ECOLOGY AND ISOLATION 



613 



tions between the sexual generations (p. 

 703). Mutations that aflFect cross breed- 

 ing between populations might be built up 

 during the parthenogenetic increase (Mayr, 

 1942, p. 192). The two aphids, Colopha 

 ulmicota and C. ^raminis, afiFord a possible 

 example of such speciation. They make 

 similar galls on their primary hosts, the elms 

 Ulmiis americana, U. fidva, and U. race- 

 mosa. The secondary hosts of C. iilmicola 

 may be grasses of the genus Ero^rostis, 

 while the secondary hosts of C. graminis 

 may be grasses of the genus heersia 

 (Patch, 1910). The ranges of the two spe- 

 cies of aphids overlap without any indica- 

 tion of geographic separation, and the 

 sexual generations seem to appear at the 

 same time. 



The primary host of the aphid, Theca- 

 hiiis affinis, is the poplar, Poptilus nigra, and 

 the secondary hosts are species of butter- 

 cups. Ranunculus. During the glacial period 

 in Europe, these plants were separated, and 

 T. affinis eyolyed a parthenogenetic form 

 confined to its secondary hosts. This geneti- 

 cally sexless form maintains itself, even 

 when the original primary host is available 

 ^Mordvilko, 19S5)\ 



Specialized plant-feeding insects should 

 show some speciation and further evolution 

 with initial isolation of populations on the 

 basis of host species rather than topo- 

 graphic separation (Brown, 1945). but the 

 data are not completely clear on this point 

 (Thorpe, 1940). Host specificity involves 

 biochemical adjustment. In some instances 

 the biochemical characters are found only 

 in a single species of host, whereas in 

 others, groups of species classified in genera 

 or other higher categories may share bio- 

 chemical characteristics and also be hosts to 

 the same species of parasite or phytophage. 

 It is probable that internal parasites in more 

 direct contact with the tissues and fluids of 

 the host are still more subject to biochemi- 

 cal factors than are external plant eaters 

 or parasites. 



Zimmerman (1938) cites the large num- 

 ber of weevils of the genus Microcrt/pto- 

 rhynchus with forty-three endemic species 

 on the small South Pacific island of Rapa. 

 These weevils are all flightless and exhibit 

 varying degrees of host-plant restriction. 

 Seventeen species were found associated 

 with a single host, but others were some- 

 times found on more than one host, al- 



though preponderantly on a single host. 

 One species occurred on nine host plants. 

 Topographic isolation is doubtless impor- 

 tant in this speciation. Probably isolation 

 on difi^erent islands separated several orig- 

 inal stocks that were later carried to 

 Rapa by strong winds. Further speciation 

 on the island then probably occurred with 

 host plants separating the evolving species. 

 Soon after emergence the female is fer- 

 tilized by a male, usually from the same 

 host plant, and new host plants are invaded 

 by crawling weevils. 



Sears (1947) says that, within the me- 

 lanica group of Drosophila, D. melanica 

 paramelanica is found in the northeastern 

 United States, D. melanica occupies the 

 south and southwest, while D. nigrome- 

 lanica overlaps both melanica and melanica 

 paramelanica in geographic range, but is 

 isolated from them by using fungi for food. 

 Also, within the quinaria group of species, 

 D. quinaria, D. palustris, D. suhpaliistris, 

 and D. transversa are found in the same 

 geographic region (northeastern United 

 States). Drosophila quinaria lives in moist 

 forests and eats fruit; D. palustris and siib- 

 palustris inhabit swampy areas and eat de- 

 cayed plants; and D. transversa lives in dry 

 deciduous forests and eats fungi. 



Kinsey (1936) states that, among gall 

 wasps (Cynipidae), "closely related spe- 

 cies are to be expected in adjacent geo- 

 graphic areas on the same or on closely re- 

 lated hosts, or in single geographic areas on 

 distinct but related hosts." The first por- 

 tion of this quotation is a restatement of 

 Jordan's rule, and expresses correlation 

 with topographic isolation. The second por- 

 tion would seem to indicate habitat isola- 

 tion. As an example of "host isolation." Kin- 

 sey cites two species of gall wasps, Ctjnips 

 echinoides and C. hibrida. He believes that 

 the morphologic, ecologic, and geographic 

 data indicate that an ancestral stock was 

 present in the Rocky Mountain region 

 where there was only a single group of de- 

 ciduous white oaks. This stock emigrated 

 into the eastern forests characterized by a 

 variety of distinct species of white oaks; 

 one species of gall wasp invaded Quercus 

 hicolor and Q. vrinus, while the other in- 

 vaded only Q. alha. Each overlaps the geo- 

 graphic range of the other species. No one 

 can be absolutely certain that C. echinoides 

 and C. hibrida originated in the same geo- 



