The ontogeny of the 300 million year old xiphosuran Euproops danae (Euchelicerata) and implications for resolving the Euproops species complex


Xiphosurans have often been considered as archaic appearing cheliceratan arthropods, with a rich fossil record. We describe here parts of the post-embryonic ontogeny of the 300 million year old xiphosuran Euproops danae (Xiphosura sensu stricto, Euchelicerata), from the Mazon Creek Lagerstätte (Upper Carboniferous), USA. Recently, the ontogeny of a closely related species, Euproops sp. from the Upper Carboniferous Piesberg quarry, Osnabrück, Germany (informally called ‘Piesproops’), has been reconstructed. This analysis has drawn characters into question that were used to differentiate E. danae from another species occurring at the same time, Euproops rotundatus from the British Middle Coal Measures. More precisely, early post-embryonic stages of Piesproops resemble E. danae; later stages resemble E. rotundatus. Based on this earlier study, the here-described reinvestigation of E. danae has been performed as the ontogenetic sequence itself may yield more reliable characters for differentiating species of Euproops. We could identify eight different growth stages for E. danae. This ontogenetic sequence shows a comparable growth to that of Piesproops, but differs markedly in the development of the opisthosomal flange. This character may serve as a basis for reliably differentiating these species. Additionally, analysing the ontogeny of further species may offer the basis for identifying heterochronic shifts in the evolution of xiphosurans.


Xiphosurans are marine chelicerates. Their most common vernacular names horseshoe ‘crabs’ or king crabs are therefore misleading. The old-fashioned, yet more appropriate term ‘sword tail’ is rarely used in the English language (and unfortunately also has a second meaning, as common name for the poeciliid fish species Xiphophorus hellerii). However, in Germany, the equivalent ‘Schwertschwanz’ is more common and not misleading. The monophyly of the group Xiphosura has been questioned recently (Lamsdell 2013). Hence, only a former ingroup now named Xiphosura sensu stricto (sensu Lamsdell 2013) remains as a monophyletic group. Xiphosura s. str. is the sister group to a group termed Dekatriata, comprising some synziphosurine species, chasmataspidids, eurypterids (all exclusively fossil) and arachnids, the modern land-living chelicerates.

Despite the challenges with the names of this group, both scientific and vernacular names, xiphosurans in the strict sense are an important group for understanding the early evolution and diversification of Chelicerata sensu stricto (sensu Chen et al. 2004). In the modern fauna, only four species are known, Limulus polyphemus, Tachypleus gigas, Tachypleus tridentatus and Carcinoscorpius rotundicauda (e.g. Shuster and Sekiguchi 2003). Yet, in former times, xiphosurans were much more species rich. Among the most commonly known fossil xiphosurans are species of Euproops, known from different localities around the world (e.g. Brauckmann 1982; Anderson 1994; Racheboeuf et al. 2002; Dunlop et al. 2008; Haug et al. 2012a). Prominent examples are the Upper Carboniferous, 300 million year old species Euproops danae from the famous Mazon Creek Lagerstätte, Illinois, USA, and Euproops rotundatus from the British Middle Coal Measures, UK. The primarily assumed relationships between the different species of Euproops as well as their validity have been challenged in recent years (Anderson 1994; Anderson and Selden 1997; Schultka 2000; Haug et al. 2012a).

Two aspects have been put forward in this respect:

  1. 1.

    Most of the species have been characterised based only on characters supposedly affected by taphonomy.

  2. 2.

    Ontogeny complicates a clear species identification due to (a) the comparison of non-corresponding ontogenetic stages, which subsequent investigations were based upon, and (b) the fact that different developmental stages may have been misinterpreted as separate species.

In more detail for (1), Anderson (1994) remarked that preservational artefacts could lead to different appearances of the specimens and may be misinterpreted as different species. In consequence, Anderson (1994) synonymised many species of Euproops, e.g. E. kilmersdonensis, with E. danae. Here, among others, compression of the prosoma seemed to have caused the misinterpretation as a separate species. Anderson considered three species as still valid: E. danae, Euproops anthrax and E. rotundatus. Schultka (2000) partly questioned the separate identity of E. danae and E. rotundatus, referring to the E. rotundatus-danae species complex. This seeming continuity was based on investigations of specimens of Euproops sp. from the Piesberg quarry near Osnabrück, Germany, leading also to aspect (2), ontogeny.

Haug et al. (2012a) reconstructed the post-embryonic ontogeny of the Euproops species from the Piesberg quarry. They noted that the opisthosoma morphology changes considerably (although chelicerates are generally interpreted as direct developers, hatching as ‘small adults’; see Waterman 1954 for general morphological constancy during post-embryonic development of extant xiphosurans). More precisely, earlier stages of ‘Piesproops’ (informally short for Piesberg-Euproops) resemble the opisthosoma morphology of E. danae, while older specimens have a similar morphology to E. rotundatus. This is not only important to understand the possible species identity of different forms of Euproops. It also indicates that heterochrony might have played an important role in xiphosuran evolution (Haug et al. 2012a). To clearly differentiate the putative different species of Euproops, several and multifaceted criteria should be used. The character ‘a bit wider prosoma’ which is used as separating character, for example, in the case of E. rotundatus versus E. danae (Schultka 2000), is in our view weak. Such differences could result from other factors, such as sexual dimorphism (as, e.g. in Limulus polyphemus: Loveland and Botton 1992; see also Hauschke and Wilde 2004; Lamsdell and McKenzie 2015). Other characters are similarly problematic. The character ‘longer genal spines’ (Anderson 1994; Schultka 2000) is highly affected by taphonomy as the distal part of the spine often remains in the matrix of the counterpart (see also Fisher 1977 for other difficulties). And finally, a ‘broad opisthosomal margin’ (Schultka 2000) seems to depend strongly on the developmental status (Haug et al. 2012a).

Therefore, for a more reliable interpretation of fossil species, taphonomy and ontogeny as well as further possible influencing factors, such as sexual dimorphism, need to be considered. While taphonomy was a central topic of the work of Anderson (1994), especially the effects of ontogeny still need to be investigated in a wider context.

Accordingly, we present here new data on supposed growth stages of Euproops danae from Mazon Creek as one step towards reconstructing the ontogeny of this species, following the holomorph approach, i.e. including data of the entire (available part of) ontogeny of a species into an analysis (Hennig 1966). We furthermore discuss the impact of our findings on xiphosuran (s. str.) diversity and evolution.

Material and methods


Specimens for this study came from the Invertebrate Paleontology collection of the Yale Peabody Museum of Natural History, New Haven (YPM IP). More than 500 fossil xiphosurans are housed in the collection; most of them can be allocated to Euproops danae, Mazon Creek. All specimens were inspected, and more than 100 specimens were photographed. Only 29 specimens were preserved well enough for allowing proper measurements.

For comparison, late embryonic and hatchling stages of Limulus polyphemus were investigated. These specimens are part of the Invertebrate Zoology collection of the Yale Peabody Museum of Natural History, New Haven (YPM IZ).

Documentation methods

The fossil specimens were photographed with a Canon Rebel T3i digital camera with an EF-S 18-55-mm lens or a MP-E 65-mm macro lens. Light was provided by a Meike FC 100 LED ring light or a Canon Macro Twin Flash MT-24EX, both equipped with polarisation filters. A cross-polarised filter was placed in front of the lens to reduce reflections and enhance the colour contrast between the specimen and the matrix (e.g. Haug et al. 2012a, b).

All specimens were documented as stereo image pairs and processed as red-cyan stereo images (e.g. Haug et al. 2012a, b). Counterparts with a negative relief were depth-inverted and some additionally colour-inverted, sometimes providing better contrast. The stereo images were prepared in Adobe Photoshop CS3 and Gimp 2.6.11.

Details of one specimen were additionally documented with a virtual surface reconstruction technique. The procedure followed Haug et al. (2013).

The extant comparative material was documented directly in its storage liquid, 70% ethanol, under auto-fluorescence conditions on a Keyence BZ-9000 microscope. The imaging and subsequent image processing followed Haug et al. (2011) and references therein.

Presentation of structural interpretation

For marking certain structures, the stereo images were processed in Adobe Photoshop CS5 via deletion of one colour channel followed by desaturation. Structures as seen on the stereo images were then marked by hand using the lasso tool. Non-stereo versions of the figures are available as electronic supplement (Supplementary Figs. 14).


Lengths and widths of the prosoma and the opisthosoma were measured on the digital photographs using Gimp (see Haug et al. 2012a, their Fig. 1). If a specimen was incomplete on one side, width was measured from the rim of the well-preserved side to the midline and then doubled. Measured lengths were displayed as scatter plots in OpenOffice. By means of these scatter plots, clusters were identified and the specimens were identified to different growth stages. Not all specimens provided all lengths, as some were strongly compressed or fragmentary, but were integrated as well based on a combination of all plots.

Fig. 1

Representatives of each growth stage of Euproops danae ordered according to size. Smallest specimen, but which was not well enough preserved to be used for the correlations (I′ YPM IP 050503). First growth stage (used for correlations) (I YPM IP 016910). Second growth stage (II YPM IP 050687). Third growth stage (III YPM IP 025590). Fourth growth stage (IV YPM IP 050835). Fifth growth stage (V YPM IP 050570). Sixth growth stage (VI YPM IP 050933). Seventh growth stage (VII YPM IP 028514). Eighth growth stage (VIII YPM IP 035153). Stereo images; please use red-cyan glasses to view. Non-stereo versions available as electronic supplement

Additionally, for each specimen well enough preserved, the relative dimensions of the opisthosomal flange were measured (see Haug et al. 2012a). For this purpose, the distance between the base of two spines and the smallest width of the epimera was measured. Subsequently, the growth groups and the epimera morphology of Euproops danae were compared with those of Piesproops (Haug et al. 2012a).


General morphology of Euproops danae

The general morphology of Euproops danae is largely similar in all investigated ontogenetic stages (Fig. 1). Details for the different developmental stages follow further below.

The body is organised into two main units: the anterior prosoma and the posterior opisthosoma (Fig. 2a). The prosomal segments form a single dorsal sclerotisation, the prosomal shield. The prosomal shield is broader than long and gently rounded along the anterior margin in dorsal view. The posterior margin of the shield is almost straight.

Fig. 2

Details of the morphology of Euproops danae. ac YPM IP 028514. a Part of fossil, dorsal view, stereo image. b Counterpart of fossil, showing ophthalmic spines, stereo image. c Same as b, ophthalmic spines colour-marked. d, e YPM IP 016912. d Image of chelicera, stereo image. e Same as d, chelicera and first walking appendage colour-marked. f, g YPM IP 016910. f Stereo image. g Same as f, structures colour-marked. h YPM IZ 058121. Limulus polyphemus embryo. Arrow points to the bulged operculum flipped anteriorly. ap1–5 walking appendages 1–5, ch chelicera, op/bgb presumed operculum or first book gill-bearing opisthosomal appendage with anterior extensions. Non-stereo versions available as electronic supplement

The postero-lateral corners of the prosomal shield are each drawn out into a posteriorly pointing spine, the genal spines. Another pair of spines, the ophthalmic spines, is located further towards the middle of the posterior shield margin. These are the prolongations of the ophthalmic ridge, which marks a central area of the shield. Well-preserved ophthalmic spines are observed in one specimen, which belongs to a very late growth stage (Fig. 2b, c). In other specimens, these are usually broken off (e.g. Fig. 2a).

Also, the segments of the opisthosoma form a single dorsal sclerotisation, the thoracetron. The thoracetron has a straight anterior margin, lateral and posterior margins rounded. Medially, a spindle-shaped axial region is apparent. A distinct ridge separates the thoracetron from a region surrounding it laterally and posteriorly, the flange. Seven segments are indicated by furrows in the axis continuing into ridges on the thoracetron and the flange. Each individual region on the flange is called an epimeron (plural: epimera). Each epimeron is drawn out latero-posteriorly into a distinct spine. Along the axial region, three distally pointing spines are apparent, more precisely their bases. Similar to the ophthalmic spines, these are usually broken off and can only be detected by intensive preparation (Fig. 2b, c).

Exceptional details

Ventral details, especially the appendages of Euproops danae, are difficult to investigate. Only with the aid of stereo images and virtual surface reconstruction, parts of appendages become visible as these are mainly preserved through relief.

The remains of the appendages usually only give a rough idea of the position of the most proximal elements of the legs. Mainly, the prosomal appendages are observable. The insertion areas of the chelicerae and the following five pairs of walking legs are positioned in a circle in the prosoma surrounding the mouth opening (Fig. 2d–g).

Of the opisthosomal appendages, only a hint of a structure could be observed (Fig. 2f, g). This structure could represent parts of the operculum or of one of the further posterior swimming appendages. The lateral regions appear lobe-like; medially, a pair of small elongations is apparent. This appendage appears flipped forward in a similar way as it can be displayed in the extant specimens of Limulus polyphemus (Fig. 2h).

In one specimen, not only the proximal region of the chelicera is preserved (Fig. 3). Here, the complete proximal element (peduncle) and the second element (proximal claw element) are preserved. The latter seems to have the elongated fixed finger still in position; only the movable finger (chelicera element 3) appears not to be preserved.

Fig. 3

Anterior appendages of Euproops danae compared with those of Limulus polyphemus. af Euproops danae, YPM IP 050472. a Close-up on anterior appendages. b Same as a, with coloured parts. c Different angle than a, but same detail. d Same as c, with coloured parts. e Virtual surface reconstruction. f Overview image of the fossil. g Close-up of anterior appendages in Limulus polyphemus, YPM IP 030358. ap1 walking appendage 1, ch chelicera, ff fixed finger, pc posterior claw element, pd. peduncle. a, c, f, g Stereo images; please use red-cyan glasses to view. Non-stereo versions available as electronic supplement

Growth stages

The smallest investigated specimen of Euproops danae measures about 5.2 mm from the anterior prosoma edge to the base of the telson. Unfortunately, it is not well enough preserved and is thus not incorporated into the graphs. However, it has clearly smaller dimensions than any other studied specimen. The prosoma width is about 8 mm (right half doubled, left side not completely preserved). The opisthosoma width is about 5 mm, the prosoma length about 3 mm, and the opisthosoma length is about 2.5 mm.

In the largest investigated specimen of Euproops danae, the prosoma is 46 mm wide and the opisthosoma is 28 mm wide. The prosoma length is not available as the anterior rim is broken, but the opisthosoma length is 18 mm. The data suggest the presence of eight more or less well-defined clusters, most likely representing growth stages (Fig. 4). The first four of these supposed growth stages appear clearly separated from each other. With further growth, the data display more variation within each supposed cluster; hence, possible stages are less distinctly separable.

Fig. 4

Scatter plots of measured lengths of Euproops danae. The different symbols represent different size classes. a Prosoma length versus prosoma width. b Prosoma width versus opisthosoma width. c Opisthosoma length versus opisthosoma width. d Prosoma length versus opisthosoma length. When combining all plots, eight growth stages can be distinguished

All graphs show an overall strict trend line, so the specimens grow steadily with respect to the measured dimensions and without large jumps. The prosoma width grows relatively faster in comparison to the opisthosoma width in the single steps, but the growth of the different body parts is steady. The growth steps from one growth group to the next one are relatively similar. The largest steps occur from group I to group II with about a 1.3 times increase of opisthosoma width. A similarly large step occurs from group IV to group V. The other steps all display about a 1.15 times increase, whereas the step between groups VII and VIII displays only about a 1.1-fold increase.

Morphological changes during ontogeny

The investigated specimens of Euproops danae remain morphologically largely the same throughout ontogeny. There is a continuous gain in size. An observable change throughout the ontogeny concerns the opisthosoma, more precisely the morphology of the surrounding flange formed by the epimera.

From growth groups I to III, the morphology of the epimera appears not to change (Fig. 5); only a general increase in opisthosoma dimensions occurs. This is presumably also the case in I′, which describes the smallest specimen, but which provides no recognisable epimera. Group IV is represented by only one specimen of which details of the epimera are not well preserved. Hence, a change in the morphology of the epimera occurs either from groups III to IV or from groups IV to V. The epimera stay thin with short spines or part of spines. Specimens of growth group V then display more elongated spines. Starting in growth stage VI, ridges occur that separate the basis of the epimera from each other. These ridges are continuous with the ridges on the central thoracetron, which mark the individual segments.

Fig. 5

Relative differences in the dimensions of the epimera between the growth stages of Euproops danae. Epimera of the smallest specimen, flipped horizontally (I′ YPM IP 050503). First growth stage epimera, flipped horizontally (I YPM IP 016910). Second growth stage epimera (II YPM IP 050687). Third growth stage epimera (III YPM IP 025590). Fourth growth stage epimera, flipped horizontally (IV YPM IP 050835). Fifth growth stage epimera (V YPM IP 050570). Sixth growth stage epimera, flipped horizontally (VI YPM IP 050933). Seventh growth stage epimera (VII YPM IP 028514). Eighth growth stage epimera (VIII YPM IP 035153). Stereo images; please use red-cyan glasses to view. Non-stereo versions available as electronic supplement

The opisthosoma width changes from a specimen of group III to group VIII from 12.8 to 28.3 mm, whereas the epimera width doubles. The relation of the distance between the base of two spines versus the smallest width of the epimera (Fig. 5) lies between 0.5 and 0.7 among all growth stages (see discussion for comparison with Piesproops).

Relative staging in comparison to the ontogeny of Piesproops

As the opisthosoma width appears to be the most reliable dimension, the data series of Euproops danae and of Piesproops were merged in a scatter plot of opisthosoma length versus opisthosoma width (Fig. 6). In this plot, also the smallest specimen is included despite its incomplete preservation for a better comparison. The scatter plot revealed that the growth groups of Euproops danae and those of Piesproops form a homogenous progression with nearly identical trend lines. As the data of Piesproops include smaller and larger specimens than those of Euproops danae and Piesproops has two more growth stages, the growth groups are shifted against each other (Fig. 7). Therefore, the smallest appropriately preserved specimen was additionally integrated as growth group I′. However, the smallest specimen of Euproops danae correlates with the third growth stage of Piesproops. Starting from the fifth growth group in both data series, size variation increases and complicates a clear separation of the growth groups. Especially growth groups 7 and 8 in the data of Piesproops scatter strongly.

Fig. 6

Scatter plot of opisthosoma length versus opisthosoma width of ontogenetic series of Euproops danae and ‘Piesproops’. Specimens of E. danae are marked with grey symbols and numbering, those of Piesproops with black symbols and numbering (based on Haug et al. 2012a; for stages 6 and 9, these values could not be measured)

Fig. 7

Comparative scheme of the epimera development and body size in different growth stages of ‘Piesproops’ (left) and Euproops danae (right). Stippled lines indicate similar epimera development; solid lines indicate similar body size. At similar size, the flange (formed from a series of epimera) is less developed in Euproops danae than in Piesproops, and the marginal spines are correspondingly slightly longer


Differentiating growth groups in Euproops danae

The supposed number of eight growth groups in Euproops danae is lower than what is known from extant xiphosuran species. Depending on the species, the sex and the rearing conditions, extant species have between 12 and 18 instars (summary in Chiu and Morton 2001). However, there seems to be no strict correlation between the final size of the adult and the number of instars. Also, the size at hatching is very different between the extant species (Waterman 1954).

In general, the growth of Euproops danae appears rather steady without any significant jumps. However, the growth rates between growth groups I and II and between IV and V are larger than the other ones. This could indicate that we miss a step between these stages as growth rates are usually rather constant in early post-embryonic development (for growth rates in extant species, see Waterman 1954). Our scatter plots show an increased scattering of the data points in later growth stages of Euproops danae (Fig. 6). This points to a higher variability between representatives of one ontogenetic stage the older the animals get. Due to this scattering, the older stages cannot be distinguished as clearly as the younger ones. Possibly also the relatively small size increase from growth stage VI to VII might be an effect of the larger variation of sizes within these groups.

One reason for this observed increased scattering might be sexual dimorphism, which is known for extant xiphosuran species, with females being larger than males (e.g. Loveland and Botton 1992; Lamsdell and McKenzie 2015). However, it is not clear if the only reason for this difference is the supposed additional moult occurring in females (Chiu and Morton 2001) or if they also grow larger between the moults (Waterman 1954; Carmichael et al. 2003). This may also be the case for Euproops danae.

Another possible reason is that the more the animals are exposed to different environmental conditions, they will differ in size gain. Arthropods can show either increased or decreased size gain at subsequent moults, depending, for example, on temperature (Block et al. 1990). Also, the maximally tolerable temperature may vary depending on the stage of development (Fields 1992). Furthermore, in certain arthropod species, additional developmental stages are optionally inserted depending on the environmental conditions (e.g. Bellinger and Pienkowski 1987; Anger 2001 and references therein). At least a certain variation in the number of moults required to reach adulthood has been reported for extant xiphosurans (Chiu and Morton 2001; Hu et al. 2015 and references therein), which may be caused by varying environmental conditions.

Functional morphology of Euproops danae

For understanding the life habits of Euproops danae, investigating the functional morphology of its body parts, especially of its appendages, is important. For its close relative Piesproops, Haug et al. (2012a) presented details of the prosomal appendages, which are relatively elongate. Schultka (2000) described walking legs and ‘pusher’ legs (with a kind of furcation distally instead of pincers) for Piesproops. The appendages described by Schultka (2000) measure about 1 mm width, which differs only little from the observation of Haug et al. (2012a) with appendage widths of ca. 0.7 mm, which is relatively slim in respect to their length.

Euproops danae most likely lived in non-marine waters, but also partly terrestrial life habits have been suggested (e.g. Fisher 1979; Brauckmann 1982; but see Anderson 1994 for several counter-arguments). The morphology of the walking legs is generally assumed to be adjusted to a benthic life style including walking on unstable substrate and crossing obstructions (Anderson and Selden 1997; Schultka 2000).

Besides the morphology of the prosomal appendages, also that of the opisthosomal appendages is crucial for the life habits of E. danae. If these animals walked on land and had in general the same opisthosomal appendage morphology as modern xiphosurans, the book gills situated on these appendages would have collapsed. This also happens in modern xiphosurans and leads to a highly reduced oxygen uptake (e.g. Reisinger et al. 1991). Thus, E. danae either performed only limited movements while being on land or it would have needed additional support structures preventing the book gills from collapsing (see discussion in Anderson 1994). Accordingly, the scenario of Fisher (1979) with E. danae climbing up plants while being on land appears rather unlikely. In our material, there appears to be part of the operculum of one of the specimens preserved, but further opisthosomal structures are lacking, so we cannot contribute new data to this discussion.

Yet, we were able to observe certain aspects of the feeding apparatus of Euproops danae. As Haug et al. (2012a) described for Piesproops, in E. danae, the prosomal appendages are also arranged in a semi-circle around the mouth opening, which roughly resembles the condition in modern xiphosurans. Just the posterior closing of the feeding apparatus, the chilaria and spines on the proximal elements were not preserved in our material. The chelicerae, of which parts of different elements are preserved, appear similarly slender as in modern representatives, and also, their positioning seems to be about the same as today. Therefore, we assume that the feeding apparatus in E. danae functioned more or less like that of modern xiphosurans.

Enrolment and microtergite in Euproops danae

Several fossil euarthropods are known to perform enrolment, such as trilobites (e.g. Feist et al. 2010), agnostines (e.g. Müller and Walossek 1987) or certain nektaspidids (e.g. Budd 1999). Also for the different species of Euproops, it has been discussed whether enrolment occurred (Fisher 1977; Waterston 1985; Anderson 1994). Within the investigated material of Euproops danae, no enrolled specimens were observed. For enrolment, the question has to be solved if the morphotype of Euproops danae in principle has the ability to perform enrolment and if there are specialised coaptive structures to secure that position. This leads to the question if or in which way the walking legs would in such a case fit under the prosoma and if there is an additional microtergite, i.e. a small tergite between the prosomal shield and the thoracetron. (The discussion about the microtergite is also important in the context of the tagmatisation of xiphosurans, which is a long-standing discussion; see Haug et al. 2012a for details.)

For enrolling extant mantis shrimp larvae (Stomatopoda, Eucrustacea), it could be shown that all appendages fit into the chamber formed by shield, pleon and telson (Haug and Haug 2014), and the width of telson and ventral gape are aligned. A similar arrangement has been observed in larval false sand crabs (Rudolf et al. 2016).

In species of Euproops, however, the prosoma is larger than the opisthosoma and the appendages appear to be rather long, so it is questionable if these animals were specialised to enrolment; yet, that does not mean that they performed it in a less specialised way. Anderson (1994) suggested that enrolment was facilitated by a microtergite between prosoma and opisthosoma. We could not observe such a microtergite that would be clearly visible; however, it might still exist but be hardly detectable in the short jointed area between the prosoma and opisthosoma. As Haug et al. (2012a) already discussed, an extreme enrolment could still not be facilitated by a single (micro)tergite, but several joints would be required to achieve complete enrolment. Only a rather flexed position would already be possible with only one joint. A very probable function of enrolment in the species of Euproops is to protect the animals from predators, especially in combination with the long prosomal spines (Anderson 1994). Whether enrolment was indeed a behaviour performed by Euproops danae remains unclear until more reliable information about the dimensions of the appendages and information about the presence of a microtergite become available.

Differentiating the species of Euproops

Even though today xiphosurans are represented by only four species, this group was apparently much more diverse in the past. Three distinct groups existed in the Carboniferous, which have been assigned different names, but can generally be recognised as Bellinurus type, Euproops type and Limulus type (the first two being probably closely related; see Anderson and Selden 1997; Lamsdell 2016). However, several species within these groups appear to have been erected on relatively weak characters, also within Euproops. As a consequence, Anderson (1994) synonymised many species of Euproops except for E. danae, E. anthrax and E. rotundatus. However, Schultka (2000) pointed out that E. rotundatus has a rather widespread distribution, but does not occur in the Upper Carboniferous of North America, while E. danae is rather common there. This is insofar striking as, apart from that, a similar composition of the Upper Carboniferous biota has been assumed for the entire American-Asian area (Remy 1975). It raises the question if E. rotundatus and E. danae could represent a single species (as partly indicated by Schultka 2000; a thorough investigation of E. rotundatus is in preparation).

Within the three remaining species of Anderson (1994), E. danae, E. anthrax and E. rotundatus, three different epimera morphologies are known. E. anthrax shows pronounced spines; E. rotundatus has a real flange with very short spines, and E. danae is somehow intermediate with prominent spines but a slight indication of a flange. Concerning these relatively small differences between the different species of Euproops, it is quite possible that they may have evolved through heterochrony, i.e. through a shift in developmental timing (Haug et al. 2012a). More exactly, the disparity in the epimera shape would result from different growth patterns, which are not related to the general opisthosomal growth. Therefore, we need to compare not only adult morphologies but also ontogenetic sequences.

Comparison of Euproops danae and Piesproops

In the combined scatter plot of Piesproops and Euproops danae, both species match perfectly in their growth rate (Fig. 6). They show an almost identical trend line, and especially the earlier growth stages of both species allow a correlation of their respective stages. They display these similar growth rates and morphology despite their differences in the appearance of the epimera and the changes of the latter. In contrast to Euproops danae, which grows rather isometrically, Piesproops indeed exhibits significant allometric morphological changes during ontogeny: The epimera are narrow and bear elongate spines with many spinules or hairs in the early stages, but later, the bases of the spines become broader and form a pronounced flange around the opisthosoma (Haug et al. 2012a). In the latest stages, the epimera no longer appear spine-like, but display a trapezoidal form (Fig. 7). Furthermore, Piesproops displays a continuous increase in the relation between the smallest width of the epimera versus the distance of the bases of two spines. Between growth groups 5 and 6, a jump occurs, from 0.9 to a much higher relation of 1.6. Additionally, in Piesproops, initial epimeric ridges were first observed in stage 4, and further developed ridges in growth stage 5.

In Euproops danae, however, the morphological changes of the epimera follow a different pattern. Stages I–III do not show a recognisable change. Their morphology roughly correlates to stages 3–4 of Piesproops, while concerning the size, they roughly correlate to stages 4–5 of Piesproops. In short, E. danae appears more juvenile in its epimera morphology. The exact epimera morphology of stage IV is unclear, but is likely similar to stage III; size-wise, it correlates to about stage 6 of Piesproops.

This is also true for the later stages. In stage V of Euproops danae, no significant changes are visible, but the flange appears to be slightly broader. Morphologically, this stage therefore corresponds roughly to stage 4 of Piesproops, although the latter already possesses initial ridges that are absent in E. danae stage V specimens. Concerning the size, stage V of E. danae correlates roughly to stages 7–8 of Piesproops.

In stage VI of Euproops danae, ridges on the flange are present. Morphologically, it therefore correlates to stage 5 of Piesproops. Concerning size, it roughly corresponds to Piesproops stage 8. Stage VII has a slightly broader epimera flange corresponding to Piesproops stage 6, while in size, it corresponds to stages 8–9. In stage VIII, the epimera flange has broadened. It has reached a condition somewhere between Piesproops stages 6 and 7, in size about stage 9.

Hence, all E. danae stages appear more juvenile in morphology when compared to similar-sized Piesproops specimens. Overall, the changes of morphology are less pronounced when compared to Piesproops.

These differences in ontogenetic pattern clearly separate both species, while single morphological characters are not sufficient for the species differentiation. Therefore, as proposed by Haug et al. (2012a), it is important to take all available parts of the ontogenetic sequence into account when describing species. For this purpose, a reconsideration of the extension of the so-called hapantotype concept would be desirable. This concept currently allows treating different life stages of an individual as a single holotype, but only for protists. The extension of this concept also to metazoans has been proposed by Williams (1980, 1986), but has not been incorporated into the nomenclatural rules yet.

While E. danae and Piesproops could now be shown to represent distinct separate species, the species identity of E. rotundatus still needs to be re-evaluated. Therefore, as a next important step, the ontogeny of E. rotundatus will be reconstructed and compared to the findings reported here (in preparation).

As mentioned above, heterochrony may have played an important role for the rise of different Euproops species. Yet, with two different patterns in two different species, it is difficult to polarise these data, i.e. it remains unclear which of these two might be more ancestral and which one more derived. Also here, more data from different xiphosuran species will be necessary.


The results of our study support the results and assumptions of Haug et al. (2012a): Species of Euproops can only be properly distinguished if the ontogeny is taken into account. However, many of these data are not available yet, so it is crucial to reconstruct the ontogenetic sequences for E. rotundatus and E. anthrax. Subsequently, other fossil xiphosuran species should also be included into this analysis as in the other xiphosuran groups, the erection of fossil species was also often based on weak characters. Additionally, the question of the presence or absence of a microtergite and with the evolution of tagmatisation in the different xiphosuran species could also be approached with these additional data. All these data should result in a much deeper insight into xiphosuran evolution and possibly elucidate why this once so diverse group is nowadays reduced to only four species.


  1. Anderson LI (1994) Xiphosurans from the Westphalian D of the Radstock Basin, Somerset Coalfield, the South Wales Coalfield and Mazon Creek, Illinois. Proc Geol Assoc 105(4):265–275.

    Article  Google Scholar 

  2. Anderson LI, Selden PA (1997) Opisthosomal fusion and phylogeny of Palaeozoic Xiphosura. Lethaia 30:19–31

    Article  Google Scholar 

  3. Anger K (2001) The biology of decapod crustacean larvae. Crust Iss 14:1–420

    Google Scholar 

  4. Bellinger RG, Pienkowski RL (1987) Developmental polymorphism in the red-legged grasshopper, Melanoplus femurrubrum (DeGeer) (Orthoptera: Acrididae). Environ Entomol 16(1):120–125.

    Article  Google Scholar 

  5. Block W, Baust JG, Franks F, Johnston IA, Bale J (1990) Cold tolerance of insects and other arthropods [and discussion]. Phil T Roy Soc B 326(1237):613–633.

    Article  Google Scholar 

  6. Brauckmann C (1982) Der Schwertschwanz Euproops (Xiphosurida, Limulina, Euproopacea) aus dem Ober-Karbon des Piesbergs bei Osnabrück. Osnabrücker Naturwiss Mitt 9:17–26

    Google Scholar 

  7. Budd GE (1999) A nectaspid arthropod from the Early Cambrian Sirius Passet Fauna, with a description of retrodeformation based on functional morphology. Palaeontology 42(1):99–122.

    Article  Google Scholar 

  8. Carmichael RH, Rutecki D, Valiela I (2003) Abundance and population structure of the Atlantic horseshoe crab Limulus polyphemus in Pleasant Bay, Cape Cod. Mar Ecol Prog Ser 246:225–239

    Article  Google Scholar 

  9. Chen J, Waloszek D, Maas A (2004) A new ‘great-appendage’ arthropod from the Lower Cambrian of China and homology of chelicerate chelicerae and raptorial antero-ventral appendages. Lethaia 15:3–20

    Google Scholar 

  10. Chiu HMC, Morton B (2001) Growth and allometry of two horseshoe crab species, Tachypleus tridentatus and Carcinoscorpius rotundicauda (Xiphosura), in Hong Kong. Asian Mar Biol 18:129–141

    Google Scholar 

  11. Dunlop JA, Penney D, Tetlie OE, Anderson LI (2008) How many species of fossil arachnids are there? J Arachnol 36(2):267–272.

    Article  Google Scholar 

  12. Feist R, Lerosey-Aubril R, Johnson R (2010) Coaptative devices, enrollment, and life habits in Paralejurus, a particular case in scutelluid trilobites. Palaeobio Palaeoenv 90(2):125–137.

    Article  Google Scholar 

  13. Fields PG (1992) The control of stored-product insects and mites with extreme temperatures. J Stored Prod Res 28(2):89–118.

    Article  Google Scholar 

  14. Fisher DC (1977) Functional significance of spines in the Pennsylvanian horseshoe crab Euproops danae. Paleobiology 3(02):175–195.

    Article  Google Scholar 

  15. Fisher DC (1979) Evidence for subaerial activity of Euproops danae (Merostomata, Xiphosurida). In: Nitecki MH (ed) Mazon Creek fossils. Academic Press, New York, pp 379–447.

    Google Scholar 

  16. Haug C, Haug JT (2014) Defensive enrolment in mantis shrimp larvae. Contr Zool 83(3):185–194

    Google Scholar 

  17. Haug JT, Haug C, Kutschera V, Mayer G, Maas A, Liebau S, Castellani C, Wolfram U, Clarkson ENK, Waloszek D (2011) Autofluorescence imaging, an excellent tool for comparative morphology. J Microsc 244(3):259–272.

    Article  PubMed  Google Scholar 

  18. Haug C, Van Roy P, Leipner A, Funch P, Rudkin DM, Schöllmann L, Haug JT (2012a) A holomorph approach to xiphosuran evolution—a case study on the ontogeny of Euproops. Dev Genes Evol 222(5):253–268.

    Article  PubMed  Google Scholar 

  19. Haug JT, Mayer G, Haug C, Briggs DEG (2012b) A carboniferous non-onychophoran lobopodian reveals long-term survival of a Cambrian morphotype. Curr Biol 22(18):1673–1675.

    CAS  Article  PubMed  Google Scholar 

  20. Haug C, Nyborg T, Vega FJ (2013) An exceptionally preserved upogebiid (Decapoda: Reptantia) from the Eocene of California. Bol Soc Geol Mexic 65:235–248

    Google Scholar 

  21. Hauschke N, Wilde V (2004) Palaeogene limulids (Xiphosura) from Saxony-Anhalt (Germany) – systematics and palaeobiogeography. Hallesch Jb Geowiss B 18:161–168

    Google Scholar 

  22. Hennig W (1966) Phylogenetic systematics. University of Illinois Press, Urbana

    Google Scholar 

  23. Hu M, Kwan BKY, Wang Y, Cheung SG, Shin PKS (2015) Population structure and growth of juvenile horseshoe crabs Tachypleus tridentatus and Carcinoscorpius rotundicauda (Xiphosura) in southern China. In: Carmichael R, Botton M, Shin P, Cheung S (eds) Changing global perspectives on horseshoe crab biology, conservation and management. Springer International Publishing, Cham, pp 167–180.

    Google Scholar 

  24. Lamsdell JC (2013) Revised systematics of Palaeozoic ‘horseshoe crabs’ and the myth of monophyletic Xiphosura. Zool J Linn Soc Lond 167(1):1–27.

  25. Lamsdell JC (2016) Horseshoe crab phylogeny and independent colonizations of fresh water: ecological invasion as a driver for morphological innovation. Palaeontology 59(2):181–194.

    Article  Google Scholar 

  26. Lamsdell JC, McKenzie SC (2015) Tachypleus syriacus (Woodward)—a sexually dimorphic Cretaceous crown limulid reveals underestimated horseshoe crab divergence times. Org Divers Evol 15:681–693

  27. Loveland RE, Botton ML (1992) Size dimorphism and the mating system in horseshoe crabs, Limulus polyphemus L. Animal Behav 44(5):907–916.

    Article  Google Scholar 

  28. Müller KJ, Walossek D (1987) Morphology, ontogeny, and life habit of Agnostus pisiformis from the Upper Cambrian of Sweden. Fossils Strata 19:1–124

    Google Scholar 

  29. Racheboeuf PR, Vannier J, Anderson LI (2002) A new three-dimensionally preserved xiphosuran chelicerate from the Montceau-Les-Mines Lagerstätte (Carboniferous, France). Palaeontology 45(1):125–147

    Article  Google Scholar 

  30. Reisinger PWM, Tutter I, Welsch U (1991) Fine structure of the gills of the horseshoe crabs Limulus polyphemus and Tachypleus tridentatus and of the book lungs of the spider Eurypelma californicum. Zool Jahrb Anat 121:331–357

    Google Scholar 

  31. Remy W (1975) The floral changes at the Carboniferous-Permian boundary in Europe and North America. In: Barlow JA (ed) The Age of the Dunkard - Proceedings of the First I. C. White Memorial Symposium, West Virginia Geological and Economic Survey, Morgantown, pp 305–352

  32. Rudolf NR, Haug C, Haug JT (2016) Functional morphology of giant mole crab larvae: a possible case of defensive enrollment. Zool Lett 2:Art 17

    Article  Google Scholar 

  33. Schultka S (2000) Zur Palökologie der Euproopiden im Nordwestdeutschen Oberkarbon. Mitt Mus Naturwiss Berl Geowiss R 3:87–98

    Google Scholar 

  34. Shuster CN Jr, Sekiguchi K (2003) Growing up takes about ten years and eighteen stages. In: Shuster CN Jr, Barlow RB, Brockmann HJ (eds) The American Horseshoe Crab. Harvard University Press, Cambridge, pp 103–132

    Google Scholar 

  35. Waterman TH (1954) Relative growth and the compound eye in Xiphosura. J Morph 95(1):125–158.

    Article  Google Scholar 

  36. Waterston CD (1985) Chelicerata from the Dinantian of Foulden, Berwickshire, Scotland. Trans R Soc Edinb Earth Sci 76(01):25–33.

    Article  Google Scholar 

  37. Williams RB (1980) Comment on report of the committee of typification of species of Protozoa. Bull Zool Nomencl 37:137–140.

    Article  Google Scholar 

  38. Williams RB (1986) Hapantotypes: a possible solution to some problems of parasite nomenclature. Parasitol Today 2(11):314–316.

    CAS  Article  PubMed  Google Scholar 

Download references


We would like to thank Susan Butts, Jessica Utrup, Eric Lazo-Wasem, Daniel Drew and Lourdes Rojas for their help in the collections of the Yale Peabody Museum, New Haven, CT, USA. Derek Briggs, Yale University, Steffen Harzsch, University of Greifswald, and J. Matthias Starck, LMU Munich, are thanked for discussions and support. We are very grateful to Joachim T. Haug, LMU Munich, for continuous help with the imaging and fruitful discussions during the entire study. CH was kindly supported by a Bavarian Equal Opportunities scholarship of the LMU and is currently funded by the German Research Foundation (DFG HA 7066/3-1).

Author information



Corresponding author

Correspondence to Carolin Haug.

Additional information

Communicated by Nico Posnien

Electronic supplementary material

Supplementary Fig. 1

Non-stereo versions of Fig. 1, otherwise identical with it. (GIF 131 kb)

Supplementary Fig. 2

Non-stereo versions of Fig. 2, otherwise identical with it. (GIF 418 kb)

Supplementary Fig. 3

Non-stereo versions of Fig. 3, otherwise identical with it. (GIF 463 kb)

Supplementary Fig. 4

Non-stereo versions of Fig. 5, otherwise identical with it. (GIF 313 kb)

High Resolution Image (TIFF 6848 kb)

High Resolution Image (TIFF 19600 kb)

High Resolution Image (TIFF 19438 kb)

High Resolution Image (TIFF 11423 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Haug, C., Rötzer, M.A.I.N. The ontogeny of the 300 million year old xiphosuran Euproops danae (Euchelicerata) and implications for resolving the Euproops species complex. Dev Genes Evol 228, 63–74 (2018).

Download citation


  • Fossilised ontogeny
  • Xiphosura sensu stricto
  • Euchelicerata
  • Heterochrony
  • Hapantotype