Millar Western Forest Products Ltd. manages a forest in west-central Alberta under a Forest Management Agreement (FMA) with the Government of Alberta. Part of Millar Western's planning process brought researchers together to develop a decision support system (DSS) for forest management planning and monitoring programs. Four modules — timber supply, biodiversity, FIRE, and WATER — were built to evaluate, with the help of indicators of sustainable forest management, current and future forest conditions predicted from computer simulations of alternative management scenarios. In the first round of assessment four management scenarios, distinct by their level of silviculture intensification and by the spatial clearcut layout pattern, were compared. Such comparison has demonstrated that (1) the current forest management scenario improved moose habitat at the expense of timber supply, (2) all scenarios had similar fire risk, (3) generated increases in peak flow and water yield of selected watersheds, and (4) slightly impoverished forest biodiversity. All scenarios were examined in light of a computer-simulated natural disturbance benchmark. This led to landscape design scenarios to reduce fire risk and balance biodiversity indicators with timber supply objectives, one of which was eventually selected for implementation. The company's monitoring and research program is also highly focused on improving DSS modules and the underlying data, hence its association with the Forest Watershed and Riparian Disturbance (FORWARD) project, which considers the effects of forest management on aquatic ecosystem indicators. Key words: decision support system, ecosystem management, forest management, natural disturbance, indicators, sustainable forest management, adaptive management.

The cumulative impacts of human and natural activity on forest landscapes in Alberta are clear. Human activity, such as forestry and oil and gas development, and natural processes such as wildfire leave distinctive marks on the composition, age class structure and spatial configuration of the forest. Also, other processes such as climate change may be slowly and subtly modifying forest dynamics and may lead to important changes over time. Given the importance and ubiquitous nature of these cumulative impacts, a forest management plan that does not adequately take such impacts into account cannot be expected to adequately manage the forest, neither its components nor its processes. In order to address the question of cumulative impacts in the context of forestry, a landscape model was designed and built in order to simulate forestry, oil and gas, climate change, wildfire, and demographic change for the Whitecourt forest management area over a long time horizon. This paper presents the model and the forest landscape states it forecasted with cumulative impacts, and evaluates the fate of some key indicators of biodiversity and forest productivity. Simulations of harvesting as the only disturbance, the nearest analogue to the current approach to forest management planning, yield results that differ greatly, in every respect, from the results of simulations of harvesting combined with other disturbance agents. The simulation of multiple disturbance agents together allows for the detection of interactions among disturbance agents, and indeed, there are important interactions between the processes of fire and oil and gas. Results also show that climate and demographic change will intensify the impact of fire on the supply of timber and other values. Also, the continued development of petroleum resources will lead to an important erosion of the forest landbase. Overall, this paper makes a strong case for cumulative impacts assessment and the use of spatial and temporal stochastic modelling in forest management.Key words: cumulative impacts, forest management, climate change, landscape modelling, APLM RÉSUMÉLes impacts cumulatifs des activités humaines et naturelles sur le paysage forestier en Alberta sont évidents. L'activité humaine, telle la foresterie et les développements pour le pétrole, et les processus naturels tel le feu laissent une marque distincte sur la composition, la structure des classes d'âge, et la configuration spatiale de la forêt. De plus, d'autres processus tels les changements climatiques influencent peut-être aussi subtilement et tranquillement la dynamique forestière et pourront mener à des changements importants avec le temps. Étant donné l'importance et l'ubiquité de ces impacts cumulatifs, un plan d'aménagement forestier ne prenant pas compte de ces impacts ne pourra aménager de façon satisfaisante ni la forêt, ses composantes, ni ses processus. Pour adresser la question des impacts cumulatifs dans le contexte de la foresterie, un modèle de paysage a été conçu et construit pour simuler...

Intensive silviculture is practiced in many parts of the world but is rare in the public forests of western Canada. Here, we make the argument that intensive silviculture could be justified in Alberta but has not been implemented due to philosophies and policy decisions by foresters from government, industry and academia. These include adherence to long rotations, management goals that are aimed at sustained total volume yield rather than economic value, limitations in the types of stands that are allowed to be regenerated and models that do not include intensive silviculture options. In Mixedwood Growth Model projections, we demonstrate the potential of intensive silviculture that includes combinations of selecting good sites and thinning to produce merchantable stands by age 50 compared to the standard rotation age of 80 with basic silviculture. There could be even more gains if forest level constraints in timber flow were removed due to the increased growth of regenerating stands. Finally, we examine the attitude and policy changes that we believe are necessary for adoption of more intensive silviculture systems on parts of Alberta’s forest landbase.

Millar Western Forest Products Ltd. completed its 2007-2016 detailed forest management plan for its forest management agreement area by engaging multi-discipline expertise to better manage forest values. Disciplines were formed into impact assessment groups, which were charged with developing indicators and targets for identified values and objectives and with constructing models to predict the impacts of forest management activities. Three different modeling approaches were used in developing the plan and the best results were obtained by integrating indicators directly into the forecasting model even when the indicators were simple versions of complex models. Integrating indicators into the forecasting model reduced the penalty cost on other indicators, permitted the model to find better solutions, provided almost immediate feedback to the plan development team, permitted more scenarios to be investigated, and provided a better understanding of the dynamics and tradeoffs. The other two techniques used in the DFMP involved the assessment of forecasted scenarios with more complex discipline specific models, which provided more detailed information about the indicators. Mots clés : prévision, planification de l'aménagement forestier, Patchworks, modélisation de l'aménagement forestier, analyse de l'approvisionnement en matière ligneuse, analyse des retombées, VOIT

No abstract