60 



predict the degree of flooding for the entire southeastern United States 

 previously have been literature -based involving many studies with vary- 

 ing research designs. 



Vegetation data resulting from this study related four vegetation 

 strata and three life forms occurring in 55 stands at 17 sites through- 

 out the southeastern United States. Hydrologic regimes were calculated 

 for a 10- to 20-year period of record for each stand. A flood tolerance 

 index (FTI) system of weighted averages based on importance values was 

 developed, and FTI numbers were calculated for various life stages of 

 each species identified in the study. 



Three hundred and twelve species were identified for each of 

 4 strata in the study including 74 tree species, 188 species of saplings 

 and shrubs, 31 species of woody vines, and 268 species of herbs and 

 woody seedlings. Comparison of the FTI numbers with two other systems 

 (Hook 1984; Reed 1988) using vegetation to estimate wetness showed gen- 

 eral agreement among the systems, especially for mature trees. 



Cluster analysis and discriminant function analysis were used to 

 evaluate the weighted averaging technique and explore the best method 

 for using the FTI numbers in predicting hydrologic regimes. 



Tree, sapling, and vine data clustered into distinct groups. Her- 

 baceous and shrub data did not group distinctively. Tree and vine 

 importance values for each cluster in a zone/sample (data taken in a 

 single zone at a site) and FTI numbers for tree data were used as inde- 

 pendent variables for the discriminant function analyses. Tree species 

 were found to be more useful than saplings, shrubs, vines, or herbaceous 

 species in predicting hydrologic zones. The tree data alone using 

 importance values provided 85 percent accuracy. Tree data alone using 



