APPLICATION OF PHYTOSOCIOLOGICAL TECHNIQUES 3OI 



under the surface of the substratum. The commonly used subclasses of hydro- 

 phytes and helophytes, aquatic and marsh plants, are not applicable to the 

 present study, and the geophytes, with their buds buried in the soil, are 

 scarcely represented in rain forests. 



Therophytes. These are the annual plants in which all parts die except 

 those that survive in seeds and fruits. The perennating bud is that of the 

 embryo in the seed. No other meristematic tissues survive. 



It is seen, then, that Raunkiaer's system meets the criteria set out: it is 

 a unitary system, being based completely on the protection afforded the 

 perennating buds. The characters are structural, essential, and adaptive. 

 Furthermore, they provide a simple basis for statistical treatment. The array 

 of percentages of the life-form classes of the flora of any area or the species 

 composing any community is referred to as the biological spectrum. Be- 

 cause each class is given a percentage representation, different spectra can 

 be compared directly. 



The percentages of phanerophytes in different floras range from zero to 

 over 90 per cent in tropical rain forests, the latter being characteristic of 

 "phanerophytic climates." Deserts have been referred to as "therophytic 

 climates" because the percentage of that class may reach more than 40. 

 Relatively high percentages of chamaephytes often occur at high latitudes 

 and altitudes, sometimes exceeding 50 per cent. Temperate forest vegetation, 

 despite its domination by trees, often has high percentages of hemicrypto- 

 phytes. Geophytes (cryptophytes) are relatively abundant in regions of 

 Mediterranean climate and in the spring flora of broad-leaf-deciduous forests. 



Because there are great variations among the biological spectra of different 

 areas and there is no inherent "yardstick" or base line for comparisons, 

 Raunkiaer devised a "normal spectrum" based on a sampling of the world 

 flora using 1,000 entities. The question is unimportant whether the normal 

 spectrum is representative of the world flora, for in any case it provides 

 a base line from which departure of the percentage of any class for any flora 

 can be ascertained. The normal spectrum has 46 per cent phanerophytes, 

 9 per cent chamaephytes, 26 per cent hemicryptophytes, 6 per cent 

 cryptophytes, and 13 per cent therophytes. Many tables of comparison of 

 biological spectra and discussions of their significance are to be found in the 

 papers of Raunkiaer and in the review by Cain previously cited. 



It is a general observation that leaves are larger the more favorable the 

 conditions of growth, and smaller in dry or cold climates. Even on a single 

 plant the more exposed leaves tend to be smaller than the protected ones. 

 There are, however, very few data available on leaf sizes, especially as a 

 characterization of a flora or community. As early as 1916 Raunkiaer published 

 on the use of leaf size in biological plant geography, and although the system 

 is relatively simple and amenable to statistical treatment of floras, little use 



