172 ANIMAL ECOLOGY 



It is very important that specimens should be killed and 

 preserved in the appropriate way, as otherwise they may be 

 useless for purposes of identification, or, at any rate, cause a 

 lot of unnecessary trouble. 



12. For constructing food-cycles there is only one method — 

 the patient collecting of all kinds of information about the 

 food and enemies of the species that are being studied. Direct 

 observation in the particular place in v^^hich you are working 

 is the best, since the food habits of animals are often very 

 variable at different times and in different places. It may be 

 convenient to summarise the various ways in which evidence 

 about animals' food may be obtained. 



(i) Watch the animal eating and, if necessary, take a speci- 

 men of its food (and of the animal itself). This is the type of 

 evidence that is most difficult to obtain. 



(2) Examine the contents of the crop or stomach or in- 

 testine. This may give good positive evidence but is useless 

 for proving a negative, e.g. remains of butterflies disappear 

 very quickly under the influence of the digestive juices of 

 birds. 



(3) Finding stores of food, etc. 



(4) From a study of the animals and plants associated with 

 it, deduce the animal's probable food. This enables the field 

 of observation to be narrowed down. For instance, the 

 writer saw a ptarmigan rise from a hillside, and on going to the 

 spot where it had been, found that a number of seeds and 

 flowers had been eaten from various plants. Since this bird 

 was the only large herbivorous animal in the region, it was 

 certain that the bird had eaten them itself. 



(5) Experiments may be made to confirm such proba- 

 biHties. 



(6) Examine excretory products, e.g. castings of owls 

 containing remains of small mammals, or droppings of terns 

 containing limbs of Crustacea. 



(7) The structure of the animal will help to narrow down 

 the field to a particular size of food. 



(8) Note any food preferences, with reference both to 

 quality and quantity. 



