TOWNSEND and GRAHAM: GROWTH AND AGE STRUCTURE OF LARVAL HERRING 



N = 210 



E 

 E 



X 



h- 

 o 



-z. 



LU 



35 

 30 

 25 

 20 



15 

 10 



Y=I0.I6+0.28X 



•jfi^**» • Y=II.29+0.2IX 



^l^/T'^*- 



n40 



35 -- 



iG 



30 E 



E 



X 



25 ^ 



'Z. 



UJ 



_l 



20 



15 



10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 



NUMBER OF OTOLITH GROWTH INCREMENTS 



Figure 7. — Growth of the second group of Atlantic herring larvae which entered the Sheepscot River estuary, 

 1978-79. This includes all larvae from the towed net samples in which the first otolith daily growth increment was 

 formed on or after 30 October. The plotted symbols indicate the collection date and the numbers in parentheses 

 indicate sample size. Regression lines were calculated for the samples collected before and after the winter period 

 of interrupted growth and the 30 January samples were therefore not included. 



mates reported by others were perhaps developed 

 from samples containing more than 1 age-group. 

 Such combinations are difficult to distinguish in 

 length-frequency data as pointed out by Das (1968, 

 1972) and Graham et al. (1972). Graham (in press) 

 reported that from one to as many as four broods of 

 larvae entered the Sheepscot estuary annually in 

 recent years. 



The winter retardation of larval growth rate 

 observed in the Sheepscot estuary (Figures 6, 7) 

 was more brief than the general slowing down of 

 growth throughout the winter reported by others 

 (Tibbo et al. 1958; Das 1968, 1972; Graham et al. 

 1972; Boyar et al. 1973). The duration of slowed 

 growth lasted only 2 or 3 wk, from the latter half of 

 January to early February, and occurred approxi- 

 mately when the second group of larvae was abun- 

 dant in the estuary, but the exact cause of the 

 retarded growth is not clear. Midwinter in general 



has been shown to be a period of stress for larval 

 herring. Chenoweth (1970) showed that the rela- 

 tive condition of Atlantic herring larvae was 

 poorest in February 1965, 1966, and 1967 and in 

 January 1968 and that these periods of low condi- 

 tion factors coincided with the high mortalities 

 reported by Graham and Davis (1971). Midwinter 

 is a time when food densities are lowest (Sherman 

 and Honey'), when water temperatures approach 

 the lethal limit (Graham and Davis 1971) and 

 when the feeding activity of the larvae is lowest 

 (Sherman and Honey 1971). Any or all of these 

 factors may have contributed to the period of re- 

 tarded growth of larvae in our study. 



In conclusion, it appears that the ages of larval 

 herring determined by the otolith growth incre- 



^Sherman, K., and K. Honey. 1970. Seasonal succession of 

 the food of larval herring in a coastal nursery area. ICNAF 

 Res. Doc. 70/72. 



129 



