Horticulture: Gardens, Orchards, and Propagation

Difficulty: easy to start, and the results arrive on a scale of years. Prerequisites: a sharp knife, water, and land you will still have access to in a decade. Agriculture covers the field crops; this page is the garden and the orchard.

Why it matters

Agriculture and crop rotation is about staple calories at field scale — rotation, soil, water, and tools. That is where the bulk of the food comes from, and it is deliberately about grain.

This page is about everything else you eat, and about how you keep a good plant once you have one. Vegetables, fruit, herbs, and the intensive cultivation that feeds a household from a small area. It also carries the single technique without which good fruit cannot exist as a category, which is grafting.

The reason it deserves separate treatment is that horticulture works by different rules from field cropping: small areas worked intensively, high value per square metre, perennial plants that take years to establish and then produce for decades, and — crucially — reproduction by cloning rather than by seed.

Part 1: The thing most people get wrong — seed does not preserve a variety

This is the central fact of the page and it is genuinely counterintuitive.

Most fruit trees do not come true from seed. Plant the pips from an excellent apple and you get seedlings that are genetically a fresh shuffle, mostly mediocre, frequently inedible. The good apple was a chance seedling that somebody noticed. If it had been left to reproduce sexually, it would have vanished within a generation.

Every named apple, pear, plum, cherry, fig, and grape variety in history is a clone, maintained by taking a piece of the original plant and growing it on. Some are astonishingly old — varieties propagated continuously for centuries are genetically the same individual as the tree someone first noticed.

So horticulture divides cleanly:

• Reproduce by seed where the plant comes reasonably true and you want the variation: most annual vegetables, grains, and herbs.
• Reproduce vegetatively (clone) where you have selected an individual worth keeping: fruit trees, vines, soft fruit, potatoes, and most perennials.

Part 2: Vegetative propagation — cloning without special tools

Cheapest first.

• Division. Dig up a clumping perennial and pull or cut it into pieces, each with roots and shoots. Rhubarb, many herbs, and most clump-forming plants. Nearly always works.
• Runners and offsets. Some plants do it for you — strawberries put out runners that root themselves; bulbs and tubers make offsets. Detach and replant.
• Layering. Bend a low branch to the ground, wound the underside slightly, peg it down, and heap soil over it. It roots while still attached to and fed by the parent, then you sever it. The most reliable method for a beginner, because the cutting cannot dry out or starve while it roots. Air layering does the same on a branch too high to bend, by packing damp moss around a wounded section and wrapping it.
• Hardwood cuttings. Take pencil-thick dormant shoots in winter, roughly 20–30 cm, cut square at the base and sloped at the top so you can tell which way up, and push two-thirds into a trench of gritty soil. Willow, poplar, currants, gooseberries, figs, and grapes root easily this way; many other species do not.
• Softwood cuttings. Green growing tips in the growing season. Root faster but wilt easily — keep them humid and shaded, and reduce the leaf area by trimming the leaves in half so they lose less water than the cutting can replace.
• Root cuttings. Sections of root laid in soil. Works for a smaller set of species.

Willow water: willow shoots steeped in water are a traditional rooting aid, and willow does contain compounds related to root-promoting hormones. Treat it as a cheap possible improvement rather than a proven one — the plants that root easily will root without it.

Part 3: Grafting — the technique that makes orchards possible

Joining a piece of one plant (the scion, chosen for its fruit) onto the root system of another (the rootstock, chosen for vigour, size, soil tolerance, or hardiness), so they grow together as one plant.

Why it matters beyond cloning: the rootstock controls the tree. A dwarfing rootstock gives a small tree that fruits young and can be picked from the ground; a vigorous one gives a large long-lived tree. The same scion variety on different rootstocks gives you completely different orchards. It also lets you put a tender variety on a hardy root, or grow a fruit on soil its own roots would not tolerate.

The one principle that decides success

The cambium layers must meet. Cambium is the thin green living layer just under the bark, between bark and wood; it is the only tissue that divides and knits the two pieces together. Align that thin layer on scion and stock, hold it firmly, and keep it from drying, and the graft takes. Miss it — align the bark edges by eye without thinking about the cambium — and nothing happens no matter how neat the cut looks.

Everything below is a different way of getting cambium to meet cambium.

Methods

• Whip and tongue. For scion and stock of similar thickness. Make a long sloping cut on each, then a small downward slit in each face so they interlock like a tongue and groove. Maximum cambium contact, mechanically self-holding, and the standard bench graft.
• Cleft graft. For a stock much thicker than the scion, including re-working an established tree. Saw the stock off, split the cut end, and insert wedge-shaped scions at the edges so their cambium lines up with the stock's cambium near the bark — not in the middle of the split, which is the usual beginner's error.
• Whip / splice graft. Two matching slanted cuts bound together. Simplest.
• Bud grafting (budding). Insert a single bud with a shield of bark under a T-shaped cut in the stock's bark, in late summer when the bark lifts easily. Extremely economical — one scion shoot yields many buds — and it is how nursery fruit trees are mass-produced.
• Approach graft. Grow both plants side by side, wound facing surfaces, bind them together, and sever the unwanted parts once united. Slow, nearly foolproof.

Doing it

1. Take scion wood while dormant, in winter, from healthy, vigorous one-year-old shoots. Store it cool and slightly damp — buried in moist sand or wrapped in damp moss in a cold place. It must stay dormant and not dry out.
2. Graft in late winter or early spring, as the stock is waking but the scion is still dormant. That difference is what drives the union.
3. Cut cleanly with a very sharp blade. A ragged or crushed cut will not knit. See sharpening in woodworking and joinery.
4. Bind firmly with raffia, waxed linen thread, or strips of cloth.
5. Seal all cut surfaces with grafting wax — beeswax, resin, and a little tallow, warmed — or clay and dung. Drying out is the commonest cause of failure, more than any error in cutting.
6. Match compatible species. Apple onto apple or crab apple; pear onto pear or quince; stone fruit onto related stone fruit. Wildly unrelated species do not unite, and claims of grafting anything onto anything are folklore.
7. Remove growth from the rootstock below the graft, repeatedly, or it will outcompete the scion and you will end up growing the rootstock.

Part 4: Seed saving, done properly

For everything you do grow from seed, saving it well is free and compounds.

• Know how the crop pollinates. Self-pollinating crops — peas, beans, lettuce, tomatoes — come reasonably true and are easy to save. Cross-pollinating crops — brassicas, squash, maize, beets, onions — readily cross with other varieties and with wild relatives, and the seed will not be what you planted.
• Isolate cross-pollinators: distance, timing (plant varieties so they flower at different times), or physical covering with hand pollination. Growing two varieties of squash side by side and saving seed is how you get neither.
• Save from many plants, not one. Taking seed from a single individual narrows the genetic base fast and brings on inbreeding problems in cross-pollinating crops.
• Select deliberately, and select for the right thing. Save from plants that performed well and stayed true to type. Note the trap: if you eat your best fruit and save seed from the leftovers, you are selecting backwards. The same logic as the breeding sections in agriculture and animal husbandry.
• Remember biennials. Carrots, beets, cabbages, and onions flower in their second year, so you must overwinter the plant to get seed at all.
• Dry thoroughly, store cool, dry, and dark, in sealed vessels against damp and vermin. Label everything — see mathematics and record-keeping.
• Test germination before sowing season. Count out a sample onto damp cloth, keep it warm, and count how many sprout. Knowing you have 40% viability lets you sow thicker instead of losing a season to a bare bed.
• Seed longevity varies enormously by species — some remain viable for many years, others for barely one. Rotate stock and grow out old varieties before they are lost.

Part 5: The garden itself

• Concentrate effort. A small, intensively worked, heavily manured bed close to the house outproduces a large neglected one and is easier to protect, water, and harvest. Site it where you pass it daily.
• Raised or deeply dug beds with heavy compost warm earlier, drain better, and let roots go deep, so plants can be spaced closer.
• Succession sowing. Sow small amounts repeatedly rather than everything at once, or the whole crop matures in one week and then there is nothing.
• Intercropping. Fast and slow crops together, or the maize-beans-squash arrangement described in agriculture.
• Rotate the garden too. Brassicas, alliums, roots, and legumes each have their own soil pests and diseases, and clubroot and onion white rot persist in soil for years once established.
• Mulch to hold moisture and suppress weeds.
• Water at the roots, in the morning or evening. Watering foliage in hot sun wastes most of it to evaporation and encourages leaf disease.

Extending the season — genuinely valuable, because the hungry gap is late winter and early spring:

• Cold frame — a low box with a translucent lid, facing the sun.
• Hotbed — a pit packed with fresh manure and straw, topped with soil. The manure ferments and produces heat for weeks, warming the bed from below. This is a real, effective, fuel-free technology for raising early seedlings and was the backbone of market gardening before glasshouses.
• Cloches and covers for individual plants; walls that store and re-radiate heat, which is why fruit was traditionally trained against them.
• Windbreaks and hedges, which raise effective temperature and cut damage.

Part 6: Orchards

• Site for air drainage. Cold air flows downhill and pools; a frost pocket at the bottom of a slope will kill blossom in spring and cost you the crop. Mid-slope is usually best.
• Plant for pollination. Many fruit trees are self-incompatible and need a different, compatible variety flowering at the same time nearby to set fruit. A lone tree of a single variety may flower magnificently and fruit hardly at all. Insect pollinators do the moving — see the beekeeping section in animal husbandry.
• Prune with a purpose: remove dead, damaged, diseased, and crossing wood first; open the centre so light and air reach the fruit; and keep the tree to a size you can pick and protect. Most fruit is borne on wood of particular ages, which differs by species — learn the local practice rather than cutting by instinct.
• Train against walls and wires (espalier, cordon, fan) where space is tight or the climate marginal.
• Thin the fruit in heavy years. It sounds wasteful and it gives larger fruit, less branch breakage, and reduces biennial bearing, where a tree exhausts itself with a huge crop and then produces almost nothing the following year.
• Graze or mow beneath, and keep fallen fruit cleared, since it harbours pests and disease over winter.
• Plant for your grandchildren. A standard fruit tree may take five to ten years to crop properly and then produce for generations. This is the longest feedback loop in this repository, which is exactly why it needs written records.

Part 7: Pests and disease, honestly

• Hygiene first. Clear fallen fruit and diseased material, and do not compost diseased plants unless the heap genuinely gets hot — see agriculture.
• Rotation and diversity break pest cycles. Monoculture concentrates the problem.
• Physical barriers — netting, fleece, grease bands on trunks, collars, and simply picking pests off by hand — are reliable and need no chemistry.
• Encourage predators. Birds, hedgehogs, ground beetles, and parasitic wasps do a great deal of work if you leave them habitat.
• Resistant varieties are the durable answer, and they come from the seed saving and selection above.
• On companion planting: some pairings have plausible mechanisms — a strong smell masking a host plant, a plant drawing pests away as a trap crop, a legume fixing nitrogen for a neighbour. Much of what circulates is untested folklore. Treat it as worth trialling and recording rather than as established fact, and run the comparison properly: two beds, one variable, written results.
• Sulfur dusted on vines controls powdery mildew and is still used — see sulfur. Note the hazards there.

Failure modes

Era notes

Grafting is ancient — described in Greek and Roman agricultural writing and practised in China well before that, and Roman authors already discuss rootstock choice. Layering and cuttings are older still and probably arose wherever people noticed a fallen branch had rooted. Systematic dwarfing rootstocks and named, characterized rootstock series are a much later development, largely 19th and 20th century. Hotbeds heated by fermenting manure were standard market-garden practice in Europe well before glasshouses, and the intensive Parisian market gardens of the 19th century produced remarkable yields from small areas using little more than manure, glass, and labour.

Your knowledge advantages: the cambium rule, which explains why grafts take and turns a mystery craft into a repeatable one; the fact that fruit does not come true from seed, and therefore that any good chance seedling must be cloned immediately or lost; rootstock as a deliberate control on tree size and soil tolerance; pollination compatibility, so an orchard is planned rather than hoped for; isolation distance and saving from many plants in seed saving; germination testing before the sowing season; cold-air drainage in siting; and the discipline of trialling folk practices as comparisons rather than adopting or dismissing them wholesale.

Sources and limits

Species-specific detail — which rootstock suits which soil and climate, pruning by fruiting habit, pollination groups, isolation distances, and local pests and diseases — varies enormously and is exactly where local horticultural knowledge and current regional guidance outrank a general description. Plant-health and quarantine rules restrict moving plant material, scion wood, and soil in many jurisdictions, and moving material is a common way diseases spread. Nothing here is a substitute for regional horticultural advice. Quantitative and historical claims on this page are not yet individually cited — see the coverage note in SOURCES.md.