458 Succession and Fluctuation 



is considered to be a maximum, then analysis of the same records for 

 the Canadian lynx yield an average of 5.8 years for the period of the 

 cycle (Cole, 1951). Lynx populations in Finland oscillate with an 

 average of 3.5 years for the cycle. The 10-year cycle may be inter- 

 preted as the average distance between "high peaks" of a basic cycle 

 with maxima 3, 3, and 4 years apart, of which the maximum after the 

 4-year interval is larger than the others (Siivanen, 1948). 



Possible cycles of other periods have been reported and are sum- 

 marized by Hutchinson and Deevey ( 1949 ) . The long-haired rat in 

 western Queensland, Australia, appears to display an 11-year cycle. 

 Irruptions of the chinch bug {Blissus leticopterus) between 1823 and 

 1940 occurred with an average period of 9.6 years for the cycle, but 

 individual periods varied from 7 to 13 years and the outbreaks are not 

 considered to be truly cyclic. The population of guano-producing sea 

 birds off Peru undergo catastrophic reductions about every 7 years 

 when the warm current "El Nino" moves farther south than usual, as 

 described in Chapter 5. 



Of a different nature is the periodicity of certain other insects in 

 which the cycles are controlled by the simultaneous emergence of 

 the adults after long, but often regular, periods of dormancy. Per- 

 haps the most spectacular is the 17-year cicada (or "locust") Magi- 

 cicada septendecim which in "locust years" swarms over the country- 

 side but during the intervening years remains unseen as a nymphal 

 stage in the ground. During one irruption in an oak-maple forest near 

 Chicago these cicadas occurred at a density of 500,000 per hectare ( 50 

 per sq m), equivalent to 77 kg of dry tissue per hectare (Strandine, 

 1940). 



Fluctuation in growth of certain trees as revealed by the width of 

 their growth rings often appears to be definitely periodic. The records 

 studied by Douglas (1936) revealed a recognizable cycle of 9i/(> to 10 

 years. The growth of hemlock trees in Pennsylvania, traced for over 

 230 years and of Douglas firs in Utah traced for about 1500 years 

 oscillated with average cycles of almost exactly 3 years (Schulman, 

 1948). Nevertheless, during the period 1703 to 1939 when the data 

 for both tree species can be compared, the fluctuations in the two 

 series appear to be completely independent. 



Causes of Fluctuation 



Discovery of the causes of fluctuations in communities is one of the 

 most desirable objectives of ecological research but also one of the 

 most difficult. We have seen that changes in the size of population 



