Origin of Biomels and Succession 259 



the area by the dominant trees. The large number of both bromeUad 

 and mosquito genera and species occurring in and restricted to 

 this biome is testimony of the long evolutionary persistence of the 

 physical environmental conditions necessary for their existence. 



It is possible that the understory club mosses in tlie present-day 

 tropical rain forests have had an even longer unbroken history in 

 their ecological situation. Species of Lycopodium may trace back 

 both phylogenetically and ecologically to their possible progenitor 

 Lycopodites, a fossil genus associated with tropical rain forests of 

 Pennsylvanian age. If so, the club moss component of the under- 

 story has evolved little, even though in a taxonomic sense the 

 dominant tree species have changed radically many times. 



Certain understory herbs of the northern boreal coniferous forest 

 illustrate the independence of dominant and subdominant species 

 in a geographic transect as opposed to a time transect. In the 

 Sierra Nevada mountains of western North America this forest 

 comprises the red fir association overwhelmingly dominated by red 

 fir Abies magnifica and having small percentages of four other 

 large dominant trees, Tinus monticola and contorta, Abies concolor, 

 and Tsiiga mertensiana, in almost every stand (Costing and Bil- 

 lings, 1943). The dominant species have only a comparatively limited 

 distribution outside the association. Seven small subdominant herbs 

 are especially characteristic of the floor of this forest. Of these only 

 Sarcodes sangiiinea is restricted to the vicinity of the Sierra Nevada 

 (Fig. 114). Pirola dentata var. Integra and P. picta extend north- 

 ward and eastward to British Columbia and Wyoming; Coral- 

 lorrhiza macuJata and Pterospora andromedea are ti'anscontinental 

 through the boreal coniferous forest; and Pirola secunda and Chima- 

 phila timbellata extend throughout practically the entire boreal 

 coniferous forest biome in both North America and Eurasia. Thus 

 four of the seven characteristic subdominants of the red fir associa- 

 tion studied by Costing and Billings (1943) have ranges greatly 

 exceeding those of any of the dominant species of the association. 



From these examples it would seem that, as long as the com- 

 munity dominants were of the same physical type, taxonomic 

 changes in dominants have no effect on the general selection pres- 

 sure exerted on understory species. This means that, as long as 

 ecological constancy prevailed, the evolution of the dominants and 

 the evolution of the subdominants would be independent. If a 

 rare dominant tree species became endowed with some biological 

 advantage resulting in that species supplanting another tree in 

 the community hierarchy of dominants, this change would have 



