Development of an expert-based model for improved biofouling risk assessment

[摘要] This report seeks to enhance the Department of Agriculture’s existing biofouling risk assessment tools, which have been developed to underpin a proposed regulatory approach to biofouling on vessels entering Australian waters. The Invasive Marine Species Programme is in the process of implementing new biofouling management arrangements, and a risk assessment system (the Marine Growth Risk Assessment or MGRA) has been developed to accompany these. The MGRA will operate alongside existing quarantine pre-arrival processes, and consider information about vessel biofouling management such as presence and age of antifouling coating, history of recent hull survey and marine growth inspection, evidence of seawater pipe work treatment, duration of stay in overseas locations, and duration in Australian ports and coastal waters. However it does not have the capacity to identify high risk pathways based on compatibility between Australian ports, the temperature and salinity tolerance and distribution of invasive marine species of concern and the last ports visited by internationally arriving vessels. Here, we report on a project seeking to improve existing risk assessment methods by undertaking expert elicitation on various incursion scenarios, and statistical analysis of outputs.The project involved twenty nine national and international experts each rating the relative risk of thirty-five comparative scenarios. These ratings were used to construct a statistical model emulator summarising the expert’s views. The summary of these is as follows.The estimated impact of anti-fouling coat condition on relative risk was large, as expected. Compared to a new anti-fouling coat, having anti-fouling at 50% of its service life increased the risk by a factor of about 2. Similarly, compared to a new anti-fouling coat, having anti-fouling at 100% of its service life increased the risk by a factor of approximately 4.Moving from an enclosed river port to a coastal port lowered the estimated risk by 54% compared to moving between either coastal ports or river ports.The environmental similarity of ports was inferred to be an important cue by experts, as evidenced by the effect of relative temperature and salinity differences between donor and recipient ports. Each degree increase in the absolute difference in mean sea surface temperature from the port of origin changed the relative risk by 0.923 (c. 8% decrease). Similarly, each unit change in the absolute difference in mean water salinity from the port of origin altered the relative risk by 0.97 (a 3% decrease).Each additional day of berthing increased the relative risk by 1.033 (a 3.3% increase).There was no evidence that experts consistently factored in similarity in tidal range, or voyage duration, into their estimates.The model emulator developed is well-suited to risk ranking applications. It cannot be applied to determine absolute risk and therefore, given a threshold, whether a journeys risk is acceptable or not. If there was a requirement to do this, further analysis using the marginal rate of introductions would be needed. Thus the model as it stands could directly support compliance targeting. To apply it to primary estimation of risk would require both additional analysis as well as work to integrate it logically with the MGRA tool.

[发布日期] 2015-06-30 [发布机构] CSIRO

[学科分类] 地球科学(综合)