The Photoperiodic Response of Chrysanthemums

Anyone who has tried to get all his or her blooms to be at their peak when the annual shows are held knows how difficult it can be to make chrysanthemums bloom on schedule.  Long stretches of hot or cold weather can play havoc with the development of blooms.  Delaying the booming of buds that develop too early is possible by pinching these early budding crowns and letting the plants bloom on later crowns.   However, making the plants bloom earlier is more difficult.  In particular, if your plants are being grown in a microclimate that is on the average a few degrees cooler than that which occurs in the gardens of the growers that set the show dates, you could consistently have plants that bloom many days or even weeks later than the show date.

On a personal note, my wife, who does ikebana (Japanese flower arranging), would be very pleased to have the blooms arrive earlier, and hence be available for a longer period of time. This talk describes a search for a simple technique to do this, based on the use of near infrared light.   I will begin with a description of the way in which chrysanthemums use red and near infrared light to control the blooming time that is called photoperiodic response.

All late blooming chrysanthemums and some of the early blooming cultivars are thought to have their blooming time triggered by the lengthening night, which occurs after June 21st.   Let us begin our discussion of this phenomenon by recalling the natural length of daylight and darkness at our latitude (near the 49th parallel.

Figure 1 shows the sunrise and sunset times and the difference, which gives the hours of light as a function of date here.   Also shown are the hours of darkness, which, is thought to be the important variable for mums.

Figure 1 (note that the jogs in April and October are changes between daylight saving and standard times)

First observe that the hours of sunlight peak on about the first day of summer (June 21) when the hours of sunlight are about 16, and that this drops to about 8 hours in mid winter.  The hours of darkness become a minimum of 8 hr in June and a maximum of 16 hr in midwinter.  The trigger for bud formation is the drop in daylight hours after this peak, or rather the rise in the hours of darkness that occurs in late June.  As we shall see there is a real experimental distinction between shortening-days and lengthening-nights and it is the latter that is the key.

The other important factor is the wavelength of the incident light.  Light from the sun has the spectral distribution of a hot body at about 5800 degrees C, however, by the time it comes through the atmosphere some of the radiation has been taken out of it by atmospheric gases so that the spectrum is as shown by the solid line shown in Figure 2.

Figure 2

That line shows the spectrum of the direct overhead sun.  You can see that it is principally the visible part of the spectrum from 400 to 700 nm that gets to the plants almost unscathed.  The UV is mostly removed by the ozone layer, and much of the infrared is taken out by water and carbon dioxide in the atmosphere.  

The dashed curve is interesting.  It is the spectrum of the light seen under a canopy of leaves.   The little bit of green radiation left is why plants look green.  If our eyes could see the far red (FR) the leaves would look a bright rich dark red because that is the major part of the transmitted radiation bouncing off and passing through the leaves.  Some insects appear to have the ability to see such FR radiation.

Notice the two bumps in the red region of the solar spectrum.  These lie at 660 nm and 730 nm, and are labeled R (for Red) and FR (for Far Red).  Plants have chosen to make use of these two wavelengths by making a molecule called Phytochrome that in one of its forms absorbs light in the red and in another form absorbs light in the near infrared.   Its molecular structure is shown in the next diagram (it is a linear tetrapyrrole attached to a protein labeled Cys in Fig. 3).

Figure 3
Phytochrome (PR)

PR absorbs red light (at 660nm) and is thus converted to the species PFR that simply has the D ring rotated as shown by the arrow in the diagram.   PFR absorbs in the near infrared (at 730nm) and re-converts it to PR.   Because of the mixture of red and near infrared light in sunlight (see Figure 2), plants end their day with about 60 percent PFR and 40 percent PR.  PR is considered to promote flower formation, while PFR may be an inhibitor of flower formation.  Such a 60/40 ratio of PFR/PR provides more than enough PFR to stop the plant from blooming.   However, in the dark the PFR slowly reverts to PR.  These processes are summarized in the following table:

PR  is converted very rapidly by red light to  PFR
PFR  reverts very slowly in the dark to  PR
PFR  can be converted very rapidly by Far Red (FR) light to  PR
Table 1

In mid summer, with its long days, there is not sufficient time dring the short nights to removed enough of the bloom-inhibiting PFR.

The phytochrome-mediated response to the photoperiod occurs in the leaves, but the phytochrome cannot move and flowering occurs in the apical meristems.  It has therefore been suggested that a protein, called florigen, is mobile and can transfer this information to the apical meristems.  This necessity for florigen is illustrated in experiments described in Figure 4.

Figure 4

The experiments illustrate the fact that imposing long nights on a single leaf, or on one part of a grafted pair of plants can cause the entire plant (or both plants in the second case) to bloom.  Florigen has recently been identified as a protein produced by a gene called Flowering locus T, or FT.

The consequences of these processes are well illustrated by the following additional experiments:

  1. A 16 hr Day + an 8hr Night yields only vegetative growth.
  2. An 8 hr Day + a 16 hr Night results in flowering.
  3. An 8 hr Day + an 8hr Night + 10 seconds of Light + 8hr Night yields only vegetative growth  (i.e. interrupting the long night stops flowering).
  4. An 8 hr Day + 10s Dark + 8hr Day + 8hr Night still yields only vegetative growth.  (i.e. interrupting the long day does nothing).
  5. An 8 hr Day + 8hr Night + 10s Red + 8hr Night yields only vegetative growth  (i.e. interrupting the long night with red light stops flowering as well.)
  6. An 8 hr Day + 8hr Night + 10s FR + 8hr Night produces flowers  (i.e. interrupting the night with FR still leads to flowering).
Table 2

From these experiments it is clear that the longer night-period that initiates flowering must not be interrupted by visible light.  That is, a very brief period of visible light inserted during a long night inhibits flowering.   This effect of the inserted light is called night-break.  As many of you probably know, we growers need to avoid this possibility that can take the form of lights from a passing car or a streetlight.   The reason for this effect is clear.  Any brief exposure to white or red light in the middle of the night undoes the previous slow conversion of PFR to PR that has been occurring in the first half of the night.

This means that the trigger for blooming must be sensitive to the relative amounts of the two forms of phytochrome.

The above description is much oversimplified.

  1. There appear to be at least 5 separable phytochromes labeled A to E.  Not all are present in a given plant.  P(A) is unstable and decomposes constantly, but P(B) does not.  The others have not been well characterized for mums.
  2. The effect of a night-break depends on what part of the night the shot of visible light is applied.  This and other more complex experiments have led scientist to propose that circadian rhythms are involved..

OK, where does this leave us?  Commercial growers have used opaque sheets to cut out the light in the early evening and provide a well-controlled onset of long nights.  For controlling the blooming time of late-blooming mums, the instructions given by W. D. MacConnachie for covering with blackout cloth reads as follows: the plants have to be in darkness for fourteen hours daily for at least fourteen days at the end of August or the beginning of September to initiate flower buds.  However, it is not a variable that most amateur growers feel they can use easily since the cover must not only be very complete, but heating of the greenhouse by the sun can be a big problem.  Some amateur growers have been known to bring their potted plants into a light-tight shed in the early evening to simulate a long night.  Not only is this impossible with early blooming cultivars that are usually grown in the ground, but the process can be back breaking with 10 or 12 inch pots.

As I stared at the photochemical reactions that I described above, another possibility occurred to me:

Why not just give the plants a shot of FR light after sunset?  If you look at the inter-conversion of PR and PFR in Table1 above, you will see that, in view of the experimental evidence in Table 2, with as little as a one minute exposure to FR radiation one can very quickly convert a significant amount of PFR that has formed during the day back into PR.  The irradiated plants should be thus fooled into thinking that they have already had a long night.

Unfortunately, there appears to be no inexpensive commercially available light bulb to do this with, and filters that cut out the red and transmit the FR are not readily available.   However, if you look at Figure 2, you will note the way that the light becomes rich in FR radiation (relative to RED) when it passes through a canopy of leaves.  The ratio of FR radiation at 730 nm to radiation at 660 nm changes from about 1:1 to about 40:1.  A layer of leaves provides an excellent filter that removes red light at 660nm while transmitting FR light at 730 nm.  Noting that the conversion between PR and PFR is extremely rapid, removing one molecule of PFR with each absorbed photon.  I decided to cover the front of a spotlight with leaves, and simply walk it around the plants spending about 1 second over a given set of plants.  I have tested the system with a spectrometer, and found that with a layer of 3 leaves, almost no red light comes through, and the FR is only slightly diminished.  The physical arrangement has a 75 watt flood light placed in a glass container (a pyrex coffee pot in my case) with darkened (spray painted) sides.  Plastic containers are useless for this purpose since they all soften and collapse in a few minutes).   The leaves were held between two layers of chicken wire and pressed against the bottom of the glass vessel containing the floodlight.  Rhubarb leaves were ideal for this purpose because of their size.

In 2006, I applied 20 seconds of FR to 70 early and late blooming cultivars for 2 weeks starting the FR radiation on about June 26.   When compared with plants that received no extra FR radiation, the irradiated plants showed no observable effect.  I thought that perhaps the natural trigger of shortening days that occurred between June 21 and 26 was sufficient and hence my treatment had little effect.

In 2007,I began irradiating half of the earlies and half of the lates on the 14th of June, stopping on July 26.   As in the previous year I could see no difference between the controls and the irradiated plants.   It was at this point that I discovered a reference to earlier attempts to observe an effect of FR radiation on the blooming of chrysanthemums.  The conclusion of those studies were the same as mine, i.e. no effect.   H. A. Borthwick; H. M. Cathey (1962).

However, I was surprised by the very early appearance of buds on many of the earlies and lates even before I began irradiating them.  Buds formed on such late blooming varieties as 12 Rayonante, 12 Susan Etheridge plants and even some Jimmy Mortram and Jennie Hapgood cultivars.   This puzzled me for a few weeks.  Finally a possible explanation occurred to me.   Several rows of mums in my garden are shielded from a street light by a laurel hedge and a cedar hedge both of which I have just trimmed back (the other rows are shielded from the light by a wooden shed).  The hedge leaves probably provide a filter that removes most of the visible component from the street light, but not the FR band at 730nm, because the effective thickness of the hedge is about 4 or 5 leaves.  Since this FR radiation continues all night, it must be effective at converting PFR during the entire circadian cycle and hence makes the plants feel that they are experiencing long nights.  It would be interesting if any other growers have observed such an effect.  Next year, I will place a light bulb on the other side of my hedge in a region that is currently not exposed to the light from the street lamp.

Three important experimental observations have come to my attention more recently.

  1. The light response of plants occurs only on the lower leaves of the plants and not on the top leaves.  This may have been a problem in early experiments.
  2. For Late blooming mums the photoresponse occurs only in late August when the nights have become sufficiently long.  Any irradiation will therefore have to occur close to that time to be effective.
  3. Although Early blooming mums show some photoperiodic response, they are very much affected by temperature, and for example can be made to bloom by having their nighttime temperature rise above 15 degrees for a few nights in a row.