The spread of a biological invasion in space and time: Modelling active and passive surveillance

Abstract

Invasive species are an important threat to global biodiversity and cause considerable economic losses. Modelling the spread of invaders can assist in mitigating the impacts of biological invasions by allowing us to identify strategies that are most likely to be effective in slowing or reversing their spread. In many situations, the main constraint to controlling or eradicating invaders is finding them rather than eliminating them after they are located. Once an invasion is found it can be treated and killed with a high probability of success. Searching large areas actively is expensive and therefore enlisting the help of the public through 'passive surveillance' is increasingly being used by pest-management agencies. The roles of active and passive surveillance and their interaction are investigated here using a spatially-explicit simulation model of the spread of an invasive species. The landscape is represented as a raster map consisting of square cells. Each cell in the landscape is characterised by various attributes, including habitat suitability and ownership type (private or public). The probability that a given site will be invaded depends on both habitat suitability and the number of propagules landing on it. Dispersal of propagules across the landscape is assumed to follow a Cauchy kernel. Long-distance dispersal may also occur independently, such as when propagules are transported by road or water. An invasion may be detected as a result of a report from the public or through active searching by a pest-control agency. Over time, the pest control agency uses passive detections, repeat searches and information about cell attributes to undertake additional searches in an attempt to eradicate the invader. The model is applied to a hypothetical invasion. Measures of success, such as cost and probability of eradication, are incorporated as fitness measures within an evolutionary algorithm that identifies optimal search and control strategies. Strategies are defined in terms of search effort applied per cell, the size of the neighbouring radius that is searched when an infestation is discovered, and the number of repeat visits to previously treated sites. Results demonstrate that increases in passive detection can reduce eradication costs and increased the probability of eradication. Although it is impossible to ensure that the global optimum is identified for a given scenario, the evolutionary algorithm helps identify quasi-optimal solutions that may be difficult to find through trial and error.