Measuring soil erosion by field plots: Understanding the sources of variation

Abstract

Soil erosion plots of different types and sizes are widely used to investigate the geomorphological processes related to soil erosion. This field method has provided a variety of results, depending on the characteristics of the plots, on their suitability to reflect the ecosystem's characteristics and on the objectives of each particular research. The coupling of real soil loss at patch and slope scale within a landscape and the values obtained by field plots depend, among other things, on how good the methodology performs over a set of ecosystem properties, such as those related with temporal and spatial scale issues, disturbance and representation of natural conditions, and the ability to account for the complexity of ecosystem interactions. Here, we present a review of (i) the advantages and limitations of the use of field plots to measure soil erosion; and (ii) the potential sources of variation in the results obtained due to a lack of harmony between methodological conditions and the processes operating in the environment at different scales.

As regards the spatial and temporal scale of measurements, topics such as the exhaustion of available material within closed plots in long term measurements, the different erosion processes operating (and measured) at different spatial scales and the problems and alternatives of extrapolation of the results from larger to smaller scales, are the main causes of variation between measurements.

The disturbance and inadequate representation of natural conditions, such as the heterogeneity, continuity and connectivity of factors and processes, are also sources of variation in the results of specific measurements. In short, the key factor is the difficulty to encapsulate the complexity of system interactions and to represent these interactions by means of field plots. The complexity concept is translated in the connectivity of water and sediment fluxes in the landscape and the interaction between processes and patterns of vegetation and surface components operating across scales.

Introduction

During the last twenty years many projects have been carried to quantify erosion rates out in different type of ecosystems. The results have provided a large database that allows the specific achievements to be described at a more conceptual level. Reflections of this type were presented by several authors working in different aspects of soil erosion processes and at different scales (Lavee et al., 1998, Puigdefábregas et al., 1999, Cammeraat, 2002, Calvo-Cases et al., 2003, Cammeraat, 2004, Brazier, 2004, Puigdefábregas, 2005, de Vente and Poesen, 2005, Boix-Fayos et al., 2005, Boix-Fayos et al., in press), and in extended reviews on collected soil erosion data (Cerdà, 2001, Romero Díaz, 2002). All such reflections are important since they permit us to learn from past experience and to delineate future bases of research.

Analysis of all the data available points to a large variation in the erosion rates measured and also to the large variation in methods used. Fig. 1 for example shows just some of the types and size of plot used in our own research.

Part of this variation can be explained by the variability of natural conditions (natural variability) and partly by the disturbance caused by the experimental design of field measurements (measurement variability). These concepts were previously introduced by Bagarello and Ferro (1998) and Nearing et al. (1999).

In general, there is a demand for knowledge of the soil erosion processes occurring in plots of different sizes and of factors that determine natural variability, as a basis for obtaining soil loss data of good quality (Bagarello and Ferro, 2004). Among the causes of measurement variability, human disturbance in collecting the data has been reported to be an important source of variation (Zobisch et al., 1996), because of the differences in accuracy and consistency of the way of operating by human hands.

Several issues are crucial as far as the suitability and performance of a field experimental design is concerned: (i) temporal and spatial scales, (ii) representation of natural conditions, (iii) the disturbance of natural conditions and (iv) account of the complexity of ecosystem interactions (Table 1). In this paper the limitations and advantages of using plots are approached from the perspective of these issues, all of which overlap and have as their basis two concepts: spatial scale and thresholds, and it is these concepts which determine the behaviour of the processes involved in the erosion response.

This paper sets out to examine the suitability of using field plots to measure soil loss, taking into consideration the functioning of natural processes, as influenced by the characteristics of the plots themselves. The objectives of the paper are to review the main sources of variation in data due to the use of field plots to measure soil erosion and to analyse the processes responsible for this variation.