21 Phylogeny of the Holometabolous Insects

Back

The Most Successful Group of Terrestrial Organisms

Michael F. Whiting

345

The radiation and diversification of the holometabolous insects

stand as two of the grandest events in all of evolutionary

history, representing an unprecedented explosion in

species coupled with extensive anatomical and physiological

specialization. The defining characteristic for Holometabola

is complete metamorphosis: every insect in this group,

with rare exception, passes through an egg, larval, pupal, and

adult stage. This is in contrast to the non-holometabolous

insect groups in which juveniles have more or less the same

form as the adult, live in the same environment, and exploit

similar resources. Although it has never been thoroughly

tested, it is thought that the evolution of complete metamorphosis

was the key innovation allowing these insects to partition

habitats between adults and juveniles, resulting in a

wider range of niches that could be occupied by the nascent

species. And occupied they have. Holometabola includes well

more than one million species representing roughly 80% of

all described insect species and just more than half of the total

number of described species on Earth today (Kristensen 1999,

Wilson 1988). The immense size of this group and their

unique morphological specializations present a serious challenge

to phylogenetic systematics. However, current research

is providing new insight into the evolution and diversification

of this, the most successful group of terrestrial organisms,

and in the past few years researchers have finally begun

to unravel the Tree of Life for holometabolous insects.

Holometabola appear to be a true evolutionary group in

the sense that all members of Holometabola can trace their

evolutionary history back to a single ancestor (i.e., Holometabola

are monophyletic). This is evidenced by the fact

that all members of Holometabola undergo complete metamorphosis,

and that they have some other distinct morphological

characteristics shared by no other insect groups

(Kristensen 1999, Whiting 1998a). For instance, holometabolans

are the only insects in which the larval eyes disintegrate

and the adult eyes develop de novo during the last

immature stage. The developing wings in the larvae of holometabolous

insects are kept inside the body until the larval-pupal

molt, whereas in other insect groups the developing wing

appears on the outside of the body in early nymphal stages.

In fact, the group Holometabola is often called Endopterygota

(internal-winged) because of this feature. Likewise, external

genitalia do not appear until the penultimate (larval-pupal)

molt. In addition, phylogenetic analysis of DNA sequence data

consistently supports the monophyly of Holometabola. With

the possible exception of the group Neoptera (winged insects),

there is no other major group of insects whose monophyly is

more strongly supported than that of Holometabola.

Holometabola are composed of 11 major living lineages,

each of which is also a monophyletic group (with one exception,

described below). Entomologists have given each

of these lineages the taxonomic ranking of an order, but the

number of species within each of these orders is drastically

unequal, reflecting both the morphological specialization

and the differential success of particular groups (table 21.1).

The majority of holometabolous insect species are placed

346 The Relationships of Animals: Ecdysozoans

within four megadiverse orders: approximately 500,000

species of beetles (order Coleoptera), 160,000 species of

bees, wasps, and ants (order Hymenoptera), 150,000 species

of flies (order Diptera), and 150,000 species of butterflies

and moths (order Lepidoptera). Additional species are

added to each of these orders on almost a daily basis, and

it is clear that we have only scratched the surface of species

diversity within these groups. The remaining seven orders

are less diverse, although they include some of the most

peculiar and specialized forms. These include caddisflies

(order Trichoptera) with roughly 7000 species, lacewings

(order Neuroptera) with 6000 species, fleas (order Siphonaptera)

with ~2400 species, twisted-winged parasites

(order Strepsiptera) with 532 species, scorpionflies (order

Mecoptera) with 500 species, dobsonflies and alderflies

(order Megaloptera) with 270 species, and snakeflies (order

Raphidioptera) with 205 species. There are good morphological

characters to support the monophyly of most of

these groups, and for well more than a century any newly

described insect with complete metamorphosis could be

easily assigned to one of these living lineages.

What we do not know, however, is the exact pattern of

phylogenetic relationships among each of the 11 holometabolous

insect orders. A child can tell a beetle from a wasp from

a butterfly, but even the entomologically erudite is left pondering

which two insects are most closely related. A few

hypotheses of interordinal phylogenetic relationships will be

presented below, but there many unanswered questions still

remain. Likewise, relationships within each of the holometabolous

insect orders are often obscure, although major insights

are being made each year. This chapter focuses on what we

think we know about holometabolan phylogeny, what relationships

are more dubious, and pinpointing major gaps in

our knowledge of holometabolan phylogeny.

Interordinal Phylogeny

Many hypotheses have been presented for phylogenetic relationships

among the holometabolous insect orders over the

past century; these reflect the general difficulty of reconstructing

the evolutionary history of this important insect group

and the variety of opinions on the matter. Summaries of the

most influential and current hypotheses are presented in figure

21.1. Boudreaux (1979; fig. 21.1A) and Hennig (1981;

fig. 21.1B) presented phylogenies based on different interpretations

of morphological characters. Both of these workers

compiled and discussed evidence for insect phylogeny

based on morphological (anatomical) data, but because neither

presented any formal analyses of these data, it remained

unclear how well a particular phylogenetic tree was supported

by the underlying data. Boudreaux placed Strepsiptera

+ Coleoptera as the most primitive holometabolan lineage

and then argued for the placement of Hymenoptera at the

base of the remaining orders. However, the questionable

morphological data he presented coupled with the particular

twist he put on the interpretation of these data (e.g., arguing

that the most common morphological feature must be

the most primitive feature), leave his conclusions unsatisfying.

Hennig was influential in the development of phylogenetic

theory and is widely considered the father of modern

phylogenetics, although he was also challenged by his attempts

to provide a complete view of insect ordinal relationships.

Hennig was uncertain as to the placement of

Hymenoptera and Siphonaptera but argued for a sistergroup

relationship between Strepsiptera and Coleoptera, and

associated Trichoptera + Lepidoptera with Diptera +

Mecoptera. Kristensen is the most influential morphological

worker in recent memory, and his summaries of insect ordinal

phylogeny (Kristensen 1975, 1981, 1991, 1995, 1999)

provide excellent commentary on the wide variety of morphological

evidence that has been garnered to support different

phylogenetic hypotheses. In his most recent summary

(Kristensen 1999; fig. 21.1C), Holometabola are divided into

two main divisions. The Coleoptera + Neuropterid lineages

(Neuroptera, Megaloptera, and Raphidioptera) form one

division, and the remaining orders are placed in a second

division (Hymenoptera + Mecopterida), with uncertainty as

to the position of the enigmatic Strepsiptera (more on this

below). Recently, Beutel and Gorb (2001) added a suite of

morphological characters associated with the tarsi of insects

and proposed a phylogeny that agrees with Kristensen

(1999) except for the position of Strepsiptera as sister group

to Coleoptera.

Although a few attempts had been made from a molecular

standpoint to decipher holometabolan phylogeny

(Carmean et al. 1992, Chalwatzis et al. 1996, Pashley et al.

1993), Whiting et al. (1997) was the first the presentation

of a formal analysis of morphological data in combination

with extensive DNA sequence data for Holometabola. These

Table 21.1

Holometabolous Insect Orders and Common Names.

Order Common name

Coleoptera Beetles

Neuroptera Lacewings, antlions, owlflies

Megaloptera Alderflies, fishflies, dobsonflies

Raphidioptera Snakeflies

Hymenoptera Bees, wasps, ants

Trichoptera Caddisflies

Lepidoptera Butterflies, moths, skippers

Mecoptera Scorpionflies

Siphonaptera Fleas

Strepsiptera Twisted-winged parasites

Diptera Flies

Nannomecoptera Nannochoristid scorpionflies

Neomecoptera Snow fleas (Boreidae)

Phylogeny of the Holometabolous Insects 347

Figure 21.1. Previous phylogenetic hypotheses of relationships among holometabolous insect

orders. (A) Boudreaux (1979), based on morphology. (B) Hennig (1981), based on morphology.

(C) Kristensen (1999), based on morphology. (D) Whiting et al. (1997) and Wheeler et al. (2001),

based on morphology and DNA. (E) Whiting (2002c), based on extensive sample of DNA sequences.

(F) Summary tree representing current state of knowledge. Dashed lines represent

uncertain relationships.

data consisted of 176 morphological characters coded across

Holometabola and outgroups, and portions of the 18S ribosomal

DNA (rDNA) molecule (~1000 nucleotides) and 28S

rDNA (~400 nucleotides). Wheeler et al. (2001) expanded

this study to include all hexapod orders and used a new

analytical tool that obviates the need to generate a multiple

alignment of the DNA sequence data before phylogenetic

reconstruction (i.e., optimization alignment). Both studies

largely concurred in their view of holometabolan phylogeny

(fig. 21.1D). These results were surprising in three ways: (1)

they suggested a sister-group relationship between the enigmatic

Strepsiptera and Diptera; (2) they demonstrated a close

association of fleas with a family placed within the scorpionflies

(Mecoptera); and (3) although their topology is

largely congruent with those trees presented by Kristensen,

their results indicate that many holometabolan interordinal

relationships are not particularly well supported. Whiting

(2002b, 2002c) performed more extensive molecular analyses

based on the entire 18S rDNA gene for roughly three times

more holometabolan species than in earlier studies. Although

this increased species sampling helped resolve some relationships

(e.g., better support for Neuropterida), the general

pattern of relationships provided by this single molecule is

in some cases different than those found with morphology

(fig. 21.1E).

So what do these studies tell us? All workers agree that

there are two well-supported relationships among the holometabolous

insect orders (table 21.2). The first is a sistergroup

relationship between Lepidoptera and Trichoptera to

form a group called Amphiesmenoptera. This relationship

is supported by more than 15 morphological characters, including

the female heterogamy (essentially, females possess

the XY chromosome) and the presence of scales or hairs on

the wing surface between veins (Hennig 1981, Kristensen

348 The Relationships of Animals: Ecdysozoans

1997, Whiting et al. 1997). This group has been found in

every DNA phylogenetic analysis to date (Chalwatzis et al.

1996, Wheeler et al. 2001, Whiting 2002c, Whiting et al.

1997) and is considered the best-supported sister-group

relationship in all of insect ordinal phylogeny. Second, all

hypotheses agree that the orders Neuroptera, Raphidioptera,

and Megaloptera should be placed in a single group called

Neuropterida. The monophyletic grouping of the neuropterids

is supported by a series of specializations associated

with the female ovipositor (Mickoleit 1973), and this group

is also consistently recovered in phylogenetic analyses based

on DNA sequence data (Wheeler et al. 2001, Whiting 2002b,

2002c). Molecular data consistently support a sister-group

relationship between Megaloptera and Raphidioptera, which

agrees with some morphological evidence (Wheeler et al.

2001, Whiting 2002c). An alternative hypothesis is that

Megaloptera and Neuroptera are sister groups based on the

presence of aquatic larvae, found in all Megaloptera and one

primitive family of Neuroptera (Nevrorthidae), although the

vast majority of neuropterans are terrestrial, with the exception

of the more derived spongillaflies (Aspцck et al. 2001).

Beyond Neuropterida and Amphiesmenoptera, the picture

becomes murky and the hypotheses more controversial.

This is largely because most of the holometabolous insect

orders are so highly specialized that it becomes difficult to

unravel the morphological clues required to determine phylogenetic

affinity. Very often the morphological evidence

presented to support hypothesized relationships consists of

only one or two characteristics that are not universally shared

by members of those groups, and the homology among these

characters is questionable. Moreover, different specialists

have different interpretations of morphology leading to dramatically

different estimates of phylogeny.

Current morphological analyses suggest that Holometabola

may be divided into two major groups: Coleoptera +

Neuropterida and Hymenoptera + Mecopterida (= Trichoptera

+ Lepidoptera + Mecoptera + Siphonaptera + Diptera).

The position of the enigmatic Strepsiptera is discussed below.

The sister-group relationship between the lacewings and

the beetles is supported by specific modifications of the ovipositor

(Kristensen 1991) and characters associated with the

base of the hind wing in these insects (Hцrnschemeyer 2002;

fig. 21.1C). The monophyly of Mecopterida is supported by

the presence of a muscle that is attached between the thorax

wall (i.e., pleuron) and a hardened structure at the base of

the wing (i.e., first axillary sclerite; Kristensen 1999), although

this character is not present in the wingless fleas.

Within Mecopterida, Lepidoptera and Trichoptera form the

group Amphiesmenoptera (as discussed above), and Diptera

+ Mecoptera + Siphonaptera form another group. Morphological

data combined with molecular data suggest that fleas

actually are an offshoot of one scorpionfly lineage. Boudreaux

(1979) placed Hymenoptera as one of the most basal members

of Holometabola (fig. 21.1A) but did not provide a

convincing argument to support this position. Kristensen

(1991, 1999) argues that Hymenoptera should be placed

as sister group to Mecopterida, based on two characters

associated with the form of the larvae and one based on a

particular modification of the sucking pump in the adult

insect (Kristensen 1999).

DNA sequences are presently being generated to try and

provide independent estimates of ordinal phylogeny, and

although these data have provided new insight into some of

the more nebulous questions, the overall view of ordinal

phylogeny is still under construction. From a molecular

standpoint, the problem has been that the few DNA markers

that are commonly used in insect ordinal phylogeny are

not informative for all portions of the phylogeny, so additional

gene regions need to be investigated to provide a more

robust estimate of the holometabolan branches of the Tree

of Life. The hope is that these additional data will provide

new insights in the patterns of diversification across Holometabola.

Although the picture is not yet clear, the current

DNA data have pointed to some very interesting relationships.

For instance, data from four independent genes suggest

that the fleas are sister group to the snow scorpionflies

(Boreidae), a family of scorpionflies that live on the snow and

are closely associated with moss (Whiting 2002a). Once the

molecular data suggested this relationship, a reevaluation of

morphology demonstrated that this is a plausible hypothesis.

Morphological features supporting this relationship

include the presence of unusual spines in the gut (proventriculous;

Schlein 1980), multiple sex chromosomes

(Bayreuther and Brauning 1971), a series of specializations

associated with the female ovaries (Bilinski et al. 1998), and

the ability to jump via a similar mechanism. These data suggest

that fleas did not evolve from a group of flies, as has been

proposed (Byers 1996), but rather were living on the snow

and then shifted to mammal burrows where they became

obligate, external parasites. An additional mecopteran lineage

of small and obscure insects (Nannochoristidae) is the

most primitive group of Mecoptera, based on both molecular

(Whiting 2002a) and morphological data (Simiczyjew

2002, Willmann 1987). These findings indicate that Mecoptera

are not monophyletic and that if the Siphonaptera are

to be retained as a recognized order, it must be subdivided

Table 21.2

Superordinal Groups in Insect Phylogeny.

Superordinal name Groups included

Neuropterida Neuroptera + Megaloptera +

Raphidioptera

Mecopterida Lepidoptera + Trichoptera + Siphonaptera

+ Diptera + Strepsiptera

Amphiesmenoptera Lepidoptera + Trichoptera

Antliophora Mecoptera + Siphonaptera + Diptera +

Strepsiptera

Halteria Strepsiptera + Diptera

Phylogeny of the Holometabolous Insects 349

into additional insect orders. Given that current classification

does not allow non-monophyletic groups to be formally

named, it is necessary to recognize the additional orders

Nannomecoptera (for Nannochoristidae) and Neomecoptera

(snow scorpionflies; fig. 21.1F). Hinton (1958) was the first

to present a series of morphological characters to elevate

snow scorpionflies to their own order, Neomecoptera.

The most perplexing question in holometabolan phylogeny,

and the one that has received the most attention in

recent years, has been the controversy surrounding the

placement of Strepsiptera. This is an unusual group of insects,

members of which spend most of their lives as obligate

internal parasites of other insects. From a morphological

standpoint, the adult females are so highly reduced and

larvalike that they leave no clues as to their phylogenetic

position. The males are highly derived with unusual eyes,

mouthparts, and other structures and are so specialized that

it has been very difficult to assign them to any particular

phylogenetic group. This perplexing amalgamation of morphological

reduction in females and extreme modification in

males, combined with unusual biology and larval characteristics,

has challenged systematic placement of this group for

more than two centuries. Strepsiptera were associated with

Coleoptera, either as a member of Coleoptera (Crowson

1960) or as sister group to Coleoptera, based on wing morphology

and function (Kathirithamby 1989, Kristensen 1981,

1991, Kukalova-Peck and Lawrence 1993). Detailed examination

of this evidence, however, suggests that these characters

are based on mistaken descriptions of strepsipteran

wing morphology and function (Beutel and Haas 2000,

Kinzelbach 1990, Pix et al. 1993, Whiting 1998b). Current

DNA sequence data strongly support a sister-group relationship

between Strepsiptera and Diptera to form a group called

Halteria (Wheeler et al. 2001, Whiting 2002c, Whiting et al.

1997, Whiting and Wheeler 1994). This result has been

challenged as a methodological artifact of a particular mode

of data analysis (Huelsenbeck 1997), although, as has been

argued elsewhere (Sidall and Whiting 1999, Whiting 1998a)

that these criticisms are off the mark. If Strepsiptera are sister

group to Diptera, then the similarities in the form and

function of their modified wings might be attributed to evolution

via shifts in development, providing new insights into

how organisms can evolve in leaps and bounds across evolutionary

time. Nonetheless, Diptera + Strepsiptera is still

controversial, and additional data are needed before this relationship

is universally accepted.

In summary, current DNA sequence data support the

monophyly of most of the holometabolous insect orders,

in agreement with morphology. DNA also supports the

superordinal groups Amphiesmenoptera, Neuropterida,

and Halteria and the relationship among Mecoptera and Siphonaptera

as described above. DNA has not, however, been

successful at confirming the relationships hypothesized

by morphology, such as Mecopterida, Hymenoptera +

Mecopterida, or Coleoptera + Neuropterida. A tree summarizing

the current state of affairs in holometabolan phylogeny

(fig. 21.1F) indicates that further work is needed to

elucidate the more ancient patterns of holometabolan evolution

and diversification.

Coleoptera (Beetles)

Beetles are widely considered the most successful group of

organisms, with estimated numbers of species ranging from

500,000 to several million (Hammond 1992). Coleoptera

appears to be a well-supported monophyletic group characterized

by the presence of front wings that are rigid, hardened,

and typically cover the entire abdomen (elytra), as well

as 20 morphological features unique to this group (Beutel

and Haas 2000). Ironically, all molecular studies to date suggest

that beetles do not form a natural grouping of species

(Caterino et al. 2002, Wheeler et al. 2001, Whiting 2002b,

Whiting et al. 1997), but this is probably more indicative of

the inadequacy of the current DNA evidence rather than

substantial evidence of coleopteran paraphyly.

Coleoptera are divided into four major lineages that are

treated as suborders: Archostemata, Myxophaga, Adephaga,

and Polyphaga (fig. 21.2). Except for the basal placement of

Archostemata, relationships among the other three suborders

are controversial. Morphological evidence places Adephaga

as sister group to Myxophaga + Polyphaga (Beutel and Haas

2000), but recent molecular analyses suggest that Adephaga

are sister group to Polyphaga, with Myxophaga placed at their

base (Caterino et al. 2002). Archostemata include four small,

living families, although this group was more extensive formerly,

as shown by the fossil record. Archostematan larvae

are wood borers, and the monophyly of this suborder is

supported by some discrete adult and larval characteristics.

Myxophaga also include four families of small to minute semiaquatic

beetles, and overall this group appears to be well

supported based on a series of morphological features (Beutel

and Haas 2000). Myxophaga and Archostemata account for

less than 1% of the living beetle diversity.

Adephaga include ~30,000 species in a dozen families

and comprises ~10% of beetle diversity. This group includes

tiger beetles, ground beetles, whirligigs, predaceous diving

beetles, wrinkled bark beetles, and others. The monophyly

of this suborder also appears to be well supported, although

relationships among the constituent families are more controversial

and are focused on whether the aquatic taxa (Hydradephaga,

six families) and terrestrial taxa (Geadephaga, six

families) form two distinct lineages within this suborder. A

recent molecular analysis suggests that the aquatic taxa are

monophyletic and proposes a phylogeny for the 12 families

(Shull et al. 2001).

The suborder Polyphaga includes the vast majority of beetle

diversity, with at least 300,000 described species from more

than 100 families. In polyphagan beetles, the lateral side of the

prothorax (pleuron) is not externally visible, making the pro350

The Relationships of Animals: Ecdysozoans

Figure 21.2. Summary phylogeny

of beetles (Coleoptera).

thorax appear as a single dorsal plate that wraps around the

lateral sides of the prothorax. It appears likely that adoption of

a plant feeding lifestyle in these beetles early in angiosperm

evolution correlates with the great number of species in some

of the major beetle lineages (Farrell 1998). Detailed phylogenetic

relationships among most families are unknown, and this

is large part because of the overwhelming diversity of anatomical

features in this group and the enormous number of species the

systematist must deal with. The monophyly of some families is

in doubt, but work by a number of beetle specialists has provided

a glimpse of polyphagan phylogeny (Crowson 1960,

Lawrence and Newton 1982, 1995, Lawrence et al. 1995).

Polyphaga are divided into four major lineages, Scarabaeiformia,

Elateriformia, Bostrichiformia, and Cucujiformia, although

relationships among these lineages are largely unknown. Scarabaeiformia

include three superfamilies: Scarabaeoidea (13 families,

including scarabs, stag beetles, dung beetles, bess beetles),

Hydrophiloidea (four families, including water scavenger

beetles and hister beetles), and Staphylinoidea (seven families,

including carrion beetles and the extremely large family of rove

beetles). Elateriformia include five superfamilies, phylogenetic

relationships among which are largely unknown. This group

includes Scirtoidea (four families, including marsh beetles and

fringe-winged beetles), Dascilloidea (two families, including

soft-bodied plant beetles and cedar beetles), Buprestoidea (one

family, the metallic wood-boring beetles), Byrrhoidea (12 families,

including pill beetles, riffle beetles, water-penny beetles),

and Elateroidea (16 families, including click beetles, net-winged

beetles glowworms, fireflies, soldier beetles, etc.). Bostrichiformia

are composed of two superfamiles: Derodontoidea (one

family, tooth-necked fungus beetles) and Bostrichoidea (six

families, including skin beetles, twig borers, and spider beetles).

Cucujiformia are the largest and most diverse beetle lineage,

including the vast majority of plant-eating beetles. The monophyly

of this group is supported by a specialized type of

malpighian tubule (essentially, the insect kidney) and is composed

of six superfamilies. Lymexeloidea (one family, ship timber

beetles), Cleroidea (seven families, including checker beetles

and soft-winged flower beetles), Cucujoidea (31 families, including

flat bark beetles, lizard beetles, pleasing fungus beetles,

ladybugs, etc.), Tenebrionoidea (26 families, including darkling

beetles, blister beetles, antlike flower beetles, tumbling

flower beetles, etc.), Chrysomeloidea (four families, including

long-horn beetles and leaf beetles), and Curculionoidea (nine

families, including weevils and bark beetles). Given the enormous

size of Coleoptera, it may take half a century to construct

Phylogeny of the Holometabolous Insects 351

a phylogeny as detailed as those currently available for most

vertebrate groups.

Neuropterida (Lacewings, Snakeflies,

Alderflies, Dobsonflies)

Neuropterida are composed of three closely related orders:

Neuroptera (17 families), Megaloptera (two families), and

Raphidioptera (two families). Adults have large, separated

eyes, mandibulate mouthparts, and multisegmented antennae.

Collectively, this group includes individuals that exhibit

a broad range of morphological and biological diversity, and

the living species are remnants of what were once more diverse

lineages, as evidenced by their rich fossil record (Aspцck

et al. 2001). As larvae, many neuropterans are voracious

predators of other insects, especially the brown and green

lacewings and the antlions. Other families have become more

specialized, including the spider egg-sac predation in the

mantis lacewings (Mantispidae) and the freshwater spongefeeding

spongillaflies (Sisyridae).

The monophyly of Neuroptera is supported chiefly by the

larvae possessing piercing, sucking tubes modified from the

primitive chewing mouthparts. In addition, the anterior intestinal

tract is not connected to the posterior intestinal tract

in the larvae, such that they are unable to pass solid waste until

the insect becomes an adult and the gut is fully connected

(Aspцck et al. 2001). The monophyly of Megaloptera is supported

by the presence of lateral, segmented tracheal gills in

larvae that allows the larval insect to respire underwater. The

monophyly of Raphidioptera is supported by an elongated

neck and a pronotum that wraps around the lateral (pleural)

regions of the thorax (Wheeler et al. 2001). There has been a

suggestion that the megalopteran alderflies (Sialidae) may be

sister group to the snakeflies (Raphidioptera), rendering the

Megaloptera paraphyletic (Stys and Bilinksy 1990), but this

interpretation is not widely accepted (Aspцck et al. 2001). As

discussed above, there is a debate as to the phylogenetic relationships

among these orders, with the molecular data strongly

arguing for Megaloptera + Raphidioptera, as well as some morphological

characters (Whiting 2002b, 2002c), versus some

revised morphological characters arguing for Megaloptera +

Neuroptera (Aspцck et al. 2001).

Relationships among neuropteran families have been historically

controversial and have most recently been investigated

quantitatively by Aspцck et al. (2001) and Aspцck (2002).

According to Aspцck, Neuroptera are divided into three main

lineages: antlion-like lacewings (Myrmeleontiformia), lacewinglike

(Hemerobiiformia), and Nevrorthiformia, including one

obscure family (Nevrorthidae; fig. 21.3). The Myrmeleontiformia

include antlions (Myrmeleontidae), owlflies (Ascalaphidae),

spoon-winged lacewings (Nemopteridae), and two

additional, rather obscure families. This group is supported

by wing and larval characteristics and is one of only two wellsupported

relationship across neuropteran phylogeny. There

is debate as to the relationships within Myrmeleontiformia,

particularly regarding the position of Psychopsidae and

Nymphidae.

Hemerobiiformia consist of 11 families, including brown

and green lacewings (Hemerobiidae and Chrysopidae), dusty

wings (Coniopterygidae), mantidflies (Mantispidae), spongillaflies

(Sisyridae), and other groups. The monophyly of this

group is questionable, although the “dilarid clade,” including

Dilaridae, Mantispidae, Rhachiberothidae, and Berothidae, is

well supported by characteristics associated with the larval

head capsule. With the exception of the dilarid clade, relationships

among the constituent families within this group are also

questionable. The Nevrorthiformia include an obscure group

of lacewings with aquatic larvae that have been placed as the

most primitive group within Neuroptera, although this is certainly

open to further investigation.

One of the more interesting questions in neuropteran evolution

has been the suggestion that Neuroptera were derived

from an aquatic ancestor. This hypothesis is based on a phylogenetic

topology where the entirely aquatic Megaloptera are

sister group to Neuroptera, and the aquatic Nevrorthidae are

the most basal neuropteran lineage (Aspцck et al. 2001). If it

turns out that Megaloptera and Raphidioptera are indeed sister

groups, as indicated by current molecular data, or that

Nevrorthidae are not the most basal lineage, then the aquatic

origin hypothesis will be left without much merit. Clearly,

Figure 21.3. Summary phylogeny of Neuropterida, including

Megaloptera (alderflies and dobsonflies), Raphidioptera

(snakeflies), and Neuroptera (lacewings, antlions, owlflies, etc.).

Dashed lines represent uncertain relationships.

352 The Relationships of Animals: Ecdysozoans

there is a need to further investigate phylogenetic relationships

among these interesting insects.

Hymenoptera (Sawflies, Bees, Wasps, Ants)

Hymenoptera are currently composed of ~150,000 described

species, but when all the undescribed species are added, the

group may be twice this size (Kristensen 1999), putting it

on par with Coleoptera. Hymenopterans are found within

most terrestrial ecosystems and play a vital role in pollination

of flowering plants and as predators and parasites of

other insects, with ants alone forming a major component

of tropical ecosystems. Hymenopterans range in size from

microscopic parasites of insect eggs to very large bees and

wasps. This group is characterized by the presence of specialized

hooks that join the hind wings to the forewings (hamuli),

absence of notal coxal muscles, and the presence of a

unique reproductive mode known as haplodiploidy.

Hymenoptera have been traditionally divided into two

groups: Symphyta (sawflies and allies) and Apocrita (bees,

wasps, and ants; fig. 21.4). In Symphyta, the thorax is three

segmented and broadly joined to the abdomen, and the wing

venation is relatively complete. Most of the members of this

group are external feeders on foliage and have an ovipositor

that is somewhat sawlike, hence the common name

“sawflies.” Comparative morphological work suggests that

Symphyta as a whole are not monophyletic, but Tenthredinoidea

(five sawfly families) and Megalodontoidea (two

families, web-spinning sawflies) are monophyletic (Ronquist

et al. 1999, Schulmeister et al. 2002, Vilhelmsen 1997).

The xyelid sawflies are considered the most primitive of all

Hymenoptera, and morphological data suggest that the parasitic

wood wasps (Orussidae) form a sister group to Apocrita

(Ronquist 1999), although molecular data suggest other alternatives

(Dowton and Austin 1999).

The monophyletic Apocrita contain the vast majority of

hymenopteran species diversity. In contrast to Symphyta, in

Apocrita the first abdominal segment (propodeum) is fused

to the thorax to form a mesosoma, and the second abdominal

segment (and sometimes the third) is constricted to form

a petiole, the threadlike waist seen in wasps, bees, and ants.

Traditionally, Apocrita are divided into the parasitic and

aculeate wasps (Rasnitsyn 1988), and although Aculeata are

clearly monophyletic, Parasitica include a large number of

lineages whose phylogenetic relationships are largely unknown.

Within the paraphyletic “Parasitica,” Evaniomorpha

are composed of a diverse number of lineages, including

stephanid wasps, ceraphronid wasps, and ensign wasps, and

this group is probably not monophyletic. There are, however,

some well-established groupings within Parasitica, some

of which have undergone formal phylogenetic investigation,

including Cynipoidea, Chalcidoidea, Platygastroidea, and

Ichneumonoidea (Rasnitsyn 1988, Ronquist et al. 1999).

Chalcidoidea include 20 families of very small wasps (0.5–3

mm) that are primarily the parasites of other insects, attacking

chiefly the egg or larval stage of the host. Cynipoidea are

composed of five families of mostly minute wasps that are

primarily gall makers. Ichneumonoidea include three families

of relatively large wasps that are parasitoids of other insects.

All of these groups have a large number of species, and

phylogenetic relationships among most of the constituent

species remain virtually unknown.

Aculeatans are hymenopterans in which the ovipositor

has been modified into a stinger. Aculeata consists of three

major lineages: Chrysidoidea, Vespoidea, and Sphecidae +

Apoidea. Chrysidoidea (cuckoo wasps and allies) include

seven families, and the basic phylogenetic relationships

among these groups are moderately well understood (Carpenter

1999). Vespoidea (ants, vespid wasps, sphecid wasps,

spider wasps, velvet ants, etc.) are a diverse assemblage of

lineages composed of roughly 10 families. Phylogenetic

analysis suggest, among other things, that ants are sister

group to vespid and scoliid wasps and that bees (Apoidea)

evidently arose from a single lineage of sphecid wasps (Brothers

1999, Brothers and Carpenter 1993). Given the minute

size of many hymenopterans, and the vast diversity of this

group as a whole, completing the hymenopteran branch of

the Tree of Life will take many years.

Lepidoptera (Butterflies, Moths, Skippers)

Lepidoptera are a large group of primarily terrestrial insects

characterized by having wings with a dense covering of setae

in the more primitive groups and scales in the more advanced

groups. Although the current estimate of described lepi-

Figure 21.4. Summary phylogeny of bees, wasps, and ants

(Hymenoptera).

Phylogeny of the Holometabolous Insects 353

dopteran species is approximately 150,000, the total number

of extant species may be as high as 500,000, making

Lepidoptera the largest lineage of primarily herbivorous

animals (Kristensen and Skalski 1999). When most people

think of Lepidoptera, they think of two groups: butterflies

and moths. Although the butterflies are certainly the most

popular and well-known lepidopterans, which do indeed

form a monophyletic group, they are only a splash in the

bucket of lepidopteran diversity. The vast majority of lepidopteran

species represent an almost infinite variety of small,

drab moths from multiple evolutionary lineages, and the key

to unraveling the story of lepidopteran evolution lies in deciphering

the phylogeny of moths. Over the last 30 years,

extensive morphological studies of the more primitive Lepidoptera,

and some more recent molecular studies, have led

to a relatively well-established hypothesis of phylogenetic

relationships among the more primitive moth groups (Davis

1986, Krenn and Kristensen 2000, Kristensen and Skalski

1999, Wiegmann et al. 2002). However, phylogenetic relationships

among the more advanced Lepidoptera, and the

more detailed relationships at the family level that include

some of the major species radiations, are still unresolved and

in need of further phylogenetic investigation.

Kristensen (1999) recognized 46 lepidopteran superfamilies

and presented a phylogeny based on a compilation of

morphological data. Although the monophyly of most of

these superfamilies is relatively well established, superfamilial

relationships, particularly among the more derived groups,

are very tentative. Lepidopteran phylogeny can be envisioned

as a comb (fig. 21.5), where a succession of morphological

modifications across a few small groups eventually gave rise

to a body type that allowed the organisms to radiate in bursts

of speciation events. The first three basal lineages (Micropterigoidea,

Agathiphagoidea, Heterobathmioidea) comprise

very primitive moths that have retained mandibles and associated

muscles for chewing, along with an unmodified,

inner pair of lobes (glossa) on the labium, or insect “lower

lip.” These insects are detritivores, feeding primarily on plant

debris in the soil, or are miners, boring into the seeds or leaves

of gymnosperms. The mandibulate moths probably reflect

very closely the morphology of the trichopteran-lepidopteran

ancestor and lack many of the modifications of the more

advanced lepidopterans.

The first major evolutionary innovation in lepidopteran

morphology was the reduction and loss of the chewing mandibles

in the adult insect, which were replaced by extension

and fusion of the inner lobes of the labium to form a coilable,

sucking proboscis typical of most Lepidoptera. This morphological

shift rendered the adults of all higher lepidopterans

dependent exclusively on fluid nutrients, which opened a

new niche that these insects were uniquely suited to exploit.

Hence, a shift from a gymnosperm feeding, mandibulate

moth to that of an angiosperm nectaring, proboscis-bearing

moth allowed higher lepidopterans to diversify concomitantly

with their angiosperm hosts (Kristensen 1997) and is

largely the reason why this is such a diverse and successful

group. More than 99.9% of all lepidopteran species possess

these sucking tubes and collectively are placed in the group

Glossata, named after their possession of the glossa modified

into the all-important proboscis. A proboscis that is

adapted for nectar feeding should be long and flexible and

should have particular sensory equipment allowing for control

of probing movements and the detection of concealed

nectar in elongated corollae (Krenn 1998). The development

of the proboscis did not occur as a single evolutionary event,

however, but a succession of gradual transformations leading

to the refinement in sensory equipment and muscle control

occurred as lepidopterans diversified. The most primitive

glossatans (Eriocranioidea, Acanthopteroctetoidea, and

Lophocoronoidae) have a relatively simple proboscis with

limited movement due to a lack of true intrinsic musculature

(Nielsen and Kristensen 1996). The group Myoglossata

possesses true intrinsic musculature of the proboscis as well

as advanced sensory organs for the more efficient detection

of nectar in flowering plants.

Two other evolutionary changes in morphology have

played a key role in the evolution and diversification of Lepidoptera.

The first was a shift from the forewings and hind

wings being approximately the same size with a similar pattern

of venation (“homoneuran” condition), to a condition

in which the hind wing is smaller than the forewing, and

has certain veins fused together. This latter group is termed

Heteroneura, meaning “different veined.” The myoglossatan,

Figure 21.5. Summary phylogeny of butterflies and moths

(Lepidoptera).

354 The Relationships of Animals: Ecdysozoans

“homoneuran” groups include ghost moths and their allies

(Neopseustoidea, Hepialoidea, and Mnesarchaeoidea). The

second major evolutionary change was a shift from a single

genital pore to a double genital pore in Lepidoptera females.

Primitive Lepidoptera females exhibit the typical insect condition

of having a single genital orifice that is used for copulation

and egg deposition. In the more advanced lepidopterans

(group Ditrysia), there is one orifice for copulation (on the

eighth ventral abdominal segment) and a separate orifice for

egg laying (abdominal segment 9–10), with an internal communication

between sperm receiving and oviduct systems.

The heteroneuran, non-ditrysian groups consist of four major

lineages (Nepticuloidea, Incurvaroidea, Palaephatoidea, and

Tischerioidea), including leaf miners, yucca moths, and fairy

moths, but these groups are sparse in species numbers relative

to Ditrysia. Roughly 98% of all lepidopterans belong to

Ditrysia, and there are no major species radiations before the

development of this unique reproductive system (Kristensen

and Skalski 1999).

Phylogenetic relationships among the ditrysian lineages are

more difficult to ascertain, in large part because of the extensive

modifications in morphology and the explosion of

species numbers. The primitive Ditrysia consist of four lineages

(Tineoidea, Gracillarioidea, Yponomeutoidea, and

Gelechioidea), including clothes moths, bagworms, and diamondback

moths. The more advanced ditrysians (Apoditrysia)

are characterized by the presence of specific modifications

of the endoskeletal structure of the second abdominal segment

(Kristensen and Skalski 1999). Within Apoditrysia

is the group Obtectomera, which is characterized by the

abdominal segments 1–4 being immovable and the wings

being appressed next to the body while in the pupal stage.

The non-obtectomeran, apoditrysian moths consist of eight

lineages, including clearwing moths, carpenter moths, plume

moths, and totrticid moths. Phylogenetic relationships among

these lineages, some of which are very large with more than

10,000 described species, are almost entirely unknown. The

obtectomeran moths can be divided roughly into two groups:

“Microlepidoptera” and Macrolepidoptera. The obtectomeran

microlepidoptera consist of six lineages, the largest of which

includes the pyralids or snout moths, and relationships among

these lineages are unknown, although it is likely that as a whole

these microlepidopterans are not monophyletic. Macrolepidoptera,

as the name indicates, include the large moths and

butterflies that have broad wings and a unique elongation on

a portion of the wing base associated with the hinge (first axillary

sclerite). This group includes the most spectacular lepidopteran

species, including silkworm moths, tiger moths,

geometrid moths, noctuids, skippers, and butterflies. Within

Macrolepidoptera, there are three major radiations (noctuid

moths, geometrid moths, and butterflies with more than

20,000 species each), one moderate-sized radiation (silkworm

lineage and allies), and four relatively minor lineages. One

group, Noctuoidea, has more than 30,000 described species

and represents by far the largest radiation of any lepidopteran

group, and getting a handle on even the basic diversity of this

group is a daunting task. So, although a basic skeletal structure

of lepidopteran phylogeny exists, the real challenge in lepidopteran

systematics for the next century will be to flesh out

the phylogenetic relationships of these diverse groups in more

detail.

Trichoptera (Caddisflies)

Trichoptera are a large group of semi-aquatic insects whose

larvae are found in lakes, streams, and rivers around the

world and form a major component of most freshwater ecosystems.

Trichopteran adults have a mothlike appearance

but with hair rather than scales on the wings, three- to fivesegmented

maxillary palps, and three-segmented labial

palps. As discussed above, a sister-group relationship between

Trichoptera and Lepidoptera is well established, but

trichopterans lack the sucking, tubelike mouthparts characteristic

of Lepidoptera. Like lepidopterans, caddisflies are capable

of spinning silk from specially modified salivary glands,

and the diversity of ways this silk is used probably accounts

for the success of the order as a whole (Mackay and Wiggins

1978). Trichoptera includes approximately 10,000 species

placed within 45 recognized families, and the group is quite

diverse in terms of the aquatic microhabitats and trophic

niches occupied by the species (Morse 1997a).

Phylogenetic relationships within Trichoptera are somewhat

controversial, although ongoing research is providing

new insights on the evolution of this group. Current classifications

recognize three major suborders that are largely

characterized by different ways in which silk is used by the

larvae (fig. 21.6). Annulipalpia (retreat-makers) include nine

families, and these caddisflies make fixed retreats or capture

Figure 21.6. Summary phylogeny of caddisflies (Trichoptera).

Dashed lines represent uncertain relationships.

Phylogeny of the Holometabolous Insects 355

nets under rocks, logs, and other objects in streams, rivers,

lakes, and ponds. All retreat makers possess a ringlike (annulated)

last segment of the maxillary palp. Integripalpia are

the largest group of caddisflies (33 families), and this group

includes species that make mobile, tubelike cases. These

tube-making caddisflies use silk to attach small rocks, sticks,

and other material to form a case that they carry around with

them as they move, and can retract their heads and thorax

inside the case for protection as needed.

Spicipalpia (cocoon-makers) are composed of four families,

including free-living and predaceous caddisflies

(Rhyacophilidae and Hydrobiosidae), caddisflies that make

a small purselike case (Hydroptilidae), and the tortoise-case

and saddle-case caddisflies (Glossosomatidae). Although the

monophyly of both the retreat making group and the tube

making group appears well supported by morphological

(Morse 1997b) and molecular data (Kjer et al. 2002), the

monophyly of the diverse cocoon makers is still debatable.

Previous phylogenetic hypotheses have included all possible

ways of arranging these three groups (Ross 1967, Weaver

1984, Wiggins and Wichard 1989), but the most recent data

suggest that the retreat maker group is the most basal suborder,

with the remaining caddisflies (Spicipalpia and Integripalpia)

forming a monophyletic group (Kjer et al. 2002).

Relationships within retreat-makers are still unclear. Kjer

et al. (2002) recognize four distinct lineages (Stenopsychidae,

Philopotamidae, Hydropsychidae, and the remaining families),

although relationships among these lineages and even

the monophyly of each of these lineages is in need of additional

investigation. As mentioned above, the cocoonmakers

may be paraphyletic, but each of the four families

composing this group is probably monophyletic. There appears

to be two distinct lineages within the tube-case makers:

Plenitentoria (12 families) and Brevitentoria (21 families).

Specific familial relationships within Plenitentoria have been

suggested by Gall (1994), but current molecular data have

not been robust enough to examine this hypothesis in detail.

Brevitentoria may consist of two lineages (Leptoceroidea

and Sericostomatoidea), but again the monophyly of these

groups and relationships within them still require further

investigation (Kjer et al. 2002, Scott 1993, Weaver and Morse

1986).

Mecoptera (Scorpionflies, Hangingflies)

Mecoptera (in the broad, classical sense) are a small but

morphologically diverse insect order with approximately 600

extant described species placed in nine families and 32 genera

(Penny 1997, Penny and Byers 1979). The common name

for this group is derived from the fact that the male 9th abdominal

segment of one family (Panorpidae) is enlarged and

bulbous and curves anterodorsally, resembling the stinger of

a scorpion. This group is not monophyletic because fleas are

sister group to snow scorpionflies (Boreidae), and the nannochoristid

scorpionflies are probably the most basal lineage.

As discussed above, both of these groups are deserving of

ordinal status (fig. 21.7).

Mecoptera include seven families, two of which—Panorpidae

(true scorpionflies) and Bittacidae (hangingflies)—

contain 90% of mecopteran species. The remaining five

families are much less diverse, but they include groups that

exhibit a wide degree of morphological specialization from the

wingless Apteropanorpidae, to the earwig flies (Meropeidae),

to the fossil-like eomeropid scorpionflies. Mecoptera have a

very well documented fossil history and are among the most

conspicuous part of the insect fauna of the Lower Permian.

The monophyly of each mecopteran family is well established

by morphological and molecular data (Byers 1991, Kaltenbach

1978, Whiting 2002a, Willmann 1987).

A number of phylogenetic hypotheses have been presented

for relationships, and each has resulted in somewhat different

conclusions. Kaltenbach (1978) presented Mecoptera subdivided

into three suborders, Protomecoptera (Meropeidae +

Eomeropidae), Neomecoptera (Boreidae), and Eumecoptera

(remaining families), but did not present a specific phylogeny

for these taxa. In a comprehensive analysis of mecopteran

morphology from extinct and extant taxa, Willmann (1987,

1989) presented a phylogeny in which Nannochoristidae are

the basalmost taxon, with Panorpidae + Panorpodidae forming

the most apical clade. This phylogeny was not the result

of a formal quantitative analysis of a coded character matrix,

Figure 21.7. Summary phylogeny of scorpionflies (Mecoptera)

showing the relative positions of fleas (Siphonaptera). The snow

scorpionflies (Boreidae) and nannochoristid scorpionflies are not

members of the true scorpionfly lineage (Mecoptera) but are

given their own ordinal status. Hangingflies (Bittacidae) are either

the sister group to Panorpodidae or at the base of Mecoptera.

356 The Relationships of Animals: Ecdysozoans

but Willmann did provide an explicit explanation of the

characters supporting each node of the phylogeny. Whiting

(2002a) sequenced four genes across multiple representatives

of Mecoptera and performed a preliminary analysis in which

Bittacidae appeared as sister group to Panorpodidae. However,

inclusion of additional data suggests a more basal

placement of Bittacidae and a sister-group relationship between

Panorpidae and Panorpodidae, more in line with

the phylogeny presented by Willmann. The phylogeny of

Mecoptera stands as probably the best-known phylogeny

within Holometabola.

Siphonaptera (Fleas)

Fleas are laterally compressed, wingless insects that possess

mouthparts modified for piercing and sucking. They have

highly modified combs and setae on their body and legs to

help stay attached to their vertebrate hosts, and their hind

legs are modified for jumping. There are approximately 2400

described flea species placed in 15 families and 238 genera

(Lewis and Lewis 1985). Fleas are entirely ectoparasitic, with

~100 species as parasites of birds and the remaining species

as parasites of mammals (Holland 1964). Flea distribution

extends to all continents, including Antarctica, and fleas inhabit

a range of habitats and hosts from equatorial deserts,

through tropical rainforests, to the arctic tundra. Fleas are

of tremendous economic importance as vectors of several

diseases important to human health, including bubonic

plague, murine typhus, and tularemia (Dunnet and Mardon

1991).

From a phylogenetic standpoint, Siphonaptera are perhaps

the most neglected of holometabolous insect orders.

Although we have a reasonable knowledge of flea taxonomy

at the species and subspecific level, and a relatively good

record of their biology and role in disease transmission, phylogenetic

relationships among fleas at any level have remained

virtually unexplored. Classically, the major obstacle in flea

phylogenetics has been their extreme morphological specializations

associated with ectoparasitism, and the inability of

systematists to adequately homologize characters across taxa.

The majority of characters used for species diagnoses are

based on the shape and structure of their extraordinarily

complex genitalia, or the presence and distribution of setae

and spines. Although these characters are adequate for species

diagnoses, they are of limited utility for phylogenetic

reconstruction. There is no generally accepted higher classification

for Siphonaptera, and several classifications published

in recent years have significantly conflicting treatments

of superfamilial relationships (Dunnet and Mardon 1991,

Lewis and Lewis 1985, Mardon 1978, Smit 1979, Traub and

Starcke 1980, Traub et al. 1983).

Molecular data are beginning to provide a more complete

view of flea phylogeny (Whiting 2002a) and Whiting (unpubl.

obs.; fig. 21.8). These data support the monophyly of the families

Certaophyllidae, Ischnopsyllidae (bat fleas), Rhopalopsyllidae,

and Stephanocircidae. The Leptopsyllidae are

paraphyletic, but the superfamilial group Ceratophylloidea

is monophyletic. Pulicidae are paraphyletic, but the subfamilies

that comprise this family (Pulicinae and Tunginae) are

each monophyletic. These data suggest that about half of the

families are paraphyletic (e.g., Chimaeropsyllidae, Hystrichopsyllidae,

Pygiopsyllidae, Leptopsyllidae, Pygiopsyllidae, and

Ctenophthalmidae), although 5 out of 20 subfamilies that

could be assessed with these data are monophyletic. Collectively,

these data suggest that many of the flea families are

artificial assemblages of species, and certain families that have

been used as a catchall for a wide range of divergent taxa (e.g.,

Ctenophthalmidae) are almost certainly paraphyletic groups,

suggesting that family-level revision of this group is warranted.

However, at the subfamily level, the current groupings

more closely reflect phylogenetic relationships. It is still

unclear which flea group is most primitive, and further data

are required to refine current phylogenetic estimates.

Diptera (Flies)

Diptera are a major order of insects with approximately

125,000 species currently described, but the actual number

of extant species is probably at least twice this number.

Dipterans are easily distinguished from other insects by the

modification of the hind wings into organs (halteres) used

for balance during flight. Mouthparts range from lapping to

biting and sucking, and flies have had a tremendous impact

on humans owing to their transmission of deadly diseases

Figure 21.8. Summary phylogeny of fleas (Siphonaptera).

Dashed lines represent uncertain relationships.

Phylogeny of the Holometabolous Insects 357

such as malaria. Higher level phylogenetic relationships

within Diptera have probably received more attention than

those of any other holometabolous insect order, and yet relationships

among the major constituent groups continue to

elude entomologists. The current state of dipteran phylogeny

is outlined in an outstanding recent review by Yeates and

Wiegmann (1999).

Diptera have traditionally been divided into two major

groups (fig. 21.9): long horned (Nematocera, flies with long

antennae) and short horned (Brachycera). Recent research

demonstrates that although the short-horned flies form a

monophyletic group, the long-horned flies are a large assemblage

of ancient lineages, which as a whole are probably not

monophyletic (Yeates and Wiegmann 1999). The long-horned

flies are generally divided into six major groups, but phylogenetic

relationships among these groups are not well resolved.

Ptychopteromorpha contains two families (Tanyderidae and

Ptychopteridae), including primitive and phantom craneflies.

The Culicomorpha are composed of 8 families and contains

all of the blood-sucking primitive flies, including mosquitoes,

black flies, biting midges, and midges. This is a wellsupported

monophyletic group based on features associated

with the modified larval mouthparts used for filter feeding.

Blephariceromorpha include three families, and all of these

midges have specially modified prolegs in larvae for attaching

to the substrate in fast flowing streams. Bibionomorpha

are composed of five families, including march flies, fungus

gnats, and gall midges, but the monophyly of this group

based on morphological characters is questionable. Tipulomorpha

are a large group containing two cranefly families,

and Psychodomorpha contain five families, including moth

flies, sand flies, and wood gnats. The monophyly of both of

these two groups is also questionable.

Brachycera, the short-horned flies, are a well-supported

monophyletic group based on reduction in antenna size,

modifications of the larval head capsule, and specific mouthpart

specializations. This group is composed of four infraorders,

Stratiomyomorpha (soldier and xylomyid flies),

Tabanomorpha (horse flies, snipe flies, and athericid flies),

Xylophagomorpha (xylophagid flies), and Muscomorpha,

which includes the vast majority of fly families. A recent

comprehensive morphological analysis suggests that Tabanomorpha

are sister group to Xylophagomorpha, with Stratiomyomorpha

at its base, and that this group is in turn sister to

Muscomorpha (Yeates 2002). Nemestrinoidea (small headed

and tangle-vein flies) are thought to contain the most basal

members of Muscomorpha, although there is some evidence

that they should be placed within the Tabanomorpha.

Asiloidea are composed of six families, including robber flies,

flower-loving flies, mydas flies, stiletto flies, and bee flies, and

the monophyly of this group is supported by a particular configuration

of spiracles in the larvae. The group Empidoidea,

Figure 21.9. Summary

phylogeny of flies (Diptera).

358 The Relationships of Animals: Ecdysozoans

dance flies and long-legged flies, is sister to Cyclorrhapha, a

large lineage of flies that have a reduced larval head capsule

and feeding structures, and pupation occurs within a specially

formed puparium. Cyclorrhapous Diptera were traditionally

divided into two groups: Schizophora and Aschiza;

however, the latter is not monophyletic but rather a compilation

of at least 10 distinct families assigned to the “lower

Cyclorrhapha.” These include flower flies, big-headed flies,

humpback flies, flat-footed flies, spear-winged flies, and

phylogenetic relationships among these groups are controversial.

Schizophora contain at least 75 families and comprises

the majority of family-level diversity within Diptera.

Schizophoran flies emerge from the puparium by the inflation

of a membranous head sac, the ptilinum. Schizophora

are traditionally divided into two groups: Acalypteratae and

Calypteratae. Acalpyteratae include a wide variety of families,

including thick-headed flies, stilt-legged flies, fruit

flies, picture-winged flies, leaf miner flies, and many others,

and the monophyly of this group is not well established.

Calypterate flies, on the other hand, are a very well-supported

monophyletic group, and these flies have the lower lobe of

the front wing (calypter) well developed. Calypteratae are

composed of three superfamiles: Hippoboscoidea (primarily

ectoparastic flies that are blood feeders), Muscoidea (house

flies, dung flies, and others), and Oestroidea (flesh flies, bot

flies, house flies, tachinid flies). The monophyly of each of

these subgroups appears relatively well supported, but relationships

within each of these subgroups deserve further

scrutiny. In short, there is an obvious need for further

investigation into the relationships of long-horned flies,

primitive short-horned flies, lower Cyclorhappha, and

acalypterate flies.

Strepsiptera (Twisted-Winged Parasites)

Strepsiptera (twisted-winged parasites) are a cosmopolitan

order of small insects (males, 1–7 mm; females, 2–30 mm)

that are obligate insect endoparasites. The order is composed

of ~550 species placed within eight extant and one extinct

family (Kathirithamby 1989). Strepsiptera derive their common

name from the male front wing, which is haltere-like,

and early workers considered it to be twisted in appearance

when dried specimens were examined. All members of this

group spend the majority of their life cycle as internal parasites

of other insects and, consequently, have a highly specialized

morphology, extreme sexual dimorphism, and a

unique biology. The adult male strepsipteran is free-living

and winged, whereas the adult female is entirely parasitic

within the host, with the exception of one family (Mengenillidae)

where the female last larval instar leaves the host to

pupate externally. Strepsiptera parasitize species from seven

insect orders: Zygentoma, Orthoptera, Blattaria, Mantodea,

Hemiptera, Hymenoptera, and Diptera. In one family

(Myrmecolacidae), the males are known to parasitize ant

hosts whereas the females are parasites of Orthoptera. The

life cycle of most strepsipteran species is unknown, and only

a few species have been studied in detail.

The difficulty of placing this group among the other insect

orders was described above. Investigation of phylogenetic

relationships among strepsipteran families has not

received the same attention as the ordinal placement of this

group. Kinzelbach (1971, 1990) used adult morphological

features to investigate this group, but he did not perform a

formal quantitative analysis of these data. Recently, Pohl

(2002) used characteristics of the first instar larvae and standard

analytical techniques to infer phylogenetic relationships.

The phylogeny he produced is somewhat different from that

presented by Kinzelbach, but the overall pattern is the same.

Strepsiptera are divided into two main lineages: the

primitive Mengenillidia and the more advanced Stylopidia

(fig. 21.10). The former lineage includes one extinct and one

living family and is characterized by presence of robust mandibles,

a single genital tube in the female, specific characteristics

associated with a vein in the hind wing (MA1 broad),

and a primitive type of larvae (Pohl 2002). In this group, the

female leaves the host to pupate, in contrast to Stylopidia,

where the female remains within the body of the host during

the pupal and adult stage. Stylopidia can be further

distinguished by the females possessing multiple genital

openings and the hind wing in males with only a remnant

of the MA1 vein. Relationships within the Stylopidia are less

known. Current data suggest that Corioxenidae is the most

primitive family in this group, but further investigation is

necessary to fully resolve relationships among the members

of this unusual insect order.

Future Prospects

Entomologists have long been humbled by the immense size

of Holometabola, and understanding the pattern of diversification

among its constituent lineages has largely eluded

scientific investigation for well more than two centuries. A

clear view of the Holometabola branch of the Tree of Life is

just beginning to emerge. Entomologists are a long way from

exhausting the usefulness of morphological data for reconstructing

holometabolan phylogeny, and for many groups

Figure 21.10. Summary phylogeny of twisted-winged parasites

(Strepsiptera).

Phylogeny of the Holometabolous Insects 359

further investigation of anatomical similarities is bound to

reveal a treasure trove of useful information. The advent of

molecular systematics in the past decade brought with it not

only a new set of tools with which to infer phylogeny, but

also the ability to take a broad-stroke look at Holometabola

in a new way, by selecting a few exemplars from a large range

of diverse groups for molecular screening. Even the best

current efforts in insect molecular systematics will seem

primitive by tomorrow’s standards, and it is clear that, like

morphology, molecular systematics has not yet reached the

pinnacle of usefulness in insects.

Many challenges still remain in unraveling the evolutionary

history these insects: the challenge to catalog the immense

number of species that are members of this group; the challenge

to train a new generation of entomologists in insect

morphology and systematics; the challenge to find novel

genetic markers that better track the phylogeny of these lineages;

and the challenge to overcome the computational limitations

of organizing and analyzing the mountains of data

emerging on insect phylogeny. But for the first time we are

beginning to see a surge of researchers zeroing in on unraveling

the complete phylogenetic structure of Holometabola,

tossing their whole arsenal of tools into the fray and providing

exciting new insights into the most wondrous event in

evolution: the diversification of insects and the evolution of

their most successful group, Holometabola.

Acknowledgments

I thank J. Cracraft and M. Donoghue for the invitation to speak

at the Tree of Life symposium, and M. Terry, H. Ogden, K.

Jarvis, J. Cherry, J. Robertson, and A. Whiting for assistance

with the manuscript. This work was supported by National

Science Foundation grants DEB-9806349 and DEB-9983195.

Literature Cited

Aspцck, U. 2002. Phylogeny of the Neuropterida (Insecta:

Holometabola). Zool. Scr. 31:51–56.

Aspцck, U., J. D. Plant, and H. L. Nemeschkal. 2001. Cladistic

analysis of Neuroptera and their systematic position within

Neuropterida (Insecta: Holometabola: Neuropterida:

Neuroptera). Syst. Entomol. 26:73–86.

Bayreuther, K., and S. Brauning. 1971. Die Cytogenetik

der Flohe (Aphaniptera). Chromosoma (Berl.) 33:19–

29.

Beutel, R. G., and S. N. Gorb. 2001. Ultrastructure of attachment

specializations of Hexapods (Arthropoda): evolutionary

patterns inferred from a revised ordinal phylogeny.

J. Zool Syst. Evol. Res. 39:177–207.

Beutel, R. G., and F. Haas. 2000. Phylogenetic relationships of

the suborders of Coleoptera (Insecta). Cladistics 16:103–

142.

Bilinski, S., J. Bьnnig, and B. Simiczyjew. 1998. The ovaries of

Mecoptera: basic similarities and one exception to the rule.

Fol. Histochem. Cytobiol. 36:189–195.

Boudreaux, H. B. 1979. Arthropod phylogeny with special

reference to insects. John Wiley and Sons, New York.

Brothers, D. J. 1999. Phylogeny and evolution of wasps, ants,

and bees (Hymenoptera, Chrysidoidea, Vespoidea, and

Apoidea). Zool. Scr. 28:233–249.

Brothers, D. J., and J. M. Carpenter. 1993. Phylogeny of

Aculeata: Chrysidoidea and Vespoidea (Hymenoptera).

J. Hymenopt. Res. 2:227–304.

Byers, G. W. 1991. Mecoptera. Pp. 696–704 in Mecoptera (I. D.

Naumann, P. B. Carne, J. F. Lawrence, E. S. Nielsen, J. P.

Spradberry, R. W. Taylor, M. J. Whitten, and M. J. Littlejohn,

eds.). CSIRO, Melbourne University Press, Melbourne.

Byers, G. W. 1996. More on the origin of Siphonaptera. J. Kans.

Entomol. Soc. 69:274–277.

Carmean, D., L. S. Kimsey, and M. L. Berbee. 1992. 18S rDNA

sequences and holometabolous insects. Mol. Phylogenet.

Evol. 1:270–278.

Carpenter, J. M. 1999. What do we know about chrysidoid

(Hymenoptera) relationships? Zool. Scr. 28:215–231.

Caterino, M. S., V. L. Shull, P. M. Hammond, and A. P. Vogler.

2002. Basal relationships of Coleoptera inferred from 18S

rDNA sequences. Zool. Scr. 31:41–49.

Chalwatzis, N., J. Hauf, Y. V. Peer, R. Kinzelbach, and F. K.

Zimmerman. 1996. 18S ribosomal RNA genes of insects:

primary structure of the genes and molecular phylogeny of

the Holometabola. Ann. Entomol. Soc. 89:788–803.

Crowson, R. A. 1960. The phylogeny of Coleoptera. Annu. Rev.

Entomol. 5:111–134.

Davis, D. R. 1986. A new family of montrysian moths from

austral South America (Lepidoptera: Palaephatidae), with a

phylogenetic review of the Monotrysia. Smithson. Contrib.

Zool. 434:1–202.

Dowton, M., and A. D. Austin. 1999. Models of analysis

for molecular datasets for the reconstruction of basal

hymneopteran relationships. Zool. Scr. 28:69–74.

Dunnet, G. M., and D. K. Mardon. 1991. Siphonaptera.

Pp. 125–140 in Siphonaptera (P. B. C. I. D. Naumann, J. F.

Lawrence, E. S. Nielsen, J. P. Spradberry, R. W. Taylor,

M. J. Whitten, and M. J. Littlejohn, eds.). CSIRO and

Melbourne University Press, Melbourne.

Farrell, B. D. 1998. “Inordinate fondness” explained: why are

there so many beetles? Science 281:555–558.

Gall, W. K. 1994. Phylogenetic studies in the Limnephiloidea,

with a revision of the world genera of Goeridae (Trichoptera).

Phylogenetic studies in the Limnephiloidea, with a revision of

the world genera of Goeridae (Trichoptera). University of

Toronto, Toronto.

Hammond, P. 1992. Species inventory. Pp. 21–25 in Species

inventory (B. Groombridge, ed.). Chapman and Hall,

London.

Hennig, W. 1981. Insect phylogeny. Academic Press, New York.

Hinton, H. E. 1958. The phylogeny of the panorpoid orders.

Annu. Rev. Entomol. 3:181–206.

Holland, G. P. 1964. Evolution, classification, and host

relationships of Siphonaptera. Annu. Rev. Entomol. 9:123–

146.

Hцrnschemeyer, T. 2002. Phylogenetic signficance of the wingbase

of the Holometabola (Insecta). Zool. Scr. 31:17–30.

Huelsenbeck, J. P. 1997. Is the Felsenstein zone a fly trap? Syst.

Biol. 46:69–74.

360 The Relationships of Animals: Ecdysozoans

Kaltenbach, A. 1978. Mecoptera (Schnabelhafte, Schnabelfliegen).

Mecoptera (Schnabelhafte, Schnabelfliegen). Walter de

Gruyter, Berlin, New York.

Kathirithamby, J. 1989. Review of the order Strepsiptera. Syst.

Entomol. 14:41–92.

Kinzelbach, R. K. 1971. Morphologische Befunde and

Facherfluglern und ihre phylogenetische bedeutung

(Insecta: Strepsiptera). Zool. 119:1–256.

Kinzelbach, R. K. 1990. The systematic position of Strepsiptera

(Insecta). Am. Entomol. 36:292–303.

Kjer, K., R. J. Blahnik, and R. W. Holzenthal. 2002. Phylogeny

of caddisflies (Insecta, Trichoptera). Zool. Scr. 31:83–91.

Krenn, H. W. 1998. Proboscis sensilla in Vanessa cardui

(Nymphalidae: Lepidoptera): functional morphology and

significance in flower-probing. Zoomorphology 118:23–30.

Krenn, H. W., and N. P. Kristensen. 2000. Early evolution of

the proboscis of Lepidoptera (Insecta): external morphology

of the galea in basal glossatan moths lineages, with remarks

on the origin of pilifers. Zool. Anz. 239:179–196.

Kristensen, N. P. 1975. The phylogeny of hexapod “orders.” A

critical review of recent accounts. Z. Zool. Syst. Evol.

Forsch. 13:1–44.

Kristensen, N. P. 1981. Phylogeny of insect orders. Annu. Rev.

Entomol. 26:135–157.

Kristensen, N. P. 1991. Phylogeny of extant hexapods. Pp. 125–

140 in Phylogeny of extant hexapods (P. B. C. I. D.

Naumann, J. F. Lawrence, E. S. Nielsen, J. P. Spradberry,

R. W. Taylor, M. J. Whitten, and M. J. Littlejohn, eds.).

CSIRO, Melbourne University Press, Melbourne.

Kristensen, N. P. 1995. Fourty [sic] years’ insect phylogenetic

systematics. Zool. Beitr. N. F. 36:83–124.

Kristensen, N. P. 1997. Early evolution of the Lepidoptera +

Trichoptera lineage: phylogeny and the ecological scenario.

Mem. Mus. Natl. Hist. Nat. 173:253–271.

Kristensen, N. P. 1999. Phylogeny of endopterygote insects, the

most successful lineage of living organisms. Eur. J. Entomol.

96:237–253.

Kristensen, N. P., and A. W. Skalski. 1999. Phylogeny and

paleontology. Pp. 7–25 in Phylogeny and paleontology

(N. P. Kristensen, ed.). De Gruyter, New York.

Kukalova-Peck, J., and J. F. Lawrence. 1993. Evolution of the

hind wing in Coleoptera. Can. Entomol. 125:181–258.

Lawrence, J. F., and A. F. J. Newton. 1982. Evolution and

classification of beetles. Annu. Rev. Ecol. Syst. 13:261–290.

Lawrence, J. F., and A. F. J. Newton. 1995. Families and

subfamilies of Coleoptera (with selected genera, notes,

references and data on family-group names). Pp. 779–1006

in Families and subfamilies of Coleoptera (with selected

genera, notes, references and data on family-group names)

(J. Pakaluk and S. A. Slipinski, eds.). Muzeum i Instytut

Zoologii, PAN, Warszawa.

Lawrence, J. F., S. A. Slipinski, and J. Pakaluk. 1995. From

Latreille to Crowson: a history of the higher-level classification

of beetles. Pp. 87–154 in From Latreille to Crowson: a

history of the higher-level classification of beetles

( J. Pakaluk and S. A. Slipinski, eds.). Muzeum i Instytut

Zoologii, PAN, Warszawa.

Lewis, R. E., and J. H. Lewis. 1985. Notes on the geographical

distribution and host preferences in the order Siphonaptera.

J. Med. Entomol. 22:134–152.

Mackay, R. J., and G. B. Wiggins. 1978. Ecological diversity in

the Trichoptera. Annu. Rev. Entomol. 24:185–208.

Mardon, D. K. 1978. On the relationships, classification,

aedeagal morphology and zoogeography of the genera of

Pygiopsyllidae (Insecta : Siphonaptera). Aust. J. Zool. Suppl.

Ser. 64:1–69.

Mickoleit, G. 1973. Uber den ovipositor der Neuropteroidea

und Coleoptera und seine phylogenetische Bedeutung

(Insecta, Holometabola). Z. Morphol. Tiere 74:37–64.

Morse, J. C. 1997a. Checklist of the world Trichoptera.

Pp. 339–342 in Checklist of the world Trichoptera

(R. W. Holzenthal and O. S. Flint, Jr., eds.). Ohio

Biological Survey, Columbus.

Morse, J. C. 1997b. Phylogeny of Trichoptera. Annu. Rev.

Entomol. 42:427–450.

Nielsen, E. S., and N. P. Kristensen. 1996. The Australian moth

family Lophocoronidae and the basal phylogeny of the

Lepidoptera–Glossata. Invert. Taxon. 10:1199–302.

Pashley, D. P., B. A. McPheron, and E. A. Zimmer. 1993.

Systematics of holometabolous insect orders based on 18S

ribosomal RNA. Mol. Phylogenet. Evol. 2:132–142.

Penny, N. D. 1997. World checklist of extant Mecoptera

species. World checklist of extant Mecoptera species.

California Academy of Sciences, San Francisco, CA.

Penny, N. D., and G. W. Byers. 1979. A check-list of the

Mecoptera of the World. Acta Amazon. 9:365–388.

Pix, W., G. Nalbach, and J. Zeil. 1993. Strepsipteran forewings

are haltere-like organs of equilibrium. Naturwissenschaften

80:371–374.

Pohl, H. 2002. Phylogeny of the Strepsiptera based on morphological

data of the first instar. Zool. Scr. 31:123–134.

Rasnitsyn, A. P. 1988. An outline of evolution of the hymenopterous

insects. Orient. Insects 22:115–145.

Ronquist, F. 1999. Phylogeny of the Hymenoptera (Insecta): the

state of the art. Zool. Scr. 28:3–11.

Ronquist, F., A. P. Rasnitsyn, A. Roy, K. Eriksson, and

M. Lindgren. 1999. Phylogeny of the Hymenoptera: a

cladistic reanalysis of Rasnitsyn’s (1998) data. Zool. Scr.

28:139–164.

Ross, H. H. 1967. The evolution and past dispersal of the

Trichoptera. Annu. Rev. Entomol. 12:169–206.

Schlein, Y. 1980. Morphological similarities between the skeletal

structures of Siphonaptera and Mecoptera. Pp. 359–367 in

Proceedings of the International Conference on Fleas (R.

Traub and H. Starcke, eds.). A. A. Balkema, Rotterdam.

Schulmeister, S., W. C. Wheeler, and J. M. Carpenter. 2002.

Simulatenous analysis of the basal lineages of Hymenoptera

(Insecta) using sensitivity analysis. Cladistics 18:455–484.

Scott, K. M. F. 1993. Three recently erected Trichoptera families

from South Africa, the Hydrosalpingidae, Petrothrincidae

and Barbarochthonidae (Integripalpia: Serocostomatoidea),

with a cladistic analysis of character states in the twelve

families here considered as belonging to the Sericostomatoidea,

by F.C. de Moor. Ann. Cape Prov. Mus. 18:293–

354.

Shull, V. L., A. P. Vogler, M. D. Baker, D. R. Maddison, and

P. M. Hammond. 2001. Sequence alignment of 18S

ribosomal RNA and the basal relationships of adephagan

beetles: evidence for the monophyly of aquatic families and

the placement of Trachypachidae. Syst. Biol. 50:945–969.

Phylogeny of the Holometabolous Insects 361

Sidall, M. E., and M. F. Whiting. 1999. Long-branch abstractions.

Cladistics 15:9–24.

Simiczyjew, B. 2002. Structure of the ovary in Nannochorista

neotropica Navas (Insecta: Mecoptera: Nannochorsittdae)

with remarks on mecopteran phylogeny. Acta Zool. 83:61–

66.

Smit, F. G. A. M. 1979. The fleas of New Zealand (Siphonaptera).

J. R. Soc. N. Z. 9:143–232.

Stys, P., and S. Bilinksy. 1990. Ovariole types and the

phylogeny of hexapods. Biol. Rev. Cambr. Philos. Soc.

65:401–429.

Traub, R., and H. Starcke, eds. 1980. Fleas: Proceedings of

the International Conference on Fleas. A. A. Balkema,

Rotterdam.

Traub, R. M., M. Rothschild, and J. Haddow. 1983. The

Ceratophyllidae, key to the genera and host relationships.

Academic Press, New York.

Vilhelmsen, L. 1997. The phylogeny of lower Hymenoptera

(Insecta), with a summary of the early evolutionary history

of the order. J. Zool. Syst. Evol. Res. 35:49–70.

Weaver, J. S., III. 1984. The evolution and classification of

Trichoptera, Part 1: the groundplan of Trichoptera.

Pp. 413–419 in The evolution and classification of

Trichoptera, Pt 1: The groundplan of Trichoptera

( J. C. Morse, ed.). Dr. W. Junk, The Hague.

Weaver, J. S., III, and J. C. Morse. 1986. Evolution of feeding

and case-making behavior in Trichoptera. J. N. A. Benthol.

Soc. 5:150–158.

Wheeler, W. C., M. F. Whiting, Q. D. Wheeler, and J. M.

Carpenter. 2001. The phylogeny of the extant hexapod

orders. Cladistics 17:113–169.

Whiting, M. F. 1998a. Long-branch distraction and the

Strepsiptera. Syst. Biol. 47:134–138.

Whiting, M. F. 1998b. Phylogenetic position of the Strepsiptera:

review of molecular and morphological evidence. Int. J.

Morphol. Embryol. 27:53–60.

Whiting, M. F. 2002a. Mecoptera is paraphyletic: multiple genes

and phylogeny of Mecoptera and Siphonaptera. Zool. Scr.

31:93–104.

Whiting, M. F. 2002b. Phylogeny of the holometabolous insect

orders based on 18S ribosomal data: when bad things

happen to good data. Pp. 69–84 in Phylogeny of the

holometabolous insect orders based on 18S ribosomal data:

when bad things happen to good data (R. DeSalle, W. C.

Wheeler and G. Giribet, eds.). Birkhauser, Basel.

Whiting, M. F. 2002c. Phylogeny of the holometabolous insect

orders: molecular evidence. Zool. Scr. 31:3–16.

Whiting, M. F., J. C. Carpenter, Q. D. Wheeler, and W. C.

Wheeler. 1997. The Strepsiptera problem: phylogeny of the

Holometabolous insect orders inferred from 18S and 28S

ribosomal DNA sequences and morphology. Syst. Biol.

46:1–68.

Whiting, M. F., and W. C. Wheeler. 1994. Insect homeotic

transformation. Nature 368:696.

Wiegmann, B. W., J. C. Regier, and C. Mitter. 2002. Combined

molecular and morphological evidence on the phylogeny of

the earliest lepidopteran lineages. Zool. Scr. 31:67–81.

Wiggins, G. B., and W. Wichard. 1989. Phylogeny of pupation in

Trichoptera, with proposals on the origin and higher classification

of the order. J. N. A. Benthol. Soc. 8:260–276.

Willmann, R. 1987. The phylogenetic system of the Mecoptera.

Syst. Entomol. 12:519–524.

Willmann, R. 1989. Evolution und Phylogenetisches System der

Mecoptera (Insecta: Holometabola). Abh. Senckenb.

Naturforsch. Ges. 544:1–153.

Wilson, E. O. 1988. The current state of biological diversity.

Pp. 3–18 in The current state of biological diversity (E. O.

Wilson, ed.). National Academy Press, Washington, DC.

Yeates, D. K. 2002. Relationships of extant lower Brachycera

(Diptera): a quantitative synthesis of morphological

characters. Zool. Scr. 31:105–121.

Yeates, D. K., and B. M. Wiegmann. 1999. Congruence and

controversy: toward a higher-level phylogeny of the Diptera.

Annu. Rev. Entomol. 44:397–428.

This page intentionally left blank

VIII

The Relationships of Animals: Deuterostomes

This page intentionally left blank

Навигация

• Главная
• Книги
• Статьи
• Обратная связь
• Поиск