340 



Fishery Bulletin 9 1(2). 1993 



The ICNAF larval herring program included the 

 entire Gulf of Maine region but concentrated on the 

 Nantucket Shoals/Georges Bank area (Lough & Bolz 

 1979a, Lough et al. 1985). The 33 surveys conducted 

 between 1971 and 1976 were used in our analysis. 

 ICNAF participants deployed the bongo at 50m/min 

 to a maximum depth of 100 m and retrieved it in a 

 smooth oblique profile at lOm/min. Towing speed was 

 3.5 kn at stations spaced at 28-37 km intervals on a 

 standard grid pattern (Fig. la). Additional stations in 

 areas of high larval abundance were sampled after 

 1974. Lough & Bolz (1979b) provided a detailed ac- 

 count of sampling operations and cruises for the ICNAF 

 time-series. 



MARMAP surveys, conducted at monthly to bi- 

 monthly intervals, covered the continental shelf from 

 Cape Hatteras, North Carolina to Cape Sable, Nova 

 Scotia. Cruise activities were described by Sibunka & 

 Silverman (1984, 1989). MARMAP surveys fished the 

 61cm bongo to a maximum depth of 200 m with payout 

 at 50m/min and retrieval at 20m/min. Ship speed var- 

 ied between 1 and 2 kn to maintain a 45° angle in the 

 towing wire. Station intervals ranged from 15 to 45 km 

 (Fig. lb). 



The herring recovery study initiated in 1988 em- 

 ployed MARMAP sampling methods. Cruises were con- 

 ducted at about monthly intervals and occupied 151 

 stations in the Nantucket Shoals/Massachusetts Bay/ 

 Georges Bank study area (Fig. lc). 



Data analysis 



We adjusted the number of larvae caught at each sta- 

 tion in the time-series to reflect day/night/twilight catch 

 differences (see Morse 1989). Larvae were partitioned 

 into size-intervals to depict changes in distribution pat- 

 terns over time. When the partitioned size-intervals of 

 4.0-7.9 mm, 8.0-12.9 mm, 13.0-17.9 mm, and >17.9mm 

 were corrected for shrinkage (see Theilacker 1980), 

 they approximated age-groups of <2 wk, 2-5 wk, 5-8 wk, 

 and >8wk, respectively (Lough et al. 1982). Surveys 

 were grouped by multi-year periods that represented 

 time-intervals between the most apparent changes in 

 spawning patterns during the 20 yr period. Figures 

 showing mean numbers of larvae per 10 m 2 surface 

 area were based on methods recommended for the delta 

 distribution (Pennington 1983). Distributions of lar- 

 vae grouped by age over time are discussed in relation 

 to circulation patterns shown in Fig. 1 d. 



Results 



Changes in distribution and abundance of herring lar- 

 vae during the 20 yr time-series reflected four contrast- 



ing spawning patterns that spanned successive multi- 

 year intervals. Larvae were abundant on both Nan- 

 tucket Shoals and Georges Bank, but not so in Massa- 

 chusetts Bay, during 1971-75. Conversely, we found 

 low to no measurable numbers of larvae on Nantucket 

 Shoals and Georges Bank during 1976-84, a period 

 when spawning activity increased in Massachusetts 

 Bay. Spawning resumed on Nantucket Shoals in 1985. 

 During 1985-87, larval abundance peaked in Massa- 

 chusetts Bay but the center of spawning activity shifted 

 to Nantucket Shoals. 



Herring were again spawning on the western half of 

 Georges Bank in 1988. The dramatic increase in spawn- 

 ing activity on Nantucket Shoals and renewed spawn- 

 ing on western Georges Bank during 1988-90 elevated 

 larval abundance estimates to their highest levels dur- 

 ing the closing years of the time-series (Fig. 2). 



When the time-series began in 1971, Georges Bank 

 had been the principal spawning grounds for herring 



INT. 1 



INT. 2 



INT. 3 INT 4 



CO 

 3 



j- c\j co *r in <o 





co a> o -— ryco^mcof^-coojo 

 .-^^cocpcococooococococoo) 

 aia>o>oioo)OiO)0)ooCT)a>o)cj>o> 



YEAR 

 Figure 2 



Changes in abundance of Atlantic herring Clupea haivni;us 

 larvae by subarea during the 20yr time-series. Intervals i Int i 

 represent multi-year periods that reflect the most apparent 

 changes in spawning patterns. 



