Background:Soil erosion is still identified as the main cause of land degradation worldwide,threatening soil functions and driving several research and policy efforts to reverse it.Trees are commonly associated to some of the most successful land-use systems to achieve soil protection goals,but the extent to which forest ecosystems reduce erosion risks can largely depend on management decisions and associated silvicultural practices.Optimization tools can assist foresters in solving the complex planning problem they face,concerning the demand for different,and often conflicting,ecosystem services.A resource capability model(RCM),based on a linear programming approach,was built and solved for a forest landscape management problem in Northwest Portugal,over a 90-years planning horizon,divided in 10-years periods.Results:Timber provision and soil erosion were found to be in trade-off.The management alternatives included in the model were proven to be sufficiently flexible to obtain the desired level of timber yield,both in volume and even distribution along the planning horizon,while ensuring lower levels of soil loss estimates(below 35 Mg·ha-1·year-1).However,under climate change conditions,compatible with an increasing greenhouse gases emission scenario,potential landscape soil erosion may be enhanced up to 46Mg·ha-1·year-1 in critical periods.Conclusions:Soil conservation concerns in landscape-level forest management planning can be addressed by LP-based optimization methods.Besides providing an optimal management solution at landscape level,this approach enables a comprehensive analysis of the RCM,possible trade-offs and potential changes towards uncertainties.