Rapid detection of incursions in wild host populations is essential for effective management of transboundary animal disease events. The authors developed an adaptive, search-theoretic surveillance framework and field-tested it in New South Wales, Australia, targeting wild ungulate populations as potential hosts. The objective was early detection of incursions while explicitly accommodating high uncertainty about where and when an incursion might occur.
The framework was built around three operational principles: accommodate uncertainty in disease incursion risk, update search priorities regularly using observed surveillance data and estimated detection probabilities, and maintain a flexible, modular structure that can respond to changing information or operational conditions over time. These principles aimed to balance focused, risk-based effort with the capacity to detect unexpected events outside predicted high-risk areas.
A coarse statewide risk map provided a weakly informative prior describing spatial variability in expected incursion risk. This risk map was loosely focused on foot-and-mouth disease virus (FMDv) as an exemplar transboundary pathogen. The map guided initial allocation of search effort but was intentionally coarse to reflect uncertainty and to avoid over-concentrating surveillance in narrowly defined locations.
Search priorities were updated every three months. Updates used realised surveillance effort and estimated detection probabilities over the preceding 12 months. This mechanism meant that even persistently high-risk areas could have low present search value if they had recently been intensively searched, thereby encouraging spatial coverage and reducing redundant sampling in the same cells.
Field surveillance activities collected blood and swab samples from wild pigs (Sus scrofa). Over a two-year evaluation and refinement period the program conducted 110 sampling occasions and sampled 1,964 individual animals. The field activities were designed to simulate FMDv surveillance operations, although FMDv-specific serological tests were not available at the time of sampling.
Operational deployment consistently concentrated effort in areas with high search value. At least 74% of sampled spatial cells were in the highest risk class defined by the framework. Surveillance system sensitivity, estimated across five successive updating cycles, ranged from 0.86 to 0.93 and increased as operational procedures were refined during the evaluation period.
The surveillance activities were intended to emulate FMDv detection operations; however, FMDv serological tests were not available during the study period. Where details on test characteristics or exact detection-probability calculations are needed, those specifics were not reported in the source and therefore are not described here.
Although the program was designed around FMDv incursion risk, it met a secondary objective of detecting unexpected events. Specifically, the program detected Japanese encephalitis virus in wild pigs before detections were recorded in humans and domestic animals. This outcome demonstrates the framework’s capacity to identify emergent or unpredicted threats while maintaining a risk-focused approach.
By combining risk-based surveillance with adaptive updating of search priorities within a modular structure, the framework offers a flexible and generalisable method for early detection of transboundary and emerging animal disease incursions in wildlife. The adaptive updating—using three-month cycles informed by the previous 12 months of surveillance—helped balance focused searches with spatial coverage and reduced redundant effort in intensively searched areas.
The field test in New South Wales showed that an adaptive, search-theoretic surveillance framework can be implemented at scale in wildlife populations, concentrate effort in high-value search areas, and achieve high estimated surveillance sensitivity (0.86–0.93). The approach also proved capable of detecting unexpected pathogens, illustrating utility for both planned transboundary disease surveillance and broader early-warning roles under conditions of high uncertainty.