Evolution Icon Evolution
Intelligent Design Icon Intelligent Design
Paleontology Icon Paleontology

Fossil Friday: Hagfish and Lampreys Overturn Scenarios of Vertebrate Phylogeny and Evolution

March 15, 2024, 6:49 AM
Photo: Mesomyzon mengae, Tiouraren via Wikimedia Commons, CC BY-SA 4.0 DEED.

This Fossil Friday we look into the two most primitive groups of vertebrates, i.e., the hagfish and lampreys, which comprise less than 1 percent of all living vertebrate species while all others belong to the clade of jawed vertebrates called Gnathostomata. They have an eel-like body and lack paired fins and jaws. The image features the fossil lamprey Mesomyzon mengae from the Early Cretaceous of China.

Hagfish and Their Fossil Record

Hagfish or slime eel are a poorly known group of very primitive jawless marine vertebrates, which lack a vertebral column and lens eyes, but possess a remarkable strategy to escape predators: they can tie themselves in a knot and work through a slippery slime cover that is quickly produced by their skin glands. This slime contains packets of tightly coiled keratin fibres that expand explosively 10,000 times in less than half a second, which is certainly a marvel of biotechnological engineering. Because of their very simple anatomy, hagfish were often considered to be either the sister group to all other vertebrates or the sister group to lampreys (see Myers 2008). Indeed, “the phenotypic traits of hagfishes, especially the lack of some vertebrate-defining features …, have been interpreted as hagfishes exhibiting a more ancestral state than those of all other vertebrates” (Oisi et al. 2013). For a long time the poor state of knowledge on hagfish embryology represented a significant obstacle for evolutionary biologists (Ota & Kuratani 2006, Kuratani & Ota 2008), but finally new data “revealed that some apparently primitive morphological traits can be regarded as artifacts deriving mainly from fixation conditions” (Ota & Kuratani 2008). Oops!

The first and oldest putative fossil hagfish was described by Bardack (1991) from the Upper Carboniferous (Pennsylvanian) of Illinois, and named Myxinikela siroka (also see Bardack 1998). It mainly differed from its blind modern relatives by the well developed eyes, which suggests that the simple eye spots in modern hagfish are not a primitive character state but due to secondary reduction (Gabbott et al. 2016). According to Bardack it showed no similarities to lampreys and/or gnathostomes that are absent in modern hagfish, thus arguably supporting the primitive absence of such similarities and therefore a sister group relationship with all other vertebrates (Bardack 1991).

However, more recently Miyashita (2020) described a second specimen of this taxon and found more similarities with lampreys such as a cartilaginous branchial basket and a well-defined midline finfold. This rather supports a closer relationship of hagfish with lampreys. Miyashita et al. (2019) described a third fossil hagfish species Tethymyxine tapirostrum from 100-million-year-old Late Cretaceous limestone of Lebanon. The new genus was named Tethymyxine because it lived in the ancient Tethys ocean, of which today the Mediterranean Sea is the only remaining vestige. The authors attributed this remarkably well-preserved fossil hagfish to the modern family Myxinidae as a crown group representative closely related to the living genus Rubicundus. Their comprehensive cladistic analysis also confirmed the monophyly of Cyclostomata and resolved the position of fossil hagfish and fossil lampreys. Actually, the authors even could document the presence of keratin fibres that are characteristic for the hagfish slime (Wood 2019).

Gilpichthys greenei was described by Bardack & Richardson (1977) from the Pennsylvanian of Mazon Creek in Illinois. It was first attributed to hagfish by Janvier (1981, 1993, 1996a, 2008), which was rejected by Bardack (1998) but concurred by Mallatt (2023). Miyashita et al. (2019) recovered it on the stem lineage of lampreys, while a reanalysis of the modified data set by Miyashita et al. (2021, also quoted as in prep. by Miyashita 2020) found it as most basal representative of the hagfish lineage.

The presumed fossil hagfish Myxineidus gononorum was described by Poplin et al. (2001) from the Upper Carboniferous of Montceau-les-Mines in France. A reexamination with microCT by Germain et al. (2014) suggested that it may rather be a lamprey, which was also confirmed by the cladistic analysis of Miyashita et al. (2019).

Lampreys and Their Fossil Record

Lampreys are more diverse than hagfish and include freshwater, anadromous, and fully marine forms, as well as parasitic and predatory species (Brownstein & Near 2023).

Because of their numerous assumed primitive traits, lampreys have generally be considered to be an “evo-devo model” (Kuratani et al. 2002), a “model for vertebrate evolutionary research” (Xu et al. 2016), and a “key species to study the evolution of morphological characters at the dawn of Craniates” (Osório & Rétaux 2008), which “serve as an important point of comparison for studies of vertebrate evolution” (Green & Bonner 2014), and are “of critical importance to understanding the evolution of the vertebrates” (Evans et al. 2018). Many other scientists agreed that “studies of cyclostome development can thus help us to understand when, and how, key aspects of the vertebrate body evolved” (Shimeld & Donoghue 2012; also see Shimeld & Holland 2000). On the other hand, Gai & Zhu (2012) acknowledged that “due to the large morphological gap between extant jawless and jawed vertebrates, the living cyclostomes actually provide little information about the profound reorganization of the vertebrate skull and body plan that occurred along with the origin of the jaw”. This may be the reason why comparative morphology and developmental biology inspired very different evolutionary scenarios concerning the origin of the vertebrate jaw (Denison 1961, Mallatt 2008).

Also some studies suggest that apparent primitive characters of lampreys could be secondarily derived reversals or convergences. A good example is the shared presence of an endostyle in non-vertebrate chordates and lampreys, as precursor of the vertebrate thyroid. While the classical evolutionary hypothesis interpreted this as a retained primitive state (plesiomorphy) in lampreys, a new evo-devo study (Takagi et al. 2022) proposed “an alternative hypothesis where the lamprey endostyle could be obtained secondarily in crown lampreys.” Everything goes in evolutionary biology.

The first fossil lamprey was described as Mayomyzon pieckoensis by Bardack & Zangerl (1968; also see Bardack & Zangerl 1971) from the Upper Carboniferous (Pennsylvanian) Mazon Creek concretions of Illinois. Some years later Janvier & Lund (1983) described an even older fossil lamprey Hardistiella montanensis from the Lower Carboniferous (Mississipian, Namurian) of Montana (also see Janvier et al. 2004). From the same locality a second undetermined specimen of fossil lamprey, which may be conspecific with Hardistiella montanensis, was briefly described by Lund & Janvier (1986). With Mesomyzon mengae a third fossil lamprey taxon was described from the Early Cretaceous Jehol biota of China (Chang et al. 2006). Some years later even fossil ammocoete larvae of Mesomyzon were discovered (Chang et al. 2014), which suggested that “the three-phased life cycle in lampreys emerged essentially in their present mode no later than the Early Cretaceous” (also see Evans et al. 2018). A recent study by Brownstein & Near (2023) recovered Mesomyzon within the crown group of living lampreys, and suggested that the “vast majority of lamprey diversity originated in the past 100 million years”.

In the same year as Mesomyzon, the fossil lamprey Priscomyzon riniensis was described from the Devonian of South Africa (Gess et al. 2006). It is about 35 million years older than the previously oldest lampreys from the Carboniferous of North America, but more similar to modern lampreys, including first evidence for an oral disc with circumoral teeth and a branchial basket. That older forms exhibit a more modern morphology than younger forms is not exactly what we would expect to find according to Darwinian evolution, and has to be explained away with ghost lineages. Last year, Huang (2023) described the large Mesozoic lamprey Caeruleum miraculum from the Early Cretaceous Jiufotang Formation in the Hebei province of China. With a body length of 240-462 mm it is about the same size as many larger modern lampreys. Just recently a second smaller species Caeruleum gracilis was described from the same layers by Huang et al. (2024). Yanliamyzon occisor and Y. ingensdentes are two fossil lamprey species from the Middle Jurassic of the Daohugou Beds in China, described by Wu et al. (2023). In spite of their Jurassic age they are already very similar to large modern lampreys with flesh-eating adaptations, “challenging the conventional wisdom” about their evolutionary origins.

With Pipiscius zangerli a further putative fossil lamprey has been described from the Carboniferous Mazon Creek locality by Bardack & Richardson (1977). It was attributed to lampreys by Janvier (1993, 1996a), which was rejected by Bardack (1998) as unlikely. However, Gabbott et al. (2021) suggested that Pipiscius could be just a synonym of the lamprey Mayomyzon from the same locality. On the other hand, the cladistic analysis by Miyashita et al. (2019) recovered Pipiscius in a much more basal position as sister group to Euconodonta+Cyclostomata. The enigmatic Xidazoon that was described from the Lower Cambrian of China by Shu et al. (1999) as putative agnathan similar to Pipiscius, was latter shown to belong to the enigmatic animal phylum Vetulicolia.

Almost 80 years ago, White (1946) described with Jamoytius kerwoodi a new chordate from the Silurian of Scotland and considered it as a basal member of “the stock from which all the craniates, Gnathostomes as well as Agnatha, were probably derived”. Ritchie (1960, 1968) provided new evidence and reinterpreted Jamoytius as an unspecialized member of the fossil agnathan group Anaspida and “on or near the line leading to living cyclostomes”. Ritchie (1984) discussed the conflicting interpretations of Jamoytius, which have “been the subject of considerable controversy” and thought that new “discoveries strengthen the evidence for a link between Jamoytius and the petromyzontids.” Janvier & Lund (1983) and Janvier (1996b) had rather considered it as sister group to all other fossil and modern lampreys. Donoghue et al. (2000) recovered Jamoytius as sister group to the fossil agnathan clade Anaspida in the stem group of gnathostomes. Newman and Trewin (2001) again placed it closer to lampreys. Newman (2003) described numerous specimens of Achanarella trewini as another assumed jamoytiiform from the Middle Devonian of Scotland, but it was later transferred to the anaspid family Euphaneropidae. The phylogenetic analysis by Shu et al. (2003) placed Euphanerops and Jamoytius as sister groups to Anaspida and not closely related to lampreys. Sansom et al. (2010a) provided a thorough redescription and rejected both a closer relationship with lampreys or with fossil anaspids, because their phylogenetic analysis placed Jamoytius and its relatives (Euphanerops and Achanarella) as most basal sister group to gnathostomes and their fossil agnathan stem group (incl. Anaspida). The latter view was also confirmed by Janvier & Sansom (2016). However, the phylogenetic analyses of Miyashita et al. (2019, 2021) resolved Euphanerops and Jamoytius as well as the enigmatic Lasanius problematicus from the Silurian of Scotland within Anaspida, and Anaspida as sister group to Cyclostomata, which was also confirmed by the most recent analysis of Reeves et al. (2023). We will certainly not have to wait long for a new study that again claims otherwise.

Phylogenetic Incongruence

Donoghue et al. (2000) and Janvier (2010) provided a good summary of the history of the views on the phylogenetic relationships between hagfish, lampreys, and the jawed vertebrates (gnathostomes). Since the early work of the French zoologist Duméril (1806), hagfish and lampreys were considered as most closely related in a group called Cyclostomata or Agnatha, mainly based on the shared eel-like body, absence of jaws and paired fins, a large notochord, and the presence of horny teeth on a lingual apparatus.

Subsequently, in the 1970s—1990s many experts (e.g., Løvtrup 1977, Janvier 1978, 1981, 1993, 1996a, Forey 1995) considered lampreys as closer related to jawed vertebrates, thus rejecting Cyclostomata as an unnatural (paraphyletic) group, mainly based on anatomical and physiological similarities between lampreys and gnathostomes. Mallatt & Sullivan (1998) remarked that “A consensus has emerged over the last 20 years that lampreys are the sister group to the gnathostomes and the hagfishes represent an ancient, basal lineage. This hypothesis has essentially displaced the classical hypothesis of monophyly of the cyclostomes (lampreys plus hagfishes).” This was often interpreted in terms of hagfish being a suitable model for the early vertebrate body plan and lampreys larvae for the early gnathostome body plan (Janvier 1981).

During the 1990s and early 2000s molecular studies restored the Cyclostomata hypothesis by again strongly supporting a closer relationship of hagfish with lampreys (e.g., Stock & Whitt 1992, Lé et al. 1993, Forey & Janvier 1993, 1994, Mallatt & Sullivan 1998, Kuraku et al. 1999, Hedges 2001, Mallatt et al. 2001, Delarbre 2000, 2002, Furlong & Holland 2002, Takezaki et al. 2003, Kuraku & Kuratani 2006, Mallatt & Winchell 2007), which is yet another instance of the common conflict between morphological and molecular data (also see Forey 1995 and Miyashita et al. 2019). 

Janvier (2008) remained unconvinced and mentioned that “these results leave morphologists and physiologists greatly perplexed, and most of them now turn back to the old ad-hoc explanation of cyclostome ’degeneracy’”. He also cautioned that “very few authors raise the question of the assumptions underlying the methodologies involved in their analysis”. Near (2009) said that “one of the most problematic issues in vertebrate phylogenetics is the disagreement between phenotypic and molecular inferences regarding the relationships among hagfishes, lampreys, and gnathostomes.” Near tried to resolve the conflict with an extensive analysis of morphological and molecular data but only added to the confusion: his “maximum parsimony analyses of 115 phenotypic characters combined with 4,638 rRNA sites and more than 10,000 amino acids each result in monophyly of lampreys and gnathostomes”, while “on the other hand, Bayesian analyses of the combined data sets support monophyly of hagfish and lampreys”. He concluded that “the fact that addition of relatively few phenotypic characters can alter phylogenetic inferences of cyclostome monophyly obtained from molecular data sets, and the inability of simulated data sets to recover key nodes in the craniate phylogeny provide reasons to view the strong support for cyclostome monophyly inferred from molecular data sets with a measured degree of skepticism.”

Just a few days ago media headlines claimed that a new study by Bedois et al. (2024) shows that “sea lamprey’s brain development is remarkably similar to that of humans“ (News Staff 2024). This might suggest to a casual reader that there is new support for a closer relationship of lampreys with gnathostomes. However, the actual study endorses the mainstream consensus about cyclostome monophyly. But here is the problem: their primary data indeed rather support a paraphyly of cyclostomes with hagfish in a more basal position. The authors looked into the gene regulatory networks that control the patterning of the vertebrate brain. They specifically studied certain gene families that are called Cyp26 and Aldh1a. Here is what they found in their own words:

We did not retrieve a Cyp26A1 homolog in the hagfish genome, which may reflect an incomplete assembly or that Cyp26A1 was lost in hagfish … While lamprey Cyp26B1/C1a and Cyp26B1/C1b appear to group with the two hagfish Cyp26C1-like genes, we are unable to infer any 1:1 orthology between cyclostomes and jawed vertebrates … For the Aldh1a complement, the tree shows clear jawed vertebrate clades for Aldh1a1 and Aldh1a2. However, it does not resolve clear relationships between the cyclostome and jawed vertebrate Aldh1a genes, since the cyclostome Aldh1a genes do not group with either Aldh1a1 or Aldh1a2 clades, nor with each other. [Emphasis added.]

Thus, the most parsimonious interpretation of the Cyp26A1 evidence would rather favor cyclostome paraphyly, while the other gene families were ambiguous or not informative. That the authors still endorse cyclostome monophyly in spite of their own conflicting data, without even bothering to discuss the issue, shows how much theory and consensus thinking trump data in evolutionary biology nowadays.

The monophyly of Cyclostomata was also supported by some morphological studies of the feeding apparatus (Yalden 1985, Mallatt 1997), the development of the neural crest (Kuratani & Ota 2008), the embryonic development of the head (Oisi et al. 2013), and some extensive cladistic studies of morphological characters based on fossil evidence (Miyashita et al. 2019, Miyashita 2020, Reeves et al. 2023). Heimberg et al. (2010) “have demonstrated that morphological data are indecisive on the question, and that both protein-coding and non-coding molecular data unequivocally support cyclostome monophyly” (Benton et al. 2015). Heimberg et al.’s analysis with microRNA data was widely considered to have settled the issue and even moved Philippe Janvier (2010) to reconsider. However, it was strongly critiqued by Thomson et al. (2014), who commented that their “results indicate that the capacity of miRNA data to resolve the tree of life has been overstated, and we urge caution in their application and interpretation”. Nevertheless, Janvier & Sansom (2016) again tentatively accepted cyclostome monophyly but mentioned that “the morphology of extant cyclostomes is not currently in accordance with interpretations of cyclostome monophyly as supported by extensive molecular evidence”. Miyashita et al. (2019) used fossil data to “potentially resolve the morphological–molecular conflict at the base of the Vertebrata”. They “showed that morphology-based analysis converged onto molecular inferences when characters are coded nonindependently”, but of course treating characters as non-independent is yet another way to fudge the data to fit the theory, because they certainly would not feel compelled to do so if there would be no conflict between morphological and molecular data. Other scientists even emphasized that “character independence is a basic assumption of most phylogenetic inference methods” (O’Keefe & Wagner 2001).

The cladistic study by Donoghue et al. (2000: Fig. 17), which also included other fossil jawless vertebrates (e.g., conodonts) and a large number of characters, found ambiguous support for both alternative hypotheses, but favored cyclostome paraphyly with a closer relationship of lampreys and jawed vertebrates. The phylogenetic analysis by Shu et al. (2003), which was mainly based on fossil anatomy data, also supported cyclostome paraphyly. The same resulted from a cladistic analysis of re-weighted morphological characters and fossil data by Gess et al. (2006) and the unconstrained phylogenetic analysis by Sansom et al. (2010a), which shows the persistent conflict between morphological and molecular data. Fujimoto et al. (2013) confirmed that “this discrepancy between the molecular and morphological data has been a source of contention regarding the evolution of the early vertebrates, and there was no consensus on the phylogenetic position of the hagfish for about three decades.” Their study found a non-parsimonious evolution of Dlx genes in hagfish and suggested that “this pattern is reminiscent of the non-parsimonious evolution of its morphological traits, including its inner ear and vertebrae, which indicate that this group is an early-branching lineage that diverged before those characters evolved.”

But even the molecular data do not agree: The analysis by Rasmussen et al. (1998) of hagfish mitochondrial DNA and its comparison with lampreys and gnathostomes, clearly “placed the hagfish as a sister group of Vertebrata sensu stricto, i.e., the lamprey and the gnathostomes.” The analysis of 7S RNA of hagfish and lampreys by Gürsoy et al. (2000) provided “additional support to the notion of a sister group relationship between Petromyzontiformes and gnathostomous vertebrates to the exclusion of Myxiniformes” and thus “an additional argument against the Cyclostomata hypothesis.”

A careful analysis of the comparative morphology of hagfish, lampreys and jawed vertebrates reveals a highly incongruent pattern with two mutually exclusive but equally well-supported hypotheses for the relationship of lampreys (Mickoleit 2004: 27–28):

The cyclostome hypothesis of a sister group relationship of lampreys and hagfish as clade Cyclostomata is supported by:

The vertebrate hypothesis of a sister group relationship of lampreys and jawed vertebrates as clade Myopterygii Janvier 1978 (= Vertebrata s.str.), with hagfish as primitive outgroup, is supported by (also see Janvier 1981, Forey 1984: Table 1, Forey 1995: Table 2, Bardack 1991):

  • eyes with three pairs of extrinsic (extra-ocular) muscles
  • inner ear with two semicircular canals (only one in hagfish, but Higuchi et al. 2019 provided evidence that this could be a secondarily derived condition)
  • separate dorsal and caudal fins
  • unpaired fins with radial muscles
  • dorsal neural arch elements above notochord
  • intestine with spiral valve (spiral fold)
  • external innervation of the heart by the Nervus vagus ramus intestinalis
  • pronephric kidney no longer functioning as excretory organ in adults
  • kidneys with collecting tubules
  • osmoregulation

Whichever of the two alternative theories is correct, the evidence for the other alternative has to be explained away with ad hoc hypotheses (see Forey & Janvier 2000). Some advocates for the vertebrate theory for example suggested that gnathostomes evolved from paedomorphotic (neotenic) ammocoetes larvae, which lack some of the typical traits of adult cyclostomes. Other suggested that the cyclostome characters were part of the history of gnathostomes and later reduced. Advocates of the cyclostome theory have to assume that similarities between lampreys and gnathostomes are either convergences or secondarily reduced in hagfish. None of these options is really resonating with the Darwinist prediction of a congruent nested hierarchy of similarities. Therefore, it is hardly surprising that the “monophyly or paraphyly of cyclostomes (hagfishes and lampreys) has been the subject of heated debate for over a century” (Gai & Zhu 2012).

Kardong (2018) wrote in the 8th edition of his standard textbook on vertebrate comparative anatomy and evolution.

Morphological data suggest that cyclostomes are not a monophyletic group but instead that hagfishes are more basal and lampreys alone are the sister group of jawed vertebrates (e.g., Gess et al., 2006). However, extensive gene-sequence data suggest otherwise, namely that hagfishes and lampreys form a natural monophyletic group, which is the living sister group to jawed vertebrates (e.g., Mallatt and Winchell, 2007). Sometimes combining data sets (morphological plus molecular) can help, but other times, as with cyclostomes and gnathostomes, a large data set (e.g., molecular) can dramatically swamp the effect of otherwise revealing information in a small data set (e.g., morphological), thereby artificially altering the phylogenetic interpretation (e.g., Near, 2009). Sometimes fossils can help resolve such conflicts by revealing ancient suites of traits that favor one view over another. Unfortunately, when the earliest lampreys turn up in the Devonian, they are already very similar to modern lampreys. That is not surprising because cyclostomes likely diverged from other vertebrates much earlier.

Red Blooded Vertebrates Evolved Twice?

The incongruence goes even further: jawless and jawed vertebrates are sister groups and both possess hemoglobin as an oxygen transport protein. This would suggest that hagfish, lampreys, and gnathostomes inherited this trait from their common craniate ancestor. However, the study by Hoffman et al. (2010) found that hemoglobin was independently invented in both lineages by cooption of paralogous genes. The press release says “that red-blooded vertebrates evolved twice” (University of Nebraska-Lincoln 2010). The authors concluded that “this example of convergent evolution of protein function provides an impressive demonstration of the ability of natural selection to cobble together complex design solutions by tinkering with different variations of the same basic protein scaffold.” God forbid to consider the obvious alternative of intelligent design.

New Fossils Overturn Previous Views

Just a few years ago a seminal study by Miyashita et al. (2021) described new fossil evidence that completely overturned previous well-established views on the phylogeny and evolution of vertebrates (also see Miyashita 2021a, 2021b). This traditional view was based on the striking similarity between invertebrate lancelets (Branchiostoma or Amphioxus) and lamprey ammocoete larvae, which are both functionally blind, eel-like filter-feeders that burrow in sand. The obvious assumption was that this represents the ancestral condition for vertebrates, which was modified in adult cyclostomes into a sucking mouth disk with spikes and modified in gnathostomes into biting jaws. Another implicit assumption is that the filter feeding larvae would be only retained in lampreys, but secondarily reduced in the evolution of hagfish and gnathostomes. This was the textbook scenario that generations of biology students, including myself, learned at university. Miyashita (2021a) admitted that “once fortified with historical inertia, just-so stories are difficult to interrogate. In this case any ammocoete-driven narrative hinges on one prediction: ammocoetes must extend deep into the vertebrate tree. Though seemingly straightforward, this prediction is hard to test.“

Well, that all changed when Miyashita et al. (2021) described larvae from the oldest Paleozoic lampreys, i.e. Priscomyzon riniensis from the 360-million-year-old Devonian of South Africa. These fossil lamprey larvae were not at all looking like ammocoetes but rather like small adults with a sucker mouth that suggests a leach-like way of life. The scientists could also document similar morphology in fossil larvae of other Paleozoic lampreys such as Pipiscius zangerli and Mayomyzon pieckoensis. Consequently, the authors suggested that filter-feeding larvae were a later innovation in modern lampreys, and also suggested a new view on vertebrate evolution, with armored jawless fish like the Early Paleozoic ostracoderms as the ancestral type for both cyclostomes and gnathostomes. This might resonate with the finding of the expression of enamel proteins in Cyclostomata and fossil evidence (Simonetta et al. 1999), but of course, it also comes with its own set of problems as ostracoderms do not appear in the fossil record before the Upper Ordovician (Sansom et al. 2015) and Anaspida as putative stem cyclostomes not before the early Silurian, while the split between cyclostomes and gnathostomes is mostly dated back into the Cambrian (see below).

The press releases of the involved universities were quite revealing: Dyani (2021) commented that the “fossil lamprey larvae from Makhanda, South Africa overturn textbook assumptions on vertebrate origins … So it’s time to rewrite the textbooks!” Gess (2021) makes the same point and says that “this discovery means that it’s time to rewrite the textbooks and revisit scientific understanding about what modern lampreys reveal about vertebrate origins.” New fossils rewrite an evolutionary just-so story, we have heard this before over and over again and reported about it many times at Evolution News. Caldwell (2021) commented “Long-accepted theory of vertebrate origin upended by lamprey fossils”. The lead researcher Tetsuto Miyashita is quoted as saying that they have “basically removed lampreys from the position of the ancestral condition of vertebrates”. While the ammocoete larvae are very different from adult lampreys, the researchers could document complete lifecycles of Paleozoic lampreys that had larval stages that just looked like small adults with well-developed eyes and toothed sucker mouths. This suggests that the simple worm-like, blind and filter-feeding ammocoete larvae of modern lampreys do not represent an ancestral primitive condition at all, but rather a more recent innovation, and therefore cannot serve as model for vertebrate evolution.

The Empire Strikes Back

Mallatt (2023) directly responded to the sensational study by Miyashita et al. (2021) and vehemently disputed their claim “that the filter-feeding larval lamprey is a new evolutionary addition to the lamprey life-cycle and does not provide information about early vertebrates, in contrast to the traditional view that this ammocoete stage resembles the first vertebrates.” He criticized that the traditional view cannot be so easily dismissed and that the new proposal “was not tested in a statistically significant meaningful way”. In Mallatt’s view the four known genera of Paleozoic lampreys could rather “comprise a clade of dwarf lampreys (whose common ancestor had lost its original filter-feeding larva)”. However, this is of course just an ad hoc assumption, and it is arguably less parsimonious than Miyashita’s interpretation. Nevertheless, Mallat correctly pointed out how similar the filter-feeding structures are in lamprey ammocoetes larvae and in the invertebrate chordates such as lancelets and tunicate tadpole larvae, which strongly suggests a symplesiomorphic homology. He also provided an in-depth review on how the morphology of lamprey larvae is helpful for reconstructing early vertebrates and stem gnathostomes, and offered a revised traditional scenario of vertebrate evolution centered on filter-feeding.

There is a bummer, though: the apparent similarity of the filter-feeding mouth between lancelets and the ammocoete larva of lampreys had already been debunked as a convergence by Yasui & Kaji (2008), who’s work was totally ignored by Mallatt (2023). Those authors had concluded:

The evolutionary history of the vertebrate mouth has long been an intriguing issue in comparative zoology. When the prevertebrate state was considered, the oral structure in adult lancelets (amphioxus) was traditionally referred to because of its general similarity to that of the ammocoete larva of lampreys. … Here, we review how the lancelet mouth has been interpreted in the study of evolution of the vertebrate mouth, as well as recent advances in chordate studies. With this background of increased knowledge, our innervation analysis supports the interpretation that the morphological similarity in the oral apparatus between ammocoetes and lancelets is a homoplasy caused by their similar food habits.

The Dubious Origin of Jaws

Lancelets, hagfish, and lampreys possess a velum as water-pumping flap, which is believed to be a derivative of the mandibular arch and thus a putative homologue or even evolutionary precursor of the jaws in gnathostome vertebrates (Janvier 1993, 1996b). Such a homology was strongly rejected by Yasui & Kaji (2008) but again strongly advocated by Yokoyama et al. (2020), who also did not bother to even mention the former work. Mallatt (1996, 2023) likewise rejected the homology in the sense of jaws being derived from a velum, but did not exclude the possibility that both are independently derived from the same ancestral structure. Kuratani et al. (2002) admitted that “there are also other enigmatic features in the lamprey where we cannot determine whether they are primitive or advanced (like velum); this is a problem which again raises the question of the phylogenetic relationships between the lamprey, hagfish, and gnathostomes, as well as the origin of the jaw.” A recent developmental study by Root et al. (2023) found no direct homology between lamprey mucocartilage and gnathostome joint tissue of the jaws, which “has important implications for the Cooption Hypothesis of jaw evolution”. Evolutionary scenarios built on uncertain homologies are speculations built on speculations, which could also be called a house of cards.

Missing Link or Rotten Corpse?

Usually, evolutionary paleobiologists assume that fossils, which exhibit a set of primitive character states compared to their modern relatives, do represent stages of the early ancestral stem lineage of this group. This is also the case for numerous assumed early chordates and vertebrates from Cambrian sediments. However, experiments showed that apparent primitive states of soft-tissue characters easily originate as an artifactual result of the decay process. Several studies on decaying bodies of lancelets, hagfish, lampreys and their ammocoete larvae showed that these successively look more and more like primitive chordates as the decay process progresses (Sansom et al. 2010b, 2011, Gabbott et al. 2021). It was therefore suggested “that this decay filter also affects other groups of organisms and that ‘stem-ward slippage’ may be a wide-spread but currently unrecognized bias in our understanding of the early evolution of a number of phyla” (Sansom et al. 2010b). In other words: many evolutionary scenarios and assumed missing links may just be artefacts of rotting corpses of animals with completely modern anatomy.

Abrupt Appearance in the Cambrian Explosion

When did cyclostomes, i.e. the group that includes hagfish and lampreys, originate and did they originate gradually as suggested by Darwinism? Most experts agree that cyclostomes appeared in the Cambrian, but the actual fossil record begins abruptly in the Devonian and Carboniferous without any gradual stem group succession. However, the existence of putative fossil stem vertebrates (e.g., Haikouichthys and Myllokunmingia, as well as maybe yunnanozoans, see Tian et al. 2022), in the Cambrian as well as molecular clock estimates suggest that cyclostomes may represent yet another group that originated during the burst of biological creativity in the so-called Cambrian Explosion.

Stock & Whitt (1992) estimated that the lineages of hagfish, lampreys, and gnathostomes may have diverged more than 500 million years ago. Donoghue et al. (2000) concluded that “myxinoids and petromyzontids were in existence at least by the early Cambrian”. Also other studies suggested a Cambrian origin of cyclostomes (e.g., Yasui & Kaji 2008, Janvier 2010, Brownstein & Near 2023, Mallatt 2023), while some placed “the split between cyclostomes and gnathostomes occurring close to the Cambrian-Ordovician boundary ~493 Ma, consistent with previous estimates (Kuraku et al., 2009b)” (Marlétaz et al. 2024). Kuraku & Kuratani (2006) had estimated the split between hagfish and lampreys at 470-390 Mya, which would also be consistent with a Cambrian origin of stem cyclostomes. Based on fossil data, Benton et al. (2015) dated the origin of crown group Cyclostomata between maximally 636.1 Mya and minimally 358.5 Mya (also see Fossil Calibration Database), which “encompasses the possibility that putative Cambrian vertebrates, such as Haikouichthys, Myllokunmingia (Shu et al., 1999), and Zhongjianichthys (Shu, 2003), are crown-cyclostomes”. Miyashita et al. (2019) found that “Bayesian estimates of divergence placed the crown cyclostome node from the earliest Cambrian (Terreneuvian) to Early Silurian (Wenlockian) times [95% highest posterior density interval (HPD): 536.31–428.44 Mya] with median near the Cambrian–Ordovician boundary (485.57 Mya)”, while the cyclostome stem group likely originated earliest Cambrian (Miyashita et al. 2021: Fig. 4) to the late Cambrian (Wu et al. 2023: Fig. 3). Yu et al. (2024: Fig. 2) suggested that vertebrates originated in the Early Cambrian and the split between cyclostomes and gnathostomes occurred in the Late Cambrian. The most recent study by Marlétaz et al. (2024) rather dated the cyclostome-gnathostome split in the Early Cambrian, thus right within the main pulse of the Cambrian Explosion.

As we have seen, hagfish and lampreys present various challenges to mainstream evolutionary biology. Their fossil record as well as their incongruent pattern of anatomical similarities is better explained by intelligent design. This is especially true for the very unique and highly complex feeding apparatus of cyclostomes or the hagfish slime glands, which represent a genuine marvel of engineering.

References