Crop choices

One way to increase crop yields is to simply grow different crops.

Even if we only measure yields in terms of calories & protein (see why) Why is ranking only based on protein and calories?
It's true that humans need a lot more than just protein and calories. We also need vitamins, minerals, fiber, and omegas (3 and 6). But in a world where farm land is scarce (which is today's world, in the sense that both hunger and deforestation are happening), protein and calories are of greatest concern. Here's why:
- There is almost always an overabundance of fibre in crops regardless. There are parts of plants we don't eat because they are so fiberous.
- Nearly every crop in its natural form contains some vitamins and minerals.
 - Absolute worst case, if there are any missing nutrients, they can be supplemented/fortified with almost zero land footprint.
 - But as you will see in the generated scenario below, optimizing for calories & protein actually improves the range of vitamins & minerals anyway.
- Omegas 3 & 6 are a genuine concern. I'll add these into the analysis when I get the time. The next version of the nutrition calculator will include omegas 3 & 6. per unit of farmland, it turns out that the top-yielding crops are incredibly varied between countries - vegetables, fruits, grains, pulses, nuts, seeds, or even spices - depending on the region.

See the full chart for every country and every food crop. Note: the page is quite big and may be slow to load.

For the rest of this page, the word "yield" is shorthand for "calorie and/or protein yield".

Imagined scenario

Using only existing crop land (no deforestation):

Suppose every country grew mostly its top-yielding crops. We don't want a pure monoculture(...)( as this would probably actually be bad for yields in the long run ), but suppose that high-yield crops are grown much more often than low-yield crops.

What might the global food supply look like? I decided to do some data analysis to find out.

Technical description of the analysisTechnical description of the analysis

Using data on average local yields for each crop in each country:
For each country:

• Each crop is given an amount of land proportional to the square of its yield score.

• The yield score is based on both calories and protein (human-edible only).
• Weighting: 2000 calories is equivalent to 60 grams of protein.

• Total amount of farmland (all crops combined) is the same as the status quo.

I generated the scenario by writing Code:food1.sql, which uses data from FAO (crop production/yields) and USDA (nutrition data).

Some key findings

• Globally, total crop yields (calories) would be almost doubled.
• In Africa, total crop yields (calories) would be tripled.
• The food supply would be more nutritionally complete. iWhen comparing the nutritional content of the global food crop supply:
  per 2200 kalories of food:
   - The status quo does not meet the daily recommendations for potassium, calcium, vitamin A, vitamin D, vitamin E, vitamin K, vitamin B5 (pantothenate) and vitamin B12.
   - The new scenario would meet the daily requirements for all nutrients except vitamin D (normal for food in general) and vitamin B12 (normal for unfermented plant-based foods).
  For more details, see the #Nutrition calculator links section below.
• There would be far more fruits & vegetables (which vary widely by region).
• However, some food choices are very questionable - such as ridiculously large amounts of garlic & chili peppers.

Maps - food crop production:

BeforeAfter

For a detailed breakdown by country, see this page.

Awkward findings

Extremely high vegetable production

It might be strange to think of garlic and chili peppers as being sources of protein, but in some places they are so high-yielding that they technically could be. This might not work so well in practice, as you'd have to eat a whole bowl of garlic cloves or hot peppers just to get a noticeable amount of protein. The pungency of these food crops is probably a factor in why they can be so high-yielding (very few pests will eat them, perhaps).

Other vegetables can also be high-yielding - in some regions, the sheer volume of them could provide calories & protein. But in order for a person to consume enough, the vegetables would probably have to be dehydrated and ground into flour. (...)( Vitamins may be lost from this, but minerals wouldn't. ) (...)( In the case of garlic and chili powders, we'd need to actually remove some of the flavor in order to be able to eat more / use as flour. This would require a whole food-engineering project. )

This opens up a whole world of culinary experimentation - making foods from mostly vegetable flours (plus some fresh vegetables and other ingredients).

However, we still don't know if this could actually help feed the world or not. Some questions remain unanswered:^{[RESEARCH needed]}

• Vegetables are normally a lot more expensive than grains and pulses, per dry mass;
  • Is this due to a high labor-intensity of harvesting? If so, scaling up vegetable production could be a major problem. (...)( If it's labor-intensive enough, the whole idea should be scrapped; otherwise it would just promote wage-slavery. )
  • Or is it mainly due to the perishability of fresh vegetables? If so, dried non-perishable vegetable powders could be inexpensive and not a problem.
    • Hopefully this is the case. One piece of evidence for it: Some countries can make sugar cheaply from sugar beets (a root vegetable).
• Are there any technical impediments that would make this whole thing a bad idea?
• How to encourage farmers in "the right" regions to grow "the right" veggies to maximize calorie/protein yields?
  • Important because this particular optimization is not the norm(...)( there are probably a whole lot of economic reasons why ), according to the data(...)( Some of the foods that are abundant in the #Imagined scenario, actually have quite low "world-average" yields currently (see /chart, row "World"). This means the crops are "grown in the wrong places", in a way. ). Doing the vegetable-powder idea with today's configuration of who-grows-what-where would actually make the whole feeding-the-world-situation worse, not better. (...)( Sorry for the awkward phrasing - I'm not sure if better terminology even exists. )

Importance of imports & exports

In the imagined scenario, some countries produce one or two foods almost exclusively. This wouldn't make a balanced diet. But since each country produces different foods, the overall global food supply is very balanced. Even on a continent-wide scale, there's enough variety of foods to make balanced diets.

In other words, countries would have to trade with each other a lot. This would require more shipping as food would be less local. But still, overall carbon emissions would probably be lower, because of all the deforestation prevented (see below). Keep in mind that food transport is only a small percent of food's environmental footprint. And it could be smaller still if foods are processed locally and shipped in their more concentrated forms.

But also, keep in mind that this whole analysis was done with only country-level data. It could very well be that within each country, different crops are optimal in different regions. This is especially likely in any country that has multiple types of land. So in many cases, a country probably could produce a diverse local diet while still maximizing calorie & protein yields.

Nutrition calculator links

The links below show the crops as foods in the nutrition calculator, as food per day per capita.

	World:	before	after
	Africa:	before	after
	Asia:	before	after
	Europe:	before	after
	North America:	before	after
	Oceania:	before	after
	South America:	before	after

Note:

• This is primary production, which is significantly higher than what people actually eat.
• Not all foods can fit in the nutrition calculator, so the totals appear slightly less than the actual totals.
• The bar graphs overflow a lot. To make it easier to visually compare nutrients, try increasing the "Time Scale" to several days.

Using the crop choices to prevent deforestation

The #Imagined scenario above, the total farmland stays the same but food production vastly increases. This was chosen only for simplicity sake.

In actual fact, we don't have to increase food production that much(...)( the existing global food supply would be more than enough if we didn't waste so much food, and if we didn't feed so much food to farm animals ). We could use the high-yielding crop choices as a way to need less land for agriculture. By reducing the demand for land, we could get rid of the economic pressures that cause deforestation. And besides just forests, it would become easier to preserve wild ecosystems in general.

Africa

Of all the continents, Africa currently produces the least amount of food crop per capita. This is probably a significant cause of Africa's high hunger rates (although technically it doesn't have to be)Primary production is still higher than the average person's nutritional needs, but just barely, so any food waste/loss or crop-fed livestock lower the supply to below sufficiency.. In the crop-choices scenario imagined, Africa's food crop production(...)( in terms of calories and nutrients in general ) is effectively tripled(...)( which is almost certainly "overkill" but important for illustrating the extent to which crop choices can make a difference ). This is achieved by reallocating farmland: less land for grains, more land for sugarcane, chilis, bananas, potatoes, and garlic.

• Interpretation 1: Africa would benefit from growing more of these high-productivity crops instead of grains. Perhaps there's a historical reason for growing grains - maybe European colonialists thought it was a good idea because grains get high yields in Europe. In that sense, if Africa switched to the other (more productive) crops, it can be a form of decolonization.
• Interpretation 2: Sugarcane etc. gets higher yields but only because corporations invest more money & fertilizer into growing it. First-world countries want sugar, and their spending power makes it happen (highly-valued currency). The same economics doesn't care about Africa's grains, because grains can be grown "at home" in rich countries. Perhaps if African grain farmers had more access to resources(...)( fertilizer? something else? depends on the specific case; this would be a whole topic in itself ), grains would yield just as many calories as sugarcane.
• Interpretation 3: None of this really matters, because the whole crop-choices scenario is too hypothetical, and involves unrealistic dietary choices.

These interpretations are opposed to each other. The current dataset can't tell us which one (if any) is true. If you have some insight, join the discussion.

More considerations

Monoculture vs polyculture

Be careful not to interpret these results as "This one crop is the best, let's just grow that and nothing else!" It could be that some crops depend on each other for their high yields. I recommend growing a mix of top crops, and especially include legumes for nitrogen fixing.

Local variation

If a country is geographically big enough, different parts probably have different top-yielding crops (due to different soil/climate/etc). The current dataset doesn't have any info on this - crop yields are only a country-level average.

Social, economic, and technological factors

We can't prove that the top-yielding crop is always better-suited for the local region. In some cases, perhaps there are just more resources invested into it, such as fertilizer. In which case, some currently-less-productive crops could be just as good if given a chance. The current dataset has no info on this.

See also

• Food - the main page which includes other possible ways to improve the global food supply.