Use of Field Experiments in Soil Erosion Research

Experiments have a long tradition in geomorphology because of the need, or desire, to reduce complexity and thus simulate otherwise unattainable conditions. Sensu stricto, experiments involve controlled procedures that are carried out with the aim of testing a hypothesis. The challenge for designing an experiment in order to meet the expectations of the researcher lies in identifying its location in the triangle of precision versus realism versus generalism (e.g., Kuhn, 2014). By their own virtue, experiments aim to be more precise than nature, or rather, what can be observed in nature. This necessitates a reductionist, or simplified approach, which can frequently place the representativeness of the experiment, and thus the general relevance of the results to real-world scenarios into question (e.g., Bryan 1990). For instance, selecting a suitable plot size and rainfall intensity to perform rainfall simulations illustrates this challenge and will be one of the issues discussed in the case study presented in this chapter. In environmental sciences, the term “experiment” is commonly used in a somewhat wider context than has typically been applied in the more traditional, or “natural,” sciences. The reason for this is the compromise that has to be made in order to meet a position in the precision–realism–generalism triangle that generates an answer to the question being addressed by the experiment (Slaymaker, 1991). Further limitations arise from the field situation, as these can typically limit the level of control when compared to a laboratory experiment. Therefore, three broad aims of field experiments in geomorphology can be identified: 1. Actual experiments that aim to test one or more hypotheses on the interaction of one or more components within a landscape system. 2. Process-rate measurements in the field in order to quantify a conceptual model or test the relevance of a particular process in a given landscape system.