ECOLOGY AND ISOLATION 



607 



practically all diverging populations, sev- 

 eral factors contribute at the same time or 

 in close succession to the prevention of 

 gene flow across the population borders. 



In order to detect the eflFect of one iso- 

 lating factor, it is best to find closely re- 

 lated species separated by this factor and 

 no other. It is seldom possible, however, to 

 find such ideal examples. More often one 

 can only eliminate some of the possible 

 isolating mechanisms in a given case, leav- 

 ing several factors that cannot be treated 

 independently. The isolating mechanisms 

 that cannot be separated in one instance 

 may often be separated in other paired spe- 

 cies, however, so that the factors appear 

 with fair distinction after studying a large 

 number of cases. Isolating mechanisms 

 that partially separate subspecies deserve 

 particular study, for the initial mechanisms 

 of speciation are seen more clearly in such 

 populations. The method of analyzing spe- 

 ciation is of necessity mainly inductive 

 from wide evidential data rather than from 

 clean-cut experimentation. 



Varying degrees of partial isolation may 

 be found that produce a quantitative effect 

 upon the rate of gene flow between popu- 

 lations. Highly important evolutionary phe- 

 nomena may be correlated with such dif- 

 ferences in genetic interchange, even 

 though complete reproductive isolation is 

 not achieved (pp. 602 and 603). 



In spite of overlaps and intergradations 

 of the isolating factors, one perceives a 

 tendency toward a temporal sequence in 

 the operation of several isolating mecha- 

 nisms. First, there is a segregation of genet- 

 ically similar populations. Second, there 

 is a genetic reorganization within each 

 population. Third, permanent isolation is 

 achieved through intrinsic mechanisms that 

 prevent all gene flow between populations, 

 whether the extrinsic barriers continue to 

 exist or not. There are many exceptions to 

 this chronology, such as chromosome muta- 

 tion in polyploid speciation, but in general 

 this sequence of events is a common oc- 

 currence. Under certain circumstances, 

 some separable factors may have closely 

 similar evolutionary consequences. For ex- 

 ample, a gross topographic barrier without 

 differences in habitat may result in specia- 

 tion in about the same time as a micro- 

 spatial habitat separation with strong se- 



lection pressures in the different environ- 

 ments. 



SPATIAL ISOLATION 



If no sharp external barriers of a topo- 

 graphic or ecologic nature separate por- 

 tions of a large, widely distributed popula- 

 tion, distance alone will nevertheless pre- 

 vent separated individuals from interbreed- 

 ing. With some mutation pressure or varia- 

 tion in gene frequencies, genes filter slowly 

 from one local population to another and 

 may commonly result in geographic varia- 

 tions with no sharp lines of demarcation 

 (Geoclines, p. 626). The density of the in- 

 terbreeding populations is an important 

 factor. A sparse population inhabiting a 

 large region might produce a local in- 

 breeding effect similar to that in a denser 

 local population partially isolated by var- 

 ious extrinsic or intrinsic mechanisms. 



Thompson (1931a) showed that there is 

 a correlation between water distance and 

 morphological divergence wathin a species 

 of fish in the rivers of Illinois. Sumner 

 (1932) stated that he had never compared 

 two local collections of the same species 

 of deer mouse (Peromyscus) from points 

 at all remote from one another without 

 finding significant differences between 

 them. Dobzhanslcy and Epling (1944) give 

 a thorough analysis of spatially isolated 

 races of Drosophila and define races as 

 "populations of a species which differ in 

 incidence of one or more variable genes or 

 chromosome structures." Physiological dif- 

 ferences as well as morphological distinc- 

 tions between spatially separated popula- 

 tions are to be expected (Baily, 1939; Du- 

 Shane and Hutchinson, 1944). 



Absolute distance in space is not neces- 

 sarily correlated with population diver- 

 gence. Timofeeff-Ressovsky (1940) and 

 Mayr (1942, p. 242) show that the range 

 of the individual of the species is of great 

 importance in the breeding structure of the 

 population (see Vagihty, p. 213). Widely 

 separated populations of the teal. Anas 

 crecca, actually interbreed. Young birds 

 taken from the nest and banded in England 

 were found nesting the next vear from Ice- 

 land and the northern Urals to central 

 France and central Russia. Pair formation 

 in many migrating ducks occurs in the 

 wintering range so that a male hatched in 

 Maine may meet a Nebraska female ii) 



