Hermit Crabs: Are They Really Polyphyletic?

Author(s): Patsy A. McLaughlin

Source: Journal of Crustacean Biology, Vol. 3, No. 4 (Nov., 1983), pp. 608-621
Published by: {brill} on behalf of The Crustacean Society
Stable URL: http://www.jstor.org/stable/1547956
Accessed: 07-06-2016 12:51 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

The Crustacean Society, Brill are collaborating with JSTOR to digitize, preserve and extend access to
Journal of Crustacean Biology

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
JOURNAL OF CRUSTACEAN BIOLOGY, 3(4): 608-621, 1983

HERMIT CRABS-ARE THEY REALLY POLYPHYLETIC?

Patsy A. McLaughlin

ABSTRACT

The independent lineages of the two superfamilies of hermit crabs, the Coenobitoidea and
Paguroidea, were originally hypothesized from evidence gained in the study of larval devel-
opment of a very few species. Subsequent studies have shown that the proposed relationships
among thalassinoids-coenobitoids and paguroids-galatheoids are not as probable as once
believed. The question of hermit crab phylogeny (including the lithodids) is re-examined
using cladistic methodology. As a result of this analysis, the families Pomatochelidae, Di-
ogenidae, Coenobitidae, Paguridae, Parapaguridae, and Lithodidae are reunited under the
superfamily Paguroidea, infraorder Anomala.

Hermit crabs are most familiar as intertidal species that often inhabit brightly
colored gastropod shells in tidepools, or semiterrestrial species such as Coenobita
clypeatus (Herbst), a form used prominently in the pet crab craze of a few years
ago. However, these decapod crustaceans, in reality, are highly diverse in habitat
and form (cf: Russell, 1962). Of the more than 800 species in 86 genera and six
families, only 12 species are semiterrestrial; the remainder are marine. They range
from polar to tropical seas, and from the supratidal to the abyss.
Early classifications of decapod crustaceans consisted of three major subdivi-
sions: the Macrura, for forms with long abdomens; the Anomura, for forms with
variable abdomens; and the Brachyura, for forms with short abdomens. While
most subsequent schemes (see Burkenroad, 1981, for review) have retained hermit
crabs (including Lomis and the lithodids, as well as more conventional hermit
crabs), galatheoids, and hippoids in the Anomura (or Anomala, fide de Saint
Laurent, 1979b; McLaughlin, 1983), the taxonomic interpretation of several other
taxa, particularly the thalassinoids, has been more ambiguous, resulting in either
their inclusion in or exclusion from the Anomura (e.g., Henderson, 1888; Bor-
radaile, 1907; Bouvier, 1940; Balss, 1957; Waterman and Chace, 1960; Burken-
road, 1963, 1981; Glaessner, 1969; de Saint Laurent, 1979b; Bowman and Abele,
1982).
During a study of hermit crab larval development, MacDonald et al. (1957)
described the larvae of 10 British hermit crabs, viz., one diogenid, eight pagurids,
and one lithodid. They also included the descriptions of nine pagurids and 11
diogenids from published accounts. Previous classifications had separated the
pagurids from the lithodids as distinct families within the Paguroidea, whereas
pagurids and diogenids had been separated only at the subfamilial level. MacDonald
et al. found only one character by which pagurid and lithodid larvae could be
distinguished, but six characters that they believed clearly separated pagurids from
diogenids. Furthermore, they noted the apparent similarities found by Hart (1937)
between diogenid and upogebiid-type thalassinid larvae and by Gurney (1942)
between pagurid and galatheid larvae. To these larval characters they added one
adult character, the position of the bases of the third maxillipeds. In pagurids
(which at that time included parapagurids) and lithodids, the bases of the third
maxillipeds are widely separated by a sternal plate. Using sparse data, MacDonald
et al. (1957) reported a similar condition for galatheids. In contrast, coenobitids
and diogenids, like thalassinids, presumably all lack a broad sternal plate, so that
these maxillipeds are approximate basally. On this evidence these authors con-
cluded that there appeared to be a separation between dardanid (=diogenid) and
608

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 609

Classification
of

MacDonald, Pike, & Williamson

Superfamily Coenobitoidea
Family Pylochelidae (=Pomatochelidae)
Family Diogenidae
Family Coenobitidae
Family Lomisidae (= Lomidae)
Superfamily Paguroidea
Family Paguridae (= Paguridae + Parapaguridae)
Family Lithodidae

ZCenobitoidea Paguroidea

other other
Thalassinoidea Galatheoidea

Upogebiinae . Galatheidae

3rd maxillieds maxillipeds

approximate \ widely separated
Ancestral ANOMURA

Fig. 1. Classification of MacDonald et al. (1957) illustrating the proposed independent lineages of
the Coenobitoidea and Paguroidea.

pagurid groups, representing a major division in the evolution of the Anomura,

and that one line of each of these groups had independently adopted the gastropod
shell as a protective device. Convergent adaptation to this particular habitat and
associated way of life within these two lines had resulted in the similarity of form
that had led to their assignment in the same broadly defined family. MacDonald
et al. thus pictured hermit crab phylogeny as two lines (Fig. 1) that should not
only be separated into different families, but into different superfamilies. They
proposed a classification that resulted in the erection of two superfamilies, equiv-
alent in rank to the other major taxa of the Anomura, as follows: within the
Coenobitoidea four families were recognized, the Pylochelidae (=Pomatochelidae,
fide Miyake, 1978), Lomisidae (=Lomidae, fide Glaessner, 1969), Diogenidae,
and Coenobitidae. Their Paguroidea contained the families Paguridae and Lith-
odidae. De Saint Laurent (1972) separated the Parapaguridae from the Paguridae,
and more recently McLaughlin (1983) removed the Lomidae from the Coeno-
bitoidea.
Although this polyphyletic origin for the hermit crabs has continued to be
reflected in most subsequent classifications (e.g., Kaestner, 1970; Miyake, 1978;
McLaughlin, 1980; Bowman and Abele, 1982), its justification, based on larval

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
610 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

characters, has been seriously undermined. Dechance and Forest (1958) pointed
out that neither of the diogenid genera Diogenes and Paguristes, upon which
MacDonald et al. (1957) based their diagnosis of diogenid larval characters, could
be considered typical of the Diogenidae as a whole. In subsequent studies, Pike
and Williamson (1960) found that the larval characters of the diogenids showed
considerable diversity, and that some characters possessed by the diogenids also
were common to some galatheoids. Furthermore, they suggested that characters
commonly shared by paguroids and galatheoids might have developed indepen-
dently in the two groups. However, as they considered the differences between
the larvae of paguroids and thalassinoids to be much greater than those between
coenobitoids and thalassinoids, they continued to propose independent lineages
of the two superfamilies. De Saint Laurent-Dechance (1964), Williamson and von
Levetzow (1967), and Provenzano (1971) all reported that in several characters
parapagurid larvae more closely resembled diogenid, than pagurid larvae. Pro-
venzano (1967) acknowledged that the uniformity of characters among pagurid
larvae was not as constant as previously had been believed, and subsequently
(Provenzano, 1971) described two distinct larval groups within the Paguridae.
Gore (1979) reported heterogeneity in larval characters among genera of the
Galatheidea, and Van Dover et al. (1982) found a distinct and uniform devel-
opment of the first zoeal scaphognathite in the Paguridae, Lithodidae, Diogenidae,
and Coenobitidae. From the studies of Williamson and von Levetzow (1967), it
would appear that the same type of development occurs in the Parapaguridae.
Gore and Scotto (1983) described the larvae of the pagurid Phimochirus holthuisi
(Provenzano) as intermediate between pagurid-type larvae on the one hand and
diogenid-coenobitid-parapagurid larvae on the other. Nothing is known of larval
development in the Pomatochelidae.
If morphological characters of the adults are considered, the case for polyphyly
becomes even less tenable. The relationship of the bases of the third maxillipeds
to the sternal plate is not the same in paguroids and galatheoids, and the position
of the bases of these maxillipeds in the latter superfamily is highly variable. For
example, among species of the rather heterogeneous genus Munidopsis the max-
illipeds may be widely separated or nearly approximate (R. Gore, personal com-
munication). Contrarily, the third maxillipeds of Galathea rostrata A. Milne
Edwards are as approximate basally as those of the diogenid hermit crab Paguristes
depressus Stimpson.
Perhaps more important is the interpretation of the relationship of the Tha-
lassinoidea to the other anomalan superfamilies. De Saint Laurent (1979b), while
acknowledging certain similarities between the larvae of a small group of thal-
assinoids and the anomalans, cited a number of larval and adult characters that
clearly separate the Thalassinoidea from the Anomala. In fact, she suggested that
although the Axiidae branch of the Thalassinoidea clearly is a monophyletic group
(de Saint Laurent, 1979a), the three small families that she grouped under the
Gebiidea may not be (de Saint Laurent, 1979b). Burkenroad (1981) reached a
similar conclusion regarding the separation of the thalassinoids from the Anomala,
although he expressed some doubt about the lack of homogeneity within the
former group.
Inherent in the MacDonald et al. (1957) classification is its phylogenetic im-
plication that all the unique, intrinsic characters shared by the coenobitoids and
paguroids are not homologous, but instead homoplastic. Wiley (1981) defined a
homologous character as one occurring in two or more taxa that is also found in
the common ancestor of those taxa. Likewise, a character (or linear sequence of
character states) is homologous if its state in one taxon is directly or sequentially

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 611

derived from its state in other taxa. There are no absolute criteria for identifying
homologies but four types of evidence are generally recognized: topography, bio-
chemistry, development, and structure (e.g., Remane, 1956; Hecht and Edwards,
1977; Eldredge and Cracraft, 1980; Wiley, 1981). Regrettably, any analysis of
homologous characters in decapod crustaceans is difficult. Topography has limited
value because relative positions of elements are generally constant; comparative
biochemical data are virtually nonexistent; and only limited developmental data
are available. Consequently, determinations of homologous characters, for the
most part, must be made using adult morphological structures. In a pair of ho-
mologous characters, the character that allegedly arose earliest in time is the
plesiomorphic (primitive) character, whereas the character evolving directly from
its pre-existing homologue is the apomorphic (derived) character (Wiley, 1981).
A character state that is interpreted as derived and shared (synapomorphic) among
two or more taxa is an indication of their monophyletic relationship (Hennig,
1966; Hecht and Edwards, 1976, 1977; Bonde, 1977).
A major problem is the determination of the polarities (cf Maslin, 1952) of
pairs of homologous characters. One criterion frequently applied is the "com-
monality principle" (cf. Eldredge, 1979; Crisci and Stuessy, 1980). However, as
pointed out by Watrous and Wheeler (1981), the use of "common = primitive"
can lead to a number of serious errors. Instead, following the recommendations
of these authors, as well as those of Ross (1974), Stevens (1980), and Wiley (198 1),
I have used "out-group comparisons" for determining character polarity. That is,
for a given character with two or more states within a group, the character state
occurring in related groups is assumed to be plesiomorphic; the alternative state
is assumed to be apomorphic.
Before testing the hypothesis of monophyly in hermit crabs, it is necessary to
address the proposition made by MacDonald et al. (1957) that a number of
characters unique to hermit crabs were convergent adaptations that resulted from
the use of gastropod shelters. One study that can provide some evidence is that
of Messing (1979) who compared the morphological adaptations of a shell in-
habiting tanaid (Peracarida), Pagurapseudes largoensis McSweeney, with the di-
ogenid Clibanarius tricolor Herbst and the pagurid Pagurus maclaughlinae Gar-
cia-G6mez. Messing found that coiling or flexing of the abdomen was the most
obvious character exhibited by both hermit crabs and P. largoensis; however,
major morphological differences were seen between the tanaid and the two hermit
crab species. While the abdomens of the hermit crabs are membranous, with
obsolete segmentation and pronounced neuromuscular asymmetry, abdominal
segmentation in the tanaid is retained, calcification decreases gradually posteriad,
and only moderate asymmetry is apparent. Cheliped asymmetry occurs in both
the hermit crabs and P. largoensis; however, in the latter only the mature male
is heterochelous. Only the second pereiopods in P. largoensis are developed as
ambulatory legs, whereas the second and third are ambulatory in the hermit crabs.
The major component of movement in the tanaid is lateromedial, in contrast to
an anteroposterior movement in the hermit crabs. The last five pairs of pereiopods
are reduced in P. largoensis, while only the last two pairs are reduced in hermit
crabs. Although corneous scales or spines occur on segments of these reduced
appendages, they are uniformly restricted to the propodus and dactyl of the ter-
minal pair, and the propodus of the preceding pair in both diogenid and pagurid-
type hermit crabs. In contrast, such structures in P. largoensis occur on the pos-
teroventral surfaces of the meri, carpi, and propodi of the more anterior pairs
and on the carpi of the last pair. The uropods of P. largoensis are filiform as in

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
612 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

other tanaids, whereas both endopods and exopods of virtually all hermit crab
uropods are provided with rasps of corneous scales.
It is clear from Messing's work, that similarities in habit may induce tendencies
toward similar types of adaptation. Whether these same habit similarities can
account for the development of identical adaptations in the two hermit crab
superfamilies is open to question.
As pointed out by Wiley (1981) and Brusca (in manuscript), out-group analyses
require the acceptance of some higher level of universality before statements
regarding relationships of lower taxa can be made. Although there is disagreement
among carcinologists as to whether the traditional subordinal subdivisions of the
Decapoda are monophyletic, there is general agreement that the ordinal taxon is
monophyletic (e.g., Glaessner, 1969; de Saint Laurent, 1979b; Burkenroad, 1981;
Schram, 1981; Bowman and Abele, 1982). De Saint Laurent (1979b) and Bur-
kenroad (1981) also have presented convincing arguments for monophyly of the
Anomala. Therefore, to test the hypothesis of monophyly among hermit crabs,
the closely related Galatheoidea, Hippoidea, and Lomoidea, as well as the prim-
itive reptant Decapoda, can be used for out-group comparisons. It is beyond the
scope of this discussion to examine the phylogenetic relationships among the
major taxa of the Anomala, other than to demonstrate the synapomorphies that
fundamentally unite them and those that may set hermit crabs, as a group, apart.
In the analysis that follows, a series of synapomorphic characters will be ex-
amined (Table 1), and both in-group and out-group comparisons made in order
to develop the hypothesis that polyphyly need not necessarily be invoked to
establish phylogenetic relationships among hermit crabs. The assessment of these
relationships is diagrammed in Fig. 2. The character, position of the bases of the
third maxillipeds, has been excluded from consideration because of the difficulty
in determining its polarity. The extinct glypheoidean palinurans have been con-
sidered central to the origin and evolution of the Decapoda (Glaessner, 1960,
1969). Living palinurid decapods typically have the bases of these appendages
contiguous; however, in the recently discovered "living fossil" glypheoid (cf
Forest and de Saint Laurent, 1975, 1976, 1981; Forest et al., 1976) the bases of
the third maxillipeds are separated. In some of the Galatheoidea, Lomoidea, and
Hippoidea these appendages are widely separated as in the Paguridae, Parapa-
guridae, and Lithodidae, whereas in other galatheoids they are as contiguous as
those of the Pomatochelidae, Diogenidae, and Coenobitidae.
Five synapomorphic characters set the hermit crabs, galatheoids, lomoids, and
hippoids apart from other reptant decapods. These and other synapomorphies
are numbered in the sequence in which they appear in Table 1 and Fig. 2: (1)
protection of the epistome; (2) no diaresis on the uropodal exopod; (3) markedly
reduced 5th pereiopods; (4) mobility of the last thoracic somite; and (5) inner-
vation of the first abdominal somite by a pair of ganglia attached to the posterior
end of the thoracic ganglionic mass.
Five segments in the antennal peduncle typically have been described as the
plesiomorphic state in reptant decapods, with the apomorphic sequence being a
reduction in number through fusion. However, Schmidt (1915) described the
musculature for six segments plus the flagellum in the crayfish Astacus astacus
(Linnaeus), and Forest and de Saint Laurent (1981) have reported an apparently
homologous condition in the glypheoid Neoglyphea inopinata Forest and de Saint
Laurent. McLaughlin (1974, 1983) has reported a "supernumerary" segment in
the antennae of hermit crabs and the lomoid Lomis hirta (Lamarck), between the
typical second and third segments. Snodgrass (1952) remarked that the identity

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 613

Table 1. Survey of states of selected morphological characters of hermit crabs.

Character Plesiomorphic state Apomorphic state

1. Epistome Unprotected Protected

2. Uropodal diaresis Present on exopod Absent from exopod

3. 5th Pereiopods Normally developed Markedly reduced
4. 8th Thoracic somite Immobile Mobile
5. Innervation of 1st 1 st abdominal Ganglia attached to tho-
abdominal somite ganglia racic ganglionic mass
6. Antennal segments 5 plus 5 or fewer

7. Uropods and telson Forming tailfan Not forming tailfan

8. Dactyls of 2nd pereiopods
9. Dactyls of 3rd pereiopods
10. Dactyls of 4th pereiopods
11. Ocular acicles
Sharp, corneous, clawlike
Sharp, corneous, clawlike
Sharp, corneous, clawlike
Absent
Flattened, calcified,
paddle-shaped
Flattened, calcified,
paddle-shaped
a) Flattened, calcified,
paddle-shaped
b) Forming part of chelate
or subchelate structure
Present

12. Uropodal rami Unspecialized Specialized

13. 4th Pereiopods Normally developed Reduced

14. Pleopods (3-5) Paired Unpaired

15. Abdominal calcification Strong, somites distinct Weak, integument often mem-
branous, somites poorly
defined or obliterated

16. Abdominal development Normal (not reduced), Usually reduced and/or

straight twisted

17. Chelipeds Symmetrical Asymmetrical

18. Handedness Symmetrical a) Right-handed
b) Left-handed
19. 1st Abdominal somite Distinct Fused to last thoracic
somite

20. Osmoregulation Marine adapted Terrestrial adapted

21. Antennules Well-developed setose Flagella compressed and
flagella truncate

22. 3rd Maxilliped Flagella well developed Flagella reduced or

absent

23. 2nd Maxilliped Flagella well developed Flagella reduced or

absent

24. 1st Maxilliped Flagella well developed Flagella reduced or

absent

25. Abdominal terga Entire Divided

26. Epistomial spine(s) Absent Present

27. Pleopodal rami 9 Straight Crossed or twisted

28. Uropods Present Absent

29. Supplemental tergal Absent Present

calcification
30. Carcinization Absent Present in varying degrees

of the antennal segments in the hippoid Emerita talpoida (Say) was difficult to
determine. In the representatives of Emerita, Albunea, and Lepidopa examined
in the course of this investigation, the segmentation has proved to be very similar
to that of hermit crabs. Therefore, it may be postulated that for the character,
antennal segmentation (6), the plesiomorphic state is more than five segments or

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
614 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

a>

0a i a E 4

n - ?. -o .2..-

24

\ \ \ \ \\f18b a18a

\ IOa9~~~~ IIO ~ 19

~\ \ "/~~~~ 7 13

112

Fig. 2. Cladogram of the families of hermit crabs and out-groups. Only apomorphies are indicated.
Numbering corresponds to characters in Table 1 and text.

alternatively, with a subdivided third segment which gives the impression of an

additional small segment between the true second and third segments. The gal-
atheoids, then, with five or four entire peduncular segments, exhibit the apo-
morphic state. The remaining anomalan taxa can be united as a sister group by
the character, reduction of the uropods and telson (7), which in the plesiomorphic
state form a tailfan as is seen in the reptant decapod morphotype and the galath-
eoids.
The Lomoidea is a monotypic superfamily (McLaughlin, 1983) represented
only by Lomis hirta. This species possesses no recognized unique apomorphies
of its own, although it shares presumably derived homologous characters with
several unrelated anomalan taxa. For example, its ocular peduncles and antennal
flagella strongly resemble specialized characters of certain hippoids. In the de-
velopment of an epistomial spine, a relationship with either the Parapaguridae
or Pomatochelidae might be proposed; however, no other characters would sub-
stantiate such a relationship. L. hirta has undergone carcinization, as have the
Porcellanidae (Galatheoidea) and the Lithodidae, but this specialization is known
to have arisen independently several times in the Decapoda. The phylogenetic
relationship of the Lomoidea to the other major taxa is unclear.
The Hippoidea are set apart from their sister group, the hermit crab families,
by a number of apomorphies, three of which are pertinent to the present discussion.
In the primitive state, the ambulatory legs of reptant decapods terminate in sharp,
typically corneous, clawlike dactyls. This plesiomorphic state is observed in the
second (8) and third (9) pereiopods of all hermit crabs, whereas the dactyls of

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 615

these appendages in hippoids have become flattened, calcified, and somewhat

paddle-shaped. The dactyls of the fourth pereiopods (10) exhibit apomorphies in
both sister groups. In the hippoids the dactyls again are flattened, calcified, and
somewhat paddle-shaped (1Oa), while in most hermit crabs they form, together
with the propodi, subchelate or chelate structures, the ventro-external surfaces of
which each carry a rasp of corneous scales (lOb).
Several characters are both unique and common among hermit crabs. Of these,
three synapomorphies clearly set the hermit crabs, as a group, apart from both
the primitive reptant decapod morphotype and the other anomalan taxa. The first
is the development of ocular acicles (11). Contrary to the statement of Makarov
(1938, 1962), ocular acicles are present in the Lithodidae, although they are
reduced in size and there is a tendency toward loss in some genera. A similar
condition is observed in the Parapaguridae, where the acicles are well developed
in Parapagurus, but reduced or lost in the other three genera (de Saint Laurent,
1972). The structures identified by McLaughlin (1980, fig. 49) for a typical al-
buneid are not acicles, but rather, calcified portions of the hippoid ocular plate.
The second shared character is the development of specialized endopodal and
exopodal rasps on the uropods (12). An exception is seen in the Lithodidae, where
the uropods have been completely lost. A similar condition occurs in the semi-
terrestrial coenobitid Birgus latro (Linnaeus) where a substantial reduction in the
size of the uropods occurs when the animal reaches adulthood (Wolff, 1961). The
third synapomorphy involves a reduction in the size of the fourth pereiopods (13),
resulting in an appendage not much longer than the fifth and never used in
locomotion. As previously indicated, the fourth pereiopods are also specialized
by being provided with a rasp of corneous scales. An exception again is seen in
the Lithodidae, where the fourth pereiopods are well developed, unspecialized,
and ambulatory as in the out-groups. It is postulated that this is a result of the
carcinization undergone by members of this family, which has allowed them a
much greater mobility. This condition is also approached in B. latro, which during
early terrestrial stages uses small shells of land gastropods and has typically reduced
and modified fourth pereiopods. With growth to maturity, Birgus abandons its
shell shelter, undergoes a moderate degree of carcinization, and takes up a free-
living existence. With these changes the fourth pereiopods grow and take on an
ambulatory function (Harms, 1932, 1938; Wolff, 1961).
The Pomatochelidae, while not distinguished by an apomorphy of their own,
are set apart from their sister group, the remaining hermit crab families, by
retaining the primitive state in several characters. These include: (14) paired
pleopods, as opposed to the loss of pleopods (except gonopods), at least on one
side of the abdomen as seen in all other hermit crabs; (15) retention of a moderately
strongly calcified abdomen with clearly defined somites versus the synapomorphic
sequence of calcium loss and varying degrees of obliteration of external segmen-
tation; (16) well-developed straight abdomen, as opposed to a tendency toward
a reduced and/or twisted abdomen; and (17) cheliped symmetry versus a tendency
toward marked asymmetry.
The Paguridae-Lithodidae-Parapaguridae are separated from their sister group
Diogenidae-Coenobitidae by two synapomorphies. The first is "handedness" (18).
As previously indicated, the plesiomorphic state is cheliped symmetry. Both sister
groups exhibit a tendency toward asymmetry, but in the Paguridae-Lithodidae-
Parapaguridae group the tendency is toward "right-handedness" (18a), while in
the Diogenidae-Coenobitidae group it is toward "left-handedness" (18b). The
second is the fusion of the first abdominal somite with the last thoracic somite
in the Paguridae-Lithodidae-Parapaguridae (19). In the Diogenidae-Coenobitidae

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
616 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

these two somites are distinct, as they are in the Pomatochelidae and the out-
groups.
The characters that separate the Coenobitidae from the Diogenidae appear to
be primarily synapomorphic adaptations to a semiterrestrial life as seen in the
two genera of the former family. These include: morphological and physiological
changes in osmoregulatory mechanisms (20); the development of compressed and
truncate antennular flagella (21); reduction in the flagella of the third (22) and
second (23) maxillipeds; and the virtual loss of the flagellum of the first maxilliped
(24).
The Paguridae-Lithodidae are separated from their sister group Parapaguridae
by the division of the abdominal terga into distinct pairs of plates (25). In the
Parapaguridae the primitive entirety of the terga is retained. The Parapaguridae
also possess apomorphies that set them apart. These include the character, loss
of the flagellum of the first maxilliped (24), a character also exhibited by the
Coenobitidae. De Saint Laurent (1972) suggested that this character is associated
with specialized habitats such as terrestrial or deep-sea. She pointed out that the
galatheoid genus Munidopsis, a deep-sea dwelling taxon like the Parapaguridae,
also has undergone a loss of the flagellum of the first maxilliped. This loss must
be considered an example of convergence rather than a suggestion of a relationship
among the three taxa. A second apomorphic character that distinguishes the
Parapaguridae from the Paguridae-Lithodidae is the presence in the former taxon
of labral and/or interantennular spines on the epistome (26). Again convergence
is presumed, as a similar spine has been reported by de Saint Laurent (1972) for
a species of the Pomatochelidae, and by McLaughlin (1983) for the Lomoidea.
A synapomorphy apparently unique to the Parapaguridae is the tendency of the
female pleopods to become crossed or twisted (27).
The Lithodidae are separated from their sister group the Paguridae by the
following synapomorphies: (28) the complete loss of the uropods; (29) the de-
velopment of supplemental calcified plates laterally adjoining the paired plates of
the abdominal terga; and (30) carcinization, which has arisen independently in
other groups as well, but is virtually complete in the Lithodidae. No single apo-
morphy defines the Paguridae as a whole, although several groups within the
family exhibit unique characters. It is possible that, as suggested by Gore and
Scotto (1983), the Paguridae may be polyphyletic. Only when intrafamilial rela-
tionships are better known will this be determinable.
From the foregoing discussion the evidence for monophyly among hermit crabs
appears convincing. If the hypothesis of monophyly is accepted, the separation
of the families of hermit crabs into two distinct superfamilies is unjustified. There-
fore, it is proposed that the use of the superfamilial name Coenobitoidea be
suppressed and the six families of hermit crabs be united under the superfamily
Paguroidea as listed below. Data are not yet available to permit adequate eval-
uations of inter- and intrasuperfamilial relationships; thus strict cladistic inter-
pretations of the cladogram (Fig. 2) are not warranted at this time.

Infraorder Anomala
Superfamily Galatheoidea
Superfamily Lomoidea
Superfamily Hippoidea
Superfamily Paguroidea
Family
Pomatochelidae
Diogenidae

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 617

Coenobitidae
Paguridae
Parapaguridae
Lithodidae

EVOLUTIONARY PATHWAYS

With our presently limited knowledge, one can only speculate as to the pathways
taken by this unusual group of decapod crustaceans. Early carcinologists tended
to consider the pomatochelids a link between pagurids and the macruran-like
thalassinoids, with the primitive condition in true hermits being one of complete
asymmetry (e.g., Milne Edwards and Bouvier, 1891; Alcock, 1905). It was thought
that the hermit crabs' predisposition to gastropod shells, most of which are dex-
trally coiled, resulted in body asymmetry, decalcification, and loss of pleopodal
appendages on the right side of the abdomen, together with the abdominal twist
and specializations of the uropods and fourth and fifth pereiopods (e.g., Perez,
1934; Makarov, 1938, 1962; Chapple and Hearney, 1974; Neville, 1977). That
all four pomatochelid genera are symmetrical (i.e., have paired pleopods on all
five abdominal somites, straight, moderately well-calcified abdomens, symmet-
rical, well-developed tergal plates, and equal chelipeds) and do not live in gas-
tropod shells might be cited as evidence of this proposition. However, pomato-
chelids have the same types of uropodal and pereiopodal specializations as do
their gastropod shell-dwelling relatives. Boas (1926a) showed that the pomato-
chelids really were primitive, symmetrical hermit crabs; however, he and most
subsequent carcinologists continued to consider symmetry in other genera such
as Cancellus, Paguropsis, and Xylopagurus a secondarily acquired adaptation to
a noncochlean habitat (e.g., Boas, 1926a, b; Rabaud, 1941; Russell, 1962). To
these and other carcinologists, symmetry is indicated by the presence of equal
chelipeds, paired pleopods, and symmetrical uropods (e.g., Wolff, 1961; Proven-
zano, 1963). Pilgrim (1973) has shown that there is no correlation in hermit crabs
between cheliped size and the sign of asymmetry. Similarly, in other decapods,
such as Uca in the Ocypodidae, asymmetry of the chelipeds is not concomitant
with body asymmetry. All larval hermit crabs studied to date have complete body
symmetry through the zoeal stages. The fixation of cheliped asymmetry is variable.
For example, although most coenobitid glaucoth6es have symmetrical chelipeds,
those of Diogenes have a much larger left than right, while those of Petrochirus
have a slightly larger right (Provenzano, 1968). Nor is the absence of paired
pleopods necessarily indicative of body asymmetry. Nearly all glaucothoes have
paired pleopods on the abdominal somites. As previously indicated, the transi-
tional series from plesiomorphic to apomorphic states appears to be one of loss.
In the diogenid genus Paguropsis and in females of Cancellus, unpaired pleopods
may occur on either the right or the left side of the abdomen; pleopods are
completely lost in males of Cancellus. In all other genera, although unpaired
pleopods are restricted to the left side, there is a tendency toward either reduction
in size, or number, or complete loss of these appendages, particularly in males
where they serve no egg-bearing function. Nevertheless, several of these same
genera have retained a straight, although often uncalcified, abdomen and have
symmetrical uropods (e.g., McLaughlin, 1982). As in the pomatochelids, these
pagurids typically do not occupy gastropod shells, but instead protect their ab-
domens in crevices in coral and rock, worm tubes, Dentalium shells, wood, bam-
boo, etc. Moreover, they too are provided with the specialized uropodal and

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
618 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

pereiopodal rasps. While it is conceivable that secondary symmetry has evolved

repeatedly among hermit crabs, a more parsimonious hypothesis would be one
in which a retention, at least in part, of basic symmetry has occurred in some of
the otherwise more advanced genera.
Any hypothesis of evolutionary pathways in the paguroids necessarily must be
based on extant species because the fossil record consists of only a few chelipeds.
I therefore suggest that the pathway of paguroid evolution has been first through
the practice of concealment and subsequent protection of the abdomen by some
external means, followed by the modification and specialization of the uropods
and fourth and fifth pereiopods to maintain such protection. Following these
adaptations, decalcification of the abdominal integument and posterior part of
the cephalothorax began. Because of the space constraints provided by the external
shelters, it would be reasonable to speculate that these paguroids found the re-
striction of egg carrying to one side of the abdomen advantageous. The mobility
afforded by the use of a gastropod shell as a shelter presumably has led to its
adoption by so many hermit crabs.

ACKNOWLEDGEMENTS

The author wishes to acknowledge gratefully the very valuable assistance provided by R. C. Brusca
(Allan Hancock Foundation), and the critical reviews and helpful comments made by L. G. Abele
(Florida State University), D. L. Felder (University of Southwestern Louisiana), R. H. Gore (Academy
of Natural Sciences, Philadelphia), J. Haig (Allan Hancock Foundation), and several anonymous
reviewers. Many of their suggestions have been incorporated.

LITERATURE CITED

Alcock, A. 1905. Anomura. Fasciculus 1. Pagurides.-Catalogue of the Indian decapod Crustacea

in the collection of the Indian Museum 2: 1-197. Indian Museum, Calcutta.
Balss, H. 1957. Decapoda: Systematik.-In: H. G. Bronn, Klassen und Ordnungen des Tierreichs,
ed. 2, 5: 321-480.
Boas, J. E. V. 1926a. Ziir Kenntnis symmetrischer Paguriden.-Biologiske Meddelelser 5: 1-52.
1 926b. Zur Kenntnis des Einsiedlerkrebses Paguropsis und seiner eigenartigen Behausung. -
Biologiske Meddelelser 5: 1-23.
Bonde, N. 1977. Cladistic classification as applied to vertebrates.-In: M. K. Hecht, P. C. Goody,
and B. M. Hecht, eds., Patterns in vertebrate evolution. Pp. 741-804. Plenum Press, New York.
Borradaile, L. A. 1907. On the classification of the decapod crustaceans. -Annals and Magazine of
Natural History 19: 457-486.
Bouvier, E. L. 1940. D6capodes marcheurs.-In: Faune de France 37: 1-399. Paul Lechevalier,
Paris.
Bowman, T. E., and L. G. Abele. 1982. Classification of the Recent Crustacea.-In: L. G. Abele,
ed., The biology of Crustacea. 1: 1-27. Academic Press, New York.
Burkenroad, M. D. 1963. The evolution of the Eucarida (Crustacea, Eumalacostraca) in relation to
the fossil record.-Tulane Studies in Geology 2: 3-16.
1981. The higher taxonomy and evolution of Decapoda (Crustacea). -Transactions of the
San Diego Society of Natural History 19: 251-268.
Chapple, W. D., and E. S. Hearney. 1974. The morphology of the fourth ganglion of the hermit
crab: A light microscope study.-Journal of Morphology 144: 407-420.
Crisci, J. V., and T. G. Stuessy. 1980. Determining primitive character states for phylogenetic
reconstruction. - Systematic Botany 5: 112-135.
Dechance, M., and J. Forest. 1958. Les glaucothoes de Catapaguroides timidus (Roux) et de Cli-
banarius erythropus (Latreille). Remarques sur le stade post-larvaire des Pagurides. - Bulletin de
la Societe Zoologique de France 83: 274-293.
Eldredge, N. 1979. Cladism and common sense. -In: J. Cracraft and N. Eldredge, eds., Phylogenetic
analysis and paleontology. Pp. 165-198. Columbia University Press, New York.
, and J. Cracraft. 1980. Phylogenetic patterns and the evolutionary process.- In: Methods
and theory in comparative biology. Columbia University Press, New York. Pp. 1-349.

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 619

Forest, J., and M. de Saint Laurent. 1975. Presence dans la faune actuelle d'un repr6sentant du
groupe m6sozoique des Glyph6des: Neoglyphea inopinata gen. nov., sp. nov. (Crustacea Decapoda
Glypheidae). -Comptes Rendus Hebdomadaires des Seances de l'Acad6mie des Sciences, Paris
281: 155-158.
,and . 1976. Capture aux Philippines de nouveaux exemplaires de Neoglyphea inopinata
(Crustacea Decapoda Glypheidae). - Comptes Rendus Hebdomadaires des Seances de l'Acad6mie
des Sciences, Paris 283: 935-938.
, and . 1981. La morphologie externe de Neoglyphea inopinata, espece actuelle de
Crustac6 Decapode Glyph6ide.-R6sultats de Campagnes MUSORSTOM. I.-Philippines (18-
28 mars 1976). Memoires ORSTOM 91: 51-84.
, and F. A. Chace, Jr. 1976. Neoglyphea inopinata: A crustacean "living fossil" from
the Philippines.-Science 192: 884.
Glaessner, M. F. 1960. The fossil decapod Crustacea of New Zealand and the evolution of the order
Decapoda.-New Zealand Geological Survey, Paleontological Bulletin 31: 1-63.
1969. Decapoda.-In: R. C. Moore, ed., Treatise on invertebrate paleontology, Part R
Arthropoda 4, 2: R400-R651. University of Kansas and the Geological Society of America,
Lawrence, Kansas.
Gore, R. H. 1979. Larval development of Galathea rostrata under laboratory conditions, with a
discussion of larval development in the Galatheidae (Crustacea Anomura).-Fisheries Bulletin,
United States 76: 781-806.
, and L. E. Scotto. 1983. Studies on decapod Crustacea from the Indian River region of Florida
XXVII. Phimochirus holthuisi (Provenzano, 1961) (Anomura: Paguridae): The complete larval
development under laboratory conditions, and the systematic relationships of its larvae.-Journal
of Crustacean Biology 3: 93-116.
Gurney, R. 1942. Larvae of decapod Crustacea.-Ray Society, London. Pp. 1-306.
Harms, J. W. 1932. Die Realisation von Genen und die konsekutive Adaptation. 2. Birgus latro L.
als Landkrebs und seine Beziehungen zu den Coenobiten.-Zoologische Wissenschaften fur Zool-
ogie 140: 167-290.
1938. Lebensablaufund Stammesgeschichte des Birgus latro L. von der Weihnachtsinsel.-
Jenaische Zeitschrift fuir Naturwissenschaft 71: 1-34.
Hart, J. F. L. 1937. Larval and adult stages of British Columbia Anomura.-Canadian Journal of
Research 15: 179-220.
Hecht, M. K., and J. L. Edwards. 1976. The determination of parallel or monophyletic relationships:
The proteid salamanders-a test case. -American Naturalist 110: 653-677.
,and . 1977. The methodology of phylogenetic inference above the species level.-In:
M. K. Hecht, P. C. Goody, and B. M. Hecht, eds., Major patterns in vertebrate evolution. Pp.
3-51. Plenum Press, New York.
Henderson, J. R. 1888. Report on the Anomura collected by H. M. S. Challenger during the years
1873-1876.-Report on the Scientific Results of the Voyage of the H.M.S. Challenger, Zoology
27: 1-221.
Hennig, W. 1966. Phylogenetic systematics.-University of Illinois Press, Urbana, Illinois. Pp. 1-
263.
Kaestner, A. 1970. Invertebrate Zoology, 3, Crustacea.-Wiley Interscience, New York. Pp. 1-523.
MacDonald, J. D., R. B. Pike, and D. I. Williamson. 1957. Larvae of the British species of Diogenes,
Pagurus, Anapagurus and Lithodes (Crustacea, Decapoda). -Proceedings of the Zoological Society
of London 128: 209-257.
McLaughlin, P. A. 1974. The hermit crabs (Crustacea Decapoda, Paguridea) of northwestern North
America. -Zoologische Verhandelingen 130: 1-396.
1980. Comparative morphology of Recent Crustacea.-W. H. Freeman and Company, San
Francisco. Pp. 1-177.
1982. Revision of Pylopagurus and Tomopagurus (Crustacea: Decapoda: Paguridae) with
the descriptions of new genera and species: Part III, Agaricochirus McLaughlin, Enallopagurus
McLaughlin, and Enallopaguropsis McLaughlin.-Bulletin of Marine Science 32: 823-855.
1983. A review of the phylogenetic position of the Lomidae (Crustacea: Decapoda: An-
omala).-Journal of Crustacean Biology 3: 431-437.
Makarov, V.V. 1938. Rakoobraznyey. Anomura. Crustac6s Decapodes Anomures.--In: A. A. Sha-
takel'berg, ed., Fauna SSSR (n. ser.) 16: 1-324.
1962. English translation of Makarov, 1938. Crustacea, Anomura.-Israel Program for
Scientific Translation, Jerusalem. Pp. 1-283.
Maslin, P. P. 1952. Morphological criteria of phylogenetic relationships.-Systematic Zoology 1:
49-70.

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
620 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 3, NO. 4, 1983

Messing, C. G. 1979. Aspects of the biology ofPagurapseudes (Crustacea: Tanaidacea) in southeastern

Florida.-Ph.D. dissertation, University of Miami. Pp. 1-241.
Milne Edwards, A., and E. L. Bouvier. 1891. Sur les modifications que subissent les Pagures suivant
l'enroulement de la coquille qu'ils habitent. -Bulletin de la Societe Philomatique de Paris 1890-
1891: 151-153.
Miyake, S. 1978. The crustacean Anomura ofSagami Bay.-Hoikusha Publishing Company, Tokyo.
Pp. 1-200 + 161 pp. in Japanese.
Neville, A. C. 1977. Symmetry and asymmetry problems in animals. -In: R. Duncan and M. Weston-
Smith, eds., The encyclopaedia of ignorance. Pp. 331-345. Pergamon Press, Oxford.
Perez, C. 1934. Exposes de biologie zoologique. I. Les pagures ou Bernards l'ermite (un example
d'adaptation).-Actualites Scientifique et Industrielle, Paris 101: 1-33.
Pike, R. B., and D. I. Williamson. 1960. Larvae of decapod Crustacea of the families Diogenidae
and Paguridae from the Bay of Naples.-Pubblicazioni della Stazione Zoologica di Napoli 31:
493-552.
Pilgrim, R. L. C. 1973. Axial skeleton and musculature in the thorax of the hermit crab, Pagurus
bernhardus [Anomura: Paguridae].-Journal of the Marine Biological Association of the United
Kingdom 53: 363-396.
Provenzano, A. J., Jr. 1963. The glaucoth6e state of Dardanus venosus (H. Milne Edwards) (De-
capoda: Anomura).-Bulletin of Marine Science of the Gulf and Caribbean 13: 11-22.
1967. The zoeal stages and glaucothoe of the tropical eastern Pacific hermit crab Trizopagurus
magnificus (Bouvier, 1898) (Decapoda; Diogenidae), reared in the laboratory.-Pacific Science
21: 457-473.
1968. The complete larval development of the West Indian hermit crab Petrochirus diogenes
(L.) (Decapoda, Diogenidae) reared in the laboratory.-Bulletin of Marine Science 18: 143-181.
1971. Zoeal development of Pylopaguropsis atlantica Wass, 1963, and evidence from larval
characters of some generic relationships within the Paguridae. -Bulletin of Marine Science 21:
237-255.
Rabaud, E. 1941. Recherches sur l'adaptation et le comportement des Pagures.- Archives de Zoologie
Exp6rimentale et G6enrale 82: 181-285.
Remane, A. 1956. Die Grundlagen des naturalischen Systems der vergleichenden Anatomie und
Phylogenetik.-K. G. Geest und Portig, Leipzig. Pp. 1-364.
Ross, H. H. 1974. Biological systematics.-Addison-Wesley Publishing Company, Reading, Mas-
sachusetts. Pp. 1-345.
Russell, E. S. 1962. The diversity of animals. An evolutionary study.-Acta Biotheoretica
13(supplement 1): 1-151.
Saint Laurent, M. de 1972. Sur la famille des Parapaguridae Smith, 1882. Description de Typhlo-
pagurusforesti gen. nov., sp. nov., et de quinze esp6ces ou sous-especes nouvelles de Parapagurus
Smith (Crustacea, Decapoda).-Bijdragen tot de Dierkunde 42: 97-123.
1979a. Sur la classification et la phylog6nie des Thalassinides: definitions de la superfamille
des Axioidea, de la sous-famille des Thomassiniinae, et de deux genres nouveaux (Crustacea
Decapoda).-Comptes Rendus Hebdomadaires des Seances de l'Academie des Sciences, Paris
(D) 288: 1395-1397.
1979b. Vers une nouvelle classification des Crustac6s Decapodes Reptantia.-Bulletin de
l'Office National des Peches de Tunisie 3: 15-31.
Saint Laurent-Dechanc6, M. de 1964. D6veloppement et position systematique du genre Parapagurus
Smith (Crustacea Decapoda Paguridae). I. Description des stades larvaires. -Bulletin de 1'Institut
Ocanographique (Monaco) 64: 1-26.
Schmidt, W. 1915. Die Muskulatur von Astacusfluviatilis (Potamobius astacus L.). Ein Beitrag zur
Morphologie der Decapoden. -Zeitschrift fir Wissenschaftliche Zoologie 112: 166-251.
Schram, F. R. 1981. On the classification of Eumalacostraca. -Journal of Crustacean Biology 1: 1-
10.
Snodgrass, R. E. 1952. The sand crab Emerita talpoida (Say) and some of its relatives. -Smithsonian
Miscellaneous Collections 117: 1-34.
Stevens, P. F. 1980. Evolutionary polarity of character states.-Annual Review of Ecology and
Systematics 11: 333-358.
Van Dover, C. L., F. R. Factor, and R. H. Gore. 1982. Developmental patterns of larval scatho-
gnathites: an aid to the classification of anomuran and brachyuran Crustacea.-Journal of Crus-
tacean Biology 2: 48-53.
Waterman, T. H., and F. A. Chace, Jr. 1960. General crustacean biology.-In. T. H. Waterman,
ed., The physiology of Crustacea. 1: 1-33. Academic Press, New York.
Watrous, L. E., and Q. D. Wheeler. 1981. The out-group comparison of character analysis. -Sys-
tematic Zoology 30: 1-11.

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms
 McLAUGHLIN: HERMIT CRAB PHYLOGENY 621

Wiley, E. O. 1981. Phylogenetics. The theory and practice of phylogenetic systematics. -John Wiley
and Sons, New York. Pp. 1-439.
Williamson, D. I., and K. G. von Levetzow. 1967. Larvae of Parapagurus diogenes (Whitelegge)
and some related species (Decapoda, Anomura).-Crustaceana 12: 179-192.
Wolff, T. 1961. Description of a remarkable deep-sea hermit crab with notes on the evolution of
the Paguridea.--Galathea Report 4: 11-31.

RECEIVED: 18 January 1983.

ACCEPTED: 28 February 1983.

Address: Department of Biological Sciences, Florida International University, Miami, Florida 33199.

Announcement

"How to Know the Freshwater Crustacea," by Joseph F. Fitzpatrick, Jr.,

University of South Alabama, has been published by Wm. C. Brown Com-
pany Publishers as a volume in "The Pictured Key Nature Series." The
single copy price is $10.95. For further information, contact the publisher
at 2460 Kerper Boulevard, Dubuque, Iowa 52001.

This content downloaded from 159.178.22.27 on Tue, 07 Jun 2016 12:51:27 UTC
All use subject to http://about.jstor.org/terms