The Invasion Of Land

Insects are principally terrestrial organisms. Indeed, despite frequent colonization of freshwaters by mayflies, dragonflies, diving beetles, predatory water bugs, various midges, and other groups, they are all terrestrial organisms by original design. The transition to land took place in the ancestor of insects and their closest relatives, the Entognatha. The freshwater life-histories of immature mayflies (Ephemeroptera), dragonflies and damselflies (Odonata), and many other insects evolved later. The occurrence of marine, stem-group Hexapoda suggests that the invasion of land occurred independently in the Myriapoda and Hexapoda (Figure 3.23).

3.26. A representative pauropod, showing the eyeless head (above, right), the distinctively branched antennae, plumose setae, and the terminal trunk segments (below, right). Not to same scale, body length 1.1 mm.

Terrestrialization also occurred independently in the Crustacea (Isopoda), Cheliceriformes (Chelicerata), Tardigrada, and Onychophora (Euonychophora). So when did all of these groups depart from the waters and first explore the terrestrial biosphere? The earliest assemblages of terrestrial arthropod fossils are from the Late Silurian (Jeram et al., 1990). However, fossilized trackways of arthropods on land are known from the Early to mid-Ordovician (Sharpe, 1932; Johnson et al., 1994; MacNaughton et al., 2002), tens of millions of years ear lier. These fossilized tracks are not of insects but instead appear to be early cheliceriforms. Indeed, the earliest evidence of insects is from the Early Devonian. Early terrestrial tracks document the presence of various arthropod lineages and support the view that insects themselves originated in a terrestrial environment. It is interesting to note that the arthropods comprised the earliest known terrestrial animals. The fact that these early land animals were all predatory indicates that the selective pressure for terrestrial living was perhaps not an herbivorous diet of land plants. Instead, early, amphibious arthropods may have ventured on to land as part of their reproductive cycle; they may have sought temporary refuge on land from predators lurking in coastal waters or to feed upon worms and other animals feeding on microbial and algal mats growing at the water's edge.


The epiclass Hexapoda consists of the entognathous hexa-pods and the true insects (Table 3.3). The group is supported by the fusion of the second maxillae to form a labium (convergent with Symphyla); the loss of an articulating endite on the mandible; fixation, at least primitively, of the number of abdominal segments to 11; and loss of jointed abdominal appendages (Kristensen, 1991). Early, nonterrestrial "hexa-podous" arthropods are known from the earliest Devonian of Germany (Bartels, 1995; Briggs and Bartels, 2001; Haas et al., 2003). These fossils lack the true hexapod condition of 11 abdominal segments (or less) and loss of appendages on the tenth abdominal segment. Instead, these fascinating marine organisms have loosely differentiated thoracic and abdominal tagma, with numerous abdominal segments bearing appendages (Figure 3.27). Like true Hexapoda, they show a reduction to a single pair of antennae (although preservation in some of these fossils is a bit ambiguous), as well as well-developed appendages on the "thoracic" segments in comparison to the trunk. Rather than include such stem-group marine forms into the Hexapoda, we prefer to consider them as members of superclass Panhexapoda, a larger clade containing Hexapoda (i.e., Entognatha and Insecta), and these stem-lineage, hexapodous, marine organisms (Figure 3.23; Table 3.2). Certainly exploration of the Devonian and latest Silurian, both terrestrial and marine, will give us our most profound insights into the origination and differentiation of the hexapods.

The first major dichotomy among hexapods is the division into Entognatha and Ectognatha, the latter more widely known as the Insecta. These two divisions were recognized as early as 1888 by Grassi, but their defining features were best established by Hennig (1953, 1969, 1981).


Three orders (each sometimes given the rank of class) are included in this group: the Collembola (springtails), the Pro-tura (proturans), and the Diplura (diplurans). All have generally edaphic lifestyles and, except for Collembola, are not widely encountered. As implied by the name, the principal feature of this group is the development of entognathy, in which the mouthpart appendages are recessed within a

3.27. A reconstruction of the Early Devonian marine panhexapod, Devonohexapodus bocksbergensis. Devonohexapodus and other marine panhexapods are stem groups to terrestrial hexapods (Entognatha and Insecta). Redrawn from Haas et al. (2003).

TABLE 3.3. Hierarchical Classification of Epiclass Hexapoda

Epiclass HEXAPODA Class Entognatha Order Diplura Ellipura Order Protura Order Collembola Class Insecta (= Ectognatha)

gnathal pouch on the head capsule. More primitive lineages (e.g., Symphyla, Diplopoda) as well as other hexapods all have ectognathous mouthparts. Entognathy is also unique in that during embryogenesis lateral folds of the head (called plica orales) form over the buds of the mouthparts (Tuxen, 1959). The plica orales grow downward to fuse with the base of the labium (at the postmentum). The result is the formation of the gnathal pouch that entirely encloses the mandibles and maxillae. The jaws lie essentially horizontal, and the mandibles are long and narrow, with their mono-condylic (i.e., singly articulated) bases near the back of the head and often sunken into the posterior wall of the gnathal pouch. Throughout hexapods the maxillae are protrusible and retractible, owing to the articulations between the cardo and stipes, and entognaths have the derived feature of protrusible/retractible mandibles as well. This movement is achieved by a set of dorsal muscles that originate on the head capsule (Tuxen, 1959). As can be imagined, the tentorium has been radically and uniquely rearranged so as to accommodate this distinctive style of mouthparts. Without question, entognathan mouthparts are highly specialized and derived and partly define the monophyly of Entognatha. Other defining features of the group include the reduced or completely absent compound eyes (although this may be convergent owing to similar edaphic lifestyles), reduced Malpighian tubules, and elongate, saclike ovarioles (except Japygidae, which is convergently more similar to Insecta).

Within the Entognatha the Collembola appear to be more closely related to the Protura and are sometimes together referred to as Class Ellipura (Börner, 1910). Defining features of the Ellipura include the absence of cerci and the presence of simple papillae in place of Malpighian tubules, paired ovarioles developed as elongate sacs, and a linea ventralis (Tuxen, 1958, 1959). The linea ventralis is a longitudinal groove that runs along the middle of the ventral part of the body, which has lateral crests and extends from the opening of the labial glands caudad onto the neck membrane in Pro-tura and to the preabdominal tube in Collembola. They are further characterized by unsegmented (i.e., monomeric) tarsi and simple claws.

Although the Diplura are similarly entognathous, this order has at times been placed as sister to the Insecta, thereby leaving Entognatha paraphyletic (e.g., Kukalova-Peck, 1987, 1991; Koch, 1997; Kraus, 1998). A relationship between Diplura and Insecta (i.e., the "Euinsecta"), however, is only weakly supported, and recent morphological and molecular studies have further recognized a monophyletic Entognatha (e.g., Bitsch and Bitsch, 1998, 2000; Carapelli et al., 1998; Frati et al., 1998; Wheeler et al., 2001; D'Haese, 2002). Those who support a Diplura + Insecta relationship base their argument on the divided ovarioles, epimorphic development, and paired claws common to both groups. Furthermore, such authors cite slight differences in the development of entognathy in Diplura relative to Ellipura, whereby the plica orales extend to the labium but remain differentiated from it by a longitudinal sulcus (e.g., Ikeda and Machida, 1998). Frequently a small sclerite, called the admentum, forms between the prementum and the plica orales, a further difference between Ellipura and Diplura. However, such differences likely merely represent unique features in diplurans as an elaboration upon the standard entognathous condition and not independent derivations of entognathy.

Fertilization is indirect in the Entognatha, and as such the genitalic structure is impressively simple, consisting externally merely of gonopores used for either depositing a spermatophore or receiving one. This feature has not been considered of general phylogenetic importance, though the simplified gonopore may represent a further defining feature of this group. Perhaps the most interesting aspect of the Devonian marine panhexapods comes from the structure of the genital appendages described for Devonohexapodus bocksbergensis (Haas et al., 2003). A pair of abdominal segments near the apex of this tagma are similar to primitive ectognathan genital segments, showing apparent gonapophyses; thus they are similar to a primitive ovipositor that in Insecta appears on the eighth and ninth abdominal segments. If such structures truly existed in basal lineages of Panhexapoda rather than being derived at the origin of Hexa-poda or Insecta, then the formation of a rudimentary ovipositor is phylogenetically more primitive than once believed. Furthermore, it suggests that the complete loss of genital appendages in the Entognatha is a secondary reduction and, thus, is a derived trait uniting these lineages rather than being a vestige from a more distant ancestry.


The Protura are rarely encountered, minute hexapods that are overall rather simple in their morphology (Figure 3.28). Approximately 500 species of Protura are distributed across all zoogeographic regions. Tuxen (1963) explored the relationships among the genera recognized at that time and monographed the world species the following year (Tuxen, 1964). Numerous species have been subsequently added (e.g., Tuxen, 1967b), and a new revision of the order is needed. More recent cladistic studies have focused on the suprageneric classification of proturans (e.g., Yin, 1983, 1984), but have not been widely followed; see also Francois (2003). Perhaps the premier feature of the order is that, while hexapodous, the proturans are functionally tetrapods. The anterior legs are directed forward and are not used in locomotion; they are instead lifted above the ground to function as sensory appendages and are, in fact, covered with sensory structures. The types and distributions of these sensory sen-sillae are of taxonomic importance in the group. Other defining features of the Protura include 12 abdominal segments; no antennae; rudimentary appendages on the first three

3.28. A proturan. The forelegs are not used in walking but are modified into sense organs, functioning like antennae.

abdominal segments; eversible vesicles at the apices of the abdominal appendages; the gonopore positioned on the eleventh abdominal segment; a transverse sclerite (sometimes considered vestigial and fused gonocoxae) in the genital chamber; a pair of lateral, genital plates (sometimes considered parameres) in the male; reduced deutocerebrum of the brain; partial fusion of the ganglia in the ventral nerve cord; and no peritrophic membrane in the gut. These are highly modified arthropods. As in Collembola the cerci are lacking, but proturans do not molt after sexual maturity (springtails, diplurans, and primitive insects do). In addition, developing Protura add segments between molts, starting with nine abdominal segments and progressing to the full compliment of 12 as in myriapods and possibly retained from an atelocer-atan ancestor.

The order is currently divided into two superfamilies: the Eosentomoidea and the Acerentomoidea, each with two families (although see Yin, 1983, 1984, for an alternative familial classification). The superfamilies differ from each other by the occurrence of spiracles, tracheae, and a striate band on the eighth abdominal segment. Eosentomoidea lack the striate band, while possessing tracheae and spiracles; Acerentomoidea have just the opposite. The features defining Acerentomoidea are notable, derived characters and justify its recognition as a monophyletic group. However, the eosen-tomoids are based solely on the absence of acerentomoid synapomorphies and are certainly paraphyletic (although Yin, 1983, 1984, suggests the complete opposite).

Proturans occur in moss, rotting wood, soil, and leaf litter, where they are believed to feed on mycorrhizal fungi. The biology of proturans is poorly understood, and fossils of the order are entirely lacking. Indeed, their minute, soft bodies would not easily fossilize. Amber preservation would be ideal, but proturans are not arboreal and would therefore not readily encounter resin.

Collembola: The Springtails

The most familiar of all the entognaths are understandably the springtails (Figures 3.29, 3.30), which are also the most diverse and commonly encountered lineage of Entognatha with about 6,000 species. Collembola live in diverse habitats worldwide; from caves, to alongside fresh or marine waters, to soil and decomposing vegetation. Most species feed on fungal matter, decomposing debris, and fecal material of other invertebrates or will prey on microorganisms. A few species feed on fresh plant material. The most recent major account of the Collembola is that of Hopkin (1997) and the phylogenetic studies of Lee et al. (1995) and D'Haese (2002). Christiansen and Bellinger (1998) have treated the North American fauna; Greenslade (1994), the Australian fauna; and Mari-Mutt and Bellinger (1990), the Neotropical fauna.

The order is universally supported as monophyletic and is easily characterized by the reduction of the abdomen to six segments (although owing to partial fusion it sometimes appears to have even fewer). They also have short, typically four-segmented antennae (the fourth segment is sometimes subsegmented); thoracic sterna divided into lateral basister-nites by the linea ventralis; legs with tibiae and monomeric tarsi fused to form a tibiotarsus; a pair of eversible vesicles at the apex of a ventral tube (called the collophore) on the first abdominal segment (Figure 3.29); and the location of the gonopore on the fifth abdominal segment. Despite this impressive suite of derived traits, the hallmark character of the Collembola is their "spring." On the third and fourth abdominal segments are interlocking structures that form a spring mechanism, allowing the springtails to propel themselves into the air. Although hardly a form of controlled flight, the spring is an effective means of escaping predation. The

3.29. Scanning electron micrographs depicting typical features of springtails (Collembola) based on species of Poduridae. In the center at top is the opening to the gnathal pouch, in which the mouthparts reside (as in all Entognatha). The other central images depict defining features of the order: the collophore and the "spring," the latter formed of the furculum and the retinaculum. Not to same scale.

3.29. Scanning electron micrographs depicting typical features of springtails (Collembola) based on species of Poduridae. In the center at top is the opening to the gnathal pouch, in which the mouthparts reside (as in all Entognatha). The other central images depict defining features of the order: the collophore and the "spring," the latter formed of the furculum and the retinaculum. Not to same scale.

actual moving portion of the spring is the furculum, formed from fused abdominal appendages on the fourth abdominal segment (Figure 3.29). The furculum is ventrally located and has a broad base called the manubrium that bears paired, frequently elongate, finger-like processes at its apex called the dens (themselves sometimes bearing small processes at their own apices called mucrones). The furculum can recline into a small "lock," the retinaculum, which is located on the third abdominal sternum. Some lineages have lost the spring. Other features of the order include the presence of compound eyes (although these are lost in some families) and the absence of tracheae except in the suborder Symphypleona, which have a single pair of spiracles in the collar and a rudimentary tracheal system.

Springtails are presently classified into three suborders (Arthropleona, Neelipleona, and Symphypleona) (Figure 3.30),

3.30. Representative springtails (Collembola). Scanning electron micrographs; not to same scale.

one of which is definitively paraphyletic. The Arthropleona is a paraphyletic assemblage of families from which the Sym-phypleona and Neelipleona are derived, the latter itself likely derived from among the symphypleones (perhaps allied to the Sminthuridae). Arthropleona have primitively elongate bodies with relatively complete abdominal segmentation and the mouth typically opening anterior to the ocelli (i.e., the heads are prognathous, or with the mouthparts held forward). By constrast, the Neelipleona and Symphypleona

(comprising the Neopleona) have globular bodies with the first four abdominal segments fused, sometimes also fused with the meso- and metathoracic segments. In addition, the mouth typically opens ventral to the ocelli (i.e., the heads are hypognathous). Neelipleona is poorly understood and consists of about 25 minute species in a single family (Neelidae), which are blind and live in caves or in the soil. The neeli-pleones differ from the Symphypleona, from which they are certainly derived, by the short antennae (being shorter than

Entognathous Mouthparts
3.31. Reconstruction of Rhyniella praecursor, the earliest fossil of Entognatha, from the Early Devonian chert of Rhynie, Scotland. The entognathous mouthparts are well preserved, and the remains of a col-lophore and furculum indicate it was a collembolan.

the head), absence of ocelli, absence of bothriotrichia, and presence of sensory regions on the abdomen. Conversely, the symphypleones have long antennae, ocelli, bothriotrichia, while lacking the sensory regions on the abdomen, all primitive traits relative to neelids.

By stark contrast to Protura and Diplura, the springtails have an extensive fossil record. Indeed, one of the oldest hexapods is a springtail. Rhyniella praecursor from the Early Devonian (Pragian) Rhynie Chert of Scotland is a rather typical collembolan (Hirst and Maulik, 1926; Tillyard, 1928b; Scourfield, 1940a,b; Massoud, 1967; Whalley and Jarzem-bowski, 1981; Greenslade and Whalley, 1986) (Figure 3.31). Although at one time placed in its own family (e.g., Paclt, 1956), it has since been recognized as being most similar to the arthropleone family Isotomidae (perhaps the most basal of all collembolan families) (Greenslade and Whalley, 1986). Unfortunately, there is a gap in the fossil record of the order of nearly 300 MY. The next oldest springtails are in ambers from the Cretaceous (Christiansen and Pike, 2002a,b; Simon-

Benito et al., 2002) and are also quite common in Cenozoic ambers (Christiansen, 1971; Mari-Mutt, 1983; Lawrence, 1985), which are mostly represented by Arthropleona but also include Symphypleona. Neelipleona are unknown in the fossil record. While several fossil species have been described, the phylogenetic implications of these taxa have not yet been explored.


The diplurans consist of two groups of rather divergent lineages: the suborders Campodeomorpha and Japygomorpha. The campodeomorphs (Figure 3.32) have multisegmented cerci and a movable, mandibular prostheca, which is a process near the molar surface developed either as a sclerite or fringe. The japygomorphs lack the mandibular prostheca and have unsegmented, forcipate cerci, similar to the cercal forceps of earwigs (Figure 3.33). Species of both lineages live in soil, rotting wood, or leaf litter but otherwise differ in their biology. The campodeomorphs are generally not aggressive and are mostly herbivorous. Japygomorphs are fiercely predatory, principally victimizing small insects and other invertebrates, subduing them by grasping them with their maxillae or their impressive cercal forceps. The genus Hetero-japyx is particularly interesting because some species behave like antlions, burying themselves head-down into the soil with only the apices of the forceps extending above the ground. Once an unsuspecting insect approaches, the Het-erojapyx seizes the prey with its forceps, emerges from the soil, and consumes its victim.

Like all entognaths, diplurans have external fertilization. Females deposit eggs in small clumps within rotting wood, vegetation, or cracks in the soil surface. Interestingly, diplu-rans can be subsocial, with females guarding their eggs and immatures for several molts, just as in many earwigs. However, this maternal devotion can sometimes lead to unfortunate consequences as japygomorphs are at times cannabilistic, with the young devouring their mother when they grow. Development, in contrast to other entognaths, is epimorphic, with relatively little change in postembryonic stages aside from the number of antennal segments or alternations in chaetotaxy. However, like all primitive hexapods, molting continues after adulthood, with up to 30 molts recorded for some Campodea.

Dipluran monophyly has not been robustly supported in the past but has been consistently recovered by rigorous and recent studies of basal hexapods (e.g., Bitsch and Bitsch, 2000). Aside from the unique form of entognathy previously discussed, all Diplura have a monocondylic articulation between the trochanter and femur and between the femur and tibia (these articulations are dicondylic in almost all other Hexapoda). Additional features of the order include the absence of eyes (both ocelli and compound eyes) and the presence of panoistic ovarioles (except campodeomorphs,

3.32. A campodeid dipluran (Entognatha), with the sternal styli and eversible vesicles indicated. Scanning electron micrograph, length 2.2 mm.

which are more like that of Ellipura), similar to primitive insects. Molecular studies concentrating on basal hexapods (e.g., Carapelli et al., 1998; Frati et al., 1998) have also found a monophyletic Diplura (in addition to a monophyletic Entognatha). The presence of paired pretarsal claws in Diplura is tantalizingly similar to the same condition seen in insects. Similarly, most diplurans have the gonopore recessed into a pouch between the eighth and ninth abdominal segments (the same position of the insectan gonopore), although some taxa have it developed between the seventh and eighth abdominal segments. The antennae are long, moniliform, and multisegmented. Major classifications of the order have been provided by Pages (1997), although his system has proven to be rather unstable (e.g., Bitsch and Bitsch, 2000), and the more conservative classifications of Paclt (1957) and Conde and Pages (1991), which are more widely employed.

The fossil record of diplurans is exceptionally poor given that they presumably evolved in the Early Devonian judging

3.33. The cercal forceps of a japygid dipluran are similar to those in earwigs and serve a similar purpose: subduing prey. Scanning electron micrograph.

3.34. A campodeid dipluran in Early Miocene Dominican amber. Fossil diplurans are extremely rare. M2232; length 2.6 mm.

3.33. The cercal forceps of a japygid dipluran are similar to those in earwigs and serve a similar purpose: subduing prey. Scanning electron micrograph.

from their phylogenetic position. The only records of cam-podeomorphs are from the Tertiary, with a few specimens in Baltic (Middle Eocene) and Dominican (Early Miocene) ambers (Figure 3.34). Japygomorphs are also known from the Tertiary (in Dominican amber and Pliocene deposits of Arizona) but records also extend into the Mesozoic, albeit based on few specimens. Typical japygomorphs have been described from the Lower Cretaceous Santana deposits of Brazil (Bechly, 2001; Wilson and Martill, 2001). These represent the oldest, definitive members of the order but are remarkably similar to modern taxa (also indicative of an ancient origin for Diplura). The most controversial fossil is Testajapyx thomasi from the Upper Carboniferous of Mazon Creek. This fossil has been described as having well-developed compound eyes, relatively externalized mouth-parts, a series of reduced abdominal appendages (leglets), among other enigmatic traits (Kukalova-Peck, 1987). However, the preservation of the fossil is quite poor, it has not been reexamined by additional entomologists to evaluate these very unusual features, and as such it cannot be conclusively considered a dipluran (e.g., Bitsch, 1994; Kristensen, 1995).

3.34. A campodeid dipluran in Early Miocene Dominican amber. Fossil diplurans are extremely rare. M2232; length 2.6 mm.

0 0

Post a comment