Coprolites (paleofeces, fossil feces or fecal pellets) have been considered as intriguing fossils1, 2, 3, 4 with low preservation potential5, and have gained a place in the multi-disciplinary research spectrum6, 7, 8. They can be interpreted as biogenic structures that are identified as a trace fossils, which are also known to be a vital tool to not only obtain paleobiological inferences9, 10, 11, 12, but also played an important role in preserving traits of behavior, the ancient trophic relations, the feeding habits and dietary, the coprophagy, the producer’s digestive tract biology and structure, followed by palynomorph evidence, bacterial residues, parasitism and DNA fragments13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30. The findings of coprolite have been recorded throughout the Phanerozoic in every continent since the 17th century, but its ichnosystematic position and classification are rather new and controversial. Also, spiral coprolites have been studied and well described, but the status of crocodilian origin coprolites has been under investigation for more than 15 years.
Crocodilian coprolite studies have been known since 1832 (see Supplementary Table 1 for full list). Robert31 reported some coprolites from the Eocene of France and deduced that it could be of crocodilian origin. In the last two centuries, a total of 26 studies have announced the findings of crocodilian coprolites from 15 countries between the Early Cretaceous and the Holocene. The oldest known possible crocodilian coprolites were recorded in Belgium32, 33, while the most recent reported coprolites belong to Crocodylus niloticus in the Sahara Region, Africa34. Looking at trace fossil records, 9 Mesozoic localities have reported on the presence of crocodilian coprolites, to which 8 of them are from the Cretaceous and one from the Triassic (Pseudosuchia). On the other hand, crocodilian coprolite localities are much abundant during the Cenozoic with 16 localities. Ironically, 11 localities are Paleogene aged, including Na Duong coal mine and one of them is from Nanxiong (Paleocene), China which was extensively studied by Young35.
In this research, we described the biogenic structures, composition and inclusions of 55 well-preserved coprolites, which herein are attributed to crocodilian coprolites that are found in the Late Eocene of Na Duong Formation, Lang Song Province, Northern Vietnam (Fig. 1). The discovery of the new ichnospecies provided new information towards the study of the crocodilian paleobiology based of Na Duong coal mine. It is also the first quantitative analysis on crocodilian coprolites during the Eocene and this has significantly contributed to our understanding on the variations of Na Duong coprolite, followed by the factors that contributed to such enigmatic preservation. Beside these, the study also highlights the multi-disciplinary approach in understanding paleobiology and paleoecology records in trace fossils. The study further aimed to recreate a window snapshot of the Oligocene-Eocene Na Duong flora and fauna niche and deduce possible factors that led to its disappearance.
Systematic ichnology
The use of a formal ichnotaxonomy for vertebrate bromalites has been an indecisive matter due to its relevance of usage, despite being mentioned in the works of Bertling et al. (36, p.265) that the “The need to name trace fossils…has unambiguously been accepted for decades”. A prominent example is those of widely applied in the studies for fossilized eggs. In coprolite work, the earliest informal naming can be seen in the works of Buckland1, 37. Works of Duffin38, Laojumpon et al.39, Milàn40 etc. have given good examples on how ichnotaxonomy can be applied for vertebrate bromalites. However, the reluctance to apply such a binomial scheme to bromalites (coprolites) is thought to be due to the occurrence and misconception that modern feces are not be distinguishable because of their variability, despite having clear evidences that extant taxa can be distinctly differentiated based on their feces41, 42, 43, 44. But there are crucial needs to consider the opinions of some workers, especially for Chin (in 45) who opted to avoid such nomenclature for coprolites for reasons of : (1) preservation of coprolites can be varied due to taphonomy and diagenesis; and (2) diets play a vital role in producing different morphologies for the same producer.
In this work, our opinions are in accordance to Hunt and Lucas7, 45, and that we strongly suggest using nomenclature for ichnotaxonomy of vertebrate coprolites, rather than just relying on morphotypes, owing to both philosophical and practical reasons, and as well as to extend their utilities. As per required by the ICZN (International Commission on Zoological Nomenclature)46, trace fossil nomenclature has to follow a number of determined rules. Since it has been generally agreed that trace fossils are not the actual remains of plants or animals, but are the actual remains of plant or animal behavior. As for this, trace fossils are named by using ichnogenus and ichnospecies, despite the fact that the coprolites cannot be assigned to a producer at the family level or even higher taxa. Three main rules indicated that: (1) Be certain that the fossil is a trace fossil; (2) Trace fossil should not be named the same as the organism that produced it; and (3) Follow all the other rules of taxonomy as you are naming an organism. In accordance to ICZN Article 66. Application:
Article 66.1. An ichnotaxon at the genus-group level proposed after 1999 must have a type species fixed for its name to be available. If established before 2000 it does not require a type species; however, one may have been, or may be, fixed in accordance with Article 69 (see also Article 13.3.3).
Chame44 for instance, a wildlife biologist, collected extant feces and made a catalogue that could easily distinguish feces and its producer. Meanwhile, Hunt and Lucas45 gave binomial names for Mammuthus and Hyaena coprolites with their producers. On the basis of these premises, we followed Hunt and Lucas’45 concept of coprolite nomenclature.
Hereinafter, we are introducing a nomenclature for crocodilian coprolites for the first time in fossil records, as they showcased very distinctive morphologies and structures that can be uniquely attributed towards them. On a side note, we are fully aware of the Rule No. 2 of ICZN and would hereby strongly advocate the prospects of having coprolites to be named after the organism that produced it.
Crococopros ichogen. nov. Halaçlar et al., 2022 (This article).
Type species. Crococopros naduongensis ichnosp. nov.
Included species. Known only from the type ichnospecies.
Etymology. Generic name composed of Croco, derived from the Latin crocodilian for the possible producer of the coprolites and the Greek kopros for “dung” (the letter k has been altered to c to compensate the overall name).
Distribution. Upper Bartonian – Late Priabonian (34–39 MA), late Middle to Late Eocene of Na Duong coal mine, Northern Vietnam47.
Diagnosis. They are generally fusiform, oval in shape and have smoothly concave ends. Hunt et al.48 and Krause and Piña12 noted these features, and Milàn49 showed the similarities in some extant crocodilian feces and coprolites, which have been referred to as crocodilian. Generally, one end is more flattened than the other end. Hunt et al.48 reported longitudinal striations for Alococopros (48, Fig. 3A) but Milàn49 did not mention longitudinal striations on extant crocodilian feces, which literally means that without having these features, they can be similar to Crococopros. On the surface and visible infrastructure of Crococopros, there are no remains of any large or complete bone residues, but CT scans showed some fragmentary remnants that could be attributed to bone fragments. A strong digestive system is a hint that it could be of crocodilian origin48, 12, 49. Milàn et al.40 mentioned the circumferential zone of constriction for crocodilian feces and some Crococopros specimens have the same features (Fig. 2, Plate 6). Krause and Piña12 showed similar feature on crocodilian coprolite as well (12, Fig. 5.3 and 5.4).
Discussion. Milàn et al.40 described a circumferential constriction mark on extant crocodilian feces and some coprolites from the Middle Eocene to the earliest Oligocene of Denmark. Such described feature is common on Crococopros. Other examples are from Krause and Piña12, where 28 crocodilian coprolites were studied and their measurements were compatible to Crococopros. A prominent detail to be noted is that Crococopros does not have the longitudinal striations as those observed in Alococopros. Also, most Crococopros specimens are biometrically larger than Alococopros. Costacoprus is characterized by the prominent exterior feature of the coprolite, which has a narrowly spaced (2–3 mm) transverse, and rounded ‘ribs’ (50, Fig. 2R-S). Crococopros doesn’t have these features and measured slightly larger than Costacoprus. On the other hand Eucoprus (51, Fig. 4; 52, Fig. 3E-V) are quite similar to Crococopros, except the fact that Crococopros are larger than Eucoprus in size, and that Crococopros has circumferential zone of constriction.
Crococopros naduongensis ichnosp. nov. Halaçlar et al., 2022.
(Fig. 2, Plate 1 to 4; Supplementary Fig. 2)
(urn:lsid:zoobank.org:act:xxxxxxx)
Holotype. IVPP V 27941/46 (Fig. 2, Plate 4c)
Referred specimens. IVPP V 27941/1 to IVPP V 27941/55 (Fig. 2, Plate 1 to 4; Supplementary Fig. 2)
Type locality. Na Duong coal mine, Northern Vietnam.
Type horizon. Na Duong Formation.
Etymology. Name after the locality, Na Duong where the type specimen originates.
Distribution. As for ichnogenus.
Description. Holotype is 201 mm long, 58 mm maximum diameter, and with 42 mm second diameter. It’s one end is flattened while another end is tapered with circular cross-sections (Fig. 2, Plate 4c). It has a circumferential constriction mark that coincides with a bend at an angle of approximately 120° to 150° (Fig. 2, Plate 6e), which is similar to extant crocodilian feces (49, Fig. 1). The surface of the Crococopros naduongensis ichnogen. et ichnosp. nov. is covered by a tiny fragile layer and inside is massively homogenized. Some of them are composed of several single, fused concavo-convex fecal units after the tiny layers.
Chin53 mentioned that if a large number of coprolites from a given locality denotes an unbiased sample and recurring size classes, this could be an indication that the specimens could have been produced by different taxa or age groups. Crococopros naduongensis ichnogen. et ichnosp. nov. can be classified into 5 ichno-variations based on their morphologies (Fig. 2, Plate 1 to 5).
Variation A (VarA) - Most abundant variation of Crococopros naduongensis ichnogen. et ichnosp. nov. with Number of Specimens (herein to be known as NS) of 35. The average length is 88 mm and the average maximum diameter is 38 mm. It is medium-sized and slightly flattened. One end is more tapered than the other (Fig. 2, Plate 1).
Variation B (VarB) - Large, flattened and both ends are tapered with NS of 3. Differences between the maximum diameter and the second diameter are greater than other variations (Fig. 2, Plate 2).
Variation C (VarC) - Second most abundant variation of Crococopros naduongensis ichnogen. et ichnosp. nov. with NS of 5, the average length is 84 mm, the average maximum diameter is 46 mm and the average second diameter is 39 mm. It is oval in shape, the cross-section is circular and both ends are tapered (Fig. 2, Plate 3).
Variation D (VarD) - Large and massive coprolites, one end is significantly tapered than the other end with NS of 4. Its cross-section is circular. This variation has strong circumferential constriction mark. Its average length is at 148.5 mm, the average maximum diameter is 63.5 mm and the average second diameter is 46 mm (Fig. 2, Plate 4).
Uncategorized Group - NS is 8 and this group includes one thought to be regurgitalite specimen, but CT scan observation did not show any residues, hence, we consider it as a taphonomically distorted coprolite (Fig. 2, Plate 5).
Discussion. Crococopros naduongensis ichnogen. et ichnosp. nov. has the same morphology and size as extant crocodilians’ feces (12, Fig. 6—6; 49, Fig. 1). Milàn49 collected 17 feces from 10 extant crocodilian species and his results shed light on the relationship between the variations of Crococopros naduongensis ichnogen. et ichnosp. nov.
Discussion I - Inter-ichnospecies of Crococopros naduongensis ichnogen. et ichnosp. nov. Biometrically, Crococopros naduongensis ichnogen. et ichnosp. nov. is not compatible with Alococopros. In addition, Crococopros naduongensis ichnogen. et ichnosp. nov. doesn’t have longitudinal striations and this puts it in a different ichnotaxon from Alococopros.
Young’s35 coprolites were morphologically and biometrically similar to Crococopros naduongensis ichnogen. et ichnosp. nov. The surface is covered by a tiny fragile layer and both insides are massively homogenized. There is a high possibility that Young’s coprolites could have belonged to the same ichogenus with Crococopros naduongensis ichnogen. et ichnosp. nov..
Note
Young’s35 Nanxiong coprolites, which are from the Nanxiong Basin, Guangdong Province, South China (also known as Nanhsiung, Kwangtung) and Na Duong coprolites are very similar in morphology and biometry, as both localities belong to the Paleogene and both localities have recorded 3 different crocodilian species at the same time. Future biogeochemical and biometric data from the Nanxiong coprolites are needed to extend this hypothesis.
Discussion II - Intra-ichnospecies variation of Crococopros naduongensis ichnogen. et ichnosp. nov. and possible producer of Crococopros naduongensis ichnogen. et ichnosp. nov. (morphotypes of Crococopros naduongensis ichnogen. et ichnosp. nov.). Böhme et al.47 primarily described Na Duong crocodilian and have divided them into 3 morphotypes. Brevirostrine type is around 2 meters and they constitute 2/3 of all Na Duong crocodilian; a longirostrine crocodilian type is more than 6 meters and they occupy 1/3 of all Na Duong crocodilian, while the last one is a second longirostrine gavialoid type where it is more than 6 meters and just one specimen was found in the Na Duong coal mine.
Milàn49 argued that the feces of Gavialis gangeticus is hard to be preserved in nature because after the animal defecated in the water body, their feces easily get separated and dispersed. Taking this information as a crucial clue, we deem that longirostrine gavialoid types’ feces couldn’t be preserved and hence we have inconclusively decided that their availability can’t be assured in the Na Duong fossil collection, or at least they cannot be placed as sufficient enough to make a variation. Judging from this, we can deduce that Crococopros naduongensis ichnogen. et ichnosp. nov. was most probably produced by those of brevirostrine crocodilian type and longirostrine crocodilian type.
The brevirostrine type crocodilian of Na Duong is the smallest and the most abundant crocodilian in the Na Duong Herpeto-fauna. Hence, it can be matched with VarA, which is the smallest and the most abundant variation of Na Duong coprolites. VarA has a flattened cross-section and it is the same with VarB, despite it being larger than VarA. Hence, with this, we deduce that the possible producer of VarB is the large-sized brevirostrine type crocodilian. It is represented by 4 specimens.
VarD is the largest variation of Crococopros naduongensis ichnogen. et ichnosp. nov. and it should be produced by a 6-meter-long longirostrine crocodilian type. In Fig. 4, it is shown that a 450 cm long Crocodylus cataphractus can create a minimum of 5 cm diameter long feces but a 250 cm extant crocodilian can only create a maximum of 4 cm diameter feces. So, having said that, VarC, which has more than 4 cm Wm could possibly be produced by a longirostrine crocodilian type. Moreover, VarC and VarD have circular cross-sections, and with such similarities, it is possible to deduce that the producer of VarC is a juvenile longirostrine crocodilian type.
Milàn’s49 work showed a 450 cm long Crocodylus cataphractus that can produce feces bigger than 4 cm. With this information, according to Fig. 4, we deduced that some of Crococopros naduongensis ichnogen. et ichnosp. nov. specimens that are larger than 4 cm in diameter were possibly produced by longirostrine crocodilian type.
Copro-phylogeny of Archosaurs. In recent years, a number of archosaur coprolite studies have been conducted, which projected the importance of trace fossil records. In this work, we introduced the term ‘copro-phylogeny’. Copro-phylogeny can be understood as the production of coprolite nomenclature by using ichnogenus and ichnospecies, which is represented in a cladogram. Although phylogeny is always related to the origin and relationship of a species to one another (branching), such relationship cannot be described or presented directly here in the field of trace fossils. However, we believed with the addition of more coprolite ichnotaxonomy work in the future, there could be more unique findings with relation to the coprolite.
At the present moment, in a preliminary mode, we would like to suggest a revision for the ichnogenus Eucoprus (Fig. 5). This is due to the fact that Hunt and Lucas50 mentioned that Eucoprus might have been produced by basal archosaur during the Mesozoic. Later on they found similar coprolites from the Eocene of Kazakhstan and the workers classified them into the same ichnogenus. Here, we argue that Eocene age Kazakhstan specimens might belong to Crococopros because they have quite a similar morphology as to of a small sized crocodilian in the Eocene of Kazakhstan. Meanwhile, Costacoprus and Alococopros are morphologically different, and should have been produced by other reptilians and not of any crocodilian origin. A detailed comparison will not be discussed in this study.