From “Neurotoxin Resistance in Snakes Around the World” ( ScienceDaily, Mar. 19, 2012), we learn,
… snakes from different regions of the world have evolved a similar, remarkable resistance to a deadly neurotoxin.
So they can eat poisonous amphibians:
Pfrender and colleagues found species of snakes in North, Central and South Americas and Asia that are able to feed on amphibians that secrete a deadly neurotoxic poison, tetrodotoxin or TTX. These snakes have similar mutations in a key sodium-channel gene that makes them highly resistant to TTX. These mutations prevent TTX from blocking the sodium channels in muscle, which would otherwise immobilize the snakes by paralyzing nervous and muscle tissue.
“The key finding is that adaptive evolution is constrained by the functional properties of the genes involved in these evolutionary responses,” Pfrender said. “While there are many possible mutations that can improve fitness, in this case resistance to the neurotoxin TTX, many of these mutations have a cost because they change the normal function of the genes. So, when we look at multiple species that have independently adapted to TTX, we see a very similar, and limited, set of mutations involved. The story is one of repeated evolutionary change that occurs through a limited set of changes at the molecular level.”
In short, they all found the one right solution. What would really be interesting is to find out how long it took them. What if the time frame in which they addressed the problem was actually quite short?
By the way, Simon Conway Morris is the go-to guy for convergent evolution.