Seeds and cuttings I collect around North Carolina, what they turned out to be, and how to grow them for a NC Piedmont native-plant nursery.
Fothergilla major · Hamamelidaceae
White bottlebrush spikes with no petals at all. A rare rocky-ridge shrub that is only obvious for three weeks in April.
Community photos of Fothergilla major to study before I head out: leaf, flower or bract, ripe fruit, bark, whole plant.
Native but genuinely rare. Global G3, NC state rank S3, state status SR-T (Significantly Rare Throughout), tracked by the NC Natural Heritage Program. Vascular Plants of NC: rare over its fairly wide range, and in most sites you may see only a handful to a few dozen plants. 24 NC counties hold natural occurrences. Person is one of them but sits right on the edge of the far northeastern Piedmont gap where there are no records at all. iNaturalist has zero Fothergilla observations of any kind in Person, Caswell and Granville. Durham and Orange counts are large but are campus and street plantings
Best bet: Cuttings, clearly. Seed carries double dormancy: about 6 months warm-moist then 3 months cold, and Clemson puts germination at 9–15 months. Softwood cuttings taken June through mid-August root readily with hormone, which is how NCBG propagates their own plant. Layering also works on a suckering clump. Do not take wild material from the known sites
Native to the NC Piedmont, so hardiness is a non-issue, but the wild sites are thin acidic rocky soil on cool aspects and it has low drought tolerance. In a Roxboro garden that means an amended, sharply drained acidic bed with consistent moisture, morning sun and afternoon shade. Full sun gives the best bloom and fall color but only if water is reliable. Heavy unamended Piedmont clay in full afternoon sun will not hold it. Slow to medium growth
Each look-alike with the one character that settles it. Tap a photo to zoom.
Statistically the most likely thing to find. Weakley notes nearly all cultivated Fothergilla here is this pentaploid hybrid, and there is no published key character to split it from F. major. Provenance decides it: no wild pentaploids exist, so a mulch ring, bed edge, row spacing or trailhead kiosk means hybrid. Backup is size class, 3–6 ft waist-to-shoulder for the hybrid against 6–12 ft for wild F. major
Coastal Plain only, north to Johnston and Beaufort counties, and every Piedmont plant is planted. Flower Carolina near Saxapahaw sells it, so expect it in yards there. Blades 1.0–6.2 cm wide with a width-to-length ratio under 0.5, base not V-cordate, 2–3 ft tall, and it flowers before the leaves rather than with them
The one actually growing alongside it, and vegetatively near-identical. Lateral veins stay buried in the blade or are exposed under 1 mm at the margin, against 2–3 mm in F. major. Backup: flowers are yellow with four crinkly strap petals in Oct–Dec, the leaf base is distinctly asymmetric, the margin is crenate all round, and it reaches small-tree size at 15–30 ft
Closest leaf match on this list: same family, same 2–4 in strongly veined leaves, grayish beneath. Flowers separate them. Corylopsis hangs pale yellow bell-shaped 5-petaled flowers in dangling racemes on bare branches; Fothergilla holds a stiffly erect white petal-less spike. Backup: Corylopsis is dentate around the whole margin with a tooth per vein, and the blade is thin and papery rather than leathery
Same family, blooms late Mar–Apr in the same window, and planted in huge quantity across the Piedmont. Evergreen to semi-evergreen ends it: leaves in January means not Fothergilla. Backup: four wavy strap petals, margin entire to finely toothed, and the common cultivars carry burgundy foliage and hot-pink flowers
Both put a white cylinder on a Piedmont shrub. Itea's raceme is a long slender arching 3–6 in cylinder of individual 5-petaled star flowers in May–June; Fothergilla's is a short stiff erect 1–3 in thimble of naked stamens in April. Backup: Itea leaves are elliptic, 2–3x longer than wide, finely serrate to the base, green beneath. Don't use moisture to separate them, since F. major also grows on riverbanks and seepy margins
The tightest leaf match in the right habitat: pale beneath with a whitish waxy coating, margin mostly smooth but sometimes serrate near the tip, and it carpets the same dry ridgetops. Height and fruit decide it. V. pallidum is a knee-high 2–3 ft colonial with pendulous urn-shaped flowers and blue-black berries in June–July; F. major is a 6–12 ft multi-stem shrub with a dry two-horned capsule
The alder the common name refers to, and summer leaf photos of it get called Fothergilla constantly. Alder carries woody cone-like strobiles and long drooping catkins almost year-round; Fothergilla never has either. Backup: alder is green beneath with no bloom, doubly serrate right to the base, base cuneate and symmetric, buds stalked with exactly two scales, and it sits on streambanks and swamp edges rather than dry ridges
Three separate problems here, and I should be honest about all of them before planning a trip.
First, only one species is in play. F. gardenii is a Coastal Plain pocosin plant and never occurs wild in the Piedmont. F. parvifolia and F. milleri, both split out in 2020, don't reach NC. So the target is F. major and nothing else.
Second, nearly everything I will actually meet is planted, and most of that isn't even a species. Weakley states it plainly: almost all cultivated Fothergilla here is F. ×intermedia, the pentaploid hybrid, sold as 'Mount Airy'. Durham shows 51 iNaturalist records and Orange shows 176, and every one I checked sits on a campus, a street, or the Duke Gardens. Zero credible wild records in either county.
Third, it is rare. G3, S3, SR-T, tracked by NHP. Person County has one record ever, LeGrand 2005, with the locality withheld precisely because of that status. Caswell and Granville have zero. Treat any wild population as look-don't-take — the known sites are a county park, a land-trust preserve, and a restricted Plant Conservation preserve, and none of them permit collecting.
So this is a low-hit-rate hunt locally. Worth doing anyway, because a Person County relocation would be real data, and because the habitat signature is specific enough to search deliberately.
Season matters more here than for almost anything else on the list. April is the month. 62 of 134 NC records fall in April, against 6 in August. In bloom it is unmistakable at thirty yards: erect white spikes on still-bare or just-leafing branches, honey-scented, on a 6–12 ft multi-stem shrub. Out of bloom it is a generic Hamamelidaceae shrub that needs a lens.
Leaf width is the fastest character. Weakley's key turns on it: some leaves on a branch exceeding 5.2 cm wide means F. major. The blade is roughly 2–4 in long and almost as wide, leathery and quilted, with coarse single teeth confined to the outer half and an entire lower half. Base V-cordate or V-rounded, with the lowest pair of secondary veins diverging in a V. Underside grayish-white and glaucous with stellate hairs along the veins.
The lens character is vein exposure at the margin. In F. major a lateral vein runs exposed along the margin for 2–3 mm before it enters a tooth. In witch-hazel it stays buried, or shows under 1 mm.
In flower, April to May: a terminal spike 3–6 cm, white, no petals at all. What looks like petals is 14–32 stamen filaments 7–16 mm long with yellow anthers. Nothing else in this flora does that.
In fruit, July to October: a brown woody capsule 5.5–13 mm with two curved horn-like styles, in a short terminal head. Two seeds each.
Bark on mature stems is smooth and light gray. Branching is zig-zag. Buds are naked and densely stellate-hairy with 2–4 subtending scales, which is a family character, not a genus one.
Witch-hazel is the real problem and it grows in the same habitats. Darke says its vegetative state is easily mistaken for Fothergilla, and Vascular Plants of NC says the same. Use the marginal vein test, then confirm with flower season: yellow strap petals in October to December is witch-hazel; white petal-less spikes in April is Fothergilla.
The hybrid is the other real problem, and it can't be settled morphologically. Ranney's 2007 separation was done by flow cytometry. On the trail, context does the work. Bed edge, mulch ring, uniform habit, near a building or parking area is hybrid. Ragged clump on a north-facing bluff a half mile from a road is a candidate.
Everything else falls out fast. Opposite leaves rules out viburnum, dogwood, hydrangea, buckeye. Compound leaves rules out yellowroot and devil's walking stick. Evergreen rules out loropetalum, distylium and mountain laurel. Fleshy fruit rules out serviceberry, chokeberry, blueberry and huckleberry. Catkins or woody cones rules out alder, hazelnut, birch and hop-hornbeam.
Two worth naming individually. Corylopsis is the closest leaf match anywhere on this list, and it separates on flowers, not leaves: yellow dangling bells against a white erect brush. Lyonia ligustrina is the winter trap, since its persistent dry capsules read like Fothergilla's, and it is a high-severity grayanotoxin plant, so don't handle it casually.
Capsules ripen July to October and dehisce ballistically, so timing is tight. Take them late August into September while still tan-green and firm, and bag or net the fruiting spike rather than planning a return trip. Two seeds per capsule, 5–8 mm, reddish brown, glossy, pointed at the apex.
Fruit set is unreliable. A 2026 Montgomery herbarium sheet reads "very common here; most plants knee-high; no fruits found." Piedmont colonies sucker and are often clonal, and clonal patches set little seed.
Seed is a bad route regardless. Double dormancy: about six months warm-moist, then three months cold. Clemson puts germination at 9 to 15 months.
Cuttings are the route that works. Softwood to semi-hardwood, June through mid-August, June best, with rooting hormone. NCBG's greenhouse manager propagates their own plant this way and sells the result at native plant sales every other year. Layering works on a suckering clump.
Given the conservation status, the sane path to nursery stock is cuttings from a garden plant of known provenance, or a purchase from a regional native nursery, rather than anything taken from the wild.
Person County has no pin to walk to. The one specimen has its locality withheld. What I can do is search the habitat: north-facing rocky bluffs and steep shaded slopes above the Hyco, Mayo and Flat River drainages, and rocky ridgetops in the Person County game lands. Cool aspect, thin acidic soil, resistant rock, ridgeline or outcrop light. April, with the search image being white spikes against a bare understory.
For actually seeing the plant, Cane Creek Mountains Natural Area at Snow Camp is the answer. Alamance County Parks, open to the public, trailheads on Bass Mountain Rd and Mt Hermon Rock Creek Rd, 987 ft high point, about 14.75 mi of trail. Sixteen records from 2021 through 2026 cluster within 40 m of 35.9525 / -79.4197 on the Northern Approach Trail. One observer logged 6–8 plants in full bloom on 27 March 2022. It is 5.5 mi west of Saxapahaw and 39 mi from Roxboro.
Worth noting for the Haw River trips: there are zero Fothergilla records anywhere in the Haw River Trail corridor at Saxapahaw. Cane Creek is close by but it is a different landform.
Second and third choices are White Pines Nature Preserve in Chatham, on the 100-ft bluffs at the Deep and Rocky confluence, where Steve Hall photographed a plant on a rock ledge facing the Rocky River on 17 October 2024; and the Uwharrie NF populations in Montgomery, which have the best abundance in the Piedmont but sit 88 mi away with the Black Ankle Bog site restricted.