A Pleistocene Fight Club revealed by the palaeobiological study of the Dama-like deer record from Pantalla (Italy)

Taxonomy, variation, and biochronology

The fossils described herein represent one of the most valuable and best-preserved samples of “Dama-like” deer from the European Early Pleistocene. The systematics of these forms has been essentially based on the morphology of the antlers and teeth, with less attention paid to the skull (due to the rarity of well-preserved finds) and postcranial bones.

The Pantalla sample shows a combination of characters allowing an unambiguous attribution to ‘P.’ nestii, a species reported confidently so far in the early Late Villafranchian of Italy (several sites) and in the Georgian Homo-bearing locality of Dmanisi (Supplementary Table S1). Based on the literature^6,8,12,38, these characters include: four-pointed antler with elongated, slender, and tubular beam; basal tine branching off at a certain distance from the burr forming an acute angle; well-developed middle tine; terminal bifurcation oriented normal to the sagittal plane; cranium with large orbits, preorbital fossae, and ethmoidal vacuities; relatively elongated neurocranium with flat parietals; caudally-oriented pedicles; molarized P2-P3; presence of cingula in upper molars; enlarged i1; un-molarized p4. However, some characters observed in the Pantalla specimens (e.g., rostral edge of the orbit reaching the level of M2; elongated metapodials) do not fit the revised diagnosis of ‘P.’ nestii by Croitor¹². The latter author considers nestii as the earliest species of the genus Cervus based on similarities with the extant red deer especially in cranial morphology^12,22,23. However, in our opinion, his conclusions are biased by relying mostly on the skull IGF 243 of ‘P.’ nestii from Upper Valdarno^6,8, which is heavily deformed and belongs to a juvenile individual (see below for details on ontogenetic variation in ‘Pseudodama’).

A broader look at the entire record of ‘P.’ nestii reveals that this species displays a mosaic of characters between Dama and Cervus, but also that the shared characters with Dama are prevalent (as already pointed out by Azzaroli⁸). The Pantalla sample allows to substantiate these conclusions very well. Our CT-based comparisons between the crania from Pantalla and those of extant red deer and fallow deer (Fig. 3) highlight some morphological similarities with the former, including a relatively longish neurocranium with steep forehead and deep preorbital fossa. On the other hand, ‘P.’ nestii from Pantalla clearly shows Dama-like cranial characters, such as a marked interfrontal crest, horizontal zygomatic arch, high maxilla below the orbit, muzzle more inclined ventrally and less cylindrical in overall shape, sub-horizontal upper cheek tooth row (i.e., the occlusal margin of the row is approximately straight in buccal view), apical surface of the pedicle more inclined dorsocaudally, and overall morphology of the antlers, which in rostral view diverge, rather than converge as in the red deer (Fig. 2).

Likewise, the teeth from Pantalla, have a mixture of Dama and Cervus characters although the former are prevalent. All the premolar characters (the complete absence of a lingual grove on P4, the presence of a cingulum on the distolingual wall of P4, the presence of a small paraconid in p2, the entoconid more aligned with the mesiodistal axis in p3-p4, and a weak mesial cingulum on p4) and most of the lower molar characters are Dama-like. The upper molar features are instead more reminiscent of Cervus being either intermediate between the morphology of the latter and that of extant Dama or even matching Cervus (see Supplementary Table S6 and below).

The postcranial remains from Pantalla appear more similar to Dama than to Cervus. Of the 23 morphological characters by Lister³⁹ which are present in the preserved bones (axis, metacarpal, tibia, astragalus, calcaneum, cubo-navicular, metatarsal, phalanx I, and phalanx II), 21 scores as fallow deer and only two as red deer (details in Supplementary Table S7).

A mixed character suite between Dama and Cervus are revealed also by our palaeoneurological analysis. The brain of ‘P.’ nestii shows Dama-like size and Cervus-like morphology with a prominent cerebellum and a dorsoventrally flattened cerebrum. The latter character is clearly noticeable in ‘Pseudodama’ and Eucladoceros, is less evident in extant Cervus, and is missing in Dama. The hypothesis that depressed and longish cerebra represent a primitive character in Cervini (at least in Pleistocene European forms) is supported by our preliminary data and agree with Azzaroli⁸.

Most interestingly, the two crania from Pantalla actually show some remarkable morphological differences. The neurocranium of 337643 is more lengthened (i.e., more Cervus-like), albeit this shape might be taphonomically modified by the lateral compression of the specimen. This morphology fits that observed in some other ‘P.’ nestii specimens such as IGF 1403 from Olivola (Italy), while the relatively shorter and more rounded neurocranium of 337655 resembles that of other specimens such as IGF 1404 also from Olivola. Moreover, 337643 shows a stronger nuchal crest than 337655. These differences may be related to ontogenesis (see the advanced age of 337643 based on tooth wear). In several cervid species including fallow deer, aging leads to morphological changes in the neurocranium, which tends to elongate and flatten and shows a more developed nuchal crest, probably as a response to the support of larger and heavier antlers^18,38. Similarly, in 337643, the pedicles are apparently closer to one another due to their thickening—an expected condition for an old individual as the distance between the pedicles tends to decrease with age⁸—and markedly shorter than wide. Our comparative data on European Dama-like deer show that the pedicle section can be highly variable both within and between species, although a general trend of laterolateral flattening (i.e., oval shape with major axis oriented anteroposteriorly) can be traced through time (Supplementary Fig. S3), probably as a result of the development of wide, laterally-projecting palmated antlers (in extant deer, D. dama is among those with the heaviest antlers relative to body size^40,41). Therefore, the Pantalla sample on the one hand confirms the variation in cranial morphology already observed for ‘P.’ nestii^6,8, on the other hand it supports the affinities between this species and the fallow deer. The presence of Cervus-like features especially in cranial morphology may be interpreted as plesiomorphic characters which, associated with some characters of the dentition and of the brain, suggest a basal position of ‘Pseudodama’ in the evolutionary history of the Cervini. This hypothesis may be tested in the future through phylogenetic analyses, currently made difficult by the lack of sufficiently well-preserved material of some species of ‘Pseudodama’ (e.g., ‘P.’ lyra, ‘P.’ perolensis).

Compared with other specimens of ‘P.’ nestii^6,8, the sample from Pantalla shows some plesiomorphic characters including a high ratio between the premolar and molar lengths, i.e., 0.77–0.82 (n = 2) for upper teeth (LP/LM) and 0.68–0.69 (n = 3) for lower teeth (Lp/Lm). These values are closer to the basal forms of ‘Pseudodama’, such as ‘P.’ lyra from Montopoli (LP/LM = 0.73, n = 1; Lp/Lm = 0.64, n = 2) and ‘P.’ rhenana from Saint Vallier (LP/LM = 0.75, n = 9; Lp/Lm = 0.68, n = 18; data from Valli⁴²), than to ‘P.’ nestii from Olivola and Upper Valdarno (LP/LM = 0.72, n = 10; Lp/Lm = 0.63, n = 17). Other putatively plesiomorphic features of the sample from Pantalla are all those that approach it morphologically to Cervus (see Supplementary Table S6), i.e., the strong development of lingual conids and stylids in lower molars (Char. 4³⁹) and of buccal cones and styles in upper molars (Char. 1³⁹), the lack of a clear step between 2nd and 3rd lobe of m3 (Char. 11³⁹), the strong lingual cingulum on upper molars (Char. 3³⁹), and the lack of the horizontal turning of the buccal columns of upper molars (Char. 4³⁹—the so-called buccal “cingulum”⁴³). The strong lingual cingulum on upper molars is constantly present in the earliest species of the ‘Pseudodama’ group, ‘P.’ pardinensis⁹, and still present, although extremely rare, in ‘P.’ lyra from Montopoli, ‘P.’ rhenana from Saint Vallier and Senèze, ‘P.’ perolensis from Peyrolles, and ‘P.’ nestii from Olivola. However, this feature is back less rare in ‘P.’ nestii from Upper Valdarno and ‘P.’ farnetensis from Selvella, suggesting a certain polymorphism at this stage. The lack of buccal “cingulum” is a constant in the earliest ‘Pseudodama’ populations (Lower Valdarno, Saint Vallier, Senèze), the buccal “cingulum” appearing, although rare, in ‘P.’ perolensis from Peyrolles and ‘P.’ nestii from Olivola and Upper Valdarno but becoming more common only in later ‘P.’ farnetensis, ‘P.’ vallonnetensis, and constant in Dama.

The above affinities between the Pantalla deer and the early representatives of ‘Pseudodama’ support the idea that the age of the assemblage may be close to the beginning of the Late Villafranchian (ca. 2.1–2.0 Ma), as already suggested based on the occurrence of Leptobos merlai⁴⁴ and a primitive form of Equus stenonis³⁵. Thus, the ‘P.’ nestii sample described herein may represent one of the earliest occurrences of the species in Europe.

Palaeoecological and palaeoethological inferences

The Pantalla sample is also noteworthy as it allows opening a window into the behaviour of these extinct deer. The anomalies found on the two male crania are probably the result of different traumas during their life.

Deer are well known for the intense fights they engage in during the rutting season using their antlers, as a result of an escalation of a broad repertoire of threats and displays⁴⁵. Mineralized antlers are solid structures able to withstand the vehemence of the fight⁴⁶, whereas growing antlers are extremely fragile and any contact with a solid object may result in a serious injury^47,48 that may jeopardize the bearer’s ability to compete with conspecifics and, consequently, its dominance status⁴⁹. Accidents are inevitable in the life of a deer and, in case of the suffered damage not leading to the breakage of the growing beam and consequent loss of its distal part, the antler may continue its growth although, in case of a severe lesion, at a crooked angle⁴⁵. Thus, if the antler was just cracked and the broken part was held together by the velvet and periosteum, with the blood supply still being guaranteed, the damaged beam would just present a conspicuous swelling around the area of fracture (i.e., a fracture callus)^45,50 and a change in the axis of orientation. These features match those seen in the left beam of 337655, which shows a fracture callus between the basal and middle tines corresponding to a change in the orientation of the beam.

The supernumerary tine of the right antler of the same individual can be interpreted as the result of a trauma, too. Considering the delicate nature of the growing antlers and the non-negligible risks of occurrence of an injury, it is safe to believe that the right antler has undergone a light traumatic event (most likely concerning the pedicle) at some early stage of its growth. In fact, it is known that limited injuries could result in the growth of supernumerary tines, even in atypical positions⁵¹, as it has been documented in other deer species (e.g., reindeer⁵², sambar⁵³). It is therefore reasonable to hypothesize that both antler anomalies of 337655 derive from traumas suffered by the deer during the antler growth, when the velvet was still present. It is not known whether the two injuries happened at the same time or in two different events. In fact, it cannot even be said that the two events took place during the same season. While the breakage of the left beam must have occurred in the year of the animal’s death (i.e., during the velvet period preceding the period of hard antler in which the individual died), the development of the supernumerary tine on the right may be the result of a trauma suffered in a previous year. This is due to the fact that when unilateral trauma affects the generative region of the antler (i.e., the pedicle area), abnormalities such as supernumerary tines can reappear in next antler cycles even in more intensified forms⁵⁴, as in the case of 337655 in which the extra-tine is extremely long.

The bone anomaly on the right squamosal of 337643 is also likely the outcome of an injury. Although the external portion was artificially smoothed during the preparation of the specimen, the outer and inner morphology matches that of a callus related to the healing of a major lesion and probable intracranial abscessation. Post-traumatic inflammatory processes are known to cause erosion or pitting of cranial bones in deer⁵⁵ and can be triggered by many factors (e.g., wounds and abrasions of the pedicle⁵⁶), among which violent sexual competition among males with hard antlers is considered one of the most common^55,57. The advanced healing of the injury shown by 337643 suggests that it was not the cause of death, but rather that the individual survived a long time after the trauma albeit with the brain partially compressed by the callus.

The six mandibles recovered at Pantalla, all coming from the same bone accumulation hence reasonably referable to a single deer population, represent several age classes, from calves as young as a few months up to very old individuals (i.e., over 15 years; Supplementary Table S4). Unfortunately, no mandible can be safely associated with the two male crania, although 337631 may belong to the same individual as 337643 based on advanced wear and size. Interestingly, the three most significant cranial remains (crania 337655 and 337643 and frontal bone fragment with basal antler base 337625) belong to adult males, which probably died during the hard antler period (i.e., rutting season: 337655 and 337625) or shortly after (i.e., 337643). The absence of females (at least among the remains with certain sex attribution) contrasts with the population structure in the extant fallow deer, in which females represent on average 75% of the herd⁵⁸. However, the relative abundance of males may increase up to 50% in the rutting season^59,60. Therefore, in spite of the relatively low number of fossils available, based on the age and sex structure of the palaeopopulation and by analogy with the extant fallow deer, the most plausible hypothesis is that the Pantalla deer died during or immediately after the rutting season (Fig. 5).

Source: Ecology - nature.com