A comparative analysis of the pore system in COST 622 volcanic soils by means of water retention measurements and image analysis

The quality and quantity of the soil pore system govern life in many terrestrial ecosystems. In this regard the case of volcanic soils, and especially Andosols, is of major interest since the unique properties of these soils, such as high water retention, low bulk density, high smeariness, etc., mainly depend on the overall organisation of the pore system. Given such importance, this chapter addresses the characterisation of the porous system of the COST action 622 reference soils by means of twowell established techniques, namely the water retention curve and image analysis. This work will also attempt to address the complex issue of comparing the results obtained by these two methods. Water retention measurements on volcanic soils were carried out, especially in soils of the circum-Pacific region, by Misono et al. (1953) for Japanese soils, Colmet-Daage et al. (1967, 1970) for soils from the Carribean, Central America and South America. Most of these measurements were performed on few points of the water retention curve, generally 30 and 1500 kPa (Maeda et al. 1977, Nanzyo et al. 1993). In Europe, only recently, detailed water retention characteristics were measured on volcanic soils, although not emphasizing their distinct properties. Ciollaro and Romano (1995) combined geostatistics and an inverse method to derive hydraulic properties on a cultivated volcanic soil transect; Basile et al. (2003a) applied the hysteresis concept to the fieldlaboratory comparison. More recently few authors took into account the distinct behaviour and characteristics of volcanic soils: Basile and De Mascellis (1999) reported results on the irreversible drying effect on water retention and transport parameters, Basile et al. (2003b) applied a water flow deterministic model to evaluate risk of debris flow triggering, Armas-Espinel et al. (2003) related hydraulic properties to specific andic properties of cultivated soils, Fontes et al. (2004) compared hydraulic properties in volcanic soils with different methods, and Ritter et al. (2004) analysed measurement strategies for the inverse optimization of the hydraulic properties of a volcanic soil. Soil pores range in size over several orders of magnitude. Many measurement methods are available in order to quantify pore size distribution, each giving best results in a specific pore size range. Nitrogen sorption at –196°C (Sills et al. 1973), for example, is a well-established method for determining specific surface area and pore size distribution in the range below 20 nm. Mercury intrusion porosimetry can rapidly provide pore size distribution ranging from 10 nm to 100 mm (Pagliai 1988). Soil water retention measurements allow estimation of equivalent pore size distribution between 200 nm to 1 mm. Image analysis methods can also be applied after using very different techniques (fluorescent resin impregnation, x-ray Tomography, NMR, etc.) which allow pore space to be visualized. In this latter case size limits depend on which technique is used. Image analysis methods are generally laborious and time consuming but they can provide details about shape and arrangement of pore space that are not given by other methods (Marshall et al. 1996). Nevertheless, a pore system cannot be adequately characterized by using a single method (Lawrence 1977). Only a comparison of different types of measurements can give more complete representation of the soil structure. Moreover, measurements of the pore size distribution in volcanic soils taking into account their distinctive properties are scarce (Mele et al. 2000). In the present chapter pore size distributions from water retention curves and from image analysis of resin impregnated soil blocks are compared for the representative soils of the COST action 622.