Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 Contents lists available at ScienceDirect Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo The role of climate change in the extinction of the last wild equids of Europe: Palaeoecology of Equus ferus and Equus hydruntinus during the Last Glacial Period Flavia Strani a, b, *, Daniel DeMiguel c, a, d a Departamento de Ciencias de La Tierra, Universidad de Zaragoza, 50009 Zaragoza, Spain b Dipartimento di Scienze della Terra, Sapienza - Università di Roma, 00185 Rome, Italy c ARAID Foundation, Universidad de Zaragoza, 50009 Zaragoza, Spain d Institut Català de Paleontologia Miquel Crusafont, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Spain A R T I C L E I N F O A B S T R A C T Keywords: The last European wild equids —Equus ferus and Equus hydruntinus— were among the large mammals (or Palaeoenvironment megafauna) that became extinct during the Late Quaternary Extinction Event disappearing from Europe during Equid Ecology the Holocene. The role that the combined action of the major climatic changes of the Pleistocene/Holocene Dental mesowear Transition and the human activities played in their extinction is not fully understood. The reduction of steppe- Dental microwear Pleistocene like biomes in Europe following the increase of mean global temperatures during the Holocene is usually LGP regarded as the main event that triggered their disappearance, as both equids display typical morphological adaptations for open grasslands and grass eating. However, E. ferus and E. hydruntinus are found, often even co- occurring, in both warm and cold phases of the Late Pleistocene. Thus, the investigation of their niche occupation can help decipher whether their ecology and specialised dietary adaptations were the main reason of their decline following the early Holocene global warming. Here we investigate the feeding strategies of the two equids by studying their long- and short-term dietary adaptations through examination of their patterns of dental wear. As expected, dental mesowear points to a highly abrasive diet concordant with a grazing feeding behaviour for both E. ferus and E. hydruntinus which is consistent with their specialised morphological adaptations for open habitats. In contrast, dental microwear suggests a somewhat degree of plasticity in diets, as both species display microscopic features commonly recorded in modern mixed feeders. Such a flexibility may be the reason for which they could have persisted even when open grasslands were not the dominant landscape. Our findings provide a new line of evidence supporting the idea that human activities (e.g., competition with the first do­ mestic forms brought from Eurasia and Africa) may have played a larger role in the extinction of some mega­ fauna groups than climate change per se. 1. Introduction mammoths) to persist into the Holocene in fragmented or isolated populations (Guthrie, 2004). Among these, wild horses (Equus ferus) and The Late Quaternary Extinction (LQE) event led to the worldwide the hydruntine or European wild ass (Equus hydruntinus) survived until disappearance of several megafaunal mammal species (Koch and the mid-late Holocene when they disappear from both fossil and Barnosky, 2006). Multiple explanations have been proposed, suggesting archeological records (Boulbes and Van Asperen, 2019). The decline and that environmental changes at the end of the last Pleistocene glaciation, eventual disappearance of these last wild equids in Europe is usually human activity, or a combination of both, played a crucial role in this associated with the reduction of steppe biomes due to the development loss of biodiversity (Barnosky et al., 2004; Koch and Barnosky, 2006; of warmer conditions after the Pleistocene/Holocene Transition, as well Faith and Surovell, 2009; Lyons et al., 2016). In the European continent, as to the arrival of domestic forms brought by human populations from this extinction event did not occur uniformly due to the presence of Western Eurasia and Africa (Kimura et al., 2010; Crees and Turvey, several habitat refugia which allowed some large mammals (e.g., woolly 2014; Boulbes and Van Asperen, 2019; Librado et al., 2021). * Corresponding author at: Departamento de Ciencias de La Tierra, Universidad de Zaragoza, 50009 Zaragoza, Spain. E-mail address: flavia.strani@unizar.es (F. Strani). https://doi.org/10.1016/j.palaeo.2023.111564 Received 13 March 2023; Received in revised form 11 April 2023; Accepted 11 April 2023 Available online 13 April 2023 0031-0182/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 According to palaeogenetic data (Bennett et al., 2017), E. hydruntinus preferences and can provide information on fossil taxa niche occupation. became extinct in Europe during the Bronze Age while the tempo and The Late Pleistocene Iberian locality of La Carihuela (Granada, mode of the extinction of E. ferus is still unclear with genetic data sug­ Spain) yielded a rich collection of E. ferus and E. hydruntinus fossil ma­ gesting a substitution by domestic horses (E. caballus) which spread from terial dated during the Last Glacial Period (LGP) which is comprised of Western Eurasian steppes around 2200–2000 BCE (Librado et al., 2021). several dentognathic remains. By analysing dental wear patterns which There are several hypotheses on the primary role of climate change records two different timeframes, it is possible to investigate whether or human activities (hunting, domestication, overkill) or both on the these equids exhibited a pure grazing behaviour or, instead, displayed a extinction of the European megafauna during the late Quaternary more flexible diet which may have allowed them to adapt even to less (Stuart, 2014). Due to morphological adaptations to live in open grass­ favorable landscapes due to climatic shifts. lands and grass eating (i.e., hypsodont dentition, cursorial locomotion), it is assumed that the raise in global mean temperatures after the Last 2. Material and methods Glacial Maximum with the subsequent reduction of dry and cold grasslands, was the main trigger leading to the extinction of wild equids 2.1. Carihuela Cave (Boulbes and Van Asperen, 2019). Both E. hydruntinus and E. ferus are, however, consistently recorded (or even associated) in several Middle Carihuela Cave is one of several caves located 45 km northeast of and Late Pleistocene localities also during warm interglacial phases Granada (Andalusia, southern Spain) in the Píñar river valley (Fig. 1). (Rivals et al., 2009; Boulbes and Van Asperen, 2019; Sánchez-Hernán­ The cave is comprised of six chambers which have been excavated since dez et al., 2020; Mecozzi and Strani, 2022). Understanding the palae­ the 1950s’ (Spahni, 1955) unearthing a rich Late Pleistocene and Ho­ oecological adaptations of these equids and how they exploited locene archaeo-palaeontological collection consisting of lithic artifacts available plant resources emerges, therefore, as crucial to shed light on and fossil vertebrates (also recording human remains belonging to both the role played by climatic and environmental changes in their disap­ Neanderthals and anatomically modern Homo) (García-Sánchez, 1960; pearance in Europe. Vega-Toscano et al., 1988). The cave has been divided into twelve Dietary behaviour of fossil species can be inferred from different Lithostratigraphic Units dated between 117 ± 41 ka and 1250 ± 60 proxies; specifically dental wear patterns provide data on the direct ef­ years (Vega-Toscano et al., 1988; Fernández et al., 2007 and references fects of ingested items (e.g., food, dust, grit) produced over a long therein) which have yielded several mammal remains, among which (mesowear) or short (microwear) period on tooth morphology (Fortelius those of wild horses and hydruntines are especially abundant (Fernán­ and Solounias, 2000; Solounias and Semprebon, 2002). Both dental dez et al., 2007; Samper Carro, 2010; Nacarino-Meneses et al., 2017). meso- and microwear patterns are thus strongly linked to feeding No information is available about the exact stratigraphic position of Fig. 1. La Carihuela geographical location (approximate coordinates: 37◦ 26′ 56” N, 3◦ 25′ 47” W). Map generated from contours.axismaps.com. Contour interval: 500 m. 2 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 the equid fossil material examined in the present study, however, chewing on a “clean” bolus (Ackermans et al., 2018). Fernández et al. (2007) and Samper Carro (2010) report that the ma­ In this study, the traditional mesowear analysis originally limited to terial of E. ferus (=E. caballus cf. germanicus in Fernández et al., 2007) upper second molars (M2) described by Fortelius and Solounias (2000) and E. hydruntinus (Samper Carro, 2010) has been collected from Units was extended to the whole upper molar row (M1-M3) and to the upper X–VI, dated using thermoluminescence approaches approximately be­ fourth premolar (P4) following Kaiser and Solounias (2003) and DeM­ tween 70 and 37 ka (Vega-Toscano et al., 1988; Fernández et al., 2007 iguel et al. (2012). Occlusal relief (high or low) and cusp shape (sharp, and references therein) (Fig. 2). The material is currently stored at the rounded, or blunt) of the apex of the paracone or metacone were Museum of the Institut Català de Paleontologia Miquel Crusafont (ICP, examined and converted to a single mesowear score (MWS) following Sabadell). the “mesowear ruler” developed for scoring dental mesowear on fossil equids by Mihlbachler et al. (2011). The method is based on seven cusp types (numbered from 0 to 6), ranging in shape from high and sharp 2.2. Dental mesowear (stage 0) to completely blunt with no relief (stage 6). Additionally, a “stage 7” is given to teeth with a convex cusp apex. Of a total of 117 For herbivorous ungulates, dental mesowear is a direct signal of a teeth, 85 specimens were scored using this method (Equus ferus N = 72; species’s diet representing the cumulative effects of ingested items on Equus hydruntinus N = 13). Teeth belonging to either juvenile or old the dental surface that are produced over a long period of time compared individuals were excluded from this analysis as mesowear is sensitive to to the lifespan of the animal (years or months; Fortelius and Solounias, the ontogenic age of the animals (Fortelius and Solounias, 2000). Raw 2000; Strani et al., 2018a, 2018b, 2019a). Attrition (tooth-to-tooth mesowear data are provided as Supplementary Information (Table S1). contact) and abrasion (tooth-to-food contact) are the main factors that determine the tooth wear. High levels of attrition produce sharper cusps and a higher dental occlusal relief, while high levels of abrasion lead to 2.3. Dental microwear blunter cusps and low occlusal relief. Modern ungulates which feed on soft plant resources (e.g., leaves, twigs, fruits, buds) usually display Food and other material consumed during the last few days prior to higher levels of attrition while grazing animals whose diet is comprised the death of an animal, leave microscopic patterns on tooth surfaces mostly of grasses, show a higher degree of abrasion due to abrasive silica which are produced during the mastication process, recording the so- contents present in monocotyledonous plants (phytoliths) (Fortelius and called “last supper effect” (Grine, 1986; Strani et al., 2018c). These Solounias, 2000). The effects of other particles ingested while feeding at microwear features were examined following the standard protocol ground level (e.g., dust or grit) may also influence the tooth wear (Kaiser developed by Solounias and Semprebon (2002). Teeth were cleaned et al., 2009) although it has been observed that some ruminants are with acetone and alcohol, moulded with high-resolution silicone capable of “washing away” exogenous particles in the rumen and then (vinylpolysiloxane) and casts were created using clear epoxy resin. Casts Fig. 2. Chronostratigraphy and lithostratigraphy of La Carihuela with correlated vegetation information (modified from Fernández et al., 2007). Highlighted: units where Equus ferus and Equus hydruntinus fossil remains have been recovered (from Fernández et al., 2007 and Samper Carro, 2010). Abbreviations: Lithostratigraphic Units (LU), Pollen Zones (PZ). Silhouette image of E. ferus by Flavia Strani, silhouette image of E. hydruntinus from PhyloPic.org (Public Domain). 3 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 were then examined with a stereomicroscope equipped with a camera (Semprebon and Rivals, 2007). In extant ungulates, grazers have (Leica S9i) under incident light to reveal microfeatures on the enamel. 0.0–22.2% of individuals with scratches between 0 and 17, mixed Specimens displaying badly preserved enamel, taphonomic distortions feeders have 20.9–70.0% of individuals with scratches between 0 and or major post-mortem alterations (e.g., fresh features made during the 17, and leaf-dominated browsers have 72.7–100.0% of individuals with collecting or storing process) were excluded from the analysis following scratches between 0 and 17 (Semprebon and Rivals, 2007). Scratch King et al. (1999) and Weber et al. (2021). Of a total of 117 teeth, 42 textures were also converted into a Scratch Width Score (SWS) to specimens (Equus ferus N = 27; Equus hydruntinus N = 15) were suitable simplify representation of the data: ‘0’ to teeth with predominantly fine for this analysis. scratches per tooth surface, ‘1’ to those with a mixture of fine and coarse Upper molars (M1–M3) and fourth upper premolars (P4) were types of textures, and ‘2’ to those with predominantly coarse scratches. selected for this analysis as they are the ones more involved in the Individual scores for a sample were then averaged to get the SWS and mastication process. Second upper molars (M2) were preferably compared to data of extant taxa (data from Rivals, 2012 and references selected. The anterior lingual blade of the paracone of upper check teeth therein). Raw microwear data are provided as Supplementary Infor­ and the posterior buccal blade of the protoconid of lower molars was mation (Table S2). preferably sampled. If badly preserved, other facets where selected. Following Solounias and Semprebon (2002), microwear features were 2.4. Statistical analysis observed at 35× magnification (Fig. 3), quantified in a standard square area of 0.16 mm2 and divided into five categories: small pits, large pits, Discriminant analyses were performed to examine the resolution of fine scratches, coarse scratches, and gouges and also recording the both mesowear and microwear variables applied to the fossil taxa. For presence of cross scratches. The IC Measure software (ver. 2.0.0.286; the mesowear analysis, the percentages of high relief, rounded and blunt The Imaging Source Europe GmbH) was used to record microwear cusps were used as independent variables and two dietary (conservative features. and radical) classifications were used as grouping variables (Fortelius Mean number of scratches and mean number of pits can be used to and Solounias, 2000). For microwear, 1) average number of pits, 2) discriminate between herbivorous ungulates with a browsing (i.e., ani­ average number of scratches, 3) percentage of individuals with pre­ mals feeding mostly on ligneous plant parts, bushes, leaves, and fruits), dominantly fine scratches, 4) percentage of individuals with predomi­ grazing (i.e., animals feeding mostly on grasses), or mixed (i.e., animals nantly coarse scratches, 5) percentage of individuals with a mixture of feeding on both food types) diet (Solounias and Semprebon, 2002; fine and coarse scratches, 6) percentage of individuals with >4 large DeMiguel et al., 2008; DeMiguel et al., 2010). The percentage of in­ pits, 7) percentage of individuals with >4 cross scratches, and 8) the dividuals with scratch numbers falling in a low scratch range (%0–17) percentage of individuals displaying gouges, were selected as indepen­ has also been used to discriminate browsers, grazers, and mixed feeders dent variables and dietary classifications of modern taxa (modified from Fig. 3. Photomicrographs of enamel surfaces at 35× magnification of selected teeth. Equus ferus: (A), specimen 54, left M1/M2; (B), specimen 77, right P4; and (C), specimen 59, right M1/M2. Equus hydruntinus: (D), specimen 10b, right M2; (E), specimen 11b, left M1/M2; and (F), specimen 7b, left M1/M2. Scale bars: 500 μm. 4 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 Solounias and Semprebon, 2002) were used as grouping variables. An­ specimens of each taxon show <17 scratches (E. ferus %0–17 = 3.7; alyses were performed using IBM SPSS Statistics 24. E. hydruntinus %0–17 = 6.7; Table 1). Scratches are predominantly fine (E. ferus SWS = 0.1; E. hydruntinus SWS = 0.2; Table 1) with similar 3. Results lengths in both taxa (E. hydruntinus Average SL = 121.1 μm; E. ferus Average SL = 126.9 μm; Table S2). Coarse scratches are generally longer 3.1. Dental mesowear than fine ones (Table S2) and E. ferus displays a higher percentage of cross scratches than E. hydruntinus (%XS = 63 and 47 respectively; The Equus ferus mesowear pattern is characterized by a predomi­ Table 1). Pits are generally small (E. ferus Average PS = 59 μm; nance of low occlusal relief (79.2%) and rounded cusps (70.8%), though E. hydruntinus Average PS = 64.3; Table S2) although a higher per­ a few specimens display also sharp (19.4%) or blunt cusps (9.7%) centage of individuals displaying large pits is recorded in E. hydruntinus (Table 1). Equus hydruntinus displays a similar pattern (%L = 69.2; %R = if compared with E. ferus (%LP = 20 and 3.7 respectively; Table 1). 69.2), but contrary to that observed in E. ferus, none of the specimens Gouges are recorded in few individuals (E. ferus %G = 11.1; exhibit blunt cusps and there is a higher incidence of sharp cusps E. hydruntinus %G = 13.3; Table 1). (30.8%) (Table 1). Both patterns result in a relatively high score (MWS While these microwear patterns point to a somewhat high abrasive = 3.5 for E. ferus and 3.0 for E. hydruntinus) (Table 1) which points to a diet, the marked predominance of fine textured scratches is seldom predominance of abrasion over attrition in the wear process that is observed in modern grazing ungulates which instead display a pre­ consistent with grazing diets. Few individuals display a combination of dominance of coarser scratches especially in those species which feed low occlusal relief and sharp cusps (%L + S = 15.3 for E. ferus and 15.4 mostly on C4 grasses (Solounias and Semprebon, 2002). Abundant fine for E. hydruntinus) (Table 1) a mesowear signal correlated with a diet scratches are recorded in some modern mixed feeders which feed on C3 based on grasses but also including a component of open environment grasses such as the Sumatran serow Capricornis sumatraensis, the wapiti browse such as aridity-resistant shrubs (Fortelius and Solounias, 2000). Cervus canadensis, the takin Budorcas taxicolor or in the llama Lama These results are in agreement with the discriminant analysis, which glama who inhabits non-forested areas and feeds in alpine grasslands provides a dietary discrimination of 74.1% of extant taxa correctly (Solounias and Semprebon, 2002). classified according to both classifications (68.5% and 74.1%, respec­ Discriminant analysis performed using microwear variables classifies tively, in cross-validation). Also, bivariate diagrams show that both E. ferus and E. hydruntinus as a seasonal and a meal-by-meal mixed E. ferus and E. hydruntinus are classified as grazers (Table 1; Fig. 4). feeder, respectively (72.2% correctly classified modern taxa, 47.7% in cross-validation; Table 1, Fig. 6). When extant seasonal and non- 3.2. Dental microwear seasonal mixed feeders are grouped together in a single dietary cate­ gory (named as mixed feeder), both fossils are classified as mixed feeders Both E. ferus and E. hydruntinus display a high number of scratches (77.3% correctly classified modern taxa, 61.4% in cross-validation; (E. ferus AS = 29.9; E. hydruntinus AS = 25.9; Table 1), concordant with Table 1, Fig. 6). the consumption of a high amount of abrasive items. The average numbers of scratches and pits are consistent with those observed in extant ungulates with a diet characterized by a high intake of grasses (grazers and grass-dominated mixed feeders) (Fig. 5). Only two Table 1 Summary of dental mesowear and microwear analysis. Mesowear Predicted Diet Taxon N %High %Low %Sharp %Round %Blunt %L + S MWS CONS RAD Equus ferus Mean 72 20.8 79.2 19.4 70.8 9.7 15.3 3.5 Grazer Grazer SD 1.5 CV 0.4 Equus hydruntinus Mean 13 30.8 69.2 30.8 69.2 0.0 15.4 3.0 Grazer Grazer SD 1.6 CV 0.5 Microwear Predicted Diet Taxon N AP AS % %G % % SWS %FS (0 %MS (1 %CS (2 Type Type* LP XS 0–17 SWS) SWS) SWS) Equus ferus Mean 27 12.9 29.9 3.7 11.1 63.0 3.7 0.1 92.6 7.4 0.0 Seasonal Mixed Mixed Feeder Feeder SD 5.8 7.3 CV 0.5 0.2 Equus Mean 15 17.5 25.9 20.0 13.3 46.7 6.7 0.2 80.0 20.0 0.0 Meal-by-Meal Mixed Mixed hydruntinus Feeder Feeder SD 8.6 7.1 CV 0.5 0.3 Abbreviations (mesowear): number of specimens measured (N); percentage of specimens with high (%High) and low (%Low) occlusal relief; percentage of specimens with sharp (%Sharp), rounded (%Rounded) and blunt (%Blunt) cusps; percentage of individuals with low occlusal relief and sharp cusps (%L + S); mesowear score (MWS); standar deviation (SD); coefficient of variation (CV). Predicted diets according to a conservative (CONS) and radical (RAD) mesowear classification. Abbreviations (microwear): average number of pits (AP); average number of scratches (AS); percentage of individuals with >4 large pits (%Lp); percentage of in­ dividuals with gouges (%G); percentage of individuals with >4 cross scratches (%XS); scratches width score (SWS); percentage of specimens with between 0 and 17 scratches (%0–17); percentage of individuals with predominantly fine scratches (%FS); percentage of individuals with a mix of fine and coarse scratches (%MS); percentage of individuals with predominantly coarse scratches (%CS). * Extant seasonal and non-seasonal mixed feeders are grouped together in a single dietary category “mixed feeders”. 5 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 Fig. 4. Bivariate diagrams based on discriminant analysis performed with mesowear data: (A), conservative classification; (B), radical classification. Group centroids: browsers (0); mixed feeders (1); and grazers (2). Extant ungulates data from Fortelius and Solounias (2000). Fig. 5. Bivariate plot of the average number of pits versus average number of scratches in extant ungulates (data modified from Solounias and Semprebon, 2002) and fossil equids from La Carihuela. Silhouette image of E. ferus by Flavia Strani, silhouette image of E. hydruntinus from PhyloPic.org (Public Domain). 6 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 Fig. 6. Bivariate diagrams based on discriminant analysis performed with microwear data. (A), classification using extant ungulates dietary types as grouping variables: fruit browsers (0); leaf browsers (1); seasonal mixed feeders (2); meal-by-meal mixed feeders (3); and grazers (4). (B), classification grouping extant seasonal and non-seasonal mixed feeders in a single dietary category (mixed feeders): fruit browsers (0); leaf browsers (1); mixed feeders (2); and grazers (3). Square symbol indicates group centroids. Extant ungulates data from Solounias and Semprebon (2002). Table 2 Summary table of dietary adaptations of E. ferus and E. hydruntinus from selected Late Pleistocene European localities. Taxa Locality Country Diet Mesowear Reference Microwear SWS Reference Equus ferus Gough’s Cave UK grazer Saarinen et al. (2016) grazer Rivals and Lister (2016) Equus ferus Kent’s Cavern UK grazer Saarinen et al. (2016) grazer Rivals and Lister (2016) Equus ferus La Carihuela Spain grazer this study mixed feeder 0.1 this study Equus La Carihuela Spain grazer this study mixed feeder 0.2 this study hydruntinus Equus ferus Canyars Spain grazer Rivals et al. (2017) grazer 0.92 Rivals et al. (2017) Equus ferus Arbreda I Spain grazer Sánchez-Hernández et al. grazer 0.5 Sánchez-Hernández et al. (2020) (2020) Equus ferus Teixoneres Cave IIIa Spain grazer Sánchez-Hernández et al. grazer 0.55 Sánchez-Hernández et al. (2020) (2020) Equus ferus Teixoneres Cave IIIb Spain grazer Sánchez-Hernández et al. grazer 0.55 Sánchez-Hernández et al. (2020) (2020) Equus Teixoneres Cave IIIa Spain grazer Sánchez-Hernández et al. grazer 0.63 Sánchez-Hernández et al. hydruntinus (2020) (2020) Equus Teixoneres Cave IIIb Spain grazer Sánchez-Hernández et al. mixed feeder 0.64 Sánchez-Hernández et al. hydruntinus (2020) (2020) Equus ferus Salzgitter Lebenstedt Germany grass dom-mixed Rivals et al. (2009) browser 1.1 Rivals et al. (2009) feeder Equus ferus Abric Romanı KLM Spain grazer Rivals et al. (2009) grazer 1.1–2.0- Rivals et al. (2009) 0.0 Equus Emine Bair Khosar Ukraine grazer Cirilli et al. (2022) hydruntinus Cave Equus ferus Aufhausener höhle Germany grazer Saarinen et al. (2016) Equus ferus Portel-Ouest F France grazer Rivals et al. (2009) mixed feeder 1.1 Rivals et al. (2009) Equus Kabazi II Ukraine grazer Ramírez-Pedraza et al. grazer 1.2 Ramírez-Pedraza et al. hydruntinus (2020) (2020) Equus ferus Villa Seckendorff Germany grazer Saarinen et al. (2016) Equus Villa Seckendorff Germany grazer Cirilli et al. (2022) hydruntinus Equus ferus Wallertheim F Germany grass dom-mixed Rivals et al. (2009) grass dom-mixed 0.5 Rivals et al. (2009) feeder feeder Equus ferus Payre D France grazer Rivals et al. (2009) mixed feeder 1.1 Rivals et al. (2009) Equus ferus Taubach Germany grazer Saarinen et al. (2016) browser 1 Rivals et al. (2009) 7 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 4. Discussion employed by sympatric E. ferus and E. hydruntinus populations at least during the most arid phases of the LGP, feeding on the same resources at 4.1. Diet of E. ferus and E. hydruntinus from La Carihuela and different times of the day and with E. hydruntinus exploiting a larger area comparison with species from other Late Pleistocene sites in Europe compared to E. ferus. Both E. ferus and E. hydruntinus from La Carihuela display a long-term 4.2. The role of climatic changes and habitat alterations in the extinction grazing behaviour, in agreement with that observed in Equus pop­ of wild equids in Europe ulations from other Late Pleistocene localities across (Southern, Western and Central) Europe (from MIS 5 to MIS 2) (Rivals et al., 2009; Sánchez- The reduction of steppe-like biomes in Europe following the warm­ Hernández et al., 2020; Cirilli et al., 2022) (Table 2), and with their well- ing phase at the beginning of the Holocene is considered the main factor known (morphological) adaptations (such as cursorial limbs, long met­ that led to the extinction of E. ferus and E. hydruntinus in the continent apodials, hypsodont cheek teeth, lophodont occlusal patterns, etc.) for (Boulbes and Van Asperen, 2019). Results from the La Carihuela sample open habitats (MacFadden, 2005; Cirilli et al., 2022). as well as from other European localities, however, indicate that despite Regarding short-term feeding behaviour, apparently both equids the fact that wild equids possessed marked adaptations for open grass­ were able to also exploit plant resources other than grasses. While the lands feeding primarily on grass, they could adapt to feed also on other high frequency of scratches and the low number of pits are consistent types of plant resources displaying a certain dietary flexibility whether with a diet characterized by high levels of abrasion, the predominance of or not cold, dry, open habitats were the predominant element of the finely textured features are unusual for a strict grazing behaviour. European landscapes. E. ferus even displayed a short-term browsing Modern grazers (as well as bark eaters and fruit browsers) tend in fact to behaviour during MIS 5e (Rivals et al., 2009), one of the warmest in­ display coarse scratches (Solounias and Semprebon, 2002), while fine terglacials of the Pleistocene characterized by global mean temperatures scratches are commonly observed in some extant ungulates displaying comparable to projections for the current climate change, and wetter intermediate diets (Solounias and Semprebon, 2002), such as the four- conditions than during the Holocene in Central Europe (Rohling et al., horned antelope Tetracerus quadricornis (considered a grazer in Solo­ 2007; Dabkowski and Limondin-Lozouet, 2021). unias and Semprebon, 2002, although dietary studies by Kunwar et al., Pollen records for the Lithostratigraphic Units X–VI of La Carihuela 2016 on living populations point to a mixed diet). which yielded E. ferus and E. hydruntinus fossil remains, point to an Interestingly, this finding is not recorded in E. ferus and alternation between pine forests (Pinus-dominated open or closed E. hydruntinus from other Late Pleistocene Iberian localities, which woodland) and grasslands (Fernández et al., 2007) (Fig. 2). The absence mostly display a SWS value above 0.5 (Rivals et al., 2009; Sánchez- of oaks (Quercus) and deciduous trees across Units X–VI indicate Hernández et al., 2020). A less abrasive short-term feeding behaviour in generally cold climatic settings although the record of meso­ E. ferus is recorded in other Late Pleistocene European localities, where thermophytes in Unit VI suggests warmer conditions around 40 ka microwear patterns points to a mixed (at Portel-Ouest and Payre in (Fernández et al., 2007). The lack of information on the exact strati­ France and Wallertheim in Germany during MIS 5e-d and MIS 4) or even graphic provenance of the examined sample does not allow us to browsing (at Taubach and Salzgitter Lebenstedt in Germany during MIS correlate the studied material with a specific vegetation phase, however 5e and MIS 3) diet (Rivals et al., 2009) (Table 2). Individuals ascribed to it appears that E. ferus and E. hydruntinus from La Carihuela could suc­ E. cf. ferus from another Iberian locality, the Middle Pleistocene Vall­ cessfully feed on the same resources and avoid competitive displacement paradìs Estació layer EVT3, also display microwear patterns which are despite these repeated fluctuations between forest- and grassland- consistent with a mixed feeing behaviour (Strani et al., 2019b), although dominated landscapes. It is possible that the mixed feeding short-term this interpretation should be tentatively taken as the sample consists of diet recorded is due to a sample consisting of specimens that may only two specimens. The available data for E. hydruntinus from other come from different units and thus lived under different climatic European sites suggests that this equid usually displayed a grazing conditions. feeding behaviour although microwear patterns consistent with a less Nevertheless, obtained results suggest that Late Pleistocene wild abrasive diet are recorded at Teixoneres Cave during the MIS 3 equids could in fact adapt to warmer and more humid conditions, and (Sánchez-Hernández et al., 2020). that while environmental changes certainly affected their distribution The discrepancy between mesowear and microwear results indicates “setting the stage” for their eventual extinction (e.g., locking them in that while both wild horses and hydruntines were well adapted to feed habitat refugia such as the Iberian Peninsula; Schmitt and Varga, 2012), on grasses, they could have displayed also a certain degree of flexibility probably the arrival of and competition with domestic forms and other in their dietary behaviour when necessary, in order to consume less livestock, may have actually triggered and led to the disappearance of abrasive plant resources. wild equids in Europe as observed between modern Przewalski’s horses Finely textured scratches have been observed by Solounias and and free-ranging domestic ungulates (Sietses et al., 2009). Semprebon (2002) in taxa which feed on high altitude habitats where C3 grasses grow or which eat less abrasive fresh or wet grasses. La Carihuela 5. Conclusions cave is located at 1020 m above sea level (Fernández et al., 2007) close to the Baetic System mountain ranges of Sierra Nevada, Sierra de Baza The role that major climatic changes and human activity played in and Sierra de María (Fig. 1). Thus, the abundance of fine scratches may the extinction of the last European wild equids during the Pleistocene/ be also a result of the fact that both equids fed on C3 grasses in moun­ Holocene Transition is not well understood. We report for the first time a tainous areas. dietary reconstruction of E. ferus and E. hydruntinus from the Late Among modern wild (or feral) equids, Przewalski horses (Equus Pleistocene locality of La Carihuela (Iberian Peninsula), providing in­ przewalskii) and kulans (Equus hemionus kulan) from Central Eurasia live formation on the niche partitioning that these equids exhibited in in steppe-biomes in habitat and climatic conditions that are comparable southern Europe during the LGP. Dental mesowear and microwear to those recorded in most European regions during the glacial phases of indicate that both wild horses and hydruntines occupied the same niche the LGP. In Przewalski horses and kulan populations with overlapping sharing similar long- and short-term adaptations as grazers that could home ranges, it has been observed that time-differentiation of area usage periodically shift their diet to also include less abrasive plant resources. allowed the two species to live in sympatric conditions, with major This, albeit limited in time, dietary flexibility is also recorded in other overlapping (>60% of shared home range) occurring only during spring Late Pleistocene European localities suggesting that both wild equids and kulans showing higher mobility than Przewalski horses (Bahloul were capable of adapting to the reduction of steppe-like biomes. Holo­ et al., 2001). Similar niche partitioning mechanisms may had been cene climatic changes may thus have not been the main trigger of their 8 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 disappearance. Human activity, specifically the spreading of domestic DeMiguel, D., Azanza, B., Morales, J., 2010. Trophic flexibility within the oldest Cervidae lineage to persist through the Miocene Climatic Optimum. Palaeogeogr. horses and donkeys from Eurasia may have played a larger role in the Palaeoclimatol. Palaeoecol. 289, 81–92. https://doi.org/10.1016/j. extinction of wild equids in the continent competing with them for the palaeo.2010.02.010. same resources. We also highlight the importance of a comprehensive DeMiguel, D., Quiralte, V., Azanza, B., Montoya, P., Morales, J., 2012. Dietary behaviour microwear analysis when reconstructing the diet of fossil ungulates and competition for vegetal resources in two early Miocene pecoran ruminants from Central Spain. Geodiversitas 34, 425–443. https://doi.org/10.5252/g2012n2a10. including both quantitative (number of features) and qualitative (fea­ Faith, J.T., Surovell, T.A., 2009. Synchronous extinction of North America’s Pleistocene ture’s texture) information to discriminate between feeding behaviours. mammals. Proc. Natl. Acad. Sci. U. S. A. 106, 20641–20645. Supplementary data to this article can be found online at https://doi. Fernández, S., Fuentes, N., Carrión, J.S., González-Sampériz, P., Montoya, E., Gil, G., Vega-Toscano, G., Riquelme, J.A., 2007. The holocene and Upper Pleistocene pollen org/10.1016/j.palaeo.2023.111564. sequence of Carihuela Cave, Southern Spain. Geobios 40 (1), 75–90. García-Sánchez, M., 1960. Restos humanos del Paleolítico Medio y Superior y del Neo- Eneolítico de Píñar (Granada). Trabajos Inst. Bernardino Sahagún Antropol. Etnol. Declaration of Competing Interest 15, 17–72. Grine, F.E., 1986. Dental evidence for dietary differences in Australopithecus and Paranthropus: a quantitative analysis of permanent molar microwear. J. Hum. Evol. The authors declare the following financial interests/personal re­ 15, 783–822. lationships which may be considered as potential competing interests: Kaiser, T.M., Solounias, N., 2003. Extending the tooth mesowear method to extinct and extant equids. Geodiversitas 25 (2), 321–345. Flavia Strani reports financial support was provided by Italian Guthrie, R., 2004. Radiocarbon evidence of mid-Holocene mammoths stranded on an Paleontological Society. The corresponding author (FS) and co-author alaskan Bering Sea island. Nature 429, 746–749. https://doi.org/10.1038/ (DDM) previously served as Reviewers for Palaeogeography, Palae­ nature02612. Kaiser, T.M., Brasch, J., Castell, J.C., Schulz, E., Clauss, M., 2009. Tooth wear in captive oclimatology, Palaeoecology. wild ruminant species differs from that of free-ranging conspecifics. Mamm. Biol. 74 (6), 425–437. Data availability Kimura, B., Marshall, F.B., Chen, S., Rosenbom, S., Moehlman, P.D., Tuross, N., Sabin, R. C., Peters, J., Barich, B., Yohannes, H., Kebede, F., Teclai, R., Beja-Pereira, A., Mulligan, C.J., 2010. Ancient DNA from Nubian and Somali wild ass provides Raw data are provided as Supplementary Material insights into donkey ancestry and domestication. Proc. Royal Soc. B. 278, 50–57. https://doi.org/10.1098/rspb.2010.0708. King, T., Andrews, P., Boz, B., 1999. Effect of taphonomic processes on dental microwear. Acknowledgments Am. J. Phys. Anthropol. 108, 359–373. Koch, P., Barnosky, A., 2006. Late Quaternary extinctions: state of the debate. Annu. Rev. This research was funded by “Borsa di Studio SPI 2020” granted by Ecol. Evol. Syst. 37, 215–250. Kunwar, A., Gaire, R., Pokharel, K.P., Baral, S., Thapa, T.B., 2016. Diet of the Four- the Italian Paleontological Society to F.S. This work is part of project horned Antelope Tetracerus quadricornis (De Blainville, 1816) in the Churia Hills of PID2020-116220GB-I00 funded by MCIN/AEI/10.13039/ Nepal. J. Threat. Taxa. 8, 8745. https://doi.org/10.11609/jott.1818.8.5.8745-8755. 501100011033/, with funding also by the Government of Aragon (ref. Fortelius, M., Solounias, N., 2000. Functional characterization of ungulate molars using the abrasion-attrition wear gradient: a new method for reconstructing paleodiets. E33_20R). F.S. is a “Juan de la Cierva Formación” fellow (ref. FJC2020- Am. Mus. Novit. 3301, 1–35. 042982-I) funded by the Agencia Estatal de Investigación of the Spanish Librado, P., Khan, N., Fages, A., Kusliy, M.A., Suchan, T., Tonasso-Calvière, L., Ministry of Science and Innovation (MCIN / AEI / 10.13039 / Schiavinato, S., Alioglu, D., Fromentier, A., Perdereau, A., Aury, J.-M., Gaunitz, C., Chauvey, L., Seguin-Orlando, A., Der Sarkissian, C., Southon, J., Shapiro, B., 501100011033) and the European Union “NextGenerationEU / PRTR” Tishkin, A.A., Kovalev, A.A., Alquraishi, S., 2021. The origins and spread of domestic program. Authors thank Editor-in-Chief H.J. Falcon-Lang and two horses from the Western Eurasian steppes. Nature 598, 634–640. https://doi.org/ anonymous reviewers for their comments that greatly improved the 10.1038/s41586-021-04018-9. quality of the manuscript. Lyons, S.K., Miller, J.H., Fraser, D., Smith, F.A., Boyer, A., Lindsey, E., Mychajliw, A.M., 2016. The changing role of mammal life histories in Late Quaternary extinction vulnerability on continents and islands. Biol. Lett. 12. References MacFadden, B.J., 2005. Fossil Horses-evidence for Evolution. Science 307, 1728–1730. https://doi.org/10.1126/science.1105458. Mecozzi, B., Strani, F., 2022. Equids from the late middle pleistocene to early holocene of Ackermans, N., Winkler, D., Schulz-Kornas, E., Kaiser, T., Müller, D., Kircher, P., the Apulia Peninsula (southern Italy): Reassessment of their taxonomy and Hummel, J., Clauss, M., Hatt, J., 2018. Controlled feeding experiments with diets of biochronology. Geodiversitas 44 (2). different abrasiveness reveal slow development of mesowear signal in goats (Capra Mihlbachler, M., Rivals, F., Solounias, N., Semprebon, G., 2011. Dietary change and aegagrus hircus). J. Exp. Biol. 221 (21). evolution of horses in North America. Science 331 (6021), 1178–1181. Bahloul, K., Pereladova, O.B., Soldatova, N., Fisenko, G., Sidorenko, E., Sempéré, A.J., Nacarino-Meneses, C., Jordana, X., Orlandi-Oliveras, G., Köhler, M., 2017. 2001. Social organization and dispersion of introduced kulans (Equus hemionus Reconstructing molar growth from enamel histology in extant and extinct Equus. Sci. kulan) and Przewalski horses (Equus przewalski) in the Bukhara Reserve, Rep. 7 (1). Uzbekistan. J. Arid Environ. 47, 309–323. https://doi.org/10.1006/jare.2000.0714. Ramírez-Pedraza, I., Rivals, F., Uthmeier, T., Chabai, V., 2020. Palaeoenvironmental and Barnosky, A., Koch, P.L., Feranec, R.S., Wing, S.L., Shabel, A.B., 2004. Assessing the seasonal context of the late Middle and early Upper Palaeolithic occupations in causes of late Pleistocene extinctions on the continents. Science 306, 70–75. Crimea: an approach using dental wear patterns in ungulates. Archaeol. Anthropol. Bennett, E.A., Champlot, S., Peters, J., Arbuckle, B.S., Guimaraes, S., Pruvost, M., Bar- Sci. 12 https://doi.org/10.1007/s12520-020-01217-9. David, S., Davis, S.J.M., Gautier, M., Kaczensky, P., Kuehn, R., Mashkour, M., Rivals, F., Schulz, E., Kaiser, T.M., 2009. Late and middle Pleistocene ungulates dietary Morales-Muñiz, A., Pucher, E., Tournepiche, J.-F., Uerpmann, H.-P., Bălăşescu, A., diversity in Western Europe indicate variations of Neanderthal paleoenvironments Germonpré, M., Gündem, C.Y., Hemami, M.-R., 2017. Taming the late Quaternary through time and space. Quat. Sci. Rev. 28, 3388–3400. https://doi.org/10.1016/j. phylogeography of the Eurasiatic wild ass through ancient and modern DNA. PLOS quascirev.2009.09.004. ONE 12, e0174216. https://doi.org/10.1371/journal.pone.0174216. Rivals, F., 2012. Ungulate feeding ecology and middle Pleistocene paleoenvironments at Boulbes, N., Van Asperen, E.N., 2019. Biostratigraphy and palaeoecology of European Hundsheim and Deutsch-Altenburg 1 (eastern Austria). Palaeogeogr. Palaeoclimatol. Equus. Front. Ecol. Evol. 7, 301. Palaeoecol. 317–318, 27–31. Cirilli, O., Machado, H., Arroyo-Cabrales, J., Barrón-Ortiz, C.I., Davis, E., Jass, C.N., Rivals, F., Lister, A.M., 2016. Dietary flexibility and niche partitioning of large Jukar, A.M., Landry, Z., Marín-Leyva, A.H., Pandolfi, L., Pushkina, D., Rook, L., herbivores through the Pleistocene of Britain. Quat. Sci. Rev. 146, 116–133. Saarinen, J., Scott, E., Semprebon, G., Strani, F., Villavicencio, N.A., Kaya, F., Rivals, F., Uzunidis, A., Sanz, M., Daura, J., 2017. Faunal dietary response to the Bernor, R.L., 2022. Evolution of the Family Equidae, Subfamily Equinae, in North, Heinrich Event 4 in southwestern Europe. Palaeogeogr. Palaeoclimatol. Palaeoecol. Central and South America, Eurasia and Africa during the Plio-Pleistocene. Biology 473, 123–130. https://doi.org/10.1016/j.palaeo.2017.02.033. 11, 1258. https://doi.org/10.3390/biology11091258. Rohling, E.J., Grant, K., Hemleben, Ch., Siddall, M., Hoogakker, B.A.A., Bolshaw, M., Crees, J.J., Turvey, S.T., 2014. Holocene extinction dynamics of Equus hydruntinus, a Kucera, M., 2007. High rates of sea-level rise during the last interglacial period. Nat. late-surviving european megafaunal mammal. Quat. Sci. Rev. 91, 16–29. https://doi. Geosci. 1, 38–42. https://doi.org/10.1038/ngeo.2007.28. org/10.1016/j.quascirev.2014.03.003. Samper Carro, S.C., 2010. Caracterización sistemática de los équidos de La Carihuela Dabkowski, J., Limondin-Lozouet, N., 2021. Comparison of temperature and humidity (Piñar, Granada). In: Moreno-Azanza, M., Díaz-Martínez, I., Gasca, J.M., Melero- during MIS 11 and MIS 5e interglacials with the Holocene using stable isotopes in Rubio, M., Rabal-Garcés, R., Sauqué, V. (Eds.), Cidaris, número 30, VIII Encuentro de tufa deposits from northern France. Quat. Res. 107, 147–158. https://doi.org/ Jóvenes Investigadores en Paleontología, volumen de actas, pp. 283–291. 10.1017/qua.2021.66. Sánchez-Hernández, C., Gourichon, L., Soler, J., Soler, N., Blasco, R., Rosell, J., Rivals, F., DeMiguel, D., Fortelius, M., Azanza, B., Morales, J., 2008. Ancestral feeding state of 2020. Dietary traits of ungulates in northeastern Iberian Peninsula: did these ruminants reconsidered: earliest grazing adaptation claims a mixed condition for Cervidae. BMC Evol. Biol. 8, 13. https://doi.org/10.1186/1471-2148-8-13. 9 F. Strani and D. DeMiguel Palaeogeography, Palaeoclimatology, Palaeoecology 620 (2023) 111564 Neanderthal preys show adaptive behaviour to local habitats during the Middle Calabrian transition in Central Italy. Palaeogeogr. Palaeoclimatol. Palaeoecol. 499, Palaeolithic? Quat. Int. 557, 47–62. https://doi.org/10.1016/j.quaint.2020.01.008. 102–111. https://doi.org/10.1016/j.palaeo.2018.03.021. Saarinen, J., Eronen, J., Fortelius, M., Seppä, H., Lister, A.M., 2016. Patterns of diet and Strani, F., Profico, A., Manzi, G., Pushkina, D., Raia, P., Sardella, R., DeMiguel, D., 2018c. body mass of large ungulates from the Pleistocene of Western Europe, and their MicroWeaR: a new R package for dental microwear analysis. Ecol. Evol. 8, relation to vegetation. Palaeontol. Electron. 19.3.32A, 1–58. 7022–7030. https://doi.org/10.1002/ece3.4222. Schmitt, T., Varga, Z., 2012. Extra-Mediterranean refugia: the rule and not the Strani, F., Pushkina, D., Bocherens, H., Bellucci, L., Sardella, R., DeMiguel, D., 2019a. exception? Front. Zool. 9, 22. https://doi.org/10.1186/1742-9994-9-22. Dietary adaptations of early and Middle Pleistocene equids from the Anagni Basin Semprebon, G.M., Rivals, F., 2007. Was grass more prevalent in the pronghorn past? An (Frosinone, Central Italy). Front. Ecol. Evol. 7 https://doi.org/10.3389/ assessment of the dietary adaptations of Miocene to recent Antilocapridae fevo.2019.00176. (Mammalia: Artiodactyla). Palaeogeogr. Palaeoclimatol. Palaeoecol. 253, 332–347. Strani, F., DeMiguel, D., Alba, D.M., Moyà-Solà, S., Bellucci, L., Sardella, R., Madurell- Sietses, D.J., Faupin, G., de Boer, W.F., de Jong, C.B., Henkens, R.J.H.G., Usukhjargal, D., Malapeira, J., 2019b. The effects of the “0.9 Ma event” on the Mediterranean Batbaatar, T., 2009. Resource partitioning between large herbivores in Hustai ecosystems during the Early-Middle Pleistocene transition as revealed by dental National Park, Mongolia. Mamm. Biol. 74, 381–393. https://doi.org/10.1016/j. wear patterns of fossil ungulates. Quat. Sci. Rev. 210, 80–89. https://doi.org/ mambio.2009.04.001. 10.1016/j.quascirev.2019.02.027. Solounias, N., Semprebon, G., 2002. Advances in the reconstruction of ungulate Stuart, A.J., 2014. Late Quaternary megafaunal extinctions on the continents: a short ecomorphology with application to early fossil equids. Am. Mus. Novit. 3366, 1–49. review. Geol. J. 50, 338–363. https://doi.org/10.1002/gj.2633. Spahni, J.C., 1955. Vestiges néanderthaliens de Píñar. Bull. l’Assoc. Française L’étude Q. Vega-Toscano, L.G., Hoyos, M., Ruiz-Bustos, A., Laville, H., 1988. La séquence de la 52, 540–548. grotte de la Carihuela (Píñar, Grenade): chronostratigraphie et paléoécologie du Strani, F., DeMiguel, D., Sardella, R., Bellucci, L., 2018a. Resource and niche Pléistocene Supèrieur au Sud de la Péninsule Ibérique. In: Otte, M. (Ed.), L’Homme differentiation mechanisms by sympatric early Pleistocene ungulates: the case study de Néandertal, L’Environnement, vol. 2. ERAUL, Liège, pp. 169–180. of Coste San Giacomo. Quat. Int. 481, 157–163. https://doi.org/10.1016/j. Weber, K., Winkler, D.E., Schulz-Kornas, E., Kaiser, T.M., Tütken, T., 2021. The good, the quaint.2017.08.064. bad and the ugly – a visual guide for common post-mortem wear patterns in Strani, F., DeMiguel, D., Bellucci, L., Sardella, R., 2018b. Dietary response of early vertebrate teeth. Palaeogeogr. Palaeoclimatol. Palaeoecol. 578, 110577. Pleistocene ungulate communities to the climate oscillations of the Gelasian/ 10