each. There were 19 treatments to test the additivity of the method. Treatments 1 to 7 were corn, barley, sorghum, soybean meal, peanut meal, cottonseed meal and rice bran, respectively. Treatments 8 to 19 were 12 mixed feeds formulated with 2 or more than 2 kinds of feed ingredients at different proportions. Each treatment contained 4 replicates with 1 digestion tube per replicate. The released amount of reducing sugar of each treatment was determined after the feed was digested with simulated digestion system to mimic the digestion of pigs. The results showed as follows:when the sample volume ranged from 0.2 to 0.8 g, a significant linear relationship (R²=0.999 2) was observed between the total released amount of reducing sugar and sample volume, and the relative released amount of reducing sugar ranged from 559.56 to 582.70 mg/g DM with a coefficient of variation (CV) of 1.66%. However, the released amount of reducing sugar of 1.0 g sample was 5.37% lower than the mean of 0.2 to 0.8 g sample. The CV of intra-batch, inter-batch and total for the released amount of reducing sugar were all less than 1.68%, and the inter-batch maximum relative difference among repeated determinations were 0.68%, 1.50%, 1.39% and 0.29% for barley, peanut meal, rice bran and corn-soybean mixture, respectively. The determined values were greater than the calculated values of the released amount of reducing sugar in 12 mixed feeds (P<0.05). However, the linear regression model of calculated values and determined values of the released amount of reducing sugar was consistent to the line of y=x (P=0.480 5 for intercept; P=0.514 1 for slope). It is concluded that a significant linear relationship is found between sample volume and the released amount of reducing sugar when the sample volume ranges from 0.2 to 0.8 g; the repeatability and additivity of the released amount of reducing sugar of feed ingredients determined with simulated digestion method for pigs can meet the basic requirements of quantitative analysis.