616 



ECOLOGY AND EVOLUTION 



species, and concludes that natural condi- 

 tioning of this type might spUt a popula- 

 tion into separate groups attached to par- 

 ticular aninial hosts or food plants, with 

 consequent prevention of interbreeding. 



Mayr (1942, pp. 199, 215; 1947) takes 

 the position that habitat isolation within 

 the same region is not the initial cause of 

 the origin of sympatric species (also see p. 

 659; Lack, 1944, 1946). He beUeves that 

 all or nearly all sympatric species origi- 

 nated as geographic rather than ecologic 

 isolates and that their ranges later came to 

 overlap. He postulates that any ecologic 

 distinction between species may assist in 

 keeping them isolated, but did not origi- 

 nally separate them and that any ecologic 

 differences within the same area without 

 geographic separation would be swamped 

 through interbreeding. Thorpe (1945) con- 

 tends, however, that the conclusion that 

 geographic isolation always precedes other 

 kinds of isolation is premature (see also 

 Crombie, 1947). 



Wright (1945; see also p. 603) has 

 pointed out that the intiagroup competition 

 between almost but not completely isolated 

 populations within large species gives 

 the best conditions for adaptive evolution 

 of species as wholes (Fig. 229). Surely 

 these partially isolated groups are com- 

 monly separated by ecologic as well as 

 geographic factors. Partial habitat isola- 

 tion probably often results in diflFerences 

 in population numbers (p. 602) with their 

 effect upon gene frequency and gene fixa- ■ 

 tion. Competition between populations in 

 the same habitat might favor adaptive 

 divergence (Crombie, 1947). If selection 

 pressures guide these partially isolated 

 groups toward ecologic adaptive diver- 

 gence, the hybrids between the two 

 successful populations would occasion- 

 ally be selectively eliminated (p. 623). 

 Selection would maintain intragroup fertil- 

 ity, but would not always maintain inter- 

 group fertility, and might even favor genes 

 producing genetic isolation if advantageous 

 combinations were swamped through inter- 

 breeding between groups adapted to differ- 

 ent environments. Thus habitat isolation 

 could initiate a process of speciation (pp. 

 616 and 676). 



Mayr (1942, 1947) emphasizes the geo- 

 graphic component in habitat isolation, 

 while we are here emphasizing the ecologic 



component. The Une of division between 

 sympatric and allopatric distributions of 

 closely related species is arbitrary in a 

 great many instances. Habitat differences 

 are also often microgeographic. However, 

 it is well to separate these two factors m 

 discussing isolating mechanisms, even 

 though they overlap and usually act to- 

 gether, because selection in different habi- 

 tats and separation in physical space may 

 independently influence divergent specia- 

 tion. 



In summary, we cannot say that closely 

 related species always replace each other 

 ecologically, but we can say that in some 

 instances there is a tendency for them to 

 do so. Closely related sympatric species 

 are usually divergent in some important 

 adaptive characters. Habitat separation is 

 surely an important isolating mechanism 

 (not so important in general evolution as 

 is topographic isolation). It plays a signif- 

 icant role in speciation of certain types of 

 organisms, particularly those with narrow 

 biotic relations such as are found among 

 phytophages and parasites. Initial habitat 

 isolation is difficult to illustrate, but is 

 theoretically possible. 



CYCLIC ISOLATION 



Adjustments to variation in seasonal fac- 

 tors such as temperature or light may bring 

 about reproductive isolation between 

 closely related populations. Single muta- 

 tional changes are not likely, however, to 

 induce the separation of two distinct popu- 

 lations with nonoverlapping temperature 

 ranges (Mailer, 1942). Polygenic charac- 

 ters are practically always involved in such 

 adaptational changes, and selection is likely 

 to be a factor in balancing the genetic com- 

 plex. 



Banta and Wood (1928) reported a 

 mutation in Cladocera that raised both the 

 upper and lower hmits of temperature 

 toleration by 10° C. Such a dramatic 

 change, which might conceivably separate 

 natural populations in the same area, would 

 probably have to be bal'viced with other 

 external and internal adjustirronts, so that 

 the emergence of seasonal differences 

 would probably necessitate gradual and 

 long evolution. Seasonal divergence may 

 have become established with geographic 

 isolation in many cases, but it is also pos- 

 sible that selection of populations at two 



