144 AXLMAL ECOLOGY 



to carry- through their breeding activities successfully. But in 

 anv case the problem remains. 



22. The consequence of this is that as soon as a food animal 

 sinks below a certain degree of abundance its enemies either 

 stan-e or turn their attention to some other source of food. 

 Fluctuations in numbers have therefore a potent influence 

 upon the food-habits of animals. In fact, if several important 

 kev-industr}- species become suddenly ver\' abundant or very 

 scarce, the whole food-cycle may undergo considerable changes, 

 if only temporarily. The various automatic balanced systems 

 which exist ^^ill tend to bring the numbers, and therefore the 

 food-habits, back in the long run to their original state. x\nother 

 way of putting it is that the favourite food of an animal is not 

 usually the most abundant one. Many animals have a definite 

 scale of food preferences, depending upon qualit}', but if a 

 favourite one becomes suddenly common the animal will 

 abandon its previous food ; and conversely, if a common food 

 becomes suddenly scarce it may still continue to seek it up 

 to a certain point, beyond which taste must give way to 

 necessity'. These variations in food-habits may lead to corre- 

 sponding variations in the habitat to which the animal resorts. 

 This is shown by the follo\^ing example of the habits of the 

 common rook {Corviis frugilegus) , which is meant only to 

 illustrate the idea, and not to prove anything final about the 

 rook itself. During a particularly dull railway journey the 

 writer noted dowii the number of pasture fields and of ploughed 

 fields recently sown, and at the same time counted the number 

 of fields of each kind which contained flocks of rooks feeding. 

 The figures obtained were as shown in the table : 



It will be seen that although there were comparatively few 

 ploughed fields they appeared to be much more sought after 

 than the pasture land (assuming for the sake of argument that 

 the figures are sufficiently large to prove this — which they are 

 not). But the absolute number of rook-flocks on pasture 



