A Study of Cannonball Trees in Thailand: Hood Staminodes are Larger than Ring Stamens but only Germination of Staminal Ring Pollen can be Stimulated by Exogenous Sucrose

Abstract

The size of the staminodes and ring stamen of Couroupita guianensis (cannonball tree) grown in the Nonthaburi province, Thailand as well as some characteristics of their pollen were investigated. The staminodes were clearly larger than the ring stamens. This finding is at variance with other previous studies. Viability staining showed that almost all the cannonball tree pollen from hood staminode were not viable but about 85% of the ring stamen pollen were viable. When both types of cannonball tree pollen were cultured on modified Mercado medium, hood staminode pollen did not germinate whereas the sucrose concentrations in the medium had a promotive effect on germination of ring stamen pollen. About 65% of the ring stamen pollen germinated on the medium supplemented with 20% sucrose. These pollen studies were in agreement with other similar studies on cannonball trees, suggesting that the relative sizes of the male reproductive organs of cannonball trees may not be related to fertility of their pollen.

Full text

167
KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
A Study of Cannonball Trees in Thailand: Hood Staminodes are Larger than Ring
Stamens but only Germination of Staminal Ring Pollen can be Stimulated by Exogenous
Sucrose
Kitti Bodhipadma, Sompoch Noichinda, Kiatphaibool Permchalad, Saranya Changbandist and
Saowaros Phanomchai
Division of Agro-Industrial Technology, Faculty of Applied Science, King Mongkut’s University of Technology
North Bangkok, Bangsue, Bangkok, Thailand
Leupol Punnakanta
Faculty of Environment and Resource Studies, Mahidol University, Salaya, Nakhon Pathom, Thailand
David W.M. Leung*
School of Biological Sciences, University of Canterbury, Christchurch, New Zealand
* Corresponding author. E-mail: david.leung@canterbury.ac.nz DOI: 10.14416/j.ijast.2016.06.003
Received: 1 May 2016; Accepted: 21 June 2016; Published online: 29 June 2016
© 2016 King Mongkut’s University of Technology North Bangkok. All Rights Reserved.
Abstract
The size of the staminodes and ring stamen of Couroupita guianensis (cannonball tree) grown in the Nonthaburi
province, Thailand as well as some characteristics of their pollen were investigated. The staminodes were
clearly larger than the ring stamens. This finding is at variance with other previous studies. Viability staining
showed that almost all the cannonball tree pollen from hood staminode were not viable but about 85% of the
ring stamen pollen were viable. When both types of cannonball tree pollen were cultured on modified Mercado
medium, hood staminode pollen did not germinate whereas the sucrose concentrations in the medium had a
promotive effect on germination of ring stamen pollen. About 65% of the ring stamen pollen germinated on the
medium supplemented with 20% sucrose. These pollen studies were in agreement with other similar studies on
cannonball trees, suggesting that the relative sizes of the male reproductive organs of cannonball trees may not
be related to fertility of their pollen.
Keywords: Couroupita guianensis, Hood staminode, Lecythidaceae, Pollen germination, Pollen fertility,
Ring stamen, Sterile pollen
Please cite this article as: K. Bodhipadma, S. Noichinda, K. Permchalad, S. Changbandist, S. Phanomchai,
L. Punnakanta, and D. W. M. Leung, “A study of cannonball trees in Thailand: hood staminodes are larger
than ring stamens but only germination of staminal ring pollen can be stimulated by exogenous sucrose,”
KMUTNB Int J Appl Sci Technol, vol. 9, no. 3, pp. 167–173, July–Sept. 2016.
Research Article
1 Introduction
Cannonball tree (Couroupita guianensis Aubl.) is one
of the cauliflorous species in the Lecythidaceae family.
The name of this plant came from the fruit size and
shape that look like a cannonball. Cannonball tree is a
good indicator for healthy ecosystem. It is frequently
grown in the tropical and subtropical regions and
currently considered as threatened medicinal tree
species [1], [2]. Various parts of this plant had been
used for several therapeutic advantages. Leaf, flower
and fruit extracts showed anti-inflammatory effects,
antioxidant and anticancer properties and antimicrobial
activities, respectively [3]–[5].
Besides, cannonball tree flowers are of great
scientific interest, for example, having two types of

168
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
male reproductive organs, the hood staminodes and
ring stamen. There is no study to examine the floral
characters of the cannonball trees grown in different
countries. Therefore, it is not known if variation in their
floral characters might be present in different geographical
regions although genotype-and-environment interactions
in plants are not uncommon [6]. The objectives of
the present study were to present findings about the
characteristics of the hood staminodes and ring stamen
of the cannonball trees grown in Thailand. Also, the
shape, size, viability and in vitro germination of the
two types of pollen produced were investigated.
2 Materials and Methods
2.1 Plant materials
Blooming cannonball tree flowers (Couroupita
guianensis Aubl.) were collected from the Wat
Sanghathan, Muang diatrict, Nonthaburi province,
Thailand at about 9 A.M. in summer (April) as it
was previously reported that the peak of cannonball
tree flower anthesis presenting pollen was around this
time of the day [7]. During harvest, the staminal disc
and androecial hood (Figure 1) were gently separated
and each placed into a plastic bag to avoid mixing up
the two types of pollen therein.
2.2 Determination of form, number and length of
stamen
Parts of the staminal disc and androecial hood were
initially observed under a stereomicroscope (EMZ-TR,
Meiji Techno Co., Ltd.) and photos were taken using
a digital camera (Olympus C-760). Then, the ring
stamen and hood staminode (Figure 2) were detached
from staminal disc and androecial hood, respectively.
After placing on the glass slide, ring stamen and hood
staminode were observed under a light microscope
(ML2000, Meiji Techno Co., Ltd.) and photos were
also obtained using a digital camera. The number
of ring stamen and hood staminode per flower were
counted from randomly selected 30 cannon ball tree
flowers whereas the length of ring stamen and hood
staminode (upper and lower parts) were measured from
120 replications of each structure.
2.3 Pollen shape, size and viability
Ring stamen or staminal ring pollen and hood staminode
pollen were separately removed from the respective
structures, and placed on a glass slide for observation of
their shape and size without any staining. Subsequently,
both types of pollen were stained with 1% (w/v)
acetocamine solution before their shape, size and viability
were observed. Pollen stained red was considered to be
viable pollen while unstained pollen was a non-viable
one. Pollen observation before and after staining was
made under a light microscope and shape of pollen had
been estimated by using the P/E ratio [8]. Data from
this experiment were collected from 30 replications.
Figure 1: Position of staminal disc and androecial hood
on natural cannonball tree flower (scale bar = 1 cm).
Figure 2: Position of ring stamen and hood staminode
(upper part and lower part) on natural cannonball tree
flower (scale bar = 1 cm).

169
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
2.4 Observation of pollen germination
Germination of both types pollen was assessed by
using modified Mercado et al. medium [9] containing
0.1 mM boric acid and 1 mM CaCl2 and different
concentration of sucrose (0, 5, 10 and 20% w/v). This
medium was adjusted to pH 5.7, gelled with 0.9%
(w/v) agar and autoclaved at 121°C and 15 psi for
20 minute before use. Staminal ring pollen and hood
staminode pollen were separately brushed over the
surface of medium and were incubated at 25±2°C
for 24 hours in the dark. The percentage of pollen
germination and tube length were determined under
a light microscope and obtained from 30 replications.
2.5 Data analysis
Statistical analysis of mean differences in number and
length of ring stamen and hood staminode, pollen size
and viability, pollen germination and tube length were
performed using independent samples t-test or Tukey’s
test at P < 0.05 level.
3 Results
3.1 Form, number and length of stamen
The hood staminode of cannonball tree had longer
filaments and anthers than the ring stamen as observed
under a stereomicroscope and light microscope (Figure 3).
The ring stamen had more stamens/flower (839.9±1.37)
than the hood staminode (439.53±1.36 stamens/
flower). The lengths of the ring stamen, upper and
lower parts of the hood staminode were also found
to be statistically different (2.84±0.31, 6.15±0.63 and
8.30±0.71 mm, respectively).
3.2 Shape and size of pollen
Before any staining and observed under a light
microscope, the shape of both staminal ring and hood
staminode pollen was prolate with P/E ratio about 1.7
(Figure 4). Besides, the staminal ring pollen appeared
as monads while the hood staminode pollen remained
as tetrads (Figure 5). After staining, both types of pollen
had the same shape which was spheroidal (Figure 6).
Additionally, the stained staminal ring pollen still
appeared as monads and the stained hood staminode
pollen remained as tetrads. The diameter of the ring
stamen pollen was greater than that of the hood
staminode pollen (Table 1).
Figure 3: Male reproductive structures of cannonball
tree: ring stamens on staminal disc (A), hood staminodes
on androecial hood (B) anther on filament of ring
stamen (C) and anther on filament of hood staminode
(D). A and B: scale bar = 0.2 mm; C and D: scale bar
= 0.1 mm.
Figure 4: Morphology of staminal ring pollen (A) and
hood staminode pollen (B) of cannonball tree before
staining (scale bar = 20 μm).
Figure 5: Pollen tetrads (arrow pointed) inside the
anther from hood staminode of cannonball tree (scale
bar = 20 μm).
(A) (B)
(A) (B)
(C) (D)

170
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
3.3 Pollen viability and in vitro germination
A majority (about 85%) of the staminal ring pollen
were shown to be viable but less than 1% of the hood
staminode pollen were viable (Table 1). The effect
of different concentrations (0, 5, 10 and 20%, w/v)
of sucrose on the germination of the staminal ring
pollen and hood staminode pollen was investigated.
None of hood staminode pollen germinated on the
different sucrose-containing media (Table 2). In the
absence of added sucrose, the staminal ring pollen also
showed no germination, but with increasing sucrose
concentrations, pollen germinability and pollen
tube lengths were considerably improved (Table 2).
The highest percentage of germination (about 65%)
of staminal ring pollen was achieved on medium
supplemented with 20% sucrose (Figure 7).
Table 1: Viability and size of staminal ring and hood
staminode pollen of cannonball tree after staining
Type Viability (%) Diameter Size (µm)
Staminal ring pollen 85.14 ± 9.80a 30.67± 1.56a
Hood staminode pollen 0.90 ± 2.39b 24.00 ± 2.14b
Data are means of 30 replications ± SD. Values marked by different
letters in a column are significantly different (P < 0.05).
Table 2: Germination percentage and tube length of
staminal ring and hood staminode pollen of cannonball
tree grown on modified Mercado et al. medium [12]
supplemented with various concentrations of sucrose
Type Sucrose
(%)
Germination
(%)
Tube Length
(mm)
Staminal ring
pollen
0 0 ± 0d 0± 0d
5 33.97 ± 13.79c 0.15±0.03c
10 51.08 ± 13.66b 0.22±0.05b
20 65.18 ± 12.01a 0.40±0.10a
Hood staminode
pollen
0 0 ± 0d 0± 0d
5 0± 0d 0± 0d
10 0± 0d 0± 0d
20 0± 0d 0± 0d
Data are means of 30 replications ± SD. Values marked by different
letters in a column are significantly different (P < 0.05).
4 Discussion
Up to now, only a few angiosperm plant species are
known to have two types of male reproductive organ:
stamen and staminode in their flowers. Most zygomorphic
flowers of Lecythidaceae contain staminal disc, ligule
and androecial hood [10]. In the androecial parts, there
are numerous stamens over the ring and hood structures
but, in some plants of this family such as Lecythis
corrugata and L. pisonis, anthers had not been found on
the hood of the androecium [11], [12] suggesting that
staminodes may or may not include pollen producing
structure. For cannonball tree (Couroupita guianensis
Aubl.), the floral organ consists of ring stamens and
hood staminodes. Normally, staminodes are smaller
than the fertile stamens [13] but, in the present study
on the cannonball trees grown in Thailand, the size
of the cannonball tree staminodes was clearly larger
than the ring stamens (Figure 3). The reason for this
Figure 6: Morphology of staminal ring pollen (A) and
hood staminode pollen (B) of cannonball tree after being
stained with 1% (w/v) acetocamine solution (scale
bar = 20 μm).
Figure 7: Staminal ring pollen of cannonball tree
germinated on modified Mercado et al. medium (1994)
[12] supplemented with 0% (A), 5% (B), 10% (C) and
20% (D) sucrose (scale bar = 50 μm).
(A) (B)
(A)
(C)
(B)
(D)

171
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
variation is not known and remains to be elucidated.
This result was, however, similar to the observations
of some plants such as Penstemon, Jacaranda and
Digomphia in which their staminodes were greatly
developed, prominent and bigger than the stamens,
and appeared to play a crucial role in the pollination
ecology of those species [13], [14]. Interestingly, the
number of cannonball tree ring stamens was found to
be greater than that of the hood staminodes. It would
be interesting to determine, in future research, if this
might possibly influence the chance of pollination and
fertilization of this plant.
In the case of cannonball tree, the anthers of the
ring stamen or hood staminode produced pollen that
were indistinguishable in shape. Nevertheless, the
size of the staminal ring pollen was larger than hood
staminode pollen which unlike the former were not
viable (Table 1). Similarly, the staminode pollen of
Commelina coelestis and C. dianthifolia had
considerably lower viability than pollen of stamen
[15]. Besides, it was interesting that pollen from hood
staminode retained as groups of four pollen grains
(tetrads) whereas pollen from ring stamen was monad.
The type of tetrad formation or arrangement inside the
anther from the hood staminode of cannonball tree was
tetrahedral. Interestingly, aggregated pollen, especially
tetrad, was commonly found in primitive angiosperm
species such as Typha latifolia and Annonaceae [16],
[17]. It was also likely that the anther of hood staminode
but not ring stamen of cannonball tree may lack some
enzymes such as callase or β-1,3-glucanase to free
the pollen grains which is known to be required for
degradation of callose to release pollen grains from
tetrads in many plants [18].
Here, the present result suggested that hood
staminode pollen were likely to be not fertile. This
result corroborated the previous idea that in dimorphic
flowers with two types of stamens, one of which is the
feeding stamen producing fodder pollen for pollinators
and another one was fertilization stamen giving normal
pollen for safe gamete transport [19]. Thus, hood
staminodes of cannonball tree are of colorful appearance
and seem to be more attractive than ring stamen (Figure 2).
For staminal ring pollen, there was no germination
found on modified Mercado et al. medium [9] without
added sucrose, suggesting that without a supply of
energy, fertile pollen of cannonball tree was unable
to germinate. When the concentration of this sugar
increased, germination percentage and pollen tube
length also rose in response. Furthermore, sucrose
concentration obviously had an effect on the pollen
tube growth of cannonball tree and the maximum
length was observed on an artificial medium containing
20% sucrose. This was the same as in the study of
sucrose effect on eucalyptus and okra pollen germination
[20], [21]. Even so, the requisite optimum sucrose
concentration in the artificial medium for pollen
germination could vary among different plant species,
for example, Calotropis procera ssp. hamiltonii and
Nymphaea nouchali var. versicolor required 30% and
5% sucrose, respectively [22], [23].
5 Conclusions
Unlike previous studies, in the present study on the
cannonball trees grown in Thailand, the size of the
cannonball tree hood staminodes was found to be
clearly larger than the ring stamens. The reason or this
variation is not known. At present, the possibility that
genotype-environment interactions might be associated
with this remains to be investigated. It would be
necessary to investigate further if this is evident also
in the cannonball trees grown in other geographical
regions. Nevertheless, the pollen produced by the hood
staminodes were different from that produced by the
ring stamens as the former were sterile and could not
germinate even in the presence of promotive sucrose
concentrations. These pollen studies were in agreement
with other similar studies on cannonball tree not from
Thailand, suggesting that the relative sizes of the male
reproductive organs of cannonball trees may not be
related to fertility of their pollen. Also, the sucrose
requirement for in vitro germination of pollen seems
to be the same from trees grown in other geographical
regions. From a possible evolutionary perspective,
the tetrad hood pollen, and not the ring monad
pollen, resembled those produced by some primitive
angiosperms.
References
[1] S. K. Sharma, “Cauliflory & cannonball tree,”
Science Reporter, vol. 48, no. 6, pp. 53–55, 2011.
[2] Y. Rai, “Early seedling growth status of threatened
medicinal tree species Couroupita guianensis
Aubl. in district Meerut, (U.P.) India,” International

172
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
Journal of Innovation and Scientific Research,
vol. 8, no. 2, pp. 252–255, 2014.
[3] M. M. G. Pinheiro, S. B. O. Fernandes, C. E.
Fingolo, F. Boylan, and P. D. Fernandes, “Anti-
inflammatory activity of ethanol extract and
fractions from Couroupita guianensis Aublet
leaves,” Journal of Ethnopharmacology, vol. 146,
pp. 324–330, 2013.
[4 ] M. Premanathan, S. Radhakrishnan, K. Kulangiappar,
G. Singaravelu, V. Thirumalaiarasu, T. Sivakumar,
and K. Kathiresan, “Antioxidant & anticancer
activities of isatin (1H-indole-2,3-dione), isolated
from the flowers of Couroupita guianensis Aubl,”
Indian Journal of Medical Research, vol. 136,
pp. 822–826, 2012.
[5] N. A. Al-Dhabi, C. Balachandran, M. K. Raj, V.
Duraipandiyan, C. Muthukumar, S. Ignacimuthu,
I. A. Khan and V. S. Rajput, “Antimicrobial,
antimycobacterial, and antibiofilm properties of
Couroupita guianensis Aubl. fruit extract,” BMC
Complementary and Alternative Medicine, vol. 12,
pp. 242, 2012.
[6] D. L. Des Marais, K. M. Hernandez, and T. E.
Juenger, “Genotype-by-environment interaction
and plasticity: exploring genomic responses of
plants to abiotic environment,” Annual Review
of Ecology, Evolution, and Systematics, vol. 44,
pp. 5–29, 2013.
[7] S. Yarsick, N. X. de Enrech, N. Ramirez, and
G. Agostini, “Notes on the floral biology of
Couroupita guianensis Aubl. (Lecythidaceae),”
Annals of the Missouri Botanical Garden, vol. 73,
no. 1, pp. 99–101, 1986.
[8] M. Hesse, H. Halbritter, R. Zetter, M. Weber, R.
Buchner, A. Frosch-Radivo, and S. Ulrich, Pollen
Terminology: An Illustrated Handbook. New York:
SpringerWein, 2009, pp. 16–17.
[9] J. A. Mercado, R. Fernández-Muñoz, and M. A.
Quesada, “In vitro germination of pepper pollen in
liquid medium,” Scientia Horticulturae, vol. 57,
no. 4, pp. 273–281, 1994.
[10] L. B. V. Matta and V. V. Scudeller, “Lecythidaceae
Poit. in the Tupé Sustainable Development
Reserve, Manaus, Brazil,” Brazilian Journal of
Botany, vol. 35, no. 2, pp. 195–217, 2012.
[11] G. T. Prance and S. A. Mori, “Lecythidaceae,”
in Flowering Plants ∙ Dicotyledons (Celastrales,
Oxalidales, Rosales, Cornales, Ericales), vol. 6,
K. Kubitzki, Ed. Berlin: Springer, 2004, pp. 221–232.
[12] S. A. Mori and G. T. Prance, “The “sapucaia”
group of Lecythis (Lecythidaceae),” Brittonia,
vol. 33, no. 1, pp. 70–80, 1981.
[13] G. Dieringer and L. R. Cabrera “The interaction
between pollinator size and the bristle staminode
of Penstemon digitalis (Scrophulariaceae),”
American Journal of Botany, vol. 89, no. 6, pp. 991–
997, 2002.
[14] E. Guimarães, L. C. di Stasi, and Rde. C. Maimoni-
Rodella, “Pollination biology of Jacaranda
oxyphylla with an emphasis on staminode function,”
Annals of Botany, vol. 102, no. 5, pp. 699–711,
2008.
[15] W. C. Hrycan and A. R. Davis, “Comparative
structure and pollen production of the stamens and
pollinator-deceptive staminodes of Commelina
coelestis and C. dianthifolia (Commelinaceae),”
Annals of Botany, vol. 95, pp. 1113–1130, 2005.
[16 ] V. A. Berdnikov, O. E. Kosterin, and V. S. Bogdanova,
“Mortality of pollen grains may result from
errors of meiosis: study of pollen tetrads in Typha
latifolia L,” Heredity, vol. 89, pp. 358–362, 2002.
[17] J. Lora, M. Herrero, and J. I. Hormaza, “Microspore
development in Annona (Annonaceae): differences
between monad and tetrad pollen,” American
Journal of Botany, vol. 101, no. 9, pp. 1508–1518,
2014.
[18] H. Z. Liu, G. S. Zhang, W. W. Zhu, Q. S. BBa,
Niu N, J. W. Wang, S. C. Ma, and J. S. Wang,
“Relationship between male sterility and β-1,3-
glucanase activity and callose deposition-related
gene expression in wheat (Triticum aestivum L.)”,
Genetics and Molecular Research, vol. 14, no. 1,
pp. 574–584, 2015.
[19] Z. Luo, D. Zhang, and S. S. Renner, “Why two
kinds of stamens in buzz-pollinated flowers?
Experimental support for Darwin’s division-of-
labour hypothesis,” Functional Ecology, vol. 22,
pp. 794–800, 2008.
[20] B. M. Potts and J. B. Marsden-Smedley, “In vitro
germination of Eucalyptus pollen: response to
variation in boric acid and sucrose,” Australian
Journal of Botany, vol. 37, no. 5, pp. 429–441,
1989.
[21] M. J. Baloch, A. R. Lakho, H. Bhutto, and M. Y.
Solangi, “Impact of sucrose concentrations on
in vitro pollen germination of okra, Hibiscus

173
K. Bodhipadma et al. / KMUTNB Int J Appl Sci Technol, Vol. 9, No. 3, pp. 167–173, 2016
esculentus,” Pakistan Journal of Biological
Sciences, vol. 4, no. 4, pp. 402–403, 2001.
[22] T. Ali and S. I. Ali, “Effect of sugar concentration
on pollinium germination in some members of
Asclepiadaceae,” Pakistan Journal of Botany,
vol. 28, no. 2, pp. 161–165, 1996.
[23] K. Bodhipadma, S. Noichinda, P. Thaiyanto,
and D. W. M. Leung, “Morphology, viability,
and germinability of pollen from two forms of
Nymphaea nouchali var. versicolor, a day-blooming
waterlily,” ScienceAsia, vol. 39, pp. 214–218,
2013.