ECOLOGY OF ISLE ROYALE. 131 



ciples. It will also be noted that these regions do not closely correspond 

 with current fannal areas, although there is a very close correlation 

 in some cases. An avian formation may, in general terms, be consid- 

 ered the analogue of a vegetational formation, although this does not 

 imply that they necessarily have the same boundaries. 



As the literature treating of the vegetation of these areas is extensive 

 and scattered, a few papers will be cited as an index to others: — 



1. 4rid Deserts; Bray, '06; Coville and MacDougal, '03. 



2. Grasslands or Plains; Clements, '05; Pound and Clements, '00. 



3. Southeastern Hardwoods; Cowles, '01; Harper, '06; Transeau, '05. 



4. Eastern Canadian Conifers; WMtford, '01; Transeau, '03, '05-'06; Ganong, 



'03, '06; Harvey, '03. 



5. Rocky Mountain and Pacific Conifers; Whitford, '05; Gray and Hoolier, '81; 



Piper, '06; Young, '07. 



6. Alpine; Merriam, '90, '99; Coville, '93; Fernald, '07. 



These environmental unit areas as found to-day, ai-e the result of 

 many successions which, in some cases at least, reach rather far back 

 into the past. This is because some occupy ancient land areas, such 

 as much of the Southeastern Hardwood ai'ea. On the other hand, some 

 occupy relatively new regions, that is, at least with regard to the dom- 

 inant factors now in control, as in the glaciated part of North America 

 and on the Coastal Plain. So far as the present is concerned such rela- 

 tions clearly show that these areas are only the end results of extensive 

 past changes oi' successions which represent the terminal branches and 

 cross sections of development. It is to the study of such regions and 

 associations that we must turn for the fundamental organization or 

 associational relations of the various elements which compose not only 

 the environments but also the associations of animals. 



In order to make as definite as possible the structural and ecological 

 characteristics of these formations, certain general relations are here 

 formulated. Throughout this paper it should be remembered that the 

 individual birds and associations of given areas form the units of com- 

 parison. Such a distinction is necessary because many species show con- 

 siderable geographic variation' in habits and in the habitats frequented. 

 The writer clearly recognizes the risks and difficulties of such an at- 

 tempt. They are deliberately put in their present form to invite criti- 

 cism and qtMUfication from field workers. It is desirable to know the 

 validity of these formations, their internal ecological relations and dy- 

 namic tendencies, their relation to dominant environmental influences, 

 etc. A complete list is not attempted, and some of the statements may 

 be only fragments of larger generalizations; but it is just such relations 

 as these which will develop if the entire subject is considered critically 

 and synthetically. Some of the leading characteristics of these larger 

 environmental units and their avian formations may be briefly outlined' 

 as follows: — 



1. The dominance of a limited number of physical conditions or 

 •complexes, as climate, topography, vegetation, animals, etc., in a given 

 area produces the primary environmental units and formations. 



2. Secondary environmental dominance is shown by a secondary avian 

 assoeiation. Thus in the Northeastern biotic center there is a second- 

 ai'y dominance due to water basins in the forest area. 



