288 Origin of Biomes and Succession 



to suppose that they would be able to initiate growth on adjacent 

 bare areas having even more rigorous conditions. 



(2) New colonizing communities contain only a few of the 

 species present in their respective parent communities. Examples 

 of this thesis are found in the large genera of North American 

 grasses associated with early stages of succession in temperate 

 deciduous forest and adjacent grassland biomes. Stipa and Aristida 

 have many species in both mesic and xeric communities, Panicum 

 has a very large number in mesic communities but few in xeric 

 ones, and Festuca and Paspalum are restricted almost entirely to 

 mesic communities. If these five genera originated as mesic types 

 (as seems likely), then originally several species of Stipa and Aris- 

 tida spread into and colonized previously uninhabited areas, while 

 only a few of Panicum and none of Festuca and Paspalum did so. 

 In these same communities grass-feeding leafhoppers (Cicadellidae) 

 parallel the grass examples. In particular, the genera Deltocephahis 

 and Hebecephalus have many species in mesic communities but 

 few or none in xeric ones, whereas the genera Flexamia and Athy- 

 sanella have many species in mesic communities and also a large 

 number in xeric communities. 



Another example of reduced number of species in colonizing 

 communities is found along the western edge of the North Ameri- 

 can temperate deciduous forest. Eastward roughly from the west- 

 ern range limit of Quercus alba this forest contains many species 

 of trees, but westward the number of tree species drops sharply. 

 In western Minnesota and the Dakotas, the climax is predomi- 

 nantly Quercus macrocarpa, often occurring in almost pure stands. 

 Furthermore, on a transect west from Missouri the wooded areas 

 contain chiefly Gleditsia, Juglans, and Celtis. The trees on the 

 edge of these western areas are those able to exist under more 

 xeric conditions than their eastern neighbors and undoubtedly rep- 

 resent the encroachment of the deciduous forest upon the tall-grass 

 prairie biome. The small number of species in these trul)' colonizing 

 communities contrasts strikingly with the larger numbers found 

 in more eastern and more mesic communities. 



(3) Communities of a later successional stage may not become 

 adapted to as great ecological extremes as communities of an ear- 

 lier successional stage. Axelrod points out that in California the 

 more shrubby chaparral occupies more xeric sites than the live-oak 

 woodland, although the chaparral community is also a subclimax 

 stage following burning of the live-oak woodland. Axelrod ofters 

 fossil evidence that earlier in the Cenozoic the chaparral may have 



