Fossil Friday: The Abrupt Origins of Lagomorphs and Rodents
This Fossil Friday features the skull of the fossil rabbit Palaeolagus haydeni from the Oligocene of Nebraska, as we look today into the origins of two orders of placental mammals: the Lagomorpha and Rodentia, which together form the clade Glires within the supergroup (cohort) of Euarchontoglires. Lagomorpha includes the 110 living species of rabbits, hares, and pikas, while rodents with 40 percent (2.600 species) of all living mammal species represent the most diverse mammalian order (Kay & Hoekstra 2008). Rodents, for example, include squirrels, beavers, hamsters, mice, and rats, as well as guinea pigs, gundis, and porcupines.
In my article last week I listed numerous studies that supported Euarchontoglires, of which the vast majority also supported the monophyly of Glires (e.g., Meng & Wyss 2005, Springer et al. 2007b, and most recently Gupta & Suggett 2022), while there were only few dissenters (e.g., Arnason et al. 2002). The molecular studies by Li et al. (1990), Graur et al. (1996), and Misawa & Janke (2003) suggested that Lagomorpha is more closely related to primates than to rodents, but these views were unanimously rejected by later studies (see Honeycutt & Adkins 1993, Douzery & Huchon 2004, Kjer & Honeycutt 2007, and Zhou et al. 2015). This is hardly surprising as rodents and lagomorphs were already classified in a common order Glires by Carl von Linné in the first edition of his famous Systema Naturae (Linnaei 1735), which basically represents the very beginning of the scientific classification of animals and predates any kind of evolutionary thinking. It was Gidley (1912) who separated these two mammalian groups in the distinct orders Lagomorpha and Rodentia and did not consider them as closely related. This view was endorsed in a famous article titled “What, if Anything, Is a Rabbit?” by Wood (1957), and it was also endorsed by Van Valen (1964) and McKenna (1975). Even though for a while “the monophyly of Glires has been one of the most controversial issues in higherlevel mammalian systematics” (Meng et al. 2003), subsequent works increasingly supported the Glires concept (e.g., Simpson 1945), so that Douzery & Huchon (2004) responded to Wood’s classical question with an article titled “Rabbits, if Anything, Are Likely Glires.” Cox & Hautier (2015) therefore concluded in the foreword to their monograph on the evolution of rodents that “at present, the Glires concept is stronger than ever.”
A Cretaceous Origin?
Many molecular clock studies proposed a Cretaceous origin of Glires and its two main branches (e.g., Li et al. 1990, Kumar & Hedges 1998, Springer et al. 2003, Janečka et al. 2007, Kay & Hoekstra 2008, Meredith et al. 2011, Suárez et al. 2011, Foley et al. 2016). Archibald et al. (2001) also argued for a Cretaceous origin of Glires based on fossil data, but these were later all refuted as not representing eutherian mammals at all (Wible et al. 2005, 2007, Halliday et al. 2015, Velazco et al. 2022). Other studies rather supported an early Paleogene origin of Glires and its divergence into the lagomorph and rodent lineages (Huchon et al. 2002, Douzery et al. 2003, Meng et al. 2003, Meng 2004, Asher et al. 2005, AMNH 2006, Horner et al. 2007, Wu et al. 2012, O’Leary et al. 2013, Phillips & Fruciano 2018, Upham et al. 2019: fig. 4), to make the theory better fit with the actual fossil record. Meng et al. (2003) revealingly remarked:
Apparently, the molecular clock hypothesis tends to push rodent divergences at various levels of the group deeper into history than what the fossil record indicates. The discrepancy is enormous.
Well, such enormous discrepancies between theoretical predictions and empirical data should give reason to pause and to question the adequacy of the theory.
So let’s have a look at the actual fossil record of Glires. Here is an up-to-date classification of all relevant higher taxa of gliriform mammals with their stratigraphic ranges and earliest fossil representatives (extinct groups are marked with +; age ranges are derived from PaleoDB):
Rodentiamorpha sensu Wyss & Meng 1996)
+Sinomylus zhaii (58.7-55.8 mya)
+Eurymyloidea (= “Mixodontia” partim) (61.7-48.6 mya)
+Eurymylidae (61.7-48.6 mya)
+Heomys orientalis (61.7-58.7 mya)
+Eomylus (55.8-48.6 mya)
+Decipomyidae (55.8-48.6 mya)
+Rhombomylidae (58.7-55.8 mya)
Rodentiaformes (sensu Wyss & Meng 1996)
+Alagomyidae (55.8-48.6 mya)
Rodentia
+Ischyromyidae (58.7-30.8 mya)
+Acritoparamys atavus (56.8-55.8 mya)
+Asiaparamys shevyrevae (58.7-55.8 mya)
+Paramys adamus (56.8-55.8 mya)
+Franimys (55.8-50.3 mya)
PaleoDB also lists these invalid families among Glires:
+Aspalacidae (48.6-40.4 mya) (belongs to the living rodent family Muridae according to McKenna & Bell 1997)
Ierboidae (2.588-0 mya) (an obsolete name for a group of living rodents)
Lagostomyidae (a synonym for the living rodent family Chinchillidae)
Ochtodontidae (an incorrect spelling of the living rodent family Octodontidae)
Concerning these datings it must be mentioned that there is some uncertainty concerning the geological layers in China that yielded some of the oldest fossils like Heomys and Mimotona (Li 1977). Benton et al. (2015) commented as follows:
The Wanghudun Formation, Qianshan Basin, China is of debated age. First, it was interpreted as belonging to the Paleocene Shanghuan Asian Land Mammal Age (Dashzeveg and Russell, 1988; Li and Ting, 1993[sic]). However, Missiaen (2011) suggests that the Wanghudun Formation may be closer to the Nongshanian ALMA. We use this younger estimate for age of the Wanghudun Formation for two reasons: it is the more current interpretation, and it is younger, and so more conservative. The marine correlate of the Nongshanian ALMA is the Thanetian, with a minimum bound of 56 Ma ± 0.0 Myr = 56 Ma (Gradstein et al., 2012).
Wang et al. (2016) provided a synopsis of the Paleocene stratigraphy of the region. They agreed that the Wanghudun Formation correlates with the Nongshanian ALMA, which they date to a Middle Paleocene age correlative to a middle Tiffanian Ti1-4a NALMA (North American Land Mammal Age). According to Secord (2008) the Ti1-T4a interval dates to 61.57-58.85 mya, which agrees with the 61 mya estimate for Mimotona by López-Torres et al. (2020).
Glires
The most obvious synapomorphic characteristic of Glires is a pair of enlarged ever-growing incisors (Martin 1999, Fostowicz-Frelik 2017). The first Glires appeared in the Early Paleocene of East Asia, close to the K/Pg boundary, shortly followed by the appearance of mimotonids and eurymyloids as first representatives of the lagomorph and rodent lineages respectively (Li & Ting 1985, 1993, Novacek 1986, Wyss & Meng 1996, Van Valen 2002, Meng & Wyss 2005, Janis et al. 2008, Lacher et al. 2016, Fostowicz-Frelik 2020, 2022).
Previously, mimotonids, and eurymyloids where often together classified in a group called Mixodontia (Dashzeveg & Russell 1988, Averianov 1994, Lopatin 2003, Lopez-Martinez 2008), but this group was mostly abandoned as non-monophyletic when cladistic classification became mainstream consensus, or it was redefined to only include the rodent-related eurymylids (Li & Yan 1979, Dashzeveg et al. 1998) as was originally proposed by Sych (1971) when he named Mixodontia.
Numerous earlier studies placed the African elephant shrews (Macroscelidea) with rodents and/or lagomorphs as well as the fossil anagalids in a clade called Anagalida (McKenna 1975, Szalay 1977, Novacek 1986, 1992, Novacek & Wyss 1986, McKenna & Bell 1997, Meng & Wyss 2001, and Rose 2006). This was challenged by the results of modern molecular systematics and phylogenomics in the late 1990s, when the order Macroscelidea was transferred to a different supergroup called Afrotheria, which became the general consensus among the experts.
Nevertheless, the family Anagalidae, which comprises rabbit-like burrowing mammals known from Paleogene deposits of China and Mongolia (López-Torres & Fostowicz-Frelik 2018), are still considered as stem group representatives of Glires (Van Valen 2002, Meng et al. 2003, Meng 2004, Rose 2006, Janis et al. 2008; but see Fostowicz-Frelik 2017 and López-Torres & Fostowicz-Frelik 2018). Originally, Anagalidae was described and classified by Simpson (1931) among insectivores as relative of treeshrews and elephant shrews, which was still endorsed by Van Valen (1967). According to Nessov et al. (1998) the Anagalida also include Wania chowi from the earliest Paleocene of Anhui in China (Wang 1995), which dates right after the Chicxulub impact that killed off the non-avian dinosaurs and thus could represent the oldest fossil record of gliriforms. Unfortunately, this attribution was never supported by a proper phylogenetic study and dates to a time when Anagalida was used in an obsolete way for a non-monophyletic group. Therefore, Wang et al. (1998, 2016) considered the ordinal affinity of Wania as indeterminate, so that we cannot use this taxon as a safe fossil dating for the age of Glires.
There are two other groups of Paleogene small insectivorous mammals that may belong to the stem group of Glires: One is the enigmatic order Apatotheria, which only includes the family Apatemyidae from the Paleocene-Oligocene of North America and Eurasia. Of this family we not only know the usual teeth and skull fragments but even completely preserved skeletons of Heterohyus from the famous Eocene Messel locality in Germany (Koenigswald 1987). They are characterized by a highly specialized dentition with a single hypertrophied lower incisor, which has been interpreted by Koenigswald (1987) as a adaptation similar to the niche of woodpeckers or the Malagasy Aye-aye lemur. Apatotherians disappeared with a mass extinction event at the Eocene-Oligocene transition (contra Kemp 2005) that is known under the French term “Grande Coupure” and was caused by an abrupt global cooling. In spite of the availability of complete skeletons there has been little agreement on apatemyid relationship (Gingerich & Rose 1982) and they have been mostly associated with primates or with insectivores, but also with rodents and even the extinct carnivore order Cimolesta (Rose 2006). However, the main modern study (Silcox et al. 2010) on the relationship of this group instead found reasonable support for a basal position in Euarchontoglires and weak support for a sister-group relationship with Glires. On the other hand, the cladistic study by Halliday et al. (2015) placed apatemyids as sister group to bats, thus in a different supergroup (Laurasiatheria).
The second group of possible stem Glires is another enigmatic family called Arctostylopidae from the Late Paleocene and Early Eocene of East Asia and North America, which is mainly known from their teeth. According to Halliday et al. (2015) “the relationships of Arctostylopidae are extremely poorly understood (Zack, 2004), but this group has been thought to be related to Glires, Notoungulata, or Artiodactyla”. Cifelli et al. (1989) considered them as possible relatives of lagomorphs. Rose (2006) still endorsed the notungulate affinity, but Missiaen et al. (2006, 2012) showed that similarities with South American notungulates arose independently and instead found synapomorphies with gliriforms. They suggested that arctostylopids may be most closely related to the poorly known early Paleocene family Astigalidae, which was sometimes attributed to Anagalida. Halliday likewise found some support for a relationship with Glires.
Finally, we have to briefly discuss an extinct family called Zalambdalestidae (94.3-61.7 mya), which lived in the Late Cretaceous and Early Paleocene of Asia and had a locomotion similar to modern lagomorphs but teeth like insectivores. They have been associated with Glires, Anagalida, or at least Euarchontoglires by several studies (e.g., Van Valen 1964, Szalay & McKenna 1971, McKenna 1975, 1994, Szalay 1977, Novacek & Wyss 1986, McKenna & Bell 1997, Dashzeveg et al. 1998, Archibald et al. 2001, Meng & Wyss 2001, Fostowicz-Frelik & Kielan-Jaworowska 2002, Averianov & Archibald 2005, Kemp 2005, Janis et al. 2008, Fostowicz-Frelik 2016, 2017, López-Torres & Fostowicz-Frelik 2018). However, Meng (2004)found neither evidence for a relationship of zalambdalestids and Glires, nor for any close relatives of Glires in the Cretaceous. Rose (2006) tentatively still classified Zalambdalestidae in Anagalida with Glires, but mentioned that the phylogenetic position is uncertain and that they could rather be basal eutherians. More recent studies indeed placed Zalambdalestidaeoutside of crown group Eutheria (Wible et al. 2005, 2007, Halliday 2015, Halliday et al. 2015, Velazco et al. 2022), thus not even as genuine placental mammals. The main evidence for such a basal position is the presence of an epipubic bone, which is retained in monotremes and marsupials but reduced in all living placental mammals. Even Van Valen (2002) now doubted his earlier belief that zalambdalestids and Glires are related.
Lagomorphs
The total group that includes crown group Lagomorpha and their stem group like Mimotonidae (Li & Ting 1985, 1993, Averianov 1994, Wyss & Meng 1996, Van Valen 2002, Rose 2006, Dawson 2008) has been named Duplicidentata after the retained possession of two pairs of upper incisor teeth. Lagomorphs originated in the Paleocene of East Asia and later diversified in Asia and North America (Lopez-Martinez 2008, Ge et al. 2013, Lacher et al. 2016). The fossil record of lagomorphs includes 78 genera and 234 species from the Paleocene to the Pleistocene (Lopez-Martinez 2008). The most up-to-date phylogenetic tree and classification of fossil lagomorphs has been provided by Lopatin & Averianov (2020). The oldest fossil record of stem lagomorphs is the insectivorous genus Mimotona from the Paleocene of China (Li 1977, Dashzeveg & Russell 1988), which belongs to the extinct clade Mimotonidae. Actually, Mimotona could represent the oldest known crown group placental mammal known at all (Halliday 2015). Second most ancient record is another mimotonid genus Gomphos from the Early to Middle Eocene of China and Mongolia (Meng et al. 2004, 2005, 2009, Asher et al. 2005) with an estimated age of about 55 million years (AMNH 2006). Several studies (Meng & Wyss 2001, Meng 2004, Rose et al. 2008, Lopatin & Averianov 2020) did not recover Gomphos in a clade with other Mimotonidae but placed it in an even more basal position as sister group to all other lagomorphs.
Dawsonolagus antiquus from the Early Eocene of Inner Mongolia in China has been suggested as transitional form between mimotonids and lagomorphs because of its mosaic combination of characters of both groups (Li et al. 2007). Therefore, Lopatin & Averianov (2020) suggested Eulagomorpha as name for clade that includes Dawsonolagus and crown group Lagomorpha. A possibly even somewhat more ancient eulagomorph is Arnebolagus leporinus from the Early Eocene (Bumbanian, 55.8-48.6 mya) Bumban Member of Naran Bulak Formation in Mongolia (Lopatin & Averianov 2008, 2020). Of similar age is the earliest Strenulagidae Aktashmys montealbus from the terminal Early Eocene of Kyrgyzstan (Averianov & Lopatin 2005, Lopatin & Averianov 2006).
The oldest fossil record of crown group Lagomorpha seems to be distinctive, small ankle bones from Early Eocene (ca. 53 mya) deposits of Gujarat in India (Rose et al. 2008). The fact that crown group and stem group lagomorphs appear together at about the same place and time is one of the many inconvenient facts of the fossil record that do not square well with a Darwinian narrative.
Rodents
The total group that includes crown group rodents and their stem group representatives such as Eurymyloidea has been named Simplicidentata after the possession of only a single pair of upper incisor teeth. Rodents first appeared in the Late Paleocene of East Asia but quickly spread to Europe and North America (Wilson 1951, Dawson & Beard 1996, Dawson 2003). The most basal stem group rodent seems to be Sinomylus, which has a single pair of incisors but uniquely retained the second upper premolar (McKenna & Meng 2001, Rose 2006). However, some cladistic studies recovered Sinomylus as a sister group to Lagomorpha+Rodentia (Rose et al. 2008, Asher et al. 2005, 2019), while Lopatin (2003) suggested that Sinomylus belongs to Eurymyloidea and is closer related to Lagomorpha. Incongruence, anybody?
The extinct Eurymylidae from the Early Paleocene to Middle Eocene of East Asia include the most basal and oldest stem group representatives of rodents (Li & Ting 1985, 1993, Wyss & Meng 1996, Van Valen 2002, Rose 2006, Rose et al. 2008). Some genera that are usually classified within Eurymylidae (McKenna & Bell 1997), and even in the subfamily Eurymylinae, have been suggested by some experts to be distinct enough to be better classified in the separate families Decipomyidae (Dashzeveg et al. 1998) and Rhombomylidae (Ting et al. 2002, Huang et al. 2004). Of course, as always in phylogenetics there is considerable conflict in the results and some studies recovered eurymylids as stem lagomorphs rather than stem rodents even though with only weak support (e.g., Lopatin 2003, Marivaux et al. 2004, Asher et al. 2005). A total outlier is the cladistic study by Dashzeveg et al. (1998), who got Decipomyidae as sister group of Lagomorpha, Mimotonidae and Rhombomylidae as successive sister groups of Rodentia (incl. Alagomyidae), and the remaining eurymylids and Gomphos as stem Glires. There is also conflict in the views about the composition of Eurymylidae: Meng et al. (1994) and Wyss & Meng (1996) considered the genus Heomys as closer related to Alagomyidae+Rodentia then to Eurymylidae, while Asher et al. (2005) considered Heomys as member of a monophyletic Eurymylidae but that as closer related to Lagomorpha. Last but not least, Meng et al. (2003), Asher et al. (2019), and López-Torres et al. (2020) also found Heomys in monophyletic Eurymylidae, but the latter as closer related to Rodentia, in agreement with the majority of other experts. Should we have any confidence in this result, – you are not supposed to ask such heretical questions.
The extinct Alagomyidae have been found in Late Paleocene and Early Eocene deposits of North America and Asia (McKenna & Bell 1997, Meng et al. 2007). They were often classified in Rodentia, because they share the typical single pair of incisors with rodents (Halliday et al. 2015). However, they seem to be advanced stem group representatives as sister group of Rodentia (Rodentiaformes) rather than crown group representatives (Meng et al. 1994, Meng & Wyss 1994, 2001, Dawson & Beard 1996, Van Valen 2002, Dawson 2003, Meng 2004, Rose 2006, Janis et al. 2008, Fostowicz-Frelik 2020). Tribosphenomys from the Late Paleocene of China and Mongolia is particularly rodent-like (Meng et al. 1994, Meng & Wyss 2001). Nevertheless, Meng et al. (2003) and López-Torres et al. (2020) has Tribosphenomysas more basal than Paramys in the stem group of rodents, unlike most other studies. Finally, at least one study (Marivaux et al. 2004) did not support the monophyly of Alagomyidae, but had Alagomys more closely related to rodents than Tribosphenomys. Disagreement and incongruence everywhere you look!
The earliest fossil record of crown group rodents could be the family Ischyromyidae (= Paramyidae) from Late Paleocene to Early Oligocene deposits of North America and Eurasia (Matthew 1910, Jepsen 1937, Anderson 2008). They may be closer related to the squirrel group of rodents (Simpson 1945, Van Valen 2002, Janis et al. 2008, Kay & Hoekstra 2008, Asher et al. 2019), even though other workers rather considered them as stem rodents (e.g., Fostowicz-Frelik 2020). The earliest fossils of ischyromyids are species like Asiaparamys shevyrevae from the Latest Paleocene of Kazakhstan (Nessov 1987, Averianov & Martin 2001) as well as Acritoparamys atavus, Paramys adamus, and Franimys sp. from the Latest Paleocene of north-western USA (Wood 1962, Carroll 1988, Dawson & Beard 1996, Anderson 2008, Beard & Dawson 2009). Genera like Paramys (Matthew 1910, Jepsen 1937, Wood 1962) and Ischyromys were arboreal mouse-like animals, while other ischyromyid genera like Manitsha were terrestrial and 1.5 times larger than a beaver.
More or less contemporaneous with stem rodents there is also the first record of crown group rodents from the Ctenodactyloidea such as the genus Tamquammys from the earliest Eocene (ca. 56 mya) Erlian Basin of Nei Mongol in China (Fostowicz-Frelik et al. 2018).The ctenodactyloids are a primitive group of living rodents that includes gundis and the Laotian rock rat, a living fossil (Biello 2006).
The Issue of Incongruence
There is a final curious fact about rodents that must be mentioned because it perfectly illustrates the important issue of incongruence between morphological and molecular/genetic data that is unexpected under Darwinism: early molecular data were interpreted as evidence that rodents are not monophyletic (Graur et al. 1991, 1992, Li et al. 1992; also see D’Erchia et al. 1996), which was extremely surprising as no morphologist ever doubted the monophyly of rodents. Molecular biologist Dan Graur, who is well known to ID proponents as “The Vigilante Who Wants to Retain the Myth of Junk DNA,” mentioned his weird idea that guinea pigs are not rodents at a public lecture at my university in Tübingen, Germany, when I was still a student. I vividly remember how we systematic biologists, educated in the great German tradition of classic comparative morphology, had nothing but ridicule and scorn for such nonsensical results of early molecular phylogenetics. It was not just that those guys suggested that rodents are diphyletic, but ideas about the diphyly of well-established groups became a kind of popular fad for molecular biologists: for example, they also suggested that well-defined groups such as bats and whales are diphyletic, and more recently that the insectivorous Eulipotyphla are diphyletic (Kjer & Honeycutt 2007) which we will discuss next in this article series.
It goes without saying that none of these ideas stood the test of time (Hasegawa et al. 1992, Novacek 1992, Honeycutt & Adkins 1993, Luckett & Hartenberger 1993, Sullivan & Swofford 1997, Janis et al. 2008, Zhou et al. 2015). Other apparently crazy ideas suggested by molecular systematics, such as the non-monophyly of taxa like Articulata (annelids and arthropods) and Tracheata (myriapods and insects), indeed became the new orthodoxy, even though there is still some courageous resistance from the “indomitable village” of classical morphologists (e.g., Scholtz 2002, Wägele & Kück 2014). The same holds for the reverse case, where the traditional view, based on anatomical and fossil evidence, that pinnipeds are diphyletic (walruses and sea lions derived from a bear-like ancestor, and true seals derived from mustelids), was overturned by molecular studies (Koretsky et al. 2016, Hafed et al. 2020). Even though there obviously is rampant conflict between molecular and anatomical data, and between molecular clock datings and the actual fossil record, which even led scientists to question the “adequacy of morphology for reconstructing the early history of placental mammals” (Springer et al. 2007a), few scientists are courageous enough to conclude from such conflicting evidence that we may have to question the Darwinian framework of evolutionary biology. To be clear: this does not mean that common descent is refuted. Rather it means that the evidence strongly suggests that a blind and random process is no causally adequate and sufficient explanation for the diversity and complexity of life.
Anyway, we can conclude from the above discussion of the fossil record of Glires that the earliest stem Glires as well as stem lagomorphs and stem rodents already appeared about 61 million years ago in Early Paleocene of East Asia. Based on Benton et al. (2015) the Fossil Calibration Database took a bit more conservative approach and dates the minimum age of Glires to 56 million years, based on the phylogenetic assignment (Meng et al. 2003, Asher et al. 2005) and younger Thanetian dating (see above) of the genera Mimotona and Heomys.
Since we have now completed our review of the supergroups (cohorts) Xenarthra, Afrotheria, and Euarchontoglires, we will move on to the final of the supergroups of placental mammals, which has been named Laurasiatheria. We will start with insectivores (Eulipotyphla) as its most basal order, but this will be in March, because due to deadlines for the rest of this month Fossil Friday will only feature photos of some nice fossil dragonflies.