ISHS Acta Horticulturae 14: Symposium on Flower Regulation in Florist Crops, 14: 157-166 (1969)
R. Moe and T. Kristoffersen
Department of Floriculture and Greenhouse Crops, Agricultural College of Norway, Vollebekk, Norway


When the light intensity is at minimum in the winter months roses are subjected to 2-3°C, and hardly any growth is detected. As the light improves, the temperature is gradually increased, and the flowers are cut in the early spring. However, many shoots fail to flower (blind shoots) especially in the first flush. This is a great problem in commercial production of roses, determining the economy of the production. Kamp (1948) and Zink (1950) reported increased blindness with decreased light intensity, while Hubbell (1934b) was not able to detect more blindness at reduced light. The temperature seemed to have only minor effect (Zink 1950).

Some of the buds may remain dormant for a long period of time after pruning. This may also happen after the flower is cut at the normal stage for marketing. Removing leaves adjacent to dormant buds, has been demonstrated to break dormancy. Kamp (1948) found the proportion of blind shoots to increase simultaneously while Durkin (1965) detected less blindness with defoliation. The rate of growth and flower production is positively correlated with the solar radiation (Post and Howland 1946, Chandler and Watson 1954, Farmer and Holley 1954), and the rate of development is positively correlated with the temperature (Boodley and Seeley 1960).

Abnormal colouring of the petals is frequently experienced. 'Baccara' tends to become very dark (blackening) at times. Low temperature has generally resulted in higher anthocyanin content in the plants (Harborne 1967). It is evident that light is required for the synthesis of anthocyanins (Shisa and Takano 1964), and that the nutrition of the plant may modify the colour intensity (Abernathie 1960, Lindstrom and Markakis 1963, Shisa and Takano 1964).

The quality of the cut flower, the strength and the length of the stem, and the size, shape, and colour of the flower, is of great importance in commercial production of roses. However, the literature offers no precise information on the effect of the climatic conditions on the growth and development of the rose shoot. With the intention to work out the optimal temperature regime for roses, several experiments were conducted which will be reported on in the present paper.


One year old plants of the cultivar 'Baccara', grafted on Rosa canina were cut back above two dormant buds, and two branches remained on each plant. Two plants were planted together in 10 l plastic buckets in perlite and peat (3:1), and supplied daily with a complete nutrient solution. Eight plants were included in each treatment where not otherwise stated. The experiments were conducted in a phytotron with natural or artificial light (Philips TL 33), where the temperature was maintained constant (± 0.5°C). Fluctuating temperature was established by moving the plants (trucks) at 0700 and 1600. The air velocity was approximately 0.5 m per sec., and the water vapour pressure deficit of the air corresponded to 5 mm of Hg. The air was renewed 10 times per hour.


Differentiation of the floral organs

One hundred and forty uniform shoots were selected and pinched above the first 5-leaflet leaf. They were subjected to 15 and 21°C (70 at each temperature), and natural light June 23. Five shoots were randomly selected at regular intervals and the apical meristem was studied by the means of a binocular. Eight stages in the flower differentiation were detected. They are presented in figure 1.

The rate of development at 15 and 21°C is illustrated in figure 2. Stage (2) appeared 1 week earlier at 21 than at 15°C. From stage (2) to stage (7) the rate of development was almost the same. From stage (7) on a more rapid development was detected at 21 than at 15°C. The length of the shoot in relation to the developmental stage was identical at the two temperatures. The shoots elongated slowly until the stamens and pistils were differentiated (stage (5) and (6)). Figure 3 shows that a rapid elongation occured at this stage.

The effect of temperature and light intensity

Three experiments were carried out in order to study the effect of light intensity and temperature on the growth and development of the flowering shoot. In the first one the plants were grown at different temperatures and light intensities from April 20 to October 5, as indicated in table 1. The constant temperature treatments in full sunlight were repeated the following year, starting February 6, and 16 plants were included in each treatment. At the same time a third experiment was conducted where 20 plants were subjected to 2000, 6000, 8000, and 10 000 lux artificial light for 16 hours per day at 12 and 21°C. In all experiments the number of days from pinch to flowering, stem length (cut above the first 5-leaflet leaf), number of leaves, the size, number and the colour of the petals was recorded.

The number of days from cut back to flowering in natural light and constant temperature is shown in figure 4. At a constant temperature, 24-27°C, maximum rate of development was obtained. Higher day or night temperature resulted in significantly more rapid development than lower temperature. The effect of night temperature was larger than the effect of day temperature when these were applied for 15 and 9 hours respectively, as can be seen in table 1. The number of days from cut of the first to the second, and from the second to the third flower was cut on the same shoot, was close to the same number of days from cut back to the first flower was cut. The effect of the temperature followed the same pattern as described for the first flush. Figure 5 illustrates the rate of development of the flowers in the second flush at different light intensity. Reduced light intensity did not affect the flowering in the first flush, but in the second flush flowering was delayed. In artificial light, plants receiving 2000 lux flowered surprisingly at the same time as those receiving 10 000 lux.

The number of flowers yielded per treatment was highly affected by the temperature. At constant 12, 15, 18, 21, 24, and 27°C 16, 25, 38, 40, 37, and 46 flowers were cut per treatment (8 plants) respectively during 5.5 months. Reduced light intensity degraded the production of flowers. In 50 and 35 per cent of full sunlight at 18°C the plants yielded 27 and 20 flowers, while 38 flowers were recorded in full sunlight.

The length of the flowering shoot was significantly reduced with increasing temperature (p=0.05), as can be seen in figure 6 for the constant temperatures. The number of leaves per flowering shoot was, however, not affected by the temperature, and as a consequence the length of the internodes decreased with increasing temperature. At the low temperatures (12-15°C) the stems were very heavy and they decreased rapidly in thickness at increasing temperatures. The best quality was obtained at 18°C. At higher temperatures the colour of the flower was too pale. At temperatures below 18°C the petals had a dark bluish tint, and many of them did not open in the regular way ('hard' buds).

The number and the size of the petals decreased significantly with increasing temperature, as is illustrated in figure 7 (p=0.01). At all temperatures the width of the petals was larger than the length. In artificial light, the number of petals were not affected by the light intensity (2000, 6000, and 10 000 lux), and approximately 60 petals per flower were detected at 21°C both in the winter and in the summer in the greenhouse. However, the size of the petals increased with increasing light intensity as is shown in figure 8.

The outer petals were frequently somewhat abnormal. They often had green areas in the central parts, but this was not found to be affected by the temperature.

When studying the anatomy of the petals, one layer of epidermic cells was detected on each side, and the cuticula was lacking on the upper side. The epidermal cells were conic on the upper (inner) side, and cubic on the lower (outer) one. At low temperatures (12°C) these cells were larger and more pointed than at higher temperatures. The mesophyl was only little differentiated and did not contain red pigments. The colour of the petals was very temperature sensitive, increasing with decreasing temperatures. At 12°C black areas were detected, especially on the lower side. Generally, this side was more intense coloured than the upper one. The best colour from a commercial point of view was obtained at 18°C. The colour was not affected by the light intensity at low temperature (12°C), but it increased with increasing light intensity at higher temperatures.

The size and the shape of the leaves varied with the temperature, as demonstrated in figure 9. Large leaves were observed, and especially the younger leaves, thorns, and shoots were reddish at low temperatures. More thorns were present at high than at low temperature, but they became larger the lower the temperature.

Blind shoots

When the roses were cut back February 6, and subjected immediately to different temperatures, it was discovered that blindness was affected by temperature. At constant 12°C as much as 87 per cent of the shoots failed to flower, and at 24°C 57 per cent of all shoots were blind.

When the roses were cut back April 20, and grown at different light intensities, it was revealed that blindness also is affected by light intensity, reduced light increased the number of blind shoots as figure 10 demonstrates. Less blindness was detected in summer than in early spring. Increasing temperatures resulted in decreased blindness. At 12 and 27°C constant temperature for example, 40 and 15 per cent of the shoots respectively were blind.

Short days during the spring could have influenced the blindness. However, this was not confirmed when the plants were exposed to artificial light for 16 hours per day, as can be learned in figure 11. It seemed likely then that blindness is primarily induced by low temperature. In order to test this hypothesis, another experiment was set up where the plants were subjected to 9 or 24°C for 1, 2, or 4 weeks immediately after cut back in February. These temperature treatments were succeeded by 12°C until flowering. Figure 12 indicate that the plants were suspectable to low temperature in the early development of the shoots, and that blindness may partly be overcome by raising the temperature above 20°C for a few weeks after cutting back.


The informations gained on the flower differentiation in 'Baccara' correspond by and large with experience in other cultivars (Hubbell 1934a, Rouffa and Gunckel 1951, Lindstrom 1956). When the shoot is cut, the axil bud will start to elongate, and the flower is initiated (Hubbell 1934a, Laurie and Bobula 1938, Lindstrom 1956, Carpenter and Watson 1965). The literature is, however, not conclusive on the number of days from cut to visible flower primordia (stage 2). This may be due to differences in the climatic conditions. In the present study this stage was detected 5 days after cut back at 21°C and after 11 days at 15°C. At this stage, the length of the shoot was approximately 10 mm long, independent of temperature. It has been stated that there has to be a certain activity in the plant before flowers may be initiated (Carpenter and Watson 1965). If this just is a question for respiration or activation of growth regulators remain to be shown.

The differentiation of the flower and the leaves (figure 2 and 6) was found to be relatively little dependent upon temperature, and this is in agreement with the requirement for differentiation in general.

Laurie and Bobula (1938) and Rauh and Reznik (1951) are of the opinion that stamens are differentiated before pistils. In 'Baccara' the petals are differentiated from the edge toward the center of the meristem. In double flowers the transition from petals to stamens is rather diffuse. The appearance of the petal and stamen primordia is very similar. The anthers were visible only 2 days after differentiation of the pistils and it is likely that pistils and stamens are initiated almost at the same time, and it may be questioned if stage 5 and 6 then can be separated.

It was manifested that the rate of elongation of the shoot was very slow until stage 7 in the flower differentiation was reached. This may be due to the fact that the presence of a flower bud is required before rapid elongation (Lang 1961). It is assumed that elongation of the flower stem is correlated with the auxin synthesis in ovaries and anthers, and Lang states further that the auxin production may stop at the time of anthesis or that growth inhibitors are produced. The rate of stem elongation (cm per week) was almost the same at 15 and 21°C in the present study, but anthesis was reached later at the low than at the high temperature, and consequently the flowering shoot became taller at 15 than at 21°C.

The effect of light intensity on the flower production was only small in short term experiments. However, in long term experiments the yield of flowers is significantly reduced, when the light intensity is limited. Post and Howland (1946), Chandler and Watson (1954), and Farmer and Holley (1954) observed less vegetative growth at low than at high light intensity, and they presumed that this was the reason for reduced crop. In the present study, however, the main cause of reduction of flower numbers at decreased light, was the increase in blindness with decreasing light intensity.

The colour of the petals is of great economic interest. Shisa and Takano (1964) demonstrated that the red anthocyanins are not synthesized at all at 30°C in the cultivar 'Crimson Glory'. In the present study 'Baccara' became very dark at low temperature (12°) and this was probably due to high anthocyanin concentration, as has been found in other cultivars (Abernathie 1960, Shisa and Takano 1964). The colouring of the cultivar 'Masquerade' is completely supressed at low light intensity (Shisa and Takano 1964), and in the present study interaction between light intensity and temperature was detected. This is probably correlated with the net photosynthesis in the plant.

Smaller flowers were obtained at high than at low temperature due to fewer and smaller petals. This is in agreement with earlier studies with 'Ma Perkins' and 'Crimson Glory' (Semeniuk 1964, Shisa and Takano 1964) In 'Baccara' all the petals were differentiated at stage 5 and 6 (figure 2), 3-4 weeks after cut back. At low temperature (12°C) a great many malformed flowers (bullheads) were present, while bullheads were never observed at high temperature (21°C). The temperature has therefore to be adjusted properly in order to obtain the desired number of petals and normal development of the flower.

Dormant buds and blind shoots represent a great problem in commercial production of roses, and specific effects of temperature and light was learned in the present study which may be beneficial to the growers. Blindness was visible approximately 30 days after cut back, and blind shoots became only 25-30 cm long while normal ones were twice as tall. The shoots were 38-40 cm when stage 5 was reached, and the blindness was probably induced before this stage. Hubell (1934a), Lindstrom (1956) and Zolotovitch et al. (1964) are of the opinion that blindness is caused by abscission of the flower organs. Abscission layer was detected before the stamens were differentiated (Lindstrom 1956). This phenomen, however, occur in many plants (Addicott and Luch 1955). Reduced auxin content has been detected in leaves before the abscission layer is formed. Addicott et al. (1964) have isolated abscission inducing substances in young cotton plants and these are now called abscisin II or dormin. It remains to prove how light and temperature affect the production of abscisin in 'Baccara', but it is likely that the production is stimulated by low temperature and low light intensity. The effect of high temperature after cut back, may prevent abscisin production and maintain optimal concentration of auxin and other growth regulators.