food availability reduction at high-canopy closure 

 levels (fig. 2). This index of food availability is more 

 easily obtained than foliage volume and appears to 

 be as useful an indicator of food availability for 

 gleaners. 



This study also illustrates the problems associated 

 with using Forest Survey data for potential wildlife 

 habitat assessment. One problem is in using vari- 

 able-radius plots that sample trees proportional to 

 basal area to get information on snags for wildlife. 

 A study is under way to compare snag information 

 obtained with variable-radius plots to actual snag 

 counts. 



Habitat variables were tested by comparing differ- 

 ences in bird use of different foliage volume and 

 snag density conditions. However, differences due 

 to study area overshadowed any differences within 

 study area. The threefold increase in bird numbers 

 from Kaibab to Trumbull can possibly be explained 

 by the greater diversity, patchiness, and layering of 

 the vegetation, the importance of which was re- 

 ported by Langelier and Garton (1986). Reduced 

 numbers of gleaners at Kaibab support the findings 

 of Medin (1985) that gleaners are less numerous on 

 logged plots where foliage volume has been reduced. 



The difference due to study area was observable 

 but was difficult to detect using Forest Survey 

 methodology, except that Trumbull plots had 

 slightly higher mean canopy cover and food availa- 

 bility than Kaibab plots. There are two possible 

 reasons. First, a subset of Forest Survey plots were 

 selected and supplemental plots were located to 

 provide a wide range of foliage and snag conditions 

 for this study, thus overriding the inherent differ- 

 ences in the study areas. Second, analysis of addi- 

 tional Forest Survey plot data indicated that both 

 sites have similar percentages of plots on which 

 evidence of logging, occurrence of thickets, and mul- 

 tiple vegetation layers were reported. Forest 

 Survey data are somewhat insensitive to subtle dif- 

 ferences in vegetation structure and distribution, 

 due to the general nature of the variables measured. 

 The difference between areas was related to details 

 of horizontal and vertical distribution of the vegeta- 

 tion to which the birds were apparently more at- 

 tuned but which would require much more detailed, 

 time-consuming methodology to identify. 



There are problems with using animal numbers 

 to assess habitat suitability. Van Horne (1983) 

 reported that density may sometimes be a mislead- 

 ing indicator of habitat quality. Density may reflect 

 temporary or recent conditions rather than long- 

 term conditions, social dominance may induce high 

 densities in poor habitats, and censuses may be 

 obtained in noncritical seasons. Diem and Zeveloff 

 (1980) found that the movements of birds can reflect 



local perturbations that may be temporary re- 

 sponses to short-term environmental factors and 

 that are impossible to monitor. Therefore, high 

 correlation should not be expected when comparing 

 habitat suitability to bird numbers. 



Lack of a high degree of correlation between bird 

 numbers and habitat variables does not mean that 

 the other variables included in the study would not 

 be useful indicators of wildlife habitat potential. 

 Even though the value of snags was not demon- 

 strated in this nonbreeding season study, the impor- 

 tance of snags for cavity-nesting birds and other 

 animals in general is well known (Thomas and 

 others 1979). Recent publications emphasize snag 

 management and Forest Service guidelines for 

 retention of snags (Langelier and Garton 1986; 

 Morrison and others 1986). A survey of snag densi- 

 ties on other northern Arizona Forest Sui^^ey plots 

 indicates that snag densities are rarely at recom- 

 mended levels. None of the sites in this study con- 

 tained recommended (Cunningham and others 

 1980) snag densities. Morrison and others (1986) 

 concluded that current guidelines for large snag 

 retention are appropriate, but even under snag 

 management strategies, recommended densities are 

 not being met. Although snag densities alone will 

 probably not be a useful indicator of abundance of 

 many bird species, general information on snags 

 from statewide inventories would be useful for 

 monitoring snag densities and, consequently, habi- 

 tat potential, on all forest lands. 



SUMMARY 



Forest Survey projects throughout the country 

 collect and maintain statewide multiresource data 

 bases that may be useful for assessing and monitor- 

 ing elements of wildlife habitat or trends over time 

 in general forest conditions that have relevance for 

 wildlife populations. One problem is that Forest 

 Survey data are primarily collected to generate re- 

 gional timber resource statistics, whereas most 

 wildlife habitat research is done on a site-specific 

 level. However, the State and regional level at 

 which Forest Survey operates is also the level at 

 which important information on the condition and 

 diversity of the Nation's resources must be moni- 

 tored and at which program funding takes place. 



This study involving bird numbers illuminated 

 some of the problems involved in habitat assess- 

 ment. It also provided some evidence that Forest 

 Survey data could be used as a starting point in 

 identifying key issues or problem areas that need 

 more site-specific studies. There is potential to use 

 Forest Survey data to assess the capacity of the 

 forests of an area to support insectivores or cavity- 



6 



