Behavioral Ecology Vol. 10 No. 3: 312-316
© 1999 International Society for Behavioral Ecology
The architecture of termite mounds: a result of a trade-off between thermoregulation and gas exchange?
Theodor Boveri Institut, Lehrstuhl für Tierökologie und Tropenbiologie, Am Hubland, D-97074 Würzburg, Germany
Address correspondence to J. Korb. E-mail: jkorb{at}biozentrum.uniwuerzburog.de
Received 27 April 1998; revised 8 August 1998; accepted 3 November 1998.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
We examined the influence of gas exchange on the architecture of termite mounds. In Comoé National Park (Côte d'Ivoire), Macrotermes bellicosus builds, as an adaptation to ambient temperature conditions, differently shaped mounds in the shrub savanna and the gallery forest. Previous studies suggested that there might be a constraint that limits the degree of thermal insulation of the interior (i.e., nest) of the mounds in environments with relatively low ambient temperatures. This factor causes, in proximate terms, suboptimal low nest temperatures and ultimately leads to reduced reproductive success in the gallery forest. In this study, we examined whether the necessity for gas exchange might constrain mound architecture. We measured CO2 concentrations in the air channels of mounds in different habitats and under manipulated temperature regimes. During both the dry and the rainy season we found higher CO2 concentrations in mounds of the gallery forest than in mounds of the savanna. Additional measurements in forest mounds, architecturally resembling those of the savanna due to an experimental increase in ambient temperatures, revealed lower CO2 concentrations than unmanipulated mounds in this habitat. Generally, concentrations were higher during the rainy season compared to the dry season and lower during day than during night. Summarizing these results we present a model that illustrates this trade-off between thermoregulation and gas exchange under different temperature regimes. Both factors together result in different mound architectures under different environmental temperatures and may finally limit the distribution of this species.
Key words: gallery forest, gas exchange, Ivory Coast, mound architecture, savanna, thermoregulation, termites.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Behavior is considered an important pacemaker in evolution (Futuyma, 1990
). Behavioral
flexibility of individuals might allow them to widen the range of habitats
that can be settled, thereby extending the geographical range of the species.
Nevertheless, there are often relatively narrow (e.g., physiological)
constraints that limit the kinds of behavioral responses needed to cope with
specific selection pressures (Wehner,
1997). The fungus-cultivating Macrotermitinae have
elaborate thermoregulatory mechanisms within the nests to yield a constant
nest temperature of 30°C and humidity near saturation year-round. These
are necessary conditions for cultivation of their fungi (Termitomyces
spp.) (Wood and Thomas, 1989).
In this study, we examined the constraints imposed on mound building in
Macrotermes bellicosus (Smeathman) in two different but adjacent
habitats, one of which represents the margin of the species' range. These
constraints seem to interfere with the ability of the termites to regulate
their nest temperature.
M. bellicosus builds different mounds under different
environmental conditions (Collins,
1979;
Lüscher,
1956). In Comoé National Park,
Côte d'Ivoire, colonies seem to adjust their
mound architecture to different, albeit adjacent, habitats. In the savanna,
mounds reach high densities and are highly structured with many ridges and
thin walls (classified according to
Grassé and
Noirot, 1961, as "nids en
cathédrale"). In contrast, in the
gallery forest mounds are more dome-shaped with hardly any protruding
structures and have thick walls (classified according to
Grassé and
Noirot, 1961, as "nids en
dôme"). These differences seem to be
caused by different environmental temperature regimes in the two habitats
(Korb and Linsenmair, 1998a).
In the cooler, but thermally more stable gallery forest, M.
bellicosus colonies insulate the interior of the mounds, where their
nests (i.e., the royal cell with the cultivated fungi and most termites) are
located. Thus, they reduce loss of heat, produced in the nest by the
metabolism of the termites and their fungi, to the exterior. In contrast, in
the thermally more suitable but more fluctuating savanna, mounds with high
surface complexities and thin walls occur. Heating experiments confirmed the
thermal significance of the different architectures: mounds in the gallery
forest have higher heat capacities, and therefore higher thermal inertia, than
mounds in the savanna (Korb and Linsenmair,
1998b).
Despite these adaptations, colonies in the forest cannot completely
compensate for the low environmental temperatures. They have suboptimal nest
temperatures (i.e., below 30°C; Korb
and Linsenmair, 1998a) and ultimately reduced reproductive success
compared to colonies in the savanna (Korb J and Linsenmair KE, in
preparation). This suggests that there might be a constraint that limits the
possibility to further insulate the nest to reduce loss of internally produced
heat to the exterior. The mound surface also functions as gas-exchange area
between the interior and the exterior of the mound (mounds have no ventilation
holes in the walls), so the necessity for sufficient gas exchange may be the
decisive constraint. The present study tested this hypothesis by investigating
the role of gas exchange in mound architecture.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Study area
This study was done in Comoé National Park (8°45' N, 3°47' W; elevation about 250 m), Côte d'Ivoire. The park is characterized by a steep climatic gradient with dry Sudan savanna in the northeast and moist Guinea savanna in the south (for more details, see FGU-Kronberg, 1979
;
Poilecot, 1991). The study site was located within the Guinea savanna and consisted of open to dense shrub savanna and gallery forests along the main rivers. The area is characterized by marked dry and rainy seasons of unpredictable lengths. Typically, the rainy season lasts from April to October, peaking in August and September, with a relatively dry period in June and July (total annual rainfall 700-1100 mm; Korb J, unpublished data).
Measuring CO2 concentrations
We registered the CO2 concentrations inside the air channels of
M. bellicosus mounds with a height of at least 2.0 m. Concentrations
in the center of the nest could not be determined because it was not possible
to obtain gas samples from this region without damaging the mounds
considerably. All measurements were taken at a height of 1 m above soil level
at all four cardinal compass directions (direction had an influence on the
concentration of CO2; Korb,
1997), once during the day and once during the night. All mounds
of a measuring series were examined in succession on the same day.
Measurements during the rainy season were done after a rain which had lasted
for at least 1 h in the early morning.
To perform each measurement, we drilled a small hole into the mound wall, placed a tube into it, and tightly sealed the hole with a piece of plastic foil. The air from the air channel was pumped at a constant flow rate of 1 ml/s to a nondifferential infrared gas analyzer (BINOS 100.1 modified for field work by adaption to 12 V and integrated installation of a filter against fine dust; Fisher Rosemount, Hanau, Germany). Additional filters were installed to remove dust. The air was also pumped through a tube with silica gel to remove water. This was necessary because water also absorbs infrared. After a lag time of 30 s, to exclude external air entering the system during the installation, we registered the CO2 concentrations until the readings were constant. The gas analyzer was calibrated with a gas mixture of known CO2 concentration (1.5% CO2 in O2) before each series of measurements and air pressure was set constant at 1024 mbar. Thus, calibration drift was < 1% of the measured value. Temperature fluctuations could cause a measurement error of at most 2% of the measured value, assuming the most extreme temperature change of 10°C during the measurements.
For each mound, CO2 values of the four cardinal directions were pooled to eliminate any directional effects. All data were analyzed with nonparametric statistics. The results were corrected with Bonferroni factor when the same data were analyzed repeatedly. We considered p values >.2 nonsignificant.
Comparison between different environmental conditions
During the dry season in 1996 we compared the CO2 concentrations
of eight cathedral-shaped mounds in the shrub savanna and five dome-shaped
mounds in the gallery forest. We also examined six mounds in the forest
habitat for which ambient temperatures had been experimentally increased since
1994, when shading trees had been cut down. These manipulated mounds showed
significantly increased surface complexities compared to unmanipulated mounds
of the forest. Thus they resemble mounds in the savanna
(Korb and Linsenmair,
1998a).
Comparison between different seasons
To investigate seasonal variation in the concentrations of CO2
in the mounds, we examined six mounds in each habitat during the dry and the
rainy season in 1997. Except for one mound, we could not remeasure the
temperature of manipulated mounds in the gallery forest because the vegetation
had grown back. Moreover, we had to choose different mounds for the two
seasons because some colonies had died in the meantime or were inaccessible
during the rainy season. Therefore we included in the analyses only different
but similar mounds, thus representing independent samples.
Furthermore, we compared the CO2 concentrations in the dry season of 1996 and 1997 to see if there were differences between years.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Comparison between differing habitats
During both day and night, the air channels of mounds in the shrub savanna had significantly lower concentrations of CO2 compared to mounds in the gallery forest (Mann-Whitney U test: day: n1 = 5, n2 = 8, U = 6.0, p =.045; night: n1 = 5, n2 = 8, U = 4.0, p =.019; Figure 1). Mounds in the forest with experimentally increased ambient temperatures, thus resembling mounds in the savanna, had lower CO2 concentrations than unmanipulated mounds in this habitat (Figure 1). These differences were significant during the night, but not during the day (Mann-Whitney U test: day: n1 = 5, n2 = 6, U = 7.0, p =.178; night: n1 = 5, n2 = 6, U = 4.0, p =.044).
|
Comparison between different seasons
In the savanna during both day and night, CO2 concentrations
were significantly higher in the rainy season than in the dry season
(Mann-Whitney U test, day: n1 = 6,
n2 = 6, U = 0.0, p =.006; night:
n1 = 6, n2 = 6, U = 0.0,
p =.006; Figure 2). In
contrast, in the gallery forest CO2 concentrations did not differ
between the seasons (Mann-Whitney U test: day: n1
= 6, n2 = 6, U = 11.0, ns; night:
n1 = 6, n2 = 6, U = 12.0, ns;
Figure 2). During both seasons,
mounds in the gallery forest had significantly higher CO2 values
than mounds in the savanna, except during the dry seasons during night
measurements (Mann-Whitney U test: dry season: day,
n1 = 6, n2 = 6, U = 3.0,
p =.045; night, n1 = 6, n2 =
6, U = 5.0, p =.123; rainy season: day,
n1 = 6, n2 = 6, U = 1.0,
p =.012; night, n1 = 6, n2 =
6, U = 2.0, p =.027). The CO2 concentrations were
significantly higher during the night than during the day, regardless of
season and habitat (Wilcoxon-test for paired samples: n = 24,
Z = -2.87, p =.012).
|
The only manipulated mound in the gallery forest had higher CO2 values during the rainy season than during the dry season and lower CO2 values during both seasons than unmanipulated mounds in the same habitat (dry season: day, 0.32%, night, 0.41%; rainy season: day, 1.12%, night, 1.34% CO2).
Comparison between 1996 and 1997
In both habitats, the CO2 concentrations did not differ during
the dry season between 1996 and 1997 (Mann-Whitney U test; savanna:
day, n1 = 6, n2 = 8, U =
12.0, ns; night, n1 = 6, n2 = 8,
U = 19.0, ns; gallery: day, n1 = 5,
n2 = 6, U = 7.0, ns; night,
n1 = 5, n2 = 6, U = 7.0,
ns).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
In this study, we examined whether gas exchange might be an antagonistic factor to thermoregulation that influences the architecture of M. bellicosus mounds and finally acts as a factor limiting the distribution of this species.
In the gallery forest, mounds are only found in relatively open and
therefore warmer stands (vegetation cover < 25%). There, the dome-shaped
mounds have thick walls and high thermal inertia, thus decreasing loss of heat
to the microclimatically suboptimal environment
(Korb and Linsenmair, 1998a).
In contrast, mounds in the warmer savanna are highly structured with thin
walls. Corresponding with mound architecture, our results revealed that mounds
in the forest had about five times higher concentrations of CO2 in
the air channels than mounds in the savanna. CO2 concentrations in
savanna mounds during the dry season were comparable to those recorded in
mounds of M. jeanneli (0.35±0.14% CO2;
Darlington et al., 1997). The
higher CO2 concentrations during the night than during the day in
all mounds seem to be the result of ventilation mechanisms that are much more
effective during the day than during the night (Korb J and Linsenmair KE, in
preparation).
A causal relationship between ambient temperature (and therefore mound
structure) and CO2 concentration seems to be confirmed by the
temperature-manipulated mounds in the forest. Termites in mounds that had been
experimentally exposed to ambient temperatures similar to those in the shrub
savanna for 2 years had responded by building more structured mounds
(Korb and Linsenmair, 1998a).
These mounds had lower CO2 values, which amount to only one-third
of those in unmanipulated mounds. (The lack of significance during the day was
most probably due to the small sample size.) These large differences cannot be
explained by differences in other CO2 sources (i.e., by differences
in the respiration of roots or other organisms), as Macrotermitinae have
fungistatic and microbiotical substances that inhibit the growth of other
foreign microorganisms (Sands,
1969; Wood and Thomas,
1989). Furthermore, it seems unlikely that this consistent pattern
was caused by respiration of soil organisms or roots which might differ
between habitats, but not between manipulated and unmanipulated mounds in the
gallery forest. In addition, uninhabited mounds from both habitats had
CO2 concentrations of 0.03% (Korb J, unpublished data). Thus they
are identical with the concentrations of the air, and respiration of roots do
not increase CO2 concentrations substantially in our study area (in
contrast to Darlington et al.,
1997). This suggests that the necessity of insulating the nest in
a cooler environment results in a decreased capacity to exchange respiratory
gases, thus leading to elevated concentrations of CO2 compared to
nests in warmer savanna habitats.
Although termites can, at least for some time, survive under high
concentrations of CO2, a colony, withup to 6 million individuals
(Lüscher,
1961) and its cultivated fungi, needs a sufficient gas exchange.
Lüscher
(1961) measured CO2
concentrations of up to 2.8% in the center of M. bellicosus nests,
and Matsumoto (1978) recorded
concentrations of up to 5.2% in the center of other Macrotermitinae nests. Our
CO2 values cannot be directly compared with these values. We
measured concentrations in the air channels (i.e., at the exchange area),
whereas the values from the Lüscher and
Matsumoto studies were taken from the nest area itself (i.e., from the places
of production of CO2). Elevated concentrations of CO2
lower the metabolism of the cultivated fungi
(McComie and Dhanarajan, 1990)
and therefore reduce both the availability of high-quality food and the
production of heat. Lüscher
(1955) estimated that a
medium-sized nest with about 500 l of air and 2 million individuals needs to
exchange at least 240 l O2 or 1200 l air per day. More recently,
Darlington et al. (1997)
calculated that a mature M. michaelseni colony, a comparable species
to M. bellicosus in Kenya, produces about 400-500 l CO2
per day.
Therefore, under suboptimal temperature regimes, there seems to be a trade-off between thermoregulation and gas exchange. Benefits from increased thermal insulation are counteracted by costs in elevated levels of CO2 and vice versa. According to our view, this trade-off proximately results in a compromise causing both suboptimally low nest temperatures and increased CO2 levels. Ultimately it reduces reproductive success compared to colonies in the savanna (Korb J and Linsenmair KE, in preparation).
Thus, according to our data, mound architecture is an adaptation to long-term ambient temperature conditions. Daily ambient temperature cycles and short-term temperature fluctuations are not considered because they must be regulated by other more flexible mechanisms. Under suboptimal temperatures, mound architecture seems to be the result of a trade-off between thermoregulation and gas exchange, as is summarized in Figure 3. In this model the surface area of a mound (y-axis) is determined by the two factors thermoregulation and gas exchange, both dependent on ambient temperature (x-axis). The curves show the mound surface areas that yield optimal and lethal nest temperatures and CO2 concentrations with increasing ambient temperature. The shapes of the curves were chosen according to the following considerations. Without regulation, an increase in ambient temperature would lead to an increase in nest temperature. This results, over a given temperature range, in a linear increasing metabolism, which further increases nest temperature. Thus, nest temperature should increase exponentially with ambient temperature due to the positive feedback of the two factors, the heating of the nest and the increased metabolism, and CO2 concentrations should increase linearly with ambient temperature due to one factor, increased metabolism. Correspondingly, termites should counteract these influences by adapting the mound surface area in the following way: (1) below the optimal ambient temperature, surface area should decline exponentially to reduce loss of heat to the environment and to yield a constant temperature, and (2) surface area should increase linearly with increasing ambient temperature to guarantee the necessary gas exchange to yield constant CO2 concentrations. The area between the curves represents surfaces under different ambient temperature regimes where the colonies survive.
|
The habitats of Comoé National Park are
incorporated in this model accordingly: in the shrub savanna, the ambient
temperatures are on a long-term basis appropriate for M. bellicosus
(Korb and Linsenmair, 1998a).
On this basis, colonies do not need to insulate their nests to reduce loss of
heat to the environment. Thus an increase in surface area should have no
thermal disadvantages, and the surface areas of the mounds should mainly be
determined by gas exchange. In contrast, in open stands of the gallery forest,
ambient temperatures are on the long-term basis below optimal nest
temperatures. Thus, termites should build compact mounds to reduce loss of
internally produced heat to the environment. Simultaneously, a sufficient gas
exchange is essential because high concentrations of CO2 reduce the
metabolism (and therefore the production of heat) of the fungi
(McComie and Dhanarajan,
1990). The surfaces are a compromise that is suboptimal concerning
both factors, as discussed above.
In dense regions of the gallery forest, ambient temperatures seem to be below a critical threshold: M. bellicosus colonies cannot survive because if they decrease mound surface area sufficiently to yield minimal heat loss, the surface area would be too low to allow the necessary gas exchange (and vice versa). Thus, the necessity for a sufficient gas exchange seems to be a constraint that limits the distribution of this termite species to open stands in the gallery forest.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
We thank Kouakou Kouadio Abdoulaye for assistance in the field and U. Grafe, J. Darlington, and one anonymous referee for comments on the manuscript. Financial support was provided by the German Academic Exchange Service (DAAD; D/931 15574), Volkswagen-Stiftung (Az I/64 102), and the University of Würzburg. The ministeries Eaux et Forêt and Recherche Scientifique de la Republique de Côte d'Ivoire gave permission to conduct research in Côte d'Ivoire (autorisation spéciale de recherche no.811; autorisation de recherche MRS/DPRF/AYK/j N°126/0).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Collins NM, 1979. The nests of Macrotermes bellicosus (Smeathman) from Mokwa, Nigeria. Insect Soc 26:240-246.
Darlington JPEC, Zimmerman PR, Greenberg J, Westberg C, Bakwin P,1997 . Production of metabolic gases by nests of the termite Macrotermes jeanneli in Kenya. J Trop Ecol 13:491-510.
FGU-Kronberg, 1979. Technische Zusammenarbeit Bundesrepublik Deutschland—Elfenbeinküste: Gegenwärtiger Status der Comoé-und Tai-Nationalparks sowie des Azagny-Reservats und Vorschläge zu deren Erhaltung und Entwicklung zur Förderung des Tourismus. Band II: Comoé-Nationalpark. Teil 1: Bestandesaufnahme der ökologischen und biologischen Verhältnisse. Kronberg: FGU-Kronberg.
Futuyma DJ, 1990. Evolutionsbiologie. Basel: Birkhäuser Verlag.
Grassé PP, Noirot C,1961 . Nouvelles recherches sur la systématique et l'éthologie des termites champignonnistes du genre Bellicositermes Emerson. Insect Soc 8:311-359.
Korb J, 1997. Lokale und regionale Verbreitung von Macrotermes bellicosus (Isoptera; Macrotermitinae): Stochastik oder Deterministik? Berlin: Wissenschaft & Technik Verlag.
Korb J, Linsenmair KE, 1998a. The effects of temperature on the architecture and distribution of Macrotermes bellicosus (Isoptera, Macrotermitinae) mounds in different habitats of a West African Guinea savanna. Insect Soc 44:51-65.
Korb J, Linsenmair KE, 1998b. Experimental heating of Macrotermes bellicosus (Isoptera; Macrotermitinae) mounds: what role does microclimate play in influencing mound architecture? Insect Soc 44:335 -342.
Lüscher M, 1955. Der Sauerstoffverbrauch bei Termiten und die Ventilation des Nestes bei Macrotermes natalensis (Haviland). Acta Trop 12:289-307.[Medline]
Lüscher M, 1956. Die Lufterneuerung im Nest der Termite Macrotermes natalensis (Haviland).Insect Soc 3:273-276.
Lüscher M, 1961. Air-conditioned termite nests. Sci Am 205:138-145.
Matsumoto T, 1978. Population density, biomass, nitrogen and carbon content, energy value and respiration rate of four species of termites in Pasoh Forest Reserve. Malay Nat J 30:335-351.
McComie LD, Dhanarajan G, 1990. Respiratory rate and energy utilization by Macrotermes carbonarius Hagen (Isoptera, Termitidae, Macrotermitinae) in Penang, Malaysia. Insect Sci Appl 11:197-204.
Poilecot P, 1991. Un Écosystème de Savane Soudanienne: Le Parc National de la Comoé (Côte d'Ivoire). Paris: UNESCO.
Sands WA, 1969. The association of termites and fungi. In: Biology of termites I (Krishna K, Weesner FM, eds). New York: Academic Press; 495-524.
Wehner R, 1997. Sensory systems and behaviour. In:Behavioural ecology: an evolutionary approach , 4th ed (Krebs JR, Davies NB, eds). Oxford: Blackwell Scientific;19 -41.
Wood TG, Thomas RJ, 1989. The mutualistic association between Macrotermitinae and Termitomyces. In: Insect-fungus interactions (Wilding N, Collins NM, Hammond PM, Webber JF, eds). New York: Academic Press; 69-92.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Korb and K. E. Linsenmair Ventilation of termite mounds: new results require a new model Behav. Ecol., September 1, 2000; 11(5): 486 - 494. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||