abrupt changes or steep gradients in the substratum, by reactions 

 of plants to shading, or by fire, grazing, or cultivation. Delimita- 

 tion is more difficult when the gradients are gradual, and abrupt 

 changes in the vegetation do not appear. However, as the gradient 

 changes, due, for example, to a decrease in soil moisture, some 

 species become less numerous and others become more so, creat- 

 ing an effect comparable to the gradual changing of color in the 

 spectrum, in which measured portions have been delimited as 

 blue, yellow, red, etc. 2°° Similarly, the vegetation on a gradient 

 (a cline) can be quantitatively segregated into units on the basis 

 of measurements of numerical abundance, cover, and frequency, 

 especially of the constant and characteristic species (see p. 125). 200 

 For example, in the lower part of a cline Agropyron smithii and Stipa 

 viridula may each average 25 per cent or more in foliage cover 

 and 100 per cent in frequency, and in the same part Bouteloua 

 gracilis and Stipa comata may each rate less than 10 per cent in 

 cover and 40 per cent in frequency. As one ascends the slope the 

 former pair gradually decreases in cover and frequency, while 

 the latter increases, until the figures may be reversed, thus form- 

 ing a different stand. The demarcation of the two stands may be 

 placed where the cover of the second pair exceeds that of the first, 

 and where the frequency of the second pair rises above 80 per 

 cent while that of the other pair falls below this percentage. The 

 portion of the cline where the data for all four species are closely 

 similar may be considered as another stand, or a transitional zone. 



Transition zones, as a rule, are readily recognized by the pres- 

 ence in them of species from each of the adjacent stands, and often 

 by changes in structure and greater heterogeneity. It is usually 

 advisable to delay analysis of these zones until a frame of refer- 

 ence of well-marked stands has become available. It is also good 

 procedure to make a reconnaissance study of an extensive area 

 of vegetation first before beginning intensive analysis of the con- 

 stituent stands, so that a general idea of the different kinds of 

 stands or types of vegetation may be gained. 



Description of a stand, as of other entities, should be based upon 

 an analysis which reveals as many of the intrinsic qualities as pos- 

 sible. It appears obvious that vegetation should be characterized 



172 • Classification of Communities 



