Wild populations of pinto abalone (Haliotis kamtschatkana) in Washington Statehave declined by 97% since 1992, despite a fishery closure since 1994. No recruitment has been detected recently, indicating probable reproductive failure due to low densities. A pilot programme placed a total of over 11,000 hatchery-origin juveniles, age 18-22 months, at 10 sites in the San Juan Islands.2. Observed (naive) year 1 survival averaged 10.2% (0-23% range) and was most influenced by site compared with lineage or size-at-outplant. Families survived in the approximate proportions that they were outplanted, and there was little support for an effect of size-at-outplant on survival.
Detection was lowdue to the small chance of sighting individuals on complex substrate. When derived from repeated sampling, an upper bound on naive detection rate averaged 0.38 and increased with size. When derived from a closed capture-recapture model, average detection was estimated at 0.19. 4. Growth was highly variable and confounded with detection, but an average 3.4% of detected outplants across all sites (0-7.5% range) had reached reproductive size in 2017. 5. A state-space model of exponential population growth was modified to account for imperfect detection and yielded an estimated density of abalone for each survey. Seven out of eight sites included in the model remained above a target abalone density of 0.3 m −2 throughout the project. 6. The majority of tagged abalone made little net movement over weekly and annual timescales, although some emigration likely reduced survival estimates.7. The restoration programme is transitioning from a pilot phase to a production phase, including optimization of hatchery and outplant processes. Existing wellperforming sites will receive additional cohorts every 4-5 years to maintain aggregation densities. New sites will replace poorly performing ones, although this is hampered by a poor understanding of the mechanisms behind site performance.