4 



Background: 



Tlic analysis of ecological comiiuinitics must in- 

 diulf a ineasurcnicnt of animal |jo|)iilations that the 

 role |)laye(l by i-ach sjifcies may be \iropeT\y evalu- 

 ated. The ecologist should also be able to determine 

 i|uantitati\ely the iihundance of species at difTerent 

 times and different places. It is not sufficient in eco- 

 logical research to indicate that a species is abundant, 

 common, or rare; abundance must be expressed in 

 such objective terms as lend themselves to statistical 

 manipulations. In spite of their fundamental imjxjr- 

 tance, available methods for measuring population 

 size are only moderately satisfactory and are in need 

 of vast improvement (Balogh 1958, Davis in Mosby 

 1960). 



Indices of abundance are sometimes used; for 

 instance the number of individuals or songs observed 

 per hour, per day, or per trip; per cent (frequency) 

 of samples in vk-hich the species was recorded ; num- 

 ber of nests, dens, tracks, or fecal pellets per unit 

 area; amount of food or bait consumed per unit of 

 time, and so forth. Under certain conditions of uni- 

 form habitat and weather, random distribution of 

 individuals, and uniform conspicuousness of the ani- 

 mals, indices are useful for demonstrating differences 

 in population size within a single species as functions 

 of time or space, but they are seldom accurate enough 

 to allow comparisons between different species. 

 There have been various attempts to correlate rela- 

 tive indices with absolute abundance (Hendrickson 

 1939, Bennett et al. 1940, Cahalane 1941, Baum- 

 gartner 1938, Emlen ct al. 1949, Eberhardt and Van 

 Etten 1956), but the results have been usually un- 

 satisfactory (Clapham 1936, Dice 1952). In most 

 types of ecological research, the aim should be to de- 

 termine absolute abundance or the actual number or 

 biomass of a species in an area of known size. The 

 difficulty in doing so is no greater than in correcting 

 relative indices for all the variables that are involved. 



STRIP CENSUSES 



Measurement 

 of Populations 



Tliis method is one of counting all indi- 

 viduals of birds and larger mammals seen on each 

 side of a line of travel over a measured distance. 

 Sometimes the count is made only of animals ob- 

 served within a definite distance from the line of 

 travel. In other cases, the effective width, and hence 

 the area, over which the animals are being censused 

 is computed as twice the average distance at which 

 each species is first observed. This makes possible a 

 quick survey of large areas in any kind of terrain, 

 but is subject to inaccuracies of individuals omitted, 

 especially as the distance from the trail increases 

 (Hayne 1949a); differences in conspicuousness of 

 different species or individuals e.\hibiting atypical or 

 unusual behavior; and variations in visibility as one 



31 



