Building a small production greenhouse

When we rebuilt the house one thing we decided to add was a small greenhouse. We picked a spot with great sun exposure and a convenient pre-existing wall and built a small glass and granite building. The same architect that designed the house remodel gave it some nice touches too.

We added a metal table with automatic watering above it on the taller side, which is where we do most of our reproduction work and left the smaller side for mostly potted plants. In the middle we built a wood walkway between the two doors.

The last change we've done is to add some benches to the smaller side and extend the watering system to that side as well.

We've been extremely successful at doing all sorts of reproductions using this greenhouse. For example we start all our tomatoes from seed we collect from the previous year's production (that's only possible because we use heirloom varieties from the The Real Seed Catalogue and not the usual hybrids).

Last season we had a tomato plant grow out of the ground by chance. We let it ride and it fruited well past the normal season (we ate the last tomatoes in January). After that much success we decided to start planting some tomato plants on the greenhouse ground to get a small but out-of-season crop. We used the low side of the greenhouse and added a watering line on the ground as tomato plants supposedly don't like to be watered from above. That also went well but it was always a small crop. That eventually made us want a much bigger greenhouse to have much more actual production instead of nursery...

After shopping around quite a bit we found a local supplier that would build a 14 by 4 meter poly tunnel for a very reasonable price. We decided to set it on the edge of our forest garden in a spot with good sun exposure and that also helped with bringing some immediate wind cover to a somewhat exposed part of the field. The construction was all done in a single day last summer. The builders brought most of the structure pre-built with them and the day's work was mostly assembly. It started early in the morning with the setting of the posts. This was done extremely quickly with a gasoline powered drill. They were pre-prepared metal posts with a cylindrical concrete base that was buried for stability.

After all the posts were level and straight the pre-bent arches where slid into place and bolted on.

Once that was done they dug two trenches on either side of the structure. This was done by hand and was by far the most grueling part of the job, especially since the sun was high and hot by then.

Once that was done they unwrapped the main length of plastic and stretched it on top of the structure.

After the plastic was over the structure they stretched it firmly and held it in place by placing it in the two ditches simultaneously and filling them back up with the soil that had been taken out and pressing it in stepping on it. Once it was stretched they bolted it to the end arches. The next step was to assemble the frames for the doors on both sides. We opted for a single door on one side and a double one on the other so we could fit the electric buggy in if we wanted to.

Both the single and double doors have a window above it that can be closed or held in place in one of two positions. Here the winter position is shown, giving some ventilation but not facilitating the egress of hot air. The summer position is swung the other way round so the hot air rising is naturally driven out. All the doors and windows have rounded corners but are set into square frames. This leaves enough of a gap that even if everything is closed (for the coldest part of winter) there is a decent amount of ventilation.

After the build was complete we drove the electric buggy through the double doors bringing a very large amount of compost from our large piles. After spreading it evenly we planted the first crop, using cardboard to create walking paths in the middle and sides.

The cardboard worked well at controlling weeds but started to waste away relatively quickly so we used some weed barrier fabric to cover the paths instead.

To water the greenhouse we started out with a simple square lawn sprinkler driven by a water timer. This worked fine in the beginning but once the plants started growing the sprinkler would often get covered by plants and not water everything evenly. We've now installed a new overhead system. We'll do a separate post detailing that.

We've been having great success with all sorts of vegetables so far. The first tomato plants have just been added and we'll see how that goes. This is not just for annual crops though. There is a selection of tropical plants in pots about to be put in. The selection includes banana, annona, papaya, mango and a few other smaller stuff. The banana may be pretty marginal but we think we have a shot at getting a crop from most of these even if it's not every year. And that's even before exploring some of the passive heating options out there (long but very interesting video).

Building a forest garden

We've been considering the forest gardening concept for a while now. Martin Crawford's book on the subject is an amazing introduction, tutorial and reference. The idea is to run an agricultural system as close to a natural forest as possible. The suggested benefits are to be able to produce at good levels per acre with much lower inputs of energy and work. We have now picked a few spots on the farm where we are planning on building forest gardens to use the space more efficiently in producing things we value.

A lower energy way to farm

In most climates if you leave a piece of land alone for enough time it will eventually turn into a forest. An empty plowed field is as far away from a forest as you can get so a lot of energy is required to push back against nature and keep it that way. Conversely a forest garden aims to be the closest thing to a forest you can get while still remaining productive, so it should take the least amount of energy to keep it that way, without turning into a full forest.

When we got the farm there was a part of it that had been left unattended for 10+ years and had grown some pretty large oak trees. After the initial clearing of the still heavy underbrush and a few more years of letting the trees grow we now have a nice spot to take an afternoon nap that never needs any weeding.
The problem with that space is that it doesn't produce any valuable products. If instead of oak trees we had chestnut trees, we would be getting something out of it but only a single small crop a year. So the usual response to this has been to create highly intensive agricultural systems that operate on a yearly cycle. In Portugal that was usually a mix of annual crops that have a beneficial profile in terms of nutrient use/generation (pumpkin, corn and beans) and then a few fruit trees made to grow very high to leave the field beneath them alone and some vines along the borders. Here's the state of the typical field this time of year:
You'll notice the bales of hay that have just been harvested from the field leaving it barren as well as the vines on the edge of it. This particular field doesn't have any trees on it but the one just to the left of it right now is fully planted with corn around a large apple tree in its center. To keep this field in this state you need to add a lot of energy to it. If you don't plow or cut yearly oak seeds from nearby forest will take hold and after a few years the field will turn into a forest.

The vines and the trees are the only perennials in the traditional system. The main sources of food however are the annual crops that require a lot of energy input not only from the annual plowing but also from the fertilizers you must add to the field to compensate for the fact that these crops are not adding enough nutrients back to the soil. Traditionally this was done by rotating crops and adding animal manure, both things that consume resources, either because you need to forfeit production from the field or have crops to feed the animals that provide you with the manure.

Forest gardening tries to use much less energy by solving the energy input problem in two different ways. First it uses perennial instead of annual plants so the energy needs are lower as you're not having to rebuild the whole plant every year from scratch and can instead steadily accumulate "plant capital". And second, it plays with the plant mix to get a whole ecosystem going, just like in the pumpkin+corn+beans example but with a much more diverse ecosystem. This whole ecosystem is much easier to maintain since it uses plants of various sizes and types to fill every niche crowding out weeds. After you've established the system you just need to do small yearly cuttings to keep it in balance without tipping over into a dense forest. Since you're only pushing nature a little bit instead of all the way to a barren field this takes much less effort and/or energy.

Getting there in baby steps

Right now we're still in the planning stages of our own forest gardens at the farm. Our base idea for a full blown garden is simple. On the north side the farm limit is a county road. From that road to the core of the property we'll build a wedge type ecosystem where the larger species are in the back and the smallest in the front. This will improve the sun exposure as the large trees in the back won't shade the smaller trees, bushes, grasses, etc in the front. We're still in the process of picking out what all the species should be. In the meantime we've taken some of the concepts and started applying them at a lower scale.

First we've started looking into which species to add in places where we've added watering like the recent installation in the kiwis. Weeds will grow very fast when given water so the best way to not have to cut them is to just crowd them out with something you actually want. In another spot in the farm where we added irrigation to some vines this happened naturally with some wild strawberry growing strongly and taking over from the weeds.

Then we started looking into how to improve the ground coverings so that we don't have as many weeds growing. There are plenty of ground cover mixes you can try. Crawford's book lists a few and the one that caught our eye was a mix of three species:

  • Trifolium repens (white clover)
  • Lolium perenne (dwarf perennial ryegrass)
  • Festuca rubra (creeping red fescue)

Crawford suggests you mix them as seeds in a 3kg+6kg+2kg proportion for each acre and sow them anywhere from April to August. So far we've only been able to get our hands on the white clover and are still looking for bulk sources of the other two. The idea is that this kind of mix will last very long without maintenance (the book says 10+ years) and still provide beneficial value. In the case of this mix that would be fixing nutrients like nitrogen into the ground while also producing flowers that bees will use. We've also used thyme as it covers the ground reasonably well and is a great herb for seasoning.

Finally the last thing we're considering doing in the immediate future is to pick the main species of large tree to put on the north side of the property. Since those are the ones that take the longest to grow it makes sense to plant them there early enough so that we can then add the rest of the garden over the years. The Italian Alder is highly recommended in the book as a good nitrogen fixer that grows relatively quickly.


We're still very much in the beginning when it comes to shaping the garden and yet the forest gardening concept has reinforced a lot of the things we had already been doing. When we took over the farm there was not a lot of diversity and we've been slowly increasing it by bringing in a lot of exotic or even native plants that were no longer present. We also fought the local tendency to cut down everything and start from scratch. For example we've been slowly cutting down trees only as they've become a nuisance, meanwhile they grow large enough to produce firewood. We've build quite a stockpile of firewood over they years because of this.

What we haven't done enough of is building the garden with enough layers and density so as to crowd out invasive species. We still have a lot of weeds and yet are always complaining that we're running out of space to try new things. Hopefully over the next few years we will have a lot more plants, of a lot more species with a lot less work in weeding and cutting.

Further reading