46 CIRCULAR 3 3 3, U.S. DEPARTMENT OF AGRICULTURE 



ments are not very reliable, because of weathering of the bark, 

 insect action, fire damage, checking, etc. Unless the cutting was 

 comparatively recent, allowance should be made for errors approx- 

 imately as follows : Stumps less than 6 inches in diameter, one-half 

 inch; stumps from 6 to 18 inches in diameter, about an inch; and 

 stumps more than 18 inches in diameter, about 2 inches. Because of 

 variation among species, systematic and reliable observations oh this 

 subject are needed. 



TREE DESCRIPTION AND CLASSIFICATION 



Sometimes need arises for sample-plot data of other kinds than 

 those ordinarily collected. Consideration should aways be given 

 to the desirability of collecting from plots not only data that are 

 useful from the standpoint of the current study but other data 

 that ma} 7 amplify our knowledge of the growth characteristics of 

 trees and the reactions of trees to site factors. 



Such supplemental data include those classifying trees according 

 to width, length, fullness, density, and shape of crown; color of 

 foliage; type and degree of injury; abnormalities of form; seed 

 production ; and sprouting capacit} 7 . The width, length, and density 

 of a tree's crown, and the color of its foliage, may indicate its vigor 

 and rate of growth. Records on such characteristics may be valuable 

 as a means of relating appearance of crown to growth characteris- 

 tics and behavior of the individual tree. Since stem diameter is 

 closely correlated with crown width, data on crown width may con- 

 tribute toward development of a method of estimating timber on 

 the basis of aerial photographs. 



In many cases it is desirable to develop a herbarium for a plot 

 or plot series, particularly if cover conditions are to be changed in 

 the course of the experiment. (A herbarium is of course an essen- 

 tial part of the record of each experimental forest.) Phenological 

 observations, particularly on flowering and fruiting, are an im- 

 portant feature of many plot records. At research centers, such as 

 experimental forests, phenological observations should be considered 

 an important and necessary activity. 



General terms should be avoided, and for each study complete 

 definitions of the various terms needed should be adopted at the 

 time of plot establishment. This is especially important in connec- 

 tion with describing the condition and appearance of individual 

 trees, for the reason that in later years other workers may remeas- 

 ure the plot and without adequate definitions may not be able to 

 record data in a comparable manner, to reconstruct the stand, or 

 to interpret the recorded data properly. 



Many of the descriptive data can be recorded in the field by 

 appropriate symbols or abbreviations. In many cases, especially 

 when it is possible that data will be put on punch cards for ma- 

 chine tabulation and analysis, there is much to recommend the use 

 of a numerical code, such as that given in the appendix (p. 73), 

 instead of symbols or abbreviations. 



TREE CLASSIFICATION 



The position that the crown of a tree occupies in the forest canopy 

 has an important bearing on the tree's growth and development and 



