Development of Experimentally Orphaned Termite Worker Colonies of Two Reticulitermes Species (Isoptera: Rhinotermitidae)

Abstract

ABStrACt Survival and caste differentiation were observed under controlled con-ditions in orphaned experimental colonies of the subterranean termites Reticulitermes grassei and R. santonensis. Worker colonies had different sizes (30, 50, 100, 200 and 300); after 12 and 32 months the differentiation of colony members in other castes was observed. twelve months after orphan-ing, 80% of the colonies had survived. For the two species, a mean number of one soldier was observed in 7 colonies and between 1 and 3 nymphs were present in 18 colonies, whatever the initial number of workers. In 53% of the surviving colonies, the differentiation of secondary reproductives occurred and they produced viable offspring. The external morphology of R. grassei male reproductives did not differ significantly from those of workers or nymphs. Thirty-two months after orphaning, colonies with an initial number of 30 workers were comprised of secondary reproductives and their offspring. In termite species, caste differentiation pathways and thus the caste system are highly flexible. Therefore, our results show that a small number of subterra-nean termites could establish a new colony within few years and thus invade a new habitat, for example in urban areas.

Full text

1015
Development of Experimentally Orphaned Termite
Worker Colonies of Two Reticulitermes Species
(Isoptera: Rhinotermitidae)
by
A. Pichon1, M. Kutnik1,2, L. Leniaud1, E. Darrouzet1, N. Châline1,3, S. Dupont1
& A.-G. Bagnères1*
ABSTRACT
Survival and caste dierentiation were observed under controlled con-
ditions in orphaned experimental colonies of the subterranean termites
Reticulitermes grassei and R. santonensis. Worker colonies had dierent sizes
(30, 50, 100, 200 and 300); aer 12 and 32 months the dierentiation of
colony members in other castes was observed. Twelve months aer orphan-
ing, 80% of the colonies had survived. For the two species, a mean number
of one soldier was observed in 7 colonies and between 1 and 3 nymphs were
present in 18 colonies, whatever the initial number of workers. In 53% of the
surviving colonies, the dierentiation of secondary reproductives occurred and
they produced viable ospring. e external morphology of R. grassei male
reproductives did not dier signicantly from those of workers or nymphs.
irty-two months aer orphaning, colonies with an initial number of 30
workers were comprised of secondary reproductives and their ospring. In
termite species, caste dierentiation pathways and thus the caste system are
highly exible. erefore, our results show that a small number of subterra-
nean termites could establish a new colony within few years and thus invade
a new habitat, for example in urban areas.
Keywords: subterranean termites, small colonies, survival, neotenics dif-
ferentiation, urban habitat.
1 IRBI - UMR CNRS 6035, Université François Rabelais, Faculté des Sciences et Techniques, Parc de
Grandmont, 37200 Tours, France.
2 CTBA, Allée de Boutaut, BP 227, 33028 Bordeaux Cedex, France.
3 LEEC - UMR CNRS 7153, Université Paris 13, 99 avenue J.-B. Clément, 93430 Villetaneuse,
France.
* Corresponding author, e-mail : bagneres@univ-tours.fr

1016 Sociobiology Vol. 50, No. 3, 2007
INTRODUCTION
In eusocial insects, the colony is a social unit constituted of individuals,
which are organised in dierent castes according to their behaviour, and/
or their morphology. Colonies can be classied on the basis of the mode
of foundation, the number of reproductives and the number of nest units
forming a colony (Pamilo et al. 1997). e social organization of colonies
(i.e. number of individuals, relatedness and reproductive skew among them)
can vary within species or even within populations, and ecological and social
factors can induce variations in the social organization of colonies (Ross &
Keller 1995). In some Hymenopteran species, an example of such variation
within a population is a dierence in queen number per colony. Adoption of
young queens can be favored when habitat saturation is high (e.g. Myrmica
sulcinodis (Nylander): Pedersen & Boomsma 1999) or when ecosystems are
characterized by interspecic competition and territoriality (e.g. Ectatomma
tuberculatum (Olivier): Hora et al. 2005).
Unlike others social insects like bees, wasps or ants, Isopteran species have
a highly plastic social organization. Termites are hemimetabolous insects and,
except for the primary reproductives, all individuals are immature. ey are
organised in castes (workers, soldiers and nymphs developing into reproduc-
tives) which are an example of larval polyphenism (Nijhout 2003; Korb &
Katrantzis 2004). e developmental pathways are not always irreversible
(Noirot 1989). In some species of Termopsidae and Kalotermitidae, nest
and food resources are situated in the same piece of wood (Abe 1987) and
workers have a exible development; they can switch to another caste or even
develop regressively into a former instar, which is a unique developmental
pattern (Korb & Katrantzis 2004). In numerous termite species, dierent
social organizations occur within populations, as a result of the mode of
foundation but also as a result of the exibility of individual developmental
patterns. Colonies can thus exhibit various breeding systems: foundation
by one couple of primary reproductives (a queen and a king) or by several
queens (polygyny) or kings (polyandry) (Fisher et al. 2004). If one of them
dies (or both), immature individuals can become secondary reproductives:
they have a larval morphology but their sexual organs are functional (Noirot
1956; Büchli 1958). ese individuals are named ergatoid or nymphoid

1017
Pichon, A. et al. — Development of Orphaned Subterranean Termites
neotenics when they develop from workers or nymphs respectively. Within
a colony, they can be distinguished from workers or nymphs by a longer
abdomen, darker pigmentation, slight sclerotisation and the presence of
eyes and ocelli (Weesner 1965; Plateaux & Clément 1984; Krishna 1989;
Serment & Tourteaux 1991; orne 1996, 1998). When neotenics replace
the missing parent they are named replacement reproductives. Sometimes
neotenics can also develop in the presence of primary reproductives and are
named supplementary reproductives (orne 1996; Roisin 2000). In several
termite species, the social organization can vary within a population but also
within a colony during its lifespan, if the death of one or several primary
reproductives occurs.
Some studies have focused on the origin of replacement reproductives
and the ability of colonies to survive an orphaning. Lenz and Runko (1993)
orphaned several eld colonies of the Australian termite Coptotermes lacteus
(Froggatt): primary queens were quickly replaced by nymphoid neotenics.
Colonies produced a high number of nymphs all-year round. As the original
colonies' survival was low, it was probably due to a strategy to disperse more
alates. Pawson and Gold (1996) have investigated the caste dierentiation
in orphaned colonies of the American subterranean species Reticulitermes
avipes (Kollar), R. virginicus (Banks) and R. hageni (Banks) at 5 worker
densities. Half of the colonies survived and their growth was ensured by
the high reproduction potential of replacement reproductives. e social
structure could vary between species: R. virginicus colonies produced more
reproductives than did colonies of R. avipes and R. hageni.
In Europe, a large majority of termite species are subterranean and belong to
the genus Reticulitermes. Among them, R. grassei Clément and R. santonensis
(Feytaud) have populations in natural environments in the south-western
area of France and they cause severe damage to buildings in urban areas.
e social structure of R. grassei and R. santonensis colonies within French
populations has been analysed in natural and urban local conditions (DeHeer
et al. 2005; Dronnet et al. 2005). e presence of colonies in urban areas is
certainly due to human activity, particularly the transport of soil and infested
wood. It is generally assumed that fractions of colonies are at the origin of the
populations in built-up areas and their development can be achieved because

1018 Sociobiology Vol. 50, No. 3, 2007
of the exibility of the social structure within populations, particularly the
dierentiation of workers and nymphs into secondary reproductives.
e purpose of the present study was to evaluate the ontogenic potentialities
of small orphaned groups of workers of the species R. grassei and R. santonensis
isolated from their respective colonies. As in Pawson and Gold (1996), the
development of new reproductives and their brood was observed in groups
with dierent sizes of workers. Our rst results were obtained over a period
of 12 to 32 months. When neotenics and their ospring were observed, we
collected them and tried to assess the number of parental pairs producing
the new ospring with microsatellite markers.
MATERIAL AND METHODS
Collection of termites and rearing conditions
A colony of R. santonensis was collected in April 2000 on the island of
Oléron (Charénte Maritime, France) and maintained in the laboratory until
the experiment started. e colony of R. grassei was collected in July 2003
at Grenade-sur-l’Adour (Landes, France). In both cases the termites were
initially kept in the original pieces of wood they had been feeding on in the
eld. Species identities were conrmed using cuticular hydrocarbon pro-
les (Bagnères et al. 1991; Clément et al. 2001, data not shown) and DNA
sequences of the mitochondrial cytochrome oxydase II gene (Kutnik et al.
2004, data not shown). For the experiments, approximately 4100 workers of
5th to 7th instars were collected from each colony.
Orphaning experiment
For each species, ve colony sizes were tested: 30, 50, 100, 200 and 300
workers, respectively named type 30, 50, 100, 200 and 300. Six articial
colonies (replicates) for each type were prepared, thus 30 articial colonies
per species. For each articial colony, the termites were transferred in a plas-
tic box (120 x 90 x 50 mm) containing 150 g of moistened Fontainebleau
sand and a piece of poplar wood (20 x 20 x 20 mm). A new piece of wood
was placed in the box before half of the food resource was consumed. e
relative humidity was maintained at 80% inside each plastic box. During the
experiment, articial colonies were maintained under constant darkness, at
room temperature.

1019
Pichon, A. et al. — Development of Orphaned Subterranean Termites
Aer 12 months, we evaluated the number of surviving articial colonies
and the proportion of surviving workers (initially placed in the box and
named ‘old’ workers) within these colonies. We recorded the numbers of new
nymphs, soldiers and secondary reproductives as well as their eggs, young
larvae and new workers when present in articial colonies. All termites from
each articial colony were transferred in a new plastic box (120 x 90 x 50 mm)
with Fontainebleau sand and a piece of poplar wood, and maintained in the
conditions described previously.
Aer 24 months, almost all R. grassei articial colonies of type 200 and
300 were surviving and several secondary reproductives with a relatively
high number of eggs and young larvae were observed (cf. results). us we
chose these articial colonies to evaluate if females were sexually mature and
to estimate the number of reproductives which produced the ospring. e
other surviving articial colonies (type 30 to 100 R. grassei replicates and R.
santonensis articial colonies) were kept in the same conditions to evaluate
the development of small groups of termites 32 months aer orphaning.
erefore, the experiment was divided in two parts:
I) e surviving R. grassei articial colonies of type 30, 50, 100 and all the
surviving articial colonies of R. santonensis were maintained in the condi-
tions of the experiment. Aer eight months (32 months aer beginning),
all the termites from surviving articial colonies were collected, their caste
determined, and ospring number evaluated.
II) e development of female neotenics was observed in R. grassei sur-
viving articial colonies of type 200 and 300. e female neotenics were
isolated and their abdomens were dissected under a dissecting microscope.
eir ovarian development was observed and the spermathecae were collected
and attened in a drop of Beadle buer (128.3 mM NaCl, 4.7 mM KCl,
2.3 mM CaCl2) and xed in ethanol. e presence of sperm was observed
via uorescence microscopy aer DAPI staining for nuclei was performed
(Darrouzet et al. 2002).
To evaluate the number of secondary reproductives which had reproduced,
we genotyped females and their ospring. us, for each R. grassei articial
colony of type 200 and 300, between 5 and 40 eggs or larvae from 1st to 3rd
instar were collected with female neotenics. Ospring were frozen along with
neotenic heads at -20°C.

1020 Sociobiology Vol. 50, No. 3, 2007
Microsatellite analysis
Genotypes of R grassei neotenics and their ospring (eggs and larvae) were
examined in the surviving articial colonies: 6 of type 200, named 200-1 to
200-6 and 5 of type 300, named 300-1 to 300-5.
DNA was extracted from heads of neotenics, eggs and larvae by a Chelex®
extraction method (Walsh et al. 1991). Samples were crushed, then mixed
with 200 µl of a 5% Chelex® solution and 3 µl of a 1% proteinase K solution.
Aer 1 hour of incubation at 56°C, samples were homogenized for 10 sec
and placed at 96°C during 15 min. Aer a centrifugation at 8000 g for 3
min, 100 µl of the supernatant was removed and puried with a solution of
absolute ethanol. We examined the microsatellite genotypes for 6 loci: Rf 6-1,
Rf 5-10, Rf 21-1, Rf 11-1, Rf 11-2 and Rf 24-2 (Vargo 2000, DeHeer et al.
2005). PCR was conducted in a total volume of 10 µl, containing 3 mM or 1.5
mM MgCl2, 2 µM of reverse primer and 0.5 or 1 µM of forward primer, 0.04
units of Taq DNA polymerase and 1 µl genomic DNA. One primer of each
pair was uorescence-labelled. e PCR conditions were 40 cycles of 1min
denaturation at 94°C, annealing at 57°C for 1min and elongation at 72°C for
15 sec. en PCR products were denaturated at 94°C in a blue-bromophenol
and formamide solution and run in a 6% denaturated polyacrylamide gel using
a LiCor automated sequencer (1500 V). Microsatellite alleles were detected
through their uorescence and scored using the computer program GENE
PROFILER 4.03 (Scanalytics, Inc.).
Individuals from each articial colony were strongly related, thus the link-
age disequilibrium and the deviations from Hardy-Weinberg equilibrium
were analysed using a resampling method. A single individual per articial
colony was randomly chosen; a total of 20 data sets were created. Analyses
were performed using the soware GENEPOP v3.4 (Raymond & Rous-
set 1995). General descriptive statistics, such as observed versus expected
heterozygosity were calculated for each articial colony, using the soware
GDA (Lewis & Zaykin 2001).
To determine if one or several reproductives had reproduced we classied
articial colonies as simple or extended families. In simple families, ospring
are produced by one pair of reproductives and observed ospring genotype
frequencies did not dier from those expected under Mendelian segregation
of alleles from two parents. In extended families, genotype frequencies within

1021
Pichon, A. et al. — Development of Orphaned Subterranean Termites
colonies are not consistent with being produced by one pair of reproductives.
Signicance of the dierence was determined by a G-test (p<0.05 when o-
spring were produced by more than two reproductives) (Bulmer et al. 2001,
Goodisman & Crozier 2002, Vargo 2003, DeHeer & Vargo 2004).
Data analysis
e dierences between the proportions of surviving workers, the number
of neotenics and their ospring were tested using a Kruskall-Wallis test or a
Mann-Whitney U test within R. grassei and R. santonensis articial colonies
of type 30, 50, 100, 200 and 300. To perform these tests, we made the as-
sumption that all the females observed in the articial colonies had mated
and reproduced. us, the mean number of ospring (eggs and larvae) was
estimated per female. All the analyses were performed with the soware
STATISTICA version 6 (StatSo France, 2003).
RESULTS
Development of articial colonies aer 12 months
Survival
Twelve months aer the beginning of the experiment, the number of arti-
cial colonies with surviving termites was 27/30 for R. grassei and 23/30 for
R. santonensis. Considering only the articial colonies with surviving termites,
the mean percentage of R. grassei workers initially placed in the box, dened
as ‘old’ workers, varied from 48.5% to 76% between articial colonies (Fig. 1).
e mean percentages of ‘old’ workers in articial colonies of R. santonensis
were 17.1% to 44% (Fig. 1). For both species, the proportion of surviving ‘old’
workers was more important when the initial group size was small (Kruskall-
Wallis test, α = 0.05): p = 0.006 in R. grassei articial colonies (n = 27) and
p = 0.031 in R. santonensis articial colonies (n = 23).
Development of nymphs and soldiers
In R. grassei articial colonies, 1 or 2 nymphs were observed in 13 articial
colonies (table 1). In 5 R. santonensis articial colonies (only type 200 and
300), 1 to 3 nymphs had dierentiated (table 1). e presence of soldiers was
rst noticed between 3 or 4 weeks aer we orphaned the articial colonies.
One soldier (or white soldier) was observed aer 12 months in 4 articial
colonies of R grassei and in 3 articial colonies of R. santonensis (Table 1).

1022 Sociobiology Vol. 50, No. 3, 2007
Because of the very small numbers
of nymphs and soldiers produced in
the articial colonies, no statistical
test could be performed to evaluate
a dierence related to the sizes of
articial colonies.
Development of neotenics
and their ospring
Neotenics of R grassei and R.
santonensis were first observed
respectively 6 and 5 months aer the beginning of the experiment. In the
R. grassei articial colonies, all the secondary reproductives observed were
females. eir number ranged from 0 to 3 per articial colony and increased
with colony size (Kruskall-Wallis test, n = 27, p = 0.000) (Fig. 2). When
we analyzed the percentage of female neotenics (per 100 workers), values
obtained for R. grassei articial colonies were signicantly dierent among
types (Kruskall-Wallis test, n = 27, p= 0.001). In R. grassei articial colonies
of type 100, 200 and 300, the mean ospring per female ranged from 82 to
211 eggs and larvae (Fig. 3). e mean ospring recorded in type 300 articial
Fig. 1. Survival (mean ± SE) in R . grassei (white) and R. santonensis (grey) articial colonies 12 months
aer orphaning. Dierent letters indicate signicant dierences within species (p < 0.05).
Table 1. Number of nymphs and soldiers per articial
colony in R. grassei and R. santonensis. Number of
articial colonies with nymphs or soldiers are in
brackets.
R. grassei R. santonensis
Type nymphs soldiers nymphs soldiers
30 1, (2)
50 1-2, (2) 1, (1)
100 1, (3) 1, (1) 1, (1)
200 1, (1) 2-3, (2) 1, (2)
300 1-3, (5) 1, (2) 1-2, (3) 1, (1)

1023
Pichon, A. et al. — Development of Orphaned Subterranean Termites
Fig. 2. Mean number (±SE) of R. grassei (white) and R . santonensis (grey) female neotenics per
articial colony, 12 months aer orphaning. Dierent letters indicate signicant dierences within
species (p < 0.05).
Fig. 3. Mean ospring per female (±SE) in R. grassei (white) and R. santonensis (grey) articial
colonies 12 months aer orphaning. Dierent letters indicate signicant dierences within species
(p < 0.05)

1024 Sociobiology Vol. 50, No. 3, 2007
colonies was signicantly higher than
in type 100 and 200 articial colonies
(Mann-Whitney U test, p < 0.05).
In R. santonensis articial colo-
nies, the number of female neoten-
ics ranged from 1 to 4 per colony
and was not signicantly dierent
between types of articial colonies
(Kruskall-Wallis test, n = 23, p >
0.05) (Fig. 2). Male neotenics were
also observed in one articial colony of type 30 and in type 100, 200 and
300 articial colonies (table 2). Male and female neotenics were observed in
the same articial colonies and their respective numbers were not dierent
(Mann-Withney U test, p> 0.05). e mean ospring per R. santonensis
female was between 24 and 50 eggs and larvae (Fig. 3).
Development of R. grassei articial colonies of type 30, 50 and
100 and all R. santonensis articial colonies aer 32 months.
Survival
Aer 32 months of orphaning, the number of R. grassei articial colonies
with living termites was 5/6 in type 30, 6/6 in type 50 and 4/6 in type 100. e
articial colonies of type 30 and 50 had a similar proportion of ‘old’ workers
(i.e. original workers): 34% and 38% respectively (Fig. 4). is percentage
highly decreased in type 100 articial colonies (6%).
e number of surviving articial colonies of R. santonensis was very low.
Only 9/30 articial colonies had living termites, all from type 30 and 50. e
mean percentage of ‘old’workers within these articial colonies was 44 and
25% respectively (Fig. 4).
Development of nymphs and soldiers
e development of nymphs and soldiers occurred only in R. grassei articial
colonies (type 30, 50 and 100). One or two nymphs and one white soldier
were observed in two articial colonies of type 30. Two articial colonies of
type 100 had one and ve nymphs.
Table 2. Mean number of female and male neotenics
per articial colony in R. santonensis series 12 months
aer orphaning.
Neotenics Neotenics
Articial colony size (females) (males)
30 1.5 0.5
50 2 0
100 1.4 1.5
200 2 2.67
300 2.75 1.5
Mean ± SE 1.93 ± 0.53 1.23 ± 1.03

1025
Pichon, A. et al. — Development of Orphaned Subterranean Termites
Development of neotenics and their ospring
e R. grassei type 30, 50 and 100 articial colonies were able to support
the dierentiation of neotenics and the production of their ospring. In
three type 30 and type 50 articial colonies, one or two female neotenics
and one male neotenic per articial colony were recorded. e four type 100
termite articial colonies had a mean number of 1.25 males and 2.5 females.
e numbers of neotenics were not signicantly dierent among the types
and the number of females was not signicantly higher than the number of
males (Kruskall-Wallis test, n = 4, p > 0.05). e mean ospring (eggs and
larvae) per female was between 2 and 98 individuals, but brood size was not
related to the original size of the articial colonies (Kruskall-Wallis test, n
= 10, p>0.05).
In R. santonensis groups, 1 or 2 female neotenics were observed in 3 type
50 articial colonies and the brood production was much reduced, around
12 larvae per articial colony.
Development of female and male neotenics in R. grassei type 200
and 300 articial colonies.
Twenty-four months aer the beginning of the experiment, we tried to
observe male neotenics in the articial colonies of R. grassei with female
Fig. 4. Survival (mean ± SE) in R. grassei (white) and R. santonensis (grey) articial colonies 32
months aer orphaning.

1026 Sociobiology Vol. 50, No. 3, 2007
neotenics. Only one articial colony (type 200) contained 3 male neoten-
ics. e ospring production in the other articial colonies could occur
through female parthenogenesis, as observed in orphaned colonies of R.
speratus (Kolbe) (Hayashi et al. 2003, Hayashi et al. 2006). Male neotenics
were present but they did not have the morphological characteristics usually
observed on neotenics. Females from type 200 and 300 articial colonies
were dissected. Within type 200 articial colonies, 11 female neotenics were
observed and dissected. Among them, 7 had mature eggs in variable quanti-
ties. Due to a technical problem during dissections we were able to dissect
the spermathecae of 9 females (instead of 11) and 6 of them contained stored
sperm at the time we took them from the articial colonies. Within the type
300 articial colonies, we identied and dissected 10 female neotenics; ve
of them had mature eggs. During dissections of spermathecae, one was dam-
aged. us spermathecae of 9 females were dissected and we observed that all
stored sperm. We could conclude from these observations that in type 200
and 300 articial colonies of R. grassei, several neotenic females had mated
with neotenic males, but some females did not produce eggs at the time they
were isolated.
Genotyping
In R. grassei articial colonies of type 200 and 300, only 2 loci were found
to be polymorphic: Rf6-1 and Rf21-1, with 2 alleles each. We scored between
1 and 4 neotenics and between 5 and 40 eggs or larvae from each articial
colony. e resampled data sets showed that none of the two loci were in
linkage disequilibrium or had shown deviations from the Hardy-Weinberg
equilibrium.
In some articial colonies (for the locus Rf6-1, colony 200-5 as example),
genotypes of female neotenics were homozygous and some of their ospring
were heterozygotes. From this observation we conrmed the reproduction
of females with male neotenics.
In 83.3% of type 200 articial colonies, the ospring was produced by a
single pair of reproductives (simple families, G-test p> 0.05) (Table 3). e
proportion of simple families is 60% in type 300 articial colonies. In a ma-
jority of R. grassei articial colonies, more than one female neotenics were
observed but all females did not reproduce.

1027
Pichon, A. et al. — Development of Orphaned Subterranean Termites
DISCUSSION
From previous studies, we know that a Reticulitermes colony can comprise
from 50000 up to 1 million individuals in natural and urban habitats (Howard
et al. 1982; Paulmier et al. 1997). e presence of R. grassei and R. santonensis
in urban habitats is certainly related to anthropic factors. Only a fraction
of a colony could be introduced through the transport of wood or plants
with a lump of soil and then invade an entire area. Studies on the breeding
system of these species in urban environment showed that all R. santonensis
colonies and 51.8% of R. grassei colonies were headed by multiple secondary
reproductives (Dronnet et al. 2005; DeHeer et al. 2005).
Nests of Reticulitermes species are subterranean and foragers collect the
cellulose in log-woods situated on the ground. Foragers can adjust their food
uptake to the characteristics of the colony (numerous dependant individuals);
food quantity and quality can have an eect on the caste composition of a
colony (Lenz 1994). In our experimental design, a new piece of poplar wood
was supplied regularly. We considered that food resource, temperature and
relative humidity were not limiting factors to the development of articial
colonies, as in human housing.
From our preliminary results, we observed that under laboratory condi-
tions, a group of 30 workers can survive over a long period (24 months) and
produce secondary reproductives and a new generation of termites. e colony
Table 3. Numbers (N) of neotenics, eggs and larvae genotyped, observed (Ho) and expected (He)
heterozygosities, structure of the family obtained by G-test (p< 0.05: extended family) in each R.
grassei articial colony of type D (200 termites) and E (300 termites).
Colony N He neotenics Ho neotenics N (eggs) N (lar vae) He ospring Ho ospring Family type
200-1 2 0.250 0.250 10 0.252 0.278 simple
200-2 2 0.250 0.250 20 4 0.473 0.595 simple
200-3 4 0.428 0.500 20 1 0.198 0.164 extended
200-4 2 0.250 0.250 10 1 0.331 0.394 simple
200-5 2 0.250 0.250 5 0.278 0.300 simple
200-6 2 0.583 0.750 40 0.507 0.396 simple
300-1 2 0.500 0.500 20 0.328 0.418 extended
300-2 3 0.381 0.417 2 25 0.326 0.315 extended
300-3 3 0.333 0.333 25 0.333 0.411 simple
300-4 1 - - 12 0.475 0.633 simple
300-5 3 0.167 0.167 20 0.378 0.485 simple

1028 Sociobiology Vol. 50, No. 3, 2007
growth is relatively slow but it could be sucient to invade an urban habita-
tion within few years. Some authors consider that a group of 50 termites is
sucient to establish a colony (Serment & Tourteaux 1991).
e survival of the R. santonensis orphaned colonies was signicantly low.
is result is in contrast with observations made frequently on several eld
colonies of R. santonensis. e high mortality within articial colonies could
be due to the preservation of the colony in laboratory conditions during a
long period, and thus, to a decrease of its strength. However, in both species,
a majority of the articial colonies were active twelve months aer the start
of the experiment and several dierences were observed among dierent
sizes of articial colonies. e workers of articial colonies with small sizes
seemed to survive better aer 12 months than articial colonies with 200 or
300 termites. Considering all the plastic boxes used in the experiment had
the same volume, the density of individuals was variable and could have an
inuence on the mortality of colony members.
In their natural habitat, Reticulitermes soldiers usually represent between
1 and 3% of the colony members, although higher values have been reported
(Grace 1996; Forschler & Jenkins 1999; Long et al. 2003). In the present
study, a maximum number of one soldier per articial colony was observed,
whatever the colony size was. ey appeared within 3 or 4 weeks aer the
orphaning. e development time from a worker to a soldier is usually longer
than 4 weeks (Büchli 1958; Lainé & Wright 2003). erefore, we can suppose
that some workers had initiated their dierentiation in soldiers before the
orphaning happened. In a similar study on orphaned colonies of R. hageni,
R. virginicus and R. avipes, Pawson and Gold (1996) have recorded up to
4 soldiers per colony and their number was increasing with worker density.
However, the low number of soldiers in our experiment could be explained
by the local conditions: stable temperature and humidity, lack of predators
and/or competitors. Moreover, the dierentiation of dependant individuals
as soldiers could happen at the expense of colony growth.
A small number of nymphs were observed 12 months aer orphaning.
Nymphs are immature and dependent individuals; they can dierentiate
into adults (primary reproductives) or into nymphoid neotenics (secondary
reproductives). Miyata et al. (2004) have observed in a study on the pattern
of neotenic dierentiation in R. speratus that ergatoid dierentiation required

1029
Pichon, A. et al. — Development of Orphaned Subterranean Termites
a longer time than nymphoid dierentiation. Büchli (1958) hypothesized
that the dierentiation of Reticulitermes workers into neotenics was dicult
because workers could not accumulate enough resources to have a rapid dif-
ferentiation. In our experiment, these nymphs would probably dierentiate
into neotenics.
We observed male neotenics of R. grassei that were not morphologically
distinguishable from a worker or a nymph several months aer the orphan-
ing. us, male neotenics could be sexually mature and be able to inseminate
females before the external morphological characteristics, used generally to
identify male neotenics (darker pigmentation, longer abdomen, presence of
eyes, etc.), appear. Further studies are needed to conrm the presence of what
could be named ‘hidden’ male neotenics.
e number of female neotenics observed in the articial colonies, aer the
orphaning, was variable: from 1 up to 4 per colony. ese values were similar
to the number of reproductives produced by R. speratus orphaned colonies
(Watanabe & Noda 1991) but lower than the number of secondary reproduc-
tives observed in R. virginicus or R. avipes orphaned colonies (Pawson &
Gold 1996). More reproductives had dierentiated at high densities of workers
than at low densities. Similar results were found with orphaned colonies of
R. hageni, R. virginicus and R. avipes (Pawson & Gold 1996).
From several studies, it was assumed that numerous replacement repro-
ductives have a greater egg production than the primary queen (Miller 1969;
Nutting 1970; Myles 1999). However, several authors consider that the
queen fecundity was superior to those of neotenics (Noirot 1990; orne
1996, 1998; Lainé & Wright 2003). In our study, the female fecundity was
variable among articial colonies. Nevertheless, the values estimated were in
the same order than found by Pawson & Gold (1996).
We tried to assess the number of reproducing females in some R. grassei
articial colonies. e loci tested were not highly polymorphic. However, we
performed G-tests to determine if the brood was produced by one parental
pair or by several parents. Simple families (one pair of reproductives) occurred
in half of the articial colonies. Colonies could be headed by the female neo-
tenic who rst dierentiated during the experiment, which could dominate
the sexual development of other females. e other female neotenics could
be dierentiated but their ovocyte maturation could be delayed. Dissections

1030 Sociobiology Vol. 50, No. 3, 2007
revealed that several female neotenics did not have mature ovocytes. In other
articial colonies, the brood was produced by several reproductives but the
exact number of females reproducing could not be evaluated. Further stud-
ies with more polymorphic colonies could determine the exact number of
females producing the new generation and thus, the mean fecundity could
be evaluated more precisely. Moreover, other experiments could investigate
the mechanisms of a possible dominance of one female neotenic (whether
the rst to dierentiate or not).
As we chose to collect the rst results aer one year of the experiment, so
as not to disturb the colonies too frequently, we did not obtain results on
the egg and larval development times or on the aggression between workers
and newly dierentiated neotenics or among reproductives, as studied previ-
ously in other termite species (Lenz & Barrett 1982; Lenz & Runko 1993;
Roisin 1993, 2000).
Because of the low number of colonies used in this rst study, we could
not evaluate the variation of caste dierentiation and reproduction strategies
within and among the two species. However, termites of the genus Reticu-
litermes have exible developmental pathways which allow the persistence
of colonies in absence of primary reproductives. A single pair of secondary
reproductives in a colony with 100 workers can produce a new generation.
erefore a special attention must be paid to the human-mediated transport
of termites and to the control strategies of fractionate colonies.
ACKNOWLEDGMENTS
We are extremely grateful to D. Limousin for his help with the microsatel-
lite analysis. We thank Louise Brinkworth for the nancial support provided
by DowAgroSciences for Magdalena Kutnik’s PhD.
REFERENCES
Abe, T. 1987. Evolution of life types in termites. In: Kawano, S., J. H. Connell & T. Hidaka
[eds]. Evolution and Coadaptation in Biotic Communities. University of Tokyo Press,
Tokyo, Japan: 125-148.
Bagnères, A.-G., A. Killian, J.-L. Clément & C. Lange. 1991. Interspecic recognition among
termites of the genus Reticulitermes: evidence for a role for the cuticular hydrocarbons.
Journal of Chemical Ecology 17: 2397-2420.

1031
Pichon, A. et al. — Development of Orphaned Subterranean Termites
Büchli, H. 1958. L’origine des castes et les potentialités ontogéniques des Termites européens
du genre Reticulitermes Holmgren. Annales des Sciences Naturelles, Zoologie 11:
263-429.
Bulmer, M. S., E. S. Adams & J. F. A. Traniello. 2001. Variation in colony structure in the
subterranean termite Reticulitermes avipes. Behavioral Ecology and Sociobiology 49:
236-243.
Clément, J.-L., A.-G. Bagnères, P. Uva, L. Wilfert, A. uintana, J. Reinhard & S. Dronnet.
2001. Biosystematics of Reticulitermes termites in Europe: morphological, chemical
and molecular data. Insectes sociaux 48: 202-215.
Darrouzet, E., J. Huignard & C. Chevrier. 2002. Eect of dierential host exposure
on reproduction in the solitary ectoparasitoid Eupelmus vuilleti. Entomologia
Experimentalis et Applicata 103: 73-81.
DeHeer, C. J. & E. L. Vargo. 2004. Colony genetic organization and colony fusion in the
termite Reticulitermes avipes as revealed by foraging patterns over time and space.
Molecular Ecology 13: 431-441.
DeHeer, C. J., M. Kutnik, E. L. Vargo & A. G. Bagneres. 2005. e breeding system and
population structure of the termite Reticulitermes grassei in Southwestern France.
Heredity 95: 408-415.
Dronnet, S., M. Chapuisat, E. L. Vargo, C. Lohou & A. G. Bagneres. 2005. Genetic analysis
of the breeding system of an invasive subterranean termite, Reticulitermes santonensis,
in urban and natural habitats. Molecular Ecology 14: 1311-1320.
Feytaud, J. 1946. Le peuple des Termites. “ue sais-je?”, Presses universitaires de France,
Paris. 127 p.
Fisher, M. L., R. E. Gold, E. L. Vargo & A. I. Cognato. 2004. Behavioral and genetic analysis
of colony fusion in Reticulitermes avipes (Isoptera: Rhinotermitidae). Sociobiolog y
44: 565-576.
Forschler, B. T. & T. M. Jenkins. 1999. Evaluation of subterranean termite biolog y using
genetic, chemotaxonomic, and morphometric markers and ecological data: a testimonial
for multidisciplinary eorts. Trends in Entomology 2: 72-80.
Gene Proler 4.03 for windows: genotyping and DNA fragment analysis soware.
Scanalytics.
Goodisman, M. A. D. & R. H. Crozier. 2002. Population and colony genetic structure of
the primitive termite Mastotermes darwiniensis. Evolution 56: 70-83.
Grace, J. K. 1996. Temporal and spatial variation in caste proportions in a northern
Reticulitermes avipes colony (Isoptera: Rhinotermitidae). Sociobiology 28: 225-231.
Grassé, P. P. 1954. Origine et répartition géographique des Termites français. Annales de
l’Ecole Nationale d’Horticulture de Montpellier 24: 1-10.
Hayashi, Y., O. Kitade & J. Kojima. 2003. Parthenogenetic reproduction in neotenics of the
subterranean termite Reticulitermes speratus (Isoptera: Rhinotermitidae). Entomological
Science 6: 253-257.

1032 Sociobiology Vol. 50, No. 3, 2007
Hayashi, Y., H. Miyata & O. Kitade. 2006. Parthenogenesis by neotenic reproductives of
Reticulitermes speratus (Isoptera: Rhinotermitidae) from various regions of Japan.
Sociobiology 48: 849-860.
Hora, R. R., E. Vilela, R. Fénéron, A. Pezon, D. Fresneau & J. Delabie. 2005. Facultative
polygyny in Ectatomma tuberculatum (Formicidae, Ectatomminae). Insectes Sociaux
52: 194-200.
Howard, R. W., S. C. Jones, J. K. Mauldin & R. Beal. 1982. Abundance, distribution, and
colony size estimates of Reticulitermes spp. (Isoptera: Rhinotermitidae) in southern
Mississippi. Environmental Entomology 11: 1290-1293.
Korb, J. & S. Katrantzis. 2004. Inuence of environmental conditions on the expression
of the sexual dispersal phenotype in a lower termite: implications for the evolution of
workers in termites. Evolution & Development 6: 342-352.
Krishna, K. 1989. Order Isoptera: Termites. In: Borror, D. J., C. A. Triplehorn & N. F.
Johnson [eds]. An introduction to the study of insects. Saunders College Publishing,
Philadelphia, PA.:234-241.
Kutnik, M., P. Uva, L. Brinkworth & A.-G. Bagnères. 2004. Phylogeography of two
European Reticulitermes (Isoptera) species: the Iberian refugium. Molecular Ecology
13: 3099-3113.
Lainé, L.V. & D. J. Wright. 2003. e life cycle of Reticulitermes spp. (Isoptera: Rhinotermitidae):
what do we know? Bulletin of Entomological Research 93: 267-278.
Lenz, M. 1994. Food resources, colony growth and caste development in wood-feeding
termites. In: Hunt, J. H. & C.A. Nalepa [eds]. Nourishment and Evolution in Insect
Societies. Westview Press, Oxford, England: 159-209.
Lenz, M. & R. A. Barrett. 1982. Neotenic formation in eld colonies of Coptotermes lacteus
(Froggatt) in Australia, with comments on the roles of neotenics in the genus Coptotermes
(Isoptera: Rhinotermitidae). Sociobiology 7: 47-59.
Lenz, M. & S. Runko. 1993. Long term impact of orphaning on eld colonies of Coptotermes
lacteus (Froggatt) (Isoptera: Rhinotermitidae). Insectes Sociaux 40: 439-456.
Lewis, P. O. & D. Zaykin. 2001. Genetic Data Analysis: Computer program for the analysis
of allelic data. Version 1.0: http://lewis.eeb.uconn.edu/lewishome/soware.html.
Long, C. E., B. L. orne & N. L. Breisch. 2003. Termite colony ontogeny: a long-term
assessment of reproductive lifespan, caste ratios and colony size in Reticulitermes avipes
(Isoptera: Rhinotermitidae). Bulletin of Entomological Research 93: 439-445.
Miller, E. M. 1969. Caste dierentiation in the lower termites. In: Krishna, K. & F. M.
Weesner [eds]. Biology of termites. Academic Press, New York, NJ.: 283-310.
Miyata, H., H. Furuichi & O. Kitade. 2004. Patterns of neotenic dierentiation in a
subterranean termite, Reticulitermes speratus (Isoptera: Rhinotermitidae). Entomological
Science 7: 309-314
Myles, T. G. 1999. Review of secondary reproduction in termites (Insecta: Isoptera) with
comments on its role in termite ecology and social evolution. Sociobiology 33: 1-43.
Nijhout, H. F. 2003. Development and evolution of adaptative polyphenisms. Evolution &
Development 5: 9-18.

1033
Pichon, A. et al. — Development of Orphaned Subterranean Termites
Noirot, C. 1956. Les sexués de remplacement chez les termites supérieurs (Termitidae).
Insectes Sociaux 3: 145-158.
Noirot, C. 1989. Social structure in termites societies. Ethology Ecology and Evolution 1:
1-17.
Noirot, C. 1990. Sexual castes and reproductives strategies in termites. In: Engels W. [ed].
An Evolutionary Approach to Castes and Reproduction. Springer-Verlag, Berlin,
Germany: 5-35.
Nutting, W. L. 1970. Composition and size of some termite colonies in Arizona and Mexico.
Annals of the Entomological Society of America 63: 1105-1110.
Pamilo, P., P. Gertsch, P. oren & P. Seppä. 1997. Molecular population genetics of social
insects. Annual Reviews of Ecology and Systematics 28: 1-25.
Paulmier, I., B. Vauchot, A.-M. Pruvost, C. Lohou, M. Tussac, M. Jéquel, J.-L. Leca & J.-L.
Clément. 1997. Evaluation of two populations of Reticulitermes santonensis De Feytaud
(Isoptera) by triple mark-recapture procedure. 28th Annual Meeting of e International
Research Group on Wood Preservation. 25-30 May 1997, Whistler, BC, Canada.
Pawson, B.M. & R.E. Gold. 1996. Caste dierentiation and reproductive dynamics of
three subterranean termites in the genus Reticulitermes (Isoptera: Rhinotermitidae).
Sociobiology 28: 241-251.
Pedersen, J. S. & J. J. Boomsma. 1999. Eect of habitat saturation on the number and
turnover of queens in the polygynous ant, Myrmica sulcinodis. Journal of Evolutionary
Biology 12: 903-917.
Plateaux, L. & J.-L. Clément. 1984. La spéciation récente des termites Reticulitermes du
complexe lucifugus. Revue de la Faculté des Sciences de Tunis 3: 179-206.
Raymond, M. & F. Rousset. 1995. An exact test for population dierentiation. Evolution
49: 1280-1283.
Roisin, Y. 1993. Selective pressures on pleometrosis and secondary polygyny: a comparison
of termites and ants. In: Keller L. [ed]. ueen Number and Sociality in Insects. Oxford
University Press, New-York, NJ.: 402-421.
Roisin Y,. 2000. Diversity and evolution of caste patterns. In: Abe, T., D. E. Bignell & M.
Highashi [eds]. Termites: Evolution, Sociality, Symbioses, Ecology. Kluwer Academic
Press, Dordrecht, Netherlands: 95-119.
Ross, K.G. & L. Keller. 1995. Ecology and evolution of social organization: insights from
re ants and other highly eusocial insects. Annual Review of Ecology and Systematics
26: 631-656.
Serment, M.-M. & O. Tourteaux. 1991. Les termites. In: Les cahiers du CTBA. CTBA,
Paris, ISSN 0528-4929.
StatSoft France (2003). STATISTICA (logiciel d’analyse de données), version 6.
www.statso.com.
orne, B.L. 1996. Termite Terminolog y. Sociobiology 28: 253-263.
orne, B.L. 1998. Biology of subterranean termites of the genus Reticulitermes. In: NPCA
Research Report on Subterranean Termites, Part I. National Pest Control Association,
Virginia, U.S.A.

1034 Sociobiology Vol. 50, No. 3, 2007
Vargo, E.L. 2000. Polymorphism at trinucleotide microsatellite loci in the subterranean
termite Reticulitermes avipes. Molecular Ecology 9: 817-829.
Vargo, E.L. 2003. Hierarchical analysis of colony and population genetic structure of the
eastern subterranean termite, Reticulitermes avipes, using two classes of molecular
markers. Evolution 57: 2805-2818.
Walsh, P. S., D. A. Metzger & R. Higuchi. 1991. Chelex 100 as a medium for simple extraction
of DNA for PCR-based typing from forensic material. BioTechniques 10: 506-513.
Watanabe, H. & H. Noda. 1991. Small-scale rearing of a subterranean termite, Reticulitermes
speratus (Isoptera: Rhinotermitidae). Applied Entomolgy and Zoology 26: 418-420.
Weesner, F. M. 1965. e Termites of the United States: a handbook. e National Pest
Control Association, Elizabeth, N.J., U.S.A. 70 p.