Female Drosophila fruit Hies crowtied into small 

 liottles do not lay as many eggs as tliey do when not 

 crowded. This has heen attributed to the competition 

 of females for space, and to frequent disturbing con- 

 tacts with other flies, so that they do not feed ade- 

 (|uately. It could also be attributed to their energy 

 being dissi]>ated and to their ovipositing being too 

 often interrupted (Pearl 1932, Bodenheimer 1938. 

 Ciiiang and Hodson 1950). On the other hand, the 

 reduction in fecundity may not be so much a result of 

 disturbance as one of reduction both in the quantity 

 and t|uality of food that is available per individual 

 (Robertson and Sang 1945). 



ICxperimental studies of populations of the flour 

 beetle Triholium confusioit show that, as they in- 

 crease in size and modify the flour in which they live, 

 there is a decrease in the number of eggs deposited, 

 an increase in the length of the larval period before 

 pupation, an increase in larval and pupal mortality, 

 and a decrease in the weight of both the pupae and 

 adults. Apparently these effects are produced partly 

 by decreased fecundity of the individual females and 

 partly by cannibalism of larvae and adults upon the 

 eggs and pupae, presumably induced by accumulation 

 of excreta and deterioration of the food supply. When 

 the modified flour is replaced by fresh flour at 48- 

 hour intervals, the rate of reproductivity rises, even 

 when the beetle populations become very large ( Park 

 1934, 1938, Park and Woollcott 1937, Hammond 

 1938-39, Rich 1956). Similar effects of crowding 

 on weight, length of developmental period, and mor- 

 tality have been demonstrated in sheep blow-flies 

 (Uliyett 1950) and in Drosophila (SokolofT 1955). 

 In the sheep blow-fly, artificial destruction of a large 

 per cent of emerging adults brings an increased 

 length of adult life, increased fecundity rate per fe- 

 male, and an increased total number of offspring 

 produced (Nicholson 1954). 



Overcrowding of pink salmon in small impound- 

 ments causes retention of many eggs within the fe- 

 male at spawning, and perhaps also mechanical in- 

 jury to the eggs already deposited from excessive 

 stirring of the gravel (Hanavan and Skud 1954). 



A study of the European bird, the great tit, car- 

 ried on for five years in 16 different areas, revealed 

 a striking inverse relation between density and fecun- 

 dity (Kluijver 1951). The average number of eggs 

 laid per pair of birds per season varied from 13 to 

 20 at population densities of 8 to 12 pairs per 40 

 hectares (100 acres), to only 7 or 8 at population 

 densities of 9 to 19 pairs per 40 hectares. The per- 

 centage of pairs having second clutches during a year 

 varied between 40 and 100 at population densities of 

 less than 16 pairs, but decreased to less than 10 at 

 higher densities. Similar results have been obtained 

 on the North American house wren (Kendeigh 

 and Baldwin 1937). Apparently the lowered fecun- 



TABLE 16-2 ReUHon of reproduction to donjity of laboratory 

 mice during a four-month period (after Retzlaff 1939). 



Number Litters Offspring Offspring 

 of per per per 



Groupings groupings female Utter female 



2<f, 2$ 



B<f, 8$ 

 n(f, 12? 



7.8 

 7.2 

 6.6 

 7.4 

 6.4 



35.2 

 31.1 

 26.2 



dity at iiigh pojiulation densities in these cases is in 

 part the result of frequent disturbance and conflicts 

 resulting from the crowding of territories and in 

 part to less food available per pair on the smaller- 

 sized territories ( I^ck 1952). The non-breeding 

 population of birds is doubtless high only when the 

 breeding population is sufficiently dense that it oc- 

 cupies all of the most favorable territories. 



In the vicinity of Ithaca, New York, during three 

 years of population increase, the average number of 

 embryos per pregnant female in the meadow vole was 

 6 to 6.2, but in the year of decline following the peak 

 only 4.5 to 5.5 (Hamilton 1937a). In California, the 

 litter size of the montane vole declined as the popula- 

 tion increased, and small litter sizes continued, as in 

 New York, during the following decline in the popu- 

 lation (Hoffman 1958). There is evidence that as 

 snowshoe rabbit populations build up in the upswing 

 of a cycle, litters are larger and more frequent than 

 during the ensuing downswing (MacLulich 1937, 

 Rowan and Keith 1956, Green and Evans 1940). 



Experimental studies of populations of the labora- 

 tory mouse lend support to high population as a cur- 

 tailing influence on reproduction (Table 16-2). With 

 an increase in the number of mice crowded into cages 

 of uniform size there was a decrease in the number 

 of litters produced, in the size of each litter, and in 

 the total number of young. At the higher densities 

 there was considerable fighting, resulting in serious 

 wounds and even death for some individuals. A so- 

 cial hierarchy was established, and it appeared that 

 only the despots at the top of the bite order were able 

 to reproduce at a normal rate. Those at the bottom 

 of the order produced few young or none at all 

 (Crew and Mirskaia 1931, Retzlaff 1939, Crowcroft 

 and Rowe 1957). It has been shown that in the 

 coccid insect Lepidosaphes ttlmi a decrease in fecun- 

 dity at high population levels was not a result of a 

 decrease in number of eggs laid by fertile females but 

 of an increase in the percentage of females that were 

 sterile (Smirnov and Polejaeff 1934). 



An experimental population of house mice was 

 established in a large enclosure with cover and water 

 supplied in excess but with food allotments held to 

 a constant daily amount. The population increased in 



Regulation of population size 223 



