eastern Maine coast, but eggs deposited on the underside of the 

 egg mat develop at the same rate as those at the top. Two 

 implications of these preliminary observations are that eggs 

 spawned during a single night probably hatch during a very short 

 period of time and that most mortality occurs at hatching and/or 

 in the larval stage. 



DISCUSSION 



Spawning at the Little Machias Bay site in September, 1986 

 produced a relatively large egg bed with several characteristics 

 which distinguished it from other fall-spawned egg beds in the 

 Gulf of Maine reported on by other authors. First of all, the 

 egg mat was fairly thick, measuring from 1-3 cm or up to 20-30 

 eggs deep; as a result, egg density and abundance were also high. 

 At other deep-water sites in the Gulf of Maine, egg mats 1-2 cm 

 and 4-5 cm thick have been reported on Georges Bank and Jeffreys 

 Ledge (Caddy and lies, 1973; McCarthy et al . , 1979). However, at 

 most of the spawning sites surveyed in the Jeffreys Ledge area, 

 egg cover was considerably thinner; sometimes the egg layer was 

 only a single egg thick. Messieh et al. (1987) reported maximum 

 thickness of over 5 cm for portions of an egg bed 1 . 1 km 2 in size 

 located in 15-20 m on Fisherman's Bank, Prince Edward Island 

 (Gulf of St. Lawrence) surveyed in September, 1986. Five smaller 

 spawning sites were located in the same area in 1985 in depths of 

 10-25 m; egg deposition reached thicknesses of 20-30 eggs or 4.5 

 cm (Messieh & Rosenthal, 1986). Mean egg abundance estimates 

 were extremely variable, ranging from 4.7 x 10^ to 7.9 x 10°/m 2 

 at the five sites. Maximum abundance estimates at LMB in 1986 

 reached 6.3 x 10 6 /m 2 at one station and averaged 4.4 x 10 6 /m 2 in 

 depths of 31-34 m. McKenzie (1964) reported that Atlantic 

 herring egg carpets in shallow water (11-13 m) off southwest Nova 

 Scotia reached 3.25 cm. Egg density estimates at the LMB site 

 reached maximum values in excess of 400 eggs/cc. Bowers (1969) 

 reported values of only 30-165 eggs/cc in egg layers up to 9 

 layers deep at deep-water spawning sites near the Isle of Man in 

 the English Channel. 



Secondly, despite a fairly dense concentration of eggs at 

 the LMB site, egg mortality was extremely low and there was no 

 evidence of significantly different egg development rates at any 

 position in the egg mass. In contrast, Cooper (personal communi- 

 cation) reported that 50-70% of the eggs in the lower portion of 

 an egg mat 4-5 cm deep near Jeffreys Ledge were non-viable. 

 Other authors have reported high egg mortalities and low hatching 

 success in egg masses reaching 16-20 layers or 5 x 10° eggs/m 2 

 (Galinka, 1971) as well as a reduced hatching rate (Hourston et 

 al., 1984; Taylor, 1971). However, these studies were conducted 

 with Pacific herring spawning in very shallow water. However, 

 egg bed surveys conducted recently in the Gulf of St. Lawrence in 

 10-25 m (see above) also reached extremely high egg mortalities 

 and retarded and abnormal embryonic development. In 1986, 

 mortality rates reached 95% in egg mats which averaged only 2 cm 

 thick due to poor circulation and lack of oxygen. "Mass 



273 



