A Logistic Model of Mortality in Thinned and Unthinned Mixed Conifer Stands of Northern Idaho
Published mortality models have not always done an adequate job of predicting the probability of mortality following thinning. This paper describes a mortality model based on a series of thinned and unthinned sample plots located in northern Idaho, some of which have been measured for 70 years. Independent variables are diameter at breast height (dbh), stand basal area, annual diameter growth, relative dbh, and species. Thinning results in changes in three of these variables: the stand basal area, relative dbh, and diameter growth rate. Model performance is evaluated by considering how well the model fits the development data set, how well the model fits an independent data set, and how well the model fits biological reality. In each evaluation the selected model shows a marked improvement over the performance of the mortality model used in previous releases of the Stand Prognosis Model (Wykoff et al. 1982). The model is shown to perform well in both thinned and unthinned stands. Forest Sci. 32:989-1000.
No Supplementary Data
No Article Media