Mortality models for major boreal mixedwood species in Alberta

This thesis focused on mortality model development for three major boreal mixedwood species in Alberta: trembling aspen (Populus tremuloides Michx.), white spruce (Picea glauca (Moench) Voss), and lodgepole pine (Pinus contorta Dougl. ex Loud. var. latifolia Engelm).; An individual tree survival function was developed for each species using a generalized logistic function. Measurement interval length entered the logistic equations as an exponent to overcome the problem of unequal measurement intervals. Unknown parameters were estimated using the maximum likelihood method. The newly developed functions were compared to the old ones previously used by the Mixedwood Growth Model (MGM). Both fitting and validation results confirmed that the new functions performed better than the old ones. MGM simulations further supported this conclusion.; The self-thinning concept was used to develop a maximum size-density relationship between quadratic mean diameter and stand density. Several functions were compared first by fitting an average size-density relationship and two equally good functions were selected based on model fitting and prediction statistics. They were then refitted with a specially defined loss function to derive the maximum size-density relationship. One final function was chosen as a stand level constraint on the empirical survival functions in MGM. It was found that site quality did not affect this maximum size-density relationship and separation of this relationship by species was not possible.; Besides the maximum size-density relationship, a group of constraining factors were also developed to further constrain the empirical survival functions. One factor was based on the idea that there is a maximum basal area a stand can produce and this maximum basal area is different for different species. This factor was developed to prevent excessive yields for very dense stands. Another factor was developed to reduce the survival rates of very old trees based on a defined quadratic mean diameter limit for each species. The third factor was based on the average height growth rate of the top height trees in a stand and was developed to break up short-lived aspen and lodgepole pine stands at old ages. Imposing these factors on the empirical survival functions assures ecologically reasonable performance of MGM when extrapolated beyond the typical data range used for model development.