Ideas for Polycultures / Companion planting

Often we see these tables "companion planting" in the internet. I wonder if they are really true or just some myth.

First let me say I am a firm supporter of companion planting. Plants can provide shade and climbing support (corn in three sisters), keep humidity (squash in the three sisters) and fix nitrogen (beans in the three sisters). In other cases they repel pests and the roots do not compete: onions/garlic and carrots.

The brassica family (cabbage family) gains advantage of having marigolds, herbs and garlic planted around them. I wonder if we could also have a nitrogen fixing vegetable for the brassica, which are heavy feeders. It would have to be a non-invasive legume. Perhaps peas or lentils. Not other kinds of more aggressive climbing beans or broad beans. Any suggestions? Anyone has tried other companions for the brassicas?

How about for celery? Once I tried a polyculture with kohlrabi and leeks and they worked fine.

How about potatoes? Potatoes do not need much attention in terms of watering or fertility, but they produce a lot of lush growth, and therefore they would need to have a companion that is tall and does not compete much for its needs and root space. Perhaps also fast yielding, since potatoes are quite quick to produce a harvest. And also preferably some cool weather crop. Otherwise I thought of beans, amaranth or corn as some possibilities. Leeks could be an option. Any has experimented with companions for potatoes?

Another difficult one is finding companions for squash or pumpkins. It must be something tall. Corn is a obvious choice. But other than it, what other possibility? Amaranth and okra are also tall and drought tolerant, but not very quick growing.

Finally how about cereals? What could be a companion for rye or wheat? Perhaps a nitrogen fixing crop?
And how about a companion for amaranth?

Anyone has experimented with companions for these crops?

My idea is to create a list of the ideal companion or companions in a two crop polyculture:

Corn - Beans - Squash
Brassicas - Onions/garlic
Carrots - Onions/garlic
Celery - Leeks
Brassicas - nitrogen fix crop ?
Potatoes - ?
Amaranth - ?
Rye- ?

For broccoli, cabbage, collard greens, kale, or anything else in that family goes well with almost any allium. When the brassicas are harvested tomatoes and peppers can take their place. For anything in the cucurbit family (squash, melons, watermelons, and cucumbers) goes well with corn, sunflowers, amaranth, and runner beans.
For grains, sow with clover, and about a week to 10 days before harvest sow lettuce, radish, and other fast growing crops in the standing grain. When the grain is harvested return the straw to the field and the lettuce and radish will sprout through the mulch.

All of these things can be sown amongst trees. Preferably fruit or nut trees. And if you use mulches and refrain from tilling you can allow for volunteers hopefully creating less work for you the farmer/gardener.

You can replace or add amaranth to the three sisters with corn. So amaranth beans and squash, and maybe additionally corn.

I just polyculture everything together. Grains in clumps, herbs dotted around, fruit trees here and there, veggies spread about and root veggies everywhere. Its as Paul mentions it "polyculture 50+."

i kinda think companion plantings are myths, but polycultures are real...that's my belief

Thanks for your suggestions.

I still miss a companion to make a polyculture for a small potato field. Perhaps some broad beans.

What about grains? I would like to have a polyculture of rye or perhaps a nitrogen fixing crop, that is edible. Maybe peas? It is for a small field of rye that I have.
Anyone tried something like that?

Paulo,Hi

Here's the list I have for a potato guild...I found it on the internet somewhere..haven't tried it,yet

horseradish
hyssop
beets
cabbage
onions
oregano
alyssum
parsnips
clover
geranium
chamomile
confrey
flax
eggplant
beans
petunia
dill

I look forward to hearing how your guild in the potato field works

All the best
Kate

With annuals I would simply just create a seed mix and distribute throughout the planting area. Allow the plants to self seed and in a couple seasons the plants will find the most ideal locations for themselves and find the best companions as well. Grains can be done in patches or in orchards. Polyculture for grain will not be as complex but they can also be sown in a field of perennial herbs, culinary or medicinal.

For the potatoes, I would include beans.

Potatoes help repel Mexican Bean Beetles.
And beans help repel Colorado potato beetles.

Last spring I tilled a huge space of grass and just threw all of my leftover seeds from all the past 5 years along with a mix of anything I thought would grow from the bulk bin at the grocery store. In all, it was a mix of about 60 varieties. By this method I was able to selectively thin things out to meet my needs. All the greens, beans, flowers, clovers, squashes and beets (along with so many others) did well. Slower growing things like carrot and onions just couldn't compete with all the others, so they just lagged. That being said, I had no real pest issues to deal with. I did find that bok choi makes a great "catch crop" for flea beetles. I think this method was a good way to see how many interactions are created when nature has the chance to make the choices.
I've made up some new mixes for this year, and I'm planning on doing three different ones in separate areas to see what happens. If I can keep the chickens and pigs clear of all of it I'll have something to report back on in the fall.

Paulo Bessa wrote:Often we see these tables "companion planting" in the internet. I wonder if they are really true or just some myth.

First let me say I am a firm supporter of companion planting. Plants can provide shade and climbing support (corn in three sisters), keep humidity (squash in the three sisters) and fix nitrogen (beans in the three sisters). In other cases they repel pests and the roots do not compete: onions/garlic and carrots.

The brassica family (cabbage family) gains advantage of having marigolds, herbs and garlic planted around them. I wonder if we could also have a nitrogen fixing vegetable for the brassica, which are heavy feeders. It would have to be a non-invasive legume. Perhaps peas or lentils. Not other kinds of more aggressive climbing beans or broad beans. Any suggestions? Anyone has tried other companions for the brassicas?

How about for celery? Once I tried a polyculture with kohlrabi and leeks and they worked fine.

How about potatoes? Potatoes do not need much attention in terms of watering or fertility, but they produce a lot of lush growth, and therefore they would need to have a companion that is tall and does not compete much for its needs and root space. Perhaps also fast yielding, since potatoes are quite quick to produce a harvest. And also preferably some cool weather crop. Otherwise I thought of beans, amaranth or corn as some possibilities. Leeks could be an option. Any has experimented with companions for potatoes?

Another difficult one is finding companions for squash or pumpkins. It must be something tall. Corn is a obvious choice. But other than it, what other possibility? Amaranth and okra are also tall and drought tolerant, but not very quick growing.

Finally how about cereals? What could be a companion for rye or wheat? Perhaps a nitrogen fixing crop?
And how about a companion for amaranth?

Anyone has experimented with companions for these crops?

My idea is to create a list of the ideal companion or companions in a two crop polyculture:

Corn - Beans - Squash
Brassicas - Onions/garlic
Carrots - Onions/garlic
Celery - Leeks
Brassicas - nitrogen fix crop ?
Potatoes - ?
Amaranth - ?
Rye- ?

I had good results last year with fava beans and potatoes (potatoes in a row, favas in the 'corners' between the rounded potato plants), and also some onions and carrots. Sometimes i'd make a line across the potato bed, sometimes cluster them in the corners. I have never had excess fava beans so planting them along with potatoes makes for good staple production.

I saw a listing that said to avoid planting potatoes with squashes, tomato, sunflower, and curcurbit;
it recommended horseradish, beans, corn, cabbage, marigold, limas (and by extension other beans?), eggplant (as a trap crop for potato beetle).

As a rule I don't tend to interplant the same family (nightshades for example) though I might alternate them down a bed if I want, say, a long bed of onions but only a few plants of eggplant, pepper, or tomato. Often the same family will tolerate similar conditions, or attract similar pollinators.

I also had a good bed a few years back with tomatoes, basil, carrot, onion, cilantro, and borage (to boost pollination). so many tomatoes I had to re-stake it twice, and didn't come near to harvesting it all.
I suspect beans would work OK there too - I usually do bush-beans so I don't need to deal with stakes. For nitrogen, we tend to apply compost and green mulches, plus I let clover and other legumes self-seed and don't weed it out unless it's crowding other plants.

Peas didn't thrive around the onions, as predicted.... but then I am not necessarily great at getting peas to thrive anyway. I think I've been planting them too late.

-Erica

I'm going to second the opinion that companion planting is mostly BS but polycultures are real.

My biggest distrust of companion planting is that a lot of the traditional companions are seasonally incompatible. "tomatoes love carrots"? Tomatoes are semi tropical and carrots are cool weather. If they love each other so much then they are doomed to a Romeo & Juliet type of relationship because they aren't going to grow together except when their seasons are overlapping a little. Someone is going to chime in with "carrots are biennial", i just know it. Let me revise. "They aren't going to grow together" unless you are saving carrots for seed or don't mind woody carrots.

The only time i can see companion planting working is when we are very practical about it and consider the physical structures of the plants. Beans climbing up corn? Thats real. Tomatoes giving shade to lettuce in summer? Absolutely. One of my friends once observed that his mint and rosemary grew well together because the mint sucks up all the water and it is otherwise too dry for rosemary to thrive. Similarly, I can definitely see the wisdom behind growing stuff that has shallow roots next to stuff that has deep roots or nitrogen loves next to light feeders (score one for tomatoes and carrots).

This whole "x plant seems to prosper next to y plant" is something i like to call "placebo companion". We observe improved growth because we are expecting to.

dan long wrote:

This whole "x plant seems to prosper next to y plant" is something i like to call "placebo companion". We observe improved growth because we are expecting to.

There is a way to quantify this. The concept of Land Equivalence Ratio (LER). If the LER is greater than 1, then the plants are good companions, making better use of the soil than they would if grown in monoculture. However, if the LER is less than 1, something detrimental is going on and they are not good companions.

If you dig through Google Scholar, searching on pairs of possible crops and including LER in the search, there are all sorts of interesting studies that have been done in out of the way places. Unfortunately, not enough pairs have been studied, let alone combinations of 3 or more plants.

John Elliott wrote:

dan long wrote:

This whole "x plant seems to prosper next to y plant" is something i like to call "placebo companion". We observe improved growth because we are expecting to.

There is a way to quantify this. The concept of Land Equivalence Ratio (LER). If the LER is greater than 1, then the plants are good companions, making better use of the soil than they would if grown in monoculture. However, if the LER is less than 1, something detrimental is going on and they are not good companions.

If you dig through Google Scholar, searching on pairs of possible crops and including LER in the search, there are all sorts of interesting studies that have been done in out of the way places. Unfortunately, not enough pairs have been studied, let alone combinations of 3 or more plants.

sounds like a good opportunity for the *mad scientists* of this board to do some experiments. One could do a few sets 4x4 beds: "Y bed" "x bed" and "xy bed". I did very poorly in statistics. Does anyone know how many beds one would need to prove that results are statistically relevant? Since it's not all that detrimental if one does some irrelevant companion planting, 90% confidence is probably sufficient. What do you think?

Following John Eliott's suggestion, I did a search on google scholar. The last sentence of the summary was as follows:

"I conclude that garden-scale intercropping can offer advantages over monoculture, but these are not achieved simply by combining certain compatible companion species. Crop densty, ratio, and relative planting times all affect the way that companion species interact with one another and their environment."

Simply put: companion planting for maximum yield is much more complicated than "x likes y" (at least for Brussels sprouts or tomatoes with basil).

That being said: it couldn't hurt, right? Just don't get so caught up on what should grow next to what that you sacrifice time in the garden for time spent wringing your hands about what should grow with what.

dan long wrote: Does anyone know how many beds one would need to prove that results are statistically relevant? Since it's not all that detrimental if one does some irrelevant companion planting, 90% confidence is probably sufficient. What do you think?

I think you have now found the motivation to learn more statistical methods. The first place to start is with the Wiki on Factorial design of experiments. If you are OK with that, then there are several web pages that describe the process in more detail. You can even download slide presentations and lessons.

Of course, nowadays we don't just have materials to read through, there are all sorts of video presentations as well:



I wonder if Prof. Hunter softened his "pilot plant and production floor" outlook over the years and applied these observations to the natural world.

In answer to your question, each planting bed is one trial. If you have an experiment with 3 factors, you need 2^3=8 trials to get one data value. If you want to get average and standard deviation on those data values, you need 3 replicates data points. Now we are up to 24 trials. As you can see, the number of beds necessary to get this kind of data rapidly increases.

Try blueberries with potatoes.

Something I haven't yet looked into but was introduced to from a long term gardener was to think about growth and feeding needs of the plants growing together, too. I suspect it may play a large part in when plants 'like' to grow together or not, in some ways.

Basically, having the various plants set up so that their peak nutrient needs do not overlap. Like, say, one species ripening fast, one medium, and one slowly so that they use nutrients at different times but also get whatever benefits they can from each other - like shade, say. It would lessen the competition for nutrients, which allows the plants to exist together without one taking over.

I believe the three sisters has this, with the beans dying off, and leaving the extra nitrogen in the soil, just when the corn and squash are feeding more heavily. Not certain on that, but I believe so. Also, beans used traditionally in the three sisters in the SW USA hate 'wet feet' and the squash readily suck up any extra water, so that takes care of that need, as well.

shauna carr wrote:

I believe the three sisters has this, with the beans dying off, and leaving the extra nitrogen in the soil, just when the corn and squash are feeding more heavily. Not certain on that, but I believe so. Also, beans used traditionally in the three sisters in the SW USA hate 'wet feet' and the squash readily suck up any extra water, so that takes care of that need, as well.

I think that the beans were just plowed in at the end of the year. Personally, my beans didnt even start producing until after the corn was fully mature and now they are still climbing all over the dead corn stalks. That being said, they could have been using a faster bean and slower corn than i am.

John Elliott wrote:


In answer to your question, each planting bed is one trial. If you have an experiment with 3 factors, you need 2^3=8 trials to get one data value. If you want to get average and standard deviation on those data values, you need 3 replicates data points. Now we are up to 24 trials. As you can see, the number of beds necessary to get this kind of data rapidly increases.

And it's understandable why this isn't done much on small scale. Who would need/want 24 beds of any given vegetable? Then again... Maybe potatoes... hmmmm.

Appreciate the link to the video!