APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 2003, p.

60186024
0099-2240/03/$08.000 DOI: 10.1128/AEM.69.10.60186024.2003
Copyright 2003, American Society for Microbiology. All Rights Reserved.

Vol. 69, No. 10

Axial Dynamics, Stability, and Interspecies Similarity of Bacterial

Community Structure in the Highly Compartmentalized Gut of
Soil-Feeding Termites (Cubitermes spp.)
Dirk Schmitt-Wagner,1 Michael W. Friedrich,2 Bianca Wagner,2 and Andreas Brune1*
kologie, Fachbereich Biologie, Universita
Mikrobielle O
t Konstanz, 78457 Konstanz,1 and Max Planck
Institute for Terrestrial Microbiology, 35043 Marburg/Lahn,2 Germany
Received 10 March 2003/Accepted 30 July 2003

The highly compartmentalized gut of soil-feeding termites is characterized by pronounced axial dynamics in
physicochemical conditions and microbial processes. In a companion paper (D. Schmitt-Wagner, M. W.
Friedrich, B. Wagner, and A. Brune, Appl. Environ. Microbiol. 69:60076017, 2003), we demonstrated that the
variety of physicochemical conditions in the different gut compartments of Cubitermes spp. is reflected in the
diversity of the respective intestinal microbial communities. Here, we used molecular fingerprints of 16S rRNA
genes of the bacterial community, obtained by terminal restriction fragment length polymorphism (T-RFLP)
analysis, to describe the axial dynamics of the bacterial community structure in the different gut sections.
Comparison of the T-RFLP profiles with the predicted terminal restriction fragments of the clones in clone
libraries of the gut segments in Cubitermes orthognathus confirmed that all hindgut sections harbored distinct
bacterial communities. Morisita indices of community similarity, calculated by comparing the different patterns, revealed large differences between the bacterial communities of soil, gut, and nest material and also
among the individual gut sections. By contrast, comparison of the homologous gut segments of different
Cubitermes species indicated that the three termite species investigated possessed a similar, gut-specific
microbiota that remained comparatively stable even during several months of maintenance in the laboratory.
A recent study combining clone analysis and T-RFLP to
investigate archaeal community structure in the gut of the
soil-feeding termite C. orthognathus provided evidence for pronounced differences among the microbiota not only between
the gut and the ingested soil, but also among the different
compartments of the intestinal tract (9). Based on the information on the phylogenetic diversity of the bacterial microbiota in the major gut compartments of C. orthognathus described in the companion paper (27), the present study
employed T-RFLP analysis to follow changes in the bacterial
community structure of ingested soil during gut passage. Comparing the terminal restriction fragment patterns of the homologous gut segments of three different species of Cubitermes,
the study also addressed questions regarding the specificity of
the gut microbiota for the compartments and the temporal
stability of gut microbial communities during maintenance of
termites under laboratory conditions.

The hindgut of soil-feeding termites is highly compartmentalized and characterized by pronounced axial dynamics of
oxygen, intestinal pH, and redox potential (1, 3, 4, 14) and of
microbial processes such as hydrogen production, methanogenesis, and reductive acetogenesis (26, 29). Recent studies
have provided strong evidence for an important role of the
special physicochemical gut conditions and the pronounced
intestinal microbiota in the chemical and microbial transformation of organic matter and microbial biomass during gut
passage (11, 12, 13).
A study of the phylogenetic diversity and axial distribution of
microorganisms in the intestinal tract of Cubitermes orthognathus revealed that this variety of physicochemical conditions
is reflected in the diversity of microbial communities in the
different gut compartments and provided first evidence for the
presence of a specific intestinal bacterial community (27).
Although sequencing and phylogenetic analysis of cloned
16S rRNA genes provides information on the different phylotypes in a given community, the effort and costs involved in this
approach limit investigations of changes in the structure of
complex microbial communities over space and time. Molecular fingerprinting methods such as denaturing gradient gel
electrophoresis (8, 21, 22) and terminal restriction fragment
length polymorphism (T-RFLP) (9, 18, 19) analyses avoid
these problems and allow the comparison of microbial communities in a larger number of samples, which has made these
methods well-established tools in microbial ecology.

MATERIALS AND METHODS

Termites and DNA extraction. Cubitermes orthognathus Emerson was collected
in grassland near Busia, and Cubitermes ugandensis Fuller was collected in a
glade of the Kakamega rainforest; both sites are located in the highlands of
Western Kenya. Species were identified by Julius Muli, National Museums of
Kenya. Cubitermes niokoloensis was collected in a dry savannah near Kolda,
Senegal, and identified by Alain Brauman, Institut de Recherche pour le Developpement, Dakar, Senegal. In addition, partial sequences (650 bp) of the
mitochondrial cytochrome oxidase II (COII) gene were determined by PCR of
DNA extracts from termite heads with previously described primers (28).
Voucher specimens of soldiers and workers preserved in alcohol are available
from the corresponding author. Whole nests were transported to the laboratory
in their country of origin, where they were cut into pieces; termites were distributed into polypropylene containers containing fragments of the nest and soil
collected at about 3 m from the nest. The termites were allowed a few days to

* Corresponding author. Mailing address: Department of Biogeochemistry, Max Planck Institute for Terrestrial Microbiology, Karlvan-Frisch-Strasse, 35043 Marburg, Germany. Phone: 49-6421-178101. Fax: 49-6421-178-999. E-mail: brune@staff.uni-marburg.de.
6018

VOL. 69, 2003

BACTERIAL COMMUNITY STRUCTURE IN TERMITE GUTS

reconstruct and fix the nest fragments in the containers before being transported
to the laboratory in Konstanz.
Termites were kept at room temperature in the dark. The containers were
inspected regularly, and parts of the nest material were removed and replaced
with fresh soil; moisture was controlled by spraying the surface of the nest
material with water. For the experiments, only worker caste termites were used.
DNA was extracted within a week after collection or at the times indicated
below. For this purpose, termites were dissected with sterile, fine-tipped forceps,
and guts were separated into six sections (27) (Fig. 1).
Pools of 10 to 20 gut sections each and aliquots (1 g) of soil samples and nest
material were extracted with a direct lysis protocol modified after that of More
et al. (20) that involves bead beating, as described previously in detail (10). DNA
was purified from the supernatant by consecutive ammonium acetate, isopropanol, and ethanol precipitation steps. To remove humic substances, the extracts
were passed through spin columns filled with polyvinylpolypyrrolidone as described previously (24). DNA concentrations were determined fluorimetrically
with Hoechst dye 33258 and a DyNA Quant 200 fluorimeter (Amersham Pharmacia Biotech, Freiburg, Germany) as recommended by the manufacturer. In
the case of C. orthognathus, the DNA extracts used for T-RFLP analyses were
from the same batch of termites as those used for cloning (27).
T-RFLP analysis. The procedure used for T-RFLP analysis followed that
described in detail by Chin et al. (5), with minor modifications. Bacterial 16S
rRNA gene sequences were amplified with 6-carboxyfluorescein-labeled primer
27F and primer 907R (5-CCG TCA ATT CCT TTR AGT TT-3 ; Escherichia
coli positions 907 to 926) (16). PCR (30 cycles) was carried out as described
previously (10), except that the annealing temperature was 52C. Purified PCR
products were quantified spectrophotometrically. PCR amplicons were digested
for 3 h at 37C in 0.5-ml reaction tubes containing 50 ng of DNA, 2.5 U of MspI
(Promega), 1 g of bovine serum albumin, and 1 l of 10 buffer in a total
volume of 10 l. Samples were analyzed on an ABI 373 sequencer (Applied
Biosystems) with a GeneScan-1000 ROX standard (5).
The total peak height of a T-RFLP profile was defined as the sum of the peak
heights of all peaks of 50 bp. The relative peak height of a given peak was
determined by dividing its peak height by the total peak height of the profile.
When determining the number of distinct terminal restriction fragments in a
given profile, only terminal restriction fragments with a relative peak height
larger than 1% of the total peak height were taken into account.
Community similarity. The Morisita index (IM) of community similarity (equation 1) (15), where is Simpsons index of dominance (calculated separately for
each community), ni is the number of individuals of species i, and N is the total
number of individuals sampled, was used.
IM

2 n i 1n i 2
1 2N1N2

(1)

Simpsons index of dominance (equation 2) describes the probability that two

randomly selected individuals from a community will be of the same species,
where s is the total number of species in the community.

nini 1

i1

NN 1

(2)

The Morisita index ranges from 0 to 1, with 0 indicating that no species are
shared between the two communities and 1 indicating complete identity. Because
the index takes species abundance into account, communities that contain the
same species but have different species abundance will have an index value of less
than 1. These equations were adapted to T-RFLP data by considering each
terminal restriction fragment a separate species and peak height a measure of
species abundance (7).

RESULTS
T-RFLP fingerprinting. The T-RFLP fingerprints obtained
for the different hindgut segments of Cubitermes orthognathus
(Fig. 1) confirmed the high diversity of the bacterial community already evidenced by the clone analysis (in the companion
paper [27]). There were striking differences among the fingerprint pattern of the different gut segments, including the midgut, and also to that of the parent soil. In the P1 and especially

6019

in the P4 sections, the high microbial diversity was reflected by

a large number of peaks in the T-RFLP profile, whereas in the
P3 and the P5 sections, only a few fragment lengths dominated
the patterns (Fig. 1).
With the predicted lengths of the terminal restriction fragments for the clonal sequences, a correlation between the
peaks in the T-RFLP patterns and the clones in the clone
library was established. The 102 clones in the clone library
yielded 69 different predicted terminal restriction fragments,
and almost every predicted terminal restriction fragment was
represented in the T-RFLP fingerprints (Table 1). Vice versa,
almost all major peaks in the T-RFLP patterns could be assigned to clones in the clone libraries (Fig. 1). Some peaks
matched the predicted terminal restriction fragments of more
than one clone (e.g., four P1 clones with different sequences
had the same predicted terminal restriction fragment of 106
bp). Often, closely related clones had a very similar predicted
terminal restriction fragment length (1- to 2-bp difference),
which were not well separated in the T-RFLP patterns (e.g.,
the predicted terminal restriction fragment length of seven
clones of the P1 cluster ranged from 106 to 108 bp).
The large phylogenetic diversity of low GC content (LGC)
clones was reflected in 46 different predicted terminal restriction fragments among a total of 69 clones (Table 1). Most
major peaks in the profile of the P1 segment could be assigned
to clones from the termite-specific clusters within the LGC
bacteria. Terminal restriction fragments of 106 to 108 bp and
299 to 300 bp, representing all nine clones of the P1 cluster,
were present only in the profiles of the P1 segment and the
midgut, which is in agreement with the absence of such clones
in the clone libraries of the posterior gut sections. In contrast,
terminal restriction fragments of 281 to 283 bp, matching the
predicted terminal restriction fragments of LGC clones in the
Sporobacter cluster (27), were present in the profiles of all
segments but represented the largest peak and also the largest
number of LGC clones in P3.
The profile of the P4 segment was characterized by a strong
increase in terminal restriction fragment length diversity and a
significant shift to longer fragment sizes, which were not
present in the P3 profile. Also, a number of shorter terminal
restriction fragments were not present in the P3 profile. The
largest peak of 204 bp matched two LGC clones from the P4
clone library (Table 1). Several other peaks in this profile
matched the predicted terminal restriction fragments of clones
from the Cytophaga-Flexibacter-Bacteroides phylum, the spirochetes, or different subgroups of the Proteobacteria. However,
two of these peaks could not be clearly assigned because clones
from other phyla had similar or identical predicted terminal
restriction fragments.
The T-RFLP profile of the P5 segment was dominated by
two major peaks. The terminal restriction fragment of 81 bp,
which was also present in the profile of P4, and the following
smaller peaks (82 to 96 bp) matched a group of eight clones
belonging to the Cytophaga-Flexibacter-Bacteroides group from
the P5 section clone library (Table 1), which clustered exclusively with clones derived from the guts of other termites (27).
The other major peak in the P5 profile, representing a terminal
restriction fragment of 497 bp, matched a clone (P5-3) belonging to the -Proteobacteria. This peak was also present in the P4

6020

SCHMITT-WAGNER ET AL.

APPL. ENVIRON. MICROBIOL.

FIG. 1. Terminal restriction-fragment length polymorphism (T-RFLP) profiles of 16S rRNA gene fragments, amplified from DNA extracts of
different gut segments of C. orthognathus and of the soil from the collection site. PCR products obtained with primers 27F (labeled) and 907R were
digested with MspI. Terminal restriction fragment lengths of major peaks (i.e., whose peak height represented more than 2.5% of the sum of all
peak heights in the respective terminal restriction fragment pattern) that matched the predicted terminal restriction fragments of clones in the
respective clone library are marked in bold and labeled with the phylum of the associated clone (for abbreviations, see Table 1). For orientation,
the sizes of several unassigned terminal restriction fragments are also indicated.

VOL. 69, 2003

BACTERIAL COMMUNITY STRUCTURE IN TERMITE GUTS

6021

TABLE 1. Predicted terminal restriction fragment lengths after MspI digestion and phylogenetic affiliation of
the 16S rRNA gene clonesb in the different clone libraries of C. orthognathus
Terminal fragment
length (bp)

Clone library (gut section)

P1

69
81
82
84
87
90
91
92
93
96
106

107
108
131

P3

P4

P318

P523
P527
P516
P525
P522

P325
P33

P13
P117
P128
P129
P12
P14
P18

P530
P512

P311
P35

P529

P42
P45

P111
P122

P110

P411
P39
P321

P31
P320

206
210
216
218
219
227

P422
P46
P416

P423
P426
P412

P125

Phylogenetic
groupa

LGC
CFB
CFB
CFB
CFB
LGC
CFB
CFB
CFB
CFB
LGC

LGC

135
139
139
140
141
152
153
161
163
166
166
171
180
183
188
204

P5

P330

P47

P57
P528
P511
P519
P520

P518
P52

P526
P514
P515
P521

LGC
Delta
LGC
Spiro
Beta
CFB
Beta
Alpha
LGC
Spiro
Unident.
LGC
Plancto
LGC
LGC
LGC
LGC
LGC
LGC
LGC
LGC
LGC
LGC
LGC

Thermal fragment
length (bp)

275
278
281
282

Clone library (gut section)

P1

P16
P17
P19

P118
P124

292

P15
P113

303
309
310
373
385
402
421
430
446
464
487
488
489
490
490
497
499
512
542
575

P119
P11
P114
P115
P121
P123

P130

P329

P510
P524

P36
P37
P317
P323
P324
P34

P312

Phylogenetic
groupa

LGC
LGC
LGC
LGC
Spiro
LGC

LGC
P51
P56
P430

LGC
LGC
LGC
LGC
LGC
LGC

P328

P315

P424

Unident.
LGC
LGC

P415

Spiro
LGC
LGC
LGC
Beta
LGC
LGC
LGC
LGC
Gamma
Beta
LGC
Delta
LGC
LGC
CFB

P414
P314

P116
P120
P126

P5

P54

P322

283

300

P4

P32

282

293
295
296
297
299

P3

P326

P319

P55

P49
P48
P429
P425
P41
P44

P59
P58
P53

P513

LGC

a
Phylogenetic groups: LGC, gram-positive bacteria with low GC content; CFB, Cytophaga-Flexibacter-Bacteroides; alpha, beta, gamma, and delta, various subgroups
of Proteobacteria; Spiro, Spirochetes; Plancto, Planctomycetales; Unident, unidentified.
b
Clones in italics could not be assigned to any of the terminal restriction fragments in the profile of the respective gut section.

profile, where it matched another clone belonging to the -Proteobacteria (P4-29) closely related to clone P5-3 (27).
The T-RFLP profiles of the soil sample and the nest material were at least as diverse as those of any gut section. However, only a few peaks were shared between soil and midgut,
and the peaks with identical terminal restriction fragment
lengths changed tremendously in their relative abundance.
Terminal restriction fragments representing LGC clones from
the termite-specific clusters (106 to 108, 283, and 300 bp) were
not present or gave only minor peaks in the soil fingerprints.
Also, the characteristic terminal restriction fragments matching the termite-specific clones of the Cytophaga-FlexibacterBacteroides group (81, 82, 140, and 542 bp) were not present in

the soil sample or the midgut, which indicated that the gut
bacterial community is not merely a reflection of that found in
the soil and also differs in the individual gut compartments.
Morisita index of community similarity. Morisita indices
were calculated as a measure of community similarity between
different pairs of T-RFLP profiles. In most cases, the profiles
of neighboring gut sections were more similar than those of
sections that were not directly connected to each other (Table
2). However, in no case did the Morisita indices exceed a value
of 0.5, underlining that the community structure in neighboring
segments also differed considerably, even if their T-RFLP profiles shared individual peaks. Morisita indices between gut
segments and soil were very low, which substantiates that the

6022

SCHMITT-WAGNER ET AL.

APPL. ENVIRON. MICROBIOL.

TABLE 2. Morisita indices of community similarity between the

soil from the collection site, the different gut sections of
C. orthognathus, and the nest materiala
Sample

Soil
Crop
Midgut
P1
P3
P4
P5

Crop

Midgut

P1

P3

P4

P5

Nest

0.27

0.26
0.01

0.12
0.17
0.32

0.09
0.11
0.17
0.28

0.17
0.13
0.22
0.35
0.29

0.14
0.25
0.04
0.10
0.15
0.49

0.76
0.15
0.34
0.16
0.15
0.14
0.08

The boundaries of the individual segments are depicted in Fig. 1 of the

companion paper (27).

shift from the soil to the gut community takes place in the
foregut and is completed in the midgut. It is interesting that
there was also little similarity in the profiles of the rectum (P5)
and the nest material, indicating that the microbial community
also changes strongly between rectal contents and nest material (which is constructed from soil and feces). The highest
Morisita index (0.76) was found when the T-RFLP profile of
the soil was compared to that of the nest material.
Low similarities between T-RFLP profiles of the individual
hindgut segments were also obtained with Cubitermes ugandensis and Cubitermes niokoloensis (Table 3), confirming that each
segment is colonized by a specific bacterial community. When
the profiles obtained for the hindgut segments of C. orthognathus were compared with those of the other two species
(Table 4), the Morisita indices were much higher than those
between the individual gut segments of the same termite,
which indicates that homologous hindgut segments harbor similar microbial communities. The highest value (0.97) was obtained for the P3 segments of C. ugandensis and C. niokoloensis, which indicates an almost identical profile.
T-RFLP analysis was used also to assess the temporal stability of the bacterial communities in the different gut segments
of C. ugandensis. Profiles were obtained directly after collection and after 2 and 5 months of storage of the termites in the
laboratory, and Morisita indices were calculated for the samples of each gut section collected at different times (Table 5).
Shifts in community structure were evident in the case of the
P1 and P4 segments and seemed to occur mostly during the

TABLE 3. Morisita indices of community similarity between the

four major hindgut segments of Cubitermes ugandensis and
Cubitermes niokoloensis.

C. ugandensis
P1
P3
P4
C. niokoloensis
P1
P3
P4

Morisita index
P3

P4

P5

0.14

0.28
0.32

0.15
0.21
0.34

0.12

Morisita index

Segment and species

Morisita index

Species and segment

TABLE 4. Morisita indices of community similarity in homologous

hindgut segments among three different species of Cubitermes

0.08
0.19

0.15
0.15
0.36

P1
C.
C.
P3
C.
C.
P4
C.
C.
P5
C.
C.

C. ugandensis

C. niokoloensis

orthognathus
ugandensis

0.89

0.43
0.48

orthognathus
ugandensis

0.69

0.67
0.97

orthognathus
ugandensis

0.57

0.40
0.58

orthognathus
ugandensis

0.70

0.41
0.54

first 2 months after collection. However, the high similarity of

the P3 profiles documents considerable stability of the microbial community of this segment over time.
DISCUSSION
The results of this study clearly demonstrate that the gut of
soil-feeding termites (Cubitermes spp.) harbors a specific bacterial microbiota and that structural and functional differences
of the individual gut segments are reflected by changes in the
composition of the bacterial community. The combined results
of the molecular fingerprints obtained in the present study and
the results of clone analysis and fluorescence in situ hybridization presented in the companion paper (27) provide strong
support for the presence of specific bacterial populations in the
different gut compartments. This is in agreement with a previous cultivation-independent study that demonstrated the
presence of specific populations of methanogens in the different gut segments of C. orthognathus (9).
Soils and intestinal tracts are environments with highly complex microbial communities. There is good evidence for several
thousand microbial species in a single soil sample (25, 30), and
the gastrointestinal tract of pigs contains almost 400 different
bacterial phylotypes when a 16S rRNA gene sequence similarity of 97% is used as a differential criterion (17). Therefore,
a cloning approach may suffice to characterize microbial communities at the phylogenetic-group level, but cannot be used to
document detailed differences in community structure with the
necessary resolution.
Fingerprinting techniques, on the other hand, are able to
TABLE 5. Morisita indices documenting changes in community
structure within the individual hindgut segments of C. ugandensis
during maintenance of the termites in the laboratorya
Time interval (mo)

02
25
05

Morisita index
P1

P3

P4

P5

0.77
0.88
0.62

0.94
0.92
0.93

0.50
0.84
0.50

0.80
0.80
0.76

a
Termites were collected from the field in May 2001 and analyzed directly
after collection and after 2 and 5 months of captivity, during which the termites
were fed only soil from the collection site.

VOL. 69, 2003

BACTERIAL COMMUNITY STRUCTURE IN TERMITE GUTS

generate an image of all gene fragments amplified from the

community rather than just a clonal subset. Moreover, the
integrative nature of the method, i.e., the ability to sum all
clones of a phylogenetic cluster within the same restriction
fragment, helps to reduce the complexity to a manageable
format and offers the possibility of comparing a large number
of samples.
Since clones from different phylogenetic groups could have
identical first restriction sites, which could result in identical
terminal restriction fragment lengths, a single peak in a TRFLP profile does not necessarily represent only one phylogenetic group. For this reason, T-RFLP should always be combined with sequence data from clone analysis (9) and, if
necessary, performed with more than one restriction enzyme.
The advantages of this approach are documented by the combined results of the present study and its companion (27). The
fingerprint patterns of the different gut sections of C. orthognathus and the other Cubitermes spp. document that the compositions of the complex microbial communities of the individual segments consist of different phylotypes. However, only
when the individual terminal restriction fragments were assigned to their corresponding clones in the clone libraries did
it become apparent that the pattern of P1 is caused mainly by
bacteria from one phylum (the LGC group), whereas that of
the P4 segment is due the presence of bacteria from a variety
of phyla.
Presence of gut-specific microbiota. Mathematical approaches such as the Morisita index provide an objective measure of community similarity in different samples (15). This
procedure not only allows a peak-to-peak pairwise comparison
of the presence and relative abundance of terminal restriction
fragments in the different patterns but is also insensitive to
differences in the absolute signal between the profiles. Although originally introduced into ecological research to compare communities at the species level, it has been adapted for
use with T-RFLP profiles by treating the individual terminal
restriction fragments like different species (7). Although the
operational definition has its weaknesses (one peak may represent more than one phylotype), it is nevertheless a powerful
tool to compare complex community profiles with a single
index value.
The composition of the microbial soil community and also of
the nest material is fundamentally different from that in the
intestinal tract of soil-feeding termites. The Morisita indices
show distinct differences between soil and gut bacterial communities. Also, the microbiota colonizing the nest material,
which is largely constructed of soil and feces (6), resembles
that of the soil rather than that of the rectal contents, adding
further evidence for the presence of a gut-specific microbiota.
Community structure in different gut compartments. In
general, the differences between the T-RFLP profiles of soil
and anterior hindgut sections are enormous. Extremely low
Morisita indices were obtained between soil and crop or midgut samples, which indicated that the changes in community
structure are not due only to the high alkalinity in the anterior
hindgut of Cubitermes spp. (pH 12 [4]), but probably also
due to digestion of soil microorganisms in the anterior gut
regions of soil-feeding termites, as discussed in the companion
paper (27).
The passage of the gut contents through the intestinal tract

6023

is relatively fast; transit times of 36 to 48 h have been reported

for the soil-feeding termite Procubitermes aburiensis (2). It is
possible that minor populations of allochthonous bacteria are
specifically enriched and give rise to transient populations in
those gut compartments that provide favorable environmental
conditions. On the other hand, the soil collected from the
vicinity of the nest will have been processed by soil-feeding
termites for many years, and the presence of gut bacteria in soil
material has to be expected, especially in the case of the LGC
clones, many of which may form endospores. To our knowledge, transfer of microbial symbionts by proctodeal trophallaxis, which is well established in lower termites (23), has not
yet been documented for soil-feeding Termitinae. However,
even an inoculation of the intestinal tract via fecal material
contained in the food soil would not contradict the existence of
autochthonous microbial populations, represented by bacterial
lineages found (so far) only in the gut of soil-feeding termites.
Also, the relative stability of the community composition
over time, as documented for C. ugandensis in this study, supports the presence of gut-specific bacterial communities, especially in the P3 segment. Nevertheless, the fact that the largest
changes in community structure occurred during the first 2
months of maintenance in the laboratory underlines that it is
important to perform such studies with material preserved
immediately after collection.
Interspecies comparison of gut communities. Further support for the presence of a specific gut microbiota is provided by
the observation that the profiles of homologous gut segments
in different soil-feeding termite species from different geographic locations were more similar to each other than the
profiles of the different gut segments within a single termite
species. The specificity of the intestinal microbiota in soilfeeding termites is obviously independent of the geographic
location and the soil of the collection site. Although it is not
possible to make correlations between T-RFLP profiles and
specific genotypes in the absence of a clone library, it is reasonable to assume that the large number of identical terminal
restriction fragments reflects the presence of bacteria from
phylogenetically related groups and therefore shows a specificity of the microbiota of soil-feeding termites without the
restriction of a termite species barrier.
ACKNOWLEDGMENTS
This study was supported by the Deutsche Forschungsgemeinschaft
and by the Max Planck Society.
We thank Hamadi Boga (Jomo Kenyatta University of Agriculture
and Technology, Nairobi, Kenya), Lucie Rogo and Nixon Onyimbo
(International Centre of Insect Physiology and Ecology, Nairobi),
Wanja Kinuthia and the late Julius Muli (National Museum of Kenya),
and Alain Brauman (Institut de Recherche pour le Developpement,
Dakar, Senegal) for help with termite collection and identification and
Karen A. Brune for critically reading the manuscript.
REFERENCES
1. Bignell, D. E., and P. Eggleton. 1995. On the elevated intestinal pH of higher
termites (Isoptera: Termitidae). Insect Soc. 42:5769.
2. Bignell, D. E., H. Oskarsson, and J. M. Anderson. 1980. Distribution and
abundance of bacteria in the gut of a soil-feeding termite Procubitermes
aburiensis (Termitidae, Termitinae). J. Gen. Microbiol. 117:393403.
3. Brune, A. 1998. Termite guts: the worlds smallest bioreactors. Trends Biotechnol. 16:1621.
4. Brune, A., and M. Ku
hl. 1996. pH profiles of the extremely alkaline hindguts
of soil-feeding termites (Isoptera: Termitidae) determined with microelectrodes. J. Insect Physiol. 42:11211127.

6024

SCHMITT-WAGNER ET AL.

5. Chin, K. J., T. Lukow, and R. Conrad. 1999. Effect of temperature on

structure and function of the methanogenic archaeal community in an anoxic
rice field soil. Appl. Environ. Microbiol. 65:23412349.
6. Collins, N. M. 1989. Termites, p. 455472. In H. Lieth and M. J. A. Werger
(ed.), Ecosystems of the world 14B: tropical rain forest ecosystems, biogeographical and ecological studies. Elsevier, Amsterdam, The Netherlands.
7. Dollhopf, S. L., S. A. Hashsham, and J. M. Tiedje. 2001. Interpreting 16S
rDNA T-RFLP data: application of self-organizing maps and principal component analysis to describe community dynamics and convergence. Microb.
Ecol. 42:495505.
8. Ferris, M. J., G. Muyzer, and D. M. Ward. 1996. Denaturating gradient gel
electrophoresis profiles of 16S rRNA-defined populations inhabiting a hot
spring microbial mat community. Appl. Environ. Microbiol. 62:340346.
9. Friedrich, M. W., D. Schmitt-Wagner, T. Lueders, and A. Brune. 2001. Axial
differences in community structure of Crenarchaeota and Euryarchaeota in
the highly compartmentalized gut of the soil-feeding termite Cubitermes
orthognathus. Appl. Environ. Microbiol. 67:48804890.
10. Henckel, T., M. Friedrich, and R. Conrad. 1999. Molecular analyses of the
methane-oxidizing microbial community in rice field soil by targeting the
genes of the 16S rRNA, particulate methane monooxygenase, and methanol
dehydrogenase. Appl. Environ. Microbiol. 65:19801990.
11. Ji, R., and A. Brune. 2001. Transformation and mineralization of 14C-labeled
cellulose, peptidoglycan, and protein by the soil-feeding termite Cubitermes
orthognatus. Biol. Fertil. Soils 33:166174.
12. Ji, R., A. Kappler, and A. Brune. 2000. Transformation and mineralization of
synthetic 14C-labeled humic model compounds by soil-feeding termites. Soil
Biol. Biochem. 32:12811291.
13. Kappler, A., and A. Brune. 1999. Influence of gut alkalinity and oxygen status
on mobilization and size-class distribution of humic acids in the hindgut of
soil-feeding termites. Appl. Soil Ecol. 13:219229.
14. Kappler, A., and A. Brune. 2002. Dynamics of redox potential and changes
in redox state of iron and humic acids during gut passage in soil-feeding
termites (Cubitermes spp.). Soil Biol. Biochem. 34:221227.
15. Krebs, C. J. 1998. Ecological methodology. Benjamin Cummings, Menlo
Park, Calif.
16. Lane, D. J., B. Pace, G. J. Olsen, D. A. Stahl, M. L. Sogin, and N. R. Pace.
1985. Rapid determination of 16S ribosomal RNA sequences for phylogenetic analyses. Proc. Natl. Acad. Sci. USA 82:69556959.
17. Leser, T. D., J. Z. Amenuvor, T. K. Jensen, R. H. Lindecrona, M. Boye, and
K. Mller. 2002. Culture-independent analysis of gut bacteria: the pig gastrointestinal tract microbiota revisited. Appl. Environ. Microbiol. 68:673
690.
18. Liu, W. T., T. L. Marsh, H. Cheng, and L. J. Forney. 1997. Characterization

APPL. ENVIRON. MICROBIOL.

19.
20.

21.
22.

23.

24.
25.
26.

27.

28.
29.

30.

of microbial diversity by determining terminal restriction fragment length

polymorphisms of genes encoding 16S rRNA. Appl. Environ. Microbiol.
63:45164522.
Marsh, T. L. 1999. Terminal restriction fragment length polymorphism (TRFLP): an emerging method for characterizing diversity among homologous
populations of amplification products. Curr. Opin. Microbiol. 2:323327.
More, M. I., J. B. Herrick, M. C. Silva, W. C. Ghiorse, and E. L. Madsen.
1994. Quantitative cell lysis of indigenous microorganisms and rapid extraction of microbial DNA from sediment. Appl. Environ. Microbiol. 60:1572
1580.
Muyzer, G. 1999. DGGE/TGGE, a method for identifying genes from natural ecosystems. Curr. Opin. Microbiol. 2:317322.
Muyzer, G., E. C. de Waal, and A. G. Uitterlinden. 1993. Profiling of complex
microbial populations by denaturing gradient gel electrophoresis analysis of
polymerase chain reaction-amplified genes coding for 16S rRNA. Appl.
Environ. Microbiol. 59:695700.
Nalepa, C. A., and C. Bandi. 2000. Characterizing the ancestors: paedomorphosis and termite evolution, p. 5375. In T. Abe, D. E. Bignell, and M.
Higashi (ed.), Termites: evolution, sociality, symbiosis, ecology. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Ramakrishnan, B., T. Lueders, R. Conrad, and M. Friedrich. 2000. Effect of
soil aggregate size on methanogenesis and archaeal community structure in
anoxic rice field soil. FEMS Microbiol. Ecol. 32:261270.
Sandaa, R.-A., V. Torsvik, . Torsvik, F. L. Daae, T. Castberg, and D. Hahn.
1999. Analysis of bacterial communities in heavy metal-contaminated soils at
different levels of resolution. FEMS Microbiol. Ecol. 30:237251.
Schmitt-Wagner, D., and A. Brune. 1999. Hydrogen profiles and localization
of methanogenic activities in the highly compartmentalized hindgut of soilfeeding higher termites (Cubitermes spp.). Appl. Environ. Microbiol. 65:
44904496.
Schmitt-Wagner, D., M. F. Friedrich, B. Wagner, and A. Brune. 2003. Phylogenetic diversity, abundance, and axial distribution of microorganisms in
the intestinal tract of soil-feeding termites (Cubitermes spp.). Appl. Environ.
Microbiol. 69:58075817.
Thompson, G. J., L. R. Miller, M. Lenz, and R. H. Crozier. 2000. Phylogenetic analysis and trait evolution in Australian lineages of drywood termites
(Isoptera, Kalotermitidae). Mol. Phylogenet. Evol. 17:419429.
Tholen, A., and A. Brune. 1999. Localization and in situ activities of homoacetogenic bacteria in the highly compartmentalized hindgut of soilfeeding higher termites (Cubitermes spp.). Appl. Environ. Microbiol. 65:
44974505.
Torsvik, V., J. Goksoyr, and F. L. Daae. 1990. High diversity in DNA of soil
bacteria. Appl. Environ. Microbiol. 56:782787.