Modularizing Polyculture - The Lego Method

Hey y'all, I have been following the site for some time now but this is my first post on the forum. I am really interested in exploring the concept of modularizing polyculture in the same manor that children build complex creations with simple Lego building blocks. I can't seem to find any concrete literature on the subject and wanted to start an open conversation to see where it goes.

In a nutshell, the video below is a basic concept of what I am envisioning but in a more digestable, usable format than video, something like an info-graphic, like the "Hardiness Zone Map" which is easy to understand, even children can use the map to plan a garden...


Obviously every parcel of land is different and unique from any other parcel of land but generalities do exist, that is why we have conventions such as the "Hardiness Zone Map" which give you a rough idea of how a plant will fair in a particular geogrpahical region. It would be super awesome if there was a similar convention/tool for polyculture, so that anyone could look at their particular region and know typically what polyculture combinations would be well suited to the region. It seems like bits and pieces of this concept exist out there but there is not an aggregated version that is easy to use. About the only project I found that is addressing this opportunity was on Github (https://github.com/Permageeks/Polyculture-Design-Tool) but it looks to be an abandoned project. What if there was an Open Source tool that would allow a user to enter their location and then generate a planting plan based on the user's location and plant preferences?

So...
What are your thoughts about this? Does this exist already and I just don't know how to use google? What other features would be useful for such a tool? Anyone interested in starting such a project?

Joe,
I like this idea quite a bit, I have also looked for something similar with no real success.

At present, each project and location seem to be addressed as a one-off design. This is for good reason, as you note every site is different and the specific needs/requirements/goals are different. However, there are clearly some reoccuring elements that are incorporated by different designers for similar projects in similar locations. From what I've seen, this is based on what they like, what has worked, and what they want to experiment with.

Developing some type of "Lego Method" would make sense, if nothing else for a starting point that could be tailored to the specific site and project. I would think such an approach would be a huge plus not only for those of us amateurs, but also for real-live permaculture designers. Particularly those who take on projects in vastily different areas.

Suppose you took the idea off with a different twist: Let it be an academic exercise used in the teaching of permaculture. As an academic exercise, it could be one step in the process of a final design. Consider the following steps,

1. Identify the sorts of things that grow natively in a region, and what grows natively together. Around here, black locust and black walnut grow natively together.
2. Calculate the various needs of each of the plants. For example, each black walnut tree needs two black locusts to support it with nitrogen. This follows from Crawford's forest gardening book. If you surround a black walnut with four black locusts, you can then overlap two of them with the next nut tree in a line.
3. Weave the black locusts together with the walnuts in a mosaic-type pattern, a lego-pattern if you will, that can then be repeated ad-infinitum.
4. Add more variety, perhaps by replacing black walnuts with persian walnuts, pecans, or oak trees. Note that all of these trees are known for having deep tap roots. They also all produce edible nuts, although you might wish to be careful with your choice of oak. And, if that wasn't all, they are all valuable trees for their wood, both as firewood, fence-post wood, or as timber.
5. You could then add vines that would climb the black locust trees. This was another of Crawford's ideas. Consider Hardy Kiwi vines for this purpose. These vines need a little bit of shade while growing up, and a strong support--such as the black locust.
6. Repeat for the rest of the layers. (I haven't gotten that far ... although I have tried adding pawpaws, mulberrys, elderberries, currants, raspberries, and more.)

You now have a sample lego building block to start from. This is where the academic exercise would end.

7. Apply this to a piece of land, adjusting for property borders, boundaries, wind direction, water flow, etc. This would be the practical, as applied, exercise or final piece of the design.

Would this accomplish what you are looking for? And, for those more experienced than I, can you even get this far in any design without paying attention to water, water flow, ponds, etc.?

Love it. Lots of wonderful ideas being shared which is exactly what I was hoping for.

@Matt: you made a statement that really hones in on the idea of "Lego-izing" (yes I just made that word up) polyculture, "At present, each project and location seem to be addressed as a one-off design. This is for good reason, as you note every site is different and the specific needs/requirements/goals are different. However, there are clearly some reoccuring elements that are incorporated by different designers for similar projects in similar locations". One-off designing is inefficient but sometimes required due to specific site characteristics. Reducing the amount of one-offs and utilizing the reoccurring elements seems to me to be the key to this project. Also, I very much liked your idea that the Lego-izing of polyculture could serve as, "a starting point that could be tailored to the specific site and project".

@Dan: awesome idea! you said "Weave the black locusts together with the walnuts in a mosaic-type pattern, a lego-pattern if you will, that can then be repeated ad-infinitum" this is essentially eluding to using tessellations, which is what mother nature uses all the time, i.e. honeycomb, tree bark, flower petals, turtle shells, snake scales, etc. Tessellations add function, strength and structure simply through their form. I think this is a great jumping off point for the Lego-ization of polyculture. You also made a good point that their are external factors besides geographic location to consider and adds to the requirements and characteristics of successfully Lego-izing polyculture, "...can you even get this far in any design without paying attention to water, water flow, ponds, etc.?". I would say that any good design needs to account for all the water characteristics you mentioned as well as solar, soil, air, etc. However, I believe that these characteristics can be generalized in the same way the "Hardiness Zone Map" is generalized. I imagine that a good info-graphic for communicating what types of poly-culture work best in a particular region would at a minimum have to account for temp, rainfall, and solar exposure which could all be overlay-ed in some meaningful way.

So far the responses have been awesome and I have re-summarized the characteristics of Lego-izing polyculture below...

1. Focus on simplicity
2. Eliminate wasteful one-off designing yet maximizes gains
3. Use tessellations which add function, strength and structure through the nature of their form
4. Don't account for all parcel characteristics but instead focus on a few key indicators such as temp, rainfall, solar exposure
5. Serve as a starting point that could be tailored to the specific site and project

I envision that a user of this hypothetical Lego polyculture system would be able to enter their location and maybe a few optional characteristics such as "full sun", "shady", "lots of water", "dry", etc and get a planting plan for that particular region with site specific recommendations based off of the optional inputs.

I would love for this discussion to continue and for this project to take form and for other folks to add their comments, thoughts, opinions, experiences, etc...

It would be great if it was that easy!

There is also the question of scale to take into consideration.

Can you scale up something that works in a domestic situation to make it work in a farm?
Can something that works on a farm be scaled down for a household application?

I like the idea of being able to have an off the shelf solution and think it would be a great starting point but I figure it would need a fair amount of tweaking to get it to work.

Personally I like flow charts and dichotomous keys for this sort of thing but big posters are good too:)

The idea of the database to end all databases has been batted around for years. It is a brutally rigorous concept given the number of relationships, so it hasn't been done. Good design documentation would be a starting point.

Mollison in a early PDC documented by Barking Frogs Permaculture talked about four kinds of forests... Wild forests for wildcrafting and wildlife made mostly of natives. Food forests designed for yield. Forage forests designed to feed stock, and fuel forests, designed to produce biofuel. You can of course hybridize (a wild forest canopy with a food forest understory...) , but this might provide a nice high level framework as you start thinking about bringing a system to market using repeated patterns.

Jacke brings up patches in his Edible Forest Gardens, advancing the concept of guilding into a more structural view.

We need to design disturbance (i.e. management) into the component (if for no other reason to support more labor efficient harvest and management. The disturbance regime is a major piece of a component design, following the critical role of disturbance in defining the structure and composition of natural system.

Following this, I think of three kinds of patches in my food forests (driven by my climate): 1) patches that you plan on cutting in may and using as mulch (maybe early spring yield.. mint etc.) 2) patches that get some tending (receive some of the mulch from the cut patches). and 3) patches that get severely distrubed and resown (weed gardens with root crops, or for chickens or pigs.)

Thanks for starting a design oriented thread.

One of my current mulch patches in full sun is build of mint, comfrey, oregano, and I am experimenting with which lupine can take annual cutting. I produces a lot of biomass by early summer, herb cuttings, and then is cut down and used to mulch a sub-shrub/forb area to reduce competition (sage, rosemary, iris, good king henry, echinacea and garlic).

Mark Livett wrote:
There is also the question of scale to take into consideration.

Scale also has a lot to do with the canopy strata you are working in... a ground layer patch can be 10s of square meters, while a forest canopy patch might be 1000s of meters.
I think scale also has to do with management regime.. as you scale up you need to think about MUCH less labor per square meter, and so disturbance becomes less 'surgical' and more broad swath. As treatment becomes less responsive to the system, you loose some control. This risk/fear of loss of control can lead back to more single yield rectilinear engineered systems.

Regarding scale, the example I gave above that discussed creating a forest of nut and black locust trees was proposed for the following scenario: suppose you have between ten and one hundred acres that you need to turn into a permaculture landscape in short order. For the sake of discussion, let's say these acres were a chemical farmer's field or an animal pasture before. Perhaps the land was denuded as a result of a forest fire. For whatever reason, there's nothing on it today. The land is now yours. You have minimal dollars to work with. You also have minimum time/energy for these acres--you want to focus elsewhere. In other words, we're not talking zones one, two, or three--these are the zones you would have farthest from your center of energy. You want something that takes care of itself, that heals itself, that works itself, but you don't want to put a lot of effort into it.

Given that scale, I figure the acres in question can be covered in one season for less than $100/acre of startup cost. (Space the black locusts on a 62' grid, nuts in between the black locusts--the acres add up pretty quickly.) Planting on such a large scale wouldn't be too difficult, but nursing your young trees and bush hogging the fields during that first year or two until the system starts taking care of itself might be a bigger problem. You may need to rent the bush hogger, and do some careful work near your planting holes to make sure what water you do get doesn't run off every time it does rain.

Paul Cereghino wrote:The idea of the database to end all databases has been batted around for years. It is a brutally rigorous concept given the number of relationships, so it hasn't been done. Good design documentation would be a starting point.

I've looked into the practicality of this a little bit. There is an abundance of data about specific species, but non of it is formatted in any uniform or practical way. I am sure there are thousands of beneficial relationships that are yet to be discovered. Acceleration of those discoveries would be dramatically advanced with a uniform database. There is probably a billion dollars worth of increased GDP lying under that solution, but it is a difficult one to solve because of the politics.

Everbody looks at the various problems different ways. PC people, vs. biologists, vs. environmental modelers, vs. conventional farmers etc. It is mostly the same information, but none of it works together.

Both the FDA and the EPA have grant programs, as well as the various state governments. The important thing is standardization. Your best bet would be start a dialog with a standards orginization like the ISO, and start developing a format for the database, while consulting the Agri departments of various universities. The trick to this project is getting the various institutions on board, so _they_ populate most of the data. Like I said, it is mostly a political problem, not a technical one.

what I did in Permaculture regarding "plants" was to make a list of every plant that I thought might grow in our area that I didn't find totally repulsive in growing..and then I alphabetized the list.

then I took the largest thickest spiral bound notebook I could find and i started with A at the beginning..and looking up everything I could on say Apples, or Asparagus..

I would make notes of everything I could find out...number of days from planting, type of soil, sun or shade, nutrients needed, nutrients supplied, height, spread, best time to plant, when it would be harvested..years to harvest of things like trees, differences of different varieties (such as things like summer squash, winter squash, spaghetti squash, pumpkins, etc.) things that will cross, thigns that won't cross,. ph, best mulches, no mulches, best fertilizers, better without fertilizers..prune or not, every tiny little thing I could think of..

Honestly i ended up having to do this 3 times getting larger notebooks each time..to where now I'm happy with mine..I have photographs or tags from the trees or plants I have planted taped or glue in also.

I have gone from the tiny little plants to trees and even have some non food things listed in the book and herbs and beneficial weeds..etc.

i have also listed some things I haven't grown ..may..or may not...and also some wish list items that I would hope to grow..some failures..etc.

I have also tried to use some colored pens or markers to highlight some things to make them easier to find....like I have clay soil so if something loves clay soil it might be written "loves clay" in red in every plant that loves clay..etc..

I also have a lot of shade..so shade lovers are highlighted. I've been trying to concentrate on protein plants, so I've noted protein contents..etc..

I'm sure this isn't what you were seeking but it sure helps me a lot.

I have a similar list, but I didn't make it myself. Martin Crawford publishes one in his Creating a Forest Garden book that I have been referencing. Indeed, the plant data offered in that book probably consumes nearly a half of the volume. Admittedly, it's not all my climate and I'm not certain of where to get all the plants listed there, but he has more plants listed in his reference list than I can get a practical grip on. However, this is only part of the puzzle of a lego method to permaculture design. Just having the database doesn't answer the question of what to put next to what, what polycultures to try, how far apart to try various different items, what the mixture ratios should be for nitrogen, potassium, and phosphorous producing plants, and so forth. (The rest of the volume develops those principles.) It's not an end all, but it is the one formal list I have been working from.

I have the benefit of this texas database has associated species at the bottom of plant info. Sub edible forms of wild species and I have a database. Texas has a diverse climate and this database could be useful for anyone below 6000' from zones 7a to 9b.

That TX database has been very helpful for me to figure out what to plant beneath my Cedar Elms, among my Texas Persimmons and Gum Bumelias, etc.

For what it's worth, the USDA has a neat search tool that allows you to search plants by type, rainfall requirements, pH requirements, coppice ability, nitrogen fixation, and a ton more categories...


http://plants.usda.gov/adv_search.html

There are some really great ideas flowing and I can see the original seed idea of Lego-izing polyculture starting to sprout. There seem to be 2 themes emerging from the original post, 2 branches if you will. One branch focusing on a simple info-graphic and the other branch focusing on a comprehensive relational database that accounts for all parcel characteristics. Perhaps a simple info-graphic would have to ultimately rely on a comprehensive database, perhaps not, or at least every characteristic need to be accounted for.

In the hopes to spark more discussion regarding the creation of a simple info-graphic similar to the USDA "Hardiness Zone Map" I propose the info-graphic below to you...

Here is the link in case the image doesn't embed in the post http://postimage.org/image/et9nhf851/

Essentially, I took the basic color spectrum, "ROYGBIV", and color coded 7 Groups, A-G. By going through this exercise I started to think that maybe this should be placed into a series of concentric circles to indicate layers radiating out from a core group with each group building upon the next. The groups are completely undefined at this point but maybe Group A would be something like hardy grasses and cover crops that fix nitrogen and "till" the dirt for you; and these should be planted everywhere in the garden because it is the core group A and maybe each grouping requires 2 or 3 different species to be present to add diversity, resilience and strength. Group B might be focused on deep rooted plants that bring water and nutrients up from the deep and should be planted in ample amounts around the garden. At the other end of the spectrum, Group F, could be water/resource intensive or specific plants such as demanding fruit trees that need lots of room to grow comparatively and. The groupings seem to be a simple way to account for characteristics in a easy to read manor. The groupings are completely open for suggestion. Maybe 7 is too many groups or maybe its too little, but I think ultimately simplicity is best and more easily digested, plus the #7 is cool

Addressing the other branch of this thread, I think that a comprehensive relational database is a really ambitious and super, awesome sauce idea and would be a huge asset to the world. With the cool "Big Data" spelunking tools out there like Tableau it would be pretty damn impressive to be able to play around with such a theoretical data set in such a visually compelling way. The most challenging part would be populating this theoretical database and forming a solid data model. There seems to be an abundance of narrowly focused data out there that could initially seed the database and relationships could be added at a later time. Or maybe the way to solve this problem is through crowd-sourcing and creating a website where users enter in data for their plot and that data could be generalized over a geographic area once enough data has been built up. Or maybe it could be a blend between research data and crowd-sourced observational data, at the very least this approach would offer a chance to compare predicted to expected results for a particular region.

In general lots of great ideas have been shared and I encourage more to share their thoughts about Lego-izing polyculture so that ultimately a child would be able to plan a thriving polyculture garden.

This problem is not as hard as we are trying to make it seem.

Most of the data is already there, we just need someone to build a centralized place that is user friendly.

So 1st we have to define what user friendly is. Who is our user base.
Mainly just tomato/vegetable growers on less than 9,000sqft, followed by a few families who just want to provide everything on 2 or so acres. Out of the 10,000 permaculture less than 5 of them is a big farmer with 15 or more acres selling making a living selling apples/oranges.
So our userbase is at best a family on 2 acres.

What is such a family looking for:
They want to provide nuts, fruits, vegetable, firewood, honey, fish, chicken, red meat, milk cheese. Do we want to also include how to make a house/sewer/solar/heating system?

How do we get them to that goal. 1st we have to find out where they currently are.
Such a family is most likely not in the wilderness on forest land, they are in suburbia/old level compacted farmland.

We should 1st build this simple system then only after we have done that move to more loftier level such as desert land/etc


Designing for more complex systems answer questions like.
1)How much land are you going to manage (2 acre or less)
2)How much man/machine hours do you have to manage it (0.5hrs/day/acre to maintain)
3)What layers of the food forest are you focusing on. (12ft or less aka fruits and vegetables, quality vs quantity)
4)How much earthwork/fertilization do you have to do (rainfall distribution, soil type, budget)

Who is computer savy that would be available to help out in this. Not necessarily asking for someone to lead it.
DB/php/css/graphics people.

Lets make a running list of providers of good data

We can overlay location on google map or maybe http://www.mytopo.com/maps/index.cfm with their awesome terrain data is even better.
I know that we can pull climate data from http://planthardiness.ars.usda.gov/PHZMWeb/# or use these guys to calculate our own http://help.wunderground.com/knowledgebase/articles/129139-how-do-i-pick-plan-or-level-

We can also check out soil type, however I dont know if they have a API something to check out. http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx

We make our own solar calculator or use
http://www.wunderground.com/calculators/solar.html

Plant data from: perennialvegetables.org/perennial-vegetables-for-each-climate-type/cool-maritime/,

Cultivar data even better than plant data: http://www.onegreenworld.com//product_info.php?cPath=1_15&products_id=279

Vendorlist data: http://www.onegreenworld.com//product_info.php?cPath=1_15&products_id=279 (PNW) , http://www.tradewindsfruit.com/fruits_ornamentals_by_hardiness.htm (seeds only, zone 10) , http://www.toptropicals.com/cgi-bin/store/store.cgi?group=fruit (zone10), http://www.sln.potsdam.ny.us/2013lcataloglores2.pdf (zone 4) , nativeseeds.org (seeds only, SW)

Hey, there are already several initiatives that are well underway providing the database part of the service you're talking about. Don't reinvent the wheel -- help them out!

Practical Plants at http://www.practicalplants.org/ is a great start, but they need volunteers to finish entering the data. Natural Capital Plant Database, currently at http://oldsite.permacultureresearch.us has gone for-profit and is not as open to updates, but the quantity & quality info they have accumulated is excellent, on a par with PFAF.org but optimized for the central USA. (disclosure: I helped write the software for this latter, years ago, when I didn't know much about Web programming... but I can't claim any credit for the data.)

As for the lego blocks... last year I took a first stab at writing the necessary software to suggest plants for a given plot of land, given its hardiness zone, water & soil type, etc. plus what other plants are already planted there. The result is effectively a guild builder app that works on mobile devices as well as the Web, so you can tell a client in the field what else they should plant in a location to support the plants they have. But I got busy with other projects most of this year and didn't have time to refine it. You can see an un-refinded example here: http://interdependentweb.com/designs/black_walnut_guild

I'd love to roll the functionality into another database such as Practical Plants so that I don't have to re-enter the data for each plant. I'd also love to add a graphical component so you can play around with arranging the plants and their connections, using this Javascript module: http://www.jsplumb.org/jquery/draggableConnectorsDemo.html . Perhaps a Kickstarter is in order?

That is an awesome looking jquery applet, and good links that you provided.

I am not trying to reinvent the wheel. I would much rather connect to other website through API.
However someone who is really good at PHP might not know of a good source for any soil data.
Or there might be a better/easier one to connect to, and as of such it is good to have multiply data source, thus the list.

If the PracticalPlant guys will allow you to it would be awesome. If not it would still be good if we could connect or import their data.
Sanitizing it is still alot of work though.

Another good reason to import or own data is because at any point someone else data could move from open source to close source.

Things are looking good, we have one more computer guy who might be able to help.
Is there anyone else, or do you know of anyone else who might have a couple free hours, please spread the word.

I was trying to figure out how to get 1/4 to 1/3 spacing/layout for N-fixers.

This is what help me.
http://en.wikipedia.org/wiki/File:Bond_brick_hexagonal_tiling.png
http://en.wikipedia.org/wiki/File:Hexagonal_tiling_4-colors.png
http://en.wikipedia.org/wiki/File:Uniform_tiling_333-t012.png

So does anyone know a equation to get this.
I would also like to see this built into the software/webapp.

Really great information everyone! Especially Bengi you are on a roll dude! You also bring up some good points about first defining what easy to use means and who the user base would be, you said, "So 1st we have to define what user friendly is. Who is our user base?". I would say that a user friendly interface would be simple and familiar. That is why I am suggesting an info-graphic like the USDA "Hardiness Zone Map" which is easy to read and comprehend due to the color coding which provides a sense of temperature, red is hotter temps and cooler temps are blue. The map doesn't address every situation or but provides the user a good understanding of what plants are well suited for their location. Ultimately we are trying to answer the question of "What should I plant in my garden?" based off of the user's location. The user base would consist of mainly those interested in polyculture but the user base could potentially balloon if it was simple and easy to use so that even a child could plan a garden using it. Lots of potential.

So more on the running list of features data providers.

Plant data at http://www.practicalplants.org/
Degree Growing days per month based on location. http://www.weather.com/outdoors/agriculture/growing-degree-days/02121
This is good for vegetables. It would also be good to have a cultivar list esp for short season (50day tomatoes) and for arid regions (110F, low water, deep tap root).


quote=S Bengi]Who is computer savy that would be available to help out in this. Not necessarily asking for someone to lead it.
DB/php/css/graphics people.

Lets make a running list of providers of good data

We can overlay location on google map or maybe http://www.mytopo.com/maps/index.cfm with their awesome terrain data is even better.
I know that we can pull climate data from http://planthardiness.ars.usda.gov/PHZMWeb/# or use these guys to calculate our own http://help.wunderground.com/knowledgebase/articles/129139-how-do-i-pick-plan-or-level-

We can also check out soil type, however I dont know if they have a API something to check out. http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx

We make our own solar calculator or use
http://www.wunderground.com/calculators/solar.html

Plant data from: perennialvegetables.org/perennial-vegetables-for-each-climate-type/cool-maritime/,

Cultivar data even better than plant data: http://www.onegreenworld.com//product_info.php?cPath=1_15&products_id=279

Vendorlist data: http://www.onegreenworld.com//product_info.php?cPath=1_15&products_id=279 (PNW) , http://www.tradewindsfruit.com/fruits_ornamentals_by_hardiness.htm (seeds only, zone 10) , http://www.toptropicals.com/cgi-bin/store/store.cgi?group=fruit (zone10), http://www.sln.potsdam.ny.us/2013lcataloglores2.pdf (zone 4) , nativeseeds.org (seeds only, SW)

After someone enters their zip code. It would be nice to be able to choice plants based off, USDA/Sunset plant zone, rainfall amount and distribution, AHS/growing degree days, soil type and topo/drainage pattern.

Enter what they would like to plant 10ft fruit/nut trees, vs 1ft vegetables, mix

They could then enter how much space they have to garden.

Based on this we could display a layout with appropriate guild/summer vs winter crop/root competition/N-fixers (25-40%)/lbs of vegetable/fruit produced.

They could then click on each square and change displayed choice things around.
Maybe a scale to pick how exotic they want to be. (Irish potatoes vs Chinese yam)

It would also be nice to select squares that shaded vs full sun.

Maybe tell the app,must have plants.

All of this can be done with data on just a few plant(400?).

Nuts (
Winter no processing store-able fruits (7) apple, persimmons/etc
Winter no processing store-able vegetable (30) roots, cabbage family, beans, squash/etc, rosemary
Early spring crops (5) silverberry, asparagus, favabeans?,

Once we start the app we can always add more logic/functionality to it.

Figuring an average planting space of 1000 sqft against a fence (100*10).
We could plant 3 rows. One for vines, 1 for fruits and nut, one for root/vegetable crops.

Thats around 24 sub-trees/vines and 100 vegetables spots. Thats alot of data to display.

The plant database above have about 74,000 plant even after factoring all the site limits
My guess is that we would on average have 30,000 possible plants. how do we limited this down.

Benji the links that you are providing are really great. I particularly like the Practical Plants website and their motto "Know the spot, but not the plant?" and they seem to have provided an answer to the question of what to plant based on location. You brought up a good point about limiting the available options and my response would be that there must be a purpose beyond pure aesthetics for a plant to be in the garden. A plant must provide food, medicine, fiber, fuel, soil nutrients, be food for an animal living on the land or attract pollinators. Also I would contend that any plants should ideally be well suited for the environment and consume minimal resources. Ideally a garden would consist of native species. Almost all climates are capable of providing fruit and vegetables. Even deserts can sustain olive and fig trees, with a little human intervention, Geoff Lawton's "Greening the Desert" project highlights this. Also another method to limiting possible combinations would be to focus on well known combinations such as the "Three Sisters", squash, maize and beans, at least initially.

Have you ever used Tableau Public? If not please check it out, I think you would really dig it. Its a data visualization tool that is really simple to use but incredibly powerful. Check out this visualization of food deserts in California I made using government data...
http://permies.tableausoftware.com/views/fooddeserts/ofpeoplewithlowaccesstoasupermarketorlargegrocerystore?:embed=y
Areas colored brown have a high number of people with low access to fresh food, which is considered a food desert, while green colored areas have a low number of people with low access to fresh food. I think that if we had a data set to work with we could use Tableau to make some really compelling visualizations. Thoughts?

All of the 74,000 plants are useful as you described and not just for looking pretty and that 74,000 doesn't even list cultivars so that number would go uo by at least 10.
However like you said it would make more sense to start off with a very limited amount of plants (400) that are recognizable to most if not everyone.
Ben Stalling was the one who brought up Practical Plants/Plant for the future I had all but forgotten about them.

So I think that if we had the graphics of Practical Plant but only the limited data of onegreenworld, nativeseed, perennialvegetables.org thing would be off to a nice start.

I checked out the Tableau+ Public, you made a cool map overlayed with food production per capita.
The website itself has a diverse set data provider and gives a user alot of ways to display that data and apply logic.

In my head I was not going to give users a "map" that they could easily explore by zoom/pan.
Instead the very 1st thing they had to do was enter zipcode and that was all they would ever see plant data for.
However I could see where we give them a bigger map listing data only for a worst case senario. 12inch rain, zone 5, only a 20ft x 20ft garden, etc.

As long as you get 12 inches of rain or capture 12 inches of rain (river/stream bed, rain barrel, graywater, etc), you can grow alot of stuff.
Have you read dryland farming, its a very good book. With deeper root cultivars and wider spacing we can still grow in desert.

S Bengi wrote: This problem is not as hard as we are trying to make it seem.

Quite. It is actually much harder.

Broad adoption is contingient on having people with many different backgrounds find the provided information useful, AT THE SAME TIME. Taxonomy by itself is a huge problem. Any given organism with have a latin name, several common names in each of dozens of languages, and possibly brand and patent names. That is before you even consider defining relationships to other organisms.

If you want broad adoption, you will need to have some sort of standard that makes sense to to most of the disciplines that will be providing your source data. This is not just a data storage and access problem, but a linguistic one as well. Everybody has a name, and subsequently everybody thinks they understand naming systems. The truth is quite different, and if you need any proof you just need to look at the corrosive politics that sarrounds the Internet DNS.

If you were to define a spec for the database, one of your first steps is to define WHO you want to use it. Get to the end of that document, and then consider how those people approach solving problems in their discipline and you will just be starting to understand the scope of the undertaking. At this point, (if you get that far) you will, A: start beating yourself with a hammer, for even considering taking on this project. B: narrow the scope of the project to make it MUCH smaller, and thus have just another of the many incompatable databases that already exist. C: start looking at well established systems that already support _some_ of the features you want, and try to figure out how to bastardize them into doing what you do want.

I applaud the enthusiasm. I do code, and I would be willing to help (a little here and there) if any of the underlying tools you choose jive with stuff I have experience with. But you need to be aware, if your goal is to achieve broad adoption, then this project is going to be EXPENSIVE. I would say $1M wouldn't be an unreasonable funding goal. Referring back to my earlier post, in terms of gain to the national and world economies, such a system would certainly return dividends for many decades to come, and so that kind of funding might be achievable. If I was a student at a University that had a well respected AG program, I would seriously consider getting some fellows from the I.T. side of the house on board, and solicit some grants.

What your talking about really, is taking the the database that the FDA already maintains for their seedbank, and dramatically expanding it to support modern scientific research methods, some of which would related to permaculture in the form of a standardized polyculture data repository.

Just a few thoughts.

We should limit the user base to
(1) 20ftx20ft vegetable gardeners..........
(2) someone who want to add 20+ fruit trees to their lawn along the fence............
(3) homestead(2 acres)
(3a) intensive 1 acre food forest (180 trees @ 15ft center + 1living fence)
(3b)1 acre goat/sheep/dwarf cow pasture (+1 living fence +4 huge trees for shade +4 ponds for water) I would not expand the scope beyond this .........

According to the research done by biointensive we could entirely feed a vegan family of 4 on just 1/2 acre (5,000sqft x4) So a 2 acre homestead for a family is more than enough. Anything above that is unsustainable. And most 100+ acre farmers are only going to sell three or so crops, so if anything they are even easier to design for.

http://growbiointensive.org/grow_main.html


I envision someone going to the site picking one of those option plus climate data and then we give them a layout.
If someone click on option 1, they are then presented with a 20ftx20ft garden they could then change the shape rectangle or square, bigger (30x20) smaller,the option of vegetable that we give them
If someone click on option 2, they get option 1 plus also fruit/nut trees.
If someone click on option 3, they get vegetable, +fruit/nut, +honey, +fish pond, +chicken/etc, +cattle

If the users want technical data let them visit the FDA seedbank database. All I want to present is something graphical and pretty.
Something along the lines of this, except it comes pre populated with plants, swales, ponds based on terrain data, climate zone, soil type etc . http://www.homestyler.com/designer

Now inside the app if they select a Walnut tree a mulberry tree automatically comes with it.
If they they right click on a tomato it will give them days to maturity(100days) yield, etc

I agree that bio intensive seems like the way to go and that the system should be geared towards less than 10 acres. Also taxonomy is indeed a linguistic challenge but i believe that the system should initially focus on common species and later add more exotic species. I can't seem to find any downloadable databases to work with. If someone knows where I can get my hands on a database I have some basic visualizations that I want to start putting together using Tableau.

This seems to be an area where a lot of people have trouble. Really, polycultures are the easiest thing to implement. People always love to make things more difficult than they need to be. To start you need a good seed mix. This mix will do four things: make biomass, fix nitrogen, attract or repel insects, and loosen the soil with either a deep taproot or a deep fibrous root system thus mining nutrients. I like to have at least 3 different plants for each category giving me a minimum of 12 plants in the seed mix. However my current seed mix has over 30 plants. This mix is intended to rapidly improve the soil while also feeding animals in a paddock shift system. Here is my current seed mix:
mustard
burdock
alfalfa
lamb's quarter
fava bean
sweet clover
lupine
landino clover
buckwheat
hairy vetch
daikon
black-eyed peas
comfrey
sun flower
yarrow
borage
chamomile
dandelion
turnip
bee balm
lavender
mullein
pea (pisum arvitiuse)
black locust
acacia mearnsii (black wattle)
red alder
empress tree
siberian pea shrub
stinging nettle
chard
maximillian sunflower
sorghum

All of these plants fit into the above four categories. Most of them are annuals but there are a few perennials and trees. to plant simply broadcast seed throughout the garden/farm area. If you don't have a lot of seed to work with you can sow more carefully but don't do straight rows sow in a sinusoidal (wave like) pattern. This increases edge and creates diversity. The more diversity there is the less problems you will have with pests, disease, and nutrient deficiencies.

If you proceed the broadcasting of your seed mix with earth works (swales, terraces, raised beds, and ponds/dams) the need for watering should be eliminated or minimized. You can now add your "production" crops like tomatoes, melons, and lettuce. To plant these crops amongst your seed mix simply chop and drop or if your productions crops are too weak to complete you can cover an area with burlap, place a couple of inches of compost on top, plant and then cover with a thick Ruth Stout type mulch. Make sure to plant plenty of trees, shrubs, and perennial veggies. Trees make up the heart and soul of a food forest so don't be afraid to go crazy on the fruit trees. If you are farming for a livining fruit trees serve 3 purposes 1. they of course produce fruit for you to eat 2. they produce food for live stock, and 3. they can be sold ( a large bearing fruit trees can easily sell for a few hundred dollars).

So there it is, polyculture can be as easy as you like it to be. To review 1. earthworks to eliminate watering and conserve/create soils 2. create and broadcast a seed mix with at least 12 members from the four categories. Match them to your climate. 3. add your normal crops like tomatoes, garlic, and leeks (whatever you like). If the plants can't compete with the polyculture chop and drop or use burlap, compost, and mulch to create an area for plants that can't compete.

James Colbert wrote:This seems to be an area where a lot of people have trouble. Really, polycultures are the easiest thing to implement. People always love to make things more difficult than they need to be. To start you need a good seed mix. This mix will do four things: make biomass, fix nitrogen, attract or repel insects, and loosen the soil with either a deep taproot or a deep fibrous root system thus mining nutrients. I like to have at least 3 different plants for each category giving me a minimum of 12 plants in the seed mix. However my current seed mix has over 30 plants. This mix is intended to rapidly improve the soil while also feeding animals in a paddock shift system...So there it is, polyculture can be as easy as you like it to be. To review 1. earthworks to eliminate watering and conserve/create soils 2. create and broadcast a seed mix with at least 12 members from the four categories. Match them to your climate. 3. add your normal crops like tomatoes, garlic, and leeks (whatever you like). If the plants can't compete with the polyculture chop and drop or use burlap, compost, and mulch to create an area for plants that can't compete.

James I think you hit the nail on the head! Seed mixes create diversity and resiliency with minimal effort. I am totally down with working smarter not harder. So I think you have really steered the direction of this post in a new way. The solution seems to be to making a seed mix that is suited to the location of the user and then planting fruit trees and annual vegetables throughout the land. I really like the idea of using a wave pattern to broadcast the seeds to increase edge areas as well. James what fruit trees have you planted with the seed mix you listed in your post? What hardiness zone is your seed mix designed for?

Perhaps the way to Lego-ize polycutlure would be to have the user input their location then be presented with a list of fruits and vegetables well suited for their location. The user could then select which fruits and vegetables they like and a seed mix could be generated based off of their selections. Simple. Love it!

Probably a better question is to ask what plants will not work. The list i provided will grow in many areas. I am currently in zone 7 I have long hot summers (over 100) and cold winters with snow. Perhaps 3 general seed mixes for dry, temperate, and cold environments are needed. But I think if you start with the list I provided your will do well in most locations. I have grown apples, pears, persimmons and many more with this mix or part of it. Really it doesn't really matter.

If you use the four category method I detailed you should be fine. Just pick 3 different plants for each category. Ideally each plant will fit in multiple categories. For example black wattle trees are nitrogen fixers, they grow fast and produce a lot of biomass , they have a deep root system and they attract beneficial predatory insects. If you don't know which plants will work in your local simply do a google search. Sepp Holzer provides seed lists in his book for cold climates. The list I gave you will work in temperate climates so all we need is a dry climate mix which I am sure you can find in something by Mollison.

James Colbert wrote:Perhaps 3 general seed mixes for dry, temperate, and cold environments are needed...If you use the four category method I detailed you should be fine. Just pick 3 different plants for each category

James thanks for narrowing this down even more. I agree 3 seed mixes for dry, temperate and cold environments seems to be the simplest solution to Lego-izing polyculture. What is really cool is that you have suggested a Lego strategy of your own with the four category method by choosing 3 plants from each category. This allows users to easily build a garden to their own specifications that accounts for their local climate.

Does anyone else have experience with the method that James has suggested? If so what has been your experience? Please detail the Good, the Bad and the Ugly. Could this system be optimized even further? Looking forward to everyone's suggestions! Thanks for the enthusiasm and candid responses.

This has been a great post.

What I got so far.

Plant Database
Grid layout to get 1/4 to 1/3 N-fixers in a food forest
List of support/pasture plants (1. N-fixers, 2. biomass, 3. pest control, 4. Aerating root
List of regular vegetable cultivar for low water usage, short time to yield,
List of perennial vegetables
List of common Nut/fruit/shurb/vine trees including n-fixer
so our database is is now full.

Local Condition
TopoMap
Plant hardiness Zone
Rainfall
Length of growing season
Soil Type

What Level of permaculture
Vegetable
Fruit/Nut Lawn
Homesteader.

I think we have most of what is needed defined now we just need to build the app (database, then logic, then display/user interface).

I think the general aim is a noble one, and there is scope within these ideas for a much needed and potentially very useful tool - something which lowers the bar to entry for newcomers by providing a basic set of building blocks to use and build upon, and perhaps even makes life a little easier for people with experience. Something which provides tools to help people design. For that, there is always need.

However, I think a lot of people have a dream of technology solving the complexity of permaculture: that is, that with a clever enough application and big enough database, you can enter your location, your gradient and orientation, click a button and have a system designed for your ready to plant. I'm very wary of this goal for two reasons.

Firstly, is it possible? To attempt to model the complexities of a permaculture design is to attempt to model nature. The human mind is an incredible tool for mapping associations in a way which a computer program is not. How can one provide a computer program with all the knowledge of subtle variations in soil quality and acidity; with frost patterns and natural shelter; with the areas where a certain pest thrives or the family dog likes to dig; all information which comes easily to the human mind, and which it can use to build a map of incredible complexity. To provide tools to help the mind is a wonderful goal, but to attempt to outsource the complex planning decisions that a human mind is so good at to a computer is, in my opinion, folly.

Secondly, do many people actually want such an application? Planning and designing and learning the skills to build and maintain a designed natural system are, for me, the raison d'etre for my interest in permaculture. The idea that my forst garden or no-dig polyculture vegetable garden should be an impersonal affair designed by a computer is awful. The joy is in the design; the incremental improvements; the lessons and the skills learned.

What are your guys' thoughts on that? I'm totally supportive of efforts to build new tools, and am eager to be onboard as long as the goal isn't a mind-outsourcing application.

Either way, I'd be happy to work with you in providing access to Practical Plants data via our API. It's an open data source and it will always stay that way. Our API is the standard Semantic Media Wiki ASK API, but at the moment our data model is undocumented so it can be a bit bewildering if you don't know your way around our system. I'd like to work on fixing that by completing some documentation, so this could be a good opportunity for me to do that.

At the moment our plant interactions and polycultures datasets are in their nascence, so there's not much Practical Plants can provide to this project there yet. Once we've completed a drive to do some article cleanups, our next drive might be to work on importing interactions data and setting up some well established polycultures. Maybe that's an area for collaboration - if you're going to have to enter it somewhere, you could join in our effort to get it up on Practical Plants.

Anyway, looking forward to seeing where this goes. Happy holidays!

Andru (PracticalPlants.org)

I am glad that your database exist and that we can use your api it does make life easier.
If your were to think of permaculture in a similar vein as linux I would like this app to be like ubuntu.
Where it will not magically build you a server cluster, if you dont know anything, but it is something my grandmother could use without compiling source code.

So the app will give the user a good enough template and the user will have to redefine the system.
The app will start out very simple and hopefully as the years go by, it will gain more complexity. (eg. computer vision to locate existing shade and arrange citrus vs spinach)
The app would mainly just focus on 1/10 acre city dweller looking to plant some veggies or add some trees to a lawn. Not a 10 acre plot with multiple ponds.

S Bengi - have you come accross growstuff.org? They're a new group building an open source social gardening platform with garden planning and tracking features. They're mostly veggy gardners, afaik the platform doesn't have any permaculture specific functionality, but otherwise it sounds like you're describing a lot of what they're building. Maybe you should float some of the permaculture planning ideas with them and see if it can be merge... Web developers interested in gardening and especially permaculture are a rare breed, the more we divide the less we're likely to achieve

Anyway, I love where this thread starts: the idea of creating sets of permaculture components like lego bricks, which the user can integrate into their garden: tessellating permaculture components. I'm really into that idea; I think it's genius is in it's simplicity, and I'd be totally up for throwing some weight behind it. It doesn't need to be a hugely complex application, it just needs to be a tool to guide the designer. I can imagine an application which works as such:

* User enters the plants they know they want in their design
* User enters the type of plants they want in their tiles (annual/biennial vegetables, small perennials, shrubs, trees)
* User enters the tile size they want to work with (eg. diameter in meters)
* The application generates a list of recommended plants, and the user can set their preference on a sliding scale of 1-5 or so
* The application creates a set of individual tiles which tesselate. The user can either print these out as they are, or they can drag them into a design, and print that.
* The user can optionally save and share their tiles with friends and so on
* The user can feed-back on the tiles they used, providing information on any problems encountered, to help improve future tiles.

Things to consider about even a simple application like this:
* Data requirements are high - it will require a solid data foundation of chemical, spacial, nutritional, animal (and so on, the ways in which plants interact are endless) plant interactions OR a very complex database of multitudes of human designed polycultures (for every climate zone)
* Possibility for error - if these tiles are automatically generated using plant interaction data, it's very possible the data may match plants which do not perform optimally as a polyculture. Even human designed polycultures must be well tested in a variety of environments before you can be sure it works.
* Feedback - user feedback is always tricky. Perhaps their tile totally failed because they planted or managed it badly, because it was a bad year, because they were unlucky, or because the plants in the polyculture don't interact as well as planned. Picking out the signal from the noise when working with a large number of automatically generated tiles will be very difficult.


Ha, the more I write the more I find problems with the idea... I wonder whether it wouldn't be better to just come up with a website where people can create their own tiles manually and provide ongoing information about their success, allowing others to upvote or downvote the tile design accordingly, and to plant in themselves and provide their own feedback on it as a design. Eventually the site would generate a list of the best tile-based polycultures hand-tested by a community of gardeners, which others could use in their design.

Actually, I prefer the latter idea. It'd also be much simpler to build leveraging existing open source technology.

Anyway, those are my 2c. Excuse my rambling. Let me know if you find anything interesting here.