Investigation into the microbiological causes of epizootics of Pacific Oyster larvae (Crassostrea gigas) in commercial production

An investigation was undertaken into the cause of Pacific oyster (Crassostrea gigas) larvae epizootics occurring in a commercial hatchery located at Bicheno, Tasmania. An extensive monitoring survey was conducted at the hatchery to characterise the microbiological environment in the immediate vicinity of the larvae associated with different production outcomes. The surveys were performed over a 12-month period and included eight different production runs. Seven of the eight production runs terminated in disease incidence with larvae exhibiting disease symptoms consistent with bacterial infection previously described as bacillary necrosis, caused primarily by pathogenic Vibrio spp. Using production data records, physiochemical data, dissolved nutrient analysis, bacterial cultivation, TRFLP fingerprinting and 16S rRNA gene clone library analysis, two separate investigations were undertaken. In the first investigation microbial communities in each compartment of the larvae tank (water column, larvae, biofilm) as well as inputs into the tank (algae feed, seawater, and eggs) were characterised in both an intensive flow-though system and low intensity batch system, in order to understand the microbial ecological context in which disease occurred. Temporal variability of microbial communities was measured as an indicator of system stability. It was shown that microbial communities of the larvae and water column varied primarily with larvae age and sampling period and that the most likely cause of variability with sampling period was variability in the seawater. Altered culture conditions changed the microbial communities of the water column but larvae communities were shown to be largely resistant to change experienced in the water column. Larvae microbial communities were closely related to the indigenous microbial communities of the egg. Thus formation of the indigenous microbial community during spawning and fertilisation may be a control point for management of the microbial composition of the larvae and potentially for managing disease incidence. The presence of predominant non-typical marine species of the genera, Sphingomonas and Ramlibacte, in eggs and larvae samples, indicated a non-marine source of contamination occurring during spawning and fertilisation. The second investigation characterised the microbial environment associated with the emergence of disease 3 symptoms, and the underlying cause of disease. There was no predominant characteristic microbial community in the larvae or water column associated with disease and no recognised bacterial pathogens were detected using culture-independent methods of assessment. Vibrio population numbers peaked with the emergence of disease symptoms but remained only a minor component of the total population as indicated by prevalence in 16S rRNA gene clone libraries and next generation sequencing. Larvae aggregative behaviour near the tank bottom prior to the development of definitive disease symptoms indicated a non-microbiological primary cause of disease, or a microbiological etiology that occurred below the detection limits of 16S rRNA gene-based analyses used. Following the monitoring study an investigation was undertaken into the effect of environmental stressors on the susceptibility of larvae to bacterial challenge. Larvae were exposed to different levels of copper for 24 and 48 h before being challenged with three different bacterial species. It was shown that sub-lethal levels of Cu decreased larvae activity and increased larvae susceptibility to bacterial attack under some conditions. Relative to sub-lethal levels larvae exposed to the lethality threshold of 25 ppb Cu had higher activity levels and higher survival rates in subsequent bacterial challenge, which may indicate induction of the so called heat-shock response. Larvae behaviour was modified at Cu levels as low as 2.5 ppb, which indicates that behaviour could be used as a sensitive biomarker of Cu stress and potentially other forms of chemical stress. The behavioural response to different concentrations of Cu was non-linear and differed with duration of exposure, indicating that behavioural assessments should be made across a range of concentrations and also across a 24 – 48 h time period