At the time of sampling, swells were approximately 15-20 cm, develop- 

 ed over a short (7 km) fetch, and waves were estimated at 10-15 cm (SOR 

 Team data). Mixing energy was thus low, so that one would not expect 

 the mechanical entrainment of large droplets of oil. 



The slight losses of the lower weight molecules indicated by the 

 fluorescence spectra, is similar to Clark's observation of losses in the 

 surface mousse (section 5.2); indicating that the surface mousse is the 

 likely source of the water column contamination. The great compositional 

 similarity of the water column contamination with this mousse (probably 

 even greater than that shown for whole oil) indicates two possibilities: 

 accidental contamination of the sample jar by surface mousse, or the 

 presence of micro-dispersed oil in the water column (the water soluble 

 fraction would not be similar to whole oil). Accidental contamination 

 of the sample jars would be expected to result in highly variable con- 

 centrations. The nearly identical, low-level concentration values 

 observed (one hour apart) would be an unlikely coincidence if resulting 

 from accident, although additional data or more than two samples are 

 needed to state this conclusively. Oil in micro-dispersed form, however, 

 was also observed during this time period adhering to and being ingested 

 by zooplankton (Chapter 4). The sedimentation trap results (Chapter 11) 

 further suggest particulate oil in the water coli 



L umn . 



5.4.3 Conclusion 



It can be concluded that concentrations of Tsesis oil of 50 to 

 60 (Jg/1 were observed by fluorescence at \ and 1 m depths in low mixing 

 energy conditions, below a captured slick which was fully emulsified and 

 which had weathered for 2 to 5 days. This oil was most likely micro- 

 dispersed particles from the surface mousse. The oil in the water 

 column (as with the surface mousse) bore great similarity in composition 

 with the cargo oil, except for slight losses of the lighter molecules. 

 This similarity can be attributed to the rapid emulsif ication reported 

 by Clark (section 5.2), which retarded evaporative losses and other 

 forms of physical weathering. 



92 



