Elsevier

CATENA

Volume 53, Issue 2, 1 September 2003, Pages 97-114
CATENA

Influence of rock fragments on the water retention and water percolation in a calcareous soil

https://doi.org/10.1016/S0341-8162(03)00037-7Get rights and content

Abstract

The water retention properties of a calcareous soil containing rock fragments have been determined in the laboratory thanks to pressure plate measurements done on both the fine earth and the rock fragments from the soil. The available water content (AWC) has been calculated from these data. We have shown that when the rock fragments are neglected, the AWC can be overestimated by 39%. When we do not neglect their volume but when their hydraulic properties are not considered, the AWC can be underestimated by 34%. By using a reservoir model, we have also calculated the effect of rock fragments on water percolation to groundwater. Depending of the climatic characteristics of the year, the underestimation of percolation when we neglect the rock fragments can reach up to 14.9% and the overestimation when we neglect their hydraulic properties can be equal to 15.8%. These findings emphasise the role of the rock fragments on the water supply in stony soils.

Introduction

Soils containing rock fragments represent about 30% of the Western Europe surface area and even 60% in the Mediterranean zone (Poesen and Lavee, 1994). Due to workability and trafficability problems, they are less used for agronomic production and, as a consequence, they have been studied less frequently. Nevertheless, the presence of rock fragments modifies (i) the soil physical properties: available water content, infiltration and runoff susceptibility, for example; (ii) the soil chemical properties: carbon content or nitrogen content; and (iii) the agronomical characteristics like the yields. Most studies dealing with stony soils usually do not take into account the rock fragments, even if their abundance cannot be neglected. As a consequence, the soil properties are not correctly evaluated, overestimated or underestimated, when only the fine fraction of the soil is considered (Ugolini et al., 1998).

As far as available water content (AWC) is concerned, Coile (1953) first showed that the calculated water available for plants is higher if we take into account the rock fragments, thus illustrating the water retention properties of the rock fragments themselves (Coile, 1953). However, AWC of soils containing rock fragments depends on several parameters: (i) the origin of the rock fragments, (ii) the volumetric percentage of the rock fragments, (iii) the size and the porosity of the rock fragments, and finally (iv) the position of the rock fragments. Indeed, the origin of the rock fragments has a large influence on the water holding capacity: when rock fragments come from chalk, it can be larger than 90% of the saturation range; whereas it is nearly equal to zero when the rock fragments come from basalt fragments (Poesen and Lavee, 1994). The volumetric percentage of the rock fragments in the soil usually has a negative influence on the AWC: the available water for plants generally decreases when the percentage of rock fragments increases but this is not always true, for example for soils developed on schist (Hanson and Blevins, 1979). The size and the porosity of the rock fragments play an important role in their water retention properties as well: the smaller the rock fragments, the more weathered (Childs and Flint, 1990) and the more they can store water Hanson and Blevins, 1979, Poesen and Lavee, 1994. Finally, the position of the rock fragments, on the surface or inside the soil profile, free on the surface or embedded in the fine earth, has a large influence on the total water properties of the soil, especially because the rock fragments modify the evaporation conditions at the surface Jury and Bellantuoni, 1976, Perez, 1998. It has been shown that a 5 cm gravel mulch can reduce the annual evaporation by 85% (Kemper et al., 1994) and that under Mediterranean conditions, the rock fragments can keep a higher water content in the soil than when they are not present (Danalatos et al., 1995). On the contrary, under strong evaporation conditions, the high calorific characteristics of the rock fragments leads to heating of the soil and therefore to a decrease of its water content.

As far as hydrodynamic characteristics of soils containing rock fragments are concerned, two types of studies have been conducted: on the one hand, some authors have discussed the effect of rock fragments on infiltration as the complementary part of runoff on erosion studies; and on the other hand, they have tried to measure or model directly the hydraulic conductivity of stony soils.

The presence of rock fragments at the soil surface can either increase or decrease infiltration (Brakensiek and Rawls, 1994). Grant and Struchtemeyer (1959) have shown in laboratory runoff experiments that the removal of rock fragments leads to a decrease in infiltration because part of the porosity has disappeared and the soil surface is not prevented from sealing by these rock fragments (Grant and Struchtemeyer, 1959). In fact, the effect of rock fragments on the soil surface strongly depends of their position: when free on the surface, they generally prevent the soil from sealing and the infiltration increases, but embedded in the surface, they participate to the establishment of a continuous crust inhibiting infiltration and reinforce runoff (Poesen and Lavee, 1994). As for water retention properties, the size and the form of rock fragments has to be highlighted: generally, the infiltration in stony soils increases for small rock fragments but there exists a threshold (Valentin and Casenave, 1992). Above this threshold, infiltration decreases because of the less accessible surface for water flow (Valentin, 1994). Furthermore, the more spherical rock fragments are, the lower the saturated hydraulic conductivity (Dunn and Mehuys, 1984). Finally, most of the studies dealing with infiltration in soils containing rock fragments were restricted to arid or Mediterranean zones.

Laboratory experiments on disturbed soils containing rock fragments have shown that soil saturated hydraulic conductivity can be estimated from the fine earth saturated hydraulic conductivity and the volumetric percentage of the non-porous rock fragments (Mehuys et al., 1975). The presence of non-porous fragments only reduces the available area for the flow and increases the tortuosity of the water flow because pockets under the rock fragments have to be wetted (Gras, 1972). But even if tortuosity increases, the creation of new voids in the stone–earth contact, useful for water flow, can increase the saturated hydraulic conductivity with increasing rock fragments content (Ravina and Magier, 1984). Based on the heat transfer theory, a formula has been calculated for a homogeneous medium containing non-porous spherical inclusions (Peck and Watson, 1979) to calculate the hydraulic conductivity of a stony soil from the hydraulic conductivity of the fine earth and the volumetric percentage of rock fragments:Ksoil/Kfe=2(1−Rv)/(2+Rv)where Ksoil (resp. Kfe) represents the hydraulic conductivity of the soil (resp. of the fine earth) and Rv is the volumetric fraction of the rock fragments. A first comparison with experimental data has shown that this formula overestimated the hydraulic conductivity for high water content (Bouwer and Rice, 1984) and it has then be simplified to:Ksoil/Kfe=1−Rmwhere Rm represents the mass fraction of the rock fragments (Brakensiek et al., 1986). Nevertheless, whatever equations are used to estimate the saturated hydraulic conductivity in stony soils, they deal only with non-porous rock fragments and, even in that case, no general relation has been found. For soils containing porous rock fragments, Gras (1972) has pointed to the air encapsulation phenomenon inside the porous fragments which modifies the soil hydraulic conductivity. Furthermore, to our knowledge, no experimental studies has dealt with hydraulic conductivity outside the saturation range. Indeed, field experiments in stony soils are quite impossible because it is really difficult to install either tensiometers or lysimeters, for example. Direct measurement of the infiltration are still a problem. Moreover, as already mentioned by Mehuys et al. (1975), sampling of undisturbed soil cores containing rock fragments remains difficult. Laboratory measurements of unsaturated hydraulic conductivity are therefore not possible. In that context, the calculation of infiltration in a soil containing rock fragments with a deterministic model that would use explicit unsaturated hydraulic conductivity data seemingly remains elusive up to the present.

The aim of our study was to illustrate the role of rock fragments in a soil containing calcareous rock fragments, developed within a temperate climate. We first present field and laboratory experimental data that allow us to discuss the effect of the nature of rock fragments on the available water content. Then, we attempt to calculate water percolation in the same soil by using a reservoir model, in order to discuss the effect of the presence of rock fragments on water percolation.

Section snippets

Experimental data

The study area is located in Beauce, south of Paris, France, where we have selected a clay loam soil containing rock fragments, developed on a calcareous substratum. The depth of the soil depends on the slope orientation. Three types of soils have already been described in this area (FAO et al., 1998):

  • clay loam soils developed on cryoturbated materials (Haplic Calcisol);

  • clay loam soils developed on calcareous parent material (Haplic Calcisol);

  • clay loam soils containing rock fragments developed

General characterisation of the two phases

The volumetric percentage of rock fragments is high in the surface horizon (0–27 cm) of the two sites: 34.2% in site A and 21.9% in site B (Table 2). This percentage is higher in the lower horizons and reaches 56.9% at 55–80 cm depth in site A. In site B, the proportion reaches 47.3% in the last soil horizon at 50 cm depth. Nevertheless, the size distribution of rock fragments varies on the two sites: at the soil surface, rock fragments belong mainly to the 2–20 mm fraction in site A, whereas

Conclusion

Our experimental and modelling work has shown the effect of rock fragments in a calcareous stony soil from France. The use of a reservoir model has enabled us to discuss the influence of rock fragments in stony soils.

When the rock fragments are neglected and the soil is considered only as fine earth, the available water content is overestimated and, as a consequence, percolation is underestimated. This percolation underestimation can reach values larger than 10%, depending on the actual

References (28)

  • S. Childs et al.

    Physical properties of forest soils containing rock fragments

  • T.S. Coile

    Moisture content of small stone in soils

    Soil Science

    (1953)
  • C. Coutadeur et al.

    Influence de la phase caillouteuse sur la réserve en eau des sols. Cas des sols de Petite Beauce du Loiret

    Etude et Gestion des Sols

    (2000)
  • N.G. Danalatos et al.

    Rock fragments: II. Their impact on soil physical properties and biomass production under Mediterranean conditions

    Soil Use and Management

    (1995)
  • Cited by (167)

    View all citing articles on Scopus
    View full text