In 1985 and 1986 many greenlip abalone Haliotis laevigata near Edithburg in the St Vincent Gulf died. The protozoan parasite, Perkinsus olseni, was suspected to be the cause. We showed that Perkinsus olseni was seasonally abundant in greenlip from the edge of the die-back area. It was also common in three other species of abalone, H. rubra, H. cyclobates and H. scalaris, and was recovered from 4 species of bivalves from the same area (Barbatia pistachio, Chlamys bifrons, Katylesia rhytiphora and Pinna bicolor).

In July 1987, 140 healthy abalone were transplanted from Hardwicke Bay to Stansbury at the centre of the die-off in an attempt to recolonize the area. This was done with the help of commercial abalone divers. In the following October, a subsample indicated the transplants were uninfected and were growing well. However, by March 1998, many of the abalone had recently died and the remainder were heavily infected.

In July 1989, 195 greenlip from Tipara Reef were tagged and transplanted to the same site. A subsample the following March indicated that there was good growth, very little infection, and no sign of recent deaths. The epizootic had evidently passed and these abalone were successfully recolonising the area. However, a follow-up sample a year later - in April, 1992 - indicated that many of the animals had become infected though there was no sign of recent deaths.

The die-back area is near the northern limit for greenlip abalone in the Gulf of St Vincent. Laboratory experiments showed that abalone infected with Perkinsus died more frequently than uninfected abalone and that stress such as high temperature may predispose the abalone to disease. The time of the original die-back corresponded to warm winters on the Yorke Peninsula.

Laboratory and field observations suggest that the parasite is common in greenlip in late summer. During winter abalone are able to contain and eliminate the infection. We conclude that the parasite is widespread around the Yorke Peninsula in hosts other than greenlip, and that greenlip near the upper limit of their temperature range are likely to contract fatal infections.