Validation of Predictions 



Verification of prediction equations is an arduous task because many sources of 

 variation in inventorying standing trees and weights of downed material after cutting 

 are difficult to control. Nonetheless, predictions were compared against inventoried 

 weights in three stands dominated by a single species--ponderosa pine, Douglas-fir, and 

 lodgepole pine. Crown and bole weights were predicted before cutting from complete 

 stand inventories on areas approximately 0.15 acre in size. D.b.h., crown ratio, and 

 crown class were used in the predictions. After cutting, the fresh slash was inten- 

 sively inventoried using the planar intersect method (Brown and Roussopoulos 1974). 



For slash less than 3 inches in diameter, predicted weights were less than in- 

 ventoried weights by 15, 22, and 37 percent of inventoried values. For all slash, 

 predicted weights varied from 4 percent more to 15 percent less than inventoried 

 weights. Some of the discrepancies were traced to biases in the test; thus, differ- 

 ences between predicted weights and actual weights would be less than indicated by our 

 test. The work in conducting a field verification test and the inconclusiveness of 

 comparing two estimates based on sources of variation that are difficult to control 

 makes this type of verification unappealing. The most productive verification would 

 be additional crown weight sampling. 



Considering the standard errors of estimate for the crown and bole weight equations 

 and the verification test, most estimates of slash weight from crowns and unmerchantable 

 bole tips for a stand of trees should be within 20 percent of the true mean. Occasion- 

 ally estimates can be expected to deviate from the true mean by as much as 50 percent. 



Crown Bulk Density 



Bulk density of live crowns influences crown fire potential, interception of rain- 

 fall, interception of forest fire retardants, infrared detection of forest fires, and 

 other phenomena. Quantifying bulk densities should help understand how tree species 

 affect these phenomena and assist in analytical modeling of' tree crown influences. 



For dominants, bulk densities for foliage and all branches of live crowns ranged 

 from 0.04 to 0.14 lb per cubic foot (fig. 13). The lowest bulk densities were displayed 

 by western redcedar and western larch, probably largely because of the open crown nature 

 of these species (Harlow and Harrar 1950) . Crowns of whitebark pine had the greatest 

 bulk densities, probably because the sample trees were relatively old and slow grown, 

 and possessed short thick branches. Subalpine fir and Engelmann spruce also had high 

 bulk densities, probably because branches were densely distributed within narrow crowns. 



Bulk densities for foliage of live crowns averaged one-half of the bulk densities 

 for entire crowns (foliage and branchwood) . The ratios of foliage bulk density to 

 entire crown bulk densities ranged from 0.36 for ponderosa pine to 0.61 for Engelmann 

 spruce. Species having high crown bulk densities also had high foliage bulk densities. 

 Conversely, species having low crown bulk densities also had low foliage bulk densities. 



Bulk densities for foliage computed using crown volumes, excluding foliage-free 

 cavities, differed only slightly from bulk densities based on crown volumes including 

 foliage-free cavities. Because foliage-free cavities were a small part of crown 

 volumes, only bulk densities based on the entire live crown volume are presented here. 

 Foliage-free cavities of the pines and larch, which are generally more intolerant and 

 retain foliage for only 1 to 3 years, were larger than cavities of the other species. 



25 



