Coprolite

Abstract

Coprolites have been studied for almost 200 years. Key workers include William Buckland, who first identified and named coprolites, Eric Callen, who recognized the importance of human coprolites and Paul Martin, who sparked interest in Pleistocene coprolites and their palynology. Bromalite are a grouping of trace fossil related to feeding that include regurgitalites, consumulites and coprolites. It is important that coprolites be given binomial names (an ichnotaxonomy). The earliest probable vertebrate coprolites are Ordovician in age, but invertebrate coprolites date back to the Early Cambrian. Both vertebrates and invertebrates have important coprolite fossil records through the remainder of the Phanerozoic. The producers of a few distinct invertebrate coprolites can be recognized at the family level, but many cannot even be attributed to a Phylum. Vertebrate coprolites can usually be assigned to a producer at the ordinal level. Coprolites have diverse utility in invertebrate and vertebrate paleobiology and other fields of study.

3.2.4 Human Coprolites

Human coprolites, as identified through nematode egg analysis (Kenward & Hall, 1995), may have a presumed hydroxyapatite matrix similar to that of dog coprolites, but differ by normally being a yellowish colour (e.g. Courty et al., 1989). They also often have a different content reflecting their omnivorous diet (both bone and plant remains occur), and they do not typically contain much clastic mineral material or coarse bone. As such, they are far more likely to include plant food fragments, such as legume testa and cereal material, including articulated phytolith-rich bran; thin sections of mummified intestinal contents found solely leguminous seed cases in some Chilean mummies (Macphail et al., 1990; unpublished reference material) (Fig. 16, 17). Here the cellulose of the seed cases is still birefringent unlike the humified cellulose found in herbivore dung – herbivores being better adapted to digest cellulose compared to humans. Probable intestinal remains that autofluoresce in blue light have been found in an Iron Age grave in Norway where bands of vivianite can also mirror wooden coffin remains, and it appears that these residues contribute to body stains in the soil (see Keeley et al., 1977).

Fig. 16. Human coprolite (Viking Age Coppergate site, York), identification confirmed by nematode egg studies (Andrew Jones, York Trust). Detail of legume testa as a food remain within the apatite groundmass; the hook-like tissue cells making up the testa are diagnostic (Ann Butler, Institute of Archaeology, UCL, personal communication) (PPL, scale bar length 25 µm).

Fig. 17. Human coprolite (Middle Saxon settlement site Maiden Lane, London, UK), identification confirmed by parasite egg study (Claire de Rouffignac, Museum of London Archaeological Service, personal communication). The sample exhibits voids from trapped gas; dark black and reddish-brown colours relate to iron and manganese oxide staining associated with unidentifiable but likely food plant traces embedded within the coprolite (PPL, scale bar length 1 mm).

Latrines contain mineralized cess, again with a probable hydroxyapatite composition (Fig. 18). Such waste can contain embedded cereal material of dietary origin. Fine fragments of this type of latrine waste often occur as a regular component of Roman and Medieval beaten floors in the United Kingdom. Dumped toilet waste and seepage from cess pits, like modern day sewage sludge, produce yellow amorphous infillings and nodules of Fe–Ca-phosphates, sometimes with neoformation of vivianite (see Karkanas & Goldberg, 2010, this book). It is common to find phosphate-embedded charcoal, both in latrine waste and in fields, implying that such material, which may also include embedded ash, was used as a “night soil” fertilizer worldwide (Henning & Macphail, 2004; Goldberg & Macphail, 2006).

Fig. 18. Lower fill of a cess pit (Norman cess pit, Monkton, Kent, UK). (A) Cess preserved as amorphous yellow material (probably calcium phosphate); mineralized cess (formed in aerobic conditions) is yellow, mainly optically isotropic although some birefringent materials can also be present in addition to probable hydroxyapatite (PPL, scale bar length 1 mm). (B) Same field in blue light fluorescence microscopy, showing strong autofluorescence of the material.

3.2.4 Human Coprolites

Human coprolites, as identified through nematode egg analysis (Kenward & Hall, 1995), may have a presumed hydroxylapatite matrix similar to that of dog coprolites but differ by normally being a yellowish colour (Courty et al., 1989; Goldberg et al., 2009a; Macphail, 2016b). They also often have a different content reflecting their omnivorous diet (both bone and plant remains occur), and they do not typically contain much clastic mineral material or coarse bone. As such, they are far more likely to include plant food fragments, such as legume testa and cereal material, including articulated phytolith-rich bran; thin sections of mummified intestinal contents found solely leguminous seed cases in some Chilean mummies (Macphail, 2016b) (Figs. 15 and 16). Here the cellulose of the seed cases is still birefringent, unlike the humified cellulose found in herbivore dung – herbivores being better adapted to digest cellulose compared with humans. Body stains and probable intestinal remains that autofluoresce in blue light have been found in an Iron Age grave in Norway where bands of vivianite can also mirror wooden coffin remains; both amorphous yellow (Ca-P) and reddish brown (Fe-Ca-P) body stains occur in the pelvic area of a warrior found in a boat grave near the Gokstad Mound (Macphail et al., 2013; Viklund et al., 2013).

Figure 15. Human coprolite (Viking Age Coppergate site, York), identification confirmed by nematode egg studies (Andrew Jones, York Trust). Detail of legume testa as a food remain within the apatite groundmass; the hook-like tissue cells making up the testa are diagnostic (Ann Butler, Institute of Archaeology, University College London, personal communication) (PPL, scale bar length 25 μm).

Figure 16. Human coprolite (Middle Saxon settlement site Maiden Lane, London, UK), identification confirmed by parasite egg study (Claire de Rouffignac, Museum of London Archaeological Service, personal communication). The sample exhibits voids from trapped gas; dark black and reddish brown colours relate to iron and manganese oxide staining associated with unidentifiable but likely food plant traces embedded within the coprolite (PPL, scale bar length 1 mm).

Latrines contain mineralised cess, again with a probable carbonate hydroxylapatite composition (Fig. 17). Such waste can contain embedded cereal material of dietary origin, as well as sphagnum moss that was most likely used as toilet paper. Fine fragments of this type of latrine waste often occur as a regular component of Roman and Medieval beaten floors in the United Kingdom. Dumped toilet waste and seepage from cess pits, like modern day sewage sludge, produces yellow amorphous infillings and nodules of Fe-Ca-phosphates, sometimes with neoformation of vivianite (Macphail, 2016b; see Karkanas & Goldberg, 2018). It is common to find phosphate-embedded charcoal, both in latrine waste and in fields, implying that such material – which may also include embedded ash – was used as a ‘night soil’ fertiliser worldwide (Henning & Macphail, 2004; Goldberg & Macphail, 2006).

Figure 17. Lower fill of a cess pit (Norman cess pit, Monkton, Kent, UK). (A) Cess preserved as amorphous yellow material (probably calcium phosphate); mineralised cess (formed in aerobic conditions) is yellow, mainly optically isotropic although some birefringent materials can also be present in addition to probable hydroxylapatite (PPL, scale bar length 1 mm). (B) Same field in BLF, showing strong autofluorescence of the material.

Abstract

Spotted hyaena (Crocuta crocuta) coprolites from four British Pleistocene sites were analysed for their pollen content. At the two open sites, the palynology was compared to that of the surrounding sediments. The results provide palynological data, supported by other lines of evidence, enabling reconstruction of human environments as well as providing insights into the taphonomic complexities of incorporation of pollen into coprolites. Pollen presence and preservation appear to be closely related to mammalian behaviour and post-depositional processes. Geological age does not seem to be a significant factor, as samples from two of these sites are amongst the earliest known from the Pleistocene to provide viable coprolite-derived pollen counts.

Diet

Genetic analysis of desiccated feces – also known as coprolites – can be used to identify which organism produced the feces and what plants and animals formed its diet. Despite the corrosive nature of mammalian digestive processes, a significant amount of amplifiable DNA can be recovered from both recent and ancient coprolites. For aDNA studies, bulk screening of plants in coprolites presents a fascinating opportunity to better understand the diets of past mammals. For example, van Geel et al. (2008) tested feces found in the digestive tract of a 22 500-year-old Siberian woolly mammoth for the rbcL gene and recovered ten DNA sequences which were consistent with the plant macrofossil and microfossil remains in the coprolite. Bulk screening of ancient human coprolites is also feasible; however, apart from a handful of studies (e.g., Rasmussen et al., 2009), this approach has been arguably underutilized by the archaeological community. Indeed, the analysis of plant DNA within a coprolite can be readily tested in conjunction with the DNA of the individual who produced the stool (cf. Gilbert et al., 2008; Rasmussen et al., 2009).

Before attempting to recover plant DNA from ancient coprolites, researchers should be aware of a number of noteworthy limitations and caveats. First, coprolites frequently contain various chemical compounds which impede DNA extraction and inhibit PCR amplification. Second, it has recently become recognized that plant DNA which is not part of an organism’s diet can be incorporated in coprolites through inadvertent swallowing of airborne and waterborne pollen (cf. Reinhard et al., 2008). This occurrence may complicate interpretations of past diets but conversely may aid environmental reconstructions by providing a snapshot of animal’s habitat shortly before its death. Another critical factor in bulk-specimen research is the database used for sequence queries. In order to determine the taxonomic origin of a genetic sequence, computational searches must be made against DNA databases, such as GenBank, in order to locate the most likely candidate. Thus, taxonomic identifications are fundamentally limited by the scope of the database. On the other hand, it is possible to revisit old data as online databases grow larger. For example, Reinhard et al. (2008) reanalyzed plant sequences recovered from prehistoric human coprolites from Texas first reported by Poinar et al. (2001). With the aid of an expanded database, they found that some previously unidentified sequences matched plants with medicinal and hallucinogenic properties. A final concern in coprolite DNA research is that it suffers from stochasticity in the finding of samples, thereby limiting choice of time periods and species that can be studied.

Southern California

Mychajliw et al. (2020) examined packrat coprolites preserved in asphalt deposits at Rancho La Brea (RLB), Los Angeles, California. These fecal remains date from 50 ka to recent. The authors employed multiple lines of research, including radiocarbon dating, body size reconstructions, stable isotope analysis, scanning electron microscopy, and sediment analyses. Unfortunately, packrat droppings contain few if any identifiable plant macrofossils. However, plant macrofossils were preserved in the same stratigraphic layers as the rodent coprolites. Prior studies of RLB plant material have been hampered by a lack of stratigraphic control for plant assemblages, and the plant macrofossils in the midden studied by Mychajliw et al. (2020) represent stratigraphically intact material that could be assessed as a coherent assemblage.

Coprolite Studies

The study of ancient human feces, preserved in dry caves as coprolites, has yielded some interesting information on insects and their interactions with Amerindians (see COPROLITE ANALYSIS). Some arthropods were consumed as food. These include cicadas and grasshoppers that were eaten by the Anasazi at Mesa Verde, Colorado, and ants and termites that were apparently eaten by Paleoindians at the Dirty Shame Rockshelter in Oregon. Other insects, such as fleas, lice, and mites, were external parasites that were either accidentally or deliberately consumed.

The remains of fly maggots and their pupal exuviae (the skin cast from the larva as it pupated) have been used in some New World sites to help determine mode and timing of burial. A study of fly remains from a Late Prehistoric Arikara Indian burial site at Leavenworth, South Dakota, placed the time of burial between late March and mid-October (the flies' active season in South Dakota).

Case Studies – Dietary Residues from Burials

As noted above, Shafer et al. presented the first modern paleoethnobotanical analysis of fecal residues (coprolites) discovered within a burial. Shafer et al. presented multidisciplinary methods for the recovery of microscopic and macroscopic remains that related to diet and medicine. Control sediment samples were taken above the pelvic girdle, grave floor, and from the burial fill. Macroscopic evidence of finely ground maize cupules was found in the coprolites. The coprolites contained high concentrations of maize pollen. Pollen from a medicinal plant, willow, was also found in high concentrations in the coprolites (Table 1 and Figure 3). Pollen concentration adds another dimension to pollen data. In addition to knowing the types of percentages of pollen in sediments, pollen concentration provides an estimate of the number of pollen grains present per gram or milliliter of sediment. This is illustrated by the fact that maize percentages are not statistically different between the samples from the Mimbres burial. However, the pollen concentration shows that compared to the average concentration of 292 pollen grains per gram for the control samples, the sacral coprolites contained 59 280 grains per gram.

Figure 3. The pollen spectra from a coprolite and control samples from a Mimbres burial.

Later, Reinhard et al. presented field and laboratory methods for recovering dietary residues from pelvic girdle sediments from a single burial. They recommended paleoethnobotanical study of burials as a nondestructive method to identify diet and season of death. In this case study, they emphasized the need to analyze multiple control samples to identify contamination. In this case, multiple samples were taken from a stratigraphic column within the pelvic girdle (Figure 4). The burial was excavated into a prehistoric house floor and there were a variety of contexts to sample for comparative evaluation. These included the burial fill, the house floor, a storage pit, the area of the lower legs, and the area of the skull. Seeds were found only in the sacrum (Figure 5) and 345 seeds were found in 4.3 g of sediment from the sacrum. No seeds were recovered in the 50-g control samples. Pollen aggregates (clumps of pollen of the same type) were found only in the sacrum. The seeds and pollen aggregates indicated that pigweed seeds, goosefoot seeds, sunflower seeds of two species, panic grass, and two other species were eaten. Pollen aggregates of grass, pigweed and/or goosefoot type, sunflower type, and ragweed type were found. Therefore, the pollen and seed evidence show that the deceased ate a variety of wild plants before he died. The pollen spectra comparison shows an elevated percentage of grass pollen in the sacrum compared to the other samples (Table 2 and Figure 6).

Figure 4. The pelvic girdle sampling strategy proposed by Reinhard et al. in their analysis of an ancestral Anasazi burial. Samples 1 and 2 are control samples taken from a stratigraphic column within the pelvic girdle. Sample 3 is a sediment sample that should be collected from the anterior surface of the sacrum, where dietary remains from the large intestine are trapped during decomposition.

Figure 5. Seeds from the ancestral Anasazi burial. The upper left shows goosefoot and pigweed seeds. The upper right shows wild sunflower seeds. The lower images are of unknown seeds. This shows that well-preserved plant foods can be recovered from burial sediments.

Table 2. Macrobotanical data from selected Arizona burials (Berg 2002)

Taxon	Burial A		Burial B		Burial C
	Control	Sacral	Control	Sacral	Control	Sacral
Maize fragments	30	254	11	189	13	0
Agave fragments	5	424	17	491	3	35
Cactus seed	0	0	1	75	0	0
Cotton seed	2	0	8	302	0	105
Plant epidermis	0	0	13	264	6	0

Figure 6. The pollen spectra from sacrum sediments and control samples from the ancestral Anasazi burial.

Berg expanded these methods to cemeteries in Arizona (Table 3) and Denmark and successfully demonstrated that the multidisciplinary approach presented in the first three case studies is extremely powerful when applied to detecting patterns of medicine and diet in groups of interments. He also formalized a consistent sampling strategy that can be applied to most burial excavations (Figures 7 and 8).

Table 3. Pollen data from the burial analyzed by Reinhard et al. (1992)

Taxon	Sacrum	Burial floor	Pithouse floor	Storage pit
Pine	4	10.5	3	10.5
Juniper	2	17	8	8
Wild grass	42.5	10	4	4.5
Sunflower type	6.5	3	1.5	5.5
Ragweed type	15.5	4.5	33	13.5
Sage brush		0.5	4.5	7
Cheno-am	18.5	47.5	26	36.5
Rocky Mountain bee plant		2	17	3
Squash				0.5
Mormon tea		1	1	2.5
Euphorbia type		0.5
Bean family	7	0.5		1
Globemallow	2.5			3
Prickly pear				0.5
Gilia	1
Oak		0.5		1.5
Willow		0.5
Greasewood		0.5
Solanaceae				0.5
Maize				1
Spruce		0.5		1

Figure 7. Berg (2001) formalized a sampling strategy for macroscopic flotation and pollen sampling for burials. He proposed that macrofossil sampling be focused on flotation samples from the pelvic area and from the grave fill near the head (red areas). He recommended that pollen control samples be recovered from long bone areas and a single control sample in the pelvic girdle. The dietary pollen sample should be recovered from the anterior surface of the sacrum (blue areas).

Figure 8. Macroscopic remains from Berg's analysis of three Arizona burials. Burial A shows an elevated frequency of maize and agave fragments in the sacrum sample compared with the control sample. In burial B, increased numbers of maize, agave, cactus, cotton, and plant epidermis macrofossils were found in the sacral sample relative to the control sample. Agave and cotton seed were evident in the burial C sacral sample.

5.1 Phylogenetic Classification

In the phylogenetic classification, traces are called after their makers, for example, a Phoronis burrow, the coprolite of Tyrannosaurus rex, or more informally, a bird footprint or buffalo wallow. This classification is the most intuitive and is also potentially the most informative, and it is used in virtually all treatments of modern traces (e.g., Eiseman et al., 2010; Elbroch, 2003; Elbroch and Marks, 2001). The aspects of a trace that are useful in identifying its maker may be called its bioprint (Rindsberg and Kopaska-Merkel, 2005). However, in the absence of the organisms and distinctive bioprint, traces cannot always be identified correctly as to their maker even in modern environments, especially those that are most unfamiliar to us, such as the deep sea (Ekdale, 1977; Wetzel, 1981; Wetzel and Uchman, 2012). In ancient examples, it is even more difficult to link a trace fossil with its maker. The two are only rarely preserved together, and even where a body fossil is found within a trace fossil, a logical case must be constructed to test whether the organism made the trace instead of simply harboring or being deposited there.

Another problem is that trace fossils commonly do not contain the right kind of data to allow them to be classified in a manner even roughly paralleling that of their biological makers. At the present state of the art, it is not possible to say that all Diplocraterion belong to any one clade, on the contrary. Neither is it possible to classify all the trace fossils made by a single organism together as one taxon, because they can be very diverse, a point that is discussed further under Section 6.

The trace fossils of some organisms are more apt for biological classification than others, especially in cases where the trace fossils are complex and are well preserved. The effort is worthwhile for an increasingly long list of trace fossils, notably, trilobite trails and resting traces (Cruziana and Rusophycus;Seilacher, 1970), the footprints of fossil vertebrates, the communal burrows of social insects (Genise, 2000), arthropod trackways (Minter et al., 2007), and many kinds of borings, including those of sponges, bryozoans, and bivalves (Bromley, 1970). In some cases, specialists have gone so far as to classify trace fossils in parallel with modern biological species, for example, boring bryozoans (Pohowsky, 1978) and vertebrate trackways. In the case of boring bryozoans, the borings accurately represent the shapes of several internal organs; moreover, the animals die and decay within minutes so that the standard technique for studying them is to remove the tissues as a matter of course before study of the borings (J.D. Soule, Allan Hancock Foundation, University of Southern California, Los Angeles, oral communication, 1982). This is a more ambiguous case than that of vertebrate trackways, which at best can represent only a small part of an animal's body and behavior, but which nevertheless often have been harnessed as representatives of the missing animal itself.

It is always worthwhile to ask what made a trace fossil, but at present the makers of most trace fossils cannot be identified unambiguously. Fossil footprints represent only a small part of the morphology of the animal. Even in the most convincing case, that of boring bryozoans, this is an error. It is true that, a few decades ago, boring bryozoans could be best classified by the morphology of the boring; comparison between living and fossil specimens was thus facilitated. But as molecular analysis replaces morphology in the study of living organisms, the classification of modern and fossil borings will become disconnected again. Fossil borings lack the necessary DNA.

Parasites have been a part of human existence throughout evolutionary history. By recovering evidence of parasites from artifacts, human remains, coprolites, and archaeological sediments, this relationship between humans and parasites can be tracked. Because human-specific parasites evolved and move with their host populations, prehistoric transhumance can be retraced. People responded to the symptoms of parasitic disease by developing a broad pharmacopoeia. When remedies could not control parasite infection, disease threatened the vitality and existence of some communities. The relationship between urbanization and the emergence of parasitic disease as an unavoidable aspect of human existence has been detailed by archaeoparasitology. The development of archaeoparasitology in the future is enhanced by molecular biology and other new biochemical applications of medicine to archaeology.