146 Species and Species Change 



According to one of Wright's models ( 1949 ) , the best chance of 

 adaptive change would occur if the species were divided into local 

 populations having only limited breeding between them. The dis- 

 tribution of H. borealis through much of Mexico and southwestern 

 United States fits this model well, occurring primarily in springs 

 which may be 50 to 100 miles apart. It cannot, however, be the 

 only critical factor because the ranges of the other species of HeJi- 

 copsyche in the area have exactly the same type of distribution. 

 Perhaps this discontinuity factor plus large population size caused 

 the unique evolutionary development of H. borealis. If so, then 

 the root of this development must have started with whatever cir- 

 cumstances produced the initial relatively large populations of H. 

 borealis. There is no clue to these circumstances. 



In the Helicopsyche borealis example there is no suggestion as 

 to the specific genetic change involved. In the American plant 

 genus Clarkia, an unusual extension of climatic tolerance in C. 

 lingtilata is associated with additional chromosomal units in the 

 genome (Lewis, 1953c). Lewis suggested changes in quantitative 

 gene dosages as the probable basic cause of this change in ecological 

 tolerance. 



REDUCTION IN GENETIC COMPLEXITY 



The total genetic complexity of the species would seem to be re- 

 lated to the ecological divergence encompassed in the total range. 

 Starting from this premise, any physical, chemical, or biotic change 

 which reduced the range in an ecological sense would reduce the 

 complexity of the gene pool. As a hypothetical example involving 

 a physical change, a terrestrial species might occupy the annual 

 rainfall gradient from 20 to 40 inches, with its range bounded by 

 a mountain range on the 20 inch side and by an ocean on the 40 

 inch side. If the mountains become more elevated and their rain 

 shadow more intense, the rainfall gradient would move out over 

 the ocean so that the species would have available only the rainfall 

 gradient 20 to 30 inches. This would eliminate the genetic factors 

 having a high selection factor on the former 30 to 40 inch part of 

 the rainfall gradient and presumably would reduce the complexity 

 of the gene pool and the interactions contributing to homeostatic 

 balance. 



Under these conditions of decreasing homeostatic complexity, 

 smaller and smaller genetic changes should theoretically produce 

 phenotypic changes sufficiently great to have a selective value. 

 As a result, in the example above genetic changes could become 



