habitat type. For example, Douglas-fir is considered shade tolerant on habi- 

 tat types where it is the climax species, but is shade intolerant where it is 

 not a climax species. 



Mechanical or burn site preparation generally decreases the probability 

 of advance regeneration, with burn site preparation decreasing the probabil- 

 ity more than mechanical. Because this study was retrospective, we could 

 not determine whether advance regeneration was deliberately destroyed or 

 whether none was present and site preparation was done to increase subse- 

 quent regeneration success. 



The probability of advance regeneration decreases over time. This could 

 be due to mortality over time or poor response to release. Trees that respond 

 poorly to release could, with time, become subordinate to a subsequent tree 

 of the same species. 



Probability of 



Subsequent 



Regeneration 



Trees up to 3 years old at the time of harvest were recorded as subsequent 

 regeneration. This was for four reasons. First, there could be some error in 

 aging small trees; accuracy was ±1 year. Second, the year of harvest entered on 

 historical records was usually the year of completion and, if it took more than 

 1 year to harvest the stand, some areas would have been harvested before 

 others. Third, a 3-year-old tree would not have suffered major physiological 

 setbacks from suppression by overstory trees. Fourth, any trees planted im- 

 mediately after harvest would appear to be advance regeneration. 



Equations for the probability of subsequent best trees are given in appen- 

 dix B, table 13. The occurrence of subsequent regeneration differs from that 

 for advance regeneration. Mechanical or bum site preparation increases 



20 



