GENERAL PROPERTIES OF POPULATIONS 



269 



Hares banded in precensus period 

 Other hares present in precensus period 



_ Banded hares trapped in census period 

 Other hares trapped in census period 



Applied to the actual case this becomes; 



948 _ 167 

 X 254 



,. _ (948) (254) 



^^ (167) ^^^^ 



948 + X = Total population 

 948 + 1442 = 2390 hares; the total population 



The authors believe that "... by this 

 method we obtained a fairly accurate esti- 

 mate of the population of the middle of the 

 piecensus period; i.e., about February 1." 



There are two major sources of error in 

 this method. If the marking incapacitates 

 the animals in any way, they may not re- 

 distribute themselves as do unmarked forms. 

 This apparently did not obtain in this case, 

 since the rabbits were marked by small 

 metal bands clipped to the ears. It might 

 hold true, say, in insects if spray paints are 

 employed. The other error is introduced 

 when deaths and migrations occur between 

 samplings. This would be particularly 

 significant in short-lived and highly motile 

 animals. This error is discussed by Jackson 

 (1933, 1936, 1939), who introduces cer- 

 tain corrections that can be used to take it 

 into account.* 



Dennis Chitty of the Bureau of Animal 

 Population at Oxford University, who has 

 recently worked intensively with the mark- 

 ing method, has kindly given us permission 

 to quote the following statement as an 

 evaluation of this method: "The value of 

 the marking technique is limited by two 

 general considerations: (1) the behaviour 

 of marked animals may not be representa- 

 tive of the whole population and death 

 rates may differ too much between marked 

 samples; (2) even when all theoretical re- 

 quirements are satisfied, the error of the 

 estimates may be rather considerable and 

 cannot at present be evaluated. Where 



• An application of the marking method ap- 

 plied to the determination of population densit}' 

 of sheep blowflies (Lticilia cuprina) and in- 

 cluding statistical analysis of the findings has 

 been presented by Gilmour, Waterhouse, and 

 Mclntyre (1946). The authors conclude that 

 the major error in their method was of the 

 magnitude of 20 per cent. 



these limitations can be overcome the tech- 

 nique is simple and useful." 



6. Indirect methods. Population size 

 is sometimes estimated by using the fre- 

 quency of some manifestation of the popu- 

 lation as a criterion of its size. This is not 

 a census method sensu stricto, but, rather, 

 an index of abundance, and its applica- 

 bility is limited as is its accuracy. Prop- 

 erly, it should be used as a supplement to 

 sampling; i.e., to substantiate or invahdate 

 an estimate of population size gained by 

 another method. This point made, it re- 

 mains for us to list some of the indirect 

 techniques which have been used without 

 attempting to evaluate them. The list that 

 follows is a summary of certain of these 

 "manifestations:" 



1. The number of shed antlers (deer popu- 

 lations ) . 



2. The number of fecal pellets and drop- 

 pings (derived from "Scatology"). (For discus- 

 sions of this method, see MacLuHch's 1937 

 work on the varying hare and Scott's 1941 

 study on the red fox. For a report overen- 

 thusiastic to the point of humor, see Seton, 

 1925. ) 



3. Pelt records, particularly those of the 

 Hudson's Bay Company of Canada. These 

 yield data for many mammals whose hides are 

 marketable. Difficulties encountered in using 

 such records are discussed ably by Elton 

 (1942). 



4. Amount of food consumed. 



5. The frequency of prey remains in the 

 stomach contents of predators. (For a de- 

 servedly critical discussion see Errington, 

 1937.) 



6. Vocalization frequency; bird calls, mam- 

 mal noises, and the like. (See discussion of L. 

 Tinbergen's work, p. 374. ) 



7. The frequency and pattern of tracks and 

 trails. 



8. The amount of oxygen consumed or 

 carbon dioxide produced by a defined popula- 

 tion ( See Howe and Oxley, 1944. ) 



7. Combination of several methods. In 

 those cases in which a total census cannot 

 be taken or in which the sample is not con- 

 sidered adequate, the careful investigator 

 may combine several of the above tech- 

 niques in getting at population size. The 

 final derived figure is based on the correla- 

 tion that exists between the several reliable 

 methods. One of the better illustrations of 

 such application is MacLulich's (1937) 

 study of the varving hare. This investiga- 

 tion, centering about population fluctua- 



