Distribution and Survival of Herring Larvae (Clupea pallasi Valenciennes) in British Columbia Waters

Through study of the distribution and survival of herring larvae of one of the main populations in British Columbia waters, it is concluded that (a) the principal cause of death in Pacific herring larvae is traceable to their passive transport by inshore water currents to the open sea, and (b) the number of larvae that are not so transported in any particular year determines the strength of the resulting year-class.Greatest concentration of larvae after hatching was found in inshore waters near the spawning grounds, and as they were carried seaward the numbers remaining decreased at an approximately constant rate. The movement of larvae, as noted through sampling, was related to the pattern of water currents, as determined by oceanographic survey. The extensive movement of larvae offshore was explained through seaward moving currents being strongest (a) in waters in which the largest larval broods were present, and (b) at the surface of the water where larvae were heavily concentrated at night. Predation, lack of food, and conditions of temperature and salinity were not considered important mortality factors in inshore areas.Sampling indicated that few larvae carried to sea returned to inshore waters. Furthermore, several pieces of evidence suggested that most larvae carried offshore did not survive. The failure to find either larvae offshore or nearly-metamorphosed juveniles near the mouth of Barkley Sound appeared significant, particularly since schools of the latter were regularly found inshore near the spawning grounds. The factors directly causing death in offshore waters were not determined, but it was suggested that death may be connected with the high salinity of the open sea. It was concluded that the strength of herring year-classes was largely determined by the numbers of larvae that remained in inshore waters during the critical period of larval life.Meteorological factors were considered capable of causing annual variations in current flow and hence in year-class strength, and the delayed effect of snow on river discharge appeared to be of greater importance than the more immediate effect of rainfall. No single meteorological factor, however, seemed to have a dominant influence.