ArticlePDF Available

Validation of methods for extraction of DNA and species identification from seized Helmeted Hornbill (Rhinoplax vigil) casques


Abstract and Figures

Helmeted hornbills (Rhinoplax vigil, J.R. Forster, 1781) are ‘Critically Endangered’ due to illegal hunting for their casques which are carved and traded for ornamental purposes. DNA species identification techniques can aid enforcement efforts, and validated wildlife forensic techniques for the species identification of R. vigil are needed. Here we tested multiple methods for sampling and extracting DNA from R. vigil casques and a validated a previously published assay using cytochrome B (cytB) primers to identity species and origin of traded casques. Phenol-chloroform: isoamyl alcohol extractions resulted in samples with higher quantity and quality of DNA than those extracted using the commercial Qiagen DNeasy blood and tissue kit. Samples collected from the caudal side of the casque yielded higher DNA quantity and quality than rostral and lateral sides, regardless of sampling method. We then assessed the repeatability, reproducibility, robustness, sensitivity, specificity, and phylogenetic resolution of a previously published species identification assay. We confirm the ability of this method to phylogenetically distinguish between R. vigil and closely related hornbills with high bootstrap support (99%). We also report the first genetic evidence of illegally traded R. vigil in Hong Kong using confiscated casques and provide more reference samples of R. vigil for future work. Overall, we provide multiple protocols for sampling and extracting DNA, and a validated species identification assay for amplifying DNA from R. vigil casques with potential to aid law enforcement in illegal wildlife crimes.
Content may be subject to copyright.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
Please cite this article as: Chloe E.R. Hatten, Forensic Science International: Animals and Environments,
Available online 5 December 2022
2666-9374/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (
DNA analysis and validation for species identication of seized helmeted
hornbill (Rhinoplax vigil) casques
Chloe E.R. Hatten
, Yuli S. Fitriana
, Tracey-Leigh Prigge
, Mohammad Irham
Hari Sutrisno
, Abinawanto
, Caroline Dingle
School of Biological Sciences, The University of Hong Kong, Hong Kong Special Administrative Region of China
Museum Zoologicum Bogoriense, National Research and Innovation Agency (BRIN), Indonesia
Department of Biology, Faculty of Mathematic and Natural Science, The University of Indonesia, Indonesia
Helmeted hornbill
Species identication
Wildlife forensic science
Helmeted hornbills (Rhinoplax vigil, J.R. Forster, 1781) are ‘Critically Endangereddue to illegal hunting for their
casques which are carved and traded for ornamental purposes. DNA species identication techniques can aid
enforcement efforts, and validated wildlife forensic techniques for the species identication of R. vigil are needed.
Here we tested multiple methods for sampling and extracting DNA from R. vigil casques and a validated a pre-
viously published assay using cytochrome B (cytB) primers to identity species and origin of traded casques.
Phenol-chloroform: isoamyl alcohol extractions resulted in samples with higher quantity and quality of DNA than
those extracted using the commercial Qiagen DNeasy Blood and Tissue kit. Samples collected from the caudal
side of the casque yielded higher DNA quantity and quality than rostral and lateral sides, regardless of sampling
method. We then assessed the repeatability, reproducibility, robustness, sensitivity, specicity, and phylogenetic
resolution of a previously published species identication assay. We conrm the ability of this method to
phylogenetically distinguish between R. vigil and closely related hornbills with high bootstrap support (99%). We
also report the rst genetic evidence of illegally traded R. vigil in Hong Kong using conscated casques and
provide more reference samples of R. vigil for future work. Overall, we provide multiple protocols for sampling
and extracting DNA, and a validated species identication assay for amplifying DNA from R. vigil casques with
potential to aid law enforcement in illegal wildlife crimes.
1. Introduction
Helmeted hornbills (Rhinoplax vigil, J.R. Forster, 1781) have been
listed as ‘Critically Endangered on the IUCN Red List of Threatened
Species since 2015. This species is at high risk of extinction across its
Southeast Asian range in Thailand, Myanmar, Malaysia, Indonesia, and
Brunei due to illegal hunting to supply trade of the birds ‘casque [1].
Unlike other hornbill species, the casque of R. vigil is solid, consisting of
internal cancellous bone and an external solid layer, which can be
carved similarly to elephant ivory [2,3]. This anatomical structure
functions as a defence organ in the species aerial-jousting combat
behaviour, as well as other high-tension impacts [4,5].
International trade for R. vigil casques and products has been regu-
lated under Appendix I of the Convention on International Trade in
Endangered Species of Wild Fauna and Flora (CITES) since 1975 [1], yet
illegal trade continues. On the black market, these casques are traded to
be carved into ornaments and jewellery that often fetch a higher price
than elephant ivory per gram [6]. National use and trade in R. vigil
casques for ornamental purposes among traditional native Southeast
Asian cultures is not a new phenomenon [7]. However, international
demand has escalated in recent years with nearly 6000 individual
casques and skulls seized by authorities across Asia between 2011 and
2021, with most seizures occurring in Indonesia and China, including
Hong Kong [1,8,9]. This relatively recent upsurge in the illegal trade of
this species has drawn the attention of both national and international
enforcement bodies, with one recent smuggling case in Hong Kong
resulting in a penalty of 32 monthsimprisonment; the longest prison
sentence ever recorded for wildlife trafcking in Hong Kong [10].
Conservation strategies and action plans across Southeast Asia have
been established, highlighting the necessity for the development and use
* Correspondence to: School of Biological Sciences, Kadoorie Biological Sciences Building, The University of Hong Kong, Pokfulam Road, Hong Kong Special
Administrative Region of China.
E-mail address: (C. Dingle).
Contents lists available at ScienceDirect
Forensic Science International: Animals and Environments
journal homepage:
Received 17 May 2022; Received in revised form 4 November 2022; Accepted 4 December 2022
Forensic Science International: Animals and Environments xxx (xxxx) xxx
of tools, including molecular forensic analyses, to prevent further spe-
cies decline [11,12].
For jurisdictions adhering to CITES, conrming the species identity
of traded products is essential for determining if a crime has been
committed. Helmeted hornbill casques are morphologically distinct
from other hornbills [13]. However, casques and casque products can be
traded and seized raw or worked [14], and although fake and real
casque material can be distinguished using microscopy [15], genetic
analysis to conrm the species as R. vigil may still be required for
prosecution. DNA species identication techniques have been applied to
wildlife crime investigations across many different types of traded
wildlife species and their parts and products [1619]. Traded wildlife
products such as skins, horn, fur, and feathers are often processed,
treated, or exposed to oxidative damage. Therefore, genetic analyses on
these wildlife parts and products are often limited by the lack of good
quality DNA. As many different types of potentially illegal wildlife parts
and products appear across seizures and markets, the most suitable ge-
netic methods vary by species, parts, or structural components of the
tissue in question [20]. This means that tests must often be tailored for
the species and part/product of concern [21].
Mitochondrial DNA (mtDNA) markers such as cytochrome oxidase
subunit 1 (COI) or cytochrome B (cytB) are commonly used to identify
species due to the low intraspecic variation and high interspecic
variation of these genes [22]. In their recent paper, Ouitavon et al. [23]
demonstrated that DNA recovery from R. vigil casques was possible and
developed primers targeting an 800-bp region of cytB to identify the
species from R. vigil casques. They showed that this mtDNA region is
informative for species identication, key for investigative purposes. In
their paper, the authors stressed that the next steps should be to include
more samples from across the speciesrange and to validate the methods
for application to forensic casework. Validation is an important part of
ensuring a test is reliable and consistent for wildlife forensic analysis
[24]. Validating the genetic marker of a DNA species identication assay
encourages uniformity and enables the sharing and co-analysis of data
within and between laboratories to aid transnational investigations
[18]. Validation steps to test the limits, robustness, repeatability,
reproducibility, sensitivity, and specicity of the assay in question are
important for improving the level of standardisation [18,2527].
Although previous studies have applied genetic methods to R. vigil [23,
2830], there are no validated tests currently available for potential use
in wildlife forensic cases.
Building on the Ouitavon et al. [23] results, our study had two main
objectives. First we aimed to optimise methods of sampling and
extracting DNA from R. vigil casques by determining whether DNA
quantity and quality, PCR amplication, and sequencing was affected by
1) extraction method, 2) sampling technique, and 3) sampling location
along the casque. Second, using the primers from the Ouitavon et al.
[23] study, which amplify a gene region commonly used for specic
identication [31], we aimed to validate the species identication assay
for potential wildlife forensic application. Ultimately, we aim to build on
previous work to provide validated and optimised methods for DNA
extraction of mtDNA from R. vigil casques and sequencing for species
identication, which is critical for wildlife forensics applications.
2. Material and methods
2.1. Sample materials
For this study, we sampled a total of 13 casques for DNA analysis.
These casques were obtained from Hong Kong SAR and Indonesia from
seizures and museum specimens (Table 1). R. vigil is the only species in
its genus and is readily identiable using morphological characteristics
of its unique solid casque [3,13,27]. R vigil has a solid cylindrical casque
of 1 2-inch thickness, which Manger Cats-Kuenen [3] describes as a
thick horn layer. This part is typically called the casque in more
recent literature, identifying it as the part which is typically carved [2,
14,32,33] (Fig. 1). The rest of the casque has a brick-red covering of
around 2 4 mm in thickness. Additionally, the R. vigil casque does not
extend beyond the plane of attachment to the upper mandible. This is
unlike other hornbill casques, including those of closely related species
to R. vigil in the Buceros genus, which extend rostrally over the beak [3].
These morphological characteristics were used by experts to conrm the
identity of all casques used in this study.
2.2. Experimental workow & sampling techniques
We sampled the six Hong Kong casques (identications: HK-01 HK-
06) to determine whether DNA quantity and quality was affected by:
extraction method (Test 1), sampling method (Test 2), or location on the
Table 1
Selected case information on suspected R. vigil casques seized in Hong Kong SAR
used in this study, and specimen information about R. vigil casques tested in the
National Research and Innovation Agency (BRIN), Indonesia, used as part of the
validation study. Casques HK-01 to HK-06 were seized by the Customs and
Excise Department and accessed through the Agriculture, Fisheries, Conserva-
tion Department of the Hong Kong SAR. Casques ID-01 to ID-04 were seized by
the Directorate General for Law Enforcement, Ministry of Environment and
Forestry of The Republic of Indonesia. Casques ID-05 to ID-07 were museum
specimens held at the Museum Zoologicum Bogoriense, Indonesia.
Specimen ID Origin/seizure/export
Collection Date
Rhinoplax vigil
Indonesia 2016 Seizure
Rhinoplax vigil
Malaysia 2012 Seizure
Rhinoplax vigil
Malaysia 2012 Seizure
Rhinoplax vigil
Malaysia 2012 Seizure
Rhinoplax vigil
Indonesia 2016 Seizure
Rhinoplax vigil
Indonesia 2016 Seizure
Rhinoplax vigil
West Kalimantan, Indonesia 2016 Seizure
Rhinoplax vigil
West Kalimantan, Indonesia 2016 Seizure
Rhinoplax vigil
West Kalimantan, Indonesia 2016 Seizure
Rhinoplax vigil
West Kalimantan, Indonesia 2016 Seizure
Rhinoplax vigil
Kalimantan, Indonesia 1939 Museum
Rhinoplax vigil
Kalimantan, Indonesia 1939 Museum
Rhinoplax vigil
Kandang Ampat, West
Sumatra, Indonesia
1932 Museum
Fig. 1. Helmeted hornbill (Rhinoplax vigil) casque showing the three sampling
locations (caudal, lateral, rostral) tested in this study. We collected samples at
each of the three regions using two sampling methods (drilling and shaving).
See main text for details.
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
casque from which the sample was collected (‘sampling location, Test
3). To collect casque material for Test 1, we drilled the caudal side of the
six casques for each extraction method (total n=12) using a Dremel®
4000 drill with a high-speed cutter engraving bit at 10,00015,000 RPM
[17, 22, 25]. Prior to sampling, we rst cleaned each casque sampling
area with ethanol and sterilised water and then placed the cleaned
casque on a metal tray which was sterilised with absolute ethanol and
bleach between each sample collection. At least 2 mm of the surface
layer was removed and discarded to avoid sampling any outer material
which may be contaminated with other DNA (human or otherwise) and
then 30 mg of powdered material was collected from each casque. Clean
paper was placed beneath the sampling locations on each casque to
collect and transfer the material to separate autoclaved 1.5 mL Eppen-
dorf tubes. The paper was changed between individual sample collec-
tions to avoid cross-contamination. Samples were then extracted by two
commonly used DNA extraction methods to determine which method
produced the highest yield and quality of mtDNA (see Section 2.3. DNA
Extraction section below for details). The extraction method with the
highest DNA concentration was then used to extract samples tested
under Tests 2 and 3.
For Test 2, we tested two sampling methods: the ‘drillingmethod as
described above, and a second ‘shavingmethod. Drilling provides an
efcient method to obtain homogenised material deeper into the casque,
therefore obtaining potentially better-quality DNA [34]. However, using
a scalpel presents a relatively cheap and simple method for collecting
material that is less invasive than drilling if wanting to preserve the
casque (e.g., for a museum collection). Furthermore, drilling at a high
power has the potential to heat up the material and cause damage to the
DNA. Shaved casque material was obtained using a sterile scalpel to rst
shave and discard at least 2 mm of the surface layer casque material.
30 mg of material was collected from each casque. Procedures for col-
lecting casque material after sampling and to avoid contamination fol-
lowed methods above. The drilling method was performed as described
To determine whether samples collected from different locations on
the casques yielded different quantity or quality DNA (Test 3), we
collected samples from three different locations of the casques. Casques
are often seized detached from the skull. As the caudal side of the casque
attaches to the living tissue of the birds skull [3], we hypothesised that
DNA concentration would be greater in this area than across other
sampling locations. However, casques are sometimes seized with their
skulls attached, and sampling this region may not be possible. To test
whether sampling from different locations on the casque yields higher
quantity/quality DNA, we collected samples from the caudal, lateral and
rostral sides of each casque (Fig. 1). Samples were collected from each of
the three locations using both the shaving and the drilling methods on
all six casques, meaning the total sample size was 36 (6 casques x 3
locations x 2 sampling methods) for each of Test 2 and Test 3.
2.3. DNA extraction
We tested two DNA extraction methods on six casques (HK-01 HK-
06) to determine which method yielded the highest quantity of DNA: a
commercial extraction kit (Qiagen DNeasy Blood & Tissue Kit; ‘QBT)
and a phenol-chloroform: isoamyl alcohol (‘PCIA) method. One nega-
tive control was used per extraction run. Samples were extracted using
the commercial kit following the manufacturers protocol, but volumes
of lysis buffer were doubled during the initial lysis stage to accommo-
date for the 30 mg of casque material. The lysis buffer included 10
freshly made 1 M dithiothreitol (DTT) and 0.4 mg proteinase K (20
L of
20 mg/mL), for improved cell lysis [35,36]. Samples were incubated
overnight for a 12-hr period at 56 C. An additional wash with 400
absolute ethanol was added to remove additional salts and help pre-
cipitate the DNA, prior to eluting the DNA in a nal elution of 75
To extract DNA using PCIA, we adapted a protocol from Coffroth
et al. [37]. Samples were initially incubated in 0.6 mL CTAB buffer with
0.4 mg proteinase K and 10
L freshly made 1 M DTT for 12 hrs at 65 C.
Chloroform: isoamyl alcohol (CIA, 24:1, 0.6 mL) was added, vortexed,
and centrifuged at 13 K RCF for 7 mins. The aqueous layer in each tube
was then transferred to a clean tube. Phenol-chloroform: isoamyl
alcohol (PCIA, 25:24:1, 0.6 mL) was used to extract the DNA and
centrifuged at 13 K RCF for 7 mins. The nucleic acid-containing aqueous
stage was then transferred again into a clean tube. Ice-cold 95% EtOH
(1 mL) was added to precipitate the DNA at 80 C for 1 hr. After
centrifugation at 13 K RCF for 30 mins, each pellet was washed twice
with 0.5 mL 70% ETOH, dried in a speed-vac (Labconco CentriVap®
Benchtop Vacuum Concentrator) and resuspended in 30
L of 1 x TE
buffer (I0 mM Tris-HC1 pH 8.0 and 1 mM EDTA).
2.4. DNA quantity & quality
DNA sampling and extraction methods were evaluated using four
criteria to determine extraction success: DNA quantity (concentration;
L) and DNA quality (DNA purity (nm)), DNA fragmentation
(visualised on agarose gels), and PCR amplication of the cytB target
region). Three samples from each casque were sequenced to test the
effectiveness of the primers on the individual specimens. DNA concen-
tration and purity was assessed using a UVVIS Nanodrop spectropho-
tometer 2000 (Thermo Scientic). This standard DNA quantication
method uses ultraviolet (UV) spectrophotometry to measure the absor-
bance of the sample at specic wavelengths. DNA purity was determined
by the ratios of absorbance at 260, 280 and 230 nm. A260 (absorbance
at 260 nm) measures the length of the nucleic acid present, A280
(absorbance at 280 nm) measures proteins, and A230 (absorbance at
230 nm) measures chemical contaminants such as salts, which may
contaminate the sample and inhibit the PCR amplication process.
A260/A280 and A260/A230 ratios were calculated to indicate purity of
the sample, with optimum ratios in the range of 1.8 2.0 and 1.4 2.0
respectively [38,39]. Extracted DNA was run through a 2% agarose gel
in SB buffer at 120 V for 30 mins to visually detect the presence of DNA
fragmentation. MtDNA was then amplied by PCR using cytB primers.
In order to determine whether the different extraction and sampling
methods yielded sufcient DNA for PCR amplication and sequencing,
we used the published primers (hhcytb800F and hhcytb800R) from [23]
to sequence an 800-bp fragment of the mtDNA cytB gene from the DNA
extracted from casques HK-01 HK-06. Polymerase chain reaction
(PCR) was conducted in 20
L reactions containing 1 x Biotechrabbit
LyoHotStart 2x PCR mastermix, a hotstart mastermix we have optimised
to work successfully in our laboratory extracting DNA from wildlife
parts and products, 0.3
mol/L primer, 1.5 mg/mL BSA, and 20 50 ng
of DNA sample, using the Applied Biosystems Veriti 96-Well Thermal
Cycler. PCR amplication for the cytB primers was conducted with an
initial denaturing step at 95 C for 2 mins followed by 30 cycles at 95 C
for 30 s, 55 C annealing temperature for 30 s, 72 C for 1 min, and a
nal extension step at 72 C for 4 mins. The activation, cycle number,
and nal extension of our PCR protocol differed from those of [23] to
accommodate our hotstart PCR mastermix. PCR amplicons were sepa-
rated by electrophoresis through a 2% agarose gel in 1 x TAE buffer at
80 V for 120 mins. Laboratory work was conducted in separate pre- and
post-PCR workspaces using dedicated reagents and equipment, and one
PCR negative control was included per PCR run. A positive control was
included using DNA from a blood sample of Rhyticeros undulatus (Ap-
pendix I in Supplementary Material), a species of hornbill not closely
related to R. vigil [29]. This sample was collected as part of a routine
veterinary procedure conducted by the Ocean Park Corporation, Hong
Successfully amplied PCR products were sent to the Centre for
PanorOmic Sciences (CPOS) at the University of Hong Kong for Sanger
sequencing on an Applied Biosystems (ABI) 3730xl DNA Analyzer.
Amplicons were sequenced using both forward and reverse primers.
DNA sequences and base-pair ambiguities were visually edited and
examined on Geneious 10.2.3 Bioinformatics Software (https://www.
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
4 and compared with published references through the
NCBI Basic Local Alignment Search Tool (BLAST) application on Gene-
ious to conrm species identi-
cation. Multiple alignments with the obtained sequences and published
references were conducted on Geneious.
2.5. Statistical analyses
To check if DNA concentrations extracted from the R. vigil casques
were normally distributed, Shapiro-Wilk tests were run, and the ho-
mogeneity of variance across the studied groups was veried with
Levenes test. As all DNA concentrations were not normally distributed
(Shapiro-Wilk <0.05), non-parametric analyses were used to determine
any signicant differences between samples. Mann-Whitney U tests
were applied to compare DNA concentrations across different extraction
methods and sampling techniques, and Kruskal-Wallis rank sum tests to
compare DNA concentrations across different sampling locations. All
data preparation, statistical analyses, and gure making was performed
in R version 1.4.1106 [40].
2.6. Species identication test validation
The species identication assay (following [27]) was validated based
on the following characteristics: repeatability, reproducibility, robust-
ness, sensitivity, specicity, and phylogenetic resolution. To test the
repeatability and reproducibility of the assay, the PCR assay was per-
formed by a different analyst using the same PCR reagents and condi-
tions on the same six Hong Kong casque samples. The hhcytb800F/R
primers were also tested on seven R. vigil casque samples at BRIN lab-
oratory, Indonesia (identications: Rhinoplax vigil ID-01 ID-07).
Conscated casques in Indonesia were sampled using PCIA with the
caudal-drill sampling method, and whole museum specimens were
sampled using the rostral-drill method. PCR conditions and reagents
were the same as those used for the six Hong Kong samples.
If a PCR thermocycler is not correctly calibrated, temperatures inside
the machine could vary from the programmed temperatures. To test the
robustness of the PCR protocol to variabilities in temperature between
PCR machines, we tested the PCR protocol using three different
annealing temperatures (the standard 55 C and ±1.5 C). We per-
formed this analysis on three different PCR thermocyclers in the same
lab, using samples from three Hong Kong casques in each run (3 ma-
chines x 3 temperatures x 3 samples per run =27 amplications). Using
this setup, we were also able to test repeatability across different ma-
chines. Finally, we conducted PCR using a mastermix made up of
different reagents (Thermo Scientic Dreamtaq PCR Master Mix) on
three casques (1 per sample), to assess the robustness of the method to
changes in the PCR reaction mix.
The sensitivity of the assay (or ‘limit of detection, [25]) was tested
using different concentrations of DNA. We conducted this test using
three serial dilutions of DNA from samples collected from two Hong
Kong casques: 1 in 10 (1 ng), 1 in 100 (0.1 ng), and 1 in 1000 (0.01 ng),
quantied using the Qubit 4.0 Fluorometer.
The ability to conrm species identication depends on the amount
of variability between the target species and other closely related species
at the targeted DNA region [41]. Within-species sequence variation
should be lower than inter-specic variation at this marker (therefore
leading to a ‘barcoding gap), and this can be assessed by comparing the
target sequence to other sequences in a reference database [42]. We
therefore performed a BLAST search and calculated the sequence
divergence between the target sequences both to conspecic and the
rst two interspecc sequences in the database.
To test specicity of the cytB primers for R. vigil, DNA from known
blood, feather, or tissue samples representing closely related [29]
and/or potential contaminating taxa in our laboratory were amplied
from hornbill species: R. undulatus (from the Ocean Park Corporation,
Hong Kong), Buceros bicornis and Berenicornis comatus (from the Leisure
and Cultural Services Department of the Hong Kong SAR government)
and from a snake sample of Naja atra (from the Kadoorie Farm and
Botanic Garden) and Manis tricuspis (pangolin) scales (from the AFCD).
One human (Homo sapiens) buccal swab was also tested, representing a
common contaminant amongst laboratories (obtained with permission
from author T. Prigge). For these species, 30 ng of template DNA was
used in the PCRs. A mixed sample test was also performed to determine
whether the presence of human DNA, a common contaminant in any
laboratory, and pangolin DNA, a common contaminant in our own
laboratory, would cause any inhibition of the reaction or affect the
amplication of the target hornbill DNA in any way. This experiment
was conducted by mixing R. vigil DNA with human DNA, and separately
R. vigil DNA with pangolin DNA, at ratios 1:1, 1:5, 1:20. The PCR success
of all reactions were checked using agarose gel electrophoresis and
Sanger sequencing (as above).
2.7. Phylogenetic analysis
Ouitavon et al. [23] demonstrated that the cytB marker obtained was
able to distinguish between R. vigil and closely related hornbill species.
Their study primarily included samples from Thailand, with one
sequence from Sarawak (Borneo). We built on this result by conducting a
similar analysis using additional sequences from samples of known
origin in other parts of the species range (particularly through the
addition of samples known or presumed to have originated from Borneo)
and from seizures of unconrmed origin. We included sequences from
all 13 casques used in this study (6 seized in Hong Kong, 7 Indonesian
samples), seven R. vigil sequences from [23] generated using the same
primers, and two other R. vigil sequences from GenBank [28,29]
(Table 1). Ouitavon et al. [23] included ve other hornbill species in
their phylogeographic tree. We increased the sample size of sequences
from the two most closely related species to R. vigil (B. bicornis and
B. rhinoceros) to examine the potential for R. vigil to be distinguished
from a wider variety of sequences from closely related species as these
are likely to have the least differences from R. vigil. As a member of the
sister group to Bucerotiformes [43], we used a species of Hoopoe (Upupa
epops) in the order Upupiformes as an outgroup for the model, following
Viseshakul et al. [28]. Only published sequences from GenBank were
used (see Table S1). Using these sequences, we conducted a maximum
likelihood phylogenetic analysis in MEGA X software [44]. A heuristic
nearest-neighbour-interchange (NNI) approach was used with 1000
bootstrap replicates [18,45]. Based on the lowest AICc value output of
the model selection analysis in MEGA X, a Hasegawa-Kishino-Yano
(HKY) model was used [46] with gamma distributed evolutionary rates.
3. Results
DNA from samples collected from all 13 casques was successfully
extracted, amplied, and an 800-bp fragment of cytB sequenced (see
Fig. S1 for example gel image). All DNA concentrations and A260/280
and A260/230 values for the extraction and sampling tests on casques
HK-01 HK-06 are shown in Table S2. CytB sequences matched R. vigil
sequences in GenBank (99.5 99.9% identity) (accession no.:
GU257918, [28]), and the positive control had a 99.8% pairwise iden-
tity match with R. undulatus (accession no.: KJ456173, [47]).
3.1. Extraction method
DNA concentrations from samples extracted using PCIA were
signicantly higher than those extracted using the QBT kit (mean ±sd,
PCIA: 127.347 ±95.821 ng/
L, QBT: 9.625 ±10.849
L, Mann-
Whitney U: p <0.05, total n=12, Fig. 2). DNA recovered by PCIA
were also more pure (mean ±sd, A260/A280: 1.800 ±0.164, A260/
A230: 1.405 ±0.553) than those extracted by QBT (mean ±sd, A260/
A280: 0.927 ±0.248, A260/A230: 0.162 ±0.027). DNA smears were
visualised on an electrophoresis gel, showing DNA fragmentation across
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
all samples extracted using QBT and PCIA (Fig. S2).
3.2. Sampling method & sampling location
We obtained signicantly higher DNA concentrations from casque
samples when using the shaving method (mean ±sd: 189.191
±506.921 ng/
L) than the drilling method (mean ±sd: 92.876
±67.407 ng/
L, Kruskal-Wallis p <0.05, total n=36, Fig. 3A). There
was one sample for which we obtained an anomalously large DNA
concentration 2155.76 ng/
L from HK-04 (shaved, caudal). We re-
sampled and re-extracted from this specimen using the same method
which resulted in a DNA concentration of 3592.930 ng/
L, with A260/
A280 and A260/A230 ratios of 2.011 and 2.070, respectively. Removing
this sample from the analysis resulted in DNA concentrations from the
drilling method being signicantly higher (mean ±sd: as above) than
the shaving method (mean ±sd: 73.511 ±130.769 ng/
L, Mann-
Whitney U, p <0.05, B). However, due to the small sample size used
in this study, the signicance of these results should not be over inter-
preted. Ultimately, sufcient DNA was obtained from all sampling
methods tested to generate sequences useful for species identication.
Gel electrophoresis showed DNA fragmentation for all samples.
Including the sample with the anomalously high DNA concentration,
drilled samples were less contaminated (e.g., with proteins or phenols,
mean ±sd, A260/A280: 1.688 ±0.297, A260/A230: 1.174 ±0.503)
than the shaved samples (mean ±sd, A260/A280: 1.465 ±0.317,
A260/A230: 0.913 ±0.680). The anomalous DNA sample was among
the most pure of all samples analysed (A260/A280: 2.009, and A260/
A230: 2.071).
Including the anomalous sample, DNA concentrations in extractions
sampled from the caudal side of the casque were signicantly higher
than from the lateral and rostral sides (Kruskal-Wallis p =0.05, Fig. 4A),
although this was likely due to the large outlier from the shaving method
(Table S2). Removing this sample meant that there were no strong sig-
nicant differences between the sampling locations (Kruskal-Wallis
p=0.1, Fig. 4B).
3.3. Validation
For the repeatability and reproducibility tests for the sequencing
assay, DNA re-extracted from all six Hong Kong casques by a second
analyst, and from all seven casques in Indonesia, was successfully
amplied and sequenced. DNA amplication was also successful across
the three different PCR thermocyclers. These three tests demonstrate the
repeatability and reproducibility of the assay by different analysts and
different thermocyclers. In our test for robustness which involved testing
our PCR protocol using different annealing temperatures, PCR ampli-
cation was successful at all three annealing temperatures (55 C
±1.5 C) across all three thermocyclers with no qualitative differences
across the gels. DNA was successfully amplied and sequenced from
PCRs conducted with different reagents, showing robustness of the
method to changes in the PCR reaction.
In the sensitivity (‘limit of detection) test, both Hong Kong casques
were successfully amplied with DNA concentrations as low as 0.01 ng/
Fig. 2. Boxplot shows that DNA concentrations (ng/
L) of R. vigil casque
samples HK-01 HK-06 were signicantly higher when extracted via phenol-
chloroform: isoamyl alcohol (PCIA) than the Qiagen DNeasy Blood & Tissue
Kit (QBT). Dots show individual casque data, boxes represent the upper and
lower quartile ranges, boxplot lines show the median.
Fig. 3. Boxplots show that DNA concentrations (ng/
L) of R. vigil casque samples HK-01 HK-06 were signicantly higher when sampled using the shaving method
compared to the drilling method (p <0.05, A). However, when one sample with an anomalously high DNA concentration was removed from the analysis, DNA
concentrations were signicantly higher using the drilling method (p <0.05, B). Dots show individual casque data, boxes show the upper and lower quartile ranges,
boxplot lines show medians.
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
L. We were able to successfully sequence R. vigil PCR products from
extracts as low as 0.01 ng/
L; sequences had at least 99.5% pairwise
identity with accession no.: GU257918.
In the specicity test, we found that the primers also amplied
R. undulatus, B. bicornis and B. comatus successfully in addition to the
R. vigil samples (Table 2). The species amplied were veried through
high percent sequence similarity with the same species in Genbank (Top
match for each species: Rhyticeros undulatus (KJ456173, [47]); Buceros
bicornis (NC_038201, [48]); Berenicornis comatus (GU257917, [27]). The
snake N. atra and pangolin M. tricuspis DNA did not amplify with the
cytB primers. DNA from H. sapiens provided a faint amplication band
on the gel around 1000-bp, but this was not sequenced successfully.
Given that the primers we used were likely to amplify DNA from other
hornbill species, we calculated barcoding gaps between R. vigil and the
next to highest ranking species in our BLAST results (Table 2). We found
1) high percentage similarity between the R. vigil target sequence and
the other intra-specic sequences on GenBank (% similarity 96.7
99.9%), and 2) a barcoding gap of at least 9.9% between the top R. vigil
match and each of the next two highest ranking species matches, Buceros
bicornis and Aceros nipalensis (Table 2).
To further test the specicity of the assay, when mixing hornbill and
pangolin DNA, hornbill DNA was amplied and sequenced successfully
even when at a 1:20 ratio (hornbill DNA to pangolin DNA). When mixing
hornbill and human DNA, hornbill DNA was amplied and sequenced at
both a 1:1 and 1:5 ratio. PCR amplication was successful, but
sequencing failed at a 1:20 ratio (hornbill DNA to human DNA). This is
likely due to the low concentration of the PCR product that resulted from
this mix. These mixed tests show that hornbill DNA can be successfully
detected and sequenced when in quite low ratios to other common
contaminants in the laboratory, suggesting that these contaminants
would be unlikely to cause any inhibition of the reaction or affect the
amplication of the target hornbill DNA. No contamination was seen in
any of the extraction or PCR negative controls.
3.4. Phylogenetic analysis
In the phylogenetic analysis, all R. vigil sequences formed a distinct
clade from the other hornbill species (Fig. 5), with a bootstrap value of
99%. When compared to a previously published R. vigil sequence
(GU257918) [28], seized suspected R. vigil casques and museum speci-
mens from Hong Kong and Indonesia displayed percentage identity
matches ranging from 99.5% to 99.9%. Across all 22 R. vigil sequences,
we found 32 variable sites, with 3 being parsimony informative. Samples
from or suspected to be from Thailand and Sumatra (blue and green,
Fig. 5) formed a clade distinct from the clade including samples from or
suspected to be from the island of Borneo (red and yellow, Fig. 5), with
74% bootstrap support. Pair-wise distances between the Thailand/Su-
matra and the Borneo group on MEGA X, resulting in distance: 0.01, se:
0 within the two groups, and distance: 0.00698, se: 0.00245 between the
two groups. There were 29 variable sites within the Borneo group, none
of which were parsimony informative, and 1 variable site, which was
parsimony informative, within the Thailand/Sumatra group. There were
two diagnostic bases at sites 196 and 478 (Borneo: T, Thailand/Sumatra:
C at both sites).
4. Discussion
Forensic species identication of helmeted hornbill casques is
important for wildlife crime investigations. Of the most common case-
work questions wildlife forensic science often addresses [49], the most
important for R. vigil is the identication of the species for CITES
member states like Hong Kong and Indonesia, whereby a violation of
national law enacting CITES regulation is achieved through the smug-
gling of a CITES Appendix listed species. Here we build on previous work
by Ouitavon and colleagues [23] by optimising and validating a range of
methods to sample material from R. vigil casques, extract DNA. These
methods yield a fragment of the cytB region which can be used for
identication of species and origin for potential enforcement use. The
fragment length (800 bp) generated is longer than commonly used for
barcoding applications, but longer fragments can provide stronger spe-
cies conrmation and are also potentially more informative for
Fig. 4. DNA concentrations (ng/
L) of R. vigil casque samples HK-01 HK-06 were signicantly higher when sampled on the caudal side of the casque than the
lateral and rostral sides (p <0.05, A). However, when one sample with anomalously high DNA concentrations was removed from the analysis, DNA concentrations
did not vary signicantly between samples (p =0.1, B). Dots show individual casque data, boxes shows the upper and lower quartile ranges, boxplot lines
show medians.
Table 2
Pairwise sequence divergence between our target sequence (from HK-01) and
both intra-specic and inter-specic sequences from Genbank. Match rank in-
dicates the top intra-specic and inter-specic match from our BLAST results.
The barcoding gap is calculated as the percent similarity of the rst species and
minus the percent similar of the second and third species.
Species Genbank
accession #
% similarity with
target sample HK-
Barcoding gap
with sample HK-
1 Rhinoplax
2 Buceros
90.0% 10%
3 Aceros
89.9% 10.1%
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
geographic assignment of samples [42].
4.1. Extraction methods
Building on previous work which has shown that DNA can be
extracted, amplied, and sequenced from R. vigil casques [23,30], we
show that while phenol-chloroform methods yielded higher quantities of
DNA from extractions, both phenol-chloroform and commercial
extraction methods yield sufcient DNA for PCR amplication and
sequencing. The PCIA method tested in this study yielded the highest
DNA concentrations, which were relatively high compared to other
keratin sources of DNA including molted feathers (range =1 81 ng/
[50]) and echidna quills (mean 24.2 ng/
L, [25]). As well as being more
cost-efcient, DNA extractions using PCIA and CTAB methods have
often yielded samples with higher DNA concentrations and better DNA
purity ratios than commercial kits [5153]. However, commercial kits
are often less labour intensive and easier to use, making them suitable
for wildlife forensic application. This is important when considering the
capacity of different forensic laboratories and the often, rapid
turn-around necessary for analyses to be presented in court [54,55].
Therefore, these results - showing that both methods can extract DNA
sufcient to amplify 800-bp of mtDNA - are promising for analysts using
different methods in their laboratories.
4.2. Sampling methods & sampling location
Our results show that the DNA quantity and quality of powdered
casque samples taken via drilling was better overall than of samples
taken by shaving the casque surface. This has been shown in previous
studies where material sampled deeper into the wildlife product, and
therefore closer to the source of cell generation, often provides better
quality and quantity DNA [34]. However, a few of the shaved samples
from the caudal site also yielded high DNA concentrations. As such, the
caudal method resulted in the top three highest DNA concentrations,
which included both shaved and drilled samples. This is likely inu-
enced by the direction of growth of the casque, which is thought to grow
rostrally [3]. The anomalous results using the shaved, caudal method
may have been due to this anatomical aspect. High quantities of DNA
were extracted from some rostral sides. These high values could be due
to other non-R. vigil DNA contaminants which would lead to high
Nanodrop readings, or inuenced by other factors such as sample age,
preservation treatments added to the sample, or stochastic effects [56,
57]. Regardless of the DNA concentrations and purity ratios, all samples
were successfully sequenced.
Fig. 5. Helmeted hornbill (Rhinoplax vigil, highlighted) species identication from closely related sister species, based on a maximum likelihood (HKY +G model)
analysis of the 761-bp cytB region (without primers). Bootstrap support values are shown above the branches (expressed as percentage of 1000 replicates). Species
name, identications given in this study, reported and conscated origins, and accession numbers are shown, with U. epops as an outgroup.
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
4.3. Species identication test validation
CytB is widely used to distinguish between species in wildlife fo-
rensics and phylogenetic contexts, as it displays sequence diversity with
relatively low levels of intraspecic variation [22,31]. In a previous
study, this marker was shown to distinguish R. vigil from other hornbill
species [23]. Our study builds on this original study with the addition of
R. vigil samples from other parts of their range (particularly Indonesian
and Malaysian regions of Borneo) and from seized samples of uncon-
rmed origin. We also included additional samples from closely related
species B. bicornis and B. rhinoceros. Our results support the ndings
from [23] and provide more evidence that the target cytB region pro-
vides sufcient resolution to distinguish between R. vigil and the most
closely related other hornbill species, B. bicornis and B. rhinoceros. R. vigil
again formed its own unique clade, and with the additional samples we
see low levels of intraspecic sequence variation between individuals
originating or conscated from Borneo, and those originating or
conscated from Thailand and Sumatra. The phylogenetic split observed
between Borneo and Thailand/Sumatra has been seen in other bird
species [58,59], where conspecics in Borneo are genetically distinct
from those in other parts of Southeast Asia. However, due to our small
sample size, our use of a single molecular marker, and the lack of true
origin for many of our samples, we cannot infer true intraspecic
provenance without further study including reference samples of known
geographic provenance.
Our study demonstrates the repeatability, reproducibility, robust-
ness, sensitivity, specicity, and phylogenetic resolution of the species
identication assay. The results from the second analyst in Hong Kong,
the repeated assay on specimens in Indonesia, and the repeated tests on
different thermocyclers all demonstrate the repeatability and repro-
ducibility of the assay. The low limit of detection of the hornbill DNA,
even when mixed with other species, suggests high sensitivity of the
assay. The assay was robust to changes in annealing temperature, PCR
reagents, samples, analysts, and equipment, meaning that this protocol
will provide reliable results under various conditions across labora-
tories. Although the primers in this study amplify other hornbill genera
(as shown in the specicity test), minimum sequence divergence be-
tween species was signicantly higher than the 3% threshold suggested
between different species using 658-bp of mtDNA (COI gene) [60]. The
reliability of species identication tests ultimately depends on the se-
quences available in reference databases [61], so where possible, good
quality, veried samples should be used to conrm the species identity.
Ideally, the validity of this study would have benetted from more
known reference samples of R. vigil, however, the lack of known samples
for species identication and geographic origin is a common limitation
to species-specic forensic tests for endangered species [61]. As there
are no known R. vigil in captivity [11], obtaining known reference ma-
terial is difcult.
5. Conclusion
In this study we show that a range of methods can be used to extract
and sample DNA from R. vigil casques, and we have validated a previ-
ously published DNA species identication test of R. vigil [23] which
phylogenetically distinguishes R. vigil from sister hornbill species with
high statistical support. Validating this assay provides further support
for these previously published primers to be used for identifying R. vigil
in wildlife forensic applications. We also present the rst genetically
identied, illegally traded R. vigil casques in Hong Kong, and provide
additional known-origin reference samples of R. vigil for future work.
The data from this study will hopefully aid enforcement efforts requiring
forensic DNA species identication for cases involving R. vigil.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
We thank the Agriculture, Fisheries and Conservation Department of
Hong Kong SAR, and the Directorate General for Law Enforcement,
Ministry of Environment and Forestry of the Republic of Indonesia for
providing permission for us to use conscated specimens in our
research. We also thank the veterinary staff at the Ocean Park Corpo-
ration for providing us with the blood sample of R. undulatus. Funding
for this project was provided by the Ocean Park Conservation Founda-
tion, Hong Kong (Grant # BD02.1819), the National Geographic Society
Species Recovery Grant (Grant # NGS-50728C-18), the PUTI Universitas
Indonesia grant contract no. NKB-2006/UN2. RST/HKP.05.00/2020,
and the Research Center for Biology under the National Priority
Research Project No.02.002 Scientic Recommendation on CITES
Wildlife Trade 2018 2019. We are also grateful for all the helpful
discussions on this project that we received from Dr Kanita Ouitavon and
her team at DNP WiFOS, Bangkok, Thailand.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.fsiae.2022.100058.
[1] C. Beastall, C.R. Shepherd, Y. Hadiprakarsa, D. Martyr, Trade in the Helmeted
Hornbill Rhinoplax vigil: the ‘ivory hornbill, Bird. Conserv. Int. 26 (2) (2016)
[2] N.J. Collar, Helmeted Hornbills Rhinoplax vigil and the ivory trade: the crisis that
came out of nowhere, BirdingASIA 24 (2015) 1217.
[3] C.S.W. Manger Cats-Kuenen, Casque and bill of Rhinoplax vigil (Forst.) in
connection with the architecture of the skull, Verh Kon Ned Akad Wet, Afd
Natuurkel, NV Noord-Holland Uitgev, Maatschappij, Amsterdam, (1961).
[4] M.F. Kinnaird, Y.Y. Hadiprakarsa, P. Thiensongrusamee, Aerial jousting by
Helmeted Hornbills Rhinoplax vigil: observations from Indonesia and Thailand, Ibis
145 (3) (2003) 506508,
[5] D. Naish, Book review: The ecology and conservation of Asian hornbills: farmers of
the forest by Kinnaird and OBrien, University of Chicago Press, (2007) 315 pp.,
Hist. Biol. 27(7) (2015) 954956. doi: 10.1080/08912963.2014.919757.
[6] Environmental Investigation Agency (EIA), Seeing ‘red the often hidden colour
of wildlife contraband, (2015). URL:
often-hidden-colour-of-wildlife-contraband (accessed 8 December 2022).
[7] R.E. Kane, Hornbill Ivory, Gems Gemol. 17 (2) (1981) 9697.
[8] Environmental Investigation Agency (EIA), Mapping seizures to aid conservation of
imperilled helmeted hornbill, (2017). URL:
ases/mapping-seizures-aid-imperilled-helmeted-hornbill/ (accessed 8 December
[9] C.E.R. Webster, The Critically Endangered Helmeted Hornbill (Rhinoplax vigil):
Threats & Conservation, PhD Thesis, The University of Hong Kong, Hong Kong,
[10] Government of Hong Kong Special Administrative Region (HKSAR), Man jailed for
smuggling helmeted hornbill beaks, press release July 30, (2019). URL: https
:// (accessed 8
December 2022).
[11] A. Jain, J.G.H. Lee, N. Chao, C. Lees, R. Orenstein, B.C. Strange, S.C.L. Chng, W.
Marthy, C.A. Yeap, M. Rao, Helmeted Hornbill (Rhinoplax vigil): Status Review,
Range-wide Conservation Strategy and Action Plan (20182027), IUCN SSC
Hornbill Specialist Group, (2018). doi:10.13140/RG.2.2.25467.16168.
[12] B.D. Adji, R.K. Sari, A. Nugroho, S. Ratnaningsih, D.S. Chandradewi, A. Santosa,
& N. Hardianto, The National Action Plan and Conservation Strategy of The
Helmeted Hornbill (Rhinoplax vigil) in Indonesia, (2019). URL: https://pdf.usaid.
gov/pdf_docs/PA00XFGS.pdf (accessed 8 December 2022).
[13] A.C. Kemp, M. Woodcock. The Hornbills: Bucerotiformes, Oxford University Press,
New York, 1995.
[14] K. Krishnasamy, B. Leupen, O.C. Or, Observations of the Helmeted Hornbill Trade
in Lao PDR, TRAFFIC, Southeast Asia Regional Ofce, Petaling Jaya, Selangor,
Malaysia, (2016).
[15] J. Liang, H. Li, T. Lu, J. Zhang, M. Shen, J. Zhou, Identication characteristics of
natural and imitation hornbill ivory, J. Gemmol.. 34 (1) (2014) 4249.
[16] D. Carde˜
nosa, J. Quinlan, K.H. Shea, D.D. Chapman, Multiplex real-time PCR assay
to detect illegal trade of CITES-listed shark species, Sci. Rep. 8 (1) (2018) 110,
[17] K.M. Ewart, G.J. Frankham, R. McEwing, L.M. Webster, S.A. Ciavaglia, A.M.
T. Linacre, D.T. The, K. Ovouthan, R.N. Johnson, An internationally standardized
C.E.R. Hatten et al.
Forensic Science International: Animals and Environments xxx (xxxx) xxx
species identication test for use on suspected seized rhinoceros horn in the illegal
wildlife trade, Forensic Sci. Int. Genet 32 (2018) 3339,
[18] K.M. Ewart, A.L. Lightson, F.T. Sitam, J. Rovie-Ryan, S.G. Nguyen, K.I. Morgan,
A. Luczon, E.M.S. Anadon, M. De Bruyn, S. Bourgeois, K. Ouitavon, A. Kotze, M.S.
A. Bakar, M. Salgado-Lynn, R. McEwing, DNA analyses of large pangolin scale
seizures: species identication validation and case studies, Forensic Sci. Int. Anim.
Environ. 1 (2021), 100014,
[19] K.I. Morgan, K.M. Ewart, T.Q. Nguyen, F.T. Sitam, K. Ouitavon, A.L. Lightson,
A. Kotze, R. McEwing, Avoiding common numts to provide reliable species
identication for tiger parts, Forensic Sci. Int. Rep. 3 (2021), 100166.
[20] M. Staats, A.J. Arulandhu, B. Gravendeel, A. Holst-Jensen, I. Scholtens, T. Peelen,
T.W. Prins, E. Kok, Advances in DNA metabarcoding for food and wildlife forensic
species identication, Anal. Bioanal. Chem. 408 (17) (2016) 46154630, https://
[21] M.K. Moore, I.L. Korneld, Best Practices in Wildlife Forensic DNA, in: J.E.
Huffman, J.R. Wallace (Eds.), Wildlife Forensics: Methods and Applications, John
Wiley & Sons, Ltd., West Sussex, (2012), pp 201236. doi:10.1002/
[22] A. Linacre, S.S. Tobe, An overview to the investigative approach to species testing
in wildlife forensic science, Investig. Genet.. 2 (1) (2011) 19,
[23] K. Ouitavon, R. McEwing, K. Penchart, K. Sri-aksorn, V. Chimchome, DNA recovery
and analysis from helmeted hornbill (Rhinoplax vigil) casques and its potential
application in wildlife law enforcement, Forensic Sci. Int. Anim. Environ. 2 (2022),
[24] A. Linacre, L. Gusm˜
ao, W. Hecht, A.P. Hellmann, W.R. Mayr, W. Parson, M. Prinz,
P.M. Schneider, N. Morling, ISFG: recommendations regarding the use of non-
human (animal) DNA in forensic genetic investigations, Forensic Sci. Int. Genet. 5
(5) (2011) 501505.
[25] A.E. Summerell, G.K. Frankham, P. Gunn, R.N. Johnson, DNA based method for
determining source country of the short beaked echidna (Tachyglossus aculeatus) in
the illegal wildlife trade, Forensic Sci. Int. 295 (2019) 4653,
[26] K.M. Ewart, A.L. Lightson, F.T. Sitam, J.J. Rovie-Ryan, N. Mather, R. McEwing,
Expediting the sampling, decalcication, and forensic DNA analysis of large
elephant ivory seizures to aid investigations and prosecutions, Forensic Sci. Int.
Genet. 44 (2020), 102187,
[27] M.K. Moore, B.W. Baker, T.L. Bauman, M. Burnham-Curtis, E.O. Espinoza, C.
S. Ferrell, G.J. Frankham, K. Frazier, J.L. Giles, D. Hawk, J.J. Rovie-Ryan, R.
N. Johnson, T. Knott, Irving, Korneld, C. Lindquist, W.D. Lord, K.L. Morgan, R.
C. OBrien, R. Ogden, T.L. Prigge, P. Schwenke, F.T. Sitam, P. Trail, E. Wictum, P.
J. Wilson, B.C. Yates, L.M.I. Webster, The society for wildlife forensic science
standards and guidelines, Forensic Sci. Int. Anim. Environ. 1 (2021), 100015,
[28] N. Viseshakul, W. Charoennitikul, S. Kitamura, A. Kemp, S. Thong-Aree,
Y. Surapunpitak, P. Poonswad, M. Ponglikitmongkol, A phylogeny of frugivorous
hornbills linked to the evolution of Indian plants within Asian rainforests, J. Evol.
Biol. 24 (7) (2011) 15331545,
[29] J.C.T. Gonzalez, B.C. Sheldon, N.J. Collar, J.A. Tobias, A comprehensive molecular
phylogeny for the hornbills (Aves: Bucerotidae), Mol. Phylogenet. Evol. 67 (2)
(2013) 468483,
[30] Y.S. Fitriana, M. Irham, H. Sutrisno, A molecular genetic approach for sex
determination on helmeted hornbill (Rhinoplax vigil) casque: a forensic casework,
BIO Web Conf. 19 (2020) 00020,
[31] W. Parson, K. Pegoraro, H. Niederst¨
atter, M. F¨
oger, M. Steinlechner, Species
identication by means of the cytochrome b gene, Int. J. Leg. Med.. 114 (1) (2000)
[32] R. Kaur, S. Singh, R. Ramli, H. Hassan, A. Majinun, M. Ancrenaz, M. Ratag, H.
A. Rauf, J. Jamal, A.S.A. Rauf, M.D.M. Yunus, M.A. Arbani, K. Asmar, M.A. Amir,
Observations at a nest of Helmeted Hornbill Rhinoplax vigil in Borneo, Malaysia,
Forktail 34 (2019) 6873.
[33] A.C. Kemp, Family Bucerotidae (hornbills), in: J. del Hoyo, A.D. Elliott, J. Sargatal
(Eds)., Handbook of the Birds of the World 6, Lynx Edicions, Barcelona, (2001), pp
[34] C.K. Harper, G.J. Vermeulen, A.B. Clarke, I. Jacobus, A.J. Guthrie, Extraction of
nuclear DNA from rhinoceros horn and characterization of DNA proling systems
for white (Ceratotherium simum) and black (Diceros bicornis) rhinoceros, in: Forensic
Sci. Int. Genet., 2013, pp. 428433,
[35] T.L. Fulton, S.M. Wagner, B. Shapiro, Case study: recovery of ancient nuclear DNA
from toe pads of the extinct passenger pigeon, in: B. Shapiro, M. Hofreiter M.
(Eds.), Ancient DNA, Humana Press, (2012), pp. 2935.
[36] S.M. Billerman, J. Walsh, Historical DNA as a tool to address key questions in avian
biology and evolution: a review of methods, challenges, applications, and future
directions, Mol. Ecol. Resour. 19 (5) (2019) 11151130,
[37] M.A. Coffroth, H.R. Lasker, M.E. Diamond, J.A. Bruenn, E. Bermingham, DNA
ngerprints of a gorgonian coral: a method for detecting clonal structure in a
vegetative species, Mar. Biol. 114 (2) (1992) 317325,
[38] K.A. Boesenberg-Smith, M.M. Pessarakli, D.M. Wolk, Assessment of DNA yield and
purity: an overlooked detail of PCR troubleshooting, Clin. Microbiol. Newsl. 34 (1)
(2012) 16,
[39] S. Sloan, C.J. Jenvey, D. Piedrata, S. Preston, M.J. Stear, Comparative evaluation
of different molecular methods for DNA extraction from individual Teladorsagia
circumcincta nematodes, BMC Biotechnol. 21 (1) (2021) 113,
[40] R. Core Team, R: A language and environment for statistical computing, R
Foundation for Statistical Computing, Vienna, Austria, (2021). URL: https://www.
[41] C.P. Meyer, G. Paulay, DNA barcoding: error rates based on comprehensive
sampling, PLoS Biol. 3 (12) (2005), e422,
[42] N. Vasiljevic, M. Lim, E. Humble, A. Seah, A. Kratzer, N.V. Morf, S. Prost, R. Ogden,
Developmental validation of Oxford Nanopore Technology MinION sequence data
and the NGSpeciesID bioinformatic pipeline for forensic genetic species
identication, Forensic Sci. Int. Genet. 53 (2021), 102493,
[43] S.J. Hackett, R.T. Kimball, S. Reddy, R.C. Bowie, E.L. Braun, M.J. Braun, J.
L. Chojnowski, W.A. Cox, K.L. Han, J. Harshman, C.J. Huddleston, B.D. Marks, K.
J. Miglia, W.S. Moore, F.H. Sheldon, D.W. Steadman, C.C. Witt, T. Yuri,
A phylogenomic study of birds reveals their evolutionary history, Science 320
(5884) (2008) 17631768,
[44] S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, MEGA X: molecular evolutionary
genetics analysis across computing platforms, Mol. Biol. Evol. 35 (6) (2018) 1547,
doi: 10.1093/molbev/msy096.
[45] S. Guindon, F. Delsuc, J.F. Dufayard, O. Gascuel, Estimating maximum likelihood
phylogenies with PhyML, in: D. Posada (Ed.), Bioinformatics for DNA Sequence
Analysis, Humana Press, 2009, pp. 113127,
[46] M. Hasegawa, H. Kishino, T. Yano, Dating the human-ape split by a molecular
clock of mitochondrial DNA, J. Mol. Evol. 22 (1985) 160174,
[47] T.D. Price, D.M. Hooper, C.D. Buchanan, U.S. Johansson, D.T. Tietze, P. Alstr¨
U. Olsson, M. Ghosh-Harihar, F. Ishtiaq, S.K. Gupta, J. Martens, B. Harr, P. Singh,
D. Mohan, Niche lling slows the diversication of Himalayan songbirds, Nature
509 (7499) (2014) 222225,
[48] Y. Chen, C. Li, H. Yan, H. Xiao, S. Chen, The complete mitochondrial genome
sequence of Buceros bicornis (Bucerotiformes: Bucerotidae), Conserv. Genet.
Resour. 10 (3) (2018) 287290,
[49] R. Ogden, N. Dawnay, R. McEwing, Wildlife DNA forensics - bridging the gap
between conservation genetics and law enforcement, Endanger. Species Res. 9 (3)
(2009) 179195,
[50] S.B. de Volo, R.T. Reynolds, M.R. Douglas, M.F. Antolin, An improved extraction
method to increase DNA yield from molted feathers, 2008, Condor 110 (4) (2008)
[51] G. Eslami, S. Khalatbari-Limaki, M.H. Ehrampoush, M. Gholamrezaei, B.
Hajimohammadi, A. Oryan, Comparison of three different DNA extraction methods
for Linguatula serrata as a food born pathogen, Iran. J. Parasitol. 12 (2) (2017) 236.
PMID: 28761484.
[52] N. Gumi´
nska, M. Płecha, H. Walkiewicz, P. Hałakuc, B. Zakry´
s, R. Milanowski,
Culture purication and DNA extraction procedures suitable for next-generation
sequencing of euglenids, J. Appl. Phycol. 30 (6) (2018) 35413549, https://doi.
[53] Z. Suadi, L.M. Bilung, K. Apun, A.A. Azmi, Efciency of traditional DNA extraction
method in PCR detection of porcine DNA in meat mixtures, J. Teknol. 82 (5) (2020)
[54] R. Ogden, J. Mailley, A review of wildlife forensic science and laboratory capacity
to support the implementation and enforcement of CITES. CITES CoP17 Doc. 25
Annex 4. A review commissioned by the Secretariat of the Convention on
International Trade in Endangered Species of Wild Fauna and Flora and undertaken
by the United Nations Ofce on Drugs and Crime, (2016). URL:
ites/default/les/eng/cop/17/WorkingDocs/E-CoP17-25-A4.pdf (accessed 8
December 2022).
[55] G. Khedkar, C. Khedkar, B. Prakash, A. Khedkar, D. Haymer, DNA barcode based
identication of a suspected tiger skin: a case to resolve mimicry, Forensic Sci. Int.
Rep. 1 (2019), 100027,
[56] C.F. Bengtsson, M.E. Olsen, L.Ø. Brandt, M.F. Bertelsen, E. Willerslev, D.J. Tobin,
A.S. Wilson, M.T.P. Gilbert, DNA from keratinous tissue. Part I: hair and nail, Ann.
Anat. 194 (1) (2012) 1725,
[57] J.B. Stetz, T. Seitz, M.A. Sawaya, Effects of exposure on genotyping success rates of
hair samples from brown and American black bears, J. Fish. Wildl. Manag. 6 (1)
(2015) 191198,
[58] E.Y. Ng, K.M. Garg, G.W. Low, B. Chattopadhyay, R.R. Oh, J.G. Lee, F.E. Rheindt,
Conservation genomics identies impact of trade in a threatened songbird, Biol.
Conserv. 214 (2017) 101108,
[59] V.L. Chua, Q. Phillipps, H. Chuan Lim, S.S. Taylor, D.F. Gawin, M.A. Rahman, R.
G. Moyle, F.H. Sheldon, Phylogeography of three endemic birds of Maratua Island,
a potential archive of Bornean biogeography, Rafes Bull. Zool. 63 (2015)
[60] P.D. Hebert, A. Cywinska, S.L. Ball, J.R. DeWaard, Biological identications
through DNA barcodes, Proc. R Soc. B 270 (1512) (2013) 313321, https://doi.
[61] N. Ahlers, J. Creecy, G. Frankham, R.N. Johnson, A. Kotze, A. Linacre, R. McEwing,
M. Mwale, J.J. Rovie-Ryan, F. Sitam, L.M.I. Webster, ‘ForCytDNA database of
wildlife species, Forensic Sci. Int. Genet. Suppl. Ser. 6 (2017) e466-e468. https://
C.E.R. Hatten et al.
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
Pangolins are the most trafficked mammal in the world, and all eight species are listed under CITES Appendix I. DNA-based wildlife forensic techniques are recognized as an important component of investigating a pangolin seizure. In particular, determining the species of pangolin in a seizure will 1) confirm the presence of pangolin to establish the legality of any trade, and 2) ensure appropriate laws are applied to their fullest extent in a prosecution. Furthermore, valuable intelligence data, such as determining the geographic provenance of samples, can be produced through analysis of pangolin seizures. Despite the immense scale of the pangolin trade, standardized wildlife forensic techniques for testing pangolin seizures are in their infancy. To address this, here, we present a standardized genetic marker suitable for species identification of all eight pangolin species, and outline practical strategies for sampling large-volume pangolin scale seizures. We assessed the repeatability, reproducibility, robustness, sensitivity and phylogenetic resolution of this species identification test. Critically, the assay was tested in four wildlife forensic laboratories involved in testing pangolins. Additionally, we demonstrated the test's utility to conduct geographic provenance analysis of Phataginus tricuspis samples. We analysed five large-volume pangolin scale seizures in Malaysia, which elucidated key target species, poaching hotspots, and trafficking routes. Phataginus tricuspis was the most commonly identified species (88.8%) from the seizure samples, and 84.3% of these P. tricuspis individuals were likely sourced from western central Africa. We expect the implementation of the techniques presented in this paper will improve enforcement of pangolin trafficking crimes.
Full-text available
Background The purpose of this study was to develop a reliable DNA extraction protocol to use on individual Teladorsagia circumcincta nematode specimens to produce high quality DNA for genome sequencing and phylogenetic analysis. Pooled samples have been critical in providing the groundwork for T. circumcincta genome construction, but there is currently no standard method for extracting high-quality DNA from individual nematodes. 11 extraction kits were compared based on DNA quality, yield, and processing time. Results 11 extraction protocols were compared, and the concentration and purity of the extracted DNA was quantified. Median DNA concentration among all methods measured on NanoDrop 2000™ ranged between 0.45–11.5 ng/μL, and on Qubit™ ranged between undetectable – 0.962 ng/μL. Median A260/280 ranged between 0.505–3.925, and median A260/230 ranged − 0.005 – 1.545. Larval exsheathment to remove the nematode cuticle negatively impacted DNA concentration and purity. Conclusions A Schistosoma sp. DNA extraction method was determined as most suitable for individual T. circumcincta nematode specimens due to its resulting DNA concentration, purity, and relatively fast processing time.
Full-text available
Species identification of non-human biological evidence through DNA nucleotide sequencing is routinely used for forensic genetic analysis to support law enforcement. The gold standard for forensic genetics is conventional Sanger sequencing; however, this is gradually being replaced by high-throughput sequencing (HTS) approaches which can generate millions of individual reads in a single experiment. HTS sequencing, which now dominates molecular biology research, has already been demonstrated for use in a number of forensic genetic analysis applications, including species identification. However, the generation of HTS data to date requires expensive equipment and is cost-effective only when large numbers of samples are analysed simultaneously. The Oxford Nanopore Technologies (ONT) MinIONTM is an affordable and small footprint DNA sequencing device with the potential to quickly deliver reliable and cost effective data. However, there has been no formal validation of forensic species identification using high-throughput (deep read) sequence data from the MinION making it currently impractical for many wildlife forensic end-users. Here, we present a MinION deep read sequence data validation study for species identification. First, we tested whether the clustering-based bioinformatics pipeline NGSpeciesID can be used to generate an accurate consensus sequence for species identification. Second, we systematically evaluated the read variation distribution around the generated consensus sequences to understand what confidence we have in the accuracy of the resulting consensus sequence and to determine how to interpret individual sample results. Finally, we investigated the impact of differences between the MinION consensus and Sanger control sequences on correct species identification to understand the ability and accuracy of the MinION consensus sequence to differentiate the true species from the next most similar species. This validation study establishes that ONT MinION sequence data used in conjunction with the NGSpeciesID pipeline can produce consensus DNA sequences of sufficient accuracy for forensic genetic species identification.
Full-text available
Tigers are killed to supply a demand for many wildlife products despite a ban on commercial international trade. As populations decrease, products from substitute species (i.e. lions and leopards) have been fraudulently sold as tiger. DNA forensic techniques are needed to definitively identify tiger in order to secure prosecutions although this is complicated by the presence of numts. Therefore, we have developed and validated a CO1 genetic marker that preferentially amplifies the mtDNA CO1 region and excludes the nuclear CO1 pseudogene, which we expect to be of use in tiger forensic casework.
Full-text available
Helmeted Hornbill (Rhinolax vigil) is the only hornbill that equipped with solid casque made from keratin for both males and females. The demand for casque in the black market was huge and resulted in IUCN status leaped up from vulnerable to critically endangered. We received a total of 68 confiscated helmeted hornbill casques. As part of the casework and the objectives of the study, we determined to reveal the sex status of those casques and the best methods to work with keratinous material. Molecular methods to determining sex in birds rely on the CHD gene located on male and female chromosomes ZZ and ZW, respectively. We optimized laboratory protocols for genetic sexing using three independent sets of primers P2/P8, 2550F/2718R, and CHD1F/CHD1R to amplify regions of the sexlinked CHD-Z and CHD-W genes. The CHD1F/CHD1R determined sex 80.88% of samples. The 2550F/2718R were quite successful, sexing 51.47% of samples. In contrast, the P2/P8 only identified the sex around 20.58% of samples. These results showed that CHD1F/CHD1R works the most effective for sexing the casques with 52.9% females, 27.9% males, and 19.1% unidentified. Therefore, the most accurate and suitable primers are CHD1F/CHD1R, 2550F/2718R, and P2/P8, respectively for keratinous samples.
Full-text available
The illegal ivory trade continues to drive elephant poaching. Large ivory seizures in Africa and Asia are still commonplace. Wildlife forensics is recognised as a key enforcement tool to combat this trade. However, the time and resources required to effectively test large ivory seizures is often prohibitive. This limits or delays testing, which may impede investigations and/or prosecutions. Typically, DNA analysis of an ivory seizure involves pairing and sorting the tusks, sampling the tusks, powdering the sample, decalcification, then DNA extraction. Here, we optimize the most time-consuming components of this process: sampling and decalcification. Firstly, using simulations, we demonstrate that tusks do not need to be paired to ensure an adequate number of unique elephants are sampled in a large seizure. Secondly, we determined that directly powdering the ivory using a Dremel drill with a high-speed cutter bit, instead of cutting the ivory with a circular saw and subsequently powdering the sample in liquid nitrogen with a freezer mill, produces comparable results. Finally, we optimized a rapid 2 -h decalcification protocol that produces comparable results to a standard 3-day protocol. We tested/optimised the protocols on 33 raw and worked ivory samples, and demonstrated their utility on a case study, successfully identifying 94% of samples taken from 123 tusks. Using these new rapid protocols, the entire sampling and DNA extraction process takes less than one day and requires less-expensive equipment. We expect that the implementation of these rapid protocols will promote more consistent and timely testing of ivory seizures suitable for enforcement action.
Since 2011 the demand from China for the keratin casque from helmeted hornbills, so called red ivory, has increased significantly according to recent studies and has the potential to drive this species to extinction. Wildlife DNA Forensics is the field of science tasked not with expanding academic knowledge but with providing evidence for court in relation to wildlife crimes or for providing robust intelligence information to enforcement agencies in relation to trade routes for illegal wildlife products. In this pilot study, we examine the potential to recover DNA from the casques of the helmeted hornbill and evaluate how this genetic information could be used to better inform investigations into the illegal trade of helmeted hornbills.
Since 2011 the demand from China for the keratin casque from helmeted hornbills, so called red ivory, has increased significantly according to recent studies and has the potential to drive this species to extinction. Wildlife DNA Forensics is the field of science tasked not with expanding academic knowledge but with providing evidence for court in relation to wildlife crimes or for providing robust intelligence information to enforcement agencies in relation to trade routes for illegal wildlife products. In this pilot study, we examine the potential to recover DNA from the casques of the helmeted hornbill and evaluate how this genetic information could be used to better inform investigations into the illegal trade of helmeted hornbills.
Through the advancement of biotechnology, DNA-based methods are the most effective techniques in species identification, as they are rapid and have higher stability in harsh conditions compared to protein-based methods. This study was conducted to determine the efficiency of the traditional DNA extraction method, phenol/chloroform/isoamyl alcohol (PCIA), and comparing it with the commercially available kit by evaluating the purity, concentration, and suitability for amplification of porcine DNA in raw chicken and beef mixtures. The quantity and quality of the DNA extracts were assessed using a UV-Vis spectrophotometer. Polymerase chain reaction (PCR) was performed using species-specific primers targeting mitochondrial DNA cytochrome b (cyt b) gene of chicken (227-bp), beef (274-bp), and pork (398-bp), to confirm the template usability and quality of the DNA extracts. High DNA concentrations and purity were obtained from meat samples extracted using the PCIA method. The visualization of pork DNA on 2% agarose gel was able to detect pork contamination in raw meat mixtures up to minute proportion (1%). The existence of pork in chicken and beef was indicated with the presence of a specific 398-bp DNA band. Thus, the PCIA method can be recommended as a cost-effective and an excellent alternative to more expensive extraction kits in detecting pork DNA in raw meat mixtures.