RETENTION OF LARVAL HERRING WITHIN THE 

 SHEEPSCOT ESTUARY OF MAINE 



Joseph J. Graham^ 



ABSTRACT 



This paper demonstrates a system of larval movements that retains larval herring within 

 an estuary despite its seaward residual flow. Data are reviewed from a 2^ factorial 

 design using buoyed and anchored nets in a narrow channel of the estuary. Three factors 

 were involved: location, tidal phase, and depth. Catch rates of larvae relative to these 

 factors suggested that the larvae maintained their position in the upper portion of the 

 estuary by (1) occupying the landward net tidal flow near the bottom, (2) moving up- 

 ward upon reaching the limit of their landward penetration (3) then seaward in the net 

 flow near the surface, and (4) finally descending again into the landward net flow. 



Larval herring, Clupea harengus harengus Lin- 

 naeus, move up the Sheepscot estuary of Maine 

 primarily in the autumn and spring. During 

 these seasons the larvae accumulate in the up- 

 per portion of the lower estuary where they are 

 retained despite the residual seaward flow of the 

 estuary. Statistically significant differences in 

 the distribution of larval herring according to 

 depth and location and the interaction of these 

 two factors with tide suggested that the larvae 

 were using tidal flows to migrate (Graham and 

 Davis, 1971). This paper examines the data, 

 upon which the statistical results were based, 

 to demonstrate a system of larval movements 

 that retain the larvae within the estuary. 



MATERIALS AND METHODS 



Buoyed and anchored nets were fished (Fig- 

 ure 1); their construction and validity in sam- 

 pling larval herring is noted briefly in this paper 

 and is available in detail in another paper (Gra- 

 ham and Venno 1968) . Four nets were attached 

 to one line which was buoyed at the surface and 

 anchored to the bottom to sample at predeter- 

 mined depths (0, 10, 15, and 20 m). A flow 

 meter measured the amount of water strained, 



^ National Marine Fisheries Service, Northeast Fish- 

 eries Center, Boothbay Laboratory, W. Boothbay Harbor, 

 ME 04575. 



from its central position in the mouth of each 

 net. The mouth opening of the net was 0.5 m 

 in diameter and the length of the net was 1.9 m. 

 Mesh diameter was 0.75 mm. Filtration effi- 

 ciencies of the net ranged from 80 to 90% at 

 current speeds above 15 cm/sec and 100% with- 

 in speeds from 15 to 3 cm/sec. Below 3 cm/sec 

 the impeller of the flow meter does not revolve. 

 Tests simulating sampling conditions showed 

 that clogging did not change these filtration eflfi- 

 ciencies. 



The nets fished approximately 6 hr on the 

 flood and 6 hr on the semidiurnal ebb tides. They 

 were set at slack water during dusk, retrieved 

 and reset at the end of the tidal stage during 

 slack water. Two lines of nets were set at the 

 landward end of the estuarine channel and two 

 were set at the seaward end (Figure 2). The 

 sampling area within the estuarine channel was 

 9 km long, 275 m wide, and about 20 m deep 

 throughout its length. The distance between 

 sets at the seaward end of the channel was 2 km, 

 that between the landward sets was 3.2 km, and 

 the distance between the two locations was 3.8 

 km as measured from the interior sets. 



For a given line of nets catch rates were 

 determined for the two nets above mid-depth 

 and for the two nets below mid-depth. In each 

 case, the total amount of water strained for the 

 two nets above and below mid-depth was 



Manuscript accepted December 1971. 



FISHERY BULLETIN: VOL. 70, NO. 2. 1972. 



299 



