Transposability of pedotransfer functions for estimating water retention of Algerian soils

Testing the efficiency of a pedotransfer function (PTF) outside of its development data-set is one of the best ways for assessing its robustness. An important question which remains unanswered is how transposable are PTFs to other agropedoclimatic contexts? Models developed and validated in a particular pedoclimatic context have been relatively little tested in other contexts. Particularly, no studies have been conducted until now to evaluate the PTFs for Algerian soils. In this study, eight (8) PTFs most frequently cited were considered. We used them to evaluate soil water retention at field capacity (FC) and wilting point (WP) on a set of 134 samples collected in the low Cheliff. The calculated Akaike information criterion and root mean square errors values showed that the Rawls, and Ghorbani Dashtaki et Homaee type 1 models were the best in estimation of soil water retention at FC (-709.795, 0.070 cm(3)cm(-3)) and WP (-733.480, 0.064 cm(3)cm(-3)). The poorer performances were presented by the PTFs developed on soils from Europe or United States where the organic matter values were much higher than the Algerian soils. However, the transposability of the PTFs formed from data spread from a wider area, produce more accurate predictions than those built from local data.