74 BASES AND CRITERIA. 



in climax areas wherever inequalities of surface structure and so forth occur. 

 The term alternation is applied to two types of structure, one in which the 

 same dominant or subdominant recurs from place to place, the other in which 

 two or more alternate over the same area. The first kind is usually serai, the 

 second is typical of associes or associations, and also of socies and societies. 

 Recurring alternes are clear-cut indicators of the same set of conditions, and 

 are of the greatest value. Striking examples are found in the burn alternes 

 of aspen or lodgepole in the Rocky Mountains. Alternating dominants or 

 subdominants are Ukewise indicators of their respective habitats. As indi- 

 cators, they are naturally less sharply set off from the related dominants, 

 but this is compensated by the evidence afforded of the degree of their equiv- 

 alence (plate 12, a). 



Layers. Layers are best known in forests and the term has usually been 

 restricted to the subordinate communities in them (Hult, 1881; Clements, 

 1916 : 15). With the increasing study of root-systems and their competitive 

 relations, it seems desirable to recognize root-layers as well as shoot-layers. 

 Our knowledge of the former is still rudimentary, but it is possible that they 

 are more general and significant than the well-known layers of woody com- 

 munities. It is almost axiomatic that a layer of either type will have a double 

 indicator value. It indicates the general equivalence with reference to the 

 controlling factor of all the important species in it. Conversely, it denotes 

 the dissimilarity of the adjacent layers and marks a certain stage in the 

 progressive modification of the controlling factor from its point of maximum. 

 Layers also serve to indicate the course of serai development, in that they are 

 generally absent during the initial stages. They appear during the medial 

 stages and usually reach a maximum in the subclimax or climax, often dis- 

 appearing in woody communities as they become mature. As a consequence, 

 the presence of several layers indicates more or less optimum conditions as to 

 water or Ught or both (plate 12, b). 



Root-layers are regularly determined by water-content, though soil-air 

 and perhaps solutes also must sometimes be taken into account. In saline 

 soils they are due to differences in the salt-content acting through its effect 

 upon water-content, except where the salts are chemically injurious. As to 

 water-content, root layers may be a response to the physical distribution as 

 determined by penetration and evaporation, or to the ecological consequences 

 of competition. In the great majority of soils, both causes play a part (cf. 

 Cannon, 1911; Weaver, 1919). Many communities show a striking correla- 

 tion between the demands of the shoot and the root-position. This is often 

 expressed in the corresponding development of root and shoot as well. It is 

 best exemplified in the desert scrub, in which the tall shrubs are most deeply 

 rooted, the undershrubs 'ess deeply, the perennial herbs still less deeply, while 

 the low annuals of the rainy season are rooted only in the first few inches. 



The obvious relation of shoot-layers is to light, though water-content and 

 humidity must sometimes be taken into account also. The best development 

 of layers is found in well-lighted forests with a Ught intensity between 0.1 and 

 0.02. The midsummer values are rarely conclusive, however, as the layers 

 tend to develop in the order of increasing height, with the result that each 

 layer receives the maximum during its period of major activity. Each layer 

 thus has two indicator values, one when it is uppermost and another when it 



