Introduction
Adulis, on the coast of present-day Eritrea, was an important hub during the rise of cross-ocean maritime trade, connecting ships, cargoes, and ideas from Egypt, Arabia, and India (Burstein, 2002; Munro-Hay, 1982; Seland, 2008). Trade peaked between the fourth and seventh centuries CE, propelling the rise and expansion of the Aksumite kingdom, but its occupation history extends, at minimum, to the first millennium BCE (Zazzaro et al., 2014). Corroborating this archaeological record are written accounts that draw attention to the importance of Adulis as one of the foremost sources of African animals or animal products during the Hellenistic period (323–31 BCE). In
Echoing this account is the first-century
Figure 1.
Strabo’s reference (17.1.40) to the worship of cynocephali at Hermopolis Magna makes clear that the animal in question is the hamadryas baboon (
The sanctuary and temple complex featured several 35-tonne statues of
Figure 2.
Present-day distributions of the six baboon species, major mitochondrial clades, and provenance of samples analysed in this study.
(a) Overview of species distributions according to the IUCN (2020) and coloured by species (red:
Though fragmentary, this historiography points to Adulis as a commercial source of mummified baboons in Ptolemaic catacombs, such as those at Saqqara and Tuna el-Gebel (Goudsmit and Brandon-Jones, 1999; Peters, 2020) [or those of their progenitors if Ptolemaic Egyptians maintained captive breeding programs; (von den Driesch et al., 2004)]. At the same time, these accounts invite questions focused on the source of pre-Ptolemaic baboons recovered from Gabbanat el-Qurud, Egypt (Lortet and Gaillard, 1907) and dated to ca. 800–540 BCE (Richardin et al., 2017), a span that corresponds to the 25th Dynasty and Late Period of Egyptian antiquity. If these older specimens can be traced to Eritrea, and by extension Adulis, then they have the potential to extend the time depth of Egyptian–Adulite trade by as much as five centuries.
Mummified baboons have been investigated morphologically, revealing species-level taxonomic assignments as well as individual details, such as age, sex, and pathological condition (Boessneck, 1987; Brandon-Jones and Goudsmit, 2022; Goudsmit and Brandon-Jones, 1999; Peters, 2020). Such data are telling, but insufficient for determining fine-scale geographic origins. Recent oxygen and strontium stable isotope evidence suggests that mummified hamadryas baboons were imported from a region encompassing northern Somalia, Eritrea, and Ethiopia (Dominy et al., 2020), a level of geographic precision with limited practical value. Another limitation concerns the captive breeding of some animals. For instance, stable isotopes can reveal a lifetime in Egypt but not the geoprovenance of the source population, as shown for olive baboons from the Ptolemaic catacombs of North Saqqara (Dominy et al., 2020). The analysis of ancient DNA (aDNA) recovered from baboon mummies and compared to the current distribution of baboon genetic diversity has the potential to provide more detailed insights on the geographic origin of baboons in ancient Egypt. To explore this possibility, we sequenced the mitochondrial genome (mitogenome) of a mummified baboon to infer its geographic origin through phylogenetic assignment.
Gabbanat el-Qurud
In
Ottoni et al., 2019 sampled dental calculus from 16 individuals in this same assemblage and reported the preservation of ancient microbial DNA in a subset of six. Their success motivated us to extract DNA from the remaining tooth material of ten individuals (Table 1, Supplementary file 1). In addition, we obtained samples (skin, bone, or tooth) from 21 modern historic specimens of baboons available in museum collections and representing the northeast African distribution of
Table 1.
Information on samples analysed in this study.
Taxon | Origin | Museum ID | Country | Latitude | Longitude | MitoClade | AccNo | Reference |
---|---|---|---|---|---|---|---|---|
| MNHN | MO-1972–357 | ETH | 9.320 | 42.119 | G3-X | OQ538080 | This study |
| SMNS | SMNS-Z-MAM-001034* | ETH | 11.500 | 39.300 | G3-X | OQ538076 | This study |
| MfN | ZMB_Mam_025647_(2) | ETH | 14.164 | 38.891 | G3-X | OQ538079 | This study |
| SMNS | SMNS-Z-MAM-000960 | ERI | 15.783 | 38.453 | G3-X | OQ538078 | This study |
| NHMUK | ZD.1910.10.3.1 | SOM | 9.933 | 45.200 | G3-X | MT279063 | Roos et al., 2021 |
| MfN | ZMB_Mam_012808 | ETH | 9.314 | 42.118 | G3-X | OQ538089 | this study |
| Wild | ETH | 8.968 | 38.571 | G3-X | JX946196 | Zinner et al., 2013 | |
| MfN | ZMB_Mam_042543_(1) | ETH | 9.593 | 41.866 | G3-Z | OQ538084 | this study |
| MfN | ZMB_Mam_074849 | DJI | 11.589 | 43.129 | G3-Z | OQ538085 | this study |
| MNHN | MO-1972–359 | ETH | 6.998 | 40.478 | G3-Z | OQ538086 | this study |
| SMNS | SMNS-Z-MAM-001288 | SDN | 19.110 | 37.327 | G3-Y | OQ538081 | this study |
| Wild | ERI | 15.011 | 38.971 | G3-Y | JX946201 | Zinner et al., 2013 | |
| SMNS | SMNS-Z-MAM-007509† | - | - | - | G3-Y | OQ538082 | this study |
| MHNL | 51000172 | EGY | - | - | G3-Y | OQ538083 | this study |
| SMNS | SMNS-Z-MAM-000584 ‡ | SDN | 13.460 | 33.780 | G3-Y | OQ538075 | this study |
| Wild | TNZ | 7.347 | 37.165 | G1 | JX946199 | Zinner et al., 2013 | |
| MNHN | ZM-MO-1977-5 | SOM | 3.243 | 45.471 | G1 | OQ538088 | this study |
| NHMUK | ZD1929.4.27.2 | COD | 0.800 | 26.633 | J | MT279061 | Roos et al., 2021 |
| NHMUK | ZD1929.4.27.1 | COD | 1.183 | 27.650 | J | MT279062 | Roos et al., 2021 |
| Wild | 19GNM2220916 | TNZ | 4.679 | 29.621 | J | MG787545 | Roos et al., 2018 |
| SMNS | SMNS-Z-MAM-032128 | SSD | 4.281 | 33.555 | J | OQ538087 | this study |
| SMNS | SMNS-Z-MAM-000583 | SDN | 13.333 | 32.729 | J | OQ538077 | this study |
| MfN | ZMB_Mam_074869 | CMR | 5.533 | 12.317 | F | OQ538071 | Kopp et al. in prep |
| Wild | NGA | 7.317 | 11.583 | F | JX946198 | Zinner et al., 2013 | |
| MfN | ZMB_Mam_074887 | CMR | 9.328 | 12.946 | F | OQ538069 | Kopp et al. in prep |
| MfN | ZMB_Mam_074885 | NGA | 7.298 | 10.318 | F | OQ538064 | Kopp et al. in prep |
| MfN | ZMB_Mam_074883 | CMR | 6.334 | 9.961 | F | OQ538072 | Kopp et al. in prep |
| Wild | SEN | 12.883 | 12.767 | E | JX946203 | Zinner et al., 2013 | |
| NHMUK | ZD.1947.586 | SLE | 8.917 | 11.817 | E | MT279064 | Roos et al., 2021 |
| MfN | ZMB_Mam_075043 | TGO | 9.260 | 0.781 | D | OQ538066 | Kopp et al. in prep |
| MfN | ZMB_Mam_011198 | TGO | 6.228 | 1.478 | D | OQ538067 | Kopp et al. in prep |
| Wild | CIV | 8.800 | 3.790 | D | JX946197 | Zinner et al., 2013 | |
| MfN | ZMB_Mam_007396_(1) | TGO | 6.950 | 0.585 | D | OQ538065 | Kopp et al. in prep |
| NHMUK | ZD.1939.1022 | NER | 17.000 | 7.933 | D | MT279065 | Roos et al., 2021 |
| NHMUK | ZD.1939.1020 | NER | 17.683 | 8.483 | D | MT279066 | Roos et al., 2021 |
| MNHN | ZM-MO-1960-476 | TCD | 20.344 | 16.786 | K | MT279067 | Roos et al., 2021 |
| MNHN | MO-1996-2511 | CAF | 3.905 | 17.922 | K | OQ538068 | Kopp et al. in prep |
| NHMUK | ZD.1907.7.8.11 | CAF | 8.000 | 20.000 | K | MT279068 | Roos et al., 2021 |
| MNHN | MO-1996-2510 | CAF | 4.966 | 18.701 | K | OQ538070 | Kopp et al. in prep |
| Wild | ZAF | 24.680 | 30.790 | B | JX946205 | Zinner et al., 2013 | |
| Wild | TNZ | 11.261 | 37.514 | B | JX946200 | Zinner et al., 2013 | |
| ZMB | 12.591 | 30.252 | C | JX946202 | Zinner et al., 2013 | ||
| Wild | 04MNM1300916 | TNZ | 6.119 | 29.730 | H | MT279069 | Roos et al., 2021 |
| Wild | ZAF | 34.456 | 20.407 | A | JX946204 | Zinner et al., 2013 | |
| Wild | 24UNF1150317 | TNZ | 7.815 | 36.895 | MT279060 | Roos et al., 2021 | |
| FJ785426 | Hodgson et al., 2009 |
AccNo, GenBank accession number; NHMUK, Natural History Museum, London; MNHN, Muséum National d'Histoire Naturelle, Paris; MfN, Museum für Naturkunde, Berlin; SMNS, State Museum of Natural History Stuttgart; MdC, Musée des Confluences, Lyon.
*
Mislabelled in museum records as
†
Unclear provenance ‘Somaliland’ (not equal to present-day Somaliland).
‡
Misidentified provenance ‘Abyssinia’ as Ethiopia in museum records.
Results
Mitogenomes from mummified and historic specimens
We discarded seven historic samples and nine mummified samples from our analysis due to insufficient DNA content, sequencing failure, or low coverage and sequencing depth (Supplementary file 1). Thus, our results are based on the newly generated mitogenomes of 14 historic and 1 mummified individual (Table 1). In total, we obtained 896,025,770 raw sequence reads, with a mean of 34,462,530 (± SD 27,945,321) raw sequence reads per sample. On average, 95.5% of reads survived trimming and a median of 9934 (range: 244–2,722,354) reads per sample mapped to the reference mitogenome. After removal of duplicates (duplication level median: 25.1%; range: 2.5–92.6%), a median of 7398 (range: 237–497,458) mapped reads per sample resulted in the median final sequencing depth of 26× (range: 0.21–2952×). After exclusion of samples with low quality, the final dataset had a median final sequencing depth of 37× (range: 16–2952×), with a median of 0.4% undetermined sites (range: 0–1.7%) and a median breadth of coverage of at least 3× of 99.3% (range: 97.4–100%) (Supplementary file 1). All these metrics differed considerably depending on sample age (historic versus mummified) and DNA concentration (Figures 3 and 4). Capture enrichment strongly increased the number of mapped reads and final mean coverage compared to the shotgun approach (Figures 3 and 4). GC content of sequences was 40–50% (Figure 5) in the same range as the reference genomes.
Figure 3.
Comparison of DNA concentration and amount of distinct mapped reads.
A higher DNA concentration produces a higher number of distinct mapped reads. Capture enrichment additionally increases the number of distinct mapped reads. Circles and triangles depict the different sequencing approaches, enrichment, and shotgun, respectively; size is related to the final coverage of the mitogenome; colours represent the different sample types and sequencing approaches (yellow: shotgun sequencing of the mummified sample, MHNL 51000172; blue: shotgun sequencing of historic sample; purple: capture enrichment of historic sample; green: capture enrichment of mummy sample).
Figure 4.
Overview of sequencing success for museum and mummy specimens.
Mean (± SD) final coverage of the mitogenome is shown for each sample (with abbreviated museum ID). Circles and triangles depict the different sequencing approaches, enrichment and shotgun, respectively; colours represent the different sample types and sequencing approaches (yellow: shotgun sequencing of mummy sample; blue: shotgun sequencing of historic sample; purple: capture enrichment of historic sample; green: capture enrichment of mummy sample). Dashed line shows the cut-off limit 10× for mean final coverage; samples below were excluded from final analyses.
Figure 5.
Distribution of GC content in historic samples and mummified samples.
The sequencing reads of the mummified sample (MHNL51000172) exhibit C to T and G to A misincorporations at 5′ and 3′ ends, reaching frequencies of 3.3 and 1.6% at the first/last position of the read (Figure 6). Mapped reads of the mummified sample agreed to median of 99.2% (IQR 1.6%) when focussing on the 125 sites that exhibited fixed differences between subclades and differed at three sites from the variant found in its subclade (Figure 7a). When focussing on the 37 sites that are fixed in the subclade of attribution of the mummified baboon but differed in its consensus sequence, mapped reads agreed to a median of 97.3% (IQR 3.1%) (Figure 7b).
Figure 6.
DNA damage plot for the sample of the mummified baboon MHNL 51000172 from 5′ and 3′ read ends, showing mean frequencies of C to T substitutions (red), G to A substitutions (blue), deletions (grey), and insertions (yellow) over the first/last 25 positions.
Figure 7.
Barplots showing the bases of mapped reads for the sample of the mummified baboon MHNL 51000172 at sites that (a) exhibit fixed differences among northeastern subclades and (b) are fixed in subclade G3-Y but differ in the consensus sequence of the mummified baboon.
Sites are named according to their position and the base in the G3-Y consensus sequence and coloured by base. Bases are colour-coded (A: red; C: blue; G: yellow, T: green).
Phylogenetic mapping
Phylogenetic trees inferred from maximum likelihood (ML) and Bayesian inference (BI) revealed identical topologies with generally strong node support (100% bootstrap support [BS] and posterior probability [PP] 1.0) and clearly defined geographic clades (Figure 8, Figure 8—figure supplement 1). These mitochondrial clades did not directly mirror species assignments. Within the northeastern baboons, the central olive baboon clade J from Democratic Republic of the Congo, Tanzania, South Sudan, and southern Sudan diverged first, followed by northern yellow baboons of clade G1 including a sample from Somalia. Hamadryas baboons formed clade G3, which also included olive baboons from the region. Clade G3 contained three subclades: subclade G3-Z comprised hamadryas baboons from Ethiopia and Djibouti; subclade G3-X comprised hamadryas and olive baboons from Ethiopia, Eritrea, and Somalia; and subclade G3-Y comprised hamadryas and olive baboons from northeastern Sudan and Eritrea. The mummified baboon from Gabbanat el-Qurud (MHNL 51000172) was located in subclade G3-Y, closely related to samples from Eritrea and northeastern Sudan.
Figure 8.
Phylogeny of baboons based on complete mitochondrial genomes as inferred from maximum likelihood analysis.
Figure 8—figure supplement 1.
Phylogeny of baboons based on complete mitochondrial genomes under Bayesian inference.
Outgroups (
The median-joining haplotype networks differentiated samples within clade G3 in greater detail and in a more precise geographic context (Figure 9, Figure 9—figure supplement 1). They revealed the same three subclades within the G3 clade. The HVRI and the cyt
Figure 9.
Median-joining haplotype network of northeastern baboons based on 644 HVRI sequences (176 bp).
The analysed baboon mummy sample resolves in clade G3-Y (depicted in red, black arrow). Circle colour reflects species and country of origin (‘Arabia’' comprises samples from Yemen and Saudi Arabia, ‘Strait’ comprises samples from near the Bab-el-Mandab Strait, i.e. southern Eritrea, Djibouti, northern Somalia).
Figure 9—figure supplement 1.
Median-joining haplotype network of northeastern baboons based on 137 cyt
The analysed baboon mummy sample resolves in clade G3-Y (depicted in red, arrow). Circle colour reflects species and country of origin (‘Arabia’ comprises samples from Yemen and Saudi Arabia, ‘strait’ comprises samples from near the Bab-el-Mandab Strait, i.e. southern Eritrea, Djibouti, northern Somalia).
Discussion
We succeeded in sequencing the mitogenomes of 14 historic baboons from northeastern Africa and a mummified baboon recovered from Gabbanat el-Qurud, presenting the first genetic data of a mummified baboon from ancient Egypt to date. DNA of the mummified baboon shows
Our phylogenetic analysis of the newly generated mitogenomes in combination with published mitochondrial sequence data produced tree topologies in agreement with those of prior studies, with three well-supported clades across the northeastern distribution of
A mummified hamadryas baboon from Gabbanat el-Qurud (MHNL 51000172) yielded sufficient aDNA to produce a complete mitogenome, which fell unequivocally in subclade G3-Y (cf. Kopp et al., 2014b). Haplotype networks allowed us to further refine subclade G3-Y, which consists of
Yet, this baboon predates the reign of Ptolemy I by centuries, presuming it is contemporaneous with another baboon (MHNL 90001206) in the same assemblage, ca. 800–540 BCE (Richardin et al., 2017). Thus, our findings raise the possibility that Adulis already existed as a trading centre or entrepôt during the 25th and 26th dynasties of Egypt. Although speculative, and expressed with due caution, our reasoning would extend the antiquity of Egyptian–Adulite trade by as much as five centuries.
Arguing for pre-Ptolemaic contact between Egypt and Adulis is fraught in the absence of corroborating material evidence—but even so, the archaeological record is not entirely silent on the prospect. Manzo, 2010 and others (Zazzaro et al., 2014) reassessed the ceramic tradition at Adulis and developed a chronology that stretches to the early second millennium BCE, the deepest levels of which contained a fragment of blue glass with yellow inlays similar to Egyptian glass from the New Kingdom (Fattovich, 2018). In Egypt, contact with the Eritrean lowlands is attested by trade goods dating to ca. 1800–1650 BCE or earlier, including potsherds, obsidian, and fragments of carbonized ebony (Fattovich, 2018; Lucarini et al., 2020). Discovered at Mersa Gawasis, a Middle Kingdom harbour, these artefacts appear to align the prehistory of Adulis with the fabled Land of Punt (Bard and Fattovich, 2018; Manzo, 2010; Manzo, 2012), an enigmatic toponym scattered across scant and disconnected records (Cooper, 2020).
Punt existed in a region south and east of Egypt, and was accessible by land or sea. For Egyptians, Punt was a source of ‘marvels,’ particularly incense, but also baboons, that drove bidirectional trade for 1300 y (ca. 2500–1170 BCE) (Tallet, 2013). Some scholars have described this enterprise as the beginning of economic globalization (Fattovich, 2012), whereas others view it as the earliest maritime leg of the spice route (Keay, 2006), a trade network that would shape geopolitical fortunes for millennia. The global historical importance of Punt is therefore considerable, but there is a problem—its location is uncertain, in part because the toponym fades from view. From the early first millennium BCE, there are no further records of Egyptians in Punt or of Puntites visiting Egypt. There are, however, two incomplete inscriptions that mention Punt in a narrative context, and both are attributed to the 26th (Saite) Dynasty (Betrò, 1996; Cavasin, 2019). One of these, the Defenneh stele, describes an expedition to Punt that was saved from dying thirst by unexpected rainfall on ‘the mountains of Punt’ (Meeks, 2003). The Defenneh stele is a testament to the efforts of Saitic pharaohs to revive maritime commerce on the Red Sea (Lloyd, 1977), while also raising the possibility of renewed trade with Punt. It is perhaps no coincidence that the Saite dynasty (664–525 BCE) exists squarely within the radiometric date range of hamadryas baboons from Gabbanat el-Qurud.
Punt, like Adulis, was a source of baboons for Egyptians, a history that raises the possibility of using baboons as a tool for testing geographic hypotheses. Recently, Dominy et al., 2020 used stable isotope mapping methods to determine the geoprovenance of mummified baboons from Thebes (modern-day Luxor) and dated to the (late) New Kingdom. Their results pointed to present-day Ethiopia, Eritrea, or Djibouti, as well as portions of Somalia, an area that corroborates most scholarly views on the location of Punt (Breyer, 2016; Kitchen, 2004), but see Meeks, 2002; Meeks, 2003; Tallet, 2013. Here, we used aDNA to show that at least one baboon from the 25th Dynasty or Late Period of Egyptian history—a span that coincides with the last known expeditions to Punt, but predates Greco-Roman accounts of Adulis as a source of baboons—can be traced to Eritrea. Thus, our findings appear to establish primatological continuity between Punt and Adulis. Such a conclusion must be viewed with caution, but it bolsters recurrent conjecture among some historical archaeologists: that Punt and Adulis were essentially the same trading centre from different eras of Egyptian antiquity (Doresse, 1959; Fattovich, 2018; Kitchen, 2004; Massa, 2021; Phillips, 1997; Sleeswyk, 1983).
At minimum, our results reinforce the view that ancient Egyptian mariners travelled great distances to acquire living baboons. A great strength of this conclusion is that it is based on distinct but complementary methods, but of course, the sample size is paltry and limited to
Future directions
Direct radiocarbon dating of MHNL 51000172 and other baboons from Gabbanat el-Qurud is an urgent priority, in part because doing so would put these specimens into conversation with those from the catacombs of Tuna el-Gebel. The oldest gallery at Tuna el-Gebel, Gallery D, is dated to the 26th Dynasty and contains a single species of baboon:
Materials and methods
DNA extraction and sequencing
DNA damage and degradation is expected from ancient (mummified) and nineteenth/early twentieth-century museum specimens. We therefore analysed mitochondrial DNA (mtDNA), which is available in higher copy numbers than nuclear DNA and holds greater potential for success when sample quality is poor. We analysed complete mitogenomes because they are effective for reconstructing robust mitochondrial phylogenies of modern baboons and have proven to indicate the geographic origin of the corresponding sample reliably (Roos et al., 2021; Zinner et al., 2013). Recent advances in sequencing technologies allow the successful sequencing of mitogenomes either with a shotgun sequencing approach or, for samples with very low DNA quality and quantity, with a capture enrichment approach (Schuenemann et al., 2017; Shapiro and Hofreiter, 2012).
We extracted DNA with a specific column-based method aimed at the recovery of short DNA fragments following established protocols and necessary precautions for the analysis of aDNA (Dabney et al., 2013a; Rohland et al., 2004; Roos et al., 2021). In particular, samples from mummified specimens were extracted separately and in a dedicated aDNA laboratory to prevent cross-contamination. Concentration of DNA extracts was measured on a Qubit fluorometer (Life Technologies, Singapore) and quality checked on a Bioanalyzer (Agilent, Santa Clara, USA) or Tapestation 2200 (Agilent). All samples were initially sequenced with a shotgun approach. Samples with DNA extract concentrations below 4.5 ng/μl or final mitogenome sequencing depth below 10×, and with enough remaining DNA extract, were enriched for mtDNA with a capture approach.
For the shotgun approach, sequencing libraries were prepared with the NEBNext Ultra II DNA Library Prep Kit (New England BioLabs, Frankfurt, Germany) according to the manufacturer’s instructions without prior fragmentation. Library concentration and quality were assessed with the Qubit Fluorometer and Bioanalyzer and molarity was estimated via qPCR with the NEBNext Library Quant Kit (New England BioLabs). Libraries were single indexed with NEBNext Multiplex Oligos (New England BioLabs) with 5–11 PCR cycles and cleaned up with the kit’s beads.
For the capture enrichment approach, RNA baits (myBaits custom Kit, Arbor Biosciences, Ann Arbor, USA) were designed for the mitogenome of
Sequencing was performed with 24 libraries per lane (23 samples + pooled negative control to monitor contamination) on an Illumina HiSeq4000 (50 bp, single-end read) at the NGS Integrative Genomics core unit of the University Medical Center Göttingen, Göttingen, Germany, or on a NovaSeq6000 SP flow cell (100 bp, paired-end read) at the Max Planck Institute for Molecular Genetics, Berlin, Germany. Capture enrichment libraries were reloaded and sequenced a second time to increase the number of reads.
Mitogenome assembly
Raw sequencing reads were demultiplexed and adapters trimmed at the sequencing facilities. We performed subsequent sequence processing on the central high-performance computing cluster bwForCluster BinAC. We checked read quality with FastQC 0.11.8 (Andrews, 2010), trimmed and filtered reads with Trimmomatic 0.39 (Bolger et al., 2014) using the settings ILLUMINACLIP:TruSeq3-PE.fa:2:30:10 MINLEN:30 SLIDINGWINDOW:4:20 LEADING:20 TRAILING:20, AVGQUAL:30, and confirmed adequate quality of trimmed reads again with FastQC. Reads were mapped with Burrows Wheeler Aligner (BWA) backtrack 0.7.17 (Li and Durbin, 2009) using default settings independently to each of the seven different mitogenomes of representatives of the northern baboon clades (
We augmented our dataset with published mitogenomes of baboons (Roos et al., 2021) and
For a more fine-scale geographic representation, we further included published sequence data from the northeastern part of the baboon distribution of two different mitochondrial markers with differing resolution: the cytochrome
We assessed contamination by checking mismatches of the mapped reads from the mummified sample at sites in the mitogenome that (i) are distinct between northeastern subclades (125 fixed differences) and (ii) are fixed in subclade G3-Y (considering all samples but the mummified baboon) but differ in the consensus sequence of the mummified sample (37 sites).
Phylogenetic reconstruction
To identify the phylogenetic affiliation of the newly investigated samples, we reconstructed phylogenetic trees based on the final dataset of 46 mitogenomes (alignment length: 16,628 bp) using ML and BI methods with W-IQ-Tree 1.6.12 (Nguyen et al., 2015; Trifinopoulos et al., 2016) and MrBayes 3.2.7 (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003), respectively. We treated the mitogenome as a single partition, the optimal substitution model for phylogenetic reconstructions was detected to be TN + F + I + G4 (Tamura and Nei, 1993) under the Bayesian information criterion and GTR + F + I + G4 (Tavaré, 1986) under the Corrected Akaike Information Criterion with Modelfinder (Kalyaanamoorthy et al., 2017) as implemented in W-IQ-Tree. The ML tree was reconstructed with 10,000 ultrafast bootstrap replications (Hoang et al., 2018) applying the TN + F + I + G4 model. The BI tree was reconstructed applying the GTR + I + G model and using four independent Markov chain Monte Carlo runs with 1 million generations, a burn-in of 25%, and sampling every 100 generations. To ensure convergence, the Potential Scale Reduction Factor was checked to be close to 1 for all parameters. We visualized phylogenetic trees with the R package ggtree 3.4.2 (Yu et al., 2017) and adopted clade nomination of Roos et al., 2021 and Kopp et al., 2014b.
Haplotype networks
To determine the mitochondrial clade of origin of the analysed samples more precisely, we reconstructed median-joining haplotype networks (Bandelt et al., 1999) with Popart 1.7 (Leigh and Bryant, 2015) for both the HVRI (n = 644, 176 bp) and the cyt
Geographic maps
Geographic maps were created in R. We obtained species distribution shapefiles from IUCN (Gippoliti, 2019; Sithaldeen, 2019; Wallis, 2020a; Wallis, 2020b; Wallis et al., 2020; Wallis et al., 2021), river, lake and coastlines from Natural Earth (https://www.naturalearthdata.com) via rnaturalearth 0.1.0 (Massicotte and South, 2023).
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2023, Grathwol et al. This work is published under https://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Adulis, located on the Red Sea coast in present-day Eritrea, was a bustling trading centre between the first and seventh centuries CE. Several classical geographers—Agatharchides of Cnidus, Pliny the Elder, Strabo—noted the value of Adulis to Greco-Roman Egypt, particularly as an emporium for living animals, including baboons (
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer