Popij] :3 



Field Sampling 



In field sampling, the usual procedure is to take a number of small 

 cores from an area of land and to mix than together to form a com- 

 posite bulk-sample. A true picture of the eelworm population could 

 only be obtained by the separate estimation of an infinite number of 

 small samples. Some idea of the relationship of bulk sampling to the 

 true population of land can be obtained by considering a transect 

 (Figure 1). XY shows the variation in eelworm population per unit 

 area along the transect j AB, the true mean of these values; AiB^, the 

 deviation of the bulk-sample mean from the true meanj and A2B2» the 

 laboratory estimate of the bulk-sample mean. The last may depart 

 considerably from the true mean of the population and gives no indi- 

 cation of the point to point variability of the population. 



Figure 1 



Relationship between within-field, or plot variability, 

 bulk sampling and laboratory estimation of bulk-sample. 



Ai 

 A 



Position or transect: 



XY = variations in population density along transect. 



AB = true mean of XY. 



A^Bi = true bulk-sample mean. 



A2B2 = laboratory estimate of true mean. 



During the last twenty years' work upon soil populations, little 

 information has been obtained upon which to decide the depth of 

 sampling, the nimber of cores per unit area, or the size of cores 



