comparison of timber volume lost as shown by the 

 reports and that compiled by the survey can be 

 made. 



After all adjustments of the data derived from 

 the individual fire reports had been made, esti- 

 mated annual averages of area deforested and 

 timber volume lost were compiled by forest type 

 and county. These averages were then summar- 

 ized by Forest Survey units. 



Methods in Growth Phase 



General procedure followed in the Forest Sur- 

 vey growth study and detail of the results obtained 

 have been previously presented in the station's 

 Forest Survey Report No. 78, "Forest Growth in 

 the Ponderosa Pine Region of Oregon and Wash- 

 ington." Nevertheless several significant details 

 of methodology are included here in the belief 

 that they may contribute to facility in interpreta- 

 tion of results given in the text of the present 

 report. 



Gross Growth Study in Virgin Stands 



This study was undertaken primarily to appraise 

 the prospects of net growth, under management, 

 from the region's extensive virgin timberland areas. 

 In so doing 323 growth samples were taken dur- 

 ing the summers of 1935 and 1936 throughout 

 the virgin saw-timber types. Plots were so located 

 on type maps in the office as to sample with reason- 

 able uniformity the principal virgin stands in the 

 region. Number of samples taken within a type 

 was determined roughly by the type's extent and 

 economic importance. 



The samples, taken in strip form, consisted of 

 stand tallies by tree class, growth rates by tree 

 class being determined from increment cores of 

 sample trees. The sample unit consisted of 1 day's 

 work for the crew of three — usually 16 acres but 

 varying from 8 to 32. Average annual growth rate 

 was based upon the period 1900 to 1935 or 1936, 

 depending on date of sampling. Growth rate of 

 ponderosa pine is characteristically cyclic. On the 

 basis of Keen's tree-ring analysis (9) supplemented 

 by additional region-wide investigation as a part of 

 the survey growth phase, these periods were selected 

 as ones during which actual growth approximated 



the normal rate. The growth rates employed were 

 the gross rates made by trees that survived to the 

 date of measurement. Growth of trees that grew 

 for part of the increment period, but succumbed 

 before measurement, could not be included. On 

 the other hand it is believed that the average incre- 

 ment of the trees that survived was superior to 

 that made by those that did not. Biases involved 

 by these two factors tend to be compensating. 

 Average annual increment of poles that grew into 

 saw-timber size during the growth period was of 

 course included. 



Sample stand tallies were converted to stock 

 tables by use of height curves, prepared for each 

 sample, and regional volume tables. For the 

 sample trees grouped into broad classes 10 ratios 

 of volume as of 1900 to volume as of 1935 or 1936 

 were computed for each sample. These ratios 

 were applied to the appropriate portion of the 

 stock table to compute estimated stand volume 

 as of 1900. This value was subtracted from 

 volume found at time of survey; the remainder 

 divided by years in the growth period and acres 

 in the sample gave a measure of the average 

 annual growth per acre made by the sample stand. 



Methods of Analysis 



Preliminary study of the samples indicated sig- 

 nificant differences in growth rate by forest type 

 and by the various stand variables shown below 

 by type. Alinement charts were constructed for 

 each type as a means of quickly estimating growth 

 corresponding to any combination of stand 

 variables encountered. The charts for the pine 

 woodland type were constructed by the empirical 

 alinement chart method (4), those for all other 

 types were made by the mathematical-graphical 

 method of successive approximation (3). 



Making Mortality Estimates 



As employed in the growth estimates, mortality 

 in virgin stands was dismissed by assuming loss 



10 The grouping into Dunning tree classes (5) approxi- 

 mated that employed by Meyer {11). The Keen tree- 

 classification system (8) widely adopted in the region 

 recently was not devised until after completion of the 

 growth-phase field work. 



91 



