• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Juliformian Millipedes From the Lower Devonian of Euramerica: Implications for the Timing of Millipede Cladogenesis in the Paleozoic

Posted on: Saturday, 22 July 2006, 06:00 CDT

By Wilson, Heather M

ABSTRACT-

Two new xyloiuloid millipedes (Diplopoda: Chilognatha: Juliformia) are described and placed in the new family Gaspestriidae: Gaspestria genselorum n. gen and sp. from the Emsian of Qubec and New Brunswick and Sigmastria dilata n. gen. and sp. from the Pragian of the Midland Valley of Scotland. These new millipedes extend the stratigraphic range of xyloiuloid millipedes, which previously were only described from the Pennsylvanian of Europe and North America. All xyloiuloid millipede families are placed within the superfamily Xyloiuloidea which is left incertae sedis within Juliformia due to a lack of preservation of diagnostic characters that would allow placement within an extant order. The presence in the Lower Devonian of the superorder Juliformia, universally agreed among diplopod taxonomists to represent the most derived clade of Diplopoda, indicates that most of millipede cladogenesis leading to high-rank extant clades had to have occurred by this time, much earlier than previously indicated by the fossil record. A stratocladogram for Myriapoda is constructed, and in combination with data from the plant fossil record and nuclear protein-encoding genes, new hypotheses regarding timing of millipede high-rank cladogenesis in the Paleozoic are formulated. These include terrestrialization of Diplopoda no later than the Ordovician along with the origin of the lineage leading to Penicillata and Arthropleuridea, followed by a period of relative stasis until the Middle Silurian, at which point there was a rapid radiation of Diplopoda, producing most of the high-rank clades by the Lower Devonian.

INTRODUCTION

WHEN PALEOZOIC millipedes (Myriapoda: Diplopoda) were first described, it was thought that they were sufficiently different from extant millipedes so as to preclude placement of these fossil taxa within Diplopoda. This prompted Scudder in 1882 to erect Archipolypoda, a separate order of Myriapoda coordinate with Diplopoda which, through subsequent additions, eventually came to contain almost all Paleozoic millipedes. For example, it was thought that the trunk segments of many Paleozoic millipedes had not fully coalesced into diplosegments, the "autapomorphy par excellence of the class Diplopoda" (Enghoff, 1990, p. 10). This misconception was first introduced into the literature by Peach (1882) who described Archidesmus macnicoli as having alternating small and large haplosegments, each bearing a single pair of appendages, as opposed to having diplosegments with a smaller prozonite and larger metazonite as in many extant millipedes. It has since been demonstrated that A. macnicoli has bona fide diplosegments (Wilson and Anderson, 2004), indicating that Peach (1882) misidentified the prozonite of each trunk pleurotergite as a separate segment. It was also thought that some Paleozoic millipedes, such as Euphoberiia Meek and Worthen, 1868 from the Czech Republic, had longitudinally divided sternites, a character thought to be primitive relative to extant millipedes (Fritsch, 1899). However, it turns out that euphoberiids do not have divided sternites (Burke, 1979; Wilson and Anderson, 2004), thus invalidating the remaining character state that had been used to establish Archipolypoda.

If the majority of Paleozoic millipedes belonged to the sister taxon of extant millipedes, then the implication is that most of cladogenesis leading to extant high-rank clades of Diplopoda occurred in the Late Paleozoic and into the Mesozoic. As of the publication of the Treatise on Invertebrate Paleontology in 1969, Archipolypoda was still considered to be a separate class of Myriapoda coordinate with Diplopoda (for a review of the composition of Archipolypoda, see Wilson and Anderson, 2004). As various taxa within Archipolypoda were revised, it became clear that not only was Archipolypoda not a natural group, but that many of the taxa it contained could be either placed within extant orders of Diplopoda or closely allied to them. For example within Pennsylvanian taxa, Amynilyspes Scudder, 1882 was removed and placed within the subclass Pentazonia and Xyloiulus Cook, 1895 was allied with the juliform order Spirobolida (Huffman, 1969). The superorder Juliformia currently has wide support as the most derived clade of Diplopoda (e.g., Enghoff, 1984; Sierwald et al., 2003). The presence of Juliformia in the Pennsylvanian meant that the timing of millipede high-rank cladogenesis needed to be reassessed, with most major clades arising before the Pennsylvanian. Most recently, Wilson and Anderson (2004) provided a revised diagnosis for Archipolypoda as a superorder within Diplopoda. Chilognatha comprised of three orders (Euphoberiida, Archidesmida, and Cowiedesmida) and demonstrated that even the oldest-known millipedes from the Middle Silurian could be accommodated within Chilognatha. Thus all known Paleozoic millipedes can be encompassed within the taxonomic framework developed for extant millipedes.

Xyloiuloid millipedes have been previously described from the Pennsylvanian of North America (Dawson, 1860; Scudder, 1890; Hoffman, 1963) and Europe (Fritsch, 1883; Anderson et al., 1999). New xyloiuloid millipedes are described herein from Qubec and New Brunswick, extending the stratigraphie range of the group back into the Lower Devonian. The presence of derived millipedes in the Lower Devonian, together with the trace- and body-fossil record of Diplopoda along with sequence data from nuclear protein-encoding genes, is used to formulate a new hypothesis for the timing of millipede high-rank cladogenesis in the Paleozoic. This timing has significant implications for understanding the timing and rate of evolution of complexity in terrestrial ecosystems.

GEOLOGICAL SETTING

Qubec.-One of the specimens of Gaspestria genselorum n. gen. and sp. was collected by Patricia and William Gensel at Locality W of Gensel and Andrews (1984) within the Cap-auxOs Member of the Battery Point Formation on the Gasp Peninsula, Qubec. This locality has been dated to the late Emsian using spores, falling within the sextantii Subzone of the annulatus-lindlarenisis Zone of Richardson and McGregor (1986). The millipede specimen was found in association with the plant genera Renalla Gensel, 1976 and Prototaxites Dawson, 1859. The Battery Point Formation occurs within strata that were deposited in a foreland basin to the west of the Acadian Orogen in response to the early stages of the Acadian Orogeny, and forming part of the Catskill clastic wedge (Rust. 1984; Lawrence and Rust, 1988). Cant and Walker (1976) suggested a braided river model of deposition for the Battery Point Formation based on evidence of lateral migration of in-channel bars and reconstructions of paleoflow directions. The Cap-aux-Os Member was interpreted by Lawrence and Rust (1988) as having been deposited by the meander belts of moderate to high sinuosity rivers that flowed across a muddy, vegetated coastal plain next to a marine embayment. The paleolatitude was approximately 10-20S in Emsian times (Ziegler. 1988; Scotese and McKerrow. 1990) and the climate was likely tropical (Witzke, 1990).

Terrestrial arthropod remains are rare within the Battery Point Formation. Several fragmentary specimens of the millipede Eiwrthropleura were described from the Battery Point Formation near Gros Cap aux Os by Wilson (1999). A single specimen of an archaeognathan hexapod was described from macerates of the Battery Point Formation by Labandeira et al. (1988); however, the fossil nature of this specimen was subsequently questioned by Jeramet al. (1990).

New Brunswick.-Additional specimens of Gaspexiria genselorutn were collected by Patricia and William Gensel from Locality B of Gensel and Andrews (1984) along the Restigouche River near Dalhousie Junction. New Brunswick, possibly within the Campbellton Formation. At this locality, G. genseltirum was found in association with the plant genera Ottcampsa Andrews et al., 1975. PsHophyton Dawson, 1859. Leclercqia Banks el al., 1972, and Kaulangiophyttm Gensel et al., 1969. This locality has been dated using spores to the late Emsian as part of the Grandiospora Subzone of the annulatus- lindlarensis Zone, as designated by McGregor (1973, 1977). The outcrops near Dalhousie Junction have not been recently mapped or stratigraphically assessed, although these sediments were considered to be equivalent to the Campbellton Formation by geologists working at the turn of the last century (Ami, 1900). Thus accurate assignment to existing formations and members and correlation with other Devonian strata in New Brunswick and Qubec is not possible. However, it is apparent that these sediments are younger than the early Devonian outcrops at Atholville (Locality Q of Gensel and Andrews, 1984; Atholville Member of the Campbellton Formation of Gamba, 1990), as well as those in the Point Ia Nim Formation of Gamba (1990) and Locality M of Gensel and Andrews (1984). The Campbellton Formation is a coarsening-upward alluvial-lacustrine sequence (Wilson et al., 2004). and the sediments of Locality B represent portions of the distal depositional sequence of a braided streambed that arose due to tectonic activity associated with a volcano tha\t was located south of the town of Campbellton (Gamba and Rust, 1989; P. Gensel, personal commun.. 2003).

Scotland.-A single xyloiuloid millipede, Sigmastria dilata n. gen. and sp., is known from the Dundee Formation of the Arbuthnott group at Carmyllie near Arbroath in Scotland. Spore assemblages from the Arbuthnott group were originally thought to indicate an early Pragian age (e.g., Richardson, 1967; Holland and Richardson, 1977), however, these are now considered to belong within the micmrnatus- newportenxis Spore Assemblage Biozone, indicating an early Lochkovian age (Richardson et al., 1984; Wellman, 1993; Lavender and Wellman, 2002). The Dundee Formation is dominated by planar and crossbedded fluvial sandstones, probably deposited in channels and as braid bars. However, drainage was periodically blocked to produce lacustrine conditions where clastic, carbonate, and organic laminates were deposited (Trewin and Davidson, 1996). In the Lower Devonian the topography of the Scottish portion of the Old Red Sandstone Continent was controlled in large part by contemporaneous normal faults that gave rise to major fault scarps and rift valleys. The Midland Valley Rift contained two major parallel river valleys, separated by volcanic uplands. The northern valley, in which the sediments of the Arbuthnott group were deposited, contained a large river system from Scandinavia with flow towards the southwest. The valleys were bounded by alluvial fans along escarpments close to both the Highland Boundary and Southern Upland faults (Mykura, 1983; Bluck. 2000). The myriapod fauna known from other localities within the Dundee Formation includes the archipolypodan millipede Archidesmus macnicoli and Kampecaris forfarensis Peach, 1882, a myriapod of uncertain taxonomic status (Almond, 1985; Wilson and Anderson, 2004).

SYSTEMATIC PALEONTOLOGY

Specimens studied are housed in the following institutions: New Brunswick Museum of Natural History (NBMG), St. John, New Brunswick; Muse d'Histoire Naturelle de Miguasha (MHNM). Miguasha. Qubec; National Museum of Scotland (NMS), Edinburgh: United States National Museum of Natural History (USNM), Washington, DC; and the Yale Peabody Museum of Natural History (YPM), New Haven. Connecticut.

Superorder JULIFORMIA Attems, 1926

Order INCERTAE SEDIS

Superfamily XYLOIULOIDEA Cook, 1895 new status

Included families.-Xyloiulidae Cook, 1895; Nyraniidae Hoffman. 1969; Plagiascetidae Hoffman, 1969; Gaspestriidae Wilson new family.

Diagnosis.-Paleozoic millipedes with tergites, pleurites, and stemites fused into complete rings; trunk rings ornamented with longitudinal striations; 40-50 trunk rings in adults. Ozopores present. Length of legs equal to half the height of the trunk rings or less.

Discussion.-Cook (1895) erected the family Xyloiulidae to include the striated millipedes that had been described from the Upper Carboniferous of Joggins, Nova Scotia, by Dawson (I860) as Xylobius (a name that was preoccupied by a beetle and changed to Xylaiulus by Cook). Hoffman (1969) established two new families, Nyraniidae and Plagiascetidae. to include the genera Nyranius Hoffman, 1963 and Plaxia.scetus Hoffman. 1963, respectively. These three families were distinguished by Hoffman (1969) based on the diameter of the prozonite relative to the metazonite and the distribution of the striations. In the Plagiascetidae, the striations are restricted to the ventrolateral surface of each metazonite and the prozonite is smooth and unornamented. In the Xyloiulidae and Nyraniidae. the trunk rings are completely striated, however in the Nyraniidae the striations of the prozonite are finer and more closely spaced relative to the metazonite. In the Nyraniidae. the diameter of the metazonite is distinctly greater than that of the prozonite, contrasting the Xyloiulidae and Plagiascetidae in which the diameter of the metazonite is only slightly greater.

Cook (1895) considered the xyloiuloids to be close to the extant Stemmiulida. based mainly on the ornament of the trunk rings. However, Stemmiulida have free sternites as opposed to complete, fused trunk rings found in the xyloiuloids. Hoffman (1969) placed Xyloiulidae. Nyraniidae, and Plagiascetidae within the extant order Spirobolida. Due to a lack of preservation of diagnostic characters that would allow placement of xyloiuloid millipedes within an extant order (such as the position and form of the genitalia and the form of the gnathochilarium). they are here left as order incertac sedis within the Juliformia (Julida + Spirobolida + Spirostreptida).

Ozopores are included in the diagnosis of Xyloiuloidea because they may be preserved in some specimens of Gaspestria genselorum n. gen. and sp., are prominently preserved in one of the syntypes of Xylobius mazonus Scudder, 1890 [USNM 38036; personal observation and Scudder (1890), pi. 37, fig. 11], and universally present in Juliformia, although they may not be readily discerned even in some extant taxa.

Family GASPESTRHDAE new family

Type genus.-Gaspestria new genus.

Included genera.-Sigmastria new genus.

Diagnosis.-Trunk rings with smaller prozonite and slightly inflated metazonite; prozonite unomamented; entire surface of metazonite ornamented with fine longitudinal striations.

Occurrence.-Lower Devonian (Lochkovian-Emsian) of the Old Red Sandstone Continent (Scotland, Qubec, and New Brunswick).

Genus GASPESTRIA new genus

Type species.-Gaspestria genselorum n. gen. and sp., by monotypy.

Diagnosis.-Numerous (10 mm^sup -1^) fine, parallel, longitudinal striations on metazonite of trunk rings. At least 40 trunk rings when mature.

Etymology.-After the Gasp Peninsula and the Latin stria, groove, referring to the cuticular ornament of the trunk rings.

GASPESTRIA GENSELORUM new species

Figures 1, 2

Fossil milliped from Battery Point Formation, SHEAR ET AL., 1996, p. 555, fig. 1.

Diagnosis.-As for genus.

Description.-Holotype MHNM 01.120 preserves 40 trunk rings in lateral aspect (Figs. 1.3, 1.4, 1.5, 2.1), NBMG 10089 preserves approximately 40 trunk rings in lateral aspect (Fig. 1.1, 1.2), NBMG 10090 preserves 17 trunk rings in lateral aspect (Fig. 1.6), and NBMG 10091 preserves 39 trunk rings in lateral aspect, reconstructed as approximately 40 mm long (Fig. 1.7). Head unknown. Tergites, pleurites, and stemites fused into rings (Fig. 2.1). Metazonite of trunk ring slightly wider then prozonite (Fig. 1.6), with closely spaced (10 per mm), fine longitudinal to slightly oblique striations covering entire surface (Figs. 1.4, 1.5, 2). Prozonite unomamented (Fig. 1.6). Small, circular depressions preserved at midheight on some trunk rings of MHNM 01.120 (Fig. 2.1) and NBMG 10091, possibly representing ozopores. Legs slender, approximately half the height of the trunk rings (Figs. 1, 2).

Etymology.-After the collectors, Patricia and William Gensel.

Type.-Holotype MHNM 01.120.

Other material examined.-NBMG 10089, 10090, 10091.

Occurrence.-Lower Devonian (late Emsian) of the Cap-auxOs Member of the Battery Point Formation, Gasp Peninsula, Qubec, and the ?Campbellton Formation, near Dalhousie Junction on the Restigouche River, New Brunswick.

Discussion.-A reconstruction of Gaspestria genselorum is presented in Figure 2.2. The form of the preanal ring, head, and collum are speculative and based on extant juliform millipedes. The number of trunk rings is also speculative as no complete specimens are known.

Genus SIGMASTRIA new genus

Type species.-Sigmastria dilata n. gen. and sp., by monotypy.

Diagnosis.-Prozonite smooth, unomamented; metazonite inflated posteriorly, ornamented with widely spaced (5 mm^sup -1^), slightly sinusoidal, longitudinal striations that are bounded anteriorly by a fine, transverse groove.

Etymology.-After the Greek sigma, the letter 's,' and the Latin stria, groove.

SIGMASTRIA DILATA new species

Figure 3

Diagnosis.-As for genus.

Description.-Holotype NMS G. 1957.1.5149 preserves series of eight articulated trunk rings in lateral aspect in a mudstone (Fig. 3). Stemites and legs not preserved. Trunk rings 2.2 times tall as wide. Prozonite smooth, unornamented. Metazonite twice as long as prozonite, inflated posteriorly giving trunk outline slightly moniliform appearance (Fig. 3.1). Metazonite ornamented with widely spaced (5 per mm), slightly sinusoidal longitudinal striations beginning at fine transverse groove (Fig. 3.2). Posterior margin of metazonite incomplete. No obvious ozopores.

Etymology.-After the Latin dilatus, spread out, enlarged, referring to the posterior dilation of the metazonites.

Type.-Holotype NMS G.1957.1.5149.

Occurrence.-Dundee Formation of the Arbuthnott group at Carmyllie near Arbroath, Scotland; Lower Devonian, Lochkovian.

TIMING OF MILLIPEDE CLADOGENESIS

Millipedes, as detritivores, are an important component of modern terrestrial ecosystems, playing an integral role in litter breakdown and nutrient cycling. While microorganisms are the main agents of litter catabolism, saprotrophic invertebrates enhance energy and nutrient flux rates through microbial populations. Millipede feeding activities increase the surface area of plant materials exposed to microbial attack, mobilize nutrients from senescent tissues, and maintain fungal and bacterial populations in an active growth phase. The amount of litter consumed by millipedes is not trivial. In one study, millipedes in the Seychelles were shown to ingest 4.5% of the standing crop of litter and 17% of the daily litter fall every 24 hours (Lawrence and Samways, 2003). In a study of millipedes in a Canadian forest, it was calculated that millipedes were consuming 36% of the annual litter fall (Carcamo et al., 2000). Bacterial populations on food leaves have been shown to increase up to a hundredfold after passage through the millipede gut, the environment of which appears to enhance growth and viability of ingested bacteria (Anderson and Bignell, 1980). Millipedes also increase the leaching of calciu\m, magnesium, nitrates, and organic carbon from litter and soil (e.g., Kaneko, 1999; Carcamo et al., 2000; Pramanik et al., 2001). Given their importance to nutrient cycling, understanding the timing of origin of high-rank millipede groups contributes to our understanding of the evolution of early terrestrial ecosystems. In the following discussion, the body and trace fossil record of millipedes is used in combination with additional evidence from the plant fossil record and nuclear- protein encoding genes to formulate hypotheses regarding the timing of millipede high-rank cladogenesis in the Paleozoic.

Evidence from phytogeny and the body fossil record.-A stratocladogram was constructed using the morphology-based diplopod phylogeny of Sierwald et al. (2003) and the morphology-based chilopod phylogeny of Shear and Bonamo (1988), which included the extinct order Devonobiomorpha and the combined morphology/molecular phylogeny of Edgecombe and Giribet (2002), together with the fossil record of Myriapoda (Fig. 4). The diplopod phylogeny of Sierwald et al. (2003) differs significantly from other published morphology- based phytogenies only in the placement of the Polydesmida. In Enghoff's (1984) phylogeny, Polydesmida occurs in a polytomy together with Julida, Spirostreptida, and Spirobolida, and in Enghoff et al.'s (1993) phylogeny, Polydesmida is the sister group to Juliformia (Julida + Spirostreptida + Spirobolida) as opposed to the sister group to Nematophora (Stemmiulida + Callipodida + Chordeumatida). However, the precise position of Polydesmida does not significantly affect the inferences that can be made from the stratocladograrn.

FIGURE 1-Gaspesiria gensetorum n. gen. and sp., Lower Devonian (Emsian). 1-6. Cap-aux-Os Member of the Battery Point Formation. Gasp Peninsula. Qubec, Canada. 1, 2, NBMG 10089, preserved in lateral aspect, anterior towards left: 1. entire specimen: 2. enlarged view of anterior: 3-5. MHNM 01.12OB, preserved in lateral aspect with anterior to right: 3. entire specimen: 4. enlarged view of midtrunk region: 5, enlarged view of anterior trunk region: 6, NBMG 10090, preserved in lateral aspect: 7. NBMG 10091. preserved in lateral aspect, anterior towards right, from Locality B of Gensel and Andrews (1984) near Dalhousie Junction, New Brunswick, Canada. Scale bars = 2 mm.

FIGURE 2-Gaspestria genselorum n. gen. and sp., Lower Devonian (Emsian). 1, Interpretative drawing of MHNM 01.120B (Fig. 1.3), av = anal valve, oz = ozopore, st = sternite; 2, reconstruction of an adult, anterior and posterior are speculative and are based on extant Juliformia.

It is widely accepted, based on morphological evidence, that the ring-forming juliform millipedes (Spirobolida + Julida + Spirostreptida) form a monophyletic clade, and that this clade represents the most derived group of millipedes (Enghoff, 1984; Enghoff et al., 1993; Sierwald et al., 2003). Previously, the oldestknown millipedes that could be assigned to the Juliformia were described from the Pennyslvanian of Euramerica, although a millipede originally identified as Euphoberia sp. by Nindel (1955) from the Visan Hainischen formation of the Erzegebirge basin in eastern Germany (Nindel, 1955) may also be a juliformian. The discovery of Gaspestria n. gen. and Sigmastria n. gen. is significant in that it moves the latest possible origin of the Juliformia back in time from the Pennsylvanian to the Lower Devonian. This is important because it necessitates that all of the extant high-rank clades of Diplopoda would had to have originated by the Lower Devonian at the latest.

The oldest-known millipede body fossils are from the Middle Silurian (Wenlock) and they belong to two clades, Archipolypoda and Zosterogrammida. Archipolypoda, as recently redefined by Wilson and Anderson (2004), span the mid-Silurian through the Pennsylvanian and comprise at least three orders: Euphoberiida, Archidesmida, and Cowiedesmida. Wilson and Anderson (2004) hypothesized that the Archipolypoda represent either helminthomorph millipedes or the sister group to Helminthomorpha, as Archipolypoda have all the characteristics of extant Helminthomorpha except their gonopods are on a different segment [segment eight as opposed to segment seven in Helminthomorpha-for a discussion of why this segmentai difference may not be taxonomically significant see Wilson and Anderson (2004)]. The Middle Silurian archipolypodan millipedes, including Pneumodesmus newmani Wilson and Anderson, 2004, Cowiedesmus eroticopodus Wilson and Anderson, 2004, and Albadesmus almondi Wilson and Anderson, 2004, are all from the Cowie Formation of the Stonehaven Group at Cowie Harbour on the northeast coast of Scotland. The zosterogrammidan millipede, Casiogrammus ichthyeros Wilson, 2005b, is from the Fish Bed Formation of the Glenbuck Group of the Hagshaw Hills Inlier in the Midland Valley of Scotland. Other representatives of the Zosterogrammida are known from the Pennsylvanian of Niany in the Czech Republic and Mazon Creek in Illinois (Wilson, 2005b). Zosterogrammida was described by Wilson (2005b) as a basal order of Chilognatha, possibly allied with the superorder Pentazonia based on the trunkring architecture and presence of divided sternites. Thus the oldest-known fossil millipedes do not belong to basal clades of Diplopoda but are quite derived, indicating that representatives of the lineages leading to Polyxenida, Arthopleuridea, Pentazonia, and possibly basal Helminthomorpha must already have been in existence by the Middle Silurian (Fig. 4).

FIGURE 3-Sigmastria dilata n. gen, and sp., Lower Devonian (Lochkovian), Dundee Formation of the Arbuthnott group at Carmyllie near Arbroath, Scotland. 1, Entire specimen in lateral aspect, anterior towards right; 2, enlarged view showing tergal ornament. Scale bars = 1 mm.

In the time elapsed between the Wenlock and the Lochkovian, considerable high-rank cladogenesis in Diplopoda must have occurred in order to produce the lineages leading to extant Colobognatha (Polyzoniida -I- Siphonophorida + Platydesmida), Nematophora (Stemmiulida + Callipodida + Chordeumatida)/Polydesmida, and Juliformia. Colobognatha is the basalmost clade of Helminthomorpha. While there are Paleozoic millipedes that may have colobognathan affinities, such as the Pleurojulida (Wilson and Hannibal, 2005), there are no Paleozoic millipedes that can be assigned to this superorder with any certainty. Mundel (1981) identified a platydesmidan millipede from the Mazon Creek Fauna, but a formal description was never published so this identification remains unconfirmed. A millipede figured by Rler and Schneider (1997, fig. 14) from the Visan of the Erzgebrige Basin in eastern Germany may also have colobognathan affinities. The oldest probable representative of the Nematophora is Hexecontasoma carinatum Hannibal, 2000 from the Pennsylvanian of Mazon Creek, which may be a callipodid. There are no known Paleozoic Polydesmida, with the first indisputable occurrence of this order in the Cretaceous (personal observation). Pleurojulus sleuri Schneider and Werneburg, 1998 from the Lower Rotliegend of Manebach in Germany is enigmatic. It was described as a pleurojulidan millipede, but the structures that authors interpreted as free pleuriies in the specimen are actually paranoia. The shape of the paranota resembles that of zanclodesmidan Archidesmida (Wilson et al., 2005), however the tergites are ornamented with a series of longitudinal ridges (reminiscent of some extant Callipodida) and each tergite has a prominent pair of ozopores. Ozopores are not known in archidesmidan Archipolypoda, but are known in euphoberiidans. As the ventral morphology of P. steuri is not known, it is impossible to determine its taxonomic affinities with any certainty, but it does not appear to represent a polydesmidan millipede.

FIGURE 4-Stratocladogram of Myriapoda combining the most recent morphological cladistic analysis of millipedes (Sierwald el al.. 2003). with the addition of extinct Paleozoic taxa, with current knowledge of the fossil record. Dashed lines denote range extensions and ghost lineages predicted from the phylogenetic tree. Note that taxon ranks are not necessarily equivalent. Arthropleuridea = Arthropleurida + Eoarthropleurida + Microdecemplicida: Archipolypoda = Archidesmida + Cowiedesmida + Euphoberiida; Pentazonia = Glomeridesmida + Glomerida + Sphaerotheriida; Colobognathu = Polyzoniida + Platydesmida + Siphonophorida; Nematophora = Stemmiulida + CaIIipodida + Chordcumatida; Juliformia = Xyloiuloidea + Spirobolida + Julida -H Spirostreplida. The numbers to the right of lhe range bars represent the following references: I, Keilbach (1982); 2, Poinar (1992); 3. Shear (1987); 4, Poinar and Edwards (1995); 5, BachofenEcht (1942, 1949); 6, Santiago-Blay and Poinar (1992); 7, Shear (1981); 8, Cockerell (1907); 9, Miner (1926); 10, Dzik (1975); II, Wilson (2001); 12, Wilson (2003a); 13, Manill and Barker (1998); 14, Grimaldi et al. (2002); 15, personal obs.; 16, Grimaldi (19%); 17, Schweigcrt and Dietl (1997); 18, Dzik (1981); 19, personal obs.; 20, Hannibal et al. (2004); 21. Mundel (1979); 22, Matthew (1894); 23, Wilson (1999); 24, Wilson (2005a); 25, Racheboeuf et al. (2004); 26, Frster(1973); 27, Hannibal and Feldmann (1981); 28, Fritsch (1883); 29, Scudder (1882); 30, Wilson and Anderson (2004); 31, Wilson (2005b); 32, Brauckmann and Kemper (1985); 33. Wilson and Hannibal (2005); 34, Rler and Schneider (1997); 35. Mundel (1981); 36, Hannibal (2000); 37, Dawson (1860); 38, Scudder (1890); 39, Hoffman (1969); 40. Nindel (1955); 41, Shear et al. (1998); 42, Payers (2003); 43. Shear and Bonamo (1988); 44, Shear and Selden (1995): 45, Wilson and Shear (2000); 46, Wilson et al. (2005); 47. this paper.

Evidence from trace fossils.-Trace fossils can provide invaluable information about the habits and distribution of organisms in space and time, providing informationfrom environments where body fossils are not preserved. Due to their terrestrial habitus, myriapod trace fossils are relatively uncommon, however trace fossils from several Cambrian-Ordovician localities have been attributed to subaerial myriapod activity, including both burrows and trackways.

Retallack (2001) described a new ichnospecies of burrow, Scnyenia heerhoweri, from a paleosol in the late Ordovician (Ashgill) Juniata Formation in Pennsylvania, first reported by Retallack and Feakes (1987), which he attributed to millipedes. Scoyenia White, 1929 is a longitudinally striated burrow with a conspicuous meniscate backfill (Frey et al., 1984), and S. beerboweri is characterized by a finely striated clayey lining, an occasional basal median groove and sinusoidal backfill, and width ranging from I to 21 mm (Retallack, 2001). Retallack (2001) hypothesized that the menisci of S. beerboweri were formed by the packing of soil behind the animal that was carried backwards by the legs, which was then capped in some cases by fecal material and compacted by backward ramming. Based on this behavioral scenario, Retallack suggested that a "borer" ecomorphotype would be the most likely to have formed the Scoyenia beerboweri traces on the basis of the sinusoidal backfills.

However, this assignment of the tracemakers to a "borer" ecomorphotype seems to be based on a misunderstanding of the behavior of this group. Hopkin and Read (1992, p. 39) described colobognathans and nematophorans as "borers," however, in Manton's (1961; 1977, p. 352-364) original description upon which this category was based, she described these groups as using a technique called "wedge-pushing." This type of burrowing requires that a millipede have an anteriorly tapered trunk and that the intersegmental joints not be incompressible as in Juliformia and Polydesmida so that each segment can be pulled into the segment in front of it to some extent. According to Manton's scenario, a wedge- pushing millipede burrows by anchoring successive segments by grasping the substrate with the legs of the segment in question. Simultaneously, the following segment is pulled forward into the anchored segment by trunk musculature. The segment that was pulled forward is then anchored and the process repeated. In this way, successively larger segments are dragged forwards, widening the crevice in which the millipede is locomoting. Thus 'borers' are not capable of physically boring down through hard substrates, but rather are crevice-wideners, an activity that does not seem likely to produce a sinusoidal backfill.

Retallack envisioned the Silurian millipedes from Stonehaven, Scotland, as being the closest-known fossil to the animal he posited as having formed Scoyenia beerboweri because he thought that they had flexible tergites and prominent sternites. While the archidesmid millipedes from Stonehaven had large, free sternites, the tergites certainly do not appear to have been particularly flexible and all had large paraterga (Wilson and Anderson, 2004). As such, these millipedes were unlikely to have been capable of forming burrows such as 5. beerboweri; moreover, it is unlikely that S. beerboweri was made by any type of millipede.

MacNaughton et al. (2002) described three types of arthropod trackways from the Cambrian-Ordovician Nepean Formation of the Potsdam Group in Ontario in sediments that they interpreted as having been deposited in an eolian dune environment. The authors suggested that the traces were made by large, homopodous arthropods with at least eight pairs of walking legs. MacNaughton et al. (2002) hypothesized that two of the trackway types, those with a central tail drag, may have been produced by euthycarcinoids. However, given that the long, gracile euthycarcinoid appendages bear an extensive fringe of setae and have a nonplantigrade structure (e.g., Gall and Grauvogel, 1964; Schram and Rolfe, 1982; Schneider, 1983; Schultka, 1991; Wilson and Almond, 2001), both features indicative of an entirely aquatic habitus, they are a curious choice given the wide range of known Cambrian arthropods to choose from (and also given the distinct possibility that the traces were produced by an undescribed arthropod, as no soft-bodied fauna is currently known from a marginal marine setting). The authors hypothesized that the third trackway type might have been made by a myriapod, however there are no features preserved in the trackway to suggest that it was made by an arthropod with appendages moving in metachronal waves.

The most convincing evidence for myriapod subaerial activity in the Ordovician comes from the English Lake District. Johnson et al. (1994) described arthropod trace fossils from the Llandeilo-Caradoc Dunnerdale and Whorneyside formations of the Borrowdale Volcanic Group. The authors proposed that the depositional environment for the sequence containing the trace fossils was freshwater lacustrine with periodic emergence, as demonstrated by the presence of mud cracks and adhesion ripples. They assigned the traces to two ichnogenera, Diplichnites Dawson, 1873 and Diplopodichnus Brady, 1947, and presented evidence, in the form of size-class distribution and the occurrence of transitions between ichnogenera within an individual trace, that both ichnogenera were produced by the same arthropod with the trace morphology dependent upon sediment water content. The Diplopodichnus trails consist of two parallel grooves while the Diplichnites trails consist of two parallel series of discrete imprints.

The morphology of the Lake District trace fossils is consistent with them having been produced by a millipede with a penicillate/ arthropleuridean body plan. Johnson et al. (1994) did not discuss the possibility that the change in trace morphology from Diplopodichnus to Diplichnites may also have been influenced by a change in gait by the producer. Wilson (2003b) found that when the extant penicillate millipede Polyxenus Latreille, 1802-1803 locomotes, it produces two different trackway morphologies dependent upon locomotion speed. At slow speeds, each leg in a metachronal wave contacts the substrate just behind the leg in front of it, creating a continuous series of imprints as the imprints from successive metachronal waves are superimposed. However, at faster speeds, all the legs in a metachronal wave contact the substrate in approximately the same place, creating a cluster of imprints. The distance between each cluster of imprints is equivalent to the stride length of the millipede. The giant Paleozoic millipede Arthropleura Jordan and Meyer, 1854 also produced both continuous and clustered trackways and, based on anatomical similarities between Polyxenus and Arthropleura, Wilson (2003b) hypothesized that the continuous Arthropleura trackways were produced at relatively low speeds while the clustered Arthropleura trackways were produced at relatively high speeds. Some of the Diplichnites trackways from the Sour Milk Gill locality in the Lake District (see Johnson et al., 1994, fig. 4) are here interpreted as having a clustered morphology in which each imprint represents the grouped imprints from one metachronal wave. If this interpretation is correct, then it is reasonable to hypothesize that the trackways were made by a millipede with a penicillate/arthropleuridean body plan. Millipedes with a more derived body plan can be ruled out because chilognath millipedes are incapable of producing clustered trackways. This is because when they locomote, propulsive legs are always diverging (see Manton, 1977, fig. 7.4) as opposed to the case in Penicillata where propulsive legs converge (see Manton, 1956, text-fig. 5). Although the propulsive legs in many centipedes converge during locomotion (see Manton, 1977, fig. 6.13), none are known to produce a clustered trackway. So we can hypothesize that the lineage leading to Penicillata/Arthropleuridea is at least as old as the Llandeilo- Caradoc (Fig. 4).

Supporting evidence from the plant fossil record.-Advances in the fossil record of plants have yielded some evidence which is useful in formulating hypotheses regarding the timing of millipede high- rank cladogenesis. Strother et al. (2004) describe cryptospores from the Middle Cambrian of eastern and western North America. They attribute these spores, which occur in both dyads and tetrads, to subaerial thallophytic plants, some of which may have been at a bryophytic grade. However, these spores have not yet been widely accepted by paleobotanists as evidence for Cambrian terrestrial plants. The oldest generally accepted evidence for land plants also takes the form of dispersed microscopic spores (cryptospores), this time from the mid-Ordovician (Llanvirn, -475 My) of Saudi Arabia (Strother et al., 1996). These spores are abundant and globally distributed from the Ordovician onwards, decreasing in abundance through the Silurian and Lower Devonian, and have been suggested to be from early relatives of the bryophytes (Wellman and Gray, 2000). This interpretation was originally controversial because there was little direct evidence of the parent plants. However, spore- containing plant fragments from the mid-Ordovician (Caradoc) of Oman have recently been described that appear to have liverwort affinities (Wellman et al., 2003), supporting the bryophyte affinity of the Ordovician-dispersed cryptospores. The first unequivocal macroscopic plant remains do not appear in the fossil record until the Silurian (Wenlock, -425 My) (Edwards and Feehan, 1980). Thus there is a significant lag of at least 50 million years between the appearance of the first dispersed terrestrial plant spores and fossils of relatively complete land-plant megafossils. During this time the vegetation was widespread, but of limited diversity and underwent very little evolutionary change, at least as indicated by spores, until the late Llandovery (Wellman and Gray, 2000). In th\e late Llandovery many types of cryptospores disappeared from the fossil record and trilete spores appeared. It has been suggested that the inception of trilete spores may reflect the first appearance of tracheophytes (Edwards and Wellman, 2001). Following this major turnover in spore types, trilete spores underwent a major diversification, by inference reflecting a diversification in early tracheophytes.

The presence of diplopod trace fossils, as discussed above, in the Ordovician (Caradoc, ~450 My) that have been interpreted as being produced subaerially (Johnson et al., 1994), suggests that myriapod terrestrialization was coeval with that of plants. It is possible to envisage a scenario in which rapid evolutionary change in myriapods associated with terrestrialization and diversification into four classes (Chilopoda, Symphyla, Pauropoda, and Diplopoda) is followed by a period of relatively slow evolutionary change in the Ordovician, mirroring that of the plants. During this period, the diplopods would be represented by only a few basal lineages. By the Wenlock we have fossil evidence for Archipolypoda and Zosterogrammida, and based on morphological phylogenies can infer the presence of stem group Penicillata and Arthropleuridea. The first appearance of millipede body fossils at the same time as a major diversification in tracheophyte plants may not be a coincidence. However, a relationship between these events is difficult to evaluate because the Ordovician through the early Silurian was a time of persistently high sea levels, and relatively fewer continental deposits are known compared to the later Silurian and Devonian. Based on the presence of juliform millipedes in the Lower Devonian, we would predict that all other millipede high-rank clades arose between the Middle Silurian and the beginning of the Devonian. This would necessitate a rapid radiation in Diplopoda in the Silurian, mirroring that of the terrestrial vegetation.

Evidence from protein-encoding genes.-A rapid radiation of Myriapoda has been invoked as an explanation for the inability of sequences from nuclear protein-encoding genes to resolve the interclass and interordinal relationships within Myriapoda (Regier and Shultz, 2001). The reasoning is that lineages that persist for long periods of time before diverging accumulate many substitutions, and thus a large phylogenetic signal is present in the genes. However, in those lineages that persist for a short time before diverging, there is a smaller phylogenetic signal in the same gene sequences due to the relatively smaller number of substitutions accumulated during the shorter time interval. Regier and Shultz (2001) attempted to resolve myriapod phytogeny using two nuclear protein-encoding genes, elongation factor-1α (EF-1α) and the largest subunit of RNA polymerase II (Pol II). They were able to demonstrate further the ability of EF-1α and Pol II for resolving higher-level arthropod relationships, but also revealed some limitations in the ability of these genes to resolve relationships between classes and within orders. They found that for Myriapoda, the phylogenetic signal within these genes is sufficient to resolve relationships at the subphylum, class, and order levels, but not at intermediate levels. Regier and Shullz (2001) hypothesized that these results were due to either insufficient taxon sampling, or alternatively, they represent a hierarchically arranged pattern of heterogeneity in phylogenetic signal within Myriapoda due to differing rates of phylogenetic diversification within the clade. In order to evaluate the "data insufficiency" and "rapid cladogenesis" hypotheses, Regier et al. (2005) increased the taxon sampling from 34 to 55 myriapods and added sequence from a third nuclear protein-encoding gene, elongation factor-2. Despite using a data matrix enhanced by additional taxa and sequence information, they found a pattern remarkably similar to the 2001 analysis: myriapod classes and orders tend to be recovered with strong to moderate support, but there are relatively few cases in which relationships among classes or orders within classes are recovered. Within Diplopoda, strong support occurred for three interordinal groups, Pentazonia (Glomeridesmida + Glomerida + Sphaerotheriida), Colobognatha (Polyzoniida, Platydesmida, Siphonophorida), and Helminthomorpha (Chilognatha exclusive of Pentazonia). However, many intraordinal relationships were recovered with strong support that are in accordance with previous morphology- based phylogenies. These results suggest that addition of taxa and sequence information does not significantly increase the phylogenetic signal useful in resolving relationships among myriapod classes and orders, but are consistent with the proposal that poor resolution is caused by heterogeneity in the timing of phylogenetic diversification: rapid radiation of classes and orders and relatively slower radiation within orders.

A pattern of diversification as hypothesized above could help explain the ability of EF-1α, EF-2, and Pol II sequences to resolve convincingly some millipede relationships and not others. During the long period of relatively slow evolutionary change during the Ordovician, many substitutions could be accumulated leading to a large phylogenetic signal. During the Upper Silurian, cladogenesis may have been too rapid for sufficient substitutions to accumulate to yield a strong phylogenetic signal.

Pisani et al. (2004) conducted a molecular clock analysis using a combination of nuclear and mitochondria! genes in order to hypothesize divergence times of major arthropod lineages. Their estimated time of divergence for millipedes and centipedes was 442 50 Ma, which they took as the minimal time (most recent) of arthropod teirestrialization, and the time of divergence for myriapods and chelicerates as 642 63 Ma, which they took as the maximal (earliest) time for myriapod terrestrialization. While the principles behind molecular clock analyses may be controversial, the results of this study are not incompatible with the hypotheses regarding the timing of high-rank millipede cladogenesis proposed herein.

WHERE ARE THE AQUATIC MYRIAPODS?

Hilken's (1998) study of the tracheal system across Myriapoda led him to propose that tracheal systems arose independently in four lineages-Chilopoda Notostigmophora, Chilopoda Pleurostigmorphora, Symphyla, and Dignatha (Pauropoda + Diplopoda)-indicating that multiple terrestrialization events occurred. Given this, why have any aquatic myriapods yet to be identified in the fossil record when several lineages of aquatic myriapods must have existed at some point? Identifying stem group Myriapoda is exceedingly difficult due to an extreme paucity of autapomorphies for the clades that have preservation potential. Possible autapomorphies of Myriapoda include a moveable cephalic endoskeleton that functions in abduction of the mandibles (Klasse and Kristensen, 2001), an independently musculated gnathal lobe, comb lamellae on the gnathal lobe (Edgecombe and Giribet, 2002), and a suite of absence characters including the loss of the following: median eyes, a crystalline cone, perforatorium in the sperm, and scolopidial mechanoreceptors (Bacetti et al., 1979; Jamieson, 1987; Ax, 1999; Paulus, 2000). A detailed discussion of each of these characters is provided by Edgecombe and Giribet (2002) and Edgecombe (2004). Of these characters, the only one with a moderate amount of preservation potential is the independently musculated gnathal lobe, which can be identified in fossils based on an articulation between the gnathal lobe and basal portion of the mandible. The only "myriapod" feature that has a decent preservation potential is the presence of a long, homonomously segmented body. However, long bodies have evolved numerous times within Arthropoda and are by no means diagnostic of Myriapoda.

Cambrian arthropods with long, homonomously segmented bodies are not lacking in the fossil record, however, and as discussed above, there are no characters preserved in these taxa that can be used to ally them with Myriapoda. Such arthropods include Meristosoma paradoxum Robison and Wiley, 1995 from the Middle Cambrian Spence Shale of Utah and Pseudoiulia cambriensis Hou and Bergstrm, 1998 from the Lower Cambrian Chengjiang Lagerstatte in the Yunnan Province, China. Pseudoiulia cambriensis has at least 31 tergites that appear to have been strongly vaulted. The appendages are not completely known, but the exopods are setiferous (this alone would not exclude P. cambriensis from the myriapod stem group, as biramous appendages are plesiomorphic for Euarthropoda). However, Hou et al. (2004) tentatively allied Pseudoiulia Hou and Bergstrom, 1998 with the "great appendage" arthropods such as Fortiforceps foliosa Hou and Bergstrm, 1997 and Jianfengia multisegmentalis Hou, 1987, as these arthropods also have long bodies, not the most robust of reasons. Budd et al. (2001) described a long-bodied arthropod, Xanthomyria spinosa, from the Upper Cambrian of East Siberia. Xanthomyria spinosa has at least 34 postcephalic tergites that they interpreted as being calcareous. Each segment has four large subconical tubercles and a pair of elongate, lateral pleural spines. The pair of paramedian tubercles together form an area which is roughly trapezoidal in shape, with the long side along the posterior margin of the tergite. Budd et al. hypothesized that X. spinosa has myriapod affinities based on the multi-segmented trunk, mineralization of the cuticle, and the morphology of the tergites, comparing the tergal morphology of X. spinosa with that of euphoberiidan millipedes. Euphoberiidan millipedes occur in the Carboniferous of Euramerica and are armed with both dorsal and lateral spines. However, from a phylogenetic perspective the resemblance would have to be circumstantial because Euphoberiida \is not a basal clade of Diplopoda (Wilson and Anderson, 2004) and there is evidence for a single terrestrialization event in Pauropoda + Diplopoda (Hilken, 1998). Number of trunk segments aside, there are more similarities in the tergal morphology of between Xanthomyria Budd, Hgstrm, and Gogin, 2001 and synziphosurines such as Maldybulakia Tesakov and Alekseev, 1998 and Willwerathia Strmer, 1969. Maldybulakia, from the Lower Devonian of central Kazakhstan, was originally described as Lophodesmus (a name preoccupied by an extant millipede) by Tesakov and Alekseev (1992) as an arthropod with possible myriapod affinities. Edgecombe (1998a, 1998b) further developed the myriapod hypothesis in descriptions of specimens of Maldybulakia from the Devonian of Australia. However, Anderson et al. (1998) suggested a synzyphosurine affinity for Maldybulakia based on the shape of tergites and the form of their anterior articular surface.

ACKNOWLEDGMENTS

For loan of material under their care I thank R. Miller, New Brunswick Museum, J. Kerr, Muse d'Histoire Naturelle de Miguasha, and L. Anderson, National Museum of Scotland. I also thank the collectors of the Canadian material, W. Gensel and P. Gensel. This work was supported by National Science Foundation grant DEB- 0075605, a New Brunswick Museum G. F. Mathew Fellowship, and the Division of Invertebrate Paleontology, Yale Peabody Museum of Natural History.

REFERENCES

ALMOND, J. E. 1985. The Silurian-Devonian fossil record of the Myriapoda. Philosophical Transactions of the Royal Society of London, 8309:227-237.

AMI, H. M. 1900. Synopsis of the geology of Canada; being a summary of the principal terms employed in Canadian geological nomenclature. Proceedings and Transactions of the Royal Society of Canada, 6:187-225.

ANDERSON, J. M., AND D. E. BIGNELL. 1980. Bacteria in the food, gut contents and faeces of the litter-feeding millipede Glomeris marginata (Villers). Soil Biology and Biochemistry, 12:251-254.

ANDERSON, L. I., M. POSCHMANN, AND C. BRAUCKMANN. 1998. On the Emsian (Lower Devonian) arthropods of the Rhenish Slate Mountains: 2. The synziphosurine Willwerathia. Palaontologische Zeitschrift, 72: 325-336.

ANDERSON, L. I., J. A. DUNLOP, R. M. C. EAGER, C. A. HORROCKS, AND H. M. WILSON. 1999. Soft-bodied fossils from the roof shales of the Wigan Four Foot coal seam, Westhoughton, Lancashire, UK. Geological Magazine, 135:321-329.

ANDREWS, H. N., P. G. GENSEL, AND A. E. KASPER. 1975. A new fossil plant of probable intermediate affinities (Trimerophyte- Progymnosperm). Canadian Journal of Botany, 53:1719-1728.

ATTEMS, C. G. 1926. Myriopoda, p. 1-402. In Handbuch der Zoologie. Vol. 4. W. De Gruyter, Berlin.

AX, P. 1999. Das System der Metazoa II. Bin Lehrbuch der Phylogenetischen Systematik. Urban and Fischer, Miinchen, 383 p.

BACETTI, B., A. G. BURRINI, R. DALLAI, AND V. PALLINI. 1979. Recent work in myriapod spermatology (The spermatozoon of Arthropoda XXXI), p. 97-104. In M. Camatini (ed.), Myriapod Biology. Academic Press, London.

BACHOFEN-ECHT, A. 1942. Ueber die Myriapoden des Bernsteins. Palaeobiologica, 7:394-403.

BACHOFEN-ECHT, A. 1949. Der Bernstein und Seine Einschliisse. Springer-Verlag, Wein, 204 p.

BANKS, H. P., P. M. BONAMO, AND J. P. GRIERSON. 1972. Leclercqia complexa gen. et sp. nov., a new lycopod from the late Middle Devonian of eastern New York. Review of Palaeobotany and Palynology, 14:19-40.

BLUCK, B. J. 2000. Old Red Sandstone basins and alluvial systems of Midland Scotland, p. 417-437. In P. F. Friend and B. P. J. Williams (eds.), New Perspectives on the Old Red Sandstone. Geological Society Special Publication, 180, London.

BRADY, L. F. 1947. Invertebrate tracks from the Coconino Sandstone of northern Arizona. Journal of Paleontology, 21:466-472.

BRAUCKMANN, C., AND M. KEMPER. 1985. Bin TausendfiiBler (Myriapoda:?Archipolypoda) aus dem Namurium B von Hagen-Vorhalle (unt. Ober-Karbon; West-Deutschland). Dortmunder Beitrage zur Landeskunde: Naturwissenschaftliche Mitteilungen, 19:65-69.

BUDD, G. E., A. E. S. HOGSTROM, AND I. GOGIN. 2001. A myriapod- like arthropod from the Upper Cambrian of East Siberia. Palaontologische Zeitschrift, 75:37-41.

BURKE, J. J. 1979. A new millipede genus, Myriacantherpestes (Diplopoda, Archipolypoda) and a new species, Myriacantherpestes bradebirksi, from the English Coal Measures. Kirtlandia, 30:1-24.

CANT, D. J., AND R. G. WALKER. 1976. Development of a braided- fluvial facies model for the Devonian Battery Point Sandstone, Qubec. Canadian Journal of Earth Sciences, 12:102-119.

CARCAMO, H. A., T. A. ABE, C. E. PRESCOTT, F. B. HOLL, AND C. R CHANWAY. 2000. Influence of millipedes on litter decomposition, N mineralization, and microbial communities in a coastal forest in British Columbia, Canada. Canadian Journal of Forest Research, 30:817-826.

COCKERELL, T. D. A. 1907. Some fossil arthropods from Florissant, Colorado. Bulletin of the American Museum of Natural History, 23:605- 616.

COOK, O. F. 1895. Introductory note on the families of Diplopoda, p. 1-9. In O. F. Cook and G. N. Collins (eds.), The Craspedosomatidae of North America. Annals of the New York Academy of Sciences. Vol. 9. New York Academy of Sciences, New York.

DAWSON. J. W. 1859. On fossil plants from the Devonian rocks of Canada. Quarterly Journal of the Geological Society of London, 15:477-488.

DAWSON. J. W. 1860. On a terrestrial mollusk, a chilognathous myriapod. and some new species of reptiles from the coal formation of Nova Scotia. Geological Society of London Quarterly Journal. 16:268-277.

DAWSON. J. W. 1873. Impressions and footprints of aquatic animals and imitative markings, on Carboniferous rocks. American Journal of Science. 105:16-24.

DZIK, J. 1975. Spiroboloid millipedes from the Late Cretaceous of the Gobi Desert. Mongolia. Palaeontologia Polonica. 33:17-24.

DZIK, J. 1981. An Early Triassic millipede from Siberia and its evolutionary significance. Neues Jahrbuch fiir Gologie und Palaonlologie. Monatshefte. 1981(7):395-404.

EDGKCOMBE. G. D. 1998a. Devonian terrestrial arthropods from Gondwana. Nature. 394:172-175.

EDGECOMBE, G. D. 1998b. Early myriapodous arthropods from Australia, Maldybulakia from the Devonian of New South Wales. Records of the Australian Museum. 50:293-313.

EDGECOMBE. G. D. 2004. Morphological data, extant Myriapoda. and the myriapod stem-group. Contributions to Zoology. 73:207-252.

EDGECOMBE. G. D., AND G. GIRIBET. 2002. Myriapod phytogeny and the relationships of Chilopoda, p. 143-168. In 3. Llorente Bousquets. J. J. Morrone, and H. Ponce Ulloa (eds.), Biodiversidad. Taxonomia y Biogeographia de Artrpodes de Mxico: Hacia una Sntasis de su Conocimiento. Prensas de Ciencias, Universidad Nacional Autmoma de Mxico, Mexico City.

EDWARDS, D., AND J. FEEHAN. 1980. Records of Cooksonia-type sporangia from late Wenlock strata in Ireland. Nature. 287:41-42.

EDWARDS, D., AND C. H. WELLMAN, 2001. Embryophytes on land: The Ordovician to Lochkovian (Lower Devonian) record, p. 3-28. In P. G. Gensel and D. Edwards (eds.), Plants Invade the Land. Columbia University Press, New York.

ENGHOFF, H. 1984. Phytogeny of millipedes-a cladistic analysis. Zeitschrift fr zoologische systematic Evolutionsforschung. 22:8-26.

ENGHOFF, H. 1990. The ground-plan of chilognathan millipedes (external morphology), p. 1-21. In A. Minelli (ed.). Proceedings of the 7th International Congress of Myriapodology. E. J. Brill, Leiden.

ENGHOFF. H., W. DOHLE, AND J. G. BLOWER. 1993. Anamorphosis in millipedes (Diplopoda)-the present state of knowledge with some developmental and phylogenetic considerations. Zoological Journal of the Linnean Society. 109:103-234.

FAYERS, S. 2003. The biota and palaeoenvironments of the Windyfield Chert, Early Devonian, Rhynie, Scotland. Unpublished Ph.D. thesis. University of Aberdeen, United Kingdom, 430 p.

FRSTER. R. 1973. Ein Diplopoden-fund aus dem Oberkarbon des Saarlandes. Neues Jahrbuch fr Geologie und Palontologie, Monatschefte. 2:67-71.

FREY, R. G., S. G. PEMBERTON, AND J. A. FAGERSTROM, 1984. Morphological, ethological and environmental significance of the ichnogenera Scoyenia and Ancorichnus. Journal of Paleontology. 58:511-528.

FRITSCH, A. 1883. Fauna der Gaskohle und der Kalksteine der Permformation Bhmens. Vol. I. F. Rivnc, Prague.

FRITSCH, A. 1899. Fauna der Gaskohle und der Kalksteine der Permformation Bhmens. Vol. 4. Prague.

GALL, J.-C, AND L. GRAUVOGEL. 1964. Un arthropode peu connue le genre Eulhycarcinus Handlirsch. Annales de Palontologie, 50:3-18.

GAMBA, C. A. 1990. Sedimentology and tectonic implications of the Pointe la Nim and Campbelllon Formations, western Chaleur Bay, Maritime Canada. Unpublished M.S. thesis. University of Ottawa, Ottawa.

GAMBA, C. A., AND B. R. RUST. 1989. Sedimentology of the Campbellton and La Garde formations of western Chaleur Bay; an initial response to Acadian tectonism. Geological Association of Canada and Mineralogical Association of Canada Annual Meeting Program with Abstracts. 14:113.

GENSEL, P. G. 1976. Kenalia hueberi, a new plant from the Lower Devonian of Gasp. Review of Palaeobotany and Palynology. 22:19-37.

GENSEL, P. G., AND H. N. ANDREWS. 1984. Plant Life in the Devonian. Praeger, New York, 380 p.

GENSEL. P. G., A. E. KASPER. AND H. N. ANDREWS. 1969. Kaulangiophyton, a new genus of plants from the Devonian of Maine. Torrey Botanic Club Bulletin, 96:265-276.

GRIMALDI. D. A. 1996. Amber. Window to the Past. Harry N. Abrams, in association with the American Museum of Natural History. New York. 216 p.

GRIMALDI, D. A., M. S. ENGEL. AND P. C. NASCIMBENE. 2002. Fossiliferous amber from Myanmar (Burma): Its rediscovery, biotic diversity, and paleontological significance. American Museum Novitates. 3361: 1-72.

HANNIBAL, J. T. 2000. Hexecontasoma. a new helminthomorph millipede (Hexacontasomatidae n. fam.) from the Mazon Creek, Illinois, fauna (Carboniferous, North America), p. 19-35. In i. Wytwer and S. Golovatch (eds.), Progr\ess in Studies on Myriapoda and Onycophora. Fragmenta Faunistica, 43 (suppl.).

HANNIBAL, J. T., AND R. M. FELDMANN. 1981. Systematics and functional morphology of oniscomorph millipedes (Arthropoda: Diplopoda) from the Carboniferous of North America. Journal of Paleontology. 55: 730-746.

HANNIBAL, J. T., A. J. LERNER. K. E. ZEIGLER. AND S. G. LUCAS. 2004. A juliform milliped from the Upper Pennsylvanian (Virgilian) Bursum Formation. Carrizo Arroyo. of central New Mexico. New Mexico Museum of Natural History and Science Bulletin. 25:211-213.

HILKEN, G. 1998. Vergleich von Tracheensystemen unter phylogenetischem Aspekt. Verhandlungen der nalurwissenschaftlichen Vereins in Hamburg (NF). 37:5-94.

HOFFMAN, R. L. 1963. New genera and species of Upper Paleozoic Diplopoda. Journal of Paleontology, 37:167-174.

HOFFMAN. R. L. 1969. Myriapoda, exclusive of Insecta, p. R572- R606. In R. C. Moore (ed.). Treatise on Invertebrate Paleontology. Vol. 2. Pt. R. Geological Society of America and University of Kansas Press, Lawrence.

HOLLAND, C. H., AND J. B. RICHARDSON. 1977. The British Isles, p. 35-44. In The Siluro-Devonian Boundary. I.U.C.S. series A, no. 5.

HOPKIN, S. R, AND H. J. READ. 1992. The Biology of Millipedes. Oxford University Press. Oxford. 233 p.

HOU, X.-G. 1987. Two new arthropods from Lower Cambrian, Chengjiang. eastern Yunnan. Acta Palaeontologica Sinica. 26:236- 256.

HOU, X.-G., AND J. BERGSTROM. 1997. Arthropods of the Lower Cambrian Chengjiang fauna. Southwest China. Fossils and Strata. 45: l-l 16.

HOU, X.-G., AND J. BERGSTROM. 1998. Three additional arthropods from the early Cambrian Chengjiang Fauna. Yunnan. Southwest China. Acta Palaeontologica Sinica. 37:395-401. pis. 1. 2.

HOU, X.-G., R. J. ALDRIDGE. J. BERGSTROM. D. J. SIVETER. D. J. SivETER, AND X.-H. FENG. 2004. The Cambrian Fossils of Chengjiang, China. Blackwell Publishing. Maiden. Massachusetts. 233 p.

JAMIESON, B. G. M. 1987. The Ultrastructure and Phylogeny of Insect Spermatozoa. Cambridge University Press, Cambridge, 320 p.

JERAM, A. J., P. A. SELDEN. AND D. EDWARDS. 1990. Land animals in the Silurian: Arachnids and myriapods from Shropshire, England. Science. 250:658-661.

JOHNSON, E. W., D. E. G. BRIGGS, R. J. SUTHREN. J. L. WRIGHT, AND S. P. TUNNICLIFF. 1994. Non-marine arthropod traces from the subaerial Ordovician Borrowdale Volcanic Group, English Lake District. Geological Magazine, 131:395-406.

JORDAN, H., AND H. v. MEYER. 1854. Ueber die Crustaceen der Steinkohlenformation von Saarbriicken. Palaeontographica A. 4:1-15.

KANEKO, N. 1999. Effect of millipede Parafontaria ionominea Attems (Diplopoda: Xystodesmida) adults on soil biologica activities: A microcosm experiment. Ecological Research, 14:271- 279.

KEILBACH, R. 1982. Bibliographie und Lisle der Arten tierischer EinschlUsse in fossilen Harzen sowie iher Aufbewahrungsorte. Deutsche entomologische Zeitschrit N. F. 29:129-286,301-491.

KLASSE, K.-D., AND N. P. KRISTENSEN. 2001. The ground plan and affinities of hexapods: Recent progress and open problems. Annales de la Socit Entomologique de France. 37:265-298.

LABANDEIRA, C. C., B. S. BEALL, AND F. M. HUEBER. 1988. Early insect diversification; evidence from a Lower Devonian bristletail from Qubec. Science, 242:913-916.

LATREILLE, P. A. 1802-1803. Histoire Naturelle. Gnrale et Particulire des Crustacs et des Insectes; ouvrage faisant suite aux oeuvres de Leclerc de Buffon. et partie du cours complet d'histoire naturelle rdig par C. S. Sonnini. Vol. 2. F. Dufart, Paris, 467 p.

LAVENDER, K., AND C. H. WELLMAN. 2002. Lower Devonian spore assemblages from the Arbuthnott Group at Canterland Den in the Midland Valley of Scotland. Review of Palaeobotany and Palynology, 118: 157-180.

LAWRENCE, D. A., AND B. R. RUST. 1988. The Devonian clastic wedge of eastern Gasp and the Acadian orogeny, p. 53-64. In N. J. McMillan, A. E Embry, and D. J. Glass (eds.), Devonian of the World. Vol. II. Canadian Society of Petroleum Geologists, Memoir 7.

LAWRENCE, J. M., AND M. J. SAMWAYS. 2003. Litter breakdown by the Seychelles giant millipede and the conservation of soil processes on Cousine Island, Seychelles. Biological Conservation, 113:125-132.

MACNAUOHTON, R. B., J. M. COLE, R. W. DALRYMPLE, S. J. BRADDY, D. E. G. BRIOOS, AND T. D. LUKIE. 2002. First steps on land: Arthropod trackways in Cambrian-Ordovician eolian sandstone, southeastern Ontario, Canada. Geology, 30:391-394.

MANTON, S. M. 1956. The evolution of arthropodan locomotory mechanisms, Pt. 5, The structure, habits and evolution of the Pselaphognalha (Diplopoda). Journal of the Linnean Society (Zoology), 43:153-187, pi. 5.

MANTON, S. M. 1961. The evolution of arthropodan locomotory mechanisms, Pt. 7, Functional requirements and body design in Colobognatha (Diplopoda), together with a comparative account of diplopod burrowing techniques, trunk musculature and segmentation. Journal of the Linnean Society (Zoology), 44:383-461.

MANTON, S. M. 1977. The Arthropoda. Clarendon Press, Oxford, 527 p.

MARTILL, D. M., AND M. J. BARKER. 1998. A new centipede (Arthropoda, Chilopoda) from the Crato Formation (Lower Cretaceous, Aptian) of N. E. Brazil. Neues Jahrbuch fur Palaontologie Abhandlungen, 207: 395-404.

MATTHEW, G. F. 1894. On the organic remains of the Little River Group, No. III. Transactions of the Royal Society of Canada, 1894:101-110, pi. 1.

McGREGOR, D. C. 1973. Lower and Middle Devonian spores of eastern Gasp, Canada. I. Systematics. Palaeontographica, 1428:1-77.

MCGREGOR, D. C. 1977. Lower and Middle Devonian spores of eastern Gasp, Canada. II. Biostratigraphy. Palaeontographica, 1638:111-142.

MEEK, F. B., AND A. H. WORTHEN. 1868. Preliminary notice of a scorpion, a Eurypterusl And other fossils, from the coal-measures of Illinois. American Journal of Science, second series, 46:25-27.

MINER, R. W. 1926. A fossil myriapod of the genus Parajulus from Flo

Source: Journal of Paleontology

More News in this Category