61 



FTI numbers was only slightly less accurate at 82 percent. All misclas- 

 sifications assigned membership to a neighboring zone. Misclassif ica- 

 tions are understandable for two important reasons. Zone 4 contains the 

 more facultative species because as wetness decreases, other environmen- 

 tal conditions begin to exert greater influence. Also, since Zone 5 is 

 so narrow compared to other zones and most species occurring in Zone 5 

 occur in greater abundance in either Zone 4 or 6 , very little difference 

 in the community structure exists between the top of Zone 4 and the 

 bottom of Zone 6 . 



The accuracy of these predictions may be somewhat inflated, 

 because hydrologic zone was a parameter used to derive species FTI 

 numbers . 



There were no regional (Gulf Coast, Lower Mississippi Valley, and 

 Atlantic Coast) differences in the accuracy of the weighted averaging 

 and predicted values. Therefore, a single FTI number calculated for 

 each species can be used to predict zones for the entire study area. 



The implication of this study is that the calculated FTI numbers 

 can be used to estimate hydrologic regimes in bottomland forest systems 

 of the southeastern United States. Trees were determined to be the most 

 reliable vegetative growth form for determining hydrologic zones. How- 

 ever, this study was conducted in relatively undisturbed areas. Because 

 trees can remain for decades following hydrologic disturbance, a modifi- 

 cation of the method using saplings and seedlings may prove to be more 

 reliable . 



Techniques used in this study to develop FTI numbers in the south- 

 eastern United States may be applicable to regions of the country that 

 have similar types of riverine forest systems. 



