LIMITING FACTORS 185 



obtainable, therefore, it would hardly be possible to calculate the minimum production 

 for the South Georgia area from the consumption of phosphate. 



The possibility that silica may be a limiting factor cannot be dismissed, for diatoms 

 are entirely dominant in the phytoplankton of Antarctic surface waters. Analyses for 

 silica content were not made during the period studied, but have been undertaken as 

 part of the programme of work that is being carried out by Mr Deacon in the far south 

 at the time of writing. It is extremely improbable that this factor would be limiting 

 except in locaHzed areas under special conditions, as Cooper (1933, Pt. 11, p. 752) has 

 recently shown that of all the essential salts this is probably the most rapidly regenerated. 

 For all that, the absence of the usual abundant diatom flora, and the dominance of the 

 small thin-walled spineless chain form of Corethron valdiviae on the exceptionally warm 

 summer survey round South Georgia leads one to think that this may sometimes be 

 a factor in Hmiting production. Some pertinent remarks by Cooper {loc. cit., p. 697) 

 show how this may have come about. He quotes Miss Stanbury's (1931) observations 

 that in diatom cultures the skeletons of the dead frustules disappear very rapidly, 

 owing to the highly alkaline media attacking the silica, and also refers to the recent 

 observations of Bachrach and Lefevre (1928, 1929) that a similar phenomenon may 

 occur in living diatoms, the skeletons of several littoral genera degenerating in cultures 

 until the cells, though still living, became scarcely recognizable. Unfortunately these 

 workers did not discuss the pR of their media. Cooper goes on to give an actual 

 instance in which an observed fall in silicate content was correlated with the succession 

 of thick-walled by thin-walled diatoms in the plankton. The analogy of the unusual 

 phytoplankton observed off South Georgia on the occasion in question is very striking, 

 as on that survey the/)H values were high, averaging 8-12 as against an average of 8-o8 

 on the spring survey when diatoms were abundant. I consider, therefore, that it is at 

 least possible that temporary reduction of silicate (complete exhaustion is not necessary, 

 vide Cooper), in conjunction with a high pR, may occasionally limit phytoplankton 

 production in the more northerly part of our area, and that in reaction to this some of 

 the usual oceanic species may disappear, while others loose their spines and assume 

 a thin-walled form of growth. The increased abundance of dinoflagellates in these 

 conditions is according to expectation. 



The silicate requirements of diatoms in the southern ocean raise several points of 

 interest. Castracane (1886, p. 15) doubted the continuity of the broad belt over which 

 the bottom deposits consist mainly of diatom ooze. This apparently encircles the globe 

 some distance to the north of the Antarctic Circle, and its probable existence was first 

 recognized by Murray (1876). The numerous bottom samples obtained by recent ex- 

 peditions leave little doubt that it is practically continuous, and is the true homologue 

 of the similar tract in the North Pacific. The action of the warm Atlantic current flowing 

 past Iceland, the greater amount of terrigenous debris, and the comparatively shoal 

 water are probably responsible for the absence of a similar belt in the North Atlantic. 

 It would seem that in the Southern Ocean the northern edge of this belt coincides 

 roughly with the Antarctic convergence, or the extreme northern limit of pack-ice ; its 



Z4 



