The first record of exceptionally-preserved spiral coprolites from the Tsagan-Tsab formation (lower cretaceous), Tatal, western Mongolia
SizesAs from the measurements, all collected coprolites vary in sizes (Table 1). The smallest and complete specimen is IVPP V 27,545 (Fig. 2D–G), and while IVPP V 27,550 (2 V-Z) is multiple time larger. The maximum length for specimen IVPP V 27,544, IVPP V 27,546, IVPP V 27,547 and IVPP V 27,549 have not been determined due to their incompleteness.Table 1 Biometrical and morphological features of spiral coprolites from Tsagan-Tsab Formation (Lower Cretaceous), Tatal, western Mongolia. Paul Rummy, Kazim Halaclar & He Chen.Full size tableSurface adhesion and marksAll specimens contained some degree of bone fragments and rhomboidal-shaped ganoid scales adhered to the coprolite surfaces (Fig. 3). Additionally, all specimens have smooth surfaces with little abrasion. The inner coil lines of specimen IVPP V 27,549 adhered with a matrix of red clay with silt (Fig. 2S–U). Only specimen IVPP V 27,550 has been seen with concentric cracks (Fig. 2V–Z). Bite marks have also been found on specimen IVPP V 27,545, in which these traces were short, parallel, shallow and isolated. They have been formed from 3 furrows of roughly 3.8 mm long and 0.3 mm deep (Fig. 4).InclusionsThrough CT scans and surface observation, we noticed that all specimens contained bone fragments and scales of varying degrees (Fig. 5). We were unable to identify the bones in detail for specimen IVPP V 27,544, IVPP V 27,546, IVPP V 27,547, IVPP V 27,548, IVPP V 27,549 and IVPP V 27,550, as they were excessive in amount and extremely fragmentary. On the contrary, for specimen IVPP V 27,545, we noticed a rather complete bone structure, such as the ribs and a segment of an infraorbital (Fig. 5H–N). SEM photograph from one random point of specimen IVPP V 27,545 yielded results of the existents of pollen grain (Fig. 6C).BoringsSurface borings of invertebrate burrowing can be seen in 2 spiral coprolites, namely IVPP V 27,547 (Fig. 2D–G) and IVPP V 27,550 (Fig. 2V–Z). CT scans revealed that the borings of specimen IVPP V 27,550 did not intrude internally, and it was the same for some of IVPP V 27,547 as well (Fig. 7). Specimens IVPP V 27,546, IVPP V 27,547, IVPP V 27,548 and IVPP V 27,549 are shown to have traces of internal borings (Fig. 5C–F).EDS analysesIn this work, in regards to Tatal’s coprolites, the mineral elements were examined by using EDS and the photos were taken with SEM. Analyses was conducted on 2 specimens (IVPP V 27,546 and IVPP V 27,545) with two sample points for each. All 4 samples showed high peaks of calcium and phosphorus. EDS results of specimen IVPP V 27,546 (Fig. 6A–B) and specimen IVPP V 27,545 (Fig. 6C–D) gave similar atomic compositions. They were mainly composed of Ca, P and O and small peaks that belong to Nb, Si, C, K, Fe and Al. We have also described a potential pollen structure under SEM image (Fig. 6C). This possible pollen structure in specimen IVPP V 27,545 (Fig. 6C) showed different atomic elements from the other EDS results, where it contained high peaks of Na and Cl.Taphonomy inferencesNo signs of abrasion were found on all of the coprolites. Coloration of the coprolites varied, thus, indicating they were buried in different sedimentary conditions. Through the shape of the coprolites, we can deduce that they have indeed spent different amounts of time or phases in water bodies before burial (see above description/discussion). Meanwhile, specimen IVPP V 27,550 showed shallow coil deepness, therefore, this indicates that it was buried rapidly after excretion.Discussion and interpretationThere are several pivotal evidences that corroborate to fecal origins of the Tsagan-Tsab Formation material: (1) basic morphology; (2) general shape and size (3) inclusions of the fecal matter; (4) high calcium and phosphorus content; (5) bioerosional scars; (6) borings and cavities; (7) concentric cracks.The fundamental puzzle in the studies of coprolite is the difficulty in identifying the potential producer, which can be due to their nature and preservation. Also, that includes the methods used to deduce them with their producer, which were done by inferring with various forms of relationship based on stratigraphy and geographical relationships, as well as on neoichnology studies7,23,54,55. Such problems similarly arose in our context as well, and the materials were collected from a stratum that were interpreted as lake deposit margins, thus, suggesting an amphibious or aquatic producer. The paleoenvironment correlates with the findings of pterosaur fossils such as the Noripterus44 or argued as ‘Phobetor’56, and the diets of these pterosaurs were dependable on the lake environment57,58,59,60. Above all, and more importantly, that the shape of the coprolite has to be intact in order to represent the shape of the internal intestine of the producer, whereby, anatomically it can lead to a certain biological aspect and digestive system of the organism. Despite these, there are on-going controversies on the origin of the spiral shaped bromalites in regards to whether or not they signify fossilized feces, or they are the cololite that was formed within the colon6,21,23,61,62.Spiral coprolites are producer of an animal with spiral intestine valves to increase the surface area of absorption, to slow down food movement in the bowel to maximise nutrient absorption, which has a significant strategy in surviving uncertain and harsh environment conditions28,63,64. Referring to past literature, it is generally agreed upon that the spiral shape is the only distinctively coprolite morphology, whereby it has been regarded as a true coprolite and can be correctly associated to the source animal, such as a range of fishes in particular6,22,52. Many primitive bony fishes (except those of teleosts), fresh water sharks (elasmobranches), coelacanths, Saurichthys, sturgeons and lungfishes are known to have the spiral valve intestine51,64,65,66. Also, Price67 suggested that the amphipolar form could have been derived from palaeoniscoids. Additionally, Romer & Parsons68 noted that the spiral valves are secondarily lost in teleost and tetrapods, while Chin69 noted a few teleosteans still possessing them.The spiral coprolites collected for this study are mainly amphipolar in shape and one in scroll. As we know, generally heteropolar spiral coprolite are produced by sharks, which have complex spiral valves62. Therefore, we can exclude those in the family of elasmobranches as the potential producers and this can also be supported by the non-marine geological settings of Tsagan-Tsab Formation. But it is also noteworthy to mention that in previous studies, some workers have conducted observations on sharks that were kept in tanks, and were not been able to find any spiral fecal pellets. The reasons given were that the sharks’ eating habits could have changed due to the tank environment, which would have differed from the natural marine environment. Also, modern day sharks are totally unrelated to the ancient Permian pleuracanth sharks6. Despite these, evidence of spiral fecal pellet can still be observed in some of the present-day fishes, such as the African lungfish Protopterus annectans, the Australian lungfish Neoceratodus forsteri, the long-nosed gar Lepisosteus osseus and the spotted gar Lepisosteus oculatus6,70,71,72. As for scroll coprolites, it is generally known to be produced by animal with longitudinal valves (valvular voluta), whereby the valves naturally rolls in upon itself , in a way that it maximises nutrient absorption8,9,17,18. Gilmore17 in his work mentioned that this type of valve must be primitive than the transverse valve (valvular spiralis), which could be a modification of the previous ones. This form is especially known to sharks of carcharhiniforms73, and it is evident that it could have been associated with sarcopterygian53, as well as anaspid and thelodont agnathans17.In this study, we recognised four new ichnotaxa for all the seven coprolite specimens. Assigning four new ichnotaxa does not conclude that the coprofauna are of four different types of animals. Considering there are two distinct morphologies, which are the amphipolar spiral and scroll, we can deduce that at least two animals can produce these coprolites. But we have to carefully consider that diverse diets at different times for the same animal can often be variable, and soft fecal materials can range disparately after defecation, as well as taphonomy influence74,75. Specimen IVPP V 27,550 is remarkably huge and its producer should be a massive animal since large animals could produce small excrement, but small animals would not be able to produce big excrement52,54. Moreover, since there are no relevant fossils fauna found in the locality, we were unable to exactly identify the specific producer, rather, we deduced with relevant sources. However, we do know that both amphipolar spiral and scroll coprolites can be attributed to certain types of fishes. As of these, we can conclude that the coprolites were produced by fishes in different sizes. Specimen IVPP V 27,545 differs from the rest by its shape and size, which makes prediction even harder, because it could be produced by either large or smaller animals.CT scans revealed that bony inclusions are evident in all of the coprolites (Fig. 5). However, except in specimen IVPP V 27,545, the bones in the rest of the coprolites are fragmentary. Specifically, bones in specimen IVPP V 27,545 are rather unaffected by the acidity of the digestive enzyme and these were evident by the presence of clusters of entire bones in the coprolite (Fig. 3A–C), as contrast to the fragmentary bones in the rests of the coprolites. Furthermore, we identified an infraorbital bone of a fish. CT scans revealed that the infraorbital bone has a sensory canal where it branches off at both ends (Fig. 5M–N). With these, we can indicate that the producer of specimen IVPP V 27,545 poorly masticated the prey and also had a rather low gut digestion for food28,55,76,77,78. Through these results, we can infer the digestive strategies of the producers were in correlation with food intake and digestion process, as discussed in Barrios-de Pedro & Buscalioni77. Specimen IVPP V 27,545 might belong to the first type of digestive strategy, whereby the producer has limited food processing in the mouth and the food stays in the digestive system for a short period of time. This strategy is regarded to be efficient in conditions where food sources are abundant and the nourishment levels are sufficient79. The rest of the coprolites possibly belong to the second digestive strategy, as the bone content is fragmentary. This suggest the producer might have limited mastication with improved digestive assimilation and longer gut time to favour better absorptions of nutrients55,80,81,82,83. The third type of digestive strategy does not imply in our study. It is also noteworthy to mention that the quantity of the inclusions is not correlated to the size of the coprolite, rather, it is dependable on the above-mentioned biological variables28,84.Carnivorous coprolites are normally composed of calcium phosphate and other organic matter, but it is important to be aware that the initial compositions are usually altered during fossilization processes33. Meanwhile, the excretion of herbivores is generally lacking in phosphates and their fossilization are mostly dependable of the mineral enrichment85. Through the morphological shape, the density of bone and scale inclusions on the surface from the CT scans, we can directly assume that these coprolites are inevitably produced by carnivorous organisms. Despite that, we still conducted SEM–EDS tests on two specimens, IVPP V 27,546 and specimen IVPP V 27,545 (Fig. 6), in order to determine its mineral content, and to prove them as a valid coprolite material because we were not able to compare these materials to any attached locality matrix at the time the study. The reason for that was because the specimens were collected almost two decades ago and they were very well-kept in the archives throughout these years. As predicted, all 4 samples gave higher content of Ca and P, thus, there is no doubt that they are indeed fossilized fecal materials. Also, in regards to the SEM–EDS on specimen IVPP V 27,545 (Fig. 6C–D), when randomly pointed to a particular structure, it yielded unusual results from the rest, in which the EDS peaks are composed of Na and Cl. At the same time, the SEM image potentially showed a pollen grain like structure. Hollocher and Hollocher86 documented a pollen image by using SEM, which brings our potential pollen image (Fig. 6C) dimensionally compatible with their sample. Although specimen IVPP V 27,545 is produced by an unidentified carnivorous vertebrate, it is common for carnivore coprolites to have plant remains within them. Also, it is known that spores and pollens are exceptionally well preserved within the encasement of calcium phosphate, which inhibits sporopollenin degradation87. Various reasons can be inferred for the presence of the pollen in specimen IVPP V 27,545, to which it could either be by accident or by preying on an herbivorous animal. Furthermore, it could also be through the adhesion on the excrement when the fecal is still fresh88. Pollens are in fact valuable information provider for paleoenvironment reconstruction, as well as for understanding the vegetation state of a particular era87,89,90,91,92. Hence, further palynology analyses are needed for future work.EDS mineral composition and coprolite coloration can be correlated to a certain degree, in which it could also explain depositional origin27. Most of the Tatal’s coprolites are pink-whitish in color, which is highly associated with the presence of calcium through its carnivorous diets93,94,95,96. The dark colors can also be due to the presence of iron or it could also be due to complete phosphatisation23,27. However, a large part of the colorations was influenced by diagenesis27,28.Traces of burrows are evident on the surface of specimen IVPP V 27,547 and IVPP V 27,550, but CT scans revealed internal traces burrowing did occur in specimen IVPP V 27,546, IVPP V 27,547, IVPP V 27,548 and IVPP V 27,549 (Fig. 5). Since not all possible burrows were dug-in, we gave the term ‘pseudo-burrow’ on those burrows that were abandoned in the early stages. For example, on all of the burrow traces in specimen IVPP V 27,547, only one traces showed burrowing holes, while the rest did not form a hole. While those specimens with internals, but without any traces on the outer surface, this can be explained by taphonomy processes, whereby the outer surface is covered with sedimentary and non-differentiable. It was reported in Tapanila et al.97, that marine bivalves are potential makers of the burrows in coprolites by expanding the diameter of the hole as they dig in, although Milàn, Rasmussen & Bonde98, reported a contradictory example, where the holes were indeed constant in diameter. In our study, we couldn’t determine if the holes were constantly in diameter or not. Numerous tiny holes were visible on all of the coprolites surface, as well as within it, and these were most probably caused by gases within the fecal matters. These holes can be called as microvoids or ‘degassing holes’, which contain gases trapped during digestion74,99,100. Microvoids are quickly filled with water when fecal matter is excreted from the animal body, thus making the fecal becoming heavy and sinking to the lake floor74.A series of three parallel furrows or bioerosional scars were evident on the surface of specimen IVPP V 27,545 (Fig. 3). Those lines only occurred once without any repetition on the rest of the surface. The information from these furrows were insufficient to deduce any potential biters, as widely discussed in the work of Godfrey & Palmer101, Godfrey & Smith102, Dentzien-Dias et al.103, and Collareta et al.104. On the other hand, deducing from the dented surface on the bitten marks, we predicted that the marks were most probably made by the biting pressures from the fish mandibles, which may indicate coprophagous behavior. The biting could have happened on the lake floor just before sedimentary deposition. Since the bitten marks are on the surface, this probably suggests unintentional scavenging and was eventually aborted during food search.In general, coprolites can be transported from the original place through various modes25 and this can be evident by the traces of abrasion51,65. However, in Tatal’s coprolites, there were little or almost no marks of abrasion. Yet again, this supports our hypothesis that these coprolites were excrements in shallow waters, such as in the lake banks with little turbulence and current, where the fecal matter was dropped in-situ after excrement. As stated in previous literature105,106, radial and concentric cracks are also evident on the surface of specimen IVPP V 27,550, therefore, these indicate that the coprolite was excreted on a very shallow environment where the water body was vastly evaporated and left for subaerial exposure before embedment. This phenomenon caused the coprolite to dehydrate through the cracking, and shrinking occurred in a low magnitude process while retaining its overall shape27,54,107. Previous authors have also discussed that the cracks could possibly be due to synaeresis under certain conditions27,54,108.It has been frequently reported in records that almost all spiral coprolite fossilization from various Phenerozoic ages have occurred in low-energy shallow marine environments54. Feces that are being excreted in this humid environment have a higher chance of preservation due to the rapid burial, as well as on the acidity level of the water bodies5,7,109,110,111. There are also several crucial factors that are involved in fecal fossilization. Among them, one of the most important criteria includes the content and composition of the fecal matter, and those of carnivorous diets tend to form coprolites than those who consumed an herbivorous diet75. As mentioned in Dentzien-Dias et al.111, there are three main stages involved in a coprolite taphonomy history, which include stages before final burial, after the final burial and after exposure. In accordance to this, we introduced the usage of phases to discuss the spiral coprolites morphologies in this study (see material and methods). The phase concept of spiral coprolites disentanglement has been widely discussed in early days by various workers6,22,70. Coprolite specimen IVPP V 27,544 and IVPP V 27,547 are considered as Phase 1, as the coils are not deep, and this can be explained as during excrement, there’s a mucosal membrane covering the surface of the fecal matter and embedment occurring rapidly, thus retaining most of its surface structure. Although there are signs of disentanglement, we predict that the uncoiling on the surface was not by natural processes, but has been caused by a breakage after on. Both of these two coprolites could have been large in actual size. Similar explanations can be given to specimens IVPP V 27,548 and IVPP V 27,550, whereby the coils are shallow, thus, classifying them as to had occurred in Phase 1. We classify specimen IVPP V 27,546 and IVPP V 27,549 as Phase 2, in which the spaces between the coils of IVPP V 27,546 were slightly separated and in IVPP V 27,549, they were strongly separated. Both of these specimens could have spent more time in water bodies before burial. Specimen IVPP V 27,545 does not provide any external information in regards of phases approach because of its non-spiral morphology. While it is also worthwhile to mention that none of them have spent sufficient time in the water bodies in order to possess the Phase 3 structure. Through these, we can also conclude that smaller coprolites are much complete while bigger coprolites tend to easily break-off. However, having mentioned that, the preservation of specimen IVPP V 27,550 is indeed valuable.Through the above morphological points, we predict that the amphipolar spiral coprolites could have belonged to groups of either prehistoric lungfishes or Acipenseriformes (sturgeon and paddlefish). Another aim of this work is to portray the existence of possible prey-predation relationships from the collected coprolites. In order to narrow down the identity of the potential producer and possibly the prey, we looked into some related fauna list from past literature. Geological settings have indicated that the Lower Cretaceous Tsagan-Tsab formation is not only recorded in the area of Tatal, but also in other regions of Mongolia as well36. There are two possibilities on the deduced prey and predator, they are either of Asipenceriformes—Lycopteriformes relationship or Asipenceriformes—Pholidophoriformes relationship. We suggest Pholidophoriformes as a much potential prey than the Lycopteriformes in the Tsagan-Tsab Formation, and the reasons will be explained thoroughly. As for the producer, we knew that Asipenceriformes are largely known from the Lycoptera-Peipiaosteus (Asipenceriformes) Fauna or the “Jehol Fauna”, as these assemblages of fishes were not only abundant in the Lower Cretaceous Yixian Formation of northeastern China, but also widely distributed over the region of eastern Siberia, Mongolia, northern China and northern Korea112. It is also noteworthy to mention that the Tsagan-Tsab formations and the Yixian formation were similar in geological age. In the same context, Jakolev35 described Stichopterus popovi (Asipenceriformes) and recorded amphipolar spiral coprolites from the Aptian lacustrine of Gurvan-Eren Formation of Mongolia , a locality that is close to Tatal. Although there are differences in the geological period of Tsagan-Tsab and Gurvan-Eren Formation, it is highly possible that Asipenceriformes existed in these areas. Furthermore, Asipenceriformes are shown to have spiral valves113, and this can be further proven with the work of Capasso64 on Peipiaosteus pani, thus, contributing to the morphology of the spiral coprolites. With these, we strongly suggest that the amphipolar spiral coprolites of Tsagan-Tsab Formation and for Gurvan-Eren Formation to belong to Asipenceriformes. As for prey, we know from existing literature that there is a close relationship between Asipenceriformes and Lycoptera, as evident in the name Lycoptera-Peipiaosteus Fauna. Yondon et al.36 reported Lycoptera middendorfii, a form of small freshwater Teleost fish from the Eastern Gobi—Tsagan-Tsab formation. But, it was clearly mentioned that Bon-Tsagan/Bon-Chagan (Fig. 1) is the westernmost locality of Lycoptera in Mongolia114. Another fact that was taken into account for the possible prey is the shape of the scales found in the inclusions, whereby Lycoptera are known for their cycloid shaped scales, while the ones in our specimens are more towards rhomboidal-shaped ganoid scales. These facts crucially eliminate the possibilities of Lycoptera for the Tsagan-Tsab fauna. With this, we further examined Jakolev35′s works and discovered the species that he described, Gurvanichthys mongoliensis (Pholidophoriformes) from the Gurvan-Eren Formation has rhomboidal-shaped ganoid scales. The size, shape of the scale and the nature of this fish fits well as a prey for the Stichopterus popovi (Asipenceriformes). Through these interpretations, we can possibly infer that the spiral coprolites in our study might have belonged to Asipenceriformes and Pholidophoriformes as the prey, which could further affirm the occurrence of prey-predator inter-relationship in the Lower Cretaceous of Tsagan-Tsab Formation.As for the sole scroll coprolite in this study, we do not intend to further deduce any detailed possibilities. Based on other works, chondricthyans origins or a sarcopterygian for scroll coprolites were suggested18,53,but such deduction is difficult to be purported in our studies as there is a lack of such fossil materials in the locality and surrounding localities. The chances of the underived producer to be a sarcopterygian is much higher than to be a chondricthyan, mainly due to its geological settings. The discovery of the single scroll coprolite can be a window opening to many paleontological questions for Tsagan-Tsab Formation. More
