GRAHAM, CHENOWETH, and DAVIS: LARVAL HERRING 



^ 



larvae appeared to grow about 1 mm every 5 

 days. Similarly, in 1964 modal lengths may be 

 traced from late September to early November 

 with an apparent larval growth of 1.3 mm every 

 5 days, but then the mode became difficult to 

 identify. 



DISCUSSION 



SEASONAL CHANGES IN ABUNDANCE 



AND LARVAL LENGTH CAUSED BY 



MORTALITY 



After larvae hatched and accumulated in the 

 coastal bays and estuaries, their abundance was 

 determined by the rate at which they died. The 

 autumn mortality was especially severe, for the 

 catch declined in the Boothbay area (Figure 6) 

 despite the addition of successive broods to the 

 area throughout the autumn. This high mor- 

 tality was also indicated by the failure of length 

 modes to persist into the winter. The number of 

 larvae in a given mode apparently was reduced 

 with time and coincidently with differential 

 growth until the larvae were too sparse in the 

 catches to form a distinguishable mode. In 

 1966 the disappearance of a mode of relatively 

 large fish from the catch caused a sharp drop 

 in the mean larval length during November 

 (Figure 7). Although the mean length in- 

 creased either with subsequent growth of the re- 

 maining larvae or with the addition of larvae to 

 the area, it remained below the mean lengths 

 of the other years throughout the winter. Win- 

 ter mortalities were not as high as those in the 

 autumn. Graham and Davis (1971) determined 

 mortalities from December to January for larvae 

 captured in the Sheepscot estuary during 1964- 

 67 and recently for mortalities for 1968-69. Es- 

 timates for the 6 years varied from 22% to 52% 

 for 15-day intervals and appeared statistically 

 reliable with the largest spread in the 0.95 fidu- 

 cial interval in 1968 (27.7-36.9%) , and the smal- 

 lest in 1966 (22.0-22.8%). 



Extensive reductions in our catch rates in the 

 late spring occurred from avoidance of the gear 

 by the larvae and their departure from our sam- 

 pling stations. Catches at the upper estuarine 



stations in the Boothbay area were larger than 

 those at the lower and outer stations. Progres- 

 sive peaks in abundance, that were present in 

 the autumn, were absent from outer to upper 

 estuarine stations in the spring. One explana- 

 tion for the lack of progression is that the larger 

 larvae moved landward so rapidly as to be un- 

 detected in the spring. Another and more likely 

 explanation is that their mortality was sufficient- 

 ly low to permit an accumulation of larvae at the 

 landward extremity of their movement where 

 numbers always greatly exceeded the number 

 of larvae moving into the area. 



Estimates of annual mortality were based on 

 winter mortality (Graham and Davis, 1971) be- 

 cause measurements of total mortality during a 

 given year were impracticable as they were in- 

 fluenced by larval movement in the autumn and 

 in the spring. For most year classes, a higher 

 winter mortality recorded in the upper end of 

 the Sheepscot estuary coincided with a smaller 

 maximum catch there in the subsequent spring. 

 Also, higher winter mortalities usually coincided 

 with a poorer condition or well-being of the lar- 

 vae (Chenoweth, 1970) for the Boothbay area 

 (Figure 12). 



The causes of larval mortality along the west- 

 ern coast of the Gulf of Maine were not deter- 

 mined, but inferences were made by Chenoweth 

 (1970) and by Sherman and Honey (1971). 

 Essentially, they suggested that winter mortal- 

 ity might be related to lower lethal temperatures, 

 inhibition of feeding by low temperatures, and a 

 scarcity of food. Sherman (personal communi- 

 cation) found in recent studies that the larval 

 guts were frequently occluded by parasites, 

 which may cause death. 



SEASONAL CHANGES IN DISTRIBUTION 

 RELATED TO LARVAL SOURCES 



After hatching, the larvae shifted their dis- 

 tribution from spawning areas to the coastal 

 bays and estuaries. These destinations were 

 apparent from our catches, but not all the sources 

 or spawning areas were determined. The sources 

 of larval herring in the Gulf of Maine, including 

 the vvestern coast, their movements from these 



317 



