18 



(1988), are cluster analysis and discriminant function analysis. These 

 two methods were applied to test the reliability of the FTI numbers. 

 FTI numbers were calculated for each species occurring in each 

 vegetation layer. A species could have three different FTI numbers at a 

 given site, depending on its growth form. For example, Quercus nigra 

 would have three different FTI numbers when present on a site as a tree, 

 sapling, and seedling. Species FTI numbers for each site were computed 

 by the following formula: 



S-i(j • IV,..) 



where 



i = the i th species 

 j a = 2.5, 3. 5... 6. 5 (hydrologic zone) 

 IVij = importance value for species i in the hydrologic zone j 

 After species FTI numbers were computed for all species in all sites, 

 the average FTI number (FTI) for each species across all sites was cal- 

 culated using the following formula: 



E. ni . FTI,, 

 J=i ij 



FTIi = 



n< 



Because vegetation was sampled between zone boundaries, midrange zone 

 numbers (e.g., 2.5 for Zone 2, 3.5 for Zone 3, etc.) for zones were 

 used in calculating FTI numbers. 



