"There is little understanding of how improved plant nutrition increases the production of dry matter" (Linder and Rook 1984). Much of the work with forest trees has been at the broad level of concentration of nutrients, rather than with detailed physiological studies of nutrient response functions, of factors controlliing 'efficiency' of nutrient utilization, and of the redistribution of nutrients from aging to developing tissues. The sustained supply of nutrients in forest ecosystems depends on processes by which nutrients are cycled from plant (in organic combinations) to soil and back to plant (in simple inorganic form). Studies of the key processes of decomposition and mineralization, and of equilibria determining nutrient availability have been hampered by lack both of appropriate chemical methods and of methods that distinguish among fractions of organic matter of varying nutritional quality. The root systems of forests must also be studied more intensively. In particular, mechanisms by which nutrients in short supply are taken up (for example, the role of mycorrhizae and of specialized systems such as proteoid roots) and the redistribution of nutrients associated with turnover of the fine root system are fields for future research.