... error and its mean squared error was almost as small asthat for model (5). Therefore, I recommend the use of model (4) for estimating the height of ponderosa pine in British Columbia.The models ... terms any of the tested models would suffice [3, 4, 8, 33]. Therefore, I chose thelog-logistic model because it was used successfully in the past for otherspecies [3, 19, 33], and because of its ... height growth trajectories and small sample sizes for some zones. A new height model for ponderosa pine is nowavailable for British Columbia. This model gives only slightly different height estimates...