干旱区生态环境知识资源中心

Surface hydrological behaviour is important in drylands because it affects the distribution of soil moisture and vegetation and the hydrological functioning of slopes and catchments. Microplot scale run-off can be relatively easily measured, i.e. by rainfall simulations. However, slope or catchment run-off cannot be deduced from microplots, requiring long-time monitoring, because run-off coefficients decrease with increasing drainage area. Therefore, to determine the slope length covered by run-off (run-off length) is crucial to connect scales. Biological soil crusts (BSCs) are good model systems, and their hydrology at slope scale is insufficiently known. This study provides run-off lengths from BSCs, by field factorial experiments using rainfall simulation, including two BSC types, three rain types, three antecedent soil moistures and four plot lengths. Data were analysed by generalized linear modelling, including vascular plant cover as covariates. Results were the following: (i) the real contributing area is almost always much smaller than the topographical contributing area; (ii) the BSC type is key to controlling run-off; run-off length reached 3m on cyanobacterial crust, but hardly over 1m on lichen crust; this pattern remained through rain type or soil moisture; (iii) run-off decreased with BSC development because soil sealing disappears; porosity, biomass and roughness increase and some changes occur in the uppermost soil layer; and (iv) run-off flow increased with both rain type and soil moisture but run-off coefficient only with soil moisture (as larger rains increased both run-off and infiltration); vascular plant cover had a slight effect on run-off because it was low and random. Copyright (c) 2014 John Wiley & Sons, Ltd.