Agerinia marandati sp. nov., a new early Eocene primate from the Iberian Peninsula, sheds new light on the evolution of the genus Agerinia

Studied material

The material studied comes from the fossil site Masia de l’Hereuet and consists of a right mandible fragment (IPS-82807) that preserves the teeth from P₂ to P₄ and the root of the P₁, and 15 isolated teeth identified as: one right and one left dP₄ (IPS-82796; IPS-82797); one left P₄(IPS-82806); two complete and one broken left M₁ (IPS-82800; IPS-82801; IPS-82802); one complete left M₂ and one fragment of a right M₂ (IPS-82805; IPS-82798); two left M₃ (IPS-82803, IPS-82804); one right dP⁴ (IPS-82814); one entire and one broken left M¹ (IPS-82808; IPS-82809); one left M² (IPS-82815); and one left M³ (IPS-82799). All this material is housed at the Institut Català de Paleontologia Miquel Crusafont, ICP (Sabadell, Spain).

Comparative sample

The material studied from Masia de l’Hereuet has been directly compared with the specimens of Agerinia roselli from Les Saleres (Spain) and Agerinia smithorum from Casa Retjo-1 (Spain), both stored at the ICP collections. In addition, the studied sample has been compared with Agerinia cf. roselli from Azillanet (France), belonging to the collections of the Université de Montpellier; and with Agerinia sp. from Condé-en-Brie (France), Donrussellia gallica, Pronycticebus gaudryi and Protoadapis curvicuspidens, which are stored at the MNHN. It has also been compared with high-quality casts of Periconodon huerzeleri, Donrussellia magna, Donrussellia provincialis, Europolemur klatti, Protadapis ignoratus, Cantius eppsi, Marcgodinotius indicus and Asiadapis cambayensis, also stored in the MNHN. Finally, comparisons with Agerinia sp. from Casa Ramón (Spain), cf. Agerinia from Rians (France), Periconodon sp. from Eckfeld Maar (Germany), Periconodon lemoinei, Periconodon jaegeri, Donrussellia lusitanica, Donrussellia russelli, Donrussellia louisi, Darwinius masillae, Europolemur koenigswaldi, Europolemur dunaifi, Europolemur kelleri, Protoadapis angustidens, Protoadapis brachyrhynchus, Protoadapis weigelti, Protoadapis muechelnensis and Cantius savagei, are based on published data.

Dental nomenclature, measurements and micrographs

The dental nomenclature used in the descriptions is that proposed by Szalay & Delson (1979). Measurements have been taken with an optic caliper “Nikon measuroscope 10” connected to a monitor “Nikon SC112”, using the criteria described by Marigó, Minwer-Barakat & Moyà-Solà (2010). Micrographs have been taken using the Environmental Scanning Electron Microscope (ESEM) at Universitat de Barcelona.

New zoological taxonomic name

The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid:zoobank.org:pub:729814E7-5509-48C1-9FF0-84D3E515D909. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central and CLOCKSS.

Phylogenetic analyses

Two phylogenetic analyses were run using a version of a character-taxon matrix of living and extinct primates as well as euarchontan outgroups that was originally published by Seiffert et al. (2005). This matrix has been successively modified (Seiffert et al., 2009) and a recent version was used by Marigó et al. (2016). The matrix analysed here (Data S1) includes 391 characters and 109 taxa for the first analysis, or 112 taxa for the second analysis (see ‘Results of the Phylogenetic Analyses’). The three taxa added in the matrix for the second analysis are Donrussellia gallica, Periconodon huerzeleri and Darwinius masillae. Their codification was taken from older versions of the matrix used by Seiffert et al. (2010) and Marigó, Minwer-Barakat & Moyà-Solà (2011). In both analyses some multistate characters were treated as ordered, and those with polymorphisms scored as intermediate states were scaled to a half-step so that transitions between adjacent “fixed” states in morphoclines were equal to one full step. Both parsimony analyses were run in PAUP 4.0b10 (Swofford, 1998) for 5,000 replicates with random addition sequence and the tree-bisection-reconnection branch-swapping algorithm. Both analyses were constrained by a molecular scaffold using a constraint tree (Data S2), and treated premolar loss as reversible.

Description

Mandible. This specimen preserves the distal part of the canine alveolus, which is placed just mesially with respect to the P₁, with no diastema between these two teeth. The P₁ root is preserved, which corresponds to a small premolar, and is aligned with the rest of the teeth on the longitudinal axis of the mandible. The mandible preserves all the teeth from P₂ to P₄ and the mesial root of the M₁; however, they are all damaged. Furthermore, on the lingual side of the mandible, below the P₁ and P₂, a slightly protruding stripe can be observed descending from the mesial to the distal part, probably indicating an unfused mandible. On the buccal side of the mandible there are three mental foramina of similar size and a tiny one mesially situated just on the bottom of the central one. The mesial-most foramen is broken and placed between the alveoli of the canine and P₁. The central foramina are situated between the root of the P₂ and the mesial root of the P₃ and the distal-most foramen is placed underneath the mesial root of the P₄.

dP₄. The trigonid is elongated, much longer than wide; it is slightly shorter and clearly narrower than the talonid. There is a broad space separating the paraconid and metaconid. The trigonid is open lingually. The paraconid is well differentiated, clearly smaller than protoconid and metaconid, and placed in the mesiolingual extreme of the tooth. The paracristid runs buccally from the paraconid and curves distally reaching the protoconid. The protoconid is located mesially with respect to the metaconid. Both cuspids are similar in height and connected by a protocristid, oblique to the buccal and lingual borders. There is no premetacristid. The cristid obliqua reaches the apex of the metaconid. The talonid basin is deep. The hypoconid is placed mesially with respect to the entoconid. The postcristid is thickened at its central part, but it does not bear a distinct hypoconulid. There is a well-marked notch that separates the preentocristid and the postmetacristid. The buccal cingulid is slightly marked and restricted to the mesiobuccal corner of the tooth, from the buccal base of paraconid to the distal base of protoconid. There are two roots.

P₂. The P₂ is slightly longer than broad, with an oval outline. It is larger than P₁ (based on the size of the roots), and clearly smaller than P₃ and P₄. The morphology is rather simple, with a single distinguishable cuspid (protoconid) and two sharp cristids directed mesially and distally from the apex to the base of the crown. At the distal end of the tooth, there is a small but distinguishable bulge centrally located. There is a weak cingulid on the mesial part of the tooth, and a more developed cingulid occupying the distolingual and the distobuccal borders. There is only one root.

P₃. There is only one P₃ known, twice long as it is broad. The crown is strongly damaged, preventing the observation of the dental morphology. However some traits can be described, particularly the presence of a protuberance on the distal end of the crown, centrally located, that does not constitute a clear cuspid. The buccal and lingual cingulids are faintly marked and restricted to the mesial and distal ends of the tooth. On the distolingual corner, the lingual cingulid encloses an incipient and shallow talonid basin. There are two roots.

P₄. The P₄ is larger than P₃. The outline is oval, somewhat more mesiodistally elongated in specimen IPS-82807 than in IPS-82806. There is a distinct paraconid in mesiolingual position. The paracristid descends from the protoconid apex, curving lingually at its end and reaching the paraconid. The metaconid, clearly distinguishable, is attached to the distolingual side of the protoconid in specimen IPS-82806 (the poor preservation of specimen IPS-82807 prevents the observation of this trait). There is a distinct cristid obliqua that runs from the joint between protoconid and metaconid to the hypoconid, which is centrally located at the distal side of the tooth (this trait is also only observed in specimen IPS-82806). The hypoconid is as high as the paraconid. The buccal and lingual cingulids, high and well marked, surround the entire base of the tooth in specimen IPS-82806; in IPS-82807 these cingulids seem less marked, but it can be due to the bad preservation of this specimen. The lingual cingulid encloses a shallow talonid basin, restricted to the distolingual corner. There are two roots.

M₁. The trigonid is shorter and much narrower than the talonid. The trigonid basin is as long as it is wide and it is open between the paraconid and the metaconid. The paraconid is well differentiated but clearly smaller than the other trigonid cuspids; it is placed on the mesiolingual corner of the tooth, closer to the metaconid than to the protoconid. The paracristid runs mesially from the protoconid and, at the mesiobuccal corner of the tooth, curves buccally reaching the paraconid. The protoconid is located in a more mesial position than the metaconid. Both cuspids are connected by a protocristid, clearly oblique to the lingual and buccal borders, that shows a well-marked V-shaped valley. In the specimens IPS-82800 and IPS-82801, there is no premetacristid, whereas IPS-82802 shows a cristid directed mesially from the metaconid, which does not reach the paraconid. In this latter specimen, the metaconid apex is slightly curved distally. The cristid obliqua reaches the trigonid wall at the level of the buccal base of the metaconid. The talonid basin is clearly deeper than the trigonid basin. The hypoconid is placed mesially with respect to the entoconid. The postcristid connects the hypoconid and entoconid, which shows several enamel swellings that do not constitute a distinct hypoconulid. The preentocristid connects to the postmetacristid closing the talonid lingually. IPS-82801 shows a small but well differentiated bulge on the middle of the preentocristid. The buccal cingulid runs from the mesiobuccal base of the paraconid to the distolingual base of the hypoconid. This cingulid is very strong on the mesial border and below the protoconid, and weaker at the level of the talonid, being interrupted below the hypoconid in IPS-82800.

M₂. The trigonid is wider than long; it is much shorter than the talonid, but similar in width, and therefore the outline of the tooth is rectangular. There is no paraconid. The paracristid runs from the protoconid, surrounding the mesial border and continuing into a premetacristid that reaches the metaconid, closing the trigonid basin lingually. The protoconid is located in a slightly more mesial position than the metaconid, so the protocristid is almost perpendicular to the buccal and lingual borders of the tooth. The cristid obliqua reaches the trigonid wall close to the lingual base of protoconid. The talonid basin is closed, moderately shallow and as wide as it is long. The hypoconid is slightly more voluminous than the entoconid and placed in a slightly more mesial position. The postcristid joins hypoconid and entoconid; at its distobuccal part, it thickens forming a tiny hypoconulid. The preentocristid and postmetacristid are connected, closing the talonid basin lingually. The buccal cingulid is weak and restricted to the base of the protoconid.

M₃. The trigonid is much shorter than the talonid and similar in width. There is no paraconid. The paracristid continues into a premetacristid reaching the metaconid, so the trigonid basin is closed. The protoconid is placed in a slightly more mesial position than the metaconid. The protocristid is nearly perpendicular to the buccal and lingual margins of the tooth. The cristid obliqua is straight and reaches the trigonid wall in a more buccal position than in the M₂. The talonid basin is closed, moderately deep and longer than wide. The hypoconid is slightly more voluminous and placed more mesially than the entoconid. Preentocristid and postmetacristid are connected, closing the talonid buccally. The hypoconulid lobe is prominent, but longer and better differentiated from the talonid basin in IPS-82803 than in IPS-82804. The postentocristid shows a slightly marked valley between the hypoconulid and the entoconid in specimen IPS-82803. The buccal cingulid is visible on the base of the protoconid. On IPS-82803 a very weak cingulid is hardly observed on the buccal base of the hypoconid.

dP⁴. The specimen is eroded, lacking the enamel. The outline is subtriangular, with the buccal side longer than the lingual side and it has three main cusps: paracone, metacone and protocone. There is no hypocone. The trigon basin is quite deep. The preparacrista is slightly marked and runs from the apex of the paracone, curving buccally, reaching the mesiobuccal corner, which is broken. The postparacrista runs distally from the paracone and connects to the premetacrista, which reaches the metacone. The postmetacrista is straight and connects the metacone to the distobuccal corner of the tooth. The protocone is slightly lower than paracone and metacone. The preprotocrista connects the protocone to a small paraconule. The preparaconule crista runs around the mesiobuccal half of the tooth, from the paraconule to the mesiobuccal corner. The metaconule is more developed than the paraconule. Three cristae run from the metaconule: a well-marked postprotocrista that connects to the protocone, a short premetaconule crista, which is directed to the metacone but does not reach its apex, and a longer postmetaconule crista (also known as lateral posterior transverse crista) that borders the distal margin of the tooth and reaches the postmetacrista at the distobuccal corner of the tooth. There is a faint anterocingulum that runs from the mesiolingual base of the protocone to the mesial half of the tooth, reaching the preparaconule crista. The postcingulum is restricted to the distolingual corner of the tooth, reaching the postmetaconule crista. There are three roots.

M¹. The outline is subtriangular, being much wider than long and with the buccal side somewhat longer than the lingual one. The trigon basin is deep. The paracone is somewhat larger than the metacone; the protocone is slightly lower than the buccal cusps, and the hypocone is clearly smaller and lower than the rest. The preparacrista is weak. From the apex of the paracone, a straight postparacrista descends distally and connects with the premetacrista that reaches the metacone apex. The postmetacrista is straight and runs from the apex of metacone to the distobuccal corner of the tooth. The paraconule is much more developed than the metaconule; it has two buccal cristae: a preparaconule crista, reaching the end of the preparacrista at the mesiobuccal corner of the tooth, and a hypoparacrista that reaches the apex of the paracone. These cristae enclose a well-developed basin mesiolingual to the paracone. The preprotocrista is well marked and connects the paraconule and the protocone. The metaconule is connected to the metacone by a hypometacrista and to the protocone by a postprotocrista. There is a small but distinct hypocone at the distolingual corner of the tooth; it is weakly connected to the base of the protocone by a very short and low postprotocingulum. There is a weak buccal cingulum restricted to the central part of the buccal border, between paracone and metacone. The anterocingulum is marked and connects to the preparaconule crista; it is longer in IPS-82809 than IPS-82808, surrounding the mesiolingual corner in the former. The posthypocone crista continues in a long and well-marked postcingulum that occupies the entire distal border, being interrupted at the distobuccal corner, without connecting to the postmetacrista. The specimen IPS-82808, shows a slightly wrinkled enamel surface at the lingual base of the protocone.

M². It is larger than the M¹, subtriangular, with the lingual side slightly shorter than the buccal side. The trigon basin is very deep. The paracone is slightly larger than the metacone and similar in height to the protocone. The preparacrista is well marked and straight; it connects the paracone to a small parastyle. The postparacrista and premetacrista are sharp and descend strongly, so their connection is placed in a very low position. The postmetacrista descends from the apex of the metacone and curves slightly towards the buccal side, reaching the distobuccal corner of the tooth. It thickens at its distal end, forming an incipient metastyle. Paraconule and metaconule are similar in size. From the paraconule, a preparaconule crista borders the mesial part of the tooth, reaching the parastyle. There is a deep basin mesiolingual with respect to the paracone, enclosed by the preparaconule and hypoparacrista. From the protocone, the preprotocrista and the postprotocrista connect to the paraconule and the metaconule, respectively. The hypometacrista is well marked and joins the metaconule with the metacone. There is a well-developed hypocone, placed in a marginal position at the distolingual corner of the tooth and protruding strongly on the outline of the molar. It is connected to the distal base of protocone by a short postprotocingulum. There is no pericone. The buccal cingulum is well developed and runs from the parastyle to the distobuccal corner of the tooth, without meeting the metastyle. The anterocingulum is also strong; it starts at the mesiolingual corner of the tooth and connects to the preparaconule crista. The anterocingulum continues into a lingual cingulum, which is crenulated at the base of the protocone. The posthypocone crista continues in a postcingulum that extends to the distobuccal corner of the tooth, without connecting to the metastyle. This cingulum encloses a very small talon basin. This specimen shows slight enamel wrinkling, especially on the trigon basin and on the distolingual base of protocone.

M³. The outline is triangular. The paracone is notably larger than the metacone and similar in height to the protocone. The trigon basin is deep. There is no hypocone. The preparacrista is curved lingually and connects the paracone to a hardly distinct parastyle. The postparacrista and premetacrista are well marked. The postmetacrista is very weak and curved buccally; it reaches the distobuccal corner of the tooth, without forming a metastyle. The conules are much less developed than in the M¹ and M²: the paraconule is a mere thickening of the preprotocrista, and the metaconule is absent. The preparaconule crista is sharp and long, reaching the parastyle. The hypoparacrista is weaker than the preparaconule crista; it descends lingually from the paracone but does not reach the paraconule. The postprotocrista connects the protocone and the metacone. The buccal cingulum starts at the mesiobuccal base of the paracone, without reaching the parastyle. This cingulum is strong at the base of paracone, but weak and discontinuous at the base of metacone. The anterocingulum starts at the base of the paraconule, surrounds the mesiolingual border of the tooth and continues in a strong postcingulum that occupies the whole distal border. This cingulum shows some bulges at its lingual base that do not constitute distinct cuspids. The enamel is slightly wrinkled on the distolingual side.

Comparisons

Comparisons with other samples attributed to Agerinia. The new species Agerinia marandati is quite similar in size and in some morphological traits to Agerinia roselli from Les Saleres (Fig. 6; Crusafont-Pairó, 1967; Szalay, 1971; Femenias-Gual et al., 2016a) such as the single-rooted P₂, the distinct paraconid on the P₄ or the lack of paraconid on the M₂. However, A. marandati clearly differs in other several traits. The root of the P₁ of A. marandati is centrally located on the mesiodistal axis of the mandible, while in A. roselli it is clearly shifted buccally (here, we consider that the most mesial root of A. roselli described by Femenias-Gual et al. (2016a) corresponds to the P₁, see Discussion). The metaconid and hypoconid of the P₄ are slightly better differentiated in A. roselli than in A. marandati. In addition, the P₄ of A. marandati lacks the distinct entoconid that is present in A. roselli. Besides, the M₁ paraconid is clearly larger in A. marandati, while it is very small in A. roselli. Finally, the distal and intermediate mental foramina of A. roselli are clearly more mesially located than in A. marandati.

Regarding Agerinia smithorum from Casa Retjo-1 (Femenias-Gual et al., 2016b), it is also very similar in size to A. marandati (Fig. 6). Furthermore, both species share several traits such as the central position of the P₁ on the mesiodistal axis of the mandible, the lack of entoconid on the P₄, the well-developed paraconid on the M₁, or the similar disposition of the distal and intermediate mental foramina, this latter being only a little more mesially located in A. marandati than in A. smithorum. Nevertheless, there are several differences between these species. The P₁ alveolus is more compressed mesiodistally in A. marandati than in A. smithorum, thus suggesting a more reduced premolar (the P₁ crown is not preserved in these species). The number of roots of P₂ is different, being double-rooted in A. smithorum and single-rooted in A. marandati. Additionally, the latter species differs from A. smithorum in the molarization of the P₄. A. marandati shows a well developed metaconid and distinct paraconid and hypoconid, while in A. smithorum the metaconid is smaller and the paraconid and hypoconid are absent. Furthermore, A. marandati lacks the paraconid on the M₂, while A. smithorum shows a tiny one. Finally, the mesial-most mental foramen is clearly more mesially located in A. marandati than in A. smithorum.

Godinot (1983) described two partial mandibles of Agerinia cf. roselli from Azillanet (MP10, France), which are similar in size or slightly larger than A. marandati. Furthermore, the M₁ from Azillanet clearly differs from that of A. marandati in lacking the paraconid.

Peláez-Campomanes (1995) documented the presence of Agerinia sp. in the fossil site Casa Ramón (MP11, N Spain), which is clearly smaller than A. marandati. Moreover, they differ in several traits. In the M₁ from Casa Ramón, the paracristid forms an acute angle in the mesiobuccal corner; this angle is obtuse in A. marandati. The M₂ of Agerinia sp. is proportionally narrower than that of A. marandati, and the paracristid of the latter is higher than in the specimen from Casa Ramón. Besides, the protocristid of A. marandati is more perpendicular to the buccal and lingual borders of the tooth than in Agerinia sp. from Casa Ramón. The M¹⁻² of Casa Ramón lacks the hypoparacrista that is well marked in the M¹ and M² of A. marandati. Finally, the hypocone is connected to the distal base of protocone by a short postprotocingulum in the M¹ and M² of A. marandati but it is isolated in Agerinia sp.

Herbomel & Godinot (2011) described some specimens from Condé-en-Brie (MP8+9, France), preliminarily assigned to Agerinia sp. This form is, in general terms, somewhat larger than A. marandati and shows slightly more bulbous cusps. The paraconid of the M₁ is larger in A. marandati than in Agerinia sp. from Condé-en-Brie, where it varies from small to moderate. The M¹ and M² of A. marandati show a hypoparacrista that connects the paracone with the paraconule, while in Agerinia sp. there is only a short hypoparacrista that does not reach the paraconule. Furthermore, the hypocone of Agerinia sp. varies in size from large to small or even absent in many M¹ and M², whereas it is well marked in the M¹ and M² of A. marandati. The pericone is absent in the upper molars of A. marandati, while in some M² from Condé-en-Brie the anterior cingulum thickens forming a real pericone. The M³ of Agerinia sp. frequently display one or two lingual cusps; on the contrary, in the M³ of A. marandati there are some bulges at the level of the connection of the anterocingulum and the postcingulum, which do not constitute distinct cusps.

There is a single M₂ from Rians (MP7, France) described by Godinot (1983), which was determined as cf. Agerinia. This specimen is clearly larger than the M₂ of A. marandati and shows some morphological differences. For instance, the difference in width between the trigonid and the talonid is much more marked in the M₂ of cf. Agerinia than in that of A. marandati, which shows a more squared outline. Furthermore, cf. Agerinia shows a well-developed paraconid that is absent in the M₂ of A. marandati. Finally, the M₂ from Rians shows an expansion of the distolingual corner and has the entoconid more distally placed with respect to the hypoconid than in A. marandati.

Comparisons with other Eurasian Notharctidae. The new species A. marandati has been compared to Periconodon and Darwinius, which together with Agerinia, form a close taxonomic group following Godinot (2015). Moreover, it has been compared with Donrussellia, the only other Euprimate genus found in the Iberian Peninsula in the early Eocene, as well as with other Eocene Eurasian notharctids, such as the genera Pronycticebus, Europolemur, Protoadapis, Cantius, Marcgodinotius and Asiadapis.

Although the taxonomy of Periconodon is uncertain, five species and a specimen without specific attribution constitute this genus (following Godinot, 1988; Godinot, 2015): P. helveticus from Ergerkingen (MP13; Rütimeyer, 1891), P. lemoinei from Grauves (MP10, Gingerich, 1977), P. huerzeleri from Bouxwiller (MP13, Gingerich, 1977), P. helleri from Geiseltal (MP13-14, Schwartz, Tattersall & Haubold, 1983), P. jaegeri from Bouxwiller (MP13, Godinot, 1988) and Periconodon sp. from Eckfeld Maar (MP13, Franzen, 2004). Regarding size, the teeth of Agerinia marandati are larger than those of P. huerzeleri and P. helveticus, similar in size to those of P. jaegeri and P. lemoinei, and shorter and broader than those of Periconodon sp. from Eckfeld Maar (Fig. 6; there are no available published measurements of P. helleri). In any case, A. marandati shows strong morphological differences with the genus Periconodon, including the lack of pericone on the upper molars, the better differentiated metaconid on the P₄, the presence of a distinct paraconid on the M₁ and of an entoconid on the M₃ (Franzen, 2004; Godinot, 2015).

Regarding Darwinius masillae from Messel (MP11, Franzen et al., 2009), it is clearly larger than A. marandati (Fig. 6) and lacks the first lower premolar. The upper molars of D. masillae lack the metaconule that is present in A. marandati; the paraconule is barely marked in Darwinius. The hypoparacrista is faint and does not reach the paraconule. Moreover, the talon basin is clearly broader in the upper molars of Darwinius than in those of A. marandati. The anterocingulum and postcingulum of D. masillae are more developed than those of A. marandati. The M₁ of Darwinius lacks a paraconid and shows a closed trigonid basin, whereas in A. marandati this tooth shows an open trigonid basin with a well-differentiated paraconid. Furthermore, the M₁ of Darwinius shows a well-differentiated metastylid that is absent in A. marandati. Finally, the mandible of D. masillae shows only one mental foramen below the P₂, whereas there are three foramina on A. marandati.

Furthermore, A. marandati clearly differs from the genus Donrussellia, which includes six species: D. gallica, D. russelli and D. louisi from Avenay, France (MP8+9, Russell, Louis & Savage, 1967; Gingerich, 1977), D. provincialis from Rians, France (MP7, Godinot, 1978), D. magna from Palette, France (MP7, Godinot et al., 1987) and D. lusitanica from Silveirinha, Portugal (MP7, Estravís, 2000). Concerning the size, A. marandati is clearly larger than Donrussellia provincialis, D. gallica and D. lusitanica and similar in size or slightly smaller than D. magna, D. russelli and D. louisi (Fig. 6). Regarding the morphology, Donrussellia clearly differs from A. marandati in the presence of a well-developed paraconid in all lower molars, whereas A. marandati only shows a well-marked cuspid in the M₁. Furthermore, Donrussellia differs from A. marandati in having a double-rooted P₂ while in the latter it is single rooted. In the M₁ of Donrussellia the trigonid is as long as the talonid, while in A. marandati it is shorter than the talonid. The M¹ and M² of A. marandati clearly differ from those of Donrussellia in having a more developed hypoparacrista. Besides, the hypocone is distinct in the M¹ and M² of A. marandati, whereas this cusp is only present in some M² of Donrussellia.

Additionally, A. marandati is smaller than Pronycticebus gaudryi from Mermerlein in France (MP20, Grandidier, 1904; Szalay, 1971). The cusps are generally more bulbous in P. gaudryi than in A. marandati. Regarding the lower teeth, they differ in the number of roots of the P₂, being double-rooted in P. gaudryi and single-rooted in A. marandati. Besides, the former species shows a paraconid in all lower molars, whereas A. marandati only displays a paraconid in the M₁. The upper teeth of A. marandati show less developed hypocone, parastyle and metastyle than those of P. gaudryi. Furthermore, the trigon basin of A. marandati is wider than in P. gaudryi, in which it is as long as it is wide. Moreover, the M³ of P. gaudryi displays a hypocone, and the hypoparacrista reaches the paraconule, whereas in the M³ of A. marandati the hypoparacrista does not join the paraconule and the hypocone is absent.

The genus Europolemur includes four species following Godinot (2015): E. koenigswaldi and E. kelleri from Messel (MP11, Franzen, 1987; Franzen, 2000), E. klatti from Geiseltal (MP13, Thalmann, 1994) and E. dunaifi from Bouxwiller (MP13, Tattersall & Schwartz, 1983; Godinot, 1988). All these species are much larger than A. marandati. The paraconid is much more developed in the M₁ of A. marandati than in Europolemur, in which this cuspid can be small or absent. Furthermore, among other differences, E. koenigswaldi and E. klatti lack the P₁ whereas A. marandati preserves it. Besides, E. klatti shows a double-rooted P₂, while this premolar is single rooted in A. marandati. The size of the hypocone is variable in the upper molars of Europolemur, being less developed in E. kelleri than in A. marandati, and more developed in E. dunaifi than in A. marandati.

The genus Protoadapis comprises six species according to Godinot (2015): Protoadapis angustidens and Protoadapis brachyrhynchus from unknown levels of the Quercy phosphorites (Russell, Louis & Savage, 1967; Gingerich, 1975), Protoadapis curvicuspidens from different sites including Grauves (MP10, Russell, Louis & Savage, 1967) and Protoadapis weigelti, Protoadapis ignoratus and Protoadapis muechelnensis from Geiseltal (MP12, Thalmann, 1994). All these species are poorly known and only represented by lower teeth, except P. curvicuspidens. In any case, there are several important differences between Protoadapis and A. marandati. The genus Protoadapis is much larger and displays more robust cusps than Agerinia. In addition, the P₃ is clearly higher than the P₄ in Protoadapis, whereas in Agerinia these premolars are more similar in height. Moreover, the paraconid of the lower molars of Protoadapis is shown as a residual cuspule, whereas A. marandati displays a well-developed paraconid on the M₁.

Several species are included in the genus Cantius, but only two are recorded from Europe: Cantius eppsi from Abbey Wood (MP8+9; Hooker, 2010) and Cantius savagei from Muntigny and Avenay (MP8+9; Gingerich, 1977). These two species show notable differences with A. marandati. Both C. eppsi and C. savagei are clearly larger and show more inflated cusps than Agerinia. Besides, Agerinia marandati differs from C. eppsi in having a paraconid only on the M₁, whereas the latter generally shows a well-developed paraconid in all lower molars. Regarding the upper molars, A. marandati shows better-developed hypocone, hypoparacrista and hypometacrista than C. eppsi. In addition, A. marandati has slightly wrinkled enamel on the M² an M³, whereas the teeth of C. eppsi have smooth enamel. Besides, C. eppsi shows a postprotocingulum, while in A. marandati it is absent. Furthermore, C. eppsi displays more developed paraconule, metaconule and lingual cingulum than A. marandati. The M₁ of Cantius savagei is broader than that of A. marandati and displays a slightly shorter talonid basin.

The asiadapine Marcgodinotius indicus from Vastan and Tadkehwar mines (early Eocene, India; Bajpai et al., 2007; Rose et al., 2007; Rose et al., 2009; Smith et al., 2016) differs from A. marandati in several traits. Regarding the size, A. marandati is larger than M. indicus (Fig. 6). Marcgodinotius has a double-rooted P₂, whereas in A. marandati this premolar is single rooted. Moreover, the P₄ of A. marandati displays small but differentiated paraconid and metaconid, whereas the P₄ of M. indicus usually lacks these cuspids. Some specimens of M. indicus show a low metaconid posterolingually placed in relation to the protoconid; on the contrary, in the P₄ of A. marandati the metaconid is higher than in Marcgodinotius and lingually attached to the protoconid. The lower molars of M. indicus display a slighly longer trigonid than those of A. marandati. Furthermore, the M₁ and M₂ of M. indicus show a small, low and buccally shifted paraconid, while A. marandati only displays a well-developed paraconid on the M₁. Furthermore, the outline of the talonid basin of A. marandati is clearly more rounded than in M. indicus. Regarding the upper teeth, the decidual P⁴ of A. marandati lacks the hypocone and the hypoparacrista that are well marked in M. indicus. The difference in length between the buccal and the lingual sides is less marked in the upper molars of A. marandati than in M. indicus. The outline of the upper molars is also different, showing concave mesial and distal borders in M. indicus. Besides, M. indicus differs from A. marandati in having a more marked buccal cingulum and styles in the M¹. The M³ of M. indicus is much wider than that of A. marandati and also differs from the latter in the more developed parastyle and in the presence in some specimens of a small and low paraconule, premetaconule and postmetaconule cristae.

Concerning Asiadapis cambayensis from Vastan mine (early Eocene, India; Rose et al., 2007; Rose et al., 2009), it is smaller than A. marandati (Fig. 6). In addition, Asiadapis lacks the first lower premolar, whereas A. marandati has a small P₁. The P₄ of A. marandati has distinct paraconid and metaconid, whereas theses cuspids are absent in some specimens of A. cambayensis. The paraconid of the M₁ is better developed in A. marandati. Furthermore, some M₂ and M₃ of A. cambayensis show a paraconid, which is absent in A. marandati. Besides, the talonid basin has a rounded outline in the lower molars of A. marandati, whereas in A. cambayensis the talonid basin is more elongated mesiodistally. Regarding the upper teeth, A. marandati shows more developed hypocone, paraconule and metaconule, especially in the M². Furthermore, A. marandati displays slightly wrinkled enamel on the M² and M³, which is smooth in A. cambayensis.