Sandflies of the Phlebotomus perniciosus complex ...hera.ugr.es/doi/15059121.pdf · Sandflies of the Phlebotomus perniciosus complex: mitochondrial introgression and a new sibling

Sandflies of the Phlebotomus perniciosus complex:mitochondrial introgression and a new sibling speciesof P. longicuspis in the Moroccan Rif

B. PESSON* , J . S . READY y, I . BENABDENNBI* , J . MARTIN-SANCHEZ z,

S . ESSEGHIR y, § , M. CADI-SOUSSI{, F . MORILLAS-MARQUEZ z and

P . D. READY y

*Laboratoire de Parasitologie, Faculte de Pharmacie, Universite Louis Pasteur – Strasbourg I, Illkirch, France, yDepartment of

Entomology, Natural History Museum, London, U.K., zFacultad de Farmacia, Universidad de Granada, Granada, Spain,§Laboratoire d’Epidemiologie et d’Ecologie Parasitaire, Institut Pasteur de Tunis, Tunisia and {Faculte de Medecine et de

Pharmacie, Universite Mohamed V, Rabat, Morocco

Abstract. The bloodsucking adult females of Phlebotomus perniciosus Newsteadand P. longicuspis Nitzulescu (Diptera: Psychodidae) are important vectors of theprotozoan Leishmania infantum Nicolle (Kinetoplastida: Trypanosomatidae) inwestern Mediterranean countries. The species status of the two phlebotominesandflies was assessed, along with the epidemiological implications. Individualsandflies from three Moroccan Rif populations were characterized morphologic-ally, isoenzymatically (by the isoelectrofocusing of alleles at the polymorphicenzyme loci of HK, GPI and PGM), and by comparative DNA sequence analysisof a fragment of mitochondrial Cytochrome b (mtDNA). By reference to thecharacter profiles of specimens from other locations, including southern Spainand the type-locality countries, the Moroccan flies were placed in three lineages:first, the lineage of P. perniciosus, which contained two mtDNA sublineages, one(pnt) widely distributed and associated with the morphology of the male typesfrom Malta, and the other (pna) associated with a P. longicuspis-like malemorphology; second, the lineage of P. longicuspis sensu stricto, including typicalforms from Tunisia; and third, a new sibling species of P. longicuspis. ThemtDNA sublineage (pnt) of typical P. perniciosus was also found in someP. longicuspis from Morocco, indicating interspecific hybridization. The typicalrace of P. perniciosus occurs in Italy as well as in Malta, Tunisia andMorocco. It isreplaced in southern Spain by the Iberian race (with the pni mtDNA sublineage).The discovery of interspecific gene introgression and a new sibling species meanthat previous records of the two morphospecies do not necessarily reflect their truevectorial roles or geographical and ecological distributions.

Key words. Phlebotomus longicuspis, morphology, Phlebotomus perniciosus com-plex, sibling species, isoenzymes, mitochondrial introgression, Morocco.

Introduction

In this report we assess the species status of various morpho-

logical forms and populations of two closely related taxa in

the Phlebotomus perniciosus species complex (Esseghir et al.,

2000) and discuss the resulting epidemiological implications.

Correspondence: Dr Paul Ready, Molecular Systematics

Laboratory, Department of Entomology, Natural History

Museum, Cromwell Road, London SW7 5BD, U.K. E-mail:

[email protected]

Medical and Veterinary Entomology (2004) 18, 25–37

# 2004 The Royal Entomological Society 25

Phlebotomus perniciosus Newstead and P. longicuspis

Nitzulescu are classified in the subgenus Larroussius

Nitzulescu (Seccombe et al., 1993) and, in the western

Mediterranean, their haematophagous females are vectors

of zoonotic leishmaniasis caused by Leishmania infantum

Nicolle (Protozoa, Trypanosomatidae). The domestic dog

and wild foxes are the reservoir hosts, and cutaneous and

visceral human disease is frequent in rural communities, both

in south-west Europe and in the Maghreb region of north-

west Africa (Killick-Kendrick, 1990; Rioux & Lanotte, 1990).

Described from Malta (Newstead, 1911), P. perniciosus is

widely distributed in the western Mediterranean Basin:

from Portugal to Croatia in Europe, and from Morocco

to Libya in north Africa, where it is found predominantly in

the sub-humid and semi-arid bioclimate zones (Rioux et al.,

1984). Phlebotomus longicuspis was described from Tunisia

(Nitzulescu, 1930), and until 1982 was recorded only from

Libya and other countries of the Maghreb (Algeria and

Morocco), where it is most frequent in the semi-arid, arid

and per-arid bioclimate zones (Rioux et al., 1984). Then

males were reported from southern Spain, in the regions

of Andalucia (Morillas Marquez et al., 1982) and Murcia

(Martınez Ortega et al., 1982).

For many years considered indistinguishable, the females

of P. perniciosus and P. longicuspis were separated by Leger

et al. (1983) according to the form and position of the dilata-

tion on each individual spermathecal duct: spherical and

basal or heart-shaped and subbasal, respectively. As for the

males, they were originally separated by the terminal struc-

ture of the aedeagi: bifurcated in P. perniciosus (Fig. 1a);

simple and tapering in P. longicuspis (Fig. 1b). However,

this character is subject to much individual variation

(Parrot, 1936b; Morillas Marquez et al., 1991) and the iden-

tity of Spanish P. longicuspis has been questioned (Morillas

Marquez et al., 1991; Collantes & Martınez Ortega, 1997).

In the course of a sandfly survey of the Rif mountains of

north Morocco in 1995 (Benabdennbi & Pesson, 1998),

both species were caught in the same traps set in Loubar,

just outside the town of Chefchaouene (Fig. 2). Although all

the females could be identified as either P. perniciosus or

P. longicuspis by the morphology of their spermathecal

ducts, all the males were at first identified as P. longicuspis.

A careful examination of the aedeagi and the finding of a

new character, the number of semi-deciduous setae on the

inner face of the coxite (Benabdennbi, 1998; Benabdennbi

et al., 1999), permitted the recognition of two distinctive

male morphs: typical P. longicuspis (LC, formerly LC B),

with each aedeagus having a straight, tapering tip (Fig. 1b),

and the coxite bearing 15–29 setae; and P. longicuspis-like

(PNA, formerly LC A), with an incurved aedeagal tip

(Fig. 1c) and 10–16 coxite setae. An isoenzyme analysis

revealed scoreable polymorphic alleles at three loci (hexo-

kinase, glucosephosphate isomerase and phosphogluco-

mutase) and unambiguously grouped the PNA males with

females of P. perniciosus (Benabdennbi et al., 1999). More

recently a RAPD study provided supporting evidence

(Martın Sanchez et al., 2000). However, the population

formed by the LC males and females, with morphology

Fig. 1. SEM images of the male aedeagi. (a) PN morphotype of

P. perniciosus; (b) LC morphotype of P. longicuspis; (c) PNA

morphotype of P. perniciosus.

26 B. Pesson et al.

# 2004 The Royal Entomological Society, Medical and Veterinary Entomology, 18, 25–37

typical of P. longicuspis, showed a deviation from

Hardy–Weinberg equilibrium in its isoenzyme genotype

frequencies, which was significant in the case of phospho-

glucomutase (Benabdennbi et al., 1999), and RAPD pat-

terns remained heterogeneous (Martın Sanchez et al., 2000).

In order to understand the reason for the genetic disequi-

librium that remained within the P. longicuspis analysed

by Benabdennbi et al. (1999), we compared individuals

captured at Chefchaouene with those from two other Rif

localities (Fig. 2), each characterized by a preponderance

of males and females with the morphology typical of one

of the two species: Taounate with 72.5% P. longicuspis, and

Ouezzane with 76.5% P. perniciosus. For this, individuals

were characterized in three ways: morphologically, by

examination of the genitalia; enzymatically, using the

three polymorphic nuclear loci that had permitted the

separation of the two species at Chefchaouene; and with

mitochondrial (mt) DNA haplotyping, by sequencing the 30

end of the Cytochrome b gene, because this marker (nor-

mally maternally inherited) had been used to relate geo-

graphical populations of many species of Larroussius in

the Mediterranean subregion, including P. perniciosus and

P. longicuspis from the countries containing the type local-

ities (Esseghir et al., 1997, 2000; Esseghir, 1998).

With the aim of placing the results from Chefchaouene in

a broader context, we also report variation in allele frequen-

cies at the three polymorphic isoenzyme loci in southern

European populations of P. perniciosus, and we compare

newly characterized mtDNA haplotypes fromMorocco and

Spain with those of the distinctive Iberian sublineage pre-

viously reported from Spain and the typical sublineage

found in Italy, Malta and Tunisia (Esseghir et al., 1997,

2000) (Fig. 2). These additional data help to establish the

stability of the diagnostic characters for P. perniciosus in all

countries except Morocco.

Materials and methods

Sandfly collections

Sandflies were captured overnight in modified miniature

CDC light traps (Madulo-Leblond, 1983) and immediately

conserved in liquid nitrogen. Most of the catches in Morocco

were made 17–25 July 1995 in three localities (Fig. 2). Two

are situated on the southern side of the Rif, at the boundary

of the sub-humid and semi-arid bioclimatic zones: to the

East, Taounate (MT samples; 500 ma.s.l.; 34�360 N,

SPAIN

EH

EG EAEM

MC

MO MT

MOROCCO

TT LC

TM

TS

TUNISIA

GZ PN

MALTA

IP

IMITALY

200 KM

Fig. 2. Geographical locations of the populations studied and type-locality countries of P. perniciosus and P. longicuspis. *¼ localities where

both P. perniciosus and P. longicuspis were collected; *¼ localities where only P. perniciosus was collected; *¼ type localities of P. perniciosus

(PN) and of P. longicuspis (LC). Key to populations (country/region/locality): EH¼Spain/Huelva/Rio Tinto and Higuera de la Sierra;

EG¼ Spain/Granada/Torvizcon, Alfacar and Loja; EA¼Spain/Almerıa/Turre; EM¼Spain/Murcia/Verdolay; IM¼ Italy/Apulia/Monte

Sant’Angelo; IP¼ Italy/Apulia/Parabita; GZ¼Malta/Gozo/Zebbug; TM¼Tunisia/Medjez Al Bab/Goubellat; TS¼Tunisia/Sidi Bouzid/

Felta; TT¼Tunisia/Beja/Tebaba; MO¼Morocco/Rif/Ouezzane; MT¼Morocco/Rif/Taounate; MC¼Morocco/Rif/Chefchaouene.

P. perniciosus complex in Morocco 27


4�400 W); and to the West, Ouezzane (MO samples;

350ma.s.l.; 34�520 N, 5�360 W). The latter is slightly more

humid, as is the third locality, Chefchaouene (MC samples;

600ma.s.l.; 35�120 N, 5�150 W) on the Mediterranean side of

the Rif. Locations and codes of additional material, captured

and cryopreserved in the same conditions, are given in Fig. 2.

Morphology

For each individual, the genitalia were dissected and cleared

in Marc-Andre solution (chloral hydrate/acetic acid), in readi-

ness for morphological identification by compound micro-

scopy (� 400–1000). The remainder of the insect was used

either for mtDNA haplotyping or for isoenzyme analysis, or

the head and thorax were separated for isoenzyme analysis,

leaving only the abdominal remains for haplotyping.

The structure of aedeagi (Fig. 1) were photographed from

males prepared as previously described (Pesson et al., 1994)

and observed with a stereoscan Cambridge 100 scanning elec-

tron microscope (LEO Microscopie Electronique, France).

Isoenzyme analysis

Isoelectrofocusing was carried out in ultrathin agarose gels

(PharmaciaTM

multiphor II, Amersham Biosciences, France)

with the ampholyte at pH4.0–6.5, according to the protocols

described by Pesson et al. (1991). The three discriminating

loci used were: hexokinase (HK, E.C.2.7.1.1), glucosepho-

sphate isomerase (GPI, E.C.5.3.1.9) and phosphoglucomu-

tase (PGM, E.C.5.4.2.2). The alleles, represented by coloured

bands on the gels, were numbered chronologically for species

of the subgenus Larroussius, with the more frequent allele

usually having been discovered first: for HK, allele 1¼ pHi

5.60, allele 2¼ pHi 5.49, allele 3¼ pHi 5.36, allele 4¼ pHi

5.63, allele 5¼ pHi 5.58, allele 6¼ pHi 5.33, and allele

7¼ pHi 5.22; for GPI, allele 1¼ pHi 5.30, allele 2¼ pHi



5.60, and allele 9¼ pHi 5.80; for PGM, allele 1¼ pHi 4.93,

allele 2¼ pHi 5.24, allele 3¼ pHi 5.10, allele 4¼ pHi 5.34,

allele 5¼ pHi 5.47, allele 6¼ pHi 5.15, allele 7¼ pHi 5.29,

and allele 8¼ pHi 5.45.

The allele frequencies, tests for deviation from Hardy–

Weinberg equilibrium, and UPGMA phenetic analysis based on

Nei’s genetic distances were all calculated using BIOSYS-2

(Swofford & Selander, 1981). In addition, GENEPOP

(Raymond & Rousset, 1995) was used to test for genotypic

linkage disequilibrium and estimate FST for each pair of

samples.

PCR amplification, sequencing and analysis of mtDNA

The protocols of Esseghir et al. (1997) were followed: to

extract genomic DNA; to amplify by the polymerase chain

reaction (PCR) the 30 end of the Cytochrome b gene (Cyt b)

using primers CB3-PDR and N1N-PDR; to purify the PCR

product by agarose-gel fractionation and binding to glass

milk (Geneclean Kit, BIO 101 Inc, U.S.A.); to sequence the

amplified fragment using each of the PCR primers and the

ABI PRISMTM

BigDye Cycle Sequencing Ready Reaction

kit (PE Applied Biosystems ¼ ABI, U.S.A.); and, to record

the DNA sequences using the ABI 373 or 377 systems and

protocols. DNA sequences were aligned using SeqEd ver-

sion 1.0.3 software (ABI) or Sequencher (Gene Codes

Corp., U.S.A.). Phylogenetic analyses were performed

with PAUP* (Swofford, 2002).

Results

Morphology

According to the aforementioned criteria, fully described by

Benabdennbi et al. (1999), the males characterized from all

countries (Spain, Italy, Malta, Tunisia and Morocco) could

be assigned to three morphotypes: typical P. perniciosus

(PN; Fig. 1a) in all countries; typical P. longicuspis (LC;

Fig. 1b) only in Tunisia and Morocco; and P. longicuspis-like

P. perniciosus (PNA; Fig. 1c) only in northernMorocco. There

were two female morphotypes: typical P. perniciosus (PN) in

all countries; and typical P. longicuspis (LC) only in Tunisia

and Morocco.

Isoenzyme analysis

Allele frequencies at the three polymorphic loci are

reported for nine populations with the PN/PNAmorphology

of P. perniciosus and for three populations with the LC

morphology of P. longicuspis (Table 1). Only some of the

isoenzyme results for the Chefchaouene (MC) flies were

published previously (Benabdennbi et al., 1999). All

P. perniciosus populations, including the PNA morphotypes,

were in Hardy–Weinberg (HW) equilibrium, and there was

no significant linkage disequilibrium except for two popula-

tions: in Ouezzane, Morocco (MO; HK-GPI, P¼ 0.010) and

in Rio Tinto, Spain (EH; GPI-PGM, P¼ 0.046). In contrast

all three populations of P. longicuspis from Morocco were

not in HW equilibrium at the PGM locus.

Nei’s genetic distances (D) and FST values were calculated

among all populations (Table 2). Nei’s D-values suggest two

groups of P. perniciosus populations, as presented graph-

ically in the UPGMA tree (Fig. 3): (1) Italy, Malta and

Morocco, where Chefchaouene (MC) is distinctive; and (2)

Iberia. The FST results indicate the same geographical struc-

ture and confirm the low gene flow observed between

Moroccan and Iberian populations (Pesson et al., 1998).

The UPGMA tree also illustrates the genetic convergence of

Moroccan P. longicuspis with the Iberian populations of

P. perniciosus, although it is necessary to caution that

P. longicuspis in Morocco does not have the genetic char-

acteristics of a single, reproductively isolated species, as

28 B. Pesson et al.


manifested by the HW disequilibrium at the PGM locus and

other indicators reported below.

mtDNA haplotypes and lineages

The last 279 nucleotides of Cytochrome b (Cyt b) were

sequenced for 91 individual sandflies from Morocco, a

single sandfly from Tunisia and 35 individual sandflies

from Spain, and the sequences were aligned with those

already obtained from other countries (Esseghir et al.,

1997, 2000). Unique sequences (or haplotypes) were identi-

fied (Table 3) by inspection of the input data matrix and the

distance matrix given by PAUP (Swofford, 2002). The haplo-

types fell into three distinct primary lineages (pern, lcus and

lcx), following a phylogenetic analysis based on genetic

distances (Fig. 4). Haplotypes were coded to denote their

lineage and the morphospecies in which they were usually

found: pern for P. perniciosus, lcus for P. longicuspis, and

lcx for a putative sibling species of P. longicuspis. Pairs of

haplotypes differed only by 1–8 nucleotides (0.4–2.9%)

within each of the three primary lineages, whereas the

range of pairwise differences (and absolute genetic dis-

tances) between these lineages was 12–18 nucleotides

(4.3–6.5%). In Table 4, the distributions of the newly

characterized Cyt b sequences from Morocco are related

to morphotype, sex and Cyt b lineage.

The first of the three primary lineages, pern, has six

diagnostic polymorphisms (at nucleotide positions 27, 49,

58, 168, 171 and 258), and these are fixed in all three of its

Table 1. Allelic frequencies at the three polymorphic isoenzyme loci characterized in nine populationsa of P. perniciosus and three

populations of P. longicuspis

Locusb Phlebotomus perniciosus Phlebotomus longicuspis

No. in sample Spain Italy Malta Morocco Morocco

Allele number EH EG EA EM IM GZc MO MT MCd MO MT MC

HK

No. 25 26 18 21 121 151 167 41 133 42 87 114

1 1.000 0.885 0.889 0.929 0.459 0.467 0.317 0.488 0.023 0.988 0.989 0.996

2 0 0.096 0.111 0.071 0.517 0.526 0.581 0.500 0.970 0 0 0

3 0 0.019 0 0 0 0.003 0.102 0.012 0.008 0.012 0 0.004

4 0 0 0 0 0 0 0 0 0 0 0.006 0

5 0 0 0 0 0.025 0 0 0 0 0 0 0

6 0 0 0 0 0 0.003 0 0 0 0 0 0

7 0 0 0 0 0 0 0 0 0 0 0.006 0

P – 1 1 1 1 1 0.874 0.347 1 1 1 –

GPI

No. 22 26 18 21 126 140 147 40 77 44 80 84

1 0.773 0.808 0.889 0.881 0.988 0.975 0.956 0.975 0.961 0.864 0.756 0.518

2 0.205 0.115 0.111 0.071 0.008 0.011 0.034 0.013 0.026 0.080 0.181 0.327

3 0.023 0.038 0 0.048 0 0.007 0.003 0.013 0.006 0.023 0.044 0.065

4 0 0 0 0 0 0.007 0.007 0 0 0 0 0

5 0 0 0 0 0 0 0 0 0.006 0 0.006 0.012

6 0 0.019 0 0 0 0 0 0 0 0.011 0 0

7 0 0 0 0 0 0 0 0 0 0.023 0.013 0.077

8 0 0.019 0 0 0 0 0 0 0 0 0 0

9 0 0 0 0 0.004 0 0 0 0 0 0 0

P 0.271 0.201 0.166 1 1 1 0.242 1 0.096 0.159 0.770 0.518

PGM

No. 22 26 18 21 124 132 173 39 114 49 88 100

1 0 0 0 0 0.004 0 0 0.013 0 0 0 0

2 0.977 0.981 1.000 0.976 0.980 0.939 0.908 0.923 0.697 0.571 0.756 0.440

3 0 0 0 0 0 0.011 0.064 0.026 0 0 0 0.005

4 0.023 0.019 0 0.024 0.012 0.042 0.020 0.038 0.285 0.184 0.193 0.290

5 0 0 0 0 0 0 0 0 0 0.245 0.051 0.265

6 0 0 0 0 0 0 0.009 0 0 0 0 0

7 0 0 0 0 0 0.008 0 0 0.013 0 0 0

8 0 0 0 0 0.004 0 0 0 0.004 0 0 0

P 1 1 – 1 1 0.383 0.367 1 0.658 < 0.001 < 0.001 < 0.001

aOrigins and their abbreviations are given in Fig. 1.bHK ¼ hexokinase, GPI ¼ glucose phosphate isomerase and PGM ¼ phosphoglucomutase; P ¼ probability of w2 value occurring by chance,

when testing for deviation from Hardy–Weinberg expectations of genotype frequencies.cSome data from Leger et al. (1991). dPNA is the only male morphotype recorded from this locality; some data from Benabdennbi et al. (1999).



sublineages, two of which were found in the Moroccan Rif

(Table 3; Fig. 4). Most P. perniciosus from the Rif had the

typical morphology (PN) and the Cyt b haplotype pern01,

which forms with haplotype pern02 the typical sublineage

(pnt) previously isolated only from Italy, Malta and

Tunisia. The typical sublineage of P. perniciosus had been

distinguished from the Iberian sublineage (pni) by two

polymorphisms, at nucleotide positions 147 and 246

(diagnostic) (Esseghir et al., 1997, 2000; Table 3). The

fixation of these polymorphisms in the Iberian sublineage

was confirmed by characterizing the Cyt b sequences of

37 P. perniciosus from four localities in southern Spain

(EH, EG, EA and EM in Fig. 2). Thirty-five flies contained

haplotype pern04 and two flies contained haplotype

pern05, which together form the Iberian sublineage in

Fig. 4. No haplotypes of the Iberian sublineage were found

in the Rif.

Haplotype pern01 of the typical sublineage was asso-

ciated in the Rif not only with the typical male and female

PN morphotypes of P. perniciosus (in Ouezzane and Taou-

nate) but also with the typical male and female LC mor-

photypes of P. longicuspis (in Chefchaouene) (Table 4). This

is consistent with interspecific hybridization and mtDNA

introgression from P. perniciosus into P. longicuspis.

Eleven males of P. perniciosus from Chefchaouene had

the PNA morphology and were characterized for Cyt b: all

had the haplotype pern06, as did two females of P. perni-

ciosus from Chefchaouene. One typical male (PN) of

P. perniciosus from Ouezzane had a similar haplotype,

pern07. The haplotypes pern06 and pern07 are considered

to form the previously unreported pna sublineage (Fig. 4),

which is distinguished from the other two sublineages of

P. perniciosus by three polymorphisms, at nucleotide posi-

tions 147, 183 (diagnostic) and 246 (Table 3). Haplotypes of

the pna sublineage were the only ones found in males with the

PNA genitalia, all from Chefchaouene, and they were not

found in any P. longicuspis sensu lato. However, the pna

sublineage is not diagnostic for males of P. perniciosus with

the PNA morphotype, because it was also found in two

typical male PNmorphotypes: one fromOuezzane,Morocco,

and the second from Tunisia (Table 3).

The second of the three primary Cyt b lineages, lcus, has

six diagnostic, fixed polymorphisms (at nucleotide positions

6, 9, 27, 54, 171 and 252; Table 3), and it had previously

been isolated only from Tunisian P. longicuspis (Esseghir

et al., 2000). It was also found only in typical male and

female LC morphotypes of P. longicuspis from the three

Moroccan localities (Table 4).

The third primary lineage, lcx, is reported here for the

first time, and is characterized by three fixed polymorph-

isms (at nucleotide positions 27, 123 and 171; Table 3). It

was found in the three Moroccan localities and, being asso-

ciated only with typical male and female LC morphotypes

of P. longicuspis, it is a marker for a sibling species.

Combined isoenzyme profiles and mtDNA in Moroccan

sandflies

Forty-nine individual sandflies originating from Ouez-

zane, Taounate and Chefchaouene were fully characterized,

morphologically, isoenzymically (at the HK, GPI and PGM

loci) and by mtDNA sequencing. Allele frequencies were

analysed at each isoenzyme locus for five populations

defined by their unique combinations of morphology and

mtDNA lineage (Table 5).

At the HK locus, the three groups of Moroccan P. long-

icuspis in Table 5 (all LC morphotypes, but each defined by

one of the three primary mtDNA lineages) were almost

monomorphic for allele 1, which also showed high frequen-

cies (0.885–1.000) in all other populations of Moroccan

P. longicuspis (LC morphotypes) as well as in Iberian popula-

tions of P. perniciosus (PN morphotypes) (Table 1). How-

ever, there were discretely lower frequencies of allele 1

Table 2. Nei’s genetic distances (above diagonal) and FST values (below diagonal) calculated from the isoenzyme data at the three

polymorphic loci for nine populations of P. perniciosus and three populations of P. longicuspis

Phlebotomus perniciosus Phlebotomus longicuspis

Populationsa EH EG EA EM IM GZ MO MT MCb MO MT MC

EH – 0.006 0.009 0.008 0.137 0.138 0.201 0.128 0.544 0.055 0.014 0.107

EG 0.0090 – 0.001 0.002 0.088 0.088 0.139 0.080 0.428 0.061 0.023 0.140

EA 0.0283 � 0.0167 – 0.002 0.076 0.077 0.125 0.069 0.403 0.061 0.027 0.152

EM 0.0236 � 0.0115 � 0.0174 – 0.090 0.091 0.145 0.082 0.437 0.052 0.021 0.139

IM 0.3649 0.2506 0.2385 0.2738 – 0.001 0.009 0.001 0.121 0.198 0.169 0.373

GZ 0.3306 0.2266 0.2152 0.2464 � 0.0007 – 0.008 0.000 0.114 0.191 0.164 0.360

MO 0.3623 0.2712 0.2642 0.2938 0.0280 0.0231 – 0.010 0.081 0.265 0.236 0.459

MT 0.3234 0.2035 0. 1987 0.2327 � 0.0027 � 0.0068 0.0210 – 0.123 0.177 0.152 0.342

MCb 0.6506 0.5881 0.5999 0.6160 0.3312 0.2908 0.2066 0.3120 – 0.561 0.548 0.802

MO 0.1415 0.1372 0.1483 0.1330 0.3771 0.3415 0.3741 0.2990 0.5940 – 0.022 0.051

MT 0.0371 0.0551 0.0723 0.0580 0.3476 0.3173 0.3597 0.2833 0.5820 0.0439 – 0.055

MC 0. 1752 0.1982 0.2184 0.2110 0.4368 0.4084 0.4301 0.3480 0.5531 0.0772 0.0952 –

aKey to populations is given in Fig. 2.bPNA is the only male morphotype recorded from this location.

30 B. Pesson et al.


(0.000–0.488) in Italian, Maltese and Moroccan popula-

tions of P. perniciosus (PN/PNA morphotypes) (Tables 1

and 5). In contrast, populations of P. perniciosus from the

last three countries were characterized by a high frequency

of allele 2 (0.500–1.000), which was fixed in the Chef-

chaouene population defined by the pna mtDNA subline-

age (Table 5) but absent from all P. longicuspis (Tables 1

and 5).

The GPI locus showed differential polymorphism among

the three primary mtDNA lineages found in P. longicuspis

in Morocco (Table 5). This locus was almost monomorphic

(0.944–1.000) for allele 1 in P. perniciosus (PN/PNA mor-

photypes) from Italy, Malta and northern Morocco, but it

was less frequent in populations of P. perniciosus from

southern Spain (0.773–0.889) and in all Moroccan popula-

tions of P. longicuspis (0.214–0.864) (Tables 1 and 5).

The PGM locus is highly polymorphic in Phlebotomus

(Larroussius) species, and in Chefchaouene it clearly differ-

entiated flies of the lcx mtDNA lineage from all other males

and females with the typical LC morphotype (Table 5).

Allele 5 was found only in the lcx mtDNA lineage, in

which allele 2 was almost absent. All Moroccan P. long-

icuspis populations were not in HW equilibrium when all

individuals were treated as belonging to one species

(Table 1), and this is consistent with the presence of a sibling

species. The frequency of allele 4 was elevated only in the

subpopulation of P. longicuspis with the lcx mtDNA lineage

and in P. perniciosus from Chefchaouene with the pna

mtDNA sublineage (Table 5). This is another characteristic

distinguishing the Chefchaouene population of P. pernicio-

sus from all others (Table 1).

Therefore, based on the PGM locus, mtDNA introgres-

sion is most likely to have occurred by hybridization

between P. perniciosus and P. longicuspis sensu stricto,

not the sibling species. However, the hybrids do have a

uniquely high frequency of allele 2 at the GPI locus

(Table 5).

It is noteworthy that there are no fixed alleles diagnostic

for either P. perniciosus or P. longicuspis from Morocco,

even after removing the sibling species (LC morphotypes,

lcx mtDNA lineage), the P. perniciosus variant (PNA/PN

morphotypes, pna mtDNA sublineage) and the putative

hybrids (LC morphotypes, pern01 haplotype of the pnt

mtDNA sublineage). When present, however, allele 2 of

HK was diagnostic for P. perniciosus and allele 2 of GPI

was diagnostic for P. longicuspis sensu lato (Table 5).

Discussion

Three species of the P. perniciosus complex in the Moroccan

Rif

The current findings indicate the presence of three phy-

logenetic species in the Moroccan Rif. One species is

P. perniciosus, which was characterized by a distinctive iso-

enzyme profile and a discrete Cyt b mtDNA lineage (com-

posed of two sublineages, typical pnt, and pna). This species

includes not only the typical male and female morphotypes

(PN) of P. perniciosus but also a male morphotype (PNA)

that can be confused with P. longicuspis (Benabdennbi et al.,

1999). The pna mtDNA sublineage is not a diagnostic

marker for a sibling species with PNA male morphotypes,

because it was also found in two typical (PN) males of

P. perniciosus fromOuezzane, Morocco, and northern Tunisia

(Tables 3 and 4). However, all the PNA males in

Distance.40 .33 .27 .20 .13 .07 .00

GZ P. perniciosus

MT P. perniciosus

IM P. perniciosus

MO P. perniciosusMC P. perniciosusEH P. perniciosus

EM P. perniciosus

EA P. perniciosus

EG P. perniciosus

MT P. longicuspisMO P. longicuspis

MC P. longicuspis

Fig. 3. UPGMA cluster analysis of the Nei’s genetic distances among nine populations of P. perniciosus and three populations of P. longicuspis,

calculated from allele frequencies at the three polymorphic isoenzyme loci (HK, GPI, PGM).



Table

3.AlignmentofmitochondrialCytbhaplotypes

(variantcharactersonly)relatedto

thespecim

ens’origins

Character

no.in

alignmenta

No.ofspecim

enswitheach

haplotypein

differentlocalities

b

Cytb

11111111112222

124555777702344667880455

Spain

Italy

Malta

Tunisia

Morocco

Haplotype

69579148235853357581131628

EH

EG

EA

EM

IPGZ

TT

TM

TS

MO

MT

MC

pern01

TAAATATTTATCTTGACATCGAAATG

12

22

18

19

pern02

TAAATAATTATCTTGACATCGAAATG

1

pern04

TAAATATTTATCTTGATATCGAATTG

14

96

4þ2

pern05

TAAATATTTATTTTGATATCGAATTG

2

pern06

TAAATATTTATCTTGACGTCGCATTG

113

pern07

TAAATATTTATCTTGACATCGCATTG

1

lcus01

CGATCACCCATTTTGGTACTGCATCA

42

18

2

lcus02

CGATCACCCATTTTGGTACTGCATCA

1

lcx01

TAGCCATCCACTTCGACACAGAATTA

21

lcx02


1

lcx03


1

lcx04


4

lcx05


5

lcx06


2

lcx07


6

lcx08


1

lcx09


1

lcx10


1

lcx11


1

Totals

16

96

62

22

34

25

21

45

aUnderlined

nucleotides

¼synapomorphiccharactersdiagnosticforeach

ofthethreemain

lineages

(pern,lcus,lcx),i.e.fixed

polymorphisms.bUnderlined

numbers¼

data

from

Esseghir

etal.(1997,2000).

32 B. Pesson et al.


Chefchaouene did have the pna mtDNA sublineage as well as

a characteristic isoenzyme profile (including a high frequency

of allele 4 of PGM), and so there is a suggestion that the

PNA aedeagus might be associated with an imperfect bar-

rier to panmixia and incipient speciation. Direct evidence

for the absence of linkage between mtDNA lineage and

male morphology was provided by finding both PN and

PNA males amongst the laboratory-reared progeny of a

single female from El Jebha, in the Moroccan Rif

(Benabdennbi, 1998). Throughout most of its range outside

Morocco, the patterns of genetic variation of P. perniciosus are

consistent with the isolation of populations in two commonly

proposed Pleistocene Ice-Age refugia (Avise & Walker, 1998).

The isoenzyme analysis (Fig. 3) indicated an Iberian cluster of

populations that is distinct from an Italy–North Africa cluster

(including the type locality in Malta), and the distributions of

mtDNA lineages matched this (Fig. 4; Esseghir et al., 1997,

2000).

The second phylogenetic species is P. longicuspis, which

was also characterized by a distinctive isoenzyme profile

Icus01 Africa LC

Icus02 Rif LC

100

0.005 substitution/sitepern05 Spain PN

pern04 Spain PN

pern07 Rif PNA

pern06 Africa PNA

pern02 Italy PN

pern01 Most PN

79

77

53

99

61

92

5751

Icx09 Rif LC

Icx11 Rif LC

Icx08 Rif LC

Icx05 Rif LC

Icx04 Rif LC

Icx06 Rif LC

Icx10 Rif LC

Icx07 Rif LC

Icx03 Rif LC

Icx02 Rif LC

Icx01 Rif LC

Fig. 4. Phylogenetic relationships among all the haplotypes of mitochondrial Cyt b. Each haplotype is coded by lineage (pern, lcus, lcx) and

number, and it is also labelled with the country or region and the morphotype (PN, PNA, LC) in which it has been found. The tree was

generated by a Neighbour joining search of genetic distances (with ties broken randomly), rooted with the lcus haplotypes as outgroups (PAUP:

Swofford, 2002). Values below some branches are bootstrap support values (100 replicates) given by the 50% majority-rule consensus tree.

A maximum parsimony analysis, using an heuristic search (PAUP: Swofford, 2002), gave a congruent 50% majority-rule consensus tree.



and a discrete mtDNA lineage (lcus). Genotype frequencies

were in HW equilibrium at all three isoenzyme loci if

specimens with the lcx mtDNA lineage were removed

from the population and treated as a third, sibling species.

Thus, mtDNA characterization permitted the separation of

P. longicuspis sensu stricto from males and females that

shared its known diagnostic morphological characters (LC

morphotypes). Benabdennbi et al. (1999) found a departure

from HW expectations at the PGM locus for P. longicuspis

caught in Chefchaouene and therefore they suspected the

presence of a subpopulation, but they had no independent

marker for separating the two. The males of P. perniciosus

(PN, PNA) and P. longicuspis (LC) were shown to have a

significantly different number of coxite hairs (Benabdennbi

& Pesson, 1998). We can now re-analyse this data, using an

analysis of variance with Scheffe test, to show that the mean

number (m) of coxite hairs in males with the mtDNA line-

age lcx (m¼ 20.3; range: 19–22; n¼ 14) is significantly less

than that in males of P. longicuspis sensu stricto with the

lineage lcus (m¼ 26.4; range: 21–31; n¼ 26).

The relatively large pairwise genetic distances (4.3–6.5%)

between the primary mtDNA lineage characteristic of

P. perniciosus (pern) and the two found only in P. longicuspis

(lcus, lcx) indicate a divergence of these three lineages about

2–3 million years ago (mya), based on a pairwise divergence

rate of 2.3% per million years (Esseghir et al., 1997, 2000).

The divergence of these mtDNA lineages may have accom-

panied allopatric speciation. However, it should be remem-

bered that divergent mtDNA lineages may arise before

speciation, because of maternal inheritance and the absence

of recombination (Avise & Walker, 1998). Therefore, the

presence in P. longicuspis of two highly divergent mtDNA

lineages does not in itself provide strong support for recog-

nizing a sibling species. The necessary, complementary sup-

port for sibling species recognition comes from the coxite

hair counts and especially from the isoenzyme analysis. This

strongly suggests that P. longicuspis sensu stricto and its

sibling species are not just phylogenetic species (character-

ized by discrete isoenzyme profiles and mtDNA lineages) but

are also biological species, with a substantial reproductive

barrier between them. The population genetics analysis,

based on isoenzymes, also indicates that P. perniciosus is a

good biological species.

The discovery of some mtDNA introgression from

P. perniciosus to P. longicuspis does not necessarily invalidate

these conclusions. However, it should be remembered that

even occasional hybridization could lead to the sharing of

epidemiologically important traits, as discussed in the next

section.

Evidence for mitochondrial introgression from P. perniciosusto P. longicuspis and its epidemiological significance

Esseghir et al. (2000) compared mtDNA (Cyt b) and

nuclear gene (EF-a) phylogenies of Larroussius species: thetwo phylogenies were incongruent for the P. perniciosus

complex, and mtDNA introgression between P. perniciosus

and P. longicuspis was suspected. The current findings pro-

vide further, compelling evidence for mtDNA introgression

between wild populations of these two species. In Chef-

chaouene, in the Moroccan Rif, some P. longicuspis mor-

photypes (LC) contained a mtDNA haplotype (pern01) of a

sublineage (pnt) that is usually found only in P. perniciosus

from Morocco, Tunisia, Malta and Italy. The presence of

the same pern01 haplotype in the two species indicates

either that introgression is still occurring or that it has

occurred within the last 125 000 years, assuming the molecu-

lar clock runs no faster than 2.3% pairwise divergence per

million years (Esseghir et al., 2000) and that the sequencing

error is not greater than 1 in 300 nucleotides. Mitochondrial

DNA is usually inherited maternally and the putative

hybrids were morphologically inseparable from, and iso-

enzymatically most similar to, P. longicuspis. Therefore,

Table 4. The number of individual sandflies with each combination of morphotype and Cyt b lineage in three Moroccan localities

Locality

Morphotypes

Cyt b

lineage/sublineageaOuezzane

(MO)

Taounate

(MT)

Chefchaouene

(MC)

LC female lcus/– 1 9 2

lcx/– 3 0 14

pern/pnt 0 0 3

LC male lcus/– 1 10 0

lcx/– 1 1 7

pern/pnt 0 0 6

PN female pern/pnt 9 0 0

pern/pna 0 0 2

PN male pern/pnt 9 1 NFb

pern/pna 1 0 NF

PNA male pern/pna NF NF 11

Total 25 21 45

aOnly the pern lineage has designated sublineages.bNF ¼ morphotype not found in the locality.

34 B. Pesson et al.


we conclude that fertile progeny usually result only from

matings between males of P. longicuspis sensu stricto and

the females of P. perniciosus or hybrids.

A comparative sequence analysis of Cyt b also demon-

strated mtDNA introgression between closely related spe-

cies within two subgenera of Lutzomyia (Marcondes et al.,

1997; Testa et al., 2002). Like the P. perniciosus complex,

both of these neotropical sandfly taxa contain many vectors

of Leishmania species, and so it is important to consider the

epidemiological implications of mtDNA introgression in

sandflies.

Mitochondrial introgression usually results from inter-

breeding. Such interspecific hybridization is not uncommon

among related insect species (Arnold, 1997), and the inde-

pendent assortment of many nuclear genes is axiomatic in

Mendelian genetics. Therefore, it should come as no sur-

prise if recently diverged species, such as P. perniciosus and

P. longicuspis, are shown to display reticulate evolution. In

these circumstances, genotypic markers as well as pheno-

typic traits and morphology will not always covary, partly

as a result of recent ancestry (with insufficient time for ‘line-

age sorting’) and partly from continuing hybridization. This

absence of character covariation is not helpful to parasitolo-

gists and applied entomologists who hope that vectorial

traits can be associated with easily identifiable taxonomic

species, which can then be targeted for control measures.

For example, much effort was made to find diagnostic

markers for the major sub-Saharan vectors of malaria

(Hill & Crampton, 1994), even though sibling species of

the Anopheles gambiae Giles complex (Culicidae) are known

to hybridize both in the wild and the laboratory (Black &

Lanzaro, 2001), and mtDNA introgression is great enough to

prevent the separation into two distinct lineages of the most

important vectors,An. gambiae sensu stricto andAn. arabiensis

Patton (Besansky et al., 1994).

Given the evidence for the occurrence of interspecific

introgression and sibling species in the P. perniciosus

complex, it should no longer be assumed that the frequen-

cies of certain vectorial traits can be directly related to

the proportions of the morphospecies previously recorded

in different geographical regions and ecological zones of

north-west Africa. For example, Rioux et al. (1984)

reported that P. longicuspis is more abundant than P.

perniciosus in the drier bioclimates of Morocco, but only

a re-examination of slide-mounted males in various

collections would indicate whether authors had confused

P. longicuspis sensu lato (morphotype LC) with P. perniciosus

(morphotype PNA). Even then, the identification of the

male morphotype LC would still leave unresolved any

ecological replacement of P. longicuspis sensu stricto by

its sibling species. Both P. perniciosus and P. longicuspis have

been found naturally infected with the same strain of Leish-

mania infantum (Esseghir et al., 2000), and now reticulate

evolution makes it less certain that ecological associa-

tions or vectorial roles are going to be linked solely to one

taxon.

Table 5. Allelic frequencies at the three polymorphic isoenzyme loci for Moroccan P. perniciosus and P. longicuspis grouped according to

morphotype (PNþPNA and LC, respectively), mitochondrial Cyt b lineage and localities

Morphotype PN PNA + PN LC LC LC

Cyt b lineage/sublineage pern/pnt pern/pna lcus/- pern/pnt lcx/-

Localities Ouezzane (MO) Chefchaouene (MC) Taounate (MT) Chefchaouene (MC) Chefchaouene (MC)

No. tested (Female þ male) 4þ 4 1þ 8 5þ 5 2þ 5 9þ 6

Locusa and alleles

HK

1 0.375 0 0.950 1.000 1.000

2 0.500 1.000 0 0 0

3 0.125 0 0 0 0

7 0 0 0.050 0 0

P 1 – 1 – –

GPI

1 1.000 0.944 0.750 0.214 0.767

2 0 0 0.200 0.786 0.067

3 0 0 0.050 0 0.067

5 0 0.056 0 0 0.033

7 0 0 0 0 0.067

P – 1 0.822 1 1

PGM

2 0.938 0.667 0.950 0.929 0.067

3 0.063 0 0 0.071 0

4 0 0.333 0.050 0 0.333

5 0 0 0 0 0.600

P 1 0.457 1 1 1

aHK ¼ hexokinase, GPI ¼ glucosephosphate isomerase, and PGM ¼ phosphoglucomutase; P ¼ probability of w2 value occurring by chance,

when testing for deviation from Hardy–Weinberg expectations of genotype frequencies.



Taxonomic significance of the findings: formal recognition of

the new sibling species?

Those researching leishmaniasis transmission need to be

aware of what information is being communicated by dif-

ferent taxonomic names. Two morphospecies, each with a

Linnean binomial, may be good phylogenetic species and,

despite occasional hybridization, may even behave most of

the time as good biological species (Arnold, 1997). In these

circumstances, it is usual to maintain the taxonomic species

concept, with identifications being based on the morph-

ology of type specimens (Otte & Endler, 1989), not least

because few laboratories have the resources for routine

genetic characterization.

Until more populations have been characterized, and

gene flow between them assessed, we do not favour creating

or synonymizing taxonomic names in the P. perniciosus

complex. Certainly, where there is reticulate evolution, the

reliance on a few characters might well mislead, as is clear

from the history of the complex. Parrot (1936a) reared

in the laboratory a small number of ‘pure strains’ of

P. perniciosus and P. langeroni var. longicuspis and, based

on this evidence for the inheritance of distinctive male

genitalia, raised the latter to full species. Soon afterwards,

however, Parrot (1936b) recorded much individual varia-

tion in the form of the aedeagus of P. perniciosus, but

insisted that the terminal bifurcation of the aedeagus was

diagnostic. Other authors accepted this, along with the later

conclusion of Parrot & Durand-Delacre (1947) that the

morphology of the aedeagus of males collected in the

Sahara (Beni Ounif-de-Figuig and Tamanrasset, Algeria)

was a variant of P. longicuspis. This variant could be the

PNA morphotype, and if so Parrot & Durand-Delacre

(1947) should have identified it as P. perniciosus. Reticulate

evolution within the P. perniciosus complex means that all

its species are unlikely to be identified by unique morpho-

logical characters or vectorial roles.

The results from Chefchaouene, in the Rif, are important

because they show that some sympatric populations of

P. perniciosus and P. longicuspis have the characteristics of

biological species. However, more extensive characteriza-

tion of populations throughout the ranges of these two

morphospecies could indicate that they would be better

treated as three informal incipient species (equivalent to

P. perniciosus, P. longicuspis and the sibling species), with

only a single taxon, P. perniciosus, being formally recog-

nized. This would be appropriate only if the results from

Chefchaouene were shown to be atypical, by demonstrating

ongoing and significant nuclear gene flow between many

sympatric populations of the three species.

Acknowledgements

The authors thank the following colleagues for help with

collecting sandflies: R and M. Killick-Kendrick (Ascot),

N. Leger and H. Ferte (Reims), G. Madulo-Leblond

(Paris), E.Martınez Ortega (Murcia) andM. C. SanchisMarin

(Almerıa). J. Llewellyn-Hughes, W. Edge and J. M. Testa

provided invaluable assistance for the mtDNA sequencing in

theNaturalHistoryMuseum,whereB.P. spentasabbatical.We

acknowledge the support of theMinistere de la Sante Publique

du Maroc (J. Mahjour), Ministere des Affaires Etrangeres

(France), the British Council (Tunis), the EC DG-XII

(Brussels; STD3 contract: TS3*CT93-0253), and the TDR

programme (UNDP/World Bank/W.H.O.).

References

Arnold, M.L. (1997) Natural Hybridization and Evolution. Oxford

University Press, Oxford.

Avise, J.C. & Walker, D. (1998) Pleistocene phylogeographic

effects on avian populations and the speciation process.

Proceedings of the Royal Society of London B, 265, 457–463.

Benabdennbi, I. (1998) Contribution a l’etude de la biodiversite chez

les Larroussius (Diptera: Psychodidae) du bassin mediterraneen.

Apports de l’analyse isoenzymatique. These de Doctorat,

Universite Louis Pasteur – Strasbourg I, France.

Benabdennbi, I. & Pesson, B. (1998) A propos de la variabilite

morphologique de Phlebotomus perniciosus (Diptera: Psychodi-

dae). Bulletin de la Societe Francaise de Parasitologie, 16, 43–50.

Benabdennbi, I., Pesson, B., Cadi-Soussi, M. & Morillas Marquez,

F. (1999) Morphological and isoenzymatic differentiation of

sympatric populations of Phlebotomus perniciosus and Phleboto-

mus longicuspis (Diptera: Psychodidae) in northern Morocco.

Journal of Medical Entomology, 36, 116–120.

Besansky, N.J., Powell, J.R., Caccone, A., Hamm, D.M.,

Scott, J.A., Collins, F.H. (1994) Molecular phylogeny of the

Anopheles gambiae complex suggests genetic introgression

between principal malaria vectors. Proceedings of the National

Academy of Science USA, 91, 6885–6888.

Black, W.C.I.V. & Lanzaro, G.C. (2001) Distribution of genetic

variation among chromosomal forms of Anopheles gambiae s.s.

introgressive hybridization, adaptive inversions, or recent

reproductive isolation? Insect Molecular Biology, 10, 3–7.

Collantes, F. & Martınez Ortega, E. (1997) Sobre la validez

taxonomica de Phlebotomus longicuspis Nitzulescu, 1931

(Diptera: Psychodidae). Boletin de la Asociacion Espanola

de Entomologia, 21, 141–146.

Esseghir, S. (1998) Phylogenie des phlebotomes mediterraneens

vecteurs de Leishmania major et de Leishmania infantum et

marqueurs moleculaires de populations. These de Doctorat,

Faculte des Sciences, Universite de Tunis, Tunisia.

Esseghir, S., Ready, P.D. & Ben-Ismail, R. (2000) Speciation of

Phlebotomus sandflies of the subgenus Larroussius coincided with

the late Miocene-Pliocene aridification of the Mediterranean sub-

region. Biological Journal of the Linnean Society, 70, 189–219.

Esseghir, S., Ready, P.D., Killick-Kendrick, R. & Ben-Ismail, R.

(1997) Mitochondrial haplotypes and phylogeography of Phle-

botomus vectors of Leishmania major. Insect Molecular Biology,

6, 211–225.

Hill, S.M. & Crampton, J.M. (1994) Review: DNA-based methods

for the identification of insect vectors. Annals of Tropical

Medicine and Parasitology, 88, 227–250.

Killick-Kendrick, R. (1990) Phlebotomine vectors of the leishma-

niases: a review. Medical and Veterinary Entomology, 4, 1–24.

36 B. Pesson et al.


Leger, N., Marchais, R., Madulo-Leblond, G., Pesson, B.,

Kristensen, A., Ferte, H., Killick-Kendrick, R. & Killick-

Kendrick, M. (1991) Les phlebotomes implique dans la transmis-

sion des leishmanioses dans l’ıle de Gozo (Malte). Annales de

Parasitologie Humaine et Comparee, 66, 33–41.

Leger, N., Pesson, B., Madulo-Leblond, G. & Abonnenc, E. (1983)

Sur la differentiation des femelles du sous-genre Larroussius

Nitzulescu, 1931 (Diptera, Phlebotomidae) de la region mediterra-

neene.Annales de Parasitologie Humaine et Comparee, 58, 611–623.

Madulo-Leblond, G. (1983) Les Phlebotomes (Diptera, Phleboto-

midae) des Iles Ioniennes. These de Doctorat es Sciences

Pharmaceutiques, Universite de Reims, France.

Marcondes, C.B., Day, J.C. & Ready, P.D. (1997) Introgression

between Lutzomyia intermedia and both Lu. neivai and Lu.

whitmani, and their roles as vectors of Leishmania braziliensis.

Transactions of the Royal Society of Tropical Medicine and

Hygiene, 91, 725–726.

Martın Sanchez, J., Gramiccia, M., Pesson, B. & Morillas

Marquez, F. (2000) Genetic polymorphism in sympatric species

of the genus Phlebotomus, with special reference to Phlebotomus

perniciosus and Phlebotomus longicuspis (Diptera, Phlebotomi-

dae). Parasite, 7, 247–254.

Martınez Ortega, E., Ward, R.D., Martin Luengo, F. & Conesa

Gallego, E. (1982) Nueva distribucion de Phlebotomus (Lar-

roussius) longicuspis Nitzulescu 1930 (Diptera, Phlebotomidae)

en Espana. Revista Iberica de Parasitologia, 42, 283–288.

Morillas Marquez, F., Guevara Benitez, D.C., Gil Collado, J. &

Ubeda Ontiveros, J.M. (1982) Presencia en Espana de Phlebo-

tomus (Larroussius) longicuspis Nitzulescu, 1930. Revista Iberica

de Parasitologia, Volume extra, 191–196.

Morillas Marquez, F., Sanchis Marin, M.C., Martın Sanchez, J. &

Acedo Sanchez, C. (1991) On Phlebotomus perniciosus News-

tead, 1911 (Diptera, Phlebotominae) in the province of Almeria

in south eastern Spain. Parassitologia, 33 (Suppl. 1), 437–444.

Newstead, R. (1911) The papataci flies (Phlebotomus) of the

Maltese Islands. Bulletin of Entomological Research, 2, 47–48.

Nitzulescu, V. (1930) Phlebotomus langeroni n. sp. et P. langeroni

var. longicuspis n. var. de Douar-Shott (Tunisie). Annales de

Parasitologie, 8, 547–553.

Otte, D. & Endler, J.A. (1989) Speciation and its Consequences.

Sinauer Associates Inc, Sunderland, Mass, U.S.A.

Parrot, L. (1936a) Notes sur les phlebotomes XX. – sur

Phlebotomus langeroni var. longicuspisNitzulescu, 1930. Archives

de l’Institut Pasteur d’Algerie, 14, 137–143.

Parrot, L. (1936b) Notes sur les phlebotomes XXI. – sur la valve

copulatrice de Phlebotomus perniciosus Newstead. Archives de

l’Institut Pasteur d’Algerie, 14, 144–147.

Parrot, L. & Durand-Delacre, R. (1947) Notes sur les phlebotomes

LVI. – Phlebotomes de Beni Ounif-de-Figuig. Archives de

l’Institut Pasteur d’Algerie, 25, 77–86.

Pesson, B., Benabdennbi, I., Schuh, C., Alves Pires, C.,

Morillas Marquez, F., Sanchis Marin, M.C., Martinez Ortega, E.,

Lucientes Curdi, J. & Gallego, M. (1998) Genetic variability

of Phlebotomus perniciosus (Diptera: Psychodidae) in the Iberian

peninsula. Acta Parasitologica Portuguesa, 5, 52.

Pesson, B., Madulo-Leblond, G., Killick-Kendrick, M., Tang, Y. &

Killick-Kendrick, R. (1994) The armature in the genital atrium

as a new taxonomic character distinguishing females of

Phlebotomus papatasi and P. duboscqi (Diptera: Psychodidae).

Annals of Tropical Medicine and Parasitology, 88, 539–542.

Pesson, B., Wallon, M., Floer, M.T. & Kristensen, A.R. (1991)

Etude isoenzymatique de populations mediterraneennes de

phlebotomes du sous-genre Larroussius. Parassitologia, 33

(Suppl. 1), 471–476.

Raymond, M. & Rousset, F. (1995) GENEPOP Version 1.2.:

population genetics software for exact test and ecumenicism.

Journal of Heredity, 86, 248–249.

Rioux, J.A. & Lanotte, G. (1990) Leishmania infantum as a cause of

cutaneous leishmaniasis. Transactions of the Royal Society of

Tropical Medicine and Hygiene, 84, 898.

Rioux, J.A., Rispail, P., Lanotte, G. & Lepart, J. (1984) Relations

Phlebotomes – bioclimats en ecologie des leishmanioses.

Corollaires epidemiologiques. L’exemple du Maroc. Bulletin de

la Societe Botanique de France, 131, 549–557.

Seccombe, A.K., Ready, P.D. & Huddleston, L.M. (1993) A

catalogue of Old World phlebotomine sandflies (Diptera:

Psychodidae, Phlebotominae). Occasional Papers on Systematic

Entomology, 8, 1–57.

Swofford, D.L. (2002) PAUP. Phylogenetic Analysis Using Parsi-

mony, version PAUP*4.0. Smithsonian Institution Press,

Washington DC, U.S.A.

Swofford, D.L. & Selander, R.B. (1981) BIOSYS-1: a FORTRAN

program for the comprehensive analysis of electrophoretic data

in population genetics and systematics. Journal of Heredity, 72,

281–283.

Testa, J.M., Montoya-Lerma, J., Cadena, H., Oviedo, M. &

Ready, P.D. (2002) Molecular identification of vectors of

Leishmania in Colombia: mitochondrial introgression in the

Lutzomyia townsendi series. Acta Tropica, 84, 205–218.

Accepted 10 October 2003



Documents

Sandflies of the Phlebotomus perniciosus complex ...hera.ugr.es/doi/15059121.pdf · Sandflies of the Phlebotomus perniciosus complex: mitochondrial introgression and a new sibling