Changes in soil water balance following afforestation of former arable soils in Denmark as evaluated using the DAISY model

Abstract

Land use change alters water and element cycles, but the changes in these cycles after conversion, for example, from cropland to forest are not fully described in hydrological and nutrient transport models, which usually describe either cropland or forest stands. In the European Union future afforestation is likely to occur on abandoned cropland, and evaluation of the future impacts of this land use change will require projections with models that include combined cropland-forest modules. This study used the agro-based DAISY model (Version 4.93) to investigate changes in the soil water balance over four decades following afforestation of a homogeneous area of former arable land on a sandy loam in Denmark. Hydrological data collected during nine hydrological years (April 2001–March 2010) were used to test the DAISY model. Monthly data on soil water content at 0–90 cm used for calibration were available from April 2001 to December 2002 for six monoculture stands of oak (age 8, 22 and 31 years) and Norway spruce (age 4, 13 and 32 years). Model performance was evaluated by considering uncertainties in model inputs using the Generalised Likelihood Uncertainty Estimation (GLUE) procedure. The GLUE estimates obtained (uncertainty bands 5% and 95%) agreed satisfactorily with measured monthly soil water content during the calibration period (April 2001–December 2002). Similarly, in the oldest oak stand, long-term monitoring observations and predictions of monthly water content were in satisfactory agreement during the period January 2003–March 2010). Sensitivity analysis showed that the DAISY model was most sensitive to the potential evapotranspiration factor and soil hydraulic parameters included in the Campbell model. Simulation results during nine hydrological years showed that 16–25% of incoming precipitation led to water recharge in the spruce stands, while the corresponding range for oak stands was 25–27%. A 35-year DAISY simulation revealed that Norway spruce consumed more water than oak, with differences in annual water recharge in the range 31–174 mm year 1 and with greater differences in rainy years (precipitation >900 mm year 1).