Farris (ms. "Diet induced variation in 

 the free amino acid complex of Sardinops 

 caerulea" ) has shown experimentally that sar- 

 dine muscle amino acids can be modified by 

 diet. 



Clearly, much more evaluative, critical 

 work must be done before these techniques are 

 ready to apply to the sardine subpopulation 

 problem in the field. 



Other attributes: It is possible that 

 studies which are being made for other reasons 

 may fortuitously prove to be of value in recog- 

 nizing and defining subpopulations of sardines . 

 One such possibility is that the fecundity studies 

 (Clark, 1934), now being extended by MacGreg- 

 or will detect subpopulations with differing 

 fecundity characteristics, if such exist. 



Another possibility is that the fall-spawn- 

 ing sardines, which spawn in waters of above - 

 normal temperatures in Sebastian Viscaino Bay, 

 may prove to belong to a distinct subpopulation . 



HYPOTHESIS OF SARDINE SUBPOPULATIONS 



From the evidence which is available and 

 which has been reviewed above, three general 

 conclusions may be drawn: 



1 . Sardines can and do move about 

 through practically the entire range for which 

 evidence is available; i e . , from Sebastian 

 Viscaino Bay, Lower California, to the Pacific 

 Northwest . (Little or nothing is known about 

 the m ovements of sardines to and from south- 

 ern Lower California and the Gulf of California.) 

 On the other hand, there are pronounced differ- 

 ences in these movements from year to year 

 and also differences in the movements of fish 

 tagged in different localities . 



2. Sardines vary between seasons, local- 

 ities and year -classes with respect to practically 

 all characteristics examined. It is not Known 

 whether these characteristics are genotypic or 

 phenotypic, but there is evidence that most, if 

 not all, of them are phenotypic . 



3. Information about the distribution of 

 sardine spawning shows that there are opportun- 

 ities for at least four space -time separations 



between spawning groups . 



What hypothesis about sardine subpopula- 

 tions can be formulated that is consonant with 

 the available information? At the present state 

 of our knowledge it is possible to advance sev- 

 eral alternative hypotheses that are perhaps 

 equally likely to be correct . Instead of advanc- 

 ing a particular hypothesis and testing it, it is 

 possible to ask a more general question, the 

 answer to which will indicate which among the 

 several possible hypotheses is the correct one. 

 This question is: "What interchange is there, 

 if any, between the several space -time spawn- 

 ing groups?" This question may be posed in the 

 affirmative (complete mixing) or the negative 

 (absence of mixing); the method of testing will 

 be the same in either case. 



METHODS OF TESTING THE HYPOTHESIS 



It is obvious that, regardless of the 

 method used to attack this problem, the crucial 

 test will be the determination of the amount of 

 mixing at spawning time. Complete mixing, 

 lack of mixing or any intermediate condition dur- 

 ing the rest of the year are not pertinent to this 

 problem . Therefore, the critical observations 

 must be made on fish collected on the spawning 

 ground. (Of course, if it turns out that there 

 are subpopulations with distinctive characteris- 

 tics on the several spawning grounds, sampling 

 at other localities will provide information about 

 movements of the subpopulations during the non- 

 spawning season.) 



What methods can be used to attack this 

 problem? One method is to collect samples from 

 the several spawning grounds and to examine 

 these to determine if there are any area -specific 

 morphological or meristic differences. But even 

 if there prove to be such differences, the question 

 of whether these are genotypic or phenotypic re- 

 mains unresolved. This approach, therefore, 

 will not now lead to conclusive results. Infer- 

 ences may be drawn from certain Kinds of 

 information (for example, variation between 

 year-classes produced at a given locality), but 

 conclusive evidence will be provided only by rear- 

 ing individuals from known parents under different 

 sets of controlled conditions. Unfortunately, 

 methods of rearing pelagic fishes are not yet 

 known . 



115 



