A pedo-transfer function (PTF) for estimating soil bulk density from basic soil data and its comparison with existing PTFs.

Introduction

Bulk density is a key parameter used in many soil hydrologic models for predicting a wide range of soil processes. The influence of many soil management practices on the soil properties that affect water movement and retention in soils requires information on the soil bulk density. However, bulk density is not always determined in routine soil surveys, as field procedure deployed for its determination is tedious, labour-intensive, time-consuming, and expensive. In addition, due to the spatial variability of soils, a large number of soil core samples is required to adequately represent a large area or a watershed. In fact, lack of this information at a regional scale has impeded both the application and improvement of the soil hydrologic models.

Experience has shown that organic carbon (OC) and texture predominately determine soil bulk density. Organic carbon and texture information is often available in soil survey campaigns. Therefore many attempts have been made to estimate soil bulk densities through some pedo-transfer functions (PTFs) based on soil OC and texture data (Curtis and Post 1964; Adams 1973; Alexander 1980; Federer 1983; Rawls 1983; Huntington et al. 1989; Manrique and Jones 1991; Bernoux et al. 1998; Tomasella and Hodnett 1998). Most of these PTFs, except the ones developed by Rawls (1983), Tomasella and Hodnett (1998), and Bernoux et al. (1998), are a function only of organic matter (OM)/OC content. Although studies conducted by Saini (1966) and Jeffrey (1970) have shown that OM has a dominating effect on soil bulk density and that it alone can be used as a good predictor of soil bulk density, it has been observed (e.g. Alexander 1980; Huntington et al. 1989; Manrique and Jones 1991) that soil texture plays a major role in controlling bulk density where OM is a minor component. Further, no attempts have been made to date to test and compare the applicability of these PTFs on an independent soil data set. In addition, no studies on the effect of varying levels of soil particle size distribution and/or OC contents (i.e. input parameters) on the errors associated with these estimation methods (i.e. sensitivity analysis) have been attempted so far for testing their applicability to a wide range of soil types.

Thus, the objectives of this study were to:

(i) develop a soil bulk density estimating PTF, based on both soil texture and OC content data; and

(ii) test/compare the applicability of both proposed and existing PTFs on an independent soil data set.

Materials and methods