Should You Quit Eating Almonds Because of the California Drought?

Many bloggers are pinning the blame for California historic drought on almond farmers and the hipsters who eat their product.

It takes 1.1 gallons of water to produce just one almond, according to a Mother Jones blog post. That led to a number of follow-up posts with headlines like “The Dark Side of Almond Use” and “Your Almond Habit is Bleeding California Dry.”

But how do almonds stack up to other foods? I wolf down a handful or two of almonds per day myself, so I thought I’d better look into this. I don’t eat meat, so almonds are a key source of fat and protein for me. I’d always thought eating plant-based foods was better for the environment. Had I been wrong all along? Would I be better off getting my fat and protein from meat?

Probably not, I’ve concluded, but I’d love for someone to check my work.

Grams of fat vs. water used

The first issue is finding out how much water goes into different types of foods. The Mother Jones post didn’t have much information, but this Treehugger post cites numbers from that are more useful for comparisons. I wasn’t able to find these specific numbers on the Water Footprint site, but I did find this table (see above), citing a study by Mekonnen and Hoekstra that I haven’t looked at yet.

Update: I’ve found the report, and it changes things a bit. doesn’t support striking through text, so I’m putting the now outdated bits in italics below.

None of them broke down the gallon per gram numbers for almonds, though, so I’m left to work that out on my own. If there are 416 almonds in a pound then it takes 457.6 gallons of water to produce that pound, which would contain 249.6 grams of fat or 1.83 gallons per gram.

Another source says there are 14 grams of fat in a serving of 23 almonds, which would be 25.3 gallons of water, or 1.81 gallons per gram.

I’m pretty sure the numbers for the almonds are wrong, because they’re so out-of-whack with everyone else’s analysis, so I’d love for someone to double check my work.

Update: Based on the Mekonnen and Hoekstra report, it takes 4,251,848.34 gallons of water to produce 2,204.62 pounds of almonds, which works out to 1,928.61 gallons per pound — about four times what I’d originally estimated. I think this was probably due to over estimating the number of almonds in a pound. I’ve since found a couple references online that say there are about 100 almonds in a pound, which would put my original numbers closer to the Mekonnen and Hoekstra numbers, which sound much more correct.

So it would actually be: 8.68 gallons per gram of fat.

Almond milk’s water footprint is more difficult to calculate. Yes, you add more water to almond milk, but unlike the water that goes into growing an almond, or some other food, you actually get to drink at least some of the added water in almond milk. Most of it, I would guess. So it’s probably not actually much worse than eating straight almonds. Update: Dairy alternative company So Delicious has an environmental footprint site that includes comparisons of milk and various non-dairy milks assorted environmental impacts, including carbon and water. But it’s not the most unbiased source and I haven’t found the source for the data it uses for non-dairy products.

Regular milk takes 33 liters per gram of fat, or 8.71768 gallons per gram.

Butter, meanwhile, takes 6.4 liters per gram of fat, which is 1.6907 gallons per gram.

Treehugger says it’s 220 gallons for a pound of avocados. A pound of avocados has 68.03 grams of fat, which is 3.23 gallons per gram.

Grams of protein vs. water used

If there are 416 almonds in a pound then it would take 457.6 gallons of water to produce that pound, which would contain 83.2 grams of protein or 5.5 gallons per gram. Based on the Food Network numbers above, it would 4.22 gallons per gram. This still sounds wrong to me.

Update: Based on the new numbers above, almonds take about 20.25 gallons of water per gram of protein.

One pound of tofu* has 36.29 grams of protein and takes 244 gallons of water to produce, or 6.72 gallons per gram.

One pound of ground beef has 65 grams of protein and takes 2,500 gallons of water to produce, or 38.46 gallons per gram.

One pound of chicken breast has 95.25 grams of protein and takes 815 gallons of water to produce, or 8.56 gallons per gram.

One pound of hard boiled eggs has 59 grams of protein and takes 573 gallons to produce, or 9.71 gallons per gram.

One pound of cheddar cheese has 113.25 grams of protein and takes 896 gallons of water to produce, or 7.91 gallons per gram.

*Soy gets a bad rap, basically because of a single study that was widely reported to have found that men who ate 56 grams of soy protein powder per day had a 19 percent reduction in their testosterone levels. A subsequent metastudy of 15 other studies found no effect. I haven’t been able to find any other study that shows an effect. It’s been hypothesized that the results of that 19% study were skewed by a single test subject who had abnormally high levels of testosterone (about double the average) going into the study. And another study found that men who ate a lot of soy protein were at a lower risk for prostate cancer. Make of that what you will. Wikipedia has a good summary. Now, that doesn’t mean necessarily mean that soy is good for you, but the claims that say soy is bad for you lack evidence. Still, lots of people are allergic to soy, so your mileage may vary.

What About Carbon?

Even if California’s drought wasn’t caused by climate change, we can expect climate change to cause and/or exacerbate droughts int he future. So, if we want to make environmentally driven dietary decisions, we also have to take carbon footprints into account. There’s been much written on the topic, so I’ll try to keep this brief.

A study by Shrink That Footprint looked at the overall consumption patterns of five different diets, and found that heavy meat eaters have the biggest carbon footprints (see graphic above).

But what’s striking is that there’s a relatively small gap between vegans and meat eaters who don’t eat beef, at least compared to the gap between meat lovers and those who don’t eat beef.


There are many other environmental factors at play here, including carbon foot print, transportation, use of fertilizers and other pollutants, etc. Some people have allergies or other condition that make vegan diets difficult, so that has to be taken into account as well.

It’s hard to make direct comparisons of nutrient content. Different cuts of meat have different properties, but for purposes of my calculations above I just treated chickens and cows as if their meat were entirely uniform. And of course there’s the question of how accurate any of these numbers are.

Also, the bioavailability of protein varies quite a bit from food to food. And there are other nutrients in play. Beef, for example, has much more bioavailable iron than soy, eggs or dairy. Also, animal products have B12 that vegans must source elsewhere, so you’d have to factor the production of those supplements into your diet. If there are health benefits to eating or not eating certain foods, that could cut down on the need for pharmaceuticals, which would have an impact as well.

I wondered if, in terms of actual bioavailable iron and protein, maybe chicken and fish might actually be more efficient than soy. But the bioavailability thing isn’t straight forward. For example, vitamin C is known to increase absorption of iron, and you’ll find that many vegetarian meals include sources of vitamin C, such as tomato, bell peppers and avocado. Companies like Beyond Meat are actively work to increase the bioavailability of nutrients in their products by mixing synergistic ingredients.

And though the health of vegan and vegetarian diets is the subject of much debate, one study found that “Although the iron stores of vegetarians may be reduced, the incidence of iron-deficiency anemia in vegetarians is not significantly different from that in omnivores,” which suggests that most of us are at least getting by.

There’s also the question of transforming plant-based protein sources like wheat and soy into meat substitutes like seitan, tofu, and various veggie patties. I asked Lindsay Wilson from Shrink That Footprint about this, and he told me that veggie protein processing has a pretty light footprint compared to meat. Beyond Meat CEO Ethan Brown told me in October that the company hadn’t done a lifecycle analysis as of last October, but was about to start one.

Again, I’ll point to the Shrink That Footprint study, since it looks at what people actually consume, on average, vs. the theoretical efficiency of individual foods, and vegans still come out with the smallest footprint.

And of course, animal rights are a big part of why people chose to go vegetarian or vegan, so environmental factors aren’t the only thing at play.

What should we actually learn from this?

There’s a pretty strong case against beef here. While almond critics like to point out that 10 percent of California’s water goes to almond farming, they don’t tend to mention that 50 percent goes towards livestock. While there’s no silver bullet answer to the drought crisis, it seems clear that best policy interventions would be those that curb beef production.

At the personal level, unless you have some medical condition that necessitates eating lots of beef, it seems hard to justify. Just cutting beef and lamb out of your diet would be almost as good as giving up meat altogether. But, as always, consult your doctor before making any sort of dietary changes.

Based on my math, it actually appears that, in terms of the amount of nutrients produced, almonds are one of the most efficient uses of resources. But I’m pretty sure I’ve either made an error or my data source is wrong, or something, because pretty much every expert out there says that almonds are really inefficient. But if my math is correct, it seems that almond production should actually be prioritized, not penalized.

Based on my math, almonds aren’t as bad as dairy milk or beef, but they certainly lag behind other alternatives, such as hazelnuts (which use almost as much water, but relatively little irrigation water) and coconut milk.

Those who want to err on the side of caution, but still want plant based alternatives, might consider diversifying their sources of fat. Personally, I’ve added pumpkin seeds and Oregon hazelnuts to my snack rotation, just to mix things up a little.

If you liked this article, please consider subscribing to my not-quite-weekly email newsletter.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.