Abstract

We compared: 1) rearing mortality, 2) size at release (mean length), 3) jack, male and female sizes, and 4) jack, female, and adult returns of coho salmon (Oncorhynchus kisutch) from three consecutive brood years reared at the Nitinat River hatchery using a conventional or a semi-natural rearing method. The semi-natural method included feeding restrictions, shading of the rearing ponds, lower rearing temperature and rearing densities, exposure to a predator and a volitional release. We found no significant effects of brood year or rearing method on rearing mortality; it was significantly lower during the marking to release phase than during the other two phases (eyed-egg to ponding, ponding to marking). Conventionally reared smolts were significantly longer. Conventionally reared males were longer. As a proportion of number of smolts released, semi-natural rearing produced 86% fewer jacks, the same proportion of females and 15% more adults. Adult production trends, described as marine survival rate (returning adults • smolt^-1) for Nitinat River Hatchery coho, and as ln recruits • female spawner^-1 for a nearby wild coho population, were similar. Jacking rates were lower in Nitinat River hatchery coho than for the nearby wild coho population. We concluded that the semi-natural rearing methodology produces adult fish more efficiently than the conventional rearing method does, and at 73% of the cost.