| dc.description.abstract | Kozak's variable exponent taper equation was fitted for balsam poplar (Populus
balsamifera L.), trembling aspen (Populus tremuloides Michx.), white spruce (Picea
glauca (Moench) Voss), black spruce (Picea mariana (Mill.) B.S.P.) and jack pine
(Pinus banksiana Lamb.) in Manitoba. Stem analysis data were collected from a total of
37 balsam poplar, 71 trembling aspen, 97 white spruce, 300 black spruce and 298 jack
pine trees. The data were collected within the Boreal Shield and Boreal Plain ecozones
in Manitoba. Four ecoregions were sampled within the Boreal Shield ecozone including
the Churchill River Upland, Hayes River Upland, Lac Seul Upland, and Lake of the
Woods ecoregions. Three ecoregions were sampled within the Boreal Plain ecozone
including the Mid Boreal Lowland, Mid Boreal Uplands and the Interlake Plain
ecoregions. Stem taper variability between ecozones and ecoregions were tested using
the F-test as the data permitted. Stem variability between site types were also tested for
black spruce and jack pine. Ecozone-specific, ecoregion-specific, site type-specific and
provincial taper equations were constructed corresponding to the results of the F-tests.
A taper equation was developed for balsam poplar ecoregion 152-154. Ecozone-specific
equations were derived for trembling aspen and white spruce. Provincial and site typespecific
equations were derived for black spruce. Ecoregion-specific and site typespecific
equations were derived for jack pine. Regional differences of stem taper were
the result of geographic, climate and site differences. The results indicated that all the
models performed quite well in predicting diameter inside bark. The residual plots of
the balsam poplar and trembling aspen taper equations showed increasing variance and a
non-normal distribution of residuals. However, the coefficients can still be considered
appropriate and the predictive ability of the models. 'Yas not affected. For each species,
the diameter inside bark ( dib) predictions of the Manitoba provincial models were
plotted with the corresponding dib predictions of the Alberta provincial equation
developed by Huang (1994). The plots indicated that the Manitoba equations performed
well and were similar to the Alberta equations. For each species, dib predictions of the
ecozone-specific, ecoregion-specific or site type-specific models were plotted with the
dib predictions of the provincial model to display the differences indicated from the Ftests.
Some of the equations were. derived with less than the minimum 60 trees
recommended by Kozak. Therefore, it is suggested that data be continually accumulated
and used to update the equations developed in this study and to further determine if
ecoregion-specific and site type-specific equations are required. Data should be
collected outside the current range of diameter at breast height (dbh) and total tree height
measurements to expand the application of the equations. | |
