135^ 



material for tracers prior to initiation of the dredging project as well as the 

 native ambient sediments from a reference area. 



Formalin is recommended to be used as a preservative for microfossils instead 

 of methanol to avoid dissolving calcareous tests. 



Multivariate statistical analysis proved to be a valuable tool to identify sample 

 grouping patterns. Clustering (ordination) and MDS methods are most strongly 

 recommended because the samples are analyzed with no bias from layer 

 classifications from core descriptions. 



For more robust statistics, more consideration should be given to the absolute 

 abundance of microfossils in the samples, especially for rare species. In 

 samples where species are common, the abundance values are reliable to ±5%. 

 The abundance of rare specimen that occur below 5 % relative abundance has a 

 greater range of uncertainty. If documentation of the abundance of rare 

 specimen is important (e.g., shelf foraminifera in the Royal River), analyzing 

 more sample trays would increase the accuracy of the results for rare species, 

 but would also increase the time it takes to examine each sample. When the 

 density of microfossils is very low, other techniques might be considered to 

 examine a larger quantity of material, such as separation with heavy liquids. 



The Portland Disposal Site Capping Demonstration Project, 1995-1997 



