Detection parameters for managing invasive rats in urban environments

• Published in: Scientific Reports
• Main Authors: Henry R. Mackenzie, M. Cecilia Latham, Dean P. Anderson, Stephen Hartley, Grant L. Norbury, A. David M. Latham
• Format: Article in Journal/Newspaper
• Language: English
• Published: Nature Portfolio 2022
• Subjects: Medicine; R; Science; Q; New Zealand; Rattus rattus
• Online Access: https://doi.org/10.1038/s41598-022-20677-8; https://doaj.org/article/f25b5674689e4414a4be615b8045a854

Abstract Effective mitigation of the impacts of invasive ship rats (Rattus rattus) requires a good understanding of their ecology, but this knowledge is very sparse for urban and peri-urban areas. We radiomarked ship rats in Wellington, New Zealand, to estimate detection parameters (σ, ε 0 , θ, and g 0 ) that describe the process of an animal encountering a device (bait stations, chew cards and WaxTags) from a distance, and then approaching it and deciding whether to interact with it. We used this information in simulation models to estimate optimal device spacing for eradicating ship rats from Wellington, and for confirming eradication. Mean σ was 25.37 m (SD = 11.63), which equates to a circular home range of 1.21 ha. The mean nightly probability of an individual encountering a device at its home range center (ε 0 ) was 0.38 (SD = 0.11), whereas the probability of interacting with the encountered device (θ) was 0.34 (SD = 0.12). The derived mean nightly probability of an individual interacting with a device at its home range center (g 0 ) was 0.13 (SD = 0.08). Importantly, σ and g 0 are intrinsically linked through a negative relationship, thus g 0 should be derived from σ using a predictive model including individual variability. Simulations using this approach showed that bait stations deployed for about 500 days using a 25 m × 25 m grid consistently achieved eradication, and that a surveillance network of 3.25 chew cards ha−1 or 3.75 WaxTags ha−1 active for 14 nights would be required to confidently declare eradication. This density could be halved if the surveillance network was deployed for 28 nights or if the prior confidence in eradication was high (0.85). These recommendations take no account of differences in detection parameters between habitats. Therefore, if surveillance suggests that individuals are not encountering devices in certain habitats, device density should be adaptively revised. This approach applies to initiatives globally that aim to optimise eradication with limited funding.