空间直观景观模型在呼中林区土壤侵蚀预测研究中的初步应用

土壤通用流失方程(USLE)已被广泛应用于大尺度的土壤侵蚀预测．在以往的土壤侵蚀研究中，由于只能获得静态的植被图，土壤通用流失方程只能用于土壤侵蚀的静态估算．空间直观景观模型能在大尺度上模拟植被动态，为土壤通用流失方程提供动态的植被因子，从而使土壤侵蚀的动态模拟成为可能．本研究结合空间直观景观模型LANDIS和土壤通用流失修正方程，以大兴安岭呼中林区为研究区。动态地模拟未来650年内有采伐和无采伐预案下的土壤侵蚀量；同时以无火无采伐预案下的土壤侵蚀为对比值．结果表明，土壤侵蚀量随时间变化呈周期性的波动，其波动程度在无火无采伐预案下最小，而在有火无采伐预案下最大；采伐对土壤侵蚀的影响没有火对土壤侵蚀的影响在空间上表现得明显，但是其累积效果则比火的影响强；降低采伐所产生的裸露土能有效降低年平均土壤侵蚀量，但是对土壤侵蚀动态变化的影响不明显；虽然采伐增加使平均土壤侵蚀量增加，但是也同时使土壤侵蚀的年际变化更趋于平稳．

Application of spatially explicit landscape model in soil loss study in Huzhong area

Universal Soil Loss Equation(USLE)has been widely used to estimate the average annual soil loss.In most of the previous work on soil loss evaluation on forestland,cover management factor was calculated from the static forest landscape.The advent of spatially explicit forest landscape model in the last decade,which explicitly simulates the forest succession dynamics under natural and anthropogenic disturbances(fire,wind,harvest and so on)on heterogeneous landscape,makes it possible to take into consideration the change of forest cover,and to dynamically simulate the soil loss in different year(e.g.10 years and 20 years after current year).In this study,we linked a spatially explicit landscape model(LANDIS)with USLE to simulate the soil loss dynamics under two scenarios:fire and no harvest,fire and harvest.We also simulated the soil loss with no fire and no harvest as a control.The results showed that soil loss varied periodically with simulation year,and the amplitude of change was the lowest under the control scenario and the highest under the fire and no harvest scenario.The effect of harvest on soil loss could not be easily identified on the map;however,the cumulative effect of harvest on soil loss was larger than that of fire.Decreasing the harvest area and the percent of bare soil increased by harvest could significantly reduce soil loss,but had no significant effects on the dynamic of soil loss.Although harvest increased the annual soil loss,it tended to decrease the variability of soil loss between different simulation years.