FISHERY BULLETIN. VOL. 79, NO. 1 



buoyed and anchored plankton nets. Shaw^ has 

 shown that when used at night there is a little, if 

 any, difference between these gear with regard to 

 catch rates or larval fish avoidance. Only the sam- 

 ples collected by the towed nets were used for 

 otolith analysis. These towed net samples were 

 collected at night on seven occasions from 24 

 October 1978 to 6 March 1979 using aim, 0.75 mm 

 mesh plankton net. One daytime sample (10 Jan- 

 uary) was taken with a 61 cm bongo net with 0.505 

 mm mesh nets on each side. Buoyed and anchored 

 net samples were collected from 5 October 1978 to 

 27 February 1979 as part of the regular larval 

 Atlantic herring monitoring program conducted 

 by the Maine Department of Marine Resources. 

 The buoyed and anchored nets consisted of six 

 lines of nets fished at four stations in the estuarine 

 channel. Each line had four 0.5 m, 0.75 mm mesh 

 nets, with a digital flowmeter mounted in each 

 net. The nets were set at dusk and retrieved at 

 dawn each sampling date, and fished approxi- 

 mately one semidiurnal tidal cycle. The buoyed 

 and anchored net samples were preserved in 5% 

 Formalin^ and length-frequency distributions 

 and catch rates for larval Atlantic herring deter- 

 mined. The characteristics and performance of the 

 buoyed and anchored nets were reported by Gra- 

 ham and Venno (1968), Graham and Davis (1971), 

 and Graham (1972). 



The larvae from the towed samples were not 

 preserved, but were sorted within 2 h of collection, 

 placed in plastic Petri dishes, and frozen fresh at 

 -18° C for future otolith analysis. The samples 

 were later thawed and each fish measured to the 

 nearest 0.5 mm. Figure 1 shows that the frozen 

 larvae shrink an average of about 1-2 mm more 

 than Formalin preserved larvae. The sagittae, or 

 largest otoliths, from both sides of the head were 

 teased onto a microscope slide under a binocular 

 microscope. The otoliths were mounted in Per- 

 mount and covered with a glass coverslip. The 

 numbers of daily growth increments in one of each 

 pair of sagittae were counted under a compound 

 microscope at 1,000 x magnification. The incre- 

 ments were counted twice and their mean number 

 computed. Only those otoliths in which there was 

 a difference between counts of 5% or less were 

 used in the analysis. These data were used to esti- 



LiJ 



o 

 a: 



UJ 



PRESERVED 

 5% Formalin 



N = 233 

 x = 370 



FROZEN 



N=2I9 

 7 = 35.8 



u 

 o 



UJ 

 Ql 



'■^Richard F. Shaw, Ph.D. candidate, Department of Ocean- 

 ography, University of Maine at Orono, Ira C. Darling Center, 

 Walpole, ME 04573, pers. commun. May 1980. 



"Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



10 

 8 

 6 

 4 



2 







12 

 10 



8 

 6 

 4 

 2 

 



LENGTH millimeters (TL ) 



Figure l. — Length-frequency comparison of 5% Formalin pre- 

 served versus frozen Atlantic herring larvae captured with 61 cm 

 bongo nets on 4 April 1980 in the Damariscotta River estuary, 

 Maine. Larvae from the starboard net were preserved in For- 

 malin and larvae from the port net were frozen. Measurements 

 were performed 2 mo later. A modified t-test (Snedecor and 

 Cochran 1967) showed that the two means were significantly 

 different (P< 0.01). 



mate daily grovv^h rates and age composition of 

 the larvae. The daily grovd;h increments in larval 

 Atlantic herring otoliths show up very clearly, 

 and only 17 of the 317 larvae examined had oto- 

 liths with increments too faint to be counted 

 accurately. 



RESULTS 



Larval Age Structure 



Changes in modal lengths of larvae during 

 autumn and winter indicate that groups of lar- 

 vae entered the estuary and subsequently lost 

 their identities through differential mortality and 

 growth, since the larvae are not known to depart 

 the estuary once established there (Graham et al. 

 1972). Length-frequency data from the buoyed and 

 anchored net samples (Figure 2) showed a tri- 

 modal length distribution (range 6-30 mm) for 

 larvae present in the Sheepscot River estuary 

 on 19 October. A large group of smaller larvae 

 entered on 2 November when only a trace of the 



124 



