Back

Multiscale modelling of invasive species control: the cane toad in Australia

Pham, A. B. D.; Geard, N.; Baker, C. M.

2022-12-08 ecology
10.1101/2022.12.07.518288 bioRxiv
Show abstract

O_LIThe cane toad (Bufo marinus) is an invasive species in Australia that has a negative impact on native species. Control methods such as trapping, fencing, and water exclusion have been devised to contain the spread of cane toads and reduce their ecological impact. However, implementing these interventions is expensive, and estimating the likely impact of a proposed intervention on spread at a large spatial scale, comprising one or more control methods, is challenging due to the lack of large-scale data and the computational cost of modelling a large number of toads. C_LIO_LITo address this challenge, we developed a multiscale model which uses individual-level data on cane toad behaviour to estimate the impact of trapping, fencing, and water exclusion when applied at scale in the Pilbara region in north-western Australia. Compared to previous work, our model allows us to explore more complex combinations and tradeoffs of control methods by utilising data sources at different scales. C_LIO_LIOur results suggest that exclusion of toads from water points is the most effective method for containing the spread of cane toads, and that trapping and fencing alone are unlikely to be sufficient. However, trapping and fencing are still useful supplementary measures in scenarios where exclusion cannot be broadly applied to a large number of water points. C_LIO_LISynthesis and applications. Our analyses highlight the importance of limiting access to sheltering and breeding sites in invasive species control. Furthermore, this study illustrates the value of multiscale computational models for exploring scenarios where parameters and calibration data are available at the scale of individuals and small groups, but management questions are framed at a much larger scale. C_LI

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.