SEASONAL WATER-BALANCE OF A SANDY SOIL IN NIGER CROPPED WITH PEARL-MILLET, BASED ON PROFILE MOISTURE MEASUREMENTS

In the Sahel, calculation of the field water balance from neutron-probe measurements is often difficult for pearl millet (Pennisetum glaucum (L.) R. Br.), which is due to the rapid drainage (D) of the sandy soils, on which it is typically grown. We present a simple method of calculating D in these soils from weekly neutron-probe data. The method divides the water balance into two phases. In the first, applicable early in the season, water flux across the maximum depth of probe measurement (Z(m)) is assumed negligible, and evapotranspiration (E) and D are calculated from the change in soil water content (theta) between the bottom of the rootzone (Z(r)) and Z(m), thus allowing calculation of unsaturated hydraulic conductivity, K(theta), from the flux across Z(r). In the second phase, when soil water starts to percolate across Z(m), D is calculated from K(theta), assuming a hydraulic head gradient of -1. The method is used to calculate a one-dimensional water balance of a pearl-millet crop grown in a deep sandy soil at two fertility levels during a season of normal rainfall. Results show that the calculated K(theta) functions compare well with those based on laboratory measurements. An acceptable estimate of drainage, and therefore E could be made. Mean cumulative E and D were, respectively, 211 and 207 mm for the unfertilized crop, and 268 and 148 mm for the fertilized crop with 440 mm of rainfall received during the crop cycle. The fertilized millet crop water balance was simulated, which compared to the calculation method resulted in an about 10% higher seasonal E and a 10% lower seasonal D. Our study shows that E can be corrected for D using a simple but accurate method, and consistent with other studies in the region indicates that rainfall is usually not the primary limiting factor to pearl-millet production.