JUNE: ATLANTIC MENHADEN GROUPINGS 



20 — 



10 - 



— 

 10 - 



o — 

 10 - 



— 

 lO- 



— 

 10 - 



— 

 10 - 



o — 



JIG - 



i 10 - 

 — 

 10 - 



o — 



10 - 

 — 



10 - 

 o — 



10 - 

 o — 



10 - 



o — 

 10 - 



— 



10 



Jl 



Ji 



A 



^t^ 



J^ 



A 



hL^ 



J}^ 



-\ — I I — I I 



15 20 25 30 35 



Avv 



A. 



JL 



A. • 



..z^ 



J^ 



Aa. 



1 1 — T — I — T" 



10 15 20 25 30 35 

 Fork Length (cm) 



n — r — 1 — I — r- 



15 20 25 30 35 



Figure 8. — Length-frequency distributions of Atlantic 

 menhaden in samples from 41 single-set purse-seine 

 catches from fall school-aggregations off North Carolina, 

 1955-62. The samples are arranged in order of decreas- 

 ing fish length. 



(Cheek et al., 1970) have confirmed and elabo- 

 rated this hypothesis. 



Briefly, the length-frequency data presented 

 show a strong tendency for Atlantic menhaden 

 to group by length. Summer schools are com- 

 prised of fish that are more nearly of the same 

 length than is found among schools within any 

 given area of the coast, or within the population 

 as a whole, and there is an increase in the aver- 

 age length of fish within schools from south to 

 north. Fall school-aggregations, which are usu- 

 ally fished only off Long Island and North Car- 

 olina, also consist of fish of similar lengths, but 



lengths within aggregations tend to be more 

 variable than within summer schools. 



SIZE OF SUMMER SCHOOLS 

 IN RELATION TO FISH LENGTH 



There was wide variation in the size (weight) 

 of summer schools in relation to the average 

 length of fish within a school. A plot of catch 

 against mean length of fish in the 275 single- 

 school catches (Figure 9) suggests a tendency 

 for larger fish to occur in larger schools. The 

 correlation between school size and mean fish 

 length (r = 0.144) differs significantly from 



(P < 0.05) , and the relationship evidently is 

 linear, since the arithmetic correlation coefficient 

 is higher than that between log school size and 

 log mean fish length {r = 0.095) , or between log 

 school size and mean fish length (r = 0.103). 

 Comparison of the arithmetic correlation coef- 

 ficients for the individual years, however, indi- 

 cated heterogeneity (x^ = 17.14; P < 0.05), so 



1 plotted the data for each year separately and 

 calculated trend lines (Figure 10). The occur- 

 rence of larger fish in larger schools is evident 

 in some years, but in other years the trend was 

 reversed, or there was no apparent relation be- 

 tween the two variables. Unequal representa- 

 tion of various length groups in different years 

 may be responsible for the apparent heterogene- 

 ity of the data, but I conclude that annual 



20 25 



Mean fork Lengtntem} 



Figure 9. — Size of 275 single-school summer purse-seine 

 catches of Atlantic menhaden plotted against the mean 

 length of the fish in catch samples, 1955-62. 



\ 



707 



