646 



ECOLOGY AND EVOLUTION 



Adjustment to a more general habitat, 

 such as a dimatic zone, may be at least 

 partially the result of directed dispersal. 

 Migratory birds may follow the seasonal 

 changes in isothermal lines in temperate or 

 mountainous regions. Nonmigratory animals 

 may slowly disperse toward favorable re- 

 gions (Goldschmidt, 1933). 



One should guard against an exaggera- 

 tion of the role of preadaptation, valid as 

 the concept may be. Even if a large part 

 of the mortality of a population is based 



stated that selective intensities efiFective in 

 evolutionary change are of the order of 1 

 per cent to one-tenth of 1 per cent in each 

 generation. 



Haldane (1932, p. 94; 1936) made a 

 mathematical analysis of the possibihty of 

 chance mutations and combinations becom- 

 ing established with and without selection. 

 If fifteen uncorrelated gene characters are 

 each present in 1 per cent of the individ- 

 uals, all fifteen would be present in only 

 one in 10^" individuals. There have not been 



Fig. 235. Nest of the termite, Constrictotermes cavifrons, in the British Guiana rain forest. 

 On the left, side view of the entire nest constructed on the smooth bark of a slanting tree. 

 Food is stored in the hanging portion of the nest. On the right, detail of the chevron-shaped 

 rain-shedding ridges extending over the surface and on the side of the nest (also above the 

 nest). Newly constructed portion is seen at the upper right. 



upon pure chance and not upon the heredi- 

 tary characters or the individual choice of 

 the organism, selective factors may still 

 guide the species in the direction of heredi- 

 tary adaptation. If one supposes that 999 

 out of every 1000 seeds happen to fall in 

 habitats where development of a plant spe- 

 cies is impossible, sorting of the remainder 

 might still be made in accordance with 

 their genetic variability and differential sur- 

 vival, an influence that would be a major 

 factor in the evolution of the plant in the 

 direction of greater fitness, both internal 

 ^nd environmental (p. 603). Fisher (1936) 



10^" individual higher plants (multicellular 

 and sexually reproducing) during geologi- 

 cal history of 10" years. Thus there is little 

 chance of all fifteen characters occurring 

 together unless sorted and selected. With 

 biparental reproduction, a favorable com- 

 bination of fifteen genes would immediately 

 be dispersed. If selection favors each char- 

 acter so that 99 per cent of the individuals 

 have these genes, 86 per cent of the popu- 

 lation would have all fifteen. It would take 

 10,000 years for such a pattern to become 

 established through natural selection in a 



