Palaeontology
A gravid fossil turtle from the early cretaceous reveals a different egg development strategy in contrast to extant marine turtles
Journal: Palaeontology
Manuscript ID PALA-10-18-4335-OA.R3
Manuscript Type: Original Article
Date Submitted by the Author: n/a
Complete List of Authors: Cadena, Edwin-Alberto; Universidad Del Rosario, Faculty of Natural Sciences and Mathematics Parra, Mary Luz; Centro de Investigaciones Paleontológicas, Parra-Ruge, Juan de Dios; Centro de Investigaciones Paleontológicas , boyaca; Cineca, boyaca Padilla, Santiago; Centro de Investigaciones Paleontológicas, Boyaca
Key words: Fossil eggs, Protostegidae, Testudines, Villa de Leyva, gravid turtle
Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online.
SIappendixVideoS1HorizontalView.mov SIappendixVideoS2CoronalView.mov SIappendixVideoS3SagitalView.mov
Palaeontology Page 1 of 29 Palaeontology
1 2 3 A GRAVID FOSSIL TURTLE FROM THE EARLY CRETACEOUS REVEALS A 4 DIFFERENT EGG DEVELOPMENT STRATEGY IN CONTRAST TO EXTANT 5 6 MARINE TURTLES 7 1,2 2 8 EDWIN ALBERTO. CADENA , MARY L. PARRA RUGE , JUAN DE D. PARRA 9 RUGE2, and SANTIAGO PADILLA BERNAL2 10 1Universidad del Rosario, Facultad de Ciencias Naturales y Matemáticas, Grupo de 11 Investigación en Paleontología Neotropical Tradicional y Molecular (PALEONEO), 12 Bogotá, Colombia; e mail: [email protected] 13 2 14 Centro de Investigaciones Paleontológicas, Villa de Leyva, Colombia; e mails: 15 [email protected], [email protected], [email protected] 16 17 Abstract: Extant sea turtles develop and lay pliable (flexible) eggs; however, it is 18 unknown whether they inherited this reproductive strategy from their closer fossil 19 relatives or if it represents an evolutionary novelty. Here, we describe the first 20 21 undisputable gravid marine fossil turtle ever found, from the early Cretaceous of 22 Colombia, belonging to Desmatochelys padillai Cadena and Parham, which constitutes a 23 representative of the Protostegidae. Using thin sectioning of one of the eggs, as well as 24 scanning electron microscopy coupled with elemental characterisation, 25 cathodoluminescence, and computer tomography, we established that Desmatochelys 26 padillai produced rigid eggs similar to those associated with some extant and fossil 27 freshwater and terrestrial turtles. At least 48 spherical eggs were preserved inside this 28 29 gravid turtle. We suggest that the development of rigid eggs in the extinct marine turtle 30 Desmatochelys padillai resulted as an adaptation for egg embryo requirements dictated 31 by the physical attributes of the nesting site. 32 33 Key words: gravid turtle, fossil eggs, Protostegidae, Testudines, Villa de Leyva. 34 35 Turtles have been a successful group of vertebrates since at least the middle Triassic 36 37 (~240 Ma), adapted to almost all types of terrestrial, freshwater, and marine 38 environments. However, how the embryological development and reproduction of extant 39 linages of turtles differ from or resemble those of their ancestors remains a poorly 40 understood aspect of turtle evolution. Evidence of reproductive strategies in fossil turtles 41 is restricted to eggs, nests, embryos, and coupling pairs (Lawver & Jackson 2014); very 42 few of these findings correspond to gravid mothers. 43 44 At present, the fossil record of gravid turtles with their eggs preserved is restricted to 45 the Upper Cretaceous (Campanian) Adocus sp. (Cryptodira, Pan Trionychia, Adocidae) 46 egg clutch and a gravid individual from the Oldman and Dinosaur Park formations in 47 Alberta, Canada preserving 26 eggs, only two of which are relatively uncrushed 48 (Zelenitsky et al. 2008), and a specimen of Basilemys variolosa (Cryptodira, 49 Nanhsiungchelyidae) from the Upper Cretaceous (late Campanian), found in the 50 Dinosaur Park Formation, Alberta, Canada, in which some eggs were found by accident 51 52 inside its carapace (Braman & Brinkman 2009). Both of these gravid fossil turtles 53 represent freshwater and/or terrestrial forms, and their egg morphologies are consistent 54 with those of extant turtles with the same life style adaptations (Lawver & Jackson 2014). 55 The egg characteristics of extant (freshwater, terrestrial, and marine) and fossil turtles 56 57 58 59 1 60 Palaeontology Palaeontology Page 2 of 29
1 2 3 (freshwater and terrestrial) have been summarised by Hirsch (1983), Schleich & Kästle 4 (1988), Nuamsukon et al. (2009), and Lawver (2017). A potential third example of a 5 6 gravid fossil turtle comes from a Late Jurassic, eurysternid? (Cryptodira, Eucryptodira) 7 from Solnhofen, Bavaria, Germany. However, investigation of the microstructure 8 revealed complete diagenetic alteration of the “eggshell”. Therefore, the authors conclude 9 that these eggs must be referred to as pseudomorphs because no details of the eggshell 10 microstructure can be ascertained (Joyce & Zelenitsky 2002; Lawver & Jackson 2014), 11 leaving this finding highly controversial. 12 13 14 Here, we describe the first undisputable gravid marine fossil turtle ever found, with 15 eggs preserved inside its visceral cavity, found in a shallow marine rock sequence 16 belonging to the Paja Formation in Villa de Leyva, a town in Colombia, South America. 17 We attribute these eggs as Testudoolithus oospp. Thus, we compare the egg development 18 strategies among the truly marine adapted extant clades Dermochelyidae and 19 Cheloniidae, and the extinct Protostegidae and suggest possible explanations for the 20 21 differences among them. 22 23 GEOLOGICAL SETTINGS 24 25 The region of Villa de Leyva has yielded important remains of fossil vertebrates, 26 invertebrates, and plants from the early Cretaceous (upper Barremian–lower Aptian, > 27 120 Ma in age), dated based on ammonoid biostratigraphy (Reboulet & Hoedemaeker 28 29 2006; Reboulet et al. 2014). All these fossils belong to the Paja Formation, which is 30 characterised by a succession of mudstones interbedded with limestones, deposited in an 31 intertidal marine complex (Forero & Sarmiento 1985). Fossil vertebrates from Villa de 32 Leyva include the pliosaurs Kronosaurus boyacensis (Hampe 1992) and 33 Stenorhynchosaurus munozi (Páramo Fonseca et al. 2016); the first dinosaur of Colombia 34 Padillasaurus leivaensis (Carballido et al. 2015); the turtle Desmatochelys padillai 35 (Cadena & Parham 2015), a basal marine protostegid; Leyvachelys cipadi (Cadena 2015), 36 37 the first sandownid turtle from South America; and several unpublished ichthyosaurs, 38 plesiosaurs, and fish. The gravid fossil turtle with its eggs reported herein was discovered 39 by Juan de Dios Parra Ruge at Monsalve hill, located between the towns of Villa de 40 Leyva and Sutamarchan (Supporting Information Fig. S1A). 41 42 MATERIAL AND METHODS 43 44 45 Fossil material and preparation 46 The fossil turtle specimen was prepared at the Centro de Investigaciones Paleontológicas 47 (CIP). The bone surface of the turtle was originally preserved in a hard orange reddish 48 layer of iron rich mudstone. To expose the bone and properly identify the taxonomy, the 49 rock matrix was removed using airscribe tools (microjack) and dental picks. 50 51 52 Light and polarised transmitted microscopy 53 To establish the microstructural preservation and mineralogical composition of the eggs 54 and the rock matrix, we took one of the four isolated eggs collected together with the 55 specimen that separated during the collection and used mechanical preparation of the 56 57 58 59 2 60 Palaeontology Page 3 of 29 Palaeontology
1 2 3 shell to produce a thin section at the Paleontological Lab, North Carolina State 4 University, Raleigh, USA. The whole egg with the rock matrix was processed after it was 5 6 embedding in Silmar SIL66 resin. We extracted bubbles using a vacuum system, with 7 subsequent refrigeration for 24 hours to harden the Silmar resin. Then, we cut the sample 8 and mounted it on a glass slide. We used a precision saw and a grinding machine to reach 9 an average thickness of 70 µm. Finally, we observed the thin section and took 10 photographs using an Olympus BX53 polarised light microscope with 2x, 4x, and 10x 11 lenses at the Paleontological Lab of Yachay Tech University in San Miguel de Urcuquí, 12 Ecuador. 13 14 15 Computer tomography scanning of the specimen 16 To establish the total number of eggs preserved inside the shell, as well as their 17 characteristics, internal morphology, and preservation, we scanned the whole specimen 18 using computer tomography (CT scanning) at the Hospital San Ignacio X ray facility, 19 Bogotá, Colombia, under the following settings: zoom factor 1.16x, 12,902 total images, 20 21 120 kV, 150 mA, and 2 mm interslice spacing (horizontal view (Data Archiving 22 Statement Video S1), coronal view (Data Archiving Statement Video S2), and sagittal 23 view (Data Archiving Statement Video S3)). We analysed the CT data using the iQ 24 VIEW software. 25 26 Scanning electron microscopy and elemental analysis (SEM-EDS) 27 We performed elemental analysis combined with high resolution imaging of the eggshell 28 29 structure and rock matrix of one of the isolated eggs using a scanning electron 30 microscope coupled with an energy dispersive X ray spectroscopy analyser, Phenom 31 ProX, at the Paleontological Lab of Yachay Tech University, San Miguel de Urcuquí, 32 Ecuador. SEM EDS allowed us to explore the mode of preservation of the eggshell 33 ultrastructure and to establish the elemental variation from the top to the base of the 34 eggshell. The eggshell of one of the internal eggs (region without application of stabiliser 35 Paraloid B 72) was carefully removed and transported to the laboratory in sterilised 36 37 aluminium foil. In the laboratory, we carefully fractured the egg fragment, and one of the 38 smaller pieces (a fresh eggshell surface) was observed and analysed under SEM EDS. 39 We mounted the sample in a holder and analysed it without adding any coating. Imaging 40 was performed at 5 kV using different magnification settings. During the entire process 41 of collecting the sample, breaking, and placing it in the SEM holder we used sterilized 42 gloves and avoid any direct contact of the sample with hands. Point and map elemental 43 44 composition were performed at 15 kV. 45 46 Reproductive traits versus phylogeny 47 To explore the potential phylogenetic role in egg development strategy between fossil 48 and extant lineages of marine turtles (Chelonioidea), we mapped information from the 49 egg shape, eggshells and clutch size on top of the simplified phylogenetic tree of Evers & 50 Benson (2018). Information for Chelonioidea was obtained as follow: for Desmatochelys 51 52 padillai (this study) and the extant sea turtles Dermochelys coriacea (data taken from 53 Solomon & Watt (1985)), for members of the crown Cheloniidae: Caretta caretta (data 54 taken from Schleich & Kästle (1988)), Lepidochelys kempii (data taken from Hirsch 55 (1983)), Chelonia mydas, and Eretmochelys imbricata (data for these two taxa taken from 56 57 58 59 3 60 Palaeontology Palaeontology Page 4 of 29
1 2 3 Sikiwat et al. (2015)). For other groups inside Testudines we took information from: 4 Vanzolini 2003; Winkler & Sánchez Villagra 2006; Zelenitsky et al. 2008; Lawver & 5 6 Jackson 2014; Lawver et al. 2015 and Erickson & Baccaro 2016. 7 8 Cathodoluminiscence analysis 9 We used the same sample analysed under SEM EDS by mounting it on a stub and 10 studying it in a TESCAN Mira 3 scanning electron microscope coupled with a 11 cathodoluminescence colour detector and backscattered electrons at 7 kV at the Centro de 12 Nanociencia y Nanotecnología (CENCINAT) of the Universidad de las Fuerzas Armadas 13 14 (ESPE), Sangolqui, Ecuador. 15 16 Institutional abbreviations. FCG CBP, Fundación Colombiana de Geobiología, Carlos 17 Bernardo Padilla, Colección Centro de Investigaciones Paleontológicas (CIP), Villa de 18 Leyva, Boyacá, Colombia. NMW, Natural History Museum, Vienna, Austria. 19 20 21 SYSTEMATIC PALAEONTOLOGY 22 Oofamily TESTUDOOLITHIDAE Hirsch 1996, sensu Jackson et al. 2008 23 Genus TESTUDOOLITHUS Hirsch 1996, sensu Jackson et al. 2008 24 TESTUDOOLITHUS oospp. 25 Figures 1–3 26 27 Referred material. FCG CBP 75 partially preserved carapace and plastron with a clutch 28 29 containing at least 48 eggs inside the visceral cavity and four isolated eggs. 30 31 Remarks. The FCG CBP 75 eggs are attributed to the Testudoolithus oogenus by: being 32 spherical–spheroidal in shape and having an eggshell thickness between approximately 33 200 and 800 µm. Even though the eggs exhibited some unique characteristics, we 34 avoided erecting a new ootaxon considering that most of them have suffered some degree 35 of diagenesis. The unique combination of characters include: spherical eggs, 32.2 to 42.8 36 37 mm in diameter; 785–820 µm eggshell thickness; shell units with a height to width ratio 38 of 1.1:1 to 1.3:1; well defined pores 180–215 µm in diameter at the eggshell surface. 39 40 Description and comparisons. 41 Specimen FCG CBP 75 resembles the holotype of Desmatochelys padillai in many 42 characteristics, including poor lateral ossification of posterior costals exhibiting naked 43 44 costal ribs, as well as developing dorsal keels at the neural series (Supporting Information 45 Figs. S3), and is similar to many extant sea turtles, such as the Caretta caretta specimen 46 NMW 31531, in that the most posterior pair of costal ribs project almost parallel to the 47 middle axis of the shell (Supporting Information Figs. S2E–H). Thus, FCB CBP 75 48 resembles other chelonioid turtles in having hyo and hypoplastral star shaped elements 49 with extremely serrate medial edges and large central fontanelles between them (Figs 50 1C–D, see also Hirayama 1997 figs 3–6). Considering all these similarities between 51 52 specimen FCG CBP 75, the holotype of D. padillai, and other chelonioids, we attribute 53 this specimen to D. padillai, and exclude the possibility that it could correspond to a 54 specimen of the sandwonid Leyvachelys cipadi (Cadena 2015), which is characterised by 55 a fully ossified carapace. Unfortunately, the preservation of the bone surface is not 56 57 58 59 4 60 Palaeontology Page 5 of 29 Palaeontology
1 2 3 optimal and does not allow clear recognition of any additional features of this specimen, 4 or clearly defined sutural contacts between the preserved bones, particularly between 5 6 costal and neural bones, or even establishing if the bones were covered by scutes and left 7 marked sulci between them. 8 Detailed comparisons between fossil ootaxa have been presented (Lawver & Jackson 9 2014; 2017; Imai et al. 2016). Four of these ootaxa resemble Testudoolithus oospp. FCG 10 CBP 75 from Desmatochelys padillai in that they are spherical to spheroidal in shape; 11 these ootaxa: 1. T. jiangi (Fang et al. 2003; Jackson et al. 2008) from the early 12 Cretaceous (Albian) of China, potentially from a terrestrially adapted turtle; 2. T. rigidus 13 14 (Hirsch 1996), from several Mesozoic and Cenozoic localities of England, Ethiopia, 15 Spain, France, and the United States, also interpreted as laid by terrestrially adapted 16 turtles; 3. T. lordhowensis (Lawver & Jackson 2016) from the middle to late Pleistocene 17 of Australia, laid by Meiolania platyceps, a meiolaniform terrestrially adapted turtle; and 18 4. T. zelenitskyae (Lawver & Jackson 2017) from the Upper Cretaceous (Campanian) of 19 the United States and Canada, from a turtle potentially related to the genus Adocus, and 20 21 potentially freshwater adapted. 22 Testudoolithus oospp. described herein differs from T. jiangi and T. zelenitskyae in the 23 shell unit height to width ratio, which is 2:5:1 to 3:1 for T. jiangi and 3.15:1 to 5.5:1 for 24 T. zelenitskyae. It differs from T. rigidus in having a thicker eggshell, 220–240 µm in 25 thickness, and in the shell unit height to width ratio, which is 2:1 for the latter. This 26 specimen is smaller than T. lordhowensis, which measures 53.9 mm in diameter. 27 28 29 RESULTS 30 31 Distribution of the eggs inside the shell 32 The eggs are preserved randomly inside the visceral cavity of the gravid turtle, from the 33 base (ventral surface) of the carapace to the top (dorsal surface) of the plastron. Some of 34 them are visible on the most dorsal and ventral surfaces of the carapace and plastron in 35 places where the bones have an extremely low degree of ossification, such as at the most 36 37 lateral portions of costal bones, at sutures between them, or at the fontanelles (Fig. 1A– 38 D). The whole turtle specimen was found fractured in several blocks, which allowed 39 observation of the eggs from different views, as well as study of their preservation 40 (Supporting Information Fig. S1B–E). Four eggs were collected separately or were 41 dislodged during preparation of the specimen (Fig. 1E). The external surfaces of the 42 eggshells exhibit a micropitted texture (Fig. 1F). 43 44 45 Microstructure and mineralogical composition of the eggshell 46 We studied a thin section of one of the eggs of Desmatochelys padillai using light and 47 polarised transmitted microscopy, which showed that the preservation of the eggshell 48 varies along the surface of the egg, with the presence of calcite crystals and others more 49 iron rich carbonates (siderite) and haematite (Fig. 2A–B). The calcitic regions of the egg 50 better preserve the details of the eggshell units, superficial pores, and a thin layer that we 51 52 interpret as the remains of the eggshell membrane. The shell units are slightly taller than 53 wide, with a height to width ratio that varies from 1.1:1 to 1.3:1 (Fig. 2C–H). Accretion 54 lines in the eggshell are very abundant at the outer portion of the shell units and at the 55 regions where the pores are located (Fig. 2D, F, H). Pores are well defined with an 56 57 58 59 5 60 Palaeontology Palaeontology Page 6 of 29
1 2 3 opening diameter between 180 and 215 µm at the eggshell surface (Supporting 4 Information Fig. S2A–D). Some of the eggs are filled with recrystallised calcite (Fig. 2A) 5 6 and/or a ferroaluminosilicate mineral, which gives them an orange reddish colour (Fig. 7 1C). 8 9 Three-dimensional exploration and egg counting 10 Computer tomography (CT scan) study of the whole gravid turtle revealed the 11 preservation of at least 48 eggs inside the visceral cavity (Fig. 3A–B). Some of the eggs 12 are internally empty, others have a moderate rim of calcite crystals, and the remaining 13 14 eggs exhibit their yolk albumen regions fully recrystallised or with some empty internal 15 pockets (Fig. 3C). Thus, we analysed each of the eggs using the magnification tool of the 16 iQ VIEW software and found that some of them underwent breaking of their eggshells, 17 with broken fragments preserved toward the inner portion of the egg. We also measured 18 all preserved eggs, and established that their diameter range varied from 32.2 mm to 42.8 19 mm (see example Fig. 3D). 20 21 CT scanning also allowed us to better outline the shell bones, particularly the costal 22 and plastral elements of the turtle. The thickness of the costal bones varied between 0.5 23 and 0.8 cm, and the plastral elements ranged from 0.4 to 0.6 cm (Fig. 3E). Additional 24 features of the shell are also visible, such as the central regions of the last series of neural 25 bones (Fig. 3F–G), portions of the pelvic girdle (Fig. 3H–I, Supporting Information Fig. 26 1B–C), and potential fragments of isolated peripheral bones (Supporting Information Fig. 27 S1D–E). 28 29 30 Eggshell fine-scale observation and elemental composition 31 We studied the eggshell of one of the eggs from the top (exterior region) to the base (near 32 the shell membrane) (Fig. 4A–B) using SEM, and we established that the eggshell lacks 33 well defined structures or arrangement of needle like crystals (aragonite) in the shell 34 units or shell membrane region, with only intercrossing crystals of calcite. Even though 35 an aragonite eggshell is considered an autapomorphy of the Testudines (Lawver 2017), 36 37 both forms of calcium carbonate (calcite and aragonite) can constitute the eggshells of 38 some extant sea turtles (Baird & Solomon 1979; Solomon & Watt 1985). Thus, it is 39 possible that the original mineralogical composition of the eggs of Desmatochelys 40 padillai was calcite instead of aragonite and that the fossil diagenetic processes that 41 affected these eggs involved recrystallisation of original calcite, transformation of calcite 42 to siderite (iron rich carbonate), and haematite growth (Fig. 4, Supporting Information 43 44 Fig. S4A–B); other fossil ootaxa exhibit similar diagenetic alterations (Lawver & Jackson 45 2014). 46 For the elemental analysis of the eggshell of Testudoolithus oospp., we used SEM 47 coupled with energy dispersive X ray spectroscopy (EDS). We observed compositional 48 variation from the top to the base of the eggshell in the sample. The top of the eggshell is 49 mostly composed of abundant concentrations of calcium, nitrogen, oxygen and carbon, 50 and traces of aluminium (Table 2, Fig. 4B–C); the latter is abundant only at the most 51 52 external layer of the eggshell. The occurrence of calcium, carbon, and oxygen support the 53 presence of calcium carbonate (calcite form, based on the geometry and morphology of 54 the crystals, resembling character 1 (Lawver 2017)). The middle portion of the eggshell is 55 exclusively dominated by calcium, carbon, oxygen, and nitrogen; again, we interpreted 56 57 58 59 6 60 Palaeontology Page 7 of 29 Palaeontology
1 2 3 this region as calcium carbonate dominated (Table 2, Figs 4B, D). The base of the 4 eggshell is dominated by carbon and oxygen, with minor contents of calcium, iron, silica, 5 6 aluminium, and magnesium (Table 2, Fig. 4B, E). This high carbon content was also 7 detected at the ventral surface of the eggshell (egg membrane surface), together with 8 some small circular bumps, which seem to represent the organic cores where the shell 9 units originated (Table 2, Supporting Information Fig. S5A–B). We hypothesise that the 10 occurrence of high carbon content at the ventral surface together with nitrogen in the 11 middle portion of the eggshell could indicate that organic remains of the original egg 12 membrane may have been preserved in some of these eggs. Richer carbon content at the 13 14 ventral surface of the eggshell (membrane inner fibrous layer) in the eggs of the extant 15 leatherback sea turtle Dermochelys coriacea has been also reported (Areekijseree et al. 16 2013). Testing this hypothesis based on specific egg antibodies and/or mass spectrometry 17 is beyond the scope of this project, but we hope it will be addressed in future studies. 18 In addition to the analysis of the eggshell composition, the rock matrix covering the 19 carapace, the plastron, and some of the eggs was also analysed using SEM EDS, which 20 21 revealed that its composition was dominated by silica, iron, oxygen, and aluminium 22 (indicating a ferroaluminosilicate mineral), without any traces of nitrogen or carbon 23 (Table 2, Supporting Information Fig. S5C–D), excluding potential contamination of the 24 entire specimen, shell, eggs, and rock matrix by an external organic source. 25 To evaluate the fossil diagenesis of the eggshell of Desmatochelys padillai studied 26 under SEM EDS, we analysed the same sample using cathodoluminescence. We found 27 that the central region (Fig. 4F) and ventral surface (Supporting Information Fig. S6A) of 28 29 the eggshell exhibited purple to cyan fluorescence corresponding to wavelengths between 30 400 and 500 nm (Fig. 4F). Chicken eggshell and some extant and fossil turtles exhibit 31 dull blue to dark green luminescence (England et al. 2006; Lawver & Jackson 2016, 32 2017) which in the case of fossils indicates minimal alteration of the original 33 composition. In contrast, bright orange to red luminescence indicates the opposite (high 34 alteration and replacement) (Lawver & Jackson 2016, 2017). 35 36 37 DISCUSSION 38 39 Differences between extant and extinct marine turtle egg development 40 Eggs of extant sea turtles exhibit highly pliable (flexible) shells, shell membranes thicker 41 than the eggshell, shell units with greater width than height, collapsible shells, and visible 42 pores (missing in Caretta caretta) (Baird & Solomon 1979; Hirsch 1983; Solomon & 43 44 Watt 1985; Schleich & Kästle 1988; Nuamsukon et al. 2009; Lawver & Jackson 2014; 45 Sikiwat et al. 2015) (Table 1). In contrast to all of these characteristics, the fossil gravid 46 specimen of Desmatochelys padillai, which was also a fully marine adapted turtle 47 (Cadena & Parham 2015), had rigid eggshells, similar to those exhibited by some extant 48 and fossil freshwater and terrestrial turtles. 49 Similarities between extinct and extant lineages of sea turtles can also be established 50 from the gravid fossil specimen of Desmatochelys padillai described herein. In addition 51 52 to the egg size and shape (circularity), the high number of eggs (52 in total preserved for 53 D. padillai) is similar to the typical numbers of eggs in extant nests of Dermochelys 54 coriacea (50 to 100) (Hendrickson 1980; Wallace et al. 2006), which is the extant taxon 55 most closely related to the Protostegidae phylogenetically (Hirayama 1998, Cadena & 56 57 58 59 7 60 Palaeontology Palaeontology Page 8 of 29
1 2 3 Parham 2015, Evers & Benson 2018). This similarity indicates that marine protostegid 4 turtles were also ecologically more r-selected vertebrates that produced many offspring, 5 6 and thus increased the possibility of survival of at least few individuals under high rates 7 of mortality due to climatic or ecological factors (Pianka 1970; Hendrickson 1980). 8 The morphology of the eggs of the fossil gravid specimen of Desmatochelys padillai 9 described herein resembles that of the rigid shelled eggs of the gravid Adocus sp. 10 specimen from the late Campanian Judith River Group of Canada (Zelenitsky et al. 11 2008). Therefore, it appears that this particular egg trait (rigid shell) was common to 12 different groups of Cretaceous cryptodiran turtles, as Adocus is more closely related to 13 14 the Trionychia than to the Protostegidae (D. padillai described here). In contrast to the 15 total of 19 eggs estimated for the Adocus sp. specimen (Zelenitsky et al. 2008), D. 16 padillai shows a total of 52 eggs preserved, similar to the numbers of eggs laid by extant 17 marine turtles mentioned above. 18 With the establishment of the egg characteristics of the most basal known 19 protostegid Desmatochelys padillai, a significant question emerges: why do the eggs of 20 21 this fully marine adapted turtle differ so much from those of extant marine turtles? We 22 hypothesise that a combination of phylogenetic and palaeoecological adaptive factors 23 could be involved. 24 25 Does phylogeny control egg morphology in marine turtles? 26 Fossil eggs of Desmatochelys padillai show that some egg features have characterized 27 Chelonoidea total group sensu stricto Evers & Benson 2018 for at least the last 125 Ma. 28 29 These are spherical eggs and large clutch >40 eggs, which we suggest here as 30 synapomorphic for the group (Fig. 5). Instead, the occurrence of rigid eggs in the oldest 31 chelonioid (D. padillai) indicates that either this characteristic is homoplastic and related 32 to particular environmental adaptations of Cretaceous marine chelonioids (Protostegidae) 33 as we discuss it in detail in the following paragraph, or that this character represents a 34 reversal in the Protostegidae linage. More gravid marine fossil turtles or fossil eggs 35 indisputably attributed to chelonioids will have to be found in order to test or support any 36 37 of these hypotheses, and to track in detail the egg reproduction strategies and changes of 38 the group during the Mesozoic and Cenozoic. 39 40 Role of palaeoenvironment in the eggs of extinct protostegid turtles 41 The development of rigid eggs in the extinct marine turtle Desmatochelys padillai could 42 have been the result of egg embryo requirements and adaptations according to the 43 44 physical attributes of the environment and nesting sites, such as water saturation and gas 45 exchange of the sand where the nest was created; the depth and temperature at which the 46 nest was located; the hydric, thermal, and respiratory properties of soil; and the type of 47 sediment at the nesting site (mineral composition, albedo, and grain size). All of these 48 factors have been reported to play key physiological roles in the development of the 49 embryo, particularly in terms of its metabolic rate, blood chemistry, and the diffusion of 50 gasses through the eggshell and membranes (Ackerman & Prange 1972; Ackerman 1980; 51 52 Hendrrickson 1980; Ackerman 1997; Bilinski et al. 2001; Wallace et al. 2006; Schneider 53 et al. 2011; Warner et al. 2011). 54 The thick and rigid structure exhibited by the eggs of the gravid specimen of 55 Desmatochelys padillai could also have favoured the development of the embryos, 56 57 58 59 8 60 Palaeontology Page 9 of 29 Palaeontology
1 2 3 providing them with a greater source of calcium, and making them more independent of 4 water conductance and environmental temperature, but maintaining their gas permeability 5 6 through the pores, as has been documented for similar eggs of some extant reptiles (Grigg 7 & Beard 1985; Packard 1999; Bilinski et al. 2001). However, it is relevant to point out 8 that maternal and abiotic controls on egg development and mortality in extant turtles have 9 shown inconsistency of selection operating on egg size, nesting phenology, and nesting 10 behaviour (Warner et al. 2010), which could also have occurred in extinct linages, such 11 as protostegid turtles. Additionally, the rigid eggs of Desmatochelys padillai could have 12 provided better protection against potential predators, such as small vertebrates and crabs, 13 14 which are typical predators of extant sea turtle nests (Spencer 2002; Spencer & 15 Thompson 2003; Bishop et al. 2011, Jackson et al. 2015) and also comprised components 16 of the palaeoecology of the environments inhabited by D. padillai during the early 17 Cretaceous. 18 19 Fossil diagenesis and potential preservation of original egg components 20 21 The luminescence colours and intensity observed in the studied eggshell of 22 Desmatochelys padillai allow us to conclude that its central and ventral surface regions 23 suffered minimal diagenetic alteration. This result favours our hypothesis that the highly 24 enriched carbon content on the ventral surface exhibited by the SEM EDS analysis may 25 be related to preservation of the remains of some of the original organic constituents, 26 which can be tested in the future with immunological, Raman spectroscopy and mass 27 spectrometry techniques. 28 29 30 Acknowledgements. We thank A. Krapf (NMW) for help with access to specimens for 31 comparisons. Special thanks to the late C. B. Padilla for his support, motivation and hard 32 work in the development of the palaeontology of Villa de Leyva and the Centro de 33 Investigaciones Paleontológicas. We also thank F. Parra for helping us in the preparation 34 of the specimen, and F. Uriza from the Hospital San Ignacio and Universidad Javeriana, 35 Bogotá for allowing us to scan this fossil. We thank K. Vizuete from the Universidad de 36 37 las Fuerzas Armadas, Sangolqui for help with the cathodoluminescence analysis. Thanks 38 to C. Jaramillo, D. Lawver and E. Prondvai for suggestions on the first draft of this 39 manuscript. R. Hirayama and another anonymous reviewer, as well as to the editors S. 40 Thomas and A. Smith are thanked for their comments and suggestions to improve this 41 manuscript. We thank S. J. Mason, MSc, from Edanz Group (www.edanzediting.com/ac) 42 for editing a draft of this manuscript. 43 44 45 Author contributions. EAC designed the study, collected data, performed the comparative 46 and analytical work, and wrote the paper. MP and JP collected and prepared the 47 specimen. SP commented and corrected the English grammar of the early stages of this 48 manuscript. 49 50 51 52 REFERENCES 53 54 AREEKIJSEREE, M., PUMIPAIBOON, M., NUAMSUKON, S., 55 KITTIWATTANAWONG, K., THONGCHAI, C., SIKIWAT, S. and CHUEN IM, T. 56 57 58 59 9 60 Palaeontology Palaeontology Page 10 of 29
1 2 3 2013. Leatherback Turtle (Dermochelys coriacea) after incubation eggshell in 4 Andaman Sea, Thailand study: Microanalysis on ultrastructure and elemental 5 6 composition. International Scholarly and Scientific Research and Innovation, 7, 820– 7 824. 8 ACKERMAN, R. A. 1997. The nest environment and the embryonic development of sea 9 turtles. 107–136. In LUTZ, P. L and MUSICK, J. A. (eds). The biology of sea turtles. 10 CRC press, 441 pp. 11 ACKERMAN, R. A. and PRANGE, H. D. 1972. Oxygen diffusion across a sea turtle 12 (Chelonia mydas) egg shell. Comparative Biochemistry and Physiology, 43, 905–909. 13 14 — 1980. Physiological and ecological aspects of gas exchange by sea turtle eggs. 15 American Zoologist, 20, 575–583. 16 BAIRD, T. and SOLOMON, S. E. 1979. Calcite and aragonite in the eggshell of 17 Chelonia mydas. Journal of Experimental Marine Biology and Ecology, 36, 295–303. 18 BILINSKI, J. J., REINA, R. D., SPOTILA, J. R. and PALADINO, F. V. 2001. The 19 effects of nest environment on calcium mobilization by leatherback turtle embryos 20 21 (Dermochelys coriacea) during development. Comparative Biochemistry and 22 Physiology Part A, 130, 151–162. 23 BISHOP, G. A, PIRKLE, F. L., MEYER, B. K. and PIRKLE, W. A. 2011. The 24 foundation for sea turtle geoarchaeology and zooarchaeology: Morphology of recent 25 and ancient sea turtle nests, St. Catherines Island, Georgia, and Cretaceous Fox Hills 26 Sandstone, Elbert County, Colorado. In BISHOP, G. A., ROLLINS, H. B. and 27 THOMAS, D. H. (eds). Geoarchaeology of St. Catherines Island, Georgia: American 28 29 Museum of Natural History Anthropological Papers, 94, 247–269. 30 BRAMAN, D. R. and BRINKMAN, D. B. 2009. Guidebook to Geology and 31 Palaeontology of Dinosaur Provincial Park, Alberta. Special Publication of the Royal 32 Tyrrell Museum, Drumheller, 82 pp. 33 CADENA, E. A. 2015. The first South American sandownid turtle from the Lower 34 Cretaceous of Colombia. PeerJ, 3, doi:10.7717/peerj.1431. 35 — and PARHAM, J. F. 2015. Oldest known marine turtle? A new protostegid from the 36 37 Lower Cretaceous of Colombia. PaleoBios, 32, doi:ucmp_paleobios_28615. 38 CARBALLIDO, J. L., POL, D., PARRA RUGE, M. L., PADILLA BERNAL, S., 39 PÁRAMO FONSECA, M. E. and ETAYO SERNA, F. 2015. A new Early Cretaceous 40 brachiosaurid (Dinosauria, Neosauropoda) from northwestern Gondwana (Villa de 41 Leiva, Colombia). Journal of Vertebrate Paleontology, 35, 1–12. 42 doi:10.1080/02724634.2015.980505. 43 44 ENGLAND, J., CUSACK, M., PATERSON, N. W., EDWARDS, P., LEE, M. R. and 45 MARTIN, R. 2006. Hyperspectral cathodoluminescence imaging of modern and fossil 46 carbonate shells. Journal of Geophysical Research, 111, G03001. 47 doi:10.1029/2005JG000144. 48 ERICKSON, J and BACCARO, F. 2016. Nest predation of the yellow spotted Amazon 49 River turtle (Podocnemis unifilis, Troschel, 1848) by the fire ant (Solenopsis 50 geminata, Fabricius, 1804) in the Brazilian Amazon. Herpetological Journal, 23, 183– 51 52 186. 53 EVERS, S. W. and BENSON, R. B. J. 2018. A new phylogenetic hypothesis of turtles 54 with implications for the timing and number of evolutionary transitions to marine 55 lifestyles in the group. Palaeontology, doi:10.1111/pala.12384. 56 57 58 59 10 60 Palaeontology Page 11 of 29 Palaeontology
1 2 3 FANG, X., LU, L., JIANG, Y. and YANG, L. 2003. Cretaceous fossil eggs from the 4 Tiantai basin of Zhejiang, with a discussion on the extinction of dinosaurs. Geological 5 6 Bulletin of China, 22, 512–520. 7 FORERO, H and SARMIENTO, L. F. 1985. La facies evaporíticas de la Formación Paja 8 en la región de Villa de Leiva. Proyecto Cretácico Contribuciones. Publicaciones 9 Geológicas Especiales del INGEOMINAS, 16, 1–16. [in Spanish] 10 GRIGG, G. and BEARD, L. 1985. Water loss and gain by eggs of Crocodylus porosus, 11 related to incubation age and fertility. In GRIGG, G., SHINE, R. and EHMANN, H 12 (eds). Biology of Australasian frogs and reptiles. Surrey Beatty and Sons, Sydney, 359 13 14 pp. 15 HAMPE, O. 1992. Ein grosswüchsiger Pliosauride (Reptilia: Plesiosauria) aus der 16 Unterkreide (oberes Aptium) von Kolumbien. Courier Forschungsinstitut 17 Senckenberg, 145, 1–32. [in German] 18 HENDRICKSON, J. R. 1980. The ecological strategies of sea turtles. American 19 Zoologist, 20, 597–608. 20 21 HIRAYAMA, R. 1997. Distribution and diversity of Cretaceous chelonioids. In J.M 22 CALLAWAY, J. M. and NICHOLLS, E. M. (eds). Ancient marine reptiles. Academic 23 Press, San Diego, pp. 225–241. 24 HIRAYAMA, R. 1998. Oldest known sea turtle. Nature, 392, 705–708. 25 HIRSCH, K. F. 1983. Contemporary and fossil chelonian eggshells. Copeia, 1983(2), 26 382–397. 27 — 1996. Parataxonomic classification of fossil chelonian and gecko eggs. Journal of 28 29 Vertebrate Paleontology, 16, 752–762. 30 IMAI, T., JINTASAKUL, P., AZUMA, Y., NODA, Y. and CHOKCHALOEMWONG, 31 D. 2016. First confirmed fossil turtle eggshells (oogenus Testudoolithus) from the 32 lower Cretaceous of Thailand. Memoir of the Fukui Prefectural Dinosaur Museum, 33 15, 1–6. 34 JACKSON, F. D., JIN, X., VARRICCHIO D. J., AZUMA, Y. and JIANG, Y. 2008. The 35 first in situ turtle clutch from the Cretaceous Tiantai Basin, Zhenjiang Province, 36 37 China. Journal of Vertebrate Paleontology, 28, 319–325. 38 — VARRICCHIO, D. J., JACKSON, R. A., WALDE, A. D. and BISHOP, G. A. 2015. 39 Taphonomy of extant desert tortoise (Gopherus agassizzii) and loggerhead sea turtle 40 (Caretta caretta) nesting sites: implications for interpreting the fossil record. Palaios, 41 30, 207–223. 42 JOYCE, W. G and ZELENITSKY, D. K. 2002. Turtle egg pseudomorphs from the Late 43 44 Jurassic of Schamhaupten, Germany. Archaeopteryx, 20, 57–62. 45 LAWVER, D. L. 2017. The evolution of reproduction within Testudinata as evidenced by 46 the fossil record. Unpublished PhD thesis, Montana State University, Bozeman, 315 47 pp. 48 — and JACKSON, F. D. 2014. A review of the fossil record of turtle reproduction: eggs, 49 embryos, nests and copulating pairs. Bulletin of the Peabody Museum of Natural 50 History, 55, 215–236. 51 52 — — 2016. A fossil egg clutch from the stem turtle Meiolania platyceps: Implications 53 for the evolution of turtle reproductive biology. Journal of Vertebrate Paleontology, 54 36, doi:10.1080/ 02724634.2016.1223685. 55 56 57 58 59 11 60 Palaeontology Palaeontology Page 12 of 29
1 2 3 — — 2017. An accumulation of turtle eggs with embryos from the Campanian (Upper 4 Cretaceous) Judith River Formation of Montana. Cretaceous Research, 69, 90–99. 5 6 LAWVER, D. L., RASOAMIARAMANANA, A. H, and WENEBURG, I. 2015. An 7 occurrence of fossil eggs from the Mesozoic of Madagascar and a detailed observation 8 of eggshell microstructure. Journal of Vertebrate Paleontology, e973030, doi: 9 10.1080/02724634.2015.973030. 10 NUAMSUKON, S. CHUEN IM, T., RATTANAYUVAKORN, S., PANISHKAN, K., 11 NARKKONG, N. and AREEKIJSEREE, M. 2009. Thai marine turtle eggshell: 12 morphology, ultrastructure and composition. Journal of Microscopy Society of 13 14 Thailand, 23, 52–56. 15 PACKARD, G. C. 1999. Water relations of chelonian eggs and embryos: Is wetter better? 16 American Zoologist, 39, 289–303. 17 PÁRAMO FONSECA, M. E., GÓMEZ PÉREZ, M., NOÉ, L. F. and ETAYO SERNA, 18 F. 2016. Stenorhynchosaurus munozi, gen. et sp. nov. a new pliosaurid from the Upper 19 Barremian (Lower Cretaceous) of Villa de Leiva, Colombia, South America. Revista 20 21 de la Academía Colombiana de Ciencias Exactas, Físicas, y Naturales, 40, 84–103. 22 PIANKA, E. R. 1970. R selection and K selection. The American Naturalist, 102, 592– 23 597. 24 REBOULET, S. and HOEDEMAEKER, P. J. 2006. Report on the 2nd international 25 meeting of the IUGS lower Cretaceous ammonite working group, the “Kilian Group” 26 (Neuchâtel, Switzerland, 2 September 2005). Cretaceous Research, 27, 712–175. 27 — SZIVES, O., AGUIRRE URRETA, B., BARRAGÁN, R., COMPANYE, M., 28 29 IDAKIEVAF, V., IVANOVF, M., KAKABADZEG, M. V., MORENO BEDMARD, 30 J. A., SANDOVALE, J., BARABOSHKINH, E. J., ÇAGLARI, M. K., FOZY, I., 31 GONZÁLEZ ARREOLAD, C., KENJOA, S., ALEXANDER LUKENEDERJ, A., 32 RAISOSSADATK, S. N., RAWSONL, P. F. and TAVERA, J. M. 2014. Report on the 33 5th International Meeting of the IUGS Lower Cretaceous Ammonite Working Group, 34 the Kilian Group (Ankara, Turkey, 31 August 2013). Cretaceous Research, 50, 126– 35 137. 36 37 ROSTAL, D. C., PALADINO, F. V., PATTERSON, R. M. and SPOTILA, J. R. 1996. 38 Reproductive physiology of nesting leatherback turtles (Dermochelys coriacea) at Las 39 Baulas National Park, Costa Rica. Chelonian Conservation and Biology, 2, 230–236. 40 SCHNEIDER, L., SILVA, A. and FERREIRA JUNIOR, P. 2011. Sedimentary 41 characteristics and their effects on hatching success and incubation duration of Caretta 42 caretta (Testudines: Cheloniidae) in Espirito Santo, Brazil. Zoologia, 28, 312–320. 43 44 SCHLEICH, H. H. and KÄSTLE, W. 1988. Reptile egg-shells: SEM Atlas. Gustav 45 Fischer, Stuttgart, 123 pp. 46 SIKIWAT, S., AREEKIJSEREE, M., SRIBUDDHACHART, K., KERDKRIANKRAI, 47 S., NAMCHOTE, S., KITTIWATTANAWONG, K. and THONGCHAI, C. 2015. 48 Comparative eggshell ultrastructure of three species of sea turtle. The 53th Kasetsart 49 University Annual Conference, 3–6 February 2015, Bangkok, Thailand. 50 SOLOMON S. E. and WATT, J. M. 1985. The structure of the eggshell of the 51 52 leatherback turtle (Dermochelys coriacea). Animal Technology, 36, 19–27. 53 SPENCER, R. J. 2002. Experimentally testing nest site selection: fitness tradeoffs and 54 predation risk in turtles. Ecology, 83, 2136–2144. 55 56 57 58 59 12 60 Palaeontology Page 13 of 29 Palaeontology
1 2 3 — and Thompson, M. B. 2003. The significance of predation in nest site selection of 4 turtles: An experimental consideration of macro and microhabitat preferences. Oikos 5 6 102, 592–600. 7 VANZOLINI, P. E. 2003. On clutch size and hatching success of the South American 8 turtles Podocnemis expansa (Schweigger, 1812) and P. unifilis Troschel, 1848 9 (Testudines, Podocnemididae). Anais da Academia Brasileira de Ciências, 75(4), 10 415–430. 11 WALLACE, B. P., SOTHERLAND, P. R., TOMILLO, P. S., BOUCHARD, S. S., 12 REINA, R. D., SPOTILA, J. R. and PALADINO, F. V. 2006. Egg components, egg 13 14 size, and hatchling size in leatherback turtles. Comparative Biochemistry and 15 Physiology Part A, 145, 524–532. 16 WARNER, D. A., JORGENSEN, C. F. and JANZEN, F. J. 2010. Maternal and abiotic 17 effects on egg mortality and hatchling size of turtles: Temporal variation in selection 18 over seven years. Functional Ecology, 24, 857–866. 19 — MOODY, M. A. and TELEMECO, R. S. 2011. Is water uptake by reptilian eggs 20 21 regulated by physiological processes of embryos or a passive hydraulic response to 22 developmental environments? Comparative Biochemistry and Physiology Part A, 160, 23 421–425. 24 WINKLER, J. D and SÁNCHEZ VILLAGRA, M. R. 2006. Testing phylogenetic 25 implications of eggshell characters in side necked turtles (Testudines: Pleurodira). 26 Zoology, 109, 127–136. 27 ZELENITSKY, D. K., THERRIEN, F., JOYCE, W. G. and BRINKMAN, D. B. 2008. 28 29 First fossil gravid turtle provides insight into the evolution of reproductive traits in 30 turtles. Biology Letters, 4, 715–718. 31 32 DATA ARCHIVING STATEMENT 33 34 35 Data for this study are available in the Dyrad Digital Repository. 36 37 Data Archiving Statement Video S1. Horizontal view of specimen FCG CBP 75 38 Data Archiving Statement Video S2. Coronal view of specimen FCG CBP 75 39 Data Archiving Statement Video S3. Sagittal view of specimen FCG CBP 75 40 http://dx.doi.org/ 41 42 43 SUPPORTING IMFORMATION 44 45 Additional Supporting Information can be found in the online version of this article: 46 47 Supporting information Fig. S1. Additional views and location of the discovery of the 48 Desmatochelys padillai FCG CBP 75 specimen and its eggs, Testudoolithus oospp. A, 49 map showing the location of Villa de Leyva in Colombia (top) and the geology of the 50 region (bottom), as well as the Monsalve hill where the fossil was found, figure modified 51 52 from Lawer & Jackson (2014). B–C, right lateroventral view of the posterior region of 53 the shell, showing the hypoplastron (red) and portions of the ilium (green); D–E, cross 54 view of one of the naturally fractured blocks that composed the fossil specimen, showing 55 details of the hypoplastron, costal bone, a potential peripheral fragment (blue), and a 56 57 58 59 13 60 Palaeontology Palaeontology Page 14 of 29
1 2 3 totally recrystallised egg (green). Abbreviations: co, costal bone; hyp, hypoplastron; ili, 4 ilium; pe, peripheral bone. 5 6 7 Supporting information Fig. S2. Detailed views of Testudoolithus oospp. eggshell 8 morphology, and Desmatochelys padillai specimen FCG CBP 75. A–B, one of the 9 portions of the egg for which a thin section was made, showing well defined shell units, 10 pores, and caverns, at 4X magnification; C–D, close up of one of the pores of the 11 eggshell, also showing well defined calcite crystals; E–F, close up of the most posterior 12 portion of the carapace of D. padillai FCG CBP 75, showing the posterior almost parallel 13 14 to the midline of the carapace projection of the naked costal rib of the most posterior 15 costal bone; G–H, close up of the most posterior portion of the shell of the extant sea 16 turtle Caretta caretta NMW 31531(Natural History Museum, Vienna, Austria), showing 17 the same pattern of the posteriorly projected naked costal rib, described in G–H for D. 18 padillai. Abbreviations: ca, cavern; cr, costal rib; po, pore; rc, recrystallised calcite. 19 20 Supporting information Fig. S3. Comparison of the holotype of Desmatochelys padillai 21 22 Cadena & Parham, 2015 and the D. padillai FCG CBP 75 specimen described here in 23 anterolateral views. A–B, D. padillai holotype exhibiting high keels along the midline of 24 the carapace. C–D, D. padillai FCG CBP 75 specimen exhibiting similar pattern of keels 25 along the midline of the carapace. 26 27 Supporting information Fig. S4. Scanning electron microscope (SEM) images and 28 elemental composition with energy dispersive X ray spectroscopy analyser (EDS) results 29 30 for the middle portion of one of the eggshells of Testudoolithus oospp. A, SEM image of 31 the calcite crystals; B, EDS elemental mapping at 15 kV for the region shown in A, 32 predominantly calcium, carbon, and oxygen, consistent with crystals of calcium 33 carbonate (calcite) and some traces of iron. 34 35 Supporting information Fig. S5. SEM images and elemental composition of the ventral 36 (visceral) surface of one of the eggshells of Testudoolithus oospp., as well as the rock 37 38 matrix that covers most of the shell bones and some of the eggs. A, ventral surface of the 39 eggshell, showing small circular bumps interpreted as the organic cores of shell units; B, 40 SEM and elemental composition analysis of a region of the ventral surface of the egg (see 41 rectangle “B” in A), details of the composition are given in the table below A, indicating 42 high carbon content. C, photograph showing the rock matrix covering most of the shell 43 bones and some of the eggs; D, SEM and EDS analysis of the rock matrix (see rectangle 44 45 “D” in D), details of the composition are given in the table below C, with absence of 46 carbon and abundant contents of silica, aluminium, and iron. 47 48 Supporting information Fig. S6. Cathodoluminescence of the ventral surface of one of 49 the eggs of Desmatochlys padillai exhibiting purple to cyan fluorescent colours 50 corresponding to wavelengths between 400 and 500 nm. 51 52 53 54 55 56 57 58 59 14 60 Palaeontology Page 15 of 29 Palaeontology
1 2 3 FIGURE CAPTIONS 4 5 6 Fig. 1. Desmatochelys padillai FCG CBP 75 and the ootaxon Testudoolithus oospp. A– 7 B, carapace in dorsal view, showing some eggs between costal bones or where they are 8 poorly preserved; C–D, plastron in ventral view, showing several eggs and cavities left 9 by the eggs; E, four isolated eggs of T. oospp.; F, close up of one of the eggs of T. oospp. 10 showing the micropitted texture of the eggshell surface. Abbreviations: cf, central 11 fontanelle; co, costal bone; cr, costal rib; ec, egg cavity; hyo, hyoplastron; hyp, 12 hypoplastron; ili, ilium; ne, neural bone. Scale bars: 10 cm horizontal bar applies for A– 13 14 D, 5 cm vertical bar applies for E, 2 cm horizontal bar applies for F. 15 16 Fig. 2. Thin section of one of the eggs of Testudoolithus oospp. A–B, whole egg view, 17 with a portion of a costal bone showing its interior and part of its cancellous bone; C–D, 18 close up of one of the regions of the eggshell (red rectangle in B), showing a remaining 19 portion of the egg membrane; E–F, close up of one of the regions of the eggshell (blue 20 21 rectangle in B), showing pores and accretion lines around them; G–H, close up of one of 22 the regions of the eggshell, showing clearly defined shell units, pores, accretion lines, and 23 a potential organic core. Abbreviations: al, accretion line; ca, calcite; cb, cancellous bone; 24 c, cavity; cv, cavern; es, eggshell; ic, internal cortex; lca, large calcite crystals; mic, 25 micrite; oc, organic core; pc, pore core; si, siderite; smr, shell membrane remain; su, shell 26 unit. Scale bars: 1 cm horizontal bar applies for A–B; each microphotograph C–H 27 includes its own scale bar. 28 29 30 Fig. 3. Computer tomography images of Desmatochelys padillai specimen FCG CBP 75 31 and its eggs, Testudoolithus oospp. (see also Data Archiving Statement Videos S1–3). A– 32 C, two horizontal slides and a close up of the yellow rectangle in B; D, CT image of one 33 of the isolated eggs using the colour mapping tool in the iQ VIEW software to visualise 34 cavities inside the egg, as well as measurements of it; E, one of the coronal slides, 35 showing the eggs, as well as portions of one of the costal and hyoplastron bones; F–G, 36 37 close up of the posterior region of the shell, where one of the horizontal slides allows 38 recognition of the posterior series of neural bones; H–I, close up of the posterior region 39 of the shell, with one slide that allows recognition of portions of the pelvic girdle bones. 40 Abbreviations: ne, neural bones; pg, pelvic girdle bones. 41 42 Fig. 4. SEM EDS imaging elemental analysis and cathodoluminescence of one of the 43 44 eggshells of Testudoolithus oospp. described herein. A, turtle eggshell microstructure 45 (modified from Lawver & Jackson (2014)). B, general view of the eggshell fragment 46 analysed by SEM without any coating, rectangles indicate the regions of images C–E 47 where elemental analysis were performed; C, top of the eggshell showing rich content of 48 aluminium; D, middle portion of the eggshell, predominantly calcium carbonate; E, base 49 of the eggshell, showing high carbon content at the egg membrane eggshell contact. F, 50 cathodoluminescence showing bright purple to cyan, corresponding to wavelengths 51 52 between 400 and 500 nm. Details of the EDS elemental analysis of the three regions of 53 the eggshell are shown in Table 2. Abbreviations: Al, aluminium; C, carbon; Ca, calcium; 54 Fe, iron; N, nitrogen; O, oxygen, oc, organic core; p, pore; sm, shell membrane; su, shell 55 unit. Darker intensity of colour indicates higher elemental concentration. 56 57 58 59 15 60 Palaeontology Palaeontology Page 16 of 29
1 2 3 Fig. 5. Distribution of reproductive traits in Testudines. Phylogenetic tree is based and 4 modified from Evers & Benson (2018). Adocus sp. was added as sister taxon of 5 6 Trionychia following Zelenitsky et al. 2008. Abbreviations: Egg shape in green (ee, 7 elongate eggs; ?e unknown egg shape; se, spherical eggs); eggshell type in purple (ps, 8 pliable shelled eggs, rs, rigid shelled eggs, ss, semipliable shelled egg); clutch size in red 9 (sc, small clutch size (less than 15 eggs), mc, medium clutch size (between 15 and 40 10 eggs), lc, large clutch size (more than 40 eggs); NG, Neogene; Q, Quaternary. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 16 60 Palaeontology Page 17 of 29 Palaeontology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig. 1. Desmatochelys padillai FCG-CBP-75 and the ootaxon Testudoolithus oospp. A–B, carapace in dorsal 46 view, showing some eggs between costal bones or where they are poorly preserved; C–D, plastron in 47 ventral view, showing several eggs and cavities left by the eggs; E, four isolated eggs of T. oospp.; F, close- 48 up of one of the eggs of T. oospp. showing the micropitted texture of the eggshell surface. Abbreviations: cf, 49 central fontanelle; co, costal bone; cr, costal rib; ec, egg cavity; hyo, hyoplastron; hyp, hypoplastron; ili, 50 ilium; ne, neural bone. Scale bars: 10 cm horizontal bar applies for A–D, 5 cm vertical bar applies for E, 2 cm horizontal bar applies for F 51 52 145x192mm (300 x 300 DPI) 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 18 of 29
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig. 2. Thin section of one of the eggs of Testudoolithus oospp. A–B, whole-egg view, with a portion of a 46 costal bone showing its interior and part of its cancellous bone; C–D, close-up of one of the regions of the 47 eggshell (red rectangle in B), showing a remaining portion of the egg membrane; E–F, close-up of one of the 48 regions of the eggshell (blue rectangle in B), showing pores and accretion lines around them; G–H, close-up 49 of one of the regions of the eggshell, showing clearly defined shell units, pores, accretion lines, and a 50 potential organic core. Abbreviations: al, accretion line; ca, calcite; cb, cancellous bone; c, cavity; cv, cavern; es, eggshell; ic, internal cortex; lca, large calcite crystals; mic, micrite; oc, organic core; pc, pore 51 core; si, siderite; smr, shell membrane remain; su, shell unit. Scale bars: 1 cm horizontal bar applies for A– 52 B; each microphotograph C–H includes its own scale bar. 53 54 55 159x231mm (300 x 300 DPI) 56 57 58 59 60 Palaeontology Page 19 of 29 Palaeontology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 20 of 29
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig. 3. Computer tomography images of Desmatochelys padillai specimen FCG-CBP-75 and its eggs, 46 Testudoolithus oospp. (see also Data Archiving Statement Videos S1–3). A–C, two horizontal slides and a 47 close-up of the yellow rectangle in B; D, CT image of one of the isolated eggs using the colour mapping tool 48 in the iQ-VIEW software to visualise cavities inside the egg, as well as measurements of it; E, one of the 49 coronal slides, showing the eggs, as well as portions of one of the costal and hyoplastron bones; F–G, close- 50 up of the posterior region of the shell, where one of the horizontal slides allows recognition of the posterior series of neural bones; H–I, close-up of the posterior region of the shell, with one slide that allows 51 recognition of portions of the pelvic girdle bones. Abbreviations: ne, neural bones; pg, pelvic girdle bones. 52 53 182x304mm (300 x 300 DPI) 54 55 56 57 58 59 60 Palaeontology Page 21 of 29 Palaeontology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig. 4. SEM-EDS imaging-elemental analysis and cathodoluminescence of one of the eggshells of 46 Testudoolithus oospp. described herein. A, turtle eggshell microstructure (modified from Lawver & Jackson 47 (2014)). B, general view of the eggshell fragment analysed by SEM without any coating, rectangles indicate 48 the regions of images C–E where elemental analysis were performed; C, top of the eggshell showing rich 49 content of aluminium; D, middle portion of the eggshell, predominantly calcium carbonate; E, base of the 50 eggshell, showing high carbon content at the egg membrane-eggshell contact. F, cathodoluminescence showing bright purple to cyan, corresponding to wavelengths between 400 and 500 nm. Details of the EDS 51 elemental analysis of the three regions of the eggshell are shown in Table 2. Abbreviations: Al, aluminium; 52 C, carbon; Ca, calcium; Fe, iron; N, nitrogen; O, oxygen, oc, organic core; p, pore; sm, shell membrane; su, 53 shell unit. Darker intensity of colour indicates higher elemental concentration. 54 55 209x264mm (300 x 300 DPI) 56 57 58 59 60 Palaeontology Palaeontology Page 22 of 29
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Page 23 of 29 Palaeontology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Fig. 5. Distribution of reproductive traits in Testudines. Phylogenetic tree is based and modified from Evers & 21 Benson (2018). Adocus sp. was added as sister taxon of Trionychia following Zelenitsky et al. 2008. 22 Abbreviations: Egg shape in green (ee, elongate eggs; ?e unknown egg shape; se, spherical eggs); eggshell 23 type in purple (ps, pliable-shelled eggs, rs, rigid-shelled eggs, ss, semipliable-shelled egg); clutch size in red 24 (sc, small clutch size (less than 15 eggs), mc, medium clutch size (between 15 and 40 eggs), lc, large clutch 25 size (more than 40 eggs); NG, Neogene; Q, Quaternary. 26 67x27mm (300 x 300 DPI) 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 24 of 29
1 2 3 Table 1. Comparisons between characteristics of eggs of some extant marine turtles, with 4 the inclusion of Desmatochelys padillai and its ootaxon Testudoolithus oospp. Data for 5 6 Lepidochelys kempii taken from Hirsch (1983), for Caretta caretta from Schleich and 7 Kästle (1988), for Chelonia mydas and Eretmochelys imbricata from Sikiwat et al. 8 (2015). 9 10 Species Egg shape Egg size Shell Shell unit Shell 11 (mm) morphotype, height to thickness 12 13 composition width (µm) 14 ratio 15 Lepidochelys kempii Spherical 38 Pliable, egg 1:2 120-400 16 (Kemp’s ridley) membrane 17 >>eggshell 18 (aragonite) 19 Caretta caretta Spherical 39-42 Pliable, egg Loose 80-140 20 21 (Loggerhead) membrane crystallic 22 eggshell mass 23 (aragonite) 24 Chelonia mydas Spherical - Pliable, egg 1:2 200 25 membrane 26 27 eggshell 28 (aragonite or 29 calcite) 30 Eretmochelys imbricata Spherical - Pliable, egg 1:3 235 31 membrane 32 eggshell 33 (aragonite) 34 35 Desmatochelys padillai Spherical 32.2-42.8 Semipliable- 1.1:1-1.3:1 785-820 36 (Fossil protostegid) rigid, egg 37 membrane 38 < 1 2 3 Table 2. Elemental analysis of the eggshell of one of the eggs of Testudoolithus oospp. 4 obtained from SEM coupled with an energy-dispersive X-ray spectroscopy (EDS). 5 6 7 Element Atomic Weight Stoichiometry 8 Symbol Concentration Concentration Concentration 9 Top (exterior region) of the eggshell 10 O 53.53 49.68 11 Ca 9.25 21.50 19.91 12 13 N 21.36 17.35 45.96 14 C 15.39 10.72 33.12 15 Al 0.47 0.74 1.02 16 Middle of the eggshell 17 O 54.06 49.50 18 Ca 10.30 23.63 22.43 19 20 N 20.80 16.67 45.29 21 C 14.83 10.19 32.28 22 Base (near shell membrane) of the eggshell 23 O 53.24 52.63 24 C 37.75 28.01 80.72 25 Fe 1.76 6.08 3.77 26 27 Ca 1.58 3.91 3.38 28 Si 2.17 3.77 4.64 29 Al 2.05 3.42 4.39 30 Mg 1.45 2.17 3.10 31 Eggshell ventral surface 32 O 46.04 47.44 33 34 C 46.74 36.15 86.61 35 Ca 4.43 11.44 8.21 36 Si 1.57 2.83 2.90 37 Al 1.23 2.14 2.28 38 Rock matrix covering the shell and some eggs 39 O 61.24 38.82 40 41 Fe 16.53 36.57 42.64 42 Si 19.35 21.53 49.91 43 Al 2.89 3.09 7.45 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 26 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Page 27 of 29 Palaeontology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 28 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Page 29 of 29 Palaeontology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Palaeontology Page 30 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology Page 31 of 29 Palaeontology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Palaeontology