GRAHAM, CHENOWETH. and DAVIS: LARVAL HERRING 



pling areas may be deduced from the residual 

 surface currents. In winter, the direction of 

 these currents is usually offshore; in spring, the 

 currents are often directed inshore. Thus, lar- 

 vae swept offshore in the winter might be re- 

 turned to the coast in the spring. However, we 

 did not detect any concerted movement by the 

 larvae offshore in the winter and suspect that 

 the assumption that larvae are transported si- 

 milarly to a particle of water is often an oversim- 

 plification. The factors controlling the move- 

 ment and retention of the larvae in shoal water 

 must be investigated to understand the possibil- 

 ities of their transport. 



ANNUAL CHANGES IN ABUNDANCE 

 AND THE FISHERY 



90r 



r 80 



UJ 



o 



K 

 liJ 



a. 



70 



£ 60 



t- 

 (/) 



in 



V) 



< 



-> 50 



o 



IT 



< 



40- 



• 66 



• 68 



1 67 



1 65 



.64 



20 



30 40 50 60 



LARVAL MORTALITY (PERCENT) 



70 



The goal of this research was to choose an 

 estimate of larval abundance or its correlate that 

 could be used to predict the annual recruitment 

 of immature 2-year-old herring to the sardine 

 fishery of the western coast of the Gulf of Maine. 

 Three different types of estimates were chosen: 

 (1) winter mortality, (2) condition in the win- 

 ter, and (3) maximum abundance in the spring. 

 Sampling should continue over a number of 

 years to determine whether the relations between 

 the three measures have substance and whether 

 any one or all of them are pertinent to predicting 

 the abundance of 2-year-old herring. 



A tentative comparison between the percent- 

 age of 2-year-old fish taken in the fishery with 

 the percentage of winter mortality of the cor- 

 responding year class during 1964-68 is shown 

 in Figure 13. Years of low mortality were usu- 

 ally related to subsequent greater percentage of 

 2-year-old fish in the fishery. 



Estimates of winter condition are important 

 because they provide an insight as to the cause 

 of larval mortality. Because winter condition 

 correlates with winter mortality, larval deaths 

 are probably caused by debilitating factors such 

 as disease, starvation, or parasitism. But, lack 

 of agreement during 1969 (Figure 12) would 

 involve other factors as well, such as predation 

 or sudden and transient effects of man's activities 

 within the coastal environment. 



Figure 13. — Comparison between the percentage of 2- 

 year-old herring captured in the Maine sardine fishery 

 and the winter larval mortality for a given year class. 

 Age composition of the fishery was determined by John 

 E. Watson (personal communication). The catch also 

 includes fish from the adjacent Canadian coast. 



Estimates of abundance in the spring are nec- 

 essary as well as winter mortality estimates be- 

 cause the number of larvae surviving until spring 

 theoretically depends upon the initial number of 

 larvae present by the end of autumn. During 

 each of the years in which winter mortality was 

 estimated in this study, the autumnal abundance 

 was reduced to approximately the same level by 

 early winter. During years of very successful 

 hatching, the autumn mortality might not be suf- 

 ficient to reduce the number of larvae to a level 

 common to that of previous years. The sub- 

 sequent spring abundance would then be deter- 

 mined by the initial number of larvae present in 

 early winter as well as the winter mortality. The 

 estimates of spring abundance in the Boothbay 

 area and in the coastal area between Cape Small 

 and Penobscot Bay (Figure 14) did not agree 

 with the estimates obtained with buoyed and 

 anchored nets in the Sheepscot for year classes 

 1964-65 (Figure 12, bottom panel). We do not 

 understand the reason for this difference. 



To date, monitoring of the winter mortality 



319 



