How many places rely on entirely fossil water sources?
I don't have a worldwide total for this, but the bits that I do know tell me that more of it does than doesn't. In the USA's 48 contiguous states, for example, the whole middle and western two thirds or so of it, including its most famous farmland, is naturally too arid for farming without wells, except along the coasts of Oregon & Washington. Water levels in the region's underground aquifers have been going down steadily, forcing wells to be redrilled or go dry. Texas & California, which also already divert disproportionate amounts of surface water from out-of-state, have had their well water levels descend by more than a hundred feet, which is a threshold also passed by India. One Indian aquifer has descended more than 30 feet just since 1980. A milder case spread over parts of several states but centered in Nebraska has lost about a tenth of its supply since the 1950s. In Asia overall, wells supply about three fourths of all agricultural water. Dried dead wells of various ages and the remains of the towns that once lived around them are a common feature of Middle Eastern and African deserts.
And will they not be able to replace them with renewable, if needed?
It's the other way around. Wells replaced the renewable sources when the perpetually increasing demand from more and more and more and more people wanting to eat and drink made the renewable sources simply not enough anymore.
This is the reason we use fertilizer.
No, it isn't. Out of 57 elements in our bodies that we get from the soil (60 minus C, O, and H), we use fertilizer to try to partially replace about a dozen. We don't use fertilizer or anything else to do anything at all about the rest. Historically, the response to a depleted field has been to just abandon it and move on. Much of what is applied is lost as runoff anyway before getting incorporated into the soil in a way that can be used, and by far most of it is made from mined rocks such as limestone and evaporite salts, not organic matter cycles, so they'll run out anyway.
If you farm a land heavily without using any form of fertilizer, you will typically "mine" it dry of nutrients in less than a decade, but much of the worlds agricultural land has been used for centuries or even millennia, so it is possible to establish a renewable cycle.
No. The depletion simply progresses more quickly in some places and more slowly in others, and in some cases another form of damage such as physical erosion or salination/salinization will ruin the land before depletion gets a chance to, but having it
not happen is just simply not a possibility within the laws of physics. Taking stuff away and not putting it back can only yield results in one direction.
Any sensible farming scheme has the avoidance of erosion as a central part.
Not avoidance. Reduction.
It remains fertile and un-eroded.
It is eroded. No farmed land isn't. It's just not eroded all the way down to nothing yet, so there's still some left. Similarly, the soil in the region I grew up in is several feet thick and still supporting farms for now, but that's only what's left after it has
lost several feet. I don't delude myself that the fact that it's not all gone yet means it will last forever. It's still only going away, not sitting still or being replaced from anywhere else.
On the "fertile" part, I don't know about the nutrient depletion rate there, but the difference between having already run out of something and being in the process of running out of it, as mentioned in the last paragraph, applies to nutrients as well as it does to the soil itself. Also, it's actually possible for the nutrient concentrations to stay the same if erosion works on the soil faster than depletion does, as long as the depleted part is what gets washed away and the part that's left behind hasn't yet had anything extracted from it (so each time you go back you're measuring the elements' concentrations in a slightly different sample from what was there before). Different lands suffer these problems at different rates, so one problem can strike when another hasn't yet.
There are ways to handle that.
If by "handle" you mean slow it down, yes, but slowing it down is only slowing it down. If by "handle" you mean completely stop or reverse it, please tell Australia and the southwestern USA about this miraculous breakthrough. You'll go down in history alongside Borlaug (and have a chance to become richer than some small countries).
