Discussion 



The data compiled in this study do not fit the curve of a normal distri- 

 bution. In the frequency of occurrence of mortalities, there is some resemblance 

 to the Poisson distribution, but the fit is poor. The data represent simple 

 empirical discoveries in a particular location over an appreciable length of 

 time, and are valid in expressing what occurred under recorded conditions. 

 Only very limited previous studies of this type have been conducted, and from 

 these only very limited conclusions can be drawn. Controlled experiments 

 conducted cooperatively by the Naval Ordnance Laboratory and this departm.ent 

 (1948) indicated that the effective killing range of underwater explosions 

 (using 30 to 300 pounds of explosive) usually does not exceed 100 yards from 

 shot point. Observations by Coker and Mollis (1950) confirm this lethal 

 distance estimated under conditions wherei the weight of charges of similar 

 explosives ranged from 250 , to 1200 pounds. No accurate estimates were 

 obtained for the explosions observed in the course of this report, bit much 

 the same pattern of kill occurred. Cole (1948) describes the "slick" of 

 darkened water, which moves across the surface as an enlarging circle from 

 the site of an underwater explosion, and which indicates the points to which 

 the shock wave has advanced. Tins phenomenon was seen re^ijular'.y, and gave 

 a rough means of measuring the range of die mortality. Complicating factors 

 are present in this study which account for the peculiar distribMti<:>n, which 

 must be considered. 



First, the pounds of fish killed jln any particular explosion is dependent 

 upon the number of fish present, and the sizes of those fish. Second, several 

 species of fish share the area, each of which has a characteristic and 

 different abundance (in both number and weight) from season to season, and 

 from year to year. Third, the distribution of fish in any area is dependent 

 to a considerable extent on their behavior and ecological relationsJnps. Tide, 

 temperature, salinity, oxygen, weather conditions, even time of day a1 1. 

 exert infh'e'rces on the movements and behavior of fish. 



AlthoMgh the characteristic feeding habits and movements of the several 

 specir-s o( (ish encountered in this study apparently have some lelAtionship to 

 the inaj.T'Kude of Uieir respective mortalities, there are other factors wh'ch 

 must be considered. For example, would those species t^'pically inhabit the 

 expJo.-vivp test area in equaJ concentrations? Specifically, wo..''das m^oy 

 croa'^ev and spot be feeding over the bof torn in the test area as in oth'er areas? 

 Would stuped bass, river herring or menhaden be distributed >n the same 

 manner over the shot point as in the rest of the river? Do nhys:<,HJ barriers 

 exist which would tend to exclude or limit certain species? rJa,?i> (.l'^47) in a 



17 



