Exploring the molecular diversity of terrestrial nemerteans (Hoplonemertea, Monostilifera, Acteonemertidae) in a continental landmass

Abstract

We report the finding of terrestrial nemerteans of the family Acteonemertidae from the northern part of Spain. The specimens were studied using molecular data from the nuclear ribosomal 18S rRNA and the mitochondrial cytochrome c oxidase subunit I genes. The 18S rRNA data strongly suggest the presence of three species of terrestrial nemerteans in the Iberian Peninsula, an old landmass where terrestrial nemerteans had not previously been reported. These specimens originate from primarily undisturbed forests across a distance of c. 1000 km. Cytochrome c oxidase subunit I data also indicate the presence of multiple lineages of Iberian terrestrial nemerteans. Nonetheless, the pattern obtained from this marker is obscured most probably by deep genetic divergences. This molecular diversity, at least in some of the clades, suggests that the Iberian species are not the result of recent introductions, as proposed for other terrestrial nemerteans found in Europe. Our data also touch upon the question of a single origin of terrestriality in nemerteans, a hypothesis rejected by both data sets. Nevertheless, terrestriality seems to have had a single origin in the family Acteonemertidae.

Full text

© 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters • Zoologica Scripta,
37
, 3, May 2008, pp 235–243
235
Mateos, E. & Giribet, G. (2008). Exploring the molecular diversity of terrestrial nemerteans
(Hoplonemertea, Monostilifera, Acteonemertidae) in a continental landmass. —
Zoologica
Scripta, 37
, 235–243.
We report the finding of terrestrial nemerteans of the family Acteonemertidae from the
northern part of Spain. The specimens were studied using molecular data from the nuclear
ribosomal 18S rRNA and the mitochondrial cytochrome
c
oxidase subunit I genes. The 18S
rRNA data strongly suggest the presence of three species of terrestrial nemerteans in the
Iberian Peninsula, an old landmass where terrestrial nemerteans had not previously been
reported. These specimens originate from primarily undisturbed forests across a distance of
c
. 1000 km. Cytochrome
c
oxidase subunit I data also indicate the presence of multiple lineages
of Iberian terrestrial nemerteans. Nonetheless, the pattern obtained from this marker is
obscured most probably by deep genetic divergences. This molecular diversity, at least in some
of the clades, suggests that the Iberian species are not the result of recent introductions,
as proposed for other terrestrial nemerteans found in Europe. Our data also touch upon the
question of a single origin of terrestriality in nemerteans, a hypothesis rejected by both data
sets. Nevertheless, terrestriality seems to have had a single origin in the family Acteonemertidae.
Corresponding author:
Gonzalo Giribet, Department of Organismic and Evolutionary Biology
and Museum of Comparative Zoology, 26 Oxford Street, Cambridge, MA 02138, USA. E-mail:
ggiribet@oeb.harvard.edu
Eduardo Mateos, Departament de Biologia Animal, Facultat de Biologia, Universitat de Barcelona,
Avenue. Diagonal 645, 08028 Barcelona, Spain. E-mail: emateos@ub.edu
Blackwell Publishing Ltd
Exploring the molecular diversity of terrestrial nemerteans
(Hoplonemertea, Monostilifera, Acteonemertidae) in a
continental landmass
E
DUARDO
M
ATEOS
& G
ONZALO
G
IRIBET
Submitted: 29 October 2007
Accepted: 13 December 2007
doi:10.1111/j.1463-6409.2008.00324.x
Introduction
Although a ubiquitous group of animals inhabiting marine,
limnic and terrestrial ecosystems, the members of the phylum
Nemertea are among the most poorly characterized inverte-
brates. One major problem in species delimitation is their
plastic body and our lack of understanding of their phenotypic
plasticity and convergent colouration patterns (e.g. Strand &
Sundberg 2005a,b). Another key aspect is the small size and
cryptic lifestyle of most of the members of the phylum, often
of nocturnal habits. These problems are even more exacerbated
in the dull-coloured, seldom-studied, terrestrial and limnic forms.
Recent morphological cladistic analyses of polystiliferous
nemerteans (Maslakova & Norenburg 2001) and heterone-
merteans (Schwartz & Norenburg 2001) lacked resolution,
perhaps due to imprecise and inadequate taxonomic descriptions
of species and their morphological features. This is obviously
a consequence of a limited number of characters and
extensive homoplasy (e.g. see Schwartz & Norenburg 2001).
Therefore, during the past decade, phylogenetic studies of
nemerteans have been mostly undertaken from a molecular
point of view (Sundberg & Saur 1998; Sundberg
et al
. 2001;
Turbeville 2002; Thollesson & Norenburg 2003; Strand &
Sundberg 2005a,b). But to date, molecular data on terrestrial
nemerteans remain unexplored.
Terrestrial nemerteans are known from many localities
around the globe, although they are mostly confined to
islands, including several in the Pacific and Indian oceans, as
well as the Caribbean, Bahamas and Mascarene Islands
(Gibson & Moore 1998; Moore
et al
. 2001). Until 1969, all
terrestrial nemerteans had been placed in the genus
Geonemertes
Semper, 1863, which was considered related to the supralittoral
genus
Acteonemertes
Pantin, 1961. Pantin (1969) established
several groups of species, including an Australian group
(
Geonemertes dendyi
,
G. australiensis
and
G. hilii
) and the
pelaensis
group (with
G. pelaensis
,
G. arboricola
and
G. rodericana
), among
others. Moore & Gibson (1981) revised the genus
Geonemertes
and established the modern genera of terrestrial nemerteans,
standardizing the anatomical characters that diagnose the

Diversity of Iberian terrestrial nemerteans
•
E. Mateos & G. Giribet
236
Zoologica Scripta,
37
, 3, May 2008, pp 235–243 • © 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters
genera, and establishing the new genera
Antiponemertes
Moore
& Gibson, 1981,
Argonemertes
Moore & Gibson, 1981,
Katechonemertes
Moore & Gibson, 1981 and
Pantinonemertes
Moore & Gibson, 1981. They also recognized two additional
genera,
Acteonemertes
Pantin, 1961 and
Leptonemertes
Girard,
1893. The genus
Geonemertes
was therefore restricted to the
species
G. pelaensis
Semper, 1863 and
G. rodericana
(Gulliver,
1879). It was not until later that the genera were grouped into
higher categories, establishing a Group I comprised of
Geonemertes
and
Pantinonemertes
; and a Group II that includes
Acteonemertes
,
Argonemertes
,
Antiponemertes
,
Katechonemertes
and
Leptonemertes
(Gibson & Moore 1985; Moore & Gibson
1988b). Members of Group I are nowadays classified in the
family Prosorhochmidae Bürger, 1895, which also included
the genera
Prosorhochmus
Keferstein, 1862 and
Prosadenophorus
Bürger, 1890, and the relationships amongst its species have
been investigated by Sundberg (1989c). Members of the
Group II were first included in the family Plectonemertidae
Gibson, 1990, along with the freshwater genera
Campbellone-
mertes
Moore & Gibson, 1972 and
Potamonemertes
Moore &
Gibson, 1973 and with the marine genus
Plectonemertes
Gibson, 1990 (Moore & Gibson 1988a; Sundberg 1989a,b;
Gibson 1990).
A subsequent morphological cladistic analysis of monos-
tiliferous hoplonemerteans (Crandall 2001) suggests that
Plectonemertes
represents a monotypic family, and that the
five ‘Group II’ terrestrial genera (
Acteonemertes
,
Argonemertes
,
Antiponemertes
,
Leptonemertes
and
Katechonemertes
) form a
distinct family, separated from the genera
Campbellonemertes
and
Potamonemertes
, which represent either a single distinct
family or two monotypic families. Later, Chernishev (2005)
erected the family Acteonemertidae Chernishev, 2005 for the
five genera constituting Group II.
To date 13 species of fully terrestrial nemerteans are
recognized, although some upper littoral species can also be
found in fully terrestrial environments (Gibson
et al
. 1982;
Moore
et al
. 2001). Interestingly, most records of terrestrial
nemerteans are from islands (e.g. Hickman 1963; Anderson
1980; Howarth & Moore 1983; Moore & Gibson 1985;
Gibson & Moore 1998), both oceanic and continental in
origin. Some of the terrestrial nemertean species have been
believed to have become extinct (Moore
et al
. 2001) due to
anthropomorphic effects such as habitat loss. Another
anthropomorphic effect is the involuntary introduction of
species into new environments, something that has also been
said for terrestrial nemerteans.
In contrast with the number of detailed morphological
analyses that have considered terrestrial nemerteans
(Sundberg 1989a,b,c; Crandall 2001), molecular data are
scarce. The first molecular sequences for a terrestrial nemertean,
Argonemertes australiensis
, were published by Maxmen
et al
.
(2003), although these sequences were not employed in the
context of nemertean systematics. In their molecular study of
the genus
Tetrastemma
based on 18S rRNA sequence data,
Strand & Sundberg (2005b) used sequences of three land
nemertean specimens, two of
A. australiensis
(one from the
study of Maxmen
et al
. 2003) and one of
Antiponemertes
novaezealandiae
, which formed a monophyletic group —
although
A. australiensis
was not monophyletic. They do not
group with the genus
Prosorhochmus
either. The genus
Pantinonemertes
was also included in another molecular
analysis based on four markers (the nuclear 28S rRNA,
histone H3 and the mitochondrial 16S rRNA and cytochrome
c
oxidase subunit I) (Thollesson & Norenburg 2003), but did
not group with the genus
Oerstedia
, then considered a
member of the family Prosorhochmidae (but see Chernishev
2005). The same species was used for a broader analysis of
metazoans employing the novel sodium–potassium ATPase
α
-subunit gene (Anderson
et al
. 2004). No members of
Groups I and II have yet been analysed together in a molecular
analysis of nemerteans, and therefore the hypothesis of their
unique origin (Pantin 1969) — indicating a single colonization
event of land — remains untested.
In this paper, we apply a molecular approach to investigate
the species delimitations and relationships among multiple
lineages of terrestrial nemerteans collected around the world,
but placing emphasis on several populations of Iberian
nemerteans, a diverse region and one of the so-called 25
biodiversity hotspots of the World (Myers
et al
. 2000), but
from which terrestrial nemerteans were previously unknown.
Materials and methods
Specimens
Live specimens of terrestrial nemerteans were collected
from seven Spanish localities, photographed, and preserved
immediately in 96% EtOH (Table 1 and Fig. 1). Although we
kindly received specimens of
Pantinonemertes
sp. and
A. australiensis
from Per Sundberg, we were not able to amplify their DNA
and therefore in our analyses the genus
Pantinonemertes
is
limited to a GenBank sequence. Terrestrial nemerteans of
Group I (family Prosorhochmidae) are therefore limited to a
Pantinonemertes
cytochrome
c
oxidase subunit I sequence and
to a sample of
G. pelaensis
from Bermuda, kindly provided
by Wolfgang Sterrer. Sampling within Group II (family
Acteonemertidae) is broader, although representatives of the
two monospecific genera
Acteonemertes
and
Katechonemertes
were not studied, and our sequences of
Antiponemertes
are
restricted again to an 18S rRNA sequence from GenBank for
which there is no COI counterpart (Table 2).
Outgroup species were selected among other monostiliferan
hoplonemerteans, including some freshwater species. The
heteronemertean
Lineus bilineatus
was used to root the
hoplonemertean tree. It is however, important to stress that
the sampling was not designed to study hoplonemertean

E. Mateos & G. Giribet
•
Diversity of Iberian terrestrial nemerteans
© 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters • Zoologica Scripta,
37
, 3, May 2008, pp 235–243
237
Table 1 Localities for the Iberian terrestrial nemerteans sampled in Spain (see Fig. 1).
Code Locality Coordinates Vegetation cover Altitude (m)
Loc1 Fragas do Eume Natural Park, Pontedeume, A Coruña 43°25′1.9′′ N, 8°3′48.8′′ W Humid Atlantic forest 60
Loc2 Katazpegi river forest Lesaka, Navarra 43°16′10.5′′ N, 1°44′57.2′′ W Humid Atlantic forest 80
Loc3 Tximista river forest Etxalar, Navarra 43°14′1.77′′ N, 1°39′34.8′′ W Humid Atlantic forest 80
Loc4 Serra del Corredor, Can Rimbles creek river forest, Canyamars, Barcelona 41°35′38.7′′ N, 2°27’52′′ E Mediterranean evergreen oak forest 250
Loc5 Serra del Corredor, Canyamars creek river forest, Canyamars, Barcelona 41°36′18.7′′ N, 2°27′23′′ E Mediterranean evergreen oak forest 260
Loc6 Montseny Natural Park, Tordera river forest, Montseny, Barcelona 41°44′0.0′′ N, 2°24′54.6′′ E Mediterranean evergreen oak forest 310
Loc7 Montseny Natural Park, Montseny, Barcelona 41°45′31.4′′ N, 2°23′56.7′′ E Mediterranean evergreen oak forest 350
Table 2 List of specimens, voucher number, locality (except for GenBank sequences not generated by the authors) and amplified fragments
used in the analyses. Families of eumonostiliferan nemerteans follow Chernishev (2005). See Table 1 for Spanish (NT) locality codes.
Family Species Voucher code Locality 18S rRNA COI
Lineidae
Lineus bilineatus
AToL000087 Sweden DQ279932 DQ280014
Prostomatidae
Prostoma eilhardi
No voucher Barcelona, Spain U29494
Prostomatidae
Prostoma graecense
AY928356
Tetrastemmatidae
Tetrastemma laminariae
AY791979
Amphiporidae
Amphiporus
sp. DNA100367 Massachusetts, USA AF119077 EU255601
Oerstediidae
Oerstedia striata
AY928354 AY791972
Oerstediidae
Oerstedia dorsalis
AY210448 AY791971
Oerstediidae
Oerstedia zebra
AJ436912
Oerstediidae
Oerstedia venusta
AJ436911
Prosorhochmidae
Geonemertes pelaensis
DNA102574 Bermuda EU255578 EU255602
Prosorhochmidae
Pantinonemertes
sp. AJ436914
Acteonemertidae
Antiponemertes novaezealandiae
AY928345
Acteonemertidae
Argonemertes australiensis
DNA100425 Tasmania, Australia AF519235 AY428840
Acteonemertidae Nemertean C NT000003 Loc1 EU255579 EU255603
Acteonemertidae Nemertean B NT000014 Loc5 EU255580 EU255604
Acteonemertidae Nemertean C NT000024 Loc1 EU255605
Acteonemertidae Nemertean C NT000025 Loc1 EU255606
Acteonemertidae Nemertean C NT000026 Loc1 EU255607
Acteonemertidae Nemertean C NT000027 Loc1 EU255608
Acteonemertidae Nemertean B NT000031 Loc4 EU255581 EU255609
Acteonemertidae Nemertean B NT000039 Loc5 EU255582 EU255610
Acteonemertidae Nemertean B NT000040 Loc5 EU255611
Acteonemertidae Nemertean B NT000042 Loc4 EU255583 EU255612
Acteonemertidae Nemertean B NT000044 Loc4 EU255584 EU255613
Acteonemertidae Nemertean B NT000046 Loc4 EU255585 EU255614
Acteonemertidae Nemertean B NT000047 Loc4 EU255586 EU255615
Acteonemertidae Nemertean B NT000048 Loc4 EU255587 EU255616
Acteonemertidae Nemertean C NT000049 Loc7 EU255588 EU255617
Acteonemertidae Nemertean B NT000050 Loc6 EU255589 EU255618
Acteonemertidae Nemertean B NT000051 Loc6 EU255590 EU255619
Acteonemertidae Nemertean B NT000054 Loc4 EU255591 EU255620
Acteonemertidae Nemertean B NT000059 Loc1 EU255592 EU255621
Acteonemertidae Nemertean C NT000060 Loc1 EU255622
Acteonemertidae Nemertean B NT000062 Loc1 EU255623
Acteonemertidae Nemertean C NT000068 Loc2 EU255593
Acteonemertidae Nemertean C NT000069 Loc3 EU255594 EU255624
Acteonemertidae Nemertean C NT000070 Loc3 EU255595 EU255625
Acteonemertidae Nemertean A NT000072 Loc1 EU255596 EU255626
Acteonemertidae Nemertean A NT000073 Loc1 EU255597 EU255627
Acteonemertidae Nemertean A NT000074 Loc1 EU255598 EU255628
Acteonemertidae Nemertean A NT000075 Loc1 EU255599 EU255629
Acteonemertidae Nemertean A NT000076 Loc1 EU255600 EU255630

Diversity of Iberian terrestrial nemerteans •E. Mateos & G. Giribet
238 Zoologica Scripta, 37, 3, May 2008, pp 235–243 • © 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters
phylogeny, but to test the lineage diversity of the terrestrial
species.
Molecular data
Molecular data were generated primarily for two markers,
the ribosomal nuclear gene 18S rRNA and the mitochondrial
protein-encoding gene cytochrome c oxidase subunit I (COI
hereafter) because a wealth of comparative data for other
hoplonemertean species exists for these two markers and
because they provide good resolution at different time
frameworks, including species level (Sundberg et al. 2001;
Thollesson & Norenburg 2003; Strand & Sundberg 2005a,b;
Sundberg & Strand 2007). Although the mitochondrial
ribosomal gene 16S rRNA has also been broadly used in
nemertean systematic studies (Envall & Sundberg 1998;
Sundberg & Saur 1998; Thollesson & Norenburg 2003;
Sundberg & Strand 2007), we were unable to obtain sequences
for more than just a few individuals and therefore we
preferred to exclude it from the current analyses.
18S rRNA was amplified in three overlapping fragments
using primer pairs 1F–4R, 3F–18Sbi and 18Sa2.0–9R, following
standard protocols. For COI amplification, we used primer
pair LCO1490–HCOoutout, which amplifies a fragment
of 794 bp. In a few cases we used a reverse primer at 5′ of
HCOoutout, such as HCO or HCOout. For primer sequences
and amplification conditions see, for example, Edgecombe &
Giribet (2006).
The amplified samples were purified using the QIAquick®
PCR Purification Kit (QIAGEN, Valencia, CA, USA), labelled
using BigDye® Terminator v3.0 (Applied Biosystems, Foster
City, CA) and sequenced with an ABI 3730 DNA Analyser
(Applied Biosystems) following manufacturer’s protocols.
Primers used in the sequencing reaction correspond to those
used in the amplification step.
Chromatograms obtained from the automatic sequencer
were read and ‘contig sequences’ were assembled using the
sequence editing software Sequencher 4.7 (Gene Codes
Corporation, Ann Arbor, MI) and further manipulated in
MacGDE 2.2 (Linton 2005). 18S rRNA was analysed as nine
fragments (1763–1790 bp for the non-terrestrial species;
1782–1964 bp for the terrestrial nemerteans). COI was
analysed as a single fragment. All new sequences have been
deposited in GenBank under the accession numbers EU255578–
EU255630 (Table 2).
Phylogenetic analyses
The 18S rRNA data set was initially analysed with direct
optimization (Wheeler 1996) in POY v. 4.0.0 rc 2318 (Varón
et al. 2007) under a parameter set where indel openings cost
3, base transformations cost 2 and indel extensions cost 1 (De
Laet 2005). The search strategy consisted of 100 replicates
of Wagner addition with SPR and TBR branch swapping,
holding up to 10 trees during swapping for each replicate.
The implied alignment (Wheeler 2003; Giribet 2005) was
then used to conduct an equivalent tree search in PAUP* v. 4.0
(Swofford 2002), with 100 Wagner addition trees followed by
TBR branch swapping and holding up to 100 trees during
swapping. One topology was selected to represent branch
lengths. Nodal support was assessed via 1000 jackknife
replicates, with a probability of deletion of characters of e–1
(Farris et al. 1996).
The COI data set was analysed both in POY and PAUP* — it
required no indel events — under the same specifications used
for the 18S rRNA data set. Both data sets readily converged
on the same solution, not requiring more sophisticated tree
search strategies (e.g. see Goloboff 1999; Giribet 2007).
The analysis files and implied alignment for the 18S rRNA
marker are available from the corresponding author upon
request.
Results and discussion
Morphospecies
The individuals collected (Fig. 1) were assigned to three
morphospecies (A, B and C) on the basis of external morphology,
pigmentation and eye number and disposition (Table 2). According
to these criteria, morphospecies A (length 8–25 mm) strongly
resembles A. dendyi (Dakin, 1915), morphospecies B (length
4–9 mm) resembles Leptonemertes chalicophora (Graff, 1879) and
morphospecies C (length 4–8 mm) is similar to depigmented
L. chalicophora (see Moore & Moore 1972) (Fig. 2).
18S rRNA data set
The POY analysis resulted into seven trees of 1212 weighted
steps, with 7% of the replicates yielding trees of minimal
Fig. 1 Distribution map of the new records of land nemerteans in the
Iberian Peninsula. Numbers for localities correspond to those of
Table 1.

E. Mateos & G. Giribet •Diversity of Iberian terrestrial nemerteans
© 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters • Zoologica Scripta, 37, 3, May 2008, pp 235–243 239
length. One of the implied alignments, analysed with PAUP*
(under equal weights — nonlinear indel costs cannot be
implemented in static alignments), yielded optimal trees of
649 steps in 100% of the replicates. The strict consensus of
all the trees from both analyses is identical, and one of the
fundamental trees is shown in Fig. 2.
18S rRNA characterization of the 23 new terrestrial
nemertean specimens plus two additional species from
GenBank showed considerable sequence variation, consistent
with a pattern of six species of terrestrial nemerteans and at
least two independent origins of terrestriality. One of the
clades includes a single terrestrial nemertean, G. pelaensis
from Bermuda, our single representative of the family
Prosorhochmidae. All the other terrestrial nemerteans,
including A. australiensis, An. novaezealandiae and the Iberian
samples (NTs) form a well-supported clade (100% jackknife
frequency; JF hereafter) that is sister to the genus Oerstedia
(100% JF), strongly suggesting monophyly of Group II. The
Iberian specimens appear in three clades, each supported with
100% JF, which grossly correspond to our initial assessment of
morphospecies A, B and C, although the latter two appear
somehow mixed (NT59, NT68 and NT69). The A clade
corresponds to the largest forms collected in Fragas do Eume
(Galicia, NW Iberian Peninsula; NT72–NT76), and none of
the members of this clade appear in the other samples
populations. Another clade includes most members of our
form C, from localities in Fragas do Eume, Etxalar (Navarra,
North Iberian Peninsula), and Montseny (Catalonia,
Northeastern Iberian Peninsula), spanning c. 1000 km along
the northern margin of Spain. This clade shows the broadest
distribution of any of the lineages examined, and it is well
supported as sister to the following clade. A third clade,
highly differentiated molecularly, includes most members of
our form B and a few members of the form C from localities
in Barcelona and Navarra. The members of this third clade
have multiple long insertions in 18S rRNA, which make the
group easily diagnosable molecularly.
The phylogenetic pattern that we observe for the Iberian
forms, forming three distinct and highly apomorphic clades
with little internal differentiation is a strong indicator of the
presence of three species of terrestrial nemerteans in the
Iberian Peninsula, none of which correspond to the species
A. australiensis or An. novaezealandiae. Morphologically, the
Iberian species most closely resemble the descriptions of
A. dendyi (A) and L. chalicophora (B and C), but until we are
able to compare our specimens to the types or obtain genetic
data from specimens from the type localities, we prefer not to
assign a species identity to either clade.
Fig. 2 One of the shortest phylograms at 649 steps obtained with PAUP* based on the implied alignment generated in POY. The different
fundamental trees only differ in the internal arrangement of the three clades of Iberian specimens. Numbers on branches indicate jackknife
v
alues. Letters after specimen codes indicate morphotypes assigned to specimens. Images show the main morphotypes surveyed. All
photographs by E. Mateos.

Diversity of Iberian terrestrial nemerteans •E. Mateos & G. Giribet
240 Zoologica Scripta, 37, 3, May 2008, pp 235–243 • © 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters
COI data set
The COI analysis in PAUP* yielded trees of 1624 steps in 74%
of the replicates. After swapping the trees to completion,
2952 trees were stored. The strict consensus of all these trees
and the phylogram representation of one of them are shown
in Fig. 3.
COI data appear more complex than the 18S rRNA data
set, because nodal support is mostly restricted to shallower
nodes. Again, G. pelaensis appears as an independent lineage,
and another Prosorhochmidae, Pantinonemertes sp., also
represented in the analyses, is not its sister group. The members
of Group II appear in two clades, one highly supported with
the other hoplonemerteans, but with no resolution in the
deepest branches. The remaining NT sequences form two
main clades, one well supported with sequences NT39–
NT42, which are samples from Barcelona of form B, but
these sequences appear unambiguously in the third clade of
the 18S rRNA analysis. Another clade includes samples from
multiple localities and postulated morphotypes, and with a
large sequence divergence. In contrast, the members of the
NT samples that clustered with the remaining monostiliferan
hoplonemerteans show little genetic divergence, perhaps
with the exception of NT74. Therefore, the members of
clade 1 in the 18S rRNA analysis including sequences NT72–
NT76, from Fragas do Eume, with identical 18S rRNA
sequences, shows p-distances between 0.016 and 0.282,
although four of the pairwise comparisons have p-distances
> 0.210. This value contrasts with those of the other two
clades that range between 0.000 and 0.010 (clade containing
sequences NT14, 31, 44, 50, 54) or 0.000 and 0.007 (clade
composed of NT3, 24–27, 49, 59, 60, 62, 70).
It is interesting to see that the COI sequences of the two
members of Group I (G. pelaensis and Pantinonemertes sp.) do
not form a clade with the other terrestrial species, although
the lack of support in the deep branches of the tree prevents
us from providing stronger evidence for the multiple origins
of terrestrialization in land nemerteans. Only G. pelaensis was
included in the 18S rRNA analysis, because we were not able
to obtain specimens of Pantinonemertes (the COI sequence
used is from GenBank).
While this study represents a first glimpse into the molecular
diversity of terrestrial nemerteans, we show the existence of
three or four lineages of terrestrial nemerteans belonging to
Group II in the Iberian Peninsula, an area where no land
nemerteans had been previously described, and only a few
citations exist for freshwater species (see Giribet & Carranza
1994). We hope that this study stimulates research on terrestrial
cryptic soil fauna, and particularly on land nemerteans.
Our study, especially the 18S rRNA data set, provides the
first test for the monophyly of land nemerteans using molecular
Fig. 3 A, B. —A. Strict consensus of 2952 most parsimonious trees (1624 steps) for the COI data set. —B. Phylogram of one of the 2952 most
parsimonious trees (1624 steps) for the COI data set. Letters after specimen codes indicate morphotypes assigned to specimens. Numbers on
branches indicate jackknife values above 80%.

E. Mateos & G. Giribet •Diversity of Iberian terrestrial nemerteans
© 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters • Zoologica Scripta, 37, 3, May 2008, pp 235–243 241
data and further corroborates the usefulness of this marker
for nemertean species-level identification and systematics.
From the six putative species included in this analysis, all the
members of Group II form a clade, including A. australiensis,
An. novaezealandiae, and the Iberian forms, which most
probably include members of the genera Leptonemertes and
Argonemertes. Crandall (2001) proposed that this Group II of
genera constituted a new family, being formalized as
Acteonemertidae by Chernishev (2005). Our results agree
with the scheme proposed by Chernishev (2005) and the
inclusion of the Group II terrestrial nemertenas in the family
Acteonemertidae. From a systematic point of view, 18S
rRNA shows A. novaezealandiae as sister to the other species,
including morphospecies A, which may be a member of the
genus Argonemertes. This contrasts with the morphological
analyses of Sundberg (1989a,b), which suggest Argonemertes
to be more derived (making An. paraphyletic) while L.
chalicophora was found to be sister to the other species. These
results would require further testing with more species.
Biogeography and distribution of terrestrial nemerteans
The fact that land nemerteans are found in many remote
islands, some of which contain no fresh water, has led some
authors to postulate an independent origin of genera from
marine ancestors (Moore & Gibson 1985), a hypothesis
supported by the occurrence of some New Zealand Acteonemertes
in the supralittoral, vs. a fully terrestrial occurrence in other
localities, or the presence of marine species of Pantinonemertes
(Roe 1991). However, the fact that some species are found on
islands across vast distances, such as the cases of G. pelaensis
(e.g. see Gibson & Moore 1998) or A. dendyi (e.g. see Moore
et al. 2001), may also indicate that land nemerteans can
disperse easily, perhaps accompanying migrant birds, since
some species may overcome desiccation by forming cocoons.
Although our data reject a single origin of terrestrial species,
our current sampling does not contradict a single origin of
the members of the family Acteonemertidae, a result that
should be further tested by the inclusion of more ingroup and
outgroup species, but that so far corroborates the morphological
cladistic analyses of Sundberg (1989a,b).
With respect to their distribution, Moore & Gibson (1985)
discussed the lack of nemerteans on continents, with most
species occurring on islands. Australia is an exception, but
most other continental species are often thought to be
introduced, since they are found in gardens or anthropogenic
habitats (but see the case of Pantinonemertes californiensis,
which seems to be a native species to California). Contrary to
this expectation, the Iberian specimens were all found
primarily in undisturbed humid forests. Locality 1 (Fragas do
Eume Natural Park) is a protected area of humid Atlantic
forest with the Pedunculate Oak (Quercus robur) as the
dominant species. Localities 2 and 3 (Navarra) are not
protected, but include Atlantic humid forest with the European
Beech (Fagus sylvatica) as the dominant tree species. These
three localities are not directly affected by human activity and
their fauna is often considered to be autochthonous. Localities
6 and 7 are located on small creeks and riverbeds inside the
Montseny Natural Park, with Evergreen Mediterranean Oak
forest (the Holm Oak Q. ilex is the dominant tree species).
Anthropogenic activity is also minimal in these localities.
Localities 4 and 5 (Serra del Corredor) have similar but more
xeric conditions, and with a slightly higher anthropogenic
influence, although by no means can this area be considered
to be secondary forest. Therefore, it is highly plausible that
the Iberian populations of terrestrial nemerteans are auto-
chthonous, and not introduced. This may actually be reflected
in the large COI divergence found in at least some of the
sampled clades (e.g. Fragas do Eume). If the diversity of
terrestrial nemerteans were to be explained by recent
introductions, genetic diversity should be much lower, reflecting
a bottleneck effect, unless multiple introductions were postu-
lated, which seems unlikely. Similar p-distance values have
been found in other soil invertebrates, which are considered
to constitute very old lineages (e.g. Boyer et al. 2007).
Other aspects of the ecology and distribution of Iberian
acteonemertids are also of interest. At least one of the
localities, Fragas do Eume N. P., in the NW Iberian Peninsula,
is home to multiple lineages of terrestrial nemerteans, as
revealed by both markers. This is also probably true for other
localities, such as Montseny and Etxalar, although it would be
better to obtain further evidence from these two sites. On the
contrary, samples from Serra del Corredor are restricted to a
single clade in the 18S rRNA analysis, even though they
appear in three clades in the COI data set. Finally, the
distribution ranges of the 18S rRNA lineages are also drastically
different. Clade I is restricted to specimens from Fragas do
Eume, while clade II includes specimens that span the entire
northern part of the Iberian Peninsula, from the provinces of
A Coruña to Barcelona. It may also be premature to decide
whether this is due to sampling artefacts, as for example, all
the members of clade I were collected during the same
sampling time, while samples from other dates in the same
locality yielded specimens from the other clades.
General conclusions
Our data provide a first glimpse into the terrestrial nemertean
fauna in an old continental landmass, the Iberian Peninsula,
showing that terrestrial nemerteans can be found in continental
undisturbed forests, and are not confined to islands and
marginal or anthropogenic environments. Furthermore, COI
data indicate that at least some lineages have been in the area
for a long time and they are not the result of a single coloni-
zation by few individuals. The presence of multiple lineages
and large molecular variation among groups of specimens

Diversity of Iberian terrestrial nemerteans •E. Mateos & G. Giribet
242 Zoologica Scripta, 37, 3, May 2008, pp 235–243 • © 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters
studied suggests a fundamental lack of knowledge of this
component of our soil fauna. It is still too early to attempt to
understand species ranges and patterns of distribution and
biogeography at a global scale, but we hope that this study
opens a new door in advancing our knowledge on nemertean
biology.
Acknowledgements
Joey Pakes (Harvard University) provided invaluable
assistance with the molecular work. Bob Mesibov, Per
Sundberg and Wolfgang Sterrer generously made specimens
available. Miquel Vila assisted with fieldwork. Greg Edgecombe
and an anonymous reviewer provided helpful comments that
improved our manuscript. EM obtained a fellowship from
AGAUR (Generalitat de Catalunya) for a 1-month visit to
the Museum of Comparative Zoology.
References
Anderson, R. (1980). The status of the land nemertine Argonemertes
dendyi (Dakin) in Ireland. Irish Naturalists’ Journal, 20, 153–157.
Anderson, F. E., Córdoba, A. J. & Tholleson, M. (2004). Bilaterian
phylogeny based on analyses of a region of a region of the sodium–
potassium ATPase α-subunit gene. Journal of Molecular Evolution,
58, 252–268.
Boyer, S. L., Baker, J. M. & Giribet, G. (2007). Deep genetic divergences
in Aoraki denticulata (Arachnida, Opiliones, Cyphophthalmi): a
widespread ‘mite harvestman’ defies DNA taxonomy. Molecular
Ecology, 16, 4999–5016.
Chernishev. A. V. (2005). System of families of enoplan nemerteans
of the order Eumonostilifera (Nemertea: Enopla). Russian Journal
of Marine Biology, 31, 27–33.
Crandall, F. B. (2001). A cladistic view of the Monostilifera
(Hoplonemertea) with interwoven rhynchocoel musculature: a
preliminary assessment. Hydrobiologia, 456, 87–110.
De Laet, J. E. (2005). Parsimony and the problem of inapplicables in
sequence data. In V. A. Albert (Ed.) Parsimony, Phylogeny, and
Genomics (pp. 81–116). Oxford: Oxford University Press.
Edgecombe, G. D. & Giribet, G. (2006). A century later — a total
evidence re-evaluation of the phylogeny of scutigeromorph
centipedes (Myriapoda: Chilopoda). Invertebrate Systematics, 20,
503–525.
Envall, M. & Sundberg, P. (1998). Phylogenetic relationships and
genetic distances between some monostiliferous interstitial
nemerteans (Ototyphlonemertes, Hoplonemertea, Nemertea)
indicated from the 16S rRNA gene. Zoological Journal of the Linnean
Society, 123, 105–115.
Farris, J. S., Albert, V. A., Källersjö, M., Lipscomb, D. & Kluge, A.
G. (1996). Parsimony jackknifing outperforms neighbor-joining.
Cladistics, 12, 99–124.
Gibson, R. (1990). The macrobenthic nemertean fauna of Hong
Kong. In B. Morton (Ed.) Proceedings of the Second International
Marine Biological Workshop: The Marine Fauna and Flora of Hong
Kong and Southern China, Hong Kong, 1986 (pp. 33–212). Hong
Kong.
Gibson, R. & Moore, J. (1985). The genus Prosorhochmus Keferstein,
1862 (Hoplonemertea). Journal of Zoology, London, 206, 145–162.
Gibson, R. & Moore, J. (1998). Further observations on the genus
Geonemertes with a description of a new species from the Philippine
Islands. Hydrobiologia, 365, 157–171.
Gibson, R., Moore, J. & Crandall, F. B. (1982). A new semi-terrestrial
nemertean from California. Journal of Zoology, London, 196, 463–
474.
Giribet, G. (2005). Generating implied alignments under direct
optimization using POY. Cladistics, 21, 396–402.
Giribet, G. (2007). Efficient tree searches with available algorithms.
Evolutionary Bioinformatics, 3, 1–16.
Giribet, G. & Carranza, S. (1994). Dos nuevas localidades para un
nemertino de agua dulce del género Prostoma Dugès, 1828
(Hoplonemertini: Tetrastemmatidae) en la Península Ibérica.
Graellsia, 50, 170–172.
Goloboff, P. A. (1999). Analyzing large data sets in reasonable times:
solutions for composite optima. Cladistics, 15, 415–428.
Hickman, V. V. (1963). The occurrence in Tasmania of the land
nemertine, Geonemertes australiensis Dendy, with some account of
its distribution, habits, variations and development. Papers and
Proceedings of the Royal Society of Tasmania, 97, 63–75.
Howarth, F. G. & Moore, J. (1983). The land nermetine Argonemertes
dendyi (Dakin) in Hawaii (Nemertinea: Hoplonemertinea:
Prosorhochmidae). Pacific Science, 37, 141–144.
Linton, E. W. (2005). MACGDE: Genetic Data Environment for
MacOS X. Ver. 2.0. Software available at <www.msu.edu/~lintone/
macgde/>.
Maslakova, S. A. & Norenburg, J. L. (2001). Phylogenetic study of
pelagic nemerteans (Pelagica, Polystilifera). Hydrobiologia, 456,
111–132.
Maxmen, A. B., King, B. F., Cutler, E. B. & Giribet, G. (2003).
Evolutionary relationships within the protostome Phylum Sipuncula:
a molecular analysis of ribosomal genes and histone H3 sequence
data. Molecular Phylogenetics and Evolution, 27, 489–503.
Moore, J. & Gibson, R. (1981). Further studies on the evolution of
land and freshwater nemerteans: generic relationships among the
paramonostiliferous taxa. Journal of Zoology, London, 216, 1–20.
Moore, J. & Gibson, R. (1985). The evolution and comparative
physiology of terrestrial and freswater nemerteans. Biological
Reviews, 60, 257–312.
Moore, J. & Gibson, R. (1988a). Further studies on the evolution of
land and freshwater nemerteans: generic relationships among
paramonostiliferous taxa. Journal of Zoology, London, 216, 1–20.
Moore, J. & Gibson, R. (1988b). Marine relatives of terrestrial
nemerteans: the genus Prosadenophorus Bürger, 1890 (Hoplonemertea).
Hydrobiologia, 156, 75–86.
Moore, J. & Moore, N. W. (1972). Land nemertines of Madeira and
the Azores. Boletim do Museu Municipal do Funchal, 26, 31–44.
Moore, J., Gibson, R. & Jones, H. D. (2001). Terrestrial nemerteans
30 years on. Hydrobiologia, 456, 1–6.
Myers, N., Mittermeier, R. A., Mittermeier, C. G., da Fonseca, G.
A. & Kent, J. (2000). Biodiversity hotspots for conservation
priorities. Nature, 403, 853–858.
Pantin, C. F. A. (1969). The genus Geonemertes. Bulletin of the British
Museum (Natural History), Zoology, 18, 261–310.
Roe, P. (1991). Aspects of the biology of Pantinonemertes californiensis,
a high intertidal nemertean. Hydrobiologia, 266, 29–44.
Schwartz, M. L. & Norenburg, J. L. (2001). Can we infer heterone-
mertean phylogeny from available morphological data? Hydrobiologia,
456, 165–174.

E. Mateos & G. Giribet •Diversity of Iberian terrestrial nemerteans
© 2008 The Authors. Journal compilation © 2008 The Norwegian Academy of Science and Letters • Zoologica Scripta, 37, 3, May 2008, pp 235–243 243
Strand, M. & Sundberg, P. (2005a). Delimiting species in the
hoplonemertean genus Tetrastemma (Phylum Nemertea): morphology
is not concordant with phylogeny as evidenced from mtDNA
sequences. Biological Journal of the Linnean Society, 86, 201–212.
Strand, M. & Sundberg, P. (2005b). Genus Tetrastemma Ehrenberg,
1831 (Phylum Nemertea) — a natural group? Phylogenetic
relationships inferred from partial 18S rRNA sequences. Molecular
Phylogenetics and Evolution, 37, 144–152.
Sundberg, P. (1989a). Classification of the family Plectonemertidae
(Nemertea) — a phenetic comparison. Zoological Journal of the
Linnean Society, 97, 57–68.
Sundberg, P. (1989b). Phylogeny and cladistic classification of the
paramonostiliferous family Plectonemertidae (Phylum Nemertea).
Cladistics, 5, 87–100.
Sundberg, P. (1989c). Phylogeny and classification of terrestrial
nemerteans — the genera Pantinonemertes Moore & Gibson and
Geonemertes Semper. Zoological Journal of the Linnean Society, 95,
363–372.
Sundberg, P. & Saur, M. (1998). Molecular phylogeny of some
European heteronemertean (Nemertea) species and the monophyletic
status of Riseriellus, Lineus, and Micrura. Molecular Phylogenetics and
Evolution, 10, 271–280.
Sundberg, P. & Strand, M. (2007). Genetics do not reflect habitat
differences in Riseriellus occultus (Heteronemertea, Nemertea)
from Spain and Wales. Marine Biology Research, 3, 117–122.
Sundberg, P., Turbeville, J. M. & Lindh, S. (2001). Phylogenetic
relationships among higher nemertean (Nemertea) taxa inferred
from 18S rDNA sequences. Molecular Phylogenetics and Evolution,
20, 327–334.
Swofford, D. L. (2002). PAUP* 4.0: Phylogenetic Analysis Using Parsimony
(*and Other Methods), Version 4. Sunderland, MA: Sinauer Associates.
Thollesson, M. & Norenburg, J. L. (2003). Ribbon worm relationships:
a phylogeny of the Phylum Nemertea. Proceedings: Biological
Sciences, 270, 407–415.
Turbeville, J. M. (2002). Progress in nemertean biology: development
and phylogeny. Integrative and Comparative Biology, 42, 692–703.
Varón, A., Vinh, L. S. & Wheeler, W. C. (2007). POY 4.0.0 rc 2318
American Museum of Natural History. <http://research.amnh.org/
scicomp/projects/poy.php>.
Wheeler, W. C. (1996). Optimization alignment: the end of multiple
sequence alignment in phylogenetics? Cladistics, 12, 1–9.
Wheeler, W. C. (2003). Implied alignment: a synapomorphy-based
multiple-sequence alignment method and its use in cladogram
search. Cladistics, 19, 261–268.