The measurement of soil loss from fractional acre plots

Abstract

The use of runoff plots in "Observational" and "Experimental" studies is reviewed and discussed. The basic requirements for a sound experimental design in each type of study are described. It is noted that most studies have been inadequately designed. The principal shortcomings have been the absence of replication and randomisation of treatments, and the inherit assumption that bias is unimportant. The paper identifies several possible sources of bias which must be accounted for before plot data can be extrapolated to field conditions. The equipment most commonly used in plot studies is briefly reviewed. It is suggested that the plot method is probably a fairly crude one, and the use of precision equipment may only give precise measurement of inaccurate values.