NATURAL SELECTION 



685 



perish, but he also says, contrary to our 

 opinion, that the instinct to emigrate can- 

 not have been produced by natural selec- 

 tion (see pp. 642, 645, 671). 



Another aspect of the evolution of popu- 

 lations deserves attention. Pearl (1930a) 

 stated that somatic differences that do not 

 rest upon a genetic basis would have no 

 evolutionary significance. Suppose that an 

 emigrating lemming did not differ geneti- 

 cally from a nonemigrating one (this has 

 not been proved or disproved so far as we 

 know). If the sacrifice of the emigrating 

 individuals had survival value to the popu- 

 lation as a whole, emigrating behavior 

 might well evolve under natural selection of 

 the whole system. The genetic pattern 

 might produce emigrating behavior only at 

 certain environmental thresholds that would 

 behavioristically differentiate the individ- 

 uals that emigrate from those that remain 

 on the breeding grounds. Populations with- 

 out this genetic characteristic would per- 

 ish because the emigration would not 

 diminish the population in conformity to the 

 food supply (pp. 286, 706). The popula- 

 tion in which some individuals show emi- 

 grating behavior under adverse conditions 

 would survive and perpetuate the genetic 

 pattern because of the sacrifice of the emi- 

 grants. The tendency in some species of 

 locusts or grasshoppers to develop solitary 

 and emigrating phases (p. 543), one of 

 which regularly invades new territory 

 where it ultimately perishes, may be the 

 result of evolution involving the sacrifice 

 of large numbers of the population. 



The number of pollen spores produced 

 by pines, which depend on the random 

 distribution of pollen by wind, is much 

 greater than the number of spores per 

 given unit produced by an insect-pollinated 

 plant, such as the yucca or the tulip tree. 

 Fishes that spread their eggs at random 

 and take no care of the young lay many 

 more eggs per fish than do fishes that make 

 nests and protect their young. Birds that 

 are subjected to a greater mortality rate 

 tend to have larger clutch-sizes (Moreau, 

 1944; see also p. 701). These balanced 

 interrelationships of a whole population 

 to its environment are best understood as 

 the result of evolutionary adaptation 

 through natural selection of population 

 units (p. 684). 



Wright (1932, 1948a) postulated that 



the breeding population size together with 

 a given breeding structure may have an op- 

 timum for evolutionary advance— large 

 enough to promote variability and to allow 

 an effective selection pressure, and small 

 enough to allow a certain random fluctua- 

 tion of gene frequencies. With a genetic 

 control of population numbers, size of 

 population might become characteristic of 

 surviving species if other factors remain 

 fairly constant. Baker (1947) states that 

 "those factors which reduce the reproduc- 

 tive capacity of a strain are per se selected 

 against," a conclusion with which we are 

 not in agreement. 



There may be an evolutionary trend in 

 the direction of a smaller reproductive po- 

 tential associated with increased shelter and 

 protection (p. 274), as in tree-nesting birds. 

 Or the trend may be in the opposite direc- 

 tion. A larger reproductive potential is often 

 associated with greater vicissitudes in the 

 life cycle, as in the parasitic roundworms 

 (Fig. 251) as compared with free-Uving 

 roundworms (Baylis, 1938), and in tape- 

 worms (Fig. 250) as contrasted with free- 

 living flatworms. Mutualism within the 

 population may be more beneficial in large 

 colony populations, as may be seen in the 

 more specialized termite and ant societies in 

 contrast with their primitive ancestral socie- 

 ties or with their solitary ancestors (p. 272). 



There seems to be a general tendency 

 for the population numbers (or biomass) of 

 social insect colonies to increase in the evo- 

 lution of vegetarian and scavenger types, 

 and to decrease in those that have evolved 

 social exploitation (i.e., thief ants, slave- 

 making ants, or socially parasitic ants, bees, 

 and wasps), while the population sizes of 

 the colonies of predatory species are 

 roughly between those of these other feed- 

 ing types. 



The evolution of increase or decrease in 

 size (number of cells) of an organism, and 

 size of a population (number of individ- 

 uals) may result from somewhat similar 

 evolutionary forces. Both trends in either 

 individual organisms or populations may 

 result in adaptation to and even control 

 over environmental fluctuations in particu- 

 lar instances. 



Those groups that exhibit cyclomorphosis 

 indicate clearly the result of selection on 

 species populations. Cyclomorphosis (p. 

 118) is a term usually used for cyclic 



