Harvest log update, February 2012

I've updated the harvest log website. I haven't posted about the harvest log since last April. Some unorganized thoughts follow:

• Our average daily calorie harvest hasn't changed much, still right about 670 calories per day. We still have perhaps 10-20 pounds of honey and maybe 3 pounds of fennel seed which we harvested but haven't weighed and entered into the database yet, which will boost the calories.
  
• We've been harvesting fewer ounces of food per day than before, but with similar total calories, meaning we're harvesting more calorie crop - mostly a result of increased honey and egg harvest.
  
• None of the fruit trees we've planted yielded last year, except the medlar in heavy shade with a little over 2 pounds of fruit. We did get twice as many cherries from our front yard seedling cherry as we did the previous year. We were a bit surprised and disappointed that the persimmon tree which gave 8 1/2 pounds in 2010 didn't yield in 2011. Our region had poor fruit harvests in general due to weather causing poor pollination. Hopefully this year the fruit trees will really start to produce, after 5 years in the ground!
  
• Great yield from our goumi, which gave 12 1/2 pounds.
  
• Much better yield from the strawberry patch, which gave 17 pounds in 2011 compared to 6 in 2010. Probably because Tulsey thinned out the plants in winter of 2010/11. The patch is about 80 square feet.
  
• Bad raspberry year (11 pounds compared to 31 in 2010), as most of the plants in our original planting died (probably from root rot or some other disease), plus we didn't water the living patches enough to get a good fall harvest.
  
• Our english walnut and two or three of our four hazels set some nuts, but all were taken by squirrels and jays. Unfortunately, we didn't eat any squirrels from the yard this year to offset that loss. We did get one tiny chestnut!
  
• Our harvest of greens dropped off a lot last year, partly because we were wrapped up in the house project, and partly because we didn't irrigate much in summer and the greens suffered a lot.
  
• Fairly large garlic harvest last year, of 18 pounds. (Shallots were a near total failure.) I'd like to see that tripled next year.
  
• Sunchoke harvest has dropped by 50% due to making fewer fires, greatly affecting the winter calorie harvest. I think I'm going to break down and just use our gas stove and a sleeping bag (to act as a "haybox") to start cooking the sunchokes.
  
• We approximately doubled our fennel seed harvest, to 5 pounds. Very successful calorie crop.
  
• More eggs this past year, as the hens we purchased in May 2010 laid heavily from December 2010 through November 2011 before slowing down. We've had a larger flock than would be at all sustainable for this yard, and have been gradually culling the flock for ongoing meat harvest.
  
• Harvesting a lot more nettles this winter season than last year. Delicious!
  
• Light skirret harvest this year. I think last year I harvested a bunch of plants that had been in the ground 2 or 3 years and got very big, then replanted those crowns. I didn't water the skirret much this past season, and many of them grew in part to full shade. I harvested almost all those plants this winter as 1 year old, relatively scrawny roots. Definitely better to get some sort of multi-year mixture going so you can always harvest older plants.
  
• Light mashua & yacon harvests this year, again because of lack of water. Almost all the mashua plants died down in the summer; I didn't expect any roots from them at all, and was pleasantly surprised at how many we did get considering how sad they were.
  
• Good oca harvest this year! Two patches did poorly (lack of water again); one patch to the north of one of our persimmon trees in polyculture with yellow asphodel & lily did very well.
  
• Nice teaser autumn olive and grape harvests! Neither huge, but larger than the few dozen berries or grapes from last year.
  
• Pretty good potato harvest, though still not as good as I'd have liked - lack of water stunted or killed many plants.
  
• We ate lots of fuki stalk, and sold or gave away several divisions, barely managing to keep the growth of the patch in check. Great vigorous perennial vegetable.
  
• So far our asparagus is a very poorly yielding crop in terms of calories per space it takes up--only 280 calories from maybe 10 plants using maybe 30 square feet? Our solomon's seal gave 2/3 the calories from a similar area but growing in heavy shade on the north wall of our house, under timber bamboo, with lungwort, lovage, and wood sorrel in there as well. And we didn't even harvest as much of the solomon's seal as we could have.
  

Ecoroof Grant Report

Here's the report I wrote for the City of Portland's Bureau of Environmental Services, which gave us a $5 per square foot grant to implement our ecoroofs. You can also download a PDF (1 MB) of this report.

Update: addendum to the original report as a blog post or as a small PDF download.

Introduction

Project Summary

We implemented four ecoroofs on our residence at 4510 NE Going St, covering a total of 1000 square feet with ecoroof soil mixes sourced from Philips Soil Products in depths ranging from 3.5" to 8". We implemented and planted two in October 2010, and two in June 2011. We have another 660 square feet of house roof we deemed too difficult to beef up adequately to support an ecoroof. We roofed this area in metal, and it all drains onto the sunspace, front porch, and carport roofs.

About Us

Tulsey Latoski and Norris Thomlinson have experimented at this site since 2006 practicing sustainable, low-work food production via a food forest, perennial vegetable garden, and chickens and bees.

Read more about our projects in general at http://farmerscrub.blogspot.com

Read ecoroof-specific blog posts at http://farmerscrub.blogspot.com/search/label/Ecoroof

View food harvest logs from our ecoroofs at http://discountpermaculture.com/cgi-bin/harvest.py?ecoroof=1

Email us at norristh@gmail.com

Design Goals

• Food production - We planned the ecoroofs for production of food crops either naturally adapted to our seasonal rains, or drought tolerant to make it through the summer.
  
• Reasonably low maintenance - We hope for minimum irrigation requirements, no more often than once every week or two. Once the perennial plants have fully established they shouldn't require much weeding.
  
• Human hang-out areas - We included space for humans to spend time eating, reading, or watching the ecoroof or the rest of the neighborhood below.
  
• Bird & insect habitat - Our food producing, perennial plants provide a diversity of flowers for insects through several seasons, and various seeds for birds to eat.
  
• Potential rabbit or chicken fodder - We envision rabbits potentially grazing on the roofs with human supervision. We planted a few plants which can either serve as human food or be cut and dropped to the chickens below.
  

Four Roofs - Details

Sunspace

392 ft² (13' 3" x 29' 7"): A newly expanded room whose roof we rebuilt from scratch. This roof slopes south with 1/12 pitch, and we constructed it with an ecoroof in mind. The west end receives full sun year round; the east end receives dappled morning shade from black locust trees from late spring through mid fall. It holds 5.5" of intensive B-4 soil mix.

~290 ft² of metal roof drains into this roof, evenly distributed along a ~28' line, depositing into the upper end of the ecoroof.

Since we have the quickest and easiest access from our kitchen to this roof, we planned it as our zone of most frequent harvest, concentrating leaf and flower crops here for frequent picking. The central 2' wide path extending the length of the roof doubles as a sitting area. The south, lower end of the roof overlooks the back yard, where our chickens free range, allowing for possible harvest and dropping of fodder to the chickens below. Rabbits may eventually range here, but they would require a ramp to get to it from their likely dwelling area on the garage roof.

Front Porch

136.5 ft² (7' 3" x 18' 10"): The front porch roof slopes north with about 1/12 pitch. It receives full sun year round. It holds 8" of intensive B-4 soil mix.

To access this roof, we have to walk from the sunspace roof up and over 20' of metal roof, so we planned this roof for less frequently harvested crops such as root crops, seeds, and berries.

~265 ft² of metal roof drains into this roof, with about 2/3 of that evenly distributed along a ~19' line dropping water from above into the upper end of the ecoroof. The other 1/3 coming into the ecoroof meets the soil perpendicular to the slope, allowing very little infiltration; this water mostly runs straight down the side of the soil to a drainage pipe directing it to the gutter.

Garage

245.3 ft² (11' 6" x 21' 4"): The garage roof slopes south with about 1.5/12 pitch. It receives afternoon shade from the house and from late spring through mid fall much of the roof receives dappled to heavy shade the rest of the day from black locust trees. It holds an average of 1.5" of extensive-E soil mix, but we created mounds of soil 3 - 3.5" high in between paths and areas of no soil.

We have a hang-out area for three or four people to gather and sit together in the sun, plus the path lower in the roof in the shade of the locusts allows one person to sit. As with the sunspace, the south end of this roof overlooks the chickens and could be used to grow fodder plants, though the thin soil depth limits the possibilities.

Carport

227.2 ft² (11' 9" x 19' 4"): The carport roof slopes east with about 1/12 pitch. It adjoins the garage roof. It receives afternoon shade from the house. ~85 square feet of metal roof drops its water into the upper edge of this roof, with about 75% dropping into a single spot.

Two paths run the length of the roof, allowing for sitting in the sun or during the afternoon in the shade of the house. It holds an average of 1.5" of extensive-E soil mix, but we created mounds of soil 3 - 3.5" high in between paths and areas of no soil. The thin soil depth and mostly full sun exposure doesn't allow for much more than succulents and Alliums.

Structure & Layers

Structural Engineering

We worked with Ken Safe and Jeff Hartman at Miller Consulting Engineers to determine the necessary structural modifications to support a minimum additional ecoroof weight of 35 pounds per square foot (psf), allowing 5.5" of intensive soil mix:

Sunspace

We had already built the sunspace with 2x12 joists on 16" centers, spanning ~12', sheathed with 7/8" tongue & groove OSB. The north wall of the sunspace is a standard 2x4 stud wall, with a 2x12 ledger attached with lag bolts to carry the joists. The south wall is a window wall, with multiple windows 34" wide with 2x6 studs between them on 3' centers. A 6x8 header spans the windows and rests on 6x6 posts (one 9' and one 12' span between posts).

Miller determined that the 2x6 studs between the windows were too weak to handle the load from the 6x8 header, and the header couldn't make the full 9' and 12' spans on its own. They recommended the retrofit of adding a 2x6 LVL to both the inside and outside face of the header to stiffen it up. They also had us add SDS screws to attach the 2x12 ledger to the house wall, as the existing lag screws weren't strong enough. Because the window wall had too few areas of plywood sheathing to provide adequate shear strength, they had us add plywood to the interior north wall of the room, calculating that the shear load could be transferred via the OSB roof sheathing to that interior wall.

Front porch

Our porch roof had existing 2x6 joists on 16" centers, spanning 67", sheathed with 1/2" plywood. One end of the joists hung from a 2x6 ledger nailed to the house studs; the other end rested on a 4x6 beam supported by 4x4 posts. Miller determined that we needed to use a 4x12 beam instead of the 4x6, and 4x6 posts set in poured concrete pads instead of the 4x4s on pre-cast pier blocks. They also had us add SDS screws to attach the 2x6 ledger to the house wall. These changes permitted 50 psf.

Garage

Our garage roof has 2x6 joists on 24" centers spanning 10' 1", sheathed with 1/2" plywood. One end of each joist hangs from a 2x6 ledger lag bolted to the house; the other end rests on a 2x4 wall. Miller determined we would need one extra joist between each existing set for a final spacing of 12" on center, and we would need to strengthen the 2x4 wall.

Carport

Our carport roof has 2x6 joists on 24" centers spanning 11' 2", sheathed with 1/2" plywood. One end of each joist hangs from a 2x6 ledger lag bolted to the house; the other end rests on a 4x6 beam support by 4x4 posts. Miller determined that we needed to add two joists between each existing set for a final spacing of 8" on center, and do something to strengthen the 4x6 beam, such as adding metal C-beams. The 2x6 ledger against the house should have SDS screws added to attach to the house studs.

Scale-Down of Garage & Carport

We originally planned to implement the garage and carport roofs similar to the sunspace and front porch, with at least 5.5" of intensive soil mix to support food crops. However, these two roofs were built right up to the property line in the past, so to put ecoroofs requiring permits on these structures would have triggered requirements to bring various aspects up to code. We didn't want to deal with that, so we decided instead to implement very light ecoroofs of 30% of the allowed dead load value. Therefore, we did not add any joists or strengthen the beams for these roofs.

Layers

From bottom to top, the ecoroof layers consist of:

• Sheathing (1/2" plywood on all roofs except the sunspace with its 7/8" OSB)
  
• Feltex (light-weight substitute for tar paper)
  
• EPDM pond liner (45 mil Firestone Pondgard. We purchased sheets large enough to fit onto each roof without having to join multiple pieces together, so as to avoid potential leak spots.)
  
• Rotting wood (on sunspace and front porch roofs, to act as a physical dam slowing water down as it works down the roof, and to hold and store water and nutrients. Though the wood was already rotting and soft, we placed a thin layer of soil mix under the wood as an extra precaution to protect the pond liner.)
  
• Soil mix (Intensive on sunspace & front porch; extensive on garage & carport)
  

We figured that the roofs had sufficient slope (1 or 1.5 in 12) to move water via gravity through the soil mix, so we didn't include a separate drainage layer.

We created "raised beds" by using 2x6 and 2x8 boards around the edges of the roofs, running the pond liner up and over before capping the boards with metal rake edge protecting the edge boards and the sheathing, extending down at least 2" into the fascia boards attached under the sheathing. We secured the edge boards with 4"x4" right angle brackets, and placed scrap pond liner pieces or foam padding over the exposed metal to prevent the main pond liner layer from being damaged by the brackets.

Overflow

The front porch roof already had a gutter attached, so we worked with that for our overflow drainage. We lifted the lower "raised bed" edge board an inch off the surface of the decking, then cut slits in the pond liner to allow water to run under the board and into the gutter. We placed filter fabric all along the slit with a layer of river rock to retain soil.

For the other three roofs, we cut holes at the bottom edge of the roof through the decking, large enough to allow a 1.5" diameter PVC or ABS pipe to fit through. We cut the pond liner in an "X" pattern over the drainpipe, folded the flaps down into the pipe, and secured and caulked it with P&L Roof & Flashing Sealant. (For the garage & carport roofs we inserted a plastic ring to help hold the flaps against the inner wall of the pipe.) We used one hole each for the garage and carport roofs, and two holes for the sunspace roof.

Over each drainage hole, we placed a ~10" diameter coffee can with holes drilled or cut out all around the sides of the can. We cut one hole in the bottom of the can to match the hole cut through the pond liner. We wrapped each can with filter fabric then a ring of river rock to minimize loss of soil, and caulked the bottom of each can to the pond liner to secure it and prevent soil from getting under the can. We painted each can with rustoleum.

The excess water from the roofs drains to different places:

• Sunspace: waterfalls into three bath tub ponds, which then overflow away from the house
  
• Front porch: waterfalls into a large pond constructed of the scrap pond liner pieces left over from the four ecoroofs
  
• Garage and Carport: trees and shrubs near their respective downspouts
  

Plants

Sunspace & Front Porch Planting Plan

The deep soil of the sunspace and front porch roofs supports a relatively broad palette of plant species, and hopefully allows for productive cropping. We designed these roof plantings for polycultures of edible plants providing nearly 100% soil coverage throughout the year. Mostly we aimed for each patch to include an evergreen ground cover with evergreen or deciduous plants rising above.

For ground covers, we planted Arctostaphylos uva-ursi, Fragaria chiloensis, Rubus calycinoides, Viola odorata, Campanula portenschlagiana, C. poscharskyana, C. cochlearifolia, Gaultheria shallon, G. procumbens, Vaccinium angustifolium, V. vitis-idaea, Valerianella locusta, and Sedum telephium. For taller plants, we planted many Allium species including garlic and elephant garlic, Astragalus canadensis, Linum perenne, Hemerocallis sp, Agastache foeniculum, Asphodeline lutea, Chenopodium bonus-henricus, Sedum spectabile, Anthriscus cerefolium, Papaver somniferum, Oenothera biennis, several ephemeral bulbs in the Camassia, Triteleia, Brodiaea, and Erythronium species, and a few miscellaneous others. See our Ecoroof Final Planting Plan blog post for full details.

Carport & Garage General Plan

We didn't design the carport & garage roofs in as much detail, since we only had about 3.5" of soil depth to work with. We obtained numerous cuttings of Sedums and other succulents, mostly of unknown species from similarly thin-soiled, dry conditions. We also planted several Allium cernuum plants, one Fragaria chiloensis, a large Origanum vulgare, and a large unknown species of Thymus. In areas of the garage roof which receive heavy summer shade from the black locust trees, we planted Viola odorata and Campanula glomerata, since the protection from the sun may allow a wider diversity of plants to grow in that area despite the thin soil.

Mid-summer report

For an ongoing record of harvests from the ecoroofs, visit http://discountpermaculture.com/cgi-bin/harvest.py?ecoroof=1

Sunspace & Front Porch

Our one plant of Vaccinium moupinense died within a month of being planted. Our seeds of Lepidium peruvianum (old seed), Valerianella locusta, and Papaver somniferum never germinated. Otherwise, the plants on the sunspace and front porch, planted in October 2010, survived the winter and now flourish to a greater or lesser extent. The wet spring and summer this year have sustained growth with no irrigation on our part except for a few recently added plants, and occasional spot watering of some of our more valued experiments (Vaccinium angustifolium, V. vitis-idaea, Gaultheria shallon, G. procumbens, Chenopodium bonus-henricus, Akebia, and Astragalus canadensis).

The Fragaria chiloensis has impressed us with its rapid growth and precocious berry production; this species may make sense as the primary ground cover, since it stays evergreen, grows low, fills in gaps between taller plants very quickly, and tastes delicious! Large swaths of the Arctostaphylos uva-ursi died off following its transplantation from our yard below, but the portions that survived have made a few berries. Unfortunately, the berries of this species don't taste very exciting so it makes an inferior ground cover in our food-focused system. The Rubus calycinoides is establishing fairly slowly, though a few plants have produced flowers. The Vaccinium angustifolium is producing a few berries.

The garlic and elephant garlic seem to have done very well, producing numerous scapes followed by reasonably sized bulbs. We haven't weighed them all yet, but it looks like a very good yield. The other Alliums are establishing fairly well, but with much less vigor so far.

The Chenopodium bonus-henricus has produced a tiny amount of seed; we'll need to wait until next year to assess the production potential of established, mature plants. The Linum perenne made numerous flowers but only a handful across all the plants set seed; we're waiting anxiously to evaluate convenience of seed harvest and their taste. We're disappointed that the Papaver somniferum didn't germinate, as we would have enjoyed that as a seed crop. We have one Oenothera biennis plant flowering profusely, which should result in a reasonable number of seeds for ourselves or for the chickens.

Hemerocallis (daylily) is proving itself very tough, already producing numerous flowers for harvest.

Many other plants have flowered over the last two months, providing an ongoing diversity of blooms and making the roofs pleasant hang-out spaces and valuable for foraging insects.

Carport & Garage

We didn't plant the carport and garage roofs until late June 2011. A month later the cuttings and plants seem to be establishing well.

Implementation

We found it fairly straight forward to implement everything. For each roof, once we had the structural supports in place as designed by our structural engineer, we removed all the old asphalt roofing, tar paper, and roofing nails. We swept up all the dirt and debris to create a clean surface. In places with more than 1/8" gap between plywood we added shims so the pond liner wouldn't get stretched down into the crevice.

Once we had the plywood surface cleaned up, we laid the feltex on the decking in the same manner as tar paper. Then we placed our pond liner, running it under the flashings of the roofs above (sunspace and front porch) or up the wall of the house (garage and carport). We worked it up and over the "raised bed" edge boards, and cut off the excess. We adjusted the liner to minimize any bubbles in the middle of the roof, and folded the extra material at the corners.

The hardest part of placing the liner was dealing with the wood stove chimney projecting through the sunspace roof. We cut an oval hole about 2/3 the size of the chimney flashing, and worked the pond liner down over the chimney, making small cuts as needed to get the pond liner down to the roof. We had to make sure the pond liner stayed 2" away from the actual chimney, so we could only bring it up the flashing to that point. It proved difficult to cut the hole in exactly the right place, so we wound up with a slit in the pond liner extending upslope from the chimney for a few inches. We protected that by adding scrap pieces of pond liner, caulked to the chimney flashing underneath the storm collar and to the main layer of pond liner. To minimize water approaching from upslope, we placed two pieces of plastic to divert water to either side of the chimney. (We also initially placed a ring of drainage pipe and river rock around the chimney, but removed them later when we suspected that rainfall was splashing off those and getting under the storm collar.)

Next we placed the rotting wood for the sunspace & front porch, then soil for all the roofs. And finally, of course, we planted the plants!

Maintenance

Irrigation

Since we planted the sunspace & front porch roofs last October, and have had a wet spring with rains extending into June, as of July 12th we've only watered a few spring-planted additions and (perhaps unnecessarily) some of our more valued experimental plants (see "Plants" section above for details.) Since we completed the carport & garage plantings at the end of June, we expect to water two or three times a week to allow establishment.

The metal roofs condense some water during humid summer nights. We don't know yet whether that will provide any meaningful moisture input, but we hope that the plants at the upper edges of the sunspace and front porch roofs will benefit.

We expect to provide occasional (perhaps once a week) irrigation in future summers to maximize crop production, though certainly we have the option to not irrigate and just accept whatever harvests are possible.

Harvests

Most of the future maintenance should be simply harvesting greens & flowers two or three times a week, plus seasonal harvest of root crops like garlic, camassia, and yellow asphodel.

Further experimentation

We'll adjust the crops planted based on how well they perform. If new plants suggest themselves as good candidates, we'll try adding them.

Weeding

Hopefully most unwanted plants will be excluded by the establishment of a solid canopy of desired plants. After that, we just have to keep those desired plants in balance, which may mean rearranging some polycultures or selectively harvesting greens of certain plants more heavily to set them back.

Fertilization

The sunspace and front porch roofs will require ongoing fertilization to replenish nutrients taking during harvest. We can easily accomplish this by occasional application of urine during harvest trips.

Lessons Learned

Early loss of silt

The runoff water from the sunspace and front porch roofs obviously carried a lot of silt for at least two weeks after the beginning of the fall rains. Perhaps the ideal time to install soil and plant would be mid spring, so that plant roots could grow quickly while the soil was still moist, but without heavy enough rains to carry off so much silt and presumably fertility.

Excess water flows

Due to the extra water coming from our existing metal roofs, we had two problem spots. On two occasions of the heaviest rainfall last winter, the water flowing onto the sunspace roof backed up enough to get past the flashing and into the interior wall of the house. We added two drainpipes, one towards each end of the ecoroof, buried and running under the path directly down the roof. The pipes are wrapped in filter fabric with the upper ends protruding past the upper end of the soil for water to easily enter. This now allows excess water to safely drain away.

On the front porch roof, moderate rains caused a stream of water to overflow the lower edge of the "side channel" where some of the metal house roof meets the edge of the ecoroof perpendicular to the ecoroof slope. We built up a higher "dam" with metal caulked to the existing edge, which now allows excess water to enter a drainpipe running from that spot towards the gutter. The picture on the left shows eroded soil piled up against the original, shallow edge at the far right.

Gutter vs Drainhole

We found integration with the front porch gutter to be more awkward than the drainholes we created for the other roofs. We made a mistake by not caulking the pond liner down to the feltex along the slit we cut for discharge into the gutter, so water initially wicked back upwards under the pond liner and leaked through nail holes in the sheathing until we corrected the problem. The gutter will require more ongoing maintenance to keep it free of leaves than will the drainholes with their small cans.

Surprising absorption of rainfall events

We've been impressed by how much of the rainfall the roofs can hold before discharging anything into the overflow, especially the sunspace & front porch which receive so much extra water from the rest of the house. We haven't made precise measurements and observations, but it seems that if the roofs dry out a bit, they can fully absorb at least a .25" rainfall.

Difficult to buy low-cost plants

We were disappointed to find that we couldn't source low-cost sedums or other ecoroof plants. The wholesalers with good prices apparently only sell to retailers, not directly to end users, even if you can meet the minimum bulk requirements. Luckily, we found friends who allowed us to take cuttings of their sedums.

Costs

With the help of friends, we did all the labor ourselves, so we only had to pay for construction materials, soil mix, plants, structural engineering, and the permit. We located used material as much as possible via the Rebuilding Center, craigslist, etc. Our total cost was about $5400:

• $700 - Structural engineering
  
• $94 - Permit
  
• $180 - Concrete (front porch pier pads)
  
• $129 - Dump fees for old roofing
  
• $1023 - Lumber - structural posts & beams, sheathing, edge boards, etc. (We would have bought some of this lumber anyway for the sunspace, but it wouldn't have needed to be so beefy had we not put the ecoroof on it.)
  
• $108 - Nails & fasteners
  
• $157 - Brackets (to attach edge boards)
  
• $403 - Rake edge to protect edge boards & match existing metal roofing theme
  
• $148 - Feltex (light-weight tar paper equivalent)
  
• $849 - Pond liner
  
• $692 - Soil mix
  
• $906 - Plants & seeds
  
• $15 - Drain pipe for water overflow
  

Perennial polyculture: New Designs

This is part three of a three part series on perennial polycultures:

1. Designs: a five year review
   
2. Species profiles
   
3. New designs
   

Introduction

Following the massive failure of our original polyculture designs (see part one), I spent some time this past winter utilizing my hard-won knowledge of our successful perennials to try again. I didn't design anything for greens, since we have more than enough already established and coexisting quite nicely. I focused instead on root crops, which have proved more difficult to just plant here and there for a few reasons (not all reasons apply to all root crops):

• Soil disturbance damages adjacent perennials or roots of woody plants
  
• I lose track of where I planted odd plants after they've gone dormant, so can't harvest the roots
  
• Roots (especially those starting from small tubers or seeds) get outcompeted by other perennials
  

In this post I'll present the polycultures we're trying this year. I'll also mention a few other ideas I've had but haven't tried to implement. Refer to part two of this series, species profiles, for individual plant characteristics, presented in roughly the same order in which they appear in polycultures in this post.

Polycultures

Implementing

Garlic & Skirret

Garlic & skirret in foreground volunteer burdock & more skirret in background

This bicrop makes use of different time niches with these two root crops. I planted garlic at about 9" between bulbs, with skirret between with 9" spacing to other skirrets and 4.5" to neighboring garlics. Garlic grows through the winter and has died down in early summer by the time skirret has gotten big enough to begin competing for light. Garlic bulbs should lift out easily without disturbing the skirret, and skirret's drought tolerance allows for non-irrigation of the garlic while the bulbs dry down.

This patch could be replanted in the same place year after year by harvesting all the skirret in September and October, then replanting garlic cloves and skirret crowns. Or move the garlic to a new patch and harvest the skirret as needed through the winter, replanting the skirret crowns into that new patch as they become available.

Skirret, Day Neutral Strawberry, & Oniony Thing

This patch combines the need to thin strawberry plants with the need to thoroughly dig up skirret roots. It works because you can leave skirret in the ground for two years before harvesting, and because we have day neutral strawberries which should be thinned in late fall or over the winter. (June bearing strawberries should be thinned after they bear their crop in mid summer). I planted four rows 15" apart in a 5' bed, with skirret and strawberries alternating in their rows every 8". (16" between strawberries, 16" between skirrets, 8" from a skirret to the closest strawberry.)

I planted some sort of evergreen oniony thing between the rows to have some winter growth of its edible leaves. Originally I planned for the ongoing disturbance of the skirret & strawberries to prevent the oniony thing from getting swamped out. But as of mid June, the oniony thing is dominating the area and I'm aggressively harvesting their leaves to open up space for the strawberry and skirret!

This fall I would harvest skirret from two of the four rows (rows 1 and 3), replanting skirret crown divisions after harvest. This will wipe out most or all of the strawberries in those two rows. Next year the strawberries left in the undisturbed rows 2 and 4 will recolonize rows 1 and 3, while the skirrets in rows 2 and 4 grow for a second year.

Next fall I would harvest the skirret in rows 2 and 4, wiping out those strawberries and replanting the skirret crowns. Then follow the same pattern in the future, harvesting two rows each year, such that each row is harvested every other year. This should keep the strawberries from crowding themselves out, as a natural byproduct of thoroughly digging the soil to harvest the skirret. I would adjust the size of the skirret crown divisions in future years to integrate well with the strawberry growth rate--smaller if the skirret is outcompeting the strawberries, or larger if the skirret is getting swamped.

Skirret, Oca, & Potato

This patch uses time niches to some effect, though it doesn't have any winter evergreens.

Oca and potatoes alternate in rows with 16" from one oca to the next potato, (32" from oca to oca and 32" from potato to potato.) I spaced two rows 30" apart in a 5' bed. I then planted one row of skirret halfway between the oca/potato rows, with skirret on 12" spacing within its row.

Skirret and potatoes grow vigorously early in the season, with oca putting on growth more slowly. We'll harvest potatoes July through September, with skirret still providing shade for the oca in the heat of the summer. With the cooler cloudier weather in September, the oca vegetation should quickly fill out to use up the space left behind by the potatoes. We'll dig all the oca tubers out after the first frost, and harvest skirret as needed through the winter. We can either reimplement the same polyculture in the same bed, or rotate it to other beds to prevent disease problems with the potatoes.

Oca, Asiatic Lily, & Yellow Asphodel

Confusing mess w/unplanned strawberry etc

Utilizes different time and height niches. These are planted in an understory wedge to the north of a young persimmon. At 6' tall the persimmon casts minor shade. Oca is planted on 30" centers with asphodel surrounding it on 10" centers. One lily is planted in the center of each oca "triangle". I don't have enough asphodel propagated yet, but eventually their density could be increased to about 6" between plants.

Asphodel grows from fall through winter til early summer, making its roots available for harvest while the oca is still small. The oca and asphodel provide ground cover for the lily, which grows above them. Harvest all oca after first frost, and harvest lilies as needed through the winter.

Lilies and asphodels can be harvested with fairly minor soil disturbance, so the main conflict might be the effects of oca harvest on the asphodel roots with their new-ish growth going into winter.

We planted this polyculture into an area somewhat invaded by strawberries, and with remnant camassia and weeds including dandelion & popweed. We may have trouble with the strawberries especially, since we don't have a strong ground cover or weed excluding element.

Yellow Asphodel, Good King Henry, & Violet

Utilizes height & time niches. Violet should be an evergreen (we're using Viola odorata) for permanent low ground cover and winter greens, with the yellow asphodel and good king henry (GKH) growing up through it. The GKH begins growing late in the spring, but the other two plants should help suppress early weeds, and the asphodel will then die down in summer for the GKH to fill out further. We should be able to harvest the asphodel roots in the summer with minimal disturbance to the GKH.

I planted GKH about 2' apart, and would eventually like to have asphodel at 6-8" spacing filling all the interior area. We don't have enough asphodel plants yet for full density, so they're more sporadic for now. The violets will fill in wherever they find gaps.

Jerusalem Artichoke, Mashua or Groundnut, & Chinese Artichoke or Creeping Bellflower

Jerusalem artichokes with small chinese artichoke underneath

This polyculture has a core structure but multiple possible plants to plug into the different niches. It mimics the well known three sisters guild of corn, beans and squash, which Eric Toensmeier has proposed morphing into the perennial guild of jerusalem artichoke, groundnut, and chinese artichoke. This polyculture makes use of above ground space niches, but not of time niches, since these root crops require heavy disturbance for harvest in fall through early spring. With the possible exception of the creeping bellflower, they should all benefit from the regular ground disturbance and loosening of the soil.

We're retaining jerusalem artichoke as the vertical element; we had an existing 100 square foot patch. However, we've never had much success growing ground nuts here, so we only planted 3 or 4 which survived from last year, instead mostly planting mashua on 3' centers as the vining element to climb the jerusalem artichokes. We can easily supply nitrogen via our urine so we don't require the leguminous groundnut for nitrogen fixation.

For the ground cover layer, we're trying about half a dozen fast-spreading chinese artichoke in half the patch, with creeping bellflower 1-2' apart as another vigorous, shade tolerant root crop in the rest of the area.

Our patch gave us about 100 pounds of jerusalem artichokes last year (1 pound per square foot). It makes sense to knock back the jerusalem artichoke production a bit in favor of more root diversity, and hopefully the total yield of roots will increase while we're at it.

Brief Mention

Oca & Tomatillo / Ground Cherry

Ocas & tomatillo at bottom tree collard and mashua at top not part of guild

Inspired by oca-testbed's oca & tomato bi-crops, I've planted 3 ocas, 2 annual ground cherries, and 2 tomatillos with 10" between each oca and its neighboring ground cherry or tomatillo (20" from one ground cherry or tomatillo to the next).

Squash & yacon

I planted some squash seeds at 6' centers and yacon halfway between at the 3' mark. The yacon should grow tall enough to hold its own by the time the squash reaches it, to share the space niche a bit. It may work somewhat as a time niche, too, as squash often dies back in early to mid fall with powdery mildew, while the yacon can keep growing until frosts kill it.

Not Implementing

Oca & squash

We created an accidental time niche bicrop a few years ago when a squash covered up some oca for most of the summer, but started dying back with powdery mildew in early fall, allowing the oca to explode in growth and fill out the space. We didn't get much of an oca yield--but I wasn't experienced enough at that time to pay close attention to frost and harvesting all the oca promptly. So maybe we got some roots but they rotted? Or maybe the squash didn't allow the oca to grow well enough to produce roots? I'd like to try this again with squash on 6' centers and two or three ocas at the 3' point in between. Or try combining it with the squash & yacon polyculture, with the squash and yacons spaced further apart to allow oca some breathing room between the larger plants. (See oca-testbed's polyculture mound of yacon, oca, and chinese artichoke.)

June bearing strawberry & summer root crop

I've tried to design a polyculture which combines digging some root crop with the need to thin June bearing strawberries in late summer, after they've finished cropping for the year. I've had a much harder time with this than with day neutral strawberries (see my polyculture with skirret above), since very few root crops can be harvested in the summertime after two years of growth to allow the alternating row harvest method. Strawberries fill out quite well by mid spring, creating a lot of competition for anything shorter than they are, limiting the ability to sow seeds or plant small divisions at the beginning of the growing season. Further, the root crop can't be allowed to outcompete the strawberries too badly -- we have some burdock in our patch, and we have to keep harvesting the huge leaves (we do eat the leaf stalks) or the strawberries get totally covered up!

Spring ephemeral bulbs such as Camassia, Triteleia, Brodiaea, or Erythronium might work for the row harvest method, especially if you establish a solid patch first, then add strawberries later. Or instead of harvesting a full row at at time, you could do a distributed harvest of the thickest clumps of bulbs, disturbing patches of strawberries here and there while eating the largest ephemeral bulbs and leaving the small ones behind to regrow quickly the following spring. Or try root crops whose seeds can germinate in the autumn, overwinter as a small plant, and grow quickly in spring: black salsify (Scorzonera hispanica) or dandelions for full row harvest after 1.5 years, or black salsify, dandelions, or parsnips for distributed patch harvest the summer after they've been sown.

Perennial polyculture species profiles

This is part two of a three part series on perennial polycultures:

1. Designs: a five year review
   
2. Species profiles
   
3. New designs
   

Introduction

In Perennial polyculture designs, I mentioned inadequate knowledge of plants as one barrier to successful implementation of my designs. We've now gained enough knowledge of a few perennial plants to try again. With this mouthful of a title, I give the relevant design characteristics of perennial vegetables we're trying in polycultures this year. Unless otherwise noted, I consider these plants primarily root crops, though some have secondary uses like edible flowers or leaves. By the way, for more information on many of these unusual root crops, including some polyculture experiments, see:

• Oca testbed - nice details on tomato & oca bicrops, plus lots more on oca and some on other roots
  
• Radix Root Crop Research and Ruminations - pioneering work with many roots I haven't tracked down or in some cases even heard of before
  
• Obligatory link to Plants for a Future database.
  

Species characteristics

Garlic (Allium sativum)

• Harvest time: July or August
  
• Harvest process: Lift all bulbs, which doesn't require much soil disturbance. Store in cool dark place.
  
• Planting process: Plant cloves in October or November, which doesn't disturb soil, but cloves shouldn't be disturbed by other digging after being planted.
  
• Generally doesn't require any watering, and in fact shouldn't receive water in July so the bulbs can dry out for long term storage.
  
• Vegetative growth: small amount of leaves over the winter, growing more actively in spring. Die back in July with the summer drought.
  

Skirret (Sium sisarum)

• Harvest time: October through April. Stores in the ground all winter.
  
• Harvest year: can harvest after one year of growth, or leave in the ground for multiple years for more and larger roots, which seem less prone to having a woody core.
  
• Harvest process: Dig large chunk of soil from around plant (up to 2.5' diameter with older plants), pull out crown. Cut off roots, optionally divide and replant crown. Can be difficult to find in late winter after stalks have rotted away.
  
• Drought & shade tolerant
  
• Vegetative growth: Appears mid-season (April), grows fairly quickly to 3-5' tall (depending on age of plant), and dies down in early fall (October). Dense, casting heavy shade.
  

Garden strawberry (Fragaria x ananassa)

• Use: berries
  
• Harvest time: summer through fall
  
• Best with full sun and adequate water through growing season.
  
• Benefits from thinning (following the annual harvest for June bearing varieties, and in the winter for day neutral varieties)
  
• Vegetative growth: semi evergreen to 1' high, spreading quickly from runners
  

Potato (Solanum tuberosum)

We haven't actually grown potatoes very much, but my understanding so far:

• Harvest time: June onwards. Can probably store in the ground through the winter?
  
• Harvest process: Seems to need a fairly thorough excavation of the soil, especially to find all the roots. We had many potatoes return this year from last year's plantings, so this seems to have potential as an overwintering perennial. Disease buildup normally demands crop rotation.
  
• Supposedly fairly drought and shade tolerant.
  
• Vegetative growth: Starts growing in late March or April. Reaches 3' wide? Yields supposed to improve with hilling up soil onto the lower stem.
  

Oca (Oxalis tuberosa)

• Harvest time: after the first light frost(s) kill the foliage, but before hard freezes damage the tubers, which often dwell close to the soil surface.
  
• Harvest process: Dig out all the roots, generally concentrated at the center, though some may form where foliage touches the soil. Store in cool dark place for the winter, replanting some tubers in spring. In mild climates, tubers missed at harvest time may resprout on their own in spring.
  
• Appreciates some shade during heat of the summer, but may not tolerate too heavy competition. We tried it as a ground cover amongst other plants one year, and they swamped it out with very little root yield.
  
• Vegetative growth: Leafs out in April or May, generally low growing to about 1', though can clamber up other plants if it needs to gain access to sunlight. Doesn't put on much growth until September, when cooler cloudier weather kicks in, at which time the foliage explodes to 3-4' wide.
  

Asiatic lily (Lilum sp)

• Harvest time: Late fall through early spring.
  
• Harvest method: Compact bulb, so fairly easy to harvest with minimal soil disturbance. Leave large offsets behind to regrow. Can be difficult to find in late winter after top stalks have rotted away.
  
• Like their feet in shade and top growth in sun, thus well suited to combination with a low ground cover.
  
• Vegetative growth: comes up in early to mid spring, grows to 3-4' tall, and dies down in fall.
  

Yellow asphodel (Asphodeline lutea)

• Harvest time: Can definitely harvest in the summer, and I think we can dig roots year round.
  
• Harvest process: Medium soil disturbance, concentrated around center of each plant. The plants make numerous offsets, and each plant has multiple thin roots. So you can harvest some entire plants and leave/replant others, or cut some of the roots off of each plant and replant them all for slower/less vigorous regrowth. Not sure yet of the best method.
  
• I think it prefers full sun, and probably doesn't need any irrigation.
  
• Vegetative growth: somewhat sparse, to about 1.5' tall on flowerless plants, or 3' tall on single flower spike. Doesn't seem to compete all that well with other plants, so may do best with a low growing ground cover for weed exclusion. From the Mediterranean, so well adapted to our summer drought by dying down in mid summer and coming back with fall rains, staying green through the winter.
  

Good King Henry (Chenopodium bonus-henricus)

• Uses: leaf crop (spinach substitute) and seed crop (used like quinoa, though smaller).
  
• Harvest time: leaves throughout season. Seeds over a period of about three months, July through September.
  
• Harvest process: seeds require periodic picking every week or two as different seed stalks ripen at different times.
  
• Somewhat shade & drought tolerant, though I assume for optimum seed production we should give it full sun and summer water.
  
• Vegetative growth: comes up in late spring, reaching about 1.5' tall and perhaps a bit wider. In our back yard where chickens roam, we have a fairly pure stand of GKH and it does fine. In our front yard, (because chickens aren't eliminating the other plants? or because the GKH hasn't dominated the root zone yet?) the GKH gets crowded out by the early spring growth of nipplewort (Lapsana communis), lemon balm (Melissa officinalis), and wood avens (Geum urbanum) and I've had to weed the bed this year. Like the yellow asphodel, would probably benefit from a low growing weed suppressing ground cover.
  

Jerusalem artichoke / Sunchoke (Helianthus tuberosus)

• Harvest time: October through April
  
• Harvest process: Major soil disturbance. Will always miss some roots so that it will regrow the next year. Can be difficult to find roots in late winter after stalks have rotted.
  
• Drought and shade tolerant
  
• Vegetative growth: appears mid spring, reaches 8+' tall. Dies off in late fall.
  

Mashua (Tropaeolum tuberosum)

• Harvest time: In theory, should harvest after the first light frost(s) kill the foliage, but before hard freezes damage the tubers. Roots seem much hardier than oca and yacon, though, such that they might store OK in the ground through the winter.
  
• Harvest process: Fairly thorough excavation to dig all roots, storing in cool dark place for the winter. We usually miss some of the roots so that it regrows the next year.
  
• Appreciates some shade during heat of the summer
  
• Vegetative growth: Vining to at least 10' tall, climbing other vegetation. Appears in mid spring and grows at a fairly steady rate until frosts.
  

Groundnut (Apios americana)

(c) 2004 Steve Baskauf

We haven't had much success with this plant, but we haven't totally given up on it yet..

• Harvest time: dormant season, late fall? through late spring?
  
• Harvest method: Extensive excavation required. Leave some tubers behind to regrow.
  
• Fixes nitrogen
  
• Vegetative growth: vining, scrambling up surrounding vegetation. In our climate, appears around June and disappears in September or October.
  

Chinese artichoke (Stachys affinis)

We've only grown these for one year, so I base this mostly on the literature:

• Harvest time: Dormant season (fall through early spring)
  
• Harvest process: Probably requires major soil disturbance, especially since I don't think dead plants leave woody stalks behind to mark their spots. Will regrow in spring from tubers you missed.
  
• Moderate moisture requirements?
  
• Vegetative growth: appears in early or mid spring, gets about 1' to 1.5' tall, runs quickly (mint family) and forms good ground cover. Not sure exactly when it dies down.
  

Creeping bellflower (Campanula rapunculoides)

• Harvest time: I think we can get usable roots year round.
  
• Harvest process: Fairly major soil disturbance. Need to dig towards the center of the clump to find usable roots. I suspect it will keep regrowing vigorously after each harvest with no need to deliberately leave roots behind.
  
• Very shade tolerant, and competes well with other vegetation.
  
• Vegetative growth: Leaves appear in early spring and make a good source of greens while others are somewhat scarce. Plant reaches about 3' tall and runs vigorously. I think that without irrigation it responds to our summer drought by going somewhat dormant, resuming growth in the fall before dying back for good over the winter.