Skill guide · Water & Food

Growing in a Food Forest

Building a Food Forest covers the design: the seven layers, the establishment timeline, and which permaculture claims actually hold up. This guide picks up once that design is in the ground: how to actually plant the tree itself, what to plant around it and how far apart, how and when to prune the fruiting layers with the right tool, how to spot the pests and disease that actually show up, how to tell fruit is ready, what polyculture pest management can and can't do, and how to stagger harvest across layers that mature on completely different timelines.

Download this guide as PDF

Planting the tree itself: the step guild design skips over

Building a Food Forest covers layout and spacing; this is the actual mechanics of getting a tree into the ground so it survives to be worth building a guild around.

Dig the hole no deeper than the root system itself, planting too deep is the single most common cause of a new tree failing, and about two to three times as wide as the roots' spread for a bare-root tree, or two to three times the width of the container for a potted one, the same ratio either way. The goal is a hole depth equal to the distance from the trunk's flare (where it visibly widens into the roots) to the bottom of the root mass, so the flare sits right at, or very slightly above, the final soil surface once backfilled; a grafted tree's graft union (the visible bulge or scar low on the trunk) should sit 4–6 inches above the soil surface, or 2–3 inches for a dwarf or semi-dwarf rootstock. Backfill with the same native soil that came out of the hole, broken up but not heavily amended, research consistently finds trees establish just as well in unamended native soil, and firm it gently as you go to remove large air pockets without compacting it hard. Water in thoroughly right after planting to settle the soil around the roots.

Staking is the exception, not the default: most young trees don't need it, and a little trunk movement in the wind actually builds the reaction wood that makes a trunk structurally strong over time. Stake only on a genuinely exposed, windy site or with an extremely dwarf rootstock that can't support its own top growth yet, and remove the stake within a year once the tree can stand on its own.

Guild composition: what actually goes around a fruit tree

A guild is a small cluster of plants placed around a central fruit tree so each does a real, specific job: something to fix nitrogen into the soil, something to attract pollinators or the insects that eat pests, something to physically compete out weeds, and something to deter a specific animal pest.

JobReal examplesWhat it actually does
Nitrogen-fixer White clover, blue wild indigo, goumi or sea buckthorn at the drip line Bacteria in these plants' root nodules convert atmospheric nitrogen into a form the soil, and the tree, can use
Pollinator/predator attractor Yarrow, anise hyssop, mountain mint, borage Draws in bees for fruit set, and predatory or parasitic insects that keep aphids and other pests in check
Groundcover/weed suppressor White clover, low-growing herbs Physically occupies the soil surface a weed would otherwise colonize, and holds moisture
Animal deterrent Daffodils at the inner ring Toxic to voles and deer, planted close to the trunk specifically to discourage them from girdling bark or roots

Comfrey shows up in almost every published guild design, and it earns its place for a real reason: it's an excellent chop-and-drop mulch plant, cut back several times a season and dropped at the base of the tree as green mulch. That's a genuinely useful practice on its own. It is not, as Building a Food Forest already covers, evidence of comfrey mining nutrients from deep soil layers, a specific mechanism the scientific literature doesn't actually support. Plant it for the mulch, not the folklore.

Where each ring actually goes: keep herbaceous guild plants at least 24 inches out from the trunk itself, closer than that competes directly with the tree's own root flare. Animal-deterrent plants like daffodils go in the innermost ring, 6–12 inches from the trunk on a young tree. Most of the working guild, nitrogen-fixers, attractors, and groundcover, sits from there out to the tree's drip line, roughly 6–15 feet from the trunk depending on the tree's eventual size; that's the zone the tree's own roots actively forage in, so it's also where a guild does the tree the most good. Plant the guild at the same time as the tree or within its first season, rather than waiting for the tree to establish first; an empty ring just becomes bare soil a weed colonizes before the guild ever gets planted.

Using the establishment years for the staples this site actually recommends

Building a Food Forest's own labor curve shows years 0–3 with full sun still reaching the ground between young, not-yet-productive canopy and understory trees, exactly the gap What to Grow's recommended annual staples can use instead of sitting bare or going to weeds. Corn, dry beans, and winter squash, three of that guide's five recommended crops, are the classic Three Sisters combination: corn gives the beans something to climb, the beans fix nitrogen the whole planting uses, and squash's broad leaves shade out weeds and hold soil moisture, the same real, evidence-backed pairing covered in full in Row Crop & Raised Bed Gardening. Plant it in the sunny gaps between young trees during exactly the establishment years when a food forest isn't feeding anyone on its own yet.

Potatoes fit the same establishment-year logic in the root layer, in loose soil between young trees rather than fighting the developed root systems mature trees eventually have. Cabbage and other brassicas are less of a fit here: they're a true annual bed crop most years, better suited to the row-crop or raised-bed method than woven into a perennial guild, though a bed of them in a sunny establishment-year gap works exactly like it would anywhere else. None of the five recommended staples belongs in the mature system's own layers once canopy closes; that's the tradeoff Building a Food Forest's labor curve already describes, years of open ground giving way to a closed, low-labor perennial canopy that doesn't have room for row crops anymore.

Secondary crops: one belongs here permanently, the rest just for establishment

What to Grow's six secondary crops split the same way the five staples do, mostly establishment-year fillers, with one real exception.

Culinary and medicinal herbs are the exception, and they're not really a workaround here at all: herbaceous plants, including herbs, are already one of the seven real layers in Building a Food Forest's own design. Perennial herbs (oregano, thyme, chives, many mints) belong in the mature system permanently, doing guild work (pollinator attraction, groundcover) and kitchen work at the same time. Annual culinary herbs (basil, cilantro) behave like the establishment-year annuals below instead, useful in the sunny gaps while they last, then replaced as canopy closes.

Carrots, onions, and garlic follow potatoes into the root layer during the establishment years, for the same reason: loose, undisturbed soil between young trees rather than a fight with mature root systems later. Sunflowers and tomatoes or peppers follow corn and squash into the sunny establishment-year gaps, useful annual production while the canopy fills in, and equally not something the mature, closed system has room or light for afterward. None of these five change the underlying tradeoff Building a Food Forest already describes: real annual production for a few years, in exchange for a low-labor perennial system once the canopy actually closes.

What the evidence actually says about polyculture and pests

The honest picture is real but more limited than permaculture writing sometimes implies. Reviews of the research find that diverse plantings genuinely reduce specialized pest loads and slow the spread of insect-transmitted plant viruses, largely through two real mechanisms: a pest has a harder time finding its one preferred host plant in a mixed planting than in a monoculture block of it, and a more diverse planting supports more of the predator and parasitoid insects that eat pests. That's a real, evidence-backed effect, not folklore.

What the same research is honest about: results are inconsistent crop to crop, some companion plants can themselves harbor pests or disease rather than deterring them, and the mechanism is genuinely more complex and less predictable than a simple pairing chart suggests. Treat guild planting as a real, useful layer of pest management, not a substitute for watching your own trees and stepping in when something specific shows up.

Pruning the fruiting layers

Prune fruit trees and cane berries while they're dormant, after the coldest part of winter has passed but before new growth starts, typically late January through early March in northern Utah. Pruning too early in fall or winter, or too late once growth has resumed, both increase the tree's exposure to cold damage or disease. Dormant-season pruning has three real advantages: most tree diseases are inactive too, so cuts are less likely to introduce infection; the tree isn't putting energy into leaves you're about to remove; and bare branches make the actual structure, and any dead or crossing wood, easy to see.

Finding the branch collar in practice: look for a slightly raised ridge in the bark where the branch meets the trunk, and often a subtle swelling just below it, that's the collar. The branch bark ridge (where the bark itself rises up in a line mirroring the branch's angle) marks the top edge of the correct cut; angle the cut down and away from the trunk from there, just outside the collar's swelling, never flush against the trunk itself.

Branch diameterRight tool
Up to 1 in. Bypass hand pruners (the curved-blade type that slides past a lower blade like scissors, not the flat anvil type, which crushes tissue)
Up to 2 in. Bypass loppers, the long-handled version of the same cutting action
Up to 3 in. or a bit more A pruning saw
Larger than 3 in. Consider whether the cut is really necessary; a cut this large on a home fruit tree is a bigger wound than most guild-scale trees need or recover from quickly

A light summer pass, removing water sprouts (the vertical, non-fruiting shoots that spring up on established wood) and anything blocking light to ripening fruit, is useful in addition to dormant pruning, not instead of it; avoid heavy cuts late in the growing season, which pushes new growth that won't harden off before winter.

Spotting trouble before it costs the harvest

Three problems account for most home fruit-tree losses.

ProblemHow to identify itWhat to do
Aphids Small green, black, or grayish insects clustered on new growth and leaf undersides, sticky honeydew and curling or yellowing leaves are the visible tell even before you spot the insects A strong water spray dislodges light infestations; insecticidal soap for heavier ones. A healthy guild's predator-attracting plants help keep this in check on their own over time
Coddling moth Small (roughly 3/8 in.) grayish-brown moths and, in the fruit itself, brownish threadlike tunnels toward the core with crumbly frass (insect droppings) at the entry hole, that damage inside the fruit is usually the first sign a grower actually notices Pheromone traps hung at bloom time catch males and give an early warning; picking up and destroying dropped fruit promptly breaks the larval cycle before it completes
Fire blight Blackened, wilted shoot tips that curl into a distinctive hook shape, sometimes called a shepherd's crook, plus dark, sunken cankers on branches that can ooze Cut at least 8-12 inches below the visible damage, into clean wood a year or more old, since the bacteria travels ahead of what's visibly infected; disinfect pruning tools between every single cut with rubbing alcohol or a bleach solution, not just between trees. Prune during dry weather if at all possible, wind and rain are how the bacteria actually spreads from an infected cut to the next one. There's no cure once a branch is infected, only removal

Knowing when the fruit is actually ready

Unlike an annual vegetable, most tree fruit doesn't have one obvious cue, and getting it wrong in either direction (picking too early, or leaving certain fruit on the tree too long) genuinely costs flavor and storage life.

FruitReal ripeness cue
Apples Lifts free from the branch with a gentle upward twist rather than needing to be yanked; background skin color has shifted from green to yellow, and seeds inside are turning brown to black rather than white
Pears Don't rely on feel or color, ripe and unripe pears are both still firm on the tree and most varieties don't change color. Instead, cup the fruit and lift gently; if the stem releases easily, pick it. Pears are meant to finish ripening off the tree, not on it, left too long on the branch they turn mealy and lose flavor
Peaches No green left around the stem, the sun-facing side has colored to a dusky red and the shaded side to yellow; the fruit gives slightly to a gentle squeeze and releases with a light upward twist rather than a hard pull

Harvest staggers across the layers, by design, not by accident

A single-crop annual garden bunches its harvest into a few weeks. A mature food forest spreads it across the whole season almost automatically, because the layers themselves mature on different timelines: herbaceous and groundcover plants often produce something within the first year, berry shrubs typically bear in their second or third year, and canopy fruit and nut trees take considerably longer to reach real production. Within a single season, that same spread repeats: early greens and herbs come on before the shrub layer's berries ripen, which finish well before the canopy's fruit does. Track roughly what's ready by layer and month for your own site after the first couple of seasons; a mature system rewards knowing its own actual rhythm more than following a generic calendar.

Warning

A guild doesn't mean 'plant it and walk away'

Guilds reduce maintenance relative to a bare fruit tree in turf, they don't eliminate it. Comfrey and other chop-and-drop plants still need cutting back on a schedule to actually deliver their mulch value; nitrogen-fixing shrubs still need occasional pruning to keep them from shading out the tree they're meant to support; and a guild that's colonized by an aggressive spreader (mint is the classic example) needs the same intervention any garden does when one plant takes over. Budget real, if reduced, ongoing attention, per the labor curve in Building a Food Forest, not zero.

Sources