Physiology and Behavior 66: 731-735, 1999
ENERGY CONSUMPTION OF
TERMITE COLONIES
OF Nasutitermes ephratae
(ISOPTERA: TERMITIDAE)
Roldan Muradian, Solange Issa and Klaus Jaffe'
Dpto.
Biología de Organismos
Universidad Simón Bolívar
Apartado 89000,
Caracas 1080, Venezuela
Key words: Colony energetics, respiration,
termites, castes
Abstract:
MURADIAN,
R., S. ISSA AND K. JAFFE. Energy consumption
of termite colonies of Nasutitermes ephratae (Isoptera:Termitidae). PHYSIOL. BEHAV. xxx:
xxx-xxx, 1998. Measurements of CO2 production of whole termite
colonies showed that respiration intensity varies slightly during the daily cycle, that the relation between standard metabolism and
colony size is best described with a negative exponential, taking into account
the proportion of the various castes. Larger colonies were more efficient in
their energy use, suggesting that energetic considerations may contribute in
explaining the maintenance in evolution of complex societies.
1 to whom correspondence should be
addressed
Introduction
The
respiratory rate reflects the total sum of all energetic costs of the metabolic
processes involved for the maintenance and activity of the organism and may
also serve to estimate the effort of the organism in covering the demands
imposed by the environment (9). Very little is known about the demands on
energy consumption imposed by social life. Studies with ant colonies (6) have
shown that energy requirements may serve to characterize key aspects of social life.
Theoretically,
three outcomes may be expected when comparing per capita energy consumption
(EC) of isolated organisms with those forming a society (6):
a) EC is independent to the type of
social life : null hypothesis.
b) EC decreases with socialization as
social life optimize the utilization of
resources : optimization
hypothesis.
c) EC increases with socialization as
social life implies increased complexity which in turn demands increased energy
consumption: thermodynamic hypothesis
The
hypothesis commonly assumed as true by default is hypothesis a. In the case of
termites, the scant data on whole colony respiration available (15 for
example), does not allow testing of these hypothesis. Despite this lack of
data, hypothesis a has been assumed implicitly as valid for termites (8,14, 15 for example). In addition, the thermodynamics
regulating the formation of social units may differ to that controlling the
growth of established societies. These may depend on the proportion of
organisms participating in social activities, on the existence of castes and on
the specific social structure. In the case of ant colonies, the transition from
isolated individuals to ever larger groups of ants is analogous to a phase
transition in terms of EC and seems to fulfill the thermodynamic hypothesis.
Experimental data showed that once a critical number of individuals has been achieved, EC decreases exponentially, fulfilling
the optimization hypothesis (3). The mechanisms causing this optimization of EC
with colony size are not clear but have been related to increasing proportions
of inactive ants (3) and to a greater synchronization of their activity (10).
In
order to test for these hypotheses in termite colonies, interfering variables
have to be controlled, the most important of which are temperature, intake of
nutrients and caste composition of the colony. The respiration
of isolated workers are known to be very sensitive to temperature (8)
and caste composition may affect EC. Hebling-Beraldo
and Mendes (4) showed with ants that EC does vary with castes and that this
variation can not be explained solely by their differences in size. Tschinkel (12) also found EC differences among castes in Solenopsis invicta,
where large major workers had the highest EC.
Here
we aim to measure EC in fully functional termite colonies of different sizes in
order to test the hypotheses presented above, taking into account variations in
EC along the daily activity cycle and the proportion of the various castes
forming the colony.
Materials
and methods
Termite
colonies of Nasutitermes ephratae (35
nest) were collected in FONAIAP - Miranda, Hacienda Padrón,
The
C02 production was measured using an infrared spectrometer (Li-Cor) with constant air flux. Twenty-four hours before
measurements, termite nests were placed into a hermetically sealed respiration
chamber 60 cm wide, 100 cm long and 60 cm in height, made of aluminum and
Plexiglas. Termites were offered food and water, a fixed photoperiod of 12
hours, and temperature fluctuated between 26 and 31 0C. The air
flowing through the chamber was pumped from the outside of the building. Mean
C02 concentration of fresh air was measured independently and
assessed at 353,6 ppm. The respirometer was calibrated before each measurement. Larger
nests were submitted to higher air flows so that absolute CO2
concentrations in the chamber remained below 370 ppm, as
ambient oxygen concentrations are known
to affect respiration rate in termites (13). Thus, air flux was fixed for each
experiment but varied depending on colony size between 30.5 and 131 ml/s.
The
measurements were performed for each nest on two different but consecutive
days. Each measurement took 24 h and consisted of continuously monitoring
atmospheric pressure, air temperature, air flux and CO2
concentrations of the air coming in and out of the respiration chamber. With
the help of the differential infrared detector and a special computer software
(Datacan V) the difference between the CO2
of fresh air (going into the chamber) and that exiting the respiration chamber
was calculated, taking into account the atmospheric pressure, air flux and
temperature. The air current leading to the spectrometer was dried with barium
hydroxide.
At
the end of the last C02 measurement the nests were cut into small
pieces and vigorously shaken to extract the termites. Termites remaining in
unbroken pieces of nest were extracted manually. The termites with some
remaining small pieces of nest material, were kept in
a freezer at -5 oC. This material was then
mixed uniformly and we took six samples (8 ml) from each nest. The material and
the samples were dried in an oven at 60 oC.
Each sample was carefully examined and the various castes counted in detail
before drying them. The castes were grouped as follows: worker nymphs having a
cephalic capsule of less than 0.1 cm in length, soldier nymphs having a
cephalic capsule of 0.07 cm, and adult sexuals,
workers and soldiers, which were recognized by their characteristic morphology.
The dry material was weighed and the biomass of each caste in each nest was
then calculated by extrapolating the data for the proportions of the weight of
each caste in the sample, to the total weight of each sample, to the total
weight of the nest. Data of 17 nests were used as we discarded nests containing
reptile eggs. Controls (measuring the noise in CO2 -production
measurements) consisted in pieces of nest structures representing about 80 % of
each nest, from 10 of the nests used above, after the termites had been
extracted, placed with food and water in the respiration chamber, and submitted
to the same experimental measurement for CO2 production as described
above.
Data are presented as the normalized
respiration intensity (NR) expressed as ml of CO2 production per
gram of dry biomass. Values of NR used were the average of the values for a
whole day (NR mean), the maximum values for a daily cycle (NR max), the minimum
values for a daily cycle (NR min) and the difference between NR min and NR max
(NR dif).
Results
The
controls showed that on average, empty termite nests with wood and water
produced 7.67 x 10-4 ± 1.71 x 10-4 ml of CO2/h.
If these values are compared to the average respiration of live termite nests
(270.8 ± 212.7 ml of CO2/h) we may assume that nest respiration is
accomplished mainly by the termites and we may confidently dismiss the
contribution of other nest components to the respiration values measured.
Comparing values of
temperature and CO2 production of the whole termite colonies, showed
no significant correlation between both (r = 0.11, p = 0.52), in the
temperature ranges used in the experiments.
Figure 1 shows the daily
variation in NR for the 17 nests studied. Termites showed a maximal respiration
activity at 22 h and a minimum at 11 h, showing that maximal activity was at
night.
Figure 2 shows the data for NR
(maximum, minimum and average) for each nest during the daily cycle. Each data
point is the mean of two measurements of two different days. Results show an
inverse exponential relation between NR and colony size.
The difference between maximal
and minimal NR of the colony in the daily cycle (NR dif) is a gross indicator
of differences in whole colony activity between maximum and minimum activity
levels during the daily cycle. This difference correlated negatively with
colony size (r = -0.757, p = 0.001), suggesting that larger colonies have
smaller NR dif (Figure 3).
Table
1 gives the values of "r' and "p" for the best fit to the
different models (Quasinewton method of Statistica). Model 1 assumes a constant relationship
between NR and the colonies biomass. Model 2 assumes that NR diminishes
linearly with colony size, Models 3-6 assume an inverse exponential relation
between NR and colony size, where models 4,5 and 6
decompose the total biomass into the different components of the colony. In
model 4 the NR is the daily average, in model 5 the NR is the minimum of the
daily cycle, and in model 6 the NR is the maximum of the daily cycle. The best
fit was obtained with models 4-6 and the worst with model 1, suggesting that NR
correlates inversely with colony size and that the various castes contribute
differently to NR.
Discussion
The
results clearly suggest that hypothesis “b” holds for termites. That is, energy
consumption per capita in termite colonies decreases with increasing colony
size.
Other
results give us additional information about the EC - colony size relationship
in termite colonies. The lack of correlation between CO2 production
of termite colonies and temperature suggests that termite colonies regulate
their temperature and are thus unaffected by small changes of environmental
temperature. As can be appreciated in Figure 1, the circadian cycle of NR has a
sinusoidal form, with a minimum respiration during the day and a maximum at
night. Most Nasutitermitinae are nocturnal
foragers. As termites had wood and
sufficient space in the respiration chamber to forage, it is probable that these differences
in NR are due, at least partially, to differences in foraging activity.
NR measures the total sum of
all activities which require oxygen. These activities include basal metabolism,
locomotion, growth, food processing, reproduction, etc. In the case of social
species, it has been assumed that the energetic requirements of the society are
just the sum of the basic requirements of its individuals. Our results do not
support this view; NR values and NR differences between maximum and minimum
values of the daily cycle decrease with colony size. We do not know, however,
what the activities requiring energy are during minimum and maximum activity
levels. NR min does
not represent the basal energy requirements of the colony, but we may safely
assume that the difference between NR min and NR max includes the foraging
activity of the colony, and that the daily NR variations are less than 8.5 % of
the total NR.
It
is interesting to note that NR decreased with colony size independently of the
type of measurement (i.e. NR dif, NR mean, NR min and NR max). Different
explanatory hypothesis can be proposed:
a) All colonies need to performed roughly the same amount of work and thus, in
larger colonies, this work is proportionally lower than in small ones. That is,
larger colonies might have larger reserves of inactive workers, or individual
workers work less than in smaller colonies.
b) The degree of individual efficiency
increases with colony size. For example, specialization, and thus efficiency,
of temporal and morphological castes increases with colony size.
c) Miramontes
and DeSouza (7) showed that the survival of
individual termites increased with increasing number of workers forming
experimental aggregates. By comparing their results with computer simulations,
they suggested that this phenomena was related to
social interactions, which become more efficient in larger groups. A similar
explanation may be applied to our results.
d) Larger nests have a larger
proportion of termites in areas of the nest with low air ventilation, which
will reduce their respiration rate (13). This may be an indirect way for large
colonies to reduce their per capita energy consumption.
Clearly, detailed behavioral
studies are needed to understand the phenomena uncovered here and to determine
which of the putative mechanisms proposed above are used by termites
colonies. Thermodynamic studies only allow to detect
basic phenomena, but are useless in explaining causal relationships. However,
results in Table 1 suggest that the contribution of the various castes to the
NR is different as has been reported previously (4, 5, 12).
Thus, not only colony size is important for explaining NR but also the specific
caste composition will affect NR.
The
inverse relationship found between colony size and per capita respiration rate, i.e. curves in
Table 1, which are derived from data for colony respiration, cannot be
extrapolated to individual termites or small groups of termites. For example,
the 100 workers mentioned above, measured by Damaschke
and Becker (2), weight
about 0.03 gr. We may thus calculate, using formula 4 of Table 1, that NR mean
= 29 ml/g(dw)/h, which if 16
times higher than what was reported by these authors (8). An inverse
exponential relationship between NR and the size of the social system, starting
from specific minimal sizes of social system, has been reported previously for
ants (3,6) and for the relation between energy consumption and number of
inhabitants among human settlements (1 and included references). The fact that
a similar relationship applies to termite colonies, as shown here, suggest that
this pattern may be a general characteristic of social systems. Fonck and Jaffe (3) found that in very small groupings of
ants the inverse exponential relationship does not appear, but only after a
threshold ant group size, which is specific for each species, does the inverse
exponential relation becomes evident. A similar phenomenon was evidenced by
studying total energy consumption of human settlements (1). In this study, we
considered that groupings of a few isolated termites were too artificial to
merit experimental study, as termites die very fast when isolated, but similar
patterns may also be present in termites.
It
seems trivial to suppose that social life has adaptive advantages, the more so
when resources are limiting, as they seem to be more efficient in exploiting
them. At the same time social behavior may decreases parasitism and predation
risks. In addition, life expectancy increases with diminishing standard
metabolism. All these factors increase individual fitness with increasing
colony size. Here we showed that energy efficiency is also achieved with
increased colony size. Although, energy efficiency is not necessarily achieved
with social life (6), once social life exists, larger societies seem to be
better in using energy than smaller ones. The strong effect that colony size
has on NR, suggests that by pure energetic considerations, large societies
should be favored in biological evolution. There must be a limit to colony
size, as we were unable to locate in the field much larger nests than those
studied, but this limit may or may not be determined by energetic
considerations.
In
summary, our results seem to be consistent with previous data reported in the
literature, and factors such as temperature, daily rhythms or caste composition
of termite colonies do not explain by
their own the results obtained. It seems thus safe to assume that termite
colonies optimize their use of energy as colonies get larger. We propose that
this optimization is a general principle of social systems as it seems to apply
also in ant societies (3,6) and in human
settlements(1).
References
1- Cabrera,
S.; Jaffe, K. 1998. On the energetic cost of human
societies: energy consumption as an econometric index. Interciencia
23: 350-354.
2- Damaschke,
K.; Becker, G. 1964. Korrelation der Atmungsintensitát
von Termiten zu Anderung der Impulsfolgefrequenz
der Athmosphare. Z. Naturforschung 19b, 157-160.
3- Fonck, C.;
Jaffe, K. 1996. On the Energetic cost of Society.
Physiology and Behavior 59: 713-719.
4- Hebling-Beraldo,
M.; Mendes, E. 1981. The respiratory metabolism of the castes
of two leaf cutting ants, Atta laevigata and A. sexdens.
Comp. Biochem. Physiol.
684: 241-247.
5- Jaffe, K.; Hebling-Beraido,
M.J. 1993. Oxygen consumption and the evolution of order: Negentropy
criteria applied to the evolution of ants. Experientia
49, 587-592.
6- Jaffe, K.; Fonck,
C. 1994. Energetics of Social Phenomena: Physics
Applied to Evolutionary Biology. Il Nuovo Cimento 16: 543-553.
7- Miramontes,
0.; DeSouza, O. 1996, The
Nonlinear Dynamics of Survival and Social Facilitation in Termites. J. Theor. Biol. 181 : 373-380.
8- Peakin,
G.J.; Josens, G. 1978. Respiration and energy flow.
In: Production ecology of ants and termites, M.V. Brian ed. Cambridge Univ.
Press,
9- Peters, R. 1983. The
Ecological Implications of Body Size. Cambridge
University Press, xii + 329.
10- Sole, R.V. ;
Miramontes, O.; Goodwin, B.C. 1993. Emergent behavior
in insect societies: Global oscillations, chaos and computation. In: H. Haken and A.
Mikhailov eds. Interdisciplinary Approaches to
Nonlinear Complex Systems, Springer Verlag, Berlin,
p. 77-88.
11- Thorne, B.L. 1985. Numerical and biomass caste proportions in colonies of the termite Nasutitermes corniger and N. ephratae.
lns. Soc. 32: 411-426.
12- Tschinkel,
W. 1993. Sociometry and sociogenesis of colonies of the fire
ant Solenopsis invicta
during one annual cycle. Ecol. Monogr. 63:
425-457.
13- Wheeler, G.S.; Tokoro,
M.; Scheffrahn, R.H.;
14- Wiegert,
R.G. 1970. Energetics of the nest-building termite Nasutitermes
costalis (Holmgren), on a
15- Wiegert,
R.G.;
Table 1. Regression coefficients between NR and colony biomass using
different models for calculating the regression.
MODEL r p
1:
NR mean = 5.285 0.00
1.00
2:
NR mean = 8.8 - 0.07B 0.475 0.054
3:
NR mean = 2.3 + 13.9 e(- 0.04B)
0.590 0.013
4:
NR mean = 3.5 + 26.1 e(- 0.8n - 42.7s - 0. 18N + 0.8S) 0.636 0.006
5:
NR min = 3.2 + 25.7 e(- 0.8n - 43.8s - 0.18N + 0.81S) 0.641 0.006
6:
NR max = 3.7 + 26.8 e(- 0.8n - 41.4s - 0.17N + 0.78S) 0.634 0.006
NR
= Normalized respiration rate; B= Total biomass; n= Biomass of worker nymphs;
s= Biomass of soldier nymphs; N= Biomass of adult workers; S= Biomass of adult
soldiers.
Figure
1 : Daily respiration pattern of termite colonies. Values
represent means and standard errors of NR mean from 34 measurements.
Figure
2 : NR values for nests of different size. Each nests was measured twice. A biomass of 1 g of dw represents approximately 2075 soldiers or 1067 workers
with a dw/fw ratio of respectively 0.31 and 0.24.
Figure
3 : NR dif values for nests of different size. Each
data represents the mean of two measurements.
