A phylogenetic analysis of Tubificinae and Limnodriloidinae
(Annelida, Clitellata, Tubificidae) using sperm and somatic
characters
Blackwell Science, Ltd
ROBERTO MAROTTA, MARCO FERRAGUTI & CHRISTER ERSÉUS
Accepted: 24 April 2002
Marotta, R., Ferraguti, M. & Erséus, C. (2003). A phylogenetic analysis of Tubificinae and
Limnodriloidinae (Annelida, Clitellata, Tubificidae) using sperm and somatic characters. —
Zoologica Scripta, 32, 255 – 278.
The spermatozoa and the sperm aggregates of 13 species belonging to four genera (Smithsonidrilus, Limnodriloides, Thalassodrilides, Doliodrilus) in the tubificid subfamily Limnodriloidinae
(Annelida, Oligochaeta) were studied and compared with the spermatozoal patterns already
described in the subfamily Tubificinae. Two characters considered exclusive for the Tubificinae
were found in the more spermatologically variable Limnodriloidinae: the production of two
kinds of spermatozoa, eusperm and parasperm, and the presence, in the spermathecae, of sperm
aggregates formed by a combination of the two sperm types. A parsimony analysis was performed on the spermatozoal data of the species examined and compared with that based on
the somatic characters of the same species: a critical revision of the already codified eusperm
characters was carried out and the ultrastructure of parasperm was used as a new subset of spermatozoal characters. In a total evidence approach, a further parsimony analysis was run using
a matrix combining both sets of characters. This analysis suggested that the double sperm
line and the sperm aggregates composed of both eusperm and parasperm may well be homologous
in tubificines and limnodriloidines. It thus supported the previous notion that Tubificinae and
Limnodriloidinae are closely related and indicated that these subfamilies may be sister taxa.
R. Marotta, M. Ferraguti, Department of Biology, University of Milano, 26 via Celoria,
I-20133, Milano, Italy. E-mail: roberto.marotta@unimi.it
C. Erséus, Department of Invertebrate Zoology, Swedish Museum of Natural History,
Box 50007, SE-10405 Stockholm, Sweden.
Introduction
Tubificidae is a large and cosmopolitan group of aquatic
oligochaete worms living in freshwater, brackish water
and marine habitats. The family comprises about 1000
species at present and is divided into five subfamilies:
Rhyacodrilinae Hrabe, 1963, Phallodrilinae Brinkhurst, 1971,
Limnodriloidinae Erséus, 1982, Telmatodrilinae Eisen, 1885
and Tubificinae Eisen, 1885 (Erséus 1990).
In recent years, however, phylogenetic analyses based on
different character types have suggested that this is not a
phylogenetic classification: Tubificidae itself and some of its
constituent subfamilies are probably paraphyletic groups
(Erséus et al. 2000, 2002). Morphological (Erséus 1990;
Brinkhurst 1994), spermiological (Ferraguti & Erséus 1999)
and molecular data (Erséus et al. 2000, 2002) suggest that the
Rhyacodrilinae is a paraphyletic taxon, and that the Naididae,
another oligochaete family, is also a tubificid group.
The Tubificinae and Limnodriloidinae are spermiologically interesting as they are the only euclitellates known to
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
possess a double sperm line (Ferraguti 2000), producing
two sperm types that differ in their structure and function —
eusperm and parasperm (Healy & Jamieson 1981).
Ultrastructural studies on species belonging to five different tubificine genera have shown that their different sperm
models are variations on the same structural theme. Moreover, in all tubificines examined, parasperm and eusperm are
grouped to form, in their spermathecae, characteristically
organized rod-shaped spermatozeugmata (Ferraguti et al.
1989, 2002).
Although morphological studies have suggested that the
Limnodriloidinae is the sister group of the Tubificinae–
Telmatodrilinae assemblage (Erséus 1990; Brinkhurst 1994),
there is limited knowledge concerning its sperm ultrastructure. So far, we have spermatological data for only two
species, Smithsonidrilus hummelincki (Righi & Kanner 1979)
and Thalassodrilides ineri (Righi & Kanner 1979), and some
limited information for S. tuber (Erséus 1983) [reviewed
in Ferraguti (2000)], but this already suggests a higher
255
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
spermatological variability inside this subfamily than in the
Tubificinae. While S. hummelincki has a double sperm line,
which to date is the only documented case of dichotomous
spermatogenesis in euclitellates outside the Tubificinae,
T. ineri produces only one sperm type. Their spermatozeugmata also differ considerably: T. ineri, considered a highly
apomorphic form (Righi & Kanner 1979), has a large, single
spermatozeugma in each spermatheca and it is formed by
only euspermatozoa which are in close contact with a large
proteinaceous structure; in the spermathecae of S. hummelincki, in contrast, there is a mixture of several small
spermatozeugmata, each one formed by either eusperm or
parasperm (Erséus & Ferraguti 1995).
Light microscopical observations of limnodriloidine
species belonging to the large genus Limnodriloides revealed
a great variation in the shape and organization of the sperm
aggregates, from poorly organized sperm bundles to rodshaped arrangements foreshadowing the more complex tubificine spermatozeugmata (Erséus 1982).
The phylogenetic meaning of the tubificine-like spermatozeugmata among tubificids is problematic. In his cladistic
analysis, Erséus (1990) considered the presence of this type of
spermatozeugma as the only synapomorphy of Tubificinae
and Limnodriloidinae, but noting that this ‘is an assumption
… somewhat weak because the detailed structure of sperm
aggregates in various taxa is poorly known’. Brinkhurst
(1994), however, regarded the tubificine-like spermatozeugmata as an autapomorphy for the Tubificinae; while the
sperm of limnodriloidines were interpreted as ‘bundles’ only.
In this study, we compared the sperm morphology and
organization of 13 species belonging to four limnodriloidine
genera: Limnodriloides, Smithsonidrilus, Thalassodrilides and
Doliodrilus; the sperm ultrastructure of 11 of these species is
described for the first time. Our aims were to establish
whether a double sperm line is present in more than one
species of Limnodriloidinae, and to describe the ultrastructure of the various sperm aggregates inside this subfamily,
to understand if they are useful as phylogenetic markers
and if they are models en route towards the tubificine
spermatozeugmata.
Using the spermatozoal characters, our purpose was also to
improve the cladistic analysis of tubificids, with particular
attention to the relationships inside the Limnodriloidinae
and between this subfamily and Tubificinae.
Within the Euclitellata, the ultrastructure of spermatozoa
has proved useful for phylogenetic assessment, not only at
higher (Jamieson et al. 1987; Ferraguti & Erséus 1999) but
also at lower taxonomic levels, among species of Enchytraeus
(Westheide et al. 1991) and Branchiobdellidae (Cardini et al.
2000). On the other hand, due to homoplasy among many
sperm characters, the relationships within Tubificidae were
poorly resolved in a study by Erséus & Ferraguti (1995).
256
For this reason, we carried out a critical revision of the
already codified eusperm characters (Erséus & Ferraguti
1995; Ferraguti & Erséus 1999) and used, for the first time in
an analysis of oligochaete phylogeny, the ultrastructural
characters of parasperm. These sperm types may be less constrained by the biology of fertilization and thus reveal more
of the phylogenetic structure than eusperm.
The results of the parsimony analysis based on spermatozoal morphology are compared with a parallel phylogenetic
analysis of the same species but based only on morphological
characters, and with an analysis performed on a data matrix
combining both characters sets (Kluge 1989).
Materials and methods
Specimen collection and microscopical techniques
Specimens of S. luteolus (Erséus, 1983) were collected in
subtidal muds and sands in Belize, in April 1993. Material
from S. sacculatus (Erséus, 1983), L. tenuiductus Erséus, 1982,
L. uniampullatus Erséus, 1982 and L. armatus Erséus, 1982
was collected in subtidal coral reef sand, at Heron Island,
Great Barrier Reef, Australia, in April 1994. Individuals of
S. hummelincki (Righi & Kanner, 1979), S. westoni Erséus, 1982,
L. monothecus Cook, 1974, L. barnardi Cook, 1974, T. gurwitschi
(Hrabe, 1971) and T. bruneti Erséus, 1990 were collected from
intertidal and subtidal sediments at Lee Stocking Island, near
Great Exuma, Bahamas, in April 1999. A new species of
Doliodrilus [described separately by Wang & Erséus (2003)]
was collected from brackish water sediments in Hainan,
southern China, in March 2000. Specimens of Isochaetides
arenarius (Michaelsen, 1926) were collected in Lake Baikal, in
July 1996. The last mentioned material was provided by
P. Martin, all other worms were collected by C. Erséus.
The specimens were fixed in a cacodylate-buffered
paraformaldehyde–glutaraldehyde mixture in a saturated
solution of picric acid (SPAFG; Ermak & Eakin 1976). They
were sent in SPAFG to Milan, where they were washed
overnight in 0.1 M cacodylate buffer (pH 7.4), postfixed in
1% osmium tetroxide in the same buffer for 2 h, washed in
distilled water, en bloc stained for 2 h in the dark in 2% aqueous
uranyl acetate, dehydrated in a graded ethanol series and
embedded in Spurr’s resin.
Sections were cut with an LKB Ultratome III or V or with
a Reichert Ultracut E. Thick (about 0.5 µm) sections were
cut and stained to localize sperm funnels and spermathecae,
and thus mature spermatozoa and spermatozeugmata, using
a light microscope; thin sections were stained with lead citrate, carbon coated and observed with a JEOL 100SX transmission electron microscope.
Taxa
The ingroup and outgroup taxa used for the phylogenetic
analysis are listed in Table 1. They cover a range of marine
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Table 1 List of ingroup and outgroup
tubificid taxa included in the parsimony
analyses. The genera are grouped according
to their current subfamilial position.
Taxon
Reference
Ingroup taxa
Tubificinae
Tubifex tubifex
Clitellio arenarius
Isochaetides arenarius
Limnodriloidinae
Smithsonidrilus westoni
Smithsonidrilus hummelincki
Smithsonidrilus luteolus
Smithsonidrilus sacculatus
Limnodriloides armatus
Limnodriloides uniampullatus
Limnodriloides tenuiductus
Limnodriloides monothecus
Limnodriloides barnardi
Thalassodrilides ineri
Thalassodrilides bruneti
Thalassodrilides gurwitschi
Doliodrilus sp.
Outgroup taxa
Rhyacodrilinae
Rhizodrilus russus
Monopylephorus limosus
Phallodrilinae
Pectinodrilus molestus
Coralliodrilus rugosus
and brackish water tubificids belonging to the subfamilies
Limnodriloidinae and Tubificinae.
The Rhyacodrilinae + Phallodrilinae assemblage emerged
as the sister group of Limnodrilinae + Tubificinae in cladistic
analyses based on morphological (Erséus 1990), and combined spermatozoal and morphological characters (Erséus &
Ferraguti 1995). Thus, the outgroup taxa for rooting the
trees (see Table 1) were selected from these other subfamilies,
the rhyacodrilines Monopylephorus limosus Hatai, 1898 and
Rhizodrilus russus Erséus, 1990 (see Ferraguti et al. 1994)
and the phallodrilines Pectinodrilus molestus (Erséus, 1988)
and Coralliodrilus rugosus, Erséus, 1990 (see Erséus &
Ferraguti 1995).
Characters and character states
In this study two different sets of characters were used,
referred to as the ‘spermatozoal’ and ‘somatic’ characters.
The characters are ‘operationally equivalent to putative
synapomorphies’, i.e. ‘conjectural, based on similarity’ statements of primary homology (de Pinna 1991).
Spermatozoal characters. Twenty-nine spermatozoal characters, all treated as unordered, were considered (Appendix 1 —
Fig. 7). Of these, 13 (characters 1–13) refer to euspermatozoa
and 13 (characters 14–26) deal with paraspermatozoa. Three
other characters pertain to another level of sperm organization,
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
Braidotti & Ferraguti (1982); Braidotti et al. (1980); Ferraguti et al. (1989)
Ferraguti & Ruprecht (1992); Ferraguti et al. (1994)
Ferraguti et al. (2002)
Present study
Erséus & Ferraguti (1995); present study
Present study
Present study
Present study
Present study
Present study
Present study
Present study
Ferraguti et al. (1989)
Present study
Present study
Present study
Ferraguti et al. (1994)
Ferraguti et al. (1994)
Erséus & Ferraguti (1995)
Erséus & Ferraguti (1995)
i.e. the presence/absence of a double sperm line (character
27), the various ways in which the spermatozoa aggregate to
form spermatozeugmata (character 28), and the presence/
absence of junctions between the tails of the parasperm
within the spermatozeugmata (character 29).
Somatic characters. Twenty-nine somatic characters were considered, of which five can be treated as ordered (Appendix 2).
Three characters are associated with the morphology of the
chaetae, two to the morphology of the oesophagus and one to
the presence of coelomocytes around the digestive tube. All
the other 23 characters concern the morphology of the
reproductive system; five of them deal with the morphology
of the spermatheca, the others with the different parts of the
male apparatus.
Parsimony analysis
Partitioned somatic and sperm data analysis. Each set of characters was analysed using PAUP version 4.0b8 for 32-bit Microsoft WINDOWS (Swofford 1998). Strict parsimony analysis was
used to determine character congruence as a test to evaluate
conjectures of homology, i.e. primary homologies. Accelerated transformation optimization (ACCTRAN) was used for
tracing character evolution.
The exact branch-and-bound search algorithm was
selected, with the following options: furthest addition
257
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
sequence (ADDSEQ = FURTHEST), save all minimal trees
found during the branch-and-bound search (MULTREES
= YES) and collapse a branch if its minimum possible length
is zero (COLLAPSE = MINBRLEN).
Congruence and data support. To evaluate the statistical significance of character incongruence among different data partitions, for the same group of taxa, the incongruence length
differences (ILD) test (Mickevich & Farris 1981; Farris
et al. 1995) was used.
To assess the incongruence between eusperm and parasperm data, the ILD test was performed, culling, prior to the
analysis, all taxa with only one sperm line. One thousand ILD
replicates, using branch-and-bound searches, were used to
estimate the null distribution.
Including all taxa, the incongruence was examined between spermatozoal and somatic characters. Five hundred
replicates, using heuristic search, were used to estimate the
null distributions.
To evaluate the support for tree topologies, 500 bootstrap
replicates were subjected to a separate heuristic search with
tree bisection-reconnection branch swapping, 10 random
addition sequences were calculated for each bootstrap replicate (ADDSEQ = RANDOM; NREPS = 10) and majority
rule bootstrap trees were calculated (Felsenstein 1985).
Combined analysis. The spermatozoal and somatic characters
were combined in a single morphological data set (Appendix
3) and were considered, in a first analysis, all unordered. In a
second analysis, the spermatozoal character 5 (acrosome
vesicle withdrawn) and the somatic characters 35 (peripheral
position of spermathecal pores), 40 (entrance of vas deferens
into atrium), 45 (position of prostatic pad in atrial wall), 49
(length and shape of atrial ampulla) and 52 (atrial duct sac)
were considered as ordered, as, in each of these cases, a particular transformation series, reflecting an order of relative
similarity, could be postulated.
Results
Sperm morphology
A great variation in sperm models and organization characterizes the Limnodriloidinae. This variability concerns the
presence or absence of a double sperm line, the ways in which
the sperm aggregate to form different spermatozeugmata,
and the morphology of the individual spermatozoa.
Euspermatozoa. The euspermatozoa of all observed species
share their general structural plan with other oligochaetes,
i.e. they are filamentous cells with a tapering, tubular, terminal acrosome, an extremely elongate nucleus which may be
straight, spirally flanged or spiral, a cylindrical sometimes
spiral midpiece, interpolated between nucleus and axoneme,
with two to six radially adpressed mitochondria, and a long
flagellum (Jamieson 1981); they also have a characteristic
basal cylinder (Ferraguti 1984a) and two modifications of the
central axonemal apparatus referred to as ‘tetragon fibres’
and a ‘prominent central sheath’ (Ferraguti 1984b). For a further understanding of the sperm descriptions, please refer to
Appendix 1 (Fig. 7).
Smithsonidrilus species. The euspermatozoa of S. hummelincki,
S. luteolus (Fig. 1A) and S. westoni (Fig. 1E) have an acrosome
formed by a characteristic thin acrosomal tube (length 0.3–
0.5 µm) of uniform thickness and ending, at its posterior
extremity, in an internal thickening, a limen ( Jamieson 1978).
The anterior portion of the acrosomal tube contains an acrosome vesicle that protrudes anteriorly contacting apically the
plasma membrane. From the posterior margin of the acrosomal vesicle, a thin secondary tube starts, and it ends, inside
the acrosomal tube, near the limen. An acrosome rod passes
through the secondary tube to end in an invagination of the
acrosome vesicle, but without protruding from the upper
margin of the acrosomal tube (Fig. 1A and E). The nucleus is
a long cylinder showing regions with different shapes in the
various species: it is apically corkscrew-shaped or twisted,
Fig. 1 A–W. Eusperm of six species belonging to the tubificid subfamily Limnodriloidinae. For a correct interpretation of the details, please refer
to Appendix 1 (Fig. 7). A–W. A–D. Smithsonidrilus luteolus. —A. Acrosome (× 82 000). —B. The twisted apical portion of the nucleus (× 52 000).
—C. The twisted basal nuclear portion ( × 35 000). —D. Cross-sectioned midpiece showing the four mitochondria ( × 67 500).
E–G. Smithsonidrilus westoni. —E. Acrosome (× 75 000). —F. The corkscrew-shaped apical portion of the nucleus (× 26 000). —G. The basal
portion of the nucleus is straight and the mitochondria are roundish (× 35 000). —H. Cross-sections of three tails of S. luteolus showing both
prominent central sheath (top) and tetragon fibres (bottom, left) as modifications of the central axonemal apparatus (× 50 000). I–L. Smithsonidrilus
sacculatus. —I. Acrosome (× 82 000) [compare with (A, E)]. —J. Longitudinal section of the loosely corkscrew-shaped nucleus ( × 45 000).
—K. Coiled thread-shaped basal portion of the nucleus: the basal cylinder is particularly evident (asterisk) (× 55 000). —L. Cross-section of the
tail showing the tetragon fibres (arrowhead) as modification of the central axonemal apparatus ( × 50 000). M–P. Limnodriloides armatus.
—M. Acrosome (× 82 000). —N. Apical portion of the nucleus (× 22 500). —O. The straight basal portion of the nucleus and midpiece
(× 55 000). —P. Cross-sectioned midpiece showing the three mitochondria and a flagellum with tetragon fibres (× 50 000). Q–S. Limnodriloides
tenuiductus. —Q. Acrosome (× 82 000). —R. The apical portion of the nucleus is an elongated coiled thread (× 28 000). —S. The basal
portion of the nucleus is a simple coiled thread (× 35 000). T–V. Limnodriloides monothecus. —T. Acrosome (× 82 000). —U. The loosely twisted
apical portion of the nucleus (× 27 500). —V. Longitudinal section of the basally straight nucleus (× 55 000). —W. Cross-sections of two tails
of L. barnardi showing the tetragon fibres as modifications of the central axonemal apparatus (× 50 000).
258
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
259
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
basally twisted or straight, and its diameter decreases considerably from the basis to the convex apex (Fig. 1B,C,F and G).
The midpiece has four parallel and roundish mitochondria
(Fig. 1D and G) and the axoneme of the tail shows regions
with a prominent central sheath and the tetragon fibres in
sequence (Fig. 1H).
The euspermatozoon of S. sacculatus differs from those of
the other three Smithsonidrilus species studied in that its acrosome (Fig. 1I) is formed by a thicker and shorter acrosome
tube (length only about 0.2 µm), which shows uniform thickness and with an acrosome vesicle almost completely external
to the tube (acrosome vesicle withdrawal about 20% only), and
an acrosome rod that runs inside the acrosome invagination
and protrudes from the upper margin of the acrosome tube
(Fig. 1I). The nucleus ends with a flat apex and does not notably
decrease in diameter from the basis to the apex (Fig. 1J and K).
The midpiece is formed by five roundish mitochondria and
finally the flagellum presents only tetragon fibres (Fig. 1L).
Limnodriloides species. The euspermatozoa of L. barnardi, L.
monothecus, L. tenuiductus and L. uniampullatus have a straight
acrosome with an acrosome tube (length 0.2–0.3 µm) of uniform thickness and ending, at its posterior extremity, in a
limen (Fig. 1Q and T). The acrosome tube contains a thin
secondary tube, starting from the inner margin of the acrosome vesicle and ending near the limen, and an acrosome rod
protuberant from its upper margin (Fig. 1Q and T). The
nucleus is a long cylinder showing regions with different
shapes: it is apically twisted, coiled thread-shaped or corkscrewshaped, basally straight or coiled thread-shaped, and its
diameter is constant from the basis to the slightly convex
nuclear apex (Fig. 1R and U). Limnodriloides tenuiductus and
L. uniampullatus have an acrosome vesicle situated almost
completely outside the acrosomal tube, and a flat nuclear
apex (Fig. 1Q). The midpiece has four roundish mitochondria.
The central apparatus of the axoneme shows only tetragon
fibres (Fig. 1W).
The euspermatozoon of L. armatus differs from those of
the other Limnodriloides species observed. It has a longer
acrosome tube (length about 0.4 µm); the acrosome vesicle is
deeply withdrawn inside the acrosome tube, and the acrosome rod does not protrude from the anterior extremity of
the acrosome tube (Fig. 1M). The nucleus ends with a convex
nuclear apex and its diameter decreases consistently from
the basis to the apex (Fig. 1N and O). There are only three
roundish mitochondria in the midpiece and the central apparatus of the axoneme shows tetragon fibres and a prominent
central sheath in sequence (Fig. 1P).
Thalassodrilides and Doliodrilus species. The euspermatozoa
of T. bruneti, T. gurwitschi and T. ineri have a straight acrosome
with a characteristic long acrosome tube (length about
0.6 µm) that is basally thick, but then abruptly thinner. The
tube has no limen and the acrosome vesicle is almost completely outside it. The thin secondary tube starts from the
inferior margin of the acrosome vesicle, ends far from the
base of the acrosomal tube, and thus there is a wide empty
space at the base of the acrosome, the basal chamber. The
acrosome rod passes through the secondary tube and enters
the invagination of the acrosome vesicle, and at the same time
it protrudes from the apical extremity of the acrosome tube
(Fig. 2A; the protrusion is not visible in this particular section
showing a slightly tangential view). The nucleus is a coiled
thread, but basally straight or twisted, and has a slightly
convex apex; its diameter is constant from the basis to the
apex (Fig. 2B and C). The midpiece contains four roundish
mitochondria. Tetragon fibres are present in the central
axonemal apparatus (Fig. 2D).
The euspermatozoon of Doliodrilus sp. has a very slender
acrosome (length/width ratio about 7) formed by an acrosomal
Fig. 2 A–R. Eusperm and parasperm in Limnodriloidinae and Tubificinae. For an interpretation of the details, please refer to Appendix 1
(Fig. 7). A–D. The euspermatozoon of Thalassodrilides gurwitschi (Limnodriloidinae). —A. The acrosome with the long acrosome tube basally
thick but then (arrow) abruptly thinner (× 60 000). —B. Longitudinal, apical section of the nucleus showing its coiled thread shape (× 40 000).
—C. The nucleus is basally straight (× 40 000). —D. Cross-section of the tail showing the tetragon fibres (× 67 500). E–H. The euspermatozoon
of Doliodrilus sp. (Limnodriloidinae). —E. The acrosome: the acrosome tube has a helical ridge (arrowhead) (× 60 000). —F. Longitudinal,
apical section of the nucleus showing its coiled thread shape × 40 000). —G. The basal portion of nucleus is loosely twisted (× 26 000).
—H. Cross-sectioned midpiece showing the four mitochondria (× 80 000). I–L. The euspermatozoon of Isochaetides arenarius (Tubificinae). —I. The
acrosome (× 60 000). —J. The nucleus is gently twisted in its apical portion (× 40 000). —K. The basal portion of the nucleus is straight
(× 40 000). —L. Cross-sectioned midpiece showing the five mitochondria (× 67 500). M–O. The parasperm of Isochaetides arenarius as an
example of tubificine parasperm. —M. The short and straight nucleus has, at the top, a simple acrosome (arrow) (× 22 500). —N. A longitudinal
section of the midpiece and the basal portion of the tail shows the reduced basal cylinder (arrow) at the base of the axoneme and the plasma
membrane separated from the axoneme for a long tract [compare with (Q), bottom] (× 26 000). —O. The midpiece has three mitochondria;
cross-sectioned nuclei show an oval outline. An arrow points to a cross-section of parasperm acrosome (× 40 000). P–R. Limnodriloidine
parasperm. —P. Longitudinal section of the parasperm of Smithsonidrilus sacculatus: the nucleus is straight and has no acrosome (× 8500). Inset:
apical portion of the nucleus of Limnodriloides armatus (× 35 000). —Q. Longitudinal section of the midpiece and basal portion of the tail of
S. luteolus: the basal cylinder is completely absent; the plasma membrane is separated from the axoneme for a short tract (compare with the crosssections of the spermatozeugma in Fig. 3G) (× 26 000). —R. Cross-sections of the midpieces and tails of L. armatus showing two mitochondria
and no modifications of the central axonemal apparatus (× 40 000).
260
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
261
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
tube that is basally thick, then abruptly thinner with a
characteristic helical ridge that completes 1.5 turns; it
contains a thin secondary tube and a wide basal chamber
(Fig. 2E). The nucleus, apically a coiled thread and basally
twisted, decreases considerably in diameter from the basis to
the convex apex (Fig. 2F and G). The midpiece contains four
roundish mitochondria (Fig. 2H) and only tetragon fibres are
present in the central axonemal apparatus.
Isochaetides. The euspermatozoon of Isochaetides arenarius
(Tubificinae) has been described in detail elsewhere
(Ferraguti et al. 2002), and only some pictures will be given
here for reference. The slender acrosome (length /width ratio
about 5.8) has an acrosome tube that is basally thick then is
gradually thinner and shows, at least in the spermatozoa
found inside the spermatheca, a deep withdrawal of the acrosome vesicle surrounding a long acrosomal rod (Fig. 2I). The
nucleus is a long cylinder, with a corkscrew-shaped apical
part, a twisted middle part and a straight basal portion, and
its diameter decreases considerably from the basis to the convex apex (Fig. 2I, J and K). The midpiece contains five roundish mitochondria (Fig. 2L) and finally the flagellum presents
only tetragon fibres.
Paraspermatozoa. Paraspermatozoa were observed, among
Limnodriloidinae, only in Smithsonidrilus and Limnodriloides
species. The differences between euspermatozoa and paraspermatozoa were consistent in all nine species of these genera
examined. In all of these, the paraspermatozoa are devoid of
an acrosome (Fig. 2P inset), and their short (length 5–10 µm)
and straight nuclei are elongated cone-shaped and regularly
condensed (Fig. 2P). The mitochondria are less numerous
(Fig. 2R). Finally, the tails lack a basal cylinder, and the
plasma membrane is often characteristically separated from
the axoneme (Fig. 2Q).
In all Limnodriloides species observed, the paraspermatozoa
have basally concave nuclei, each with a nearly circular outline (cross-section); and in all except L. armatus there are a
midpiece with two straight and roundish mitochondria
(Fig. 2R) and a tail with tetragon fibres; the paraspermatozoa
of L. armatus have no modifications of the central axonemal
apparatus.
In all the Smithsonidrilus species examined, the paraspermatozoa have basally convex nuclei with an irregular outline
in cross-section, except the paraspermatozoa of S. westoni and
S. sacculatus, which have, respectively, basally concave nuclei
or nuclei with an almost circular outline. The midpiece
contains three (two in S. luteolus) straight and roundish
mitochondria in all species except S. sacculatus in which the
mitochondria are spiral and elongate. The tails have no
modifications of the central axonemal apparatus, except those
of S. luteolus which has tetragon fibres.
The paraspermatozoon of Isochaetides arenarius (Tubificinae) has been described in detail elsewhere (Ferraguti et al.
2002), thus only some pictures will be given here for reference. The acrosome, 0.5 µm long, has no rod, no secondary
tube, and its acrosome vesicle lies completely external to the
tube (Fig. 2M). The cone-shaped nucleus is short (length
3.7 µm), never condensed completely, and with an oval outline in cross-section (Fig. 2M and O). The midpiece contains
three mitochondria, the overall volume of which being about
two times that of the mitochondria in the eusperm (Fig. 2O).
The tail, showing a reduced basal cylinder inside the basal
body of the flagellum, has no modification of the central
axonemal apparatus, and the plasma membrane is characteristically separated from the axoneme (Fig. 2N and O).
Spermatozeugmata. Inside each of the four different genera
studied, a great variety in the shape and the structure of
sperm aggregates in the spermathecae was observed. Thalassodrilides bruneti and T. gurwitschi are obvious exceptions as
they have no spermathecae.
Based on their structure, these different spermatozeugmata can be grouped into at least three classes:
1 Thalassodrilides ineri and Doliodrilus sp. have spermatozeugmata formed only by euspermatozoa, which are packed
together and arranged in parallel (Fig. 3C). In T. ineri, the
single spermatozeugma is connected to a peculiar proteinaceous structure (Righi & Kanner 1979; Ferraguti et al. 1989),
whereas in the spermathecae of Doliodrilus sp., the few spermatozeugmata are seahorse-shaped (Wang & Erséus 2003).
2 In the Smithsonidrilus species studied, paraspermatozoa
and euspermatozoa form separate spermatozeugmata inside
the spermathecae.
Fig. 3 A–G. Tubificine (A, B) and limnodriloidine (C–G) spermatozeugmata. —A. Cross-sectioned spermatozeugma of Isochaetides arenarius,
with the axial cylinder (a) formed by parallel eusperm and the cortex (c) formed by tightly packed parasperm (× 2200). The arrow indicates
helliptical cross-sections of parasperm nuclei. —B. Cell junctions connect the parasperm tails in the cortex of Isochaetides arenarius (× 45 000).
—C. Cross-section of the large spermatozeugma formed by eusperm inside the spermatheca of Doliodrilus sp. (× 2200). —D. Cross-sectioned
spermatozeugma of Limnodriloides monothecus, formed by both eusperm and parasperm. Note the large axial cylinder (a) and the narrow cortex
(c), formed by a few rows of parasperm tails (× 5500). The arrow indicates round cross-sections of parasperm nuclei [compare with (A)].
—E. There are no visible junctions connecting the plasma membranes of neighbouring parasperm tails inside the cortex of L. monothecus
(× 45 000) [compare with (B)]. —F. Cross-section of part of the large and poorly organized eusperm aggregate of Smithsonidrilus luteolus with
its characteristic whirlpool-like arrangement (× 3000). —G. Cross-sectioned spermatozeugma of S. luteolus formed by parasperm only (× 7000)
[compare with (A, D)].
262
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
263
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Fig. 4 A, B. Trees obtained with a parsimony analysis of a single data set. —A. Strict consensus of 56 most parsimonious trees based on the
somatic data alone (all unordered). —B. Most parsimonious tree resulting from cladistic analysis of the spermatozoal characters only (all
unordered). The numbers above the internodes are bootstrap values. Branch lengths are proportional to amount of change.
Particularly in S. hummelincki and S. sacculatus, the paraspermatozoa form small and irregular bundles and their shape
is slightly modified in comparison with those observed at the
sperm funnels, whereas the euspermatozoa form larger and
more organized spermatozeugmata. In S. luteolus, the paraspermatozoa are organized in spermatozeugmata formed by
a central axis of nuclei surrounded by a large cylinder of tails,
and the latter with plasma membranes largely separated from
the axonemes (Fig. 3G). These parasperm aggregates, which
superficially resemble the tubificine spermatozeugmata but
consist of only one sperm type, are distributed among less
organized eusperm aggregates. The latter are spermatozeugmata with a characteristic whorl-like arrangement of euspermatozoa (Fig. 3F). We found only empty spermathecae in
S. westoni.
3 In all the Limnodriloides species studied, there is only one
kind of spermatozeugmata, and it is formed by a combination
of the two sperm types. These spermatozeugmata vary in
size and shape, from spindle to club-shaped, and from stout
to slender (Erséus 1982), but they have the same general
organization. As for tubificines (Fig. 3A), it is possible to
recognize a large ‘axial cylinder’, formed by parallely
264
arranged euspermatozoa, surrounded by a narrower ‘cortex’,
formed by a few rows of helically arranged paraspermatozoa
(Fig. 3D). However, in contrast to the situation in many
tubificines (Fig. 3B), there are no visible junctions connecting the plasma membranes of neighbouring parasperm tails
inside the cortex (Fig. 3E).
4 Tubificine spermatozeugmata (here, for comparison
in Fig. 3A and B, Isochaetides arenarius) have large spermatozeugmata formed by eusperm surrounded by parasperm.
Parasperm tails are connected by prominent cell junctions.
Analysis of individual data sets
Somatic data. The branch-and-bound search on the somatic
data alone (all characters unordered) yielded 56 equally most
parsimonious trees (MPTs), all with 66 steps and a consistency index (CI) of 0.576, and a retention index (RI) of
0.728. Ordering some of the somatic characters (see above),
the analysis resulted in 15 MPTs, also with 66 steps with
CI = 0.576 and RI = 0.733. The strict consensus trees with all
characters unordered is shown in Fig. 4A. A large basal polytomy does not allow any resolution at the subfamily level,
among the Phallodrilinae, Tubificinae and Limnodriloidinae.
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Table 2 Results of the incongruence length
differences (ILD) tests for comparison
among two data partitions.
Comparison
N°MPT
L
CI
P-value for ILD test
Eusperm–parasperm characters (1–13 vs. 14–26)*
Sperm–somatic characters (1–29 vs. 30–58)
1
1
44
145
0.636
0.52
0.334
0.018
*Eight characters are uninformative due to the deletion of some taxa that have no parasperm. L, length; CI, consistency
index.
Inside this last subfamily, only the three Thalassodrilides species
with Doliodrilus sp. and a few species belonging to Smithsonidrilus and Limnodriloides form distinct groups in all trees.
Inside the Tubificinae, Isochaetides arenarius and Tubifex
tubifex (Müller, 1774) are always sister species, but the position of Clitellio arenarius (Müller, 1776) is ambiguous inside
the large basal polytomy. The majority rule bootstrap tree is
similar to the strict consensus tree: the majority of monophyla present in all MPTs have reasonable to good support.
The main difference is the lack of bootstrap support for the
three Thalassodrilides species and Doliodrilus sp.
Spermatozoal data. The parsimony analysis of the 29 sperm
characters resulted in a single MPT, with 67 steps and
CI = 0.567, RI = 0.758 (Fig. 4B). A similar tree was obtained
using only the eusperm characters. Inside the outgroups
Pectinodrilus molestus with Coralliodrilus rugosus (Phallodrilinae)
group on their own. The Tubificinae, at the basis of one of
the two ingroup clades, is paraphyletic. The monophylum
formed by the three Thalassodrilides species and Doliodrilus
sp., being the sister group of all other ingroup taxa, makes the
Limnodriloidinae paraphyletic. Inside this last subfamily, the
genera Smithsonidrilus and Limnodriloides, but not Thalassodrilides, are polyphyletic. The bootstrap analysis gave notable
support only to a few monophyla, those formed by the three
Thalassodrilides species and Doliodrilus sp., L. uniampullatus
and L. tenuiductus and, to a lesser degree, Pectinodrilus molestus
and Coralliodrilus rugosus, respectively.
Congruence. Only one of the two comparisons between the
different data sets yielded statistically significant results at the
P < 0.05 level (Table 2). In fact, sperm and somatic characters
were significantly incongruent (P-value = 0.018), whereas the
incongruence between eusperm and parasperm data was not
statistically significant (P-value = 0.334).
Combined data analysis
Spermatozoal and somatic data. Somatic and spermatozoal
characters (all unordered), used together, produced a single
MPT, with 144 steps and CI = 0.528, RI = 0.695 (Fig. 5). As
in the tree based on only the sperm characters, Pectinodrilus
molestus groups with Coralliodrilus rugosus (Phallodrilinae)
outside the basis of the tubificine–limnodriloidine ingroup.
The Limnodriloidinae stays paraphyletic: the clade formed
by Doliodrilus sp. and the monophyletic Thalassodrilides is the
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
sister group of the assemblage formed by the Tubificinae and
the other limnodriloidine taxa. The Tubificinae is a monophyletic group: Clitellio arenarius emerges as the sister taxon
to Isochaetides arenarius + Tubifex tubifex. Inside the Limnodriloidinae, Limnodriloides is paraphyletic; L. armatus is sister
to a group formed by the monophyletic Smithsonidrilus and
the other more apomorphic Limnodriloides species. The majority rule bootstrap tree is more resolved than that obtained
with the sperm characters only: there are more smaller
branches that ‘survive’ the bootstrapping. However, the
sister group position of the Phallodrilinae with respect to the
ingroup is not resolved; there is no bootstrap support for
the group Tubificinae + Limnodriloidinae. The position of
the Tubificinae vis-à-vis Limnodriloidinae is also unresolved.
Ordering some of the somatic and spermatozoal characters
(see above), the analysis resulted in two MPTs, each with 146
steps, CI = 0.521, RI = 0.694. One tree is the same as that
obtained with all characters unordered. The second MPT
differs significantly: the Limnodriloidinae is monophyletic,
i.e. the Tubificinae clade is sister to the Limnodriloidinae,
while the three Thalassodrilides species and Doliodrilus sp.
form a sister group to the rest of the limnodriloidines (Fig. 6).
However, monophyly of the Limnodriloidinae is not supported by the bootstrap analysis.
Discussion
The double sperm line
Among the five tubificid subfamilies the presence of a double
sperm line is now established not only in the Tubificinae
(Ferraguti 2000: Table 1), but also in the subfamily Limnodriloidinae: the species belonging to the genera Smithsonidrilus and Limnodriloides produce a dichotomous sperm line.
The parasperm of limnodriloidine species are more similar
to each other than the eusperm, as in the Tubificinae. In both
subfamilies, the differences between parasperm and eusperm
are associated with the same parts of the cell; in parasperm
the various acrosome structures are reduced, the nuclei are
smaller and simpler, the number of mitochondria is lower and
there are differences in the axoneme and in the plasma membrane around it.
However, there are some differences between the parasperm of Tubificinae and Limnodriloidinae. In the limnodriloidine parasperm, the acrosome is absent (Fig. 2P inset),
whereas in tubificines this structure is vestigial (Fig. 2M).
Furthermore, the nucleus is longer (about 10 µm compared
265
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Fig. 5 The single most parsimonious tree resulting from cladistic
analysis of combined sperm and somatic characters (all unordered).
The numbers above the internodes are boot-strap values. Branch
lengths are proportional to amount of change.
with 3 µm) and fully condensed in Limnodriloidinae (Fig. 2P).
The parasperm mitochondria are less numerous as in Tubificinae (Fig. 2O and R), but generally in Limnodriloidinae
they have the same volume as that of eusperm. A structure
similar to the basal cylinder (Ferraguti 1984a), which is
present, albeit reduced, in the basal bodies of tubificine
parasperm (Fig. 2N), is completely absent in the parasperm of
limnodriloidines (Fig. 2Q). Finally, in some limnodriloidine
species, the plasma membrane is not separated from the
parasperm axoneme, whereas in other species it is separated
only for a short tract (Fig. 2Q).
Tubificid parasperm show some morphological similarities
with other parasperm models, both in annelids and other
invertebrates. Among Polychaeta, for example, the parasperm of different Protodrilus species (von Nordheim 1989) and
of Siboglinum ekmani, a perviate pogonophoran (Franzén
1973), as compared with the eusperm, have a reduced
acrosome, a simpler and not fully condensed nucleus and a
simpler midpiece with larger mitochondria.
Even if ‘the “invention” of dichotomous sperm lines has
been an independent phenomenon [ … ] which occurred many
times probably as an answer to different evolutionary pressures’ (Ferraguti 2000: 163) in many invertebrates groups
( Jamieson 1987), the present analyses of combined somatic
266
and spermatozoal characters (see Figs 5 and 6) support the
idea that in tubificids the double sperm line has originated
only once, in a common ancestor of the Tubificinae, and the
genera Smithsonidrilus and Limnodriloides; i.e. the double sperm
line of tubificines and limnodriloidines may well be homologous.
In one of the trees resulting from the parsimony analysis
with some characters ordered (Fig. 6), the double sperm line
appears to be secondarily lost at the basis of the Thalassodrilides + Doliodrilus clade. Although it is difficult to envisage a
reversal to a single sperm line from such a complex model of
dichotomous spermatogenesis, a molecular analysis of 18S
rDNA suggested that the same three Thalassodrilides species
plus a species of Doliodrilus (D. tener) form a terminal branch
close to some species of Smithsonidrilus (Erséus et al. 2002).
Moreover, the secondary loss of a multiple sperm line is a
phenomenon not so unusual among the invertebrates. Inside
the Gastropoda (Mollusca) there is sperm polymorphism in
many groups belonging to various orders of Prosobranchia
(Healy & Jamieson 1981; Giusti & Selmi 1982). Inside
Prosobranchia, parasperm models diversified in the Caenogasteropoda, but they have completely disappeared in the
more apomorphic Heterobranchia (Buckland-Nicks 1998).
Also, in different species of stalk-eyed flies (Diopsidae),
sperm dimorphism is present as an ancestral character associated with functional spermathecae, and the secondary
reduction or loss of sperm storage function by spermathecae
appears to coincide with the loss of sperm dimorphism
(Presgraves et al. 1999).
Spermatozeugmata
In all species of Tubificinae studied to date, with the exception of Aulodrilus and related forms (Brinkhurst 1990), there
are typical spermatozeugmata, each formed by two sperm
types (Fig. 3A) (Ferraguti 2000). Within the Limnodriloidinae, this architecture is present only in Limnodriloides,
whereas all examined species of the genus Smithsonidrilus
produce separate spermatozeugmata of two kinds, each
formed by one sperm type. The spermatozeugmata of the
genus Limnodriloides, in fact, although more variable in shape
and dimension (Erséus 1982), share the same general architecture with those of Tubificinae; both are composed of two
different layers: an internal axial cylinder containing eusperm
and an external cortex composed of parasperm, the latter
packed together (Braidotti & Ferraguti 1982; Ferraguti
1998). However, the cortex of the Limnodriloides spermatozeugma is formed by fewer rows of parasperm tails than in
tubificines, resulting in a thinner sheath around the axial
cylinder, and the Limnodriloides parasperm tails are not
connected by a junctional complex. Inside the tubificine
spermatozeugmata, with one known exception for Clitellio
arenarius (Ferraguti & Ruprecht 1992), there is a complex
pattern of junctions inside a wider cortex (Fig. 3A and B)
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Fig. 6 One of the two most parsimonious trees resulting from cladistic analysis of combined sperm and somatic characters (characters 5, 35,
40, 45, 49, 52 ordered). The numbers on the left of the internodes are bootstrap values. Branch lengths are proportional to amount of change.
(Braidotti et al. 1980; Ferraguti 1998; Ferraguti et al. 2002). It
is interesting to note that the junctions among parasperm
tails are more complex and differentiated in the tubificine
freshwater genera (Tubifex, Heterochaeta, Isochaetides) than in
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
the marine genus Tubificoides (Ferraguti et al. 1989, 2002).
Therefore, the absence of a junctional complex in the
spermatozeugmata of the genus Limnodriloides, which is
exclusively marine, is perhaps not surprising.
267
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Within the other tubificid subfamilies Rhyacodrilinae and
Phallodrilinae, several types of eusperm bundle have been
found (Ferraguti et al. 1994; Erséus & Ferraguti 1995). They
differ from the spermatozeugmata present in Doliodrilus sp.
and T. ineri in being less structured.
The morphology of the spermatozeugmata formed by
eusperm only inside the spermathecae of Doliodrilus sp. and
T. ineri differs in many aspects in these two species. The parsimony analysis indicates that these spermatozeugmata may
even have arisen independently.
On the other hand, in spite of the differences in morphological details, the analysis suggests that the spermatozeugmata composed of both eusperm and parasperm may be
homologous in the Tubificinae and Limnodriloides. Furthermore, our analysis interprets the simpler spermatozeugmata
observed in Smithsonidrilus as the result of an apomorphic
secondary transformation of the tubificine-like spermatozeugmata in Limnodriloides.
Character analysis
To discuss the patterns of sperm and somatic characters,
under a total evidence approach, the phylogram in Fig. 6
has been selected. In it, as in previous phylogenetic hypotheses (Erséus 1990; Brinkhurst 1994), the Tubificinae and
Limnodriloidinae are sister groups. However, as noted above,
this tree is only one of two equally parsimonious trees
resulting from the analysis, and we do not judge whether it
should be regarded as the ‘preferred’ hypothesis of the two.
Spermatological character patterns. The presence of a thin secondary tube inside the acrosome (character 6) is the sperm
character that defines the Tubificinae and Limnodriloidinae
clade in relation to the outgroup. The presence of a double
sperm line (character 27) is another character state common
to this clade, but with a reversal for the terminal group comprising the Doliodrilus and Thalassodrilides.
Autapomorphic spermatological character states for the
subfamily Tubificinae refer only to parasperm, while the tubificine eusperm remain plesiomorphic. Autapomorphies are
the presence of a vestigial acrosome (character 14), the presence of a shorter nucleus (character 15) and the presence of a
basal cylinder inside the proximal region of the axoneme
(character 26). A nearly oval nuclear outline (character 18)
and the presence of a junctional complex among the parasperm tails in the spermatozeugmata (character 29) are synapomorphies for Tubifex tubifex and Isochaetides arenarius.
The presence, in the eusperm, of an acrosomal tube of
uniform thickness and basally with limen (character 2) is the
only synapomorphic state for all Limnodriloidinae except the
Thalassodrilides + Doliodrilus clade which possesses an autapomorphic basally thick, then abruptly thinner acrosome tube,
and lacking a limen (character 2). An acrosomal tube of
268
uniform thickness and with a limen seems to have evolved
convergently in Rhizodrilus russus.
Our analysis suggests that the genus Limnodriloides is either
paraphyletic (as in the example tree shown in Fig. 6) or
polyphyletic. The L. armatus eusperm is different from that
of the other Limnodriloides species, instead being similar to
tubificine eusperm. Moreover, in L. armatus, the parasperm
belong to the ‘limnodriloidine type’, but they have areas of
uncondensed chromatin in their nuclei (character 19), and a
plasma membrane expanded for a long tract around the
axoneme (character 24).
Most characters of the eusperm of the Thalassodrilides and
Doliodrilus species are plesiomorphic; the synapomorphies
shared with the ‘typical Limnodriloides’ (i.e. excluding L.
armatus) and S. sacculatus are an acrosome with a protuberant
acrosome rod (character 4) and a vesicle external to the tube
(character 5). The ordering of character 5 (acrosome vesicle
withdrawal), unlike that imposed on the somatic characters
(see above), is not reflected in the tree of Fig. 6. Character 5
changed in its evolution in a discontinuous way, from
completely inside the acrosome tube to completely outside it,
at the basis of the ‘typical Limnodriloides’ and Doliodrilus +
Thalassodrilides clade.
Autapomorphies for the ‘typical Limnodriloides’ + Smithsonidrilus clade are the presence in the eusperm of a stout
(character 1) and short acrosome (character 3). A stout acrosome occurs also as the result of convergence, in Rhizodrilus
russus. No autapomorphism defines the ‘typical Limnodriloides’
clade; but the presence of parasperm with a filiform apical
shape of the nucleus is synapomorphic for L. tenuiductus and
L. uniampullatus.
There are two synapomorphic character states for all
species of Smithsonidrilus except S. sacculatus: the acrosome
vesicle is only partially withdrawn inside the acrosome tube
(character 5) and there are parasperm with an irregular
nuclear outline in cross-section (character 18); the plasma
membrane expanded around the nuclear apex of the parasperm (character 20) characterizes only the sister species S.
hummelincki and S. westoni.
Many spermatological characters show homoplasy (convergence or reversal to ancestral states) as is evident from the
low CI value of the MPTs in Fig. 6, and this may be a common feature of sperm in lower level tubificid phylogenies
(Erséus & Ferraguti 1995). Moreover, as also shown from the
high RI value, most spermatozoal features show a low level of
homoplasy in their evolution: they are homoplasic characters
that give support for tree topology or, following de Pinna
(1991), secondary homology with a low level of generality.
Only seven characters show a high level of homoplasy in their
evolution: four of them refer to eusperm, three to parasperm.
The eusperm characters are those dealing with the shape of the
nucleus (characters 7, 8, 9) and the number of mitochondria
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
(character 10), while the parasperm ones are the presence of
transformed nuclei in the spermatozeugma (character 21),
the number of mitochondria (character 22) and the presence
of a basal body with dense material (character 23).
The trend towards an increased withdrawal of the acrosome vesicle in the evolution of the oligochaete sperm types
( Jamieson et al. 1987) was not found in this analysis. Among
the oligochaetes, inside the ‘more evolved’ tubificids, a completely withdrawn vesicle (inside the acrosome tube) is the
plesiomorphic condition, whereas a more or less external vesicle is the apomorphic state characterizing all limnodriloidine
species.
Somatic character patterns. The presence/absence of coelomocytes (character 37) and the way in which the vas deferens
enters into the atrium (character 40) are the somatic characters that identify the Phallodrilinae + Tubificinae + Limnodriloidinae clade, when the tree is rooted at the Rhyacodrilinae.
The position of spermathecal pores in the middle part of the
spermathecal segment (character 34) seems to be the only
character state common to Tubificinae and Limnodriloidinae, excluding the Thalassodrilides + Doliodrilus clade.
A pendant penis located centrally inside a well-developed
copulatory sac (character 53) characterizes the subfamily
Tubificinae, and the presence of an ‘ectoprostate’ gland
(character 41) is the synapomorphic state for Tubifex tubifex
and Isochaetides arenarius.
The presence of an atrium bearing a lobed, broadly
attached prostate gland derived from peritoneal cells
(Gustavsson 2002) is the synapomorphy for all members of
the Limnodriloidinae, and this gland communicates with a
prostatic pad inside the atrial ampulla (character 43). The
presence of paired oesophageal diverticula in segment IX
(character 39) is a state common to all Limnodriloidinae
analysed here except the Thalassodrilides + Doliodrilus clade. The
presence of a barrel-shaped, dilated portion of the oesophagus (Gustavsson & Erséus 1999) is the only autapomorphy
for the Thalassodrilides + Doliodrilus group (character 38),
while the presence of a well-developed and strongly muscular
atrial duct sac unites T. gurwitschi and T. ineri (character 52).
As was the case with the sperm characters, L. armatus
also differs in many somatic characters from the other
Limnodriloides species (the ‘typical Limnodriloides’ group). No
somatic synapomorphy supports the ‘typical Limnodriloides’ +
Smithsonidrilus clade, but the subapical entrance of the vas
deferens on the dorsal surface of the atrium (character 40)
and the presence of a pseudopenial papilla (character 54) are
autapomorphies for the ‘typical Limnodriloides’ group. Dilatation
of the inner end of the atrium (character 47) is the only state
shared by all Smithsonidrilus species excluding S. sacculatus.
Only three somatic characters show a high level of homoplasy in their evolution: the shape of the teeth of bifid chaetae
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
(character 30), the absence/presence of modified spermathecal chaetae (character 32) and the size/shape ratio of the
prostatic pad (character 44). Many others show a low level of
homoplasy. The position of spermathecal pores (character
35), the atrial ciliation (character 46), the presence of a
copulatory organ associated with an ectodermal invagination at male pores (character 53) and the presence of an ‘exterior’ spermathecal vestibule (character 58) are homoplasic
because there is no uniform character state in the outgroup
species.
Congruence test and parsimony analysis. The inclusion of parasperm characters in our data set, even if not all taxa have a
double sperm line, proved useful in the phylogenetic analysis.
Thus, the presence of parasperm constitutes an important
character shared by Tubificinae and Limnodriloidinae, and
the ultrastructure of these cells contributes to a better discrimination of one subfamily from the other. Besides, as
shown from the ILD test (Table 2), parasperm characters are
not significantly incongruent with those of eusperm ones.
This suggests the possibility that the two sperm types carry
information of a similar phylogenetic structure. On the other
hand, the ILD test shows significant incongruence between
sperm and somatic characters (Table 2). This may indicate
that both data sets contain some internal incongruence
(homoplasy) and that the phylogenetic signal (internal congruence or synapomorphies) is, in one particular data set,
stronger for some groups, while the signal is stronger for
other groups in the other data set. The result of the combined
data set, in a total evidence approach, gives more phylogenetic information than any of the separate data sets. The
sperm tree (Fig. 4B), which differs from the somatic one
(consensus tree in Fig. 4A) by being more resolved, contains
only three well-supported clades: the phylogenetic structure
in the sperm data is thus weak. The somatic tree is different
from the sperm tree and also the patterns of bootstrap support are different. There are characters in the somatic data
which give good support to groups not so strongly supported
in the sperm tree. For instance, the somatic tree has a bootstrap support of 80 for a large group, comprising all species
except Rhizodrilus russus and Monopylephorus limosus; this
group exists in the sperm tree too, but without bootstrap
support. In the somatic tree, there is strong support (94)
for L. tenuiductus + L. monothecus + L. uniampullatus, which is
a group that is ‘indicated’ in the sperm tree too, but with
S. sacculatus inserted. Smithsonidrilus hummelincki and S. westoni,
which in the sperm tree do not form a group, are located
adjacent to each other at the base of a large clade in this tree;
in the somatic trees the two species comprise a strongly
supported clade (97). In contrast, the sperm tree contains
support for groups that are not supported by the somatic data: the
group formed by Doliodrilus sp. and the three Thalassodrilides
269
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
species is supported by the sperm characters but is not
supported (although it exists) in the somatic tree. In the tree
based on the combined data with some characters ordered
(Fig. 6), the number of groups supported by bootstrap values
over 50 (values between 54 and 93) has greatly increased. A
group comprising all species except Rhizodrilus russus and
Monopylephorus limosus is well supported. The group Pectinodrilus molestus + Coralliodrilus rugosus is well supported (89), an
additive effect of the already existing but less supported
groups by the somatic and sperm characters alone (see
Fig. 4A and B). Moreover, there is a ‘trade off’ in the group
formed by L. uniampullatus + L. tenuiductus + L. monothecus,
which in the combined tree only has a bootstrap value of 78,
while the support was 94 using the somatic data alone. The
latter fact somehow had to be balanced by the ‘interference’
of S. sacculatus in the sperm data (see above). In conclusion,
even if the two data sets are significantly incongruent, they
both contribute to a more resolved and more strongly supported phylogenetic hypothesis when combined.
To conclude, our analysis, based on combined somatic and
spermatological characters, supports the notion indicated
previously (Erséus 1990; Brinkhurst 1994; Erséus & Ferraguti
1995; Ferraguti & Erséus 1999) that Tubificinae and
Limnodriloidinae are closely related, possibly sister groups
(Fig. 6). However, this result is neither strongly supported by
the data (there is no bootstrap support >50 for a clade comprising the two subfamilies), nor based on more than a few
selected species. There are about 150 species of Tubificinae
(rough estimation), and over 100 species of Limnodriloidinae
known (C. Erséus, unpublished observation), and we cannot
be certain that our study has covered more than a fraction of
the total character variation in the taxa. Nevertheless, our
analysis has shown that sperm ultrastructure exhibits great
similarities in tubificine and Limnodriloides spermatozeugmata, and that sperm characters may be important to broaden
the basis of evidence in phylogenetic reconstruction.
Acknowledgements
This research has been supported by a grant (Molecular evolution and markers for phylogenesis and adaptation) from
MURST (Rome) to MF. We are indebted to P. Martin (Institut Royal des Sciences Naturelles de Belgique, Belgium) for
providing material of Isochaetides arenarius. The third author
is grateful to K. Ruetzler (National Museum of Natural History, Washington, DC), and the directors and staff of Carrie
Bow Cay Field Laboratory (Belize), Heron Island Research
Station (Queensland, Australia), and Caribbean Marine
Research Center (Lee Stocking Island, Great Exuma, Bahamas), for providing excellent facilities for field work; to
friends and colleagues who assisted in the collection of oligochaetes (in particular, N. Dubilier, O. Giere, Z. Gong, and H.
Wang), and to the Swedish Research Council (and the former
270
Swedish Natural Science Research Council), and the Perry
Foundation, Inc. (operator of Caribbean Marine Research
Center) for financial support.
References
Braidotti, P. & Ferraguti, M. (1982). Two sperm types in Tubifex
tubifex spermatozeugmata. Journal of Morphology, 171, 123–136.
Braidotti, P., Ferraguti, M. & Fleming, T. P. (1980). Cell junctions
between spermatozoa flagella within the spermatozeugmata of
Tubifex tubifex (Annelida, Oligochaeta). Journal of Ultrastructure
Research, 73, 299 –309.
Brinkhurst, R. O. (1990). A phylogenetic analysis of Tubificinae.
Canadian Journal of Zoology, 69, 3 9 2 –397.
Brinkhurst, R. O. (1994). Evolutionary relationships among the Clitellata: an update. Megadrilogica, 5, 109 –112.
Buckland-Nicks, J. (1998). Prosobranch parasperm: sterile germ
cells that promote paternity? Micron, 29, 267–280.
Cardini, A., Ferraguti, M. & Gelder, S. R. (2000). A phylogenetic
assessment of the branchiobdellidan family Branchiobdellidae
(Annelida, Clitellata) using spermatological and somatic characters. Zoologica Scripta, 29, 347 – 366.
Ermak, T. H. & Eakin, R. M. (1976). Fine structure of the cerebral
and pygidial ocelli in Chone ecaudata (Polychaeta: Sabellidae).
Journal of Ultrastructural Research, 54, 243 – 260.
Erséus, C. (1982). Taxonomic revision of the marine genus Limnodriloides (Oligochaeta: Tubificidae). Verhandlungen des Naturwissenschaflichen Vereins in Hamburg (Neue Folge), 25, 207 –277.
Erséus, C. (1983). Taxonomic studies of the marine genus Marcusaedrilus Righi & Kanner (Oligochaeta, Tubificidae), with description
of seven new species from the Caribbean area and Australia.
Zoologica Scripta, 12, 25– 36.
Erséus, C. (1990). Cladistic analysis of the subfamilies within the
Tubificidae (Oligochaeta). Zoologica Scripta, 19, 57– 63.
Erséus, C. & Ferraguti, M. (1995). The use of spermatozoal
ultrastructure in phylogenetic studies of Tubificidae. In B. G. M.
Jamieson, J. Ausio & J. L. Justine (Eds) Advances in Spermatozoal
Phylogeny and Taxonomy (pp. 189–201). Paris: Mémoires du
Muséum National d’Histoire Naturelle.
Erséus, C., Källersjö, M., Ekman, M. & Hovmöller, R. (2002). 18S
rDNA phylogeny of Tubificidae and its constituent taxa (Clitellata): dismissal of the Naididae. Molecular Phylogenetics and Evolution, 22, 414–422.
Erséus, C., Prestegaard, T. & Källersjö, M. (2000). Phylogenetic
analysis of Tubificidae (Annelida, Clitellata) based on 18S rDNA
sequences. Molecular Phylogenetics and Evolution, 15, 381– 389.
Farris, J. S., Källersjö, M., Kluge, A. G. & Bult, C. (1995). Testing
significance of incongruence. Cladistics, 10, 315 –319.
Felsenstein, J. (1985). Confidence limits on phylogenies: an
approach using the bootstrap. Evolution, 39, 783 –791.
Ferraguti, M. (1984a). Slanted centriole and transient anchoring
apparatus during the spermiogenesis of an oligochaete (Annelida).
Biologie Cellulaire, 52, 175 –180.
Ferraguti, M. (1984b). The comparative ultrastructure of sperm
flagella central sheath in Clitellata reveals a new autapomorphy of
the group. Zoologica Scripta, 13, 201 – 207.
Ferraguti, M. (1998). The double spermatogenesis in Annelida and
Pogonophora. In K. W. Wolf (Ed) Double Spermatogenesis. The
Theory and a Collection of Experiments. Reading: Harwood Academic Publishers.
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Ferraguti, M. (2000). Euclitellata. In K. G. Adyiodi & R. G. Adyiodi
(Eds) Reproductive Biology of Invertebrates, Vol. IX. Part B: Progress
in Male Gamete Ultrastructure and Phylogeny (pp. 165–222).
Chichester: John Wiley.
Ferraguti, M. & Erséus, C. (1999). Sperm types and their use for a
phylogenetic analysis of aquatic clitellates. Hydrobiologia, 402,
225–237.
Ferraguti, M., Grassi, G. & Erséus, C. (1989). Different models of
tubificid spermatozeugmata. Hydrobiologia, 180, 73 – 82.
Ferraguti, M., Marotta, R. & Martin, P. (2002). The double sperm
line in Isochaetides (Annelida, Clitellata, Tubificidae). Tissue and
Cell, 34, 305–314.
Ferraguti, M. & Ruprecht, D. (1992). The double sperm line in the
tubificid Clitellio arenarius. Bollettino di Zoologia, 59, 349 –362.
Ferraguti, M., Ruprecht, D., Erséus, C. & Giere, O. (1994). An
ultrastructural overview of tubificid spermatozoa. Hydrobiologia,
278, 165–178.
Franzén, Å. (1973). The spermatozoon of Siboglinum (Pogonophora). Acta Zoologica, 54, 179 –192.
Giusti, F. & Selmi, M. G. (1982). The atypical sperm in prosobranch
molluscs. Malacologia, 22, 171–181.
Gustavsson, L. (2002). The development of the genital ducts in Limnodriloidinae (Tubificidae, Clitellata). Journal of Zoology, 257, 83–
91.
Gustavsson, L. & Erséus, C. (1997). Morphogenesis of the genital
ducts and spermathecae of Clitellio arenarius, Heterochaeta costata,
Tubificoides benedii (Tubificidae) and Stylaria lacustris (Naididae)
(Annelida, Oligochaeta). Acta Zoologica, 78, 9 – 31.
Gustavsson, L. & Erséus, C. (1999). Morphology and phylogenetic
implications of oesophageal modifications in the Limnodriloidinae
(Oligochaeta, Tubificidae). Journal of Zoology, 248, 467– 482.
Healy, J. M. & Jamieson, B. G. M. (1981). An ultrastructural examination of developing and mature paraspermatozoa in Pyrazus
ebeninus (Mollusca, Gastropoda, Potamididae). Zoomorphology,
112, 101–119.
Jamieson, B. G. M. (1978). A comparison of spermiogenesis and
spermatozoal ultrastructure in megascolecid and lumbriculid
earthworms (Oligochaeta: Annelida). Australian Journal of Zoology,
26, 225–240.
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
Jamieson, B. G. M. (1981). The Ultrastructure of the Oligochaeta.
London: Academic Press.
Jamieson, B. G. M. (1987). A biological classification of sperm types,
with special references to annelids and molluscs, and an example
of spermiocladistics. In H. Mohri (Ed.) New Horizons in Sperm
Cell Research (pp. 311–332). New York: Japan Scientific Society
Press, Gordon and Breach Scientific Publications.
Jamieson, B. G. M., Erséus, C. & Ferraguti, M. (1987). Parsimony
analysis of the phylogeny of some Oligochaeta (Annelida) using
spematozoal ultrastructure. Cladistics, 3, 145–155.
Kluge, A. G. (1989). A concern for evidence and a phylogenetic
hypothesis of relationships among epicrates (Boidae, Serpentes).
Systematic Zoology, 38, 7– 25.
Mickevich, M. F. & Farris, J. S. (1981). The implications of congruence in Menidia. Systematic Zoology, 30, 351– 370.
von Nordheim, H. (1989). Vergleichende Ultrastrukturuntersuchungen der Eu- und Paraspermien von 13 Protodrilus-Arten (Polychaeta, Annelida) und ihre taxonomische und phylogenetische
Bedeutung. Helgoländer Meersuntersuchungen, 43, 113 –156.
de Pinna, M. C. C. (1991). Concepts and tests of homology in the
cladistic paradigm. Cladistics, 7, 367– 394.
Presgraves, D. C., Baker, R. H. & Wilkinson, G. S. (1999). Coevolution of sperm and female reproductive tract morphology in
stalk-eyed flies. Proceedings of the Royal Society of London, 266, 1041–
1047.
Righi, G. & Kanner, E. (1979). Marine Oligochaeta (Tubificidae and
Enchytraeidae) from the Caribbean Sea. Studies of the Fauna of
Curacao and Other Caribbean Islands, 58, 44–68.
Swofford, D. L. (1998). PAUP: Phylogenetic Analysis using Parsimony,
Version 4. Sunderland, MA: Sinauer Associates.
Wang, H. & Erséus, C. (2003). New species of Doliodrilus and other
Limnodriloidinae (Oligochaeta, Tubificidae) from Hainan and
other parts of the northwest Pacific Ocean. Journal of Natural History (in press).
Westheide, W., Purschke, G. & Middendorf, K. (1991). Spermatozoal ultrastructure of the taxon Enchytraeus (Annelida,
Oligochaeta) and its significance for species discrimination and
identification. Zeitschrift fur Systematik und Evolutionforschung, 29,
323–342.
271
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Appendix 1
The 29 sperm characters and character states (Fig. 7).
Fig. 7 The 29 sperm characters and character states.
272
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Fig. 7 Continued.
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
273
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Fig. 7 Continued.
274
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
Fig. 7 Continued.
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
275
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
Appendix 2
The 29 somatic characters and character states.
30 Teeth of bifid chaetae: upper tooth about as long as, or
shorter than, lower tooth (0); upper tooth somewhat (up to
1.5 times) longer than lower tooth (1).
31 Penial chaetae, i.e. modified chaetae associated with male
pores: absent or bifid (0); present, with single-pointed tips (1).
32 Modified spermathecal chaetae, i.e. slender and ectally
grooved, single-pointed chaetae associated with spermathecal
pores: absent (0); present, at least in some individuals (1).
33 Number and position of male pores: paired, and
lateroventral or ventral (0); unpaired, midventral (1). In most
cases, state 1 refers to a median bursa that receives two atrial
ducts at its inner end.
34 Position of spermathecal pores within spermathecal
segment (in Tubificidae, generally segment X): in anterior
part of segment (0); in middle part of segment (1).
35 Peripheral position of spermathecal pores: midventral
(0); lateroventral or lateral (1); dorsal (2). (This character may
be treated as ordered.)
36 Number of spermathecae: two (0); one (1); none (2).
37 Coelomocytes: few or absent (0); numerous and
conspicuous (1).
38 Barrel-shaped, dilated portion of oesophagus, with a reticulate
blood plexus: absent, i.e. oesophagus is a simple unmodified tube,
or modified in some other way (0); present, extending at least
through most of segment IX (1). See Gustavsson & Erséus (1999).
39 Paired oesophageal diverticula: absent, i.e. oesophagus is
a simple unmodified tube, or modified in some other way (0);
present in segment IX (1). See Gustavsson & Erséus (1999).
40 Entrance of vas deferens into atrium: clearly subapical,
on ventral surface of atrium (or on anterior surface when
atrium is erect) (0); more or less on apex of (inner end of)
atrium (1); clearly subapical, on dorsal surface of atrium (2).
(This character may be treated as ordered.)
41 ‘Ectoprostate’, stalked prostate gland derived from
ectodermal (inner) epithelium of atrium: absent (0); present
(1). So far, an ectodermal origin of prostates is only known for
members of the Tubificinae (see Gustavsson & Erséus 1997).
42 Granulation of inner epithelium in atrial duct: none or
partial only (0); inner epithelium densely granulated throughout length of atrial duct (1). This character is applicable to
Limnodriloidinae only; granulation of the atrial epithelium
also occurs in some of the other taxa, but a priori statements
about which parts of their atria correspond to the limnodriloidine ‘atrial ducts’ are difficult to make. Thus, for species
outside the Limnodriloidinae, this character is coded as ‘?’.
43 ‘Mesoprostate’, prostate gland derived from peritoneal
cells: absent, i.e. neither diffuse nor lobed gland present (0);
diffuse gland present, i.e. a more or less continuous layer of
prostatic cells covering most of atrial surface (1); each atrium
bearing lobed, broadly attached, prostate gland, com276
municating with a distinct ‘prostatic pad’, a granulated body
of mesodermal origin, embedded in ventral wall of atrial
ampulla (2). Non-additive character. It has been demonstrated
that the diffuse prostates of one naidid (Stylaria lacustris) and
two rhyacodriline tubificids (including Monopylephorus
rubroniveus) are mesodermal, derived from peritoneal cells
(Gustavsson & Erséus 1997, 1999). As a deduction from these
studies, a mesodermal origin of the similarly diffuse prostates
of Monopylephorus limosus and Rhizodrilus russus is assumed
here. Gustavsson (2002) recently showed that the prostate
glands and prostatic pads of Limnodriloidinae are also
derived from peritoneal cells. State 2 of this character is fully
congruent with the differentiation of the atrium into an ental
ampulla and an ectal duct, a feature shared by all members of
the Limnodriloidinae (Erséus 1982: 267).
44 Size/shape of prostatic pad: small and round to oval, and
unless atrial ampulla is short (see state 0 of character 49),
restricted to only a part of ampulla (0); elongated, and
extending through a greater part of ampulla (1). For taxa
lacking prostatic pads, this character is coded as ‘?’.
45 Position of prostatic pad in atrial wall: pad somewhat
external on ampulla, i.e. it forces outside of atrial wall to
bulge considerably (0); pad more confined to wall of atrial
ampulla, i.e. it neither forces wall to bulge considerably nor
extends markedly into the interior of ampulla (1); pad
extending inwards, markedly reducing lumen of atrial
ampulla (2). (This character may be treated as ordered.) For
taxa lacking prostatic pads, this character is coded as ‘?’.
46 Atrial ciliation: lining epithelium of atrium with few or no
cilia (0); lining of ental part of atrial ampulla bearing cilia (1);
atrium conspicuously ciliated throughout (2).
47 Dilatation of inner end of atrium: absent (0); present, i.e.
inner end of atrium wider, and also more thin walled, than
other parts of atrium (1).
48 Curvature of inner end of atrium: absent, i.e. apical end
of atrium not markedly curved (0); present, i.e. innermost end
of atrium abruptly curved towards ventral side of worm (1).
49 Length and shape of atrial ampulla: short, more or less
spherical (0); of medium length, oval to cylindrical, and
without clearly prolonged ental part (1); long, cylindrical,
with ental end markedly prolonged in relation to extension of
prostatic pad (2). (This character may be treated as ordered.)
For taxa without differentiated atrial ampullae (and prostatic
pads), this character is coded as ‘?’.
50 Coiling of ectal end of atrium: ectal end of atrium short
or moderately elongated, but not forming coils (0); much
prolonged and coiled (1).
51 Unpaired ectal end of atria: absent, atria paired
throughout (0); ectal ends of atria merged into an unpaired
duct leading to (unpaired) male pore (1).
52 Atrial duct sac, i.e. a mesodermal sac enclosing ectal part
of atrium, making latter act as an eversible pseudopenis:
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. Marotta et al. • Phylogeny of Tubificinae and Limnodriloidinae
absent (0); present, but thin and simple, not markedly
muscular (1); well developed and strongly muscular, with
great complexity, also involving an invagination of the body
wall to take part in the pseudopenis (2). (This character may
be treated as ordered.)
53 Copulatory organ associated with an ectodermal
invagination at male pore: absent, i.e. atrium opening directly
to exterior on body wall (0); present as a more or less developed invagination (copulatory sac), into which atrium opens,
but no pendent penis developed (1); present as a pendent
penis located centrally inside a well-developed copulatory
(penial) sac (2). State 1 here may appear to overlap with state
2 of character 52, but the latter feature is interpreted as a
derived condition of the pseudopenis associated with the
atrial duct sac, i.e. not as homologous with the copulatory
organs of character 53.
54 Pseudopenial papilla, i.e. distinct papilla formed in lateral
wall of copulatory sac and bearing opening of atrium: absent
© The Norwegian Academy of Science and Letters • Zoologica Scripta, 32, 3, May 2003, pp255 – 278
(0); present (1). This character is only applicable for taxa with
state 1 of character 53 (see above). For all other taxa, this
character is coded as ‘?’.
55 Folds in wall of copulatory sac: absent (0); present (1). For
taxa without copulatory sacs (i.e. without states 1 or 2 of
character 35), this character is coded as ‘?’.
56 Copulatory sac musculature: indistinct (0); well developed, present as distinct layer enclosing copulatory sac (1).
For taxa without copulatory sacs (i.e. without states 1 or 2 of
character 53), this character is coded as ‘?’.
57 Spermathecal duct: distinct, more or less cylindrical and
well set off from spermathecal ampulla (0); not distinct, i.e.
short and not well set off from ampulla, and if distinguishable, generally triangular in shape (1).
58 ‘Exterior’ spermathecal vestibule, i.e. a secondary invagination at spermathecal pore, forming a distinct chamber
immediately outside spermathecal duct proper, and typically this
chamber is wider than the duct proper: absent (0); present (1).
277
Data matrix.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
Zoologica Scripta, 32, 3, May 2003, pp255– 278 • © The Norwegian Academy of Science and Letters
R. russus
M. limosus
P. molestus
C. rugosus
T. tubifex
C. arenarius
I. arenarius
Doliodrilus sp.
T. bruneti
T. gurwitschi
T. ineri
L. armatus
L. uniampullatus
L. tenuiductus
L. monothecus
L. barnardi
S. sacculatus
S. hummelincki
S. luteolus
S. westoni
0
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
?
1
1
1
1
0
0
0
0
0
1
0
1
0
0
0
0
0
0
?
?
1
1
1
0
1
1
?
1
1
0
0
0
2
2
2
2
2
2
?
?
0
0
0
2
0
0
?
?
0
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
3
3
3
2
2
3
1
1
0
0
2
1
1
3
3
3
3
2
3
0
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
0
1
1
1
0
1
1
1
0
1
1
1
0
0
0
1
0
0
0
1
0
1
1
1
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
2
1
1
?
1
1
2
1
1
1
1
1
2
2
2
1
1
1
1
0
0
0
0
?
0
0
0
0
0
1
1
0
1
1
1
?
?
?
?
0
0
0
?
?
?
?
1
1
1
1
1
1
1
1
1
?
?
?
?
1
1
1
?
?
?
?
0
0
0
0
0
0
0
0
0
?
?
?
?
1
1
1
?
?
?
?
1
0
0
1
1
1
1
1
1
?
?
?
?
0
0
0
?
?
?
?
0
0
0
0
0
1
1
1
0
?
?
?
?
2
0
2
?
?
?
?
0
0
0
0
0
0
1
1
1
?
?
?
?
1
1
1
?
?
?
?
1
0
0
0
0
0
0
0
1
?
?
?
?
0
0
0
?
?
?
?
0
0
0
0
0
0
1
0
1
?
?
?
?
1
1
0
?
?
?
?
0
0
0
0
0
0
0
0
0
?
?
?
?
0
1
1
?
?
?
?
0
0
0
0
0
1
1
0
1
?
?
?
?
0
?
?
?
?
?
?
1
0
0
?
0
0
1
0
1
?
?
?
?
1
1
1
?
?
?
?
1
0
0
2
2
0
2
2
1
?
?
?
?
1
0
0
?
?
?
?
0
1
1
1
1
0
0
1
0
?
?
?
?
1
1
1
?
?
?
?
0
0
0
?
0
0
0
0
0
0
0
0
0
1
1
1
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
0
3
3
3
1
0
0
1
3
3
3
3
3
2
2
2
?
?
?
?
?
1
0
1
?
?
?
?
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
1
1
0
0
0
0
0
0
0
0
1
1
0
0
0
0
1
1
0
1
0
0
0
0
1
1
?
0
?
?
0
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
?
1
?
?
1
0
2
2
2
1
0
0
1
0
0
0
0
0
0
0
0
0
2
2
0
0
1
0
1
0
1
1
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
0
?
1
1
1
1
1
1
1
1
1
1
2
2
2
2
1
1
1
1
0
0
?
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
?
?
?
?
?
?
?
0
0
1
1
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
2
2
2
2
2
2
2
2
2
2
2
2
2
?
?
?
?
?
?
?
0
0
0
1
1
0
0
0
0
0
0
0
0
?
?
?
?
?
?
?
0
1
1
0
2
1
1
1
0
1
0
1
0
0
?
2
1
0
2
?
0
0
0
0
0
0
0
0
0
1
1
1
1
0
?
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0
1
1
1
?
?
0
?
0
0
?
1
?
0
0
0
?
0
0
0
0
1
1
1
?
?
?
?
?
?
?
2
1
1
1
1
1
1
1
1
0
2
2
2
1
?
0
0
0
0
?
0
0
0
0
0
1
1
1
0
0
1
0
1
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
?
0
2
0
?
0
1
2
2
0
?
0
0
0
0
0
0
0
1
1
1
0
?
2
2
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
?
0
?
?
?
?
?
?
0
1
1
1
1
0
0
0
0
1
1
1
?
0
0
?
?
?
?
?
0
0
0
0
0
0
1
1
1
1
1
1
?
0
0
?
?
?
?
?
1
0
0
0
0
1
1
0
1
0
0
0
0
0
0
?
0
?
?
0
0
1
1
1
0
0
0
0
0
1
1
1
0
0
0
?
0
?
?
1
1
0
0
0
0
1
1
1
1
Phylogeny of Tubificinae and Limnodriloidinae • R. Marotta et al.
278
Appendix 3