EVOLUTION OF ECOLOGICAL CONCEPTS 351 



The changes have usually been regarded piecemeal, as particular changes 

 affecting particular concepts; but this conclusion will attempt a more general 

 interpretation of their meaning. 



One over-all feature of the change is the dissolution of a coherent, well- 

 ordered, deductive system for the interpretation of vegetation, an attempt at 

 deductive whole-knowledge of vegetation, and its replacement by less coherent 

 and less interdependent, inductive part-knowledges of different vegetational 

 problems. Implied in this is a change in the view of natural communities 

 themselves. These seem no longer to form an area of clear-cut, well-ordered, 

 simply defined, and neatly interdependent phenomena to which a deductive 

 system like Clements' may be appropriate. Although the deductive system is 

 an ideal of science, students of natural communities lind themselves authors of 

 an inductive science, with problems more akin to those of the social sciences 

 than those of physics and geometry. 



The lack of clear-cut orderliness in natural communities is a necessary 

 consequence of the multiplicity of factors and complexity of interrelations 

 with which the ecologist must deal. This complexity is as much a fundamental 

 circumstance of the study of natural communities as of the study of man's 

 communities. A basis of lack of simple orderliness may be found also in 

 Ramensky's principle of species individuality. Species populations are dis- 

 tributed "individualistically," and they are not organized in terms of man's 

 ecological concepts and classifications. Major problems of community classi- 

 fication, of succession and climax, and community evolution stem directly 

 from this individuality of species distributions. Many ecologists view the 

 implications of species individuality in a more conservative light than that of 

 this paper; but to the author species individuality is one major theme of cur- 

 rent changes in ecological concepts. A further aspect of the lack of simple 

 orderliness is in substantial effects of chance, of largely unpredictable factors 

 of dispersal and population interaction, on species distributions and commu- 

 nities (Palmgren, 1929; Egler, 1942; Whittaker, 1953). Species individuality, 

 multiplicity of factors, and effects of chance all contribute to giving most 

 statements about natural communities a quality of probability, not necessity, 

 partial correlation, not strict interdependence, inductive generalization for 

 which limitations and exceptions are granted, not exact prediction. And if 

 one turns to Braun's description of the eastern forests with these things in 

 mind, a view of these forests quite different from Clements' results. One is 

 struck first by their immense, their almost overwhelming, complexity of pat- 

 tern. If one seeks to view this complexity in perspective, in terms of species 

 populations in space and successional and evolutionary change and without 

 the intervention of man's ecological abstractions, then the view of the forest 

 is not one of clear and orderly associations, successions, and phylogeny. It is 

 one of a veritable shimmer of populations in space and time. 



If this population shimmer is not simply and clearly orderly, it is also not 



