32 



PLANT SOCIOLOGY 



cover were left out of consideration. This method, designed for use in 

 forestry, is suitable only for large uniform bodies of vegetation. 



A five-parted scale is better for the study of most communities. In 

 Scandinavia the so-called Hult-Sernander scale is most used. The 

 limits of the cover classes are as follows: 



1 = covering less than }{ e oi the ground. 



2 = covering }{q to }i (6.25 to 12.5 per cent); 



3 = covering }i io 14. (12.5 to 25 per cent). 



4 = covering }i to 3^ (25 to 50 per cent). 



5 = covering 3^ to 1 (50 to 100 per cent). 



The values of these cover classes are easily memorized. But on 

 account of the very wide range of the upper grades they give in many 

 cases no true idea of the degree of cover. Whereas the figures 1, 2, 3 

 are confined to the narrow limits of one-quarter of the surface, there are 

 available for the important larger classes — one-fourth to four-fourths 

 of the surface — only the two figures 4, 5 (Fig. 12a). 



Fig. 12. — Degrees of cover (hatched); (a) according to Hult-Sernander; {h) according to 



Braun-BlanqTiet. 



In actual field work the following scale has proved more satisfactory 

 (Fig. 126): 



1 = very scant (covering less than J^o of the ground surface). 



2 = covering ^2 o to M of the ground surface. 



3 = covering ^i to 3-2 of the ground surface. 



4 = covering }/>, to ^i of the ground surface. 



5 = covering ^i to }i of the ground surface. 



The exactness of the estimate is entirely satisfactory in small test 

 areas (up to 10 sq. m.). With large areas, however, the estimate is 

 difficult to carry out, especially when there is a mixture of many 

 species. 



AVhen dealing with communities whose cover shows great 

 seasonal variations, the estimate must be made repeatedly at different 

 seasons of the year. Under some circumstances it may be done at the 

 culmination of the seasonal development, but many communities, 



