TOWNSEND and GRAHAM: GROWTH AND AGE STRUCTURE OF LARVAL HERRING 



mately 10 October for the first group of larvae and 

 23 November for the second. This analysis indi- 

 cates, along with the length-frequency data (Fig- 

 ure 2) and the larval catch rates (Figure 3), that 

 there were at least two peaks in spawning effort 

 along the Maine coast, separated by about 6 wk, 

 and that the two groups of larvae entered the 

 estuary at separate times. The times of hatching 

 for each group of larvae can be approximated by 

 assuming that the first otolith daily growth incre- 

 ment is formed at the time of yolk-sac absorption 

 (Rosenberg and Lough footnote 4) and allowing 

 about 5 d for yolk-sac absorption at 10° C (Blaxter 

 and Hempel 1966). The two broods of larvae which 

 entered the estuary in November and December 

 were probably hatched, therefore, in early October 

 and mid-November, and were probably spawned 

 in late September and early November. 



Larval Growth Rates 



Growth of the larvae was examined separately 

 for each of the two major age-groups which en- 

 tered the estuary. The first group included those 

 in which the first otolith daily growth increment 

 was laid down before 30 October, which is the 

 dividing point between the two age distributions 

 discussed above. The second group of larvae in- 

 cluded those in which the first daily growth incre- 

 ment was laid down on or after 30 October. 



Both age-groups of larvae experienced approxi- 

 mately a 2-3 wk period of retarded growth. The 

 changes in growi;h rate appeared as breaks in the 

 plotted data in Figures 6 and 7. Figure 2 showed 

 also that modal lengths increased only slightly, if 

 at all, from 29 January to 27 February. The first 

 major group of larvae to enter the estuary showed 

 retarded growth (Figure 6) beginning at a length 

 of about 35 mm and about 80-100 d after 10 Octo- 

 ber, the mean date on which the first otolith daily 

 growth ring was formed (Figure 5). Thus, this 

 period of retarded growrth began during the latter 

 half of January and continued until early Febru- 

 ary. The second major group to enter the estuary 

 showed retarded grovvi;h (Figure 7) beginning at a 

 length of about 26 mm, and 50-60 d after 23 No- 

 vember, the mean date of the first otolith daily 

 growth ring for group 2 (Figure 5). This period of 

 retarded growth also began during the latter half 

 of January and continued until early February. It 

 appears, then, that these two groups of larvae, 

 which differed in age by about 6 wk and in length 

 by about 9 mm, experienced similar reductions in 



their growth rates during the same period in late 

 January and early February. Apparently the en- 

 vironment at this time was not conducive to their 

 growth. 



Assuming that growth was interrupted during 

 late January and early February for each of the 2 

 age-groups of larvae, regression lines were calcu- 

 lated for those larvae caught before the interrup- 

 tion in grovvi:h and for those caught after. The 

 larvae caught 30 January were therefore not in- 

 cluded (Figures 6, 7). The slopes and elevations of 

 the two regression lines for each age-group were 

 compared using the ^-test described by Zar (1974: 

 228-230). There was no significant difference be- 

 tween slopes for group 1 but the elevations differed 

 significantly (P<0.01). The two regression lines 

 for group 2 differed significantly in slope (P < 0.05) 

 and in elevation (P<0.01). Group 1 larvae, then, 

 grew about 2.1 mm/wk before and after the inter- 

 rupted grovvi;h period. Group 2 larvae grew ap- 

 proximately 2.0 mm/wk before this period and 

 about 1.5 mm/wk after. 



DISCUSSION 



Previous workers have reported polymodal 

 length-frequency distributions of Atlantic herring 

 larvae in the Gulf of Maine-Bay of Fundy areas 

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

 Boyar et al. 1973). Graham et al. (1972) detected 

 two broods of Atlantic herring larvae during Sep- 

 tember 1964 in the Boothbay area of the western 

 Gulf of Maine, which includes the Sheepscot River 

 estuary. The two broods were indicated by length- 

 frequency modes of 9 and 13 mm. They reported 

 that in 1965 only a single brood was detected in the 

 area initially and that a second group of smaller 

 larvae appeared in November. They suggested 

 that the variations in lengths of the larvae might 

 be attributed to the location of the Boothbay area 

 within a coastal zone of transition in hatching 

 times. Atlantic herring larvae hatch earlier in the 

 eastern coastal Gulf of Maine than in the west, 

 and may be carried westward and into the Booth- 

 bay area by coastal currents (Graham 1970; Gra- 

 ham et al. 1972). This may explain the variation in 

 modal sizes on 19 October when the buoyed and 

 anchored nets (Figure 2) captured larvae recently 

 hatched (<10 mm) and others obviously older. The 

 two groups of larvae captured in November and 

 December (Figures 2, 4) perhaps also drifted along 

 the coast before entering the estuary, since each 

 group was about 4 wk old when first sampled. 



127 



