THE BEHAVIOUR AND PHYSIOLOOY OF HERRING AND OTHER OLUPEIDS 281
herring larvae caught at sea in different a r e a and found in some
instances very low values indicating the larvae were near starvation.
b k e r (1962) quoted reports of Sardinops caerulea starving in the sea.
The general subject of critical periods has been reviewed by Marr
(1956), with new data on Sardinops caerulea,, and he concluded that this
concept shoilld be treated cautiously. Farris (1960) took this study
further and calculated the effect of different types of growth curves on
estimates of larval survival. Taking data on the abundance of larvae
of Sardinops caerulea, Sardinops naelanostida and Engraulis japonica
caught at sea, he concluded that high mortality took place before the
final absorption of the yolk, assuming that net selection, which might
allow the escape of small larvae, was unimportant. This observation
was supported by laboratory cxperiments on rearing Sccrdinops caerulea
when a high mortality took place before absorption of the yolk. While
such a mortality in the sea could be explained by predation, this was
not so in tanks and Farris suggested that a high proportion of eggs
might be imperfect, both in the sea and in tanks. This would be rather
surprising on general biological grounds. Murphy (1961) disccssed the
critical period concept for Sardiwps caerulea, suggesting how both poor
food supply and low temperatures might retard development and
prolong this period.
Bridger (1960,1961) considered the fluctuations in the abundance of
herring larvae in the southern North Sea and English Channel and
concluded that the abundance dcpendcd on the number of eggs spawned
by the older fish rather than on the total stock. This might be due to
the older fish spawning later, at a timc of better food supply for the
larvae, or it might be due to intrinsic diffcrc?nce8 in the yolk supply,
the older fish producing larger c g y ~ with rrtorc yolk. A nirriilru vir!wpoint has also been put forward ~ J Y Marty ( 1 959). Again o r ) h t h q i c d
grounds it seems somewhat unlikely that large Rcctionn of a #pawning
population should bo producing inviable larvae.
Thc irnportancs of cgg H i m on eurvival of larvao has been studied
by Blaxter and Hempel (1!163). Thoy tihowcd that within a race
larger eggs tended to produce larger larvae on hatching an wcll a8
larvae that lived longer on their yolk, Whether larger larvae arc' morc!
viable as a result of a larger mouth or greater activity remains to IJC
tested. Certainly larger larvae take larger food organisms according to
samples caught at sea (see Section IV, B). Between races the yolk
supply had a very strong effect on survival, larvae living for from
12-25 days at 8°C for instance, depending on their race. Nikitinskaya
( 1958b) found that the larvae of Clupea pallasii reared at temperatures
above normal were smaller, less well-developed and less active, as
herring larvae caught at sea in different a r e a and found in some
instances very low values indicating the larvae were near starvation.
b k e r (1962) quoted reports of Sardinops caerulea starving in the sea.
The general subject of critical periods has been reviewed by Marr
(1956), with new data on Sardinops caerulea,, and he concluded that this
concept shoilld be treated cautiously. Farris (1960) took this study
further and calculated the effect of different types of growth curves on
estimates of larval survival. Taking data on the abundance of larvae
of Sardinops caerulea, Sardinops naelanostida and Engraulis japonica
caught at sea, he concluded that high mortality took place before the
final absorption of the yolk, assuming that net selection, which might
allow the escape of small larvae, was unimportant. This observation
was supported by laboratory cxperiments on rearing Sccrdinops caerulea
when a high mortality took place before absorption of the yolk. While
such a mortality in the sea could be explained by predation, this was
not so in tanks and Farris suggested that a high proportion of eggs
might be imperfect, both in the sea and in tanks. This would be rather
surprising on general biological grounds. Murphy (1961) disccssed the
critical period concept for Sardiwps caerulea, suggesting how both poor
food supply and low temperatures might retard development and
prolong this period.
Bridger (1960,1961) considered the fluctuations in the abundance of
herring larvae in the southern North Sea and English Channel and
concluded that the abundance dcpendcd on the number of eggs spawned
by the older fish rather than on the total stock. This might be due to
the older fish spawning later, at a timc of better food supply for the
larvae, or it might be due to intrinsic diffcrc?nce8 in the yolk supply,
the older fish producing larger c g y ~ with rrtorc yolk. A nirriilru vir!wpoint has also been put forward ~ J Y Marty ( 1 959). Again o r ) h t h q i c d
grounds it seems somewhat unlikely that large Rcctionn of a #pawning
population should bo producing inviable larvae.
Thc irnportancs of cgg H i m on eurvival of larvao has been studied
by Blaxter and Hempel (1!163). Thoy tihowcd that within a race
larger eggs tended to produce larger larvae on hatching an wcll a8
larvae that lived longer on their yolk, Whether larger larvae arc' morc!
viable as a result of a larger mouth or greater activity remains to IJC
tested. Certainly larger larvae take larger food organisms according to
samples caught at sea (see Section IV, B). Between races the yolk
supply had a very strong effect on survival, larvae living for from
12-25 days at 8°C for instance, depending on their race. Nikitinskaya
( 1958b) found that the larvae of Clupea pallasii reared at temperatures
above normal were smaller, less well-developed and less active, as