What to Plant and When to Spray

Heirloom nurseries are the only place to buy most of the varieties above, and their provenance writing is often excellent and worth citing. Their disease ratings are a different matter.

We reviewed the Trees of Antiquity catalogue in September 2026, roughly fifty product pages against six extension sources, because they are the best-known heirloom nursery in the country and they carry 217 apple varieties. The history on those pages is first rate: Coxe in 1817, Downing, Thoreau on Blue Pearmain, Lee Calhoun on Magnum Bonum, Oliver Chapin's 1847 letter about Northern Spy, Jefferson writing from Paris that the French had no apple to compare with the Newtown Pippin. Much of the history in the cards above comes from them.

The disease field is another thing entirely. It is a single word, and it does not track the evidence:

• All four Limbertwigs are rated Excellent. No land grant institution has published disease data on any Limbertwig.
• Stayman Winesap is rated Excellent, with the specific claim that it "displays resistance to scab, cedar apple rust and fireblight." Purdue rates it very susceptible to scab and mildew and susceptible to rust. Kentucky and Cornell agree.
• Roxbury Russet is credited with excellent scab and rust resistance. Purdue rates it susceptible to both. The one trait Purdue does credit it with, fire blight resistance, is not claimed, though Cornell's table has no fire blight rating for it to confirm that.
• Rhode Island Greening is rated Excellent. Purdue rates it susceptible on all four diseases, including very susceptible to fire blight. Cornell agrees on scab, rust and mildew; on fire blight Cornell's sources run from moderately resistant to very susceptible, which is typical of fire blight ratings and no support for "Excellent".
• GoldRush and Liberty, the two genuinely scab-immune varieties they sell, are rated merely Good. And nowhere does the catalogue mention that GoldRush is highly susceptible to cedar-apple rust, which in Virginia is the fact that settles whether it is worth planting.

Four of the heirlooms best suited to Virginia are triploids the catalogue does not flag: Roxbury Russet, Blue Pearmain, Calville Blanc d'Hiver and Wolf River. Order any of them together without two diploids alongside and you have a planting with little viable pollen in it. Self-fertility claims are also unreliable there for Newtown Pippin, Cox's Orange Pippin and Pitmaston Pineapple.

One more thing before you order: every apple they sell is on MM.111, a 16 to 20 foot semi-standard that takes four to six years to bear. It is reasonably tolerant of fire blight, which eXtension describes as "quite tolerant", but it is not in the Geneva stocks' class and it is not replant tolerant. There is no dwarfing option and no Geneva rootstock. That is a fine tree on good ground and the wrong tree on a small lot or a replant site.

This comes up constantly with the best low-input cultivars. Sundance has the cleanest four-disease profile on the Kentucky table and almost nobody grows it. Several of the heirlooms above exist in one collection and two catalogues. The answer is that you buy the two halves separately and join them, which is how every grafted tree is made anyway.

Cummins, which grew out of the Geneva breeding programme, sells rootstock and scionwood as well as finished trees, and the arithmetic is not close. In small quantities a rootstock runs about $5.50 and a stick of scionwood $8 to $12, so a tree you graft yourself costs around $15 to $18 in parts. The same tree finished and two years old is $40 to $48. One stick of scionwood carries several buds, so it makes more than one tree.

The catch is a year and a skill. You bench-graft in late winter, usually a whip-and-tongue or cleft graft, then grow the thing on for a season before it is a plantable tree. First attempts fail at a fair rate. Extension services and orchard clubs run grafting workshops in February and March, and scion exchanges are where people trade wood off trees no catalogue lists, which is how most old local varieties still change hands.

Treat a local name as a guess. Wood cut from an old tree comes with whatever name the family or the neighbour remembers, and those names are often wrong. When Washington State University genotyped hundreds of samples sent in under a name from oral history or visual identification, about one in six matched the name, about one in three was a known variety under the wrong name, and about half matched nothing in their reference set. A match on appearance is a hypothesis. WSU's MyFruitTree service runs a DNA test on apple leaf samples for a fee, and it is the only way to know.

Ask for clean scionwood on a sensitive rootstock. Cummins charges a dollar a stick more for it and says plainly that regular scionwood may carry one of the common latent viruses. They name G.16, G.935 and G.814 as the stocks where it matters, and test by grafting onto G.16 as an indicator: the graft survives if the wood is clean. On MM.111 or a seedling stock the question is far less pressing.

Top-working is the other route. If you already have an established tree of a variety you dislike, its root system and trunk are years of growth you would otherwise wait for. Cut it back and graft the new cultivar onto the scaffolds, and you get fruit in two or three years instead of five. It is a better use of a mature, healthy, badly chosen tree than removing it.

Almost every apple sold as "scab immune" carries a single gene, Rvi6, taken from one wild crabapple in 1914. A 2024 review found Rvi6 has been bred into more than 90% of scab-resistant cultivars. Single-gene resistance is exactly the kind a fungus overcomes.

It has been overcoming it. Scab appeared on Rvi6 seedlings at Ahrensburg, Germany in 1984, and the strain was named race 6. Breakdown followed in Belgium, Poland, Spain and Sweden. The international VINQUEST monitoring project reports that Rvi6-virulent isolates now occur all over Europe and have also been observed in the USA.

In North America it has been confirmed once, at a Cornell research orchard in Geneva, New York, on 20-year-old unsprayed trees. Genetically the American virulent isolates are distinct from the European ones, which means the trait arose here independently rather than being imported. In greenhouse inoculation, Macfree produced abundant sporulating scab. In the field, most commercial cultivars held: Dayton, Jonafree, Priscilla, Redfree and Florina stayed clean. The researchers' reading is that those cultivars carry background resistance beyond the single gene. Separately, Purdue lists Prima, Pristine, Jonafree and Enterprise among cultivars where field scab resistance has broken down.

What to do with that. Do not cancel your order. No US extension service reports field failure in home plantings. Extension puts a dwarf apple's life at 15 to 20 years and a semi-dwarf's at 20 to 25, and the European breakdown took 15 to 30 years under heavy commercial pressure, so a backyard tree on a dwarfing stock has a good chance of seeing out its life before the gene fails. A full-sized tree that may live 40 years or far longer is a longer bet. But stop calling these apples immune, scout the leaves in May and June, and if you find sporulating olive-green lesions on a Liberty you now need a spray program on that tree. Autumn leaf litter treatment matters more than it used to, because shrinking the overwintering population shrinks the pool in which virulent strains arise.

The other move is to diversify by mechanism, not by name. Planting four Rvi6 cultivars is planting one gene four times. Grimes Golden's scab resistance does not come from Rvi6: Cornell, Purdue and Kentucky rate it resistant, and whatever genes are responsible, race 6 does not target them. That spreads the risk. It does not prove the resistance will last; nobody has tested it against modern scab populations the way Rvi6 has been tested. Akane and Sansa carry quantitative rather than single-gene resistance. Florina carries two genes rather than one.

One widely repeated claim the literature does not support: that planting a diverse mixture prevents resistance breakdown. VINQUEST and the 2024 review both recommend gene pyramiding and polygenic resistance; neither endorses cultivar mixtures as a resistance-management tool. Diversify anyway, but know the evidence is thin.

• Self-fruitful, one tree is enough: peach and nectarine (except J.H. Hale), tart cherry, most apricot, Stanley and Damson-type European plums, fig, jujube, medlar, and the self-fertile sweet cherries.
• Self-unfruitful, you need two: apple, most pear, most sweet cherry, most European plum, all Japanese plum, pawpaw.
• Triploid apples give no usable pollen. UMaine names them: Baldwin, Boskoop, Bramley's Seedling, Crispin, Gravenstein, Jonagold, Mutsu, Rhode Island Greening, Ribston Pippin, Roxbury Russet, Shizuka, Spigold, Wealthy, Winesap. Plant one of these and you need two other diploid varieties, not one, or all three trees underperform.
• Sweet cherries have incompatibility groups. Bing, Lambert and Napoleon are all group III and mutually sterile. Self-fertile cultivars carry a mutated allele that makes them universal pollen donors, though WSU warns Stella does not reliably pollinate Bing in some areas.
• Plums are three separate systems. European, Japanese and American hybrid plums do not cross-pollinate each other reliably. Toka is the universal donor for American hybrids. Minnesota specifically warns that La Crescent and Superior will not pollinate each other.
• Two pears that will not work together: Seckel and Bartlett are cross-incompatible. And Magness is pollen-sterile, so it needs two other cultivars.
• Distance matters more than people think. Missouri gives hard numbers: semi-dwarf apple pollinizers within 50 feet, dwarf within 20 feet. Bees work down a row and rarely cross between them, so plant partners adjacent, not at opposite corners of the lot.
• Crabapples pollinate apples well. Manchurian for early bloomers, Snowdrift for mid-to-late. But no extension source rates pollinizer crabapples for disease, and a fire-blight-susceptible crab is a permanent inoculum reservoir sitting in your orchard. Choose one with a published disease rating.
• Pawpaw needs hand pollination. Two genetically different trees minimum, and an artist's brush, because the natural pollinators are carrion flies and beetles that do not reliably show up.

1. Order in autumn for spring delivery, and earlier than you think. Good cultivar-and-rootstock combinations sell out by December. A custom graft is a different timescale again: Cummins was taking custom orders for spring 2029 in the autumn of 2026, roughly three seasons out, because someone has to bench-graft your tree and then grow it. Decide a year before you want to plant; a month ahead is too late.
2. Insist on knowing the rootstock. If a nursery cannot tell you, buy elsewhere. Big-box trees usually do not state it, which means you cannot know the mature size or whether it needs a permanent stake. Maryland Extension advises against bargain plants and against multi-graft "5-in-1" trees, where the most vigorous graft takes over and the rest die.
3. Ask whether the stock is virus-tested. Three apple viruses are latent: the tree looks fine and yields perhaps 10 to 30% less, forever. They spread only by grafting, so certification is the only control. It matters most on the virus-sensitive rootstocks, which Cummins names as G.16, G.935 and G.814, and when grafting scionwood off an old backyard tree of unknown status. "Certified nursery stock" in most states just means licensed and inspected, not virus-tested.
4. Know what the grades mean, because they are not marketing. Cummins grades a one-year whip at 5/16 to 3/8 inch caliper and about three feet, and a two-year "Fancy" at over 3/4 inch with five or more branches. The branched two-year tree fruits a year or two earlier and costs roughly a third more. For a backyard planting of three or four trees, buy the branched two-year and skip a year of waiting; for twenty, the price break on whips is worth the wait.
5. Reject on arrival anything with dried, blackened or moldy roots, a cracked graft union, sunken cankers near the union, or no label naming the rootstock. Photograph damage the day it arrives; nurseries replace on a same-season claim and never after planting.
6. Test drainage before you dig. Michigan State's spec is the only quantitative one published: the soil must drain at a minimum of 2 inches per hour. Dig a hole, fill it, and time it. Peaches are the least tolerant of wet feet and die suddenly in year four to seven.
7. On heavy clay, plant on a berm 12 to 18 inches high and 4 to 6 feet wide, built from native soil. A hole backfilled with amended soil in clay becomes a bathtub.
8. Soil test in autumn if planting in spring, so lime has months to move. Target pH 6.0 to 6.5 for pome and stone fruit.
9. Set the graft union 2 to 3 inches above the final soil line after settling, lower on burr-knot-prone stocks. If it gets buried, the scion roots and the tree becomes full size, and any fire blight resistance in the rootstock is lost because the scion's own roots carry the scion's susceptibility.
10. Do not put a new apple where an old apple died. Replant disease is a pathogen complex that costs a tree its first four years. Offset the new tree 6 to 10 feet into ground that never held a tree, which is free and works better than any amendment. If you must replant in place, use a Geneva rootstock: WSU rates G.935, G.30, G.65, G.210, G.41 and G.214 tolerant, and M.7, M.9, M.26, MM.106 and MM.111 susceptible.
11. Cage it against deer and guard the trunk against voles the day you plant. Hardware cloth 12 to 18 inches tall, quarter-inch mesh, pushed 2 to 3 inches into the soil, not tight against the bark.
12. Remove all fruit for the first two or three years. Maryland is explicit. On a very dwarfing rootstock an early crop can set the tree back permanently.
13. Check the stake ties every spring. A tie left in place girdles the trunk within two or three seasons, and that quietly kills more backyard fruit trees than any disease.
14. Map what you planted. Cultivar, rootstock, nursery, date. Ten years on it is the only way to know whether the tree that never sets fruit is a triploid, a mislabeled clone of its neighbor, or on the wrong rootstock. Move the nursery tag off the branch or it will girdle it.

Bagging keeps insects and summer diseases off the fruit with nothing sprayed after the first few weeks. Each fruitlet goes into a small paper bag once it has set, and the bag stays on until close to harvest. Japanese growers have bagged apples, pears, peaches and grapes for decades. It is slow, fiddly work, and it only makes sense on dwarf or trained trees you can reach from the ground.

The method below is the University of Kentucky's home-orchard version (Ric Bessin and John Hartman, reprinted in POMONA, Summer 2001).

1. Cover the early season first. Scab, rust and fire blight strike before the fruit is big enough to bag, and plum curculio starts cutting fruitlets at petal fall. Plant resistant varieties and keep the kaolin passes going through first cover.
2. Thin and bag together, when fruitlets are ½ to ¾ inch across, about three weeks after petal fall. Leave one fruit per cluster, six to seven inches apart.
3. Use a proper bag or make one. Two-layer Japanese apple bags are made for this. A 3-pound grocery bag cut down to about 6 inches, slit 1½ inches at the mouth for the stem, pleated shut and closed with a twist tie does the same job. Clemson University sells fruit bags with the tie built in, about 20 cents each, from mid-February to mid-June.
4. Take apple bags off about three weeks before harvest if you want red color. Left on, even red varieties ripen pale yellow.

Two things decide whether it works. Whatever is already on the fruit keeps going inside the bag: a curculio scar, a codling moth egg, a scab infection that has not shown yet. And the bag has to be snug around the stem, or insects walk in.

What growers reported. In Decorah, Iowa, David Sliwa bagged 500 apples on 25 trees in 2001 with stapled grocery bags, at about two and a half minutes a fruit. About 5% of the bags blew off, 5% of the fruit had bird damage, and only about 40% came off clean (POMONA, Summer 2002). In Ann Arbor, Michigan, Dave Strayer has used Clemson bags since 2020 on dwarf and espaliered peaches, plums, pears and the scab-resistant apples Williams' Pride, Akane and GoldRush, with no sprays at all. Most of his fruit is good to excellent. Damaged fruit has run from almost none to about 30% in his worst year, mostly curculio scars made before the bags went on; the plums he missed were all lost to curculio, and he snaps the stems on 5 to 10% of pears while bagging (POMONA, Summer 2024).

Robert Norton of Washington State University expected bagged apples to look better and taste slightly less intense, and warned that varieties with very short stems or fruit held tight on spurs, such as Macoun and Cox's Orange Pippin, are very hard to bag (POMONA, Fall 1994). These are grower reports, not trials. Bag one tree first and keep notes.

How cold is too cold depends on the bud, not the date. A swelling bud shrugs off temperatures that would kill an open flower outright. The table gives the temperature that kills 10% and 90% of flower buds when held for 30 minutes, from Washington State's critical-temperature tables. Losing 10% of the buds costs you nothing on a tree that needs thinning anyway; 90% is the crop.

Each crop was measured on a reference cultivar, and cultivars differ by a degree or two; Virginia Tech prints the apple figures as ranges for that reason. Nobody publishes these by cultivar. Tart cherry is left out because its published figures are internally inconsistent. The apple figures also appear on each bud-stage card in the spray calendar.

Losing the crop is not losing the tree. These are bud temperatures in spring. The wood survives far colder weather in midwinter: Ontario's ministry puts peach flower buds at about −13 °F and the peach tree itself not until below about −20 °F, and apple and pear wood is hardier than peach. In northern Virginia a frost in April takes the crop; it very rarely takes an apple or pear tree.

What the orchard floor does on a frost night. Penn State's figures, against bare, undisturbed, moist soil:

That puts two pieces of advice on this page in tension. The flowering understory recommended for beneficial insects is tall cover at exactly the wrong moment. Mow it short before bloom, which you are already doing so the bees are not in the spray, and let it grow back after petal fall. Wood chip mulch insulates the soil too, so rake it back from under a small tree on a night when a hard frost is forecast.

Two kinds of frost, and they behave oppositely. A radiation frost comes on a clear, still night: the ground loses heat to the sky, cold air slides downhill and pools, and a few feet of elevation or a slope with somewhere for the air to drain makes a real difference. That is the frost that site choice, bare ground and row cover protect against. An advective freeze is a cold air mass arriving on a wind above about 5 mph. Nothing pools, ground cover makes little difference, and higher ground is no warmer and can be colder. The usual advice to plant high to escape frost is advice about radiation frost only.

Look for the dam. Cold air ponds behind anything downslope that stops it draining: a hedgerow, a road embankment, a solid fence, a tree line at the bottom of the field. The pond fills uphill to roughly the height of that obstruction. Opening a gap in a downslope hedge can drain a frost pocket; planting a windbreak across the bottom of a slope can create one.

Old varieties and frost pockets. In April 1983 a freeze with 30 mph winds caught John Creech's young trees at pink, in the Kentucky mountains near Cumberland, and he wrote down how every variety cropped. Most commercial apples, Red Delicious and Stayman among them, set almost nothing, and so did Arkansas Black, Roxbury Russet, Smokehouse, Blue Pearmain and Golden Russet. Grimes Golden, Newtown Pippin, Calville Blanc and Virginia Beauty set 5 to 20% of a crop. Rome Beauty, Esopus Spitzenburg and old local apples such as Milam, Pound Pippin, Benham and Mountain Boomer set 20% or more. His explanation, which he called speculation: old homestead orchards sat in creek valleys where cold air pooled, so only late-blooming or frost-tolerant varieties kept bearing there, and the rest were cut down or regrafted (POMONA, Winter 1983). One freeze, one orchard, young trees. It is still the only variety-by-variety frost record for the Southern heirlooms we have found.

Do not spray it. A newly planted tree has no crop to protect and no disease history, and a spray program in year one is money spent on the wrong problem.

Water it instead, about five gallons a week through the first two summers. Mulch out to the drip line and keep the mulch four to six inches off the trunk. Cage it against deer and guard the trunk against voles. Remove any fruit that sets for the first two or three years so the tree builds structure instead. Start the spray program in year three, when there is fruit worth defending.

• Spraying persimmon, fig or pawpaw on a program. They do not need one.
• Buying mycorrhizal inoculant for an established tree in living soil. It is almost certainly already colonized. Spend the money on wood chips.
• A bloom biological in a cool, dry bloom. If no model shows infection risk, skip it.
• Boron, unless a leaf test came back under 20 mg/kg. The useful and toxic ranges sit close together.
• Compost tea, unless you will actually control the feedstock and brew it fresh. A badly brewed tea is worse than none.
• Fall nitrogen on a vigorous tree. It buys next year's fire blight.
• Season‑long kaolin. Use it through the curculio window and stop, because continuous use raises plant‑feeding mite numbers.

Brown marmorated stink bug has been the defining tree fruit pest of the mid‑Atlantic since about 2010, and there is no good organic answer. Kaolin helps somewhat. Perimeter plantings and trap crops help somewhat. Expect some damage and plan for it rather than reading a guide that leaves it out.

Spotted lanternfly is established in northern Virginia. It does not usually kill fruit trees, but it feeds heavily on tree of heaven, so removing that from your property line does more than anything you can spray.