184 



Table 3 - The relationship befween extent of dispersion, concentration of oil in solution and toxicity of physically- 

 dispersed oil in the sea test 



after sampling, it was not possible to measure the 

 droplet size distribution of the physically-dispersed 

 oil using this method because of the instability of the 

 oil droplets. The concentration of physically and 

 chemical]>x3ispersed oil present in each tank was 

 also measured by ultra-violet fluorescence spectro- 

 pho'.ometr)'. 



The results of these experiments are given in 

 Figure 3. For each treatment, the mortality resulting 

 from addition of dispersant to an oil film (Method 2) 

 was less than that resulting from addition of the pre- 

 mixed oil and dispersant (Method 1). For two of the 

 three concentrate dispersants this reduction in to.xi- 

 city was accompanied by an increase in the mean 

 droplet size of the dispersed oil. There also seemed to 

 be an inverse relationship between droplet size and 

 toxicity when Method 1 (premixing) was used for all 

 three concentrate dispersants. Although there was an 

 empirical link between increased droplet size and 

 reduced toxicity it is possible that the Coulter 

 counter itself may have modified the size distribution 

 of the dispersed ofl droplets or that this distribution 

 could have changed between being sampled from the 

 test tanks and analysed on the Coulter counter. In 

 this case the larger droplet size when Method 2 (dis- 

 persant added to oil film before agitation) was used 

 could be due to the production of a less stable dis- 

 persion in which some degree of droplet coalescence 

 occurred before analysis. 



The concentration of oil dispersed by each method 

 and dispersant was similar; such differences that did 

 occur may have been caused by the difficulty of 

 sampling through a surface oil film, especially in the 

 oil ortly tanks. There was no obvious relationship be- 

 tween the concentration of dispersed oil and toxicity. 



An attempt was also made to determine the cause of 

 death of the shrimps in the test. The two principle 



effects of acute poisoning by oil are chemical toxicity 

 (in which the oil hydrocarbons disrupt cellular or sub- 

 cellular functions) and physical impairment (in which 

 epithelial cells are coated and the animals are depriv- 

 ed of oxygen) (Moore and Dwyer, 1974). Nelson- 

 Smith (1972) suggested that the constituents of dis- 

 persants could increase the uptake of oil components 

 and Anderson et a/.(1974a, b) obtained evidence that 

 the soluble aromatics of an oil were responsible for 

 most of its toxic effects. Samples of shrimps from 

 some of the oil and dispersant tanks were analysed at 

 the end of the 24 h recovery period for naphthalene 

 and its C,-, C^- and C,- derivatives by the technique 

 of mass fragmentography (Law, 1978a). Concentra- 

 tions of these compounds m the tissues of the 

 shrimps which survived the experiments (0.3-0.8 /ill'' 

 total naphthalenes) appeared to be greater than those 

 in shrimps which had died (0.1-0.3 \A 1). It is 

 therefore unlikely that ingestion of these or related 

 chemicals was responsible for the mortality. A micro- 

 scopic examination of the dead shrimps revealed 

 small oil droplets in the gill cavity and a coating of oil 

 on the mouthpans and in the stomach. Little 

 difference was observed between the shrimps exposed 

 only to oil and those exposed to oil chemically- 

 dispe-^'d. It therefore seemed possible that shrimps 

 were killed by physical processes, such as suffocation 

 due to the mechanical clogging or damage of respira- 

 tory surfaces. These effects have been noted by other 

 workers (Nelson-Smith, 1972; McKeown and March, 

 1978). 



The experiments reported here have shown that the 

 toxicity of the oil in the sea test is dependent upon a 

 number of factors. A more precise determination of 

 the individual contribution of each of these factors 

 towards the toxicity of an oil or oil-dispersant dis- 

 persion was not possible during these studies which 

 dealt mainly with the validation of test procedures. 



