CLASSIFYING FOREST SITES 375 



part of it, and for any tree species for which certain elementary data 

 are available. On the principle that the rate of height growth of 

 dominant trees is a practical index of the site quality (when used with 

 trees which have passed the juvenile stage), Roth suggests choosing 

 a single uniform age (100 years) as a reference point and classifying 

 the sites in terms of the height attained at that age by dominant trees. 

 He would distinguish between tall, medium, and short species, and 

 for each of these groups would supply a set ("standard") of site 

 classes. The table used for illustration in his paper is as follows : 



Years 



I 



II 

 III 



IV 



These standards and sites within standards would, of course, be 

 based upon the performance of representative species. The table cited 

 is merely illustrative. Thus the maximum height given (110 feet) 

 may prove to be too small, or the three standards may be too few to 

 provide for all our important species (although a consideration of the 

 growth habits of these species will bring out many resemblances, mak- 

 ing it appear probable that all could be provided for by a very small 

 number of standards).^ It is not essential that every one of the species 

 composing a "standard" (i.e., reaching approximately the same maxi- 

 mum height on the best site) should be represented in every one of the 

 sites within that standard. If, for example, yellow poplar and chestnut 

 belong to the same standard (being rather similar in growth rate on 

 soils occupied by both), the chestnut would be found on all the sites, 

 from site 1 (cove) to site 4 (ridge), while the poplar would be limited 

 to the sites representing cove and lower slope conditions. Great im- 

 portance is placed by Roth upon the need for simplicity in any com- 

 prehensive plan of classification. This plan is, therefore, pretty well 

 stripped to its bare essentials. 



THE DOMINANT HEIGHT AS A SITE INDEX 



From a theoretical standpoint the true index of site is doubtless 

 the volume per acre. The possible yield per acre in cordwood or lum- 



'In the writer's opinion, however, each forest region might have its own site 

 system, related to those of other regions by the common method employed. 



