426 DISCOVERY REPORTS 



Strength of each brood in the first, second and third year of Ufe is estabUshed before it becomes 

 demersal, the factors determining the numbers in each instance being operative during the vulnerable 

 pelagic phase of the haddock's life-history (Parrish, 1956). In E. superba, it would seem, we have an 

 instance where the factors governing the abundance are operating in the reverse direction, the basis 

 from which its fantastic numbers spring being perhaps laid down during its early existence in the 

 deeps long before it reaches the surface stream. 



Simpson (1956) summarises the present position of our knowledge regarding survival in demersal 

 fishes during the pelagic phase in the following passage : 



. . . the accumulating evidence and speculation on theoretical grounds point towards the fundamental mortalities 

 being biological ones, which at some stage become density-dependent, that they are not cataclysmic but persistently 

 high while the fish is small and helpless, predation probably being the main cause. Superimposed on this mortality 

 there may be expected to occur more specialized types of mortality differing from species to species, and probably 

 from year to year, currents being important to some species in carrying the larvae to destruction, storms may destroy 

 millions of eggs by violent agitation, food shortages may lead to small survival in other species. While the former 

 type of mortality, predation, may normally be the main cause of death, both annual variations in predation and these 

 other factors may be expected to be responsible for annual fluctuations in survival. 



During its deep lifetime, it seems clear, as it hatches from the egg and grows to the Metanauplius 

 form, E. superba can be little affected by such biological or physical hazards, and in the prevailing 

 calm of its bathypelagic environment annual variations in such physical hazards as there may be are 

 likely to be small, resulting in little material fluctuation in survival from year to year. For predation 

 must be at a minimum, such currents as there are do not carry the larvae to destruction, but rather 

 the reverse, no storm can shake the eggs, and above all the Nauplii and Metanauplii do not require to 

 feed. They are of course unable to feed and so to them food shortages are of no concern, and density- 

 dependent mortalities, such as from time to time affect the multitudes of a feeding community, are 

 impossible. 



The swarming habit too, it seems, might contribute something, although at first sight it would seem 

 to place the krill at an enormous disadvantage, presenting as it does a rich and ready reward to major 

 predators such as whales, seals, penguins and fish. In its final phase, however, it must lead to the 

 liberation of astronomically large numbers of highly concentrated eggs, the females of a gravid swarm 

 (p. 250, Fig. 53), as Bargmann (1937, p. 341) has found, producing between 11,000 and 11,500 each. 

 Moreover, having regard to the sex ratio (p. 246, Table 49), in the tightly packed shoals, the prospects 

 of successful mating for all but for perhaps a negligible few of the maturing females would manifestly 

 be extremely high. In fact, as Bargmann (1945, p. 118) remarks, 'faulty implantation hardly ever 

 occurs '. 



Another, perhaps all-important, factor contributing to the sheer numerical pre-eminence of the 

 southern krill is that they, perhaps alone among Antarctic euphausians, are equipped, and equipped 

 superbly (Hart, 1934; Barkley, 1940; Marshall, 1954), to take the fullest possible advantage of the 

 enormously rich pastures of the circumpolar sea. In this respect their cleaving to a surface habitat, 

 to the optimum zone of phytoplankton luxuriance, might well contribute further both to the stature 

 of these animals and the immense scale on which they have populated the Antarctic seas. A parallel 

 instance of biological success, based on adaptation for algal feeding, is recorded by Fryer (1959) 

 among a group of rock-frequenting cichlid fishes from Lake Nyasa. ' It is significant ', he writes, ' that, 

 basing biological success on numerical abundance of species, the algal eaters are the most successful. 

 This was made abundantly clear during the work at Nkata Bay and elsewhere where algal eaters 

 predominate and where the only really common invertebrate-eater is Labidochromis vellicans'. A 

 parallel too it seems is presented by Calanus finmarchiciis, Lucas (1956^) calling attention to the 



