F-19 



In this context a swairo can be defined as a dense aggregation of individuals moving ha]>- 

 moniously and is probably the type of aggregation that will give a strong indication on an 

 echosounder. A swarm may therefore be considered as the extreme case in terms of co-ordinated 

 movement of individuals in a group. The other extreme would be independently random movements 

 by individuals. Between these two extremes there almost certainly exist concentrations that 

 are intennediate in terras of aggregation and organization. 



Swarming has already been discussed briefly in the section on zooplankton where it was 

 suggested that swarming was a result of some social behaviour pattern (as yet unidentified) 

 working in areas of hif^ average concentration (such as discussed in the section on krill 

 distribution). The section which follows is a review of information on swarm formation and 

 disintegration. 



Marr (1962) concluded on the basis of analysis of plankton net samples that 

 from hatching onwards the number of krill outside of swarms. was negligible. Although Marr 

 was well aware of the maximum density krill attained when swarming his definition of a swarm 

 was not nearly as restricted as that given above. 



Other authors have been more evasive on the point only mentioning swarms without making 

 any firm commept on the proportion of the total population present in swarms. The observa- 

 tions of Pavlov (1969) and Shust (I969) indicate that swarming is a transient phenomenon 

 related to food availability. These observations were based on the state of the gut in 

 freshly cau^t animals, a method also used by Nemoto (1968). According to Pavlov when phyto- 

 plankton is abundant the krill spend part of the day actively feeding in the surface water 

 until they are replete at which time they aggregate into swarms and sink. In one region in 

 which Pavlov made his observations this cycle occurred twice in 24 hours. The cycle of swarm- 

 ing and dispersion was repeated in other areas although without vertical migration; Pavlov 

 attributes this difference to one of size; adolescent krill, he suggests, migrate vertically 

 whereas adult krill do not. Shevtsov and Makarov (l969)» however, note that in the course 

 of vertical migration the adult krill tended to be deeper and in fact on occasions formed a 

 separate layer (detected by echosounder) beneath the adolescent krill; they make no comment 

 on the causes of swarming. 



Nakaraura (1973) identified several type:5 of krill aggregation and related their presence 

 and formation to light intensity. He found the densest svrarms at the surface on the darkest 

 ni^ts and when the ligti't intensity increased the density decreased as the animals migrated 

 down. The commonest pattern of aggregation that he identified was a layer generally within 

 the range 30 to 100 m depth or about 20 m above the thermocline. The results of Ozawa _et al. 

 (1968) conflict with these observations and indicate surface aggregations in daytime; they 

 were hovrever a result solely of visual observations. Shust (I969) considered that the level 

 of illumination (cloudiness) of the environmental factors he measured (wind, sea state, atmos- 

 pheric pressure) was the most important although he considered feeding also to be of major 

 importance. 



The suggestions of Ruud (1932), Beklemishev (I96O) and Bogdanov (1974) that turbulence 

 and eddy currents are responsible for the formation of svrarms is probably not valid on the 

 above definition of a swarm but more likely applies to the formation of high average abundance. 

 Marr (1962) and Naumov (l962a) suggested that some swarms were in part spawning concentrations. 

 Althouf^ there does not appear to be any conclusive evidence (no catches have knowingly been 

 made of krill in the act of spawning) Makarov (1973) has shovm the presence of concentrations 

 of ripe individuals during the spawning period; he does not, hoifevcr, describe the concentra- 

 tions as swarms as defined above. It has also been suggested that swarming is a result of 

 increased buoyancy due to the increase in lipid content during the summer (SurukovBkii 1967 )• 

 Presumably this would be expected to have the effect of concentrating the krill into LanfTtiuir 

 cells as described by Stavn (1971) for Daphnia although there is no evidence to suggest that 

 for krill this does in fact occur. 



