Well, the whole of the Mediterranean (the Greek and the Spanish didn't make their fleets out of the present-day mediterranean vegetation)

We're talking about food, not forestry. And Spain is a net food exporter, with a vastly larger population than it had in the days of the Armada, so I really don't see that it supports your claim.

Italy and Greece would be better examples -- both are net food importers -- but both have strongly increased their food yields over the last decades (Italy, Greece), allowing them to support vastly larger populations than in their empire years.

the Middle East (the cedars of Libanon, North Africa: the bread basket of Rome). Each time the soil in Mesopotamia became bad enough, the current empire retracted and after a few floodings it was ready to go again.

Egypt currently produces enough food for about 2/3 of its population, or about 50 million people. i.e., Egypt alone could feed 80% of the Roman Empire circa 300AD; North Africa could still be the breadbasket of Rome.

In fact, almost every major agricultural area currently supports many more people than it ever did in the distant past. Even Iraq produces millions of tons of cereals, despite its agricultural capability being degraded by successive wars and sanctions, which is most likely more than it produced in Babylonian times.

In fact, the strong trend seems to be that ancient breadbaskets are still strong producers of agricultural goods, which is directly counter to your argument.

And farming might be sustainable for millennia in its current form, that's not enough. It must be sustainable practically forever.

One way or another, farming 1,000 years from now will be very different from farming now -- the odds that we'll be at a comparable level of technology are vanishingly small.

I do agree with you that we should avoid long-term damage or degradation to our fertile lands, though.