Effects of rainfall characteristics on infiltration and redistribution patterns in revegetation-stabilized desert ecosystems

SummaryRainfall, the dominant source of water replenishment in the semi-arid sand dune area of north-western China, plays an important role in sustaining the desert ecosystem. An experiment to measure water balance associated with infiltration events was conducted on the re-vegetated sand dunes in the Tengger Desert, north-western China. The redistribution of infiltrated moisture in the course of percolation, root extraction, and evapotranspiration pathways was investigated for a period of 45 days during the growing season. Time domain reflectometry probes were inserted horizontally at 12 different depths below the ground surface in the Caragana korshinskii dwarf-shrub community to record volumetric soil moisture at hourly intervals. Rainfall events were sporadic with widely different intensities during the period of the experiment. The presence of vegetation markedly influenced the infiltration and redistribution patterns on the stabilized sand dunes. Infiltration rates varied greatly with individual rainfall quantity and antecedent soil moisture, with drier soil profile facilitating infiltration. The relationship between infiltration rate and rainfall intensity was linear, with infiltration rate at 80% the magnitude of rainfall intensity. Contrasts between the infiltration rate and cumulative infiltration varied with the feature of rainfall events of the vegetation-stabilized desert soil and the un-vegetated bare desert soil indicate that the measured precipitation alone is insufficient to explain the effective rainfall of the studied regions. At rainfall amount <8.2 mm, with rainfall intensity <0.5 mm h−1, no soil moisture was gained for the re-vegetated soil, while for the bare soil the comparable values were <6.4 mm, and <0.7 mm h−1, respectively. Root withdrawal of soil water and evapotranspiration (reaching 69–90% of precipitation) restricted the wetting front penetration for the vegetated soil. In contrast, the bare soil was prone to infiltration zone beneath the depth of 40 cm. Therefore, the effective rainfall for the vegetated soil, in terms of the soil moisture replenishment, differed from the bare soil in the study region due to the presence of vegetation.