To assure unbiased estimates and good distribution, the samples were drawn at 

 random within strata defined by land ownership and operator size class. Two owner- 

 ship classes were used--National Forest and other. Operator size class corresponded 

 to the production class of the sawmill for which the logging was being done. ' Two size 

 classes were used--small (less than 10 million board feet per year) and large 

 (10 million board feet and more). 



Past experience in Idaho, Montana, Arizona, and New Mexico indicated that about 

 44 samples distributed as shown in the tabulation on page 10 should result in 

 estimates of standard errors of the means of cubic residue volumes of about 20 percent 

 per State and for western South Dakota. 



The actual sample sizes used and standard errors achieved are shown in table 7. 

 The cubic-foot standard errors are the most meaningful to Forest Survey because survey 

 standards are expressed in cubic volumes. In the four States the variability en- 

 countered was greater than anticipated and the resulting standard errors are higher 

 than expected. In Utah the standard error is especially high but because the residue 

 volume in the State is small, the error appears acceptable. 



Trees felled for products were measured where they had been felled to determine 

 both product and residue volumes. Residue volume from other trees damaged or cut as a 

 result of felling product trees was also measured. Skidding damage to trees along skid 

 trails was determined after logs reached the landing except in those cases where crews 

 were on hand to witness damage as it occurred. 



A slightly different procedure was used for clearcut operations where there was 

 difficulty in relating felling and skidding damage to individual product trees. In 

 these cases, product trees were those whose stumps fell within a circular plot. All 

 product trees were measured within the plot and felling and skidding damage was 

 assessed on the plot area. 



Caution is recommended in applying estimates shown in this report to any State 

 subdivisions. Important interrelated factors that contribute to variation in the 

 amount of residues per unit of product volume from one area to another include stand 

 conditions (size and soundness of trees, species, stocking, etc.); markets for various 

 species, sizes, and qualities of timber; and logging costs and techniques as determined 

 by accessibility, terrain, etc. 



Table 7 . --Standard errors of ratios for logging residue volumes by net 

 cubic- and board- foot measures 



Unit of measurement 











and 



Western 









minimum top diameters 



South Dakota 



: Wyoming : 



Utah 



: Colorado 





(14 samples) 



: (12 samples) : 



(6 samples) 



: (21 samples) 







----- Percent 







Cubic foot 











4 inches 



16.6 



21 .0 



35.8 



12.0 



Board foot 











variable 



29.7 



21.5 



47.1 



21.0 



Board foot 











7 inches 



32.4 



1^2 



45.3 



21 .7 



11 



