TY - JOUR T1 - The accuracy of manual runoff and sediment sampling from erosion plots JF - Journal of Soil and Water Conservation SP - 231 LP - 233 VL - 51 IS - 3 AU - M. A. Zöbisch AU - P. Klingspor AU - A. R. Oduor Y1 - 1996/05/01 UR - http://www.jswconline.org/content/51/3/231.abstract N2 - Evaluating the effects of different soil and crop management practices on soil loss is largely based on soil erosion plot experiments. However, most soil erosion plots in developing countries are simple, low-cost installations. Runoff and soil loss are sampled manually, and their accuracy depends on the sampling procedures adopted and the diligence of the field staff. A laboratory in Kenya compared the manual sampling accuracy of field staff in charge of erosion plots at the Steep Lands Research Station of the University of Nairobi. The runoff volume was generally slightly under-estimated, but sampling accuracy was acceptable for manual sampling; the mean square error was 5.6%, with a range of 3.6-6.7%. Marked differences between the performances of the field staff were not observed. However, the accuracy of soil loss measurements varied significantly. The mean square sampling errors ranged from 4.7 to 83.0%, with an average of 41.3%, unacceptably poor accuracy. Only two of the five field staff participating in the tests produced acceptable results. To ensure adequate sampling performance, appropriate sampling methods should be designed, training of field staff in sampling techniques should be provided, and regular performance testing should be implemented. ER -