Singh A, Bicudo J R
Department of Biosystems and Agricultural Engineering, University of Kentucky 128 CE Barnhart Building, Lexington, KY 40546, USA.
Environ Technol. 2005 May;26(5):471-8. doi: 10.1080/09593332608618541.
Rapid on-farm assessment of manure nutrient content can be achieved with the use of quick tests. These tests can be used to indirectly measure the nutrient content in animal slurries immediately before manure is applied on agricultural fields. The objective of this study was to assess the reliability of hydrometers, electrical conductivity meter and pens, and Agros N meter against standard laboratory methods. Manure samples were collected from 34 dairy farms in the Mammoth Cave area in central Kentucky. Regression equations were developed for combined and individual counties located In the area (Barren, Hart and Monroe). Our results indicated that accuracy in nutrient estimation could be improved if separate linear regressions were developed for farms with similar facilities in a county. Direct hydrometer estimates of total nitrogen were among the most accurate when separate regression equations were developed for each county (R2 = 0.61, 0.93, and 0.74 for Barren, Hart and Monroe county, respectively). Reasonably accurate estimates (R2 > 0.70) were also obtained for total nitrogen and total phosphorus using hydrometers, either by relating specific gravity to nutrient content or to total solids content. Estimation of ammoniacal nitrogen with Agros N meter and electrical conductivity meter/pens correlated well with standard laboratory determinations, especially while using the individual data sets from Hart County (R2 = 0.70 to 0.87). This study indicates that the use of quick test calibration equations developed for a small area or region where farms are similar in terms of manure handling and management, housing, and feed ration are more appropriate than using "universal" equations usually developed with combined data sets. Accuracy is expected to improve if individual farms develop their own calibration curves. Nevertheless, we suggest confidence intervals always be specified for nutrients estimated through quick testing for any specific region, county, or farm.
使用快速检测方法能够快速在农场对粪便养分含量进行评估。这些检测可用于在粪便施用于农田之前直接间接测量动物粪便中的养分含量。本研究的目的是评估液体比重计、电导率仪和笔式测试仪以及Agros N仪相对于标准实验室方法的可靠性。从肯塔基州中部猛犸洞地区的34个奶牛场采集了粪便样本。针对该地区（巴伦、哈特和门罗）的各个县以及合并后的县建立了回归方程。我们的结果表明，如果为一个县内设施相似的农场分别建立线性回归方程，养分估算的准确性可以提高。当为每个县分别建立回归方程时，液体比重计对总氮的直接估算最为准确（巴伦、哈特和门罗县的R2分别为0.61、0.93和0.74）。通过将比重与养分含量或总固体含量相关联，使用液体比重计对总氮和总磷也能获得较为准确的估算（R2 > 0.70）。使用Agros N仪以及电导率仪/笔式测试仪对氨氮的估算与标准实验室测定结果相关性良好，尤其是使用哈特县的单独数据集时（R2 = 0.70至0.87）。本研究表明，针对粪便处理与管理、畜舍和饲料配给相似的小区域或地区开发的快速检测校准方程，比使用通常基于合并数据集开发的“通用”方程更为合适。如果各个农场自行绘制校准曲线，预计准确性会有所提高。尽管如此，我们建议对于通过快速检测估算的任何特定区域、县或农场的养分，始终指定置信区间。
