RELATIONS AMONG SPECIES 



369 



vironments and dominants. Hence, dominants can 

 and have changed from one small taxon to another 

 and eventually from one major taxon to another. 

 For example, tropical rain forests have changed from 

 the Mississippian lycopsid-sphenopsid-pteridosperm- 

 dominated types, to fern-cycad-conifer types, and 

 finally to the present flowering plant dominants. 

 (Pteridosperms, or seed ferns, were fern-like plants 

 that produced seeds.) However, with each replace- 

 ment of type, there still is a tendency for the subordi- 

 nate species or their evolved representatives to re- 

 main. For example, the tropics of today, areas prob- 

 ably most typical of ancient environments, are the 

 usual habitats for living representatives of ancient 

 plant groups such as psilopsids, lycopsids, sphenop- 

 sids, ferns, cycads, and primitive conifers. There- 

 fore, when subordinate plants or animals evolve or 

 disappear in direct relation to dominants it usually 

 must be the result of some intimate biological re- 

 lationship, rarely of chance alone. 



Another way of looking at these observations is 

 in relation to the persistence of the various units, com- 

 munities (serai and climax), biomes, and biome- 

 classes. Although biome-classes will not be con- 

 sidered until later, one can now point out that they 

 are more persistent than biomes and biomes are likely 

 to last longer than their communities. Particular 

 communities may last millions of years; but because 

 they are only a part of a biome, communities can be 

 lost because any of the factors causing biomes to 

 evolve also tend to destroy some of the components 

 of a biome. 



Persistence at the community level normally favors 

 serai stages over climaxes. Serai stages are more 

 flexible; they persist while climax stands are dis- 

 turbed or destroyed. This is true even though serai 

 stages regularly are grouped in an orderly sequence 

 (succession) from the community first invading an 

 area (primary serai stages) to the climax. Climaxes 

 require more time to develop; they are a later stage in 

 evolution, a stage most closely adapted to the existing 

 environment. The differences between developmental 

 stages and climaxes are emphasised at the species 

 level. Serai species often occupy larger areas than do 

 climax species. Although individual species may be 

 serai in one biome and climax in another, typically, 

 serai- restricted plants and their associated animals 

 inhabit greater expanses of the world than do climax 

 plants and their allied animals. As a consequence, 

 serai stages generally occupy more different biomes 



than do climaxes. Again, this probably reflects less 

 stringent ecological requirements on the part of serai 

 species as compared with climax species. Hence, 

 serai species might not be as specialized for success 

 in restricted environments as are climax species. 



Further differences between serai and climax 

 stages are significant in biogeographical succession. 

 The fact that serai stages normally are widely dis- 

 tributed throughout biomes, or even biome-classes, 

 in a sense causes serai communities to be more im- 

 portant biogeographically. One consequence of their 

 occupying greater space is an increase in probability 

 that serai stages will be near a new highway. The 

 opening of such highways causes them, to a greater 

 extent than climaxes, to have remote, discontinuous 

 distributions. Also, wide-ranging animals probably 

 disperse through serai stages rather than climaxes. 

 Another consequence of larger ranges leads to serai 

 stages having greater potential for evolution. This 

 potential comes from greater environmental diversity 

 throughout their ranges and more opportunity for 

 isolation of individual stands of the stages. 



To summarize, the factors that produce biomes and 

 biome-classes are complex. The factors do not form 

 identical isolated subunits. However, biomes such 

 as the Tundra can have a continuous distribution, 

 and in such cases the various geographic segments 

 of the biome are similar. It is only when discontinu- 

 ities in a biome or biome-class exist that taxonomi- 

 cally unlike communities might develop. The degree 

 of dissimilarity of isolated subunits, as in the Tropical 

 Rain Forest, occurs in direct proportion to the 

 amount of extinction, diflerential evolution, and 

 differential immigration in the various segments. 

 This means of causing dissimilarity of isolated seg- 

 ments is also the means for unique rates of change 

 in different members of a single taxon. Moreover, 

 it is this independent evolution within taxa that 

 causes biomes not to change in tola, but to evolve 

 gradually by replacement and evolution of individual 

 large taxa. Therefore, biogeographical succession is 

 the consequence of evolution of individual large taxa 

 and not of biomes per se; for this reason biomes are 

 organizations of unrelated taxa. 



BIOME-CLASS 



Further changes in a biome can come from replace- 

 ment of one or more of the climax dominants. Al- 

 though such substitution often does not alter the in- 



