Variability in soil erosion data from replicated plots

Understanding and quantifying the large, unexplained variability in soil er osion data are critical for advancing erosion science, evaluating soil eros ion models, and designing erosion experiments, We hypothesized that it is p ossible to quantify variability between replicated soil erosion field plots under natural rainfall, and thus determine the principal factor or factors which correlate to the magnitude of the variability: Data from replicated plot pairs for 2061 storms, 797 annual erosion measurements, and 53 multi-y ear erosion totals were used. Thirteen different soil types and site locati ons were represented in the data. The relative differences between replicat ed plot pair data tended to be lesser for greater magnitudes of measured so il loss, thus indicating that soil loss magnitude was a principal factor fo r explaining variance in the soil loss data. Using this assumption, we esti mated the coefficient of variation of within-treatment, plot replicate valu es of measured soil lass. Variances between replicates decreased as a power function (r(2) = 0.78) of measured soil loss, and were independent of whet her the measurements were event-, annual-, or multi-year values. Coefficien ts of variation ranged on the order of 14% for a measured soil loss of 20 k g/m(2) to greater than 150% for a measured soil loss of less than 0.01 kg/m (2) These results have important implications for both experimental design and for using erosion data to evaluate prediction capability for erosion mo dels.