I got these numbers from the USDA 2014 Agricultural Overview. I aggregated them willy-nilly and even rounded to make the addition easier. This is rough; it won’t even add up to 10 million acres. But the next time someone says ‘California grows HALF! the countries fresh vegetables and fruit’, you can say, ‘yes, and they grow that on about one-third of their irrigated acreage. So we could stop irrigating the other parts with no loss to American salads.’
The 2014 California and Nevada Vegetable Crop Summary tells me:
The top 24 fresh market vegetables: 1.58 million acres
The top 8 processed vegetables: 1.09 million acres
The rest of these values are from the USDA 2014 Agricultural Overview:
Citrus: 271,000 acres (oranges: 170,000, lemons: 46,000, grapefruit: 10,000, tangerines: 45,000)
Fruit: 217,000 acres (peaches: 44,000, prunes: 48,000, melons: 56,000, cherries: 33,000, nectarines: 21,000, apples: 15,000)
Rice: 434,000 acres
Grapes: 880,000 acres (wine: 570,000, raisin: 200,000, table: 110,000)
Corn: 520,000 acres (420,000 acres silage = fed to cows)
Hay: 1,380,000 acres (alfalfa: 875,000 acres, other hay: 500,000 acres)
Nuts: 1,365,000 acres (almonds: 860,000 acres, pistachios: 215,000 acres, walnuts: 290,000 acres)
Last year’s drought resulted in fallowing about 500,000 acres.
You can add those up in different ways, but the fresh produce that humans eat directly (both veggies, citrus, fruit, table grapes) comes to 3,300,000 acres. That is one third of our irrigated acreage. We can round up to 4 million acres to get the small specialty crops. Sixty percent of California’s acres could drop out of production, 6 million acres, with no loss of fruits and vegetables. The rice is directly consumed by humans. Then about 2 million acres of corn, alfalfa and other hay is fed to animals. Then we grow nuts for the world.
I list these numbers to refute the implicit argument that goes: California grows half the country’s fruits and veggies and therefore must get enough water to irrigate all the land in production now. If the important thing is having fresh produce, we do that on about forty percent of currently irrigated land. After that, I believe there are policy choices to be made about whether we want to use water to make wine, keep the price of meat and dairy low, or grow nuts for the world.
On the public record 
Mike Wade is so going to find a different way to willy-nilly aggregate these numbers.
I know these are rough-and-ready numbers, but how well does “fraction of irrigated land” correspond to “fraction of water used for irrigation”? Is water use per acre relatively steady, or does it vary a lot by crop?
If water application is efficient, you can guess that a crop uses about 3 to 3.5acre-feet/acre. With very precise application, that might hover around 3acrefeet-acre. Almonds are thirsty, at about 4acre-feet/acre. These days, rice is about the same. If you told me about a crop that took only 2acre-feet/acre, I’d think it was getting some subsurface water or was a very drought tolerant species. You can do deficit irrigation for a while (one or two years in a row), but it will eventually stress the trees or vines and yields will drop. Once a year or two, you have to apply water just to wash salts below the root line.
What varies more than how much crops need water is how the water is applied. Any irrigation method can be very efficient, but some methods take a hell of a lot of management to be efficient. (especially furrows. Efficient furrow irrigation is incredibly sophisticated. Moderately efficient (or outright inefficient) furrow irrigation is more common.) Well managed subsurface drip, surface drip or precision sprinklers can apply water at very nearly the crop water requirement.
tl;dr: Acreage is a decent approximation for crop water use, probably not off by more than 10-15%. Crop water needs are clustered within 15% of 3.5acrefeet/acre (roughly). Irrigation method introduces more variation than crop water need.