F-3 



6.2 Distribution 



In broad terms the overall geographical distribution of ^ superba may be taken as being 

 circunrpolar south of the Antarctic convergence (Marr 1962, Baker 1954) and is one of the 

 dominant macroplankton organisms. (The dominant components of the macroplankton in the 

 Pacific sector may be Salps and Coelenterates(Barkhatov et al. 197 3) although Mackintosh 

 (1973) thinks most of the krill there are in the East Wind Drift and therefore in the Pack 

 loe Zone). There are few records of Ej_ superba occurring north of the convergence. 

 Mackintosh (1973) described krill north of South Georgia in a tongue of cold water, Sasaki 

 et al. (1968) described a concentration at 45°S 145°E and Ball and Dunstan (1957) found fresh 

 krill in the stomsich of a humpback whale caught off Queensland. Marr (1957) considers such 

 oocurrances of minor importance. Within these very broad limits the density of krill is 

 extremely variable. Marr (1962) describes the major concentrations as occurring in the East 

 Wind Drift, Scotia Sea, Weddell Drift and South Georgia areas although Mackintosh (1973) has 

 shown the existence of large concentrations elsewhere. As. a result of recent sampling Nemoto 

 (1968) has shown the presence of large concentrations in the vicinity of the Kerguelen 

 Gaussberg Ridge and at Longitude 150°W (see also map in Mackintosh 1973). This area of con- 

 centration is also suggested hy Beklemishev (I96O, I96I) on the basis of whale catches in the 

 region. 



Because krill concentrations tend to occur in certain clearly defined areas, it has been 

 suggested that these may represent several self maintaining populations (implied by Mackintosh 

 1973) or even distinct races (Hakarov 1974). Clearly whether or not such a system is present 

 will have direct effect on the way in which this resource is managed. In this review, the 

 distribution 6f all life stages has been considered in order to lead to an overall understanding 

 of the population structure. 



Distribution of eggs and early larvae 



In an analysis of the vast collection of material collected by Discovery Investigations, , 

 Marr (I962) showed that E;_ superba spawns in the surface layers (althou^ Fraser (1936) con- 

 sidered that spawning occurred on the bottom in the shelf zone). The eggs then sink and begin 

 the processes of cleavage leading to hatching. The scarcity of cleaved eggs in the surface 

 samples and the fact that Bargnann (1937) records spawning in aquaria but without subsequent 

 cleavage implies that depth (probably pressure) may be necessary for development. Hatching 

 is thou^t to occur in deep water and the resultant nauplii moult throu^ the metanaupliar 

 stage as they rise through the water column to reach the surface as first calyptopes. This 

 description of the cycle of events is clear from Marr' s (1962) analysis. What is not clear 

 and which has been the subject of a great deal of study are the details of locality, time and 

 conditions under which these events take place. 



Althou^ Marr (I962) recorded both gravid and spent females in the oceanic surface water, 

 he considered that hatching in the oceanic deep water far away from the shelf was negligible. 

 He did however consider that the early larvae were carried in the cold bottom water away from 

 the continental shelf to undergo the developmental ascent in the oceanic deep water. This is 

 shown diagrammatically in Fig. 6.3. 



In his analysis Marr considers very carefully all the evidence both for and against this 

 explanation because it clearly has a great deal of bearing on the dynamics of distribution. 

 In this respect the important factors are the time taken over the early development and the 

 rates of flow of the cold bottom water and warm deep water. 



There is at present no direct information on early development rates for Ej_ superba . 

 Marr assumed that Ej_ superba would take about half as long again as the 15 days Mefcanyctiphanes 

 norvegica takes to reach the first calyptopis stage (p. 207). Mackintosh (1972) also considers 

 three weeks to be a reasonable estimate although his analysis is based on much the same in- 

 formation as Marr's. 





