KOGAIONIDAE IN
|
publication ID |
https://doi.org/ 10.1206/0003-0090.456.1.1 |
|
persistent identifier |
https://treatment.plazi.org/id/0B3C494E-FF9D-FFA2-FF1F-AF2A2B77FE29 |
|
treatment provided by |
Felipe |
|
scientific name |
KOGAIONIDAE IN |
| status |
|
KOGAIONIDAE IN PHYLOGENY, TIME, AND SPACE
In order to better understand the wider significance of the latest Cretaceous Transylvanian multituberculates, a brief outline of their place within Multituberculata is required. Multituberculates are members of Allotheria, a nontherian mammalian clade placed close to the origin of Theria (e.g., Kielan-Jaworowska et al., 2004). The clade Allotheria groups exclusively extinct mammals that are currently considered more derived than prototherians (hence, they are members of the crown-group Mammalia), but less derived than therians (Metatheria plus Eutheria) as well as several fossil groups leading up to the therian root (such as symmetrodonts, dryolestoids, and stem zatherians; e.g., Luo et al., 2002; KielanJaworowska et al., 2004; Yuan et al., 2013; Han et al., 2017). The latest phylogenies find that, within Allotheria, multituberculates are the sister-group of Haramiyida, a probably paraphyletic assemblage of Late Triassic–Late Jurassic mammaliaforms with a superficially multituberculate-like, policusped postcanine dentition (e.g., Yuan et al., 2013; Han et al., 2017; fig. 4). Other analyses recovered multituberculates as allotherians closely allied to Haramiyida plus Gondwanatheria ( Krause et al., 2014, 2020; but see, e.g., Huttenlocker et al., 2018, for an alternative placement of these clades), a group of mammaliforms with peculiar, hypsodont-crowned molars and which had an exclusively southern, Gondwanan distribution during the later Mesozoic and Paleogene- Miocene (e.g., Krause and Bonaparte, 1993; Gurovich and Beck, 2009; Chimento et al., 2015).
Among multituberculates, kogainonids are definitively members of the derived, probably monophyletic clade Cimolodonta (e.g., KielanJaworowska and Hurum, 2001; fig. 4A), as supported conclusively by a large number of shared derived craniodental characters (e.g., KielanJaworowska et al., 2004; Csiki et al., 2005; Csiki and Grigorescu, 2006; Smith and Codrea, 2015; Csiki-Sava et al., 2018; Smith et al., 2022). These characters include foremost the overall skull shape as well as the significantly reduced dental formula of the kogaionids ( fig. 5 View FIGURE 5 ). The dentition is reminiscent of the plesiomorphic cimolodontan condition (e.g., Kielan-Jaworowska and Hurum, 2001; Kielan-Jaworowska et al., 2004), with an upper tooth row that consists of two incisors (I2 and I3), no canine, four premolars, and two molars ( Rădulescu and Samson, 1996; Peláez-Campomanes et al., 2000; Smith and Codrea, 2015; Csiki-Sava et al., 2018; Smith et al., 2022; fig. 5C, F View FIGURE 5 ), as well as a lower tooth row with one incisor (i1), most probably only one premolar (p4; see below), and two molars ( Csiki et al., 2005; Smith and Codrea, 2015). Further cimolodontan apomorphies recognized within Kogaionidae (see Csiki-Sava et al., 2018) include the rather simple, peglike I3 (although in kogaionids, this incisor is relatively large compared to I2, and may be slightly more complex in morphology than usually found in other cimolodontans); the arcuate, relatively high p4 that protrudes slightly above the level of the molar occlusal plan, and presents a single labiodistal structure (commonly a platformlike feature that is better developed and more complex in morphology than the comparable structure in other cimolodontans), as opposed to the row of labiobasal cusps seen in the more basal plagiaulacidan multituberculates; and the presence of a prismatic (gigantoprismatic: Fosse et al., 2001) enamel structure.
The exact position of Kogaionidae within Cimolodonta remains, however, controversial ( fig. 4), in part due to the nature of the group’s fossil record (which was mainly fragmentary until very recently), but also in part to the unstable nature of higher-level multituberculate relationships as currently understood. Accordingly, kogaionids have been placed alternatively in a large polytomy with a number of other important cimolodontan clades ( Kielan-Jaworowska and Hurum, 2001; fig. 4A), as basal cimolodontans (e.g., Rougier et al., 1997, Smith et al., 2022; fig. 4B), even forming a nonmonophyletic assemblage of taxa that represent successively more derived lineages at the base of Cimolodonta ( Csiki and Grigorescu, 2006), or as closely related to the clade of dominantly large Asian–North American taeniolabidoids (e.g., Smith and Codrea, 2015; Mao et al., 2016; Csiki-Sava et al., 2018; fig. 4C, D). These controversies stem to a great extent from the peculiar and highly autapomorphic craniodental anatomy of the kogaionids ( fig. 5 View FIGURE 5 ) that spurred the recognition of a distinct multituberculate family once the first relatively well-preserved remains of a kogaionid (the partial skull of the group namesake Kogaionon ) was identified and described ( Rădulescu and Samson, 1996).
The diagnosis of the Kogaionidae has evolved significantly over the years, as a more and more inclusive membership had been identified, as well as more completely preserved remains were uncovered and described; these changes in the diagnosis (reflecting our understanding concerning the nature of Kogaionidae ) were most recently reviewed and commented upon by Csiki-Sava et al. (2018: Supplementary Appendix) as well as by Smith et al. (2022). According to our current ideas, the family is diagnosed by a series of cranial and dental features that are either autapomorphic to the clade, or else form a unique combination of characters. The skull of the kogaionids is as yet incompletely known, but at least in the case of four latest Cretaceous taxa ( Barbatodon transylvanicus , Kogaionon ungureanui and K. radulescui , and Litovoi tholocephalos ) enough of it is preserved so as to allow reasonably complete and detailed reconstructions ( figs. 5A, B, D, E View FIGURE 5 , 6A–H View FIGURE 6 ). In all of these four taxa, the skull appears to be autapomorphically constructed with a narrow and relatively long snout region followed by distal skull segment characterized by widely diverging zygomatic arches that are oriented transversaly and arise at the level of the P3/P4 contact. Although neither of these characteristics is in itself unique to Kogaionidae (see discussions in Kielan-Jaworowska et al., 2004; Csiki-Sava et al., 2018: Supplementary Appendix), their combination is, and it is this combination that sets apart the skulls of kogaionids from those of other cimolodontans. Furthermore, kogaionids appear to have relatively large, massive, and tall premaxillae ( fig. 6D, F View FIGURE 6 ). Detailed description of the known kogaionid cranial material is needed before further kogaionid apomorphies can be identified.
The most diagnostic feature of the kogaionids is, however, represented by their dentition, and especially their upper dentition ( fig. 5C, F View FIGURE 5 ). Most remarkably, the premolar row is lenghtened relative to the molar row, often amounting to twice the length of the latter. Furthermore, the longest element of the premolar row is represented by the elongated P3, although its relative development compared to P4 varies between the different taxa, from clearly longer than P4 to subequal with the latter. The P3 is often even longer than, or at least comparable in its length to, the largest upper molar (M1). In its turn, the P4 is remarkable in that it develops a unique occlusal morphology, unlike that seen in other cimolodontans, with an obliquely striking main labial cusp row seconded by a short and distally restricted lingual cusp row, a morphology first identified and recognized as diagnostic for the clade by PeláezCampomanes et al. (2000). Finally, the M1 is characterized by a unique combination between a short and wide occlusal outline, a reduced cusp count (only four cusps are present in the median row in all known kogaionids), and a well-developed lingual cusp row that extends mesially for at least half the mesiodistal length of the crown.
In the lower dentition, the apomorphically arched p4 is associated with a relatively short m1 that shows a reduced cusp formula of 3–4 cusps in each row; the relative development of the p4 compared to m1 varies between the different taxa, from 50% longer than m1 to more than twice as long. Finally, another kogaionid synapomorphy may be represented by the unequally developed, lingually reduced enamel cover of the incisors (in i1 and I2, at the least), although this feature is currently documented only in the latest Cretaceous representatives of the clade. It should be also emphasized that although our knowledge concerning the anatomy of the kogaionids is limited mainly to their craniodental morphology, recent discoveries reveal details of their postcranial anatomy as well ( Csiki-Sava et al., 2018; Vremir et al., 2018; Botfalvai et al., 2021; fig. 6I, J View FIGURE 6 ).
Due to their highly apomorphic craniodental morphology, the true affinities of the first discovered kogaionid remains, which at that moment were represented exclusively by isolated teeth, had been obscured by their fragmentary nature. Indeed, the first found kogaionid fossils, coming from the Paleocene of Western Europe ( Belgium and France) and referred to the genus Hainina ( Vianey-Liaud, 1979, 1986), were left unassigned at a higher taxonomic level ( Vianey-Liaud, 1979; McKenna and Bell, 1997), or else were tentatively included either in Cimolomyidae ( Vianey-Liaud, 1986) or in Ptilodontoidea ( Rougier et al., 1997). Meanwhile, the second identified kogaionid taxon ( Barbatodon from the uppermost Cretaceous of Romania; Rădulescu and Samson, 1986) was considered related to the Asian djadochtatherioids ( Rădulescu and Samson, 1986), or, alternatively, to the North American basal cimolodontans of the “ Paracimexomys group” ( Grigorescu and Hahn, 1987). Such widely diverging taxonomic opinions were the byproduct of the unique dental morphology of these multituberculates, with combinations of characters hard to fit into the diagnoses of previously separated taxonomic categories.
The discovery of Kogaionon , the third recognized kogaionid genus (also coming from the uppermost Cretaceous of Romania; Rădulescu and Samson, 1996; fig. 5D, F View FIGURE 5 ), finally offered the opportunity to understand the true taxonomic and systematic affinities of Hainina and Barbatodon . First Hainina ( Peláez-Campomanes et al., 2000; Kielan-Jaworowska and Hurum, 2001), then Barbatodon as well ( Csiki et al., 2005) were securely nested in the newly erected Kogaionidae , alongside Kogaionon . Recently, the family was enlarged with the description of a fourth genus ( Litovoi ) also originating from the uppermost Cretaceous of Romania ( Csiki-Sava et al., 2018; figs. 5A–C View FIGURE 5 , 6 View FIGURE 6 ).
According to a series of discoveries spread over the past 40 years, the currently known fossil record of kogaionids is restricted exclusively to Europe, and the group is often regarded as a typical European endemic clade (e.g., Csiki-Sava et al., 2015). Its stratigraphically oldest representatives are reported from the uppermost Campanian–lowermost Maastrichtian of Romania ( Vremir et al., 2014, 2015b), and they occur throughout the Maastrichtian of the southwestern Transylvanian area ( fig. 3 View FIGURE 3 ). Here, they are represented by the above-mentioned three genera ( Barbatodon , Kogaionon , Litovoi ), with Barbatodon and Kogaionon each including two species—the medium-large type species B. transylvanicus ( Rădulescu and Samson, 1986; Csiki et al., 2005; Smith and Codrea, 2015; Solomon et al., 2016) and the smaller B. oardaensis ( Codrea et al., 2014, 2017a) in the case of the former, and the larger K. ungureanui ( Rădulescu and Samson, 1996) and the smaller K. radulescui ( Smith et al., 2022) for the latter. Although Hainina has been also reported from the same area and time interval (e.g., Csiki and Grigorescu, 2000, 2002; Kielan-Jaworowska et al., 2004), this original taxonomic referral should be now reconsidered, given the now significantly improved local fossil record of the group (see below).
Besides these four nominal taxa, kogaionid remains that are as yet unidentified at a genus and/or species level have also been reported from several locations across southwestern Transylvania. Some of these remains had been tentatively referred to already known genera, but left in open nomenclature at a species level, while others might represent even further new taxa or else remain indeterminate at a lower taxonomic level due to their fragmentary and/ or nondiagnostic nature. Besides the taxonomic diversity it already reveals, the known fossil record of these Transylvanian kogaionids is also characterized by a noticeable diversity in craniodental morphology and especially in body size ( fig. 7 View FIGURE 7 ; see also below). Most remarkably, according to our current knowledge all latest Cretaceous kogaionids were restricted exclusively to the Transylvanian Landmass, and the clade can be thus identified as being strictly endemic for this area at the end of the Cretaceous ( Csiki and Grigorescu, 2002, 2006; CsikiSava et al., 2015; Smith and Codrea, 2015; Solomon et al., 2015, 2016). This could, of course, change with future discoveries of kogaionids in other Upper Cretaceous units in Europe, but for the time being their absence in otherwise well-sampled terminal Cretaceous faunas in France and Spain is striking (e.g., Csiki-Sava et al., 2015), and supportive of local endemicity in the Transylvanian area.
Starting at the very beginning of the Paleocene, kogaionids suddenly achieved a significantly wider spatial distribution, as their remains have been identified from the lowermost Paleocene of northern Spain ( Peláez-Campomanes et al., 2000), the lower Paleocene of Belgium ( Vianey-Liaud, 1979; De Bast and Smith, 2017), as well as the upper Paleocene of France, Romania, and possibly Spain ( Vianey-Liaud, 1986; Gheerbrant et al., 1999, 2003; López-Martínez and Peláez-Campomanes, 1999). Despite this important paleobiogeographical range expansion, kogaionids remained constrained to Europe; they have not been found, for instance, in any of the well-sampled Paleocene rock units of North America (e.g., Rose, 2006). Furthermore, their higher-level taxonomic diversity apparently dropped in the Paleocene compared to the latest Cretaceous; only one genus, Hainina , is recorded throughout the epoch, with several temporally and—in part—spatially nonoverlapping species: the early Danian Spanish H. pyrenaica , the sympatric but differently sized, late Danian H. belgica and H. godfriauxi from Belgium, and the Thanetian H. vianeyae from France, while a currently indeterminate Hainina sp. has also been reported from the Thanetian of Romania.
The evolutionary fate of the kogaionids dwindled during the Paleocene. While they formed almost exclusively the European multituberculate assemblages during the Danian ( Vianey-Liaud, 1979; Peláez-Campomanes et al., 2000; De Bast and Smith, 2017), kogaionids became subordinate by the early Thanetian to the neoplagiaulacid cimolodontans, introduced from North America in an as yet poorly understood mid-Paleocene dispersal event ( Vianey-Liaud, 1986; De Bast and Smith, 2017). Kogaionids appear to have survived into the latest Paleocene in some marginal regions of Europe ( Romania, Spain), where they still occured alongside more common placentals that were already represented by relatively modern groups ( Gheerbrant et al., 1999; López-Martínez and Peláez-Campomanes, 1999). In other parts of Europe, however, kogaionids were either replaced entirely by the immigrant neoplagiaulacids (e.g., sylvanicus, left p 4 in dentary UBB P-Mt3-1, Pui (site PB5), Hațeg Basin (from Solomon et al., 2016). Jehle et al., 2012), or else multituberculates disappeared altogether with the advent of the more derived, modern placental orders (e.g., Smith et al., 2014), following an important dispersal-related faunal turnover near the Paleocene-Eocene boundary (e.g., Hooker and Collinson, 2012; De Bast and Smith, 2017).
No known copyright restrictions apply. See Agosti, D., Egloff, W., 2009. Taxonomic information exchange and copyright: the Plazi approach. BMC Research Notes 2009, 2:53 for further explanation.