"Three Sisters" on the industrial scale?

[Admiral Squish]

So, I got way into studying native american cultures a while ago, and one thing that's stuck with me was how differently these cultures approached agriculture, and some frankly ingenious tricks they employed.

One of the most notable strategies being the 'Three Sisters'. Wiki link, if you wanna learn all the details. Basically, instead of monoculture fields arranged in rows, you arrange mounds in a grid or hex pattern, and in each mound you plant corn and vine-beans, with squash in the spaces between. The corn grows, the beans climb the corn stalks and add nitrogen to the soil, and the squash's broad, hairy leaves shade the soil to help it retain moisture and serve to deter certain pests. All very symbiotic.

So, my question is, could this strategy be scaled up to a modern industrial level of production? I imagine harvesting would be problematic, but would it be truly impossible to develop a harvester capable of gathering all three crops at once? I must confess, I know basically nothing about gardening/farming/agriculture, but the idea at least seems worth pursuing from my perspective.

[Rockphed]

So, I got way into studying native american cultures a while ago, and one thing that's stuck with me was how differently these cultures approached agriculture, and some frankly ingenious tricks they employed.

One of the most notable strategies being the 'Three Sisters'. Wiki link, if you wanna learn all the details. Basically, instead of monoculture fields arranged in rows, you arrange mounds in a grid or hex pattern, and in each mound you plant corn and vine-beans, with squash in the spaces between. The corn grows, the beans climb the corn stalks and add nitrogen to the soil, and the squash's broad, hairy leaves shade the soil to help it retain moisture and serve to deter certain pests. All very symbiotic.

So, my question is, could this strategy be scaled up to a modern industrial level of production? I imagine harvesting would be problematic, but would it be truly impossible to develop a harvester capable of gathering all three crops at once? I must confess, I know basically nothing about gardening/farming/agriculture, but the idea at least seems worth pursuing from my perspective.

The big reason modern agriculture uses rows instead of mounds is that it is easier to navigate a tractor down a field of rows of corn. Also, it is easier to overturn the earth with a plow into rows than any other shape. I could see it being possible to have rows of corn mixed with vine-beans and plant squash in the space between the rows. However, I'm pretty sure that vine-beans and corn have very different harvesting times. Running a mechanical processor over either of them would destroy the other. And with the squash between, manually harvesting the beans before the corn is ripe is harder than without the squash. However, depending on the squash, you can probably harvest it after you run a combine through the corn.

In other words, I think you might be able to make it work, but unless you can justify manually harvesting the beans, you are probably better off with a 3 year crop rotation.

[Trekkin]

So, my question is, could this strategy be scaled up to a modern industrial level of production?

It could, but not as efficiently as other strategies.

The Three Sisters, much like milpas, is simply a specific kind of companion planting. There are tangible benefits to companion planting, but you also run into problems trying to optimize the conditions for each type of plant and nutrify the soil to a level that can handle the requisite plant density. (It's difficult to get beans to grow up corn without their root systems competing.) It's also hard to handle cases where the crops you want to companion plant are sown or harvested at different times, since weeds will grow in the gaps.

On an industrial scale, it's simpler to rely on modern chemistry and optimize the fields for a monoculture. Given access to a chemical plant, companion planting is valuable mostly for making crops more valuable to chemophobes via a fallacious appeal to nature.

[Spojaz]

In many of the corn fields I've been in, there were soybeans growing between the rows, because planting nitrogen fixing soybeans is cheaper than fertilizer. If I remember correctly, the beans profit margin was so low that they were never harvested, just tilled under and left to decompose as nutrients for the next season. They could have pulled two crops out of the same land, but the limiting factor in American farming is the cost of manhours and machines, not efficient land use. I think the two crops have different times to be harvested anyway, so a perfect machine made to harvest it all would pull up a bunch of unripe or overripe produce.

Where everything is picked by hand, and one limiting factor is how far you have to walk to pick your fields, polyharvest farming works very well. In industrial farming, it does not add enough value to justify the specialized machines.

[Trekkin]

I didn't realize there were places where nitrogen fixing companion crops were economically preferable. Interesting.

I wonder, then, if polyharvest farming wouldn't make sense for biodiesel production. Assuming you could find the right crops, ripeness might be less important than total biomass, in which case you really could just rip everything up at once.

[cobaltstarfire]

Most of the information I've read on companion crops used in modern contexts is the same as what Spojaz has said, they are grown together, but the companion crop is usually cut down rather than harvested. Or sometimes when a field would otherwise be empty there is a crop grown in it who's soul purpose is to prevent erosion and, mop up nutrients, these are usually cut, let to sit a while, then tilled under.

One of the interesting companion crops I saw was barley? (I can't remember I was reading up on this stuff last year) where it is grown specifically to shield the desired crop from pests, weeds, and too much sun until it was bigger and stronger. I think these are sometimes called nurse crops?

[Yora]

Machine harvesting is integral to industrial agriculture and that just wouldn't work.

The next best thing is crop rotation, where you plant different plants in different seasons and don't repeatedly plant the same crop in the same spot every year. A simple method that appears to be in use for some 8000 years now.

[jayem]

Machine harvesting is integral to industrial agriculture and that just wouldn't work.

I don't know. Machine complexity of actions has increased recently. It's more feasible for a machine to do complex actions quickly than it was say 20 years ago, combined with semi-AI based observation. Almost certainly not enough to be financially viable* (especially for the big American fields and combines, for British sized fields and combines though it would be closer), but definitely heading back that way. I don't think it will be enough for corn though.

*after all experienced people are pretty quick, and lose out to the combine with corn.

[Dodom]

I didn't realize there were places where nitrogen fixing companion crops were economically preferable. Interesting.

I wonder, then, if polyharvest farming wouldn't make sense for biodiesel production. Assuming you could find the right crops, ripeness might be less important than total biomass, in which case you really could just rip everything up at once.

It's already done for forage in some places, and it does increase total yield. As long as all the plants used are edible as is and have compatible transport conditions, they can all be harvested in one go and left mixed. Cows won't mind.

[shawnhcorey]

And then there's permaculture. But it's hard to compare permaculture to monocropping since permaculture is dependant on local conditions.

[Lacuna Caster]

Machine harvesting is integral to industrial agriculture and that just wouldn't work.

The next best thing is crop rotation, where you plant different plants in different seasons and don't repeatedly plant the same crop in the same spot every year. A simple method that appears to be in use for some 8000 years now.

This depends on how sophisticated machine harvesting gets. A general-purpose agricultural android- which isn't that far away- could have the same degree of finesse as manual labour without the same labour costs (even if it probably went slower than, say, a combine harvester.)

I am curious about how things like beans and strawberries are currently harvested on an industrial scale- it seems like the latter in particular would be tremendously easy to damage?

[Trekkin]

I am curious about how things like beans and strawberries are currently harvested on an industrial scale- it seems like the latter in particular would be tremendously easy to damage?

You'd be right.

Beans are generally harvested with bean threshers, which work much like most harvesters: yank most of the plant off of the roots and spin/tumble off all the foliage. Underground beans, like peanuts, are harvested like potatoes and share lifted carrots: the foliage is cut off beforehand and the crop dug up and shaken free of dirt.

I had thought there were top lifting bean threshers, but apparently there are not.

Strawberries are picked by hand, although automatic mechanisms for removing the calyx exist. They work best with specific cultivars, though.

[J-H]

+1 for permaculture, although unfortunately it's generally a very labor-intensive process (design, install, maintain, harvest). I've seen few if any setups that allow mechanical harvesting on a practical scale.

[Grey Watcher]

I do wonder if developments in developing swarm behaviors in robots might make industrial scale Three Sisters viable. Granted, there's a number of issues: the programming of the individual units; making sure the hardware has the right combination of precision, strength, and durability to do harvesting; figuring out what to do with the newly picked crops such that the worker drones don't get weighed down to immobility, etc. But it's an interesting thought....

Machine harvesting is integral to industrial agriculture and that just wouldn't work.

The next best thing is crop rotation, where you plant different plants in different seasons and don't repeatedly plant the same crop in the same spot every year. A simple method that appears to be in use for some 8000 years now.

A bit of a tangent, but I had thought crop rotation was invented in the middle ages; prior to that we just alternated a given field between "growing" and "lying fallow." At least in Europe, because every "world" history class I've ever taken is really European-and-US history. Were my European ancestors that far behind my Indian ones? (Or was it the Chinese or the Mali or somebody else that got to call "f1rst cr0p!" on that one?)

[snowblizz]

A bit of a tangent, but I had thought crop rotation was invented in the middle ages; prior to that we just alternated a given field between "growing" and "lying fallow." At least in Europe, because every "world" history class I've ever taken is really European-and-US history. Were my European ancestors that far behind my Indian ones? (Or was it the Chinese or the Mali or somebody else that got to call "f1rst cr0p!" on that one?)

I'm not convinced saying the "3 sisers method" is superior is even correct, or suggesting medieaval Europe in this case was backwards. Most human populations learned to make effective use of what they had at hand. The biggest problems invariably comes later on when trying to apply an ingrained (ha! bonus pun) agricultural system in a place without considering if it is feasible. Most European colonist had trouble here wherever they went.

Eg Chinese rice cultivation exploded in yield and spread in the middle ages as far as it could go north basically. But it also meant quickly filling up all arable land with a very labour intesive (though productive) crop that couldn't easily be mechanised, while at the same time pushing out space for eg livestock and thus fertilizer. This meant chinese rice culture to a degree lived on the cutting edge of what the land could produce, no matter what, well into 1800s with some unfortunate results for society and people over time (population pressures, famines etc etc).

Likewise in the Americas the lack of similar big working animals meant a different take on agriculture. The 3 Sisters version is a very good way to work in a limited scale within the technological and cultural constraints existing. I know similar systems worked elsewhere too.

In Eurasia there were plenty of domesticable animals to work with and so agriculture tends towards using muscle power and combining livestock + cropping. I'm not an agronome (agrologe?) so can't say for sure, but it's likely Europe in particular didn't have a simialr trifecta to work with. They used the methods that worked in European conditions.

Ultimately all systems have different strengths and weaknesses. The point about "world" history bears considering when saying one system was more backwards than another, success can be measured different ways.

[ace rooster]

Growing nitrogen fixing plants as an alternative to fertiliser sounds good, but is actually quite inefficient. You would be growing something like 4 fields of nitrogen fixation per field to match the yields we get (biofuels exempted*). You might think that this shows our dependence on fossil fuels, but actually it more shows the low yields of nitrogen fixers. Making fertiliser from biofuel (using the haber process) is far more efficient, with the current problem being that fossil fuels outcompete biofuels.

There are definite advantages to moving away from monocultures, but nitrogen budget is not one of them.

* Biofuel production is not nitrogen limited, because there should be very little nitrogen in the fuel. The nitrogen should be tilled back into the soil, rather than leaving the system (and hence having to be replaced). The only things leaving the system should idally be hydrocarbons, which are made from water and air.

[Lacuna Caster]

I'm not convinced saying the "3 sisers method" is superior is even correct, or suggesting medieaval Europe in this case was backwards. Most human populations learned to make effective use of what they had at hand. The biggest problems invariably comes later on when trying to apply an ingrained (ha! bonus pun) agricultural system in a place without considering if it is feasible. Most European colonist had trouble here wherever they went.

...Likewise in the Americas the lack of similar big working animals meant a different take on agriculture. The 3 Sisters version is a very good way to work in a limited scale within the technological and cultural constraints existing. I know similar systems worked elsewhere too.

I don't have any particular expertise in the area aside from what I read in 1491, but my understanding was that maize/beans/squash (with lime for cooking) provided a complete complement of amino acids, and Europeans were frequently astonished by maize yields relative to wheat. (Some of that holds up today.)

On the same time-scale where we get reliable swarm robotics, we're likely to have sufficiently powerful gene-editing tools that you can just plug nitrogen-fixing genes directly into maize (or, hell, grow chicken mcnuggets on umbilicals in a vat, or something.) But I guess it's interesting to ponder.

[Rockphed]

I don't have any particular expertise in the area aside from what I read in 1491, but my understanding was that maize/beans/squash (with lime for cooking) provided a complete complement of amino acids, and Europeans were frequently astonished by maize yields relative to wheat. (Some of that holds up today.)

On the same time-scale where we get reliable swarm robotics, we're likely to have sufficiently powerful gene-editing tools that you can just plug nitrogen-fixing genes directly into maize (or, hell, grow chicken mcnuggets on umbilicals in a vat, or something.) But I guess it's interesting to ponder.

Of course, without the lime treatment, corn does not provide very many amino acids at all.

[tantric]

it should actually be the 'four sisters' with sunflowers.

i suspect you would want to use 'no till agriculture'. this would be part of sustainable agriculture, which is completely different from organic ag (organic foods are luxury items. for every vegie that makes it to the market, several more are discarded. if you really want to reduce pesticide usage GET USED TO PICKING BUGS OFF YOUR PRODUCE, DAMMIT)

one thing that will help such systems is the full domestication of ants. weaver ants are already being used to protect cashew trees from pests but we need a species of ant that will live in a contained colony, like bees in hives, that can be moved from field to field. that should be 100% insectivorous and obligatory diurnals so that the colony can be put in a field in the morning and opened all day then plugged at night and moved.

[Lacuna Caster]

One thing that will help such systems is the full domestication of ants. weaver ants are already being used to protect cashew trees from pests but we need a species of ant that will live in a contained colony, like bees in hives, that can be moved from field to field. that should be 100% insectivorous and obligatory diurnals so that the colony can be put in a field in the morning and opened all day then plugged at night and moved.

That's a great idea. I actually think that some heavily-genetically-engineered strain of social insects could be a major boon that way- particularly if the grubs could be collected as animal protein, like we get honey from beehives. I suspect that regulating their food supply would be a problem, though- if they're 100% insectivorous then it's hard to eliminate pests without the colony starving. Plenty of ants also live partly on nectar, though, and there are acacias that provide 'food pellets' to their lodgers that way.

[Rockphed]

it should actually be the 'four sisters' with sunflowers.

i suspect you would want to use 'no till agriculture'. this would be part of sustainable agriculture, which is completely different from organic ag (organic foods are luxury items. for every vegie that makes it to the market, several more are discarded. if you really want to reduce pesticide usage GET USED TO PICKING BUGS OFF YOUR PRODUCE, DAMMIT)

one thing that will help such systems is the full domestication of ants. weaver ants are already being used to protect cashew trees from pests but we need a species of ant that will live in a contained colony, like bees in hives, that can be moved from field to field. that should be 100% insectivorous and obligatory diurnals so that the colony can be put in a field in the morning and opened all day then plugged at night and moved.

I would think obligate nocturnal would be better. You set them out in the afternoon and open the hive. Then you let them go all night and plug them up in the morning. You get the benefits of a full 12 hour insect work cycle without having to do the work of moving them in the middle of the night when you would rather be sleeping.

[tantric]

you're right - nocturnal ants could get the damned slugs and wouldn't attack pollinators. it seems like a very useful technology. i wonder if i could get some venture capital on this?

[Rockphed]

you're right - nocturnal ants could get the damned slugs and wouldn't attack pollinators. it seems like a very useful technology. i wonder if i could get some venture capital on this?

Smart money in venture capitalism is to wait until they actually have a model for what they are going to build before you give them money. Then you don't build a $300 juice press that nobody wants.

[Trekkin]

Smart money in venture capitalism is to wait until they actually have a model for what they are going to build before you give them money. Then you don't build a $300 juice press that nobody wants.

There's also the question of relevant expertise, particularly for a project like this -- and when I say expertise, I mean PhDs, plural, at bare minimum, covering everything even tangentially related to the endeavor in question. Then there's the business end, which gets even more annoying, because that's where they start bikeshedding, and past all that is the legal part, where they get scared.

[Lacuna Caster]

There's also the question of relevant expertise, particularly for a project like this -- and when I say expertise, I mean PhDs, plural, at bare minimum, covering everything even tangentially related to the endeavor in question. Then there's the business end, which gets even more annoying, because that's where they start bikeshedding, and past all that is the legal part, where they get scared.

I think this is one of those situations where a government grant- or possibly some money from the Rockefeller Foundation- going to the right people might be a more suitable funding model. One of the few coherent arguments I've heard against GM crops is that corporate funding leads to monopolistic practices (e.g, infertile 2nd-gen seeds) that undercut poor farmers' autonomy.

How expensive is gene-editing, exactly? Social insects seem to have a couple of useful properties for creating minimal-impact GMOs, since all reproduction is centralised in a few individuals, and they have pretty short breeding cycles, though you'd need to be careful that they don't breed with wild populations.

[Trekkin]

How expensive is gene-editing, exactly?

Without touching on the many, many problems associated with getting these sorts of things funded and subsequently using them out in the wild: this is a question with no single nitpick-proof answer except to say that the cost of the plasmids themselves is far less than the cost of the equipment and facilities required to work with them and the ongoing cost to maintain them. This isn't even touching on the expertise required; "the right people" do not work for enthusiasm, and they generally want things in addition to money (publications, et cetera) that they won't get working for laypeople in garage labs, so you'd have to offer substantially more than even an industry salary.

For me, the cost to do a single CRISPR/CAS9 reaction (since that's what you mean by gene editing) is about $60-$70 plus media depending on the guide RNAs involved, but that's a woefully incomplete figure. The total cost for some random person to start editing ants is going to be several orders of magnitude higher, and to hire someone to do so successfully will be more expensive still.

[Lacuna Caster]

Without touching on the many, many problems associated with getting these sorts of things funded and subsequently using them out in the wild: this is a question with no single nitpick-proof answer except to say that the cost of the plasmids themselves is far less than the cost of the equipment and facilities required to work with them and the ongoing cost to maintain them. This isn't even touching on the expertise required; "the right people" do not work for enthusiasm, and they generally want things in addition to money (publications, et cetera) that they won't get working for laypeople in garage labs, so you'd have to offer substantially more than even an industry salary.

For me, the cost to do a single CRISPR/CAS9 reaction (since that's what you mean by gene editing) is about $60-$70 plus media depending on the guide RNAs involved, but that's a woefully incomplete figure. The total cost for some random person to start editing ants is going to be several orders of magnitude higher, and to hire someone to do so successfully will be more expensive still.

True, but hypothetically a sufficiently enthusiastic layperson could put themselves through college or self-educate on the topic, then use savings to buy the equipment. Probably still rather pricey, but if some mad scientist wanted to start a project in their basement, it's not quite impossible.

To be clear, I'm not nominating myself for this idea- I wouldn't know where to start. But you do occasionally get qualified professionals willing to do work at substantial discounts.

[Rockphed]

For me, the cost to do a single CRISPR/CAS9 reaction (since that's what you mean by gene editing) is about $60-$70 plus media depending on the guide RNAs involved, but that's a woefully incomplete figure. The total cost for some random person to start editing ants is going to be several orders of magnitude higher, and to hire someone to do so successfully will be more expensive still.

So a graduate student might be able to do research on editing the genome of ants and then take that to industry to get funded, but they would probably instead form a working group at a university and, over the course of 20 years, explore the depths and breadths of editing the genes of sugar ants. One of their students' students would probably be the person who would have a base ecosystem of ant-editing to be able to make domesticated, obligate-nocturnal, obligate-pestivore ants. I.e. this is technology that is 30+ years away.

Ninjaes:

To be clear, I'm not nominating myself for this idea- I wouldn't know where to start. But you do occasionally get qualified professionals willing to do work at substantial discounts.

Norman Borlaug is one of my heroes.

My Engineering 1000 professor told us to chase money right up until we got a degree. Then the money would come, but we wouldn't be working for money. Looking over Borlaug's wikipedia entry, I think he had similar ideas. When he left Dupont for Mexico, Dupont offered to double his salary, which he refused. The man kept working pretty much right up until he died because he cared.

[Trekkin]

So a graduate student might be able to do research on editing the genome of ants and then take that to industry to get funded, but they would probably instead form a working group at a university and, over the course of 20 years, explore the depths and breadths of editing the genes of sugar ants. One of their students' students would probably be the person who would have a base ecosystem of ant-editing to be able to make domesticated, obligate-nocturnal, obligate-pestivore ants. I.e. this is technology that is 30+ years away.

Something like that, yes. It's in the realm where a functioning system for funding science could see it happen in under a decade, but our present dumpster fire won't approach for fifty years or so. Thirty is a good average. Part of the problem, of course, is that the student's student would want to publish their results, at which point no industry will touch them since they can't be patented. Science can make things possible, but they need to be profitable to exist, and that generally means a guaranteed monopoly given these up-front costs. The bigger problem is the utter lack of enthusiasm for basic research or indeed anything that won't end in a new wonder drug and a cure for cancer and at least six of the reviewers' dead family members coming back to life and baking cookies for them. This project is about nine-tenths basic science, at a first approximation, which neither government nor industry are especially willing to fund; it only becomes explicitly hypothesis-driven and translational and boringly possible once all the -omic science is done.

[Rockphed]

Something like that, yes. It's in the realm where a functioning system for funding science could see it happen in under a decade, but our present dumpster fire won't approach for fifty years or so. Thirty is a good average. Part of the problem, of course, is that the student's student would want to publish their results, at which point no industry will touch them since they can't be patented. Science can make things possible, but they need to be profitable to exist, and that generally means a guaranteed monopoly given these up-front costs. The bigger problem is the utter lack of enthusiasm for basic research or indeed anything that won't end in a new wonder drug and a cure for cancer and at least six of the reviewers' dead family members coming back to life and baking cookies for them. This project is about nine-tenths basic science, at a first approximation, which neither government nor industry are especially willing to fund; it only becomes explicitly hypothesis-driven and translational and boringly possible once all the -omic science is done.

I don't know how biology works, but my Electrical Engineering department produces lots of patents. Just because they publish their results does not mean they have to forgo getting a patent. If anything, patents exist to encourage publishing of scientific advances.