Ground-truthing a model in outback WA
- There is much value in ecological modellers getting input from conservation practitioners
- We demonstrated this point by refining a model predicting the effectiveness of a waterless barrier to stop the cane toad invasion in Western Australia
- Our updated model has produced a feasible and cost-effective solution
When building models of ecological systems, there are many reasons to engage with practitioners. For starters, practitioners often have a better understanding of the system being modelled and have access to the most relevant data. Such information can improve the realism and accuracy of model predictions. Practitioners can also benefit from engagement through an increased understanding and awareness of a model and its capabilities. This two-way dialogue can improve trust in a model and increase the chance it will be adopted to support decision-making.
Despite these advantages, engagement between modellers and practitioners is often limited. We recently attempted to bridge this gap between theory and practice for one of the worst invasive species in Australia: the cane toad.
We updated an existing theoretical model predicting cane toad spread through arid regions of Australia with local knowledge of weather and land use, to find the most cost-effective location for a ‘waterless barrier’ to contain the spread of toads (Southwell et al, 2016).
A waterless barrier
The idea behind a ‘waterless barrier’ is that cane toads need access to water every 3-4 days. Because of their inability to retain water, the toads simply can’t survive without it. So, in very dry regions, we may be able to halt their spread by excluding them from permanent water sources.
In arid areas of Australia, artificial water points, such as pastoral dams and tanks, are the only permanent water sources at which toads can rehydrate and breed. If we could prevent toads from accessing these water points, by replacing dams with leak-proof tanks, we could halt the invasion dead in its tracks.
A previous study (Tingley et al, 2013) suggested that a ‘waterless barrier’ composed of around 100 ‘managed’ dams in a thin ‘corridor’ of pastoral land between Broome and Port Hedland, in Western Australia, would likely contain the spread of toads into the Pilbara.
This arid coastal strip is a gap where the Great Sandy Desert almost reaches the coast. The cane toads really have to squeeze through this narrow bottleneck to reach the Pilbara and then continue spreading through WA. This corridor is the perfect trap.
Locals and experts
While a number of NGOs and local management groups have expressed interest in the waterless barrier idea, some practitioners had reservations. Why the doubt? Were they concerned with the data and assumptions underpinning the model? Or did they just misunderstand its limitations and capabilities?
To find out, we decided to ask them. To do that, we ran a workshop in Broome with local practitioners and experts in cane toad biology. Attendees revealed that they were most concerned about the accuracy of input data going into the model, such as rainfall variability, the locations of dams, and other land uses in the corridor that might support toad populations.
In response to these concerns, we updated the previous spread model to incorporate this information. We also drove the entire length of the corridor, verifying the locations of artificial and natural waterbodies on every property between Broome and Port Hedland.
In addition, we mapped a variety of other points that could potentially provide refuges for toads, such as dwellings, homesteads, roadhouses, as well as regions of irrigation and cropping. Our fieldwork enabled us to produce the most upto-date map of permanent water and land use on pastoral land between Broome and Port Hedland.
Waterless and cost-effective
With these updated maps, we then investigated the most-cost effective location for a ‘waterless barrier’ in the corridor. This involved simulating the spread of toads through the region in the absence of management, and then testing barriers placed at 17 potential locations. An economic model estimating the upfront installation and ongoing maintenance costs of a barrier was developed to find the most-cost effective barrier location. The updated model was run with and without local knowledge of the landscape, to determine if this information affected the best barrier location.
The results of the analysis suggest that the toad invasion front could be contained by excluding toads from fewer than 100 waterbodies, at a cost of approximately AUS$4.5 million over 50 years (that’s less than $100,000 a year). This is considerably less than the amount spent on other invasive species management programs in Australia. For example, the Australian government recently spent $19 million on feral camels in central Australia over 4 years, and $35 million on the fox eradication program in Tasmania over 8 years.
Just as importantly in terms of the environmental decision science, our research demonstrates the importance of practitioner engagement during model development. Local knowledge overlooked in the original incarnation of the model – specifically knowledge about irrigation and dwellings – influenced the best place for a barrier. Our new research suggests that the idea of a waterless barrier to halt the spread of cane toads is both feasible and cost-effective, and promises big conservation outcomes. This proposal is also a win–win situation for pastoralists and conservationists, because installing leak-proof tanks improves farm productivity, while simultaneously mitigating a key threatening process for biodiversity.
Structured decision making
While there are various ways to engage practitioners, to date, there is little understanding of which approaches achieve and maintain collaboration. For our waterless barrier workshop we adopted a structured decision-making approach, which is being increasingly advocated in the conservation literature for explicitly acknowledging uncertainty, facilitating relationship building and revealing hidden agendas (Addison et al, 2013). There is no doubt that in this case, the engagement process, facilitated by structured decisionmaking, helped all parties agree on the problem and objective, while improving practitioners’ understanding of the model’s capabilities and limitations, as well as modellers’ understanding of the landscape.
Addison PFE, L Rumpff, SS Bau, JM Carey, YE Chee, FC Jarrad, MF McBride & MA Burgman (2013). Practical solutions for making models indispensable in conservation decisionmaking. Diversity and Distributions 19: 490–502. (see Decision Point #74)
More info: Darren Southwell firstname.lastname@example.org
Note: The workshop in Broome was funded by the NERP Environmental Decisions Hub.
Southwell D, R Tingley, M Bode, E Nicholson & BL Phillips (2016). Cost and feasibility of a barrier to halt the spread of invasive cane toads in arid Australia: incorporating expert knowledge into model-based decision-making. Journal of Applied Ecology http://onlinelibrary.wiley.com/doi/10.1111/1365-2664.12744/full
Tingley R, BL Phillips, M Letnic, GP Brown, R Shine & SJE Baird (2013). Identifying optimal barriers to halt the invasion of cane toads Rhinella marina in arid Australia. Journal of Applied Ecology 50: 129-137. (And see Decision Point #82)