This tool shows a probabilistic forecast of cane toad spread towards Broome and into the Pilbara region of Western Australia. The coloured dots on the map represent known waterpoints, and the colour of each dot shows the likelihood that toads will have reached that waterpoint by the year you select. The forecast is based on a computer simulation model that has been fitted to — and validated against — the real history of toad spread across northern Australia.
Cane toads cannot survive thoe northern dry season without water. In seasonally arid landscapes like the southern Kimberley and Pilbara, water is found at discrete points — bores, tanks, dams, springs, and rock holes — scattered across an otherwise dry landscape. The model treats these waterpoints as stepping stones. Toads establish at a waterpoint and survive there over the dry season. The next wet season they breed nearby and then the young toads disperse outward. If they are lucky, they find another waterpoint to survive the next dry season.
How far toads can disperse in a given year depends on rainfall. More rainy days means more nights of active movement, and so toads can travel further. The model uses the average number of rainy days recorded at each waterpoint over the past 20 years to estimate a "dispersal kernel" — essentially a probability map of how likely a toad starting at one waterpoint is to reach any other waterpoint during a wet season. This dispersal kernel was calibrated using GPS-tracked toads from the invasion front near Darwin.
Once we know how far toads are likely to travel, the model works out how many toads are expected to arrive at each unoccupied waterpoint from all the occupied waterpoints around it. A waterpoint is considered colonised when at least two toads arrive — one toad alone cannot establish a breeding population. Once colonised, a waterpoint becomes a new source of dispersing toads in the following wet season.
Toad spread is not perfectly predictable — there is genuine randomness in how many toads arrive at any given point in any given year. To capture this uncertainty, the model is run many times ( 100 replicates), each time with slightly different random outcomes for dispersal and colonisation. The colour on the map reflects the fraction of those runs in which a given waterpoint was colonised by the selected year. A waterpoint shown in dark red was reached in nearly every simulation run; a yellow waterpoint was reached in only a small proportion of runs.
The model was originally developed using data on toad spread through the Victoria River District and the Kimberley. We re-fitted it to the observed spread across the Kimberley, which gives a more precise estimate of how fast and how readily toads move through this type of landscape. The forecast shown here is initialised each year from the latest known position of the toad invasion front, as mapped by nocturnal surveys across the front.
This website is supported by the SKIP Foundation, the Western Australian Government’s Department of Energy and Economic Diversification (through the Premier’s Science Fellowship Program), and by BHP Social Investments. Data locating the invasion front each year have been collected with the assistance of Nyikina Mangala, Karajarri, and Nyangumarta Rangers.
The app was built by Ben Phillips , with help from Claude AI. It is hosted on the ARDC's Nectar Cloud.
Forecast last updated: July 2026 . toadfree.zone