45 









Table 



8 









Relative Frequency of Occurrence i 



of 



Each Sanlins Species in 



the 





Hvdrologic 



Zones along 



with ' 



rheir Cluster 



Memberships 





Species 



2 

 99.4 



3 

 0.6 



4 

 0.0 





5 

 0.0 



6 



Cluster 



SANI 



0.0 



1 



FRCA 



61.0 



21.2 



2.7 





0.0 



15.1 





ACRU 



1.4 



35.8 



35.1 





4.6 



23.0 





FRPE 



0.0 



50.0 



21.9 





21.9 



6.3 



2 



QULA 



4.7 



53.3 



36.7 





3.6 



1.8 





ULAM 



0.0 



2.0 



42.1 





32.2 



23.7 



3 



CELA 



0.0 



13.1 



47.5 





19.5 



19.9 





OSVI 



0.0 



0.0 



0.0 





71.2 



28.8 



4 



CODR 



0.0 



0.0 



1.5 





80.0 



18.5 





ULAL 



0.0 



0.0 



0.8 





17.7 



81.5 





QUNI 



0.0 



0.4 



14.2 





11.1 



74.2 





CATO 



0.0 



0.0 



0.0 





0.0 



100.0 





NYSY 



0.0 



0.0 



24.7 





22.7 



52.6 



5 



LIST 



0.0 



5.4 



24.2 





14.6 



55.8 





ILOP 



0.0 



0.0 



55.5 





5.0 



39.5 





CACA 



0.2 



1.4 



40.6 





15.7 



42.0 





The herb data also did not cluster distinctly (Figure 10) . 

 Inspection of the abundance distributions of the herbaceous species 

 reveals an absence of strong unimodal peaks that correspond to hydro - 

 logic zones; therefore, this vegetative type is less useful in classifi- 

 cation. The lack of significant clustering of the shrub and herbaceous 

 data sets was consistent with the results of ordination attempts of the 

 ground layer at Neches River sites by Mohler (1979). Mohler concluded 

 that the herbs and low shrubs were relatively unimportant components of 

 the forest when compared to the trees. Clearly, the low shrubs and 

 herbaceous species are more prone to bias due to successional processes 

 and vegetative change caused by local disturbances which may be 

 unrelated to the environmental gradient. 



