(215 tons). From 1948 to 1952 it averaged ap- 

 proximately 238 tons, while in 1953 (434 tons) 

 it was nearly twice that of 1948 (237 tons). 



Total fishing effort, expressed in terms 

 of boat-weeks, may be estimated once the catch 

 per unit effort index has been determined, by 

 dividing the total catch by the average catch 

 per boat -week. It may be seen from table 2 

 and figure 6 that fishing effort increased over 

 the period, reaching a peak in 1953. 



Discussion 



The method employed in computing the 

 catch per unit index does not completely elim- 

 inate effects resulting from factors other than 

 changes in availability of fish to the fishermen. 

 Doubtless there has been some variation in the 

 experience and efficiency of vessel crews, due 

 chiefly to the addition of new fishermen over 

 the period of expansion. The use of echo sound- 

 ers, radar, scouting planes, radio telephones, 

 and increased capacities of plants also effect 

 the apparent availability. Furthermore, the 

 amount of time that an individual vessel spends 

 in actual fishing may vary from year to year. 

 Although such factors as weather, vessel and 

 plant breakdowns, net repairs, etc., should 

 average out over a period of years, more im- 

 portant is the reduction in actual fishing time 

 during seasons when menhaden are relatively 

 more abundant. The latter would tend to influ- 

 ence the calculated availability index to a great- 

 er extent during years of relatively high 

 abundance than during years of relatively low 

 abundance . 



A point of further concern, possibly, in- 

 volves the use of arithmetical means, rather 

 than geometrical means, in computing the index 

 of abundance, since the latter is the best meas- 

 ure for averaging rates of change. Frequency 

 distributions of relatives are often skewed: if 

 the logarithms of the variate are computed, the 

 resulting frequency distributions usually prove 

 to be of the symmetric type. Various fishery 

 workers (Clark, 1939; Rounsefell, 1931) also 

 have observed that frequency distributions of 

 catches often exhibit a fairly high degree of 

 positive skewness. In the present study it was 

 foun d that when vessel catches were plotted for 

 each year, the resulting curves approximated 



normality, 

 applied. 



No system of weighting was 



Mention should also be made of the re- 

 duction in the number of fishing days for those 

 weeks in which the fishermen are paid. Al- 

 though each boat -week in the defined season 

 contains six possible fishing days (excluding 

 Sundays), fishing does not take place on the 

 Saturday ending each two week pay period. Thus, 

 some boat -weeks contain only 5 fishing days. 

 Short weeks, however, are more or less random 

 occurrences from season to season, and their 

 effect on the calculated catch per unit index 

 should average out about the same over the 

 period of years considered. 



In summary, the sources of error pointed 

 out above doubtless affect the accuracy of the 

 calculated index and must be considered in inter- 

 preting changes and trends in abundance. The 

 calculated indices provide a rough measure of 

 broad changes occurring in the fishery, and with 

 certain limitations may be considered to be pro- 

 portional to the abundance of fish on the grounds 

 during the fishing season. 



OTTER TRAWL FISHERY 



The otter trawl was introduced into the 

 waters of the western Atlantic in 1905 . Although 

 its use quickly spread in the New England fish- 

 eries, it was not generally adopted in the middle 

 Atlantic region unti 1 after the close of World 

 War I. In 1921, only 9 otter trawlers operated 

 out of New Jersey ports; by 1950 their number 

 had increased to 92 . In 1939, a resident trawl 

 fleet was established in Ocean City, Maryland. 

 Itinerant trawlers, however, had landed catches 

 at this port following the opening of the inlet in 

 1935. Only one trawler operated in Delaware . 

 prior to 1946. At present, resident vessels 

 operate from ports located in southern New Jersey, 

 Delaware, and Maryland. In addition, a fleet of 

 offshore trawlers from southern New England and 

 Virginia, numbering as many as 40 vessels in 

 certain years, fishes in these waters. Except 

 for occasional landings, however, their catches 

 are landed at home ports . In 1953, otter trawl 

 landings within the survey area exceeded 20 million 

 pounds of food fish with an estimated value of 

 about $2.0 million to the fishermen. 



12 



