664 



In figure 3, line A applies to statistics obtained 

 by area sampling. This includes private owner- 

 ships of less than 5,000 acres in all regions except 

 the Pacific Northwest, Northern Rocky Mountain, 

 and Southern Rock}' Mountain, where list samp- 

 ling was employed. Line B applies to statistics 

 obtained by list sampling. This includes all owner- 

 ships in the size range 5,000 to 50,000 acres and 

 small ownerships in the Pacific Northwest, North- 

 ern Rock}' Mountain, and Southern Rocky 

 Mountain Regions. Line B will apply also in 

 approximating the sampling errors of breakdowns 

 of all ownerships of more than 5,000 acres. 



nonsampling errors may have 

 Occurred 



In addition to sampling errors, the productivity' 

 data are also subject to other errors. These in- 

 clude possible mistakes in measurement, tallying 

 and reporting, or compilation — the same possible 

 mistakes previously mentioned in connection with 

 other groups of data. However, tl'e principal 

 factor affecting reliability of the data was the use 

 of personal judgment in appl3dng the criteria. 

 Judgment had to be used in classifying crop trees 

 and seed trees, determining prospective stocking, 

 appraising the condition of the seedbed, and so on. 

 Finally, the productivity data are dependent upon 

 the ownership data and their reliability is partly 

 dependent upon the reliability of the latter. 



The productivity criteria are believed to be 

 adequate for rating purposes. Separate criteria 

 were established for each forest type in every 

 region, and for site classes, physiographic units, 

 or localities within types as deemed necessarj-. 

 In establishing them, tables of normal stocking 

 and other technical benchmarks were adjusted to 

 conform to the conditions found on recently cut 

 lands of ownerships judged to be managed under 

 the better cutting practices. Although the re- 

 liability of the criteria is not under consideration, 

 it should be pointed out that professional judg- 

 ment was exercised in coordinating standards for 

 types common to two or more regions, and in 

 determining the need for separate criteria for site 

 classes, physiographic iinits, or localities within 

 a type. 



FOREST LAND AND TIMBER 

 OWNERSHIP 



The commercial forest land ownership data are 

 among the most reliable data in the Timber Re- 

 source Review. They are adequate for most 

 breakdowns, and their reliability is frequentl}' 

 sufficient for State analysis. The timber owner- 

 ship data, though reliable enough for use in most 

 States, are not quite as adequate. Thej' cannot be 



TIMBER RESOURCES FOR AMERICA'S FUTURE 



broken down as finely as the area data, and there 

 is no way of separating forest industry timber 

 volumes from volumes owned by nonfarm private 

 owners. 



Ownership Data Tied to Forest 

 Survey 



Total commercial forest area, total sawtimber 

 volume, and total gro^\^ng stock estimates were 

 obtained as explained above under Forest Area, 

 Timber Volume, and Growth Data. Ownership 

 procedures were aimed simply at segregating 

 these groups of data by ownership class in terms 

 of acreage, timber volume, and number of holdings. 



Most of the estimates of private ownership of 

 commercial forest land were obtained as part of 

 the Forest Survey or as part of special surveys of 

 forest area, timber volume, and growth. Public 

 forest land areas were usually secured from the 

 officials administering them. Private farm forest 

 areas were derived from Census estimates. Forest 

 industry and other private forest area was the 

 calcidated residual. The subdivision of these 

 latter data into ownerships of lumber manufac- 

 turers, pulp manufacturers, other wood manu- 

 facturers, and other private owners was accomp- 

 lished as a part of the survey of productivity of 

 recently cut lands, just described. This siirvey 

 also provided the estimates of number and area of 

 private ownerships by size class. 



Estimates of timber volume were based on 

 public records or on aerial photograph interpre- 

 tation and ground plot measurements. Regular 

 Forest Siu^vey procedures were used in most cases, i 



Sampling Errors Apply to Part of 

 Ownership Data 



Except for the area of public ownerships and 

 the numbers of medium and large private owner- 

 ships, ownership data were obtained by sampling 

 procedures. For the estimate of private com- 

 mercial forest land in the United States and 

 Coastal Alaska, the sampling error is 1.4 percent 

 (table 84). For the total number of small private 

 ownerships, the error due to sampling is 2.0 per- 

 cent, and for the commercial-forest area of these 

 ownerships it is 1.9 percent. Further breakdowns 

 of these items can be calculated by using figure 3 

 as explained in connection with the productivity 

 data. 



The sampling error of total sawtimber volume 

 is 0.8 percent, and the sampling errors for volumes 

 owned by each of the various types of ownership 

 are somewhat larger. The latter errors can be 

 read from table 84 and figure 1 in the same manner 

 as errors of other breakdowns of the forest land, 

 timber volume, and growth data. 



