The physical environment should be sampled both vertically and horizon- 

 tally. Vertical sampling should include samples from the sediment, the water 

 column, the water surface, and if the spill is particularly large or the 

 spilled substance particularly volatile, the atmosphere. Horizontal sampling 

 should be carried out in such a manner that concentration gradients of the 

 spilled substance could be established. 



Sampling of the food web should include samples of benthic organisms, 

 plankton, fish, and higher carnivores when appropriate. At least one group 

 of sessile benthic organisms such as mussels or oysters should be sampled 

 during the period of acute or short-term spill impact. 



Even though such a procedure may be of questionable scientific value, it 

 is advisable to analyze the tissues of highly visible organisms such as 

 waterfowl and fish whose carcasses often wash ashore following a spill. The 

 deaths of such organisms usually have a disproportionately large psychological 

 impact on the public. Analytical data verifying that the spilled substance is 

 in the tissues of such organisms can be of considerable value to an attorney. 



Phase II -- Period of Chronic or Long-Term Impact 



Analysis of a Limited Number of Serially 

 Collected Samples from the Spill Site 



Analytical documentation of the persistence of the spilled substance in 

 the physical environment and the food web is necessary to assess the chronic 

 or long-term effects of the spill. 



For this type of documentation serial samples should be taken over a 

 period of time, usually measured in years. Samples should be collected at 

 close intervals immediately after the spill, then less frequently as time 

 progresses. No fewer than four samples should be taken per year. This 

 number is based upon the assumption that biologists would continue their 

 assessment of chronic environmental impacts by sampling spill site biota 

 during each of the four seasons. Analysis of serial samples would also docu- 

 ment the weathering processes undergone by the spilled substance under the 

 actual environmental conditions encountered after the spill. Such documenta- 

 tion lends credence to identifications of severely weathered forms of the 

 discharged material in environmental samples made months or even years after 

 the spill . 



The initial set of samples taken to determine persistence of the spilled 

 substance in the environment should, ideally, include every sample type (not 

 e^jery sample) collected for delineation of the spill in Phase I. Then, as the 

 concentration of the spilled substance returns to the control level in any 

 particular sample type, that sample type is dropped from the next sampling 

 cycle. In all but the most severe spills, this will occur very quickly for 

 most water-column and biotic samples. Indeed, the spilled substance will 

 probably persist longest in sediments, which are proven sinks for many organic 

 (Farrington and Tripp 1977) and inorganic (Bryan 1976) pollutants, and in 

 benthic organisms, such as mussels and oysters (Stegman and Teal 1973). 



175 



