NOTE Lazzari et a!.: Physical environment and recruitment variability of Clupea harengus 
381 
bulence are two processes hypothesized to affect the 
survival of marine fish larvae (Lasker, 1975; Norcross 
and Shaw, 1984), but strong evidence linking larval 
survival to wind conditions remains inconclusive. 
Wind-related transport is believed to influence the 
recruitment of many species of marine invertebrates 
(Roughgarden et ah, 1988; Farrell et al., 1991) and 
fishes (Bailey, 1981; Heath, 1989). 
For Atlantic herring in the Gulf of Maine, the east- 
ern Maine-Grand Manan Island spawning ground 
presents a unique case in how southwesterly winds 
enhance larval transport and survival. Bigelow 
(1927) found that winds from the southwest tend to 
“build up” surface waters in the Bay of Fundy caus- 
ing an “overflow” in the shape of a westerly drift that 
increases the flow of the coastal current along the 
eastern Maine coast, i.e. against the prevailing winds. 
Herring larvae depend on these currents for dispersal 
to more productive nursery areas (Graham, 1982; 
Graham and Townsend, 1985; Townsend et ah, 1986, 
1987) because the area of extensive tidal activity in 
the Bay of Fundy and off eastern Maine (Garrett et 
al., 1978) leads to pronounced vertical mixing and 
less stratification of the water column off eastern 
coastal Maine (Yentsch and Garfield, 1981). As a re- 
sult, primary production is much lower in the north- 
eastern Gulf of Maine because the mixed layer ex- 
tends deeper than the critical depth for plankton 
production (Townsend et al., 1987). Zooplankton prey 
organisms that support larval growth and survival 
are extremely rare on this spawning ground in the 
fall, only reaching adequate densities about 100 km 
“downstream” from the spawning ground (Townsend 
et al., 1986, 1987); therefore, increased dispersal of 
recently hatched larvae that originate on the east- 
ern Maine-Grand Manan Island spawning ground 
is a mechanism that could enhance the recruitment 
of juveniles to the coastal herring stock. 
This study generally supports the Campbell and 
Graham (1991) theory that release of larvae from the 
eastern Maine-Grand Manan spawning ground is 
mediated by wind events that generate horizontal 
flows and can carry larvae out of retention areas into 
the counterclockwise residual flow that moves from 
northeast to southwest along the Maine coast. 
Bigelow and Schroeder (1953) thought that this cir- 
culation is set in motion by wind and freshwater in- 
flow and that it influences the availability of two- 
year-old herring because the fish follow the drifting 
planktonic animals on which they feed. Furthermore, 
Chenoweth et al. ( 1989) observed that larvae hatched 
in this spawning area at the same time had differ- 
ent horizontal displacements away from the spawn- 
ing area; some larvae remained in the area for up to 
a month after spawning, whereas others were trans- 
ported up to 100 km southwestward down the coast 
during the same time period. Townsend (1992) at- 
tributes this variable release of larval herring from 
the eastern Maine retention area to the intrusion of 
slope water into Jordan Basin (a deep offshore basin 
located in the northeastern Gulf of Maine), the tim- 
ing of hatching (to coincide with lunar periodicity and 
the intensity of tidal mixing), and the location of egg 
beds in relation to the front between the area of tidal 
mixing and more stratified water offshore, where the 
geostrophic flow that would pull larvae out of the 
retention area is greatest (Brooks and Townsend, 
1989). In addition, Brooks (1990) found a relation 
between wind stress and currents that suggested the 
action of a density-modulated coastal upwelling 
mechanism in which the deep inward currents over 
Lindenkohl sill respond directly to northeastward 
alongshore wind stress at times of weak stratifica- 
tion, such as occurs in fall. 
Once entrained within the coastal current, the lar- 
vae are dispersed to an overwintering area that has 
not been conclusively determined as yet. Greater 
advection of larvae from the eastern Maine-Grand 
Manan Island spawning area to the southwest as 
hypothesized by Graham ( 1982) would distribute lar- 
vae among more coastal overwintering areas. This 
distribution would improve recruitment success by 
lessening density-dependent mortality within the 
estuarine and nearshore waters shallower than 100 
m that act as a nursery area and would establish a 
carrying capacity for larvae on the Maine coast for a 
given year. Research has shown that larvae from this 
spawning ground reach at least as far south as the 
Sheepscot River in mid-coastal Maine (Graham, 
1982; Graham and Townsend, 1985; Stevenson et al., 
1989). However, recent research shows that larval 
herring overwintering “offshore” may have a higher 
survival rate than those wintering in nearshore wa- 
ters (Townsend et al., 1989; Townsend, 1992) and that 
dispersal associated with southerly winds could en- 
hance offshore transport. In either case, dispersal of 
larvae away from the eastern Maine-Grand Manan 
Island spawning area is critical for good larval sur- 
vival (Graham, 1982; Campbell and Graham, 1991; 
Townsend, 1992). Wind-induced effects on transport 
have been shown to affect the distribution and re- 
cruitment of other marine fishes (Stevenson, 1962; 
Checkley et al., 1988; Fechhelm and Griffiths, 1990; 
Koutsikopoulos et al., 1991; Castillo et al., 1993). 
Therefore, we propose that more southwesterly 
wind conditions in September increased dispersal of 
eastern Maine-Grand Manan Island larvae in 1966, 
1970, 1983, and 1988, setting up an initial situation 
of high larval survival, which, when combined with 
more southerly wind conditions through December 
