Sunday, March 10, 2013

Organic Farming, Swedish Style

Every once in a while, a paper gets published in the academic literature that makes readers (or at least, these readers) say, "what a cool set of experiments!"  Such was the case this week with an article published in the journal Acta Agriculturae Scandinavica, Section B - Soil & Plant Science.  A team from the Swedish University of Agricultural Sciences in Uppsala, Sweden took a look at the sustainability of organic agriculture on a small farm for a variety of livestock and fossil fuel demand scenarios.  The authors' test farm was 35 hectares (ha, 86.5 acre), consisting of 8 ha (19.8 acres) arable land, 5.5 ha (13.6 acres) meadow, 3.5 ha (8.5 acres) pasture, and 18 ha (44.5 acres) forest (of which 10.5 ha (25.9 acres) were also pastured).  The main goal was to to find the maximum number of people that could be supported from their organic test farm, and to see if the number changed significantly depending on livestock and fossil fuel scenario (you can see from the abstract in the link above that the livestock make more of a difference than the fossil fuel demand, as long as the fossil fuels are readily available).

Breakdown of how the test farm is distributed.  Areas are proportional but not necessarily representative of how the actual farm looks.  (If anyone wants to send us to Sweden to find out in person, please let us know in the comments section below!)  Arable land is divided into eight one-hectare (2.5 acre) plots.

The authors determined that the farm's demand for "tractive power," or "tasks that are normally done with a tractor," could be optimally met with one diesel-powered tractor and combine, three draft horses, or a combination of one tractor, one combine, and one draft horse.  In the last scenario, the tractor and combine would be used for high power-demand or time-sensitive tasks like plowing, threshing, manure spreading, and grain harvesting, while the horse would be used for most other tasks, like haying, harrowing, sowing, and vegetable harvesting.  Importantly, the combination of horse and diesel power allowed the farm to produce enough fuel on-farm to run the tractor and combine (although the authors don't seem very confident in their ability to produce biodiesel on a small scale, for some reason).  Since one goal was to keep the farm as a self-sufficient unit of production, the diesel-horse combination was especially preferred.

Another important point is that the only fossil fuel demand considered was for "tractive power."  The authors were able to validate this relatively narrow scope because they chose a crop rotation that heavily emphasized ley and green manures, such that additional fertilizers (other than the manure produced by their livestock) were unnecessary.  Specifically, the authors split their 8 ha of arable land into eight 1-ha (2.5 acre) plots, rotated through the following sequence:
  1. Alfalfa
  2. Rapeseed (similar to canola)
  3. Winter wheat undersown with Crimson clover
  4. Potatoes and other vegetables
  5. Buckwheat
  6. Oats, undersown with alfalfa
  7. Alfalfa, harvested twice
  8. Alfalfa, harvested three times
The yields obtained from these crops (in addition to the forage provided by the pasture and meadow) were used to determine the number of people and animals the farm could support.  The authors chose crops and animals with which they had experience and, thus, accurate baselines for feed consumption and growth rates.  They also made conservative assumptions on the food energy produced, assigning all vegetables to a "lettuce equivalent" and assuming their horses would eat like they were pregnant. (Seriously!)  Specifically, they chose two breeds of Swedish Mountain Cow (one large, one small), North Swedish draft horses, and unspecified breeds of sheep and poultry.  They constructed a model to determine how much livestock could be supported (after feeding the horses), assuming the cows and sheep would eat no grain and the chickens would eat waste grain supplemented with other feed (see below).  Because a cow requires many times more resources than a sheep (about 9.2 times in their system), the farm's production could support in some cases, for example, nineteen cows and seven sheep, but not twenty cows.  Thus, the authors considered a number of livestock scenarios:

  1. All large cows, with the remainder of forage going to sheep, plus chickens
  2. Same number of small cows, with the remainder of forage going to sheep, plus chickens
  3. No cows, all forage going to sheep, plus chickens
  4. All small cows, with the remainder of forage going to sheep, plus chickens
  5. Roughly equal balance of large cows and sheep, plus chickens
  6. Equal number of large and small cows, with the remainder of forage going to sheep, plus chickens

Crop scenario #6 was only considered for the combined horse and diesel powered fossil fuel scenario.  In terms of gross calorific food value (the metric used to determine the number of people the farm could support), crop scenario #1 came out on top for all three fossil fuel scenarios, but the calories came to a larger extent from milk and less from meat and eggs (to the tune of 12-13 liters of milk (over three gallons) per person per week).  Of course, it would be advisable to use some of that milk for butter, yogurt, cheese, etc.  On the other hand, if all of the non-vegetable calories shifted to meat and eggs (scenario #3), the number of people supported decreased by roughly a factor of two.  Of course, there are many other combinations of livestock and crops that could be used to vary milk-egg-meat ratio to produce meat more efficiently while yielding a reasonable amount of milk per person, but this study makes a good starting point.  Notably, in every case, per capita meat consumption well below the current global average was required.

Livestock considered in the various scenarios.  Clockwise from upper left: North Swedish horse, generic sheep, generic chicken, and Swedish Mountain Cows (Photo credit: Wikipedia for horse, sheep, chicken, and cows, respectively).  Generic photos of sheep and chicken used in part to protect their identity from the Swedish Chef.

As with any model, a number of assumptions were made.  Most were based on the conservative side of previous years' data from their research farm or nutrition values from previous studies, but one assumption that wasn't is worth mentioning.  That is, a significant fraction of the poultry feed was assumed to come from slaughter waste of the cows and sheep.  In normal operations, that probably isn't directly a good idea (or even legal).  However, the slaughter waste could be used to cultivate grubs and larvae that would make good chicken feed, so although the assumption is overly simple, it isn't entirely without merit.

One of the main conclusions of the article was that by extrapolating the results of their small organic farm system, it was possible to generate enough calories to feed seven billion people (and in the best cases, nine billion people) in each of their fossil fuel demand scenarios.  Such large extrapolations are always dangerous, but this particular claim that organic agriculture can feed the world has a lot going for it.  First, conservative estimates for yields were used, and the growing season in Uppsala, at nearly 60° N latitude (approximately the division between the northern and southern Canadian provinces), is shorter than in many other inhabited regions.  Second, the system was largely self-contained, not requiring external inputs of fertilizer or, in some cases, even fossil fuels. Although the production of rapeseed esters from the rapeseed oil (that is, biodiesel) would likely require fossil resources (for methanol to make methyl esters) or larger grain yields (to make ethanol for ethyl esters), there is an appreciable cushion of resource availability, at least for the present global population.  (For the authors' preferred scenario--combined horse and diesel power, livestock scenario #6--their farm could support 69 people compared to the 58 people required to extrapolate their results out to global scale.)  Thus, some grain could conceivably be used to produce ethanol and subsequently rapeseed ethyl esters for biodiesel, but the exact balance isn't clear from the article.  In any case, this article constitutes a very promising outlook for the future of organic farming, and the model is a useful starting point for scientists and engineers looking to plan out their homesteads (like we are)!

Have you done any similar calculations for your homestead, or do you know of any tools to help aspiring homesteaders do the same?  Tell us about it in the comments section below!


  1. I wonder about the horses...they eat a lot. Could oxen or even sheep be used for the traction needs? Seems like the horses would contribute little besides muscle power. I am wondering why not use animals that could be multipurpose for work and for their food/dairy/wool. Though I am not knowledgeable about what the needs on the farm would be or how much each of these animals can pull. They mentioned small/large cows. Maybe a smaller version of horse or even a donkey would eat less but still do the work?

    1. I would think if you could train cattle or sheep (even pigs, in terms of tilling) to do the work you needed, there's probably a good opportunity for using multipurpose animals. There are lots of examples of other animals being put to work in similar ways, especially if you don't have a ton of land to work. However, according to this comparison , horses are as good as it gets in terms of work output. Of course, it's hard to imagine anything putting out more energy than a black lab puppy, so maybe you'd have something these. ;-)

    2. Looks like the link didn't show up. This comparison:

      or search for 'world wide flood draft animal power'