Our understainding of forest production and forest growth is incomplete. Present yield tables have not predicted the recent changes in boreal and temperate regions of Europe. We argue that we need to have a causal understanding built upon a mechanistic knowledge of important processes. Using a carbon/nitrogen model as a framework, important components and processes of the forest ecosystem are discussed and the new knowledge needed is identified. Present soil chemistry may not be fully relevant in describing plant/soil relationships. It has been shown that the near-root environment deviates from the conditions described by bulk chemistry. Modelling is discussed as an important tool for increased understanding and prediction. Sustainability of forest production is usually treated from a long-term perspective. Long-term changes can lead to deficiency or even excess of mineral nutrients, which will affect the resistance of the tree or stand to drought, frost as well as attacks of insects and pathogens. Recent ideas of tree vigour or vitality in relation to insects and pathogens are reviewed as a component for understanding production stability in short- and long-term perspectives.
