The general analysis of variance table (Table 22-2) for this particular model 
is shown. From this table the partitioning of the degrees of freedom can be 
determined as well as the appropriate tests of significance determined. Two 
points can be made concerning this model. They are: 1) the variation between 
sample locations can be estimated, and 2) the systematic variations between 
sample locations can be eliminated from the study of other effects, i.e. depth. 
Assuming that an approach such as the above has been applied to 
establishing whether one or more sampling stations are required for the 
monitoring program, we now turn to a detailed analysis of a single station, 
which was a major component of the work which has been performed to date. 
341 
