Figure 1. — On a tree classified 

 as "serotinous , " 90 percent 

 or more of the cones it bears 

 are closed. Note the 

 characteristic fusiform shape 

 of the cones indicated by 

 the arrows. 



The for equation (2) is 0.703. But the "Y" values used as input for this equation 

 were, themselves, derived from "smoothed" tree estimates using equation (1) so that the 



of 0.703 contains upward bias and the standard error of estimate (13.4 thousand 

 cones per acre) is biased downward. These two biases are thought to be relatively small. 



HoLJ to collect data:--l\\e plot data required for Method 2 are little more than 

 normally obtained for inventory purposes, and trees do not have to be felled. The 

 data required are: 



1. Mean plot d.b.h. in inches (merchantable trees only), 



2. Mean plot age in years (merchantable trees only) , 



3. Number of serotinous-cone types of trees per acre. 



Compute the average plot d.b.h. and age. Examine each tree in the plot to deter- 

 mine if it is of the serotinous-cone type. Use good quality binoculars of 6 or 7 power 

 and determine the tree's cone habit. Trees included in the formula should definitely 

 bear 90 percent, or more, serotinous cones (figure 1). Current-year, immature cones 

 are not to be included. Trees not counted are those having 90 percent, or more, open 

 cones (classified as "open-coned," figure 2), and those having between 10 and 90 percent 

 serotinous cones (classified as "intermediate," figure 3). These data should not be 

 collected in wet weather when open cones are closed by hygroscopic swelling (figures 4, 

 5, and 6). However, observations can be made in the winter when snow is on the ground 

 if relative humidities are low. Field crews should be supervised for accuracy concern- 

 ing cone determination. 



Using equation (2) , compute the estimated number of serotinous cones per acre for 

 each plot. Then the average across all plots will be the estimated number of cones per 

 acre. Values from preliminary samples can be used to determine the coefficient of 

 variation and required sampling intensity as mentioned in Method 1. 



3 



