Ecological site quality, site index, and height growth of white spruce stands in the sub-boreal spruce zone of British Columbia

In order to quantify the relationship between ecological site quality and white spruce [Picea glauca (Moench) Voss] height growth, vegetation, soil, foliar nutrient, and stem analysis data were obtained from 102 white spruce-dominated stands across a wide range of sites throughout the Sub-Boreal Spruce (SBS) biogeoclimatic zone of central British Columbia. The data were analyzed using the principles and methods of biogeoclimatic ecosystem classification, statistical analysis, and growth modelling. Using vegetation classification, the study stands were organized into eight seral plant associations on the basis of differences in the floristic composition of understory vegetation. The ratio of actual/potential evapotranspiration and the depth to a gleyed layer and groundwater table were used to stratify the study stands into seven actual soil moisture regimes. Soil water saturation (implied by the seven soil moisture regimes), soil drainage, soil texture, and slope were used for qualitative characterization of three soil aeration regimes. Soil moisture and aeration regimes were then combined, and nine soil moisture-aeration regimes were recognized. Total soil mineralizable-N (kg ha-1) and C/N were used to characterize a regional soil nitrogen gradient and to stratify the study stands into five soil nutrient regimes. When understory vegetation, foliar nutrients, or white spruce siteindex were related to categorical or continuous measures of soil moisture, aeration, and nitrogen, the strength of relationships obtained indicated that these measures were meaningful estimates of ecological conditions experienced by plants. Using the recognized soil moisture, aeration, and nutrient regimes and zonal units as differentiae, the study stands were organized into seven site groups, 13 site associations, and 30 site series. Regression analysis showed that both continuous and categorical measures of ecological site quality were useful in predicting white spruce site index (0.66 5 R20.94). The semi-empirical model, based on soil moisture-aeration and nutrient regimes and limiting factor analysis, was nearly as useful for predicting site index as the best developed regression model. Independent tests indicated that (1)integration of easily measurable soil variables into synoptic variables and (2)stratification of the study stands according to soil moisture conditions improved the strength of site index-ecological site quality relationships. Depending on the availability of data, the stratified soil model (mean prediction error = 1.44 m), the soil moisture-aeration regime plus soil nutrient regime model (mean prediction error = 0.85 m), and the semi-empirical model (mean prediction error = 0.96 m) were recommended for operational application. Z ratio (the height at 60 years of age divided by the height at 30 years of age)was used to quantify white spruce height growth pattern and to stratify the study stands into seven Z-groups. Developing height growth models for each Z-group improved prediction precision compared to traditional height growth models. In spite of the changing height growth pattern among stands, neither site index nor any measure of ecological site quality was found to be the major source of the variation in the study stands. Three ecologically based, site-specific height growth models (site series group-specific model, edaphic unit-specific model, and site group-specific model) were developed. The edaphic unit- and site group-specific models, with mean error of estimate