This equation has an r-square of 0.7694 and was developed from 362 stands 

 in which 10 or more Vaoo-acre plots were sampled and on which all established 

 regeneration was recorded. The relationship between stocking and density 

 is shown in figure 17. 



The predicted ratio of stocked plots is then used to mathematically adjust 

 the probability of stocking equations given in appendix B, table 9. The ad- 

 justment is made to Sp^Xj in the probability equation 



P = (l+e-<^W)-^ 



so that the probability remains bounded in the interval [0,1]. 



Figure 18 illustrates the calibration procedure. Line A is the probability 

 of stocking curve generated from the Abies grandis series equation in ap- 

 pendix B, table 9. Line B represents the same stand that was inventoried 

 at 3 years since harvest and is more densely stocked than expected. Line C 

 represents the same stand as line A, but it is less densely stocked than ex- 

 pected at 3 years. Calibration adjusts lines B and C to coincide with the in- 

 ventories (points b and c). Adjusted predictions (lines B and C) are still 

 bounded within the interval [0,1]. 



A second part of calibration concerns creating new tree records. If the ac- 

 tual number of trees on the plot exceeds the predicted number, no new tree 

 records are created for that plot. 



Consistency of The regeneration model uses pseudo-random numbers to make unbiased, 



Predictions discrete choices from a number of possibilities; for example, the species that 



will occupy a plot. This feature mimics stochastic variation seen in nature. 

 Regeneration in a stand would vary depending on the year it was harvested 

 because of differences in weather, seed crops, and animal/insect activity 



1.0 H 



1,000 2,000 3,000 4,000 5,000 



Trees Per Acre 



Figure 17 — The relationship of trees per acre to stocl<ed plot ratio. 



34 



