Jerboas are like kangaroo mice that produce all of the water they process physiologically. They don't drink. dumb cheap longevity idea: feed jerboas as well as regular mice different hydration diets then measure longevity.
If there is a difference then there is likely a nonROS post metabolism lifespan strategy addressable with the kidneys.
I've read that even moderate decreases of water with humans results a third less muscle output. That means there might be a more/less efficient metabolic pathway cascade when a third less output or a bizarre annoying CR type thing urging / discouraging drinking water prior to being asleep when the body repairs/doesn't need to metabolize.