FISHERY BULLETIN: VOL. 70, NO. 3 



borderline statistical significance (P ^^ 0.05). 

 If these results are real, it may be inferred from 

 the positive correlations that larger schools were 

 associated with greater school numbers. The 

 negative correlation, on the other hand, suggests 

 that schools tended to be more numerous as their 

 average size decreased. 



The reasons for the apparent discrepancies 

 in the school density-size relationship in the dif- 

 ferent areas are not clear. The lack of any 

 trend in the South Atlantic Area may result from 

 variable availability of the schools. Fishermen 

 and aircraft spotters report frequent disappear- 

 ances of schools from surface coastal waters 

 early in the day and occurrences of fish inside 

 the extensive sounds that border the summer 

 fishing grounds in this area (this behavior pat- 

 tern of the fish may be associated with tidal cur- 

 rents passing through the numerous inlets and 

 their relation to the abundant plankton food sup- 

 ply inside the sounds). A possible explanation 

 for the inverse relationship between school size 

 and number in the Middle Atlantic Area is that 

 fishing captains tend to avoid the smaller fish 

 and smaller schools when fish are abundant. 

 Heavy fishing in the Middle Atlantic Area may 

 also be a contributing factor in keeping the stock 

 broken up into smaller schools. Fishing eflfort 

 in this area averaged from li/^ to 41/2 times 

 greater than in any other area and increased by 

 about 11% over the 8-year period (Nicholson, 

 1971b). There very likely is some level of fish- 

 ing effort that disrupts the normal schooling ha- 

 bits of the fish, but the interrelationships in- 

 volved appear to be much more complex than 

 can be determined from the data in hand. 



The main inference to be drawn from this sec- 

 tion is that size of summer schools is related to 

 stock abundance in each major fishing area. The 

 school density-stock abundance relationship, 

 however, needs to be clarified by analysis of data 

 collected after 1962 when the catch dropped to 

 the lowest levels since the 1930's. 



DISCUSSION 



Knowledge of the nature and consistency of 

 groupings of Atlantic menhaden is fundamental 

 to understanding the dynamics of the population. 



Results of this study demonstrated that the fish- 

 able stock occurs in schools or school-aggrega- 

 tions during most of the year. It is also known 

 that the young fish school from the time of their 

 entry into estuarine nurseries as larvae (June 

 and Chamberlin, 1959; Reintjes and Pacheco, 

 1966). Thus, schooling is one of the. basic be- 

 havior characteristics of this fish. June and 

 Reintjes (1959) provided initial evidence that 

 the fish school by size and to some extent by age 

 and also showed that schools in a given locality 

 more closely resemble each other in composition 

 than schools from diflferent localities. June and 

 Nicholson (1964) and Nicholson (1971a) de- 

 scribed in some detail the increase in average 

 length and age of the fish with increased lat- 

 itude and inferred annual north-south move- 

 ments of the fish from seasonal changes in their 

 length and age distributions. The findings of 

 this paper support the conclusions of the fore- 

 going studies and provide further insight into 

 the variability of the distribution, composition, 

 and size of summer schools and fall school-aggre- 

 gations. But one of the critical unknowns is 

 the extent to which such variability is related 

 to stock density and, ultimately, to fishing. 



Evidence presented here indicates that the 

 size of summer schools is a function of stock 

 abundance. It was also shown that an inverse 

 relationship existed between fish length and size 

 (weight) of summer schools. These findings 

 suggest that there probably is an optimum school 

 size for fish of given length that is most favorable 

 for survival. If so, there must be a level below 

 which the population must not be fished without 

 running the risk of disrupting schooling to the 

 point of inflicting irreparable damage to popu- 

 lation resilience. While these hypotheses are un- 

 provable with present data, the inference is that 

 schooling marine fishes that are sought by sur- 

 face sightings are more vulnerable from unregu- 

 lated fishing than nonschooling species that are 

 not subject to direct observation. Slight changes 

 in the nature and consistency of grouping of At- 

 lantic menhaden ought therefore to be considered 

 in assessing both short- and long-term eflfects 

 of fishing on the population. 



The inverse relationship shown between fish 

 length and the number of individuals within 



710 



