Available online at www.sciencedirect.

com

Forensic Science International 177 (2008) 66–76

www.elsevier.com/locate/forsciint

A global study of forensically significant calliphorids:

Implications for identification
M.L. Harvey a,*, S. Gaudieri a,b, M.H. Villet c, I.R. Dadour a
a
Centre for Forensic Science, M420, University of Western Australia, Nedlands 6907, Australia
b
School of Anatomy and Human Biology, University of Western Australia, Nedlands 6907, Australia
c
Department of Zoology and Entomology, Rhodes University, Grahamstown 6140, South Africa
Received 17 February 2006; received in revised form 12 September 2007; accepted 31 October 2007
Available online 4 March 2008

Abstract
A proliferation of molecular studies of the forensically significant Calliphoridae in the last decade has seen molecule-based identification of
immature and damaged specimens become a routine complement to traditional morphological identification as a preliminary to the accurate
estimation of post-mortem intervals (PMI), which depends on the use of species-specific developmental data. Published molecular studies have
tended to focus on generating data for geographically localised communities of species of importance, which has limited the consideration of
intraspecific variation in species of global distribution. This study used phylogenetic analysis to assess the species status of 27 forensically
important calliphorid species based on 1167 base pairs of the COI gene of 119 specimens from 22 countries, and confirmed the utility of the COI
gene in identifying most species. The species Lucilia cuprina, Chrysomya megacephala, Ch. saffranea, Ch. albifrontalis and Calliphora stygia
were unable to be monophyletically resolved based on these data. Identification of phylogenetically young species will require a faster-evolving
molecular marker, but most species could be unambiguously characterised by sampling relatively few conspecific individuals if they were from
distant localities. Intraspecific geographical variation was observed within Ch. rufifacies and L. cuprina, and is discussed with reference to
unrecognised species.
# 2008 Published by Elsevier Ireland Ltd.

Keywords: Calliphoridae; Forensic entomology; Blowflies; COI; Cox1; Intraspecific variation

1. Introduction majority of molecular studies, however, have used the

The advent of DNA-based identification techniques for use
in forensic entomology in 1994 [1] saw the beginning of a
proliferation of molecular studies into the forensically
important Calliphoridae. The use of DNA to characterise
morphologically indistinguishable immature calliphorids was
recognised as a valuable molecular tool with enormous
practical utility. Numerous studies have since addressed the
DNA-based identification of calliphorids [2–6]. A variety of
regions of DNA have been suggested for study including the
nuclear internal transcribed spacers (ITS) [7], mitochondrial
rRNA genes [8] and the mitochondrial control region [8]. The
* Corresponding author at: School of Biological Sciences, University of
Portsmouth, King Henry Building, King Henry I Street, Portsmouth PO12DY,
UK. Tel.: +44 23 9284 5012.
E-mail address: Michelle.Harvey@port.ac.uk (M.L. Harvey).
cytochrome oxidase I (COI or cox1) encoding region of
mitochondrial DNA (mtDNA) [1,2,4–6,9].
The COI gene holds enormous utility for species identifica-
tion. Lying within the mitochondrial genome, it has the
advantages of easy isolation, higher copy number than its
nuclear counterparts, and conserved sequence and structure
across taxa. COI has been well studied in the Insecta [10], with
its utility for distinction between closely related species of
Diptera demonstrated by the large number of COI studies of
species complexes in the Culicidae (e.g. [11]).
Calliphorid molecular taxonomic studies have focused
largely on sequencing of the COI gene and have illustrated
the ability to successfully distinguish between a wide variety of
forensically important species based largely on monophyly
[1,2,4–6,9]. The main limitation to the use of COI sequence
data has been the inability to distinguish between some closely
related species of the genus Calliphora, generally due to
incidences of para- or polyphyly. Wallman et al. [4,12] found

0379-0738/$ – see front matter # 2008 Published by Elsevier Ireland Ltd.

doi:10.1016/j.forsciint.2007.10.009
 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76 67

Table 1
Individuals used in this study, listed with locality of origin and GenBank accession number, and publication data where identified from another publicationa
Species Locality Accession no. Source
Chrysomya saffranea Broome, Australia EU418533 New sequence
Broome, Australia EU418534 New sequence
Brisbane, Australia AB112841 [6]
Chrysomya megacephala Sydney, Australia EU418535 New sequence
Perth, Australia AB112846 [6]
Perth, Australia AB112847 [6]
Pretoria, South Africa AB112848 [6]
Kitwe, Zambia AB112861 [6]
Kitwe, Zambia AB112856 [6]
KwaZulu-Natal, South Africa AB112830 [6]
Hawaii, United States EU418536 New sequence
Papua New Guinea AF295551 [32]
Kuala Lumpur, Malaysia EU418537 New sequence
Malaysia AY909052 NCBI submission
Malaysia AY909053 NCBI Submission
Chrysomya pinguis Hsintien, Taipei County, Taiwan AY092759 [33]
Chrysomya bezziana Bogor, Indonesia AF295548 [32]
Chrysomya inclinata KwaZulu-Natal, South Africa AB112857 [6]
Chrysomya chloropyga Graaf-Reinet, South Africa EU418540 New sequence
Graaf-Reinet, South Africa EU418541 New sequence
Pretoria, South Africa EU418538 New sequence
KwaZulu-Natal, South Africa EU418539 New sequence
Chrysomya putoria Kitwe, Zambia AB112831 [6]
Kitwe, Zambia AB112860 [6]
Snake Island, Botswana AB112835 [6]
Snake Island, Botswana AB112855 [6]
Sao Joao da Boa Vista, Brazil EU418542 New sequence
near Chilbre, Panama AF295554 [32]
Chrysomya marginalis Pretoria, South Africa AB112838 [6]
Pretoria, South Africa AB112832 [6]
Karoo, South Africa AB112866 [6]
Karoo, South Africa AB112862 [6]
Karoo, South Africa EU418543 New sequence
KwaZulu-Natal, South Africa AB112837 [6]
KwaZulu-Natal, South Africa AB112834 [6]
Chrysomya varipes Gladstone, Australia EU418544 New sequence
Sydney, Australia EU418545 New sequence
Adelaide, Australia AF295556 [32]
Perth, Australia AB112868 [6]
Perth, Australia AB112869 [6]
Perth, Australia AB112867 [6]
Chrysomya norrisi Wau, Papua New Guinea AF295552 [32]
Chrysomya rufifacies Perth, Australia EU418546 New sequence
Perth, Australia AB112828 [6]
Perth, Australia AB112845 [6]
Campbell Town, Tasmania EU418547 New sequence
Florida, USA AF083658 [32]
Knoxville, USA EU418548 New sequence
Oahu, Hawaii, USA EU418549 New sequence
Chingmei, Taipei City, Taiwan AY092760 [33]
Malaysia AY909055 NCBI submission
Malaysia AY909054 NCBI submission
Chrysomya albiceps Alexandria, Egypt AF083657 [32]
Pretoria, South Africa AB112840 [6]
Pretoria, South Africa AB112839 [6]
KwaZulu-Natal, South Africa AB112836 [6]
KwaZulu-Natal, South Africa AB112842 [6]
Deka, Zimbabwe AB112849 [6]
Deka, Zimbabwe AB112858 [6]
Manzini, Swaziland AB112865 [6]
Manzini, Swaziland AB112854 [6]
Manzini, Swaziland AB112851 [6]
Cochliomyia hominivorax Alfenas, Brazil EU418550 New sequence
Cochliomyia macellaria Salvador, Brazil EU418551 New sequence
68 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
Table 1 (Continued )
Species Locality Accession no. Source
Gainesville, Florida AF295555 [32]
Calliphora dubia Geraldton, Western Australia EU418552 New sequence
Perth, Australia EU418553 New sequence
20km north New Norcia, Western Australia EU418554 New sequence
Ravensthorpe, Australia EU418555 New sequence
Toodyay, Australia EU418556 New sequence
Calliphora augur Sydney, Australia EU418557 New sequence
Sydney, Australia EU418558 New sequence
Calliphora hilli Gladstone, Tasmania EU418559 New sequence
Calliphora varifrons Boddington, Australia EU418560 New sequence
Calliphora ochracea Sydney, Australia EU418561 New sequence
Sydney, Australia EU418562 New sequence
Calliphora stygia Wallaceville, New Zealand EU418563 New sequence
Kaitoke, New Zealand EU418564 New sequence
Kempton, Tasmania EU418565 New sequence
Calliphora albifrontalis 20km north New Norcia, Australia EU418566 New sequence
Perth, Australia EU418567 New sequence
Perth, Australia EU418568 New sequence
Calliphora vomitoria Montferrier-Sur-Lez, France EU418569 New sequence
Calliphora vicina Montferrier-Sur-Lez, France EU418570 New sequence
Kempton, Tasmania EU418571 New sequence
London, UK EU418572 New sequence
London, UK EU418573 New sequence
Bristol University Colony, UK AJ417702 [32]
Lucilia illustris Montferrier-Sur-Lez, France EU418574 New sequence
Langford, UK AJ551445 [34]
Lucilia ampullacea Montferrier-Sur-Lez, France EU418575 New sequence
Lucilia cuprina Gladstone, Tasmania EU418576 New sequence
Perth, Australia AB112863 [6]
Perth, Australia AB112852 [6]
Perth, Australia AB112853 [6]
Townsville, Australia AJ417710 [34]
Dorie, New Zealand AJ417706 [34]
Chiang Mai University Lab Colony, Thailand EU418577 New sequence
Tororo, Uganda AJ417711 [34]
Dakar, Senegal AJ417708 [34]
Chingmei, Taipei City, Taiwan AY097335 [33]
Honolulu, Hawaii AJ417704 [34]
Waianae, Hawaii AJ417705 [34]
Lucilia sericata Montferrier-Sur-Lez, France EU418577 New sequence
Montferrier-Sur-Lez, France EU418578 New sequence
Perth, Australia AB112833 [6]
Graaf-Reinet, South Africa AB112850 [6]
Graaf-Reinet, South Africa AB112843 [6]
Pretoria, South Africa AB112864 [6]
Pretoria, South Africa AB112859 [6]
Harare, Zimbabwe AB112844 [6]
Harare, Zimbabwe AJ417717 [34]
Nerja, Spain AJ417716 [34]
Hilerod, Denmark AJ417712 [34]
Langford, Somerset, UK AJ417714 [34]
Dorie, New Zealand AJ417713 [34]
Los Angeles, USA AJ417715 [34]
Hydrotaea rostrata Perth, Australia AB112829 [6]
a
‘‘New sequence’’ indicates sequences have not been published elsewhere and have been submitted to the public databases, with release pending publication of this
study.

difficulty in separating the C. augur/C. dubia, C. stygia/C.
albifrontalis and C. hilli/C. varifrons species pairs, creating
difficulty in geographical regions where these sister species
overlap. Our first aim was to contribute to solving such
problems by enlarging the available data set.
A number of species of forensic utility are relatively
cosmopolitan, such as L. cuprina, L. sericata, Ch. rufifacies,
Ch. megacephala and C. vicina. However, forensic entomol-
ogy is a locality-specific science and molecular studies are
generally directed to the specific fauna found in a region [e.g.
4–6]. Specimens from new localities may not exactly match
published DNA sequences, raising questions regarding
acceptable levels of variation for distinction. Our second
aim was therefore to use conspecific specimens from
 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
geographically distant localities to estimate the range of
variation found in various recognised species, and thus
evaluate the conservation of the COI gene over geographic
distances and implications for necessary sample size in
taxonomic studies of these species.
Specimens that are similar to, but do not lie within the known
range of genetic variation of, a recognised species may
represent previously-unsampled geographical variation of that
species, another species recognised by systematics but not yet
sampled for DNA, or taxonomically unrecognised cryptic
species. The occurrence of cryptic species, which appear
morphologically the same as named species but differ in their
behaviour, development or other biology, may contribute to
error in PMI estimates. For example, Ch. rufifacies is a fly with
a widespread distribution displaying variable behaviour in
different localities [6], and Wallman et al. [12] have recently
indicated the possibility of cryptic species within Ch. rufifacies
in Australia.
In molecular phylogenies, discrimination is commonly
based on the separation of monophyletic clades, or
alternately, DNA barcoding studies have led to the suggested
application of heuristic thresholds or expected percent DNA
sequence divergence in discriminating species and identify-
ing novel taxa [13]. The sampling of data over a target region
of DNA provides data for the definition of prescribed levels
of inter and intraspecific variability, and therefore identifying
novel taxa. However, numerous authors have cautioned the
use of such an approach, particularly within undersampled
taxa [23,24], where levels of intra and interspecific variation
may overlap and prevent accuracy. Our third aim was to
estimate the amount of genetic variation found within and
between recognised species, with a view to commenting on
the use of the threshold approach in the designation of novel
species and identification of species affinity of forensically
important calliphorids.
This study gathered data on a variety of forensically
significant calliphorids across their geographical distributions
and assessed the potential for the COI gene to provide
distinction between the species. Sequencing of 1167 base pairs
of the mitochondrial COI gene was conducted and phylogenetic
analysis used to represent the relationships between the taxa.
This study considered 119 flies from 28 species and 22
countries, including 47 new sequences.
2. Materials and methods
2.1. Samples
Flies were obtained from a variety of locations, either trapped by the authors
using liver-baited traps or kindly supplied by colleagues. They were identified
using traditional morphological characters. Specimens used in this study are
listed in Table 1.
2.2. DNA extraction
DNA was extracted from the flight muscles of specimens using a DNEasy
Tissue Kit (Qiagen) according to manufacturer’s instructions, with an overnight
incubation step.
69
2.3. Amplification
Approximately 1270 bp of the COI gene was amplified using the primers
C1-J-1718 (50 –30 GGAGGATTTGGAAATTGATTAGTTCC) and TL2-N-3014
(50 –30 TCCAATGCACTAATCTGCCATATTA) [14]. For the amplification of
some species, TL2-N-3014 proved problematic and therefore a degenerate
primer TL2-N-3014MOD (50 –30 TCCATTGCACTAATCTGCCATATTA) was
designed based on the sequence of Chrysomya chloropyga (accession number
AF352790), and used to amplify a number of individuals for which amplifica-
tion was not achieved with the original reverse primer.
The PCR reaction mix composed of: 1X PCR buffer (Biotools; Fisher
Biotec), 200 mM dNTPs (Biotools; Fisher Biotec), 1.5 mM MgCl2, 25 pM each
primer, 1 unit of Taq polymerase (Biotools; Fisher Biotec), 10–150 ng of
template DNA, and water added to a total volume of 50 mL. Reactions were
performed on Perkin Elmer GeneAmp PCR System 2400 and Applied Bio-
systems GeneAmp PCR System 2700 thermocyclers. Cycling conditions were:
90 s 94 8C denaturation, followed by 36 cycles of: 94 8C for 22 s, 48 8C for 30 s
and 72 8C for 80 s. A final extension period of 1 min at 72 8C was used,
followed by holding at 4 8C. Products were visualised using 1.5% agarose gels
with ethidium bromide staining and UV transillumination.
PCR products were purified using the QiaQuick PCR Purification Kit
(Qiagen), according to manufacturer’s instructions.
2.4. COI sequencing
Sequencing reactions were performed using the ABI PRISM Big Dye
Terminator 3.0 or 3.1 Sequencing Kit (Perkin Elmer), according to the
manufacturer’s protocol. Cycling conditions for the sequencing reactions were
as per the manufacturer’s recommendations, but the annealing temperature was
lowered to 48 8C. Individuals were sequenced using the external primers C1-J-
1718 and TL2-N-3014 or (TL2-N-3014MOD, where appropriate), and the
internal primers C1-J-2183 (50 –30 CAACATTTATTTTGATTTTTTGG) and
C1-N-2329 (50 –30 ACTGTAAATATATGATGAGCTCA).
2.5. Sequence analysis
Sequences were visualised using Chromas v1.43 (http://trishul.sci.gu.e-
du.au/conor/chromas.html), and alignments and editing conducted using
DAPSA [15]. Sequences were submitted to DDBJ (accession numbers in
Table 1). Additional COI sequences of some relevant calliphorids were obtained
from the publicly available DNA database (Genbank) at www.ncbi.nlm.nih.gov
(Table 1).
Phylogenetic analyses were performed using MEGA 3.0 [16], PAUP*4.0
[17] and MrBayes v.3.0b4 [18]. MEGA was used to calculate pairwise distances
and was employed in the distance analysis using the neighbour-joining method
with the Tamura-Nei model of substitution and 500 bootstrap replications. Base
frequencies, transition/transversion ratio and the gamma shape parameter were
estimated from the data using PAUP, and analyses were performed using
MrBayes v.3.0b4. These Bayesian inference analyses were conducted using
one cold and three hot chains, and the INVGAMMA model. Analyses were run
for 1,500,000 generations, sampling every 100 generations. The likelihood
scores from every 100 generations was plotted to evaluate when stationarity had
been reached. From the plots, it appeared that the burn-in phase was complete
by 50,000 generations. However, the first 1000 trees were excluded as burn-in,
this exclusion being considered to be conservative. Posterior probabilities (PP)
were calculated from the remaining trees by means of a majority rule consensus
tree produced using PAUP.
A further analysis was conducted using 13 individuals of Ch. rufifacies,
using MEGA to perform neighbour-joining analysis with 500 bootstrap replica-
tions. Hydrotaea rostrata was the assigned outgroup in all analyses, with the
exception of the Ch. rufifacies distance analysis.
3. Results
A total of 47 individuals were sequenced and aligned over
1167 base pairs of the COI gene. A further 72 additional
70 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76

Fig. 1. Bayesian inference tree constructed from 1167 bp of COI data. Posterior probabilities are indicated on nodes. Ch = Chrysomya; C = Calliphora;
Co = Cochliomyia; and L = Lucilia.

sequences were obtained from Genbank for comparative 3.1. Identification

purposes. The sequences correspond to positions 1776-2942
of the Drosophila yakuba mitochondrial genome (accession
number NC_001322). No insertions or deletions were located
over this region. Of the 388 variable positions identified, 318 of
these were considered parsimony-informative.
The muscid outgroup, Hydrotaea rostrata, was clearly
separated from the calliphorids in the Bayesian Inference tree
(Fig. 1). Furthermore, the calliphorid species were correctly
assigned to the sub-families Chrysomyinae, Luciliinae or
 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76 71

Calliphorinae, with each of the sub-families monophyletic.
Good posterior probability (PP) support for a sister-group
relationship between the Calliphorinae and Luciliinae was
obtained, but support for a monophyletic Chrysomyinae was
weak.
The genus-level arrangement accurately reflected the
affiliations of the species, and was supported by PP values
above 75%.
At the species level, all of the species has PP support of over
94% except Ch. megacephala (paraphyletic, but with a large
internal clade with 83% support), L. sericata (81%), and Ch.
putoria (69%). Most specimens were accurately assigned to
their respective species (Fig. 1). The exceptions were C. stygia
and C. albifrontalis, which were intermingled; Ch. mega-
cephala, which formed a paraphyletic grade with respect to a
monophyletic (C. rufifacies + C. albiceps) and a monophyletic
Ch. saffranea (Fig. 1) because of two Malaysian specimens
(AY909052 and AY909053); and Lucilia cuprina, which
formed two distinct clades that were collectively paraphyletic
with respect to L. sericata (Fig. 1). One clade of L. cuprina
consisted of individuals from Australia, Senegal and Uganda,
while the other represented Taiwan, Thailand and Hawaii. The
latter group were more probably related to the L. sericata clade
than to conspecific individuals from the other clade.
3.2. Intraspecific variation
Where it could be calculated, all values for intraspecific
variation (Table 2) fell below 0.8%, with the exception of L.
cuprina and Ch. rufifacies. Chrysomya rufifacies showed two
well-supported subgroups (Fig. 1). A separate neighbour-
joining analysis of 722 bp of COI data for this species allowed
an analysis including more individuals from the public
databases. A radial tree (Fig. 2) effectively illustrated the
subdivision within the cluster, with Malaysian and Taiwanese
individuals forming at least one group distinct from individuals
from Australia and the United States. Pairwise comparison
indicated that when considering the species as two clades, one
containing the Malaysian and Taiwanese individuals, and the
remaining individuals as a group, 0.94% variation was
observed.
Pairwise calculation indicated a relatively high value of
3.94% variation across the two clades of L. cuprina. However,
variation within the two clades was much smaller (Table 2).
Unfortunately, insufficient sequences were available for a more
detailed analysis.
Within each species, individuals from the same locality were
intermixed with specimens from other sites (Fig. 1), and no
geographical patterns were obvious.
3.3. Interspecific variation
Levels of interspecific variation between calliphorid species
varied from 0.23 to 13.34% (Table 3). Species pairs such as Ch.
rufifacies/Ch. albiceps and Ch. chloropyga/Ch. putoria were
separated by 3.64% and 2%, respectively. Calliphora dubia and
its sister species C. augur were clearly differentiated, yet
displayed only a 1.27% difference. Lucilia sericata differed
from L. cuprina by 2.8%, yet by only 0.93% from the Asian
clade. In general, species were separated by at least 3%, with
the exception of some closely related pairings. Chrysomya
megacephala and Ch. saffranea differed by only 0.23%.
4. Discussion
4.1. Identification of taxa

Table 2 This study showed that the COI successfully distinguished

Maximum intraspecific variation expressed as a percentage of the total of 1167
base pairs of COI data
all but four of 27 species, even when conspecific specimens
were drawn from well-separated parts of the geographical
Species Maximum %variation within species distributions. Genera were clearly separated and the subfamilial
C. albifrontalis 0.26 arrangement reflected morphological findings commonly
C. stygia 0.60 reported in most taxonomic literature, with the Luciliinae
C. ochracea 0.17
and Calliphorinae grouping together and Cochliomyia posi-
C. dubia 0.26
C. augur 0.69 tioned basally within the Chrysomyinae.
C. vicina 0.26 The sister species Ch. putoria and Ch. chloropyga are
L. illustris 0.77 Afrotropical blowflies associated with latrines and carrion,
L. cuprina 3.94 respectively. The two were long treated as synonymous
Asian clade 0.48
despite evidence of ecological and reproductive separation
Afro-Australian clade 0.30
L. sericata 0.26 [19,20]. Despite recent recognition of their status as distinct
Co. macellaria 0.34 species [20], the morphological characters used to distinguish
Ch. putoria 0.60 them are of little use with eggs and early instars [19]. Wells
Ch. chloropyga 0.51 et al. [19] sequenced the two species over 593 bp of the COI
Ch. marginalis 0.26
gene, and maximum parsimony analysis determined Ch.
Ch. rufifacies 1.37
Asian clade 0.64 chloropyga to be paraphyletic with respect to Ch. putoria. In
Americo-Australian clade 0.18 this study, with twice as many characters, both species were
Ch. albiceps 0.51 monophyletic. Apparently the increased size of the dataset
Ch. varipes 0.77 contributed valuable distinguishing nucleotide information.
Ch. megacephala 0.34
Our results confirm the identity of Ch. putoria in South
Ch. saffranea 0.18
America (Fig. 1).
72 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
Fig. 2. Radial neighbour-joining tree based on 1167 bp of COI data for Chrysomya rufifacies individuals from a variety of localities.
The sister species Ch. rufifacies and Ch. albiceps are
particularly distinct, despite their morphological similarity.
This is in line with an estimate that the rufifacies lineage is at
least 4 million years old [12]. They are usually placed together
as the sister clade to the rest of their genus (e.g. [12]), and their
position in Fig. 1 may be the cause of the low PP support for
Chrysomya in our analysis.
Chrysomya megacephala and C. saffranea are generally
regarded as morphologically and genetically similar [21,22] but
ecologically distinct [23]. In this study, their maximum
interspecific sequence variation was only 0.33% even though
COI is a relatively fast-evolving gene. By comparison, variation
within C. megacephala was 0.34% (Table 2). Explanations for
apparent paraphyly include misidentification, hybridisation and
incomplete lineage sorting. Neither hybridisation nor mis-
identification would produce the very low interspecific genetic
distance (Table 2) and tree topology (Fig. 1) found in this study.
Wallman et al. [12] compared one individual each of Ch.
saffranea (Queensland) and Ch. megacephala (New South
Wales) and reported only 0.4% variation between them across a
variety of regions, including 822 bp of the COI gene. It may be
that the mtDNA regions sequenced to date, which include
3008 bp from the COI, COII, ND4 and ND4L genes [12], are
not useful in distinguishing these species, perhaps as a result of
relatively recent speciation: they are estimated to have
originated within the last million years [12]. Since this is
probably insufficient time to complete lineage sorting, it is not a
surprise that C. megacephala is paraphyletic with respect to C.
saffranea. Funk and Omland [24] have assessed that 23% of
metazoan species are not molecularly monophyletic. Fortu-
nately, the very low variation in C. saffranea makes this species
sufficiently distinctive to be identifiable. As discussed already,
the position of the (Ch. rufifacies + Ch. albiceps) clade within
C. megacephala is probably an artefact, and does not affect the
identifiability of any of these species.
The genus Calliphora is well-represented by a clade of
endemic species in the Australasian region, but it is difficult to
separate some of these based on molecular data. Using four
Table 3
Calculated raw interspecific distances using a neighbour-joining approach with Tamura-Nei model of substitution
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
1 Ch.megacephala
2 Ch.saffranea 0.23

M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76

3 Ch.rufifacies 7.56 7.75
4 Ch.albiceps 6.9 7.07 3.64
5 Ch.chloropyga 6.12 6.04 8.79 8.23
6 Ch.putoria 5.21 5.13 8.05 7.42 2
7 Ch.marginalis 5.77 5.78 7.73 8.29 5.99 4.62
8 Ch.norrisi 5.75 5.76 8.04 10.44 7.49 6.27 6.02
9 Ch.pinguis 3.11 3.02 8.81 8.91 6.9 6.12 6.36 6.43
10 Ch.inclinata 5.4 5.32 9.28 8.59 4.2 3.73 6.02 6.95 6.13
11 Ch.bezziana 5.37 5.45 8.53 8.13 7.3 5.92 5.85 6.91 5.99 6.59
12 Ch.varipes 7.07 7.07 8.76 7.68 8.1 7.08 6.89 6.24 8.35 7.16 7.75
13 Co.hominivorax 10.25 10.23 12.81 11.71 11.29 10.95 11.93 10.96 11.93 10.91 11.16 11.97
14 Co.macellaria 8.74 8.77 12.04 11.4 10.18 8.94 9.13 8.56 8.73 8.99 9.4 10.11 8.65
15 C.stygia 12.06 12.05 13.29 13.19 12.54 11.79 12 10.95 12.78 12.02 12.51 12.33 12.17 11.56
16 C.albifrontalis 12.02 12.01 13.34 13.22 12.65 11.82 12.1 10.98 12.67 12.05 12.46 12.37 12.12 11.41 0.32
17 C.dubia 10.26 10.36 11.79 11.67 10.3 9.79 10.21 9.94 11.15 10.45 9.93 10.43 11.36 11.71 7.82 7.75
18 C.augur 10.43 10.59 11.76 11.92 10.59 10.27 10.61 10.41 11.52 10.72 10.45 10.72 11.72 12.25 8.23 8.31 1.3
19 C.vicina 10.52 10.51 11.66 11.3 10.08 9.61 10.05 10.19 10.98 9.8 9.98 11.38 10.73 12.19 9.2 9.13 7.69 8.37
20 C.vomitoria 10.85 10.63 11.8 11.75 9.88 9.34 10.12 9.53 12 9.97 10.3 11.99 11.28 11.91 8.31 8.22 6.82 7.47 4.49
21 C.varifrons 10.62 10.61 11.84 12.05 11.06 10.26 10.58 10.29 11.57 10.97 10.74 11.19 12.06 11.92 8.42 8.45 6.7 6.97 7.99 7.95
22 C.hilli 10.67 10.78 12.11 11.34 11.48 11.46 11.68 10.44 12.13 11.74 10.67 11.78 12.04 13.31 9.51 9.54 6.31 6.91 8.25 8.9 4.94
23 C.ochracea 11.84 11.83 12.14 11.24 11.73 10.94 11.46 10.68 12.45 12.19 11.7 11.99 12.43 12.97 8.59 8.56 8.03 8.49 8.82 8.74 8.25 8.06
24 L.sericata 8.79 8.82 11.72 10.97 10.44 9.71 8.69 9.05 10.18 9.99 10.19 10.29 11.18 9.64 9.85 9.81 9.06 9.61 8.67 8.46 9.53 10.59 9.98
25 L.cuprina 8.91 8.98 11.04 10.36 10.49 9.72 8.37 8.95 10.52 9.76 9.81 10.21 11.57 10.12 10.61 10.63 8.81 9.03 9.14 8.83 9.55 10.46 10.06 2.8
26 Asian clade 9.27 9.37 11.8 11 10.61 9.88 8.72 8.98 10.75 10.16 10.36 10.24 11.5 9.73 10.03 9.99 9.46 9.73 8.82 8.61 9.39 10.74 10.08 0.93
27 Afro-Australian clade 8.75 8.81 10.7 10.08 10.43 9.65 8.21 8.93 10.42 9.58 9.56 10.2 11.61 10.29 10.87 10.91 8.52 8.72 9.28 8.93 9.62 10.34 10.05 3.63 4.03
28 L.illustris 9.73 9.63 11.18 10.69 10.29 9.21 8.74 9.51 10.1 9.12 10.11 10.59 12.24 9.14 10.19 10.23 8.96 9.81 9.07 8.57 10.21 10.74 9.79 5.53 6.25 5.69 6.5
29 L.ampullacea 11.6 11.51 12.44 12.39 12.29 11.5 11.5 10.28 12.23 11.2 12.52 10.82 11.52 11.65 9.53 9.67 9.99 10.35 10.36 9.41 10.54 11.21 10.14 6.54 7.85 6.88 8.27 5.99
30 H.rostrata 16.97 17.12 15 14.42 17.65 16.89 16.53 15.16 17.66 17.94 17.23 16.46 17.24 17.03 15.89 15.97 15.7 15.6 16.64 16.22 16.09 16.27 16.12 15.74 15.97 16.03 15.94 16.69 16.06

73
74 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
genes, Wallman et al. [12] had difficulty separating members of
the C. hilli group (including C. varifrons), C. augur/C. dubia
and C. stygia/C. albifrontalis. In this study C. hilli and C.
varifrons were clearly distinguished with 4.94% interspecific
variation. It appears that the three additional genes used by
Wallman et al. [12] obscured the signal from the COI gene, but
our result may equally be due to the increase in size of the COI
region we used.
Like the previous species, C. stygia and C. albifrontalis are
morphologically similar flies that speciated less than a million
years ago [12], and appear to be present together in certain
locations [4]. The COI data in this study did not successfully
distinguish between the two species, which are mutually
polyphyletic. New regions of DNA such as the internal
transcribed spacer (ITS) regions are more useful in distinguish-
ing such closely related species (Harvey, unpublished data).
The sister species Calliphora augur and C. dubia also
formed monophyletic clades, although the latter received no PP
support (Fig. 1). Wallman et al. [12] found a comparable pattern
of PP support, and estimated the pair to have diverged just over
a million years ago. These results imply that lineage sorting of
COI in blowflies may take about two million years, and that a
faster-evolving molecular marker is needed for younger
species.
The sister species L. sericata and L. cuprina are
morphologically similar, yet on the basis of this data are
separated quite convincingly. High support was obtained for L.
sericata as distinct from L. cuprina.
4.2. Geographical variation
The molecular taxonomic facet of forensic entomology
generally accumulates genetic data for a DNA region suitable
for identifying all species of forensic importance in a specific
locality, or for global specimens of a focus taxon. Our results
show that geographical variation was not more pronounced in
species sampled from larger geographical distributions. For
example, Ch. albiceps and Ch. megacephala show levels of
intraspecific variation comparable with C. augur, C. stygia and
C. chloropyga (Table 2). Furthermore, no geographical pattern
of relationship was evident in most species (Fig. 1), except Ch.
rufifacies and L. cuprina. Part of the explanation for this may be
that most carrion-breeding flies are synanthropic, and are
spread by human activities. This is illustrated by the occurrence
of European species like C. vicina in Tasmania and L. sericata
in New Zealand, and the African species Ch. putoria in South
America (Fig. 1). This is reassuring for the use of COI
sequences in identification of blowflies, because it implies that
most of the variation within a species can be captured by
relatively localised samples.
4.3. Cryptic species
The geographical structuring of the C. rufifacies and L.
cuprina clades is linked with unusually high intraspecific
variation that may be suggestive of the presence of currently
unrecognised taxa.
Chrysomya rufifacies is primarily a tropical Australasian and
Oriental fly, but the species now also inhabits areas of the
United States [25]. The behaviour of Ch. rufifacies is variable
across its distribution [5,25]. Such variable behaviour may be a
result of ecological interaction with other carrion-colonising
species. It may also be that a species displaying a wide
geographic distribution and variable behaviour might also
display correlated genetic variation.
Based on data from mitochondrial genes, Wallman et al. [12]
suggested that there are two sibling species within Ch.
rufifacies in Australia, one in the south-east and one in the
north-east. Our Western Australian individuals seem to be part
of the south-east taxon. However, combining the COI data from
both studies (Fig. 2) showed that the Western Australian,
Tasmanian and United States individuals at most constitute
only one slightly variable clade (Table 2). Our analyses (Figs. 1
and 2) provide evidence that the Malaysian and Taiwanese
individuals formed a separate cluster from the Australian/
American clade. Clearly, the evidence for cryptic species within
Australia must come from the other mitochondrial genes
sequenced by Wallman et al. [12]. Given the seminal nature of
current molecular studies of the Calliphoridae, it is risky to
form conclusions about species status from a single-gene
phylogeny and limited sampling, but the occurrence of an Asian
clade and an Australian/American clade is also indicated by
preliminary nuclear internal transcribed spacer (ITS) data
(Harvey, unpublished data).
Lucilia cuprina is classified into two subspecies, L. c.
cuprina (Wiedemann) and L. c. dorsalis Robineau-Desvoidy
[26]. The former inhabits the New World, Asia, Indonesia, and
Oceania, while the latter is Afrotropical and Australasian
[27,28]. The two are found to readily interbreed in the
laboratory, and hybrid populations are suggested to exist in
areas of Australia [26]. The observation of two paraphyletic
clades of L. cuprina in this study could be explained by genetic
variation between subspecies. The clade from Australia and
Africa would represent L. c. dorsalis, and the Pacific
individuals L. c. cuprina.
Wallman et al. [12] suggested based on mitochondrial genes
that their Western Australian L. cuprina may be L. c. dorsalis,
while their New South Wales and Queensland individuals
represented L. c. cuprina. They rightly indicated that, given the
history of hybrid populations on the east coast of Australia, this
cannot be concluded with certainty, but suggest that cuprina
and dorsalis lineages are separate species, an interpretation
supported by the paraphyly found in their analyses and ours. In
our study, the Townsville, Queensland individual appears to be
L. c. dorsalis, so that it is likely that both taxa are resident in
Queensland and extensive sampling and study is required, given
the tendency of the variants to interbreed.
5. Conclusions
The molecule-based identification of calliphorids relies on
the location of unique stretches of DNA sequence that are
common to (at least sub-sets of) all members of the chosen
taxon, yet distinct from all other taxa. This necessitates
 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
extensive sequencing of conspecific individuals to verify the
robust nature of markers chosen as species identifiers. This
study indicates that most of the variation within a species may
be adequately captured by samples of as few as 10
geographically distant conspecific individuals. This may be
partly attributed to the high mobility of blowflies.
Some species show more intraspecific variation than
others, and the discovery of an outlying sample must address
the question of whether it is an extreme example, or if it
represents an otherwise unsampled species. It is tempting to
seek an empirical heuristic threshold [29,30], but the
differentiation of species on predetermined thresholds of
genetic divergence relies on the degree of overlap between
intraspecific variation and interspecific divergence. While
thresholds of intraspecific variation and interspecific diver-
gence have been used in other studies to delimit species
boundaries and infer novel species, this is a fraught practice
where taxa are undersampled [29], as they are in the
Calliphoridae. Moritz and Ciccero [30] state that threshold
overlap is greater where a large proportion of closely related
species are included, such as the Australian Calliphora
species. In a review of the use of percent DNA sequence
difference across the COI in insects, Cognato [31] revealed
that such thresholds may fail up to 45% of the time when
used to diagnose species due to overlap in inter and
intraspecific sequence divergences. Harvey et al. [6] and
Wells and Sperling [32] reported levels of 0.8% maximum
intraspecific variation and 3% minimum interspecific
divergence in the Calliphoridae, but in this study these
thresholds overlap.
It is suggested that forensic entomologists heed the
published cautions against the use of percent DNA sequence
divergence in species delimitation. A concerted effort in the
field to gather the necessary data to sufficiently sample the
relevant taxa, identify the limitations of the COI and locate
alternate regions for identification in recently diverged taxa will
allow the relationships revealed in molecular phylogenies to
distinguish individuals. The designation of novel species is not
likely based solely on molecular bases. Paraphyly and
polyphyly may provide an indication of some species process,
however, they may likewise represent an inappropriate gene
choice for the specific focus taxa, and further study will be
necessary. Unpublished data (Harvey) has indicated that the
COI para/polyphyly observed in some calliphorid taxa may be
resolved through an alternate gene choice, and it is such study
that indicates the importance of thorough global study of
calliphorids of interest if principles of monophyly are to be
enforced in identification.
This study has again illustrated the enormous potential for
the use of the COI gene for distinguishing between forensically
significant calliphorid species, considering these species from a
global perspective and linking allopatric populations of species
often considered in isolation. This study has supported the
existence of sibling species within L. cuprina, and illustrated
the potential unsuitability of the COI gene for distinction
between young species like Ch. saffranea and Ch. mega-
cephala, and C. stygia and C. albifrontalis, probably as a result
75
of incomplete lineage sorting. This may necessitate the
adoption of a secondary assay for distinction of such species,
perhaps the ITS regions.
Sequencing of individuals from a wide variety of localities,
and thus the pooling of data from various researchers, will
greatly contribute to the taxonomy of calliphorids on a global
scale. Molecular forensic entomology, while in its infancy, has
increased the ability to identify unknown immature calliphor-
ids. It is the extensive sampling of populations and more species
that will ultimately contribute to the increased relevance of the
field to the courtroom.
Acknowledgements
We thank numerous people for their generous donation of
flies for this study: Geoff Allen, Dallas Bishop, Amoret Brandt,
M. Lee Goff, Ana Carolina Junqueira, Gary Levot, Nicola Lunt,
Mervyn Mansell, Nolwazi Mkize, Carol Simon, Tony Postle,
Cameron Richards, Kabew Sukontason and Ruxton Villet.
Jennifer Roy provided technical assistance. Some field work
was funded by Rhodes University grants to MHV. MLH was
supported in part by Rotary International, and the American
Australian Association.
References
[1] F.A.H. Sperling, G.S. Anderson, D.A. Hickey, A DNA-based approach to
the identification of insect species used for postmortem interval estima-
tion, J. Forensic Sci. 39 (1994) 418–427.
[2] Y. Malgorn, R. Coquoz, DNA typing for identification of some species of
Calliphoridae: an interest in forensic entomology, Forensic Sci. Int. 102
(1999) 111–119.
[3] J. Stevens, R. Wall, Genetic relationships between blowflies (Calliphor-
idae) of forensic importance, Forensic Sci. Int. 120 (2001) 116–123.
[4] J.F. Wallman, S.C. Donnellan, The utility of mitochondrial DNA
sequences for the identification of forensically important blowflies (Dip-
tera: Calliphoridae) in southeastern Australia, Forensic Sci. Int. 120
(2001) 60–67.
[5] M.L. Harvey, I.R. Dadour, S. Gaudieri, Mitochondrial DNA cytochrome
oxidase I gene: potential for distinction between immature stages of some
forensically important fly species (Diptera) in western Australia, Forensic
Sci. Int. 131 (2003) 134–139.
[6] M.L. Harvey, M.W. Mansell, M.H. Villet, I.R. Dadour, Molecular identi-
fication of some forensically important blowflies of southern Africa and
Australia, Med. Vet. Entomol. 17 (2003) 363–369.
[7] S.D. Ratcliffe, D.W. Webb, R.A. Weinzierl, H.M., Robertson. PCR-RFLP
identification of Diptera (Calliphoridae, Muscidae and Sarcophagidae)—a
generally applicable method, J. Forensic Sci. 48(4) (2003) 1–3.
[8] J. Stevens, R. Wall, Genetic variation in populations of the blowflies
Lucilia cuprina and Lucilia sericata (Diptera: Calliphoridae). Random
amplified polymorphic DNA analysis and mitochondrial DNA sequences,
Biochem. Systematics Ecol. 25 (1997) 81–97.
[9] S. Vincent, J. Vian, M.P. Carlotti, Partial sequencing of the cytochrome
oxidase b subunit gene I: a tool for the identification of European species
of blow flies for postmortem interval estimation, J. Forensic Sci. 45 (2000)
820–823.
[10] D.H. Lunt, D.-X. Zhang, J.M. Szymura, G.M. Hewitt, The insect cyto-
chrome oxidase I gene: evolutionary patterns and conserved primers for
phylogenetic studies, Insect Mol. Biol. 5 (1996) 153–165.
[11] C. Garros, R.E. Harbach, S. Manguin, Systematics and biogeographical
implications of the phylogenetic relationships between members of the
funestus and minimus groups of Anopheles (Diptera: Culicidae), J. Med.
Entomol. 42 (2005) 7–18.
76 M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76
[12] J.F. Wallman, R. Leys, K. Hogendoorn, Molecular systematics of Aus-
tralian carrion-breeding blowflies (Diptera: Calliphoridae) based on mito-
chondrial DNA, Invertebrate Systematics 19 (2005) 1–15.
[13] P.D.N. Hebert, M.Y. Stoeckle, T.S. Zemlak, C.M. Francis, Identification of
birds through DNA barcodes, PLoS Biol. 2 (2004) e312.
[14] C. Simon, F. Frati, A. Beckenbach, B. Crespi, H. Liu, P. Flook, Evolution,
weighting, and phylogenetic utility of mitochondrial gene-sequences and
a compilation of conserved polymerase chain-reaction primers, Ann.
Entomol. Soc. Am. 87 (1994) 651–701.
[15] E.H. Harley, Dapsa (DNA and Protein Sequence Alignment), Department
of Chemical Pathology, University of Cape Town, Cape Town, South
Africa, 1996.
[16] S. Kumar, K. Tamura, I.B. Jakobsen, M. Nei, MEGA2: Molecular
Evolutionary Genetics Analysis software, Bioinformatics 17 (2001)
1244–1245.
[17] D.L. Swofford, PAUP* Phylogenetic Analysis Using Parsimony (*and
other methods). Version 4, Sinauer Associates, Sunderland, Massachu-
setts, 2002.
[18] J.P. Huelsenbeck, F. Ronquist, MRBAYES. Bayesian inference of phy-
logeny, Bioinformatics 17 (2001) 754–755.
[19] J.D. Wells, N. Lunt, M.H. Villet, Recent African derivation of Chrysomya
putoria from C. chloropyga and mitochondrial DNA paraphyly of cyto-
chrome oxidase subunit one in blowflies of forensic importance, Med. Vet.
Entomol. 18 (2004) 445–448.
[20] K. Rognes, H.E.H. Paterson, Chrysomya chloropyga (Wiedemann, 1818)
and Chrysomya putoria (Wiedemann, 1830) (Diptera, Calliphoridae) are
two different species, African Entomol. 13 (2005) 49–70.
[21] N. Evenhuis (Ed.), Catalog of the Diptera of the Australasian
and Oceanian Regions, Bishop Museum Special Publication, 1989 ,
p. 86.
[22] F.H. Ullerich, M. Schöttke, Karyotypes, constitutive heterochromatin, and
genomic DNA values in the blowfly genera Chrysomya, Lucilia, and
Protophormia (Diptera: Calliphoridae), Genome 49 (6) (2006) 584–597.
[23] H. Kurahashi, Dispersal of filth flies through natural and human agencies:
origin and immigration of a synanthropic form of Chrysomya megace-
phala, in: M. Laird (Ed.), Commerce and the spread of pests and disease
vectors, Praeger, New York, 1984.
[24] D.J. Funk, K.E. Omland, Species-level paraphyly and polyphyly: fre-
quency, causes, and consequences, with insights from animal mitochon-
drial DNA, Annu. Rev. Ecol., Evol. Systematics 34 (2003) 397–423.
[25] D.L. Baumgartner, Review of Chrysomya rufifacies (Diptera: Calliphor-
idae), J. Med. Entomol. 30 (1993) 338–352.
[26] K.R. Norris, Evidence for the multiple exotic origin of Australian popula-
tions of the sheep Blowfly, Lucilia cuprina (Wiedemann) (Diptera:
Calliphoridae), Aust. J. Zool. 38 (1990) 635–648.
[27] B. Montgomery, Elucidation of the evolutionary status of Queensland
coastal Lucilia cuprina (Wiedemann). Master of Science Thesis. Uni-
versity of Queensland (1990).
[28] J. Stevens, R. Wall, Species, sub-species and hybrid populations of the
blowflies Lucilia cuprina and Lucilia sericata (Diptera: Calliphoridae),
Proc. Roy. Soc. Lond. B 263 (1996) 1335–1341.
[29] C.P. Meyer, G. Paulay, DNA barcoding: error rates based on comprehen-
sive sampling, PLoS Biol. 3 (2005) 2229–2238.
[30] C. Moritz, C. Ciccero, DNA barcoding: promise and pitfalls, PLoS Biol. 2
(2004) 1529–1531.
[31] A.I. Cognato, Standard percent DNA sequence difference for insects
does not predict species boundaries, J. Economic Entomol. 99 (2006)
1037–1045.
[32] J.D. Wells, F.A.H. Sperling, DNA-based identification of forensically
important Chrysomyinae (Diptera: Calliphoridae), Forensic Sci. Int. 120
(2001) 110–115.
[33] W.-Y. Chen, T.-H. Hung, S.-F. Shiao, Molecular identification of foren-
sically important blow fly species (Diptera: Calliphoridae) in Taiwan, J.
Medical Entomol. 41 (2004) 47–57.
[34] J.R. Stevens, The evolution of myiasis in blowflies (Calliphoridae), Int. J.
Parasitol. 33 (2003) 1105–1113.