SITE CLASSES AND AVERAGE HEIGHTS OF TIMBER 291 



The result of running the five strips at regular intervals is to include within 

 type I, 30 acres, which at 30,000.board feet per acre would give 900,000 board feet. 

 In type II, 20 acres was included which at 10,000 board feet gives 200,000 board 

 feet, a total for the 50 acres run, of 1,100,000 board feet. As this is 25 per cent of 

 the area, the required factor for the tract without subdivision into types would 

 be a multiple of 4, giving an estimate of 4,400,000 board feet, an error of +10 per 

 cent caused not by errors in the strip but by failure to get the weighted average 

 stand from the strips run. 



But if while running these same strips the tally sheet had been changed wherever 

 the strip passed from one of these types to the other, and both the map of the area 

 and the corresponding estimate of the timber, or tally, had thus been separated 

 into two areas, corresponding with each of the two types, the computed estimate 

 would show that while on 30 acres 900,000 board feet was tallied the average acre 

 for type I is 30,000 board feet, but instead of this applying to three-fifths of the total 

 area, it applies only to the actual area shown to be in the type, or one-half of the total, 

 which is 100 acres, totaling 3,000,000 board feet. The less fully-stocked type in 

 the same way is shown to contain 1,000,000 board feet or a correct total for the tract 

 of 4,000,000 board feet. The 10 per cent error incurred in the first method is elimi- 

 nated. The accuracy of this area correction obviously depends first upon ability 

 to obtain by sketch a correct map of the actual areas of the different types, and 

 second, to convert this area from the map into acres by use of the proper methods 

 of map reading as explained in this paragraph. 



This system of type divisions is of especial value in mountainous regions where 

 sharp distinctions can be drawn between types coinciding with great differences 

 in the average density, volume, size and value of the timber. Under such circum- 

 stances the more valuable types would require a greater per cent of the total area 

 to be estimated, to obtain the same basis of accuracy as could be secured for the 

 less densely stocked and less valuable tracts with a smaller per cent. The type 

 divisions also are more conveniently made in large or irregular areas than where 

 estimates are separated by rectangular tracts of 40 acres. 



227. Site Classes and Average Heights of Timber. Differences 

 in the quality of the site on which timber is growing cause very great 

 differences in total volume per acre, and in the total heights of the 

 trees and stands. To quite an extent these differences are closely 

 correlated with changes in cover types, different types being found 

 on wet soils, fresh well-drained soils, and dry, shallow soils. But it 

 often happens that the same type of forest cover will extend without 

 appreciable changes in composition over a range of site quality so great 

 that it becomes necessary to subdivide the area within the type into 

 from two to three site classes, ranging from good to poor. This is 

 made necessary by the effect of site" upon the height of the trees in the 

 stand, on account of the methods usually required, of selecting sample 

 trees to measure for height. 



Heights constitute an extremely variable factor in timber estimating. 

 Not only do total heights range through limits of at least 100 per cent 

 for the same diameter, but merchantable heights, especially in old hard- 

 woods, vary still more widely. Just as, in a 100 per cent estimate, 

 the necessity for averages is eliminated, so when the height of every 



