| The focus of this study was on the development of height growth curves for ponderosa pine, Douglas-fir, western larch and lodgepole pine in western Montana.; In comparison to real growth curves, 9 of the 11 published height growth curves examined had inappropriate curve shape. This resulted in aged-related over or under predictions of 2 to 12 feet depending on the curve. These systems should not be used in western Montana for estimating height growth curves of dominant trees. The new site curves constructed for each species from the stem analysis data should provide more reliable estimates of dominant tree height growth patterns and site index (SI) in western Montana.; SI prediction models were developed which explained from 48% (ponderosa pine) to 73% (lodgepole pine) of the variation in observed SI. Standard errors about regression ranged from 9.0' (ponderosa pine) to 6.3' (lodgepole pine). The models' inputs (solar radiation, temperature and precipitation) can be easily obtained from site measurements normally taken on a forest inventory (slope, aspect, elevation) and from published state climatological summaries. Soil attributes were not significant independent variables for predicting SI. It is argued that these models are superior to most SI prediction equations due to: (1) use of variables more closely related to the fundamental facts affecting growth; and (2) use of a multiplicative model form, rather than the usual additive form. The response surfaces for the SI models show generally reasonal behavior. A larger sample and a more complex model would be required to more completely capture the dynamics of the system, however.; For all species, except western larch, soil and site variables proved to be significant courses of variation in curve shape, over and above site index. For Douglas-fir and lodgepole pine, soil and site variables were the only significant sources of variation in curve shape. Clearly, improved estimates of height growth curves should be obtained from a system which incorporates soil and site variables.; Comparable accuracy and precision were obtained by the site curve system and an alternative system in which curve shape is estimated from soil and site variables. An improved application procedure is proposed which should make the alternative system superior to the site index system. |
