This is the accepted version of the journal article: Castanera, Diego; Pascual Arribas, Carlos; Canudo, José Ignacio; [et al.]. «A new look at Crocodylopodus meijidei : implications for crocodylomorph locomotion». Journal of vertebrate paleontology, (March 2022). DOI 10.1080/02724634.2021.2020803 This version is available at https://ddd.uab.cat/record/257291 under the terms of the license 1 A new look at Crocodylopodus meijidei: implications for crocodylomorph locomotion 2 3 DIEGO CASTANERA,1* CARLOS PASCUAL,2 JOSÉ IGNACIO CANUDO,3 and 4 EDUARDO PUÉRTOLAS-PASCUAL4,5 5 6 1 Institut Català de Paleontologia Miquel Crusafont, Universitat Autònoma de Barcelona, c/ 7 Escola Industrial 23, 08201 Sabadell, Barcelona, Spain; 8 2 C. Real 65, 42002 Soria, Spain, capascual-1@telefonica.net; 9 3 Aragosaurus—IUCA, Paleontología, Facultad de Ciencias, Universidad de Zaragoza, 50009 10 Zaragoza, Spain, jicanudo@unizar.es; 11 4 GeoBioTec, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa, Caparica, 12 Portugal, eduardo.puertolas@gmail.com; 13 5 Museu da Lourinhã, Lourinhã, Portugal 14 RH: CASTANERA ET AL.— CROCODYLOPODUS MEIJIDEI AND LOCOMOTION IN 15 FOSSIL CROCODYLOMORPHS 16 17 18 19 20 21 22 * Corresponding author. Email: diego.castanera@icp.cat; dcastanera@hotmail.es 23 Color versions of one or more of the figures in the article can be found online at 24 www.tandfonline.come.ujvp. 1 25 ABSTRACT—A review of the type material of the crocodylomorph ichnotaxon 26 Crocodylopodus meijidei Fuentes Vidarte and Meijide Calvo, 2001 from the Berriasian of 27 Spain is carried out. The review allows a better characterization of this type ichnotaxon and 28 provides interesting new data on the candidate trackmakers and especially on their 29 locomotion. Three different size classes possibly related to different ontogenetic states or 30 sexual dimorphism of the same small to medium-sized crocodylomorph trackmaker are 31 distinguished. Morphological differences within the sample such as digital impression lengths 32 might be a consequence of differences in allometric growth, assuming similarities with extant 33 crocodylians. Other differences are a consequence of variation in the morphological quality 34 and mode of preservation across the sample. Some trackway features (intermediate-gauge 35 trackways with high pace angulation, absence of tail, belly or drag marks) indicate the 36 trackmakers, presumed neosuchian crocodylomorphs, were walking in a “high-walk” mode 37 with a semi-erect posture at a moderate speed. The trackmaker may have walked with more 38 erect limb posture and with the center of mass located more anteriorly than occurs in extant 39 species, albeit not as erect as quadrupedal animals such as mammals or other extinct 40 archosaurs including trackmakers of other crocodylomorph ichnotaxa (e.g., Batrachopus). 41 42 SUPPLEMENTAL DATA— Supplemental materials are available for this article at 43 www.tandfonline.com/XXXX. 44 45 INTRODUCTION 46 47 Crocodylomorphs were an abundant component of vertebrate assemblages throughout 48 the Mesozoic. For more than 220 million years from the Late Triassic to the present, 2 49 crocodylomorphs have occupied a variety of habitats. Some clades contained completely 50 aquatic or marine forms (e.g. Thalattosuchia and Tethysuchia), and others were fully 51 terrestrial (e.g. Notosuchia, Sphenosuchia and Protosuchia), whereas many others (mainly 52 Neosuchia) had a freshwater, semi-aquatic mode of life (Benton and Clark, 1988; Brochu, 53 2003; Pol et al., 2009; Bronzati et al., 2015; Wilberg et al., 2019). The crocodylomorph track 54 record is relatively scarce compared with the osteological record, and only a few ichnotaxa 55 attributed to crocodylomorphs have been described. Hitchcock (1845) was first to identify 56 crocodylomorph tracks (Batrachopus) in the Lower Jurassic of the USA.Since then, the 57 description of new tracks of extinct crocodylomorphs has increased considerably (see Milàn 58 et al., 2010; Lockley et al., 2020; Kim et al., 2020; Masrour et al., 2020 and references 59 therein) including several reports in the Iberian Peninsula (Vila et al., 2015; Segura et al., 60 2016; Castanera et al., 2021). Crocodylomorph tracks are well known from the Lower 61 Jurassic to the Cenozoic (Klein and Lucas, 2010; Lockley and Meyer, 2004; Lockley et al., 62 2010b) with three main crocodylomorph ichnotaxa being the most significant in terms of the 63 number of reports: Batrachopus (mainly Lower Jurassic - Cretaceous), Crocodylopodus 64 (mainly Upper Jurassic- Cretaceous) and Hatcherichnus (mainly Upper Jurassic-Cretaceous, 65 see Lockley et al., 2020; Kim et al., 2020; Masrour et al., 2020 and references therein). The 66 ichnotaxon Crocodylopodus meijidei from the Huérteles Formation (Soria, Spain) is the type 67 of the ichnogenus Crocodylopodus and is thus a key ichnotaxa during the Mesozoic (Fuentes 68 Vidarte and Meijide Calvo, 2001). Since its description, new materials related to 69 Crocodylopodus have been described, especially from Middle–Upper Jurassic and Lower 70 Cretaceous localities of Morocco, Spain, Korea and Iran (Avanzini et al., 2007, 2010; Abbassi 71 et al., 2015; Klein et al., 2018; Lockley et al., 2020). Recent studies are providing excellent 72 information that aids our understanding of how extant crocodylians move and can help us 3 73 interpret the stance and gait of extinct crocodylomorphs (Houck et al., 2010; Farlow et al., 74 2018a, 2018b; Hutchinson et al., 2019). Two features seen in trackways assigned to 75 Batrachopus and Crocodylopodus that differ from extant ones are their narrowness (autopods 76 located close to the trackway midline) and the absence of tail traces (Masrour et al., 2020; 77 Lockley et al., 2020 and references therein), suggesting differences in limb posture during the 78 locomotion. 79 The Lower Cretaceous Huérteles Formation in Soria is one of the key Mesozoic 80 formations to understand the crocodylomorph footprint record, since several sites with 81 crocodylomorph tracks have been reported (e.g.: Pascual Arribas et al., 2005; Hernández 82 Medrano et al., 2008). During a review of the Crocodylopodus meijidei collection in the 83 Museo Numantino de Soria (Spain), we noticed certain ichnotaxonomic issues related with 84 the original description of the type material. Furthermore, the collection includes undescribed 85 materials. The aims of this paper are multiple. Firstly, to review and describe all the 86 Crocodylopodus meijidei material housed in the Museo Nunmantino de Soria (Spain). 87 Secondly, to resolve the ichnotaxonomic issues and emend the diagnosis for the type material 88 through comparisons with other tracks assigned to Crocodylopodus, other extinct 89 crocodylomorph ichnotaxa, and extant crocodylian footprints. Thirdly, to reconstruct limb 90 posture of the trackmaker based on trackway parameters. Finally, to provide an overview of 91 the candidate trackmakers for Crocodylopodus meijidei. 92 Institutional Abbreviations—MNS, Museo Numantino de Soria (Numantine 93 Museum of Soria), Spain. 94 95 GEOGRAPHICAL AND GEOLOGICAL SETTING 96 4 97 The type material of Crocodylopodus meijidei comes from a site close to the El 98 Frontal and Fuente Lacorte tracksites (Fuentes Vidarte and Meijide Calvo, 2001; Razzolini et 99 al., 2014) located in the village of Bretún (Fig. 1), which lies within the region of Tierras 100 Altas in the northern part of the province of Soria (Spain). The area has been well known 101 from an ichnological point of view since the 1980s and especially since the geotourism 102 project “Ruta de las icnitas de Soria” (Ichnite Route of Soria) was launched (see Hernández 103 Medrano et al., 2008; Castanera et al., 2018 and references therein). Geologically, the Tierras 104 Altas region is part of the eastern Cameros Basin. A detailed description of the geological 105 setting of the El Frontal tracksite can be seen in Razzolini et al. (2014). In summary, these 106 tracksites belong to the Huérteles Formation (Fig. 1), which is included in the Oncala Group 107 as a part of depositional sequence 3 of the infill of the Cameros Basin (Gómez-Fernández and 108 Meléndez, 1994a; Quijada et al., 2013; Mas et al., 2019). This formation comprises mainly 109 siliciclastic deposits and was deposited in broad, low-gradient tidal flats, traversed by 110 meandering channels (Quijada et al., 2013; but see also Gómez-Fernández and Meléndez, 111 1994b). The age of the Huérteles Formation is Berriasian according to ostracods and 112 charophytes (Gómez-Fernández and Meléndez, 1994a; Schudack and Schudack, 2009; Mas et 113 al., 2019). The slabs that preserved the crocodylomorph tracks are siltstones to very fine- 114 grained sandstones. 115 116 MATERIALS AND METHODS 117 118 The material is housed in the MNS. The Crocodylopodus meijidei collection 119 comprises 10 slabs: 2002/96/2bis, 2002/96/3, 2002/96/4, 2002/96/5, 2002/96/6, 2002/96/7, 120 2002/96/8, 2002/96/10, 2002/96/12, 2003/92/8. The acronym MNS precedes the registration 5 121 number and refers to the museum. The word “bis” after the number is used by the museum to 122 distinguish among registration numbers. Slab 2002/96/12 preserves one set of coupled manual 123 and pedal tracks and one trackway made by different trackmakers, which are hereafter 124 referred to the registration number plus t1 and t2, respectively. Slab 2003/92/8 preserves two 125 trackways at the upper and lower surface of the slab (but it is unkown which surface is the 126 base and which one the top); these are here after referred to as 2003/92/8a and 2003/92/8b. 127 2002/96/7 and 2002/96/8 are part and counterpart, as possibly 2002/96/10 and 128 MNS2003/92/8b are as well. Fuentes Vidarte and Meijide Calvo (2001) proposed three 129 holotypes (trackways MNS2002/96/2bis, MNS2003/92/8a and MNS2002/96/4), and as 130 paratypes they proposed the set of coupled manual and pedal tracks MNS2002/96/3 and the 131 “isolated footprints in the tracksite” (see Remarks section for clarification within the context 132 of the ICZN rules). 133 We reviewed all the material housed at the MNS and took photographs of each slab 134 with a Sony Alpha 5100. From sets of 20-48 pictures we constructed photogrammetric 3D 135 models of each slab using the software 3DF Zephyr Free version 4.530 136 (https://www.3dflow.net/3df-zephyr-free/) and Agisoft Metashape Standard Edition. 137 Subsequently, the 3D models were processed in CloudCompare (v.2.7.0) in order to obtain 138 false-color depth maps. The photogrammetric 3D model of the lectotype is available for 139 download in the Supplemental data, following the recommendations of Falkingham et al. 140 (2018). 141 Measurements for individual tracks were taken (Fig. 2, Table 1, S1) for the footprint 142 length (FL), footprint width (FW), the length (LI, LII, LIII, LIV, LV) and width (WI, WII, 143 WIII, WIV, WV) of the digital impressions, interdigital impression angles (I^II, II^III, III^IV, 144 IV^V) and manual–pedal impression distance (Dm–p). The total interdigital divarication was 6 145 judged to be either low (IA < 30º), medium (30º-60º) or high (IA > 60º) on the basis of the 146 published data for crocodylomorph footprints (extant and extinct). Individual digital 147 impressions are referred to as DI, DII, DIII, DIV and DV. Trackway parameters were 148 measured for pace length (PL), stride length (SL), pace angulation (PA, center of the 149 footprint; ANG, tip of the impression of digit III), footprint rotation (FR), outer width of the 150 trackway (OW). Heteropody was determined on the basis of the heteropody index (HI), 151 calculated as HI = FL x FW of the manual impression/ FL x FW of the pedal impression x 152 100. The heteropody was accordingly considered either pronounced (HI < 35%), medium (35- 153 70%) or low (HI > 70%) on the basis of the published data for crocodylomorph tracks. 154 Masrour et al. (2020) recently characterized trackway gauge in crocodylomorphs on the basis 155 of Ar/FW, where Ar is the distance from center of the track to the midline. We have used here 156 the following categories: narrow (Ar/FW < 0.5), intermediate Ar/FW (0.5-1) and wide 157 (Ar/FW >1). Measurements were taken from the 3D models using the software ImageJ. The 158 morphological preservation (MP) of each specimen was calculated according to Marchetti et 159 al. (2019) and following their recommendations only tracks with its MP scale values higher 160 than 2 were used for ichnotaxonomy. The letters m and p are used in the description of each 161 specimen and in the tables to distinguish between the manual and pedal tracks. ML refers to 162 the trackway midline. The glenoacetabular distance was estimated following Leonardi (1987) 163 and Farlow et al. (2018b). A review of the main crocodylomorph tracks suggested the 164 following size classes on the basis of footprint length: small < 5 cm; medium 5-10 cm; large 165 10-20 cm; and very large > 20 cm. Data for comparisons among ichnotaxa were taken or 166 estimated from the descriptions and outline drawings in the original publications. 167 168 SYSTEMATIC PALEONTOLOGY 7 169 170 Ichnogenus CROCODYLOPODUS Fuentes Vidarte and Meijide Calvo, 2001 171 Emended Diagnosis—star-shaped pentadactyl manual prints with slender digital 172 impressions whose lengths vary as follows: III≥ IV=II> V=I. Pedal track with digital 173 impressions with the following length variations DIII> DII≥ DIV>DI. DIII is clearly the 174 longest and DI the shortest. Interdigital divarication varies from medium to high. Pronounced 175 to medium heteropody (HI = 30-40%). Manual tracks are laterally rotated whereas the pedal 176 tracks are slightly medially rotated. Intermediate-gauge trackway. Absence of tail, belly or 177 any other drag marks. 178 179 CROCODYLOPODUS MEIJIDEI Fuentes Vidarte and Meijide Calvo, 2001 180 (Figs 3, 6) 181 Lectotype— MNS2002/96/2bis 182 Paralectotypes— MNS2003/92/8a; MNS2002/96/4 (see descriptions in Supplemental 183 Data). 184 Referred Specimens— MNS2002/96/3, MNS2002/96/5, MNS2002/96/6, 185 MNS2002/96/7, MNS2002/96/8, MNS2002/96/10, MNS2002/96/12, MNS2003/92/8b (see 186 descriptions in Supplemental Data). 187 Locality, Horizon, and Age—Bretún, close to the El Frontal tracksite (Soria), 188 Huérteles Formation (Berriasian). 189 Diagnosis—as for the ichnogenus 190 Description— MNS2002/96/2bis. This specimen is one of the holotypes (Trackway 191 A, fig. 1 and fig. A in Fuentes Vidarte and Meijide Calvo (2001) and is the holotype 192 according to Lockley and Meyer (2004). The specimen (Fig. 3) includes four sets of coupled 8 193 manual and pedal tracks (Fuentes Vidarte and Meijide Calvo, 2001, also draw one isolated 194 manual print partially preserved at the beginning of the trackway that is not clearly identified 195 here). The tracks are preserved as true tracks (or very shallow undertracks). Digital pads 196 cannot be recognized but other details such as claw marks are clearly discernible. Some tracks 197 (e.g. 3m) still preserve part of the overlying layer inside them. The MP value is quite variable 198 (1–2.5) along the trackway, with manual-pedal set 3 (Fig. 3D, 3E) showing the highest MP 199 (2.5). This is a small- to medium-sized specimen (Pedal FL= 4.6–5.1 cm; Pedal FW = 3.3–3.9 200 cm). The manual prints are pentadactyl, star-shaped and wider than long (FL = 1.5–2.2 cm; 201 FW = 2.5–3 cm; FL/FW ratio = 0.6–0.75). The digital impressions are noticeably thin (WI- 202 WV 0.3–0.4 cm), with an apparent acuminate end in the fifth digital impressions (no clear 203 differences between DI-DIII and DIV-DV). DIII is the longest (1.7–2.2 cm), DII and DIV are 204 slightly shorter and similar (but variable) in length, whereas DI and DV are the shortest and 205 also of similar length. I^II is the lowest angle (36–41º), the other angles (II^III, III^IV, IV^V) 206 being higher and variable (43–66º). The total divarication in the manual track is high (IA = 207 209º–218º). Generally, DI-DII and DIV-DV are oriented medially/antermedially and 208 laterally/posterolaterally respectively, and DIII has an anterior orientation. DI and DV are to a 209 large extent point in opposite directions. These orientations are variable because of the 210 variability of the footprint rotation, which is lateral (15º–36º outwards) in all the manual 211 prints. No clear claw marks are identified in the manual tracks. The pedal tracks are 212 tetradactyl, subtriangular in shape, and longer than wide (FL/FW ratio =1.3–1.42). The central 213 digital impressions (DII and DIII) are longer than the lateral and medial ones. Specifically, 214 DIII is the longest (4.6–5.1 cm); DII (4.5–4.7 cm) and DIV (4.1–4.2 cm) are slightly shorter, 215 with DII clearly longer than the latter. DI is the shortest, being considerably shorter (3.2–3.5 216 cm). The digital impressions are thin and of variable width (WI-WIV = 0.3–0.6 cm). The four 9 217 of them have an acuminate end, showing clear evidence of claw marks associated with the 218 first three digits, DIV having a less acuminate end as seen in 2p and 3p (Fig. 3D, 3E). No 219 evidence of the claws digging into the substrate. The orientation of the digital impressions is 220 medial (DI and DII), anteromedial (DIII) and anterolateral (DIV), with an average total 221 divarication of 56–57º. I^II (10–15º) is the lowest angle, II^III (21–24º) and III^IV (18–24º) 222 being quite similar. The heel impression is oval to subtriangular and quite symmetric, and is 223 shallower than the anterior part of the footprint. Pedal rotation is low (8–20º) and medial 224 (inwards). No clear evidence for the presence of webbing in both manual and pedal tracks. 225 The trackway is intermediate-gauge (Ar/FW = 0.58-0.62). The trackway shows an irregular 226 gait, which might be associated with a slight change in the direction of travel or some 227 sinusoidal movement as a consequence of swaying during the walk cycle. The manual–pedal 228 impression distance is short 3.8–4.5 cm. PL shows few variations, with relatively similar 229 values between the manual and pedal tracks (9.5–11.6 cm). SL is also similar for both the 230 manual and pedal tracks (17–20 cm). Pace angulation is high but variable (PA = 118–137º 231 and ANG = 116–137º for the pedal tracks; PA = 105–140º and ANG = 102–129º for the 232 manual tracks). The heteropody varies from pronounced to medium values, with HI varying 233 from 21–33%. The lower values are related to the lower MP value of some manual prints 234 (1mMP = 1.5, showing slightly collapsed sediment). The manual-pedal track ratio is 235 approximately 1:3. The manual prints are deeper than the main area of the pedal prints, 236 especially in the first two manual-pedal sets, the anterior part of the digital impressions 237 having similar depth. There is no evidence of overprinting of the manual impression or of tail 238 or belly drag marks. The estimated glenoacetabular distances range from 11.5 to 14.2 cm. 239 Remarks—Lockley and Meyer (2004) noted that three holotypes (Rastro A, B and C 240 = MNS2002/96/2bis, MNS2003/92/8a and MNS2002/96/4, respectively) were designated in 10 241 the original description by Fuentes Vidarte and Meijide Calvo (2001) and that such a 242 procedure is not permitted by the ICZN, so they selected “Rastro A” as the holotype and 243 designated “Rastro B” and “Rastro C” as paratypes. According to the ICZN, however, the 244 holotype “can only be fixed in the original publication and by the original author” (Article 245 73.1.3). The ICZN thus recommends the designation of “a lectotype rather than (assuming) a 246 holotype” (Recommendation 73F). Accordingly, here we designate specimen 247 MNS2002/96/2bis (Fig. 3) as a lectotype. On the other hand, an “author who designates a 248 lectotype should clearly label other former syntypes as “paralectotypes” (Article 74F), and we 249 thus designate specimens MNS2003/92/8a (Fig. 4A-C) and MNS2002/96/4 (Fig. 4D-F) as 250 paralectotypes. In the original diagnosis proposed by Fuentes Vidarte and Meijide Calvo 251 (2001) and the revised diagnosis proposed by Lockley and Meyer (2004), we have found 252 some issues that have led us to propose an emended one. Although Fuentes Vidarte and 253 Meijide Calvo (2001) proposed MNS2002/96/3 and the isolated tracks in the tracksite as 254 paratypes, these are not considered here to be the paralectotypes. This is because they 1) show 255 some features that are slightly different from the lectotype and paralectotypes; 2) the MP 256 value is rather low; and 3) they are isolated manual prints. 257 Fuentes Vidarte and Meijide Calvo (2001) suggested that the Crocodylopodus meijidei 258 material was different enough to define the new ichnofamily Crocodylopodidae. Lockley and 259 Meyer (2004) also noted the differences between Batrachopodidae and Crocodylopodidae 260 (slenderness of the digital impressions on both the manual and pedal tracks, divarication 261 angles, especially in the pedal tracks, pace angulation and footprint rotation), but judged that 262 there were not enough differences to define a new ichnofamily. Accordingly, they included 263 Crocodylopodus meijidei in Batrachopodidae Lull, 1904 and synonymized Crocodylopodidae 264 with Batrachopodidae (Lockley and Meyer, 2004:177). The authors criticized the fact that 11 265 Fuentes Vidarte and Meijide Calvo (2001) had made a tri-level monospecific diagnosis for 266 ichnofamily, ichnogenus and ichnospecies. Although the procedure might be not correct, 267 Crocodylopodus meijidei does not fit in the revised diagnosis of Batrachopodidae proposed by 268 Lockley and Meyer (2004) since they differ in several features; in tracks assigned to 269 ichnotaxa within Batrachopodidae, for example, the digital impression lengths II and IV in the 270 pedal tracks are generally not subequal in length, the manual prints do not show lateral 271 rotation and the pace angulation is considerably lower (and the trackway is narrower). Kim et 272 al. (2020:5) recently proposed that differences between Batrachopus and Crocodylopodus 273 “may be explained in part by differential preservation”. Thus, many of the differences 274 between the two ichnotaxa are likely to be a consequence of differences relating to the 275 different modes of locomotion of their respective trackmackers rather than just preservational 276 factors. 277 278 DISCUSSION 279 280 Morphological Variations in the Type Material of Crocodylopodus meijidei 281 The sample shows some differences among the various specimens, which are related 282 to divergent MP values (variation from 0.5 to 2.5) and the mode of preservation of the tracks 283 (either as epireliefs or hyporeliefs). Differences in the size of various tracks across the sample 284 suggest that they were produced by different individuals. The lectotype (MNS2002/96/2bis, 285 Fig. 3) and paralectotypes (MNS2003/92/8a and MNS2002/96/4, Fig. 4) as well as 286 MNS2003/92/8b and MNS2002/96/10 (Fig. 5D-I) and MNS2002/96/12t1 (Fig. 6A-C) are 287 more or less similar in size and fall between the small and medium-sized categories. The 288 similarities among the values of these specimens raise the hypothesis that some of them could 12 289 even be parts of the same trackway or produced by the same individual. The former two have 290 a pedal FL (see Table 1 and S1) of around 5 cm, whereas the latter four have lower pedal FL 291 (close to 4 cm) but lack the preservation of the heel mark (so all of them can be considered 292 medium-sized). This similar size is corroborated by a similar pedal FW (close to 4 cm) and 293 similar FL (around 2 cm) and FW (between 2.5 and 3 cm) in the manual track among all the 294 specimens. MNS2002/96/7 and MNS2002/96/8 (Fig. 6E–6F) also fall within this medium- 295 sized category (manual FL around 2 cm). Since MNS2003/92/8a and MNS2003/92/8b are 296 preserved in the same slab, they represent two different trackways (and slightly different 297 stratigraphic levels). These data indicate that there are at least two different trackmakers that 298 fall within the medium-sized category. MNS2002/96/3 (Fig. 5A-C) is the largest specimen 299 (Pedal FL = 7 cm), whereas MNS2002/96/5 (Fig. 6D) is a manual print that is similar in size 300 to the manual track of MNS2002/96/3. Thus, there is at least one other medium-sized (but 301 larger) trackmaker. MNS2002/96/12t2, is the smallest track (Pedal FL = 2.5 cm) in the whole 302 sample, and represents one trackmaker of the small-sized category. Accordingly, there were at 303 least four different individuals: a small individual (MNS2002/96/12t2), at least two medium- 304 sized individuals (MNS2002/96/2bis, MNS2003/92/8a, MNS2003/92/8b, MNS2002/96/10, 305 MNS2002/96/4, MNS2002/96/12t1; MNS2002/96/7-MNS2002/96/8), and at least one 306 medium-sized but larger individual (MNS2002/96/3 and MNS2002/96/5). 307 The lectotype (Fig. 3, MNS2002/96/2bis), one of the paralectotypes (Fig. 4A-C, 308 MNS2003/92/8a), and the pedal impression in MNS2002/96/3 (Fig. 5A-C) are those with 309 high MP values. Interestingly, the manual tracks generally have higher MP values than the 310 pedal tracks and in many specimens are deeper (similar depth just to the anterior part of the 311 pedal impression). MNS2002/96/3 is the only specimen that has a clearly deeper pedal than 312 manual impression. Other major morphological differences among the specimens are related 13 313 to the preservation, such as the presence/absence of a heel impression, slight variations in 314 heteropody, variations in interdigital divarication angles (from medium to high), or the 315 manual impression sometimes seeming tridactyl/tetradactyl instead of pentadactyl. Specimens 316 preserved as natural casts have lower MP values; the claw marks are not clearly identified; the 317 divarication angles are higher; and the length of DI and DIV in the pedal impressions are 318 more similar, but this might be a consequence of the absence of the heel mark impression and 319 thus it is difficult to measure correctly. 320 Other considerable differences among the specimens are the relative lengths of the 321 digital impressions. Padian and Olsen (1984) warned of the possible allometric changes in 322 footprints from ontogenetic and phylogenetic aspects. Possible differences due to allometric 323 growth were proposed for C. meijidei by Lockley and Meyer (2004: 176), who, on the basis 324 of the sketches by Fuentes Vidarte and Meijide Calvo (2001), calculated a lower heteropody 325 for MNS2002/96/3 (the largest) and proposed “an allometric increase in the relative size of 326 the pes during growth”. According to our data, the heteropody in specimen MNS2002/96/3 is 327 dubious because of the low MP value of the manual print, but our estimated values are not 328 very different from the lectotype (HI = 29% and 28–33%) or from the other specimens (see 329 Table S1). What is more intriguing is that on MNS2002/96/3, the impression of digit IV is of 330 similar length to the impression of digit II, a feature not apparent in the other medium-sized 331 but smaller specimens. This specimen is slightly larger (2 cm longer in FL) than the others, 332 but these differences among the specimens could be explained by possible ontogenetic 333 differences as a consequence of allometric growth (cf. Lockley and Meyer, 2004). Notably, 334 Farlow and Britton (2000:189) pointed out possible changes in autopodial lengths with body 335 size in Alligator mississippiensis (“with increasing body size, hind limb and autopodial 336 lengths become shorter relative to the shoulder-hip length and the pes become shorter”). 14 337 Subsequently, Farlow (2018) analyzed the proportions of pedal skeletons in alligators, 338 suggesting that young alligators posess relatively longer digits and feet than adult specimens. 339 Assuming proportional changes during growth similar to those of an extant species such as 340 Alligator mississippiensis, the possibility of variations in digit proportions should be taken 341 into account when analyzing C. meijidei material. A comparison of the foot proportions 342 reveals the FL/FW ratio in the pedal impression of MNS2002/96/3 (1.25) to be slightly lower 343 than in the lectotype (1.3–1.42) but within the range of variation seen in the paralectotype 344 MNS2003/92/8a (1.19–1.3). Moreover, the other specimens with low MP values also show 345 lower values for the FL/FW ratio. Thus, we consider that these differences in relative size are 346 not very significant. A comparison of the relative lengths of the digital impressions in 347 MNS2002/96/03 (DI, DII, DIII, DIV = 0.68, 0.9, 1, 0.88), the lectotype (3p; DI, DII, DIII, 348 DIV = 0.62, 0.9, 1, 0.88) and the paralectotype (2p; DI, DII, DIII, DIV = 0.72, 0.9, 1, 0.8) 349 shows the relative lengths of the digital impressions are very similar, except for the DI. The 350 MP value of the smallest specimen (MNS2002/96/12t2) is rather low and DIV is not 351 preserved, so a comparison of the FL/FW ratio and the fourth digital impression is not 352 possible. However, the other three digital impressions (B1p; DI, DII, DIII = 0.6, 0.84, 1) show 353 some difference in DI, although smaller in DII. These differences between lengths in digital 354 impressions might be explained by allometry. 355 Another possible explanation for these differences between relative lengths of the 356 digital impressions, especially between MNS2002/96/3, the lectotype (3p) and the 357 paralectotype (2p), which are the specimens with the highest MP values, might be just 358 preservational factors. Extant crocodylian pedal prints show claw marks in digits I, II and III 359 that sometimes dig into the substrate, thus not reflecting real anatomical lengths (Farlow et 360 al., 2018b). Furthermore, the absence of a claw mark in digit IV might also produce 15 361 differences in relative digit lengths, since this digit may be less clearly marked in the 362 sediment. Other possible differences could be associated with different kinematics and 363 behavior during locomotion, but these are more difficult to analyse across the sample since 364 there are only a few short trackways. For instance, there are slight variations in the trackway 365 gauge (narrower in the lectotype than in the paralectotypes) that are common in other 366 quadrupeds (e.g. Castanera et al., 2012). The narrownes of the lectotype might be caused by a 367 slight change in direction or with the swaying as a consequence of the lateral movement 368 (Carpenter, 2009) rather than by different speed, since the stride lengths are very similar in all 369 three specimens. The manual-pedal distances are also similar among the specimens (3.5–4.5 370 cm), although in the largest specimen it is slightly larger (9 cm). An alternative explanation 371 for the size classes might be sexual dimorphism since in extant crocodylians adult males can 372 be 20–40% larger than adult females. This difference is more marked in larger than in smaller 373 species where this difference is not as pronounced or even females can be slightly larger 374 (Thorbjarnarson, 1994; Cox et al., 2007; Platt et al., 2009; Hone et al., 2020). 375 In summary, the C. meijidei type material shows a series of morphological differences 376 among specimens that can be mainly explained in terms of two different factors: 1) different 377 ontogenetic states (variations in lengths of the digital impressions and size differences); 2) 378 differences in the preservation of the specimens (e.g., natural cast vs true tracks, 379 absence/presence of heel mark impressions, absence/presence of claw marks, 380 absence/presence of certain digital impressions associated with both the manual and pedal 381 tracks). Besides, possible differences due to sexual dimorphism (size variations) and 382 kinematics and behavior during locomotion (variations in certain features or parameters) may 383 have also played a role. 384 Comparisons with Other Tracks Assigned to Crocodylopodus 16 385 A comparison of C. meijidei with other Crocodylopodus material is necessary to 386 understand possible variations (Fig. 7, Table S2). The oldest Crocodylopodus reports are from 387 the Middle Jurassic of Iran (Abbassi et al., 2015) and the Middle-?Upper Jurassic of Morocco 388 (Klein et al., 2018). Abbassi et al. (2015) reported a small-sized trackway classified as 389 Crocodylopodus isp. (Fig. 7B). This trackway differs from C. meijidei in a number of ways. 390 Firstly, the manual morphology, which is tetradactyl with the four digital impressions 391 anteriorly directed. Secondly, the pedal track differs in having a rounded heel impression. The 392 trackway also preserves tail drag impressions. Klein et al. (2018) described medium-sized 393 tracks assigned to C. meijidei (Fig. 7C) from the Middle-?Upper Jurassic of Morocco. They 394 display considerable differences in the pedal impressions, such as widely divaricated digital 395 impressions, an elongated heel, and differences in relative lengths of the digital impressions. 396 The manual prints are also star-shaped, but DI and DV are not located as posteriorly as in the 397 type of C. meijidei. Upper Jurassic beds of the Asturian basin of the Iberian Peninsula have 398 also produced material related to Crocodylopodus (Avanzini et al., 2007; 2010). Among these 399 tracks, MUJA 0101 is small-sized (Fig. 7F) and mainly differs from C. meijidei in the 400 digitigrade pedal impression, with lower FL/FW ratio, lower interdigital divarication, 401 evidence of drag marks or the absence of manual impression. MUJA 0102 is small in size 402 (Fig. 7G–7H) and main differences include the preservation of phalangeal pads in the pedal 403 impressions and the manual prints generally tridactyl, showing almost no rotation. C. meijidei 404 bears some similarities to MUJA0038, a large-sized specimen (Fig. 7D) which pedal 405 impression has digital impressions II and IV subequal in length (and shorter than DIII) and 406 extremely widely divaricated. But the heteropody of this specimen is slightly lower. Avanzini 407 et al. (2010: 243) also studied other specimens and considered that the Asturian samples 408 “show similar characteristics, which are consistent with a substrate-related morphological 17 409 variation within the Crocodylopodus meijidei ichnospecies”. Recently, Castanera et al. (2021) 410 reported one isolated pedal track from the Upper Jurassic of the Lusitanian basin (Fig. 7E) 411 assigned to Crocodylopodus isp. in which the differences from C.meijidei were in the lower 412 FL/FW ratio and the wider digital divarication angle, and slightly differences in the length of 413 the digital impressions. 414 To turn to the Cretaceous occurrences of tracks related to Crocodylopodus, Pascual 415 Arribas et al. (2005) described a large-sized crocodylomorph trackway (Fig. 7I) from the 416 same formation as C. meijidei that was subsequently classified as cf. Crocodylopodus by 417 Lockley et al. (2010a) and ?Crocodylopodus by Lockley et al. (2020). The tracks show 418 considerable differences in the pedal impressions, which have digital pads, lower FL/FW 419 ratio, are laterally rotated, have slightly lower digital divarication, and DII and DIV are 420 similar in length. No clear tail marks exist, although the authors describe some traces that 421 could be tail marks. Another large sample of Crocodylopodus tracks has recently been 422 described from the Lower Cretaceous (?Aptian) of Korea (Lockley et al., 2020). The authors 423 describe several small to medium-sized trackways (Fig. 7J–7M) that show several features 424 that are different from C. meijidei especially the FL/FW ratio (varies in the Korean 425 specimens), the relative digital impression lengths and their orientation (more anteriorly 426 oriented), heteropody (much lower in the Korean specimens), the lower pace angulation and 427 wider-gauge trackway than in C. meijidei. These trackways also show no evidence of tail drag 428 marks. Only two possible reports of Crocodylopodus tracks have been described in Upper 429 Cretaceous deposits. Simpson et al. (2010) describe a single large-sized pedal track classified 430 as cf. Crocodylopodus from the Upper Cretaceous Wahweap Formation of Utah. The authors 431 suggest that the track is indistinguishable from Crocodylopodus, although they also note 432 differences in robustness and consider the specimen to be reminiscent of Hatcherichnus as 18 433 well. Recently, Lockley et al. (2020) have proposed that it could be assigned to 434 Hatcherichnus or cf. Hatcherichnus. Noteworthy differences are the orientation of the digital 435 impressions (more anterior), relative lengths in digital impressions and the rounded heel 436 impression. Finally, Vila et al. (2015) report a single small-sized track (Fig. 7N) classified as 437 cf. Crocodylopodus from the Upper Cretaceous of Spain. This is an isolated pedal track that 438 shows similarities in the symmetrical and triangular heel impression but also shows 439 differences in that DIV seems longer than DII and has a very lateral orientation thus showing 440 high intedigital divariaction. 441 As expounded in the previous paragraphs and shown in the Table S2 there are 442 considerable differences among the tracks assigned to Crocodylopodus. Main differences are 443 in manual impression morphology (which varies from tridactyl to pentadactyl), in the pedal 444 impressions (FL/FW ratio, the length of digital impressions, heel morphology or in digital 445 divarication) and in heteropody (variation from pronounced to medium). Some of these 446 morphological variations are a consequence of possible anatomical differences that might also 447 be ontogenetically influenced, as noted above. Nonetheless, as specified by Avanzini et al. 448 (2010), many differences can be related to the state of the substrate and the preservation, such 449 as variations in digital impressions (e.g., slender or stout), absence of certain digital 450 impressions, digital divarication (higher in soft substrates), the morphology of the heel mark 451 impression (from rounded to triangular but sometimes absent) that affect variations in FL/FW 452 ratio, and the absence/presence of scale marks. Finally, other characters are linked to the 453 locomotion (see next sections). These are the narrowness/width of the trackway, 454 absence/presence of tail drag marks, lateral/medial rotation of the footprints. Taking into 455 account the differences set forth in this section, the influence of substrate, locomotion and 456 anatomical differences, and given the variation seen in the type material of C. meijidei and in 19 457 the other Crocodylopodus tracks, it is difficult to ascertain whether C. meijidei is a 458 monospecific ichnotaxon, or whether some of the tracks referred to Crocodylopodus might be 459 a different ichnospecies from that represented by type material. Thus, the variations seen 460 among the samples and in many cases the poor morphological quality of the specimens or the 461 absence of abundant material justifies previous asignments taken by other authors who have 462 classified (see references in Table S2) some tracks either as cf. Crocodylopodus isp. or 463 Crocodylopodus isp. It is noteworthy that no unequivocal reports of Crocodylopodus have 464 been reported from the Upper Cretaceous (see Table S2) and that many of the specimens not 465 classified to ichnospecies level are medium to large sized often with wide gauge trackways 466 (Table S2). All the material assigned to C. meijidei is produced by small to medium-sized 467 individuals, and are (with the exception of one report from Morocco, Klein et al. 2018) 468 restricted to the Upper Jurassic and the Lower Cretaceous (Berriasian) of the Iberian 469 Peninsula. Thus, there is the possibility that more than one ichnospecies may exist taking into 470 account differences in size, FL/FW ratio, heteropody or type of trackway although currently 471 there are not enough data (see discussion) to distinguish between them. 472 Comparisons with Other Crocodylomorph Ichnotaxa 473 Crocodylopodus is clearly distinct from the other crocodylomorph ichnotaxa. Kim et 474 al. (2020) recently summarized which crocodylomorph ichnotaxa comprise walking traces 475 and which comprise swimming traces. Crocodylopodus is clearly distinct from all the 476 ichnotaxa that represent swimming traces (e.g.: Hatcherichnus Foster and Lockley, 1997; 477 Kuangyuanpus and Laiyangpus Lockley et al., 2010a; Albertasuchipes McCrea et al., 2004; 478 Indosuchipes Rajkumar et al., 2015 and Anticusuchipes Mustoe, 2019). Of all the 479 crocodylomorph ichnotaxa, it is interesting that only Batrachopus and related ichnotaxa (e.g.: 480 Antipus Coombs, 1996 and Angolaichnus Mateus et al. 2017), Crocodylopodus and Mehliella 20 481 represent walking traces (Fig. 8). The main differences between Batrachopus (Fig. 8B) and 482 Crocodylopdus are the lower heteropody, wider pedal impression, more slender and divergent 483 digital impressions, lower lateral rotation of the pedal impressions and greater lateral rotation 484 of the manual prints in C. meijidei. Besides, in Batrachopus the digital impressions are 485 straight, and the interdigital divarication angles for the pedal tracks are very low (25º–30º in 486 the type specimen). Another notable morphological difference is the mark of digit V (when 487 present) in the pedal impression and generally narrower gauge trackway in Batrachopus 488 (Olsen and Padian, 1986; Lockley and Meyer, 2004; Masrour et al. 2020). Olsen and Padian, 489 (1986) considered Antipus a synonym of Batrachopus, although this view has not been 490 adopted by other authors (Coombs, 1996; Lockley and Meyer, 2004). C. meijidei differs from 491 Antipus (Fig. 8C) in the pedal (shorter DI impression and lower digital divarication in 492 Antipus) and manual morphology (DI-DV medially/laterally in Antipus) and the lateral 493 rotation of both manual (strongly rotated in Antipus) and pedal prints. Angolaichnus from the 494 Lower Cretaceous of Angola (Mateus et al., 2017, Fig. 8D) is also different showing a manual 495 impression that is functionally tetradactyl and plantigrade with extreme lateral rotation, and a 496 pedal impression with different digital impression lengths, DIV being the shortest. It also has 497 different interdigital divarication (lower in Angolaichnus), digit orientation (digits II and III 498 being bent slightly lateral) and higher pace angulation. C. meijidei is also clearly different 499 from Mehliella (Mehl, 1931; Lockley, 2010, Fig. 8G) which is characterized by tracks larger 500 in size, the wider trackway, with a very low pace angulation (50º), and with clear tail or belly 501 traces. Besides, it may also shows interdigital webbing traces. 502 Main differences among crocodylomorph ichnotaxa are in manual/pedal morphology, 503 lengths of digital impressions in the pes, heteropody, footprint rotation or trackway gauge. 504 Differences in certain features could be explained by substrate-related factors (e.g. slender 21 505 and divergent digital impressions, digit orientation, absence/presence of webbing). 506 Nonetheless, several differences among the ichnotaxa at the ichnogenus level are better 507 explained by anatomical factors (heteropody, digit divarication, the presence/absence of 508 digital impression V) and especially locomotor and behavior factors (rotation of the 509 footprints, mode of locomotion, swimming/subaqueous behavior, low walk/slow high walk). 510 Although Kim et al. (2020) note that differences between Batrachopus and Crocodylopodus 511 might be preservation-related, we consider that major differences between them are related to 512 different locomotor modes employed by their respective trackmakers reflected in, the 513 trackway gauge, pace angulation and footprint rotation. 514 Comparisons with Tracks of Extant Crocodylians and notes on the Mode of Locomotion 515 Regarding the tracks of extant crocodylians there are considerable differences from 516 Crocodylopodus tracks. A summary of trackway features in extant crocodylians is provided in 517 Table S3 and sheds interesting light on the interpretation of certain features. In a general 518 comparison it should be noted: manual imprints generally have DI and DV located more 519 medially/laterally (lower interdigital divarication) with claw marks in DI-DIII, a feature not 520 clearly seen in C. meijidei (preservation bias?). Pedal impressions show differences in the 521 length of the digital impressions with the central digits (DII and DIII) being the longest but DI 522 and DIV being more similar in length and slight variations in the orientation of the digital 523 impressions. Besides, several extant species show lower interdigital divarication values than 524 those of C. meijidei, these being more similar to members of Alligatoridae (higher values) 525 than to Crocodylidae (Milàn and Hedegaard, 2010; see Table S3). These variations in 526 interdigital divarication in both manual and pedal impressions might be related to the 527 development of webbing between extant crocodylians and the trackmaker of C. meijidei (it 528 would have reduced interdigital webbing) and may also explain the variations seen in 22 529 Crocodylopodus tracks from the different areas (Table S2). Trackways of extant crocodylians 530 differs from that of C. meijidei in the presence of tail, belly and drag marks, wider-gauge 531 trackways (with lower pace angulation), pedal prints deeper than the manual impressions, and 532 overprinting of the manus by the pes. 533 Some of the differences are anatomical (e.g., interdigital divarication angles, DIV in 534 the pes more laterally located, relative digital impression lengths, webbing development), 535 whereas others might be just preservational (e.g., the presence/absence of scale prints and 536 webbing, evidence of clear claw marks, the orientation of the digits). Many of the main 537 differences are associated with locomotion and possible differences in gait (e.g., pedal 538 impressions deeper than the manual, DI and heel deeper than the rest of the pedal impression, 539 lower pace angulation and wider trackways, belly and tail drag marks). The locomotion of 540 some extinct crocodylomorphs is different from that of extant taxa. For example, the earliest 541 members of Crocodylomorpha walked with an erect limb posture that fits well with their 542 terrestrial habits (Parrish, 1987; Salisbury and Frey, 2001; Molnar et al., 2015). On the other 543 hand, extant crocodylians use a variety of gaits: the belly walk, the high walk, and galloping. 544 Thus, in terms of locomotor posture, extant crocodylians fall between early sprawler reptiles 545 and erect dinosaurs and birds (Zug, 1974; Parrish, 1987; Gatesy, 1991; Reilly and Elias, 1998; 546 Hutchinson et al., 2019). Parrish (1987:396) suggested that the “sprawling stance used by 547 extant crocodylians can be viewed as a secondary adaptation to an aquatic existence”. Reilly 548 and Elias (1998:2559) pointed out that the crocodylian sprawl is not equivalent to the 549 primitive sprawling (seen in other reptiles), being “a lower version of a high walk”. 550 Accordingly, they named it “low walk” and suggested that crocodylomorphs do not change 551 from “a primitive sprawling posture to an intermediate semi-erect posture”, arguing that 552 “crocodilian low and high walk behaviors are not intermediate forms in the sprawling-to-erect 23 553 continuum”. Houck et al. (2010) summarized the features that characterized high-walking and 554 low-walking in extant crocodylian trackways. In the particular case of Crocodylopodus 555 trackways, many of these show the features described in the high-walking trackways (see 556 table 3 in Houck et al., 2010), although they also show some differences, such as a higher 557 pace angulation, the absence of tail and foot drag marks, and infrequent pes/manus 558 overprinting. 559 In recent years, advances have been made in studies of the limb posture and gait of 560 extinct archosaurs by analyzing the pace angulation of both fossilized and recent trackways 561 (Kubo and Benton, 2009; Kubo and Ozaki, 2009). Kubo and Benton (2009) argued that the 562 erect limb posture likely evolved during the Early Triassic, as the average pace angulation 563 value of the trackways underwent a major increase during this epoch. Kubo and Ozaki (2009) 564 demonstrated how pace angulation can be used to estimate limb posture and its relation with 565 the femoral abduction angle and pelvic rotation. Their analysis of locomotion in species of 566 extant crocodylians and lizards provided them how to reconstruct the limb posture in extinct 567 tetrapods. Differences in limb abduction are directly related to the pace angulation, which also 568 reflects differences in stride length and trackway width. Accordingly, an erect animal would 569 leave a trackway with a high pace angulation, whereas a sprawler would produce a trackway 570 with a low pace angulation. Kubo and Ozaki (2009) thus suggested that femoral abduction has 571 more influence on pace angulation than pelvic rotation does (high PA values cannot be 572 explained only by pelvic rotation) and that speed has less significant influence, although the 573 fact that only walking gaits were analyzed in that study may have affected the result. They 574 reasoned that “a trackway with an average pace angulation value of 120° or more could not be 575 produced by a trackmaker that is a true sprawler”. The authors also estimated that “at values 576 of 108° the predicted range of the femoral abduction angle did not include 0” (Kubo and 24 577 Ozaki, 2009:58). This implies that a trackway with an average pace angulation value of 108º 578 or less is unlikely to be produced by an animal with fully erect limbs in which the femoral 579 abduction angle is 0º. 580 Interestingly, the pace angluation values for Crocodylopodus trackways are very close 581 or higher than this threshold value of 108º (Tables 1, S1 and S2), an exception being the 582 trackways from Korea (Lockley et al., 2020, see Fig. 7J–7M). Avanzini et al. (2007, 2010) 583 already noted the high PA values for the tracks from Asturias and related them with different 584 styles of walking, during which the pace angulation is higher (and the resultant trackway 585 narrower) when speed increases. Tracks assigned to Batrachopus (Lockley et al., 2018; Kim 586 et al., 2020; Masrour et al., 2020) or included in Batrachopodidae (e.g. Angolaichnus, Mateus 587 et al., 2017) have even higher PA values than Crocodylopodus trackways. Regarding the data 588 for extant species the PA values (Table S3) are generally lower (variation between 75º and 589 120º in different species) than in Crocodylopodus trackways, the values of most of the studied 590 specimens being close to the upper range of the values of extant crocodylians. Variations in 591 pace angulation can be influenced by a series of factors such as posture, speed, body sized and 592 thus ontogeny and body mass or the hip and knee joint excursions or the lateral movement of 593 the body (Kubo, 2008; Kubo and Benton 2009; Carpenter, 2009; Kubo 2010). Kubo and 594 Benton (2009:1033) also suggested “body size could be an important factor in determining 595 pace angulation since modern sprawlers are small and modern erect animals are generally 596 relatively large”. Notably, Salisbury and Frey (2001) indicated that the greater the mass of an 597 animal, the more difficult sustained highwalking is likely to become. Thus, “large (>three 598 metre) extant crocodilians often seem reluctant to carry their own weight on land, and 599 sustained terrestrial locomotion appears to be a labour, only undertaken in moments of 600 extreme urgency or alarm” (Salisbury and Frey 2001:120). Although there are several 25 601 anatomical features that exert an influence but cannot be known directly from the trackways 602 (e.g. lateral body movement, hip and knee joint excursions), the Crocodylopodus tracks were 603 produced by small to medium-sized crocodylomorph trackmakers and ontogeny is not a factor 604 influencing the PA data in the type material since the four reported trackways have similar 605 pedal lengths, suggesting a small to medium-sized trackmaker. Despite the high PA values in 606 C. meijidei, it should be borne in mind that they are far from the values of completely erect 607 animals such as mammals, birds and other archosauromorphs, which have values generally 608 higher than 140–150º including several trackways assigned to Batrachopus (see Kubo and 609 Benton, 2009; Masrour et al., 2020). These variations in pace angulation are directly related 610 with the variations seen in trackway gauge seen in extinct crocodylomorph ichnotaxa where 611 Crocodylopodus trackways are characterized by intermediate-gauge trackways that are 612 narrower than extant crocodylians and some extinct crocodylomorph ichnotaxa such as 613 Mehliella but wider than Batrachopus (see Masrour et al., 2020). These data suggest different 614 postures during locomotion among extinct crocodylomorphs. 615 As regards the absence/presence of tail marks in crocodylomorph trackways, their 616 absence is noteworthy in all the Crocodylopodus trackways described in the literature (except 617 the one described by Abbassi et al., (2015)). Avanzini et al. (2007:151) suggested that this 618 absence of tail marks “suggests complete support of the whole-body during walking”. 619 Comparison with trackways of extant species reveals that many of the trackways described 620 have tail marks (Fig. 8F, Table S3). McCrea et al. (2004) noted that the absence of tail drag 621 marks could be a consequence of a variety of factors such as behavior, gait and the 622 consistency of the substrate, as well as the possibility that the animal was walking or wading 623 underwater (with floating tail). From a biomechanical point of view, tails provide semi- 624 aquatic tetrapods with propulsion during swimming, although they can compromise terrestrial 26 625 locomotion, as they have to be dragged. In extant crocodylians such as Alligator 626 mississippiensis the tail weight represents 28% of the total body mass (Willey et al., 2004). In 627 consequence, both the fore and hindlimbs have to counteract the tail’s braking effect (Willey 628 et al., 2004). The common presence of tail, belly and digital drag marks observed in extant 629 crocodylian footprints (Table S3, Fig. 8F) is likely to be related to this awkward high- 630 walking. On the other hand, there is no evidence of drag marks of any type (foot, claw, belly 631 or tail) in C. meijidei. 632 Another difference between Crocodylopodus and some modern trackways is that in 633 Crocodylopodus the manual tracks are deeper than the majority of the pedal impression area 634 (with the exception of MNS2002/96/3) with the heel traces generally absent or poorly 635 preserved (Fig. 7, Tables S2-S3). Lockley et al., (2020:5) suggested that this “raises questions 636 of whether the trackmakers exerted more pressure on the substrate with manus than pes”. A 637 crucial factor underlying the aforementioned differences is likely to be a different center of 638 mass between extant species and the trackmaker of Crocodylopodus meijidei. This would also 639 be in accordance with the absence of tail marks in the latter. Thus, in the high-walking extant 640 crocodylians the tail is dragged behind the body rather than elevated off the ground, so the 641 long, heavy tail causes the center of mass to lie more caudally, just in front of the pelvis 642 (Willey et al., 2004). Experiments with extant alligators have shown that body weight support 643 is concentrated over the hindlimb (51%) during locomotion, while the forelimbs and tail 644 support 37% and 12% of the remaining weight respectively (Willey et al., 2004; Grigg and 645 Kirshner, 2015). Therefore, the fact that the deeper areas in the trackways are in the manual 646 tracks and the anterior part of the pedal tracks seems to indicate that the producer of C. 647 meijidei would have had its center of mass more anteriorly located. This anterior displacement 648 of the center of mass could be explained by a reduction in the size of the tail (length or 27 649 weight) or by an increase in the mass of the anterior region of the body (Fig. 9). Another 650 possible factor that could cause a displacement of the center of mass is a different distribution 651 of weight among the limbs due to different forelimb/hindlimb length proportions from those 652 observed in extant crocodylians (see next section). In extant crocodylians and most fossil 653 crocodylomorphs, the hindlimbs are longer than the forelimbs (Iijima et al., 2018). Another 654 interesting feature of C. meijidei in comparison with extant crocodylian trackways is the near- 655 absence of overprinting of the manual impressions by the pes (see Table S3). These 656 overstepping is produced at moderate to higher speeds (Padian, 2003; Kubo, 2008; Milàn and 657 Hedegaard, 2010). 658 The described extant crocodylian tracks (Table S3) can give us an idea of the size of 659 the trackmaker of C. meijidei. These data give us an intimation of the total length of the 660 trackmaker of C. meijidei, which is around 50–80 cm for the lectotype and slightly greater for 661 the largest specimen, MNS2002/96/3 (Farlow et al. (2018b) proposed equations for predicting 662 the length of Crocodylus acutus on the basis of manual and pedal length proxies. Though 663 based on a different species, it will provide an indication of the total length and the shoulder- 664 hip length of the C. meijidei trackmaker. Estimations for the lectotype yield a total length of 665 78.71 cm based on the pedal impression length, and 58.59 cm based on the manual impression 666 length. The estimation for the largest specimen (MNS2002/96/3) would be 109.58 cm. On the 667 basis of the hindfoot (HF), Hutton (1987) proposed a ratio for the length of the Nile crocodile 668 of 1 : 14 where HF < 150 mm, and 1 : 13.5 where HF > 150 mm. Accordingly, C. meijidei 669 (the lectotype) would be around 71.4 cm long, and MNS2002/96/3 would be 98 cm long. 670 Farlow and Britton (2000) proposed that the total length is about four times the 671 glenoacetabular distances that gives consistent, or slightly smaller size (46–56.8 cm) for the 672 lectotype trackway. 28 673 In summary, the combination of PA data and the absence of tail and other drag marks 674 plus the presence of almost no overprint in Crocodylopodus meijidei trackways suggests that 675 the trackmaker was probably walking in a semi-erect posture in a high-walking mode and at 676 moderate speed when it produced the trackways. Moreover, the trackmaker would have 677 possibly had rather long forelimbs (at least in relation to the hindlimbs and the total length). 678 These data suggests that the trackmaker of C. meijidei walked in an agile way compared to 679 extant crocodylians (Fig. 9), possibly also on account of its small size and the fact that the 680 trackmaker was not a crocodylian sensu stricto (next section). 681 Candidate Trackmakers 682 It is difficult to assign the Crocodylopodus meijidei tracks to a concrete group of 683 crocodylomorphs since a synapomorphy-based approach (Carrano and Wilson, 2001) cannot 684 be pursued with confidence. This is because of the conservative morphology of autopods 685 through crocodylomorph history. Geological provenance and body size can also help to infer 686 trackmakers (Carrano and Wilson, 2001) but osteological fossils are almost absent from the 687 Berriasian Huérteles Fm. (Hernández Medrano et al., 2008; Castanera et al., 2018). Indeed, 688 the osteological crocodylomorph record of the Iberian Peninsula is almost absent in the 689 Berriasian. 690 Crocodylomorphs identified of the Berriasian are goniopholidids, bernissartiids, 691 pholidosaurids and “atoposaurids”. Berriasian crocodylomorph specimens of Europe mainly 692 come from the Purbeck Limestone Group in England (see Salisbury, 2002; Andrade et al., 693 2011 and references therein); Cherves-de-Cognac and Angeac-Charente in France (Pouech et 694 al., 2014; Martin et al., 2016; Rozada et al. 2020); the Rabekke Formation in Scandinavia 695 (Schwarz et al., 2009); and the Obernkirchen Sandstone in Germany (Salisbury et al., 1999; 696 Andrade and Hornung, 2011). Although there is no record in the Berriasian of Europe, 29 697 possible representatives of a clade of small crocodylomorphs, Gobiosuchidae, have been 698 recovered in the Kimmeridgian-Tithonian of Portugal (Buscalioni et al., 1996; Schwarz and 699 Fechner 2004, 2008) and in the Barremian of Spain (Buscalioni 2017), so their presence in the 700 Berriasian cannot be ruled out. However, aside from age, the extremely small size (less than 701 35 cm in total lenght) of these gobiosuchids (Buscalioni 2017) would dismiss them as the 702 putative trackmakers. Pholidosaurids are unlikely to be the producers of C. meijidei. They 703 were usually large-sized marine or freshwater aquatic animals with open, sagittally segmented 704 paravertebral shield and amphicoelous vertebrae; therefore, they probably had lower or no 705 capacity for sustained high-walking (Salisbury and Frey, 2001). Among the candidates, 706 despite bernissartiids also have a similarly open dorsal shield, their small body mass, within 707 the size range (60 cm estimated for B. fagesii) of C. meijidei, could have allowed them a 708 sustained high-walking and a terrestrial locomotor behaviour (Salisbury and Frey, 2001; 709 Martin et al., 2020). Goniopholidids and atoposaurids had much more rigid dorsal shields than 710 that of extant crocodylians and the other neosuchians. These shields would have restricted the 711 lateral flexion of the trunk in favour of greater stabilization of the vertebral column during 712 terrestrial locomotion. The relative length of the limbs is directly related with locomotor 713 functions and terrestrial locomotor capabilities (Iijima et al., 2018). Generally, atoposaurids 714 have relatively longer hindlimbs than extant crocodylians whereas some goniopholidids have 715 forelimbs that are longer than their hindlimbs (measurements from Tennant et al., 2016; 716 Iijima et al., 2018; Ijima and Kubo, 2019). Several atoposaurid species were probably too 717 small to have produced Crocodylopodus meijidei although some of them could have reached 718 lengths that fit the size range such as Alligatorellus with a body length 42 to 55 cm (Schwarz- 719 Wings et al., 2011). Theriosuchus is thought to be possible terrestrial taxa with a small size of 720 around 50 cm total length and relatively long limbs (Schwarz and Salisbury, 2005; Schwarz et 30 721 al. 2017), and T. pusillus is a Berriasian species. Goniopholidids are also good candidate 722 trackmakers for C. meijidei. However, if the producers of C. meijidei were members of 723 Goniopholididae, they would be juvenile individuals or belong to a small unknown species 724 (Puértolas-Pascual and Mateus, 2020), so they would have greater ease walking on land. As 725 we have already mentioned, one of the most remarkable characteristics of C. meijidei is the 726 absence of tail marks (Fig.9). From an anatomical point of view, this could be explained by 727 several factors or a combination of them such as the presence of a shorter or lighter tail or a 728 stiffer tail base to keep it elevated. Regarding length, in most of the aforementioned candidate 729 taxa the relationship between total body length and tail length cannot be compared due to the 730 incompleteness of the fossil record. Some of the most complete specimens (e.g., 731 Alligatorellus and Atoposaurus), do not seem to have a particularly shorter tail (see 732 measurements from Tennant et al., 2016), with the length of the tail about half the total length 733 of the body, a similar condition to that observed in extant crocodylians (see measurements 734 from O’Brien et al., 2019). There is also no evidence of the lightness or rigidity of the tails of 735 these taxa; therefore, the reason for the absence of tail marks in C. meijidei remains unclear 736 until more complete fossils are found. Taking into account the paleogeographic and 737 geochronological context of C. meijidei, only some non-eusuchian neosuchian taxa are the 738 candidate of the trackmaker since the oldest eusuchian Hylaeochampsa is from the Barremian 739 of England (Clark and Norell, 1992); and there is no record of protosuchians, notosuchians, 740 gobiosuchids or sphenosuchians during the Berriasian of Europe. As a consequence, the 741 producer of C. meijidei was most likely a small non-eusuchian neosuchian crocodylomorph 742 such as goniopholidid, atoposaurid or bernissartid. 743 744 CONCLUSIONS 31 745 746 Little attention has been paid to extant crocodylomorph footprints until very recently, 747 despite the fact that this is one of the few groups where we can directly compare living and 748 extinct taxa. New descriptions in recent years have provided an excellent database to shed 749 light on the reasons of morphological variations among the extinct ichnotaxa. Our review of 750 the type material of Crocodylopodus meijidei has revealed the existence of new material and 751 provides new data for the characterization of this type ichnotaxon. The C. meijidei collection 752 shows at least three different size classes, which might reflect different ontogenetic stages 753 and/or sexual dimorphism. Our analysis of the sample, plus comparison with other tracks 754 assigned to Crocodylopodus, shows high morphological variation within this ichnogenus, but 755 it is difficult to interpret whether these differences are anatomical, substrate- and 756 preservation-related or locomotion/behavior-related. Comparisons with other crocodylomorph 757 ichnogenera highlight that the main differences between them relate to trackway features and 758 therefore different locomotor patterns/behaviors (either swimming tracks or walking traces 759 with different lomotor patterns) or different body plan of the trackmakers. Several trackway 760 parameters of the C. meijidei, such as its intermediate-gauge trackway, its relatively high pace 761 angulation (values higher than 108º), the absence of tail and other drag marks and 762 overprinting of manual prints by the pes, and manual tracks and anterior part of the pedal 763 tracks deeper than the posterior part, point to a style of locomotion different from extant 764 crocodylians and from the other walking tracks of extinct crocodylomorphs (e.g., 765 Batrachopus and Mehliella). The trackmaker was a small (probably no larger than 1.10 m for 766 the largest specimens) non-eusuchian neosuchian crocodylomorph presumably a 767 goniopholidid, an atoposaurid or a bernissartid, and walked possibly with its center of mass 768 more anteriorly located and with a more erect limb posture than exhibited by extant 32 769 crocodylomorphs. Thus, the trackmaker was possibly better adapted for terrestrial locomotion 770 than modern crocodylians. 771 772 ACKNOWLEDGMENTS 773 774 DC was funded by the Alexander von Humboldt Foundation and by the Beatriu de Pinós 775 postdoctoral programme (BP2017-00195) of the Government of Catalonia’s Secretariat for 776 Universities and Research of the Ministry of Economy and Knowledge. EPP is supported by 777 postdoctoral grants and projects funded by the Fundação para a Ciência e a Tecnologia, 778 Portugal (SFRH/BPD/116759/2016 and Geobiotec UIDB/04035/2020). This study was 779 subsidized in part by Project CGL2017-85038-P of the Spanish Ministerio de Economía y 780 Competitividad-ERDF, as well as by the Aragón Regional Government (“grupo de Referencia 781 Aragosaurus: reconstrucciones paleoambientales y recursos geológicos”). M.A. Arlegui 782 Sánchez, B. Martínez, M. L. Revilla Andía, J. L. García and D. Diez Corral (Museo 783 Numantino de Soria) are gratefully acknowledged for their help with the collections housed in 784 the Museo Numantino de Soria. L. Piñuela (MUJA) provided valuable information on the 785 Late Jurassic Asturian fossils; M. Belvedere (Università degli studi di Firenze) provided 786 significant help with the use of the CloudCompare software and I. Díaz-Martínez provided 787 valuable discussion on ichnotaxonomy. R. Glasgow (Zaragoza, Spain) revised the English 788 grammar of a previous version of the manuscript. The comments of the reviewers, J.O. 789 Farlow and T. Kubo, and the editor S.W. Salisbury have helped to improve the manuscript 790 and are greatly appreciated. 791 792 LITERATURE CITED 793 33 794 Abbassi, N., S. D. O. Porchetti, A. Wagensommer, and M. G. Dehnavi. 2015. Dinosaur and 795 crocodylomorph footprints from the Hojedk Formation (Bajocian, Middle Jurassic) of 796 north Kerman, central Iran. Italian Journal of Geosciences 134(1):86–94. 797 Andrade, M. B., and J. J. Hornung. 2011. A new look into the periorbital morphology of 798 Goniopholis (Mesoeucrocodylia: Neosuchia) and related forms. Journal of Vertebrate 799 Paleontology 31(2):352–368. 800 Andrade, M. B., J. J. Hornung, R. Edmonds, M. J. Benton, and R. Schouten. 2011. A new 801 Berriasian species of Goniopholis (Mesoeucrocodylia, Neosuchia) from England, and 802 a review of the genus. Zoological Journal of the Linnean Society 163(1):S66–S108. 803 Avanzini, M., J. C. García-Ramos, J. Lires, L. Piñuela, and M. G. Lockley. 2007. 804 Crocodylomorph tracks from the Late Jurassic of Asturias (Spain). Ichnos 14:143– 805 153. 806 Avanzini, M., J. C. García-Ramos, J. Lires, L. Piñuela, and M. G. Lockley. 2010. 807 Preservational morphotypes of Crocodylopodus from the Late Jurassic of Asturias 808 (Spain). New Mexico Museum of Natural History and Science Bulletin 51: 239–244. 809 Benton, M. J., and J. M. Clark. 1988. Archosaur phylogeny and the relationships of the 810 Crocodylia; pp. 295–338 in M. J. Benton (ed.), The phylogeny and classification of 811 the tetrapods. Volume 1, Systematics Association Special Volume no 35A 812 Amphibians, reptiles, birds. Clarendon Press, Oxford, UK. 813 Brochu, C. A. 2003. Phylogenetic approaches toward crocodylian history. Annual Review of 814 Earth and Planetary Sciences 31(1):357–397. 815 Bronzati, M., F. C. Montefeltro, and M. C. Langer. 2015. Diversification events and the 816 effects of mass extinctions on Crocodyliformes evolutionary history. Royal Society 817 Open Science 2(5):140385. 34 818 Buscalioni, A. D. 2017. The Gobiosuchidae in the early evolution of Crocodyliformes. 819 Journal of Vertebrate Paleontology 37(3): e1324459. 820 Buscalioni, A. D., F. Ortega, B. P. Pérez-Moreno, and S. Evans E. 1996. The Upper Jurassic 821 maniraptoran theropod Lisboasaurus estesi (Guimarota, Portugal) reinterpreted as a 822 crocodylomorph. Journal of Vertebrate Paleontology 16(2):358–362. 823 Carrano, M. T., and J. A. Wilson. 2001. Taxon distributions and the tetrapod track record. 824 Paleobiology 27(3):564–582. 825 Carpenter, K. 2009. Role of lateral body bending in crocodylian track making, Ichnos 826 16:202–207. 827 Castanera, D., C. Pascual, J. I. Canudo, and J. L. Barco. 2018. Bringing together research, 828 geoconservation and reaching a broad public in the form of a geotourism project: the 829 Ichnite Route of Soria (Spain). Geoheritage 10(3):393–403. 830 Castanera, D., C. Pascual, J. I. Canudo, N. Hernandez, and J. L. Barco, 2012. Ethological 831 variations in gauge in sauropod trackways from the Berriasian of Spain. Lethaia, 45(4) 832 :476–489. 833 Castanera, D., E. Malafaia, B. C. Silva, V. F. Santos, and M. Belvedere. 2021. New 834 dinosaur, crocodylomorph and swim tracks from the Late Jurassic of the Lusitanian 835 Basin: implications for ichnodiversity. Lethaia, 54(2), 271–287. 836 Clark, J. M., and M. Norell. 1992. The Early Cretaceous crocodylomorph Hylaeochampsa 837 vectiana from the wealden of the Isle of Wight. American Museum novitiates 3032:1– 838 19. 839 Carpenter, K. 2009. Role of lateral body bending in crocodylian track making. Ichnos, 840 16(3):202–207. 841 Cox, R. M., M. A. Butler, and H. B. John-Alder. 2007. The evolution of sexual size 35 842 dimorphism in reptiles; pp. 38–49 in W.U., Blanckenhorn, T., Székely and D.J., 843 Fairbairn (eds.), Sex, size and gender roles: evolutionary studies of sexual size 844 dimorphism. Oxford University Press, Oxford, UK. 845 Coombs, W. P. 1996. Redescription of the ichnospecies Antipus flexiloquus Hitchcock, from 846 the Early Jurassic of the Connecticut Valley. Journal of Paleontology 70(2):327–331. 847 Falkingham, P. L., K. T. Bates, M. Avanzini, M. Bennett, E. M. Bordy, B. H. Breithaupt, D. 848 Castanera, P. Citton, I. Díaz-Martínez, J.O. Farlow, A.R. Fiorillo, S.M. Gatesy, P. 849 Getty, K.G. Hatala, J.J. Hornung, J.A. Hyatt, H. Klein, J.N. Lallensack, A.J. Martin, 850 D. Marty, N.A. Matthews, C.A. Meyer, J. Milàn, N.J. Minter, N.L. Razzolini, A. 851 Romilio, S.S. Salisbury, L. Sciscio, I. Tanaka, A.L.A.Wiseman, L. D. Xing, and M. 852 Belvedere. 2018. A standard protocol for documenting modern and fossil ichnological 853 data. Palaeontology 61(4):469–480. 854 Farlow, J.O. (ed.). 2018. Noah's ravens: Interpreting the makers of tridactyl dinosaur 855 footprints. Indiana University Press, Bloomington, Indiana, 636 pp. 856 Farlow, J.O., and Britton, A. 2000. Size and body proportions in Alligator mississippiensis: 857 implications for archosaurian ichnology. Paleontological Society of Korea Special 858 Publication 4:189–206. 859 Farlow, J.O., N. J. Robinson, M. L.Turner, J. Black, and S. M. Gatesy. 2018a. Footfall pattern 860 of a bottom-walking crocodile (Crocodylus acutus). Palaios 33(9):406–413. 861 Farlow, J.O., N. J. Robinson, C. J. Kumagai, F. V. Paladino, P. L. Falkingham, R. M. Elsey, 862 and A. J. Martin. 2018b. Trackways of the American crocodile (Crocodylus acutus) in 863 northwestern Costa Rica: implications for crocodylian ichnology. Ichnos 25(1):30–65. 864 Foster, J. R., and M. G. Lockley. 1997. Probable crocodilian tracks and traces from the 865 Morrison Formation (Upper Jurassic) of eastern Utah. Ichnos 5(2):121–129. 36 866 Fuentes Vidarte, C., and M. Meijide Calvo. 2001. Primeras huellas de cocodrilo en el Weald 867 de Cameros (Soria, España). Nueva familia: Crocodylopodidae, nuevo icnogénero: 868 Crocodylopodus, nueva icnoespecie: C. meijidei; pp. 329–335 in Colectivo 869 Arqueológico-Paleontológico de Salas (eds.), Actas de las I Jornadas Internacionales 870 sobre Paleontología de Dinosaurios y su Entorno, Salas de los Infantes, Burgos, 871 Spain. 872 Gatesy, S. M. 1991. Hind limb movements of the American alligator (Alligator 873 mississippiensis) and postural grades. Journal of Zoology, 224(4):577 –588. 874 Gómez Fernández, J. C., and N. Meléndez. 1994a. Estratigrafía de la "Cuenca de los 875 Cameros" (Cordillera Ibérica Noroccidental, N de España) durante el tránsito Jurásico- 876 Cretácico. Revista de la Sociedad Geológica de España 7(1-2):121–139. 877 Gómez Fernández, J. C., and N. Meléndez. 1994b. Climatic control on Lower Cretaceous 878 sedimentation in a playa-lake system of a tectonically active basin (Huérteles 879 Alloformation, Eastern Cameros Basin, North-Central Spain). Journal of 880 Paleolimnology 11(1):91–107. 881 Grigg, G., and D. Kirshner, (eds.). 2015. Biology and evolution of crocodylians. Csiro 882 Publishing, Clayton South, Victoria, Australia 672 pp. 883 Hitchcock, E. 1845. An attempt to name, classify, and describe the animals that made the 884 fossil footmarks of New England. Proceedings of the 6th Annual Meeting of the 885 Association of American Geologists and Naturalists 6:23–25. 886 Hernández Medrano, N., C. Pacual Arribas, P. Latorre Macarron, and E. Sanz Pérez. 2008. 887 Contribución de los yacimientos de icnitas sorianos al registro general de Cameros. 888 Zubía 23-24:79–120. 889 Hone, D., J. C. Mallon, P. Hennessey, and L. M. Witmer. 2020. Ontogeny of a sexually 37 890 selected structure in an extant archosaur Gavialis gangeticus (Pseudosuchia: 891 Crocodylia) with implications for sexual dimorphism in dinosaurs. PeerJ 8:e9134. 892 Houck, K., M. G. Lockley, M. Caldwell, and B. Clark. 2010. A crocodile trackway from the 893 Dakota Group (Cretaceous) of the Golden Area, Colorado. New Mexico Museum of 894 Natural History and Science Bulletin 51:115–120. 895 Hutchinson, J. R., D. Felkler, K. Houston, Y. M. Chang, J. Brueggen, D. Kledzik, and K. A. 896 Vliet. 2019. Divergent evolution of terrestrial locomotor abilities in extant Crocodylia. 897 Scientific Reports 9(1):1–11. 898 Hutton, J. M. 1987. Morphometrics and field estimation of the size of the Nile crocodile. 899 African Journal of Ecology 25(4):225–230. 900 ICZN 1999. International Code of Zoological Nomenclature. Fourth Edition. The 901 International Trust for Zoological Nomenclature, London, UK. 306 pp. 902 Iijima, M., and T. Kubo. 2019. Comparative morphology of presacral vertebrae in extant 903 crocodylians: taxonomic, functional and ecological implications. Zoological Journal of 904 the Linnean Society 186(4):1006–1025. 905 Iijima, M., T. Kubo, and Y. Kobayashi. 2018. Comparative limb proportions reveal 906 differential locomotor morphofunctions of alligatoroids and crocodyloids. Royal 907 Society Open Science 5(3):171774. 908 Kim, K. S., M. G. Lockley, J. D. Lim, S. M. Bae, and A. Romilio. 2020. Trackway evidence 909 for large bipedal crocodylomorphs from the Cretaceous of Korea. Scientific Reports 910 10(1):1–13. 911 Klein, H., and S. G. Lucas. 2010. The Triassic footprint record of crocodylomorphs–a critical 912 re-evaluation. New Mexico Museum of Natural History and Science Bulletin 51:55– 913 60. 38 914 Klein, H., A. Lagnaoui, G. D. Gierliński, H. Saber, J. N. Lallensack, M. Oukassou, and A. 915 Charrière. 2018. Crocodylomorph, turtle and mammal tracks in dinosaur-dominated 916 Middle–? Upper Jurassic and mid-Cretaceous ichnoassemblages of Morocco. 917 Palaeogeography, Palaeoclimatology, Palaeoecology 498:39–52. 918 Kubo, T. 2008. In quest of the Pteraichnus trackmaker: comparisons to modern crocodilians. 919 Acta Palaeontologica Polonica 53(3):405–412. 920 Kubo, T. 2010. Variation in modern crocodylian limb kinematics and its effect on trackways. 921 Crocodyle Tracks and Traces. New Mexico Museum of Natural History and Science 922 Bulletin 51:51–53. 923 Kubo, T., and M. J. Benton. 2009. Tetrapod postural shift estimated from Permian and 924 Triassic trackways. Palaeontology 52(5):1029–1037. 925 Kubo, T., and M. Ozaki. 2009. Does pace angulation correlate with limb posture?. 926 Palaeogeography, Palaeoclimatology, Palaeoecology 275(1-4):54–58. 927 Leonardi, G. (ed.). 1987. Glossary and manual of tetrapod footprint palaeoichnology. 928 Conselho Nacional de Desenvolvimento Cientifico e Tecnologico – Brasil. 929 Lockley, M. G. 2010. A solution to the Mehliella mystery: tracking, naming, identifying and 930 measuring the first crocodylian trackway reported from the Cretaceous (Dakota 931 Group, Colorado). New Mexico Museum of Natural History and Science Bulletin 932 51:157–164. 933 Lockley, M. G., and C. Meyer. 2004. Crocodylomorph trackways from the Jurassic to Early 934 Cretaceous of North America and Europe: implications for ichnotaxonomy. Ichnos 935 11(1-2): 167–178. 936 Lockley, M. G., R. Li, M. Matsukawa, and J. Li. 2010a. Tracking Chinese crocodilians: 937 Kuangyuanpus, Laiyangpus and implications for naming crocodylian and crocodylian- 39 938 like tracks and associated ichnofacies. New Mexico Museum of Natural History and 939 Science Bulletin 51:99–108. 940 Lockley, M. G., K. Cart, J. Foster, and S. G. Lucas. 2018. Early Jurassic Batrachopus-rich 941 track assemblages from interdune deposits in the Wingate Sandstone, Dolores Valley, 942 Colorado, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 491:185–195. 943 Lockley, M. G., S. Lucas, J. Milàn, J. Harris, M. Avanzini, J. R. Foster and J. Spielmann. 944 2010b. The fossil record of crocodylian tracks and traces: an overview. New Mexico 945 Museum of Natural History and Science Bulletin 51:1–14. 946 Lockley, M. G., J. D. Lim, H. D. Park, A. Romilio, J. S.Yoo, J. W. Choi, K. S. Kim, S.-H. 947 Kang, D.H. Kim, and T. H. Kim. 2020. First reports of Crocodylopodus from Asia: 948 implications for the paleoecology of the Lower Cretaceous. Cretaceous Research 949 104441. 950 Lull, R.S. 1904. Fossil footprints of the Jura–Trias of North America. Boston Society of 951 Natural History 5:461–557. 952 Marchetti, L., M. Belvedere, S.Voigt, H. Klein, D. Castanera, I. Díaz-Martínez, D. Marty, L. 953 Xing, S. Feola, R.N. Melchor, and J. O. Farlow. 2019. Defining the morphological 954 quality of fossil footprints. Problems and principles of preservation in tetrapod 955 ichnology with examples from the Palaeozoic to the present. Earth-Science Reviews 956 193:109–145. 957 Martin, J. E., J. Raslan-Loubatié, and J. M. Mazin. 2016. Cranial anatomy of Pholidosaurus 958 purbeckensis from the Lower Cretaceous of France and its bearing on pholidosaurid 959 affinities. Cretaceous Research 66:43–59. 960 Martin, J. E., T. Smith, C. Salaviale, J. Adrien, and M. Delfino. 2020. Virtual reconstruction 961 of the skull of Bernissartia fagesii and current understanding of the neosuchian– 40 962 eusuchian transition. Journal of Systematic Palaeontology 18(13):1079–1101. 963 Mas, R., M.I. Benito, J. Arribas, S. Omodeo-Salé, P. Suárez-González, I.E. Quijada, J. 964 Guimerà, L.González-Acebrón, and M.E. Arribas. 2019. The Cameros basin; pp. 205– 965 214 in C.Quesada and J.T. Oliveira (eds.), The Geology of Iberia: A Geodynamic 966 Approach. Volume 3: The Alpine Cycle. Springer, Berlin, Germany. 967 Masrour, M., M. Boutakiout, J. Herrero Gascón, J. L. S. R. De Zuazo, R. O. Martínez, and F. 968 Pérez-Lorente. 2020. Footprints of Batrachopus isp. From the Imilchil megatracksite. 969 Middle?-Upper Jurassic, central High Atlas (Morocco). Journal of African Earth 970 Sciences, 172, 103980. 971 Mateus, O., M. Marzola, A. S. Schulp, L. L. Jacobs, M. J. Polcyn, V. Pervov, A. O. 972 Gonçalves, and M. L. Morais. 2017. Angolan ichnosite in a diamond mine shows the 973 presence of a large terrestrial mammaliamorph, a crocodylomorph, and sauropod 974 dinosaurs in the Early Cretaceous of Africa. Palaeogeography, palaeoclimatology, 975 palaeoecology 471:220–232. 976 Mazin, J. M., J. P. Billon-Bruyat, P. Hantzpergue, and G. Lafaurie. 2003. Ichnological 977 evidence for quadrupedal locomotion in pterodactyloid pterosaurs: trackways from the 978 Late Jurassic of Crayssac (southwestern France). Geological Society, London, Special 979 Publications 217(1):283–296. 980 McCrea, R. T., S. G. Pemberton, and P. J. Currie. 2004. New ichnotaxa of mammal and 981 reptile tracks from the Upper Paleocene of Alberta. Ichnos 11(3-4): 323–339. 982 Mehl, M.G. 1931. Additions to the vertebrate record of the Dakota Sandstone. The American 983 Journal of Science 21:441–452. 984 Milàn, J., and R. Hedegaard. 2010. Interspecific variation in tracks and trackways from extant 985 crocodylians. New Mexico Museum of Natural History and Science Bulletin 51:15– 41 986 29. 987 Milàn, J., S. G. Lucas, M. G. Lockley, and J. A. Spielmann. 2010. Crocodyle tracks and 988 traces. New Mexico Museum of Natural History and Science Bulletin 51. 989 Molnar, J. L., S. E. Pierce, B. A. S. Bhullar, A. H. Turner, and J. R. Hutchinson. 2015. 990 Morphological and functional changes in the vertebral column with increasing aquatic 991 adaptation in crocodylomorphs. Royal Society Open Science 2(11): 150439. 992 Mustoe, G. E. 2019. Lower Eocene Footprints from Northwest Washington, USA. Part 1: 993 Reptile Tracks. Geosciences 9 (321):1–19. 994 O’Brien, H. D., L. M. Lynch, K. A. Vliet, J. Brueggen, G. M. Erickson, and P. M. Gignac. 995 2019. Crocodylian head width allometry and phylogenetic prediction of body size in 996 extinct crocodyliforms. Integrative Organismal Biology, 1(1), obz006. 997 Olsen, P. E., and K. Padian. 1986. Earliest records of Batrachopus from the southwestern 998 United States, and a revision of some Early Mesozoic crocodylomorph ichnogenera; 999 pp. 259–273 in K. Padian (ed.), The beginning of the age of Dinosaurs, faunal change 1000 across the Triassic-Jurassic boundary. Cambridge University Press, New York, USA. 1001 Padian, K. 2003. Pterosaur stance and gait and the interpretation of trackways. Ichnos 10(2- 1002 4): 115–126. 1003 Padian, K., and P.E. Olsen. 1984. The fossil trackway Pteraichnus: Not pterosaurian, but 1004 crocodilian. Journal of Paleontology 58:178–184. 1005 Parrish, J. M. 1987. The origin of crocodilian locomotion. Paleobiology 13(4): 396–414. 1006 Pascual Arribas, C., N. Hernández Medrano, P. Latorre Macarrón, and E. Sanz Pérez. 2005. 1007 Nuevo rastro de icnitas de cocodrilo en la Aloformación Huérteles de la Cuenca de 1008 Cameros. Studia Geologica Salmanticensia 41:75–91. 1009 Platt, S.G., T.R. Rainwater, J.B. Thorbjarnarson, A.G. Finger, T.A. Anderson, and S.T. 42 1010 McMurry, 2009. Size estimation, morphometrics, sex ratio, sexual size dimorphism, 1011 and biomass of Morelet’s crocodile in northern Belize. Caribbean Journal of Science 1012 45(1):80–94. 1013 Pol, D., A. H. Turner, and M. A. Norell. 2009. Morphology of the Late Cretaceous 1014 crocodylomorph Shamosuchus djadochtaensis and a discussion of neosuchian 1015 phylogeny as related to the origin of Eusuchia. Bulletin of the American Museum of 1016 Natural History 324: 1–103. 1017 Pouech, J., R. Amiot, C. Lécuyer, J. M. Mazin, F. Martineau, and F. Fourel. 2014. Oxygen 1018 isotope composition of vertebrate phosphates from Cherves-de-Cognac (Berriasian, 1019 France): environmental and ecological significance. Palaeogeography, 1020 Palaeoclimatology, Palaeoecology 410:290–299. 1021 Puértolas-Pascual, E., and O. Mateus. 2020. A three-dimensional skeleton of Goniopholididae 1022 from the Late Jurassic of Portugal: implications for the Crocodylomorpha bracing 1023 system. Zoological Journal of the Linnean Society 189(2):521–548. 1024 Quijada, I. E., P. Suárez González, M. I. Benito, and J. R. Mas. 2013. New insights on 1025 stratigraphy and sedimentology of the Oncala Group (eastern Cameros Basin): 1026 implications for the paleogeographic reconstruction of NE Iberia at Berriasian times. 1027 Journal of Iberian Geology 39(2): 313–334. 1028 Rajkumar, H. S., G. E. Mustoe, K. S. Khaidem, and I. Soibam. 2015. Crocodylian Tracks 1029 from Lower Oligocene Flysch deposits of the Barail Group, Manipur, India. Ichnos 1030 22(2): 122–131. 1031 Razzolini, N. L., B.Vila, D. Castanera, P. L. Falkingham, J. L. Barco, J. I.Canudo, P.L. 1032 Manning, and A. Galobart. 2014. Intra-trackway morphological variations due to 1033 substrate consistency: the El Frontal dinosaur tracksite (Lower Cretaceous, Spain). 43 1034 PloS ONE 9(4). 1035 Reilly, S.M., and J.A. Elias, 1998. Locomotion in Alligator mississippiensis: kinematic effects 1036 of speed and posture and their relevance to the sprawling-to-erect paradigm. Journal of 1037 Experimental Biology 201:2559–2574 1038 Rozada, L., R. Allain, R.Vullo, J. Goedert, D. Augier, A. Jean, J. Marchal, C. Peyre de 1039 Fabrègues, M. Qvarnstrom, and R. Royo-Torres. 2020. A Lower Cretaceous 1040 Lagerstätte from France: a taphonomic overview of the Angeac‐Charente vertebrate 1041 assemblage. Lethaia 54 (2): 141–165. 1042 Salisbury, S. W. 2002. Crocodilians from the Lower Cretaceous (Berriasian) Purbeck 1043 Limestone Group of Dorset, southern England. Special Papers in Palaeontology 1044 68:121–144. 1045 Salisbury, S. W., and E. Frey. 2001. A biomechanical transformation model for the evolution 1046 of semi-spheroidal articulations between adjoining vertebral bodies in crocodilians; 1047 pp. 85–134 in C.G. Grigg, F. Seebacher and C.E. Franklin (eds.), Crocodilian biology 1048 and evolution. Surrey Beatty and Sons, Chipping Norton, Australia. 1049 Salisbury, S. W., P. M.Willis, S. Peitz, and P. M. Sander. 1999. The crocodilian Goniopholis 1050 simus from the Lower Cretaceous of north-western Germany. Special Papers in 1051 Palaeontology 60:121–148. 1052 Schudack, U., and M. Schudack. 2009. Ostracod biostratigraphy in the Lower Cretaceous of 1053 the Iberian chain (eastern Spain). Journal of Iberian Geology 35(2):141–168. 1054 Segura, M., F. Barroso-Barcenilla, M. Berrocal-Casero, D. Castanera, J. F. Garcia-Hidalgo, 1055 and V. F. Santos. 2016. A new Cenomanian vertebrate tracksite at Tamajón 1056 (Guadalajara, Spain): Palaeoichnology and palaeoenvironmental implications. 1057 Cretaceous Research 57: 508–518. 44 1058 Schwarz, D., and R. Fechner. 2004. Lusitanisuchus, a new generic name for Lisboasaurus 1059 mitracostatus (Crocodylomorpha: Mesoeucrocodylia), with a description of new 1060 remains from the Upper Jurassic (Kimmeridgian) and Lower Cretaceous (Berriasian) 1061 of Portugal. Canadian Journal of Earth Sciences 41(10):1259–1271. 1062 Schwarz, D., and S. W. Salisbury. 2005. A new species of Theriosuchus (Atoposauridae, 1063 Crocodylomorpha) from the Late Jurassic (Kimmeridgian) of Guimarota, Portugal. 1064 Geobios 38(6):779–802. 1065 Schwarz, D., and R. Fechner, 2008. The first dentary of Lisboasaurus (crocodylomorpha,? 1066 mesoeucrocodylia) from the Lower Cretaceous (Barremian) of Una, Cuenca Province, 1067 Spain. Journal of Vertebrate Paleontology 28(1):264–268. 1068 Schwarz, D., J. Rees, and J. Lindgren. 2009. Lower Cretaceous mesoeucrocodylians from 1069 Scandinavia (Denmark and Sweden). Cretaceous Research 30(5):1345–1355. 1070 Schwarz, D., M. Raddatz, and O.Wings. 2017. Knoetschkesuchus langenbergensis gen. nov. 1071 sp. nov., a new atoposaurid crocodyliform from the Upper Jurassic Langenberg 1072 Quarry (Lower Saxony, northwestern Germany), and its relationships to Theriosuchus. 1073 PLoS One 12(2):e0160617. 1074 Schwarz-Wings, D., N. Klein, C. Neumann, and U. Resch. 2011. A new partial skeleton of 1075 Alligatorellus (Crocodyliformes) associated with echinoids from the Late Jurassic 1076 (Tithonian) lithographic limestone of Kelheim, S-Germany. Fossil Record 14(2):195– 1077 205. 1078 Simpson, E. L., H.L. Hilbert-Wolf, M.C. Wizevich, S. G. Lucas, E. Tester, S.E. Tindall, and 1079 J. J. Bernard 2010. A crocodylomorph track in the upper cretaceous capping sandstone 1080 member of the Wahweap Formation, Grand Staircase-Escalante national monument, 1081 Utah, USA. New Mexico Museum of Natural History and Science Bulletin 51:165– 45 1082 170. 1083 Tennant, J. P., P. D. Mannion, and P. Upchurch. 2016. Evolutionary relationships and 1084 systematics of Atoposauridae (Crocodylomorpha: Neosuchia): implications for the rise 1085 of Eusuchia. Zoological Journal of the Linnean Society 177(4):854–936. 1086 Thorbjarnarson, J.B. 1994. Reproductive ecology of the spectacled caiman (Caiman 1087 crocodilus) in the Venezuelan Llanos. Copeia, 4:907–919. 1088 Vila, B., D. Castanera, J. Marmi, J. I. Canudo, and À. Galobart. 2015. Crocodile swim tracks 1089 from the latest Cretaceous of Europe. Lethaia 48(2):256–266. 1090 Willey, J. S., A. R. Biknevicius, S. M. Reilly, and K. D. Earls. 2004. The tale of the tail: limb 1091 function and locomotor mechanics in Alligator mississippiensis. Journal of 1092 Experimental Biology 207(3):553–563. 1093 Wilberg, E. W., A. H. Turner, and C. A. Brochu. 2019. Evolutionary structure and timing of 1094 major habitat shifts in Crocodylomorpha. Scientific reports 9(1):1–10. 1095 Zug, G. R. 1974. Crocodilian galloping: an unique gait for reptiles. Copeia, 1974(2):550 – 1096 552. 1097 1098 Submitted July 17, 2020; revisions received Month DD, YYYY; accepted Month DD, 1099 YYYY 1100 1101 Figure captions (Colour version) 1102 FIGURE 1. Geographical and geological setting of the El Frontal tracksite. A, Geographical 1103 setting of the Ichnite Route of Soria (after Castanera et al., 2018). B, Geological setting of the 1104 area showing the outcrops of the Huérteles Formation (after Quijada et al., 2013). [Intended 1105 for page width] 46 1106 FIGURE 2. Measurements taken on the Crocodylopodus meijidei specimens. A, trackway 1107 parameters in the pedal impressions; B, trackway parameters in the manual impressions; C, 1108 Estimation of the glenoacetabular distance; D, parameters in individual manus-pes set. PA = 1109 pace angulation from the center of the footprint; ANG, pace angulation from tip of the 1110 impression of digit III; SL = stride length; PL = pace length; FR = footprint rotation; OW = 1111 overall width; ML = trackway midline; Dm-p = manus-pes distance; FL = footprint length; 1112 FW = footprint width; LI, LII, LIII; LIV, LV = length of each digital impression; IA = 1113 interdigital divarication angle. [Intended for page width] 1114 1115 FIGURE 3. Lectotype of Crocodylopodus meijidei, specimen MNS2002/96/2bis. A, Picture 1116 of the trackway. B, False-colour depth map of the trackway. Purple colour indicating deeper 1117 parts of the slab. C, Outline drawing of the trackway. D, Close-up picture of manus-pes set 1118 3mp. E, Close-up picture of manus-pes set 2mp. Note that both 2p and 3p only show evidence 1119 of claw mark in digital impressions I-III. Note that manual print 3m still has sediment inside 1120 the print. Scale bars equal 5 cm (A, B, C), 1 cm (D, E). [Intended for page width] 1121 1122 FIGURE 4. Paralectotypes of Crocodylopodus meijidei, specimens MNS2003/92/8a (A-C) 1123 and MNS2002/96/4 (D-F). A, Picture of the trackway. B, False-colour depth map of the 1124 trackway. Purple colours indicating deeper parts of the slab. C, Outline drawing of the 1125 trackway. D, Picture of the trackway. E, False-colour depth map of the trackway. Purple 1126 colours indicating shallower parts of the slab. F, Outline drawing of the trackway. Scale bars 1127 equal 5 cm. [Intended for page width] 1128 47 1129 FIGURE 5. Referred specimens of Crocodylopodus meijidei, specimens MNS2002/96/3 (A- 1130 C), MNS2003/92/8b (D-F) MNS2002/96/10 (G-I). A, Picture of the set of coupled manual 1131 and pedal tracks. B, False-colour depth map of the set. Purple colours indicating deeper parts 1132 of the slab. C, Outline drawing of the set. D, Picture of the trackway. E, False-colour depth 1133 map of the trackway. Purple colours indicating shallower parts of the slab. F, Outline drawing 1134 of the trackway G, Picture of two set of coupled manual and pedal tracks. H, False-colour 1135 depth map of the sets. Purple colours indicating deeper parts of the slab. I, Outline drawing of 1136 the trackway. Note that the latter two specimens are possibly part and counterpart (mold and 1137 true track of the same trackway). Scale bars equal 5 cm. [Intended for page width] 1138 1139 FIGURE 6. Referred specimens of Crocodylopodus meijidei, A-C, Specimen 1140 MNS2002/96/12. A, Picture of the trackway. B, False-colour depth map of the trackway. 1141 Purple colours indicating shallower parts of the slab. C, Outline drawing of the trackway. D, 1142 Picture of specimen MNS2002/96/5, E, Picture of specimen MNS2002/96/7 and F, Picture of 1143 specimen MNS2002/96/8. Scale bars (and coin) equal 2.5 cm. [Intended for page width] 1144 1145 FIGURE 7. Comparison of the lectotype of C.meijidei with Crocodylopodus tracks described 1146 in other areas. A, Lectotype of C.meijidei. B, Crocodylopodus isp. from the Middle Jurassic 1147 of Iran (after Abbassi et al., 2015). C, Crocodylopodus meijidei from the Middle-?Upper 1148 Jurassic of Morocco (after Klein et al., 2018). D, F, G, H, Crocodylopodus meijidei tracks 1149 from the Upper Jurassic of Asturias, Spain (after Avanzini et al., 2007, 2010). E, 1150 Crocodylopodus isp. from the Upper Jurassic of Portugal (after Castaner et al., 2020). I, cf. 1151 Crocodylopodus from the Lower Cretaceous (Huérteles Formation) of Soria, Spain (after 1152 Pascual Arribas et al., 2005). J, K, L, M, Crocodylopodus isp. from the Lower Cretaceous of 48 1153 Korea (after Lockley et al., 2020). N, cf. Crocodylopodus from the Upper Cretaceous of Spain 1154 (after Vila et al., 2015). Scale bars equal 1 cm (D, F, G, H), 3 cm (C), 5 cm (A, B, N), 10 cm 1155 (I, J, K, L, M). [Intended for page width] 1156 1157 FIGURE 8. Comparison of the lectotype of C. meijidei with walking trackways of extinct 1158 crocodylomorph ichnotaxa and a modern crocodylian trackway. A, Lectotype of C. meijidei. 1159 B, Batrachopus deweyi from the Lower Jurassic of the USA (after Padian and Olsen, 1986). 1160 C, Antipus flexiloquus from the Lower Jurassic of the USA (after Coombs, 1996). D, 1161 Angolaichnus adamanticus from the Lower Cretaceous of Angola (after Mateus et al., 2017). 1162 E, Mehliella jeffersonensis from the Cretaceous of the USA (after Mehl, 1931; Lockley 1163 2010). F, Modern trackway of Crocodylus niloticus (after Mazin et al., 2003). Scale bars 1164 equal 3 cm (B), 5 cm (A, C), 10 cm (D, F) 50 cm (E). [Intended for page width] 1165 1166 FIGURE 9. Summary of anatomical features that may explain the trackway differences 1167 between the trackmaker of C. meijidei (an indeterminate neosuchian crocodylomorph) and an 1168 indeterminate extant crocodylian. Silhouettes of crocodylomorphs are not based on any 1169 particular species. Trackway in B modified from Milàn and Hedegaard (2010). Dark colours 1170 indicating depper parts of the footprints. [Intended for page width] 1171 1172 1173 Figure captions (Black and white version) 1174 1175 FIGURE 1. Geographical and geological setting of the El Frontal tracksite. A, Geographical 1176 setting of the Ichnite Route of Soria (after Castanera et al., 2018). B, Geological setting of the 49 1177 area showing the outcrops of the Huérteles Formation (after Quijada et al., 2013). [Intended 1178 for page width] 1179 1180 FIGURE 2. Measurements taken on the Crocodylopodus meijidei specimens. A, trackway 1181 parameters in the pedal impressions; B, trackway parameters in the manual impressions; C, 1182 Estimation of the glenoacetabular distance; D, parameters in individual manus-pes set. PA = 1183 pace angulation from the center of the footprint; ANG, pace angulation from tip of the 1184 impression of digit III; SL = stride length; PL = pace length; FR = footprint rotation; OW = 1185 overall width; ML = trackway midline; Dm-p = manus-pes distance; FL = footprint length; 1186 FW = footprint width; LI, LII, LIII; LIV, LV = length of each digital impression; IA = 1187 interdigital divarication angle. [Intended for page width] 1188 1189 FIGURE 3. Lectotype of Crocodylopodus meijidei, specimen MNS2002/96/2bis. A, Picture 1190 of the trackway. B, Solid three-dimensional model of the trackway. C, Outline drawing of the 1191 trackway. D, Close-up picture of manus-pes set 3mp. E, Close-up picture of manus-pes set 1192 2mp. Note that both 2p and 3p only show evidence of claw mark in digital impressions I-III. 1193 Note that manual print 3m still has sediment inside the print. Scale bars = 5 cm (A, B, C), 1 1194 cm (D, E). [Intended for page width] 1195 1196 FIGURE 4. Paralectotypes of Crocodylopodus meijidei, specimens MNS2003/92/8a (A-C) 1197 and MNS2002/96/4 (D-F). A, Picture of the trackway. B, Solid three-dimensional model of 1198 the trackway. C, Outline drawing of the trackway. D, Picture of the trackway. E, solid three- 1199 dimensional model of the trackway. F, Outline drawing of the trackway. Scale bars equal 5 1200 cm. [Intended for page width] 50 1201 1202 FIGURE 5. Referred specimens of Crocodylopodus meijidei, specimens MNS2002/96/3 (A- 1203 C), MNS2003/92/8b (D-F) MNS2002/96/10 (G-I). A, Picture of the set of coupled manual 1204 and pedal tracks. B, Solid three-dimensional model of the set. C, Outline drawing of the set. 1205 D, Picture of the trackway. E, Solid three-dimensional model of the trackway. F, Outline 1206 drawing of the trackway G, Picture of two set of coupled manual and pedal tracks. H, Solid 1207 three-dimensional model of the sets. I, Outline drawing of the trackway. Note that the latter 1208 two specimens are possibly part and counterpart (mold and true track of the same trackway). 1209 Scale bars equal 5 cm. [Intended for page width] 1210 1211 FIGURE 6. Referred specimens of Crocodylopodus meijidei. A-C, Specimen 1212 MNS2002/96/12. A, Picture of the tracks. B, Solid three-dimensional model of the tracks. C, 1213 Outtline drawing of the tracks. D, Picutre of specimen MNS2002/96/5, E, Picture of specimen 1214 MNS2002/96/7 and F, Picture of specimen MNS2002/96/8. Scale bars (and coin) equal 2.5 1215 cm. [Intended for page width] 1216 1217 FIGURE 7. Comparison of the lectotype of C.meijidei with Crocodylopodus tracks described 1218 in other areas. A, Lectotype of C.meijidei. B, Crocodylopodus isp. from the Middle Jurassic 1219 of Iran (after Abbassi et al., 2015). C, Crocodylopodus meijidei from the Middle-?Upper 1220 Jurassic of Morocco (after Klein et al., 2018). D, F, G, H, Crocodylopodus meijidei tracks 1221 from the Upper Jurassic of Asturias, Spain (after Avanzini et al., 2007, 2010). E, 1222 Crocodylopodus isp. from the Upper Jurassic of Portugal (after Castaner et al., 2020). I, cf. 1223 Crocodylopodus from the Lower Cretaceous (Huérteles Formation) of Soria, Spain (after 1224 Pascual Arribas et al., 2005). J, K, L, M, Crocodylopodus isp. from the Lower Cretaceous of 51 1225 Korea (after Lockley et al., 2020). N, cf. Crocodylopodus from the Upper Cretaceous of Spain 1226 (after Vila et al., 2015). Scale bars equal 1 cm (D, F, G, H), 3 cm (C), 5 cm (A, B, N), 10 cm 1227 (I, J, K, L, M). [Intended for page width] 1228 1229 FIGURE 8. Comparison of the lectotype of C. meijidei with walking trackways of extinct 1230 crocodylomorph ichnotaxa and a modern crocodylian trackway. A, Lectotype of C. meijidei. 1231 B, Batrachopus deweyi from the Lower Jurassic of the USA (after Padian and Olsen, 1986). 1232 C, Antipus flexiloquus from the Lower Jurassic of the USA (after Coombs, 1996). D, 1233 Angolaichnus adamanticus from the Lower Cretaceous of Angola (after Mateus et al., 2017). 1234 E, Mehliella jeffersonensis from the Cretaceous of the USA (after Mehl, 1931; Lockley 1235 2010). F, Modern trackway of Crocodylus niloticus (after Mazin et al., 2003). Scale bars 1236 equal 3 cm (B), 5 cm (A, C), 10 cm (D, F) 50 cm (E). [Intended for page width] 1237 1238 FIGURE 9. Summary of anatomical features that may explain the trackway differences 1239 between the trackmaker of C. meijidei (an indeterminate neosuchian crocodylomorph) and an 1240 indeterminate extant crocodylian. Silhouettes of crocodylomorphs are not based on any 1241 particular species. Trackway in B modified from Milàn and Hedegaard (2010). Dark colours 1242 indicating depper parts of the footprints. [Intended for page width] 1243 1244 1245 52 TABLE 1: Measurements of the lectotype (MNS2002/96/2bis) of Crocodylopodus meijidei. MP, Morphological preservation value (Marchetti et al., 2019); FL, footprint length; FW, footprint width; FL/FW, footprint length/footprint width ratio; LI, LII, LIII, LIV, LV, digital impression length; WI, WII, WIII, WIV, WV, digital impression width; I^II, II^III, III^IV, IV^V, Total IA, interdigital divarication angles; Dm–p, manus-pes distance; HI, heteropody index; PL, pace length; SL, stride length; PA/ANG, pace angulation (PA, center of the footprint; ANG, tip of the impression of digit III); FR, footprint rotation; OW = overall width; GA, glenoacetabular distance. Ar = inner trackway width. FL, FW, LI, LII, LIII, LIV, LV, WI, WII,WIII, WIV, Dm-p, PL, SL, OW, GA, Ar in cm. I^II, II^III, III^IV, IV^V, Total IA, PA, ANG, FR in degrees (º). HI, PTR, MTR in %. Trackway MNS2002/96/2bis MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III 1p 2.5 5 3.5 1.42 3.2 4.5 5 4.2 none 0.5 0.5 0.5 0.6 none 13 21 1m 1.5 1.5 2.5 0.6 1.1 1.3 1.7 1.3 1 0.3? 0.3? 0.3? 0.3? 0.3? 38 53 2p* 2 4.6 3.3 1.39 3.5 4.5 4.6 4.1 none 0.5 0.5 0.4 0.4 none 15 24 2m 2 1.8 2.4 0.75 1.2 1.6 2 1.6 1 ? ? ? ? ? 50 59 TABLE 1. (Continued) 2.5 5.1 3.9 1.3 3.2 4.7 5.1 4.2 none 0.3 0.4 0.4 0.5 none 10 23 3m 2.5 2.2 3 0.73 1.5 1.9 2.2 1.7 1.5 0.3 0.4 0.4 0.4 0.4 36 56 4p 1.5 ? ? ? 1.8 3 3.3 ? none 0.4 0.4 0.4 ? ? ? ? 4m 1 2? 2.9? 0.68? ? ? ? ? ? ? ? ? ? ? ? ? Total FL III^IV IV^V IA Dm–p x FW HI PL SL PA ANG FR OW GA Ar Ar/FW 1p 22 none 56 4.5 17.5 21 11.5 20 none none 8-I 5.8 14.2 2.2 0.62 1m 66 61 218 4.5 3.75 21 11.5 20 none none 36-O 6.3 14.2 none none 2p* 18 none 57 4.4 15.18 28 10.5 17 137 137 20-I 7.5 11.5 2 0.6 2m 57 43 209 4.4 4.32 28 9.5 17 140 129 15-O 6 11.5 none none 3p 24 none 57 3.8 19.89 33 10 none 118 116 none 8.7 ? 2.3 0.58 3m 58 65 215 3.8 6.6 33 11.6 none 105 102 none 8.9 ? none none TABLE 1. (Continued) 4p ? ? ? 4.3 ? ? ? ? ? ? ? 7.7? ? ? none 4m ? ? ? 4.3 ? ? ? ? ? ? ? 8.8 ? none none Observations: * heel inferred A new look at Crocodylopodus meijidei: implications for crocodylomorph locomotion DIEGO CASTANERA1*, CARLOS PASCUAL-ARRIBAS2, JOSÉ IGNACIO CANUDO3, EDUARDO PUÉRTOLAS-PASCUAL3,4,5 1: Institut Català de Paleontologia Miquel Crusafont, Universitat Autònoma de Barcelona, c/ Escola Industrial 23, 08201 Sabadell, Barcelona, Spain; 2: C. Real 65, 42002 Soria, Spain. capascual-1@telefonica.net; 3: Aragosaurus—IUCA, Paleontología, Facultad de Ciencias, Universidad de Zaragoza, 50009 Zaragoza, Spain. jicanudo@unizar.es; 4: GeoBioTec, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa, Caparica, Portugal. eduardo.puertolas@gmail.com; 5: Museu da Lourinhã, Lourinhã, Portugal; * Corresponding author. Email: diego.castanera@icp.cat; dcastanera@hotmail.es SUPPLEMENTAL DATA: DESCRIPTION OF THE PARATYPES AND REFERRED MATERIAL MNS2003/92/8a. This specimen is one of the holotypes (Trackway B and fig. 2 in Fuentes Vidarte and Meijide Calvo, 2001) and one of the paratypes according to Lockley and Meyer (2004). The specimen (Fig. 4A-C) is composed of seven footprints, three manual-pedal sets, and one pedal print at the end of the trackway. The tracks are preserved as true tracks or very shallow undertracks. The MP value is high and quite constant (MP = 2–2.5), with the exception of the first manual-pedal set (MP = 1). It is a small- to medium-sized specimen (Pedal FL= 4.9–5.1 cm; Pedal FW = 3.9–4.1 cm). The manual prints are pentadactyl, star-shaped and wider than long (FL/FW ratio = 0.61– 0.7). The digital impressions are thin with an acuminate end. DIII is the longest (1.9–2.1 cm), DII and DIV are similar and slightly shorter, whereas DI and DV are the shortest and also similar in length. The digital impressions are thin (WI-WV = 0.2–0.4 cm). I^II is the lowest angle (40–43º), the angles II^III (60–65º) and III^IV (61–67º) being higher and similar, whereas angle IV^V (42–59º) is slightly lower and variable. The total divarication in the manual impressions is high (IA = 203º–234º). Generally, digital impressions I-II and IV-V are oriented medially and laterally respectively, and digital impression III has an anterior-anterolateral orientation. DI and DV are slightly opposed. The manual prints are laterally rotated (28–40º). No clear claw marks are identified. The pedal impressions are tetradactyl, subtriangular in shape, and longer than wide (FL/FW ratio =1.19–1.3). The central digital impressions are longer than the lateral and medial ones. DIII is the longest (4.9–5.1 cm), DII (4.5–4.8 cm) and DIV (4–4.1 cm) are slightly shorter, but DII is clearly longer than DIV. DI is the shortest, being considerably shorter (3.3–3.7 cm). The digital impressions are noticeably thin (WI-WIV = 0.3–0.4 cm). The four of them have an acuminate end but only 2p shows clear evidence of claw marks in the first three digital impressions and a slightly rounded distal end in DIV. The orientation of the digital impressions is medial (DI and DII), anteromedial (DIII) and anterolateral (DIV), with a high total IA angle of 60–67º. I^II (16–19º) is the lowest angle, II^III (20–28º) and III^IV (23–25º) being relatively similar but variable. The heel area is shallow and poorly preserved in all the tracks but 2p possibly shows an oval to subtriangular symmetric morphology. Pedal rotation is low (6–20º) and medial. The manual-pedal distance is short (4.3–4.8 cm). PL is very similar for manual and pedal impressions (10.2–12.5 cm). SL is also very similar for both (19–19.5 cm). The trackway is intermediate-gauge (Ar/FW = 0.8-0.84). PA is medium-high, with values of 107–109° for the pedal impressions and 120º for the manual (ANG = 110º for both). The heteropody is medium, with HI of 29–31% suggesting that the manual-pedal ratio was around 1:3. There is not a great difference between the maximum depth of manual and pedal impressions, but in both limbs the anterior part is deeper; notably, digital impressions II-III-IV are deeper than I and V in the manual prints. There is no evidence of overprinting or of tail or belly drag marks. The estimated glenoacetabular distance is 14.2 cm. MNS2002/96/4. This specimen is one of the holotypes (Trackway C and fig. 3 in Fuentes Vidarte and Meijide Calvo, 2001) and one of the paratypes according to Lockley and Meyer (2004). The specimen is composed of three manual-pedal sets although the first pedal print is not complete. The tracks (Fig. 4D-F) are preserved as a natural cast (the drawing in Fuentes Vidarte and Meijide Calvo (2001) is a mirror image of MNS2002/96/4). The three manual prints have a medium MP value (1.5-2), whereas the MP of the pedal prints is medium to low (0.5–1.5) since only the anterior part is well marked. It is a small-sized specimen (Pedal FL= 3.7–4.1 cm; Pedal FW = 3.8–3.6 cm). The manual prints are pentadactyl, star-shaped and wider than long (FL = 1.9–2.1 cm; FW = 2.6–2.7 cm; FL/FW ratio = 0.7–0.77). The digital impressions are thin and of variable width (WI-WV 0.2–0.5 cm), with an acuminate end. Manual print 3m has the widest digital impressions. DIII is the longest (1.9–2.1 cm), DII and DIV are slightly shorter, whereas DI and DV are the shortest and similar in length. I^II is the lowest interdigital angle (IA) but quite variable (35–56º), II^III (43–62º), III^IV (46–49º) and IV^V (57–59º) being higher and with similar values though also quite variable. The total divarication in the manual impression is high (IA = 200–206º). Generally, DI-DII and DIV-DV are oriented medially-anteromedially and laterally-posterolaterally, respectively; DIII has an anterior-anterolateral orientation. DI and DV are to a large extent opposed (e.g. 1m). The manual prints are laterally rotated (23º–43º). No clear claw marks are identified in the manual impressions. The pedal prints are tetradactyl and longer than wide (FL/FW ratio =0.97–1.33), although the heel is not preserved and the real lengths of the autopod cannot be calculated. DIII is the longest (3.7–4.1 cm), DII (3.6–3.7 cm) is slightly shorter, whereas DIV (2.8–3.1 cm) and DI are even shorter (2.5–2.8 cm). The digital impressions are thin, and the width varies (WI-WIV = 0.2–0.5 cm) among the different digits but also in different pedal impressions. The digital impressions have an acuminate end, with no clear evidence of claw marks. The orientation of the digital impressions is medial (DI), anteromedial (DII and DIII) and lateral (DIV), with a high (higher than the other specimens) total divarication angle of 71–89º. I^II (20–27º) is the lowest angle, II^III (22–31º) and III^IV (22–38º) being relatively similar but variable among the digital impressions. The heel area is not preserved (the total FL would be longer). Pedal rotation is very low (3–5º) and medial. The manual-pedal distance is short (3.5–4.2 cm). PL is very similar for manual and pedal impressions, with almost no variation between them (10.5–11.2 cm). SL is also very similar for the trackway (19.1–19.2 cm). The trackway is intermediate-gauge (Ar/FW = 0.68-0.88?). PA is medium-high, with 110°? for the pedal and 118º for the manual impressions (ANG =100º and 105º, respectively). The heteropody is medium, with an HI of 36–38% suggesting that the manual-pedal ratio was around 1:3. The manual prints are slightly deeper than the pedal prints although the anterior part of the pedal reaches a similar depth. There is evidence of overprinting in the third manual- pedal set, with the pedal partially overprinting DI of the manual impression. There is no evidence of tail or belly drag marks. The estimated glenoacetabular distance is 12.8 cm. MNS2002/96/3. This specimen is an isolated manual-pedal set regarded by Fuentes Vidarte and Meijide Calvo (2001) as a paratype (fig. B in Fuentes Vidarte and Meijide Calvo, 2001). The tracks (Fig. 5A-C) are preserved as true tracks or very shallow undertracks, with the pedal showing a high MP value (2.5), but the manual has a low MP (0.5) and only some digital impressions can be distinguished. It is the largest specimen in the whole sample, being clearly medium-sized (Pedal FL= 7 cm; Pedal FW = 5.6 cm). The manual print morphology is not well preserved, but an FL of 2.8 cm and FW of 4.1 cm are estimated, making the print wider than long, with a FL/FW ratio of 0.68. The pedal impression is tetradactyl, subtriangular in shape, and longer than wide (FL/FW ratio = 1.25). DIII is the longest (7 cm), whereas DII and DIV are subequal in length (6.3 and 6.2 cm, respectively) and are considerably longer than DI (4.8 cm). The digital impressions are thicker than in the other specimens, WI and WII being slightly thinner than WIII and WIV (FW = 0.6 vs 0.8 cm). The four digital impressions have acuminate ends, the first three digital impressions being more acuminate than digit IV, indicating not well-preserved claw marks in DI-DIII and their absence in DIV. The orientation of the digits (taking into account the midline of the track) is medial (DI), anteromedial (DII), anterior (DIII) and anterolateral (DIV), with a high IA (61º) increasing from I to IV. I^II is the lowest angle (13º), then come II^III (21º) and III^IV (27º). The heel area is shallower than the anterior part of the footprint, but the morphology seems to be oval to subtriangular and quite symmetric. The manual-pedal distance is 9 cm, i.e. comparatively longer than previous specimens. The heteropody is pronounced, with an HI of 29% suggesting that the manual-pedal ratio was around 1:3. The pedal print is deeper than the manual print. Fuentes Vidarte and Meijide Calvo (2001) also described in this slab an isolated partial manual track, not clearly identified in this study. MNS2003/92/8b. This trackway is preserved in the same slab as MNS2003/92/8a (Fig. 5D-F), but the tracks are preserved as natural casts. It is composed of three manual-pedal sets with a generally low-medium MP value (1–1.5), only 1m having a medium-high MP (2). It is a small-sized specimen (Pedal FL= 3.7–4.1 cm; Pedal FW = 3.4–4 cm). The manual prints are pentadactyl, star-shaped and wider than long (FL = 1.5? –1.8 cm; FW = 2.4–2.8 cm; FL/FW ratio = 0.6–0.7), but not all the digital impressions are clearly impressed. The digital impressions are thin and of similar width (0.3–0.4 cm), with an acuminate end. DIII (1.6–1.8 cm) and DIV (1.6–2.2? cm) are the longest, DII is slightly shorter (1.4–1.6 cm), and DI and DV are the shortest and similar in length (1.3–1.5 cm). The interdigital angle is quite variable. In 1m these are as follows: I^II (45º), II^III (51º) III^IV (55º) and IV^V (55º), the total IA being high (206º). DI is oriented medially, DII antermedially, DIII anteriorly, DIV laterally, and DV posterolaterally. DI and DV are in large measure opposed. The manual prints are laterally rotated (37–42º). No clear claw marks are identified in the manual impressions. The pedal prints are tetradactyl and slightly longer than wide (FL/FW ratio =0.92–1.14) because the heel is not preserved and the real length of the autopod cannot be calculated. The central digital impressions (DII and DIII) are longer than the lateral and medial ones. DIII is the longest (3.7–4.1 cm) and DII (3.6–3.9 cm) is slightly shorter, whereas DIV (2.5–2.8 cm) and DI are even shorter (2.5–2.8 cm) and are similar in length. The digital impressions are thin, with widths varying (WI-WIV 0.2–0.5 cm) among the different impressions but also in different pedal prints. The digital impressions have an acuminate end with no clear evidence of claw marks. The orientation of the digital impressions is medial (DI), medial to anteromedial (DII), anterior to anterolateral (DIII), and lateral (DIV), with a high (higher than the other specimens) total IA of 72–95º. I^II (23–25º) and II^III (21–30º) are slightly lower than III^IV (28–40º). Pedal rotation is very low, almost 0 (0–4º) and medial. The trackway is intermediate-gauge (Ar/FW = 0.75-0.94). The manual-pedal distance is short (3.5–4.4 cm). PL is very similar for both manual and pedal impressions (11–12 cm). SL is also very similar for manual (19.5 cm) and pedal impressions (18.5 cm). PA is medium-high, with 106° for the pedal prints and 110º for the manual prints. The heteropody is pronounced to medium, with a heteropody index of 27–32% suggesting that the manual- pedal ratio was roughly 1:3. The manual prints are deeper than most of the pedal print, with the exception of the anterior part, the heel area being shallower than the anterior part of the footprint. There is no evidence of overprinting, or of tail or belly drag marks. The estimated glenoacetabular distance is 13 cm. MNS2002/96/10. MNS2002/96/10 (Fig. 5G-I) is composed of two consecutive manual-pedal sets preserved as true tracks with a rather low-medium MP value (1–1.5). These two set of tracks correspond to the mold of the first two sets of the specimen MNS2003/92/8b (part and counterpart). It is a small-sized specimen (Pedal FL= 3.8–4 cm; Pedal FW = 3.6 cm). The manual prints are star-shaped but not all the digital impressions can be distinguished, so the prints look tetradactyl. They are wider than long (FL = 1.4–1.8 cm; FW = 2.2–2.7 cm; FL/FW ratio = 0.63–0.66). The digital impressions are thin, with variable widths (0.2–0.4 cm) and an acuminate end. DIII (1.5–1.8 cm) and DIV (1.5–1.7 cm) are the longest, DII is similar or slightly shorter (1.3–1.8 cm), and DI and DV are the shortest and similar in length (1.3–1.5 cm). The digital divarication angles are quite variable: I^II (42º), II^III (55º), III^IV (38–63º) and IV^V (47–50º). The total IA is high (210º). DI is oriented medially, DII anteromedially, DIII anteriorly, DIV anterolaterally, and DV laterally. DI and DV are largely opposed (e.g. 1m). The manual prints seem laterally rotated. No clear claw marks are identified in the manual impressions. The pedal prints are tetradactyl and slightly longer than wide (FL/FW ratio =1.05–1.1) because the heel is not preserved and the real length of the autopod cannot be calculated. The central digital impressions (DII and DIII) are longer than the lateral and medial ones. DIII is the longest (3.8–4 cm), DII (3.6–3.7 cm) is slightly shorter, whereas DIV (2.4–3 cm) and DI are the shortest (2.6–2.9 cm) and are similar in length. The digital impressions are thin, with widths varying from 0.2–0.5 cm among them but also in different pedal impressions. The digital impressions have an acuminate end with no clear evidence of claw marks. The orientation of the digital impressions is medial (DI), anteromedial (DII), anterior to anterolateral (DIII), and lateral (DIV), with a high (relative to the other specimens) total divarication angle of 77º–99º: I^II (25–26º) and II^III (21–31º) are slightly lower than III^IV (31–42º). The heel area is not preserved, being shallower than the anterior part of the footprint. The manual-pedal distance is short (3.4–3.5 cm). The heteropody varies from pronounced to medium; a heteropody index of 21–35% suggests that the manual-pedal ratio was around 1:3. The manual prints are deeper than most of the pedal print, with the exception of the anterior part of the pedal impressions. There is no evidence of overprinting, or of tail or belly drag marks. MNS2002/96/5. This specimen is an isolated manual track (Fig. 6D) preserved as a true track with a medium MP value (1.5). It is interpreted as a left manual impression and is a large-sized specimen (the largest manual track in the whole sample). It is pentadactyl, star-shaped and wider than long (FL = 2.9 cm; FW = 3.6; FL/FW ratio = 0.8). DIII (2.9 cm) and DIV (2.9–3.2 cm) are the longest digital impressions, DII is slightly shorter (2.6 cm), and DI and DV are the shortest and are slightly different in length (1.8 and 2.1 cm, respectively). The digital impressions are thin, with a width of 0.4 cm and with an acuminate end. The interdigital angles are quite variable, I^II (62º) being the highest, and the others roughly similar to one another: II^III (32º), III^IV (29º) and IV^V (30). The total IA is medium (153º), i.e. considerably lower than in the other specimens. DI and DV are not as opposed, showing a more medial/lateral orientation with respect to the footprint axis. MNS2002/96/6. This specimen is an isolated manual-pedal set preserved as a natural cast with a very low MP value (0.5–1), so it is not described in detail. MNS2002/96/7 and MNS2002/96/8. These specimens are part and counterpart (Fig. 6E–6F) of an isolated right manual-pedal set and are preserved as a natural casts (MNS2002/96/7) and true tracks (MNS2002/96/8), respectively. The manual impression has a medium-high MP value (2), whereas the pedal impression has a low MP (0.5) because only two digital impressions can be distinguished. The manual track is pentadactyl, star-shaped and wider than long (FL = 1.8 cm; FW = 2.4; FL/FW ratio = 0.75). DII and DIII are equal in length (1.8 cm), DIV being slightly shorter (1.5 cm), and DI and DV are the shortest and subequal in length (1.2–1.3. cm). III^IV is the highest interdigital angle (59º), whereas the others are more similar to one another: I^II (40º), II^III (41º), IV^V (50º). The total IA is high (190º), albeit lower than in many specimens. The pedal impression only shows marks of two digital impressions, possibly of digits III and IV considering their position, with DIII longer than DIV. The manual- pedal distance is short (about 3.9 cm). MNS2002/96/12. This specimen (Fig. 6A-C) is composed of different tracks preserved as natural casts. MNS2002/96/12t1 is an isolated manual-pedal set and is a small-sized specimen (Pedal FL= 3.7). The manual impression has a medium MP value (1.5), whereas the pedal has a low MP (1). The manual print shows the typical pentadactyl star-shaped morphology. It is wider than long (FL = 1.8 cm; FW = 2.1 cm; FL/FW ratio = 0.85). DIII (1.8 cm) is the longest digital impression, DII and DIV are subequal in length and slightly shorter (1.4–1.5 cm), and DI and DV are the shortest (0.9 and 1.1). The digital impressions are thin (0.3 cm in width) with an acuminate end. The interdigital angles are quite variable, I^II (51º) and IV^V (60º) being considerably higher than II^III (35º) and III^IV (38º); the total IA is high (184º). DI and DV are slightly opposed. No clear claw marks are identified in the manual prints. Pedal tracks only shows the first three digital impressions. DIII (3.7 cm) is the longest, DII (3.4 cm) being slightly shorter and DI (2.1 cm) considerably shorter. The digital impressions are thin (W = 0.3–04 cm). DIV and the heel impression are not preserved. The digital impressions have an acuminate end, with possible evidence of claw marks (e.g. DII). The divarication angle would possibly be high, since I^III is 46º. The manual-pedal distance is 4.4 cm. The manual print is deep, indeed similar in depth to the anterior part of the pedal impression. MNS2002/96/t2 is a manual-pedal set preserved as natural casts (B1p-B1m), with a low-medium MP value (1–1.5). The manual-pedal set is thought to be a right one, and is a very small-sized specimen (Pedal FL= 2.5 cm). The manual print is almost complete, pentadactyl, star-shaped and wider than long (FL = 1.4; FW = 1.8; FL/FW ratio = 0.77). In this case, DII is the longest digital impression, DIII being slightly shorter (1.4 cm), whereas DI, DIV and DV are shorter and similar to one another in length (1–1.1 cm). The digital impressions are thin and of similar width (0.2 cm), with an acuminate end. The interdigital angle is variable, I^II (30º) and II^III (36º) being considerably lower than III^IV (52º) and IV^V (58º); the total angle is high (176º). The manual track is laterally rotated with respect to the pedal print. The pedal print is partially preserved and shows three digital impressions. The central digital impressions (DII and DIII) are longer than the medial one, DIII (2.5 cm) being the longest, DII (2.1 cm) slightly shorter and DI (1.5 cm) the shortest. The digital impressions are very thin (W = 0.2–0.3 cm). The total divarication angle would have been low (I^III = 33º). The heel impression is not preserved. The manual-pedal distance is short (2.7 cm). The specimen also preserves other small tracks in the sample, which seem to be tridactyl tracks. They are possibly the impressions of two mani that left only three digits (DII- DIV). These impressions appear to be of roughly similar proportions to B1m, with DIII being longer than DII and DIV. TABLE S1: Measurements of the paratypes and referred specimens of Crocodylopodus meijidei. MP, Morphological preservation value (Marchetti et al., 2019); FL, footprint length; FW, footprint width; FL/FW, footprint length/footprint width ratio; LI, LII, LIII, LIV, LV, digital impression length; WI, WII, WIII, WIV, WV, digital impression width; I^II, II^III, III^IV, IV^V, Total IA, interdigital divarication angles; Dm–p, manus-pes distance; HI, heteropody index; PL, pace length; SL, stride length; PA/ANG, pace angulation (PA, center of the footprint; ANG, tip of the impression of digit III); FR, footprint rotation (I, inward; O, outward); OW = overall width; GA, glenoacetabular distance. Ar = distance from center of the track to the midline. FL, FW, LI, LII, LIII, LIV, LV, WI, WII,WIII, WIV, WV, Dm-p, PL, SL, OW, GA, Ar in cm. I^II, II^III, III^IV, IV^V, Total IA, PA, ANG, FR in degrees (º). HI, PTR, MTR in %. Trackway Total Dm– FL x MNS2003/92/8a MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW HI PL SL PA ANG FR OW GA Ar Ar/FW 1p 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 4.8 ? ? 12 19.5 ? ? 10I? ? 14.2 ? ? 1m 1 1.3 2.1 0.61 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 4.8 ? ? 10.2 19 ? ? 28O? ? 14.2 none none 2p 2.5 5 4.1 1.21 3.6 4.5 5 4 none 0.4 0.4 0.3 0.3 none 19 21 25 none 65 4.3 20.5 29 12.5 19.5 107 110* 12I 10.6 ? 3.3 0.8 2m 2.5 2.1 2.9 0.7 1.3 1.8 2.1 1.8 1.4 0.2 0.4 0.4 0.3 0.2 43 65 67 59 234 4.3 6.09 29 12 ? 120 110? 33O? 8.2 ? none none 3p* 2 5.1 3.9 1.3 3.7 4.8 5.1 4.1 none 0.3 0.4 0.4 0.4 none 17 20 23 none 60 4.6 19.89 31 11.5 ? 109 110 20I 10.5 ? 3.3 0.84 3m 2 2 3.1 0.64 1.6 2 1.9 1.7 1.5 0.3 0.3 0.3 0.3? 0.3 40 60 61 42 203 4.6 6.2 31 ? ? ? ? 40O 8.7? ? none none 4p* 2 4.9 4.1 1.19 3.3 4.5 4.9 4.1 none 0.4 0.4 0.4 0.4 none 16 28 23 none 67 ? 20.09 ? ? ? ? ? 6I 10.8 ? 3.3 0.8 Observations: * heel inferred. DIII estimated in 1p Trackway Total Dm– FL x MNS2002/96/4 MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW HI PL SL PA ANG FR OW GA Ar Ar/FW 1p* 0.5 ? ? ? ? 0.9 1.5 ? none ? 0.4 0.4 ? none ? ? ? ? ? 4? ? ? 11 19.2 ? ? 5I ? 12.8* ? ? 1m 2 1.9 2.7 0.7 1.2 1.6 1.9 2.2 1.5? 0.2 0.4 0.4 0.3 0.3 47 51 49 59 206 4? 5.13 ? 10.5 19.1 ? ? 23O 8.4 12.8* none none 2p 1.5 3.7 3.8 0.97 2.8 3.7 3.7 2.8 none 0.3 0.4 0.3 0.2 none 20 31 38 none 89 4.2 14.06 36 10.6 ? 110? 100 ? 11 ? 2.6 0.68 2m 2 2 2.6 0.76 1.5 1.9 2 1.7 1.4 0.3 0.3 0.3 0.3 0.2 35 62 46 57 200 4.2 5.2 36 11.2 ? 118 105 32O? 11 ? none none 3p 1.5 4.1 3.6 1.13 2.5 3.6 4.1 3.1 none 0.5 0.5 0.4 0.4 none 27 22 22 none 71 3.5 14.76 38 ? ? ? ? 3I 12.7 ? 3.2? 0.88? 3m 1.5 2.1 2.7 0.77 1.5 1.8 2.1 1.8 1.4 0.2? 0.3 0.4 0.5 0.5 56 43 46 57 202 3.5 5.76 38 ? ? ? ? 43O? 11.2 ? none none Observations: * 1p broken not complete 2p and 3p heel poorly preserved. Estimations. Manus/pes set Total Dm– FL x MNS2002/96/3 MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW HI 1p 2.5 7 5.6 1.25 4.8 6.3 7 6.2 none 0.6 0.6 0.8 0.8 none 13 21 27 none 61 9 39.2 29? 1m 0.5 2.8? 4.1? 0.68 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 9 11.48 29? Trackway Total Dm– FL x MNS2003/92/8b MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW HI PL SL PA ANG FR OW GA Ar Ar/FW 1p* 1.5 3.7 4 0.92 2.5 3.6 3.7 2.8 none 0.4 0.5 0.5 0.4 none 25 30 40 none 95 3.5 14.8 32 11 18.5 ? ? 4I 10 13 3? 0.75 1m 2 1.7 2.8 0.6 1.3 1.6 1.7 2.2? 1.5 0.4 0.4 0.3 0.3 0.3 45 51 55 55 206 3.5 4.76 32 11.5 19.5 ? ? 37O 10 13 none none 2p 1.5 3.9? 3.4 1.14 2.6? 3.6? 3.9? 2.7? none 0.3 0.4 0.4 0.3 none 23 21 28 none 72 3.8 13.26 27 12 ? 106? ? 0 11.5 ? 3.2 0.94 2m 1 1.5? 2.4 0.6 ? 1.4 1.6 1.6 1.4 ? 0.3 0.4 0.4 0.3 38? 48 51 43 180 3.8 3.6 27 12 ? 110 105 42O 10.5 ? none none 3p 1 4.1? 3.8 1.07 2.8 3.9 4.1? 2.5 none 0.4 0.5 0.4 0.2 none 23 28 37 none 88 4.4 15.58 28 ? ? ? ? 3I 12.1 ? 3.3 0.86 3m 1 1.8 2.5 0.7 1.4 1.4 1.8 1.9? 1.4? 0.3 0.4 0.4 0.3 0.3 51 59 61 50 221 4.4 4.5 28 ? ? ? ? 42O 10.5 ? none none Observations: * heel inferred Tracks Total Dm– FL x MNS2002/96/10 MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW HI 1p* 1 3.8 3.6 1.05 2.9 3.7 3.8 2.4 none 0.4 0.5 0.3 0.2 none 26 31 42 none 99 3.5 13.68 35 1m 1.5 1.8 2.7 0.66 1.5 1.8 1.8 1.7 1.3 0.4 0.4 0.3 0.2 0.2? 42 55 63 50 210 3.5 4.86 35 2p 1 4 3.6 1.1 2.6 3.6 4 3 none 0.3 0.4 0.3 0.2 none 25 21 31 none 77 3.4 14.4 21? 2m 1 1.4 2.2 0.63 ? 1.3? 1.5? 1.5? 1.4 ? 0.4? 0.3? 0.2? 0.2? ? 55? 38? 47? ? 3.4 3.08 21? Observations: * heel inferred FL Slab Total Dm– x MNS2002/96/12 MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p FW PL t1.1p 1 3.7 ? ? 2.1 3.4 3.7 ? None 0.3 0.3 0.4 ? None 23 23 ? none 4.4 ? none t1.1m 1.5 1.8 2.1 0.85 0.9 1.5 1.8 1.4 1.1 0.3 0.3 0.3 0.3 0.3 51 35 38 60 184 4.4 3.78 none t2.1p 1 2.5 ? ? 1.5 2.1 2.5 ? None 0.3 0.2 0.2 ? None 19 14 ? none ? 2.7 ? ? t2.1m 1.5 1.4 1.8 0.77 1 1.6 1.4 1.1 1.1 0.2 0.2 0.2 0.2 0.2 30 36 52 58 176 2.7 2.52 8.1? Total Track MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA 2.9- MNS2002/96/5 1.5 2.9 3.6 0.8 1.8 2.6 2.9 3.2? 2.1 0.4 0.4 0.4 0.4 0.4 62 32 29 30 153 Manus/pes set Total Dm– MNS2002/96/7 MP FL FW FL/FW LI LII LIII LIV LV WI WII WIII WIV WV I^II II^III III^IV IV^V IA p 1p 0.5 ? ? ? ? ? 1.8 1.4 ? ? 0.3 0.3 ? ? ? ? ? ? ? 3.9 1m 2 1.8 2.4 0.75 1.2 1.8 1.8 1.5 1.3 0.3 0.3 0.3 0.3 0.3 40 41 59 50 190 3.9 TABLE S2. Comparison of the Crocodylopodus tracks described in the fossil record. IA, interdigital divarication angles (in degrees (º)). HI, heteropody index. Heteropody: pronounced (HI < 35%), medium (35-70%) or low (HI > 70%). PA, pace angulation. Ar = distance from the center of the track to the midline; FW = footprint width. Trackway gauge: narrow (Ar/FW < 0.5), intermediate Ar/FW (0.5-1) and wide (Ar/FW >1).*data estimated from the drawing. Digital Age and Trackway- Previous Size FL/FW Manual impression Heel Heteropody Other different Trackway Geological IA PA gauge References Assignations category ratio morphology lengths in pedal morphology (HI) features Formation (Ar/FW) prints Lower MNS2002/96/2bis Cretaceous small- DIII> DII > DIV pronounced Intermediate C.meijidei 1.37 pentadactyl 56 subtriangular 127 This work Lectotype Huérteles medium >DI (30 %) (0.58-0.62) Fm Middle Jurassic small DIII> DII > DIV pronounced Intermediate Abbassi et al., LBP-Type 1 Crocodylopodus isp. 1.3 tetradactyl 63 rounded 108 tail present Hojedk (4.8 cm) >DI (32 %) (0.72) 2015 Fm. Middle-? DI-DV in manual Upper medium DIII> DIV > DII pronounced CDUE 728 C.meijidei 1.3 pentadactyl 90 elongated None ? print more Klein et al., 2018 Jurassic (6 cm) >DI (22 %) medially/laterally Isli Fm. Upper Avanzini et al., Jurassic small not DIII> DII > DIV not Intermediate digitigrade pes; MUJA 0101 C.meijidei 0.96 110 unknown 126 2007; Lastres (2.8 cm) preserved = DI preserved (0.85) drag marks 2010 Fm. Upper tridactyl Avanzini et al., Jurassic small DIII> DII > DIV not medium 117- Intermediate Phalangeal pads MUJA 0102 C.meijidei 1.3 DI-DV not 45 2007; Lastres (2.1 cm) >DI preserved (48%) 140 (0.8-0.96) in the pes. preserved 2010 Fm. Upper Jurassic medium DIII> DII= DIV> medium DIV oriented Avanzini et al., MUJA0038 C.meijidei 1.23 pentadactyl 170 subtriangular ? ? Lastres (8 cm) DI (40%) very laterally 2007 Fm. Upper Jurassic medium not DII > DIII> Castanera et al. SHN.(JJS).ICNO.62 Crocodylopodus isp 0.88 86 subtriangular unknown ? ? Alcobaça (7.5 cm) preserved DIV> DI 2021 Fm. digital pads; Berriasian large DIII> DII= DIV> pronounced Intermediate tail marks? Pascual et al., VALD-NV-T2 Huérteles cf. Crocodylopodus (21.4 1.09 pentadactyl 48.5 oval 114 DI (29 %) (0.65) Pes laterally 2005 Fm. cm) rotated Lower medium tridactyl to DIII> DII= DIV> subtriangular medium wide DI-DIII Lockley et al., CUE E4 C001 Cretaceous Crocodylopodus isp. 0.97 42* 89.3 (7.1 cm) pentadactyl DI to elongated (36%) (1.2) anteriorly 2020 Jinju Fm. Lower CUE Ji 3rd medium DIII> DII= DIV> subtriangular low wide DI-DIII Lockley et al., Cretaceous Crocodylopodus isp. 1.69 pentadactyl 45* 84 PCS001 (8.8 cm) DI to elongated (73%) (1) anteriorly 2020 Jinju Fm. Not Lower CUE E100516- small not calculated DI-DIII Lockley et al., Cretaceous Crocodylopodus isp. 1.1 pentadactyl DIII> DII > DI ? low (89%) 106 Cr001-1 (3.6 cm) preserved pes anteriorly 2020 Jinju Fm. incomplete Lower wide* DI-DIII Lockley et al., CUE E100516- Cretaceous medium not well (1.2 but anteriorly 2020 Cr001-2 Jinju Fm. Crocodylopodus isp. (5.4 cm) 1.74 pentadactyl DII > DIII> DI ? preserved low (81%) 105 variable) Simpson et al., Upper cf. Crocodylopodus large not DIII> DII= DIV> 2010; Wesses Canyon Cretaceous 1.4 60 subrounded unknown ? ? cf. Hatcherichnus (14 cm) preserved DI Lockley et al., Wahweap 2020 Upper Cretaceous medium not DIV oriented Serraduy Norte cf. Crocodylopodus 1.2* DIV> DII > DI 70* subtriangular unknown ? ? Vila et al., 2015 Tremp (7.1 cm) preserved very laterally Fm. LITERATURE CITED Abbassi, N., S. D. O. Porchetti, A. Wagensommer, and M. G. Dehnavi. 2015. Dinosaur and crocodylomorph footprints from the Hojedk Formation (Bajocian, Middle Jurassic) of north Kerman, central Iran. Italian Journal of Geosciences 134(1):86– 94. Avanzini, M., J. C. García-Ramos, J. Lires, L. Piñuela, and M. G. Lockley. 2007. Crocodylomorph tracks from the Late Jurassic of Asturias (Spain). Ichnos 14:143–153. Avanzini, M., J. C. García-Ramos, J. Lires, L. Piñuela, and M. G. Lockley. 2010. Preservational morphotypes of Crocodylopodus from the Late Jurassic of Asturias (Spain). New Mexico Museum of Natural History and Science Bulletin 51: 239–244. Castanera, D., E. Malafaia, B. C. Silva, V. F. Santos, and M. Belvedere. 2021. New dinosaur, crocodylomorph and swim tracks from the Late Jurassic of the Lusitanian Basin: implications for ichnodiversity. Lethaia, 54(2), 271–287. Klein, H., A. Lagnaoui, G. D. Gierliński, H. Saber, J. N. Lallensack, M. Oukassou, and A. Charrière. 2018. Crocodylomorph, turtle and mammal tracks in dinosaur-dominated Middle–? Upper Jurassic and mid-Cretaceous ichnoassemblages of Morocco. Palaeogeography, Palaeoclimatology, Palaeoecology 498:39–52. Lockley, M. G., J. D. Lim, H. D. Park, A. Romilio, J. S.Yoo, J. W. Choi, K. S. Kim, S.-H. Kang, D.H. Kim, and T. H. Kim. 2020. First reports of Crocodylopodus from Asia: implications for the paleoecology of the Lower Cretaceous. Cretaceous Research 104441. Pascual Arribas, C., N. Hernández Medrano, P. Latorre Macarrón, and E. Sanz Pérez. 2005. Nuevo rastro de icnitas de cocodrilo en la Aloformación Huérteles de la Cuenca de Cameros. Studia Geologica Salmanticensia 41:75–91. Simpson, E. L., H.L. Hilbert-Wolf, M.C. Wizevich, S. G. Lucas, E. Tester, S.E. Tindall, and J. J. Bernard 2010. A crocodylomorph track in the Upper Cretaceous Capping Sandstone Member of the Wahweap Formation, Grand Staircase- Escalante National Monument, Utah, USA. New Mexico Museum of Natural History and Science Bulletin 51:165–170. Vila, B., D. Castanera, J. Marmi, J. I. Canudo, and À. Galobart. 2015. Crocodile swim tracks from the latest Cretaceous of Europe. Lethaia 48(2):256–266. TABLE S3. Summary of the main footprint and trackway features in extant crocodylians. FL, Footprint length; TL, Total length; IA, interdigital divarication angles (in degrees (º)); FR, footprint rotation; PA, pace angulation; Ml, midline. * estimated from pictures. Note that some species names and ontogenetic states are abbreviated in the table (e.g.: Pt = Paleosuchus trigonatus; sa = subadult). Size Digital Digital Other category Manual impression Pedal Trackway Species Manual IA orientation in FR PA significant References (FL and morphology lengths in the IA features the pes features TL) pes scale small- tridactyl Pt: 94-97º wide-gauge, impressions Paleosuchus DII-DIII medium to digit IV is Cp: 75- tail marks in manus trigonatus; DI-DV more DIV shallower anterior in sized pentadactyl 20–55º curved 112º (shallower and pes. Kubo 2008 Crocodylus porosus; medially/laterally DI-DIII deeper the pedal claws in DI- anterolaterally Ts: 76- than pedal prints Tomistoma schlegelii impressions 3.3–9.2 cm DII-DIII 89º manus/pes) deeper than manual. DI-DIII with DI and the 140º–160º claw wide-gauge, heel deepest Kumagai and DI-DV more 35-45º manual and about 90º large to very pentadactyl impressions. DI-DIII more drag and tail parts of the Farlow medially/laterally webbing pedal prints (measured Crocodylus acutus large claw marks DII and DIII anterior. DIV marks, pedal print (2010) webbing (especially slightly from 15–24 cm in DI-DIII the longest, DI anterolaterally. overstepping and pes Farlow et al. (especially DIV- DII-DIV) lateral DIII) and DIV manus-pes deeper than (2018) DV) similar. the manus pes with interdigital manual webbing 45-55º large pentadactyl 180º laterally wide-gauge, increases Alligator claw marks in webbing DIV Farlow and 15.5–20 cm claw marks DI-DV more pedal 105º tail and belly from DI to mississippiensis DI to DIII especially anterolaterally Esley, 2010 in DI-DIII medially/laterally parallel to drag marks DIV (DIII-DIV) midline (sometimes not registered) small (juvenile) manual DI-DII anterior- wide-gauge, large sized DI-DIV laterally Milàn and Paleosuchus 179º (j) 86 (j) anterolaterally; 87º (j)- tail mark (j, a) (adult) pentadactyl* subequal in pedal Hedegaard palpebrosus 145º (a) 72º (a) DII-DIV 102º (a) Overprinting 4.8–13.6 cm length* parallel to (2010) anterolaterally* (a) 50 cm (j)– midline* 140 cm (a) Size Digital Digital Other category Manual impression Pedal Trackway Species Manual IA orientation in FR PA significant References (FL and morphology lengths in the IA features the pes features TL) pes medium (subadult) DI-DII anterior- overprinting, large sized DI-DIV pedal Milàn and 82º (sa) anterolaterally; 100º (sa) tail and claw Caiman latirostris (adult) overprinted overprinted subequal in parallel to Hedegaard 72º (a) DIII-DIV 96º (a) marks. belly 7.8–12.2 cm length* midline* (2010) anterolaterally* mark (a) 70 cm (sa)– 135 cm (a) medium- pedal Milàn and sized tail, belly and overprinted overprinted not clear 60º not clear parallel to 93º Hedegaard Alligator sinensis 14.7 cm claw marks midline* (2010) 150 cm medium- sized belly and (subadult) pedal Milàn and 99º (sa) claw scale marks Crocodylus johnstoni 9.2 cm overprinted overprinted not clear 46º (sa) not clear parallel to Hedegaard 66º (a) dragmarks (sa) Large-sized midline* (2010) (adult) 112 cm medium- claw drag sized Milàn and DIV not marks Crocodylus rhombifer (subadult) overprinted overprinted ? ? ? 100º Hedegaard impressed occasional tail 5.1 cm (2010) mark 80 cm manual belly and large-sized DI-DIII strongly claw Milàn and Crocodylus (subadult) overprinted 138º subequal in 36º anterolaterally* lateral, 86º dragmarks Hedegaard novaeguineae 12.5 cm length* pedal slight and faint tail (2010) 175 cm lateral mark large-sized DI-DII anterior- manual narrow- DII-DIII Milàn and (subadult) anterolaterally; laterally gauge. tail Crocodylus siamensis tetradactyl ? subequal; DI- 47º 113º Hedegaard 12.1 cm DIII-DIV pes parallel and claw DIV subequal* (2010) 140 cm anterolaterally* to ml* marks Size Digital Digital Other category Manual impression Pedal Trackway Species Manual IA orientation in FR PA significant References (FL and morphology lengths in the IA features the pes features TL) pes large-sized pedal wide gauge, Milàn and (subadult) Crocodylus overprinted overprinted ? 51º anterolaterally* slightly 94º tail, belly and Hedegaard 12.5 cm cataphractus lateral claw marks (2010) 149 cm medium manual narrower (sa), (subadult) lateral, wide (a) large sized DI-DIII pedal tail marks (a, Milàn and (adult) 42º (sa) anterior- 110º (sa) Osteolaemus tetraspis overprinted overprinted (a) subequal in parallel* sa) Hedegaard 7.5–13.6 43º (a) anterolateral* 101 (a) (a) length* (sa) Belly and (2010) cm pedal claw marks 79 cm (sa)– parallel* (a) (a) 160 cm (a) medium manual Mazin et al., DI-DII sized laterally 2003 DI-DV more anteriorly; tail and drag Crocodylus niloticus pentadactyl tetradactyl 47º* pedal about 98º (estimated medially DIII-DIV marks (about 6 parallel to from drawing anterolaterallly cm) midline of Fig. 4a) DI medially, small sized DII-DIII pedal pentadactyl DII-DIII tail and drag Padian and Caiman crocodilus 4.3 cm overprinted subequal; DI- 45* parallel to 106º* overprinted anteriorly, DIV marks Olsen, 1984 48.6 cm DIV subequal* midline laterally LITERATURE CITED Farlow, J. O. Elsey, R. 2010. Footprints and trackways of the American alligator, Rockefeller Wildlife Refuge, Louisiana. New Mexico Museum of Natural History and Science Bulletin 51: 31–40. Farlow, J.O., N. J. Robinson, C. J. Kumagai, F. V. Paladino, P. L. Falkingham, R. M. Elsey, and A. J. Martin. 2018b. Trackways of the American crocodile (Crocodylus acutus) in northwestern Costa Rica: implications for crocodylian ichnology. Ichnos 25(1):30–65. Kubo, T. 2008. In quest of the Pteraichnus trackmaker: comparisons to modern crocodilians. Acta Palaeontologica Polonica 53(3):405–412. Kumagai, C. J., Farlow, J. O. 2010. Observations on traces of the American crocodile (Crocodylus acutus) from northwestern Costa Rica. New Mexico Museum of Natural History and Science Bulletin 51:41–49. Mazin, J. M., J. P. Billon-Bruyat, P. Hantzpergue, and G. Lafaurie. 2003. Ichnological evidence for quadrupedal locomotion in pterodactyloid pterosaurs: trackways from the Late Jurassic of Crayssac (southwestern France). Geological Society, London, Special Publications 217(1):283–296. Milàn, J., and R. Hedegaard. 2010. Interspecific variation in tracks and trackways from extant crocodylians. New Mexico Museum of Natural History and Science Bulletin 51:15–29. Padian, K., and P.E. Olsen. 1984. The fossil trackway Pteraichnus: Not pterosaurian, but crocodilian. Journal of Paleontology 58:178–184.