19 Are the Crustaceans Monophyletic?

Back

Frederick R. Schram

Stefan Koenemann

319

Grasses of the Distant Past

In these times of rampant Tree of Life cultivation, it is perhaps

hard to realize that up until 20 years ago the “phylogenetic

lawn” of figure 19.1 entailed everything we as

carcinologists knew about crustacean relationships. This

poorly resolved “twig” was, and to some people still is, a

widely excepted consensus that dates back to the 1962 conference

on the Phylogeny and Evolution of Crustacea held at

the Museum of Comparative Zoology at Harvard (Dahl

1963). Moreover, this was considered state of the art for the

field during that period!

The situation within the major groups of crustaceans was

little better. Like the deer frozen in the middle of a roadway

by oncoming headlights on its way to becoming a venison

pancake, crustacean workers were overawed by the multiplicity

of body plans all too evident in their animals. Somehow,

they thought, one could not make sense of such

diversity. Other arthropod groups exhibited greater uniformity

of body plans. An insect is an insect: a head, three sets

of legs, two sets of wings, and gonopores at the end of the

abdomen. Each one of the millions of insect species conforms

in principle to this basic body plan. So, too, with other arthropod

groups: an arachnid is an arachnid, a sea spider is a

sea spider. Even myriapods, although they vary in length,

conform to similar body plans. However, such uniformity of

body plan is not true for crustaceans. Some crustaceans have

long bodies, whereas others are short. Some have gonopores

in the front of the body, most someplace in the middle (but

by no means in the same position from group to group), and

one or two even have the gonopores at the posterior end of

the body. Some bear body shields, or carapaces; others do

not. Some possess legs on every segment, but many do not.

“Entomologists may have more species,” trumpeted carcinologists,

“but we have greater diversity of body plans.” However,

there is little reflection on just what this greater diversity

really meant.

Attempts to Get “True” Trees

Morphology

Beginning in mid-1980s, long after cladistic methods of

analysis had been taken up in other arthropod groups, more

rigorous approaches to crustacean tree building began to

emerge. Sieg (1983) published the first true cladograms for

any crustacean group, namely, the Tanaidacea, albeit using

the paper-and-pencil method of Hennig. The next year,

Schram (1984) published the first numerical cladistic analysis

of eumalacostracan crustaceans, and a few years later (Schram

1986) employed computer based cladistic analyses for the

crustaceans as a whole in his book-length overview of the

subphylum.

Furthermore, based on the results of these analyses,

Schram (1986) produced a higher taxonomy of Crustacea

320 The Relationships of Animals: Ecdysozoans

that grew out of the cladistic analyses. Essentially, four classes

were recognized and defined on the base of apomorphic features:

Remipedia, Malacostraca, Maxillopoda, and Phyllopoda

(fig 19.2A). Several authorities criticized this arrangement,

viewing the disappearance or relocation of favorite higher

taxa with alarm. None of the 1986 trees were arrived at a

priori; that is, they emerged as the simplest patterns of relationships

derived from character matrixes. But some workers

took exception to them purely on the basis that the

patterns of relationships did not conform with what people

had thought before that time. These critics preferred to argue

from an evolutionary systematic viewpoint, using a few

characters to a priori judge the affinities within the crustaceans

and arrange the higher taxa accordingly. Nevertheless,

Schram (1986) had a positive effect on the field of carcinology

because in ensuing years cladistic analyses of smaller groups

within the crustaceans began to appear.

By 1990, a special conference held at Kristineberg, Sweden,

was deemed necessary to address the origin of crustaceans

and their evolution, within which an attempt was made

at generating a new computer-based cladistic analysis of

morphological features (Wilson 1992). One is tempted to

assume that because the matrix upon which this analysis was

based was the product of the deliberations of all the participants

at the conference, it would contain few errors. This

proved not to be the case (Schram 1993) and had serious

consequences for the trees derived from that analysis.

To address this last problem, Schram and Hof (1997)

undertook a more comprehensive analysis of fossil and recent

crustaceans (fig 19.2B). This paralleled an independently

conceived and carried out attempt directed at the same time

by Wills (1997). Wills achieved a “cleaner” result in that he

obtained a tree in which the four classes of Schram were

more-or-less clearly evident (fig 19.3), albeit in a slightly

different branching sequence, but his study was essentially

an ingroup analysis rooting the tree to the long-bodied

remipedes. Schram and Hof (1997) took a broader approach

using insects and myriapods as outgroups (fig 19.2B). In

addition, they tested various alternative partitions of the

database; for instance, they wanted to determine what happened

with and without fossils included (and with fossils

alone), or with and without soft anatomy (so often impossible

to assess for fossil forms). The shifting patterns of relationship

they observed (for details, see Schram and Hof 1997)

indicated great instability to the underlying data. Interestingly,

the trees of Schram and Hof (1997) did not confirm

the results uncovered by Wilson (1992), nor did it convincingly

support the four classes derived by Schram (1986).

Molecules

Meanwhile, it was inevitable that the analysis of molecule

sequences would enter into the picture. The sequencing of

various molecules to elucidate the phylogeny of many crustacean

groups proceeded apace through the 1990s. In any

analysis of this sort, it is critical to perform comprehensive

analyses with a wide array of crustaceans and molecules.

Nevertheless, these type studies focus instead on model types

or selected species and use only one or two molecules. Perhaps

then it is not too surprising that just about every con-

Figure 19.1. Phylogenetic “lawn” of crustacean relationships. This view, with unconnected lines,

prevailed for more than 20 years from the early 1960s to the 1980s. Diverse body plans could not

be reconciled, so carcinologists made little effort toward “growing” a tree. Modified from Dahl

(1963).

Are the Crustaceans Monophyletic? 321

ceivable result one could hope for was in fact obtained by

these methods (fig 19.4).

Unsurprisingly, one of the first molecules sequenced was

18S ribosomal DNA (rDNA), and studies using this molecule

remain the most comprehensive to date. They culminated in

the results (fig 19.4A) presented in Spears and Abele (1997).

These 18S rDNA data yielded a pattern of a polyphyletic Crustacea,

that is, crustaceans interspersed among other groups of

arthropods. Because Spears and Abele made a conscious attempt

to sample a wide array of taxa and several species from

each major group, these results have to be considered seriously.

Another laboratory undertook a separate but not quite

so taxonomically comprehensive an analysis (fig 19.4B) employing

elongation factor-1a (EF-1a) and later augmented

this data with RNA polymerase II (Pol II) (Regier and Shultz

1998, Shultz and Rieger 2000). The EF-1a (both alone and

with Pol II) yielded trees with at least paraphyletic crustaceans

vis-б-vis an insect/hexapod clade, but these analyses suffered

from a limited array of taxa sampled.

A final set of analyses tried to address the issue of comprehensiveness

of the character set (Edgecombe et al. 2000,

Giribet et al. 2001). As a result, these studies used total evidence

approaches, combining morphology together with

molecular data, but still examined only a limited number of

taxa. The results (fig 19.4C) stand in contrast to the above

studies in that virtual monophyly of the limited number of

crustaceans (as well as other arthropod groups) emerged.

All of these molecular data sets exhibited the phenomenon

of long-branch attraction, especially prominent in

Spears and Abele (1997) and Giribet et al. (2001), with the

terminal taxa of long branches tending to emerge in single

clades. We can summarize the results derived from these

molecular studies as follow: (1) The phylogenies do not agree

with each other. (2) The results are quite at odds with the

analysis based on morphology alone. (3) The varying patterns

of relationship within their own data depending on

different methods of analysis. Like the differences noted

between morphological analyses, the inconsistent results

Figure 19.2. (A) A phylogenetic tree for

Crustacea derived from a computer-analyzed

morphological matrix showing the hypothesized

relationships amongst four recognized

classes (modified from Schram 1986). (B) A

more comprehensive analyses of relationships

of fossil and recent crustaceans (simplified from

Schram and Hof 1998). This analysis revealed

that the Schram (1986) clade of branchiopods

and cephalocarids (and allies) is probably

polyphyletic (some of those taxa occur near the

base, whereas others are high in the tree), and

that the maxillopodans are paraphyletic (two

separate clades side by side).

322 The Relationships of Animals: Ecdysozoans

obtained from molecule sequencing are vexing. Again, this

may indicate some underlying problem with the assumptions

concerning either the nature of, or membership in, what we

have come to call Crustacea—reflecting maybe the great disparity

of body plans.

Challenge of the Cambrian

Naturally, when one discusses phylogeny, one cannot avoid

deliberating on fossils; the crustaceans are no exception. We

could categorize crustacean fossils into two types. On the one

hand, are taxa that extend throughout the fossil record and

essentially amplify aspects of the deep history of extant forms.

However, this record is uneven. For example, the background

history concerning “lobsters” is relatively well known,

whereas the situation within a group such as the copepods

is opaque (for details, see Schram 1986). On the other hand,

some fossils do not easily fit into living groups. These are

largely Paleozoic, especially Cambrian, in age, and consideration

of these species in a phylogenetic context presents real

challenges. Yet these taxa are crucial to understanding the

history of crustaceans and crustacean-like creatures because

these fossils come from a time when the basic body plans took

origin. We can delineate two groups of these fossils: the shortbodied,

micro-arthropods of the Cambrian Orsten and the

long-bodied arthropods from faunas such as the Burgess

Shale and Chengjiang (both described below). All these fos-

Figure 19.3. Comprehensive

analyses of relationships of fossil

and recent crustaceans (simplified

from Wills 1997). Note that

Wills uncovered a pattern of

relationships similar to that

shown in figure 19.2A, except

that the clade of branchiopods

and cephalocarids (and allies) is

shifted toward the base of the

tree and is paraphyletic. Orsten

and Burgess fossils appear in

different places in this tree.

Are the Crustaceans Monophyletic? 323

sils fortunately are characterized by high quality of preservation,

such that the monographs describing these animals

often resemble in their detail those for living forms.

The Orsten

Few fossil discoveries in the last century have so consistently

yielded amazing insights into the minutest aspects of Cambrian

life as have the micro-arthropods of the Orsten, or

“stink stones.” The first, and still most productive, sites

come from the bituminous limestones of Sweden, but similar

localities occur around the world. The literature is

voluminous, prolific, and detailed. The often unusual preservation

of these fossils (fig 19.5) allows insights into not

only the fine details of features such as appendage setation

(see Mьller and WaloЯek 1986b, WaloЯek and Mьller

1998, WaloЯek 1999), but also knowledge of larval forms

(Mьller and WaloЯek 1986a; WaloЯek and Mьller 1989).

In many instances these larvae can be related to adult and

subadult forms into coherent life cycles. Two basic forms

of larvae are noted by WaloЯek and Mьller (1997): those

that have four sets of limbs [antennae + three additional

limb pairs], and those that have three sets of limbs [antennae

+ two additional locomotory/feeding limb pairs = nauplius].

The insights these fossils offer into the unfolding

process of “cephalization”; that is, the formation of a distinct

head region is new.

Furthermore, another critical feature of the Orsten fossils,

which we believe has not been fully absorbed by most

authorities, is the generally short-bodied nature of their structural

plans. Orsten fossils not only are characterized by small

body size, in the range of a few millimeters but also exhibit a

relatively small number of body segments. In this, the Orsten

animals differ from the long-bodied living remipedes and also

from many of the elongate Cambrian species from the Burgess

Shale and Chengjiang faunas (fig 19.6).

Consideration of the Orsten fossils has provided us with

many hypotheses about the evolution of arthropods, and

especially crustacean body plans. Some of these, for example,

the differences between an “abdomen” and a “pleon”

(WaloЯek 1999), have received unexpected confirmation

from developmental genetics (Schram and Koenemann

2003). Other models of morphological evolution, for example,

the suggested significance of “proximal endites” (see

WaloЯek and Mьller 1997), may have to contend with alternative

hypotheses (Schram and Koenemann 2001). Nevertheless,

Orsten arthropods play a key role in reconstructing

the phylogeny of the crustaceans.

The Burgess and Chengjiang Faunas

The fossils of the Burgess Shale of British Columbia, Canada,

have long been known for their unusual preservation and

unique array of arthropod body forms (see Gould 1989). In

Figure 19.4. Disparity of

phylogenetic results from

molecular sequencing: trees

simplified from the published

versions, with crustacean clades

highlighted with open lines. (A)

Tree derived from 18S rDNA

(simplified from Spears and

Abele 1997); note polyphyletic

crustaceans, and the longbranch

attraction of a clade of

remipedes, cephalocarids, and

mystacocarids. (B) Tree derived

from EF-1a and Pol II (simplified

from Shultz and Regier

2000); note paraphyletic

crustaceans. (C) Total evidence

tree derived from eight molecular

loci and morphology

(simplified from Giribet et al.

(2001); note a more-or-less

monophyletic Crustacea (given

that long branch attractions are

undoubtedly operant in

producing the barnacle/fruit fly

clade).

324 The Relationships of Animals: Ecdysozoans

the remipede crustaceans (fig 19.6C and 19.6D) that also

have long homonomous trunks.

Clues from Developmental Genetics

The most exciting new source of information in recent years

that can be applied to issues of phylogeny is emerging from

the field of developmental genetics. This area of research is

leading to fundamentally new insights into relationships of

arthropods and our understanding of comparative anatomy.

As an example, Averof and Akam (1993, 1995) and Akam

et al. (1994) suggested that the patterns of expression of the

Hox gene complexes indicate that crustaceans and hexapods

share a common body plan. However, the Hox condition in

myriapods, chelicerates, or the near-arthropods such as tardigrades

and onychophorans was then, and still is largely

now, not known. To resolve relationships between two

groups, one needs to assess their position in reference to a

third, potential outgroup. We now know that Hox genes are

shared by all higher metazoans. Therefore, Hox genes in

broad aspect are plesiomorphic features and thus tell us little

directly about phylogenetic relationships within the groups.

For example, Cartwright et al. (1993) uncovered multiple

copies of the Hox gene complex in the horseshoe crab, Limulus.

These resemble the Hox B sequence in the mouse. The

multiple Hox clusters in horseshoe crabs could be considered

an autapomorphy of the limulids, probably. Even

though this is the only known occurrence of multiple Hox

gene complexes outside of the Chordata, no one is about to

suggest on this basis a return to the old arachnid theory for

vertebrate origins. We have to be careful with conclusions

on arthropod relationships based on Hox genes alone.

Nevertheless, individual aspects of Hox gene complexes

can be used for assessing homologies. A stunning example

concerns the examination of Hox expression in chelicerates

(Damen et al. 1998, Telford and Thomas 1998). There appears

clear evidence to indicate that the old ideas about the

lack of a deutocerebrum in chelicerate brains were wrong.

Rather, Hox gene patterns indicate that the relevant “deutocerebral

segment” is in fact present in chelicerates. As a result,

the chelicerae, those diagnostic limbs for horseshoe

crabs, true spiders, and sea spiders, are seen as homologous

to the hexapod antennae and, by extension, the crustacean

antennules (first antennae).

Schram and Koenemann (2001) examined modes of limb

formation among arthropods and compared the resulting

limb patterns with those found in fossils. In this instance,

the purported close affiliation of two well-known fossil species,

Lepidocaris rhyniensis from the Devonian and Rehbachiella

kinnekullensis from the Cambrian Orsten, to the living

brine shrimp could not be supported. The development of

the legs seen in the fossils is entirely different from that of

the living brine shrimp and its relatives, precluding close

relationship.

Figure 19.5. Some representative Cambrian Orsten “crustaceomorphs”

all drawn to the same scale. (A) Henningsmoenicaris

scutula (modified from WaloЯek 1999). (B) Rehbachiellakinnekulensis

(modified from WaloЯek 1993). (C) Bredocaris

aadmirabilis (modified from Mьller and WaloЯek 1988). (D)

Skara anulata (modified from Mьller and WaloЯek 1985). (E)

Martinssonia elongata (modified from Mьller and WaloЯek

1986b).

Figure 19.6. Some typical long-bodied arthropods. (A).

Odaraia alata, from the Cambrian Burgess Shale, Canada (from

Briggs 1981). (B) Fuxianhuia protensa, from the Cambrian

Chengjiang, China. (C) A living geophilomorph centipede

(from Meglitsch and Schram 1991). (D) Body outline of

Speleonectes lucayensis, a remipede crustacean (from Schram

et al. 1986).

recent years, the Burgess fossils have been joined by an

equally well-preserved series of fossils from the Cambrian

beds of Chengjiang, China (Hou and Bergstrцm 1997,

Bergstrцm and Hou 1998). These assemblages largely consist

of macrofossils with body sizes up to 120 mm. They

present a contrasting group of species in apposition to the

micro-arthropods of the Orsten. Unlike the short-bodied

Orsten taxa, the trunks of many (although not all) of the

Canadian and Chinese fossils have a large number of

homonomous segments each bearing a pair of limbs identical

to each other (fig 19.6A,B). In this they resemble many

modern forms, for example, the long-bodied myriapods and

Are the Crustaceans Monophyletic? 325

Straitjackets of the Past

Finally, we need to highlight one other factor that strongly

influences discussions of phylogenetic relationships. It is

orthodoxy, the dogma of long-standing assumptions that

constrain thought and have become the heavy shackles of

tradition. “Analyzing phylogeny” is an exercise closely akin

to scriptural exegesis. The weight of authority sits heavy.

Although a book could be written on the subject, discussion

of two issues will illustrate the point.

We alluded above to the generally short-bodied nature

of Orsten arthropod body plans and how the potential significance

of this has not been fully absorbed by the most

authorities. The classic theories about the direction of arthropod

evolution have featured long, equally segmented

bodies as ancestral forms that gave rise to shorter-bodied

organisms. These descendants were believed to have segments

specialized into regions (like a thorax, or abdomen,

or pleon). It is difficult to identify the ultimate source for this

idea. The long-bodied ancestor theory for arthropods is the

result of a consensus arrived at when it was first suggested

that annelid worms were the precursors of arthropods in the

“great chain of being,” or “ladder of life.” Snodgrass (1935,

1952) is only one of the more recent advocates of this view

(fig 19.7).

The effect of this fully emerges when examining morphology-

based phylogenies of arthropods such as illustrated by

Schram and Hof (1997) or Wills (1997). The latter is particularly

relevant here because Wills (1997), as mentioned above,

performed an ingroup analysis, rooting his tree to the

Remipedia, because he believed this class of crustaceans had

the most primitive body plan. Schram and Hof (1997) achieved

the same result without explicitly making the assumption of

“long-bodied = primitive.” The effect was the same because

their chosen outgroups induced polarities determined by the

use of longer-bodied forms such as centipedes.

Another example of historical constraint arises out of the

classic definition of Crustacea, namely, that they possess a

“second antenna.” When some crustaceans really do bear

a second set of antennae on the segment immediately posterior

to the “first antennae,” we then deal with a true

apomorphy. However, we cannot score any limb on that

segment as a second antenna. In actual fact, the “first postantennal

limb” is by far one of the most variable appendages

amongst all the arthropods (fig 19.8). In addition to real

antennal specializations of that limb (fig 19.8D), we also see

at that position purely locomotory limbs (fig 19.8C), limbs

that function in feeding and locomotion (fig 19.8A,B), limbs

that serve as attachment structures, limbs that assist in copulation

(fig 19.8E), and reduced limbs of uncertain function

(fig 19.8F)—and these variants are only the alternatives seen

among crustaceomorphs. To this we might add limbs modified

as a “labrum” that apparently occur in insects and myriapods.

By assigning “traditional” simple labels in character

analyses, for example, presence or absence of a labrum or

a second antenna, phylogenetically important information

is potentially disregarded. By ignoring the rich variety of

anatomy expressed on that limb we mask some possible

apomorphies that could help sort phylogenetic relationships

among crustaceomorph arthropods.

A New Analysis

In light of the above commentary, we have conducted a new,

preliminary analysis of fossil and recent crustaceomorphs.

(We prefer this term, which serves to denote all the traditional

Crustacea sensu stricto and the various fossil stem

forms.) In this we paid particular attention to incorporating

as much useful information as possible from the Cambrian

and other relevant fossil forms. In combination with morphology

of modern forms, we attempted to be as “theory free”

as we could in scoring characters. In addition, constructed a

matrix using features that have not figured prominently in

the past, if at all, for example, body plans, gonopore locations,

Hox expressions, and limb ontogeny. We polarized our

data with a hypothetical ancestor that represents a shortbodied

form. A total of 42 characters (half of them multistate)

were used for 31 end groups. A series of alternative analyses

were performed, but in the end we preferred a partially

weighted and unordered data set for which we obtained 45

trees. We took the strict consensus of those trees and ex-

Figure 19.7. Diagrammatic schema for the evolution of a longbodied

articulate (at the top), through successive degrees of

body regionalization and specialization, into an arthropod (at

the bottom). Modified from Snodgrass (1935).

326 The Relationships of Animals: Ecdysozoans

plored alternative branchings for polychotomies to find resolved

branching patterns of minimum tree length.

Our working hypothesis is presented in figure 19.9. We

have highlighted the traditionally recognized crustacean taxa.

These are groups that can for the most part be defined by concise

sets of characters related to gonopore location and other

aspects of the body plan. The tree renders three separate clades

that until now we might have placed within the Crustacea sensu

stricto. These clades often contain fossil taxa, but we did not

automatically include fossils within the highlighted clades if

doing so would obscure the definition of the clade. One of

these clades, which might be termed “Eucrustacea,” contains

the Malacostraca (lobsters, crabs, and allies), a reduced array

of maxillopodans (including copepods, barnacles, and closely

related forms), remipedes, and cephalocarids. We obtained a

sister clade to this assemblage, which we might term the

“Pancrustacea.” This includes Branchiopoda, insects, and some

Devonian and Cambrian Orsten fossil species often included

within branchiopods. We cannot, however, include those fossils

within the branchiopods because they appear to be separated

from the brine shrimp and allies by the insects. A third

crustacean clade can be seen deeper in the tree and is composed

of the fish lice (branchiurans) and the mystacocarids

allied with Skara, an Orsten genus.

The tree clearly reveals a series of individual stem taxa

and stem clades leading toward the three clades with living

crustacean forms. However, it appears that at least some of

the crustacean-containing clades also contain within them

other stem forms, for example, Pancrustacea can be interpreted

as a crown clade with Lepidocaris and Rehbachiella as

stem forms to it. The juxtaposition of insects with branchiopods

is not as startling as it might appear. Galent and Carroll

(2002) and Ronshaugen et al. (2002) have recently shown

that slight changes in the Hox gene Ultrabithorax can achieve

an astounding shift of morphology in the fruit fly, Drosophila,

and the brine shrimp, Artemia. Finally, the large clade near

the base of the present tree in figure 19.9 contains an array

of long-bodied taxa, However, it is unclear whether this clade

will persist as we expand the database to include some additional

taxa and characters, as we move toward a more definitive

analysis.

Are Crustacea Monophyletic?

What is a crustacean? The classic definition (see Schram 1986)

says that crustaceans are arthropods that have (1) a five-segmented

head including two pairs of antennae, mandibles, and

two pairs of maxillae; (2) a tendency to fuse the head segments

to form a cephalic shield and to develop from the back of the

head a posterior out-growth (carapace); (3) a tendency to regionalize

the trunk into an anterior thorax and a posterior abdomen,

or pleon; and (4) anamorphic development that

typically begins with a unique larva or ontogenetic stage termed

the nauplius, or egg nauplius. This definition has essentially

served since the time of Cuvier and Latreille in the early 1800s.

However, WaloЯek (1999), in light of his work on Orsten

arthropods, proposed some modification of this definition: (1)

the antenna (2nd antenna) with a coxa proximal to the basis;

(2) the maxillules (1st maxilla) specialized for food transport;

(3) the labrum, atrium oris, paragnaths, sternum, and feeding

areas of limbs covered with fine hairs (setae) (even in the nauplius);

(4) the trunk terminating in a conical segment that bears

the anus terminally and a pair of caudal rami; and (5) the first

larval or developmental stage as a nauplius. Under either definition,

Crustacea remain as a monophyletic crown taxon.

Therefore, results of our new analysis indicating polyphyly

are rather significant. It would appear that the Crustacea sensu

stricto do not constitute a monophyletic group. WaloЯek (1999,

and elsewhere) has argued forcefully and convincingly for a

stem group/crown group understanding of crustacean history.

In our present analysis, we do obtain stem taxa arrays. However,

we have a multiplicity of crown groups, each of which

has their fossil stem taxa. All these clades in turn form a transition

series from the root of a “crustaceomorph” clade near the

Orsten genus Cambrocaris extending to diverse terminal groups

that we have traditionally identified as “Crustacea.”

Is their any validity to the concept of a monophyletic taxon

Crustacea? We could of course redefine what it is to be a crus-

Figure 19.8. Variations of morphology and function found on

the so-called second antenna of crustaceans. (A) The mystacocarid,

Derocheilocaris typicus, a locomotory/feeding limb. (B) The

cephalocarid, Hutchinsoniella macracantha, a locomotry/possibly

feeding limb. (C) Devonian fossil species, Lepidocaris rhyniensis,

probably a feeding limb. (D) The decapod, Penaeus setiferous, a

true “second” antenna. (E) Male brine shrimp, Branchinecta

campestris, a copulatory clasper. (F) Female Branchinecta, with

rudimentary limb stub of uncertain function. All modified from

Schram (1986).

Are the Crustaceans Monophyletic? 327

Figure 19.9. Preliminary

cladistic analysis of “crustaceomorph”

arthropods. Note the

three distinct clades of “crustaceans”

highlighted with the

open lines. There is a shortthorax,

mystacocarid/fish-lice

(Branchiura) clade, typically

placed in the past within

Maxillopoda. Brachiopoda sit in

a larger clade with insects

(Pancrustacea) and other fossil

forms, which in Schram (1986)

was a part of the class Phyllopoda.

Eucrustacea, with the

remaining crustaceans that have

gonopores located midbody on

the 6th through 8th thoracic

segments, which includes what

Schram (1986) classified as

Remipedia, Malacostraca, most

of the Maxillopoda, and

miscellaneous phyllopods.

tacean, that is, seek to identify the last common ancestor

of all the currently recognized groups we refer to as “Crustacea”

and extract the ground pattern for that “ancestor.”

These features then could serve to define that monophylum.

However, in doing that we would (1) dilute our understanding

of what it is to be a crustacean (and we might ask, in so

doing would such a watered-down definition really have

any meaning?) and (2) simply leave untouched the issue of

unreconciled disparity of body plans.

We prefer to accept the results of the analysis at face value

and acknowledge that Crustacea sensu stricto are not a monophyletic

group. This has two advantages. First, it maintains

328 The Relationships of Animals: Ecdysozoans

the primacy of Bauplдne, body plans, as a raison d’кtre for

recognizing major groups of arthropods. Second, it allows

us to better reconcile the disparity of results between morphological

and molecular analyses. The message of many of

the molecular studies has been that there is a problem with

maintaining monophyly for crustaceans. By reassessing the

assumptions upon which the morphological studies have

been based, we now can begin to see the way clear toward a

grand synthesis of cladistic phylogeny that can more effectively

integrate crustaceans into the Tree of Life.

Many times morphologists summarily reject the results

of molecular phylogenies because these trees do not agree

frequently with those derived from morphology alone. The

skepticism of morphologists may be justified, because molecular

databases are often less comprehensive regarding

taxon sampling. However, morphologists ignore at their peril

the message of the molecules, which should at least compel

a reexamination of the assumptions implicit in morphological

character surveys. Conversely, molecular systematists look

askance at the morphologists and derided them for their

archaic and “subjective” approaches to phylogeny development.

However, it is hubris on the part of molecular systematists

to persist in ignoring the conflicting results of different

molecule databases and not to acknowledge the financial and

technical limits of achieving real comprehensiveness of taxon

and character sampling in a molecular framework. Obviously,

a truce is needed for real cooperation to occur and promise

the hope of a coordinated consensus on this vexing issue of

crustaceomorph relationships with other arthropods on the

Tree of Life.

Acknowledgments

We want to express special thanks to Jan van Arkel, who

prepared the graphic figures, and Joris van der Ham, who

assisted us with cladistic analysis of the Cambrian Orsten forms.

We also want to pay tribute to the many workers on Cambrian

fossil arthropods, whose careful and detailed studies published

over the years have allowed us to better “see” the deep history of

the arthropods. In addition, special mention should go to Dieter

WaloЯek, with whom we have had numerous and intense

discussions over the years, and, although we do not always

agree, these exchanges have always served as a learning process.

Finally, Trisha Spears has been a guiding light for us on matters

concerning molecular phylogenies.

Literature Cited

Akam, M., M. Averof, J. Castelli-Gair, R. Dawes, F. Falciani, and

D. Ferrier. 1994. The evolving role of Hox genes in

arthropods. Development 1994 (suppl.):209–215.

Averof, M., and M. Akam. 1993. HOM/Hox genes of Artemia:

implications for the origin of insect and crustacean body

plans. Curr. Biol. 3:73–78.

Averof, M., and M. Akam. 1995. Hox genes and the diversification

of insect and crustacean body plans. Nature 376:420–

423.

Bergstrцm, J., and X. Hou. 1998. Chengjiang arthropods and

their bearing on early arthropod evolution. Pp. 151–184 in

Arthropod fossils and phylogeny (G. Edgecombe, ed.).

Columbia University Press, New York.

Briggs, D. E. G. 1981. The arthropod Odaraia alata Walcott,

Middle Cambrian, Burgess Shale, British Columbia. Phil.

Trans. R. Soc. Lond. B 291:541–584.

Cartwright, P., M. Dick, and L. W. Buss. 1993. HOM/Hox type

homeoboxes in the chelicerate Limulus polyphemus. Mol.

Phylogenet. Evol. 2:185–192.

Dahl, E. 1963. Main evolutionary lines among recent Crustacea.

Pp. 1–15 in Phylogeny and evolution of Crustacea (H.

Whittington and W. D. I. Rolfe, eds.). Museum of Comparative

Zoology, Cambridge.

Damen, W. G. M., M. Hausdorf, E.-A. Seyfarth, and D. Tautz.

1998. A conserved mode of head segmentation in arthropods

revealed by the expression pattern of Hox genes in a

spider. Proc. Natl. Acad. Sci. USA 95:10665–10670.

Edgecombe, G. D., G. D. F. Wilson, D. J. Colgan, M. R. Gray,

and G. Casis. 2000. Arthropod cladistics: combined analysis

of histone H3 and U2 sequences and morphology. Cladistics

16:155–203.

Galent, R., and S. B. Carroll. 2002. Evolution of a transcriptional

repression domain in an insect Hox protein. Nature

415:910–913.

Giribet, G., G. D. Edgecombe, and W. C. Wheeler. 2001.

Arthropod phylogeny based on eight molecular loci and

morphology. Nature 413:157–161.

Gould, S. J. 1989. Wonderful life. Norton, New York.

Hou, X., and J. Bergstrцm. 1997. Arthropods of the Lower

Cambrian Chengjiang fauna, southwest China. Fossils Strata

45:1–116.

Meglitsch, P. A., and F. R. Schram. 1991. Invertebrate zoology

3rd ed. Oxford University Press, New York.

Mьller, K. J., and D. WaloЯek. 1985. Skaracarida, a new order

of Crustacea from the Upper Cambrian of Vдstergцtland,

Sweden. Fossils Strata 17:1–65.

Mьller, K. J., and D. WaloЯek. 1986a. Arthropod larvae from

the Upper Cambrian of Sweden. Trans. R. Soc. Edinb. Earth

Sci. 77:157–179.

Mьller, K. J., and D. WaloЯek. 1986b. Martinssonia elongata

gen. et sp. n., a crustacean-like euarthropod from the Upper

Cambrian “Orsten” of Sweden. Zool. Scr. 15:73–92.

Mьller, K. J., and D. WaloЯek. 1988. External morphology and

larval development of the Upper Cambrian maxillopod

Bredocaris admirabilis. Fossils Strata 23:1–70.

Regier, J. C., and J. W. Shultz. 1998. Molecular phylogeny of

arthropods and the significance of the Cambrian “explosion”

for molecular systematics. Am. Zool. 38:918–928.

Ronshaugen, M., N. McGinnis, and W. McGinnis. 2002. Hox

protein mutation and macroevolution of the insect body

plan. Nature 415:914–917.

Schram, F. R. 1984. Relationships within eumalacostracan

crustaceans. Trans. San Diego Soc. Natl. Hist. 20:301–312.

Schram, F. R. 1986. Crustacea. Oxford University Press, New York.

Schram, F. R. 1993. Review of: Boxshall, G. A., J.-O. Strцmberg,

and E. Dahl. 1992. The Crustacea: origin and evolution.

Acta Zool. 73:271–392. J. Crust. Biol. 13: 820–822.

Are the Crustaceans Monophyletic? 329

Schram, F. R., and C. H. J. Hof. 1997. Fossils and the interrelationships

of major crustacean groups. Pp. 233–302 in

Arthropod fossils and phylogeny (G. Edgecombe, ed.).

Columbia University Press, New York.

Schram, F. R., and S. Koenemann. 2001. Developmental

genetics and arthropod evolution: part I, on legs. Evol. Dev.

3:343–354.

Schram, F. R., and S. Koenemann. 2003. Developmental

genetics and arthropod evolution: on body regions of

Crustacea. Crust. Issues 15:75–92.

Schram, F. R., J. Yager, and M. J. Emerson. 1986. Remipedia,

Pt. 1: Systematics. San Diego Soc. Nat. Hist. Mem. 15:1–60.

Shultz, J. W., and J. C. Regier. 2000. Phylogenetic analysis of

arthropods using two nuclear protein-encoding genes

supports a crustacean + hexapod clade. Proc. R. Soc. Lond.

B 267:1011–1019.

Sieg, J. 1983. Evolution of Tanaidacea. Crust. Issues 1:229–256.

Snodgrass, R. E. 1935. Principles of insect morphology.

McGraw-Hill, New York.

Snodgrass, R. E. 1952. A textbook of arthropod anatomy.

Comstock, Ithaca, NY.

Spears, T., and L. G. Abele. 1997. Crustacean phylogeny inferred

from 18S rDNA. Pp. 169–187 in Arthropod relationships

(R. Fortey and R. Thomas, eds.). Chapman and Hall, New

York.

Telford, M. J., and R. H. Thomas. 1998. Expression of homeobox

genes shows chelicerate arthropods retain their deutocerebral

segment. Proc. Natl. Acad. Sci. USA 95:10671–10675.

WaloЯek, D. 1993. The Upper Cambrian Rehbachiella and the

phylogeny of Branchiopoda and Crustacea. Fossils Strata

32:1–202.

WaloЯek, D. 1999. On the Cambrian diversity of Crustacea.

Pp. 3–27 in Crustaceans and the biodiversity crisis (F. R.

Schram and J. C. von Vaupel Klein, eds.). Koninklijke Brill,

Leiden.

WaloЯek, D., and K. J. Mьller. 1989. A second type A-nauplius

from the Upper Cambrian “Orsten” of Sweden. Lethaia

22:301–306.

WaloЯek, D., and K. J. Mьller. 1990. Upper Cambrian stemlineage

crustaceans and their bearing upon the monophyletic

origin of Crustacea and the position of Agnostus. Lethaia

23:409–427.

WaloЯek, D., and K. J. Mьller. 1997. Cambrian “Orsten”-type

arthropods and the phylogeny of Crustacea. Pp. 139–153 in

Arthropod relationships (R. Fortey and R. Thomas, eds.).

Chapman and Hall, New York.

WaloЯek, D., and K. J. Mьller. 1998. Early arthropod phylogeny

in light of the Cambrian “Orsten” fossils. Pp. 185–231 in

Arthropod fossils and phylogeny (G. Edgecombe, ed.).

Columbia University Press, New York.

Wills, M. A. 1997. A phylogeny of recent Crustacea derived

from morphological characters. Pp. 189–209 in Arthropod

relationships (R. Fortey and R. Thomas, eds.). Chapman

and Hall, New York.

Wilson, G. D. F. 1992. Computerized analysis of crustacean

relationships. Acta Zool. 73:383–389.

Навигация