Higham C D, Horne D, Singh R, Kuhn-Sherlock B, Scarsbrook M R
DairyNZ-Policy and Advocacy, Private Bag 3221, Hamilton 3240, New Zealand; Institute of Agriculture and Environment, Massey University, Private Bag 11222, Palmerston North 4474, New Zealand.
Institute of Agriculture and Environment, Massey University, Private Bag 11222, Palmerston North 4474, New Zealand.
J Dairy Sci. 2017 Aug;100(8):6772-6784. doi: 10.3168/jds.2016-12200. Epub 2017 Jun 16.
Robust information for water use on pasture-based dairy farms is critical to farmers' attempts to use water more efficiently and the improved allocation of freshwater resources to dairy farmers. To quantify the water requirements of dairy farms across regions in a practicable manner, it will be necessary to develop predictive models. The objectives of this study were to compare water use on a group of irrigated and nonirrigated farms, validate existing water use models using the data measured on the group of nonirrigated farms, and modify the model so that it can be used to predict water use on irrigated dairy farms. Water use data were collected on a group of irrigated dairy farms located in the Canterbury, New Zealand, region with the largest area under irrigation. The nonirrigated farms were located in the Manawatu region. The amount of water used for irrigation was almost 52-fold greater than the amount of all other forms of water use combined. There were large differences in measured milking parlor water use, stock drinking water, and leakage rates between the irrigated and nonirrigated farms. As expected, stock drinking water was lower on irrigated dairy farms. Irrigation lowers the dry matter percentage of pasture, ensuring that the amount of water ingested from pasture remains high throughout the year, thereby reducing the demand for drinking water. Leakage rates were different between the 2 groups of farms; 47% of stock drinking water was lost as leakage on nonirrigated farms, whereas leakage on the irrigated farms equated to only 13% of stock drinking water. These differences in leakage were thought to be related to regional differences rather than differences in irrigated versus nonirrigated farms. Existing models developed to predict milking parlor, corrected stock drinking water, and total water use on nonirrigated pasture-based dairy farms in a previous related study were tested on the data measured in the present research. As expected, these models performed well for nonirrigated dairy farms but provided poor predictive power for irrigated farms. Partial least squares regression models were developed specifically to simulate corrected stock drinking water, milking parlor water, and total water use on irrigated dairy farms.
对于以牧场为基础的奶牛场而言，获取关于用水的可靠信息对于农民提高用水效率以及更合理地分配淡水资源给奶牛场至关重要。为了以切实可行的方式量化各地区奶牛场的用水需求，有必要开发预测模型。本研究的目的是比较一组灌溉农场和非灌溉农场的用水情况，利用在非灌溉农场测得的数据验证现有的用水模型，并对该模型进行修改，使其能够用于预测灌溉奶牛场的用水情况。用水数据是在新西兰坎特伯雷地区一组灌溉奶牛场收集的，该地区灌溉面积最大。非灌溉农场位于马纳瓦图地区。用于灌溉的水量几乎是所有其他形式用水总量的52倍。灌溉农场和非灌溉农场在挤奶厅用水量、牲畜饮用水量和渗漏率方面存在很大差异。正如预期的那样，灌溉奶牛场的牲畜饮用水量较低。灌溉降低了牧场的干物质百分比，确保全年从牧场摄入的水量保持较高水平，从而减少了饮用水需求。两组农场的渗漏率不同；非灌溉农场有47%的牲畜饮用水因渗漏而流失，而灌溉农场的渗漏量仅相当于牲畜饮用水量的13%。这些渗漏差异被认为与地区差异有关，而非灌溉农场与灌溉农场之间的差异。在先前一项相关研究中开发的用于预测非灌溉牧场型奶牛场挤奶厅、校正后的牲畜饮用水量和总用水量的现有模型，在本研究测得的数据上进行了测试。不出所料，这些模型对非灌溉奶牛场表现良好，但对灌溉农场的预测能力较差。专门开发了偏最小二乘回归模型来模拟灌溉奶牛场校正后的牲畜饮用水量、挤奶厅用水量和总用水量。
