Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 6, Issue 1, January 1956

The Causes of Flowering

page break

The Causes of Flowering


The return of spring is each year heralded by the appearance in our gardens of snow-drops, daffodils and crocuses. These flowers last for only a few weeks, and as summer approaches other kinds of flowers take their place, only to be replaced themselves by the autumn-flowering species. This unfailing sequence of spring, summer and autumn flowers is too familiar to cause much surprise, but it is a very extraordinary phenomenon, and only in the past 30 years have botanists obtained an understanding of the factors which control flowering.

It has of course been known for a very long time that most plants will flower only at certain times of the year, and that in many species, and especially in cultivated plants, ‘early-’ and ‘late-flowering’ varieties occur. Nevertheless all attempts to relate the different times of flowering to seasonal differences of temperature or light intensity had met with failure. In 1920, however, two Americans, Garner and Allard, suggested a much more effective way to control flowering. They found that a variety of tobacco called Maryland Mammoth would not flower during the summer when grown in Washington, D.C., although other varities of tobacco flowered freely under the same conditions. However, when Maryland Mammoth plants were grown in a warm glasshouse during the winter they flowered profusely, suggesting that some seasonal factor was effective in controlling the formation of flowers. Garner and Allard suspected that this factor was the length of day.

By the simple experiment of artificially altering the day-length they were able to cause many species of plants to flower at unusual times of the year. Thus Maryland Mammoth tobacco which would normally not flower during the long days of summer could be made to flower at this time by covering the plants with light-tight boxes at 4 p.m. each day and removing the boxes at 9 a.m. each morning, thus subjecting the plants to days of only seven hours. In the same way a variety of soybean called Biloxi which normally flowers very late in the summer could be made to flower in midsummer if it received only seven hours light each day. If Biloxi soybeans are grown in the winter in a warm glasshouse they will flower profusely, but if the length of winter days is extended by leaving an electric light on near the plants from about 5 p.m. to midnight they will remain vegetative. Conversely, species which normally produce their flowers in mid-summer can be made to flower even in mid-winter by extending the hours of daylight with artificial lighting.

page 2

Garner and Allard were able to classify a large number of species as either short-day plants which would flower only if the length of day was less than about 12 hours, or long-day plants which flowered only when the day-length exceeded about 12 hours. The response of plants to the relative length of day and night they called photo-periodism. While many plants remain vegetative indefinitely if grown in days of the wrong length, some others will flower irrespective of the length of day. These are called day-neutral plants.

Short-Day Species Long-Day Species Day-Neutral Species
Chrysanthemum morifolium (Chrysanthemum) Lactuca sativa (lettuce) Antirrhinum majus (snapdragon)
Perilla ocymoides Plantago lanceolata (plantain) Lycopersicon esculentum (tomato)
Senecio cruentus (cineraria) Spinacea oleracea (spinach) Taraxacum officinale (dandelion)
Soja max var. Biloxi (Biloxi soybean) Zea mays (maize)
Xanthium pennsylvanicum (cocklebur)

Although it was possible to classify plants as long-day or short-day species it soon became evident that no particular length of day could be chosen to separate the two groups. Some short-day species will flower only when the length of day is less than 12 hours, others will flower so long as the days are not longer than 16 hours. Thus for every species there is a critical day-length. Short-day plants will flower only in days shorter than the critical day-length for the species, long-day plants will flower only when the length of day exceeds this value. The important distinction seems to be that, whereas short-day plants require a certain minimum length of darkness after each light period to flower, long-day plants will flower even when growing under continuous illumination. In fact, darkness seems to inhibit flowering in long-day plants.

In their native habitats plants seem to have adapted their flowering response to the day-lengths which they normally encounter. Thus species which are native to the tropics, where day-lengths tend to be constant at about 12 hours throughout the year, are generally short-day plants, those from more temperate regions, which have longer summer days, are often long-day plants. This explains the frequent failure of plants moved to new geographical regions to flower despite the favourable temperatures in which they are grown.

Differentiation Of Flower Buds

Before discussing the research which followed the first demonstration of photoperiodism, we can briefly indicate the changes which occur as a plant page 3 passes from the vegetative to the reproductive condition. Seedlings will not flower even when they are placed in favourable day-lengths. Before a plant can respond to a photoperiodic stimulus it must have attained a certain stage of development, and this stage has been termed ‘ripeness-to-flower’. Just what is involved in the change from seedling to ‘ripeness-to-flower’ is not clearly understood, but it is in some ways similar to the change from the sexually-immature juvenile to the sexually-mature adult of animal species.

The first visible indication of flowering is a change in the shape of the apical meristem of the shoot, but this can be seen only by careful examination under a microscope. During vegetative growth the meristem, from which new leaves and stem tissues arise, is usually conical in shape and quite small, often only 0.1 mm. in diameter, but with the onset of flowering each meristem becomes flattened into a broad disc which may give rise to a single flower, e.g. poppy, or to many flowers, e.g. capitulum of daisy.

During the later stages of flower development some of the younger internodes of the stem elongate, raising the flowers above the vegetative body of the plant, and the leaves in this part of the stem are usually smaller and of simpler shape than those farther down the shoot.

Photoperiodic Induction

The importance of Garner and Allard's discovery was soon recognised and a large number of botanists, especially in America and Russia, turned their attention to elucidating the various steps involved in the photoperiodic response.

It was soon found that it was not necessary to maintain a plant in a favourable length of day for its whole life in order to cause flowering, but that a relatively brief exposure to the correct day-length was sufficient, after which the plant would flower even if it were subsequently placed under unfavourable day-length conditions. The stimulation of flowering caused by days of the correct length is called photoperiodic induction. Different species were found to differ considerably in the number of photoperiods needed to induce flowering. Chrysanthemum requires up to 30 successive short days while Biloxi soybean will flower after being exposed to only two short days. The American cocklebur, Xanthium pennsylvanicum, is remarkably sensitive to photoperiods of the correct length. It is a short-day plant and remains vegetative indefinitely in days of more than 16.5 hours illumination. However, a single short day is sufficient to induce flowering even if the plant is subsequently placed in long days. Once flowering is induced a plant may continue to flower for a very long time; a Xanthium plant after being induced by only seven short days, flowered continuously for the following 18 months even though it was kept in long days for all of this time.

Within certain limits photoperiodic induction produces a quantitative flowering response. For although two short days are sufficient to initiate flowering in Biloxi soybeans, heavier flowering follows further short-day page 4 treatment up to the point where all the vegetative apices have been converted to flowering apices.

In long-day plants that require several photoperiodic cycles for floral induction the treatment may be given on consecutive days or may be broken into parts separated by several short days. Plantago lanceolata requires 25 long days for maximum flowering response. But by breaking the long-day treatment into two periods of 10 and 15 days separated by 20 short days, maximum response is still obtained. Evidently under the correct length of day a permanent physiological change occurs in the plant. This change persists during unfavourable day-length conditions, and is augmented when the plant is again subjected to the correct photoperiod. Since the initiation of flowers at the shoot apex is an ‘all-or-none’ process, the changes caused by the inductive day-length must first build up to a threshold level before any flowers are initiated.

However, some varieties of wheat, some oats and rye, carrots, cabbages and some other plants do not flower in their first season despite the favourable day-lengths which they encounter at this time. These plants, which normally flower in their second season and are therefore biennials, require a period of several weeks of cold weather before they will flower. This is usually received during the winter, but by subjecting these plants to a period of cold in the seedling stages, or even during germination, they can be made to flower in their first season provided they encounter the correct length of day. This explains the ‘bolting’ or precocious flowering of some biennial crops when they are planted too early in the spring. The cold treatment of germinating cereal seeds, called vernalisation, so that they will flower and fruit in their first season, forms much of the basis of recent agricultural practice in Russia.

The time at which ornamental plants flower is of considerable importance to florists as out-of-season flowers may fetch high prices, and the practice of regulating the daily duration of illumination so as to control the time of flowering of their crop is now being quite extensively employed by commercial flower-growers in America and Europe. By covering such short-day plants as Chrysanthemum with heavy black cloth early each afternoon in summer they can be induced to flower several weeks before the normal time. By leaving lights switched on in the glasshouses until midnight each night during autumn, flowering of Chrysanthemum and orchids can be prevented. Plants can then be made to flower in winter when flowers are in short supply. Sugar cane, another autumn-flowering species, is in some regions prevented from flowering by artificially extending the length of autumn days by floodlighting the fields, thus lengthening the period of vegetative growth and increasing the yield of sugar.

Perception Of The Day-Length Stimulus

In early experiments the whole plant had been exposed to the favourable photoperiod, but by a series of ingenious experiments in which different page 5 parts of a plant were shielded in such a way that one part could be exposed to long days while another part of the same plant received short days, perception of the day-length stimulus was shown to be localised in newly-matured leaves, and a flowering stimulus transported from them to the apices where flowers were formed.

Fig. 1 a-f illustrate experiments which demonstrated the localisation of photoperiodic perception in Xanthium pennsylvanicum. In long days Xanthium remains vegetative (a), but flowers in short days (b). If the immature leaves and apical bud are placed in a box which can be darkened early each afternoon so that this part of the plant receives short days while the mature leaves are exposed to long days the plant does not flower (c), but if the treatment is reversed so that only the mature parts of the shoot receive short days, flowering is induced (d). Subjecting a single mature leaf to short days is sufficient to induce flowering (e), but plants from which all the mature leaves have been removed will not flower even in short days (f).

These experiments suggest that under short-day conditions a flowering stimulus is formed in the mature leaves of Xanthium and is transported to the growing apices of the shoot where it causes flowers to be initiated. The flowering stimulus, which resembles plant and animal hormones by being formed in one part of the organism and utilised, after translocation, in another part, has been called florigen but so far all attempts to extract florigen or any chemical substance which induces flowering when applied to another plant have failed. However, experiments can be carried out which allow us to determine some of the characteristics of this hypothetical substance florigen.

Subjecting a single mature leaf of a Xanthium plant to short days is sufficient to induce flowering, but many other species were found to remain vegetative after this treatment (g). Perilla ocymoides, another short-day plant, would flower after a single leaf had been photoperiodically induced only if all other mature leaves of the plant had been removed (h) or placed in a dark box so that they received no light at all (i). It therefore seems probable that florigen in some species at least is destroyed or inactivated in the presence of leaves which are not producing it. This was demonstrated conclusively by a Russian botanist called Cajlachjan, using plants of Perilla on which only one mature leaf was left. These plants flowered if the single remaining leaf was in short days (j) but not if it was exposed to long days (k). Cajlachjan then subjected each half of the leaf to a different length of day by wrapping black paper each afternoon over the part which was to receive short days. The plants flowered even when only the proximal half of the leaf was in short days, the distal half being in long days (1), but they failed to flower when the distal half of the leaf was in short days and the proximal half in long days (m). If longitudinal halves of the leaf were exposed to long and short days simultaneously the plants failed to flower (n). However, plants flowered in all those treatments where one page 6 half of the leaf was exposed to short days and the other half darkened continuously (o, p, q).

Evidently in short-day plants florigen is formed only in leaves which are exposed to short days. It travels through leaf tissue receiving short days or continuous darkness but will either not pass through leaves exposed to long days or is inactivated in them. The inhibition exerted by parts of a leaf receiving long days is only apparent when the long-day part of the leaf is between the short-day part and the terminal bud of the shoot.

We have seen that florigen moves from a mature leaf to the apex of the same shoot but we may ask whether florigen also travels from one shoot to another. To test this, two-branch Xanthium plants were used. One branch (the donor branch) was maintained under short days which induce flowering, the other (the receptor branch) was kept in long days which prevent flower formation in this species. However, both donor and receptor branches of these plants flowered (r), indicating that florigen can be translocated over considerable distance in the plant.

The translocation of florigen has also been demonstrated by grafting experiments in which a photoperiodically induced Xanthium plant (the donor) was grafted to a receptor plant kept in long days. In successful grafts the florigen was translocated across the graft union into the receptor plant which subsequently flowered even though it had never been exposed to short days (s).

A striking experiment which indicates that the florigen of long-day and short-day species is identical was carried out by grafting long-day donor plants to short-day receptors or vice versa. In each case the donor species was maintained under the appropriate inductive day-length, which was of course unfavourable for flowering of the receptor. But in each case the receptor plant flowered despite the adverse day-length conditions, as a result of the florigen it had received from the induced donor plant.

A natural case of grafting and florigen transfer is that of dodder, a parasitic plant which attaches itself to a host plant by means of haustoria which penetrate the tissues of the host. If the host is a short-day species and becomes photoperiodically induced the attached dodder plant will flower. If, however, the host is a long-day species the parasitic dodder plant will not flower in short days but only after the host begins to flower in long days.

page break
Fig. 1a-s. Diagrams which illustrate experiments on the short-day species Xanthium (a-f, r, s) and Perilla (g-q) by which localised perception and translocation of the flowering hormone through the plant were demonstrated. Full descriptions of the experimental treatments are given in the text. Unshaded areas denote parts of the plants which were exposed to long days, hatched areas were exposed to short days, and black parts were maintained in continuous darkness. Plants in treatments b, d, e, h, i, j, l, o, p, q, r, and s flowered, and flowers are shown at the apex of these plants. Plants in treatments, a, c, f, g, k, m, and n failed to flower.

Fig. 1a-s. Diagrams which illustrate experiments on the short-day species Xanthium (a-f, r, s) and Perilla (g-q) by which localised perception and translocation of the flowering hormone through the plant were demonstrated. Full descriptions of the experimental treatments are given in the text. Unshaded areas denote parts of the plants which were exposed to long days, hatched areas were exposed to short days, and black parts were maintained in continuous darkness. Plants in treatments b, d, e, h, i, j, l, o, p, q, r, and s flowered, and flowers are shown at the apex of these plants. Plants in treatments, a, c, f, g, k, m, and n failed to flower.

page 8

Before florigen will cross a graft it is necessary that the two cut surfaces should heal together in a graft union which suggests that the stimulus is able to travel only through contiguous living cells. In the leaf florigen can be made to travel through mesophyll cells by severing the veins of the leaf blade, but in the petiole and stem it probably travels in the phloem with the stream of photosynthetic products. Thus if the petiole tissue is killed by steaming but left intact, florigen cannot pass through the dead cells, even though the transpiration stream travelling in the xylem continues.

Short-Day Plants

From these experiments what general conclusions can we arrive at concerning the mechanism of flowering? It was suggested that florigen of long-day and short-day plants is identical, but it seems that the steps in its synthesis in the two groups must differ. It will therefore be best to look at each separately.

Before short-day plants will flower they must be subjected to a day-length which does not exceed a certain critical value and each short day must be followed by a night of relatively long duration. Short-day plants will not normally flower if kept in continuous darkness and it seems to be the regular rhythm of short days and long nights which causes them to flower. The necessity for light suggests that photosynthesis or one of the intermediate reactions of photosynthesis may be involved in flowering (reaction I). The relationship between photosynthesis and flowering is also suggested by the following facts. Both processes require carbon dioxide to be present in the air and both proceed only in light of high intensity. Furthermore, some plants such as potato which have large photosynthetic reserves are able to flower even when kept in darkness, and in some other short-day plants injected sugars can substitute for the light period.

For flowering to occur each favourable light period must be followed by a long period of uninterrupted darkness. From this we may infer that the substance formed in the light is utilised during the darkness. This dark reaction (reaction II) becomes effective in causing flowering only when the dark period exceeds the critical night-length for the species. Evidently the substance formed in the dark reaction is synthesised only slowly in the leaves each night, and therefore a fairly prolonged period of darkness is necessary. We know this because lengthening the days beyond the critical day-length by means of supplementary artificial illumination even of low intensity prevents flowering. The dark reaction is photosensitive, only occurring in the absence of light. Thus short-day plants will not flower even in short days if the night is interrupted by a brief period of light (reaction X). This light break need last only about one minute and be of low intensity illumination to destroy the beneficial effects of a long night. Furthermore, a light break is most effective in preventing flowering when it is given at about the middle of the dark period. Earlier or later interruption of the dark period is less effective, and if, as previously page 9 suggested, the synthesis of the dark reaction product occurs fairly slowly, we can explain this result, and also gain considerable insight into the mechanism of floral induction. A light break early in the night would destroy only the small amount of dark product already formed in the leaves; the ensuing period of darkness would still be long enough to allow more of this substance to be synthesised and flowering would occur. By the middle of the dark period a large amount of dark product would have been synthesised and this would all be destroyed by a light break at this time. The ensuing dark period would be too short to allow enough dark product to form, so flowering would not occur. But towards the end of the dark period most of the light product would have been converted to dark product and most of this would already have been translocated out of the leaves, so that a light break at this time would destroy only the small amount of dark product which still remained in the leaves and would not prevent flowering. Light therefore seems to have two opposing actions, high light intensities providing precursors and promoting flowering, low intensity light breaks destroying photosensitive products of the dark reaction and thus inhibiting flowering.

The substance formed in the dark reaction is translocated from the leaves, where it is synthesised, to the apices where it causes the initiation of flowers. During this time a further reaction must occur to render the flowering hormone light-stable (reaction III), otherwise plants requiring more than one photoperiodic cycle for induction would always fail to flower, for the hormone would be destroyed during each light period. It is thought that the hormone is light-sensitive only in the leaves and that in the stem it is converted to a light-stable form. Once the hormone reaches the apex it evidently begins to be synthesised in meristematic tissue, and its further synthesis there is independent of length of day, for once flowering has commenced it will continue under all conditions of day-length and after all photoperiodically induced leaves have been removed from the plant.

The sequence of photoperiodic induction in short-day plants can be written as follows:—
Reaction I=high intensity light reaction (photosynthesis).
II=photosensitive reaction occurring only in mature leaves in darkness.
III=photostabilisation of flower hormone in stem tissue.
X=antagonistic effect of long days, low intensity supplementary illumination, or light breaks on reaction II.

Horizontal arrows indicate reactions favourable to flowering, vertical arrows those which inhibit flowering.

page 10

Long-Day Plants

Long-day plants flower with least delay in continuous light of high intensity, periods of darkness of all durations decreasing the flowering response, until when in excess of a certain critical value darkness prevents flowering entirely. Thus darkness does not seem to play a positive role in the formation of a flowering hormone in long-day plants. As in short-day plants a high intensity light reaction is a necessary preliminary to flowering (reaction I). The flowering hormone is synthesised directly from the product of this reaction in periods of light (reaction III), and is inactivated during periods of darkness (reaction II). Thus in long-day plants darkness is inhibitory to flowering whereas in short-day plants long periods of uninterrupted darkness are necessary for flowering to occur. Long-day plants will not flower in short days for two reasons. First, in short days insufficient hormone would be synthesised to reach the threshold level, and secondly, this hormone would all be inactivated in the ensuing long dark period.

As in short-day plants the effect of darkness can be negatived by light breaks of low intensity and short duration given at the middle of the dark period (reaction X). However, the effect of the light break is opposite in the two groups because in short-day plants a light break prevents flowering while in long-day plants a light break promotes flowering. Thus again light appears to have two distinct actions, high intensity light promoting flowering, low intensity light breaks reversing the effect of darkness.

The sequence of photoperiodic induction in long-day plants can be written as follows:—
Reaction I=high intensity light reaction (photosynthesis).
III=synthesis of flower hormone in light.
II=antagonistic effect of short days, or long periods of darkness.
X=effect of light breaks or low intensity supplementary illumination antagonistic to reaction II.

Horizontal arrows indicate reactions favourable to flowering, vertical arrows those which inhibit flowering.

These schemes for the sequence of floral induction in long- and short-day plants are as yet hypothetical explanations of observed facts. The reacting substances formed under the different conditions have not yet been identified, but recognition of the several steps involved in the synthesis of a flowering hormone is a necessary preliminary to a search for the page 11 substances involved. It is also evident that the original concept of ‘florigen’ will require to be modified in the light of this more recent work.

One of the difficulties faced by workers on photoperiodism is that an experimental alteration in the length of day results in a concomitant alteration in the night-length, the combined length of the day-night cycle still being 24 hours. However, to test general hypotheses it would clearly be desirable to alter the length of one period without affecting the other, or to vary the length of both in the same direction simultaneously. In recent years several experiments in which all of the light is supplied from artificial sources have been carried out, using light-dark cycles varying from one to 72 hours. These experiments give added support to some of the present ideas, but indicate that others may have to be further modified or should be accepted only with caution. However, until more species have been subjected to this type of experiment it is not possible to assess the general validity of the results.

Other Photoperiodic Effects

So far we have considered the role of day-length only as it affects the initiation of flower primordia, but although this is the most spectacular response may other activities of the plant are sensitive to the effects of altered day-length.

The vegetative growth of both long-day and short-day species is usually more prolific under long days. In short days the root growth of many species is strikingly decreased, onions may not ‘bulb’, and potatoes may fail to form tubers. Leaf size, shape, texture and pubescence have been shown to vary greatly in many species under different lengths of day.

The onset of dormancy, development of frost resistance, and shedding of leave in deciduous species are, in part, responses to the shorter day-lengths encountered in autumn, and can be delayed by extending the daily period of illumination. Dormant plants can often be made to shoot prematurely by increasing the length of day.

The initiation of flower primordia may not always lead to the completion of the reproductive cycle, because the development of flowers and fruit may require different photoperiodic conditions from those which stimulate floral initiation. Some varieties of soybean are day-neutral and initiate flower primordia even in continuous light. However, the primordia do not develop further under these conditions. Flowers do develop but fail to open when the length of day is 16-18 hours. In day-lengths of 13-16 hours flowers open but no fruit are set, fruit developing to maturity only when the day-length is less than 13 hours. In Xanthium continuation of short days beyond the number necessary to induce flowering results in a high percentage of abortive pollen grains and the development of few fruits, many of which are empty having developed without fertilisation.

The length of day may also determine the type of flower which develops. In Viola cleistogamous flowers, which never open and are self-pollinated, page 12 develop in days of 12-14 hours, but in days of about 8 hours flowers which open and are cross-pollinated develop. Sex reversal in unisexual flowers may be brought about under altered day-lengths. In maize the terminal tassel produces only male flowers when plants are grown in days of 13-15 hours, but in days of 8-10 hours female flowers and fruit develop on this part of the plant, the formation of male flowers being completely suppressed.

The length of day is clearly a very potent factor controlling as it does many of the vegetative and reproductive phases of development. Knowledge of the way in which light affects flowering is already being put to use in agriculture and horticulture, but before we can make general use of photoperiodic principles to control more exactly the development of crop plants, we will require much more detailed information concerning the separate stages of development of each species and the chemical control of these several stages.

Reading List

Ashby, E., 1947. The control of flowering. Science News 4, 127-138.

Bonner, J., and Galston, A. W., 1952. ‘Principles of Plant Physiology. Ch. 17. The physiology of reproduction. W. H. Freeman and Co., San Francisco, U.S.A.

Garner, W. W., and Allard, H. A., 1920. Effect of relative length of day and night and other factors of the environment on growth and reproduction in plants. J. Agric. Res., 18, 553-606.

Lang, A., 1952. Physiology of flowering. Plant Physiol, 3, 265-306.

Leopold, A. C., 1951. Photoperiodism in plants. Quart. Rev. Biol., 26, 247-263.

Murneek, A. E., and Whyte, R. O., 1948. ‘Vernalisation and photoperiodism.’ Chronica Botanica Co., Waltham, Mass., U.S.A.