They are big. From the Wiki cane toad page,
The cane toad is very large; the females are significantly longer than males, reaching an average length of 10–15 cm (3.9–5.9 in). "Prinsen", a toad kept as a pet in Sweden, is listed by the Guinness Book of Records as the largest recorded specimen. It reportedly weighed 2.65 kilograms (5.84 lb) and measured 38 cm (15 in) from snout to vent, or 54 cm (21 in) when fully extended.They have spread rapidly. (From the cane toads in Australia page)
They have caused some, but not many, as best as I can tell, human fatalities, and in fact can be eaten safely if prepared correctly. But predators- including cats and dogs- are at risk. Native predators and other amphibians are at risk as well.
So back to the cat food. Is cat food toxic to the toads? No. It was actually that odd juxtaposition that led me to click over and read the story. Toads and frogs famously "can't see" immobile objects; I suspect the more accurate phrasing would be "don't appear to perceive." No, the cat food can used to bait foraging ants to the edges of water bodies during the period when the young toads are finishing their metamorphosis from tadpoles to their mature form. Australian meat ants are voracious predators, and will attack the young toads. They are also immune to the toads' poison.
Between July and September 2008, researchers studied tens of thousands of cane toads emerging from cat food-lined ponds and found that 98 percent of them were attacked by meat ants within two minutes. Of the toads that escaped, 80 percent died within a day from ant-inflicted injuries.As is pointed out in the article, this will not be enough to complete control the toad population, and the absolute mortality rate is not clear to me, but appears to be somewhere in the range of 80 to 98%. That's nothing to sneeze at.
But what really struck me was this bit:
The baby toads are less than half an inch (1 centimeter) in size, about the same as a meat ant. The aggressive ants have strong jaws and can kill even larger animals by sheer numbers.In most circumstances, that is not a stupid response. While an individual toad may be injured or die, it leads to an elimination of predators that are inclined to attack cane toads. In evolutionary terms, it benefits the toads' population. In the case of ant attacks though, while I wouldn't use the word "stupid," it clearly is a non-beneficial reaction.
"It's a pretty unequal fight," Shine said. "The toads have this terribly stupid response to attack - which is just to freeze and do nothing."
So what would non-scientific approaches predict in this situation? Nada. They don't do predictions, and they don't do testing. They just take everything for granted, and are thus useless.
What would a scientific approach predict? I'm no expert (he says, for the zillionth time), but this seems a pretty easy one to me. Chance variation will lead to a few offspring that jump like pole vaulters upon being bitten, or smelling the ants (formic acid is pretty pungent), or seeing an ant approach, or some other ant-related stimulus. These offspring will have a better chance of surviving and passing these traits on to still more offspring. So if cat food warfare is widely and consistently adopted, it seems likely that it will lead to behavioral changes in the cane toad population. Which in turn suggests that cat food technology may have an impressive impact at first, but decreasing benefits over the long term.
This has happened over and over, with insects, with germs, with rodents, with all sorts of pests. Man invents a better mousetrap, natural processes design a mouse better equipped to avoid it. That's not to imply that we should give up trying to invent better mousetraps- quite the opposite. What it does imply, though, is that our actions have broadly predictable outcomes, and it behooves us to be aware of them. Anticipating future conditions is far preferable to arriving in the future and being left with praying for divine intervention as our only option for coping with problems we ourselves have contributed to creating.