Short answer. A "hermie" is a female plant that also makes male flower parts: small yellow-green, banana-shaped anthers poking out among the pistils, or full pollen-sac clusters. Cannabis is a dioecious species with a long-established XY system [2][3]; a well-bred female stays female. Herm-prone plants are mainly breeder error: lines built on repeated selfing, careless feminising, or untested seed from "pollen chuckers". Stress (light leaks, heat, a missed dark period) reveals that weakness; it doesn't create it in sound genetics. Fix the trigger, pluck a few closed anthers in late flower or remove the plant if it's shedding pollen, and never keep it as a mother or breed from it.
Breeder's standard. A plant that throws male flowers under normal room conditions is culled from the breeding programme, along with any clone of it. It is never used as a pollen donor, a seed mother or a reversal donor. Its herm rate is recorded and disclosed with any seed from the same line. GenetixBay treats this as an ethics standard, not a preference: see record-keeping and naming ethics.
What's happening
Cannabis has been a dioecious plant, with separate males and females, for as long as people have grown it. Its sex chromosomes are old and the Y is strongly degenerated [2][3]. Separate sexes are the species' stable, normal state. An intersex female is an unwanted outlier, not the plant "doing what plants do".
Drug-type cannabis has been recorded as dioecious throughout its history: growers and writers have always distinguished male from female plants, and ancient Chinese texts already used separate names for male and female hemp [5]being reviewed. Monoecious hemp, with both sexes on one plant, exists only because 20th-century fibre breeders selected for it on purposebeing reviewed; it is not the norm and says nothing about how drug-type cannabis behaves. Selfing, too, was historically an emergency measure, used to save a line when no male was left, not a trait anyone wanted.
Why do some modern plants herm so readily, then? The GenetixBay breeder position, drawn from decades of growing and breeding:
- Repeated selfing (S1 on S1). Selfing a plant that has already shown intersex flowers concentrates whatever makes it do so. See S1s and selfing.
- Careless feminising. Reversing a female that hasn't been proven stable across several runs, or collecting pollen from a stress herm instead of a controlled reversal, puts that tendency into every seed. See feminised seed and sex reversal.
- "Pollen chuckers". People who cross whatever is in the room and release the seed without growing out and culling. Unstable genetics then spread widely under popular names.
What the research shows fits that picture. Punja and Holmes characterised hermaphroditism in commercial indoor cannabis: anthers formed in clusters in bract tissue next to the stigmas, from about week 4 of flower through harvest; anther and pollen looked like those of normal males; pollen germinated 10–30% in vitro; seeded inflorescences held 22–48 seeds; whole inflorescences converting to anthers was rare (~2 in 1000 plants) [1]. They saw this on roughly 5–10% of plants across three cultivars in one commercial facility [1]. We read that rate as evidence of poor genetics in that operation's cultivars, not as normal cannabis behaviour. The founder has grown thousands of cultivars and doesn't see it in older lineages, only in the recent wave of low-quality releases. They cite older work linking male flowers to applied chemicals (gibberellins, silver compounds, cobalt chloride) and to stresses such as altered photoperiods, interrupted darkness and low temperatures [1]. We read that as stress revealing susceptible genetics. The same light leak that makes one clone herm leaves a stable clone beside it untouched. Seed from hermaphrodite pollen gave only female progeny in the strains tested [1], which is why a herm in the room passes its tendency straight into the next generation of "feminised" seed.
Late-flower "bananas" are not normal
House position (firm): late nanners are a herm, and they're unwanted. You'll hear it in every grow forum: "a few bananas at the end are normal, the plant is just trying to seed itself." It isn't normal. A female that grows anthers in its buds late in flower is expressing male flowers: it's intersex, a herm. That holds even if it's the best plant you've ever grown, the loudest gas in the room, the keeper you've hunted for years. It's a sign of poor genetics, from bad breeding or pollen-chucking behaviour, and base lines that are prone to it got that way through bad practice early on. You can sometimes breed it out of a line with patient selection, but usually that isn't the approach: cull it, don't keep it as a mother, and don't breed from it. The founder will die on this hill.
What a late banana actually is. The yellow "banana" is a staminate flower reduced to its anthers, forming inside a pistillate inflorescence. Punja and Holmes found these anthers in clusters in bract tissue next to the stigmas, from about week 4 of flower right through to harvest, with pollen that looked like a normal male's and germinated 10–30 % in vitro [1]. Late timing doesn't make it a different thing. It's the same intersex expression, showing up when the plant is under the most pressure.
Why we call it genetics, with the evidence:
- Sex expression is heritable. In monoecious hemp, how much male and female flowering a plant shows varies between cultivars and has a genetic basis [7]; genome-wide work has mapped loci associated with sex expression [8]. That's hemp, not drug-type cannabis, but it's the same species and the same flowering machinery.
- It responds to selection. Monoecious hemp exists because 20th-century fibre breeders selected for plants that make both sexes [10]. If selection can push the species towards both sexes on one plant, careless breeding can drift a line that way, and careful breeding can keep it female.
- Herm pollen makes herm-prone "feminised" seed. Seed from hermaphrodite pollen gave only female progeny in the strains tested [1], and controlled reversal with silver does the same thing on purpose [9]. A late banana that drops pollen in your room has just made feminised seed carrying its own tendency. Selfing such plants, or reversing them for "fem" seed, is how the trait spreads through the market. See S1s and selfing and feminised seed and sex reversal.
- Stable plants don't do it under the same conditions. In one facility, intersex flowers turned up on roughly 5–10 % of plants [1]; we read that as a sign of poor genetics in those cultivars, not as a normal rate. Under the same lights, the same feed and the same long finish, a well-bred female stays female.
The honest part. Research does show that stress can trigger intersex flowers: interrupted darkness, altered photoperiods, low temperatures, and chemicals such as silver compounds or gibberellins have all been linked to male flowers on females [1][9]. Late flower is a stressful time, especially if a plant is run past its window. We don't deny that. We read it the way a breeder does: stress reveals susceptible genetics; it doesn't make a stable plant herm. The light leak or the extra two weeks was the test, and the plant failed it. What hasn't been published is a heritability estimate for intersex tendency in drug-type cannabis specificallybeing reviewed, so the genetic case rests on the hemp work, the breeding history and long breeder experience.
What to do. If it's a handful of closed anthers in the last days and you're finishing the run, pluck and monitor. But record it, don't take cuts from it, don't make seed with it, and tell the breeder. On GenetixBay, herm testing has two tiers: a 10-plant screening grow where any herm fails the line for Bay Lines, the Seed Club and featuring; then pooled community grow reports, and only at about 150 plants or more can a line earn "Verified under 2% herm rate" (ten plants can't tell 2 % from 10 %; see herm threshold). See breeders vs pollen chuckers.
How to recognise it
| What you see | What it is | Urgency |
|---|---|---|
| A few yellow "bananas" (bare anthers) poking from buds, late flower | Hermaphroditic anthers | Medium: pluck, monitor, flag the genetics |
| Clusters of round, pale green balls on stalks at nodes (pollen sacs) | Male flowers on a female | High: remove plant or branch |
| Pollen dust on leaves below | Pollen release | High: isolate immediately |
| Fresh calyx spires, no anthers | Foxtailing | See foxtails |
| Seeds in buds at harvest | Pollination happened weeks ago | Find the source |
Numbers that matter
| Parameter | Target | Notes |
|---|---|---|
| Dark period in flower | 12 h, uninterrupted, light-tight | Grower practice; light interruptions linked to sex reversal [1] |
| Light leak test | no visible light after ~10 min eyes adjusted in the dark | Grower practice |
| Canopy temperature | < ~30 °C (86 °F) | Heat stress [6]; grower practice links heat to hermies in susceptible plants |
| Typical onset | ~week 4 of flower onwards | [1] |
| Observed incidence | ~5–10% of plants, 3 commercial cultivars, one facility | [1]; not a "normal" rate for well-bred lines |
| Pollen germination | 10–30% in vitro | [1] |
| Runs a mother must stay clean before breeding | at least 2 | GenetixBay breeder's standard |
What to check first
- Light leaks. Sit in the room or tent in the dark for 10 minutes. Check door seals, LEDs on equipment, vents.
- Timer and controller logs. A night with the light on, or a skipped dark period, is the classic trigger for a susceptible plant.
- Heat and light stress at the tops.
- Other stresses: severe root problems, pests, drought, extreme pruning in flower.
- Plant history. Where did the seed come from? Is it an S1 of an S1, a feminised line from an unproven donor, or a "freebie" from someone who doesn't cull? Has this clone done it before?
- The neighbours. If one plant hermed and the rest of the room under identical conditions did not, the environment exposed it; the genetics are the problem.
How to fix it
- Isolated bananas, late flower, not open: pluck them carefully with tweezers, mist the area lightly with water to deactivate any loose pollen (grower practice), and inspect every 2–3 days.
- Many anthers, male flower clusters, or pollen shedding: remove the plant from the room (bag it first to contain pollen). Harvest early if it is near maturity.
- Fix the trigger: seal light leaks, replace faulty timers, correct heat stress. That protects the rest of the room.
- Fix the genetics: cull the clone; don't keep it as a mother. Record the event (week, trigger if any, photos) in your journal. If it came from a seed line, tell the breeder and note it publicly on the line record.
- Seeded buds: quality is reduced; seed formation diverts resources [1]. Don't save or sell the seed.
How to confirm it worked
- No new anthers on any plant over the next 7–14 days of inspection (hand lens, focus on bud interiors and node sites).
- At harvest, few or no seeds.
- Next run, with the trigger fixed and the culprit gone, a clean room.
Common mistakes
- Treating hermies as normal cannabis behaviour. They are not; they are a defect to breed out.
- Calling late-flower bananas "normal" and keeping the plant because it smokes well.
- Blaming only the light leak, and keeping the clone. The leak was the test; the clone failed it.
- Saving "free feminised seed" from a hermed bud. Every seed carries the tendency.
- Missing light leaks from equipment indicator LEDs.
- Confusing foxtails with anthers.
Evidence
- Cannabis is dioecious with an XY system: Established [2][3].
- Morphology, timing, pollen viability and progeny sex of intersex flowers: Established by one controlled study [1]. Its 5–10% incidence describes one facility's cultivars and is read here as a sign of poor genetics, not a normal rate.
- Drug-type cannabis recorded as dioecious through history; monoecious hemp a deliberate modern breeding product: Established [5].
- Stress and chemical triggers: Supported by older literature cited in [1]; specific thresholds are Grower practice.
- Herm-prone lines arise mainly from breeding practice (repeated selfing, careless feminising, no culling): GenetixBay house position, grounded in long breeder experience. Selfing was historically an emergency measure to save a line, not a desirable trait. Rule: cull herm-prone plants and don't breed from them.being reviewed
- Sex expression having a genetic basis and responding to selection: Supported in hemp [7][8], and by the deliberate breeding of monoecious hemp [10].
- "Late-flower bananas are normal": Myth (house position). They are intersex flowers [1]; stress can trigger them, which we read as stress revealing susceptible genetics.
