386 DISCOVERY REPORTS 



Demersal fishes of extensive depth range tend to show larger individuals in deeper zvater, except when 

 this tefidency may temporarily be masked by seasonal movements. A similar tendency can sometimes be 

 discerned interspecifically , but there are many obvious exceptions {such as large, exclusively littoral species) 

 if purely taxonomic units are considered. 



Most of the fish for which we have any adequate data demonstrate the first (intra-specific) part of 

 this proposition very clearly. The second part is also clear when suflicient data are to hand to permit 

 ecological rather than taxonomic groupings to be considered, and is very well shown by the depth 

 relations of Patagonian Nototheniiformes. 



Fishes of wide latitudinal range often show a gradation of size with latitude, the individuals being larger 

 towards the polar limits of the range of the species. This implies either that the stocks are different in different 

 latitudes {with different spazvning times) as has been shown for European hake; or that only the larger 

 individuals penetrate towards the polar limits of the range of the species. 



The first type of correlation between size and latitude seems to be exemplified by our Patagonian 

 hake, which consistently showed larger fishes as one proceeded south, but in the movements of which 

 little meridional trend could be detected. But to the extreme south of the species' range only a very 

 few outsize female individuals were taken. These can hardly be a breeding stock and are thought to 

 be exceptionally aged individuals that may have lost the seasonal urge to migration through reduced 

 reproductive activity. 



Macrurotius also showed the largest individuals to the southward at most seasons, but with this 

 species a considerable meridional trend of movement seems certain, so that the size latitude correlation 

 is probably a transient one, consequent upon the greater speed and strength of the larger fish, which 

 may enable them to proceed farther afield in the search for food. 



In many fishes a marked change in diet occurs zvith increasing size {age). Often this may coincide with 

 the attainment of sexual maturity and jor greatly increased migrations (cf. Harold Thompson, 1943, p- 86, 

 on Nezvfoundland Cod; Hartley, 1945, pp. n, 26, on British fresh-water fish). 



This change of diet is very well shown by the larger species of the shelf, Macruronus and hake. The 

 smaller individuals seem to feed mainly upon macroplanktonic Crustacea, the larger ones upon fish 

 and squids. Of the more definitely bottom-feeding species it is practically certain that Notothenia 

 ramsayi will be found to show an analogous change, from a diet of mysids, benthic Crustacea and 

 polychaetes to one of larger prey — fish, especially Falkland herring. Our data are insufficient to 

 establish the point beyond question. The probable coincidence of change in feeding habits with onset 

 of sexual maturity and increased migration is particularly well shown by Macruronus. 



In a large majority of fishes, females are larger than males; and this is usually sufficient to account for 

 the slight abnormalities of sex-ratio {preponderance of females) commonly encountered in net-caught 

 samples. Marked abnormalities in this or the other directioti are usually due to peculiar seasonal trends 

 towards unisexual shoaling, and are zvell known in certain elasmobranchs. 



Nearly all the Patagonian species for which we have adequate data showed the females to be larger 

 than the males. Usually the difference was small but strongly significant. The single possible exception 

 to this ' rule ' was Cottoperca gobio (Bovichthyidae) where the males seemed to be the larger. More 

 data would probably dispel this apparent anomaly. The Patagonian hake {Merluccius hubbsi) shows 

 a most unusual discrepancy in size between the sexes, the males being very much smaller than the 

 females (much more so than in European hake). With a few species strongly abnormal sex-ratios were 

 encountered that suggest possibilities of unisexual shoaling (Macruridae, Micromesistius australis). 



