Van Til L D, O'Rourke R L, Dohoo I R
Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown.
Can J Vet Res. 1991 Jul;55(3):277-84.
Regression analysis was used to determine the ability of a number of biological parameters to predict economic efficiency. Detailed feed, financial, and production records were maintained by a random sample of eighteen Prince Edward Island (PEI) swine producers (each producing over 1000 market hogs per year). Relative economic efficiency of the operations was measured using return to management and labor (RML). Of the routinely monitored biological parameters, RML on PEI farrow-finish operations was best predicted (R2 = 64.8%) by: marketed per square meter per year (p = 0.008) and marketed per sow per year (p = 0.096). Regression of fixed costs revealed that biological parameters had limited ability to predict fixed costs per hog on farrow-finish operations (R2 = 30.7%). The only parameter contributing to the prediction of the fixed cost component of RML was feeder hog density (p = 0.077). The variable cost component of RML on farrow-finish operations was predicted (R2 = 94.3%) by feed cost per kg gain (p = 0.000), and marketed per sow per year (p = 0.044). The routinely recorded biological parameters on feeder farms had only limited ability to predict RML in this study (R2 = 43.7%). The only parameter of any importance was marketed per square meter per year (p = 0.106). Prediction of the fixed cost component of RML on feeder farms (R2 = 67.4%) was best realized by measuring feeder hog density (p = 0.045). The variable cost component of RML on feeder farms was reasonably well predicted (R2 = 74.7%) by feed cost per kg gain (p = 0.012). Although this parameter is difficult to monitor from records currently maintained on most farms, it points out the need to monitor feed consumption on swine farms.
回归分析用于确定一些生物学参数预测经济效率的能力。爱德华王子岛（PEI）的18个养猪生产者（每人每年生产超过1000头上市肥猪）的随机样本保存了详细的饲料、财务和生产记录。使用管理和劳动力回报率（RML）衡量这些养殖场的相对经济效率。在常规监测的生物学参数中，PEI仔猪育肥场的RML最佳预测指标（R2 = 64.8%）是：每年每平方米上市猪数（p = 0.008）和每年每头母猪上市猪数（p = 0.096）。固定成本回归分析表明，生物学参数预测仔猪育肥场每头猪固定成本的能力有限（R2 = 30.7%）。对RML固定成本部分预测有贡献的唯一参数是育肥猪饲养密度（p = 0.077）。仔猪育肥场RML的可变成本部分可通过每千克增重的饲料成本（p = 0.000）和每年每头母猪上市猪数（p = 0.044）进行预测（R2 = 94.3%）。在本研究中，育肥猪场常规记录的生物学参数预测RML的能力有限（R2 = 43.7%）。唯一重要的参数是每年每平方米上市猪数（p = 0.106）。通过测量育肥猪饲养密度（p = 0.045）能最好地实现对育肥猪场RML固定成本部分的预测（R2 = 67.4%）。育肥猪场RML的可变成本部分可通过每千克增重的饲料成本得到较好预测（R2 = 74.7%）（p = 0.012）。尽管该参数难以从大多数农场目前保存的记录中监测，但它指出了养猪场监测饲料消耗的必要性。
