Content uploaded by Ingolf Lamprecht
Author content
All content in this area was uploaded by Ingolf Lamprecht on Mar 19, 2016
Content may be subject to copyright.
Thermologic investigations of three species of Amorphophallus
Ingolf Lamprecht
•
Roger S. Seymour
Received: 27 March 2010 / Accepted: 14 May 2010
Ó Akade
´
miai Kiado
´
, Budapest, Hungary 2010
Abstract Thermologic investigations were carried out on
three species of Amorphophallus: A. konjac, A. paeoniifo-
lius and A. titanum, all the three strongly thermogenic.
Moreover, their breeding system is described as protogy-
nous, the heat production occurs in the appendix and male
florets, no warming is seen in the female florets and pollen
is shed after the end of heat dissipation. All the three have
large, impressive inflorescences developed from big corms
and have considerable sizes. During their inflorescence,
they have a strong scent like rotting meat with carrion
smell. Amorphophallus konjac (K. Koch) has a large,
exposed appendix that produces a disgusting scent during
the day of the female phase of blooming. The appendix
produces about 3 W for several hours, and the temperature
elevation is about 2.9 K. The low temperature elevation is
attributed to a high surface area and a high evaporative heat
loss from the appendix. During the male phase of bloom-
ing, a second episode of thermogenesis occurs during the
same time of day, apparently from the male florets,
reaching a maximum of 1.6 W. Amorphophallus pae-
oniifolius (Dennst.) Nicolson has a spadix that varies
considerably from that of A. konjac and A. titanum with an
amorphous upper end of the appendix like a shrunken red
pepper instead of cone-like appendices for the two others.
It shows thermogenic temperature increases of up to
?9.1 K in the male florets and ?2.6 K for a short time in
the appendix. Amorphophallus titanum (Becc.) Becc. ex
Arcang is the largest inflorescence of the world, growing
up to 300 cm high and 250 cm across. A much smaller
plant was observed during its thermogenic period by means
of infrared (IR) thermography, IR thermometry, and ther-
mometric data logger. The temperature maximum showed
36.6 °C at ambient 24.0 °C, which means a temperature
difference of about ?12.6 K. In the morning of the next
day, all temperatures are back to ambient at about 24 °C.
Estimates of the heat production (about 74 W) were made
from the geometric data and special assumptions with
respect to the heat transfer.
Keywords Amorphophallus Heat production rate
IR-thermometry Thermogenic plants Thermography
Introduction
De Lamarck described more than 200 years ago for the first
time that thermogenic plants of the Arum lily family (Ar-
aceae) warm up their blossoms by more than 10 degrees
above the ambient and usually distribute an unpleasant
smell, to say the least. Nowadays, it is known that ther-
mogenic plants exist among 11 families of ancient angio-
sperms and cycads [1]. They were intensively investigated
in the past and up to recently for their degree of warming,
the metabolism (including the biochemical pathways for
heat generation), the repetition and period of this phe-
nomenon, its duration, the time of the day, scent production
and the interconnection with possible pollinators. Most
Electronic supplementary material The online version of this
article (doi:10.1007/s10973-010-0891-9) contains supplementary
material, which is available to authorized users.
I. Lamprecht (&)
Institute for Zoology, Free University of Berlin, Ko
¨
nigin-Luise-
Straße 1-3, 14195 Berlin, Germany
e-mail: ingolf.lamprecht@t-online.de
R. S. Seymour
Ecology and Evolutionary Biology, University of Adelaide,
Adelaide, SA 5005, Australia
e-mail: roger.seymour@adelaide.edu.au
123
J Therm Anal Calorim
DOI 10.1007/s10973-010-0891-9
experiments were performed not only in the family of
Araceae, but also on the water lily Victoria and the sacred
lotus Nelumbo nucifera. Many specific and astonishing
results were detected, among them their ability to regulate
the temperature within a small range and over hours or
days.
The most astonishing members of plants in the Araceae
family are found among the Amorphophallus which com-
prises about 170 species, growing in or at the edge of
forests in the Tropics, from West Africa to Polynesia with
the exception of the New World [2]. They possess under-
ground tubers that are sometimes cultivated, because the
corms are rich in starch, easy to harvest and a welcome
addition to nutrition. One leaf with a vertical stalk and a
horizontally leaf blade with finger-like leaflets develops
from the tuber. As is well known from other Araceae, a
spathe together with the spadix forms the inflorescence
(ensemble of many small male and female florets and
accessory structures). At the base of the spadix are the
protogynous female florets, above them the male florets,
sometimes sterile male florets and the by far larger top of
the spadix, the appendix. ‘‘Protogynous’’ means that the
female flowers are receptive at first for about 1 day, and
afterwards the male florets shed pollen, so that self-polli-
nation is excluded. Later in the year when the spathe
disappears, the female florets show mostly red berries
(W. Hetterscheid, International Aroid Society, http://www.
aroid.org/genera/amorphophallus/amintro.html).
Within the genus Amorphophallus well-known species
are A. konjac, A. paeoniifolius and A. bulbifera. But the
largest inflorescences of the world are produced by A. ti-
tanum. They originate from tubers of many tens of kilo-
grams, exceptionally more than a hundred, and are found
exclusively in Rain Forests in Sumatra. Leaves and inflo-
rescences alternate in temporal appearance, sometimes
bring a series of years with leaves only and then a
blossom appears without leaves. The leaves form ‘‘trees’’
of up to 6 m in height and endure for up to 24 months.
During this period, they produce the metabolites for a
new, even larger tuber. After some years of leaf growth,
an inflorescence appears developing up to a height of 3 m
and a diameter of about 1.5 m. Owing to their dark
brownish colour and their more than unpleasant smell like
decaying flesh or carrion, they attract pollinators, once
thought to be elephants, but now recognised as dung or
carrion beetles [3], which pollinate the protogynous
female florets before the males produce pollen. The
spectacle of inflorescence is rare and a matter for press,
television and radio: only about 75 plants have been
observed to flower since the Titan arum was detected in
1878 by the Italian botanist Odoardo Beccari [4]. Flow-
ering in Sumatra has been observed by Bogner [5] and
described in detail by Barthlott and Lobin [6].
Amorphophallus konjac (K. Koch)
Amorphophallus konjac is a common food-plant cultivated
around the world (Fig. 1). Temperatures and respiration
rate were measured between 16 and 22 September, 1995,
on one specimen that was growing in the Adelaide Botanic
Gardens. The plant was potted in a container about 50 cm
in diameter and 50 cm deep, so it was moved to a glass
house on the campus of the University of Adelaide for
measurements. The source of the plant is not known.
Temperature was measured with copper–constantan
thermocouples placed inside the appendix, at the top,
middle and bottom. Respirometry was accomplished by the
decrease in oxygen level in an air stream that flowed
through a hood over the inflorescence, through a water trap
and flow meter, and into a Taylor-Servomex model OA570
paramagnetic oxygen analyser. Air from an aquarium
pump flowed through a calibrated Fischer and Porter
rotameter at about 400–500 mL min
-1
, which was recor-
ded regularly. The hood was fashioned from thick plastic
and gas entered the bottom, where the hood was partially
sealed at the base, and was sucked out of the top to the
analyser. A timer-controlled solenoid valve switched the
Fig. 1 A. konjac blooming in the botanical garden in Berlin. Please
notice the simultaneous growth of a leaf (left) with the stalk and the
dissected leaf on top with a large diameter. More A. konjac are
growing in the background with larger or developing leaves
I. Lamprecht, R. S. Seymour
123
analyser to read air for 5 min every 20 min to control for
analyser drift. Note that neither water vapour nor carbon
dioxide was absorbed prior to analysis of oxygen. Because
water vapour condensed in the trap on the way to the
analyser, the water vapour pressure was assumed to be
saturated at analyser temperature. Considering the error in
the flow meter and the assumptions of the calculations, the
error in measurement of oxygen consumption is estimated
to be less than about 10% [7]. Respiration rate was con-
verted to heat production according to the equivalence of
21 J/mL of oxygen consumed [8].
The hood was placed over the entire inflorescence on the
days before opening of the spathe. After the spathe opened,
the hood covered only the appendix and male florets within
the floral chamber. Opportunistic observations were made
during the flowering sequence, and measurements were
made of the height of the inflorescence from the base of the
spathe to the top of the appendix. About 24 h after the
thermogenic episode of the male florets, the appendix with
male florets was cut and weighed.
Measurements of the inflorescence covered the complete
sequence of flowering, from 5 days before the first ther-
mogenic episode, when the spathe was closed, through
opening of the spathe and intense thermogenesis during the
female phase, and the final episode of thermogenesis
associated with pollen release in the male phase. Between
15 and 19 September, the inflorescence grew from 63 to
88 cm in height. On the day before opening (19 Septem-
ber), the spathe was observed to be loosening significantly.
For more than 24 h before opening, temperatures of the
appendix were not greatly elevated over hood temperature,
and the level of respiration indicated that heat production
was low (Fig. 2). During the night, the inflorescence pro-
duced about 0.3–0.4 W, but this increased to about 0.7 W
during the day, when hood temperatures were higher. On
the morning of opening day (20 September), the spathe was
closed, and little scent was apparent in the outflow of the
respirometer. However, there was an explosive rise in
thermogenesis that occurred after 10:00, continued
throughout the day and started to decline precipitously at
about 17:30 (Fig. 2). This was accompanied by a horren-
dous odour that resembled a combination of a decaying
animal and vomit. During this period, the spathe had
opened, so the hood was removed at 18:45 for examination
and repositioning over the appendix and male florets only.
Respirometry was resumed at 19:50. During the subsequent
night, heat production decreased to previously low values.
However, a second thermogenic episode began on the next
day (21 September) at about the same time (10:00), and
with a similar duration. At 12:10, the hood was removed
for sampling of exudates and odour. This time, the smell
was less powerful and different, akin to an open latrine. No
pollen was evident at 12:10, but it began to appear during
the afternoon as a brown coating of the male florets that
was confirmed to be pollen under a microscope. There were
no more thermogenic episodes, and so measurements were
discontinued on 22 September. On this date, the mass of
the excised appendix and male florets was 213 g. The
appendix was 67 cm long and 4.6–8.1 cm at the widest part
of its irregularly shaped cross section. The inside of the
appendix was hollow. The specific mass of the A. konjac
inflorescence with a mean diameter of 6.4 cm amounted to
0.1 g cm
-3
, the same value as Boecker [9] determined for
A. titanum with a comparable appendix structure (see
below).
The increase in heat production during the day before
spathe opening was apparently caused by passive heating
of the inflorescence due to higher day-time ambient tem-
peratures. An increase from 0.35 to 0.70 W at ambient
temperatures of 18 and 26 °C, respectively, would require
aQ
10
of 2.4, which demonstrates the van’t Hoff/Arrhenius
effect only. Unlike some other species that show a circa-
dian lead up to the intense thermogenic episode in the
female phase [10], this does not occur in A. konjac.
The maximum heat production during the female phase
was about 3 W at 14:00 h, resulting in a 2.9 °C rise in
middle appendix temperature (Fig. 2). In the male phase,
heat production was about 1.6 W at 14:00, but the tem-
perature increase in the male florets was unfortunately not
measured, because at the time of the study, thermogenesis
by the male florets of arum lilies was not established. It is
surprising at first that, despite a 3 W rate of heat produc-
tion, the temperature elevation of the appendix was quite
small. By comparison, a 1 W heat generation in the lotus
0.00
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
32
30
28
26
24
22
20
18
16
12.00 0.00 12.00
Time of day
Heat production/WTemperature/°C
0.00 12.00 0.00
Opening/scent
Appendix
Hood
Pollen
release
Fig. 2 Thermogenic episodes of an A. konjac inflorescence of 213 g.
Top: Rate of heat production during 3 days of blooming, showing
ambient temperature induced small increases on the 1st day, major
thermogenesis by the appendix on the 2nd day and a lesser episode by
the male florets on the 3rd day. Bottom: Temperatures of the appendix
and respirometry hood. Gaps in the record represent disruptions for
calibration and manipulation (for details see text)
Thermologic investigations of three species of Amorphophallus
123
Nelumbo nucifera can raise floral temperature about 20 K
above the environment. There are at least two explanations
for this result. First, the appendix of A. konjac has a high
surface area and is directly exposed to the environment, so
that it has little insulation for retaining heat, unlike the
heat-generating receptacle of the lotus that is surrounded
by petals. The larger the surface area of a heat-generating
object, the lower the surface temperature will be. Second,
the rate of evaporation from the appendix was undoubtedly
high in A. konjac. Liquid water was observed running
down the appendix and the interior hood walls during peak
thermogenesis. In Dracunculus vulgaris, an aroid that has a
similar, but smaller exposed appendix, the maximum
appendix heat production is 1.7 W, but the evaporative
heat loss is 2.2 W, so the temperature of the appendix can
actually be lower than ambient [11]. This demonstrates that
measurements of thermogenesis by thermometry can seri-
ously underestimate actual heat production as measured by
direct or indirect calorimetry. The purpose of thermogen-
esis in the appendices of Amorphophallus appears to be
volatilisation of scented compounds that may be associated
with secretion of water. Therefore, evaporation can be
considerable, without a great elevation of appendix
temperature.
Amorphophallus paeoniifolius (Dennst.) Nicolson
Amorphophallus paeoniifolius (Dennst.) Nicolson, called
also A. campanulatus Blume or with the trivial name, the
elephant foot-yam, is found in Madagascar, South-East
Asia and Polynesia, and is cultivated in open fields or as
intercrop in coconut gardens. Its corms are globose or
depressed-globose with a deep dip in the centre looking
like a donut. Its growth structure is similar to that of A.
titanum with one leaf on a high stalk of 150 cm and with a
blade of about 300 cm. A. paeoniifolius is an Ayurvedic
medicinal plant and also good for other pharmaceuticals,
edible and cultivated in plantations from corms or 100-g
pieces of them, because it is also rich in nutrients and a
delicacy as food [12]. Flowering is seldom seen in plan-
tation, occurring, if at all, in May or June.
The plants are regularly cultivated in the Botanical
Garden of Berlin, and flower sporadically. We could follow
the development and inflorescence between the 27 June
and 10 July 2001. The corm had a diameter of 25 cm and a
mass of 4,300 g (determined half-a-year earlier); the fresh
mass of the inflorescence was estimated to 300 g from the
known dry mass of 30.2 g. The temperature profiles were
taken by means of thermosensors inserted into different
parts of the inflorescence or with IR thermometry at its
surface. Its metabolism was estimated by monitoring the
oxygen consumption under a headspace container of 10 L.
The putrid odour of the 1st day of blooming resembles the
smell of rotting meat. Dimethyl disulphide and trisulphide
were the only components found in equal amounts [13, 14].
A. paeoniifolius was closely watched in its development
during a fortnight, beginning with a status of just emerging
from the cataphyll. All temperatures were at ambient level,
fluctuating with the sunshine in the greenhouse. Six days
later, the inflorescence started its thermogenic phase in the
morning of a rainy day with ambient temperatures of
17.2 °C. The large open inflorescence (diameter about
30 cm, height 31 cm) made a strong and astonishing
impression. The apex of the spadix was formed by a dark
brownish appendix in the shape of a brain or a shrunken red
pepper. Below this was a dense yellow zone of male florets
above a clear constriction by a ring of looser female florets.
The spotted spathe had opened and surrounded both zones
like a broad curled frill bending downwards, giving access
to the male and female florets. It was strongly wrinkled and
creased. Thermogenesis reached a maximum around 9:00
a.m., decreased around noon and vanished at 16:00. Fig-
ure 3 shows the plant in the morning of the female period
when a shiny film covered the head of the plant and gave
off a distinct, but not very strong stench. Such a film is
similar to that of A. johnsonii as described by Beath [15].
The female florets were receptive, when the ring of male
florets was at peak temperature. The males topped at
26.3 °C at the surface of the ring and 26.8 °C 3 mm inside
the ring, which was 8.5 and 9.0 K above ambient,
respectively. The appendix also warmed up a bit, with
nearly no increase above ambient on the outside, but of
2.6 K inside. A slight, but passive heat up was seen in the
Fig. 3 A. paeoniifolius on the day of thermogenesis. The brain- or
shrunken-pepper-like structure on top is the appendix, below this is
the yellow ring of the male florets, below them are the female florets
and around all is the curled frill-like spathe. The cataphyll is still
attached to the spathe. The corm with the deep dip in the centre is
seen in the soil of the pot
I. Lamprecht, R. S. Seymour
123
female florets because of heat transfer from the neigh-
bouring male florets. On the next day, all temperatures
were back to ambient values, and the ripe male florets had
long protruding yellow pollen threads.
A. campanulatus Blume is supposed to be another name
for A. paeoniifolius. Skubatz and colleagues [16] showed
by infrared thermography that A. campanulatus is among
the strongly thermogenic inflorescences of the Arum fam-
ily, with highest temperatures developing in the male flo-
rets, and no warming is seen in the appendix or female
florets. However, temperatures of the appendix are con-
siderably lower if the structure has a large surface area,
such as Dracunculus vulgaris [11], but higher if it has a
small surface area, such as Arum concinnatum [17]. At
8:45 in the morning, the male florets were 2 K warmer than
air, increasing to 6 K by 11:15. Meanwhile, the appendix
increased from about 1–5 K at 12:05. By 16:00, all tem-
peratures were back at the ambient level.
The metabolism of A. paeoniifolius was followed by
indirect calorimetry. It was performed by head-space
technique with a 10-L plastic bottle without bottom, which
was placed over the whole plant and sealed to the ground.
An electrolytic oxygen sensor (FIGARO GS Oxygen
Sensor KE-Series, UNITRONIC, Du
¨
sseldorf, Germany) on
the top of the bottle monitored the decrease of the oxygen
concentration, and the mV signal was registered in a data
logger (UNIDAN
PLUS
, ESYS, Berlin, Germany). As this
volume was not thermostatted, it followed the ambient
temperature between 17 and 26 °C. In order to adjust for
these changes, the obtained values were corrected to 22 °C
with a Q
10
value of 2.0 in both directions. Subsequently,
the oxygen consumption was transformed into heat by the
value 21 J/mL oxygen for carbohydrates. In the 1st days
before opening, the plant had low heat production rates of
approximately 140 mW which increased about sevenfold
in the female phase and endured into the male phase before
they declined to the first level. The values were taken as
approximations as they were measured in the morning
between 7:30 and 9:00 because of technical reasons. It
might be that they were more pronounced at other times.
Amorphophallus titanum (Becc.) Becc. ex Arcang
Amorphophallus titanum is regularly cultivated in the
Berlin Botanical Garden. In 2006, a specimen originating
from the Botanical Garden in Bonn started to flower and
developed an inflorescence of about 57 cm high, above a
13-cm stalk. It stopped growing when the spathe was still
closed and died away. The tuber of the present plant came
from the Palm Garden in Frankfurt/Main, Germany, and as
wild material from near Padang, Indonesia. It had a mass of
11.95 kg when replanted in a larger pot on November
2008. The emergence of a shoot was seen in the first half of
February 2009, and it became clear on the 17 April that an
inflorescence was developing. In the last 10 days before the
opening of the inflorescence, a linear increase of
4.8 cm day
-1
was observed. The final height was 131 cm,
and the spathe was 84.5 cm in diameter and 265 cm in
circumference (Fig. 4). For comparison: A 32-kg tuber in
Bonn showed a daily growth rate between 7 and
19 cm day
-1
at 30 °C[18], another cultured inflorescence
at first a rate of 11.7 cm day
-1
andsloweddowntoatotal
rate of 8.5 cm day
-1
over a period of 8 days [http://www.ftg.
org./blooms/Amorphophallusalice01.html Fairchild Tropical
Garden, Gables (Miami) FL33156 USA]. Gandiwijaja and
colleagues [19] reported about an A. titanum on Sumatra
which was potted in March, and 8 weeks later, developed
a shoot. Its growth rate was 6.2 cm day
-1
3 weeks later,
and 7.3 cm day
-1
1 week later. Its final height was
164 cm. All of these data compare quite well with growth
rates of 8.0 cm day
-1
for bamboo and of 15 cm day
-1
of
the radius of a leaf of the water lily Victoria cruziana
[20].
It was the first time that A. titanum flowered successfully
in the Botanical Garden of Berlin, which was captured by a
web cam transmitted online to the internet and discussed
there (Fig. 4). It started to flower in the afternoon of 28
April 2009, when the sun still touched it and the ambient
Fig. 4 Optical photography of A. titanum at 1:00 in the night. The
cataphyll, spathe and appendix are clearly seen. The male and female
florets are hidden deep in the cone of the spathe. Dataloggers are
behind the appendix
Thermologic investigations of three species of Amorphophallus
123
temperature was at 27 °C. The timing was as described by
Barthlott and co-workers [21] with the slow opening of the
spathe in full daylight into the evening. While ambient
temperature was declining to 24 °C, the top of the spadix
reached the maximum temperature of 36.6 °C around
20:00. In the subsequent hours, the elevated temperature
spread over the whole spadix and faded away in the early
morning of the next day (Fig. 5). Around 7:45, it was back
to the temperature of the green house. The spathe had
already closed a little bit continued shrinking. The spadix
collapsed at 5:21 at the 3rd day (2 May 2009) when the
appendix turned vertically down (see videos of the
Botanical Garden, Berlin). The blossom separated after
3 weeks from the tuber. In the middle of July, it made a
new shoot; 1 month later, it had a height of 56 cm; and just
2 months later, a stalk of 170 cm developed, topped with a
dissected leaf of 280 cm diameter.
Thermographic pictures were taken with an un-cooled
‘‘D-Infrared Camera PYROVIEW 380 L compact, DIAS
Infrared GmbH, Dresden, SN C1000102’’. A sensitive 2D-
array with 384 9 288 micro-bolometer elements was used.
Technical data were k = 8–14 lm, T
M
=-20 to 500 °C;
U
B
= 10–36 V DC, 20 VA. The spatial resolution
amounted to 1.4 mrad, the temperature resolution 2% of the
signal. The evaluation of the results was carried out by a
PYROSOFT Control Software. As the borrowed camera
was not handheld, but bought for laboratory experiments
on optical benches, it had to be fixed on a massive tripod
and could not be moved during the investigation.
The different false-colours indicate the temperature of
the specimen. The camera was not calibrated when used,
but was calibrated after the fact by correlating measured
object temperatures with apparent thermographic temper-
atures according to the equation:
Object temperature ¼ 1:29 Thermographic temperature
22:3:
The blooming of A. titanum was monitored by this fixed
IR camera (Fig. 6). The picture shows the upper part of the
appendix in the centre and part of the widely opened
spathe. As the spathe played no role in the thermogenic
episode of the plant, we concentrated on the appendix. The
male florets are not visible from above, because they are
deep in the floral chamber of this species, and we were
prevented from damaging the plant. Similarly, Barthlott
and co-workers [22] did not observe heating of the male
florets. However, this was successfully done later by
Korotkova and Barthlott [21] observing the second
thermogenic period and the pollen shedding of the male
florets on the second evening.
The heating of the appendix began around 20:00, with
the hottest point initially at the top, as previously described
[22]. The thermographic values amounted to 35.3 °C in the
top, 31.7 °C in the upper middle, and 29.8 °C in the mid-
dle. Lower positions of the appendix were covered by the
spathe and not accessible with the fixed camera. Two hours
later (22:30), the values were 35.8, 32.7 and 31.5 °C,
respectively. Shortly before midnight, the values changed
to 35.6, 31.5 and 31.2 °C and, at 01:15 to 31.8, 27.4 and
25.5 °C, respectively. Later values were monitored by the
data loggers only. Thermal images of the inflorescence
sequence were taken every 15 min and showed impres-
sively the beginning of heat evolution at the top, spreading
over the full appendix until to the early morning (Fig. 7), as
well as in a time-lapse video for the time between 20:00
and 01:30 the next day (Amorphophallus_Film_2.avi). The
heat production proceeded from the top downwards over
20
25
30
35
40
15:00 18:00 21:00 0:00 3:00 6:00 9:00 12:00 15:00 18:00
Top appendix
Middle appendix
Bottom appendix
Ambient
Time of day
Temperature/°C
Fig. 5 Temperature development determined with an IR thermom-
eter of the inflorescence of A. titanum and its environment for more
than 1 day. Temperatures of the top, middle and bottom of the
appendix and the ambient surface just outside of the spathe are
recorded
Fig. 6 IR thermography picture of the appendix of A. titanum at
20:15 h. Heat production is apparently highest in the tip of the
appendix. As the IR camera was not calibrated, correct temperatures
are shown for some points. The dark brown lines are heating pipes of
the greenhouse; the lighter blue ones show the glass construction; the
bright fan-like structures are palm-leaves in the background
I. Lamprecht, R. S. Seymour
123
the entire spadix. However, no periodic phenomena [22]
could be observed with our Titan arum. This might be
because of different sizes and ages of the plants and
varying conditions in both greenhouses.
Simultaneously, four data loggers (HOBO Temp 508-
563-9000) monitored the temperatures with the sensors
placed through plastic drinking straws in two points of the
appendix, in the air within the spathe and in the ambient
air. The first logger was inserted into the middle of the
spadix, the second lower down the appendix approximately
30 cm above the male florets. The top of the spadix
remained without a logger not to injure the plant, a pre-
caution that proved superfluous later on during the exper-
iment. As the temperature at the top could not be registered
because of security reasons, it is shown as point wise
determinations (circles) together with the curves of logger
1 and logger 4 (Fig. 8). At 19:30 the sun no longer touched
the plant so that a value around 23 °C resulted. About
20:00 the curve of the middle of the appendix began to
climb to a value of 31 °C, which remained constant until
midnight and returned steadily back to the initial value.
Superimposed are the red points of the appendix top which
culminate in the maximum of 36.6 °C and maintain this
value to 01:30 before they decline to ambient values also.
The temperature within the spathe in the time from 21:30
until 7:30 the next morning was nearly constant at 24.0 °C
in agreement with Korotkova and Barthlott [21] who saw
no warming inside the floral chamber during the blooming.
The ambient temperature at the same time was 24.3 °C,
about the same as in the spathe. From 22:00 until 00:00 the
temperature in the middle appendix remained at 30.6 °C
and that in the lower appendix area at 30.3 °C between
23:00 and 02:00 the next day. They show the slow wan-
dering of temperature through the appendix. In any case,
they were ?6.6 and ?6.3 K higher than the air inside the
cone of the spathe.
Figure 5 shows the temperatures determined with the
handheld IR thermometer during the night (Non-Contact
Infrared Thermometer TH1-300, Tasco Japan Co. Ltd.). The
emission was set to 0.95 as was usually done with plants
(Jones 2004). In the beginning, the setting sun hit the
opening plant with decreasing intensity, so that all four
curves followed the ambient temperature for a short time
(Fig. 5). However, while it dropped to 22.5 °C and then
climbed up to 24.4 °C, the bottom of the appendix was
27.4 °C and rose to 31.0 °C due to the general warming of
the inflorescence. The middle appendix increased to 29.0,
31.0 and 32.5 °C, while the tip temperature showed a small
dip around 31.8 °C and a strong increase to 36.6 °C, then a
plateau about 35.7 °C(DT = 11.3 K) and a subsequent
larger decrease. All curves met in the morning at around
24.3 °C (Fig. 5). The brim of the spathe remained at a value
of 22.6 °C, one degree below ambient in the mean. Later in
the day the appendix values fluctuated a bit and lay 1–2 K
lower than the greenhouse temperature (28.9 °C), obviously
due to evaporation despite more than 70% humidity.
Not even one of the Berlin scientists had nearer contact
with A. titanum blooming before, so that the expectations
were derived from the article of Barthlott et al. [22]. The
expected carrion smell was very strong before and around
Fig. 7 Thermographically determined temperature development in A.
titanum during the early night. The records clearly show that the heat
production starts at the top, spreads downwards along the appendix
and fades away nearly homogeneously with the exception of a slightly
warmer top
22
26
30
34
38
20:00 22:00 0:00 2:00 4:00 6:00 8:00
Temperature/°C
Time of day
Middle appendix
Ambient
Tip appendix
Fig. 8 Temperature development of A. titanum during thermogene-
sis. Temperatures of the ambient air (lower continuous line) and the
middle (upper continuous line) and top of the appendix (points
determined with an IR thermometer) are represented
Thermologic investigations of three species of Amorphophallus
123
18:00 but not monitored by us, waiting for even more. No
sign of dimethyldi- and trisulfides as the single molecules
in A. titanum smell detected by Kite and coworkers [13, 14]
could be found in the belated spectra, although these
compounds were easily traceable in Dracunculus vulgaris
with our methods (unpublished results). They went unno-
ticed even by the human nose which is by far more sen-
sitive than the applied techniques.
A weak sweet smell started around 19:00 and lasted
until 03:00 the next day. It was sensed by the nose, sampled
two times (air flow of 340 mL min
-1
; sampling duration
1 h) at 21:35 and 22:35 by an active charcoal absorber and
analysed by a coupled gas-chromatographic/mass-spectro-
metric machine (Fisons GC Model 8060; Fisons MD800
quadrupol MS). Organic compounds with a sharp, pungent
smell as well as pleasant ones used for perfumes were
found. Benzaldehyde with an almond-like odour dominated
the second odour spectrum. The odors are similar to those
of the voodoo lily (Sauromatum guttatum) whose appendix
evolves a bad smell of carrion, while the club-shaped
organs in the floral chamber have a sweet odour like fruit
salad, stimulating beetles to different activities, including
mating [23]. In the same way, Arum maculatum also pro-
duces a fruit-like smell during the heating of male florets,
which occurs both before and after heating of the appendix
which produces a dung-like smell [24]. It remains specu-
lative whether this pleasant smell here serves the same
purpose of stimulating mating activities in beetles.
As no direct or indirect calorimetric measurements by
respiration were possible due to the size of the plant, a
calculation of the energy output was made from the known
geometric and thermal data. The energy balance of the
appendix consists of the energy taken up and dissipated by
the system as well as the stored energy which must
balance:
E uptake and productionðÞE dissipatedðÞ
E storedðÞ
¼ 0:
The energy taken up reduces in our case to the infrared
radiation which is proportional to the 4th power of the
absolute temperature. The dissipated energy contains this
term also, just with slightly higher temperatures, so that it
outweighs the absorbed energy. In addition, heat is lost by
evaporation, but in the case of a green house with more
than 70% saturated humidity, is reduced. Heat dissipation
is also affected by convection and conduction. The stored
energy might be taken as zero when the temperature is
stable.
Based on the book of Park S. Nobel [25] Introduction to
Biophysical Plant Physiology, one may apply the energy
flows at a flat leaf in the steady state with slight
modifications to the appendix of A. titanum. The biological
dimensions of the appendix are approximately the follow-
ing: height from the earth surface in the container to the
tip—131 cm, minus petiole below the spathe—118 cm,
minus male and female florets—105 cm; maximum diam-
eter of the appendix—15 cm; volume may be taken in a
first approximation as that of a straight cone and calculated
to be 6,200 cm
3
, although its cross section is by far not a
circle and the surface characterized by many longitudinal
dents and grooves; specific mass after Boecker [9]is
0.1 g cm
-3
, which means that the mass of the appendix
becomes 620 g.
If we consider [25] that a stationary air layer of 0.13 cm
exists around the appendix and a measured temperature
gradient of 6.7 K between the appendix and the air in the
spathe, one calculates a heat flow of 12.2 mW cm
-2
through the cone surface of 2,480 cm
2
or a total flow of
30.3 W by conduction and convection, corresponding to
48.9 mW g
-1
for a mass of 620 g. This specific heat pro-
duction nearly doubled at the tip of the appendix.
The heat exchange with the environment via radiation is
calculated as follows: There is a linear dependence
between the radiation and the 4th power of the absolute
temperature (Stefan–Boltzmann law). For the emitted
radiation, one gets
E ¼ e
IR
r TðÞ
4
with e
IR
being the emission coefficient (about 0.95;
see [26]), r the Stefan–Boltzmann constant (5.67 9
10
-8
Wm
-2
K
-4
) and T the absolute temperature. For
T = 304 K (31 °C)—for the total appendix—one gets a
radiation of 46.0 mW cm
-2
and for the maximum
T = 309 K (36 °C)—a radiation of 49.5 mW cm
-2
. Taking
a partition of 20% top and 80% rest of the appendix, i.e.
46.7 mW cm
-2
, we obtain 115.8 W of emitted radiation for
the total appendix of 2,480 cm
2
. At the same time, the
appendix takes up 104.9 W (42.3 mW cm
-2
at T = 297 K
(24 °C)). In the difference, it emits 115.8 - 104.9 =
10.9 W more energy than it receives. If one takes the above
calculated 30.3 W for the total appendix and a temperature
difference of 6.7 K as a mean, then one obtains a total heat
export of 10.9 ? 30.3 = 41.2 W, which has to be produced
by the metabolism of the plant in the stationary state.
Until now, no evaporation was taken into account which
appears without any doubt, even at or above 70% RH.
Taking Nobel’s model [25] for one side of a mesophyllic
leaf and a temperature difference of 5 K, one gets a water
flow of 0.29 lmol s
-1
cm
-2
(at 50% RH), which results in
684 lmol s
-1
for the above given values of the appendix.
With a heat of vaporization of 10.44 kcal mol
-1
at 30 °C
(Nobel 1974, App, II) corresponding to 43.7 kJ mol
-1
, one
arrives at 29.9 W at 50% RH. With our values of
DT = 7 K and 70% RH, one obtains 25.1 W responsible
I. Lamprecht, R. S. Seymour
123
for evaporation. Taking all together, the heat production for
the appendix of A. titanum sums up to 41.2 ? 25.1
= 66.3 W, that is 107 mW g
-1
for a mass of 620 g. This
value is about 20 times higher than the value of 3.0 W
which Baumann and colleagues [27] calculated from the
CO
2
metabolism and a heat transition coefficient of
2Wcm
-2
K
-1
. There are hints that their value must be
too small, e.g. the plumes of condensation which ascend in
parallel along the appendix in a video movie by Barthlott
and coworkers [22].
It was mentioned above that the appendix of A. titanum is
not smooth, but is characterized by vertical dents and
grooves. Using the cross section given by Boecker [28] and
determining its periphery, it is larger by about 12% than a
circle with the same area. Therefore, if the surface of the
cone is larger by the same factor, that means 1.12 9
2,480 cm
2
= 2,778 cm
2
. The total heat dissipation of A.
titanum thus amounts to 1.12 9 66.3 = 74.3 W, and the
specific dissipation to 120 mW g
-1
as the mass is not
influenced by the change of the circumference.
Conclusions
Three strongly thermogenic plants of the genus Amorpho-
phallus are compared in this article, namely A. konjac, A.
paeoniifolius and A. titanum. All the three have inflores-
cences belonging to the largest and most impressive plants
in the kingdom. They consist of a spathe and an appendix,
attractive for pollinators by their dark reddish to brown
colour and even more by their obnoxious carrion-like
odour. However, at the same time, they differ in both their
spathes as well as their appendices. As long as the plants
are still developing and not in the thermogenic period, the
spathes lie close to the appendix which extends a bit over
the spathe. At their lower end, they form a floral chamber
around the female florets that are visited by the pollinators.
In the beginning of the opening day when the spathes
loosen the contact to the appendix, the differences in the
spathes become obvious. While A. konjac has a slim and
asymmetric spathe with male and female florets being
visible from above, A. titanum is of similar construction but
with an approximate horizontal upper end of the spathe.
Both male and female florets are deep down in the spathe
and not visible from outside. A. paeoniifolius is compact
with an appendix like a shrunken red pepper and a spathe
which opens wide into the horizontal and even further like
a collar in the female day, exposing male as well as female
florets.
All the three show the same periods of thermogenic
metabolism: at first, the female florets are receptive (pro-
togynous) with the male florets being inactive. It is the time
of a considerable heat production in the appendix and the
dissipation of ugly smells that attract the pollinators into
the floral chamber where they unload their pollen charge to
the female florets. Later on, the metabolic ‘‘explosion’’ of
the appendix decreases, smell becomes less intensive and
the female florets loose their receptivity. Heat production
by the male florets continues, but without pungent scent
production until 1 day later at about the same time of day,
the male florets shed pollen, mainly in the form of sticky
threads. This is the time when the pollinators are released
from their compulsory ‘‘prison’’ in the floral chamber
loaded with pollen to seek for the next inflorescence in the
female stage.
Acknowledgements We are grateful to the Director of the Botan-
ical Garden of Berlin, Prof. T. Borsch, for the permission to inves-
tigate the blooming of A. titanum, and for the technical assistance to
Dr. C. Lo
¨
hne and G. Hohlstein as well as to several gardeners; to
Dr. R. Ho
¨
lzel (Fraunhofer Institute for Biomedical Engineering IBMT,
Potsdam, Germany) for lending the IR camera and giving us technical
support; and to F. Mu
¨
ller (Institute for Applied Zoology/Animal
Ecology, Free University of Berlin, Berlin, Germany) for determining
the odour spectra. This project was supported by the Australian
Research Council, and the Alexander von Humboldt Foundation.
References
1. Thien LB, Bernhardt P, Deval MSL, Chen ZD, Luo YB, Fan JH,
Yuan LC, Williams JH. Pollination biology of basal angiosperms
(ANITA grade). Am J Bot. 2009;96:166–82.
2. Mayo SJ, Bogner J, Boyce PC. The genera of Araceae. Kew,
London: R Bot Gard; 1997.
3. Bown D. Aroids: plants of the Arum family. Portland: Timber
Press; 2000.
4. Beccari O. Untitled note. Bull R Soc Toscana Ortic. 1878;3:271.
5. Bogner J. Amorphophallus titanum (Becc.) Becc. ex Arcangeli.
Aroideana. 1981;4(2):43–53.
6. Barthlott W, Lobin W, editors. Amorphophallus titanum. Stutt-
gart: Franz Steiner Verlag; 1998.
7. Withers PC. Measurement of V(dot)O
2
, V(dot)CO
2
, and evapo-
rative water loss with a flow-through mask. J Appl Physiol:
Respir, Environ Exerc Physiol. 1977;42:120–3.
8. Wieser W. Bioenergetik. Energietransformationen bei Organis-
men. Stuttgart: Georg Thieme; 1986.
9. Boecker M. Maße und Gewichtswerte. In: Barthlott W, Lobin W,
editors. Amorphophallus titanum, Chapter 5.3. Stuttgart: Franz
Steiner Verlag; 1998. pp. 27–36.
10. Seymour RS, Schultze-Motel P. Physiological temperature reg-
ulation by flowers of the sacred lotus. Philos Trans R Soc Lond B.
1998;353:935–43.
11. Seymour RS, Schultze-Motel P. Respiration, temperature regu-
lation and energetics of thermogenic inflorescences of the dragon
lily Dracunculus vulgaris (Araceae). Proc R Soc Lond B Biol Sci.
1999;266:1975–83.
12. Pandley SK. Horticulture, vegetable science, potato and tuber
crops. nsdl.niscair.res.in.
13. Kite GC, Hetterscheid WLA. Inflorescence odours of Amorpho-
phallus and Pseudodracontium (Araceae). Phytochem. 1997;46:
71–5.
14. Kite GC, Hetterscheid WLA, Lewis MJ, Boyce PC, Ollerton J,
Cocklin E, Diaz A, Simmonds MSJ. Inflorescence odours and
Thermologic investigations of three species of Amorphophallus
123
pollinators of Arum and Amorphophallus (Araceae). In: Owens
SJ, Rudall PJ editors. Reproductive biology. Kew: Royal Botanic
Gardens; 2008. pp. 295–315.
15. Beath DDN. Pollination of Amorphophallus johnsonii (Araceae)
by carrion beetles (Phaeochrous amplus) in a Ghanaian rain
forest. J Trop Ecol. 1996;12(3):409–18.
16. Skubatz H, Nelson TA, Dong AM, Meeuse BJD, Bendich AJ.
Infrared thermography of Arum lily inflorescences. Planta.
1990;182:432–6.
17. Seymour RS, Gibernau M, Pirintsos SA. Thermogenesis of three
species of Arum from Crete. Plant Cell Environ. 2009;32:
1467–76.
18. Ittenbach S, Lobin W. Die Titanwurz: (Amorphophallus titanum).
Bot Gart Univ Bonn, Inf No. 2, Bot Gart Bonn; 2000.
19. Gandiwijaja D, Idris S, Nasution R, Nyman LP, Arditti J.
Amorphophallus titanum Becc.: a historical review and some
recent observations. Ann Bot. 1983;51:269–78.
20. Lamprecht I, Schmolz E, Blanco L, Romero CM. Energy
metabolism of the thermogenic tropical water lily, Victoria cru-
ziana. Thermochim Acta. 2002;394:191–204.
21. Korotkova N, Barthlott W. On the thermogenesis of the Titan
arum (Amorphophallus titanum). Plant Signal Behav. 2009;4:1–3.
22. Barthlott W, Szarzynski J, Vlek P, Lobin W, Korotkova N. A torch
in the rain forest: thermogenesis of the Titan arum (Amorpho-
phallus titanum). Plant Biol. 2008;1–7, ISSN 1435-8603.
23. Borg-Karlson A-K, Englund FO, Unelius CR. Dimethyl oligo-
sulphides, major volatiles released from Sauromatum guttatum
and Phallus impudicus. Phytochem. 1994;35:321–3.
24. Bermadinger-Stabentheiner E, Stabentheiner A. Dynamics of
thermogenesis and structure of epidermal tissues in inflorescence
of Arum maculatum. New Phytol. 1995;131(1):41–50.
25. Nobel PS. Introduction to biophysical plant physiology. San
Francisco: W.H. Freeman and Co; 1974. 488 pp.
26. Jones HG. Application of thermal imaging and infrared sensing in
plant physiology and ecophysiology. Adv Bot Res, Inc Adv Plant
Pathol. 2004;41:107–63.
27. Baumann H, Knoche M, Noga G. Gaswechsel sowie Verteilung
von Kohlendydraten und Mineralstoffen. In: Barthlott W, Lobin
W, editors. Amorphophallus titanum. Stuttgart: Franz Steiner
Verlag; 1998. pp. 157–166.
28. Boecker M. Florale Morphologie und Anatomie (Morphology
and anatomy). In: Barthlott W, Lobin W, editors. Amorpho-
phallus titanum, Chapter 6. Stuttgart: Franz Steiner Verlag; 1998.
pp. 37–67.
I. Lamprecht, R. S. Seymour
123
