Arthropod Structure & Development 38 (2009) 401–416 Contents lists available at ScienceDirect Arthropod Structure & Development journal homepage: www.elsevier.com/locate/asd Early development of the anterior body region of the grey widow spider Latrodectus geometricus Koch, 1841 (Theridiidae, Araneae) Yu Liu*, Andreas Maas, Dieter Waloszek Workgroup Biosystematic Documentation, University of Ulm, Helmholtzstrasse 20, Ulm, Germany a r t i c l e i n f o a b s t r a c t Article history: Received 17 November 2008 Accepted 1 April 2009 We document the early morphogenesis of Latrodectus geometricus, particularly of the anterior body region. Significant changes in the development of the external prosomal structures revealed with scanning electron microscopy (SEM) images include: (1) reorganisation of each pre-cheliceral lobe by subdivision and internalisation of its central area; (2) shortening of the ventro-median bridge connecting the pre-cheliceral lobes and its eventual disappearance; (3) appearance and expansion of a prospective mouth region between the pre-cheliceral lobes with a recessed median area surrounded by lip-like borders, the anterior lip-part developing into the hypostome; (4) reduction of the mouth region to an area around the hypostome and the lip-like latero-posterior border of the mouth opening; (5) change of the position of the mouth region from anterior to the insertions of the chelicerae to posterior to them; (6) eventual shortening of the mouth opening to a slit overhung by the hypostome; (7) origination of the prosomal shield from the anterior margin of the pre-cheliceral lobes and the tergal portions of the four posterior-most prosomal segments; and (8) expansion of a ‘ventral sulcus’ from the cheliceral to the fifth opisthosomal segment separating the sides of these segments. Embryonic features are compared across the Chelicerata and discussed briefly in a phylogenetic context. Ó 2009 Elsevier Ltd. All rights reserved. Keywords: Prosoma Spider embryos Chelicerata Phylogeny 1. Introduction Crown-group chelicerates (¼Euchelicerata, following the system proposed by Chen et al., 2004) have a specific body tagmosis not found in other euarthropods (trilobites, crustaceans, myriapods and insects): the anterior region, prosoma, comprises seven segments, and the posterior region, opisthosoma, 13 segments. The prosoma encompasses the ocular segment, and the segments carrying chelicerae, pedipalps and the other four pairs of limbs. Like the head of other arthropods, the prosoma is a character-rich region for reconstructing relationships, in this case amongst chelicerates (Anderson, 1973, p. 442). Early embryos show striking similarities between the prosoma of chelicerates and the head of other arthropods (Anderson, 1973). This provides an opportunity to compare the developing morphology and topology of structures of the anterior region of the embryos across arthropods. These embryological data can contribute to our understanding of arthropod phylogeny. Among the crown-group chelicerates, embryos of spiders have played an important role in recent studies about arthropod * Corresponding author. Tel.: þ49 731 503 1006; fax: þ49 731 503 1009 E-mail address: yu.liu@uni-ulm.de (Y. Liu). 1467-8039/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.asd.2009.04.001 evolution. For example, numerous molecular studies on the embryos of Cupiennius salei Keyserling, 1877 (Ctenidae, Entelegynae) make this animal a well-established model species (Stollewerk et al., 2001; Damen, 2002; Schoppmeier and Damen, 2005; Janssen et al., 2008; Stollewerk and Seyfarth, 2008; McGregor et al., 2008). Historically, basic studies on the embryonic development of spiders, focusing on morphogenesis, have been carried out since the 19th century (e.g. Herold, 1824; Claparède, 1862). This early research remained of interest until well into the second half of the 20th century (see Anderson, 1973). Surprisingly, subsequent to the review of Anderson (1973), interest in this issue declined significantly with the exception of a brief paper by Marechal (1994) that paid special attention to the visual system of a mygalomorph spider, Ischnothele guyanensis (Walckenaer, 1837). Existing knowledge about spider embryonic development can be briefly summarised as follows. Once cleavage has finished, the yolk-rich spider eggs enter a blastula stage, at which most of the blastoderm cells begin to converge at one side of the egg to form the germ disc (Holm, 1952, 1954; Rempel, 1957; Suzuki and Kondo, 1995). The centre of the germ disc then has a bulged appearance due to the development of the caudal cumulus underneath the germ cells (Holm, 1952, 1954; Rempel, 1957; Suzuki and Kondo, 1995; Akiyama-Oda and Oda, 2003). The caudal cumulus later on migrates to the periphery of the egg and its migrating route finally 402 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 splits the germ disc (Akiyama-Oda and Oda, 2003). The two separated parts of the germ disc migrate away from each other until the whole germinal area is band-like in shape (Holm, 1952, 1954; Rempel, 1957; Suzuki and Kondo, 1995; Akiyama-Oda and Oda, 2003). Thereafter, the germinal area is called germ band. The germ band elongates itself around the egg and stops elongating when its anterior and posterior ends almost touch (Holm, 1952; Rempel, 1957; Chaw et al., 2007; McGregor et al., 2008). In some species, the distance between the two ends is not that short, because the second half of the opisthosoma does not attach to the egg surface but folds anteriorly (e.g. Holm, 1940; Yoshikura, 1955). At this time in development, a process called ‘inversion’ (or reversion) begins. This process occurs in most spiders except those of the Mesothelae and certain species from the Orthognatha (Foelix, 1996, p. 216). The inversion in spider embryology means that the germ band, which is initially situated on one side of an egg, is separated into two halves longitudinally by a furrow appearing in the midline of the germ band. The furrow was termed a ‘ventral sulcus’ by various embryologists (see Anderson, 1973). The two halves of the germ band are then pushed away from each other by the widening of the ventral sulcus, and they finally reach a position on the lateral side of the egg (Holm, 1952, 1954; Rempel, 1957; Yoshikura, 1954, 1955, 1958, 1961, 1972; Suzuki and Kondo, 1995). Thereafter, a thin epidermal layer starts to grow on the ‘dorsal’ side of the egg to connect the two halves of the germ band. This process is called ‘dorsal closure’, at the end of which the body form of the hatchling, resp. the adult spider emerges (Holm, 1940, 1952, 1954; Rempel, 1957; Yoshikura, 1954, 1955, 1958, 1961, 1972). In order to understand the development of the prosoma of chelicerates, and to compare it with the head formation of Fig. 1. Latrodectus geometricus Koch, 1841. A. Arrangement of a Step 5 embryo growing around the egg. B. Scheme demonstrating the range of angles covered by SEM photographs obtained from the different parts of the embryo. Labels C–E correspond to the subsequent photographs. C. Anterior area ranging from posterior tip of the telson to the walking limb L2. D. Mid-level area ranging from the hypostome to the walking limb L4. E. Posterior area ranging from the walking limb L4 to the tip of the hypostome. Scale bars ¼ 100 mm. Abbreviations: CH: chelicerae, e: pedipalpal endite, EP: epithelial portion, HY: hypostome, L1–4: walking limbs 1–4, M: mouth opening, OLB1–4: opisthosomal limb buds 1–4, PCL: pre-cheliceral lobes, PP: pedipalp, TE: telson, VS: ventral sulcus, Y: yolk. Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 other arthropods, we employed scanning electron microscopy to investigate embryos of the grey widow spider Latrodectus geometricus Koch, 1841 (Theridiidae, Araneae) at an ultrastructural level. Several details found herein are first documented for spider embryos: (1) a ventro-median bridge connecting the pre-cheliceral lobes and its development; (2) three pores being arranged like the mathematical ‘because’ sign (q) in the prospective mouth region; and (3) the prosomal shield is developed from the fusion of the anterior margin of the pre-cheliceral lobes and the tergal portions of the four posterior-most prosomal segments. We discuss our data in the light of earlier investigations on the embryogenesis within the Chelicerata. Common features found in early embryos of various chelicerates are further discussed. Such discussion provides preliminary information for mapping the embryonic features into ground pattern conditions, as what has been briefly carried out by Anderson (1973) and Yoshikura (1975). 2. Materials and methods Living material of Latrodectus geometricus was kindly provided by Martin Thierer-Lutz from the born to be eaten Insektenzucht GmbH, Schnürpflingen, Germany. The selection of this species is mainly based on three reasons: (1) easy lab culture of the adults; (2) the large amount of eggs produced by the females all through the year; and (3) the embryonic development of this species has never been described before. Eggs were removed from the cocoons and fixed with Carnoy’s fixative (100% alcohol: chloroform: glacial acetic acid ¼ 6:3:1) for 30 min. The fixed samples were then rinsed two times in 90% ethanol. After the fixation, the outer chorion was, in most cases, ruptured so that it could easily be removed from the yolk resp. embryo by gentle pipetting. The inner vitelline membrane was removed manually in 0.1 M phosphate-buffered saline (PBS) by using tissue forceps under a binocular. No postfixation was applied. The dissected embryos were dehydrated through a graded series of ethanol. The samples were incubated in hexamethyldisilazane for 10 min and then air-dried overnight (Nation, 1983a,b; Giammara et al., 1987). After mounting on stubs with black wax, the samples were sputter-coated with a mixture of 403 gold and palladium, and observed under a scanning electron microscope (SEM; Zeiss DSM 962). The embryo of Latrodectus geometricus grows around the egg such that the anterior and posterior ends approach each other (Fig. 1A). The provided diagram (Fig. 1B) includes the three standard angles that we used to document photographically the anterior, median and posterior parts, respectively (Figs. 1C–E). Of these, we focus on the anterior view, as in Fig. 1C. The digital SEM images obtained were trimmed in Adobe PhotoshopÔ and arranged into plates in Adobe IllustratorÔ. We adopted mainly the terminology and staging of spider embryos introduced by Anderson (1973), Suzuki and Kondo (1995), Akiyama-Oda and Oda (2003) and Chaw et al. (2007) for embryonic terms and general chelicerate/arachnid and arthropod terminology. We use the term ‘step’ for describing every single specimen that was fixed at a certain moment of development. The term ‘stage’ is avoided because it represents a period of time. Concerning the phylogeny within Arachnida we follow the system put forward by Ax (1999), in which the taxon Scorpionida is the sister taxon to the remaining arachnids, the Lipoctena (sensu Weygoldt and Paulus, 1979). 3. Results The current work mainly presents investigations on the postgerm disc development of the anterior region of the widow spider embryos (Figs. 2A and 3–5; see the supplementary data for brief observations on the germ disc phase). The dorsal side of the embryos at selected steps is presented in order to document the formation of the prosomal shield (Fig. 5). Two diagrams are provided to explain the details (Figs. 6 and 7). In total, 16 embryonic steps are recognised with significant changes between consecutive steps (a detailed description of each step identified is included in the supplementary data). Embryonic development of the widow spider results in a hatchling (Fig. 8). We herein provide a summarised description of the major events occurring in the early development of the anterior body region of Latrodectus geometricus. A pair of lobe-like structures comprises the anterior-most part of the spider embryo. We herein use the term ‘pre-cheliceral lobes’ for Fig. 2. Latrodectus geometricus Koch, 1841. SEM pictures of a Step 1 embryo. A. Antero-lateral view demonstrating the pre-cheliceral lobes (PCL) and the following limb buds (CH, PP, L1, L2). B. Postero-lateral view demonstrating the last four pairs of prosomal limbs (L1–4) and the opisthosoma (OP). Note that the walking limb L4 is presented as a single swelling, from which the two limb buds will differentiate at the next step. Abbreviations other than used in the previous figure: An: anterior. 404 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 highlighting the location of those structures. The development of the pre-cheliceral lobes starts with the occurrence of an initially smooth area (Figs. 2A, 3A,B and 6A,B), which changes into an area with distinct structures such as median and lateral subdivisions (ms, ls in Figs. 3, 4 and 6; marked in yellow in Fig. 6; see also Table 1) and three pairs of furrows appearing successively in the central area (a, b, c in Figs. 3, 4 and 6; Table 1). Subsequently, the two subdivisions merge into each other to form the single central area of each pre-cheliceral lobe (yellow in Fig. 6). The central area later recesses or sinks underneath the anterior margin (AM in Figs. 3–6) of each precheliceral lobe through the former anterior furrow (a in Figs. 3, 4 and 6), and then becomes internalised. Eventually, the entire central area vanishes from the surface of the embryo (Figs. 4D–H and 6L–P; Table 1). The anterior margin becomes more and more prominent, and it folds slightly dorsally and posteriorly. It eventually fuses with the developing dorsal (tergal) area of the four posterior-most prosomal segments, and forms the anterior part of the prosomal shield (Fig. 5; Table 1; see the following for more details). In the earliest recognised step, only pre-cheliceral lobes are present as the anterior-most structures (Figs. 2A and 3A). A smooth sub-rectangular area between the lobes appears at the next step (Fig. 3B, red in Fig. 6B; see also Table 2). This area has a significant differentiation during later development (red in Figs. 6C–H). One feature is that a shallow depression first appears with three small pores (Figs. 3C, 6C and 7B), which are bordered anteriorly by a pair of swellings, the prospective hypostome (HY in Figs. 3C and 7B). Subsequently, the depression deepens to form the mouth opening, while the prospective hypostome rises to form a uniform lobe eventually. The posterior border of the mouth opening develops into a lip-like ridge (Figs. 3D–G, 6D–G and 7C,D). The mouth opening becomes relatively shorter, and widens to a slit. Together with the posterior ridge around it, the mouth opening becomes overhung by the developing hypostome (Figs. 3H, 6H and 7E). The hypostome eventually represents the only visible structure left of the entire initial mouth area (red in Figs. 6H–P). We did not detect the fate of the three pores after Step 4 or Step 5 (Figs. 3D and 6D), so we have no clues yet of their structural nature and function. In summary, in L. geometricus, a mouth area does not develop until the germ band has reached its full length and until the buds of the prosomal limbs have all appeared (Figs. 3A,B and 6A,B). Therefore we term the smooth area of Step 2 ‘initial mouth area’. This area and the three pores within the initial mouth region have rarely drawn attention before. Dorsally, after the central area of each pre-cheliceral lobe has internalised, the marginal area represents the major remains of each lobe (Fig. 4D, green in Fig. 6L). As the anterior part of the marginal area, i.e. the anterior margin grows dorsally and posteriorly, its anterior rim starts to get fused with the developing tergal 405 area of the four posterior-most prosomal segments (compare the open arrowhead in Figs. 5C–F). This fusion results in a shield covering the entire prosomal region. Hence, the shield is called prosomal shield, of which the anterior-most part is developed from the anterior margin of the former pre-cheliceral lobes. On the other hand, the posterior rim of the pre-cheliceral lobes secondarily develops into the anterior border of the prosomal shield (compare the black arrowhead in Figs. 5A–F; see also Figs. 3C–H, 4 and 6C–P). The complete outline of the prosomal shield is visible from Step 12 on (Figs. 4D, 5D and 6L). The tergal area of the cheliceral and pedipalpal segments is undetectable due to the forward shift of the attaching position of those appendages. In summary, formation of the prosomal shield of L. geometricus is completed before the embryonic phase ends. The prosomal shield develops from two different structures: (1) the anterior margin (AM in Figs. 3–6) of the pre-cheliceral lobes; and (2) the tergal area of the four posteriormost prosomal segments. A longitudinal furrow is present at the ventral midline, labelled ventral sulcus by many authors (e.g. Anderson, 1973), and it is present from Step 2 to Step 14 (Figs. 3B–H, 4A–F and 6B–N; see also Table 3). From the cheliceral segment to about the fourth or fifth opisthosomal segment, the embryonic body primordium is divided into two halves longitudinally by the ventral sulcus at the ventral midline (Figs. 1C–E). As the two separate halves of the embryonic median body primordium migrate to the sides, the width of the ventral sulcus is increasing, being widest at the border between prosoma and opisthosoma. In parallel, the anterior and posterior ends of the embryo are drifting apart from each other passively as a result of the width expansion of the ventral sulcus, since the length of the embryonic body primordium does not increase throughout the entire post-germ disc phase (Figs. 5A–C). At this step, the differentiation of the prosoma and the opisthosoma can be recognised by the fact that on each half of the germ band, the area at the boundary between the prosoma and the opisthosoma has been narrowed much more than the remaining area. The area at the boundary between the prosoma and the opisthosoma later on becomes a thin stalk bridging these two tagmata. A single ventral sternal (cuticular) plate with a smooth surface begins to form at the time when the two halves of the embryonic body primordium have been pushed sideways by the widening of the ventral sulcus to eventually reach their final positions. This formation occurs through an expansion of the ventral epithelial portion of the four posterior-most prosomal segments (EP in Figs. 4C,D, see also Figs. 4E–G and 6M–P). Ventral epithelial portions (EP in Figs. 3 and 6B) of the cheliceral and pedipalpal segments are visible at Step 2–Step 8 (Figs. 3B–H and 6B–H), and almost undetectable in later steps (e.g. Figs. 4A–H and 6I–P). There are many modifications of the ventral inter-limb area among the extant Fig. 3. Latrodectus geometricus Koch, 1841. SEM pictures of Step 1–Step 8 embryos. Anterior view. A. Step 1 embryo: two pre-cheliceral lobes (PCL) with smooth surface are separated by a concave region (open triangle), all limb buds are located posterior to the pre-cheliceral lobes, the buds of each limb pair approaching each other basally (asterisks); B. Step 2 embryo: enlarged concave region (open triangle) between the pre-cheliceral lobes (PCL). The ventral sulcus (VS) and epithelial portion (EP) keep the limb buds much further apart than before. C. Step 3 embryo: two subdivisions (ms, ls; highlighted also by white solid lines) differentiate in the central area of each pre-cheliceral lobe. A mouth region appears with the prospective hypostome (HY) anteriorly, three small pores (white arrowheads) and a U-shaped swelling posteriorly. The ventral sulcus starts to widen. D. Step 4 embryo: the anterior margin (AM) starts to differentiate due to the development of an anterior furrow. (a) A lateral furrow (b) is present in each pre-cheliceral lobe. The hypostome (HY) has a median furrow (white arrowhead) indicating its bilobed appearance. The area bearing three pores recesses within a single mouth opening (inset: magnified image from another specimen). E. Step 5 embryo: in each pre-cheliceral lobe, the two furrows (a, b) deepen, the median subdivisions (ms) bulge, while the lateral subdivisions (ls) flatten. The hypostome (HY) loses its bilobed appearance. F. Step 6 embryo: in each pre-cheliceral lobe, the anterior furrow (a) deepens while the lateral furrow becomes narrower than in the previous step. Note that up to this step the chelicerae (CH) are located posterior to the mouth opening (M). G. Step 7 embryo: both subdivisions in each pre-cheliceral lobe flatten. The chelicerae (CH) are para-oral. The hypostome (HY) has a straight posterior border, and the mouth opening (M) is slit-like. On the outer side of the basal portion of each pedipalp, note the newly developed cone-like projection (white circle, hypothesised to penetrate the egg shell at hatching, see Discussion). H. Step 8 embryo: in each pre-cheliceral lobe, the anterior furrow (a) is groove-like, while the lateral furrow (b) is elongated. The anterior margin (AM) is more prominent than in the previous steps (compare ‘AM’ in D–H). A newly developed furrow (c) is situated posterior to the median subdivision (ms). The mouth opening (M) is overhung by the hypostome (HY). The ventral sulcus is wider than in all previous steps (compare ‘VS’ in C–H). In all pictures, the dashed line marks the posterior margin of one of the pre-cheliceral lobes; the open star labels the pre-hypostomal region (the star in F also labels the pre-hypostomal region, but the vertical ridge here is interpreted as a shrinkage artefact); black arrows mark the paired pores close to the base of the chelicerae; white arrows point to the inner borders of the ventral sulcus; white (open) and black arrowheads mark the anterior and posterior rim, respectively, of the (prospective) anterior margin (AM) of the left pre-cheliceral lobe. Scale bars ¼ 100 mm (inset in (D) ¼ 10 mm). 406 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 arachnid taxa (see, e.g., Siewing, 1985, his Fig. 841; Westheide and Rieger, 1996, their Fig. 619). Due to insufficient information about the embryonic development of some arachnid taxa, it is difficult to draw any phylogenetic conclusions about the formation of the sternites or sternum in arachnids at the moment. The six pairs of prosomal limb buds together with the precheliceral lobes are the first recognizable external structures in early embryos of Latrodectus geometricus (Figs. 2 and 3A). At the earliest recognised step herein for L. geometricus, the left and right buds of the first five pairs of prosomal limbs are very closely apposed (Figs. 2A, 3A and 6A). The two limb buds of the sixth (¼the last) prosomal limb pair (L4 in Fig. 2B) are undifferentiated from a single swelling. At the next step, the two limb buds are clearly developed. The distance between the left and right buds of all prosomal limbs enlarges due to the development of the inter-limb sternal region of the segments and the widening of the ventral sulcus. In parallel, the limb buds of the pedipalps and following walking legs become progressively more elongated and segmented (Figs. 3B–H and 4). An endite starts to develop on the inner side of the basal (¼basipodal) podomere of the pedipalps from Step 3 onwards (Figs. 3C and 6C). All limb buds are initially post-oral (Figs. 2A and 3A), and this situation does not change until Step 7, at which the chelicerae are located on the sides of the mouth opening (Fig. 3G). In the next steps, the chelicerae are located anterior to the hypostome, and even the pedipalps finally reach a position anterior to the mouth opening (e.g. Figs. 4D–H; see also Section 4.1.2). A pair of small cone-like structures develops on the outer side of the pedipalpal podomere carrying the endite from Step 7 onwards (Figs. 3G–H, 4 and 6G–P). This structure is considered an egg tooth and is probably used by the animal to penetrate the eggshell when hatching (cf. Kästner, 1949; Yoshikura, 1955, p. 22). The egg tooth is not present in any of the hatchlings of L. geometricus (unpublished data). It is obvious that a pair of small pores (black arrow in Figs. 3D,E,G) develops close to the base of, and anterior to the chelicerae from Step 4 to Step 7. No earlier investigation has documented these pores. The functional significance of the pores remains unclear. The developmental events described above result in a hatchling (Fig. 8). At this step, the dorsal closure has been finished. The opisthosoma seems to be swollen with yolk (Fig. 8A). In the anterior region of the prosoma, the chelicerae are closely situated in front of the hypostome. The mouthparts form a pre-oral tube consisting of the hypostome anteriorly, the so-called labium posteriorly, and the pedipalpal endites laterally (Fig. 8B). A large and distinctive sternum is present posterior to the labium (Fig. 8A). 4. Discussion 4.1. Embryonic development of Latrodectus geometricus 4.1.1. Pre-cheliceral lobes In embryos of L. geometricus, each of the pre-cheliceral lobes can be recognised as two areas: (1) a marginal region that later on 407 becomes the anterior part of the prosomal shield; and (2) a central area that is subdivided into two portions that later fuse and separate from the marginal region by three furrows. The central area of each pre-cheliceral lobe finally gets internalised. Similar findings have been reported and discussed by several previous authors. Pross (1966) recognised two pairs of somites in front of the chelicerae in the embryos of the lycosid spider Pardosa hortensis Thorell, 1872. He further suggested the two pairs of somites to be two pre-cheliceral segments: the protocerebral and the deutocerebral segments (Pross, 1966, his Figs. 25a–c). Subsequently, the cheliceral segment was considered as being tritocerebral. The findings of Pross (1966) match well with our discovery of the two subdivisions in the central area of each pre-cheliceral lobe of the widow spider embryos. In histological sections, the central area (¼the subdivisions) of the pre-cheliceral lobes appears to give rise to pre-cheliceral ganglia (Figs. 20–25 in Yoshikura, 1955, 1958; Figs. 4a,b in Marechal, 1994) or nerve components (Kästner, 1950). Based exclusively on the existence of the neuro-related subdivisions, we hesitate to consider those subdivisions as representatives of segments, which normally contain also muscles and appendages (in most cases). Interestingly, two subdivisions of the central area of each pre-antennal lobe, visualised by the expression of the neurorelated genes Pax6.1 and Pax6.2, are reported in the embryos of a diplopod (Prpic, 2005). Three subdivisions were also found in the same region of collembolan (Uemiya and Ando, 1987) and thysanuran embryos (Larink, 1970; Machida, 1981). Even more, Kishimoto and Ando (1985) reported four subdivisions in the preantennal lobes of the embryos of a stonefly. In all, it turns out to be difficult to compare and homologise segments and their appendages in the anterior region of arthropods, if each subdivision found in the pre-cheliceral/pre-antennal lobes of different arthropod embryos is regarded as representing one segment. We herein intend to homologise the pre-cheliceral and pre-antennal lobes of various arthropod embryos as the most parsimonious explanation. Recently, it could be demonstrated by different methods, such as the neuro-system staining techniques, palaeontological approaches and hox gene expressions, that the chelicerae of chelicerates are homologous to the antennulae or first antennae in all the rest of the euarthropods, and both are innervated by the second neuromere, the deutocerebrum (e.g. Mittmann and Scholtz, 2003; Chen et al., 2004; Jager et al., 2006). The anterior furrow (a in Figs. 3D–H, 4A–C, 5B,C and 6D–K) has been found to be significant for the brain formation of a spider. According to the histological sections presented by Yoshikura (1955, his Figs. 20–25), the anterior furrow (‘cerebral groove’ in Yoshikura, 1955, 1958) develops into a deep depression through which the central area (¼the subdivisions) of the pre-cheliceral lobes ‘sinks’ into the yolk, the internalisation process. Another two furrows (b, c in Figs. 3–6) in each pre-cheliceral lobe help laterally and posteriorly isolating the central area from the marginal region of the pre-cheliceral lobes. A similar internalisation can also be observed in other spiders (Yoshikura, 1955, 1958, 1972) and other Fig. 4. Latrodectus geometricus Koch, 1841. SEM pictures of Step 9–Step 16 embryos. Anterior view. A. Step 9 embryo: the median and lateral subdivisions (ms, ls) of each precheliceral lobe start to merge into each other. B. Step 10 embryo: the anterior margin (AM) of each pre-cheliceral lobe is much more differentiated than in all previous steps. C. Step 11 embryo: the width of the anterior part of the pre-cheliceral region is reduced. A ridge (white arrowhead) is located between the pre-cheliceral lobes. D. Step 12 embryo: the entire pre-cheliceral region is highly reduced in size (the white arrowhead points to the highly reduced median ridge). The outline of the prospective prosomal shield is marked by a dashed line (see also Fig. 5D). Note that the former central area (ms þ ls) of each pre-cheliceral lobe has internalised and only triangular holes are left as the remains. E. Step 13 embryo: the ridge (white arrowhead) between the two pre-cheliceral lobes is highly reduced. Two slits represent the remains of the internalised central area (ms þ ls). A prospective sternum (pSTN) is developed from the growth of the former epithelial portion (see EP in C, D). F. Step 14 embryo: the anterior margins (AM) of the pre-cheliceral lobes medially fuse with each other. The remains (ms þ ls) of the internalised central area of each pre-cheliceral lobe are hole-like. G. Step 15 embryo: the hole-like remains (ms þ ls) of the internalised central area of each pre-cheliceral lobe are smaller than those in the preceding step. Two circular regions (asterisk) are situated laterally to the medially fused anterior margins (AM) of the pre-cheliceral lobes. A labium (LB) starts to develop at the anterior end of the prospective sternum (pSTN). H. Step 16 embryo: the prosomal shield is well-developed, with a clearly defined anterior border. The cone-like projection is still present on the outer side of the basal portion of each pedipalp. In all pictures, white arrows point to the borders of the ventral sulcus; white (open) and black arrowheads mark the anterior and posterior rim, respectively, of the anterior margin of the left pre-cheliceral lobe; white circles mark the cone-like structure developed on the outer side of the basal portion of each pedipalp. Note that the cone-like structures (white circles; cf. Fig. 3H) do not change their morphology from Step 9 to Step 16. Scale bars ¼ 100 mm. Abbreviations other than used in the previous figure: pSTN: prospective sternum. Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 arachnids (Kästner, 1948, 1949, 1950; Yoshikura, 1961; Weygoldt, 1975; Moritz, 1957; Brauer, 1895; Farley, 1998, 2001; Thomas and Telford, 1999), although the anterior furrow was termed differently, e.g. ‘fovea’ (Rempel, 1957), ‘semilunar groove’ (Holm, 1940, 1954; Farley, 1998, 2001; adopted also by Anderson, 1973), and ‘cerebral groove’ (Yoshikura, 1954, 1955). The lateral furrow (b in Figs. 3–6) was observed in Ischnothele guyanensis and termed as ‘lateral vesicles’ by Marechal (1994), while the posterior furrow (c in Figs. 3, 4 and 6) has never been mentioned before. In the embryos of Latrodectus geometricus, a sub-triangular median area, the postero-median bridge (light green in Fig. 6) connects the pre-cheliceral lobes (dark green in Fig. 6). This bridge retains a medial connection after the appearance of the mouth area (red in Fig. 6), but later on it is separated by the development of the mouth region and vanishes eventually (Fig. 6). The reduction of the bridge indicates the shift of the mouth region from anterior to posterior. Unfortunately, information about the bridge is so rare in existing works that further comparison cannot be made for the moment. Many small pore-like depressions distributed all over the surface of the pre-cheliceral lobes (Figs. 3D,G) have also been found on the surface of not only the pre-cheliceral lobes but also the area between the two buds of the same appendages that belong to the more posterior segments in early embryos of a scorpion (Farley, 2001, his Figs. 9–11). Those depressions were thought to suggest ‘invagination of epithelial primordia for internal structures’ (Farley, 2001, p. 73). Within the Lipoctena, data from Palpigradi, Solifugae and Ricinulei, are still missing. Information from the Pantopoda (e.g. Morgan, 1891) and the Xiphosura (e.g. Sekiguchi et al., 1982) are not sufficient to confirm the presence or absence of the structures and processes discussed above. Therefore, it can only be suggested that the subdivisions, the internalisation, and the associated furrows of the pre-cheliceral lobes belong to the ground pattern of, at least, the Arachnida. 4.1.2. Mouth region A structure that refers to what we call ‘hypostome’ in L. geometricus has been observed by all other investigators of chelicerate embryology, such as Farley (1998, 2001) in a scorpion, who took the term ‘rostrum’ for it after Yoshikura (1955) and Foelix (1996). Another term, ‘labrum’, was largely applied to our ‘hypostome’, mainly by zoologists or embryologists working on extant animals (e.g., Anderson, 1973). ‘Hypostome’ is commonly found in the palaeontological literature, established there for the pre-oral sclerotic plate between the antennulae of, e.g., Trilobita. The detailed data of mainly the three-dimensionally preserved Cambrian ‘Orsten’ arthropods (summary on the ‘Orsten’ in Maas et al., 2006) demonstrated, however, the distinction between the hypostome and the labrum on the ventral side of the arthropod head (e.g., Walossek and Müller, 1990). In fact, a ‘labrum’ can be understood as an expansion of the mouth membrane at the rear end of the (also present) hypostome. Furthermore, a labrum develops only in the in-group crustacean taxon Labrophora, hence, is proposed as one of the autapomorphies of this taxon (Maas et al., 2003; Siveter et al., 2003). Indeed stem taxa of the Crustacea ‘‘still’’ have the ‘hypostome’ only, e.g. the ‘Orsten’ forms Goticaris longispinosa Walossek and Müller, 409 1990 and Cambropachycope clarksoni Walossek and Müller, 1990 (both re-studied in Haug et al., in press), Oelandocaris oelandica Müller, 1983 (re-studied in Stein et al., 2008) and Henningsmoenicaris scutula Walossek and Müller, 1990 (re-studied in Haug et al., pers. comm). Therefore, ‘hypostome’ and ‘labrum’ are not equivalent or interchangeable, do not exclude each other, but are two different structures next to each other, in a particular euarthropod taxon only (e.g., Waloszek, 2003). In consequence, the terms ‘hypostome’ and ‘labrum’ need to be treated with care since the original terminology might be as much misleading as described structures with the same name may not be homologous (see also Maas et al., 2003; Scholtz and Edgecombe, 2005, 2006). Excluding the nature of the hypostome, the question remains what it actually is and belongs to. Our data suggest that it is a structure appearing and developing in the mouth region, hence, not part of the pre-cheliceral lobes (¼ocular or protocerebral segment), neither the sternal (¼inter-appendage) region of the antennular (¼cheliceral or deuterocerebral) segment. Its development starting after all the prosomal limb buds have developed and the lack of expression of neuro-related genes Pax6.1 and Pax6.2 in the hypostome of Glomeris marginata (Villers, 1789) embryos (Prpic, 2005) are arguments against the hypothesis raised by some authors considering the hypostome as a pair of appendages (e.g., Butt, 1960; Haas et al., 2001a,b; Boyan et al., 2002, 2003 [considering the hypostome to be related to the tritocerebral segment ¼ pedipalp segment of chelicerates]; Kimm and Prpic, 2006). The increasing distance between the hypostome and the anterior margin of the pre-cheliceral lobes suggest that the whole mouth region changes its position in ventral and posterior direction during embryonic development (Figs. 3B–H, 4A–C and 6B–K). This gives the impression that the attaching position of the chelicerae and the pedipalps ‘move’ from being post-orally located (Figs. 6B–G) to their final positions of being anterior to the hypostome and parallel to the mouth opening, respectively (Figs. 6L–P and 8). It is possible that, together with cell proliferation and apoptosis during the inversion (Prpic and Damen, 2005), the internalisation of the central area (¼the subdivisions) of the pre-cheliceral lobes motors the change of the positions of the mouth region and the prosomal appendages. A similar development can also be observed in the embryos of Limulus polyphemus Linnaeus, 1758 (Mittmann, 2004), two scorpions (Brauer, 1895; Farley, 1998, 2001), a mite (Thomas and Telford, 1999), and a pantopod (Morgan, 1891, his pl. IV, Figs. I and II). However, unfortunately, not much attention has been paid to the position changes. We conclude that the shift of the entire mouth region from anterior to posterior belongs to the ground pattern of, at least, the Chelicerata, or it is even more plesiomorphic than this level. Information about the presence of the three small pores and the latero-posterior border of the mouth opening cannot be found in earlier literature on other chelicerates. Therefore, further phylogenetic comparison across the Chelicerata in this specific respect cannot be made. 4.1.3. Formation of the prosomal shield In embryos of L. geometricus, the prosomal shield consists of the marginal region of the previous pre-cheliceral lobes anteriorly and the tergal area of the four posterior-most prosomal segments. Fig. 5. Latrodectus geometricus Koch, 1841. SEM pictures of embryos at selected steps. Dorsal view. A. Step 3 embryo: the anterior and posterior ends of the embryo are approaching each other. B. Step 8 embryo: the anterior and posterior ends of the embryo are much more separated than before. C. Step 10 embryo: the anterior and posterior ends of the embryo have the maximum distance from each other. D. Step 12 embryo: the outline of the developing prosomal shield is marked by a dashed line. E. Step 14 embryo: the anterior margin (AM) of the pre-cheliceral lobes starts to become the anterior-most part of the prosomal shield through the fusion of its anterior rim (white open arrowhead) with the tergal area of the four posterior-most prosomal segments (compare A–E). The posterior rim (black arrowhead) of the anterior margin (AM) of each pre-cheliceral lobe represents the anterior rim of the entire prosomal shield (compare A–E). F. Step 15 embryo: the formation of the prosomal shield completes, with the anterior margin (AM) of the pre-cheliceral lobes being the anterior-most part of the prosomal shield. The anterior rim (white open arrowhead) of the former anterior margin (AM) of the pre-cheliceral lobes is almost undistinguishable. The open arrow in E and F points to the posterior border of the prosomal shield. Other labels correspond to those in Fig. 4. Scale bars ¼ 100 mm. 410 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Fig. 6. A–P. Diagrams of embryos of Latrodectus geometricus in the developmental steps following the germ disc phase. Each drawing represents one of the steps (1–16) described earlier with SEM pictures. Meaning of the colours: dark green – the pre-cheliceral lobes in A, B, and the marginal region of the lobes in C–P when the differentiation of the central area (¼ms þ ls, marked as yellow) is present; light green – the postero-median bridge connecting the pre-cheliceral lobes; dark yellow – the furrows around the central area of the pre-cheliceral lobes; red – the mouth region; orange – the chelicerae; light blue – the pedipalps; grey – the tergite area of the four posterior-most prosomal segments. In each image, arrows mark major structures of each step; open arrowhead labels the anterior rim of the anterior margin (AM) of the pre-cheliceral lobes, corresponding to the white open arrowheads in Figs. 3–5. Other labels and abbreviations correspond to those in Figs. 3–5. Scale bars ¼ 100 mm. Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 411 Fig. 7. Diagrams of the mouth region of the embryos of Latrodectus geometricus Koch, 1841. A. Step 2 embryo: the mouth region is undifferentiated. B. Step 3 embryo: a pair of lobelike structures, the hypostome (HY), develops in the frontal part of the mouth region. Three pores develop posterior to the hypostome (HY). A U-shaped swelling, the lower lip, develops posterior to the pores. C. Step 4 embryo: the hypostome (HY) and the lower lip approach each other to form the mouth opening that encloses the three pores. Note that the hypostome (HY) still has a bilobed appearance. D. Step 7 embryo: the hypostome (HY) has lost its bilobed appearance. The mouth opening is slit-like, due to the approaching of the hypostome (HY) and the lower lip. E. Step 8 embryo: the mouth opening and the lower lip is overhung by the hypostome (HY), hence are invisible. Similar processes of the prosomal shield formation to what we report have also been documented for other Araneae (Yoshikura, 1954, 1955; Holm, 1940, 1954) and other arachnids such as uropygids (Kästner, 1948, 1949, 1950; Yoshikura, 1961), harvestmen (Moritz, 1957), amblypygids (Weygoldt, 1975), and scorpions (Farley, 1998, his Figs. 21 and 23). To our knowledge, comparable information has not yet been yielded for the Xiphosura and Pantopoda. Phylogenetically, the existing data indicate that the prosomal shield consisting of the anterior margin of the pre-cheliceral lobes and the tergite regions of the last four prosomal segments, and the developmental processes of the prosomal shield can be considered to be part of the ground pattern of, at least, the Arachnida. 4.1.4. Ventral sulcus and inversion The ventral sulcus and its correlated inversion process observed in embryos of L. geometricus are very common in various lipoctenans (Weygoldt, 1975; Yoshikura, 1961; Kästner, 1948, 1949, 1950; Moritz, 1957; Anderson, 1973; Thomas and Telford, 1999) with exceptional absence only in some spiders from the Mesothelae and Orthognatha (Foelix, 1996). Staining and sectioning methods on embryos of Cupiennius salei have demonstrated that neural commissures of the appendage-bearing segments are present in this area (unpublished data from Linne and Stollewerk, personal communication; see also Stollewerk et al., 2001). The ventral sulcus and the inversion previously found in other species are basically the same structure and phenomenon as what we document herein, with only very detailed differences. For example, Holm (1954) reported for the embryos of Ischnothele karschi Strand, 1907 that the ventral sulcus first appears as a median thinner area in the 7th and 8th segments, which later on extends anteriorly and posteriorly and finally divides the whole germ band into two lateral rows of segmental plates. In L. geometricus, however, according to our investigation, the ventral sulcus first appears in Step 2 embryos as a longitudinal furrow extending from the cheliceral segment back to the opisthosoma. Fig. 8. SEM pictures of a fresh hatchling of Latrodectus geometricus. A. Ventral view. B. Magnified view of the mouthparts of the specimen shown in (A). The chelicerae are located anterior to the hypostome (HY). Note the pre-oral tube formed by the hypostome (HY) anteriorly, the labium (LB) posteriorly, and the pedipalpal endites (e) laterally. Scale bars ¼ 100 mm. Abbreviations other than used in the previous figure: STN: sternum. 412 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Table 1 Summary of the development of the pre-cheliceral lobe (PCL) in Latrodectus geometricus embryogenesis. Step 0 A PMB 1 2 3 4 5 A B C D E 90 45 30 60 60 6 F 60 7 G 80 8 H – 9 I – present present present present median part shorter median part shorter median part absent completely absent – 10 J – – 11 K – – 12 L – – 13 M – – 14 15 N O – – – – 16 P – – Subdivisions of the central area a b c d d d shallow, elliptic deepened d d d d d d d median subdivision (ms) present ms more distinct than before, slightly raised ms distinctly humped d d lateral subdivision (ls) present d ls more distinct than before shallow, elliptic more flattened than ms crescentic ms even more distinct than before more distinct than before crescentic, deeper deeper d ms flattened again more flattened than before un width 50% d deep groove narrow, slit-like present un longer un un un un un un un indistinguishable indistinguishable indistinguishable – – – – – – – – – – – – ms separated from anterior margin (AM) of the un pre-cheliceral lobe (PCL) ms and ls merge to form a single, rounded triangular central area of each PCL, slightly smaller than before rounded triangular, smaller; AM partly slightly overhanging the recessed central area, particularly anteriorly central area more recessed than before, smaller, triangular, slightly more posteriorly and inwardly located against sharply rising ridge in the ventral midline central area invisible, only a small triangular hole remaining between AM and the sharply rising ridge, triangular area recessing underneath AM posterior margin of each PCL significantly shorter than before, forming a shallow valley with 2 slit-like holes separated by a thin ridge AM less sharply marked than in Step 13, holes rounded entire anterior area narrower and shorter than before, AM less clearly marked, holes smaller posterior margin of pre-cheliceral lobes small, AM being the anterior-most of prosomal shield, holes absent Columns: ‘‘step’’ – steps as described in text; ‘‘0’’ – letters corresponding to those of Figure 6; ‘‘A’’ – orientation of the long axis of the PCL relative to the long body axis; ‘‘PMB’’ – postero-median bridge; ‘‘a’’ – anterior furrow; ‘‘b’’ – lateral furrow; ‘‘c’’ – posterior furrow. Grey filling of boxes marks significant changes. Abbreviations: un ¼ unchanged; – ¼ inapplicable; d ¼ absent. In the mite Hyalomma dromedarii Koch, 1844 (Anderson, 1973) and the whip spider Tarantula marginemaculata Koch, 1840 (Weygoldt, 1975), the ventral sulcus was demonstrated as separating left and right parts of the embryonic body primordium behind the precheliceral lobes, but also even the pre-cheliceral lobes in very early embryos. The embryonic body primordium of L. geometricus completely curves around the yolk, i.e. attaching its dorsal side to the yolk (Figs. 1D,E). In the embryos of many other spiders and arachnids the terminal part of the opisthosoma posterior to the ventral sulcus folds ventrally, such as in Segestria bavarica Koch, 1843 (Haplogynae; Holm, 1940, Figs. 5 and 6), Ischnothele karschi Dönitz, 1887(Holm, 1954, Fig. 3; see also Anderson, 1973, Fig. 141d), Atypus karschi Dönitz, 1887 (Yoshikura, 1958, Figs. 6B,C), Ummidia fragaria Dönitz 1887 (Yoshikura, 1972, Figs. 2–3), Hyalomma dromedarii (Anderson, 1973, Figs. 148c–e), and Paruroctonus mesaensis Stahnke, 1957 (Farley, 2001, his Fig. 12). In xiphosuran embryos the terminal part of the opisthosoma does not fold; the embryos do not grow around the egg but develop on one side of the egg (Sekiguchi et al., 1982; Mittmann, 2004). In all, it seems that differences from our investigation on L. geometricus are due to the possibility of rather taxon-specific modifications of their development, even despite close phylogenetic relationship, such as in the araneids. Outside Lipoctena, studies by Brauer (1895) and Farley (2001) did not confirm the presence of a ventral sulcus for scorpion embryos. Also xiphosuran embryogenesis seems not to include the formation of a ventral sulcus (Sekiguchi et al., 1982; Mittmann, 2004). Data on pantopods are still needed. The embryonic development of the pre-cheliceral lobes and the existence of the ventral sulcus are unclear from the existing data on pseudoscorpions (Weygoldt, 1964). In the embryos of Archegozetes longisetosus Aoki, 1965 (Acari) the ventral sulcus is probably present (Thomas and Telford, 1999, their Fig. 5A). Further comparison cannot be made for the moment, mainly due to the difficulty in correlating the various studies, in which not only parts were omitted in the line drawings made by many previous authors. Anderson (1973), for example, termed the dorsal parts of the embryonic epidermis ‘dorsal extraembryonic ectoderm’ (embryonic cells outside the embryo is contradictory). Summing up, it is likely that the embryonic ventral sulcus is an autapomorphy in the ground pattern of the Lipoctena. Outside Chelicerata a ventral sulcus-correlated phenomenon has been reported for centipede embryos, e.g., Scolopocryptops rubiginosus Koch, 1878 (Scolopendromorpha, Myriapoda) by Sakuma and Machida (2004) and Sakuma (2005), and in Orchestia cavimana Heller, 1865 (Isopoda, Crustacea) by Ungerer and Wolff (2005). Considering the large phylogenetic gap among these taxa and focusing on the Chelicerata, assumptions of convergence appear to be most plausible to us. 4.1.5. Formation of the sternitic region and the nature of the labium In the hatchlings of L. geometricus, a triangular lobe (LB in Fig. 8B) seems to be articulated with the anterior part of the sternum and points against the hypostome. This structure is usually termed ‘labium’ by arachnologists (e.g., Foelix, 1996), and its development was mentioned by few previous authors, e.g., Yoshikura (1955). Its abaxial joint with the sternum is right at the posterior end of the insertions of the pedipalps. To our knowledge, no earlier literature has ever mentioned the development of the labium in spiders. According to our investigation, the labium first Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 413 Table 2 Summary of the development of the mouth region in Latrodectus geometricus embryogenesis. Step 0 Pre-hypostomal portion Hypostome (HY) Mouth opening Lip-like bulge lateral and posterior to mouth opening 1–2 3 A–B C d Smooth rectangle d Pair of slightly enhanced lobes connecting medially d Slightly bulged 4 D Triangle; length tripled from Step 3 Much raised, bilobed 5 E 2 triangles standing on the tips of each other Overhanging the mouth opening; bilobation absent d Present as a slightly depressed area surrounded by HY anteriorly and the lip-like bulge laterally and posteriorly 3 pores in the area being arranged as the mathematic because sign q More depressed and more enclosed by the approaching of HY and the lip-like bulge, 3 pores still present More depressed 3 pores not detected from this step onwards 6 7 F G Un Similar to Step 5–Step 6, but with a straight posterior border Un Slit-like, on level with the insertions of chelicerae Un Much less prominent 8 H 9 10 I J Un Posterior triangle more elongated than in Step 6, anterior one unchanged Posterior triangle more elongated than in Step 7, anterior triangle unchanged Indistinguishable – Slightly posterior to the insertions of chelicerae and covered by HY, not traced from this step onwards ? ? Covered by the hypostome, not traced from this step onwards ? ? 11 12 K L – – ? ? ? ? 13 M – ? ? 14–16 un – Lobe-like; with a convex tip pointing backwards and covering the mouth opening Un Located significantly posterior to the insertions of chelicerae Un Almost completely covered by the chelicerae Located on level with the insertions of pedipalps Un ? ? Slightly more elevated Un Columns: ‘‘step’’ – steps as described in text; ‘‘0’’ – letters corresponding to those of Fig. 6. Bold highlighted text marks significant changes. Abbreviations: Un ¼ unchanged; ? ¼ unknown condition; – ¼ in applicable; d ¼ absent. appears in the late embryogenesis (LB in Figs. 4G,H) as an anterior, pointed tip of the sternal plate, i.e. the labium derives from the sternal plate and, hence, has a cuticular nature. It is noteworthy that the term ‘labium’ is used differently in arthropod terminology. In entomology, ‘labium’ refers to the medially fused second maxillae in insects, hence has an appendicular origin. In crustacean terminology, ‘labium’ is used by some workers on specific in-group taxa, e.g. by ostracodologists, for the sternum as the fusion product of the sternites of the mandibular and maxillulary (1st maxilla) segments (cf. Maas et al., 2003; Maas Table 3 Summary of the development of the prosomal appendages and the ventral body region in Latrodectus geometricus embryogenesis. Step 0 Chelicerae (CH) Pedipalps (PP) Walking limbs 1–4 Ventral body region A 1 A Cone-shaped anlage 08 B Elongated, unsegmented Cellular band broader, ventral sulcus (VS) and sternal portions (ST) present Un 3 4 C D Ovoid buds with a bifid tip Un Segmented, endite medially Un First three pairs present as paired anlagen; 4th pair as a single swelling Elongated, unsegmented; 4th pair present as paired anlagen Segmented Un Narrow cellular band in the intra-appendage region of the segments 2 Cone-shaped anlage posterior to pre-cheliceral region Spherical, unsegmented Un 108 5 6 E F Tip tripartite Un Un Endite more significant Un Un 7 G H Un VS slightly wider than in Step 7 Un 9 I Cone-like structure (egg-tooth) present Egg-tooth with wrinkled surface Un Un 8 Un Un 458 10 11 12 13 J K L M Located lateral to mouth opening Longer, tips approaching HY medially ¼ longer than in Step 8, tips superposing HY laterally Located anterior to HY Un Un Un Width of VS doubled from previous step, ST un VS slightly wider than in Step 3 and widening more and more towards the posterior VS slightly wider than in Step 4 VS slightly wider than in Step 5; region posterior to the CH segment wider than pre-cheliceral region VS slightly wider than in Step 6, trapezoidal Lateral to mouth opening Un Anterior to mouth opening Located almost lateral to CH Un Un Located postero-laterally to PP Un Un Un 608 Un 14 15 N O P Un Right and left buds closer to each other than before Un Un Un 16 Un Right and left buds closer to each other than before Un VS slightly wider than in Step 9 Un Un VS un; ST forming a single plate of sternum (STN) VS un; STN more developed VS covered by STN; STN suboval, expanding over entire prosomal area ventrally Un Un 208 Un 308 Un Un Un Columns: ‘‘step’’ – steps as described in text; ‘‘0’’ – letters corresponding to those of Fig. 6; ‘‘A’’ – angle of insertions of prosomal appendage to the ventral midline. Bold highlighted text marks significant changes. Abbreviations: Un ¼ unchanged. 414 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 and Waloszek, 2005), so being again differently used. However, according to our investigation, we exclude the possibility of an appendicular origin of the labium in spiders. 4.2. Some more notes on the embryonic development of other taxa of Chelicerata (Chelicerata s. str. sensu Chen et al., 2004) Of the Pantopoda, many species hatch at the so-called protonymph stage, carrying three pairs of appendages, i.e. the chelicerae and another two (Bain, 2003; Vilpoux and Waloszek, 2003; Waloszek and Dunlop, 2002), while in others the hatchlings are ‘small adults’, such as in Pallene empusa Wilson, 1880 (Morgan, 1891). The earliest embryonic stage of P. empusa that Morgan (1891) documented shows a germ band growing around the egg, which exhibits, from anterior to posterior, a pair of pre-cheliceral lobes, a mouth region being located between the two limb buds of the chelicerae and buds of the following walking legs (Morgan, 1891, his pl. IV, Figs. I and II). This is similar to the early embryos of L. geometricus in terms of the existence of the same body elements, i.e. the pre-cheliceral lobes, mouth region and so on. However, the position of the mouth region in between the cheliceral limb buds is dissimilar to early embryos of L. geometricus but similar to older ones of Step 7–Step 8 recognised here (cf., e.g., Figs. 3D,H). This may indicate that there are more embryonic stages with recognizable external structures in P. empusa earlier than those documented by Morgan (1891). We suppose that the mouth region in the embryos of P. empusa is also initially located anterior to the cheliceral limb buds, as is the case in the embryos of L. geometricus and other chelicerates, e.g., scorpions (Farley, 2001). Studies of the embryogenesis of pantopods with protonymph hatchlings are not available yet. In embryos of Limulus polyphemus (Xiphosura), all limb buds are originally located posterior to the mouth opening, which is situated in-between a pair of pre-cheliceral lobes (Sekiguchi et al., 1982; Mittmann, 2004). The pre-cheliceral lobes in the embryos of L. polyphemus are not as prominent as those in spider embryos. Like in L. geometricus, the position of the mouth region of L. polyphemus embryos also shifts from anterior to posterior during embryogenesis. Details of the development of the pre-cheliceral lobes remain unknown. 5. Conclusions and outlook The detailed investigation on the embryonic development of Latrodectus geometricus enables a comparison between our data and those from existing literature. We are now able to characterise several features found in the embryos of various chelicerates and to further achieve new understanding of the phylogenetic status of those features. New understanding are: (1) in the pre-cheliceral lobes, the two subdivisions and their later internalisation, as well as the associated furrows are currently considered to be part of the ground pattern of the Arachnida. Investigation of such lobes in embryos of xiphosurans and pantopods is essential to further determine the status of those features. (2) The shift of the entire mouth region from anterior to posterior belongs to the ground pattern of, at least, the Chelicerata, or is even more plesiomorphic than this level. (3) The prosomal shield forming from the fusion of the anterior margin of the pre-cheliceral lobes and the tergal region of the four posterior-most prosomal segments plus the entire formation process of the prosomal shield are regarded as belonging to the ground pattern of the Arachnida. (4) It seems reasonable to consider the embryonic ventral sulcus as an autapomorphy for the Lipoctena. The embryonic morphogenesis data are able to demonstrate the initial situation and the developmental processes of various structures. Similar studies have already been done by many authors in the 19th and the first half of the 20th century. At that time, they mainly presented light-microscopic and histological investigations. With modern techniques, such work can be expanded to visualise more details of the external and internal development of arthropod embryos. We propose that better understanding of arthropod evolution can be achieved by performing such studies on various arthropods. Acknowledgements We are grateful to Martin Thierer-Lutz, Schnürpflingen and his colleagues from born to be eaten for providing us all living material used herein. Professor Ryuichiro Machida, Sugadaira, Japan and his students, and Professor Roger Farley, Riverside, USA are thanked for sharing their experience of preparing arthropod embryos for SEM observations to one of the authors (YL). Joachim T. Haug, Ulm, Viktoria Linne, Mainz, and Angelika Stollewerk, London, provided valuable, still unpublished information. We are grateful to Nigel Hughes, Riverside, and Ronald Jenner, London, for correcting the language. Two anonymous reviewers gave valuable suggestions and comments. We also thank the Central Unit for Electron Microscopy at the University of Ulm for supporting our SEM work. Gerd Mayer, University of Ulm, is appreciated for his help with ordering literature. Particular thanks are due to the EU Programme ‘‘Marie Curie Host Fellowships for Early Stage Research Training (EST)’’ for funding Yu Liu as a PhD fellow within the MOLMORPH network under this programme (contract number MEST-CT-2005020542). Appendix A. Supplemental material Supplementary information for this manuscript can be downloaded at doi:10.1016/j.asd.2009.04.001. References Akiyama-Oda, Y., Oda, H., 2003. Early patterning of the spider embryo: a cluster of mesenchymal cells at the cumulus produces Dpp signals received by germ disc epithelial cells. Development 130, 1735–1747. Anderson, D.T., 1973. Embryology and Phylogeny in Annelids and Arthropods. Pergamon Press, Oxford-New York-Toronto-Sydney-Braunschweig. Ax, P., 1999. Das System der Metazoa II. Ein Lehrbuch der phylogenetischen Systematik. Gustav Fischer Verlag, Stuttgart. Bain, B.A., 2003. Larval types and a summary of postembryonic development within the pycnogonids. Invertebrate Reproduction and Development 43, 193–222. Brauer, A., 1895. Beiträge zur Kenntnis der Entwicklungsgeschichte des Skorpions. II. Zeitschrift für Wissenschaftliche Zoologie 59, 351–433. Boyan, G.S., Williams, J.L.D., Posser, S., Bräunig, P., 2002. Morphological and molecular data argue for the labrum being non-apical articulated, and the appendage of the intercalary segment in the locust. Arthropod Structure and Development 31, 65–76. Boyan, G.S., Bräunig, P., Posser, S., Williams, J.L.D., 2003. Embryonic development of the sensory innervation of the clypeo-labral complex: further support for the serially homologous appendages in the locust. Arthropod Structure and Development 32, 289–302. Butt, F.H., 1960. Head development in the arthropods. Biological Reviews of the Cambridge Philosophical Society (London) 35, 43–91. Chaw, R.C., Vance, E., Black, S.D., 2007. Gastrulation in the spider Zygiella x-notata involves three distinct phases of cell internalization. Developmental dynamics 236, 3484–3495. Chen, J.Y., Waloszek, D., Maas, A., 2004. A new ‘‘great appendage’’ arthropod from the Lower Cambrian of China and the Phylogeny of Chelicerata. Lethaia 37, 3–20. Claparède, E., 1862. Recherches sur l’évolution des Araignées. Utrechtsch Genootschap van Kunsten en Wetenschappen. Natuurkundige verhandelingen 1. Damen, W.G.M., 2002. Parasegmental organization of the spider embryo implies that the parasegment is an evolutionary conserved entity in arthropod embryogenesis. Development 129, 1239–1250. Farley, R., 1998. Matrotrophic adaptations and early stages of embryogenesis in the desert scorpion Paruroctonus mesaensis (Vaejovidae). Journal of Morphology 237, 187–211. Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Farley, R., 2001. Development of segments and appendages in embryos of the desert scorpion Paruroctonus mesaensis (Scorpiones: Vaejovidae). Journal of Morphology 250, 70–88. Foelix, R.F., 1996. Biology of Spiders, second ed. Oxford University Press Georg Thieme Verlag, pp. 1-330. Giammara, B.L., DeVries, W.C., Baker, R.L., Dobbins, J.J., Hanker, J.S., 1987. Hexamethyldisilazane drying for rapid detection of bacteria in implant specimens. Proceedings of the Electron Microscope Society of America, 45, 878–879. Haas, M.S., Brown, S.J., Beeman, R.W., 2001a. Pondering the procephalon: the segmental origin of the labrum. Development Genes and Evolution 211, 89–95. Haas, M.S., Brown, S.J., Beeman, R.W., 2001b. Homeotic evidence for the appendicular origin of the labrum in Tribolium castaneum. Development. Genes and Evolution 211, 96–102. Haug, J.T., Maas, A., Waloszek, D. Ontogeny of two Cambrian stem crustaceans, Goticaris longispinosa and Cambropachycope clarksoni. Palaeontographica, in press. Herold, M., 1824. De generatione Aranearum in ovo. Marburg. Holm, A., 1940. Studien über die Entwicklung und Entwicklungsbiologie der Spinnen. Zoologiska Bidrag Från Uppsala 19, 1–214. Holm, A., 1952. Experimentelle Untersuchungen über die Entwicklung und Entwicklungsphysiologie des Spinnenembryos. Zoologiska Bidrag Från Uppsala 29, 293–424. Holm, A., 1954. Notes on the development of an orthognath spider, Ischnothele karschi Bös. & Lenz. Zoologiska Bidrag från Uppsala 30, 199–222. Jager, M., Murienne, J., Clabaut, C., Deutsch, J., Guyader, H.L., Manuel, M., 2006. Homology of arthropod anterior appendages revealed by Hox gene expression in a sea spider. Nature 441, 506–508. Janssen, R., Feitosa, N.M., Damen, W.G.M., Prpic, N.M., 2008. The T-box genes H15 and optomotor-blind in the spiders Cupiennius salei, Tegenaria atrica and Achaearanea tepidariorum and the dorso-ventral axis of arthropod appendages. Evolution and Development 10, 143–154. Kästner, A., 1948. Zur Entwicklungsgeschichte von Thelyphonus caudatus (Pedipalpi). 1. Teil. Die Ausbildung der Körperform. Zoologische Jahrbücher – Abteilung für Anatomie und Ontogenie der Tiere 69, 493–506. Kästner, A., 1949. Zur Entwicklungsgeschichte von Thelyphonus caudatus L. (Pedipalpi). 2. Teil. Die Entwicklung der Mundwerkzeuge, Beinhüften und Sterna. Zoologische Jahrbücher – Abteilung für Anatomie und Ontogenie der Tiere 70, 169–197. Kästner, A., 1950. Zur Entwicklungsgeschichte von Thelyphonus caudatus L. (Pedipalpi). 3. Teil. Die Entwicklung des Zentralnervensystems. Zoologische Jahrbücher – Abteilung für Anatomie und Ontogenie der Tiere 71, 1–55. Kimm, M.A., Prpic, N.M., 2006. Formation of the arthropod labrum by fusion of paired and rotated limb-bud-like primordial. Zoomorphology 125, 147–155. Kishimoto, T., Ando, H., 1985. External features of the developing embryo of the stonefly, Kamimuria tibialis (Pictet) (Plecoptera, Perlidae). Journal of Morphology 183, 311–326. Larink, O., 1970. Die Kopfentwicklung von Lepisma saccharina L. (Insecta, Thysanura). Zeitschrift für Morphologie der Tiere 67, 1–15. Maas, A., Waloszek, D., 2005. Phosphatocopina – ostracode like sister group of Eucrustacea. In: Ikeya, N., Tsukagoshi, A., Horne, D.J. (Eds.), Proceedings of the 14th International Symposium on Ostracoda, August 04 to 09, 2001, Shizuoka, Japan. Hydrobiologia 538, 139–152. Maas, A., Waloszek, D., Müller, K.J., 2003. Morphology, Ontogeny and Phylogeny of the Phosphatocopina (Crustacea) from the Upper Cambrian ‘Orsten’ of Sweden. Fossils and Strata 49, 1–238. Maas, A., Braun, A., Dong, X.P., Donoghue, P., Müller, K.J., Olempska, E., Repetski, J.E., Siveter, D.J., Stein, M., Waloszek, D., 2006. The ‘Orsten’ – more than a Cambrian Konservat-Lagerstätte yielding exceptional preservation. Palaeoworld 15, 266–282. Machida, R., 1981. External features of embryonic development of a jumping bristletail, Pedetontus unimaculatus Machida (Insecta, Thysanura, Machilidae). Journal of Morphology 168, 339–355. Marechal, P., 1994. Embryonic and early postembryonic development of Ischnothele guyanensis (Araneae, Mygalomorphae, Dipluridae), with special reference to the visual system. Invertebrate Reproduction and Development 26 (2), 133–144. McGregor, A.P., Hilbrant, M., Pechmann, M., Schwager, E.E., Prpic, N.M., Damen, W.G.M., 2008. My favorite animal: Cupiennius salei and Achaearanea tepidariorum: spider models for investigating evolution and development. BioEssays 30, 487–498. Mittmann, B., 2004. Die Embryologie des Pfeilschwanzkrebses Limulus polyphemus (Xiphosura, Chelicerata) und anderer Arthropoden unter besonderer Berücksichtigung der Neurogenese (PhD thesis, Humboldt-Universität zu Berlin). Mittmann, B., Scholtz, G., 2003. Development of the nervous system in the ‘‘head’’ of Limulus polyphemus (Chelicerata: Xiphosura): morphological evidence for a correspondence between the segments of the chelicerae and of the (first) antennae of Mandibulata. Development Genes and Evolution 213, 9–17. Morgan, T.H., 1891. A Contribution to the Embryology and Phylogeny of the Pycnogonids. Dissertation, Johns Hopkins University, pp. 1–76. Moritz, M., 1957. Zur Embryonalentwicklung der Phalangiiden (Opiliones, Palpatores) II. Die Anlage und Entwicklung der Coxaldrüse bei Phalangium opilio L. Zoologische Jahrbucher – Abteilung für Anatomie und Ontogenie der Tiere 77, 229–240. 415 Nation, J.L., 1983a. A new method using hexamethyldisilazone for preparation of soft insect tissues for scanning electron microscopy. Stain Technology 58 (6), 347–351. Nation, J.L., 1983b. Specialization in the alimentary canal of some mole crickets (Orthoptera: Gryllotalpidae). International Journal of Insect Morphology and Embryology 12, 201–210. Pross, A., 1966. Untersuchungen zur Entwicklungsgeschichte der Araneae (Pardosa horrtensis (Thorell)) und besonderer Berücksichtigung des vorderen Prosoma-abschnittes. Zeitschrift für Morphologie und Ökologie der Tiere 58, 38–108. Prpic, N.M., 2005. Duplicated Pax6 genes in Glomeris marginata (Myriapoda: Diplopoda), an arthropod with simple lateral eyes. Zoology 108, 47–53. Prpic, N.M., Damen, W.G.M., 2005. Cell death during germ band inversion, dorsal closure and nervous system development in the spider Cupiennius salei. Developmental Dynamics 234, 222–228. Rempel, J.G., 1957. The embryology of the black widow spider, Latrodectus mactans (Fabr.). Canadian Journal of Zoology 35, 35–74. Sakuma, M., 2005. Embryonic and postembryonic developments of the centipede Scolopocryptops rubiginosus L. Koch (Chilopoda: Scolopendromorpha) (PhD thesis). Sakuma, M., Machida, R., 2004. Germ band formation of a centipede Scolopendra subspinipes L. Koch (Chilopoda: Scolopendromorpha). Proceedings of Arthropodan Embryological Society of Japan 39, 41–43. Scholtz, G., Edgecombe, G.D., 2005. Heads, Hox and the phylogenetic position of trilobites. In: Koenemann, S., Jenner, R. (Eds.), Crustacea and Arthropod Relationships. Crustacean Issues, 16, pp. 139–165. Scholtz, G., Edgecombe, G.D., 2006. The evolution of arthropod heads: reconciling morphological, developmental and palaeontological evidence. Development. Genes and Evolution 216, 395–415. Schoppmeier, M., Damen, W.G.M., 2005. Suppressor of Hairless and Prenesilin phenotypes imply involvement of canonical Notch-signalling in segmentation of the spider Cupiennius salei. Developmental Biology 280, 211–224. Sekiguchi, K., Yamamichi, Y., Costlow, J.D., 1982. Horseshoe crab development studies I. Normal embryonic development of Limulus polyphemus compared with Tachypleus tridentatus. In: Bonaventura, J. (Ed.), Physiology and Biology of Horseshoe Crabs: Studies on Normal and Environmentally Stressed Animals. Alan R. Liss, Inc., New York, pp. 53–73. Siewing, Rolf (Hrsg.), 1985. Lehrbuch der Zoologie. Band 2 Systematik. Gustav Fischer Verlag, Stuttgart, New York. Siveter, D.J., Waloszek, D., Williams, M., 2003. An early Cambrian Phosphatocopid Crustacean with three-dimensionally preserved soft parts from Shropshire, England. In: Lane, P.D., Siveter, D.J., Fortey, R.A. (Eds.), Trilobites and their Relatives. Special Papers in Palaeontology, 70, pp. 9–30. 2 pls. Stein, M., Waloszek, D., Maas, A., Haug, J.T., Müller, K.J., 2008. Oelandocaris oelandica revisited. Acta Palaeontologica Polonica 53, 462–484. Stollewerk, A., Seyfarth, E.A., 2008. Evolutionary changes in sensory precursor formation in arthropods: embryonic development of leg sensilla in the spider Cupiennius salei. Developmental Biology 313, 659–673. Stollewerk, A., Weller, M., Tautz, D., 2001. Neurogenesis in the spider Cupiennius salei. Development 128, 2673–2688. Suzuki, H., Kondo, A., 1995. Early embryonic development, including germ-disk stage, in the Theridiid spider Achaearanea japonica (Bös. et Str.). Journal of Morphology 244, 147–157. Thomas, R.H., Telford, M.J., 1999. Appendage development in embryos of the oribatid mite Archegozetes longisetosus (Acari, Oribatei, Trhypochthoniidae). Acta Zoologica (Stockholm) 80, 193–200. Uemiya, H., Ando, H., 1987. Embryogenesis of a Springtail, Tomocerus ishibashii (Collembola, Tomoceridae): external morphology. Journal of Morphology 191, 37–48. Ungerer, P., Wolff, C., 2005. External morphology of limb development in the amphipod Orchestia cavimana (Crustacea, Malacostraca, Peracarida). Zoomorphology 124, 89–99. Vilpoux, K., Waloszek, D., 2003. Larval development and morphogenesis of the sea spider Pycnogonum litorale (Ström, 1762). Arthropod Structure and Development 32, 349–383. Waloszek, D., 2003. The ‘Orsten’ Window – three-dimensionally preserved Upper Cambrian Meiofauna and its contribution to our understanding of the evolution of arthropoda. Paleontological Research 7, 71–88. Waloszek, D., Dunlop, J.A., 2002. A larval sea spider (Arthropoda: Pycnogonida) from the Upper Cambrian ‘Orsten’ of Sweden, and the phylogenetic position of pycnogonids. Palaeontology 45, 421–446. Walossek, D., Müller, K.J., 1990. Stem-lineage crustaceans from the Upper Cambrian of Sweden and their bearing upon the position of Agnostus. Lethaia 23, 409–427. Westheide, W., Rieger, R. (Eds.), 1996. Spezielle Zoologie, Erster Teil: Einzeller und Wirbellose Tiere, 1. Aufl. Gustav Fischer Verlag, Stuttgart, Jena, New York, pp. 1–883. Weygoldt, P., 1964. Vergleichend-embryologische Untersuchungen an Pseudoscorpionen (Chelonethi). Zeitschrift für Morphologie und Ökologie der Tiere 54, 1–106. Weygoldt, P., 1975. Untersuchungen zur Embryologie und Morphologie der Geißelspinne Tarantula marginemaculata C.L. Koch (Arachnida, Amblypygi, Tarantulidae). Zoomorphologie 82, 137–199. Weygoldt, P., Paulus, H.F., 1979. Untersuchungen zur Morphologie, Taxonomie und Phylogenie der Chelicerata. II. Cladogramme und die Entfaltung der Chelicerata. Zeitschrift für zoologische Systematik und Evolutionsforschung 17, 177–200. 416 Y. Liu et al. / Arthropod Structure & Development 38 (2009) 401–416 Yoshikura, M., 1954. Embryological studies on the liphistiid spider, Heptathela kimurai. Kumamoto Journal of Science 3B, 41–50. Yoshikura, M., 1955. Embryological studies on the liphistiid spider, Heptathela kimurai. II. Kumamoto Journal of Science B2, 1–86. Yoshikura, M., 1958. On the Development of a purse-web Spider, Atypus karschi Dönitz. Kumamoto Journal of Science B (2): Biology 3, 73–85. Yoshikura, M., 1961. The development of the whip-scorpion Typopeltis stimpsonii Wood. Acta Arachnologica 17, 19–24. Yoshikura, M., 1972. Notes on the development of a trap-door spider, Ummidia fragaria (DOENITZ). Acta Arachnologica 24, 29–39. Yoshikura, M., 1975. Comparative embryology and phylogeny of arachnida. Kumamoto Journal of Science 12, 71–142.