Manual sampling and tank size effects on the calibration curve of plot sediment storage tanks

In many experimental soil erosion plots, runoff is collected and carried by a conveyance system to a sequence of storage tanks. If the soil loss is measured by collecting, after mixing, samples of the stored suspension, then a calibration curve between the actual mean concentration (C) and the measured concentration (Cm) in the storage tank occurs. The aim of this article was to evaluate experimentally the factors affecting the relationship between C and Cm. For a sandy loam soil, the replicated measurements of Cm (20 samples) for two values of the actual concentration (C = 5 and 25 g/L) showed that the variability of the measurements of Cm is low and confirmed the reliability of a calibration curve obtained by a single series of runs. Results from experiments carried out with a clay soil to compare the calibration curves obtained by four field workers suggested that the maximum uncertainty in the soil loss measurement due to the choice of the calibration curve should not exceed 100% of the true value. Moreover, the slope of the calibration curve was independent of both the water level in the tank and the field worker. Finally, a comparison among the calibration curves of a prototype tank and some model tanks was carried out for both a sandy loam and a clay soil in order to establish a scaled−up relationship among tanks of different size. Soil−specific and theoretically based scaled−up relationships were deduced.