The tricky thing about ecological niche crashes is that, while academically fascinating, they tend to induce high rates of mortality among bystanders. That is why I love messing with wolf-sheep-grass models. They let you tinker with semi-realistic predator-prey models — by changing start variables, cutting the food supply, etc. — without offing yourself.
Case in point: Here is a run of a wolf-sheep-grass model — wolves eat sheep, which eat grass, and both burn energy wandering around in search of food — where I made the grass regrowth time somewhat lower than the norm. It is, in a sense, a simplified model of a plausible system in which the underlying energy supply grows, but at a rate vastly slower than the energy needs of the populations supported, at first, by said energy source’s legacy size. Gosh, that sounds so familiar. I wonder why ….
As you can see, at the end of a few cycles of population boom and bust, the world returned to all grass, and both predators and prey were gone. First the prey disappeared in a paroxysm of over-consumption from an over-inflated population of wolves, and then the wolves followed the sheep over the die-off cliff, what with no longer having anything to eat, grass apparently not being an adaptation they could quickly make. You have to admit, however, the landscape sure is nice and green at the end though.