Predicting Site Index and 

 Height for Selected Tree 

 Species of Nortiiern Idaho 



Brian IVI. Steele 

 Stephen V. Cooper 



INTRODUCTION 



Given the manifold combinations of species, sites, and 

 soils occurring in northern Idaho, and that growth rate, 

 quantity, quality, and value of wood produced varies by 

 species, foresters face complex decisions regarding 

 species to favor on a given site. One important consider- 

 ation is the relative productivity of each species. One 

 useful estimate of productivity is site index: the height 

 of dominant and codominant trees in a stand at an 

 arbitrary (index) age, usually 50 years in the Western 

 United States. Site index can be used to estimate vol- 

 ume growth through yield tables or as a variable in 

 growth simulation models. Though its appropriateness 

 has been questioned (Curtis and others 1974), the site 

 index equation is often solved for height, which is then 

 used as an estimate of stand height at ages other than 

 the index age. 



One method of estimating the site index of species 

 absent from a site is the use of species on species 

 regression equations or species comparison graphs. A 

 number of such studies have been done in the East and 

 Midwest (Foster 1959; Curtis and Post 1962; Carmean 

 and Vasilevsky 1971; Carmean 1979; Carmean and Hahn 

 1983). 



Copeland (1956) first demonstrated the utility of site 

 index comparisons for the western white pine (Pinus 

 monticola) type of northern Idaho. Deitschman and 

 Green's (1965) study, also conducted in northern Idaho, 

 stands as the most comprehensive of its type for the 

 Northern Rockies. Except for western white pine, this 

 study was done before site curves were available for 

 most tree species in the interior West. Consequently, 

 western white pine's site index was predicted as a 

 function of height and age of other tree species, and 

 environmental variables; conversely, height of the other 

 species was predicted as a function of white pine site 

 index, age, and environmental variables. 



Our report presents equations for predicting the site 

 index of one species from the measured site index of an 

 associated species and equations for predicting the 

 height of one species from its age and the measured site 

 index of an associated species. 



STUDY AREA 



This study is based on data collected in the course of 

 refining the northern Idaho habitat type classification 

 (Cooper and others 1983). The study area extends from 

 the Salmon River to the Canadian border, encompassing 



the Nezperce, Clearwater, and Idaho Panhandle National 

 Forests, where the sampling effort was focused. The 

 great diversity of forest environments found in this 

 region has been described and classified by Daubenmire 

 and Daubenmire (1968) and Cooper and others (1983). 



FIELD PROCEDURES 



Our methods generally followed those advocated by 

 Pfister and Arno (1980) with respect to stand selection, 

 sampling, and data analysis, with three important excep- 

 tions. First, stands that were not well stocked or approx- 

 imately even-aged were deleted from site index analysis. 

 Second, in order to more accurately estimate habitat 

 type productivity (reflected by site index), our goal was 

 to obtain height and age (at breast height, 4.5 feet) for 

 five unsuppressed, undamaged dominant or codominant 

 trees of each species in each stand. Third, soils were 

 described according to Soil Survey Manual prescriptions 

 (Soil Survey Staff 1975). We defined a stand to be a 

 group of trees of simileir age, structure, and species com- 

 position (canopy and undergrowth), occurring on an area 

 of uniform topography, and centered on a circular 

 5,350-ft^ sampling plot. 



Site index values were estimated using equations or 

 site index curves when equations were unavailable 

 (table 1). We employed Clendenen's (1977) computer 

 algorithm of Alexander's (1967) Engelmann spruce (Picea 

 engelmannii) curves to estimate mountain hemlock 

 (Tsuga mertensiana) and subalpine fir (Abies lasiocarpa) 

 site indexes because site curves specific to species and 

 region have not been developed for these two taxa. 

 Engelmemn spruce curves were chosen for extrapolation 

 to the other two taxa because of their similarity in 

 successional roles, undergrowth tolerance, and ecological 

 amplitude. 



In selecting site trees, no spacing or crown class limi- 

 tations were imposed except that all site trees be 

 dominant or codominant. This procedure departs from 

 sampling procedures accompanying published site curves 

 only in the case of western white pine. Haig's (1932) site 

 index for white pine involved complex precriptions and 

 the sampling of some trees that would not be considered 

 site trees by our standards because of their inferior 

 canopy position or low crown ratios. According to 

 Monserud (1984a), who compared site index derived from 

 three to four dominants to that obtained by Haig's 

 specifications (13 to 20 trees), our western white pine 

 site index values could be positively biased 20 percent or 

 more. 



1 



