Modelling the Sustainability of Forest Production and Yield for a Changing and Uncertain Future

The expected growth of the human population to about 11 billion sometime within the next rotation of most northern temperate forest crops will put greatly increased and varied demands on today's forested lands. Development of the timber resources on those lands that remain dedicated to timber production must be demonstrably sustainable if forest management is to help arrest rather than aggravate the continuing deterioration of the global environment, and if managed forests are to be a carbon sink rather than a carbon source with respect to the global "green-house effect". Unfortunately, the experience-based models traditionally used by foresters cannot make accurate predictions of future forest growth, yield, and carbon balance for the altered growing conditions that are expected to accompany this increase in human numbers. These tools are therefore unsuitable as a means of assessing the sustainability of site productivity under current or anticipated future forest management practices and the expected future soil, climate, and biotic conditions. It is time for foresters around the world to confront this issue and to develop and use more ecologically-sensitive, ecosystem-level stand growth and yield models.Knowledge-based, process-simulation stand growth models have many theoretical advantages, and are the only way of predicting future forest growth and carbon budgets on a particular site in the absence of accurate data on the past growth of forests on that site. However, such models have generally had significant practical limitations as an alternative to traditional forest yield models. They have either been too simple, or, if sufficiently complex, have had unacceptably large calibration data requirements, which has limited their portability. This has restricted most process-based simulation models to research and educational applications.An alternative approach which combines both the experience-based and the knowledge-based approaches offers a more practical alternative. The combination of "historical bioassay" and process-based modelling approaches into "hybrid simulation" stand models can provide a means by which to rank the most probable outcomes and the sustainability of alternative stand-level management strategies under a variety of possible future growing conditions.The accuracy of most forest management and forest economics models ultimately depends on the accuracy of stand-level growth models. As the world experiences increasing problems of air pollution (acid rain and the greenhouse effect), soil degradation, and deforestation, there is an urgent need for foresters to use ecosystem-level growth models that are sensitive to human-induced and naturally caused environmental changes. Use of such models is a necessary prerequisite to good stewardship of forest land and our legacy to future generations.