Eurypterid

Size comparison of six of the largest eurypterids: Pterygotus grandidentatus, Pentecopterus decorahensis, Acutiramus macrophthalmus, A. bohemicus, Carcinosoma punctatum, and Jaekelopterus rhenaniae

Eurypterids were highly variable in size, depending on factors such as lifestyle, living environment and taxonomic affinity. Sizes around 100 centimeters (3.3 ft) are common in most eurypterid groups.[6] The smallest eurypterid, Alkenopterus burglahrensis, measured just 2.03 centimeters (0.80 in) in length.[7]

The largest eurypterid, and the largest known arthropod ever to have lived, is Jaekelopterus rhenaniae. A chelicera from the Emsian Klerf Formation of Willwerath, Germany measured 36.4 centimeters (14.3 in) in length, but is missing a quarter of its length, suggesting that the full chelicera would have been 45.5 centimeters (17.9 in) long. If the proportions between body length and chelicerae match those of its closest relatives, where the ratio between claw size and body length is relatively consistent, the specimen of Jaekelopterus that possessed the chelicera in question would have measured between 233 and 259 centimeters (7.64 and 8.50 ft), an average 2.5 meters (8.2 ft), in length. With the chelicerae extended, another meter (3.28 ft) would be added to this length. This estimate exceeds the maximum body size of all other known giant arthropods by almost half a meter (1.64 ft) even if the extended chelicerae are not included.[8] Two other eurypterids have also been estimated to have reached lengths of 2.5 metres; Erettopterus grandis (closely related to Jaekelopterus) and Hibbertopterus wittebergensis, but E. grandis is very fragmentary and the H. wittenbergensis size estimate is based on trackway evidence, not fossil remains.[9]

The family of Jaekelopterus, the Pterygotidae, is noted for several unusually large species. Both Acutiramus, whose largest member A. bohemicus measured 2.1 meters (6.9 ft), and Pterygotus, whose largest species P. grandidentatus measured 1.75 meters (5.7 ft), were gigantic.[8] Several different contributing factors to the large size of the pterygotids have been suggested, including courtship behaviour, predation and competition over environmental resources.[10]

Giant eurypterids were not limited to the family Pterygotidae. An isolated 12.7 centimeters (5.0 in) long fossil metastoma of the carcinosomatoid eurypterid Carcinosoma punctatum indicates the animal would have reached a length of 2.2 meters (7.2 ft) in life, rivalling the pterygotids in size.[11] Another giant was Pentecopterus decorahensis, a primitive carcinosomatoid, which is estimated to have reached lengths of 1.7 meters (5.6 ft).[12]

Typical of large eurypterids is a lightweight build. Factors such as locomotion, energy costs in molting and respiration, as well as the actual physical properties of the exoskeleton, limits the size that arthropods can reach. A lightweight construction significantly decreases the influence of these factors. Pterygotids were particularly lightweight, with most fossilized large body segments preserving as thin and unmineralized.[8] Lightweight adaptations are present in other giant paleozoic arthropods as well, such as the giant millipede Arthropleura, and are possibly vital for the evolution of giant size in arthropods.[8][13]

In addition to the lightweight giant eurypterids, some deep-bodied forms in the family Hibbertopteridae were also very large. A carapace from the Carboniferous of Scotland referred to the species Hibbertoperus scouleri measures 65 cm (26 in) wide. As Hibbertopterus was very wide compared to its length, the animal in question could possibly have measured just short of 2 meters (6.6 ft) in length. More robust than the pterygotids, this giant Hibbertopterus would possibly have rivalled the largest pterygotids in weight, if not surpassed them, and as such be among the heaviest arthropods.[14]

Illustration of subaqueous flight in Eurypterus in which the shape of the paddles and their motion through water is enough to generate lift. This type of locomotion was limited to eurypterine (with swimming paddles) eurypterids.

Illustration of Hibbertopterus, a large stylonurine (lacking swimming paddles) eurypterid

The two eurypterid suborders, Eurypterina and Stylonurina, are distinguished primarily by the morphology of their final pair of appendages. In the Stylonurina, this appendage takes the form of a long and slender walking leg, while in the Eurypterina, the leg is modified and broadened into a swimming paddle.[15] Other than the swimming paddle, the legs of many eurypterines were far too small to do much more than allow them to crawl across the sea floor. In contrast, a number of stylonurines had elongated and powerful legs that might have allowed them to walk on land (similar to modern crabs).[16]

A fossil trackway was discovered in Carboniferous-aged fossil deposits of Scotland in 2005. It was attributed to the stylonurine eurypterid Hibbertopterus due to a matching size (the trackmaker was estimated to have been about 1.6 meters (5.2 ft) long) and inferred leg anatomy. It is the largest terrestrial trackway—measuring 6 meters (20 ft) long and averaging 95 centimeters (3.12 ft) in width—made by an arthropod found thus far. It is the first record of land locomotion by a eurypterid. The trackway provides evidence that some eurypterids could survive in terrestrial environments, at least for short periods of time, and reveals information about the stylonurine gait. In Hibbertopterus, as in most eurypterids, the pairs of appendages are different in size (referred to as a heteropodous limb condition). These differently sized pairs would have moved in phase, and the short stride length indicates that Hibbertopterus crawled with an exceptionally slow speed, at least on land. The large telson was dragged along the ground and left a large central groove behind the animal. Slopes in the tracks at random intervals suggest that the motion was jerky.[17] The gait of smaller stylonurines, such as Parastylonurus, was probably faster and more precise.[18]

The functionality of the eurypterine swimming paddles varied from group to group. In the Eurypteroidea, the paddles were similar in shape to oars. The condition of the joints in their appendages ensured their paddles could only be moved in near-horizontal planes, not upwards or downwards. Some other groups, such as the Pterygotioidea, would not have possessed this condition and were probably able to swim faster.[19] Most eurypterines are generally agreed to have utilized a rowing type of propulsion similar to that of crabs and water beetles. Larger individuals may have been capable of underwater flying (or subaqueous flight) in which the motion and shape of the paddles are enough to generate lift, similar to the swimming of sea turtles and sea lions. This type of movement has a relatively slower acceleration rate than the rowing type, especially since adults have proportionally smaller paddles than juveniles. However, since the larger sizes of adults mean a higher drag coefficient, using this type of propulsion is more energy-efficient.[20]

The holotype of Palmichnium kosinkiorum, containing the largest eurypterid footprints known.

Some eurypterines, such as Mixopterus (as inferred from attributed fossil trackways), were not necessarily good swimmers. It likely kept mostly to the bottom, using its swimming paddles for occasional bursts of movements vertically, with the fourth and fifth pairs of appendages positioned backwards to produce minor movement forwards. While walking, it probably used a gait like that of most modern insects. The weight of its long abdomen would have been balanced by two heavy and specialized frontal appendages, and the center of gravity might have been adjustable by raising and positioning the tail.[21]

Preserved fossilized eurypterid trackways tend to be large and heteropodous and often have an associated telson drag mark along the mid-line (as with the Scottish Hibbertopterus track). Such trackways have been discovered on every continent except for South America. In some places where eurypterid fossil remains are otherwise rare, such as in South Africa and the rest of the former supercontinent Gondwana, the discoveries of trackways both predate and outnumber eurypterid body fossils.[22] Eurypterid trackways have been referred to several ichnogenera, most notably Palmichnium (defined as a series of four tracks often with an associated drag mark in the mid-line),[23] wherein the holotype of the ichnospecies P. kosinkiorum preserves the largest eurypterid footprints known to date with the found tracks each being about 7.6 centimeters (3.0 in) in diameter.[24] Other eurypterid ichnogenera include Merostomichnites (though it is likely that many specimens actually represent trackways of crustaceans) and Arcuites (which preserves grooves made by the swimming appendages).[23][25][26]

The supposed "gill tracts" of eurypterids have been compared to the air-breathing pseudotracheae present in the posterior legs of modern isopods, such as Oniscus (pictured).

In eurypterids, the respiratory organs were located on the ventral body wall (the underside of the opisthosoma). Blatfüsse, evolved from opisthosomal appendages, covered the underside and created a gill chamber where the "gill tracts" were located.[2] Depending on the species, the eurypterid gill tract was either triangular or oval in shape and was possibly raised into a cushion-like state. The surface of this gill tract bore several spinules (small spines), which resulted in an enlarged surface area. It was composed of spongy tissue due to many invaginations in the structure.[27]

Though the Kiemenplatte is referred to as a "gill tract", it may not necessarily have functioned as actual gills. In other animals, gills are used for oxygen uptake from water and are outgrowths of the body wall. Despite eurypterids clearly being primarily aquatic animals that almost certainly evolved underwater (some eurypterids, such as the pterygotids, would even have been physically unable to walk on land), it is unlikely the "gill tract" contained functional gills when comparing the organ to gills in other invertebrates and even fish. Previous interpretations often identified the eurypterid "gills" as homologous with those of other groups (hence the terminology), with gas exchange occurring within the spongy tract and a pattern of branchio-cardiac and dendritic veins (as in related groups) carrying oxygenated blood into the body. The primary analogy used in previous studies has been horseshoe crabs, though their gill structure and that of eurypterids are remarkably different. In horseshoe crabs, the gills are more complex and composed of many lamellae (plates) which give a larger surface area used for gas exchange. In addition, the gill tract of eurypterids is proportionally much too small to support them if it is analogous to the gills of other groups. To be functional gills, they would have to have been highly efficient and would have required a highly efficient circulatory system. It is considered unlikely, however, that these factors would be enough to explain the large discrepancy between gill tract size and body size.[28]

It has been suggested instead that the "gill tract" was an organ for breathing air, perhaps actually being a lung, plastron or a pseudotrachea.[29] Plastrons are organs that some arthropods evolved secondarily to breathe air underwater. This is considered an unlikely explanation since eurypterids had evolved in water from the start and they would not have organs evolved from air-breathing organs present. In addition, plastrons are generally exposed on outer parts of the body while the eurypterid gill tract is located behind the Blatfüssen.[30] Instead, among arthropod respiratory organs, the eurypterid gill tracts most closely resemble the pseudotracheae found in modern isopods. These organs, called pseudotracheae, because of some resemblance to the tracheae (windpipes) of air-breathing organisms, are lung-like and present within the pleopods (back legs) of isopods. The structure of the pseudotracheae has been compared to the spongy structure of the eurypterid gill tracts. It is possible the two organs functioned in the same way.[31]

Some researchers have suggested that eurypterids may have been adapted to an amphibious lifestyle, using the full gill tract structure as gills and the invaginations within it as pseudotrachea. This mode of life may not have been physiologically possible, however, since water pressure would have forced water into the invaginations leading to asphyxiation. Furthermore, most eurypterids would have been aquatic their entire lives. No matter how much time was spent on land, organs for respiration in underwater environments must have been present. True gills, expected to have been located within the branchial chamber within the Blatfüssen, remain unknown in eurypterids.[31]

Larval (left) and juvenile (right) instars of Strobilopterus (not to scale)

Like all arthropods, eurypterids matured and grew through static developmental stages referred to as instars. These instars were punctuated by periods during which eurypterids went through ecdysis (molting of the cuticle) after which they underwent rapid and immediate growth. Some arthropods, such as insects and many crustaceans, undergo extreme changes over the course of maturing. Chelicerates, including eurypterids, are in general considered to be direct developers, undergoing no extreme changes after hatching (though extra body segments and extra limbs may be gained over the course of ontogeny in some lineages, such as xiphosurans and sea spiders). Whether eurypterids were true direct developers (with hatchlings more or less being identical to adults) or hemianamorphic direct developers (with extra segments and limbs potentially being added during ontogeny) has been controversial in the past.[32]

Hemianamorphic direct development has been observed in many arthropod groups, such as trilobites, megacheirans, basal crustaceans and basal myriapods. True direct development has on occasion been referred to as a trait unique to arachnids. There have been few studies on eurypterid ontogeny as there is a general lack of specimens in the fossil record that can confidently be stated to represent juveniles.[32] It is possible that many eurypterid species thought to be distinct actually represent juvenile specimens of other species, with paleontologists rarely considering the influence of ontogeny when describing new species.[33]

Studies on a well-preserved fossil assemblage of eurypterids from the Pragian-aged Beartooth Butte Formation in Cottonwood Canyon, Wyoming, composed of multiple specimens of various developmental stages of eurypterids Jaekelopterus and Strobilopterus, revealed that eurypterid ontogeny was more or less parallel and similar to that of extinct and extant xiphosurans, with the largest exception being that eurypterids hatched with a full set of appendages and opisthosomal segments. Eurypterids were thus not hemianamorphic direct developers, but true direct developers like modern arachnids.[34]

The most frequently observed change occurring through ontogeny (except for some genera, such as Eurypterus, which appear to have been static) is the metastoma becoming proportionally less wide. This ontogenetic change has been observed in members of several superfamilies, such as the Eurypteroidea, the Pterygotioidea and the Moselopteroidea.[35]

Pterygotus depicted hunting Birkenia.

No fossil gut contents from eurypterids are known, so direct evidence of their diet is lacking. The eurypterid biology is particularly suggestive of a carnivorous lifestyle. Not only were many large (in general, most predators tend to be larger than their prey), but they had stereoscopic vision (the ability to perceive depth).[36] The legs of many eurypterids were covered in thin spines, used both for locomotion and the gathering of food. In some groups, these spiny appendages became heavily specialized. In some eurypterids in the Carcinosomatoidea, forward-facing appendages were large and possessed enormously elongated spines (as in Mixopterus and Megalograptus). In derived members of the Pterygotioidea, the appendages were completely without spines, but had specialized claws instead.[37] Other eurypterids, lacking these specialized appendages, likely fed in a manner similar to modern horseshoe crabs, by grabbing and shredding food with their appendages before pushing it into their mouth using their chelicerae.[38]

Fossils preserving digestive tracts have been reported from fossils of various eurypterids, among them Carcinosoma, Acutiramus and Eurypterus. Though a potential anal opening has been reported from the telson of a specimen of Buffalopterus, it is more likely that the anus was opened through the thin cuticle between the last segment before the telson and the telson itself, as in modern horseshoe crabs.[36]

Eurypterid coprolites discovered in deposits of Ordovician age in Ohio containing fragments of a trilobite and eurypterid Megalograptus ohioensis in association with full specimens of the same eurypterid species have been suggested to represent evidence of cannibalism. Similar coprolites referred to the species Lanarkopterus dolichoschelus from the Ordovician of Ohio contain fragments of jawless fish and fragments of smaller specimens of Lanarkopterus itself.[36]

Though apex predatory roles would have been limited to the very largest eurypterids, smaller eurypterids were likely formidable predators in their own right just like their larger relatives.[6]

Type A genital appendage of Adelophthalmus mansfieldi

Type B genital appendage of Kokomopterus longicaudatus

As in many other entirely extinct groups, understanding and researching the reproduction and sexual dimorphism of eurypterids is difficult, as they are only known from fossilized shells and carapaces. In some cases, there might not be enough apparent differences to separate the sexes based on morphology alone.[16] Sometimes two sexes of the same species have been interpreted as two different species, as was the case with two species of Drepanopterus (D. bembycoides and D. lobatus).[39]

The eurypterid prosoma is made up of the first six exoskeleton segments fused together into a larger structure. The seventh segment (thus the first opisthosomal segment) is referred to as the metastoma and the eighth segment (distinctly plate-like) is called the operculum and contains the genital aperature. The underside of this segment is occupied by the genital operculum, a structure originally evolved from ancestral seventh and eighth pair of appendages. In its center, as in modern horseshoe crabs, is a genital appendage. This appendage, an elongated rod with an internal duct, is found in two distinct morphs, generally referred to as "type A" and "type B".[16] These genital appendages are often preserved prominently in fossils and have been the subject of various interpretations of eurypterid reproduction and sexual dimorphism.[40]

Type A appendages are generally longer than those of type B. In some genera they are divided into different numbers of sections, such as in Eurypterus where the type A appendage is divided into three but the type B appendage into only two.[41] Such division of the genital appendage is common in eurypterids, but the number is not universal; for instance, the appendages of both types in the family Pterygotidae are undivided.[42] The type A appendage is also armed with two curved spines called furca (lit. 'fork' in Latin). The presence of furca in the type B appendage is also possible and the structure may represent the unfused tips of the appendages. Located between the dorsal and ventral surfaces of the Blatfüsse associated with the type A appendages is a set of organs traditionally described as either "tubular organs" or "horn organs". These organs are most often interpreted as spermathecae (organs for storing sperm), though this function is yet to be proven conclusively.[43] In arthropods, spermathecae are used to store the spermatophore received from males. This would imply that the type A appendage is the female morph and the type B appendage is the male.[16] Further evidence for the type A appendages representing the female morph of genital appendages comes in their more complex construction (a general trend for female arthropod genitalia). It is possible that the greater length of the type A appendage means that it was used as an ovipositor (used to deposit eggs).[44] The different types of genital appendages are not necessarily the only feature that distinguishes between the sexes of eurypterids. Depending on the genus and species in question, other features such as size, the amount of ornamentation and the proportional width of the body can be the result of sexual dimorphism.[2] In general, eurypterids with type B appendages (males) appear to have been proportionally wider than eurypterids with type A appendages (females) of the same genera.[45]

The primary function of the long, assumed female, type A appendages was likely to take up spermatophore from the substrate into the reproductive tract rather than to serve as an ovipositor, as arthropod ovipositors are generally longer than eurypterid type A appendages. By rotating the sides of the operculum, it would have been possible to lower the appendage from the body. Due to the way different plates overlay at its location, the appendage would have been impossible to move without muscular contractions moving around the operculum. It would have been kept in place when not it use. The furca on the type A appendages may have aided in breaking open the spermatophore to release the free sperm inside for uptake. The "horn organs," possibly spermathecae, are thought to have been connected directly to the appendage via tracts, but these supposed tracts remain unpreserved in available fossil material.[46]

Type B appendages, assumed male, would have produced, stored and perhaps shaped spermatophore in a heart-shaped structure on the dorsal surface of the appendage. A broad genital opening would have allowed large amounts of spermatophore to be released at once. The long furca associated with type B appendages, perhaps capable of being lowered like the type A appendage, could have been used to detect whether a substrate was suitable for spermatophore deposition.[47]

A reconstruction of Pentecopterus, the earliest known eurypterid. The family to which Pentecopterus belongs, the Megalograptidae, was the first truly successful eurypterid group.

Until 1882 no eurypterids were known from before the Silurian. Contemporary discoveries since the 1880s have expanded the knowledge of early eurypterids from the Ordovician period.[48] The earliest eurypterids known today, the megalograptid Pentecopterus, date from the Darriwilian stage of the Middle Ordovician, 467.3 million years ago.[49] There are also reports of even earlier fossil eurypterids in deposits of Late Tremadocian (Early Ordovician) age in Morocco, but these have yet to be thoroughly studied.[50]

Pentecopterus was a relatively derived eurypterid, part of the megalograptid family within the carcinosomatoid superfamily. Its derived position suggests that most eurypterid clades, at least within the eurypterine suborder, had already been established at this point during the Middle Ordovician.[51] The earliest known stylonurine eurypterid, Brachyopterus,[6] is also Middle Ordovician in age. The presence of members of both suborders indicates that primitive stem-eurypterids would have preceded them, though these are so far unknown in the fossil record. The presence of several eurypterid clades during the Middle Ordovician suggests that eurypterids either originated during the Early Ordovician and experienced a rapid and explosive radiation and diversification soon after the first forms evolved, or that the group originated much earlier, perhaps during the Cambrian period.[51]

As such, the exact eurypterid time of origin remains unknown. Though fossils referred to as "primitive eurypterids" have occasionally been described from deposits of Cambrian or even Precambrian age,[52] they are not recognized as eurypterids, and sometimes not even as related forms, today. Some animals previously seen as primitive eurypterids, such as the genus Strabops from the Cambrian of Missouri,[53] are now classified as aglaspidids or strabopids. The aglaspidids, once seen as primitive chelicerates, are now seen as a group more closely related to trilobites.[54]

The fossil record of Ordovician eurypterids is quite poor. The majority of eurypterids once reportedly known from the Ordovician have since proven to be misidentifications or pseudofossils. Today only 11 species can be confidently identified as representing Ordovician eurypterids. These taxa fall into two distinct ecological categories; large and active predators from the ancient continent of Laurentia, and demersal (living on the seafloor) and basal animals from the continents Avalonia and Gondwana.[49] The Laurentian predators, classified in the family Megalograptidae (comprising the genera Echinognathus, Megalograptus and Pentecopterus), are likely to represent the first truly successful eurypterid group, experiencing a small radiation during the Late Ordovician.[55]

Reconstruction of Erettopterus, a member of the highly successful Silurian and Devonian eurypterid family Pterygotidae

Eurypterids were most diverse and abundant between the Middle Silurian and the Early Devonian, with an absolute peak in diversity during the Pridoli epoch, 423 to 419.2 million years ago, of the very latest Silurian.[15] This peak in diversity has been recognized since the early twentieth century; of the approximately 150 species of eurypterids known in 1916, more than half were from the Silurian and a third were from the Late Silurian alone.[48]

Though stylonurine eurypterids generally remained rare and low in number, as had been the case during the preceding Ordovician, eurypterine eurypterids experienced a rapid rise in diversity and number.[56] In most Silurian fossil beds, eurypterine eurypterids account for 90% of all eurypterids present.[57] Though some were likely already present by the Late Ordovician (simply missing from the fossil record so far),[51] a vast majority of eurypterid groups are first recorded in strata of Silurian age. These include both stylonurine groups such as the Stylonuroidea, Kokomopteroidea and Mycteropoidea as well as eurypterine groups such as the Pterygotioidea, Eurypteroidea and Waeringopteroidea.[58]

The most successful eurypterid by far was the Middle to Late Silurian Eurypterus, a generalist, equally likely to have engaged in predation or scavenging. Thought to have hunted mainly small and soft-bodied invertebrates, such as worms,[59] species of the genus (of which the most common is the type species, E. remipes) account for more than 90% (perhaps as many as 95%) of all known fossil eurypterid specimens.[57] Despite their vast number, Eurypterus are only known from a relatively short temporal range, first appearing during the Late Llandovery epoch (around 432 million years ago) and being extinct by the end of the Pridoli epoch.[60] Eurypterus was also restricted to the continent Euramerica (composed of the equatorial continents Avalonia, Baltica and Laurentia), which had been completely colonized by the genus during its merging and was unable to cross the vast expanses of ocean separating this continent from other parts of the world, such as the southern supercontinent Gondwana. As such, Eurypterus was limited geographically to the coastlines and shallow inland seas of Euramerica.[57][61]

During the Late Silurian the pterygotid eurypterids, large and specialized forms with several new adaptations, such as large and flattened telsons capable of being used as rudders, and large and specialized chelicerae with enlarged pincers for handling (and potentially in some cases killing) prey appeared.[3][4] Though the largest members of the family appeared in the Devonian, large two meter (6.5+ ft) pterygotids such as Acutiramus were already present during the Late Silurian.[9] Their ecology ranged from generalized predatory behavior to ambush predation and some, such as Pterygotus itself, were active apex predators in Late Silurian marine ecosystems.[62] The pterygotids were also evidently capable of crossing oceans, becoming one of only two eurypterid groups to achieve a cosmopolitan distribution.[63]

Reconstruction of Adelophthalmus, the only eurypterine (with swimming paddles) eurypterid to survive the Late Devonian extinction and persist into the subsequent Carboniferous and Permian periods.

Though the eurypterids continued to be abundant and diversify during the Early Devonian (for instance leading to the evolution of the pterygotid Jaekelopterus, the largest of all arthropods), the group was one of many heavily affected by the Late Devonian extinction. The extinction event, only known to affect marine life (particularly trilobites, brachiopods and reef-building organisms) effectively crippled the abundance and diversity previously seen within the eurypterids.[64]

A major decline in diversity had already begun during the Early Devonian and eurypterids were rare in marine environments by the Late Devonian. During the Frasnian stage four families went extinct, and the later Famennian saw an additional five families going extinct.[64] As marine groups were the most affected, the eurypterids were primarily impacted within the eurypterine suborder. Only one group of stylonurines (the family Parastylonuridae) went extinct in the Early Devonian. Only two families of eurypterines survived into the Late Devonian at all (Adelophthalmidae and Waeringopteridae). The eurypterines experienced their most major declines in the Early Devonian, during which over 50% of their diversity was lost in just 10 million years. Stylonurines, on the other hand, persisted through the period with more or less consistent diversity and abundance but were affected during the Late Devonian, when many of the older groups were replaced by new forms in the families Mycteroptidae and Hibbertopteridae.[65]

It is possible that the catastrophic extinction patterns seen in the eurypterine suborder were related to the emergence of more derived fish. Eurypterine decline began at the point when jawless fish first became more developed and coincides with the emergence of placoderms (armored fish) in both North America and Europe.[66]

Stylonurines of the surviving hibbertopterid and mycteroptid families completely avoided competition with fish by evolving towards a new and distinct ecological niche. These families experienced a radiation and diversification through the Late Devonian and Early Carboniferous, the last ever radiation within the eurypterids, which gave rise to several new forms capable of "sweep-feeding" (raking through the substrate in search of prey).[67]

Reconstruction of the hibbertopterid Campylocephalus, the last known surviving eurypterid. Campylocephalus went extinct during the Permian-Triassic extinction event or sometime shortly before.

Only three eurypterid families—Adelophthalmidae, Hibbertopteridae and Mycteroptidae—survived the extinction event in its entirety. These were all freshwater animals, rendering the eurypterids extinct in marine environments.[64] With marine eurypterid predators gone, sarcopterygian fish, such as the rhizodonts, were the new apex predators in marine environments.[66] The sole surviving eurypterine family, Adelophthalmidae, was represented by only a single genus, Adelophthalmus. The hibbertopterids, mycteroptids and Adelophthalmus survived into the Permian.[68]

Adelophthalmus became the most common of all late Paleozoic eurypterids, existing in greater number and diversity than surviving stylonurines, and diversified in the absence of other eurypterines.[69] Out of the 33 species referred to Adelophthalmus, 23 (69%) are from the Carboniferous alone.[70][71] The genus reached its peak diversity in the Late Carboniferous. Though Adelophthalmus had already been relatively widespread and represented around all major landmasses in the Late Devonian, the amalgamation of Pangaea into a global supercontinent over the course of the last two periods of the Paleozoic allowed Adelophthalmus to gain an almost worldwide distribution.[57]

During the Late Carboniferous and Early Permian Adelophthalmus was widespread, living primarily in brackish and freshwater environments adjacent to coastal plains. These environments were maintained by favorable climate conditions. They did not persist as climate changes owing to Pangaea's formation altered depositional and vegetational patterns across the world. With their habitat gone, Adelophthalmus dwindled in number and had already gone extinct by the Leonardian stage of the Early Permian.[72]

Mycteroptids and hibbertopterids continued to survive for some time, with one genus of each group known from Permian strata: Hastimima and Campylocephalus respectively.[73] Hastimima went extinct during the Early Permian,[74] as Adelophthalmus had, while Campylocephalus persisted longer. A massive incomplete carapace from Permian deposits in Russia represents the sole fossil remains of the species C. permianus, which might have reached 1.4 meters (4.6 ft) in length, while originally considered to be Late Permian in age[9] other sources suggest a Kungurian-Roadian age for the specimen.[75] A fragment of eurypterid cuticle, given the species name Woodwardopterus freemanorum is known from the terminal Permian of Australia, which represents the youngest known eurypterid.[75] No eurypterids are known from fossil beds higher than the Permian. This indicates that the last eurypterids died either in the catastrophic extinction event at its end or at some point shortly before it. This extinction event, the Permian–Triassic extinction event, is the most devastating mass extinction recorded, and rendered many other successful Paleozoic groups, such as the trilobites, extinct.[76]

The most important taxonomic character in eurypterids is the morphology of the sixth pair of appendages. In most eurypterids of the suborder Eurypterina, such as Bassipterus (top), this leg is modified into a swimming paddle. In eurypterids of the suborder Stylonurina, such as Parastylonurus (bottom), it is not.

The internal classification of eurypterids within the Eurypterida is based mainly on eleven established characters. These have been used throughout the history of eurypterid research to establish clades and genera. These characters include: the shape of the prosoma, the shape of the metastoma, the shape and position of the eyes, the types of prosomal appendages, the types of swimming leg paddles, the structure of the doublure (the fringe of the dorsal exoskeleton), the structure of the opithosoma, the structure of the genital appendages, the shape of the telson and the type of ornamentation present. It is worth noting that not all of these characters are of equal taxonomic importance.[97] They are not applicable to all eurypterids either; stylonurine eurypterids lack swimming leg paddles entirely.[15] Some characters, including the prosoma and metastoma shapes and the position and shapes of the eyes, are seen as important only for the distinction between different genera.[98] Most superfamilies and families are defined based on the morphology of the appendages.[99]

The most important character used in eurypterid taxonomy is the type of prosomal appendages as this character is used to define entire suborders. General leg anatomy can also be used to define superfamilies and families. Historically, the chelicerae were considered the most important appendages from a taxonomical standpoint since they only occurred in two general types: a eurypterid type with small and toothless pincers and a pterygotid type with large pincers and teeth. This distinction has historically been used to divide the Eurypterida into the two suborders Eurypterina (small chelicerae) and "Pterygotina" (large and powerful chelicerae).[100] This classification scheme is not without problems. In Victor Tollerton's 1989 taxonomic revision of the Eurypterida, with suborders Eurypterina and Pterygotina recognized, several clades of eurypterids today recognized as stylonurines (including hibbertopterids and mycteroptids) were reclassified as non-eurypterids in the new separate order "Cyrtoctenida" on the grounds of perceived inconsistencies in the prosomal appendages.[101]

Modern research favors a classification into suborders Eurypterina and Stylonurina instead, supported by phylogenetic analyses.[102][35] In particular, pterygotid eurypterids share a number of homologies with derived eurypterine eurypterids such as the adelophthalmids, and are thus best classified as derived members of the same suborder.[103] In the Stylonurina, the sixth pair of appendages is represented by long and slender walking legs and lack a modified spine (referred to as the podomere 7a). In most eurypterids in the Eurypterina, the sixth pair of appendages is broadened into swimming paddles and always has a podomere 7a. 75% of eurypterid species are eurypterines and they represent 99% of all fossil eurypterid specimens.[15] Of all eurypterid clades, the Pterygotioidea is the most species-rich, with over 50 species. The second most species-rich clade is the Adelophthalmoidea, with over 40 species.[57]

The cladogram presented below, covering all currently recognized eurypterid families, follows a 2007 study by O. Erik Tetlie.[104] The stylonurine suborder follows a 2010 study by James Lamsdell, Simon J. Braddy and Tetlie.[105] The superfamily "Megalograptoidea", recognized by Tetlie in 2007 and then placed between the Onychopterelloidea and Eurypteroidea, has been omitted as more recent studies suggest that the megalograptids were members of the superfamily Carcinosomatoidea. As such, the phylogeny of the Carcinosomatoidea follows a 2015 study by Lamsdell and colleagues.[106]

• Kazlev, M. Alan (2002). "Palaeos - Eurypterida". www.palaeos.com. Archived from the original on 13 August 2007.