of the tree nunbers from the cut trees as a basis for determining 

 the amount cut. Asking a logger to report which trees have been cut 

 takes control of this important accounting procedure away from the 

 responsible forest officero 



Despite its several drawbacks, tree measurement is generally considered 

 to be less expensive than log scaling under some conditions, and as 

 personnel become better trained in its application, the system will 

 increase in favor and in use. 



samplhtg in tree 15Easured sales 



Administrative costs on national forest timber sales can be further 

 reduced by the use of sampling methods in tree measured sales, 

 Lexen (9_) pioneered in sample tree measuring just as in sample log 

 scaling vjhen, in 1947, he outlined a method to be used on large Forest 

 Service sales in the Rocky Mountain and Southwestern Regions. The 

 possibility had been suggested in 1945 by an investigator in the 

 Pacific Northwest after applying samipling procedures to log scaling 

 felled and bucked full-length trees in the woods, (10) 



Sample tree measurement as considered here means marking and counting 

 by species all trees to be cut on a sale but measuring only a certain 

 percentage of them. Thus, the volume of a part is multiplied by the 

 appropriate factor and becomes an estimate of the volume of the vjhole. 

 The check scale, which is necessary in most cases, can be thought of 

 as a sample within a sample vjhere a fev; trees, accurately scaled after 

 felling and bucking, provide a basis for correcting the volume derived 

 from the large sample. . In sample tree measurement, the volume must be 

 kept separate by species because of the .differing stumpage prices among 

 species in the same saleo Different sampling rates for each species 

 may be required because the total number of trees is usually markedly 

 different by species. The process necessarily requires different forms 

 and bookkeeping methods from the common scale book procedure, and a 

 strict system of randomization in the selection of sample trees is 

 essential. Unlike sample log scaling, a mechanical selection of sample 

 trees is not appropriate because trees do not grow in such an order 

 that the selection of every fifth tree, for example, could be free of 

 human choice. The marker must not know in advance which tree is to be 

 measured. Methods of accomplishing randomization for crew marking 

 and individual marking will be discussed later. 



Sample tree measurement has all the advantages of total tree measure- 

 ment, plus being less expensive. Sale areas can be marked and measured 

 considerably in advance of selling the timber. Log scaling is elimi- 

 nated. More time can be devoted to sale area inspection and supervision, 

 and the check scale sam-ple can be gradually collected at the same time. 



-5- 



