374 



POPULATIONS 



and termite eaters are the exception rather 

 than the rule. N. Tinbergen (1933) reported 

 on the food consumption of a population 

 of long-eared owls {Asio otiis) during the 

 successive winters of 1930-1931 and 1931- 

 1932, During the first winter, when the 

 vole (Microtus) was abundant, this species 

 constituted 86 per cent of the owl's total 

 food. The wood mouse (Apodemus) fur- 

 nished 7 per cent; other mammals, 2 per 

 cent; house sparrows, 2 per cent; and 

 other birds, 3 per cent. During the second 

 winter the voles were scarce, and the owls 

 turned to other prey. Voles constituted 30 

 per cent rather than 86 per cent; wood 

 mice, 15 per cent; other mammals, 7 per 

 cent; house sparrows, 30 per cent; and 

 other birds, 18 per cent. Shifts in preda- 

 tion pressure of this sort have obvious im- 

 plications for zoogeography and commun- 

 ity studies as well as for the strictly popu- 

 lational aspects. 



These comments focus on a predator 

 population. This has a corollary in terms 

 of the prey population. The general eco- 

 logical principle can be stated in this way: 

 A predator may exploit several prey spe- 

 cies, while, conversely, a prey species may 

 be exploited by several predators. The lat- 

 ter point finds illustration in Stoddard 

 (1932). He showed that the percentages 

 of bobwhite nests destroyed— that is, 

 through eating of the eggs— by natural 

 enemies over four years' observation were 

 as follows: 1924, 46 per cent; 1925, 41 per 

 cent; 1926, 38 per cent; and 1927, 34 per 

 cent. Skunks, cur dogs, house cats, cot- 

 ton rats, opossums, blue jays, crows, tur- 

 keys, snakes, the thief ant Solenopsis mole- 

 sta (which enters the egg to feed as soon 

 as it is punctured by the emerging chick), 

 man, and "unknown agencies," among 

 them foxes, weasels, and other animals, 

 were incriminated. 



Another significant point emerging from 

 Stoddard's tabulation, a principle of gen- 

 eral ecological moment, is that predation 

 is frequently directed against the imma- 

 ture stages of the prey and as such may 

 constitute an eflFective limiting factor. 



There is considerable evidence in the 

 literature suggesting that the abundance of 

 a predator is associated with the abundance 

 of its prey. This is particularly true when 

 the prey constitute the major food item in 

 the predators' diet. To the field naturalist 

 this is, of course, an old story. He fre- 



quently notes a large increase in predators 

 during years when the prey populations 

 (especially rodents) are large. This in- 

 crease may result from increased predator 

 iimnigration into the region, from height- 

 ened and more effective reproduction, or 

 from both. The phenomenon is well illus- 

 trated by Elton (1942). 



A more analytical example of this, 

 developed by MacLuhch (1937), concerns 

 the relation between the varying hare and 

 the lynx in North America. In 1905 Mac- 

 Farlane said: "The yearly catch of lynxes 

 rapidly diminishes in volume as soon as the 

 rabbits become scarce and when the latter 

 are comparatively rare a large proportion 

 of the great but now dwindling crowd of 

 lynxes suffer privation, and some actually 

 starve to death." Seton (1925) pointed out 

 that in the winter of 1906-1907 in the 

 MacKen2de River valley, when, presum- 

 ably, the rabbits were sparse, an examina- 

 tion of the stomach of twelve lynxes 

 showed no food present at all. These ani- 

 mals were dying of starvation— "mere walk- 

 ing skeletons." 



MacLulich examines this predator-prey 

 cycle in Canada (Fig. 118). He finds that 

 the varying hare has a cycle with a mean 

 duration from peak to peak (or dip to dip) 

 of 9.6 years, the lynx of 9.7 years, and that 

 the two are essentially confluent. A correla- 

 tion coefficient was calculated between rab- 

 bit and lynx records arbitrarily designated 

 "scarce," "intermediate," and "abundant" 

 for records extending from 1847 to 1934, 

 inclusive. This coefficient, with the value 

 -f 0.55 dz 0.05, indicates appreciable cor- 

 relation between population sizes of the 

 two species. "Therefore there is good 

 ground for befieving the decreases in num- 

 bers of lynxes are caused by starvation 

 when the hares disappear, or at least by 

 inability to withstand adverse circumstan- 

 ces and winter conditions on short rations" 

 (p. 102). 



One of the most noteworthy studies of 

 predation that has yet appeared is that of 

 L. Tinbergen ( 1946; see also the review in 

 English by Hartley, 1947). This is primari- 

 ly concerned with the effect that the Euro- 

 pean sparrow hawk (Accipiter nisus) ex- 

 erts against populations of the house spar- 

 row (Passer domesticus) , the chaffinch 

 {Fringilla coelehs), the great tit (Parus 

 major), and the coal tit (Parus ater). The 

 four prey species are censused or estimated 



