Part 1 · The Basics of Permaculture
Plants, Polyculture and Guilds
Polyculture and guild design, nitrogen fixing, succession and stacking, and an honest account of which diversity claims the evidence supports.
A polyculture is a planting where two or more crops share ground and time. The claim made for it is that diversity yields more than the same crops grown apart. Part of that claim is well evidenced. Part of it is folklore copied from one handout to the next, and the folklore is worth separating out, because a plan built on an untested claim fails in the field the way a plan built on a bad number fails.
What the diversity evidence shows
Gaudin and co-workers, in a 2015 PLOS ONE study of cropping sequences, found that diversification reduced yield variation and improved yield stability under variable weather. Read that carefully. The strong result is about stability, which is what a household eats in a bad year, rather than about a higher average yield. Martin and co-workers, writing in PLOS ONE in 2019, documented the opposite trend at global scale: crop diversity concentrating toward fewer species through the Anthropocene. Diversity is a risk hedge that industrial agriculture has been trading away.
Altieri's Agroecology supplies the mechanism. Polycultures and intercrops can exploit complementary niches, and biodiversity supports pest regulation and resilience. The mechanism is real, and the size of the effect depends on the species, the soil and the season. Anybody promising a fixed percentage from mixing crops is guessing.
Nitrogen fixing is a species question
Which legumes fix, and how much, is where the plan usually breaks. Common bean, Phaseolus vulgaris, is documented in the literature as inefficient at symbiotic fixation, and SARE's own cover crop chart omits it while listing cowpea at 100 to 150 pounds of nitrogen per acre. Austrian winter pea runs 90 to 150, and up to 300 in a good stand.
Two constraints the handouts leave out. Nitrate in the soil suppresses nodulation, so the field that looks richest fixes the least, and the credit shrinks toward zero on manured ground. And inoculant is strain specific: pea and vetch share Rhizobium leguminosarum bv. viciae, common bean needs R. phaseoli, lupin needs Bradyrhizobium. The wrong packet does nothing. The free check is to dig a plant, wash the roots, and cut a nodule open. Pink inside means fixing. White means buy inoculant.
Guilds, and the honesty problem
A guild is a group of species assembled to fill niches and support one another: a canopy tree, a nitrogen fixer, a mineral accumulator, a ground cover, an insectary plant. The shape of the idea is sound. It is also where this course stops repeating things.
Most specific pairing claims in circulation, comfrey under fruit trees, horseradish at the orchard edge, yarrow for pest control, are asserted and rarely measured. Where a mechanism has been documented it is usually narrower than the claim built on it. Theophrastus watched plants closely and wrote down what he saw, and even he could describe interference between species and only guess at the cause.
Succession, and stacking in space and time
Succession is the sequence ground runs through if you leave it alone: bare soil, annual weeds, perennials, shrubs, pioneer trees, woodland. A vegetable bed is succession held deliberately at the annual stage, and holding it costs work every season. A food forest runs succession forward on purpose, letting the shrub and tree layers take ground that was annual, and paying for that in years before the tree crop arrives.
Stacking means using the third and fourth dimensions. Vertical stacking is a tall crop over a short one, a vine on a trunk, a root crop under a leaf crop. Temporal stacking is a fast crop and a slow crop sharing a bed in different months. Where species genuinely differ in rooting depth, season and demand, the mixture can out-produce either alone, and the measure for that is the land equivalent ratio: the area of separate monocultures you would need to match the mixed plot. Measure it on your own ground.
Every species in a designed guild should earn its place with a stated job: fix nitrogen, accumulate minerals, cover soil, host the predators of pests, provide mulch, provide food. A plant with no job is decoration.
Exercise
Design a three-species guild and its control
Choose three species that differ in rooting depth, season, and what they take from the soil: a deep-rooted accumulator, a nitrogen-fixing legume, a shallow-rooted leaf crop. Write the job each one does and the mechanism you expect. Then mark four beds of equal size and equal soil: the full mixture, and each species alone. Plant them the same day at the same spacing, and weigh every harvest separately for one season. That gives you a land equivalent ratio and, more usefully, a reason to believe or drop your guild.
Fieldwork
Walk a hedge and list the layers
Find a mature hedgerow, a woodland edge or an old fence line, and spend an hour listing what grows there by layer: ground cover, herbaceous, shrub, small tree, canopy. Note which species sit next to which, which are climbing on which, and where the bare ground is. That hedge is a guild that assembled itself without a designer.
Failure mode
The first failure mode is planting a mixture and crediting it for what the weather did. Without a control plot in the same season, a good harvest proves nothing about the guild. The second is tighter and more dangerous: crowding the same plant family into the same ground year after year concentrates soilborne disease. Beans after beans builds seed-borne blight and anthracnose. Peas after peas builds Aphanomyces root rot, which can take the whole crop. Rotation is the control, and a polyculture is not a substitute for it.