harvey et al., 2008_a global study of forensically significant calliphorids

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

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, Australiab 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

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

majority of molecular studies, however, have used the

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

www.elsevier.com/locate/forsciint

Available online at www.sciencedirect.com

Forensic Science International 177 (2008) 66–76

* 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: [email protected] (M.L. Harvey).

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

doi:10.1016/j.forsciint.2007.10.009


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]


Pretoria, South Africa AB112848 [6]

Kitwe, Zambia AB112861 [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]


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]


Karoo, South Africa AB112866 [6]

Karoo, South Africa AB112862 [6]

Karoo, South Africa EU418543 New sequence



Chrysomya varipes Gladstone, Australia EU418544 New sequence

Sydney, Australia EU418545 New sequence

Adelaide, Australia AF295556 [32]




Chrysomya norrisi Wau, Papua New Guinea AF295552 [32]

Chrysomya rufifacies Perth, Australia EU418546 New sequence



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]



Chrysomya albiceps Alexandria, Egypt AF083657 [32]





Deka, Zimbabwe AB112849 [6]

Deka, Zimbabwe AB112858 [6]

Manzini, Swaziland AB112865 [6]



Cochliomyia hominivorax Alfenas, Brazil EU418550 New sequence

Cochliomyia macellaria Salvador, Brazil EU418551 New sequence

M.L. Harvey et al. / Forensic Science International 177 (2008) 66–76 67


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

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


Calliphora hilli Gladstone, Tasmania EU418559 New sequence

Calliphora varifrons Boddington, Australia EU418560 New sequence

Calliphora ochracea Sydney, Australia EU418561 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



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




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


Graaf-Reinet, South Africa AB112850 [6]

Graaf-Reinet, South Africa AB112843 [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.

M.L. Harvey et al. / Forensic Science International 177 (2008) 66–7668


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.

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]. MEGAwas 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



sequences were obtained from Genbank for comparative

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.

3.1. Identification

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

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.



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

This study showed that the COI successfully distinguished

all but four of 27 species, even when conspecific specimens

were drawn from well-separated parts of the geographical

distributions. Genera were clearly separated and the subfamilial

arrangement reflected morphological findings commonly

reported in most taxonomic literature, with the Luciliinae

and Calliphorinae grouping together and Cochliomyia posi-

tioned basally within the Chrysomyinae.

The sister species Ch. putoria and Ch. chloropyga are

Afrotropical blowflies associated with latrines and carrion,

respectively. The two were long treated as synonymous

despite evidence of ecological and reproductive separation

[19,20]. Despite recent recognition of their status as distinct

species [20], the morphological characters used to distinguish

them are of little use with eggs and early instars [19]. Wells

et al. [19] sequenced the two species over 593 bp of the COI

gene, and maximum parsimony analysis determined Ch.

chloropyga to be paraphyletic with respect to Ch. putoria. In

this study, with twice as many characters, both species were

monophyletic. Apparently the increased size of the dataset

contributed valuable distinguishing nucleotide information.

Our results confirm the identity of Ch. putoria in South

America (Fig. 1).

Table 2

Maximum intraspecific variation expressed as a percentage of the total of 1167

base pairs of COI data

Species Maximum %variation within species

C. albifrontalis 0.26

C. stygia 0.60

C. ochracea 0.17

C. dubia 0.26

C. augur 0.69

C. vicina 0.26

L. illustris 0.77

L. cuprina 3.94

Asian clade 0.48

Afro-Australian clade 0.30

L. sericata 0.26

Co. macellaria 0.34

Ch. putoria 0.60

Ch. chloropyga 0.51

Ch. marginalis 0.26

Ch. rufifacies 1.37

Asian clade 0.64

Americo-Australian clade 0.18

Ch. albiceps 0.51

Ch. varipes 0.77

Ch. megacephala 0.34

Ch. saffranea 0.18



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

Fig. 2. Radial neighbour-joining tree based on 1167 bp of COI data for Chrysomya rufifacies individuals from a variety of localities.



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

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

M.L

.H

arvey

eta

l./Fo

rensic

Scien

ceIn

terna

tiona

l1

77

(20

08

)6

6–

76

73


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



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

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.



[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. Schottke, 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.


harvey et al., 2008_a global study of forensically significant calliphorids

Documents