Morphological and mitochondrial DNA characters for identification and phylogenetic analysis of the myiasis-causing flesh fly Wohlfahrtia magnifica and its relatives, with a description of Wohlfahrtia monegrosensis sp. n. Wyatt & Hall

Introduction

Myiasis is the infestation of live vertebrates, including humans, with the larvae of true flies, Diptera (Zumpt, 1965). Traumatic or wound myiasis is the infestation of cutaneous tissues at sites of wounding or in body orifices (Hall & Farkas, 2000). Subcutaneous tissues can also be affected in a serious infestation. Three fly species are economically important as obligate larval parasites causing traumatic myiasis: the Old World screwworm fly (OWS) Chrysomya bezziana Villeneuve (Diptera: Calliphoridae), the New World screwworm fly (NWS) Cochliomyia hominivorax (Coquerel) (Diptera: Calliphoridae) and Wohlfahrt’s wound myiasis fly Wohlfahrtia magnifica (Schiner) (Diptera: Sarcophagidae). Taxonomic knowledge of groups containing any insect pest, including those that cause traumatic myiasis in humans and animals, is vital for their correct identification. The latter is essential in any control campaign, especially when a pest is new to a region. This was exemplified following the introduction of NWS to Libya, when correct and timely identification of the pest species led to a rapid and ultimately successful eradication campaign (Lindquist et al., 1992).

Wohlfahrtia is a genus of 26 species, including the new species described here, Wohlfahrtia monegrosensis, placed in the subfamily Paramacronychiinae of the family Sarcophagidae (Diptera) (Pape, 1996; Lehrer, 2003). Most species in this subfamily, including Wohlfahrtia, are densely covered with greyish microtomentum with a more-or-less distinct abdominal pattern of a median stripe and dark lateral spots, and character states considered autapomorphic for the subfamily are all associated with the male terminalia: sternite 5 is small, with shallow posterior emargination; tergite 6 is fused to syntergosternite 7–8; the surstyli are fused to the epandrium and the acrophallus is long and curved (Pape, 1996). Within this subfamily, Wohlfahrtia can be diagnosed by the following characters: the male is without proclinate orbital bristles; the parafacial plate is bare or has a few scattered setulae; the costal spine is at most as long as crossvein r–m, and the abdominal microtomentum is densely silvery-grey with distinct black spots that sometimes coalesce posteriorly (Pape, 1996). Most species in this genus are confined to arid parts of the Palaearctic and Afrotropical regions, especially northern Africa, the Middle East and central Asia, but one species occurs in North America.

The larval feeding habits in the genus Wohlfahrtia are diverse (Table 1). The most economically important species is W. magnifica, the larvae of which are obligate parasites that cause a serious traumatic myiasis, mainly in livestock, but also in humans and, rarely, in wildlife (Farkas et al., 1997; Hall, 1997; Hall & Farkas, 2000). Wohlfahrtia magnifica is widely distributed throughout the southern and central parts of the Palaearctic region, from Morocco in the west to China in the east. An expansion of its range onto the Mediterranean island of Crete occurred in 1999 (Sotiraki et al., 2003, 2005).

The only species of Wohlfahrtia to occur in the New World is Wohlfahrtia vigil (Walker), found in the U.S.A. and Canada, as well as the Palaearctic region (Hall & Farkas, 2000). The definition of this species has been problematic because of variable external morphology, and a number of different names have been used, mainly W. vigil and Wohlfahrtia opaca (Coquillett) in the Nearctic region and Wohlfahrtia meigenii (Schiner) in the Palaearctic region (Pape, 1996). There also appear to be regional differences in the larval feeding habitats of this species (see Discussion). Nevertheless, all forms have been synonymized under the name W. vigil by Pape (1996) because to date no reliable morphological characters have been found to distinguish Nearctic and Palaearctic specimens.

The causal agent in cases of traumatic myiasis is usually identified to species using mature larvae taken from infested wounds. It is common for the appearance and biology of the adult fly to be almost unknown among farmers and veterinarians dealing with a myiasis problem. Whereas the correct identification of third instars of W. magnifica is relatively straightforward (Spradbery, 1991), identification of the adults of Wohlfahrtia can be more complex. However, identifications of adults are required when they are captured during trapping surveys of myiasis-causing species.

The genitalia of adult males often exhibit the most reliable characters for recognizing new species and identifying individual sarcophagids (Pape, 1987). Identification of females is far less satisfactory and, for some taxa, identification can only be arrived at if the female is taken in copulo with a male. Members of the genus Wohlfahrtia have some useful external characters, which have been used in conjunction with genitalia characters to generate workable identification keys to the species. However, it is often impractical to dissect genitalia of all Wohlfahrtia specimens collected and thus, in practice, only a few specimens will be dissected and, as a result, males and females collected with them, showing the same external character states, will be given the same identification. As indicated by Salem (1938), the small number of external characters and their variability makes identification of Wohlfahrtia difficult. Characters of both male and female genitalia provided better results, but were not ideal.

The current report tests the diagnostic value of commonly used external morphological characters by subjecting data from nine Wohlfahrtia species to a phenetic analysis to determine which taxa could be unambiguously identified solely on these characters. A phylogenetic analysis of these taxa using external and genitalia morphology characters, and mitochondrial DNA analysis for some taxa, to look for subgeneric structure in the group was attempted. The main objective of all these approaches was to be able to distinguish unambiguously the males and females of the economically important obligate parasite W. magnifica from any other species in the genus. The relationship between North American and European specimens of W. vigil was examined and, in addition, the description of a new species, W. monegrosensis, from northern Spain, was undertaken. Both species are sympatric with W. magnifica and have a similar morphology; hence it is important for epidemiological studies to be able to identify W. magnifica unambiguously.

Materials and methods

Results

Male

Head The frons at the narrowest point is slightly narrower than the width of an eye, at approximately 0.3× head-width; it is silvery-grey pollinose, with the parafrontalia slightly more densely silver pollinose than the darker area between. Parafacials are approximately 1.5× the width of the antennal flagellum, with a few short setulae on the lower half towards the eye margin. Eyes are bare. Antennae are mostly blackish, but the pedicel is a paler brownish-yellow towards the tip. The flagellum is approximately twice as long as the pedicel. The arista is very short-haired; hair length is approximately equal to arista width; the arista is mostly dark in colour, but a distinct, narrow, pale yellowish-white band is usually present near the middle. The vibrissae are situated slightly above the mouth margin. The hairs of the gena and occiput are entirely black. The palpi are brownish-yellow and somewhat darkened at the tip.

Thorax The mesonotum is pale grey pollinose and shows three narrow, dark vittae when viewed from behind, one situated centrally between the rows of acrostichal setae, and the other two between the dorsocentral and intra-alar rows of setae. An additional inconspicuous and narrower pair of dark vittae is present anterior to the suture, situated between the central and lateral vittae. There are 2 + 3 pairs of dorsocentrals; the first postsutural pair is weak. Acrostichals are restricted to a strong prescutellar pair and some weak presutural setae can also be differentiated. The scutellum has three pairs of marginal setae; the apical pair is crossed, and a pair of discal setae is also present. Both anterior and posterior spiracles are dark blackish-brown. The halteres are brownish-yellow. Calypters are white.

Wings The wing membrane is hyaline with no distinct dark or coloured markings. The tegula is pale brown and the basicosta pale creamy yellow. There is a row of 5–9 setae at the base of R4 + 5 extending approximately two-thirds of the way from the base of the vein to the crossvein r–m.

Legs Legs are entirely blackish in colour. The femorae are covered with rather dense, fine hair-like setae ventrally, lacking strong bristles; some of these hairs are wavy at the tip and the longest are slightly longer than the greatest width of the femur. Slightly stronger rows of anteroventral and posteroventral setae are present on the hind femur in addition to this hairing. The hind tibia has a rather weakly developed fringe of hairs on the posteroventral surface, the strongest of which is as long as the tibia is wide. A few weak, hair-like setae can also be differentiated on the anteroventral surface.

Abdomen The abdomen is densely pale grey and dusted on the dorsal surface, with three sharply defined, shiny, dark brown spots on the posterior margins of syntergite 1 + 2 and tergites 3–5 (Fig. 1). These spots extend forward for less than half the length of the tergite except on syntergite 1 + 2, where the spot extends for most of the tergite length, and tergite 3, where it narrows and extends anteriorly, reaching the anterior margin of the segment. The dark spots on these basal tergites do not have any strong setae arising from them. The central spot on tergite 4 is also weakly extended anteriorly on some specimens; this and the central spot of tergite 5 are associated with two strong setae, whereas the large lateral spots on these tergites are associated with one seta only. The ventral surface of the abdomen has fine, wavy-tipped hairs on the sternites, which are slightly longer than the width of the hind femur. This surface is also less densely pollinose and has a dull reddish background colour, which contrasts clearly with the dorsal surface when viewed laterally (Fig. 2). The protandrial segment is orange–red, the background colour is partially obscured by grey pollinosity, and there are two transverse rows of strong setae, the anterior of which is somewhat stronger. The epandrium is a shiny orange–red. 3, 4 show the cerci of W. monegrosensis.

Genitalia The genitalia of W. monegrosensis and W. smarti Salem are extremely similar, but have the following differences. In the W. monegrosensis aedeagus (Fig. 5), the distiphallus is almost parallel-sided but is rounded and slightly broadened at the tip; the two lobes of the vesica are prominent and somewhat cup-shaped, and project roughly at right angles from the main body of the distiphallus; the lobes are finely toothed around the edges (visible at high magnifications only), and have a downward-pointing thorn-like structure immediately above them; the parameres are strongly curved with hook-like tips, and gradually narrow towards the tip; the gonopods are broad and almost parallel-sided, and somewhat rounded at the tip with a small, tooth-like structure at the centre of the distal edge. Wohlfahrtia smarti (Fig. 6) has a curved distiphallus with a membranous tip; the lobes of the vesica are narrower; the thorn-like structure above the vesica is not present; the parameres are curved, almost uniformly narrow and hooked at the tip; the gonopods are somewhat narrower than in W. monegrosensis and are slightly curved backwards, not parallel-sided, more strongly convex on the inner surface and have a slightly hooked tip.

Female

The female shares most of the male’s characters, with the following differences. The frons is somewhat wider, making it as wide or slightly wider than the width of an eye (when viewed from the front), approximately 0.33–0.35× head-width at its narrowest point. Two pairs of strong proclinate setae are seen on the upper half of the parafrontalia. The legs are without fringes of fine hairs on the femorae and tibiae; the femorae have rows of strong setae on the ventral surfaces. The long, fine hairs on the ventral surface of the abdomen in the male are absent in the female; only scattered, short adpressed setae are present. The terminalia are mostly reddish-brown. Tergite 6 is entire, not divided dorsally, and has a row of strong marginal setae. The two sclerites of tergite 7 are both triangular, rather narrow, and taper to a point dorsally; there are three marginal setae. Tergite 8 has a group of 3–4 fine setae on either side. Sternite 6 has a row of fairly strong marginal setae, which are somewhat weaker in the middle. Sternite 7 has much finer marginals (7, 8).

The body length of Wohlfahrtia monegrosensis is 6.5–10.5 mm.

Analysis of adult morphology

Species identification The morphological characters used to describe Wohlfahrtia species are usually found on the male genitalia. Nine characters were scored for the male genitalia dissected from between one and six specimens of each species, from which it was confirmed that each species has a unique combination of these characters and little intraspecific variation, none of which confuses identification. Six characters were scored for the female genitalia dissected from between one and six specimens of each species, from which it was also confirmed that each species has a unique combination of these characters and little or no intraspecific variation.

A distance measure NJ search was used to cluster all males and all females based on their score for eight external body morphology characters. Searches of males (Fig. 9) and females (Fig. 10) each returned a single tree of minimum length. Examination of both trees showed that each returned single clusters containing every individual of the same three taxa, Wohlfahrtia bella, W. magnifica and W. vigil, suggesting that at least these three taxa can be reliably identified using external body morphology alone. Inability to retrieve the other six taxa does not indicate that they are not good taxa, but merely that genitalia characters are additionally required for their unambiguous identification. Principal components analysis (PCA) of body character data confirmed the reliable separation of individuals of W. bella, W. magnifica and W. vigil for both sexes and, in addition, of W. monegrosensis for females (11, 12). The pairs of characters contributing most strongly to each component were the same for both sexes (i.e. the length ratio of the antennal segments and parafacial plate vestiture to component 1; the colour of antennal segments and presutural acrostichal setae to component 2) (Table 3). The relative spread of points for W. vigil was much greater along a diagonal axis for both sexes of this species than for other species, especially along component 1, which lends support to the suggestion that specimens of this taxa potentially belong to more than one species (11, 12). All taxa in the left side of the W. vigil‘envelopes’ were from western North America, whereas all taxa in the right side of the envelopes were from eastern North America or Europe. Support for this geographical clustering also comes from the NJ trees (9, 10).

Phylogenetic analysis All 23 characters (body and genitalia of both sexes) for each taxon were used in a single cladistic analysis: where intraspecific variation was recorded, characters were scored as multistate characters. An MP heuristic search with all characters given equal weight generated a single most parsimonious tree of 36 steps (Fig. 13) with an ensemble consistency index (CI) of 0.667 and a retention index (RI) of 0.6. The tree gives strong support to the hypothesis that W. magnifica and W. vigil are sister taxa (bootstrap support 74%) and places the new species, W. monegrosensis, as a sister taxon to this grouping, but the support for this was weaker (bootstrap support 37%). Most of the branches have very weak support, particularly the basal ones (Fig. 13).

Because genitalia provide the distinguishing characters between species, these characters were analysed in isolation. Only the female genitalia characters showed a pattern of synapomorphies that support some phylogenetic relationships amongst species, but the support was weak throughout (Fig. 14).

Molecular analysis The PCR primers amplified either a fragment containing the 3′ terminal 273 nucleotides of cytochrome b or, when the DNA was more degraded, two overlapping segments of the same sequence. In the latter treatment, the identity of 14 nucleotides near the midpoint could not be determined and this region (positions 163–176) of the alignment was then excluded from the comparative analysis. Only a single haplotype (= unique mitochondrial DNA sequence) was discovered for most species, but two haplotypes were recorded for W. magnifica and W. vigil s.l. Sequences have been deposited in GenBank under accession numbers FJ379598–FJ379604.

Pairwise differences between haplotypes ranged from 1.3% to 6.7% intraspecifically (based on a 149-bp sequence in geographical populations of W. magnifica and W. vigil s.l.) and from 9.4% to 15.6% interspecifically (based on a 259-bp sequence in all species; Tables 5 and 6).

Table 5. Alignment of the variant nucleotides of cytochrome b of the 259-bp haplotypes of selected species (Table 4). The two haplotypes of Wohlfahrtia magnifica (CB3_magn1001, CB3_magn1002) were characterized by nucleotide changes only at positions 84 and 126.

Haplotype	11111111111111111111111111122222222222222
	1122223334455566778889901112223344445567777788999901112223335555
	3495812780362824707281476951271360212450993567878036981450370240149
CB3_asch1001	ATTATTCTCCATTTTATTAACAACCATCTTTAAGATCTCTATTAACATCTACCATTACCCACTCCC-
CB3_mone1001	ATAACTCTTAGCTTTAATTTCAACCTCTTCTAAGATCTTTATCAACATTAATTATCATCTGATCCCC
CB3_bell1001	ATTATTCTCTATCACAACTACATTCTTTTTTTAGACCTCCGTTAACACCAATTTCTATTTATTCTTC
CB3_vigi1001	ACATTACCTAATCTCCTTTATTATTTCCATAAAATTCTTCACCGGTTTTACTTTTTACCTATTTTTT
CB3_magn1001	GTAATTTTTAATTACCCTTTCAATTTTTATTCAGACTATTACTAACATCTACTTTCCTCTATCTTTT
CB3_magn1002	GTAATTTTTAATTACCCTTTCAGTTTTTATTCGGACTATTACTAACATCTACTTTCCTCTATCTTTT

Table 6. Uncorrected genetic distances (‘p’) between pairs of 259-bp haplotypes of cytochrome b for selected species (Table 4).

Taxon code		1	2	3	4	5
CB3_asch1001	1	–
CB3_mone1001	2	0.09399	–
CB3_bell1001	3	0.10555	0.11923	–
CB3_vigi1001	4	0.15632	0.14615	0.15000	–
CB3_magn1001	5	0.12890	0.12308	0.11538	0.13462	–
CB3_magn1002	6	0.13660	0.13077	0.11923	0.14231	0.00769

The pairwise genetic distances between the haplotype of the new species and those of the other Wohlfahrtia species were comparable with those between most pairs of the species characterized (Table 6), which supports its description as a morphological species. Using the haplotype of Wohlfahrtia aschersoni (Enderlein) or W. bella (Macquart) as an outgroup, an MP analysis of an alignment of the first 149 bp of the cytochrome b fragment was performed using a branch-and-bound search with 1000 bootstrap replicates; this gave a 50% majority-rule consensus tree (Fig. 15) that supported an ingroup containing the haplotype of the new species (CB3_mone1001), W. magnifica and W. vigil s.l, with the latter two as sister species represented by geographical pairs of haplotypes. For W. magnifica, western (Spain/France) and eastern (Hungary/Romania/Italy/Israel/Iraq) lineages were recognized, differing by just two nucleotides, a 0.77% divergence. For W. vigil, only the first half of the cytochrome b fragment from the single North American specimen (Wv3) was sequenced. However, this 149-bp sequence was shown to have 10 nucleotide differences compared with samples with haplotype vigi1001 from Europe (Table 4). This equates to a difference of 6.7% between the European and North American haplotypes of W. vigil. The sister-species relationship of W. magnifica and W. vigil was strongly supported in the consensus tree produced by an MP analysis of the 259-bp sequence of cytochrome b, but this did not resolve the relationships among the other three species.

Discussion

Morphological analysis confirmed that only genitalia characters are diagnostic for most species of Wohlfahrtia. However, this study demonstrated that males and females of the economically important myiasis-causing species W. magnifica can be distinguished from all the other species examined using only external morphological characters. The NJ and PCA methods of statistical analysis used here grouped specimens on the basis of their overall similarity. This does not provide reliable evidence concerning evolutionary relationships, but it is useful when determining whether specimens comprise one or more distinct morphological forms. Evidence from both the molecular analysis of cytochrome b (Fig. 15) and the morphological analysis of body and genitalia characters (Fig. 13) support the hypothesis that W. magnifica and W. vigil are sister species, and the newly described species, W. monegrosensis, is closely related. Poor phylogenetic resolution based on morphology alone was not unexpected because the characters selected were taken from identification keys which, by their nature, favour species-specific characters. Hence several characters were autapomorphic, having derived states unique to a single taxon.

The analysis of molecular sequence data demonstrates that there are two geographically separated haplotypes of W. magnifica, with diagnostic characters at two nucleotide positions. Thus there is a haplotype in Western Europe (CB3_magn1002, Table 4) and another found elsewhere in Europe and eastwards to the Middle East (CB3_magn1001, Table 4). The genetic divergence does not appear to have resulted in differing pathogenicity. Strains from the Iberian peninsula appear to be just as aggressive and invasive as those from elsewhere (Hall & Farkas, 2000).

There is presently no information on the biology of W. monegrosensis and this study has examined only one-third of the described species of Wohlfahrtia. However, the morphological and molecular analyses in this study suggest that W. monegrosensis is closely related to W. magnifica and W. vigil (14, 15), both of which have parasitic larvae, although whether W. vigil is obligate or facultative is not certain (Table 1). Whether or not W. monegrosensis is also parasitic and, if so, facultative or obligate, remains to be determined. However, if it is parasitic then it does not appear to be an important species in terms of causing myiasis in larger livestock. Identification of larvae from cases of sheep myiasis in the region of discovery of W. monegrosensis showed 96% of cases (45/47) were caused by W. magnifica (M.J.R. Hall, N.P. Wyatt, J. Lucientes Curdi and J.A. Castíllo Hernandez, unpublished data, 1995). The Los Monegros region where the species was discovered and after which it is named is in the central part of the Ebro Valley, to the east of Zaragoza. It has an arid climate with only 200–400 mm of rainfall per year, frequent northwest or southeast winds that are very desiccating and high summer temperatures (to > 40 °C). Intensive cereal cultivation on the flatter lands, which are planted and left as bare gypsum soils in alternate years, has left the climax vegetation of juniper woods (Juniperus thurifera) only on the hillsides and hilltops. Wild rabbits are abundant in the area and, if W. monegrosensis is parasitic, they represent a potential host.

Although our analysis was limited by the degraded DNA from North American specimens of W. vigil, it shows a > 6% difference between the sequence structures of New World and Old World W. vigil. This was based on specimens from Europe and a specimen from New Mexico in western North America. Morphology also suggested that there might be two groups of W. vigil, one in western North America, the other including specimens from eastern North America and Europe. This combined evidence strongly supports the need for a thorough re-examination of the synonymy of W. vigil to determine whether it is indeed a single species or a complex of two or three species or subspecies (i.e. with either New World and Old World taxa, or two taxa in North America, one of which is also found in the Palaearctic region, or three taxa, two in North America and one in the Palaearctic region). In addition to the morphological and molecular evidence for diversity within W. vigil, there appear to be differences in its larval biology associated with location, first highlighted by Salem (1938). Hence, in North America, W. vigil has been recorded as an obligate, myiasis-causing parasite of mink, fox, rabbit, cat, dog (Gassner & James, 1948; Strickland, 1949; Greve, 1968), ducklings (Wobeser et al., 1981), rodents (Morrison, 1937; Boonstra, 1976; Craine & Boonstra, 1986; Schorr & Davies, 2002) and humans, especially infants, in whom it typically causes a furuncular myiasis (Ford, 1936; Rich & Knowlton, 1937; O’Rourke, 1954; Haufe & Nelson, 1957; Stabler, 1961; Stabler et al., 1962; Dong, 1977; Eads, 1979). Both described forms of W. vigil in North America have been reported to cause myiasis, W. vigil in the east and W. opaca in the west (Gassner & James, 1948), but the majority of the cases reported above (12/16) were caused by the western form. In the Palaearctic region, W. vigil has only once been reported to cause myiasis in mammals (traumatic myiasis in a human infant; Grunwald et al., 1998), but the accuracy of this identification is unclear. By contrast, its larvae have been reported as cutaneous and cavity parasites of frogs and toads (Cepelák and Artamonov, in Povolný & Verves, 1997). It is perhaps not surprising that there should be a genetic divergence between these geographically separate populations of W. vigil because they are neither synanthropic nor pests of widely dispersed livestock and, therefore, are unlikely to be moved around to any great extent by human activity. Future studies will concentrate on more robustly addressing the unresolved questions surrounding the identity of W. vigil and its synonyms.

Conflicts of interest

All authors declare no conflicts of interests.