Aust. J . Agric. Res., 1976, 27, 399-407
Temperature Effects on Flower Initiation
in Sweet Orange (Citrus sinensis)
G. I. Moss
Division of Irrigation Research, CSIRO,
Private Bag, P.O., Griffith, N.S.W. 2680.
Abstract
The effects of temperature on flower induction in sweet orange were studied over a range of shoot
and root temperatures representing the conditions found in the field in commercial growing regions.
Rooted cuttings of the orange cultivar 'Late Valencia' were used in the Canberra phytotron, with
an extended photoperiod of 16 hours and a photo-temperature of 8 hours' duration.
Following low temperature inductive conditions (5 weeks at 15/1OoC),profuse flowering occurred
at 24/19"; flowering was only slightly less at this air temperature with a root temperature of 30°,
and at an air temperature of 30125". At 36/31" few flowers were formed and these abscissed before
opening. However, when the root temperature was maintained at 22" with this shoot temperature
a few flowers were formed. In another experiment a root temperature of 30" and shoot temperature
of 24/19" reduced the number of flowers that formed (86 cf. 137). Only a small part of the effect
of high temperature in inhibiting flower initiation was due to root temperature.
Two attempts were made to induce flowering under non-inductive temperatures (27122") with
low root temperatures (15 and 11"). Flowering could not be induced under these conditions, and
the lack of flowers apparently was not due to inhibition of growth by low root temperature.
Although few flowers were formed at 27/22" without any previous inductive treatment (15
flowers per plant), profuse flowering was induced by 27/13" (141 flowers/plant), which indicated
that a low night temperature will induce flowering.
A high root temperature (29') during inductive conditions (15110") had little effect on the number
of flowers formed. It appeared that the site of flower stimulation by temperature was in the shoot
only, and that subsequent high root temperature effected floral development.
It was concluded that flowering could be induced in the field during the summer by low night
temperature, and root temperatures would probably have little influence on floral development.
Introduction
Temperature is the most important factor controlling flowering of sweet orange
in the commercial citrus-growing regions of Australia, where flowering occurs in
the spring after several months with mean temperatures below 15°C (Moss 1969).
Only at moderate temperatures (24/19")* are short days necessary for flower induction,
while at higher temperatures the photoperiod seems mainly to control the rate of
growth (Lenz 1969). Commercially this is useful, since it ensures one crop per year
as the spring blossoming is usually very heavy; fruit from the spring flush inhibits
flower formation at other times of the year (Moss 1971). However, citrus is a subtropical evergreen and under tropical conditions flowers may be formed at each
flush (there may be four flushes of growth per year) and fruit of all stages of maturity
* The first figure refers to the photo-temperature
maintained for 8 hr, the second to the temperature
during the rest of the day; in the case of long days (LDS)a further 8 hr of light was provided.
G. I. Moss
are found on the tree. Likewise in warm temperate inland areas fruit may be set
at other times if the amount of fruit set in the spring flush is small; these laterflowering flushes are of no commercial value. Little is known about the conditions
stimulating or inhibiting flowering under summer conditions, so this work was aimed
to provide basic information in this area.
Materials and Methods
Methods
Sweet orange cuttings of the cultivar 'Late Valencia', at least 1 year after
rooting, were grown in sand culture as described in previous work (Moss 1969). The
experiments were carried out in the Canberra phytotron in temperature-controlled
conditions with natural daylight. An extended photoperiod of 16 hr was used, partly
to promote a higher growth rate. No effect of photoperiod on floral development
after induction or early initiation has been detected.
Prior to the experiments plants received 4 weeks at 27/22OC, or higher, to initiate
and complete a vegetative flush to ensure that the plants would not be producing
flowers initiated by some previous regime. Sweet orange cuttings under phytotron
conditions can form flowering shoots on any mature flush, so a further aim of the
pretreatment was to ensure a uniform set of 'mature' terminal growth flushes. Where
the treatments were intended to study flower development, flower initiation was
induced by 5 weeks at 15/10'.
Table 1. Soil, leaf and air temperatures of orange plants in pots in a glasshouse as influenced by two
methods of controlling soil temperature
Temperatures ("C) were measured continuously by thermocouples, means compiled from 24 hr
readings. All conditions were measured concurrently
Temp. coadition
of soil in pot:
1. Ambient-on
glasshouse bench
2. Controlled-pot in
enclosed water-bath
3. Controlled-pot in
open water-bath
Time of day (hours):
0345
1345
Mean
0345
1345
Mean
0345
1345 Mean
Soil temp. (10 cm)
Lower leaf temp.
Upper leaf temp.
Airtemp.aboveplants
13.6
11.9
11.6
11.9
23.5
23,8
23.8
24,8
17.8
17.0
17.0
17.5
28.2
15.0
12.6
12.2
27.2
23.8
23.1
24.8
28.7
18.7
17.5
17.9
27.0
15.1
13.3
13.3
29.2
27.0
28.4
25.9
28.2
20.3
19.8
19.1
Root Temperature
Two types of equipment for controlling root temperature were compared. In
the first, root temperatures were maintained by placing sealed pots in a water-bath
with provision for pot drainage. However, the temperature of the foliage was found
to be influenced by the temperature of the water-bath when this was above the
ambient air temperature; this was especially so in a glasshouse with no forced
ventilation (Table I). The second type of bath was insulated around the sides and
on top with 25 mm of polystyrene foam and the upper surface covered by sheet
aluminium; this was used in all subsequent experiments. Where the experiment
involved controlled root temperatures, then these were monitored.
Flower Initiation in Sweet Orange
These may be of several types: single-flowered, single-flowered with leaves,
multiple-flowered, or multiple-flowered with leaves (Moss 1970). In the field these
types are significant since the majority of the fruit is set on the leafy types (Sauer
1951; Lenz 1966). The ratio of flowers to inflorescence leaves in each inflorescence was
measured as an indication of the type of inflorescences obtained as well as the mean
number of flowers per inflorescence.
Procedure and Results
Experiment 1
Aim. This was to examine the effect of temperature on subsequent flower formation
after induction.
Pretreatments. Following 4 weeks at 27/22' flowering was induced by 5 weeks
at 15/10" followed by the treatments.
Treatments. 24/19" (control); under this treatment flower development is rapid
and normal, but it does not induce further flowering.
24/19" with a root temperature of 30".
30125".
36/31".
Table 2. Effects of temperature on flower initiation following induction
All temperature treatments were carried out with 16 hr photoperiod; root temperature was
ambient unless indicated by RT. Pretreatment: 27/12" for 4 weeks followed by 15/10" for 5 weeks
Temperature
treatment
A
Flowers per plant
Flowers per Ratio of flowers to
Plants per
treatment (before first flowers opened) inflorescence inflorescence leaves
Standard error of mean.
Flowers abscissed before fully developing.
Results. Flowering was recorded at weekly intervals until petal fall, but as some
treatments resulted in abscission before opening, the number of flowers present
before the first blossoms opened was taken as the record.
The most notable effect was that 36/31" resulted in a reversal of floral development;
few flowers were formed despite an adequate floral-inductive stimulus (compare with
treatment 24/19") and those few blossoms abscissed before opening (Table 2). The
treatment 30125" resulted in only slightly fewer flowers than at 24/19', partly because
there were fewer flowers per inflorescence and the latter were of the more leafy type.
A root temperature of 30" had only a slight effect on inflorescence development, and
the results are not very different from either the 24/19" or the 30125" treatments.
Experiment 2
Aims. These were to see if the adverse effect of high temperatur
ment was due to high root temperature, to test if low root te
G. 1. Moss
inductive conditions would induce flowering, and to see if low night temperature
would induce flowering.
Pretreatment. A temperature regime of 36/31' was imposed for 4 weeks to remove
any tendency of the plants to flower.
Treatments. 15/10' for 5 weeks to induce flowering, followed by 24/19" (control).
15/10" for 5 weeks, followed by 3613 1" with a root temperature of 22".
27/22'; this treatment was intended to test how much flowering was
induced by this regime with no inductive pretreatment.
27/22" with a root temperature of 15"; no inductive pretreatment.
27/13" to see if low night temperature would induce flowering;
no inductive pretreatment.
Table 3. Shoot, root and night temperature effects on flower initiation
All temperature treatments were carried out with 16 hr photoperiod; root temperature was ambient
unless indicated by RT. All plants were initially subjected to 36/31" for 4 weeks
Temperature
treatment
Pretreatment
15/10"5weeks
-
15/10" 5 weeks
Flowers per
plant
24/19"
207+43
27/22"
15+6
27/13"
141+32
27/22", RT 15'
30+ 7
3 6 / 3 1 " , ~ ~ 2 2 ~48f-18
Inflorescences
per plant
44.0+9
4.4k1.4
27.0+5
6.2+ 1
12.0f 3.6
Vegetatwe
shoots per plant
8.3L2.5
6.5+1.0
7.2+ 1 . 4
2.3+0.7
6.320.9
Ratlo of flowers to
inflorescence leaves
1.8920.31
0.72+0.09
1.1720.07
1.04+0.13
1.19k0.14
Results. The 24/19' plants flowered profusely (Table 3); the low temperature
pretreatment induced a large number of buds to grow out, and although many were
floral a significant number (8.3 per plant) were vegetative. The inflorescences were
predominantly floral. The data were taken at the time of full bloom.
Very few inflorescences were induced at 27/22' (without any pretreatment at low
temperature) and these tended to have fewer flowers and more leaves. However,
with lower night temperatures a large number of inflorescences of the less leafy type
were produced. On the other hand, a low root temperature at 27/22' failed to stimulate
floral initiation; the number of inflorescences differed only slightly from that of
the 27/22" treatment, although there were fewer leaves on the inflorescences. One
notable feature was that fewer vegetative shoots were induced by the low root
temperature. With the 22' root temperature at 36/31" some flowers were formed,
although flowering was still suppressed compared with the 24/19' treatment, which
suggests that the role of root temperature in reversing flower initiation at higher
temperatures is relatively small.
Experiment 3
.
Aim. This was to see if low root temperature would induce flowering.
Pretreatment. Four weeks at 3613 1".
Treatments. 27122"-no control of root temperature.
27/22' and a root temperature of 11" (+2').
Flower Initiation in Sweet Orange
Five weeks after commencement of these treatments flowering was recorded; all
flower buds were removed, and the root temperature treatment discontinued.
Flowering was again recorded 33 weeks later.
Results. Only half the plants at 11" root temperature produced any flowering
shoots, and these inflorescences did not differ in the mean number of flowers or in
the flower/inflorescence-leaf ratio from those with ambient root temperature (Table 4).
Table 4. Attempt to induce flowering with low root temperature
All treatments were carried out with 16 hr photoperiod, root temperature ambient unless indicated
by RT
Init~al
treatment
No. of
Inflorescences Vegetattve
per
plants
shoots
flowermg
per plant
plant
Flowers
Ratto of flowers
per
to mflorescence
mflorescence
leaves
27/22'
27/22",
RT 11'
9
6
5.711.7A 11.2k1.9 3.120.4
3.011.1 2.4L0.5 2.910.6
27/22'
4
1.610.9
2.2k0.9 2.110.5
27/22",
11'
7
9.114.1
5.011.4
3.lk0.6
RT
A
Mean stem
dlameter
(mm)
T~me
of
recording
I
0.85f 0.06 1 4 . 3 k 0 . 8 5 weeks after
0.89L0.05 1 2 . 6 1 0 . 7 commencement
of treatment
Flowers remov1.1620.11
n.d.
ed, root temp.
discontinued,
1.21k0.13
n.d.
recorded 3;t
weeks later
I
Standard error of mean, 12 plants per treatment.
Considerably fewer vegetative shoots were produced and growth was restricted by
the low root temperature. In the later recording, more inflorescences were formed
on the plants from the low root-temperature treatment, but this was variable. One
can infer from this that low root temperature is not responsible for floral induction
when plants are maintained at temperatures below 21'.
Table 5. Effect of high root temperature on flower initiation
All treatments were carried out with 16 hr photoperiod, root temperature ambient unless indicated
by RT. All plants were initially subjected to 15/10" for 5 weeks
Treatment
No. of plants
flowering
Idorescences
per plant
Vegetative
shoots per plant
12
86 12
+
6.3k1.3
24/19", RT 31'
A SE
Flowers per
inflorescence
3.6L0.2
Ratio of flowers
to inflor. leaves
Mean stem
diam. (mm)
2.0410.21
11.210.5
of mean, 12 plants per treatment.
Experiment 4
Aim. This was to check the effect of high root temperature on floral development.
Pretreatment. After 5 weeks at 15/10", to induce flowering, plants were transferred
to the treatments.
Treatments. 24/19" with ambient root temperature.
24/19" with a root temperature of 31" (the recorded root temperature
was 30-33').
G. I. Moss
Results. The higher root temperature did not inhibit flowering, but fewer inflorescences were formed compared with the plants with ambient root temperature (Table 5).
The number of flowers per inflorescence was not affected by the high root temperature,
but these inflorescences tended to be more leafy, as indicated by the lower flower/
inflorescence-leaf ratio.
Although the time to flowering was not recorded, there appeared to be little
difference between the two treatments. One difference was that less growth was made
at the high root temperature; this is evident by the fewer vegetative shoots and slightly
smaller stem diameter found in this treatment.
Experiment 5
Aim. This was to determine the relative importance of root and shoot temperatures during induction.
Pretreatment. 30125" for 32 days.
Treatments. 15/10" with ambient root temperature.
15/10" with a root temperature of 30". (During the course of the
experiment the root temperature rose above this for one day, but
at no time did the temperature fall below 29O.)
After 38 days all plants were transferred to 24/19" with ambient root temperature.
Table 6. Flowering of late Valencia orange after 56 weeks at 15" day, 10" night temperatureA:
ambient root temperature compared with 30" controlled root temperature
Root
temperature
Replicates
Nodes per
plant
No. of plants
flowering
Inflorescences
per plant
Ambient
15
205 39
+
13
15,325.8
Flowers
per plant
Vegetative
shoots per plant
- -
A
64k27
20.523.5
Plants transferred to 24/19", root temperature discontinued.
Results. Full bloom was recorded 21 days after cessation of the low temperature
treatment. There was no apparent difference in the time to flowering of the two
treatments. There was little difference in the flowering behaviour of plants from the
two treatments (Table 6). More flowering shoots and slightly more flowers were
found after the high root-temperature treatment, but the differences are not significant
when the greater variability of flowering in the high root-temperature treatment is
considered. This indicates that root temperature is of little importance during the
stage of flower induction in sweet orange.
Discussion
Flower formation in sweet orange is sometimes considered as consisting of two
or possibly three periods of development. (1) The first requirement is that buds
must be released from 'dormancy'; (2) then there may be a period of initiation;
(3) finally floral development occurs. The present results offer the following evidence
for this scheme.
Flower Initiation in Sweet Orange
(1) A period of low temperature induces many buds to grow out, and most of
these are floral. Against the suggestion for a first period would be the following
observations.
(a) The number of vegetative shoots is not greatly influenced by environmental
treatments except low root temperature, which results in fewer vegetative
shoots.
(b) Long days, which increase the rate at which buds are released* from dormancy
(Lenz 1969), prevent flower formation.
These objections may be answered by noting that treatments such as low temperature or low night temperature also favour flowering, while long photoperiods inhibit
flowering despite their effect on bud growth.
(2) Evidence for a second period is hard to justify, as it appears that sweet orange
has an innate tendency to form flowers, and the environmental conditions affecting
flowering are those that inhibit flower formation. Supporting this would be evidence
that under 27/22" (which is not warm enough to adversely affect floral development)
with long days, some flowering took place. In this case well-established cuttings
were used, and production of inhibitors may be less as the mean age of the leaves
increases (this conjecture would need experimental verification). However, if plants
are defoliated or pruned many buds are released from dormancy yet few flowers are
formed (Moss 1973), which suggests that an induction stimulus may be produced
in the leaves. Further evidence for this was given by Ayalon and Monselise (1960),
who used cincturing and defoliation techniques to demonstrate an inductive floral
stimulus in sweet orange. As high root temperature and inductive shoot conditions
(Table 6) or low root temperature under non-inductive conditions (Table 4) had no
effect on flowering, the production of the flowering stimulus is located in the shoot.
No evidence of floral differentiation in the buds during the inductive period was
found, and other workers have reported similar results (e.g. Abbot 1935; Ayalon
and Monselise 1960). This suggests that floral differentiation takes place some time
after induction.
(3) The third period, one of floral development, is better defined, as transfer to
high temperature prevents flower formation even after a period of 'inductive' low
temperature treatment (Table 2). High temperature reversal of the flowering response
to inductive conditions has been described for Pharbitis nil-a short-day plant (King
and Evans 1969). Part of this response in sweet orange (although a lesser part) is
controlled by the roots, since high root temperature with moderate shoot temperature
slightly reduced the number of flowers formed and further reduced the number
opening (Tables 2 and 5), while moderate root temperature with high shoot temperature partly mitigated the effect of high temperature (cf. Tables 2 and 3).
Root and Night Temperature Effects
In contrast to this work are the observations of Liebig and Chapman (1963) who
compared 14, 22 and 30" root temperatures and reported more flowering with low
root temperatures. In this case the real effect was probably that flowering was
* Sweet orange grows by a series of more or less synchronized flushes (Moss 1973).
G. I. Moss
adversely affected by 30°, so as to make it appear, by contrast, that the lower
temperature had stimulated flowering. This raises the question of the validity of
results in glasshouse experiments. Open water-baths used to control root temperature will result in a higher night temperature if this is below the temperature of the
water-bath. The effect of raising the night temperature a few degrees may be more
significant than the effect of root temperature in some instances.
With other quantitative short-day plants night temperature influences the response
to day length (e.g. Zea mays; Heslop-Harrison 1961). Therefore, it is not unusual
that flowering should be freely induced in sweet orange with a 16 hr photoperiod
and low night temperature of 13", although the day temperature was 27". On the
other hand, Hield et al. (1966) found that a day temperature of 29-38" prevented
flower formation with a night temperature of 7-12".
Flower Initiation in the Field
From a practical viewpoint this study indicated that flowers may be initiated in
sweet orange under a wide range of conditions, and may be initiated under Australian summer conditions in the commercial growing regions by low night temperature. Even in January, the mean minimum temperature is often below 18" (Table 7),
so that initiation could take place under relatively warm long days.
Table 7. Relevant field temperatures fnr Griffith during midsummer (January)
Mean maximum Mean minimum
(long-term)
(long-term)
A
temp^.^ under bare soil at 20 cm
0900 hours
1500 hours
Means for 1972-1974, but 1972 figures were for eastern standard time and
1973, 1974 were for daylight saving time.
Conclusions
Flowering in sweet orange is an innate tendency, with absence of inhibitors
possibly more important than promotory effects. Evidence here and in the literature
suggests that the flowering tendency is located in the leaves and probably transmitted
to the buds, rather than the alternative hypothesis that the buds are 'programmed'
to form flowers and can be suppressed by inhibitors. Flower induction and initiation
is affected by the shoot environment, while the process of floral development can
be suppressed by high shoot temperature with a small effect due to high root
temperature.
Acknowledgments
The following persons are thanked: Mr R. Gray for producing the plants used;
Mr R. Dunstone and the staff of the phytotron; Mrs K. Morrow for recording
experiments; Dr R. W. King for constructive criticism of this work and for suggesting
experiment 5; Dr P. J. M. Sale for assistance with writing; Mr S. J. Paterson for
technical assistance.
References
Abbott, C. E. (1935). Blossom-bud differentiation in citrus trees. Am. J. Bot. 22, 476-85.
Ayalon, S., and Monselise, S. P. (1960). Flower bud induction and differentiation in the Shamouti
orange. Pvoc. Am. Soc. Hort. Sci. 75, 216-21.
Flower Initiation in Sweet Orange
Heslop-Harrison, J. (1961). The experimental control of sexuality and inflorescence structure in
Zea mays L. Proc. Linn. Soc. Lond., 172 Session, 1959-60, Pt. 1, pp. 108-23.
Hield, H. Z., Coggins, C. W., and Lewis, L. N. (1966). Temperature influence on flowering of
grapefruit seedlings. Proc. Am. Soc. Hort. Sci. 89, 175-81.
King, R. W., and Evans, L. T. (1969). Timing of evocation and development of flowers in Pharbitis
nil. Aust. J. Biol. Sci. 22, 559-72.
Lenz, F. (1966). Flower and fruit development in 'Valencia Late' orange as affected by type of
inflorescence and nutritional status. Hort. Res. 6 , 65-78.
Lenz, F. (1969). Effect of daylength and temperature on the vegetative and reproductive growth of
'Washington Navel' orange. Proc. Int. Citrus Symp. Univ. Calif., Riverside, March 1968,
Vol. 1, pp. 333-8.
Liebig, G. F., and Chapman, H. D. (1963). The effect of variable root temperatures on the behaviour
of young Navel Orange trees in a green house. Proc. Am. Soc. Hort. Sci. 82, 204-9.
Moss. G. I. (1969). Influence of temperature and photoperiod on flower induction and inflorescence
development in sweet orange (Citrus sinensis L. Osbeck). J. Hort. Sci. 44, 311-20.
Moss, G. I. (1970). Chemical control of flower development in sweet orange (Citrus sinensis).
Aust. J. Agric. Res. 21, 233-42.
Moss, G. I. (1971). Effect of fruit on flowering in relation to biennial bearing in sweet orange
(Citrus sinensis). J. Hort. Sci. 46, 177-84.
Moss, G. I. (1973). Regrowth and flowering in sweet orange after pruning. Aust. J. Agric. Res.
24, 101-9.
Sauer, M. R. (1951). Growth of orange shoots. Aust. J. Agric. Res. 2, 105-17.
Manuscript received 15 May 1975