THE ORIGIN OF SPECIES 



Restriction of Random Dispersal. Distribution maps generally indi- 

 cate that a particular species is found continuously over broad areas, but, 

 in fact, all species select in such large areas those restricted portions which 

 present suitable ecological features. A checker board is thus a better model 

 of species distribution than is the typical distribution map. For example, 

 the distribution map for the American sycamore or plane tree, Plataniis 

 occidentalis, indicates a continuous distribution over more than half of the 

 United States, from Texas in the South and Iowa in the North to the At- 

 lantic coast. But if one wishes to find natural groves of sycamores in this 

 vast area, it is necessary to look in rich bottom lands and along the banks 

 of streams. Again, distribution maps show that the eastern meadowlark 

 is found over most of the United States east of the Rocky Mountains, but 

 it will be found only in open grasslands. The various local populations of 

 most species are similarly separated by barriers of greater or lesser extent 

 of territory which they cannot utilize for one reason or another. Within 

 the broad areas of mapped distribution, there is almost no geographical 

 feature which may not prove to be a barrier to the dispersal of some 

 species. 



Even a small amount of salt water is a nearly absolute barrier to am- 

 phibians. For this reason, the Pacific Islands are usually uninhabited by 

 amphibians, except when they have been introduced by man, as in Ha- 

 waii. Salt water also separates many fresh-water fishes. For example, on 

 the Pacific coast many fresh-water streams follow more or less parallel 

 courses to the ocean. Typically, each stream will have its own subspecies 

 or even species. Although the expanse of salt water separating the mouths 

 of neighboring streams may be small, the fish do not cross it. But if the 

 flood waters of the streams join together during the rainy season, then the 

 streams share their fishes. This shows that it is actually the salt-water 

 barrier, and not a homing instinct or other factor which typically keeps 

 the fish faunas of neighboring streams separate. 



Large bodies of water are among the most effective barriers to land 

 birds. The Amazon River seems to be an absolute barrier to many birds, 

 for in species after species the subspecies on opposite banks are different. 

 Mayr has pointed out many instances in which neighboring tropical is- 

 lands are inhabited by different subspecies even though the distances 

 between them are rather short. When Darwin studied the birds of the 

 Galapagos Islands, he found 26 species of land birds and 11 species of 

 marine birds. But of these, 21 of the species of land birds were endemic, 

 while only 2 of the species of marine birds were endemic. Islands which 

 are separated only by short distances may have distinct subspecies or 

 species, but, generally, those islands which are most isolated have the 

 highest proportion of endemics. 



Many mammals are also stopped by water barriers. Thus, subspecies of 

 mice on opposite sides of major rivers are very likely to be different. The 

 zoogeographical realms have been defined primarily on the basis of their 

 mammalian fauna, and it is noteworthy that water barriers, the oceans, 

 separate many, though not all, of these realms. 



Mountains have sometimes been described as islands in a sea of low- 



272 



