O. 2008. The use of in situ volumetric water content at field capacity to improve prediction of soil water retention properties. Can. J. Soil Sci. 88: 533Á541. Most pedotransfer functions (PTF) developed over the past three decades to generate water retention characteristics use soil texture, bulk density and organic carbon content as predictors. Despite the high number of PTFs published, most being class-or continuous-PTFs, the accuracy of prediction remains limited. In this study, we compared the performance of different class-and continuous-PTFs developed with a regional database. Results showed that the use of in situ volumetric water content at field capacity as a predictor led to much better estimation of water retention properties compared with using predictors derived from the texture, or the organic carbon content and bulk density. This was true regardless of the complexity of the PTFs developed. Results also showed that the best prediction quality was achieved by using the in situ volumetric water content at field capacity after stratification by texture. Comparison of in situ volumetric water content at field capacity, with the water retained at different matric potentials as measured in the laboratory, showed field capacity to approximate 100 hPa, whatever the soil texture. Finally, the lack accuracy of PTFs that do not use the in situ volumetric water content at field capacity as predictor did not appear due to the test soils being unrepresentative of the soils used to develop the PTFs, but were instead related to poor correlations between the predictors used and the water retention properties.Key words: Pedotransfer functions, root mean square error, mean error of prediction, standard deviation of prediction, texture, bulk density, organic carbon content Al Majou, H., Bruand, A. et Duval, O. 2008. Utilisation de la teneur en eau volumique a`la capacite´au champ in situ pour ame´liorer la pre´diction des proprie´te´s de re´tention en eau des sols. Can. J. Soil Sci. 88: 533Á541. La plupart des fonctions de pe´dotransfert (FPT) de´veloppe´es durant les trois dernie`res de´cennies pour pre´dire les proprie´te´s de re´tention en eau des sols ont utilise´des caracte´ristiques de´rive´es de la composition granulome´trique, la teneur en carbone organique et la densiteá pparente comme pre´dicteurs. En de´pit du nombre e´leve´de FPT publie´es qui sont le plus souvent des classes de fonctions de pe´dotransfert ou des fonctions de pe´dotransfert continues, la pre´cision des pre´dictions reste faible. Dans cette e´tude, nous avons compare´les performances de diffe´rentes FPT de´veloppe´es a`partir d'une base de donne´es re´gionale. Les re´sultats montrent que l'utilisation de la teneur en eau volumique a`la capacite´au champ in situ comme pre´dicteur conduit a`des pre´dictions de qualite´supe´rieure a`celles enregistre´es avec des pre´dicteurs de´rive´s de la composition granulome´trique, ou avec la teneur en carbone organique et la densite´apparente quelle que soit la complexite´des FPT de´veloppe´es. Les re´...

. Prediction of soil water retention properties after stratification by combining texture, bulk density and the type of horizon. Soil Use and Management, Wiley, 2008, 24 (4) show that performance of class-PTFs can be similar to the more popular continuous-PTFs. In this study, we compare the performance of PTFs that were derived from a set of 456 horizons collected in France grouped by combinations of texture, bulk density and type of horizon (topsoil and subsoil). The performance of these class-PTFs was validated against water retained at -33 and -1500 kPa. Our results show that the best performance was obtained with 2 class-PTFs that used both texture and bulk density (texture-structural class-PTFs). They showed also that incorporation of horizon type into the PTF did not improve prediction performance. Comparison of performance at -33 and -1500 kPa showed very little difference, thus indicating no bias according to the value of water potential. Finally, the class-PTFs developed are well suited for predicting water retention properties at continental and national scales because only very basic soils data is available at these scales. A map of the available water capacity (AWC) was established for France using the 1:1 000 000 Soil GeographicalDatabase of France and an averaged AWC of 104 mm was computed for France.