McQuinn: Year-class twinning in sympatric spawning populations of Clupea harengus 
127 
son (Einarsson, 1951; Postuma and Zijlstra, 1958; 
Messieh, 1972). The comparison of hatching season with 
spawning season as determined by these two methods, 
respectively, thus provides us with a rare opportunity 
to study the reproductive interactions between sympa- 
tric seasonal-spawning herring populations. 
As with most herring populations, herring from 
western Newfoundland (Canada) are characterized 
by the periodic appearance of very large year classes 
followed by several years of relatively poor recruit- 
ment. In addition, the sympatric seasonal-spawning 
populations in eastern Canadian waters often show 
year-class twinning (Winters et ah, 1986; de 
Lafontaine et al., 1991). This phenomenon is most 
evident when a strong autumn-spawning year class 
of a given year coincides with a strong spring-spawn- 
ing year class of the following year. Twinning is rela- 
tively common with seasonal-spawning populations 
but does not always occur. It is also true that year- 
class twinning rarely occurs between successive 
spring- and autumn-spawning year classes of the 
same year. 
Winters et al. (1986) showed a weak, though sig- 
nificant, relationship between year-class strength 
(natural log scale) of autumn-spawning herring in 
eastern Newfoundland and that of spring spawners 
of the following year. An example where year-class 
twinning occurred in the western Newfoundland 
herring populations is with the 1979 autumn-spawn- 
ing and 1980 spring-spawning year classes, both of 
which were very large (McQuinn and Lefebvre 4 ). 
However, the 1982 spring-spawning year class was 
also very large but had no large autumn-spawning 
twin in 1981. Again we observed twinning with the 
1986 and 1987 autumn- and spring-spawning year 
classes. What then are the characteristics that dis- 
tinguish these year classes and that might explain 
why twinning occurred in 1979-80 and 1986-87, but 
not in 1981-82? 
Year-class twinning was first reported in herring 
by Einarsson (1952), who termed it “year-class 
strength parallelism.” He speculated that favorable 
oceanographic and feeding conditions occurring from 
the fall of one year until the following summer re- 
sulted in a parallelism in larval survival between the 
two spawning populations. However, an alternative 
explanation is that year-class twinning is simply a 
consequence of straying between sympatric spring- 
and autumn-spawning populations, i.e. significant 
numbers of individuals from a large cohort of one 
4 McQuinn, I. H., and L. Lefebvre. 1994. An assessment of the 
west coast of Newfoundland (NAFO division 4R) herring re- 
source up to 1993. Department of Fisheries and Oceans (DFO) 
Res. Doc. 94/43, 48 p. Atlantic Stock Assessment Secretariat, 
RO. Box 1006, Dartmouth, Nova Scotia, Canada B2Y 4A2. 
seasonal-spawning population subsequently spawn 
in the other season, creating a strong year class in 
both populations. Jean 5 and Graham (1962) sug- 
gested that the determination of spawning season of 
sympatric herring populations may be influenced by 
juvenile growth rates. Winters et al. (1986) presented 
data in support of their hypothesis that faster-grow- 
ing spring-spawned juveniles may become autumn- 
spawners and conversely that slow-growing autumn- 
spawned juveniles may spawn in the spring. The 
objective of the present study is to use the otolith 
characteristics and maturity stage methods to estab- 
lish whether indeed juvenile growth rates (as repre- 
sented by size at age) have an effect on the determi- 
nation of the onset of first maturation and thus the 
establishment of spawning season in Atlantic herring. 
Materials and methods 
Data for this study were collected from the west coast 
of Newfoundland herring fishery from 1982 to 1990 
(after 1990 otolith characteristics of mature herring 
were no longer determined by our agers). Samples 
were frozen and shipped to the Fisheries and Oceans 
laboratories for detailed analyses. Basic biological 
data (total length, total weight, gonad weight, and 
otolith characteristics, as well as the number of win- 
ter rings from which age was determined) were re- 
corded for all specimens. 
The hatching season was ascertained for each fish 
from otolith characteristics by applying the standard 
criteria (the size and type — opaque or hyaline — of the 
nucleus) developed by the Canadian Atlantic Fish- 
eries Scientific Advisory Committee as described by 
Cleary et al. 6 These criteria were developed from the 
rationale that rapid growth in the first summer of 
spring-spawned herring results in a small opaque 
otolith nucleus. Conversely, the slow first-winter 
growth of autumn-spawned herring results in a large 
hyaline otolith center and the first-winter ring is 
formed only in their second year (Jakobsson et al., 
1969; Postuma, 1974). Although the assignment of 
hatching season is determined subjectively from 
these criteria, consistency between agers has been 
shown to be relatively high. A comparative study was 
5 Jean, Y. 1956. A study of the spring and fall spawning her- 
ring Clupea harengus L. at Grande-Riviere, Bay of Chaleur, 
Quebec. Contribution 49 of the Department of Fisheries, 
Quebec, Quebec, Canada, 76 p. 
6 Cleary, L., J. J. Hunt, J. Moores, and D. Tremblay. 1982. 
Herring aging workshop; St. John’s, Newfoundland, March 
1982. Canadian Atlantic Fisheries Scientific Advisory Com- 
mittee (CAFSAC) Res. Doc. 82/41, 10 p. CAFSAC, Department 
of Fisheries and Oceans, P.O. Box 1006, Dartmouth, Nova Scotia, 
Canada B2Y 4A2. 
