25 



similar or different needs. There is competition, sometimes new species 

 entering a formerly occupied niche, sometimes different species seeking 

 the same food, oftentimes among members of the same species. 



Different species of fish often intermingle in the marine environ- 

 ment, particularly botton fish. 



Fish tend to prefer specific sites for certain activities, but iden- 

 tification of the causes beyond first-order estimates is extremely diffi- 

 cult, made so by the great number of variables that must be measured. 



In simple terms, the ocean is a big place and all aspects of it 

 cannot be monitored. Even if it could be well monitored, the interpre- 

 tation of feedback requires an understanding of the forces that operate 

 ecosystems, and those forces themselves deliver complex, at times impos- 

 sible to measure, feedback. 



Until this information is known, the foundations of marine science 

 must remain general. The demands on that science, clearly, cannot ask for 

 more specificity than that science provides. 



In addition, there exist complexities concerning the biological data 

 that is received through survey activities. These complexities must be 

 incorporated in any model so as to gauge bias. Some sources of bias are 

 extremely difficult to erase: 



• Any sampling method that is random or stratified in scope must, to be 

 accurate, reflect the possible variations at each sampling station. 

 Yet, these variations cannot always be judged. Determination of temper- 

 ature, salinity, etc., during a tow is relatively straightforward, al- 

 though the depth of the towed body, its swept area, and its position 



in relation to earlier tows may be somewhat in error. 



• Any specific population analysis must first identify the chosen 

 population unit. When such a population is scattered or widely 

 spearated, data are inconclusive. 



