° 1912 ] McAtee, Contents of Bird Stomachs. 449 



METHODS OF ESTIMATING THE CONTENTS OF BIRD 



STOMACHS. 



BY W. L. MCATEE. 



In his report on "The Food of Birds in India," C. W. Mason 

 expresses decided opinions on the merits of the numerical and the 

 percentage-by-bulk methods of estimating the contents of bird 

 stomachs. Regarding his own work, Mason says: "I have made 

 no statements in a general way as to the relative bulks of the food 

 taken. We see it stated repeatedly that relative bulks of food 

 taken by birds are very important in any conclusion that we may 



wish to draw from economic entomology, [ornithology?] 



Our only method for obtaining this end practically consists of a 

 complete study of the food of the birds, from specimens obtained 

 throughout the year under all climatic, physical and seasonal con- 

 ditions and even at different times during the day (this latter 

 point is certainly one of importance in some birds and possibly 

 therefore in most) . . . .Comparative bulks of foods, if expressed 

 merely as percentages, are of absolutely no value whatever, and 

 cannot give any idea as to the true economic ratio of the food of 

 the bird in question. What we want to know is the exact number 

 of grains of corn, the number of insects, etc., taken, and we must 

 not draw our conclusions from a small number of records nor 

 from a mass of records that have been accumulated at one season 

 of the year only. W^e must take a fair average." (pp. 18-19.) 



As these dicta regarding the percentage valuation of the elements 

 of bird food reflect upon the methods in continuous use by the Bio- 

 logical Survey, U. S. Department of Agriculture, since 1895, a 

 defense of those methods is in order. Arguments similar to those 

 of Mason have been made before, but always it appears by those 

 having a relatively small amount of experience in the actual exami- 

 nation of the contents of birds' stomachs. In fact Mason's argu- 

 ments are not so strongly opposed to estimation of percentages by 

 bulk, as some of his sentences by themselves seem to imply. For 

 instance he admits that the bulk method would be satisfactory 

 if a large number of stomachs, representing all localities and seasons. 



