Incorporating human behaviour into the risk-release relationship for invasion vectors: why targeting only the worst offenders can fail to reduce spread

1. Models of the spread of invasive species in transportation networks often assume a constant probability of transport across vectors, an assumption that is known to be false. The human dimension of these pathways is an important contributor to variation in vector risks. Yet, little attempt has been made to understand the effectiveness of strategies that involve changing risky behaviours. 2. We modelled the efficacy of management campaigns that aim to reduce the spread of invasive biofouling organisms by improving antifouling behaviour of boaters. We examined the influence of three factors on the effectiveness of the campaigns: (i) the proportion of the vessel population that is targeted by the measure, (ii) the average amount of change in behaviour made by individual boaters and (iii) the rate of uptake of the measure by the targeted population. 3. The modelled campaigns reduced the spread of an invasive species within New Zealand's recreational boating network by 0-56% over a 10-year period. The outcome depended strongly on the combination of the three factors examined. 4. Strategies that targeted only the 'worst offenders' caused relatively little reduction in spread unless there was very high uptake of the measure (>75% compliance) and a larger shift from current practice than could reasonably be expected. In comparison, a strategy that utilized a lower threshold for renewing the coatings and targeted a larger proportion of boaters was three times more effective, even for modest changes in current behaviour. 5. Synthesis and applications. Vector management can be effective in reducing the spread of a marine invader, but campaigns need to be well designed if they are to be effective. Their success will be determined by the shape of the 'risk-release relationship' and where along this curve the population of vectors lies. Greatest gains are likely when measures are targeted at critical transition points, where relatively small changes in behaviour lead to a considerable reduction in risk. High uptake of measures is required to reduce spread significantly. Combined 'carrot and stick' approaches that incorporate incentives for compliance are likely to be most effective at eliciting meaningful levels of change.