This is the accepted version of the article: Nacarino Meneses, Carmen; Orlandi Oliveras, Guillem. «The life history of Eu- ropean Middle Pleistocene equids : first insights from bone histology». Historical Biology, Vol. 33, Issue 5 (2021), p. 672-682. DOI 10.1080/08912963.2019.1655011 This version is available at https://ddd.uab.cat/record/239869 under the terms of the license 1 The life history of European Middle Pleistocene equids: first insights 2 from bone histology 3 Carmen Nacarino-Meneses1,2* and Guillem Orlandi-Oliveras1 1 4 Evolutionary paleobiology department, Institut Català de Paleontologia Miquel 5 Crusafont (ICP), Campus de la Universitat Autònoma de Barcelona, Bellaterra, 6 Barcelona, Spain. 2 7 Department of Biological Sciences, University of Cape Town, Private Bag X3, Rhodes 8 Gift, 7700 South Africa 9 *Corresponding author: Dr. Carmen Nacarino-Meneses 10 Address: University of Cape Town, Department of Biological Sciences, Private Bag X3, 11 Rhodes Gift, 7700 South Africa 12 Email: carmen.nacarino@gmail.com 13 ORCiD: orcid.org/0000-0003-2123-8758 14 15 1 16 Abstract 17 Evolutionary trends in body size are a central issue of study in Paleontology. 18 However, and despite body size being one of the most important life history traits of an 19 animal, iconic size-decrease trends such as the one experienced by Equus during the 20 European Pleistocene has never been analysed under a life history framework. Here, we 21 studied the metapodial bone histology of two large Middle Pleistocene species (Equus 22 mosbachensis and Equus steinheimensis) to reconstruct key life history traits that 23 correlate with body size (e.g. size at birth, growth rate), and compare them with that of 24 smaller extant Equus (Equus grevyi, Equus quagga, Equus zebra and Equus hemionus). 25 Our results show that neonatal size of these Middle Pleistocene equids fits predictions 26 from body mass scaling. We estimate a similar age of epiphyseal fusion for the 27 metapodia of E. mosbachensis and E. steinheimensis and that of extant equids. Our 28 findings also reveal that extinct equids grew at higher rates than extant Equus. This 29 result conforms to the predictions of LH theory on environments with different levels of 30 resource availability and provides a new framework of study for body size shifts on 31 European Pleistocene equids. 32 Keywords: Bone histology, body size, life history, growth rate, Pleistocene, Equus 2 33 Introduction 34 The family Equidae is of special relevance for Palaeontology and evolution 35 (MacFadden 1992; Janis 2007). The abundance of equid remains in the fossil record has 36 facilitated the tracking of evolutionary changes in the clade (Orlando 2015), positioning 37 Equidae as a classic example of macroevolution (MacFadden 2005; Janis 2007). Along 38 with modifications in dentition and number of digits (Simpson 1953; Azzaroli 1992; 39 MacFadden 1992; Strömberg 2006; Janis 2007; Cantalapiedra et al. 2017; McHorse et 40 al. 2017), this mammalian group experienced significant variations in size during its 41 evolutionary history (Simpson 1953; MacFadden 1986; Eisenmann 1991; Forsten 42 1991a; MacFadden 1992; Alberdi et al. 1995; Guthrie 2003; Ortiz-Jaureguizar and 43 Alberdi 2003; Shoemaker and Clauset 2014; Cantalapiedra et al. 2017). Generally, a 44 phylogenetical large-scale trend towards ever increasing body size is observed (Cope’s 45 law, Stanley (1973)), although several clades of the family experienced size reduction 46 (MacFadden 1986; Forsten 1991a; Forsten 1991b; MacFadden 1992; Alberdi et al. 47 1995; Alberdi et al. 1998). One of these dwarfing tendencies occurred during the 48 evolution of the genus Equus in the Old World (Forsten 1991b; Alberdi et al. 1995). A 49 size shift towards a smaller body size is described for the first monodactyl horses 50 (stenonoid horses) (Alberdi et al. 1995; Alberdi et al. 1998) that arrived in Eurasia 51 between 3.0 – 2.5 Ma ago (Lindsay et al. 1980; Azzaroli 1983) (Fig. 1). A second 52 migratory event (1 – 0.8 Ma ago) introduced the equid caballoid forms in Europe 53 (Alberdi and Bonadonna 1988), which also experienced size reduction during their 54 Pleistocene evolution in this continent (Forsten 1991b; Alberdi et al. 1995) (Fig. 1). The 55 size decrease trend continued throughout the Holocene (Forsten 1988; Forsten 1991b), 56 resulting in a smaller body size of extant Equus (Ernest 2003) in comparison to their 57 Pleistocene relatives (Alberdi et al. 1995) (Fig. 1). 58 Body size is one of the most important characteristics of an animal, as it tightly 59 correlates with its physiology (Kleiber 1932; Peters 1983; McNab 1990), life history 60 (Blueweiss et al. 1978; Western 1979; Calder 1984) and ecology (Damuth 1981; Peters 61 1983; Eisenberg 1990). Due to the close relationship between body size and the 62 ecological conditions of the ecosystem, the size shift observed in Pleistocene equids has 63 usually been related to the climatic and resource variations occurring during that epoch 64 (Forsten 1991b; Alberdi et al. 1995; Cantalapiedra et al. 2017). Demographic 65 (population density) and behavioural (social structure) characteristics of the species 3 66 have also been proposed to explain the body size differences observed between 67 European Pleistocene horses (Saarinen et al. 2016). Body size is also a key life history 68 trait that correlates with other biological traits (Peters 1983; Calder 1984; Damuth and 69 MacFadden 1990), but no previous research has investigated the body size variations 70 observed in Equus within a life history framework. In fossil mammals, the histological 71 analysis of bones is known to provide valuable insights into their life history strategy 72 (Köhler and Moyà-Solà 2009; Köhler 2010; Marín-Moratalla et al. 2011; Martínez- 73 Maza et al. 2014; Amson et al. 2015; Kolb et al. 2015; Moncunill-Solé et al. 2016; 74 Orlandi-Oliveras et al. 2016; Orlandi-Oliveras et al. 2018). The kind of bone tissue type, 75 its vascularization, and the bone growth marks (BGMs) present within a bone cortex, 76 record the pace of growth and development of the species (Chinsamy-Turan 2005; 77 Huttenlocker et al. 2013; Lee et al. 2013; Woodward et al. 2013). The detailed study of 78 these histological features, hence, allows the inference of the key life history longevity 79 (Köhler 2010; Marín-Moratalla et al. 2011; Martínez-Maza et al. 2014; Moncunill-Solé 80 et al. 2016; Orlandi-Oliveras et al. 2016), growth rate (Padian et al. 2001; Köhler and 81 Moyà-Solà 2009; Cubo et al. 2012; Amson et al. 2015; Orlandi-Oliveras et al. 2018) 82 and age at maturity (Köhler and Moyà-Solà 2009; Marín-Moratalla et al. 2011; 83 Woodward et al. 2015; Jordana et al. 2016; Nacarino-Meneses et al. 2016; Orlandi- 84 Oliveras et al. 2018; Calderón et al. 2019) in extant and extinct vertebrates. 85 In the present research, we aim to investigate whether the body size shift 86 observed in the Equus lineage is related to changes in the life history of the species. We 87 focus our study in the European Middle Pleistocene species Equus steinheimensis and 88 Equus mosbachensis. Both extinct caballoid horses were described for the first time at 89 Steinheim an der Murr and Mosbach Sands respectively, two classical Middle 90 Pleistocene localities of Germany (Adam 1954; Forsten 1999; Koenigswald et al. 2007; 91 Van Asperen 2013). Previous studies based on measurements of their dental and 92 postcranial remains have reported differences in body size for these extinct equids 93 (Eisenmann 1991; Forsten 1999). While E. steinheimensis is considered a medium-sized 94 caballoid horse (Eisenmann 1991; Forsten 1999), E. mosbachensis is one of the largest 95 equids that can be found in the Pleistocene fossil record of Europe (Eisenmann 1991; 96 Alberdi et al. 1995). From bone histology, we reconstruct their pace of life and compare 97 it with that of extant Equus, as both Middle Pleistocene equids greatly exceeded in size 98 extant representatives of the group (Alberdi et al. 1995; Forsten 1999; Ernest 2003). 4 99 Thus, we also examine here the bone histology of the extant equid species of Asiatic 100 wild ass (Equus hemionus), Grevyi’s zebra (Equus grevyi), Plains zebra (Equus quagga) 101 and mountain zebra (Equus zebra). These four extant taxa are the most appropriate ones 102 for comparison with our fossil sample, as they cover almost all habitats, body size and 103 life histories within extant wild Equus (Grubb 1981; Penzhorn 1988; Churcher 1993; 104 Nowak 1999; Ernest 2003; Novellie 2008; Moehlman et al. 2013; Kaczensky et al. 105 2015; Orlando 2015; King and Moehlman 2016). 106 Material and methods 107 Equid sample 108 We analysed postcranial fossil remains of E. steinheimensis and E. 109 mosbachensis belonging to the collections of the Staatliches Museum für Naturkunde 110 (Stuttgart, Germany) (Table 1). Specifically, isolated metapodial bones (metacarpi and 111 metatarsi) were used for palaeohistological analyses, while isolated phalanges and 112 metacarpi were analysed for body mass estimations. The material of E. steinheimensis 113 comes from Steinheim an der Murr site (Steinheim, henceforward), a late Middle 114 Pleistocene locality (Van Asperen 2013) situated in south-west Germany (Fig. 2). 115 Steinheim fossil site presents a mix of glacial and interglacial fauna (Adam 1954) due to 116 a sampling bias during the recovery of the remains and/or to the coexistence of species 117 with flexible adaptations (Pushkina et al. 2014). Although the dating of Steinheim site 118 is, hence, controversial, it is thought to generally correlate with the Holsteinian 119 Interglacial (MIS 11) (Van Asperen 2013). The sample of E. mosbachensis belongs to 120 Mosbach Sands, an early Middle Pleistocene site (Koenigswald et al. 2007) located in 121 Wiesbaden (Germany) (Fig. 2). This German classical site is divided into two different 122 stratigraphic ages based on lithological and paleontological criteria (Maul et al. 2000). 123 The E. mosbachensis sample analysed here belongs to the Mosbach 2 fauna, which is 124 known to present interglacial conditions (Koenigswald et al. 2007) and to correlate with 125 the Cromerian Interglacial (MIS 13, 15) (Maul et al. 2000; Kahlke et al. 2011). 126 We also studied an adult metapodial sample of extant E. hemionus, E. grevyi, E. 127 quagga and E. zebra (Table 1) for comparison. Specimens of E. hemionus are stored at 128 the Museum of Domesticated Animals (Halle, Germany) and lived wild in the Gobi 129 Desert (Table 1). They were found killed by poachers and collected during the 5 130 Mongolian-German Biological Expeditions (2001 – 2006) (Schöpke et al. 2012). The E. 131 grevyi individual belongs to the collections of the Zoological Institute of Hamburg 132 University (Germany) (Table 1), after living captive in the Hagenbeck Zoo (Hamburg, 133 Germany) (Table 1). Finally, E. quagga and E. zebra specimens of the study lived semi- 134 captive in the Réserve Africaine de Sigean (Sigean, France) and are part of the 135 collections of the Institut Català de Paleontologia (Barcelona, Spain) (Table 1). 136 Preparation and analysis of histological thin-sections 137 We sectioned 18 metapodia of extant and extinct equids: 4 metapodia of E. 138 steinheimensis, 3 metapodia of E. mosbachensis, 4 metapodia of E. hemionus, 1 139 metatarsus of E. grevyi, 2 metapodia of E. quagga and 4 metapodia of E. zebra (Table 140 1). To avoid damaging of the most valuable fossil specimens, only fossil metapodia 141 fragmented at the mid-shaft level were used in the present study. The preparation of the 142 histological thin-sections followed standard procedures in our laboratory (Köhler et al. 143 2012; Nacarino-Meneses et al. 2016). In all extant and extinct exemplars, a chunk of 2 – 144 3 cm was extracted from the mid-diaphysis of each bone and embedded in an epoxy 145 resin (Araldite 2020). Afterwards, this block was cut into two halves with a low speed 146 diamond saw (IsoMet, Buehler). The mid-shaft surface of each block was later polished 147 using a Metaserv®250 (Buehler) or carborundum powder and fixed to a frosted glass 148 with ultraviolet-curing glue (Loctite 358). Each sample was then cut and grounded with 149 a diamond saw (PetroThin, Buehler) and polished again using the Metaserv®250 150 (Buehler) or carborundum powder to obtain histological thin-sections of 100 – 120 μm 151 thick. Finally, histological slices obtained from fossil samples were dehydrated in 152 alcohol, bathed in a histological clearing agent (Histo-Clear II) and mounted using DPX 153 mounting medium (Scharlau). Thin-sections prepared from extant specimens were 154 covered with a mix of oils (Lamm 2013) to improve their view under the microscope. 155 Samples were analysed under polarized light in a Zeiss Scope.A1 microscope 156 and photographed with the camera incorporated on it (AxioCam ICc5). We also 157 observed the histological thin-sections under polarized light with a ¼ filter, which 158 helps on the identification and visualization of the bone tissue and the bone growth 159 marks (BGMs) (Turner-Walker and Mays 2008). Histological description of bone tissue 160 types and vascular organization follows classical bibliography (Francillon-Vieillot et al. 161 1990; de Margerie et al. 2002; Huttenlocker et al. 2013). Regarding BGMs, both 6 162 cyclical and non-cyclical features were identified in our metapodial equid sample. 163 Cyclical BGMs (CGMs) record annual cycles of growth in response to hormonal and 164 environmental cycles (Castanet et al. 1993; Köhler et al. 2012). From their study, we 165 inferred the minimum longevity of extinct equid species by counting the total number of 166 CGMs within a bone cross-section (Nacarino-Meneses et al. 2016). A non-cyclical 167 BGM related to birth (Nacarino-Meneses and Köhler 2018) was also recognized in our 168 metapodial sample. We considered this neonatal line (NL) as time zero in growth 169 reconstructions. The perimeter of each BGM, regardless of its cyclicality, was 170 calculated using Image J software and the results were plotted to reconstruct the pattern 171 of metapodial growth (Bybee et al. 2006; Woodward et al. 2013). From the obtained 172 metapodial growth curves, we estimated the timing of epiphyseal fusion in these bones 173 for each fossil species following Nacarino-Meneses et al. (2016a). 174 Adult and neonatal body size estimation 175 We analysed metacarpi and phalanges of E. steinheimensis and E. mosbachensis 176 to estimate their adult body mass, a proxy of its body size (Damuth and MacFadden 177 1990). For each species, we measured the proximal depth on the first phalanx and the 178 distal minimal depth of the lateral condyle on the third metacarpus (Eisenmann et al. 179 1988), as these linear measurements present the best correlation with body mass in 180 extant equids (Alberdi et al. 1995). Measurements were taken using a digital electronic 181 precision calliper (0.05 mm error). We used the equations provided by Alberdi et al. 182 (1995) to perform body mass estimations. These data were based on allometric models 183 expressed as the power function y = axb (Damuth and MacFadden 1990) and 184 logarithmically transformed to obtain a linear relationship (lny = lna + b lnx) (Peters 185 1983); where y is body mass, x is the measurement taken on the bone, b is the allometric 186 coefficient and a is a constant (Alberdi et al. 1995). 187 We also estimated the size at birth of extinct and extant equids from the analysis 188 of the NL found in their metapodia (Nacarino-Meneses and Köhler 2018). Alberdi et al. 189 (1995) did not found a significant correlation between the measurements taken at the 190 mid-diaphysis on the metapodia of extant Equus (measures 3 and 4 according to 191 Eisenmann et al. (1988)) and their adult body mass. Because histological slices of the 192 present study were prepared at this level of the diaphysis, we cannot use measurements 193 of the NL to infer body weight at birth. Moreover, equations of Alberdi et al. (1995) are 7 194 calculated for adult individuals. Thus, they are little useful for estimating body mass at 195 earlier ontogenetic stages (Köhler 2010). We decided, then, to represent size at birth as 196 a linear measurement instead of on terms of body mass, by using the perimeter of the 197 NL as a proxy of the neonatal size. 198 Results 199 Bone histology is generally better preserved in Steinheim (Fig. 3A, C; Fig. 4A, 200 B) than in Mosbach samples (Fig. 3B, D; Fig. 4C, D). Although several histological 201 slides from both fossil sites present several microtaphonomical alterations (Fig. 3; Fig. 202 4), it is still possible to recognize the primary bone tissue and the BGMs in all fossil 203 specimens studied (Fig. 3; Fig. 4). The description and identification of the diagenetic 204 alterations of the bone microstructure is, however, beyond the scope of the present 205 research. 206 Primary bone histology 207 Metapodial bones of E. steinheimensis and E. mosbachensis are mainly 208 composed of fibrolamellar bone (Fig. 3A–D), the same as in extant equids (Fig. 3E–H). 209 Most of the vascular canals identified in extinct species are oriented longitudinally (Fig. 210 3A, B). However, both fossil equids also present multiple circular and radial canals that 211 interconnect primary osteons (Fig. 3A, B). Therefore, we found a qualitatively higher 212 number of vascular canals in fossil (Fig. 3A, B) than in extant metapodia (Fig. 3E–H). 213 Metacarpi of E. steinheimensis do not show an external fundamental system 214 (EFS) in their outermost cortex (Fig. 3C). This avascular and lamellar bone tissue 215 (Huttenlocker et al. 2013) is recognized, however, in the metatarsi of this species and in 216 metacarpi and metatarsi of E. mosbachensis (Fig. 3D). 217 Skeletochronology 218 Up to five cyclical BGMs are identified in the metapodia of E. steinheimensis 219 (Fig. 4A, B). In E. mosbachensis, the highest number of cyclical BGMs is found in the 220 metacarpus IPS96017 (Fig. 4C). This specimen presents six cyclical BGMs within the 221 primary bone tissue (Fig. 4C). Besides, all fossil metapodia show a non-cyclical BGM 222 in the most internal cortex (Fig. 4) that corresponds to the NL observed in extant Equus 223 (Nacarino-Meneses and Köhler 2018). 8 224 From the study of cyclical and non-cyclical BGMs, we reconstruct the 225 metapodial growth pattern of E. steinheimensis and E. mosbachensis (Fig. 5). As Figure 226 5 shows, metacarpus (Fig. 5A) and metatarsus (Fig. 5B) of extinct equids follow the 227 same growth pattern as the metapodia of extant E. hemionus, E. grevyi, E. quagga and 228 E. zebra (Fig. 5). In both fossil and extant species, a decrease in growth rate at the 229 second year of life is observed (Fig. 5). Growth reconstruction also shows, however, a 230 steeper slope in fossil than in extant samples (Fig. 5), both during the phase of highest 231 growth rate and during residual growth (Fig. 5). These differences in growth rate are 232 unlikely related to sex differences, as equid bones from males and females are known to 233 grow similarly (Nacarino-Meneses et al. 2016). 234 Adult and neonatal body size 235 We estimated the adult weight of fossil equid species from measurements on 236 metacarpi and phalanges (Table 2). A mean adult body mass of 469.3 kg is obtained for 237 E. steinheimensis, while adult E. mosbachensis is predicted to weight around 607.4 kg 238 (Table 2; Fig. 1). 239 The perimeter of the NL (Fig. 4) is used here to provide a rough estimation of 240 the size at birth for the Pleistocene species studied (Table 3). As Table 3 shows, extinct 241 equids present a larger size at birth than extant Equus. The mid-shaft perimeter of the 242 metacarpi and metatarsi of a neonate E. mosbachensis measures between 83.65 and 243 90.86 mm (Table 3, Fig. 4), while that of E. steinheimensis does around 80 mm (Table 244 3, Fig. 4). Perimeters of neonate metapodia in extant species are considerably smaller, 245 as they vary between 50 – 60 mm in E. hemionus, around 75 mm in E. grevyi, and ca. 246 70 mm in E. quagga and E. zebra (Table 3). 247 Discussion 248 Evolutionary trends in body size of European extinct Equus are widely studied 249 (Eisenmann 1991; Forsten 1991b; Alberdi et al. 1995; Cantalapiedra et al. 2017), but no 250 previous research has investigated its relationship with the life history of the species. In 251 the present research, we reconstruct several life history traits of the extinct caballoid 252 horses E. steinheimensis and E. mosbachensis for the first time, providing a unique life 253 history framework for the analysis of the size shifts observed in European Pleistocene 254 equids. 9 255 Body size of European Middle Pleistocene Equus 256 Body size is usually represented by body mass (i.e. the weight of an individual) 257 because this measure does not consider the shape of the organisms (Saarinen et al. 258 2016). The exhaustive study of Alberdi et al. (1995) reported adult body mass 259 estimations for the most important Plio-Pleistocene Equus species of Europe, Africa and 260 America. Here, we expand their database by calculating the body weight of E. 261 steinheimensis. Our results show that its mean body mass is around 470 kg (Table 2), 262 which is similar to that reported for the Late Pleistocene species Equus germanicus 263 (Fig. 1) (Alberdi et al. 1995). This finding agrees with a previous study by Forsten 264 (1999), where she indicated similarities in metapodial size between E. steinheimensis 265 and latter caballoid equids such as E. germanicus. Our adult body mass estimations for 266 this equid also matches previously published data about the weight of Steinheim horses 267 (Saarinen et al. 2016). We further estimated body mass of E. mosbachensis from 268 measurements on its metacarpi and phalanges (Table 2). We obtained a mean body 269 weight of 607 kg for this early Middle Pleistocene species (Table 2). As expected, this 270 result matches previous body mass estimations of 671.15 kg for E. mosbachensis (Fig. 271 1) (Alberdi et al. 1995) that were obtained from measurements on fossil material 272 recovered from the same locality (Mosbach Sands fossil site) but housed at other 273 European collections (Eisenmann 1979). 274 Size at birth is one of the most important life history traits of mammals (Stearns 275 1992), although this information is still lacking for most extinct equid taxa. In the 276 present study, we obtained an estimation of this attribute for the Middle Pleistocene 277 equids of the analysis (Table 3) based on the perimeter of the NL (Nacarino-Meneses 278 and Köhler 2018) identified in their metapodia (Fig. 4). Our findings reveal that E. 279 mosbachensis presents the largest size at birth within the equids studied, followed by E. 280 steinheimensis, E. grevyi, E. zebra, E. quagga and E. hemionus (Table 3). E. 281 mosbachensis is also the largest Equus of the study (Table 2, Fig. 1), while the Asiatic 282 wild ass is the smallest (230 kg, Ernest 2003). E. steinheimensis (Table 2, Fig. 1), E. 283 grevyi (384 kg, Ernest 2003), E. zebra (296 kg, Ernest 2003) and E. quagga (257 kg, 284 Ernest 2003) present intermediate adult weights between E. mosbachensis and E. 285 hemionus. Therefore, our results on neonatal body size of extinct equids (Table 3) are 286 the ones expected by scaling, as they agree with the known correlation between size at 287 birth and adult body size (Blueweiss et al. 1978; Calder 1984). 10 288 Growth rate of European Middle Pleistocene Equus 289 Growth rate inferences made from bone histology are usually based on the 290 combined analysis of the bone matrix and its vascularization (Amprino 1947; de 291 Margerie et al. 2002; Chinsamy-Turan 2005; Huttenlocker et al. 2013; Lee et al. 2013). 292 Our histological study of E. steinheimensis and E. mosbachensis has not revealed 293 differences in bone tissue type and/or vascularization between both extinct species (Fig. 294 3A, B), which indicates similar rates of bone formation in these Pleistocene equids (de 295 Margerie et al. 2002; Lee et al. 2013). We have observed, however, a higher number of 296 vascular canals in their metapodia (Fig. 3A, B) than in extant Equus’s (Fig. 3E – H). 297 According to de Margerie et al. (2002), this suggests a higher rate of bone deposition in 298 the Pleistocene equids analysed, which could be generally interpreted as a higher 299 growth rate. Growth curves have also been commonly used to report differences in 300 growth rate between sexes (Marín-Moratalla et al. 2013), taxa (Erickson et al. 2001; 301 Erickson 2005; Orlandi-Oliveras et al. 2018) and even between animals under different 302 ecological regimes (Marín-Moratalla et al. 2013; Nacarino-Meneses et al. 2016; 303 Calderón et al. 2019). The slope of the metapodial growth curves obtained here for 304 extinct equids is steeper than those of E. hemionus, E. grevyi, E. quagga and E. zebra 305 (Fig. 5), suggesting higher rates of growth for E. steinheimensis and E. mosbachensis in 306 comparison with these extant species. Therefore, our metapodial growth reconstruction 307 (Fig. 5) supports the histological inference (Fig. 3A, B) of higher rates of growth in E. 308 steinheimensis and E. mosbachensis than in extant Equus. 309 Age at maturity and longevity of European Middle Pleistocene Equus 310 Estimates about the age at maturity in fossil taxa use to rely on the identification 311 of the lamellar and avascular bone tissue that is deposited in the outer cortex of the 312 bones: the external fundamental system (EFS) (Chinsamy-Turan 2005; Marín-Moratalla 313 et al. 2013; Calderón et al. 2019). The timing of deposition of this tissue in equid and 314 ruminant femora is known to correlate with the age at first reproduction of the species 315 (Jordana et al. 2016; Nacarino-Meneses et al. 2016), as it is supposed to record the life 316 history trade-off between growth and reproduction (Stearns 1992). The significance of 317 deposition of the EFS in other bones, however, is poorly known. Although several 318 authors have suggested a relationship between the presence of EFS in metapodia and the 319 skeletal maturity of the species (Martínez-Maza et al. 2014), the appearance of this 11 320 tissue type in these bones is just likely related with the cessation of periosteal/radial 321 bone growth (Nacarino-Meneses et al. 2016). Therefore, the lack of EFS in metacarpi of 322 E. steinheimensis (Fig. 3C) might just indicates a prolonged periosteal bone growth in 323 this Pleistocene species. Alternatively, inferences on skeletal maturity can be made from 324 growth reconstructions, as growth curves provide reliable information about the 325 cessation of longitudinal bone growth (Nacarino-Meneses et al. 2016). The decrease in 326 growth rate observed at the second year of life in metapodia of Pleistocene and extant 327 species (Fig. 5) suggests a similar age of epiphyseal fusion in these bones for both 328 groups of equids (Nacarino-Meneses et al. 2016). However, we cannot use this finding 329 to infer an absolute age at skeletal maturity for E. mosbachensis and E. steinheimensis 330 because metapodia fuse their epiphyses very early in ontogeny (Silver 1963). Future 331 histological studies on other long bones that finish later their growth, such as the femur 332 or the tibia (Silver 1963), will be necessary to obtain information about this biological 333 trait in the extinct taxa studied. These forthcoming studies would also shed light on the 334 onset of sexual maturity in these Middle Pleistocene species, by analysing the timing of 335 deposition of the EFS in their femora (Nacarino-Meneses et al. 2016). 336 The total number of CGMs within a bone cortex is known to broadly match the 337 age at death of an individual (Castanet et al. 1993). Hence, several studies have 338 analysed these histological features on ancient bones to estimate longevity on extinct 339 mammals (Köhler and Moyà-Solà 2009; Marín-Moratalla et al. 2011; Moncunill-Solé et 340 al. 2016; Orlandi-Oliveras et al. 2016). Because mammals present asymptotic growth 341 (Lee et al. 2013) and the identification of CGMs within the EFS is sometimes a 342 challenging issue (Woodward et al. 2013), this methodology always yields a minimum 343 longevity in this group of vertebrates (Castanet et al. 2004; Castanet 2006; Calderón et 344 al. 2019). Moreover, femur is the most valuable bone for skeletochronological studies 345 (Horner et al. 1999; García-Martínez et al. 2011; Nacarino-Meneses et al. 2016), 346 although metapodia is known to yield accurate individual age estimations in equids 347 (Martínez-Maza et al. 2014; Nacarino-Meneses et al. 2016). In our Pleistocene equid 348 sample, we have identified a maximum of 5 and 6 CGMs in the metapodial cortex of E. 349 steinheimensis and E. mosbachensis respectively (Fig. 4). These results suggest that the 350 minimum longevity of E. steinheimensis is 5 years, while it is of 6 years in E. 351 mosbachensis. Although the number of CGMs identified in the metapodial cortex of 352 these extinct equids is similar to that found in the limb bones of E. hemionus (Nacarino- 12 353 Meneses et al. 2016), the small sample size of the present research limits us to perform 354 a comparison with enough confidence between the longevity of extinct species (E. 355 steinheimensis and E. mosbachensis) and that of extant equids. Further studies with a 356 larger sample size are necessary, hence, to estimate the longevity of these Middle 357 Pleistocene species. 358 Size variation in Old World Equus: environment and life history 359 Our life history reconstruction of E. mosbachensis and E. steinheimensis 360 indicates that these Middle Pleistocene species grew at higher rates than extant horses 361 (Fig. 3 and 5). As resource availability is the main selection pressure acting on 362 individual growth rate (Palkovacs 2003), the difference in adult body size between 363 extant and Middle Pleistocene equids would likely be influenced by the resource 364 availability of each habitat. On the one hand, paleoenvironmental reconstructions on 365 Mosbach (Cromerian interglacial (MIS 13, 15), Kahlke et al. (2011)) suggest open 366 steppe landscapes interrupted by warm and humid episodes with extended forests to be 367 the main habitats at this fossil site (Maul et al. 2000; Kahlke et al. 2011). Steinheim 368 horse sample, on the other hand, is thought to have been originated in the Holsteinian 369 interglacial (MIS11) (Van Asperen 2013), which is characterized by temperate, humid 370 and forested environments (Nitychoruk et al. 2005). Specifically, Steinheim site during 371 this interglacial is thought to present a temperate humid habitat of woodland and 372 shrubland (Pushkina et al. 2014). Extant Grevy’s zebra also occurs in grasslands and 373 shrublands, but its habitat is considered as arid or semiarid (Rubenstein et al. 2016) with 374 a negative mean annual climatic water balance (Schulz and Kaiser 2013). The habitat of 375 E. zebra, who dwells in mountainous and escarpment areas (Novellie 2008), also 376 presents a negative mean annual climatic water balance (Schulz and Kaiser 2013). Thus, 377 the habitat of both E. grevyi and E. zebra can be considered as poor-resourced in 378 comparison with that of E. steinheimensis or E. mosbachensis. The drier habitat of the 379 Grevy’s and the mountain zebra probably forces them to grow at lower rates than 380 extinct Equus (Fig. 3 and 5) which, following the model of Palkovacs (2003), finally 381 results in a smaller adult body size for E. grevyi and E. zebra than for Middle 382 Pleistocene taxa. A similar reasoning applies to E. hemionus and E. quagga, which also 383 grow at lower rates than E. steinheimensis and E. mosbachensis (Fig. 3 and 5). The 384 Asiatic wild ass, on the one hand, is endemic of the Gobi desert, one of the most arid 385 habitats of the world (Kaczensky et al. 2015). Equus quagga, on the other hand, lives in 13 386 the open savannah and open woodland of several African countries (King and 387 Moehlman 2016) that are likely poor-resourced in comparison with Middle Pleistocene 388 environments. Growth rate of both E. hemionus and E. quagga, hence, is probably 389 restricted by the environmental conditions of their habitat, as it is their final body size. 390 Along with modifications in growth rate, the life history model proposed by 391 Palkovacs (2003) postulates that age at sexual maturity also influences adult body size. 392 However, and as previously mentioned, we could not obtain information about this LH 393 trait from our histological analysis due to the characteristic of the sample studied (e.g. 394 lack of femora). Following this model (Palkovacs 2003), we would expect that the fast- 395 growing Middle Pleistocene equids present an advance in their age at sexual maturity in 396 comparison with extant taxa, although future histological studies on femora (Nacarino- 397 Meneses et al. 2016) must be done to confirm this preliminary hypothesis. Interestingly, 398 the hypothesis of an earlier sexual maturation in the Middle Pleistocene species E. 399 mosbachensis and E. steinheimensis conforms to the predictions of life history theory in 400 environments where there is a high extrinsic mortality pressure (i.e. predation) (Roff 401 1992; Stearns 1992; Roff 2002). In these circumstances, theory predicts that organisms 402 achieve sexual maturity early to reduce the time of exposure to juvenile mortality 403 (Stearns 2000). Paleontological and ecological data indeed suggest differences in levels 404 of extrinsic mortality between Middle Pleistocene equids and extant species. Thus, 405 while important potential predators for equids such as felids (Lynx issiodorensis, 406 Panthera gombaszoegensis, Panthera leo fossilis, Panthera pardus, Acinonyx 407 pardinensis, Homotherium), canids (Canis lupus mosbachensis, Cuon alpinus priscus, 408 Cuon dubius stehlini) and hyaenids (Pliocrocuta perrieri, Crocuta spelaea) have been 409 found in Mosbach or Steinheim sites (Adam 1954; Kahlke 1961; Kahlke 1975), the 410 extant E. grevyi, E. hemionus, E. quagga and E. zebra are threatened by a few number 411 of carnivore species (Grubb 1981; Penzhorn 1988; Churcher 1993; Feh et al. 2001). 412 Conclusions 413 Our study infers, for the first time, several life history traits of European Middle 414 Pleistocene Equus from the histological analysis of their metapodia. Bone histology and 415 growth reconstruction performed in these bones indicate higher growth rates for the 416 extinct E. mosbachensis and E. steinheimensis in comparison with extant 417 representatives of the group (E. hemionus, E. grevyi, E. quagga and E. zebra). Their 14 418 occurrence in temperate and humid habitats during European interglacials explains the 419 higher growth rates found in these extinct equids, as they dwelt in resource-richer 420 environments than those inhabited today by the extant Asiatic wild ass and the Grevyi’s, 421 mountain and plains zebra. The resource levels of each habitat seem to be, thus, the 422 main selection pressure acting upon individual growth rate and finally determining body 423 size in Equus. The influence of other key life history traits such as age at maturity in 424 determining the adult body size of these mammals could not be assessed in the present 425 research, although some hypotheses are proposed. Our first approach to the 426 skeletochronology of these Middle Pleistocene equids suggests a similar age at 427 epiphyseal fusion for the metapodia in E. mosbachensis and E. steinheimensis and 428 extant Equus and yield a minimum age at death for these extinct species. In the present 429 study, we also provide estimates of the adult (in terms of body mass) and the neonatal 430 body size (considering the perimeter of the NL as a proxy of this size) of E. 431 mosbachensis and E. steinheimensis. Our results match previous adult weight 432 estimations for these extinct equids and show that their size at birth is the expected from 433 allometric scaling. 434 Acknowledgments 435 We thank Reinhard Ziegler for access to the collections of the Staatliches 436 Museum für Naturkunde (Stuttgart, Germany) and for permission to cut the metapodia 437 of E. steinheimensis and E. mosbachensis. We are also grateful to Renate Schafberg, 438 Thomas Kaiser and for loans and permission to cut bones from the collections of the 439 Museum of Domesticated Animals of the Martin-Luther-University Halle-Wittenberg 440 (Halle, Saale, Germany) and the Zoological Institute of Hamburg University (Hamburg, 441 Germany) respectively. We thank B. Lamglait (currently at Université de Montréal, 442 Canada) and the Réserve Africaine de Sigean (Sigean, France) for donation of aged 443 equid specimens. Manuel Fernández and Gemma Prats-Muñoz are acknowledged for 444 the preparation of the histological thin-sections of the study. We are also in debt to 445 Meike Köhler for discussion on this paper and the revision of earlier versions of the 446 manuscript. Thanks are extended to Xavier Jordana, Alexandra Houssaye and one 447 anonymous reviewer for their feedback, which resulted in significant improvements of 448 this manuscript. 449 Funding details 15 450 This work was supported by the Spanish Ministry of Economy and Competitiveness 451 under Grant CGL-2015-63777, Grant BES-2013-066335 (to C.N.-M.) and Grant 452 EEBB-I-16-10546 (to C.N.-M.); the Government of Catalonia under Grant 2014-SGR- 453 1207, Grant CERCA Programme and Grant 2016FI_B00202 AGAUR (to G.O.-O); and 454 the DST-NRF Centre of Excellence in Palaeosciences under Grant COE2019-PD03 (to 455 C.N.-M.). Opinions expressed and conclusions arrived at, are those of the author and are 456 not necessarily attributed to the CoE. 457 Declaration of interest statement 458 The authors declare that there is no conflict of interest. 459 References 460 Adam KD. 1954. Die mittelpleistozänen Faunen von Steinheim an der Murr 461 (Württemberg). Quaternaria. 1:131–144. 462 Alberdi MT, Bonadonna FP. 1988. Equidae (Perissodactyla, Mammalia): extinctions 463 subsequent to the climatic changes. Rev Española Paleontol. 3:39–43. 464 Alberdi MT, Ortiz-Jaureguizar E, Prado JL. 1998. A quantitative review of European 465 stenonoid horses. J Paleontol. 72(2):371–387. doi:10.1017/S0022336000036350. 466 Alberdi MT, Prado JL, Ortiz-Jaureguizar E. 1995. Patterns of body size changes in 467 fossil and living Equini (Perissodactyla). Biol J Linn Soc. 54:349–370. 468 Amprino R. 1947. La structure du tissu osseux envisagée comme expression de 469 différences dans la vitesse de l’accroissement. Arch Biol. 58:315–330. 470 Amson E, Kolb C, Scheyer TM, Sánchez-Villagra MR. 2015. Growth and life history of 471 Middle Miocene deer (Mammalia, Cervidae) based on bone histology. Comptes Rendus 472 - Palevol. 14(8):637–645. doi:10.1016/j.crpv.2015.07.001. 473 Van Asperen EN. 2013. Position of the Steinheim interglacial sequence within the 474 marine oxygen isotope record based on mammal biostratigraphy. Quat Int. 292:33–42. 475 doi:10.1016/j.quaint.2012.10.045. 476 Azzaroli A. 1983. Quaternary mammals and the “end-Villafranchian” dispersal event - 16 477 A turning point in the history of Eurasia. Palaeogeogr Palaeoclimatol Palaeoecol. 44(1– 478 2):117–139. doi:10.1016/0031-0182(83)90008-1. 479 Azzaroli A. 1992. Ascent and decline of monodactyl equids: a case for prehistoric 480 overkill. Ann Zool Fennici. 28:151–163. 481 Blueweiss L, Fox H, Kudzma V, Nakashima D, Peters R, Sams S. 1978. Relationships 482 between body size and some life history parameters. Oecologia. 37:257–272. 483 Bybee PJ, Lee AH, Lamm E-T. 2006. Sizing the Jurassic theropod dinosaur Allosaurus: 484 assessing growth strategy and evolution of ontogenetic scaling of limbs. J Morphol. 485 267:347–359. doi:10.1002/jmor. 486 Calder WAI. 1984. Size, Function and Life History. New York: Dover Publications. 487 Calderón T, DeMiguel D, Arnold W, Stalder G, Köhler M. 2019. Calibration of life 488 history traits with epiphyseal closure, dental eruption and bone histology in captive and 489 wild red deer. J Anat. Online ver. doi:10.1111/joa.13016. 490 Cantalapiedra JL, Prado JL, Hernández Fernández M, Alberdi MT. 2017. Decoupled 491 ecomorphological evolution and diversification in Neogene-Quaternary horses. Science 492 (80- ). 355(6325):627–630. doi:10.1126/science.aag1772. 493 Castanet J. 2006. Time recording in bone microstructures of endothermic animals; 494 functional relationships. Comptes Rendus - Palevol. 5(3–4):629–636. 495 doi:10.1016/j.crpv.2005.10.006. 496 Castanet J, Croci S, Aujard F, Perret M, Cubo J, de Margerie E. 2004. Lines of arrested 497 growth in bone and age estimation in a small primate: Microcebus murinus. J Zool. 498 263(1):31–39. doi:10.1017/S0952836904004844. 499 Castanet J, Francillon-Vieillot H, Meunier F, de Ricqlès A. 1993. Bone and individual 500 aging. In: Hall BK, editor. Bone: A Treatise, Vol. 7. Boca Raton: CRC Press. p. 245– 501 283. 502 Chinsamy-Turan A. 2005. The microstructure of dinosaur bone. Deciphering biology 503 with fine-scale techniques. Baltimore and London: The Johns Hopkins University Press. 17 504 Churcher C. 1993. Equus grevyi. Mamm Species. 453:1–9. 505 Cubo J, le Roy N, Martinez-Maza C, Montes L. 2012. Paleohistological estimation of 506 bone growth rate in extinct archosaurs. Paleobiology. 38(2):335–349. 507 doi:10.5061/dryad.j2m25n82. 508 Damuth J. 1981. Population density and body size in mammals. Nature. 290:699–700. 509 Damuth J, MacFadden BJ. 1990. Body size in mammalian Paleobiology. Estimations 510 and biological implications. Cambridge: Cambridge University Press. 511 Eisenberg JF. 1990. The behavioral/ecological significance of body size in the 512 Mammalia. In: Damuth J, MacFadden BJ, editors. Body size in mammalian 513 Paleobiology. Estimations and biological implications. Cambridge: Cambridge 514 University Press. p. 25–37. 515 Eisenmann V. 1979. Les métapodes d’Equus sensu lato (Mammalia, Périssodactyla). 516 Geobios. 12(6):863–886. doi:10.1016/S0016-6995(79)80004-2. 517 Eisenmann V. 1991. Les chevaux quaternaires européens (Mammalia, Perissodactyla). 518 Taille, typologie, biostratigraphie et taxonomie. Geobios. 24(6):747–759. 519 Eisenmann V, Alberdi MT, de Giuli C, Staesche U. 1988. Studying fossil horses. 520 Collected papers after the “New York International Hipparion Conference, 1981”. 521 Volume I: Methodology. Woodburne M, Sondaar P, editors. Leiden: E.J. Brill. 522 Erickson GM. 2005. Assessing dinosaur growth patterns: A microscopic revolution. 523 Trends Ecol Evol. 20(12):677–684. doi:10.1016/j.tree.2005.08.012. 524 Erickson GM, Rogers KC, Yerby SA. 2001. Dinosaurian growth patterns and rapid 525 avian growth rates. Nature. 412(6845):429–433. doi:10.1038/35086558. 526 Ernest SKM. 2003. Life history characteristics of placental nonvolant mammals. 527 Ecology. 84(12):3402. doi:10.1890/11-1341.1. 528 Feh C, Munkhtuya B, Enkhbold S, Sukhbaatar T. 2001. Ecology and social structure of 529 the Gobi khulan Equus hemionus subsp. in the Gobi B National Park, Mongolia. Biol 530 Conserv. 101:51–61. 18 531 Forsten A. 1988. The small caballoid horse of the upper Pleistocene and Holocene. J 532 Anim Breed Genet. 105(1–6):161–176. doi:10.1111/j.1439-0388.1988.tb00288.x. 533 Forsten A. 1991a. Size trends in Holarctic anchitherines (Mammalia, Equidae). J 534 Paleontol. 65(1):147–159. 535 Forsten A. 1991b. Size decrease in Pleistocene-Holocene true or caballoid horses of 536 Europe. Mammalia. 55(3):407–420. doi:10.1515/mamm.1991.55.3.407. 537 Forsten A. 1999. The horses (genus Equus) from the Middle Pleistocene of Steinheim, 538 Germany. In: Haynes G, Klimowicz J, Reumer JWF, editors. Mammoths and the 539 Mammoth Fauna: Studies of an Extinct Ecosystem. Deinsea 6. p. 147–154. 540 Francillon-Vieillot H, de Buffrénil V, Castanet J, Géraudie J, Meunier FJ, Sire JY, 541 Zylberberg L, de Ricqlès A. 1990. Microstructure and mineralization of vertebrate 542 skeletal tissues. In: Carter JG, editor. Skeletal biomineralization: patterns, processes and 543 evolutionary trends. New York: Van Nostrand Reinhold. p. 471–530. 544 García-Martínez R, Marín-Moratalla N, Jordana X, Köhler M. 2011. The ontogeny of 545 bone growth in two species of dormice: reconstructing life history traits. Comptes 546 Rendus - Palevol. 10(5–6):489–498. doi:10.1016/j.crpv.2011.03.011. 547 Grubb P. 1981. Equus burchelli. Mamm Species. 157:1–9. 548 Guthrie RD. 2003. Rapid body size decline in Alaskan Pleistocene horses before 549 extinction. Nature. 426(6963):169–171. doi:10.1038/nature02098. 550 Horner JR, de Ricqlès A, Padian K. 1999. Variation in dinosaur skeletochronology 551 indicators: implications for age assessment and physiology. Paleobiology. 25(3):295– 552 304. doi:10.1666/0094-8373(1999)025<0295:VIDSII>2.3.CO;2. 553 Huttenlocker AK, Woodward HN, Hall BK. 2013. The biology of bone. In: Padian K, 554 Lamm E-T, editors. Bone histology of fossil tetrapods: advancing methods, analysis, 555 and interpretation. Berkeley: University of California Press. p. 13–34. 556 Janis C. 2007. The Horse Series. In: Regal B, editor. Icons of Evolution. Westport: 557 Greenwood Press. p. 251–280. 19 558 Jordana X, Marín-Moratalla N, Moncunill-Solé B, Nacarino-Meneses C, Köhler M. 559 2016. Ontogenetic changes in the histological features of zonal bone tissue of 560 ruminants: a quantitative approach. Comptes Rendus - Palevol. 15(1–2):255–266. 561 doi:10.1016/j.crpv.2015.03.008. 562 Kaczensky P, Lkhagvasuren B, Pereladova O, Hemami M, Bouskila A. 2015. Equus 563 hemionus. IUCN Red List Threat Species. e.T7951A45. 564 Kahlke HD. 1975. The Macro-faunas of Continental Europe during the Middle 565 Pleistocene: Stratigraphic sequence and problems of intercorrelation. In: Butzer KW, 566 Isaac GL, editors. After the Australopithecines. Stratigraphy, Ecology, and Culture 567 Change in the Middle Pleistocene. The Hague: Mouton Publishers. p. 309–374. 568 Kahlke RD, García N, Kostopoulos DS, Lacombat F, Lister AM, Mazza PPA, Spassov 569 N, Titov V V. 2011. Western Palaearctic palaeoenvironmental conditions during the 570 Early and early Middle Pleistocene inferred from large mammal communities, and 571 implications for hominin dispersal in Europe. Quat Sci Rev. 30(11–12):1368–1395. 572 doi:10.1016/j.quascirev.2010.07.020. 573 Kahlke VH-D. 1961. Revision der Säugetierfaunen der klassischen deutschen 574 Pleistozän-Fundstellen von Süßenborn, Mosbach und Taubach. Geologie. 10:493–532. 575 King SRB, Moehlman PD. 2016. Equus quagga. IUCN Red List Threat 576 Species.:e.T41013A45172424. 577 Kleiber M. 1932. Body size and metabolism. Hilgardia. 6(11):315–353. 578 doi:10.1017/CBO9781107415324.004. 579 Koenigswald W V, Smith BH, Keller T. 2007. Supernumerary teeth in a subadult rhino 580 mandible (Stephanorhinus hundsheirnensis) from the middle Pleistocene of Mosbach in 581 Wiesbaden (Germany). Paläontologische Zeitschrift. 81:416–428. 582 Köhler M. 2010. Fast or slow? The evolution of life history traits associated with insular 583 dwarfing. In: Pérez-Mellado V, Ramon C, editors. Islands and Evolution. Vol. 19. Maó: 584 Institut Menorquí d’Estudis. Recerca. p. 261–280. 585 Köhler M, Marín-Moratalla N, Jordana X, Aanes R. 2012. Seasonal bone growth and 20 586 physiology in endotherms shed light on dinosaur physiology. Nature. 487:358–361. 587 doi:10.1038/nature11264. 588 Köhler M, Moyà-Solà S. 2009. Physiological and life history strategies of a fossil large 589 mammal in a resource-limited environment. Proc Natl Acad Sci U S A. 106(48):20354– 590 20358. doi:10.1073/pnas.0813385106. 591 Kolb C, Scheyer TM, Veischegger K, Van den Hoek Ostende LW, Hayashi S, Sánchez- 592 Villagra MR. 2015. Mammalian bone palaeohistology: a survey and new data with 593 emphasis on island forms. PeerJ. 3:e1358. doi:10.7717/peerj.1358. 594 Lamm E-T. 2013. Preparation and sectioning of specimens. In: Padian K, Lamm E-T, 595 editors. Bone histology of fossil tetrapods: advancing methods, analysis, and 596 interpretation. Berkeley: University of California Press. p. 55–160. 597 Lee AH, Huttenlocker AK, Padian K, Woodward HN. 2013. Analysis of growth rates. 598 In: Padian K, Lamm E-T, editors. Bone histology of fossil tetrapods: advancing 599 methods, analysis, and interpretation. Berkeley: University of California Press. p. 217– 600 251. 601 Lindsay EH, Opdyke ND, Johnson NM. 1980. Pliocene dispersal of the horse Equus 602 and late Cenozoic mammalian dispersal events. Nature. 287(5778):135–138. 603 doi:10.1038/287135a0. 604 MacFadden BJ. 1986. Fossil horses from “Eohippus” (Hyracotherium) to Equus: 605 scaling, Cope’s Law, and the evolution of body size. Paleobiology. 12(4):355–369. 606 doi:10.1017/S0094837300003109. 607 MacFadden BJ. 1992. Fossil horses. Systematics, Paleobiology, and Evolution of the 608 Family Equidae. Cambridge: Cambridge University Press. 609 MacFadden BJ. 2005. Fossil Horses - Evidence for Evolution. Science (80- ). 610 307:1728–1730. doi:307/5716/1728 [pii] 10.1126/science.1105458. 611 de Margerie E, Cubo J, Castanet J. 2002. Bone typology and growth rate: Testing and 612 quantifying “Amprino’s rule” in the mallard (Anas platyrhynchos). Comptes Rendus - 613 Biol. 325(3):221–230. doi:10.1016/S1631-0691(02)01429-4. 21 614 Marín-Moratalla N, Jordana X, García-Martínez R, Köhler M. 2011. Tracing the 615 evolution of fitness components in fossil bovids under different selective regimes. 616 Comptes Rendus - Palevol. 10(5–6):469–478. doi:10.1016/j.crpv.2011.03.007. 617 Marín-Moratalla N, Jordana X, Köhler M. 2013. Bone histology as an approach to 618 providing data on certain key life history traits in mammals: implications for 619 conservation biology. Mamm Biol. 78(6):422–429. doi:10.1016/j.mambio.2013.07.079. 620 Martínez-Maza C, Alberdi MT, Nieto-Diaz M, Prado JL. 2014. Life-history traits of the 621 miocene Hipparion concudense (Spain) inferred from bone histological structure. PLoS 622 One. 9(8):e103708. doi:10.1371/journal.pone.0103708. 623 Maul LC, Rekovets L, Heinrich W-D, Keller T, Storch G. 2000. Arvicola mosbachensis 624 (Schmidtgen 1991) of Mosbach 2: a basic sample for the early evolution of the genus 625 and a reference for further biostratigraphical studies. Senckenbergiana lethaea. 80:129– 626 147. 627 McHorse BK, Biewener AA, Pierce SE. 2017. Mechanics of evolutionary digit 628 reduction in fossil horses (Equidae). Proc R Soc B Biol Sci. 284(1861):20171174. 629 doi:10.1098/rspb.2017.1174. 630 McNab BK. 1990. The physiological significance of body size. In: Damuth J, 631 MacFadden BJ, editors. Body size in mammalian Paleobiology. Estimations and 632 biological implications. Cambridge: Cambridge University Press. p. 11–23. 633 Moehlman PD, Rubenstein DI, Kebede F. 2013. Equus grevyi. IUCN Red List Threat 634 Species.:e.T7950A21070406. doi:10.2305/IUCN.UK.2013- 635 1.RLTS.T7950A21070406.en. 636 Moncunill-Solé B, Orlandi-Oliveras G, Jordana X, Rook L, Köhler M. 2016. First 637 approach of the life history of Prolagus apricenicus (Ochotonidae, Lagomorpha) from 638 Terre Rosse sites (Gargano, Italy) using body mass estimation and paleohistological 639 analysis. Comptes Rendus - Palevol. 15(1–2):227–237. doi:10.1016/j.crpv.2015.04.004. 640 Nacarino-Meneses C, Jordana X, Köhler M. 2016. Histological variability in the limb 641 bones of the Asiatic wild ass and its significance for life history inferences. PeerJ. 642 4(10):e2580. doi:10.7717/peerj.2580. 22 643 Nacarino-Meneses C, Köhler M. 2018. Limb bone histology records birth in mammals. 644 PLoS One. 13(6):e0198511. 645 Nitychoruk J, Bińka K, Hoefs J, Ruppert H, Schneider J. 2005. Climate reconstruction 646 for the Holsteinian Interglacial in eastern Poland and its comparison with isotopic data 647 from Marine Isotope Stage 11. Quat Sci Rev. 24(5–6):631–644. 648 doi:10.1016/j.quascirev.2004.07.023. 649 Novellie P. 2008. Equus zebra. IUCN Red List Threat Species.:e.T7960A12876787. 650 Nowak RM. 1999. Walker’s mammals of the world. Baltimore and London: The Johns 651 Hopkins University Press. 652 Orlandi-Oliveras G, Jordana X, Moncunill-Solé B, Köhler M. 2016. Bone histology of 653 the giant fossil dormouse Hypnomys onicensis (Gliridae, Rodentia) from Balearic 654 Islands. Comptes Rendus - Palevol. 15(1–2):238–244. doi:10.1016/j.crpv.2015.05.001. 655 Orlandi-Oliveras G, Nacarino-Meneses C, Koufos GD, Köhler M. 2018. Bone histology 656 provides insights into the life history mechanisms underlying dwarfing in hipparionins. 657 Sci Rep. 8(1):17203. doi:10.1038/s41598-018-35347-x. 658 Orlando L. 2015. Equids. Curr Biol.:R973–R978. doi:10.1016/j.cub.2015.09.005. 659 Ortiz-Jaureguizar E, Alberdi MT. 2003. The pattern of body mass changes in the 660 Hipparionini (Perissodactyla, Equidae) of the Iberian Peninsula during the Upper 661 Miocene-Upper Pliocene. Coloquios Paleontol. 1:499–509. 662 Padian K, de Ricqlès A, Horner JR. 2001. Dinosaurian growth rates and bird origins. 663 Nature. 412(6845):405–408. doi:10.1038/35086500. 664 Palkovacs EP. 2003. Explaining adaptive shifts in body size on islands: a life history 665 approach. Oikos. 103(1):37–44. doi:10.1034/j.1600-0706.2003.12502.x. 666 Penzhorn BL. 1988. Equus zebra. Mamm Species. 314:1–7. 667 Peters RH. 1983. The ecological implications of body size. Cambridge: Cambridge 668 University Press. 669 Pushkina D, Bocherens H, Ziegler R. 2014. Unexpected palaeoecological features of the 23 670 Middle and Late Pleistocene large herbivores in southwestern Germany revealed by 671 stable isotopic abundances in tooth enamel. Quat Int. 339–340:164–178. 672 doi:10.1016/j.quaint.2013.12.033. 673 Roff DA. 1992. The evolution of life histories. Theory and analysis. New York: 674 Chapman and Hall. 675 Roff DA. 2002. Life history evolution. Sunderland: Sinauer Associates. 676 Rubenstein D, Low Mackey B, Davidson Z, Kebede F, King SR. 2016. Equus grevyi. 677 IUCN Red List Threat Species.:e.T7950A89624491. 678 Saarinen J, Eronen J, Fortelius M, Seppä H, Lister AM. 2016. Patterns of diet and body 679 mass of large ungulates from the Pleistocene of Western Europe, and their relation to 680 vegetation. Palaeontol Electron. 19(3):1–58. 681 Schöpke K, Stubbe A, Stubbe M, Batsaikhan N, Schafberg R. 2012. Morphology and 682 variation of the Asiatic wild ass (Equus hemionus hemionus). Erforsch Biol Ressourcen 683 der Mongolei. 12:77–84. 684 Schulz E, Kaiser TM. 2013. Historical distribution, habitat requirements and feeding 685 ecology of the genus Equus (Perissodactyla). Mamm Rev. 43(2):111–123. 686 doi:10.1111/j.1365-2907.2012.00210.x. 687 Shoemaker L, Clauset A. 2014. Body mass evolution and diversification within horses 688 (family Equidae). Ecol Lett. 17(2):211–220. doi:10.1111/ele.12221. 689 Silver IA. 1963. The aging of domestic animals. In: Brothwell D, Higgs E, editors. 690 Science in Archaeology: a comprehensive survey of progress and research. New York: 691 Basic Books. p. 250–268. 692 Simpson GG. 1953. The major features of evolution. New York: Columbia University 693 Press. 694 Stanley SM. 1973. An explanation for Cope’s rule. Evolution (N Y). 27:1–26. 695 Stearns SC. 1992. The evolution of life histories. New York: Oxford University Press. 696 Stearns SC. 2000. Life history evolution: successes, limitations, and prospects. 87:476– 24 697 486. 698 Strömberg CAE. 2006. Evolution of hypsodonty in equids: testing a hypothesis of 699 adaptation. Paleobiology. 32(2):236–258. 700 Turner-Walker G, Mays S. 2008. Histological studies on ancient bone. In: Pinhasi R, 701 Mays S, editors. Advances in human paleopathology. Chichester: John Wiley & Sons, 702 Ltd. p. 121–146. 703 Western D. 1979. Size, life history and ecology in mammals. Afr J Ecol. 17(4):185– 704 204. doi:10.1111/j.1365-2028.1979.tb00256.x. 705 Woodward HN, Freedman Fowler EA, Farlow JO, Horner JR. 2015. Maiasaura, a 706 model organism for extinct vertebrate population biology: a large sample statistical 707 assessment of growth dynamics and survivorship. Paleobiology. 41(04):503–527. 708 doi:10.1017/pab.2015.19. 709 Woodward HN, Padian K, Lee AH. 2013. Skeletochronology. In: Padian K, Lamm E-T, 710 editors. Bone histology of fossil tetrapods: advancing methods, analysis, and 711 interpretation. Berkeley: University of California Press. p. 195–215. 712 713 25 714 Figure 1. Size-decrease trend experienced by caballoid (circles) and stenonoid 715 (triangles) horses during the Pleistocene (data from Alberdi et al. 1995; Ernest 2003; 716 Cantalapiedra et al. 2017). Body mass estimates of Equus mosbachensis (black star) and 717 Equus steinheimensis (white star) is also shown. Black circles and triangles = extinct 718 Equus species; 75% grey triangles = zebras; 25% grey triangles = asses. 719 720 26 721 Figure 2. Map of Germany showing the provenance of the fossil sample. M = Mosbach 722 Sands; S = Steinheim an der Murr. 723 724 27 725 Figure 3. Metapodial bone histology of European Middle Pleistocene (a–d) and extant 726 Equus (e–h). (a) Metacarpal bone cortex of Equus steinheimensis (IPS96013). (b) 727 Metatarsal bone cortex of Equus mosbachensis (IPS96014). (c) Metacarpal bone cortex 728 of Equus steinheimensis (IPS96011). (d) Metacarpal bone cortex of Equus 729 mosbachensis (IPS96017) (e) Metacarpal bone cortex of Equus hemionus (IPS83877). 730 (f) Metatarsal bone cortex of Equus grevyi (IPS84963). (g) Metacarpal bone cortex of 731 Equus quagga (IPS116319). (h) Metatarsal bone cortex of Equus zebra (IPS116321). 732 CVC = circular vascular canal; EFS = external fundamental system; FLC = 733 fibrolamellar complex; LCV = longitudinal vascular canal; RVC = radial vascular canal. 734 Scale bars = 200 μm. 28 735 736 29 737 Figure 4. Bone growth marks identified in the metapodia of Equus steinheimensis (a,b) 738 and Equus mosbachensis (c,d). (a) Metacarpal bone cortex of Equus steinheimensis 739 IPS96013. (b) Metatarsal bone cortex of Equus steinheimensis IPS96010. (c) 740 Metacarpal bone cortex of Equus mosbachensis IPS96017. (d) Metatarsal bone cortex 741 of Equus mosbachensis IPS96014. White arrows = cyclical bone growth marks; black 742 arrows = non-cyclical bone growth marks (neonatal line – NL). Scale bars = 1 mm. 743 744 30 745 Figure 5. Metapodial growth of European Middle Pleistocene (continuous lines) and 746 extant (dashed lines) Equus. Bone perimeter (mm, ordinate axis) is plotted against 747 estimated age (years, abscissa axis) to obtain growth curves. (a) Growth curves obtained 748 from the metacarpus. (b) Growth curves obtained from the metatarsus. Legend is shown 749 at the bottom of the figure. 750 751 31 752 Table 1. Sample studied for bone histology. Mc = metacarpus; Mt = metatarsus; SMNS 753 = Staatliches Museum für Naturkunde Stuttgart (Stuttgart, Germany); MDA = Museum 754 of Domesticated Animals (Halle, Germany); ZIHU = Zoological Institute of Hamburg 755 University (Hamburg, Germany); ICP = Institut Català de Paleontologia (Barcelona, 756 Spain). The star (*) indicates fossil species. Collection Species Code Bone Site/Habitat Institution code IPS96010 32917/23 Mt Steinheim an der SMNS Murr IPS96011 32803/61 Mc Steinheim an der SMNS Murr E. steinheimensis* IPS96012 32803/235 Mt Steinheim an der SMNS Murr IPS96013 32803/321 Mc Steinheim an der SMNS Murr IPS96014 32850/87 Mt Mosbach Sands SMNS E. mosbachensis* IPS96016 32850/6 Mc Mosbach Sands SMNS IPS96017 32850/112 Mc Mosbach Sands SMNS IPS83876 225 Mc, Mt Gobi Desert MDA E. hemionus IPS83877 381 Mc, Mt Gobi Desert MDA E. grevyi IPS84963 7111 Mt Hagenbeck Zoo ZIHU IPS116319 A16/033 Mc, Mt Réserve Africaine ICP E. quagga de Sigean IPS116321 A15/213 Mc, Mt Réserve Africaine ICP de Sigean E. zebra IPS116322 A16/003 Mc, Mt Réserve Africaine ICP de Sigean 757 758 32 759 Table 2. Adult body weight estimation (kg) in extinct Equus. n = number of samples, 760 SD = standard deviation. Metacarpus Phalanx Mean  SD Species Locality n Mean  n Mean  SD (n) SD E. steinheimensis Steinheim an der 18 478.7  36 464.7  469.3  112.6 Murr 63.1 131.2 (54) E. mosbachensis Mosbach Sands 7 595.6  82 9 616.54  607.4  207.8 275.2 (16) 761 762 33 763 Table 3. Size at birth in extant and extinct Equus. The perimeter of the neonatal line 764 (NL), considered as a proxy of the size at birth, is shown. The star (*) indicates fossil 765 species. Perimeter of the NL (mm) Species Code Metacarpus Metatarsus IPS96010 - 81.26 E. steinheimensis* IPS96011 76.55 - IPS96013 77.59 - IPS96014 - 90.86 E. mosbachensis* IPS96016 86.02 - IPS96017 83.65 - IPS83876 60.77 60.04 E. hemionus IPS83877 56.29 53.09 E. grevyi IPS84963 - 74.15 E. quagga IPS116319 69.98 68.35 IPS116321 68.70 69.13 E. zebra IPS116322 71.92 73.29 766 767 34