Flowering Induction in Outdoor Cannabis: What Starts It and When It Arrives

At the latitude of Stuttgart or Vienna the day runs longer than 16 h only between 8 June and 4 July. A variety whose limit sits there switches from growing to flowering shortly after midsummer, stands well into flower by August and is ready to harvest from the second week of September. Another variety in the same bed, its limit a quarter of an hour lower, starts a good two weeks later. What separates the two is a property fixed in the seed: the day length at which the plant switches over. It cannot be read off a plant, and it decides whether anything ripe comes together in autumn.

How the plant knows the time has come

The plant counts darkness. It measures how long the dark runs unbroken and switches over once that stretch grows long enough. Because the nights lengthen daily after midsummer, that point arrives outdoors by itself. The trigger is not the date on the calendar.

The process is called flowering induction, and cannabis counts as a quantitative short-day plant. Short-day plant means the flower comes once the nights pass a certain length. Quantitative means the response is graded. The plant reacts more strongly the longer the night runs; there is no switch that flips at one fixed value.

Day length and night length: two measures for the same limit

Converting one into the other throws the estimate off by weeks. One figure is the critical day length: how long the sun may stand above the horizon before the plant switches over. The other is the critical night length, measured as unbroken darkness of the kind a timer produces indoors; for light-sensitive varieties it runs between 8 h and 12 h.

The two do not add up to 24 h, because twilight sits between day and darkness, and the plant still counts twilight as day. Evening twilight in the field was measured at 2.4 µmol·m⁻²·s⁻¹, a measure of how much light falls on a surface each second, and that level still acted on the plants. One variety responded to as little as 0.1 µmol·m⁻²·s⁻¹ at leaf level with delayed flowering, less than a twentieth of that.

How large the difference gets shows on the longest day of the year at the latitude of Stuttgart: 6 h 28 min pass with the sun more than 6° below the horizon, and only 4 h 30 min with it more than 12° below. The value the plant really counts as night lies between the two and closer to the smaller one. It cannot be given more precisely here, because the point at which brightness after sunset drops to 0.1 µmol has not been calculated. For planning a garden the day length is therefore the more usable figure: it stands in a solar table for any place and any date.

Where the limit of a variety sits

A trial across 27 hemp varieties found critical day lengths between 13 h 45 min and more than 15 h 30 min; of fifteen varieties grown for flower, fourteen sat at 14 h or above. Outdoor varieties with a limit around 16 h therefore sit at the upper edge of the documented range.

How finely varieties differ shows in a single number: a difference of 15 min in day length shifted the flowering start of several varieties by five to thirteen days. In the calendar that grows larger at midsummer, because the days barely shorten then. At the latitude of Stuttgart the day loses 0.87 min in early July, 2.76 min in early August and 3.50 min in mid-September. The same 15 min between varieties therefore amount to roughly 17 days in July and only four in September. Precisely at the top, where the limits of early outdoor varieties lie, small differences hit the hardest.

Two properties get mixed up here. A day-neutral variety flowers regardless of day length, so even while nights are short; four fibre-bred lines from the north did so even under round-the-clock light. An early variety stays light-sensitive and merely carries its limit high up. A grower at the latitude of Stuttgart reports that varieties counted as sativa-leaning, which take a long time to ripen, still finish on time when they switch over early enough; on that experience the sativa share works neither as a criterion for choosing nor for ruling out. What matters is the day length at which the variety switches over.

From switching over to visible flower

Anyone quoting a flowering time has to say which starting point they count, because three different events occur and only two of them hang on light. Single flowers in the angle between leaf and stem form as the plant ages, even while nights are short. They are no sign that it has switched over.

The real switch starts at the shoot tip and runs in two steps that depend on light to different degrees. The first step is the laying down of the first flowers, technically flower initiation. It still occurred in ten varieties with at least 19 % THC, the main intoxicating compound, when the day ran 14 h long, delayed by nothing to four days against a 12 h day; four of those varieties showed no delay at all. Three varieties laid down flowers even at 15 h of daylight, three hours above the usual 12 h day. Nothing came of any of those initials: the second step, the growing out into the dense flower clusters that are eventually harvested, failed to happen. That second step is called inflorescence development, and it needs the long night without exception, while the first merely prefers it.

Two things follow for planning. Above 13 h of daylight, an hour beyond the 12 h day usual indoors, the clusters came out smaller and yielded less early, even where the initials appeared on time. And time figures from indoor growing count from the day someone sets the timer to 12 h. That day does not exist outdoors: out there the nights lengthen a little each day, and the switch stretches over several days. From the initials to visible flowers it takes four to eleven days in Canadian and northern European lines, twelve to 16 days in northern Italian and northern Chinese ones, and 21 to 25 days in southern Chinese ones.

What the location contributes

How much a more southerly location buys depends on where the limit of the variety sits. For a variety at 16 h there is no window left at 44° N in which the day would be too long: the light there favours flowering all year. At the latitude of Stuttgart it is 27 days in early summer, at 52° N it is 61, at 56° N 84 and at 60° N 101. The further north, the later the day drops below the limit, and the tighter the remaining season gets.

For a variety with a lower limit the picture reverses in autumn. The closer the limit sits to the equinox, the smaller the gap between locations: on 22 September only 50 min of darkness separate 40° N and 64° N. After that the north holds the longer nights. Outdoors the same shows up as locations shifting varieties unequally: in a European fibre hemp trial the variety Markant needed 26 days less to flower in northern Italy than in Latvia, the variety Monoica 52 days less. Such trials harvest at full male flower, so what transfers from them is the timing and not the duration of flowering. A rule of thumb by latitude that leaves out the limit of the variety therefore does not hold.

Working the date backwards

Anyone wanting to know whether a variety finishes at a given location calculates from the end. Three figures are needed: the latitude, the latest harvest date still defensible there, and the flowering duration of the variety counted from one of the two events above. According to the grower quoted earlier, the usable harvest window at the latitude of Stuttgart falls in the second to fourth week of September, at the outside in the second week of October; anything later turns uncertain through frost and persistent damp.

Subtracting the flowering duration from the harvest date gives the day on which the clusters have to start growing. Subtracting from that the four to 25 days from initials to visible flowers gives the day on which the day has to be short enough. How long the day runs on that date comes from a solar table for the relevant latitude, and that figure is the limit the variety has to have at minimum. For the variety in the example above the calculation works out: switch-over from late June to early July, harvest from 8 September, which makes 65 to 81 days of flowering for a variety rated at 60 days and more.

The sowing date follows backwards from the same chain. Before any response to light comes the juvenile phase, the time after the seedling emerges during which the plant does not respond to day length at all; for fourteen European varieties it was determined at 25 to 62 days, a range of more than a month. Subtracting it from the required flowering start gives the latest date by which sowing has to be done. For a variety whose limit exceeds the longest day of the location, the juvenile phase is in fact the only pacemaker left: light never holds it back there.

A second route to choosing a variety is described and does not survive testing outdoors: selection through a genetic test that detects a known flowering trait. In plants under continuous light such a test predicted the flowering behaviour of the offspring completely. The same offspring outdoors at 42.9° N showed no relation between test result and flowering date, and a further generation at 53.3° N showed none either. Outdoors, what the plant actually does at the location is what counts. One building block of this calculation also stays open: a frost date by latitude from a weather series has not been gathered here, so the latest defensible harvest date remains a judgement made by the grower.

Calculator: dates for a specific location

Outdoors, flowering starts once the day gets short enough. The exact length differs from variety to variety. This calculator sets that value against the location and works out when flowering begins and whether the harvest will be ready in time.

Location

Location

Latitude determines day length and nothing else. It says nothing about weather, frost or fog at the site; that comes in the next step.

Harvest window

Only the grower can supply this. Three things decide when it is over, and they differ within a few kilometres:

First frost. A hollow where cold air pools overnight is hit earlier than a slope above it.

Fog and damp. Dense flowers dry badly after a wet night. Beside a river or in a foggy spot the harvest has to be in earlier than at a dry position in the same village.

Distance from the coast. By the sea autumn stays mild longer; inland it turns cold earlier. Two gardens on the same latitude end up weeks apart.

Harvest window

Variety

Variety

The flowering time on the packet applies to indoor growing and counts from the moment the timer switches. It still works as a hint whether a variety runs short or long. The switch-over day length is the figure that matters outdoors: across 27 hemp varieties tested it ran between 13 h 45 min and over 15 h 30 min, and early outdoor varieties sit around 16 h.

What the variety has to manage

Switches over at a day length of

The slider sits on the value the calculation above produces. This figure is rarely on the packet, and usually nobody knows it. That is exactly why the calculator works it out: it says what a variety has to manage so it finishes in time at the location set above. Where the value of a particular variety is known, it can be slid in here. For scale: across 27 hemp varieties tested it ran between 13 h 45 min and over 15 h 30 min, and early outdoor varieties sit around 16 h. Any change above resets the slider to the required value.

Result

What follows from the value on the slider. Set above the requirement, the variety switches over earlier than needed, and everything below gets more comfortable.

The year at this location

Sun above the horizon still light last light dark

The same variety elsewhere

Only the latitude moves, everything else stays put. The matching row is highlighted. The weather does not move with it: on the coast a late date is workable, inland at the same latitude often not.

LatitudeFlowering startsReady to harvest at the earliestHarvest window

Day lengths count from sunrise to sunset. The twilight after that appears alongside in the chart, because the plant still counts it as day; how much of it the plant registers has not been measured.

Sources