460 Succession and Fluctuation 



problem as complete a knowledge as possible must be obtained of 

 the interrelations of the species in question with all aspects of its en- 

 vironment at every stage in its life cycle. Apparently correlated con- 

 ditions must be examined over a period of time and in different parts 

 of the range. When it is feasible, the factor suspected of causing the 

 fluctuation should be tested by laboratory or field experiments. 



Many illustrations of the control of growth and reproduction by 

 individual environmental factors, or combination of factors, have been 

 given in previous chapters, and others may be found in reports dealing 

 primarily with this topic, such as those of Shelford (1951, 1951a). 

 Following extensive investigations of the fluctuations in certain forest 

 and non-forest animal populations in the Middle West, Shelford 

 emphasized the fact that environmental influences exert their con- 

 trol chiefly during sensitive periods in the life histories of the 

 species concerned. Forest invertebrates, for example, were found to 

 develop larger populations in years when rainfall is great during late 

 March, April, and May with short dry periods, whereas for certain 

 birds and mammals the conditions in February and March are more 

 significant. 



Origin of Cycles. Because of the lack of sufficiently precise data 

 we cannot at present state that the fluctuation of any population is 

 strictly cyclic in the sense that accurate prediction of the times of 

 maxima and minima can be made well into the future. Further study 

 of the possible existence of truly cyclic behavior is very much desired 

 since, in addition to the theoretical interest, prediction would permit 

 precautions to be taken against outbreaks of pests and preparations to 

 be made for glut or scarcity of fish, game, or furs. Since fluctuation 

 may be only approximately cyclic, we must include random variations 

 as a possible explanation. As possible causes of these biological 

 cycles we shall examine self-induced effects within the population, 

 random effects of many external factors, and single controlling in- 

 fluences that themselves fluctuate in a random fashion or are truly 

 cyclic. 



Regular oscillation in abundance of a single species might be pro- 

 duced in a constant environment as the result of delays in the effects 

 of excessive or favorable numbers. Such a population would alter- 

 nately overshoot and undershoot a possible equilibrium level, as de- 

 scribed in Chapter 9. When two or more species affect each other's 

 abundance reciprocally, as may occur in prey-predator or host-parasite 

 combinations, oscillations may be similarly set up, as pointed out in 

 Chapter 10. Such oscillations caused by reactions within populations 

 of one or more species are termed intrinsic cycles. 



I 

 I 



