Origin of Cycles 463 



future. Fluctuation in populations of guano-producing birds is defi- 

 nitely determined by periodic influx of warm currents. The growth 

 of Douglas fir in Utah was found to be closely correlated with precipi- 

 tation, but the growth of hemlock trees in Pennsylvania during the 

 same years with the same average cycle period exhibited no significant 

 relation to rainfall. The oscillations of the arctic fox {Alopex lagopus) 

 and of the snowy owl (Nijctca nijctea) are evidently timed by those 

 of the lemmings and mouse-like rodents on which they feed. As the 

 rodent population diminishes, the snowy owls disperse in search of 

 food, and their periodic appearance as far south as New England 

 attracts much interest among ornithologists (Gross, 1947). Maxima 

 of the Canadian lynx follow those of the snowshoe rabbit by a year or 

 so. The fact that the colored fox {Viilpes fuJva) and the marten 

 {Mnrtes americana) exhibit a 4-year cycle in the arctic and a 9- or 

 10-year cycle farther south is perhaps due to a corresponding differ- 

 ence in the periodicity of its prey in the two regions. In at least some 

 of these instances the fluctuations of the prey are not controlled by 

 the predators as in reciprocating intrinsic cycles. For example, the 

 snowshoe rabbits on Anticosti Island, where there are no lynx, fluctuate 

 with the rabbit population on the mainland where lynx abound. 



In the foregoing type of situation the apparently cyclic fluctuation 

 of the species is not due to a random eftect of many environmental 

 influences but is controlled principally by the changes in one par- 

 ticular factor. The biological cycle might then be caused either by 

 apparently cyclic random fluctuations of this controlling factor, or by 

 a truly cyclic change in this factor. A long search has been made for 

 a predictably regular cycle in the climate or in some other feature of 

 the environment that might serve as a master timer for biological 

 cycles, but no convincing and generally accepted factor has been 

 found. Sunspots were at one time thought to be correlated with cer- 

 tain population cycles since their average period is about 11 years and 

 is thus closely similar to certain cycles of growth and abundance. 

 However, the sunspot cycle is more variable than the Canadian lynx 

 cycle, for example, since the period of the former has varied from 7 to 

 17 years as against 8 to 12 years for the latter. In addition, a com- 

 parison of the two cycles from 1750 to 1935 shows that, although they 

 correspond for a nimiber of years, they subsequently become com- 

 pletely out of phase (Fig. 12.16). 



A satisfactory explanation of these intriguing cycles of growth and 

 abundance accordingly continues to elude us. These fluctuations are 

 another manifestation of the many changes going on within the com- 

 munity that have been reviewed in this chapter. We have seen that 



