344 WHITTAKER 



but in a world view there is extensive continuity and intergradation among 

 formations, as is especially evident in the tropics (Beard, 1955). This con- 

 tinuity is not easily reconciled with the view of the formation as a distinct 

 entity comparable to an organism. All these observations contribute to a 

 changed view of the formation as a vegetation unit. The formation seems not 

 a distinct, concrete vegetation unit determined solely by climate, but an 

 abstract grouping of communities of similar physiognomy and environmental 

 relations, a grouping dependent in the end on man's choice of what constitutes 

 sufficient similarity of physiognomy and environment. 



The nature of the association has been more bitterly contested and is a 

 problem many ecologists do not feel has been resolved. The essence of the 

 concept lies in the association, i.e., occurrence together, of species populations 

 to form community units which may be recognized by species composition. 

 Since associations are thought to be "natural units," very often compared 

 with species as units of classification, it has been natural to assume that 

 they are distinct in the sense of being separated from one another by well- 

 defined boundaries. The basis of the association concept was challenged by 

 Ramensky (1924) with two propositions: (1) The principle of species indi- 

 viduality — each species responds uniquely to external factors and enters the 

 community as an independent member; there are no two species which relate 

 themselves to environments and communities in quite the same way. (2) The 

 principle of community continuity — composition of the plant cover changes 

 continuously in space; sharp boundaries between communities are special 

 circumstances requiring special explanation. Ideas closely related to Ramen- 

 sky 's were expressed about the same time by Gleason (1926) and Lenoble 

 (1926, 1928); they are most familiar to Americans as the former's "indi- 

 vidualistic concept of the plant association" (Gleason, 1926, 1939). 



If the association is a "natural unit" of species populations, its unity is 

 presumably expressed in species distributions. Most, or at least some, of the 

 associated species should have closely similar distributions. Recent studies 

 in the eastern forests (Whittaker, 1951, 1952, 1956; Curtis and Mcintosh, 

 1951; Brown and Curtis, 1952) have shown that, in fact, no two species are 

 distributed alike, that species show the heterogeneity of distribution and lack 

 of organization into distinct groups of associates which Ramensky's first 

 principle asserts. Relations to other species must, to be sure, affect their 

 distributions; species distributions cannot be "independent" in the sense that 

 they are unaffected by competition and other interrelations. These interrela- 

 tions, however, do not result in the organization of species into definite groups 

 of associates; and species distributions are "individualistic" in the sense that 

 each species is distributed according to its own way of relating to the range 

 of total environmental circumstances, including effects of and interrelations 

 with other species, which it encounters. Both continuity and discontinuity of 



