Genetic diversity of African ticks (Amblyomma compressum) recapitulates genetic population origins of pangolin host (Phataginus tricuspis) from Asian seizures

Abstract

Background

Ectoparasites maintain close associations with their hosts and can complement host genetics data in studies aimed at identifying wildlife trafficking routes, source populations, and whether trafficked individuals are wild-caught or captive-bred. Despite this association, ectoparasites are rarely incorporated into wildlife forensic investigations. Using ticks from Asian seizures of white–bellied pangolin (Phataginus tricuspis) scales, we investigated ectoparasite-host associations by comparing tick genetic diversity to pangolin genetic populations of origin.

Results

We successfully genotyped 34 tick samples – identified as Amblyomma compressum – from 27 white–bellied pangolin scales, from eight seizures in Hong Kong, all originating from Africa. Tick haplotypes were mapped using estimated geographic coordinates assigned to pangolin scales through SNP-based genotyping, enabling assessment of tick haplotype diversity and distribution across the species’ range. Our study suggests concordance between tick genetic diversity and corresponding pangolin genetic clusters but not purported seizure origins. Ticks exhibited intra-specific variation within and between pangolin genetic clusters, based on the number of haplotypes identified (within pangolin genetic clusters) and observed spatial clustering (between pangolin genetic clusters).

Conclusions

These findings enhance our understanding of pangolin–tick associations at the population level and demonstrate that ectoparasites from trafficked wildlife can serve as valuable biological tags. When integrated with host population genetics, they provide additional insight into geographic origin and host health, thereby strengthening tools available for conservation and wildlife crime investigations.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12862-026-02563-3.

Introduction

The illegal wildlife trade (IWT) is a lucrative enterprise involving the movement of millions of live plants and animals, fungi and wildlife-derived products [1]. Over-exploitation driven by poaching to meet consumer demands has resulted in significant population declines across numerous species [2, 3]. Consequently, many highly trafficked species are protected by the Convention on International Trade in Endangered Species of Fauna and Flora (CITES) and the International Union for Conservation of Nature (IUCN). Identifying heavily exploited species, determining whether specimens are illegally sourced, and pinpointing poaching hotspots and trade routes are critical components of wildlife forensic investigations and preventative conservation strategies.

Several genetic tools have been developed to address questions surrounding geographic traceability of some highly trafficked species [4–9]. These approaches rely on comprehensive sampling across species’ ranges to accurately characterize genetic diversity. However, wild sample acquisition can be difficult for large, charismatic species, especially when CITES permits are required for conducting research. Additionally, some highly trafficked species, occupy large home ranges, while other species exist in small populations across fragmented habitats, thus exhibiting low genetic heterozygosity, limiting the resolution of genetic assignment methods [10–12]. To overcome these limitations, alternative biological materials—such as ectoparasites—may provide valuable complementary data, as they are often closely associated with hosts and can reflect geographic patterns of otherwise difficult-to-sample species.

Pangolins (Order Pholidata) are among the taxa most severely impacted by the illegal wildlife trade. They have been exploited for centuries for their scales, for perceived medical benefits, particularly in Traditional Chinese Medicine (TCM) [13] and meat, considered a delicacy in parts of Asia and a source of subsistence in some African regions [14]. Intensified demand for scales and meat have led to illegal and unsustainable hunting in pangolin populations, particularly in Asia. The depletion of pangolin populations across Asia led to a shift in markets and increase demand for African pangolin species [15, 16].

Ticks are common ectoparasites of pangolins and are often incidentally discovered in seizures. At least twenty tick species, spanning eight genera, have been identified from tick-pangolin associations studies [17, 18]. Some of these species are vectors of pathogens capable of causing severe disease [19] while others contribute to physical and physiological stress in their hosts [20, 21]. Ticks typically aggregate between scales, the face and ventral-thoracic regions of pangolins and can travel, undocumented, on traded scales and live pangolin specimens to new locations. Two of these species, Amblyomma javanense (Asian pangolin tick) and A. compressum (African pangolin tick) are mostly found on pangolins, thus, are considered highly specialized and possibly, co-endangered with declining pangolin populations [22–25].

Due to their close ecological association with hosts, ticks have potential utility as biological tags in wildlife forensic investigations. Their genetic diversity and geographic structuring may complement host-based genetic analyses, particularly when host populations exhibit low genetic variation. In some cases, ticks may retain sufficient genetic signal to enable geographic localization even when host-based approaches fail. Furthermore, while ticks are globally distributed, certain species exhibit restricted geographic ranges [26], providing additional spatial information at the species level. Importantly, ticks are not regulated under CITES, which reduces logistical barriers to their collection and use in research.

Geographic assignment frameworks have already been developed for several pangolin species [6, 7, 9, 27, 28]. Given that ticks are routinely recovered from pangolin scales seizures and removed from rescued animals, it is possible to test whether ticks can be used to complement pangolin geographic origin studies. In this study, we integrate genetic data from both pangolins and their associated ticks obtained from scale seizures. Building on the work of Tinsman et al. [7], we incorporated tick genetics into analyses of single nucleotide polymorphism (SNP) variation in white-bellied pangolins to evaluate whether tick genetic patterns correspond with host population assignments. We used sequence-based typing to identify tick species and assess their genetic diversity, and constructed haplotype networks to examine spatial and evolutionary relationships among ticks across seizure batches and pangolin genetic clusters. We hypothesized that tick genetic variation would exhibit spatial structuring consistent with pangolin SNP–based population assignments.

Methods

White-bellied pangolin scales identification and/or barcoding for SNP genotyping

Sample sorting and/or cleaning

We sampled pangolin scales that were trafficked from Africa to Hong Kong between 2013 and 2018. Samples from these seizures were donated to the Conservation Forensics Lab, The University of Hong Kong, by the Agriculture, Fisheries and Conservation Department (AFCD) and Kadoorie Farm & Botanic Garden (KFBG). From January–February 2025 we searched through seizures of pangolin scales and removed scales that had one or more ticks (regardless of life stage) attached to them (see Additional file 1 for details on pangolin scale seizures).

Prior to cleaning scales for total DNA extraction, we removed ticks attached to individual pangolin scales with forceps and gave them unique codes linking them to both the scales they were removed from and the seizure bag from which they were obtained. To remove old, dried pangolin tissue and dirt, we placed tick specimens in 1.5 mL Eppendorf tubes containing ultra-pure water and washed them by vortexing. This step was done three times. We then placed individual ticks on a sterile tray and scrubbed them with absolute EtOH, and a small paint brush [29]. Following this, tick samples were rinsed with absolute ethanol (EtOH) to get rid of any contaminants, after which they were left to air-dry. After photographing each tick, we used a clean scalpel to cut individual ticks into small pieces – changing the scalpel blade between samples. We transferred the portions to 1.5 mL Eppendorf tubes and used a clean pestle to crush each tick (see Additional file 2 for details on tick labels, life stage and sex).

We cleaned pangolin scales using 10% bleach (NaClO), followed by absolute EtOH using individual brushes and left to airdry. The tissue removed from each scale was placed in individual 1.5 mL Eppendorf tubes and washed with molecular grade EtOH to remove any lingering debris/contaminant. After 30–60 min we removed the EtOH and tubes containing tissue were placed in an oven at 60 °C to ensure evaporation of all EtOH.

P. tricuspis and A. compressum DNA extraction, gel electrophoresis, PCR, PCR cleanup and sequencing

To confirm species identification we extracted total DNA from individual pangolin scales’ tissue and crushed ticks using QIAGEN DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany), following the manufacturer’s instructions, with the following changes, to optimize the protocol for increased DNA yield from low quality samples, at the lysis step, we incubated pangolin tissue and tick samples overnight and adjusted the volume of Buffer AE from 200 𝜇L to 25 𝜇L for both elutions, then let the columns stand for 10 min at room temperature before each centrifugation. We included a negative control for each set of extractions to monitor possible contamination of reagents used during DNA extraction. Following DNA extraction, we assessed yield and quality of DNA using Nanodrop One (Thermofisher^TM). DNA extracts were stored at -20 °C for further analyses.

We performed polymerase chain reaction (PCR) on pangolin DNA extracts by utilizing a 307 bp region of Cytochrome B (CytB) which have been validated for pangolin species identification studies [30]. For tick DNA tick extracts, we utilized a 710 bp region of the COI gene [31] along with a 478 bp region of the 16S gene [32], which have been used previously to assess tick identity and genetic diversity (see Table 1 for primer sequences). Each reaction (total volume = 20 𝜇L) contained 2 𝜇L of DNA template, 8 𝜇L 1 x Biotechrabbit^TM Lyo Hot Start PCR Master Mix, 6 𝜇L H₂O, and 2 𝜇L each of forward and reverse primer (10 𝜇M each). The initial template denaturation profile for amplifying the three gene regions was programmed at 95 °C for 2 min. This was followed by 30 cycles of 95 °C for 30 s, 50 °C (for tick primers) and/or 57 °C (for pangolin primer) for 30 s, 72 °C for 1 min; and a final elongation step at 72 °C for 7 min.

Table 1.

Oligonucleotides used in P. tricuspis and A. compressum identification studies

Target gene	Primer	Sequence (5’- 3’)	Amplicon (bp)	References
CytB	PID_F	CCCTCYAAYATCTCHGCATGATGRAA	307	[30]
	PID_R	GCNCCTCARRADGAYATYTGTCCTCA
COI	LCO1490	GGTCAACAAATCATAAAGATATTGG	710	[31]
	HCO2198	TAAACTTCAGGGTGACCAAAAAATCA
16SrRNA	16 S + 1	CTGCTCAATGATTTTTTAAATTGCTGTGG	478	[32]
	16 S-1	CCGGTCTGAACTCAGATCAAGT

To determine successful amplification of the target region, we visualized PCR products on a 2% agarose gel in 1 x SB buffer using 5 𝜇L PCR product. After this we purified PCR products using the QIAGEN MinELute PCR Purification Kit (QIAGEN, Hilden, Germany), following manufacturer’s instructions, with the following change, we added 10 𝜇L ultra-pure water (Milli-Q) at the DNA elution step and let the column stand for 10 min before centrifugation. Purified amplicons were submitted for Sanger sequencing at the Centre for PanorOmic Sciences (CPOS), at the Li Ka Shing Faculty of Medicine, The University of Hong Kong. Chromatograms obtained from sequencing were quality checked, primer sequences were trimmed, then consensus sequences were generated and globally aligned in Geneious Prime^® v2.1 (https://www.geneious.com) [33]. To determine highest similarity (% pairwise identity), we compared consensus sequences to available CytB sequences in GenBank using the Basic Local Alignment Search Tool (BLAST) in the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/) [34].

P. tricuspis SNP genotyping cluster analysis

SNP genotyping cluster assignment

Since geographic origins of white – bellied pangolins were unknown, we utilized the SNP Type™ Genotyping Panel (Standard BioTools) developed by Tinsman et al. [7] to determine SNP genotyping cluster assignment and/or origins of the 43 pangolin scales used in the current study (see Additional file 3 for fluidigm assay details). We followed the manufacturer’s instructions (FLDM-00947 Rev 01 protocol) to process the samples and made the following change: specific target Amplification (STA) products were diluted 1:20, instead of 1:100.

Samples were loaded onto a chip 96.96 Dynamic Array™ IFC, following the priming step, for Genotyping, and run on a Fluidigm™ IFC controller HX and on Biomark HD system. We included positive controls from known genetic clusters and 2 negative (no template–NTC) controls for each run (2 runs performed), along with the test and/or unknown samples. The Genotyping Analysis Software (Fluidigm Inc.) was utilized for SNP calls, with a 90% threshold, and any call that was not clearly assigned, heterozygote or homozygote, was removed.

Confiscated P. tricuspis scales assignment and A. compressum haplotype mapping

To determine pangolin cluster assignment and geographic origins of individual scales, we utilized R packages rubias [35] and OriGen [36], and followed methods, parameters and quality control checks described by Tinsman et al. [7]. We mapped tick haplotypes, based on inferred coordinate assignments of pangolin scales in the respective SNP genotyping clusters, using the Quantum Geographic Information System (QGIS), v3.42.1-Münster [37]. We overlayed pie charts to represent each tick haplotype at these inferred coordinate assignments of white-bellied pangolin hosts.

Data processing – A. compressum genetic variability and calculated diversity estimates

Cytochrome C oxidase 1 (CO1) sequences were omitted from nucleotide diversity analysis due to failed sequences despite successful amplification. To determine intra- and inter- specific genetic variation of ticks from white-bellied pangolin scales from respective regions, we used the randomized minimum spanning tree method (RMST) to construct haplotype networks using the Pegas package in R [38]; number of and/or diversity of haplotypes, number of mutations nucleotide (π) and haplotype (h) diversity were calculated using this package. We made comparisons of tick haplotype networks generated from individual pangolin seizure bags and pangolin SNP genotyping clusters to determine which geographic information would best represent tick genetic variability.      

The estimate () function from the vegan package in R [39] was utilized to determine the Chao1 richness index to estimate the potential number of missing tick haplotypes in each white-bellied pangolin SNP genotyping cluster (Standard Error (SE) and log-transformed asymmetric 95% confidence intervals (CI) were also computed along with Chao1 richness index).

We ran a hierarchical Analysis of Molecular Variance (AMOVA), based on Kimura 2 pairwise distance matrix (K80), using the Pegas package [38] to evaluate potential source(s) – genotyping clusters and/or individual pangolins – accounting for variation in tick sequences. We also calculated pairwise distances and binomial probabilities of tick sequences across the 3 pangolin SNP genotyping clusters using the ape package [40]. A Principal Coordinates Analysis (PCoA) plot was then generated to further evaluate and/or visualize genetic and spatial relationships among the tick haplotypes, using multidimensional scaling (cmdscale).

Results

SNP genotyping cluster assignment of P. tricuspis scales

Twenty-seven white-bellied pangolin scales obtained from 8 pangolin seizure bags, for the present study. The scales were successfully assigned to three of the five white-bellied pangolin SNP genotyping clusters identified by Tinsman et al. [7]. Despite most of the seizures having reported origins (based on details from international shipment records) in Nigeria, only four scales from 2 seizure sample bags (KBG- n = 3; and KBH- n = 1) were assigned to the Nigeria genetic cluster. A total of 20 scales from the 6 seizure sample bags originating in Nigeria and from the unknown bags (G and KBT) were assigned to the Cameroon_Gabon SNP genetic cluster (CMR_GAB). Only 2 scales, 1 each from seizure sample bags KBO and KBR, were assigned to the Congo Basin SNP genotyping cluster (CAR_CON_GAB_DRC) (see Additional file 4 for SNP genotyping cluster assignments and predicted coordinates for pangolin scales).

Molecular characterization of A. compressum ticks

Partial 16SrRNA (length, 295 b.p.) sequences were obtained from 34/50 ticks removed from pangolin scales. All ticks were identified as Amblyomma compressum. The closest sequence matches to these sequences in GenBank were (1) the complete mitochondrial sequence of an A. compressum tick collected in Cameroon [41] and (2) a partial 16rDNA sequence from an A. compressum tick collected in Democratic Republic of Congo [42]. No successful sequences were obtained from the CO1 primer pair.

A. compressum haplotype diversity

Haplotype analysis of 30 A. compressum 16SrRNA sequences revealed a total of 11 haplotypes across the 8 seizures and/or 3 pangolin SNP genotyping clusters. Most of the tick haplotypes (H1 to H9) were identified in the Cameroon_Gabon SNP genotyping cluster (9/11), and the remaining haplotypes (H10 and H11), in the Nigeria (2/11) and Congo basin (1/11) SNP genotyping clusters (Fig. 1). Ten unique haplotypes were observed – 8 in the Cameroon_Gabon SNP genotyping cluster (H2 – H9) and 2 in the Nigeria SNP genotyping cluster (H10 and H11). Six of these unique haplotypes were observed only once each (singletons), in a tick individua1 (H3, H5, H6, H7 and H8 from the Cameroon_Gabon SNP genotyping cluster and H11 from the Nigeria SNP genotyping cluster. Haplotype 1 (H1) was shared with ticks from Congo basin and Cameroon_Gabon SNP genotyping clusters.

Fig. 1.

Number of Amblyomma compressum 16SrRNA sequences and assigned haplotypes with respect to white-bellied pangolin SNP genotyping clusters

The most dominant haplotype, H4, was shared among 8 ticks from 4 estimated geo-referenced locations in the CMR_GAB cluster. Other haplotypes that were identified in 3 or more ticks were H1 (n = 4), H2 (n = 4), H9 (n = 5) and H10 (n = 3).

Comparison of A. compressum genetic diversity based on seizure data and P. tricuspis SNP genotyping cluster assignments

Haplotypes H1 to H8 formed a distinct cluster from H9 to H11 (Fig. 2). Mutations ranged from 1 to 3 among H1 to H8. The two clusters were separated by 14 mutations (between H8 and H9). Haplotype H10 and H11 differed from H9 by 1 mutation.

Fig. 2.

Amblyomma compressum tick haplotype network based on (a) white-bellied pangolin scale seizure origins (based on shipment information records from law enforcement) and (b) white-bellied pangolin SNP genotyping cluster assignments (from 3/5 genetic lineages identified in Tinsman et al. [7]). Each node and/or circle in the haplotype networks represent a unique haplotype (labelled 1 to 11); node size is proportional to number of tick sequences with the same haplotype (circles in legend representing number of tick sequences); colours within each node correspond to either white-bellied pangolin scale seizure bag and/or white-bellied pangolin SNP genotyping cluster; hash lines on lines connecting nodes represent the number of mutations between haplotypes; (c) white-bellied pangolin SNP genotyping clusters to which pangolin scales were assigned

Haplotype diversity and nucleotide diversity for A. compressum tick sequences from respective SNP genotyping clusters are summarized in Table 2. Haplotype and nucleotide diversity values decreased with decreasing sample size. While high h and values were obtained for tick sequences from the Cameroon_Gabon SNP genotyping cluster, lower values were computed for tick sequences from the Nigeria SNP genotyping cluster. The Congo basin cluster had only one haplotype and thus diversity was not calculated.

Table 2.

Haplotype and nucleotide diversity of A. compressum ticks

Cluster assignment	Sample size (n)	No. of haplotypes	Haplotype diversity (h ± SD)	Nucleotide diversity (π ± SD)
Cameroon_Gabon	24	9	8.4E-01 ± 9.2E-01	1.6E-02 ± 1.3E-01
Nigeria	4	2	6.7E-01 ± 6.9E-03	1.0E-02 ± 8.2E-01

The estimated numbers of missing tick haplotypes (Chao1 index) calculated from the Cameroon_Gabon SNP genotyping cluster was 5 (standard error = 5.96; 95% confidence interval range = 9.79–40.6), while low sample sizes for Nigeria and Congo basin precluded accurate estimation (see Additional file 5 for Chao1 index results).

Pairwise distances of A. compressum tick haplotypes ranged from zero – indicating no observed mutation between sequences (e.g. H10 and H11) to 0.058 – indicating up to 17 mutations between sequences (e.g. H5 compared to H10) (see Additional file 6 for pairwise distances of tick haplotypes). In instances where more than one ticks were removed from a scale and identified as having the same life stage – e.g. larvae – sequences were assigned the same haplotype [Additional file 2: TKBTT001.1 and TKBTT001.2 (H2) and TGT004.1 and TGT004.2 (H4)]. No mutations were observed between these tick sequences, pairwise distance was zero and binomial p – value was 1.0. The opposite was observed when ticks in different life stages were removed from the same scale (Additional file 2: TGT001.1 (H1) and TGT001.2 (H4)- nymph and larva, respectively). The haplotypes were statistically distinct (p – value < 0.05) – 2 base changes were observed between the tick sequence pairs and pairwise distance and binomial p – values were 0.007 and 0.036, respectively.

Potential sources contributing to genetic variability in A. compressum tick sequences

The AMOVA results suggest that both SNP genotyping clusters and pangolin individuals contribute to observed variability in tick sequences, with geographic origin accounting for more of the observed variation (p-values < 0.05) (see Table 3).

Table 3.

Analysis of molecular variance (AMOVA) of sources of variation in A. compressum tick 16SrRNA sequences (SSD – Sum of Squared Deviations; MSD – Mean Squared Deviations; df – degree of freedom)

Source of variation	SSD	MSD	df	sigma 2	p-value
SNP genotyping cluster	7.08E-03	3.54E-03	2	6.20E-04	< 0.001
Pangolin (scales)	1.03E-02	4.31E-04	24	3.79E-04	< 0.001
Error	2.43E-05	8.11E-06	3	8.12E-06
Total	1.74E-02	6.01E-04	29

A. compressum genetic distance at population and individual level based on 16SrRNA

From the PCoA plot, PCoA1 and PCoA2 accounted for 96.4% and 1.2% of the total variability in tick sequences, respectively. Tick sequences from pangolins assigned to the CAR_CON_GAB_DRC SNP genotyping cluster grouped with tick sequences from the CMR_GAB SNP genotyping cluster. This cluster (orange on the PCoA plot) was distinct from tick sequences obtained from pangolins assigned to the Nigeria SNP genotyping cluster (Fig. 3).

Fig. 3.

Principal Coordinates Analysis (PCoA) of genetic distance of 30 A. compressum tick sequences

A. compressum tick haplotype mapping based on inferred coordinates from host estimated geo-referenced location assignments

The pangolin scales were further assigned to 13 estimated geo-referenced locations in the 3 white-bellied pangolin SNP genotyping clusters; these coordinate estimates were used to map tick haplotypes (Fig. 4). Ten of the 13 coordinates (77%) were estimated in the Cameroon_Gabon SNP genotyping cluster while the other coordinate estimates were assigned to the Nigeria and Congo basin genotyping clusters [2/13 (15%) and 1/13 (8%), respectively].

Fig. 4.

Tick haplotypes mapped using coordinates inferred from corresponding P. tricuspis scales

Two or more tick haplotypes were mapped at 5/13 host estimated geo-referenced locations in the CMR_GAB (4/13) and Nigeria (1/13) SNP genotyping clusters. The highest number of haplotypes recorded at one geo-referenced location was 4. These haplotypes, H1, H4, H5 and H9, were associated with 8 ticks, removed from 7 pangolin scales, across 3 different seizures (D, G and KBG), in the Cameroon_Gabon SNP genotyping cluster. Tick haplotypes H1, H2 and H4 were mapped at another location in the Cameroon_Gabon SNP genotyping cluster, based on estimated geo-referenced location data obtained for 2 pangolin scales removed from the same seizure bag. In this instance different haplotypes were all identified in the 3 ticks removed from the scales (n = 2 ticks, from one scale and n = 1 from tick from second scale).

Two haplotype pairs - H4_H6 and H2_H3 - were assigned to 2 separate locations, also within the CMR_GAB SNP genotyping cluster. For the H4_H6 haplotype pair, H4 was assigned to 2 tick sequences, one each from 2 seizure bags (G AND KEF) and H6 was assigned to 1 tick sequence, also from seizure bag G, but removed from a different scale. Haplotypes H2 and H3 were assigned to ticks removed from scales (n = 2 ticks; 1 tick removed from each scale) from 2 seizure bags (D and G) and assigned the same estimated geo-location. Two haplotypes, H10 and H11, were identified at one location in the Nigeria SNP genotyping cluster (3 ticks; 2 seizures).

Discussion

Our study suggests concordance between trafficked, white-bellied pangolin SNP genotyping cluster assignments (geographic origins) and the haplotype diversity of their tick parasites. By estimating the geographic origins of tick-infested pangolin scales [7], we were able to generate results suggesting that genetic variability of A. compressum ticks corresponds well to the SNP genotyping clusters of their pangolin hosts. Importantly, there was no such association between the genetic diversity of ticks and reported pangolin scale seizure origins.

Preliminary support for correlation of A. compressum genetic diversity with P. tricuspis SNP genotyping clusters

The results highlight the value of using geographic origins of highly traded species, or their parasites, determined by molecular or genomic studies, over reported host seizure origin information. Based on seizure records, most of the pangolin scales were identified as coming from Nigeria (6/8 seizures were labelled as coming from Nigeria and 2/8 from unknown locations). However, through the SNP genotyping data, we were able to assign pangolin scales to three of the five genetic populations of origin identified by Tinsman et al. [7], namely the Congo Basin, Cameroon_Gabon and Nigeria SNP genotyping clusters; none of the scales were assigned to the Cameroon_Manyu nor West Africa SNP genotyping clusters. In a related study, save for the Cameroon_Manyu SNP genotyping cluster, all other pangolin genetic lineages identified by Tinsman et al. [7] were confirmed, the authors suggesting that the Cameroon_Manyu lineage belongs to the Cameroon_Gabon lineage [4].

Most scales were assigned to the Cameroon_Gabon SNP genotyping cluster rather than the Nigeria SNP genotyping cluster (which would be related to seizure shipment information). As such, tick haplotypes correlated with pangolin SNP genotyping assignments and not reported seizure origin data. In another wildlife-related study, researchers utilized DNA to determine the actual origin of elephant ivory, by comparing unknown DNA sequences to reference sequences of known origins, dismissing seizure data origins [43]. Similarly, incorporating DNA in the pangolin-related study by Tinsman et al. [7] proved useful for determining SNP genotyping cluster assignments of unknown scales by comparing generated sequences to known reference sequences, indicating that ticks can provide geographic localization information – they form part of the initial story as to where harvesting is occurring – providing crucial information that is not possible from traditional seizure records or law enforcement intelligence.

By using information on host genotyping by population (in this case, pangolins), it is possible to make inferences into tick ecology and population dynamics. Through the generation of estimated coordinates for individual pangolin scales, we were able to map haplotypes of associated ticks, providing baseline information for understanding the dynamics of A. compressum ticks. For example, pangolin ticks are considered host specific, thus, co-endangered with their hosts; host extinction could result in co-extinction of these tick species [23] and associated symbionts, some of which are yet to be identified [44]. Given the unknown ecological consequences of ectoparasites going extinct with their hosts, conservation strategies, of which some methods have already been outlined [45, 46], must be implemented to prevent further complications. Since host specificity is an important parasitic trait, additional field surveys and studies should be conducted to provide more robust inferences on the consequences of host endangerment and/or extinction for tick ecology, population dynamics, and conservation status.

Genetic variability of A. compressum ticks using the 16SrRNA barcode region

Eleven haplotypes were identified from 30 tick sequences, 295 b.p. in length, using the mitochondrial DNA 16SrRNA barcode region. Both haplotype and nucleotide diversity were high for tick sequences from the Cameroon_Gabon SNP genotyping cluster. High haplotype and nucleotide diversity is indicative of stable populations and possible presence of multiple tick lineages with many nucleotide differences between them. However, calculations for both genetic variability measures were based on relatively short sequences generated from the 16SrRNA mitochondrial DNA (mtDNA) marker; mtDNA markers, while good for species identification, are not the best options for genetic variability studies due to sensitivity to sample size [47]. While our results suggest genetic variability among individuals and between populations, with the small sample size coupled with sampling imbalance, we were unable to make definitive conclusions regarding genetic variability in A. compressum ticks. For future studies, we suggest incorporating multi-locus and/or utilizing more robust molecular approaches such as whole genome sequencing to effectively capture variability within individuals from different populations.

Biogeography of A. compressum ticks based on P. tricuspis geographic localization

The Nigeria tick haplotypes formed a distinct cluster from Cameroon_Gabon and Congo Basin SNP genotyping clusters on both haplotype networks and PCoA plot and hierarchical AMOVA results suggest that white-bellied pangolin populations are more influential in determining tick genetic variability. These results are indicative of spatial structuring of A. compressum.

The spatial structure observed between the Nigeria and Cameroon_Gabon SNP genotyping clusters could be due to physical geographic barriers, such as river systems, forest fragmentation and mountain ranges, and climatic conditions, causing isolation of both white-bellied pangolins and A. compressum tick populations and ultimately resulting in the genetic differentiation observed in the study. Gu et al. [4] indicated that the Cameroon_Gabon SNP genotyping cluster is separated from the sister group of Nigeria and West Africa genetic clusters due to the presence of Niger River which is a biogeographic barrier; further distinction of the sister group may be due to the presence of the Volta River. Bernáthová et al., 2024 [27] documented distinct mitochondrial lineages of white-bellied pangolins in the Congo Basin possibly due to a combination of river barriers and forest fragmentation preventing gene flow between pangolin host populations. While the Nigeria SNP genotyping cluster haplotypes (H10 and H11) can be explained by clear geographic differences, H9 being present in the CMR_Gabon SNP genotyping cluster but genetically closer to H10 and H11 (both separated from H9 by one mutation each) may be due to recent divergence (estimated number of mutations between the 2 sequences based on calculated pairwise distance was 1).

Most tick haplotypes recorded in this study were identified in the Cameroon_Gabon SNP genotyping cluster, the population with the greatest number of scales and ticks. Possible explanations for variation within this cluster include, but are not limited to, the presence of river systems, such as the Sanaga river which is thought to be a driver of genetic diversity, distinct ecological zones in countries such as Cameroon [48] which have distinct habitats and climatic patterns [49], host range and movement. Haplotype H4 was observed in 8 ticks across 4 estimated locations in the Cameroon_Gabon SNP genotyping cluster. Having the highest frequency and being present in different geographical areas suggest that H4 is possibly dominant and/or stable within that population. Haplotype H9 was the second most frequent haplotype (n = 5) and could possibly be another high frequency haplotype in the region.

Limitations and recommendations

Pangolins are CITES listed and a major hurdle for research is sample acquisition. While the weight of pangolin scale shipments can sometimes weigh more than a tonne when intercepted, sample acquisition from these seized shipments can be challenging. Further, seizure samples are usually highly degraded due to how they are transported and stored [50, 51] and while scales were numerous within the seizures, not all the scales had associated ticks. Consequently, our tick sample size was relatively limited. Therefore, the sample bags do not fully represent entire pangolin scale shipments, nor do they fully capture tick and/or pangolin diversity.

Most scales and ticks were from the Cameroon_Gabon SNP genotyping cluster, so the Congo Basin and Nigeria SNP genotyping clusters were not well- represented. Since haplotype networks are reflective of sample sizes, under-sampling and imbalanced sample sizes may result in false inferences of genetic variability, particularly when numerous rare haplotypes (singletons) are generated. Given our small number of samples and imbalanced sample sizes, we were unable to effectively compare haplotype diversity among individual ticks, within the 3 white-bellied pangolin populations. A larger tick sample size with greater pangolin diversity will clarify our findings on pangolin-tick associations.

Sequence typing was the only method used for tick species identification and determining intra-species variation. While this method is useful for population genetic studies to address broad-scale temporal or geographic questions, highly conserved pre-selected loci are normally used (16SrRNA in our study) and so variation in other parts of genome may be overlooked [52]. To avoid this bias, genome-wide approaches would reveal finer resolution of tick genetic diversity to pangolin populations (but note again sample quality limitations). Reference sequences for A. compressum in Genbank are limited – at the time of this study, most of the sequences in Genbank were partial 16SrRNA sequences of unknown genetic population origins, so they were not included in our analyses. Despite the limited samples in Genbank, it may be easier to fill geographic gaps through tick sampling, than to sample CITES–listed pangolins across their respective ranges.

Conclusions

This study highlights the potential value of leveraging ectoparasites to gain additional insights into the geographic localization of host species and the genetic diversity of their parasites. Genetic data from pangolin-associated ticks potentially correlates with SNP genotyping clusters and estimated geographic origins (within respective clusters) of their hosts. With the incorporation of larger tick datasets, utilization of more robust molecular techniques and availability of higher-resolution genetic markers, we can make more definitive conclusions pangolin-ticks genetic variability and subsequently, implement ectoparasite-based tracing for trafficked wildlife. Beyond this, these “hitchhiking” ticks offer a valuable lens into potential zoonotic risks associated with the illegal wildlife trade. Ticks contribute to pangolin mortality through pathogen transmission [19–21] and frequent human contact with pangolins further elevates the relevance of these risks. The distinctiveness of the Nigeria tick haplotypes and the possible existence of high frequency haplotypes in the Cameroon_Gabon SNP genotyping cluster warrants further investigations into vector capacity within and between white-bellied pangolin SNP genotyping clusters.

Targeted studies examining tick haplotypes, vector capacity, and pathogen diversity across these SNP genotyping clusters would be particularly informative. Amblyomma compressum ticks appear to be genetically diverse, and this diversity seems to correlate with assigned, white-bellied pangolin genetic clusters, but notably not with reported seizure origins. Additionally, pangolin SNP genotyping cluster assignments in this study replicate the findings of Tinsman et al. [7] and expand current knowledge by identifying additional likely harvesting locations. More broadly, incorporating ectoparasites into population genetic studies of rare or elusive species—particularly heavily trafficked, CITES–listed taxa—can alleviate challenges associated with direct sampling. This approach offers a more rapid and efficient means of identifying trafficking routes, inferring host geographic origins, and assessing zoonotic risks based on genetic clustering, ultimately strengthening conservation efforts for endangered species.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AFCD

Agriculture, Fisheries and Conservation Department

BLAST

Basic Local Alignment Search Tool

CAR_CON_GAB_DRC

Congo Basin SNP genotyping cluster

CITES

Convention on International Trade in Endangered Species of Fauna and Flora

CMR_GAB

Cameroon_Gabon SNP genotyping cluster

CO1

Cytochrome c oxidase subunit I

CPOS

Centre for PanorOmic Sciences

CYTB

Cytochrome B

DNA

Deoxyribonucleic acid

ETOH

Ethanol

IUCN

International Union for Conservation of Nature

IWT

Illegal Wildlife Trade

KFBG

Kadoorie Farm and Botanic Garden

mtDNA

Mitochondrial DNA

NACLO

Sodium hypochlorite

NTC

No template

PCR

Polymerase Chain Reaction

QGIS

Quantum Geographic Information System

SNP

Single Nucleotide Polymorphism

STA

Specific target Amplification

TCM

Traditional Chinese Medicine

16SrRNA

16 S ribosomal ribonucleic acid

Author contributions

Conceptualization–TAB; TCB. Methodology–TAB; TCB; TLP; TBS; JT. Funding acquisition: TCB. Data curation–TAB; Formal analysis–TAB; JT. Writing (Original draft)–TAB; TCB. Writing (Review and editing)–all authors. Resources–GA; JT; LRJ; TAB; TCB; TLP; Project administration–TAB; TCB; Visualization –TAB; JT. Supervision – TCB.

Funding

This research was supported by the Research Impact Fund (R7021-20), University Grants Committee, Hong Kong Special Administrative Region (HKSAR).

Data availability

Tick barcoding sequences are available in the NCBI Genbank database (assigned accession numbers: PZPZ306247–PZ306276). Additional data are available in the supplementary data file.

Declarations

Ethics approval and consent to participate

All animal samples collected in this study were derived from seizures of (long dead) animal products within the wildlife trade. No live animals were involved in this research. Permission to use pangolin scales for the purposes of this research was granted by the Agriculture, Fisheries and Conservation Department of the Hong Kong Government.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.