We tested three hypotheses concerning the timing of spawning for a circumpolar species, capelin (Mallotus villosus), for which timing of larval emergence is known to be synchronized by physical conditions. The first hypothesis, developed from previous studies, was that spawning would be synchronized by upwelling events. Initial results from Middle Cove Beach in eastern Newfoundland indicated that spawning was not synchronized with upwelling. We next hypothesized that spawning was a function of several environmental variables. Results from logistic regression indicated that neither single-factor nor multi-factor models could explain the timing of spawning. Single variables could predict spawning in some years but no variable could reliably predict the time of spawning year after year. Finally, we hypothesized that the probability of spawning increased as a set of significant variables approached preferred levels. For capelin at Middle Cove, the set of variables that influence capelin spawning were identified as wave height, sea surface roughness and capelin abundance in the water. Thus only a combination of variables explained the timing of spawning for capelin. Preferred conditions for capelin spawning were wave heights less than 20 cm at the beach, a sea surface with a slight ripple, and an intermediate rank abundance of capelin in the water corresponding to hundreds to thousands of individuals. Capelin abundance alone was not a useful predictor. During the course of the study we observed a shift in the dates that capelin arrived and spawned at the beach. During 1987–1990 capelin spawned at Middle Cove Beach during June, but in more recent years (1991–1993) capelin did not arrive or spawn until J
展开▼
