SITE DETERMINATION AND YIELD FORECAST 11 



ference, although the difference in diameter growth may, of course, 

 be large. 



The site classification system here described has been criticized as 

 "mathematical" and "artificial," as compared with a type classification 

 upon a biological basis. From a practical standpoint this criticism is 

 believed unimportant. The basing of site determination upon whatever 

 of the upland hardwoods are dominant on an area is justified in the 

 interests of simplicity, ease of application, and relative reliability. It 

 represents the site in terms not of a single but a number of species. 

 It gives a true measure of the site as gauged by an average of the 

 height growth of several species. Different species on the same site 

 will, of course, behave differently in height growth; but until a great 

 deal more has been learned of the similarity of types in different regions 

 and their natural division into "sites within types," the present scheme 

 will furnish an immediate and simple means of classification upon which 

 growth forecasts may be made with greater assurance than without 

 such classification. 



USE OF THE SITE CXASSIFICATIOX IN FORECASTING GROWTH 



This involves two general steps : first, the determination of the site, 

 as already described, and second, the application of the available growth 

 data classified by sites. Growth data will be of two kinds, for even- 

 aged, well-stocked stands, and for stands varying in density, sge and 

 condition. 



Growth of Regular Stands. — For stands of the first sort, yield tables 

 and increment tables derived therefrom will be used. In the construc- 

 tion of yield tables for immediate use advantage was taken of a series 

 of plot measurements compiled by W. D. Sterrett in connection with 

 a study of oak but containing also a large number of measurements of 

 stands running heavily to chestnut, poplar, and other species. Of 

 these plots, 34.5 were measured in Maryland, mostly on "coaled" lands, 

 by F. W. Besley in co-operation with the U. S. Forest Service, while 

 2-5 were made by the writer in the course of the present and previous 

 studies in the Southern Appalachians. The plots were first classified 

 in terms of the general site classification by comparing dominant 

 heights ■"' according to age with the curves in the diagram. The cubic 

 yields per acre of the plots thus classified were then plotted and curved. 



^ Codominant heights were used when no measurements of dominant trees 

 were given. This was checked by comparison of codominant and dominant 

 heights when both were given. The difference was rarely found great enough 

 to influence the classification. 



