IO 



1970 



iij 

 o 



z 

 < 

 a 



z 



O 



< 



B 



4.0- 



3.0- 



2.0- 



< 

 > 



< 



o 



IO~ 



0.0- 



J' • '0' ' |j" 'A 1 ' 'j 1  b 1 ' |J ' 'a 1 ' 'J 1 ' 'o 1 ' p" 'a' ' 'j 



1974 



1975 



1976 



1977 



Figure 2.— Abundance ofMytilus edulis larvae (solid line) and water temperature (broken line) at station D7: A) 



1969-73; B) 1974-77. 



at temperatures of 10°-13°C or higher while few 

 studies have reported spawning at lower tempera- 

 tures (Table 2), which also suggests a thermal 

 threshold for spawning. The significance of this 

 threshold may be linked to gametogenesis. Bayne 

 (1965) found that mussels with fully developed 

 gametes would not spawn when held at 5°C under 

 high food concentrations. However, if temperatures 

 were raised to 12°-14°C, gametes matured and 

 spawning ensued. Similarly, Sastry (1968) found 

 that in the bay scallop, Aequipecten irradians, 

 oogonia and spermatozoa formed at 15 °C and 20 °C, 

 but that temperatures higher than 20 °C were 

 necessary for oocytes to reach a fertilizable stage. 

 Therefore, the apparent correlation between a par- 

 ticular temperature and the initiation of spawning 

 may actually reflect the maturation of gametes 

 followed by induction of spawning by any of a num- 



ber of stimuli. Given the predictable rise in temper- 

 ature each spring, this may explain the initiation of 

 spawning at approximately the same time each year. 

 Use of degree days to predict the time of spawn- 

 ing does not appear to be useful. This is due to a 

 very regular pattern of rising and falling water 

 temperatures each year. As a result, the sum of 

 degree days between spawning periods conveyed no 

 more information than did elapsed time. Newell et 

 al. (1982) arrived at a similar conclusion for mussel 

 populations in Long Island Sound. They found that 

 one Long Island Sound mussel population spawned 

 3 mo later than another, despite nearly identical 

 temperature conditions, difference in degree days 

 due solely to a difference in elapsed calendar days. 

 Bayne (1975), however, did find a relationship be- 

 tween rate of gametogenesis and degree days, but 

 not calendar days. 



998 



