SPECIAL ISSUE: NATURE’S MICROBIOME
Niphargus–Thiothrix associations may be widespread in
sulphidic groundwater ecosystems: evidence from
southeastern Romania
JEAN-FRANC
 OIS FLOT,*† 1 JAN BAUERMEISTER,*1 TRAIAN BRAD,‡ ALEXANDRA
HILLEBRAND-VOICULESCU,§ SERBAN M. SARBU¶ and SHARMISHTHA DATTAGUPTA*
*Courant Research Center Geobiology, University of Göttingen, Goldschmidtstraße 3, 37077 Göttingen, Germany, †Max Planck
Institute for Dynamics and Self-Organization, Biological Physics and Evolutionary Dynamics, Bunsenstraße 10, 37073
G€ottingen, Germany, ‡Emil Racovit
a Institute of Speleology, Strada Clinicilor 5, 400006 Cluj-Napoca, Romania, §Emil
Racovit
a Institute of Speleology, Strada Frumoasa 31, 010986 Bucures
ti, Romania, ¶Grupul de Explorari Subacvatice s
i
Speologice, Strada Frumoasa 31, 010986 Bucures
ti, Romania
Abstract
Niphargus is a speciose amphipod genus found in groundwater habitats across Europe.
Three Niphargus species living in the sulphidic Frasassi caves in Italy harbour sulphur-
oxidizing Thiothrix bacterial ectosymbionts. These three species are distantly related,
implying that the ability to form ectosymbioses with Thiothrix may be common among
Niphargus. Therefore, Niphargus–Thiothrix associations may also be found in sulphidic
aquifers other than Frasassi. In this study, we examined this possibility by analysing ni-
phargids of the genera Niphargus and Pontoniphargus collected from the partly sulphidic
aquifers of the Southern Dobrogea region of Romania, which are accessible through
springs, wells and Movile Cave. Molecular and morphological analyses revealed seven
niphargid species in this region. Five of these species occurred occasionally or exclusively
in sulphidic locations, whereas the remaining two were restricted to nonsulphidic areas.
Thiothrix were detected by PCR on all seven Dobrogean niphargid species and observed
using microscopy to be predominantly attached to their hosts’ appendages. 16S rRNA
gene sequences of the Thiothrix epibionts fell into two main clades, one of which (herein
named T4) occurred solely on niphargids collected in sulphidic locations. The other Thio-
thrix clade was present on niphargids from both sulphidic and nonsulphidic areas and
indistinguishable from the T3 ectosymbiont clade previously identified on Frasassi-
dwelling Niphargus. Although niphargids from Frasassi and Southern Dobrogea are not
closely related, the patterns of their association with Thiothrix are remarkably alike. The
finding of similar Niphargus–Thiothrix associations in aquifers located 1200 km apart
suggests that they may be widespread in European groundwater ecosystems.
Keywords: amphipods, ecology, sulphide, symbiosis, systematics, taxonomy
Received 15 April 2013; revision received 6 July 2013; accepted 13 July 2013
Introduction
Since their discovery at hydrothermal vents in the late
1970s, myriad examples of symbioses between sul-
phur-oxidizing bacteria and invertebrates have been
discovered worldwide in sulphidic marine environments
(Dubilier et al. 2008). Dark, isolated and sulphide-rich
habitats analogous to hydrothermal vents also exist in
land-locked caves, such as Movile Cave in Romania and
the Frasassi caves in Italy (Forti et al. 2002). Recently, ec-
tosymbioses between sulphur-oxidizing Thiothrix bacte-
ria and three species of the groundwater amphipod
genus Nipharguswere reported from Frasassi (Dattagupta
et al. 2009; Bauermeister et al. 2012), extending the realm
of such symbioses into nonmarine ecosystems.
Correspondence: Jean-Franc
ois Flot and Sharmishtha
Dattagupta, Fax: +49-551-5176-575; +49-551-39-7918; E-mails:
jean-francois.flot@ds.mpg.de and sdattag@uni-goettingen.de
1Joint first authors.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
This is an open access article under the terms of the Creative Commons Attribution License,
which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Molecular Ecology (2014) 23, 1405–1417 doi: 10.1111/mec.12461
The three ectosymbiotic Niphargus species in Frasassi
harbour on their exoskeleton three distinct Thiothrix
clades (T1–T3), which are predominantly attached to
hairs (setae) and spines of their legs and antennae (Bau-
ermeister et al. 2012). Clade T1 has so far only been
found on Niphargus frasassianus, a species restricted to
sulphidic locations, whereas clades T2 and T3 occur on
Niphargus species in both sulphidic and nonsulphidic
waters (T2 on Niphargus ictus and Niphargus montanarius,
and T3 on all three Frasassi-dwelling species). The three
ectosymbiont clades do not form a monophyletic group
(Bauermeister et al. 2012), and neither do their host spe-
cies (Flot et al. 2010a). The lack of congruence between
the host and symbiont phylogenies suggests indepen-
dent establishments of the symbioses and/or interspe-
cies symbiont transfer (Bauermeister et al. 2012).
Although sulphide is a potent inhibitor of mitochon-
drial electron transfer (Bagarinao 1992) that is generally
toxic to aquatic life (Theede et al. 1969; Oseid & Smith
1974; Sandberg-Kilpi et al. 1999), several niphargid
species have been reported to thrive in sulphide-rich
environments such as Frasassi (up to 415 lM sulphide;
Flot et al. 2010a) and Acquapuzza (410 lM sulphide;
Latella et al. 1999) in Italy as well as Movile Cave in
Romania (up to 500 lM sulphide; Sarbu 2000). Other
Niphargus species are found in the sulphidic cave of Me-
lissotrypa in Greece (J.-F. Flot and S. Dattagupta, unpub-
lished data) as well as in anchihaline caves in Croatia,
where sulphide is present but has not yet been quantified
(Sket 1996; Gottstein et al. 2007). This raises the question
whether Niphargus–Thiothrix associations are restricted to
Frasassi or also found in other sulphidic locations.
The Southern Dobrogea region (southwestern Roma-
nia) provides the ideal locality to start examining this
question, as it has a sulphidic aquifer that can be accessed
through artificial wells, springs and Movile Cave. Discov-
ered in 1986, Movile Cave was the first terrestrial chemo-
autotrophic ecosystem described (Sarbu & Popa 1992;
Sarbu et al. 1996; Sarbu 2000) and is one of the most thor-
oughly studied to date. It harbours two niphargid species
(Sarbu et al. 1996), and amphipods are also known to
occur in sulphidic and nonsulphidic wells and springs in
the surrounding area. Our goals in this study were (i) to
molecularly characterize the niphargids of the Southern
Dobrogea region and compare them phylogenetically
with Frasassi species and (ii) to examine them for Thio-
thrix epibionts using microscopy and molecular methods.
Materials and methods
Sampling and DNA extractions
Amphipods were collected between April 2011 and
September 2012 in the Southern Dobrogea region of
Romania (see Tables 1, S1 and S2 for information on
collection sites). Specimens were preserved in 70% eth-
anol for morphological examination and DNA sequenc-
ing, in RNAlater
 (Sigma-Aldrich, Steinheim,
Germany) for DNA sequencing and fluorescence in situ
hybridization (FISH), and in 2.5% glutaraldehyde pre-
pared in phosphate-buffered saline (PBS) for scanning
electron microscopy (SEM). A microbial mat sample
was collected in April 2011 from Airbell 2 of Movile
Cave (see Sarbu et al. 1994 for a map) and kept frozen
at 
20 °C. This mat sample was later used for explor-
ing the diversity of free-living Thiothrix found in Mo-
vile Cave.
DNA extractions for niphargid sequencing and PCR
screenings (see below) were performed as described
in Flot et al. (2010a) using Qiagen DNeasy kits,
whereas DNA extractions for 16S rRNA gene clone
library constructions followed Bauermeister et al.
(2012). DNA from the microbial mat sample was
extracted using the method of Neufeld et al. (2007).
All sequencing was performed on an ABI 3730xl
DNA Analyser.
Niphargid sequence analysis
A total of 71 amphipod specimens were analysed
molecularly (Table S1, Supporting information), com-
prising 69 niphargids and two outgroups (Synurella sp.
and Gammarus sp.). PCR amplifications and sequencing
of a fragment of the mitochondrial cytochrome c oxi-
dase (COI) gene, of the complete nuclear internal tran-
scribed spacer (ITS) and of a fragment of the nuclear
28S rRNA gene were performed as reported previously
(Flot et al. 2010a). The sequences of length-variant
heterozygotes (Flot et al. 2006) were unravelled using
the online program Champuru (Flot 2007; http://www.
mnhn.fr/jfflot/champuru), and the haplotypes of other
heterozygotes were determined using Clark’s method
(Clark 1990).
Sequences were aligned by eye in MEGA5 (Tamura
et al. 2011) for COI or using MAFFT’s E-INS-i option
(Katoh et al. 2002) for the 28S rRNA gene and for
ITS. Only the 1st and 2nd codon positions were taken
into account for COI. Maximum-likelihood (ML)
phylogenetic analyses were performed in MEGA5
under the GTR+G+I model (using all sites) and
with 1000 bootstrap replicates (Felsenstein 1985);
additional bootstrap analyses were conducted using
neighbour-joining (under the K2P model, pairwise
deletion) and parsimony approaches. The ITS
phylogenetic tree was turned into a haploweb by
adding connections between haplotypes found co-
occurring in heterozygous individuals (Flot et al.
2010b, 2011).
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1406 J . -F . FLOT, J . BAUERMEISTER ET AL.
Thiothrix epibiont detection using SEM and FISH
Ten specimens from Movile Cave and from surround-
ing sulphidic and nonsulphidic wells in the town of
Mangalia were examined for filamentous Thiothrix
epibionts using SEM (Table S2, Supporting information).
Sample preparation and analysis were done as
described previously (Bauermeister et al. 2012).
Three other individuals (JFF_12.19, JFF_12.32 and
JFF_12.39) were investigated using FISH. The Thiothrix-
specific oligonucleotide probe G123T (5′-CCT TCC GAT
CTC TAT GCA-3′) and its competitor probe G123T-C
(5′-CCT TCC GAT CTC TAC GCA-3′; Kanagawa et al.
2000) were synthesized at Eurofins MWG Operon
(Ebersberg, Germany), with G123T being 5′-fluorescent-
ly labelled with cyanine-3. Both probes were mixed in
equimolar amounts to enhance binding specificity to
Thiothrix as recommended (Kanagawa et al. 2000).
Niphargid samples were fixed in 4% paraformaldehyde
for 3 h at 4 °C and transferred to a 1:1 ethanol–PBS
solution. Several legs and antennae of each specimen
were dissected and transferred into individual tubes.
FISH experiments were carried out as described by
Amann (1995) using the 1:1 G123T/G123T-C probe mix.
Formamide concentration was 40% (following Kanaga-
wa et al. 2000) and hybridization time was set to 2 h.
Table 1 Location details and groundwater geochemical characteristics of niphargid collection sites in this study
Town
Measurement
location Latitude Longitude
Measurement
date T (°C)
EC
(lS/cm)
Eh
(mV)
H2S
(lM)
Niphargid
species
Hagieni Hagieni Spring 43°48′08.90″N 28°28′29.00″E 09.2012 18.1 1080 
245 (172) N. hrabei*,
P. ruffoi
Mangalia Movile Cave 43°49′36.38″N 28°33′43.48″E 09.2011 21.2 1071 
341 245 N. cf. stygius,
P. racovitzai
Mangalia str. Matei
Basarab 62
43°49′09.11″N 28°34′16.10″E 09.2012 19.8 1380 
266 (188) N. cf. stygius,
N. decui
Mangalia str. Matei
Basarab 74
43°49′10.61″N 28°34′ 07.90″E 09.2012 18.1 1460 
174 (126) N. decui
Mangalia str. Gheorge
Netoi 1
43°49′10.87″N 28°34′12.74″E 09.2011 18.6 1078 
263 133 N. cf. stygius
Mangalia str. Dumitru
Ana 13
43°49′23.59″N 28°34′01.45″E 09.2011 19.1 1052 
120 101 N. cf. stygius
Mangalia str. Ion Mecu 51 43°49′25.75″N 28°34′29.40″E 09.2011 19.9 1135 
89 66 N. cf. stygius
Mangalia Aleea Cetat
ii 1 43°48′53.21″N 28°35′01.84″E 05.2013 19.3 1650 
64 (48) N. cf. stygius,
P. racovitzai
Mangalia str. Maior
Giurescu 22
43°49′15.43″N 28°34′46.19″E 09.2012 18.4 2440 28 (0) N. gallicus
Mangalia str. Horia Clos
ca
Cris
an 13
43°49′18.67″N 28°34′ 23.10″E 09.2012 19.7 1450 40 (0) N. decui
Mangalia str. Delfinului 16 43°48′34.73″N 28°34′44.89″E 09.2011 19.1 1473 66 0 N. gallicus
Mangalia str. Mihai Viteazu 20 43°48′49.30″N 28°34′50.31″E 09.2011 17.4 1193 68 0 N. decui
Mangalia str. Pictor Tonitza 1 43°49′09.05″N 28°35′03.71″E 09.2011 19.0 1242 104 0 N. cf. stygius,
N. decui,
P. racovitzai
Mangalia str. Vasile Pa^rvan 16 43°48′51.25″N 28°35′07.32″E 09.2012 15.0 2166 139 (0) N. gallicus
Mangalia str. Delfinului 16 43°48′34.73″N 28°34′44.89″E 09.2011 19.1 1473 66 0 N. gallicus
Albes
ti near road 393 43°47′47.50″N 28°25′35.80″E 09.2011 13.8 1445 72 0 N. decui
Doi Mai str. Mihail
Kogalniceanu 393
43°47′25.72″N 28°34′37.10″E 09.2011 16.1 2235 56 0 N. dobrogicus,
N. decui
Dulces
ti near road 394 43°54′00.07″N 28°32′39.10″E 09.2011 15.4 1216 63 0 N. gallicus,
N. decui
Limanu corner str.
Marului/str.
Traian Vuia
43°48′10.01″N 28°31′24.83″E 09.2011 16.2 1092 69 0 N. dobrogicus,
N. decui
Vama Veche str. Mihail
Kogalniceanu 23
43°45′ 07.07″N 28°34′18.59″E 05.2013 14.3 1690 55 (0) N. dobrogicus
Vama Veche str. Plajei 100 43°45′09.10″N 28°34′38.40″E 05.2013 14.0 1760 73 (0) N. decui
H2S values in parentheses were not measured directly but inferred from measured redox potentials (taking advantage of the quasi-
linear relationship observed between these two parameters).
*This species was collected at a time when the spring was almost dry and no smell of sulphide was perceived.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1407
The niphargid appendages were subsequently trans-
ferred onto glass slides and mounted with Vectashield
(Vector Laboratories, Burlingame, CA, USA). Confocal
epifluorescence micrographs of attached Thiothrix fila-
ments were collected on a Zeiss 510 Meta scanning
microscope equipped with argon and helium–neon
lasers (wavelengths 488 and 543 nm, respectively).
Design and optimization of Thiothrix-specific PCR
primers
The Thiothrix-specific forward primer THIO714F (5′-ATG
CAT AGA GAT CGG AAG G-3′; Bauermeister et al.
2012) and the newly designed reverse primer THIO1492R
(5′-GGC TAC CTT GTT ACG ACT T-3′) were used for
constructing partial 16S rRNA gene clone libraries and
for direct PCR screening of niphargids. Using PRIM-
ROSE (Ashelford et al. 2002), THIO1492R was designed
to match nearly all publicly available Thiothrix sequences.
Gradient PCRs were performed to determine the optimal
annealing temperature for the primer pair. THIO714F
worked well for PCRs but not for sequencing. Thus,
another Thiothrix-specific forward primer (THIO718Fseq;
5′-ATA GAG ATC GGA AGG AAC A-3′) was designed
as described above and used in combination with
THIO1492R for direct sequencing of PCR products.
Thiothrix clone library construction
Partial 16S rRNA gene clone libraries were constructed
from DNA extracts of the Movile microbial mat sample
and of two niphargid specimens (AH_10.4 and
SS_10.1). PCR mixtures (50 lL) contained 19 ammo-
nium buffer (Bioline, Luckenwalde, Germany), 2 mM
MgCl2, 0.3 mM dNTP mix (Bioline), 25–30 ng of DNA
(quantified using a ND-1000 Nanodrop, PEQLAB Bio-
technology, Erlangen, Germany), 1.25 units of BioTaq
DNA polymerase (Bioline) and 25 pM each of the prim-
ers THIO714F and THIO1492R. PCRs were performed
in a Labcycler (SensoQuest, G€ottingen, Germany), with
an initial denaturation at 94 °C for 5 min, followed by
35 cycles of 94 °C for 1 min, 56 °C for 25 s, 72 °C for
2.5 min and a final extension at 72 °C for 5 min. PCR
products were checked on a 1% agarose gel. Bands of
the expected size (~800 bp) were excised and extracted
using the QIAquick Gel Extraction Kit (QIAGEN,
Hilden, Germany). PCR products were cloned and
sequenced as described previously (Bauermeister et al.
2012). Sequences were manually checked using Codon-
Code Aligner version 3.7.1.2 (CodonCode Corporation,
Dedham, MA, USA) and screened for chimeras using
Pintail version 1.0 (Ashelford et al. 2005). Putative chi-
meras were excluded from subsequent analyses. Using
MOTHUR (Schloss et al. 2009), a rarefaction curve was
created from all sequences obtained from the Movile
mat clone library in order to evaluate the number of
clones to be sequenced to cover the most abundant
Thiothrix species.
PCR detection of Thiothrix epibionts
DNA extracts obtained from the 71 amphipod speci-
mens (Table S1, Supporting information) were exam-
ined for Thiothrix DNA by direct PCR screenings. PCR
mixtures (10 lL) contained the same ingredients as
described for Thiothrix clone library construction.
Cycling conditions were as follows: initial denaturation
at 94 °C for 3 min, followed by 50 cycles of 94 °C for
45 s, 56 °C for 30 s and 72 °C for 1.5 min. PCR prod-
ucts were checked on a 1% agarose gel, and samples
revealing bands of the expected size (~800 bp) were
sequenced directly using the primers THIO718Fseq and
THIO1492R. In cases where mixtures of two overlap-
ping Thiothrix sequences were obtained from the same
PCR product, the individual sequences were resolved
by comparison with Thiothrix epibiont sequences from
the clone libraries. All sequences were assembled and
screened for chimeras as described above.
Phylogenetic analysis of Thiothrix sequences
Sequences obtained from 16S rRNA gene clone libraries
and direct PCR screenings were compared with
sequences in the public GenBank database using nucle-
otide BLAST (Altschul et al. 1990). 84 sequences (60 of
68 obtained from the clone libraries and 24 of 35
obtained by direct PCR) turned out to be closely related
to sequences of cultivated Thiothrix species and to
sequences previously obtained from Niphargus species
and microbial mats in Frasassi. 81 of these 84 sequences
(leaving out three redundant Thiothrix sequences from
samples AH_10.4 and SS_10.1 that were obtained in
both PCR screenings and clone libraries) were used for
phylogenetic analyses, together with 65 closely related
Thiothrix sequences downloaded from GenBank. All
sequences were aligned using the MAFFT version 6
multiple sequence alignment tool (Katoh & Toh 2010)
with the Q-INS-i strategy for consideration of RNA
secondary structure (Katoh & Toh 2008). The alignment
was manually refined, and a 50% consensus filter was
applied in MOTHUR (Schloss et al. 2009), resulting in
743 nucleotide positions used for phylogenetic analysis.
jModelTest version 0.1.1 (Posada 2008) was used to
determine the best-suited nucleotide model among 88
possible models following the Bayesian Information
Criterion. The selected model (TIM3+I+G) was used to
build a ML phylogenetic tree (1000 bootstrap replicates)
using PhyML 3.0 (Guindon & Gascuel 2003). The tree
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1408 J . -F . FLOT, J . BAUERMEISTER ET AL.
was rooted with an epibiont clone sequence from the
hydrothermal vent galatheid crab Shinkaia crosnieri
(GenBank accession number AB476284; Watsuji et al.
2010). In addition, neighbour-joining bootstrap values
for all nodes were calculated based on the same align-
ment using the BioNJ algorithm (Kimura 2-parameter
model; 1000 bootstrap replicates) implemented in Sea-
View version 4 (Gouy et al. 2010).
Results
Seven niphargid species were identified in Southern
Dobrogea
All 71 amphipod specimens yielded 28S rRNA gene
sequences (Table S1, Supporting information), whereas
sequencing of the ITS marker failed for one specimen
and the COI sequences of four of them were of bacterial
origin. Both the nuclear ITS and the mitochondrial COI
markers were congruent in delineating seven species
among our samples (Fig. 1); each of these species was
monophyletic and supported by bootstrap values >90%
using all three methods (maximum likelihood, distance
and parsimony). Although most species were also dis-
tinguishable using the 28S rRNA gene, two of them had
identical sequences and could not be resolved using this
marker (Fig. 2).
The seven niphargid species delineated molecularly
among our samples were further identified morpho-
logically. Six of them were already known from South-
ern Dobrogea: Niphargus cf. stygius (Schi€odte 1847),
Niphargus decui Karaman & Sarbu 1995, Niphargus dobr-
ogicus Dancau 1964, Niphargus gallicus Schellenberg
1935, Pontoniphargus racovitzai Dancau 1970 and Pon-
toniphargus ruffoi Karaman & Sarbu 1993. The seventh
species, Niphargus hrabei Karaman 1932, had never
been reported from southeastern Romania (Brad 1999).
To confirm our morphological identification of N.
hrabei, we compared its COI, ITS and 28S rRNA gene
sequences with those of one individual collected from
a side arm of the Danube River near Budapest, about
50 km away from its type locality. The sequences
obtained were very close (for COI) or identical (for
ITS and the 28S rRNA gene) to the ones from Dobro-
gea, despite a geographical distance of over 1000 km.
We were not able to obtain material from the type
locality of N. gallicus in France to verify whether it is
really the same species (as hypothesized by Dancau
1963).
Comparison of the 28S rRNA gene sequences
obtained from Southern Dobrogean niphargids with
previously published ones (Fig. 2) revealed that the
sequences of N. cf. stygius were markedly different from
those of the actual N. stygius from Slovenia, confirming
that these are distinct species. The 28S rRNA gene phy-
logeny did not support the putative sister-genus rela-
tionship between Niphargus and Pontoniphargus; instead,
P. racovitzai and P. ruffoi were nested within the genus
Niphargus and closely related to N. dobrogicus.
Thiothrix epibionts were detected on all niphargid
species from Southern Dobrogea
SEM revealed accumulations of filamentous bacteria on
two P. ruffoi specimens from Hagieni, on one N. cf. sty-
gius individual from Movile Cave and on two N. cf. sty-
gius specimens from sulphidic wells in Mangalia (Table
S2, Supporting information). The bacteria were found
attached predominantly to hairs and spines on legs and
antennae of the niphargids (Fig. 3). The Thiothrix-spe-
cific oligonucleotide probe G123T bound to filamentous
bacteria on appendages of all three niphargid individu-
als investigated using FISH (one P. ruffoi and two N.
cf. stygius).
Thiothrix-related partial 16S rRNA gene sequences
were obtained from 21 of the 71 amphipod DNA
extracts using PCR screenings with Thiothrix-specific
primers (Table S1, Supporting information). From three
of these 21 DNA extracts (samples SS_10.1, SS_10.2 and
JFF_12.39), mixtures of two overlapping Thiothrix
sequences were obtained, which were resolved by
comparison with Thiothrix epibiont sequences in clone
libraries obtained from P. ruffoi (AH_10.4) and N. cf.
stygius (SS_10.1). From 11 of the remaining 50 amphi-
pod samples, bands of the expected size (~800 bp) were
obtained on the agarose gel, but the top BLAST hits of
the corresponding sequences belonged to bacteria other
than Thiothrix (Table S1, Supporting information). The
clone library constructed from the Movile microbial mat
DNA yielded 52 Thiothrix sequences, which were used
for comparison with Thiothrix sequences obtained from
the niphargids.
The majority (86%) of Thiothrix epibiont sequences
obtained in this study fell into two clades (Fig. 4). One
of these clades (100% ML bootstrap support) contained
Thiothrix sequences from P. racovitzai, N. cf. stygius,
N. decui, N. gallicus and N. dobrogicus from both sulphi-
dic and nonsulphidic areas, and this clade was indis-
tinguishable from the T3 ectosymbiont clade of
Frasassi-dwelling Niphargus species. The other clade,
hereafter named T4, was supported by an 89% ML
bootstrap value and exclusively contained sequences
from individuals of P. ruffoi, P. racovitzai and N. cf.
stygius collected in sulphidic locations (Table 1). Four
remaining Thiothrix epibiont sequences obtained from
samples of N. gallicus, P. ruffoi and N. hrabei fell neither
within clade T3 nor T4, but either formed individual
branches in the Thiothrix tree or clustered with Thiothrix
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1409
. .
. .
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1410 J . -F . FLOT, J . BAUERMEISTER ET AL.
sequences from the Movile mat sample (this study) and
from Frasassi microbial mats (Macalady et al. 2006, 2008).
Discussion
Invertebrates harbouring ectosymbionts are common in
sulphidic marine habitats (Dubilier et al. 2008; Goffredi
2010). Well-known examples include different deep-sea
alvinocaridid shrimp (Tokuda et al. 2008; Petersen et al.
2010), hydrothermal vent crabs of the genus Kiwa (Goff-
redi et al. 2008; Thurber et al. 2011) and stilbonematid
nematodes dwelling in marine coastal sediments (Polz
et al. 1992; Bayer et al. 2009). The discovery of filamen-
tous, sulphur-oxidizing Thiothrix bacteria on Niphargus
amphipods living in the land-locked Frasassi caves
showed that ectosymbioses might also be prevalent in
sulphidic freshwater environments (Dattagupta et al.
2009). Here, we demonstrate that Niphargus–Thiothrix
associations are not restricted to Frasassi but are also
found in the partly sulphidic aquifers of the Southern
Dobrogea region of Romania. The regular presence of
Thiothrix bacteria on several geographically and phyloge-
netically distant members of the genus Niphargus makes
these relationships particularly suitable for evolutionary
studies. Moreover, as the genus Niphargus contains over
300 species distributed across Europe (Fiser et al. 2008),
it is possible that Niphargus–Thiothrix associations are
even more diverse than we have uncovered so far.
In this study, we used molecular and morphological
analyses to delineate the niphargid species inhabiting
sulphidic and nonsulphidic aquifers in Southern Dobro-
gea. The combination of ITS and COI sequence markers
proved sufficient to resolve five Niphargus and two
Pontoniphargus species with a high level of confidence
(Fig. 1) and to find their position in a phylogenetic tree
of niphargid amphipods. The less variable 28S rRNA
gene adjacent to ITS in ribosomal DNA was easier to
amplify and sequence consistently and could be used to
construct a rooted phylogeny of Niphargus (Fig. 2).
However, it was not variable enough to distinguish the
two closely related Pontoniphargus species. Contrary to
previous morphological hypotheses (Dancau 1970; Kar-
aman 1989; Karaman & Sarbu 1993), both Pontoniphar-
gus species turned out to be firmly nested within
Niphargus clade A (Fiser et al. 2008) and closely related
to N. dobrogicus. No new species was discovered, which
was surprising considering previous reports of rampant
cryptic species in amphipods in general and in the
genus Niphargus in particular (Lefebure et al. 2006, 2007;
Trontelj et al. 2009).
Direct PCR screenings revealed that individuals of all
seven niphargid species harboured Thiothrix epibionts,
most of which belonged to two distinct clades named
T3 and T4 (Fig. 4 and Table S1, Supporting informa-
tion). While T3 had previously been reported as an ec-
tosymbiont clade of three Niphargus species from the
Frasassi caves in central Italy (Bauermeister et al. 2012),
T4 is a new discovery. Both T3 and T4 were found each
on multiple individuals of different niphargid species,
and they were distinct from Thiothrix bacteria identified
in a Movile microbial mat sample (Fig. 4). Therefore, T3
and T4 may be regarded as putative ectosymbionts of
Southern Dobrogean niphargids pending future studies.
Four Thiothrix epibiont sequences clustered with Movile
and Frasassi microbial mat sequences instead of with
sequences of T3 or T4 (Fig. 4). Further, sequences
belonging to bacteria other than Thiothrix were obtained
from 11 of the examined niphargid samples (Table S1,
Supporting information). On the basis of the present
data, it is not possible to say whether these additional
bacteria represent yet unknown symbionts or free-living
bacteria that were loosely attached to the niphargid
individuals at the time of collection.
There are many similarities between the Niphargus–
Thiothrix epibioses found in Southern Dobrogea and
Frasassi. First, the Thiothrix filaments are attached to the
base of hairs on the amphipod appendages in both
cases (Fig. 3; cf. Dattagupta et al. 2009; Bauermeister
et al. 2012). Second, more than one Thiothrix clade
occurs on some Niphargus individuals (Table S1, Sup-
porting information). Third, clade T3 is present in both
sulphidic and nonsulphidic waters, whereas the other
clades (T4 in the case of Southern Dobrogea and T1-T2
in the case of Frasassi) are only abundant in sulphidic
locations (Bauermeister et al. 2012). Given that the
niphargids of Southern Dobrogea are not closely related
to the three species described from Frasassi (Fig. 2), the
parallels between the Niphargus–Thiothrix associations
found in these aquifers that are more than 1200 kilome-
tres apart from each other are particularly striking.
T3 Thiothrix were found to associate with Niphargus
in nonsulphidic locations in both Southern Dobrogea
Fig. 1 Top: Haploweb of ITS sequences of the 68 niphargid samples successfully sequenced for this marker. The two outgroups
AH_12.2 (Synurella sp.) and JFF_12.13 (Gammarus sp.) were not included in the alignment, as their sequences were too divergent. The
underlying phylogeny was obtained using a maximum-likelihood approach (model: GTR+G+I), following which connections were
added between the sequences found co-occurring in heterozygous individuals. Bottom: Haplotree of COI sequences of the 67 amphi-
pods samples successfully sequenced for this marker. The underlying phylogeny was obtained using a maximum-likelihood
approach (model: GTR+G+I). Both approaches delineated seven species (represented by different colours); bootstrap values obtained
from maximum-likelihood/parsimony/neighbour-joining are shown next to the name of each species.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1411
N. cf. tatrensis
N. tatrensis
N. virei
N. decui
N. pupetta
N. labacensis
N. aberrans
N. strouhali alpinus
N. cf. aquilex
N. longidactylus
N. pectinicauda
N. bajuvaricus
N. grandii
N. cf. tauri
N. wolfi
N. carniolicus
N. cf. tauri
N. aquilex dobati
N. cf. aquilex
N. cf. tauri
N. cf. tauri
N. cf. aquilex
N. dobrogicus
P. racovitzai + P. ruffoi
Clade A
from Fišer et al.
2008
N. kieferi
N. gallicus
N. laisi
N. ladmiraulti
N. delamarei
N. dimorphopus
N. kochianus
N. aquilex
N. aquilex
N. cf. longidactylus
N. schellenbergi
N. schellenbergi
 Synurella sp.
 Gammarus sp.
99
99
99
99
67
99
52
98
99
97
99
99
99
90
68
69
91
97
63
57
93
94
59
91
0.1
N. cf. stygius
N. montanarius
N. sp. clade 4
N. pachytelson
N. salonitanus
N. cf. salonitanus 
N. ictus
N. longiflagellum
N. dolichopus
N. subtypicus
N. stenopus
N. rejici
N. cf. abiter
N. abiter
Clade C
from Fišer et al. 2008
(without N. orcinus)
N. cf. tauri
Clade E from Fišer et al. 2008
N. cf. aquilex
N. cf. aquilex
N. sanctinaumi
N. maximus
N. costozzae
N. hrabei
N. plateaui
N. rhenorhodanensis
N. sp. Djevojacka
N. dolenianensis
N. sphagnicolus
N. tridentinus
N. lessiniensis
N. timavi
N. cf. longicaudatus
N. cf. longicaudatus
N. cf. longicaudatus
N. longicaudatus
N. cf. longicaudatus
N. frasassianus
N. pasquinii
N. auerbachi
N. rhenorhodanensis
N. lourensis
N. rhenorhodanensis
N. cf. fontanus
N. puteanus
N. kenki
N. rhenorhodanensis
N. rhenorhodanensis
N. orcinus
N. spinulifemur
N. stygius
N. sp.
N. cf. fontanus
N. fontanus
Clade B from Fišer et al. 2008
98
90
99
70
99
99
80
98
98
99
99
56
98
54
99
99
99
96
83
99
93
98
99
91
89
56
56
55
50
73
98
63
58
99
Fig. 2 Maximum-likelihood 28S rRNA gene phylogeny of the amphipods collected in the present study. This phylogeny includes
sequences from Lefebure et al. (2006, 2007), Fiser et al. (2008), Trontelj et al. (2009), Flot (2010), Flot et al. (2010a) and Hartke et al.
(2011). Filled arrows point at Niphargus sequences from the Frasassi caves in Italy, whereas empty arrows point at niphargid
sequences from the present study.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1412 J . -F . FLOT, J . BAUERMEISTER ET AL.
and Frasassi. Although Thiothrix are commonly consid-
ered sulphur-oxidizing bacteria, several heterotrophic
strains have been shown to grow in the absence of
reduced sulphur (Howarth et al. 1999; Aruga et al. 2002;
Chernousova et al. 2009). If T3 is capable of growth
without sulphide, it may be widely distributed on
niphargids throughout Europe. Our results provide
grounds for looking further into the distribution and
evolution of the Niphargus–T3 association.
Epibiotic Thiothrix filaments were previously identi-
fied on the marine amphipod Urothoe poseidonis living
in coastal sediments (Gillan & Dubilier 2004). Just as in
the case of Niphargus from Frasassi and Southern
Dobrogea, Thiothrix were found attached predomi-
nantly to hairs and spines of the legs of U. poseidonis.
Thus, it is possible that Thiothrix bacteria have a ten-
dency to associate with both freshwater and marine
amphipods and a preference to attach to their append-
ages. Whether this attachment location is advantageous
for either Thiothrix or their amphipod hosts is not yet
known. A previous study showed that the Thiothrix ec-
tosymbionts of Frasassi-dwelling Niphargus probably do
not play a role in sulphide detoxification for their hosts
(Bauermeister et al. 2013). In the present study, Thio-
thrix sequences were not amplified from DNA of 48 of
71 amphipods screened with PCR (Table S1, Support-
ing information), and Thiothrix filaments were missing
on 5 of 10 niphargid specimens examined by SEM
(Table S2, Supporting information). It seems unlikely
that all these apparently Thiothrix-lacking individuals
had lost their epibionts due to moulting. Instead, our
results suggest that the Thiothrix ectosymbionts may
not be obligate for their hosts. Whether the Thiothrix
bacteria benefit from associating with amphipods
remains to be investigated.
Conclusion
Six of the seven niphargid species identified in the
Southern Dobrogea region of Romania harbour putative
Thiothrix ectosymbionts of clades T3 and/or T4, which
are attached to their hosts in a manner similar to that
previously observed in the Frasassi caves in Italy. Thus,
Niphargus–Thiothrix associations are not restricted to
Frasassi, where they were first discovered, but may be
widespread in European sulphidic and nonsulphidic
aquifers.
Acknowledgements
Many thanks to Melanie Heinemann for logistic assistance in
the laboratory and to Cene Fiser, Radu Popa, Boris Sket and
Fabio Stoch for useful comments. We gratefully acknowledge
Mihai Baciu, Andreea Cohn, Vlad Voiculescu, Sandra Iepure
and Marjeta Konec for their help during fieldwork, the Grupul
de Explorari Subacvatice s
i Speologice (GESS) for logistic sup-
port in Mangalia, Colin Murrell and Jason Stephenson for
sending us a DNA extract from the Movile microbial mat they
collected, and Peter Borza for sending us specimens of Niphar-
gus hrabei from Hungary. We also thank three anonymous
reviewers for their valuable suggestions. This article is contri-
40 μm 40 μm
Thiothrix
filament
Gnathopod
(”grooming leg”) of
N. cf. stygius (JFF_12.33)
(b)
(d)(c)
(a)
50 μm 3 μm
Thiothrix
filaments
Pereopod
(”walking leg”)
of P. racovitzai
(JFF_12.19)
Seta of
N. cf. stygius
(JFF_12.33)
Filamentous
bacteria
Filamentous
bacteria
Gnathopod
(”grooming leg”) of
N. cf. stygius (SS_10.10)
Fig. 3 Thiothrix epibionts on niphargids from Southern Dobrogea. Panels a & b: Scanning electron micrographs (SEM) of filamentous
bacteria attached to hairs on the legs of two N. cf. stygius individuals. The filaments closely resemble Thiothrix bacteria previously
identified on Niphargus species from the Frasassi caves in Italy (Bauermeister et al. 2012). Panels c & d: Confocal epifluorescence
micrographs showing a Thiothrix-specific FISH probe (red) bound to bacterial filaments on legs of N. cf. stygius and P. racovitzai.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1413
N. frasassianus epibiont PC0810-10 (JN983605)
Thiothrix caldifontis (EU642573)
Frasassi microbial mat clone RS06101-B68 (EU101261)
Frasassi microbial mat clone CS102-B42 (JF747887)
N. gallicus (AH 12.3, Dulceşti) epibiont PCR sequence
N. ictus epibiont LV06NSB-B35 (EU884104)
N. ictus epibiont LV06NSB-B1 (EU884093)
N. ictus epibiont LC09102-21 (JN983572)
N. montanarius epibiont BUG083-97 (JN983563)
N. ictus epibiont LC09102-1 (JN983568)
N. montanarius epibiont BUG083-16 (JN983554)
N. montanarius epibiont BG08205-5 (JN983542)
N. montanarius epibiont BG08205-13 (JN983539)
Frasassi microbial mat clone WM44 (DQ133922)
N. frasassianus epibiont RS09103-3 (JN983601)
P. ruffoi (AH 10.4-29, Hagieni) epibiont clone sequence
N. frasassianus epibiont PC0810-47 (JN983603)
Movile Cave mat clone MC11MM1-57
Movile Cave mat clone MC11MM1-41
Frasassi microbial mat clone RS06101-B78 (EU101269)
N. montanarius epibiont BUG083-40 (JN983599)
Movile Cave mat clone MC11MM1-1
Frasassi microbial mat clone RS06101-B91 (EU101280)
Frasassi microbial mat clone ST09101-30 (JN983607)
Frasassi microbial mat clone ST09101-61 (JN983613)
Frasassi microbial mat clone ST09101-45 (JN983609)
Frasassi microbial mat clone RS06101-B4 (EU101239)
Frasassi microbial mat clone CS102B51 (JF747906)
Frasassi microbial mat clone ST09101-59 (JN983612)
Thiothrix fructosivorans (GU269554)
N. frasassianus epibiont RS09103-21 (JN983598)
Frasassi microbial mat clone SILK62 (EF467457)
N. frasassianus epibiont PC0810-2 (JN983602)
Frasassi microbial mat clone lka37 (EF467543)
N. frasassianus epibiont RS09103-8 (JN983597)
Frasassi microbial mat clone SILK69 (EF467473)
Frasassi microbial mat clone SILK45 (EF467508)
Thiothrix sp. CT3 (AF148516)
Thiothrix lacustris (EU642572)
Corroding concrete biofilm clone (AB255060)
N. cf. stygius (SS 10.2-2, Mangalia) epibiont PCR sequence
N. cf. stygius (SS 10.1-7, Mangalia) epibiont clone sequence
P. ruffoi (AH 10.4-4, Hagieni) epibiont clone sequence
P. ruffoi (AH 10.4-25, Hagieni) epibiont clone sequence
P. ruffoi (AH 10.4-18, Hagieni) epibiont clone sequence
P. ruffoi (AH 10.2, Hagieni) epibiont PCR sequence
P. ruffoi (AH 10.3, Hagieni) epibiont PCR sequence
P. ruffoi (AH 10.4-1, Hagieni) epibiont clone sequence
P. ruffoi (AH 10.5, Hagieni) epibiont PCR sequence
P. racovitzai (CF 10.1, Mangalia) epibiont PCR sequence
P. racovitzai (JFF 12.19, Mangalia) epibiont PCR sequence
N. cf. stygius (JFF 12.32, Mangalia) epibiont PCR sequence
N. cf. stygius (CF 10.4, Mangalia) epibiont PCR sequence
N. cf. stygius (JFF 12.39-2, Mangalia) epibiont PCR sequence
P. ruffoi (AH 10.4-38, Hagieni) epibiont clone sequence
N. cf. stygius (JFF 12.50, Mangalia) epibiont PCR sequence
Frasassi microbial mat clone ST09101-51 (JN983610)
Frasassi microbial mat clone RS06101-B48 (EU101246)
N. hrabei (SS 11.6, Hagieni) epibiont PCR sequence
N. frasassianus epibiont ST09103-6 (JN983592)
N. frasassianus epibiont RS08216-21 (JN983576)
N. frasassianus epibiont RS08216-24 (JN983579)
N. frasassianus epibiont RS08216-22 (JN983577)
N. frasassianus epibiont GB08207-15 (JN983566)
N. frasassianus epibiont ST09103-13 (JN983585)
N. frasassianus epibiont GB08207-19 (JN983567)
N. frasassianus epibiont RS09103-18 (JN983582)
Clade T4 Thiothrix
(This study)
Clade T1 Thiothrix
(Bauermeister et al.
2012)
Clade T2 Thiothrix
(Bauermeister et al.
2012)
50/60
78/80
89/93
98/98
0.02 Movile Cave mat clone MC11MM1-71
Movile Cave mat clone MC11MM1-59
Movile Cave mat clone MC11MM1-14
Movile Cave mat clone MC11MM1-26
Movile Cave mat clone MC11MM1-16
Movile Cave mat clone MC11MM1-8
Movile Cave mat clone MC11MM1-67
Movile Cave mat clone MC11MM1-49
Thiothrix unzii (NR_044655) 
Edwards-Trinity Aquifer biofilm clone EAP-SON8-WH-19B (FJ165521)
Movile Cave mat clone MC11MM1-65
Movile Cave mat clone MC11MM1-51
Movile Cave mat clone MC11MM1-43
Movile Cave mat clone MC11MM1-20
Movile Cave mat clone MC11MM1-54
Cesspool cave microbial mat clone CC1B-16S-69 (EU662339)
Movile Cave mat clone MC11MM1-19
Movile Cave mat clone MC11MM1-22
Movile Cave mat clone MC11MM1-38
Frasassi microbial mat clone SILK46 (JN983605)
Movile Cave mat clone MC11MM1-68
Movile Cave mat clone MC11MM1-66
Movile Cave mat clone MC11MM1-35
Movile Cave mat clone MC11MM1-36
Movile Cave mat clone MC11MM1-15
Movile Cave mat clone MC11MM1-12
Movile Cave mat clone MC11MM1-56
Movile Cave mat clone MC11MM1-42
Movile Cave mat clone MC11MM1-13
Movile Cave mat clone MC11MM1-23
Movile Cave mat clone MC11MM1-33
Movile Cave mat clone MC11MM1-64
Movile Cave mat clone MC11MM1-47
Movile Cave mat clone MC11MM1-62
Movile Cave mat clone MC11MM1-40
Movile Cave mat clone MC11MM1-73
Movile Cave mat clone MC11MM1-48
Movile Cave mat clone MC11MM1-10
Movile Cave mat clone MC11MM1-30
Movile Cave mat clone MC11MM1-25
Movile Cave mat clone MC11MM1-31
97/100
85/99
Edwards-Trinity Aquifer biofilm clone EAP-SON8-WH-38B (FJ165513)
Movile Cave mat clone MC11MM1-39
N. gallicus (TB 12.3, Mangalia) epibiont PCR sequence
Movile Cave mat clone MC11MM1-7477/75
56/-
67/63
85/81 100/100
52/52
52/60
100/99
Movile Cave mat clone MC11MM1-53
Movile Cave mat clone MC11MM1-37
Movile Cave mat clone MC11MM1-72
Thiothrix nivea (L40993)
Movile Cave mat clone MC11MM1-44
Movile Cave mat clone MC11MM1-69
Movile Cave mat clone MC11MM1-21
Movile Cave mat clone MC11MM1-7
Marine amphipod Urothoe poseidonis epibiont (AY426613)
Thiothrix flexilis (NR_024757)
Thiothrix disciformis (NR_024756)
Thiothrix eikelboomii (NR_024758)
Wastewater activated sludge clone AS99 (EU283396)
Activated sludge clone N1512-84 (EU104250)
Wastewater bacterial community clone OTU-28-AW (JQ624269)
N. montanarius epibiont BG08205-11 (JN983537)
N. cf. stygius (SS 10.2-1, Mangalia) epibiont PCR sequence
N. frasassianus epibiont RS09103-5 (JN983583)
N. frasassianus epibiont ST09103-28 (JN983590)
N. montanarius epibiont BUG083-46 (JN983556)
N. montanarius epibiont BUG083-93 (JN983561)
N. cf. stygius (SS 11.1, Mangalia) epibiont PCR sequence
N. cf. stygius (SS 10.1-2, Mangalia) epibiont clone sequence
N. cf. stygius (JFF 12.39-1, Mangalia) epibiont PCR sequence
N. gallicus (TB 11.6, Mangalia) epibiont PCR sequence
N. frasassianus epibiont ST09103-9 (JN983594)
N. dobrogicus (SS 10.7, Doi Mai) epibiont PCR sequence
N. decui (SS 10.8, Doi Mai) epibiont PCR sequence
N. frasassianus epibiont PC0810-9 (JN983574)
P. racovitzai (JFF 12.46, Mangalia) epibiont PCR sequence
N. cf. stygius (JFF 12.31, Mangalia) epibiont PCR sequence
*Hydrothermal vent galatheid crab Shinkaia crosnieri (AB476284)
Clade T3 Thiothrix
(This study &
Bauermeister et al. 2012)
66/-
61/82
98/100
89/88
100/100
100/100
66/67
Movile Cave mat clone MC11MM1-52100/100
Movile Cave mat clone MC11MM1-2
Movile Cave mat clone MC11MM1-18
75/82
56/-
62/63
-/66
-/67
56/65
Outgroup*
Fig. 4 Maximum-likelihood 16S rRNA gene phylogeny of Thiothrix. Sequences obtained from Southern Dobrogean niphargid samples
are in red, those contained in the Movile mat clone library in blue. Cultivated Thiothrix strains are in bold. Accession numbers of
sequences downloaded from GenBank are given in parentheses. Maximum-likelihood/neighbour-joining bootstrap values greater
than 50 are displayed next to the respective nodes.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1414 J . -F . FLOT, J . BAUERMEISTER ET AL.
bution number 122 from the Courant Research Center Geobiol-
ogy funded by the German Initiative of Excellence.
References
Altschul SF, Gish W, Miller W, Meyers EW, Lipman DJ (1990)
Basic local alignment search tool. Journal of Molecular Biology,
215, 403–410.
Amann RI (1995) In situ identification of microorganisms by
whole cell hybridization with rRNA-targeted nucleic acid
probes. In: Molecular Microbial Ecology Manual (eds Akker-
man ADL, van Elsas DJ, de Bruijn FJ), pp. 1–15, Kluwer
Academic Publishers, Dortrecht.
Aruga S, Kamagata Y, Kohno T, Hanada S, Nakamura K, Kanag-
awa T (2002) Characterization of filamentous Eikelboom type
021N bacteria and description of Thiothrix disciformis sp. nov.
and Thiothrix flexilis sp. nov. International Journal of Systematic
and Evolutionary Microbiology, 52, 1309–1316.
Ashelford KE, Weightman AJ, Fry JC (2002) PRIMROSE: a
computer program for generating and estimating the phylo-
genetic range of 16S rRNA oligonucleotide probes and prim-
ers in conjunction with the RDP-II database. Nucleic Acids
Research, 30, 3481–3489.
Ashelford KE, Chuzhanova NA, Fry JC, Jones AJ, Weightman
AJ (2005) At least 1 in 20 16S rRNA sequence records cur-
rently held in public repositories is estimated to contain sub-
stantial anomalies. Applied and Environmental Microbiology, 71,
7724–7736.
Bagarinao T (1992) Sulfide as an environmental factor and toxi-
cant: tolerance and adaptations in aquatic organisms. Aquatic
Toxicology, 24, 21–62.
Bauermeister J, Ramette A, Dattagupta S (2012) Repeatedly
evolved host-specific ectosymbioses between sulfur-oxidizing
bacteria and amphipods living in a cave ecosystem. PLoS
ONE, 7, e50254. doi:10.1371/journal.pone.0050254.
Bauermeister J, Assig K, Dattagupta S (2013) Exploring the
sulfide tolerance of ectosymbiotic Niphargus amphipods from
the Frasassi caves, central Italy. International Journal of Speleol-
ogy, 42, 141–145.
Bayer C, Heindl HR, Rinke C, Lu
̈
cker S, Ott JA, Bulgheresi S
(2009) Molecular characterization of the symbionts associated
with marine nematodes of the genus Robbea. Environmental
Microbiology Reports, 1, 136–144.
Brad T (1999) The present stage of our knowledge concerning
the spreading of subterranean amphipods in Romania. Studii
Şi Cercetări (Biologie) Bistriţa, 5, 157–164.
Chernousova E, Gridneva E, Grabovich M et al. (2009) Thiothrix
caldifontis sp. nov. and Thiothrix lacustris sp. nov., gammapro-
teobacteria isolated from sulfide springs. International Journal
of Systematic and Evolutionary Microbiology, 59, 3128–3135.
Clark A (1990) Inference of haplotypes from PCR-amplified
samples of diploid populations. Molecular Biology and Evolu-
tion, 7, 111–122.
Dancau D (1963) Niphargus gallicus Schell., amfipod subteran
nou pentru fauna R.P.R. Comunicarile Academiei Republicii
Populare Roma^ne, 13, 123–129.
Dancau D (1964) Noi contribut
ii la studiul amfipodelor subter-
ane Niphargus dobrogicus n. sp.. Lucrarile Institutului de speolo-
gie “Emil Racovit
a”, 3, 397–403.
Dancau D (1970) Sur un nouvel amphipode souterrain de
Roumanie, Pontoniphargus racovitzai, n. g., n. sp.. In: Livre du
centenaire. Emile G. Racovitza 1868–1968 (eds Orghidan T,
Dumitresco M), pp. 275–285, Academie de la Republique
socialiste de Roumanie, Bucarest.
Dattagupta S, Schaperdoth I, Montanari A et al. (2009) A novel
symbiosis between chemoautotrophic bacteria and a fresh-
water cave amphipod. The ISME Journal, 3, 935–943.
Dubilier N, Bergin C, Lott C (2008) Symbiotic diversity in mar-
ine animals: the art of harnessing chemosynthesis. Nature
Reviews Microbiology, 6, 725–740.
Felsenstein J (1985) Confidence limits on phylogenies: an
approach using the bootstrap. Evolution, 39, 783–791.
Fiser C, Sket B, Trontelj P (2008) A phylogenetic perspective on
160 years of troubled taxonomy of Niphargus (Crustacea:
Amphipoda). Zoologica Scripta, 37, 665–680.
Flot J-F (2007) Champuru 1.0: a computer software for unravel-
ing mixtures of two DNA sequences of unequal lengths.
Molecular Ecology Notes, 7, 974–977.
Flot J-F (2010) Vers une taxonomie moleculaire des amphipodes
du genre Niphargus: exemples d’utilisation de sequences
d’ADN pour l’identification des especes. Bulletin de la Societe
des Sciences Naturelles de l’Ouest de la France, 32, 62–68.
Flot J-F, Tillier A, Samadi S, Tillier S (2006) Phase determina-
tion from direct sequencing of length-variable DNA regions.
Molecular Ecology Notes, 6, 627–630.
Flot J-F, W€orheide G, Dattagupta S (2010a) Unsuspected diver-
sity of Niphargus amphipods in the chemoautotrophic cave
ecosystem of Frasassi, central Italy. BMC Evolutionary Biology,
10, 171.
Flot J-F, Couloux A, Tillier S (2010b) Haplowebs as a graphical
tool for delimiting species: a revival of Doyle’s “field for recom-
bination” approach and its application to the coral genus Pocil-
lopora in Clipperton. BMC Evolutionary Biology, 10, 372.
Flot J-F, Blanchot J, Charpy L et al. (2011) Incongruence
between morphotypes and genetically delimited species in
the coral genus Stylophora: phenotypic plasticity, morphologi-
cal convergence, morphological stasis or interspecific hybrid-
ization? BMC Ecology, 11, 22.
Forti P, Galdenzi S, Sarbu S (2002) The hypogenic caves: a
powerful tool for the study of seeps and their environmental
effects. Continental Shelf Research, 22, 2373–2386.
Gillan DC, Dubilier N (2004) Novel epibiotic Thiothrix bacte-
rium on a marine amphipod. Applied and Environmental
Microbiology, 70, 3772–3775.
Goffredi SK (2010) Indigenous ectosymbiotic bacteria associ-
ated with diverse hydrothermal vent invertebrates. Environ-
mental Microbiology Reports, 2, 479–488.
Goffredi SK, Jones WJ, Erhlich H, Springer A, Vrijenhoek RC
(2008) Epibiotic bacteria associated with the recently discov-
ered Yeti crab, Kiwa hirsuta. Environmental Microbiology, 10,
2623–2634.
Gottstein S, Ivkovic M, Ternjej I, Jalzic B, Kerovec M (2007)
Environmental features and crustacean community of an-
chihaline hypogean waters on the Kornati islands, Croatia.
Marine Ecology, 28, 24–30.
Gouy M, Guindon S, Gascuel O (2010) SeaView version 4: a
multiplatform graphical user interface for sequence align-
ment and phylogenetic tree building. Molecular Biology and
Evolution, 27, 221–224.
Guindon S, Gascuel O (2003) A simple, fast, and accurate algo-
rithm to estimate large phylogenies by maximum likelihood.
Systematic Biology, 52, 696–704.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1415
Hartke TR, Fiser C, Hohagen J, Kleber S, Hartmann R, Koene-
mann S (2011) Morphological and molecular analyses of
closely related species in the stygobiontic genus Niphar-
gus (Amphipoda). Journal of Crustacean Biology, 31, 701–
709.
Howarth R, Unz RF, Seviour EM et al. (1999) Phylogenetic rela-
tionships of filamentous sulfur bacteria (Thiothrix spp. and
Eikelboom type 021N bacteria) isolated from wastewater-
treatment plants and description of Thiothrix eikelboomii
sp. nov., Thiothrix unzii sp. nov., Thiothrix fructosivorans sp.
nov. and Thiothrix defluvii sp. nov. International Journal of
Systematic Bacteriology, 49, 1817–1827.
Kanagawa T, Kamagata Y, Aruga S, Kohno T, Horn M, Wag-
ner M (2000) Phylogenetic analysis of and oligonucleotide
probe development for Eikelboom Type 021N filamentous
bacteria isolated from bulking activated sludge. Applied and
Environmental Microbiology, 66, 5043–5052.
Karaman S (1932) Beitrag zur Kenntnis der S€usswasser-
Amphipoden. Prirodoslovne Razprave, 2, 179–232.
Karaman GS (1989) New data on genus Pontoniphargus Dancau,
1970 (fam. Niphargidae) from Romania (Contribution to the
knowledge of the Amphipoda 1999). Glasnik Republickog
Zavoda za Zastitu Prirode i Prirodnjackog Muzeja u Titogradu,
22, 79–94.
Karaman GS, Sarbu S (1993) A new species of the genus
Pontoniphargus Dancau, 1970 (Amphipoda Gammaridea,
family Niphargidae) from Romania, P. ruffoi n. sp. Bollettino
del Museo Civico di Storia Naturale di Verona, 20, 569–582.
Karaman GS, Sarbu SM (1995) Niphargus decui n. sp. (Amphi-
poda, Gammaridea, Niphargidae), a new species from
Romania. Travaux de l’Institut de Speologie Emile Racovitza, 34,
77–87.
Katoh K, Toh H (2008) Improved accuracy of multiple ncRNA
alignment by incorporating structural information into a
MAFFT-based framework. BMC Bioinformatics, 9, 212.
Katoh K, Toh H (2010) Parallelization of the MAFFT multi-
ple sequence alignment program. Bioinformatics, 26, 1899–
1900.
Katoh K, Misawa K, Kuma K, Miyata T (2002) MAFFT: a novel
method for rapid multiple sequence alignment based on fast
Fourier transform. Nucleic Acids Research, 30, 3059–3066.
Latella L, Di Russo C, de Pasquale L, Dell’Anna L, Nardi G,
Rampini M (1999) Preliminary investigations on a new
sulfurous cave in central Italy. Memoires de Biospeologie, 26,
131–1335.
Lefebure T, Douady CJ, Gouy M, Trontelj P, Briolay J, Gibert J
(2006) Phylogeography of a subterranean amphipod reveals
cryptic diversity and dynamic evolution in extreme environ-
ments. Molecular Ecology, 15, 1797–1806.
Lefebure T, Douady CJ, Malard F, Gibert J (2007) Testing dis-
persal and cryptic diversity in a widely distributed ground-
water amphipod (Niphargus rhenorhodanensis). Molecular
Phylogenetics and Evolution, 42, 676–686.
Macalady JL, Lyon EH, Koffman B et al. (2006) Dominant
microbial populations in limestone-corroding stream bio-
films, Frasassi cave system, Italy. Applied and Environmental
Microbiology, 72, 5596–5609.
Macalady JL, Dattagupta S, Schaperdoth I, Jones DS, Druschel
G, Eastman DK (2008) Niche differentiation among sulfur-
oxidizing bacterial populations in cave waters. The ISME
Journal, 2, 590–601.
Neufeld JD, Schafer H, Cox MJ, Boden R, McDonald IR,
Murrell JC (2007) Stable-isotope probing implicates Methyl-
ophaga spp and novel Gammaproteobacteria in marine
methanol and methylamine metabolism. The ISME Journal, 1,
480–491.
Oseid DM, Smith LL Jr (1974) Chronic toxicity of hydrogen sul-
fide to Gammarus pseudolimnaeus. Transactions of the American
Fisheries Society, 103, 819–822.
Petersen JM, Ramette A, Lott C, Cambon-Bonavita M-A,
Zbinden M, Dubilier N (2010) Dual symbiosis of the vent
shrimp Rimicaris exoculata with filamentous gamma- and
epsilonproteobacteria at four Mid-Atlantic Ridge hydro-
thermal vent fields. Environmental Microbiology, 12, 2204–
2218.
Polz MF, Felbeck H, Novak R, Nebelsick M, Ott JA (1992)
Chemoautotrophic, sulfur-oxidizing symbiotic bacteria on
marine nematodes: morphological and biochemical charac-
terization. Microbial Ecology, 24, 313–329.
Posada D (2008) jModelTest: phylogenetic model averaging.
Molecular Biology and Evolution, 25, 1253–1256.
Sandberg-Kilpi E, Vismann B, Hagerman L (1999) Tolerance of
the Baltic amphipod Monoporeia affinis to hypoxia, anoxia
and hydrogen sulfide. Ophelia, 50, 61–68.
Sarbu SM (2000) Movile Cave: a chemoautotrophically based
groundwater ecosystem. In: Ecosystems of the World. Subterra-
nean Ecosystems(eds Wilkens H, Culver DC & Humphreys
WF), pp. 319–343. Elsevier, Amsterdam.
Sarbu SM, Popa R (1992) A unique chemoautotrophically based
cave ecosystem. In: The Natural History of Biospeleology(ed
Camacho AI), pp. 637–666. Museo Nacional de Ciencias
Naturales, Madrid.
Sarbu SM, Kinkle BK, Vlasceanu L, Kane TC, Popa R (1994)
Microbiological characterisation of a sulfide-rich groundwa-
ter ecosystem. Geomicrobiology Journal, 12, 175–182.
Sarbu SM, Kane TC, Kinkle BK (1996) A chemoautotrophically
based cave ecosystem. Science, 272, 1953–1955.
Schellenberg A (1935) Schl€ussel des Amphipodengattung
Niphargus mit Fundortangaben und mehreren neuen Formen.
Zoologischer Anzeiger, 111, 204–211.
Schi€odte JC (1847) Unders€ogelser over den underjordiske
Fauna i Hulerne i Krain og Istrien. Oversigt over det Kongelige
danske Videnskabernes Selskabs Forhandlinger og dets Medlem-
mers Arbeider i Aaret, 1847, 75–82.
Schloss PD, Westcott SL, Ryabin T et al. (2009) Introducing
mothur: open-source, platform-independent, community-
supported software for describing and comparing microbial
communities. Applied and Environmental Microbiology, 75,
7537–7541.
Sket B (1996) The ecology of anchihaline caves. Trends in
Ecology & Evolution, 11, 221–225.
Tamura K, Peterson D, Peterson N et al. (2011) MEGA5: Mole-
cular Evolutionary Genetics Analysis using maximum likeli-
hood, evolutionary distance, and maximum parsimony
methods. Molecular Biology and Evolution, 28, 2731–2739.
Theede H, Ponat A, Hiroki K, Schlieper C (1969) Studies on
the resistance of marine bottom invertebrates to oxygen-
deficiency and hydrogen sulphide. Marine Biology, 2, 325–
337.
Thurber AR, Jones WJ, Schnabel K (2011) Dancing for food in
the deep sea: bacterial farming by a new species of Yeti crab.
PLoS ONE, 6, e26243. doi:10.1371/journal.pone.0026243.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
1416 J . -F . FLOT, J . BAUERMEISTER ET AL.
Tokuda G, Yamada A, Nakano K, Arita NO, Yamasaki H
(2008) Colonization of Sulfurovum sp. on the gill surfaces of
Alvinocaris longirostris, a deep-sea hydrothermal vent shrimp.
Marine Ecology, 29, 106–114.
Trontelj P, Douady CJ, Fiser C et al. (2009) A molecular test for
cryptic diversity in ground water: how large are the ranges
of macro-stygobionts? Freshwater Biology, 54, 727–744.
Watsuji T, Nakagawa S, Tsuchida S et al. (2010) Diversity and
function of epibiotic microbial communities on the galatheid
crab, Shinkaia crosnieri. Microbes and Environments, 25, 288–
294.
Sample collections and geochemical measurements:
J.F.F., T.B., A.H.V., S.M.S., S.D.; DNA extractions and
molecular analyses: J.F.F., J.B.; morphological identifica-
tions: J.F.F.; SEM and FISH: J.B.; manuscript drafting:
J.F.F., J.B. and S.D.; manuscript revision: T.B., A.H.V.
and S.M.S.
Data accessibility
All DNA sequences obtained in the present study were
deposited in GenBank (accession numbers KF290023-
KF290376 and KF362044-KF362049). Alignments were
deposited in Dryad (doi:10.5061/dryad.qm636).
Supporting information
Additional supporting information may be found in the online
version of this article.
Table S1 Niphargid samples analysed molecularly in this study.
Samples are colour-coded according to niphargid species.
Table S2 Niphargid samples analysed by scanning electron
microscopy (SEM). Samples are colour-coded according to
niphargid species.
© 2013 The Authors. Molecular Ecology Published by John Wiley & Sons Ltd.
NIPHARGUS–THIOTHRIX ASSOCIATIONS IN ROMANIA 1417