2.1 Ethics statement

Field observations, community engagement, and the acquisition and export of specimens were conducted under the authorization of the Congolese Institute for Nature Conservation (ICCN), the DRC government agency responsible for protected areas and with jurisdiction over their wildlife. ICCN operations are guided by the Congolese Law 08/08 of 2008 concerning Forest Classification, the National Strategy for the Conservation of Biodiversity in the Protected Areas of DRC (2012), and the Law 14/003 relative to Nature Conservation of 2014. These require ICCN to assure adequate outreach and dialogue with communities whose lands border protected areas. Information provided by residents of these settlements concerning the new Colobus monkey, C. congoensis (Likweli), was collected fortuitously or secondarily during village gatherings to raise awareness about the park. The gatherings were authorized by the ICCN, but organized with local administrative and traditional authorities. Typically, during a village visit by representatives of the ICCN, a chief or elder would call a village meeting where park rules as well as rights and responsibilities of the community members relative to the park and buffer zone were discussed. Villager attendance was not required, people were free to come and go, and could volunteer information or not. In these sessions, ICCN was particularly concerned with the communities’ perception of the buffer zone in which they lived, the composition of its fauna and how communities exploited it or were impacted by it. In this context, ICCN would ask questions about rare species including the presumed new Colobus species. For instance, the information regarding the rare occurrence of Likweli in the eastern portion of its range was acquired during community sessions explaining the park limits and the extent of the buffer zone before mobilizing community members for the participatory demarcation process in the Mituku sector of the buffer zone.Field work associated with the discovery of this species was conducted under partnership contracts between ICCN and the Lukuru Foundation (2011 – 2019) and between ICCN and the Frankfurt Zoological Society (FZS) from 2019 to present. The Lukuru Foundation led explorations of the Tshuapa-Lomami-Lualaba interfluve region (TL2) and established Lomami National Park in collaboration with ICCN, while FZS supports the ongoing co-management and protection of the park and buffer zone in partnership with ICCN.

Specimens of deceased individuals of the new colobine species were confiscated from hunters during ICCN patrols in the national park in April 2021. We acquired samples from C. angolensis killed in the buffer zone and destined for illegal bushmeat trade. Confiscations followed the ICCN Code of Conduct for park guards. Acquisition of the specimens for scientific description was authorized by the ICCN. Specimens were exported under CITES authorization and imported into the United States of America under permits by the Centers for Disease Control and Prevention and US Fish and Wildlife Services. Non-exported specimens are housed at the University of Kisangani’s Centre de Surveillance de la Biodiversité (CSB) in accordance with CSB procedures. Non-invasive field observations were made in Lomami National Park by field teams composed of project scientific staff, ICCN eco-guards, and local community monitors. Additional information on the species, including its distribution and hunting threats, was provided by residents in communities living within the buffer zone who were interviewed in the context of community conservation projects.

2.2 Local knowledge

We evaluated local knowledge on the occurrence and ecology of the new species by conducting interviews with residents and hunters familiar with the area comprising its range and vicinity. Interviews were conducted in 52 villages in the Sectors of the Mituku Basikate, Mituku Bamoya, and Walengola-Babila in Tshopo Province and in the Balanga Sector in Maniema Province. We asked interviewed individuals to describe primates that occurred in their area of activity and to provide vernacular names. For likely observations of the new Colobus species, we asked informants to provide detailed descriptions of the monkey, information on behavior and ecology, locations where they encountered them, and if they were a target for hunting. We accepted observations of the new species if the reports included descriptions of diagnostic pelage and facial features. When specific locations for sightings were identified, and when logistics permitted, we invited informants to accompany field teams to return to the site to search for the animals.

2.3. Inclusivity in global research

Additional information regarding the ethical, cultural, and scientific considerations specific to inclusivity in global research is included in the Supporting Information (S1 Text).

2.4. Morphology

2.4.3 Craniodental and skeletal measures and analysis.

Skeletal specimens of 13 extant African colobine species were sampled for comparative cranial and dental analysis (n = 182; Table 1), following the taxonomy used by the IUCN [1]. Five species of Colobus were sampled (with multiple subspecies of C. guereza and C. angolensis), excluding the recently elevated and skeletally rare Mt. Kilimanjaro Guereza, Colobus caudatus (a close relative and former subspecies of C. guereza [15]). Seven species of Piliocolobus were also sampled, as well as the monotypic Procolobus verus. Wild specimens were included whenever possible, but a few non-pathological captive specimens were included to increase the sample size for Piliocolobus kirkii. Only adult specimens were measured for craniomandibular metrics, as defined by full M³/₃ eruption. Dental metrics on erupted adult teeth in juvenile specimens were included to increase sample sizes of C. vellerosus and C. satanas. A complete list of sampled specimens, including subspecific identifications and calculated indices for each specimen, is provided in Table A in S1 File.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Comparative skeletal sample used in morphometric comparisons.

https://doi.org/10.1371/journal.pone.0349857.t001

Nine skeletal indices were calculated to compare the new colobine with extant African colobines, capturing aspects of size and shape in the cranium, mandible, and dentition (Table 2). To control for possible allometric scaling effects, Mosimann shape indices of the craniodental features were tested against the geometric mean of 17 linear cranial measurements as a proxy for specimen size [30–32]. Indices were considered allometrically influenced if they had a significant and strong correlation with size (cranial geometric mean) indicated by a Pearson’s correlation coefficient |r| > 0.5 [32–34]. Absence of a strong and significant correlation between size and shape suggests isometry of the shape variable within the sample. Indices that scale allometrically are figured as bivariate scatter plots to compare specimens at a similar size. Non-allometric (isometric) indices are presented as boxplots, and statistically significant differences were assessed among genera and species with ANOVAs and Tukey’s Honestly Significant Difference (HSD) post-hoc pairwise comparisons.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Description of craniomandibular measurements and indices used in morphometric comparisons.

https://doi.org/10.1371/journal.pone.0349857.t002

Postcranial morphology will be described in a separate, functional analysis. For the current descriptions, we provide craniodental measurements and basic limb indices (i.e., intermembral index) in the supplementary information (Table B in S1 File).

2.5 Genetics

DNA was extracted from tissue samples of three individuals of the new Colobus species and two C. angolensis angolensis individuals using the DNeasy Blood & Tissue Kit (Qiagen 69504; Germantown, MD) following manufacturer's protocols. A 4,090 base pair (bp) region of the mitochondrial genome was amplified, including the NADH3, NADH4L, NADH4, and NADH5 genes [35]. This marker has been used in previous studies to investigate mitochondrial phylogenies of African and Asian colobines [35,36] and allows for comparative analysis between the new species and all recognized Colobus species. We used PCR primers from Ting [35] and designed additional sequencing primers (available upon request) based on known African colobine templates. Sanger sequencing was performed by the Molecular Cloning Laboratories (San Francisco, California), and sequence chromatograms were inspected by eye and assembled using Geneious R11 (v11.0.5; https://www.geneious.com). Additionally, we obtained orthologous sequences from Asian and African colobine species via the National Center for Biotechnology Information GenBank database (S1 Table). Novel sequences generated in this study have been archived in GenBank (S1 Table; https://www.ncbi.nlm.nih.gov/nuccore/?term=Likweli). Only unique haplotypes were included in the final alignment.

To infer a phylogeny, we assembled and annotated the mitochondrial regions of all the samples with MitoFinder/1.4.1 [37], using a published Piliocolobus badius mitochondrial genome as reference (NC_008219) [38]. Each mtDNA region was then divided into 13 partitions: 1st, 2nd, and 3rd codon positions of four protein-coding genes and a concatenated sequence of four tRNA, which were individually aligned using MAFFT/7.407 [39]. The best evolutionary model was evaluated using jModelTest [40]. A maximum likelihood tree was constructed in RaxML [41], using the GTR + F model of substitutions and performing 1,000 rapid bootstraps.

Divergence dates were estimated using a Bayesian phylogenetic analysis conducted with MCMCTree, as implemented in PAML v4.9j [42]. We used a single sample per species or subspecies for all monophyletic groups and included one representative per clade in cases of paraphyly. We applied a fossil-calibrated prior to the Colobina–Presbytina split based upon de Vries and Beck [43], using a uniform distribution with a hard minimum bound at 8.125 million years ago (Ma) and a soft maximum bound at 15 Ma [see also 44]. To facilitate comparison with previously published divergence estimates for the Colobus clade, we also performed analyses using a secondary calibration for the African-Asian colobine split, constraining this node to 9.42–12.28 Ma based on the divergence estimate reported by Perelman et al. [45] and subsequently used as a calibration point in Roos and Zinner [36]. For all MCMCTree analyses, we employed the correlated-rates clock model, discarded the first 10,000 iterations as burn-in, and sampled every 50 iterations until a total of 20,000 post-burn-in samples were obtained.

2.6 Vocalizations

We recorded six roaring events from the new colobine species in Lomami National Park during 2020 and 2021 using a Canon digital camera (n = 4), Zoom H5 4-input digital recorder with a Sennheiser MKE 600 shotgun microphone (n = 1), and a cell phone (n = 1). It was not possible to distinguish the caller’s sex. However, roaring is more common in adult males in other Colobus species [46]. Recordings were collected at locations spaced 3–15 km apart, likely from individuals in different social groups. We could hear more than one individual calling in some recordings. Three recorded events produced vocalizations that were suitable for bioacoustic analysis. We compared recordings of the new Colobus species to C. satanas vocalizations from Lope National Park, Gabon and from Bioko Island, Equatorial Guinea. Structural components of the new Colobus species and C. satanas roars were also compared with published roars of other Colobus species [46–49].

We uploaded recordings into Audacity 3.7 software (open source, Muse Group) and applied a noise reduction filter (12 decibels). The filtered recordings were analyzed with Raven Pro software (Cornell Laboratory of Ornithology, Ithaca, New York) using methods adapted from Schel et al. [49], Charif et al. [50], and Fuller [51]. To characterize the new Colobus roars and compare them to other Colobus roars, we measured basic call features including pulse rate (pulses/second), peak fundamental frequency (Hz), and presence or absence of frequency modulation. For a more detailed comparison between high-quality recordings of the new Colobus and C. satanas roars specifically, we analyzed 23 temporal and acoustic call features (Table A in S2 File) [51].

We generated spectrograms using a 512-point fast-Fourier transform and a Hann window function to measure the morphology of the vocalizations. We generated a power-time oscillogram to determine temporal limits and a power frequency spectrum using a 1024-point fast-Fourier transform and a Hann window to determine frequency measurements and frequency limits. The noise floor in the power frequency spectra was determined for each file using a region of the audio with no signal and only background noise. We used summary statistics and principal component analysis (PCA) to compare roar phrases between the new Colobus and C. satanas.

2.7 Field surveys

2.7.2 Habitat description and use.

Habitat classification of encounters with the new Colobus species was determined by superimposing observation locations on a map of topographic and substrate classes in the park and buffer zone with the addition of the forest type as determined in the field at the location of the observation. Topographic and substrate classification of the Lomami National Park and buffer zone is based on two inputs: 1) a land cover classification of 2010 Landsat imagery, and 2) geo-referenced field data on topography, elevation, seasonal inundation, and predominant soil type (clay or sand). Topographic descriptions follow Guillocheau et al. [52] for Congo Basin Neogene landforms and Bwangoy et al. [53] for seasonally inundated sites. Habitat classes identified in the range area are described in Table 3.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Habitat classes within the range of the new Colobus species.

https://doi.org/10.1371/journal.pone.0349857.t003

2.8. Nomenclatural acts

The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix “http://zoobank.org/”. The LSID for this publication is: urn:lsid:zoobank.org:pub:AAB9196D-DC31–4172-A218-CE924AE3BC65. The electronic edition of this work was published in a journal with an ISSN, and has been archived and is available from the following digital repositories: PubMed Central, LOCKSS.

3.1 Discovery and local knowledge

The first evidence of the new species came in 2008 during exploration of the Lomami Basin by members of the Lukuru Foundation, in an area that later became Lomami National Park. Ashley Vosper and Bernard Ikembelo photographed an unidentified monkey in a high forest canopy on the east bank of the Lomami River in the Courbure Sector, named for six large meanders of the Lomami River. Although the photograph revealed only part of the animal, the field team strongly suspected that this was not any of the monkey species they already knew from the region. Following the initial sighting, nothing more was reported until November 2018, when Jean Pierre Kapale led a surveillance patrol in the Courbure Sector and photographed a black monkey with pale markings around its mouth and a white perianal patch. These characters were unlike those of any known primate species in the area. Over the next 10 months, Kapale and his team, using surveillance patrols and directed searches accompanied by local field assistants, found and photographically documented the monkey seven more times in different locations. Comparisons of the photos taken in 2008 and in 2018–2019 showed that the monkey Kapale encountered was the same species as the previously unknown monkey seen a decade earlier. Following Kapale’s reports, we conducted a review of photographs documenting primate sightings in earlier surveillance patrols by other teams. We found that the same monkey had been seen and photographed in August 2018 in an area 35 km north of Kapale’s discoveries, but that the monkey had been misidentified in the patrol report and never announced at that time.

The newly discovered Colobus species is not well known by local communities, even within the vicinity of its range. This contrasts with the extensive knowledge of other primates in these communities. Residents in only eight out of the 52 surveyed localities bordering the new species’ range reported knowledge of the primate and could accurately describe it. In one interview, a hunter who had once opportunistically killed the monkey described it but could not give it a local name. Another hunter described the species’ strong odor. Eventually, the people of the Balanga ethnic group living in the buffer zone west of the Lomami River and bordering the species’ range gave Jean Pierre Kapale the vernacular name, Likweli. The Mituku local communities, who occupy the eastern limits of the range, referred to the species as kasaba nkoni, which means “the branch shaker.” The species was described as quiet and cryptic with a localized distribution. Several informants stated that its calls, reported mostly to be heard at dawn, resembled the roars of C. angolensis.

3.2 Diagnosis and description

3.2.1 Systematics.

Family Cercopithecidae Gray, 1821

Subfamily Colobinae Blyth, 1863

Subtribe Colobina Blyth, 1863

Genus Colobus Illiger, 1811

Colobus congoensis J. Hart, Amboko, Arenson, Horton, Coates, Kapale, Koko, T. Hart, Gilbert, Sargis, and Detwiler, sp. nov. urn:lsid:zoobank.org:act:8F7FB2EC-5E42-4517-9D8C-634A2F7CBD5D

Holotype: Yale Peabody Museum, YPM MAM 17307 (Field Catalog JH30), adult male, skin and skeleton.

Paratypes: Yale Peabody Museum, YPM MAM 17306 (Field Catalog JH28) and YPM MAM 17308 (Field Catalog JH31), two adult females, skins and skeletons.

Type Locality: Lomami National Park, Democratic Republic of Congo, Tshopo Province, Biondo Sector, east bank of the Lomami River, S1.70190°, E25.13970°.

Etymology: The nominal congoensis refers to the species range limited to the Democratic Republic of Congo. Likweli, the vernacular name given to the species by communities bordering its range, is the recommended common name.

Diagnosis and Description: A small, long-tailed Colobus monkey that exhibits little sexual dimorphism in overall pelage coloration. The pelage of the limbs and torso is black. Hair length is longest on the upper shoulders and back (Color 89 in [54]), producing a sheen in some lighting and a cape-like pelt in some individuals. Body hairs are relatively short (~9–12.5 cm long in the male; ~ 6–8 cm long in females at mid-dorsum). The tail is largely black. Older individuals may show gray hairs (Glaucous Colors 79–80 in [54]) along the caudal dorsum and the tail, as seen in YPM MAM 17308, which also exhibits heavily worn dentition in the associated cranium, indicating an advanced age for this individual (Text A and Fig B in S4 File). The single male examined had a small, black terminal tail tuft in the fresh specimen, not seen on the two females. The head in both sexes is framed by long black hairs (~4–6 cm) on the forehead and sides of the face. The dark glabrous face is marked in both males and females by a prominent patch of bare skin that is pinkish to orange-cream in color, surrounding the mouth and extending from the upper lip and philtrum to the inferior portions of the nasal alae. The extent of this pale patch is variable, in some individuals covering much of the lower face below the nares and in others restricted to the circumoral region and philtrum. The bare skin over the pronounced zygomatic arches and temporal region is dark gray, contrasting with the black skin on eyelids and around the eyes, and giving the face a mask-like appearance (Fig 2). Ear pinnae are large and black with crenulated and folded margins. Both sexes have a white perianal patch (9 × 5 cm in the male; 12 × 6 cm in the female) between pale orange callosities. The patch is largely glabrous in the females whereas it is sparsely covered with fine white hairs in the male (Fig 3, S2 Table). The scrotum and penile sheath are dark gray, and the penis is pinkish white. Field measures are given in S3 Table.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Colobus congoensis and C. satanas in life.

(A & B) C. congoensis, Lomami National Park, Democratic Republic of Congo. Photo credits: (A) Daniel Rosengren, (B) Bravo Bofenda. (C & D) C. satanas, Lope National Park, Gabon. Photo credits: (C) Martin Royele/Royele Safaris, (D) Barna Takats. Photos used with permission.

https://doi.org/10.1371/journal.pone.0349857.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Colobus congoensis perianal patch.

(A) Adult male (YPM MAM 17307). (B) Adult female (YPM MAM 17308).

https://doi.org/10.1371/journal.pone.0349857.g003

The cranium of C. congoensis is airorhynchous with a narrow and elongated basicranium (Fig 4; Figs A-C in S4 File). The interorbital pillar is broad, and in profile forms a continuous curve with a nearly vertically oriented frontal squama. The anterior portion of the neurocranium displays modest frontal “bulging” or “saddle-shape” in lateral view. The neurocranium is globular with widely spaced temporal lines originating from the lateral margin of the supraorbital torus. The male individual YPM MAM 17307 has slightly raised temporal lines and a moderately developed nuchal crest; females (YPM MAM 17306 and YPM MAM 17308) are more gracile (see Text A in S4 File and Figs A-C in S4 File for further descriptions and photographs of individual specimens). The rostrum is highly abbreviated with moderate subnasal prognathism, and the malar is shallow. The orbits are bounded superiorly by a thin torus restricted to the lateral ~3/4 of the orbit. The mandibular corpus is shallow and deepens slightly posteriorly. The gonion is clearly expanded in the male individual but not greatly expanded in the female individuals. The symphysis is sloping, unmarked by mental ridges or a median mental foramen. The corpus lacks lateral buttressing in the form of prominentia laterales. The ramus is tall and vertically to slightly posteriorly oriented, with a shallow coronoid notch.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Holotype of C. congoensis compared with other African colobines.

Left to right: C. congoensis holotype male skull (YPM MAM 17307), Colobus satanas male skull (AMNH M-236358), Colobus angolensis male skull (AMNH M-52189), and Piliocolobus oustaleti male skull (AMNH M-52340). Top row: anterior view of cranium; middle row: right lateral view of cranium; bottom row: right lateral view of mandible.

https://doi.org/10.1371/journal.pone.0349857.g004

Dentally, the molars are typically colobine, characterized by tall cusps, low crowns, and minimal basal flare with distinct mesiobuccal clefts in the lower molars. No accessory cuspules are observed on the molars. The upper P³ lacks a protocone, as is typical of African colobine monkeys, but the P⁴ is relatively small compared to that of other species of Colobus. Although only one male specimen is known, male upper canine size is relatively small, similar to that of C. satanas males (Fig 4; Fig 5E). The female canines are also relatively small and not masculinized, unlike the female canines of C. guereza. Overall canine dimorphism, however, is within the range of other Colobus species (Table A in S1 File). The lower P₄ metaconid is subequal to or slightly taller than the protoconid. The upper incisors are small and typically colobine, unlike the enlarged incisors of C. satanas. The lower incisors have enamel on the lingual aspect and a distinct prong on the distal margin of the lateral incisors.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Craniomandibular indices among African colobines and within Colobus.

(A) In cranial flexion (airorhynchy), C. congoensis is more similar to Colobus than Piliocolobus of similar size. (B) The petrous pyramid is shortened in C. congoensis, more similar to that of Colobus or Procolobus than to the elongated petrous of Piliocolobus. (C) In palate shape, C. satanas and C. congoensis share a relatively short and broad palate. (D) A relatively shallow mandibular corpus differentiates C. congoensis and C. satanas from other species of Colobus. (E) Male relative canine size is small in C. satanas and the single male specimen of C. congoensis. Skull indices are described in Table 2.

https://doi.org/10.1371/journal.pone.0349857.g005

3.2.3 Comparisons to other Colobus species, with special reference to C. satanas.

In life, C. congoensis differs from all other members of its genus by its striking piebald facial pattern, extensive white perianal patch, smaller body mass, and small cranium. Measurements of the available specimens suggest limited body mass sexual dimorphism in adults, although the available data on body weight includes a young adult male and two gravid females, likely skewing any body mass dimorphism ratio (Fig 6A & 6B; Table B in S1 File; S3 Table). Although additional body mass data will clarify the range of sexual size dimorphism, the available data indicate that C. congoensis is smaller in overall body mass than other Colobus species but exhibits moderate to low body mass and canine dimorphism within the range of other Colobus species, most similar to C. satanas (Fig 6A & 6B; Table B in S1 File; Text A in S4 File). S2 Table provides a comparison of selected features of the external morphology of C. congoensis with that of other species in the genus Colobus.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Craniodental indices differentiating C. congoensis and C. satanas.

In both females (A) and males (B), C. congoensis cranial size is smaller than all other species of Colobus. (C) C. satanas has relatively broader central incisors compared to other species of Colobus, which C. congoensis lacks. (D) In lower molar shape, Colobus has shorter molars than Procolobus and Piliocolobus, and within Colobus, C. satanas has uniquely mesiodistally shortened, buccolingually broad lower molars. C. congoensis is differentiated from C. satanas and other species of Colobus in exhibiting elongated lower molars. (E) C. congoensis lacks the enlarged P⁴ characteristic of Colobus relative to other African colobines. Skull indices are described in Table 2; MD = mesiodistal.

https://doi.org/10.1371/journal.pone.0349857.g006

Among species of Colobus, C. congoensis exhibits notable similarities to Colobus satanas (Figs 2, 5). The nearly uniform black pelage covering the torso is similar in both species, in striking contrast to the white and black markings of all other Colobus species (Fig 2; Fig D in S4 File). Both C. congoensis and C. satanas have crenulated ear pinnae, markedly so in C. satanas. Prominent, long black hairs frame the forehead and sides of the face in both species. Fig 2 shows a number of external morphological features of C. congoensis and C. satanas as seen in life.

In the skull of C. congoensis and C. satanas, the frontal squama rises superiorly, and in lateral view, is continuous with the nasal profile unbroken by the glabella. The frontal of C. congoensis variably exhibits a slight bulging similar to the “saddle-shape” seen in C. satanas (sensu Groves [55]), although it is not as pronounced (Fig 4). Palate shape is allometric across all African colobines (Pearson’s r = 0.65; p < 0.001) but not within Colobus (Pearson’s r = 0.39; p < 0.001), and both C. congoensis and C. satanas have significantly shorter palates than most species of Colobus (ANOVA F_{(5, 88)} = 29.54; p < 0.001; Tukey’s HSD p < 0.05 for all pairwise comparisons with C. congoensis and C. satanas except C. polykomos, see also Table A in S4 File; Fig 5C). The mandibular corpus is shallow in both C. congoensis and C. satanas especially relative to C. guereza and C. polykomos (ANOVA F_{(5, 76)} = 7.39, p < 0.001; Tukey’s HSD p < 0.05, see also Table A in S4 File; Fig 5D). The male canine is relatively small for both C. satanas and the single male of C. congoensis, though this difference is not statistically supported in pairwise comparisons for C. congoensis, likely due to the small sample sizes (ANOVA F_{(5, 38)} = 2.61, p < 0.05; Fig 5E); C. satanas is significantly different from C. angolensis and C. polykomos (p < 0.05; Table A in S4 File). Though some of these craniodental comparisons between C. congoensis and other Colobus species are not statistically significant due to small sample sizes for C. congoensis, it consistently falls closest to the mean for C. satanas among living Colobus taxa in these distinctive features (Fig 5).

Although most similar to C. satanas among Colobus species, C. congoensis also exhibits a number of other features that are distinct from C. satanas, and in some cases, from all other species of Colobus. From field measurements and observations, C. congoensis is smaller in overall body mass and significantly smaller in cranial size compared to all other Colobus species, including C. satanas (combined sex ANOVA F_{(5, 79)} = 6.81, p < 0.001; Tukey’s HSD p < 0.001; see also Table A in S4 File; Fig 6A & 6B). In the pelage, the facial skin of C. congoensis has pale patches around the mouth and orbits, whereas in C. satanas this part of the face is entirely black (Fig 2; S2 Table). The mid-dorsum hair of C. congoensis is shorter on average (~5–12.5 cm) than that seen in C. satanas, which has longer mid-dorsum hair (12–16.5 cm; Fig D in S4 File). In C. satanas, a long tuft of hairs is present at the proximal base of the tail (extending ~15–18 cm distally from the base of the tail), whereas the hair on the rest of the tail is short (~1.5–3 cm). The tail of C. congoensis lacks a similar tuft at the base, and the hairs are only slightly longer at the base of the tail (~2–3 cm) than along the rest of the tail (~1–1.5 cm). Colobus congoensis has a distinct and extensive white perianal patch not found in C. satanas. In other Colobus species, any perianal white hair patches are typically more restricted to areas near and surrounding the callosities (S2 Table). Additional descriptions of C. congoensis pelage and comparisons to C. satanas are provided in Text A in S4 File.

Further comparisons of craniodental features corroborate the distinctiveness of C. congoensis from C. satanas, although small sample sizes again prevent statistical significance in some comparisons (Fig 6). Qualitatively, the frontal bulging in lateral view that is noticeably present in C. satanas (i.e., “saddle-shape” of [55]) is less pronounced in C. congoensis (Fig 4). Colobus satanas is distinct from other Colobus species in having enlarged upper central incisors and mesiodistally short, buccolingually broad lower molars, and C. congoensis lacks both of these features (Fig 6C-6D). In I¹ breadth, C. satanas has significantly broader incisors than C. angolensis and C. guereza (ANOVA F_{(5, 78)} = 5.21; p < 0.001; Tukey’s HSD p < 0.01, see Table A in S4 File for all pairwise comparisons; Fig 6C). Although C. congoensis is not statistically significantly differentiated from C. satanas, it does exhibit relatively narrow incisors and falls at the bottom of the range of variation for C. satanas (Tukey’s HSD p = 0.56; Fig 6C). Colobus as a genus typically has shorter molars than Procolobus and Piliocolobus (ANOVA F_{(3, 156)} = 37.96, p < 0.001; Tukey’s HSD p < 0.05), and within Colobus, C. satanas has uniquely mesiodistally shortened, buccolingually broad lower molars (ANOVA F_{(5, 82)} = 16.82, p < 0.001; Tukey’s HSD p < 0.001 for all except C. vellerosus; Table A in S4 File). Colobus congoensis has elongated molars (p < 0.05 compared to C. satanas, C. polykomos, and C. vellerosus), and it is differentiated from Colobus as a genus (p < 0.05), resembling Piliocolobus in this index. In addition to molar shape, C. congoensis is differentiated from the rest of Colobus in exhibiting a small P⁴ relative to M¹ length, more similar to that of Procolobus and Piliocolobus (ANOVA F_{(3, 156)} = 17.05, p < 0.001; Tukey’s HSD Colobus vs. all groups, p < 0.01, Table A in S4 File; Fig 6E). In molar shape and relative P⁴ size specifically, C. congoensis is more similar to Piliocolobus and Procolobus than other members of Colobus.

A multivariate assessment of these indices demonstrates the distinctiveness of C. congoensis relative to C. satanas and all other species of Colobus (see Tables B-D and Figs E & F in S4 File). Phenetic distances between C. congoensis and C. satanas are greater than between any other pair of Colobus species (Fig E and Table C in S4 File) or pair of subspecies (Fig F and Table D in S4 File).

3.3 Genetics

Of the three C. congoensis mitochondrial DNA sequences that were generated in this study, two had unique haplotypes and were included in the phylogenetic analyses. The final alignment consisted of 46 African colobine samples, including representatives of the six currently recognized Colobus species, 10 species of Piliocolobus, and Procolobus verus, together with seven samples from Asian colobines (S1 Table). Our phylogenetic analyses of mitochondrial DNA sequences recover the C. congoensis samples as a monophyletic clade that is sister to C. satanas; together, the C. congoensis–C. satanas clade forms the sister to all other Colobus species (Fig 7; Fig A in S5 File). Mitochondrial sequences from two C. angolensis individuals from the buffer zone of Lomami National Park form a monophyletic clade that is the sister to a C. angolensis clade that includes one individual from Tanzania and two of unknown provenance (Fig A in S5 File).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Estimated divergence dates of African colobines using a fossil calibration.

MCMCTree divergence date estimates using available mitochondrial data and a fossil-calibrated prior from de Vries and Beck [43]. The error bars on the nodes represent 95% highest posterior density intervals. The dashed gray line highlights the inferred divergence date between C. congoensis and C. satanas. Colobus congoensis is highlighted by a red rectangle.

https://doi.org/10.1371/journal.pone.0349857.g007

Using a calibration point on the African–Asian colobine split from a recently published set of curated primate fossil calibrations [43], the split between C. congoensis and C. satanas is estimated at 5.0 million years ago (Ma; 95% highest posterior density [HPD] = 4.3–5.8 Ma; Fig 7, Table A in S5 File). The divergence of the sympatric Lomami C. angolensis angolensis population from other C. angolensis populations is estimated at 3.4 Ma (95% HPD = 2.9–4.0 Ma; Fig 7, Table A in S5 File). When using the estimated divergence date from Perelman et al. [45] as the calibration on the Colobinae node, the split between C. congoensis and C. satanas is estimated at 4.1 Ma (95% HPD = 3.6–4.7 Ma), and the split between the Lomami C. angolensis angolensis population and other C. angolensis populations is estimated at 2.8 Ma (95% HPD = 2.4–3.3 Ma; Fig B and Table A in S5 File).

3.4 Vocalizations

Roaring calls were heard at all times of the day, but particularly in the morning. The pale coloration around the mouth produces a distinctive visual display in roaring C. congoensis (S1 Video). Multiple individuals were observed calling together on several occasions. In one observation of a social group (n = ~10 individuals), we heard an intense bout of rapidly repeated roars during an attack by a crowned eagle (Stephanoaetus coronatus). These observations show that C. congoensis, like other Colobus species, roars in varied situations, including when under threat from potential predators [46,56].

We analyzed C. congoensis roars from three audio files, all of which included more than one individual calling during the event. In the field, C. congoensis roars are similar to those of sympatric C. angolensis but can be distinguished by the rapid pulse rate of roars and a distinctive snort between roar sequences. Snorts by C. angolensis are often produced alone and independent of roaring [46]. A video recording of roaring C. congoensis is provided in S1 Video.

3.4.1 Comparisons within the genus Colobus.

Oates and Trocco [48] and Oates et al. [47] compared features of roars of five Colobus species: C. guereza, C. polykomos, C. satanas, C. angolensis, and C. vellerosus. They identified three characteristics within roar phrases that could be used to group species: pulse rate (pulses per second), peak fundamental frequency (Hz), and the presence of frequency modulation. We compared C. congoensis roars with these five species plus C. caudatus (reported as C. guereza caudatus in [47,48]). Colobus guereza, C. caudatus, and C. vellerosus roars had a slow pulse rate and low fundamental frequency. Colobus angolensis and C. polykomos roars had an intermediate pulse rate and intermediate or high fundamental frequency. Colobus satanas roars had a rapid pulse rate and high fundamental frequency with frequency modulation. Colobus congoensis roars are structurally closest to those of C. satanas with rapid pulse rate, high fundamental frequency, and frequency modulation across the call (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Comparative summary of roar features for Colobus species.

https://doi.org/10.1371/journal.pone.0349857.t004

3.4.2 Comparison of C. satanas and C. congoensis roars.

For our comparative analysis of C. congoensis and C. satanas roars, we used 86 C. congoensis roar phrases from one audio file with at least two individuals calling. We extracted 101 C. satanas roar phrases from two audio files with at least three individuals calling. Roars by C. congoensis and C. satanas have two dominant frequency bands and discrete pulses within each phrase. Two roar phrase variants were visually identified using spectrograms: a pronounced roar considered to be a ‘primary roar’ and a less pronounced roar considered to be a ‘secondary roar’ (Fig 8). Primary roars were longer in duration and exhibited frequency modulation across the call. Secondary roars were shorter in duration with less frequency modulation than the primary roar preceding it. The intensity of frequency modulation varied among primary phrases in both species. Primary roars were either isolated or started a sequence of phrases that alternated or ended with a secondary roar. In several instances, a primary roar by C. congoensis was followed by two secondary roars. Primary roars for both species were almost always preceded by a snort vocalization. Secondary roars in C. satanas were separated from primary roars by a snort, whereas C. congoensis did not snort between primary and secondary roars. Audio recordings of roaring C. congoensis and C. satanas are provided in S1 Audio and S2 Audio, respectively.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Spectrogram of C. congoensis and C. satanas.

Spectrogram of C. congoensis (top) and C. satanas (bottom) roars. Primary roars in both species are preceded by a snort and show frequency modulation. Secondary roars show little to no frequency modulation and are preceded by snorts in C. satanas, but not in C. congoensis. Audio files that include the calls depicted in the spectrograms can be found in S1 Audio (C. congoensis) and S2 Audio (C. satanas).

https://doi.org/10.1371/journal.pone.0349857.g008

Colobus congoensis roars were higher in frequency on several acoustic features and had a larger formant dispersion between dominant frequency bands compared to C. satanas roars. Primary roars of C. congoensis consistently exhibited a negative slope, whereas those of C. satanas displayed an arched contour of varying intensity in the fundamental frequency. However, secondary roars for C. congoensis and C. satanas did not differ in their slope. Primary roars were similar in duration and rise time, but in secondary roars, duration and rise time were shorter in C. congoensis compared to C. satanas. Pulse rate was similar for both species across both call variants. Using principal component analysis (PCA), both call variants formed species-specific clusters for C. congoensis and C. satanas (Figs A & B in S2 File, and Tables A-E in S2 File). Summary statistics for acoustic features of roar calls are reported in Table A in S2 File.

3.5 Distribution, ecology and behavior

The known range of Colobus congoensis is limited to an area of 1,700 km² in the provinces of Tshopo and Maniema in east-central DRC. Its range extends from the right bank of the Lomami River, east into the adjacent Lualaba basin to the Lilo River, a tributary of the Congo (Lualaba) River. The northern known limit is at approximately 1°20’ South Latitude. The southern limit is approximately 2° South Latitude (Fig 1). Colobus congensis is not known from the west bank of the Lomami River despite the occurrence of apparently suitable habitat. Elevations in the range area vary from 390 to 500 meters above sea level (a.s.l.), with observations of the species recorded from 406 to 490 meters a.s.l. Annual rainfall recorded at two sites bordering the north and south of the C. congoensis range averaged 1,870 mm and 1,720 mm, respectively. Rainfall seasonality is bimodal: October through December and March through April are the wettest months with monthly rainfall averaging 201–269 mm. June and July are the driest months, with monthly rainfall averages of 34–111 mm.

Colobus congoensis is currently known from 114 detections made between 2018 and 2022 (Fig 9, S4 Table). Colobus congoensis were observed visually on 89 encounters. During mixed-species associations, observers estimated the number of C. congoensis in 62 of the visual encounters. Twenty-five detections were by vocalizations only or by vocalization with little visual contact.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. Observations of Colobus congoensis, 2018–2022.

Following its discovery in November 2018, an additional eight observations of C. congoensis were made through 2019. Beginning in 2020, surveillance patrol coverage, directed searches, and dawn call surveys were increased in the range area, and line transect surveys initiated. One hundred five of 114 observations were recorded from 2020 through 2022. Surveillance patrol surveys were deployed on a 5 × 5 km grid. Grids with high coverage had at least 3 patrols per year and an accumulation of 30 km of patrol track. Low coverage patrol grids had lower annual and/or total patrol track coverage. Line transect coverage for survey blocks A–F is given in Table 5. Villages reporting knowledge of C. congoensis are restricted to the buffer zone bordering the Lomami National Park.

https://doi.org/10.1371/journal.pone.0349857.g009

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Line transect inventory results.

https://doi.org/10.1371/journal.pone.0349857.t005

Colobus congoensis is an uncommonly encountered monkey within its range (Fig 9). Eight other primate species were documented in the range area, including seven monkey species and the Bonobo, Pan paniscus (S5 Table). Detection rates within the range area (detections / survey effort) for C. congoensis were evaluated for all survey methods, except for 25 detections made on directed searches. Between 2020 and 2022, 3,063 kilometers of terrestrial surveillance patrols were deployed within the C. congoensis range. Twenty-nine of a total of 36 C. congoensis detections by surveillance patrols were made during this period, for an overall encounter rate of 0.01 detections per km of patrol recce. Surveillance patrol detections were recorded in just 20 of 76 patrol grid cells (5 × 5 km) covered by patrols within the C. congoensis range. Colobus congoensis detection rates were low on riverine patrols; 790 km of riverine coverage from 2020 through 2022 yielded only 5 encounters of C. congoensis. Line transect inventories, totaling 1,285 km across all 6 survey blocks, yielded a total of 38 C. congoensis observations out of a total of 709 detections of all primates (Table 5). Colobus congoensis was detected during transect inventories only in survey blocks A (11 detections) and B (27 detections), comprising 0.033 and 0.041 detections per km of transect for the two blocks, respectively, and representing 5.4% of all primate detections (Table 5). Dawn primate call surveys yielded patterns of C. congoensis detection rate and distribution similar to those produced by transect inventories. Colobus congoensis were recorded on only 10 of 351 call surveys within the species’ range over the 3-year period 2020–2022 (Fig A in S3 File).

Colobus congoensis are concentrated toward the western limits of the known range, with 104 of 114 detections recorded in pediment forest within 10 km of the Lomami River. Within this zone, observations of C. congoensis are further concentrated in the vicinity of four of the six major meanders constituting the dominant feature of the Lomami River’s course in the Courbure Sector of the Lomami National Park. East of this band, and stretching into the Congo (Lualaba) Basin, the landscape gives way to a mosaic of upland and sende forests. Colobus congoensis were reported from only two locations in the sende zone; both were islands of upland forest west of the Lilo River that are not subject to seasonal inundation.

Observed group size of C. congoensis ranged from 1 to 20 individuals, with an average of 6.2 and a median size of 5.5 individuals. Colobus congoensis shows a consistent tendency to move in mixed-species groups of primates. Of 62 observations where group composition was determined, C. congoensis was reported in a polyspecific association on 45 observations, with group compositions ranging from 2 to 6 species. Two other colobines present in the landscape were recorded in mixed groups with C. congoensis: C. angolensis was observed in association with C. congoensis five times and Piliocolobus parmentieri was observed once. Adult C. congoensis with young were observed in February, April, July, August, and September. In all cases, the young animals had the full black coloration and piebald facial pattern of adults. Colobus congoensis were observed primarily in the canopy or middle level of mature forest. Two exceptions were reported. In one event, two C. congoensis of a monospecific group of 5 individuals descended to the ground briefly before moving back up into the middle level. In the second event, a group of approximately 10 C. congoensis under attack from a crowned eagle (Stephanoaetus coronatus) moved from the canopy to the understory.

4.1 The place of C. congoensis in the radiation of the genus Colobus

Colobus congoensis represents the sister taxon to the Black Colobus (C. satanas), long recognized to be an unusual member of the genus characterized by its all-black pelage, unique cranial morphology, and dietary ecology [48,55,57,58]. Our description of a new species most closely related to C. satanas thus represents a significant expansion of known Colobus diversity, with implications for understanding the evolution of the genus and of African colobines, more broadly. Our molecular analysis is consistent with previous phylogenies in positioning C. satanas (here, together with C. congoensis) as sister to a monophyletic clade of black-and-white Colobus species (C. guereza, C. angolensis, C. polykomos, C. vellerosus, and C. caudatus) [19,32,35,36,59].

Our fossil calibrated mitochondrial clock analysis estimates the most recent common ancestor of C. congoensis and C. satanas existed ~4.27–5.78 Ma. In an additional analysis that used the same secondary calibration point as Roos and Zinner [36], we obtained younger divergence estimates by approximately 1 million years (3.44–4.73 Ma). This difference is likely due to limitations of applying molecular dating methods to small mitochondrial datasets and differences in calibration approaches. Future research on comparative whole genome data (nuclear and mitochondrial) will provide more accurate divergence date estimates for the Colobus clade [e.g., 60,61].

A Pliocene age split between the C. congoensis–C. satanas lineages represents the deepest split between sister species within the genus Colobus, similar in age to the oldest species-group divergences within other speciose cercopithecid genera (i.e., Macaca and Cercopithecus) [59,61]. This divergence is additionally substantially older than between any subspecies among African colobines (Fig 7). The antiquity of this split, combined with distinct pelage and craniodental features, indicates that C. congoensis is an independently evolving lineage that deserves to be recognized as a novel taxon apart from its phylogenetically closest relative, C. satanas, at the species level [62]. Under a phylogenetic species concept [63], the specific status of C. congoensis is corroborated by several unique, diagnosable features relative to both its sister species, C. satanas, and all other species of Colobus (Figs 5 & 6, Figs E & F and Tables C & D in S4 File; see also formal diagnosis above and summary below). Structural analysis of loud call vocalizations identifies similarities between C. congoensis and C. satanas that distinguish them from all other Colobus species [47,48] but also highlights distinctive features that further separate C. congoensis in vocalization sequence, frequency modulation, and formant dispersion. Morphological similarities shared by C. congoensis and C. satanas that are not present in other Colobus species include the all-black pelage and the unusual saddle-shaped frontal bones. Colobus congoensis, however, is distinguished by its striking piebald facial skin markings; the white perianal patch; unique aspects of its dental morphology in relative size of the upper incisor, premolar proportions, and molar shape; and its overall small body size and cranial size. Multivariate analyses of phenetic distances also confirm the morphological distinctiveness of C. congoensis relative to C. satanas and all other Colobus species (Figs E & F and Tables C & D in S4 File).

For many aspects of its external appearance, C. congoensis is reminiscent of Asian colobine genera such as Trachypithecus and Presbytis, particularly in the light colored skin around the mouth, eye rings, and the presence of a crest/crown on the top of the head, raising the question of whether C. congoensis retains some pelage features from the common ancestor with Asian colobines. Similarities in external features to Trachypithecus and Presbytis have also been noted for Procolobus verus, which is the sister group to Piliocolobus [64]. These similarities suggest that both Piliocolobus and the other Colobus species are independently more derived in external appearance compared to Procolobus and the C. satanas–C. congoensis clade. Derived features in Colobus include the striking black and white coloration and elaborate manes, capes, and tail tufts seen in many members of the genus (S2 Table) [58,65,66].

Dental differences between C. congoensis and its congeners, particularly C. satanas, may also suggest a distinctive dietary ecology. Colobus species are characterized by mesiodistally shorter molars, greater basal flare, relatively larger premolars, and reduced canine base area dimorphism relative to Piliocolobus, and these features have been related to a greater degree of seed predation and mature leaf consumption within the genus [31,67–72]. Even relative to other Colobus species, C. satanas is highly granivorous, with seeds making up half or more of its diet [67,71]. The enlarged upper incisors, particularly mesiodistally shortened and buccolingually broad molars, and higher basal flare of C. satanas have been linked to its extensive seed predation [68,69,72]. On the other hand, C. congoensis lacks these distinctive dental features: it has small incisors, smaller premolars, and mesiodistally elongated molars. In these latter two features, it is also distinguished from all other species of Colobus and is more similar to Piliocolobus. This suggests that C. congoensis is likely not as granivorous as C. satanas and may rely on seeds even less than all other species of Colobus, with a dietary profile more similar to Piliocolobus species.

4.2 Habitat specialization in C. congoensis

Colobus congoensis is limited to upland forests, and in particular those on pediments with deep clay soils. The species’ occurrence outside of this habitat is highly localized. Upland clay substrates support some of the most diverse forests in Lomami National Park, with significantly higher tree species diversity than in forests on upland white sand soils or on seasonally inundated substrates [73,74]. This suggests that C. congoensis may be limited by habitat quality, and that nutrient poor and/or less diverse forests, including those occurring on white sand substrates, are not suitable habitats.

At a finer geographic scale, the highly variable spatial distribution of canopy tree species in the terra firme, clay-substrate forest [74] suggests that local floristic composition may be further shaping the localized distribution of C. congoensis. Repeated observations of C. congoensis in the vicinity of the same location suggest that C. congoensis groups resample selected sites over time. Colobus congoensis is rare in the sende zone, which occurs at the eastern limits of the species’ range. We recorded only two occurrences in this zone, both on well drained islands of sandy clay where local informants led us to monkeys that they reported seeing regularly. Both these areas were also used on a temporary basis for shifting cultivation by people based in settlements east of the Lilo River (Fig 10).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 10. Colobus congoensis distribution and habitat.

Colobus congoensis observations (2018–2022) shown relative to habitat classes (see Table 3 for habitat descriptions) within the known range.

https://doi.org/10.1371/journal.pone.0349857.g010

Fragmented or geographically restricted distributions are not uncommon among primates, and are shaped by factors such as habitat requirements, biogeographic barriers, and faunal community composition. Among African colobines, for instance, populations of Piliocolobus oustaleti persist in small islands of gallery forests isolated in savannas on the escarpment overlooking Lake Albert in the DRC [75]. Similarly, in Lomami National Park, Piliocolobus parmentieri has a patchy distribution in the Lomami-Lualaba interfluve that appears to be associated with selection for islands of suitable habitat. Like P. parmentieri, the distribution of C. congoensis follows a comparable pattern in its rarity and localized distribution, suggesting that habitat specialization is most likely driving the fragmented distributions of these taxa.

4.3 Biogeography and range restriction of C. congoensis

Colobus congoensis is limited to an unusually small geographic range currently estimated at 1,700 km² (Fig 1), while most other known species of the genus have more extensive ranges exceeding 60,000 km² [76]. The range of C. congoensis is notably separated from its closest phylogenetic relative, C. satanas in western central Africa, by at least 1,200 km. Both the limited distribution of C. congoensis and the geographic separation from its sister species raise questions about the historical biogeography of the genus Colobus and local biogeographic factors limiting its range.

A west African origin for both the Colobus and Procolobus–Piliocolobus clades, and by extension, African colobines as a group, has been proposed on the basis of the deep divergences between C. satanas and its congeners and between Procolobus and Piliocolobus [77,78]. However, C. satanas does not represent the westernmost extension of the Colobus range (Fig 1b), and the position of C. congoensis as sister to C. satanas in the Congo Basin further complicates this picture. The distribution of Colobus species does not follow a straightforward west-east phylogeographic pattern, and thus the geographic origins of the genus are difficult to discern. Complicated west-east biogeographic dispersal scenarios have also been hypothesized for several other African catarrhine primates, including Cercocebus mangabeys [79], chimpanzees [80,81], some guenon clades [82–84], and notably, red colobus monkeys [35,85]. Identifying the origins of the C. satanas–C. congoensis clade as either within or outside the Congo Basin remains obscure. Regardless, the relative evolutionary isolation of C. congoensis may suggest it is a relictual member of a previously more diverse radiation within the Congo Basin, possibly confined to its small current range by competitive exclusion (e.g., by Piliocolobus species) and/or by habitat specialization.

Molecular estimates for the origins of several colobinan lineages date to the late Miocene or early Pliocene [35,45,59,61]. Our mitochondrial clock analyses estimate the split between C. satanas and C. congoensis at ~4.1–5.0 Ma. The antiquity of this split indicates that the divergence of the two taxa was unlikely to be directly due to Pleistocene glacial cycles and the formation and expansion of forest refugia, as proposed for other speciose cercopithecid radiations such as the guenons (tribe Cercopithecini), Cercocebus mangabeys, or Piliocolobus [35,48,77,79,85]. Rather, riverine dynamics and the evolution of central African geomorphology likely played a role in the differentiation of C. satanas and C. congoensis and the current restriction of the C. congoensis range [86,87].

While the current course of the Congo River dates to the end of the Eocene or beginning of the Oligocene, 30–40 million years ago [88], the configuration of the Congo’s central basin tributary watersheds developed more recently in the late Pliocene through the Quaternary [89]. North-south oriented drainages with sources on the periphery of the Central Basin, such as the Lomami River, are thought to have developed over the past 2 million years [89]. The appearance of dissected pediments, including those occupied by C. congoensis, are dated to the late Pliocene [52, Plate 2]. Collectively, this suggests that the modern configuration of the C. congoensis range may have been in place by the late Pliocene or early Pleistocene, postdating the estimated age of the split between C. satanas and C. congoensis.

The rivers that bound the C. congoensis range to the west and east appear to represent biogeographic barriers for the species through different mechanisms. The Lomami River, which defines the western boundary of the C. congoensis range, borders the species and subspecies ranges of several other primates in the Lomami National Park and buffer zone region (S5 Table). With an average width of over 200 meters, the Lomami River appears to represent a significant barrier to dispersal. At the eastern limit of the C. congoensis range, the Lilo River is less than 50 meters wide, and in some places is covered by canopy traversable by arboreal primates. However, the Lilo River and all other major tributaries within the Lomami-Lualaba interfluve in the C. congoensis range have “elbows,” right-angle changes of course that are characteristic of captured or “beheaded” rivers (Fig 1) [90]. All these rivers are also characterized by depositions of white sands and seasonal inundations in regions that were occupied by the former course of the captured flowage. Nutrient-poor, white-sand soils and their associated vegetation have been recognized as ecological barriers for a number of plant and animal taxa in Amazonia, and in some cases have been demonstrated to promote allopatric speciation [91–93]. Such habitats have not previously been identified as geographic barriers limiting primate ranges in Africa, as seems to be the case for C. congoensis. The effect of these nutrient-poor habitats in the Lomami-Lualaba interfluve in defining the range limits of C. congoensis would be better understood with an assessment of the species’ diet, as the relationship between nutrient limitation and primate distribution and abundance has been shown to be mediated through diet choice and food availability [67,94–96].

4.4 Conservation status

We propose a provisionary IUCN Red List classification for C. congoensis of Endangered (EN). Colobus congoensis is a rare, geographically restricted monkey that is not well known even by local communities in the vicinity of its range. Mean group size and encounter rates on surveys are low, and C. congoensis sightings are highly localized and mostly limited to a narrow area in the vicinity of the Lomami River. According to local informants, it appears that the species was not specifically targeted by hunters in the past, nor has its range been reduced in living memory. Most of the known locations of C. congoensis occur within the protected Lomami National Park.

Butynski and de Jong [97] list seven predictors of heightened risk of extinction in colobines, of which three are pertinent to C. congoensis at present: dependence on high canopy, old growth forest, and a body size greater than 5 kg. Among the remaining predictors, habitat loss, human population growth, and increased hunting in the range can be projected as threats to this species over the upcoming decades. Human population growth in DRC is estimated at 3.2% annually, one of the most rapidly growing national populations on the continent [98]. Population growth is one of the most important drivers of forest loss. Over 3,300 km² of primary forest is projected to have been lost in the DRC between 2015 and 2025 [99]. Loss of primary forest cover in Tshopo Province, which includes a significant portion of the C. congoensis range, is estimated at 4% annually [100].

The C. congoensis range in the buffer zone of Lomami National Park is located in an area where the local economy is dependent on hunting, fishing, and shifting cultivation. Furthermore, the area is characterized by extensive seasonally inundated forests, reducing the availability of terra firme islands within the sector, preferred by both C. congoensis and humans. Expansions of human populations from the current areas of settlement are ongoing, with at least 15 new villages added within and adjacent to the range of C. congoensis in the Lomami National Park buffer zone between 2015 and 2023 (Lomami National Park, unpublished data). If these demographic and land use changes continue, we anticipate significant reductions of C. congoensis populations and loss of habitat in the buffer zone over the upcoming decades.

Effective protection of Lomami National Park emerges as the most important requirement for the conservation of C. congoensis. The discovery of C. congoensis was preceded by the discoveries of two previously unknown or little known cercopithecids in the Lomami National Park and buffer zone: Cercopithecus lomamiensis [28] and Chlorocebus dryas [101–103]. The latter was previously known from a single small range area over 400 km to the west.

The known primate fauna of the Lomami National Park and its buffer zone totals 15 species, including 8 species endemic or near endemic to the DRC (S5 Table). Although many of these primate species are also found outside of the TL2 interfluves, the discovery of C. congoensis along with the novel species Ce. lomamiensis and the very poorly known Ch. dryas points to a gap in scientific knowledge of the biodiversity in this region. These discoveries confirm the exceptional importance of the TL2 landscape for primate biogeography and evolution in Congo’s Central Basin. The Lomami National Park is one of the most important protected areas for primate conservation in Central Africa.