Can sibling species of the Drosophila willistoni subgroup be recognized through combined microscopy techniques?

ABSTRACT

In several arthropod groups, male genitalia is the most important feature for species identification, especially in cryptic species. Cryptic species are very common in the Drosophila genus, and the Neotropical Drosophila willistoni species group is a good example. This group currently includes 24 species divided into three subgroups: alagitans, bocainensis and willistoni. There are six sibling species in the willistoni subgroup – D. willistoni, D. insularis, D. tropicalis, D. equinoxialis, D. pavlovskiana and D. paulistorum, which is a species complex composed of six semispecies – Amazonian, Andean-Brazilian, Centroamerican, Interior, Orinocan and Transitional. The objective of this study was to characterize male genitalia of the willistoni subgroup, including the D. paulistorum species complex, using scanning electron microscopy and light microscopy. We also tried to contribute to the identification of these cryptic species and to add some comments about evolutionary history, based on male genitalia characters. Despite being cryptic species, some differences were found among the siblings, including the Drosophila paulistorum semispecies.

Keywords:
Cryptic species; Drosophila willistoni subgroup; Drosophila paulistorum complex; Male genitalia; Semispecies

Introduction

Male terminalia is one of the most important traits used to identify cryptic species, which are very common in the Drosophila Fallén genus. The Neotropical Drosophila willistoni species group is a good example of cryptic speciation. This group includes 24 species, which are divided into three subgroups: alagitans, bocainensis and willistoni (Bachli, 2015); the last of them showing various taxonomic levels with successive degrees of reproductive isolation (Robe et al., 2010). The willistoni subgroup includes six sibling species: D. willistoni Sturtevant, 1916, D. equinoxialis Dobzhansky, 1946, D. insularis Dobzhansky, 1957, D. tropicalis Burla and Da Cunha, 1949, D. pavlovskiana Katritsis and Dobzhansky, 1967 and D. paulistorum Dobzhansky and Pavan, 1949. These species are almost morphologically indistinguishable based on external morphology, exhibit varying degrees of premating isolation and usually do not cross-hybridize (Ehrmann and Powell, 1982). Within the subgroup, Drosophila paulistorum is a species complex, or also referred as a superspecies, composed of six semispecies (Dobzhansky and Spassky, 1959; Perez-Salas et al., 1970). The willistoni subgroup also shows taxonomic differentiation at the subspecies level: D. willistoni differentiates into the willistoni and quechua subspecies (Ayala and Tracey, 1973); D. tropicalis contains the tropicalis and cubana subspecies (Townsend, 1954); and D. equinoxialis is divided into the equinoxialis and caribbensis subspecies (Ayala et al., 1974).

According the review of Cordeiro and Winge (1995), the sibling group is still in an active process of speciation and all levels of this process can be observed. The authors suggest two steps of speciation: incipient isolation, represented by the subspecies. The second step of speciation is exemplified by the semispecies, which show several degrees of reproductive isolation ranging from complete isolation to the presence of fertile offspring (Cordeiro and Winge, 1995) which was observed in the crossings of the Transitional semispecies with the Andean-Brazilian and the Centroamerican semispecies (Ehrman, 1961, 1965).

D. willistoni has the broader distribution of the group (Fig. 1), spanning from Central Mexico and Florida to Southern Brazil and Northern Argentina, and from the Atlantic to the Pacific Ocean (Dobzhansky and Powell, 1975; Ehrmann and Powell, 1982), even in areas of human disturbance (Valiati and Valente, 1996). D. willistoni is uninterruptedly distributed over this area, except in deserts and high altitudes (Ehrmann and Powell, 1982). Other sibling species have narrower distributions within the distribution of D. willistoni, except D. insularis, which is endemic to Saint Kitts and Saint Lucia of the Antilles Islands (Dobzhansky et al., 1957), and D. pavlovskiana, which has been found only once in Guyana (Spassky et al., 1971) and has not been collected since then (Fig. 1).

Fig. 1
Geographic distribution of the D. willistoni species subgroup, according to Spassky et al. (1971), Winge (1971), Dobzhansky and Powell (1975), Ehrmann and Powell (1982), Santos and Valente (1990).

The D. paulistorum semispecies occur from Southern Brazil to Central America (Guatemala) and Trinidad (Dobzhansky and Spassky, 1959) (Fig. 2). According to Dobzhansky et al. (1964), when the semispecies’ territories overlap, they apparently do not interbreed. Previous studies suggested that the differences among morphological, physiological and ecological traits within the semispecies are too small to distinguish each semispecies (Pasteur, 1970; Perez-Salas and Ehrmann, 1971) and could only be recognized by examination of the gene arrangements on their chromosomes (Kastritsis, 1967; Rohde et al., 2006) and by crossing tests (Perez-Salas and Ehrmann, 1971).

Fig. 2
Geographic distribution of the D. paulistorum superespecies, according to Spassky et al. (1971), Winge (1971), Dobzhansky and Powell (1975), Ehrmann and Powell (1982), Santos and Valente (1990).

Despite being a traditionally studied group with an exciting evolutionary history, there is a lack of studies concerning the morphology in the early stages of development and also in adults. Some morphological studies of adults were made after the species were described. Burla et al. (1949) presented illustrations of maxillar palpi, vaginal plates, spermatechae and hypandria of D. willistoni, D. tropicalis, D. equinoxialis and D. paulistorum. Hsu (1949) briefly described D. willistoni and D. equinoxialis male genitalia, in addition of some species of the alagitans and bocainensis subgroups. Malogolowkin (1952) provided a very detailed description of the male and female genitalia of D. willistoni, D. equinoxialis, D. tropicalis, D. paulistorum and some species of the bocainensis subgroup. Spassky (1957) shown illustrations of hypandria, aedeagi, surstily and prensiseta of five sibling species – D. willistoni, D. tropicalis, D. equinoxialis, D. paulistorum and D. insularis. Pasteur (1970) made a biometrical comparison of four semispecies of D. paulistorum regarding wing and tibia lengths, wing-to-tibia ratios, wing size and number of prensiseta on surstylus. Vilela and Bächli (1990) redescribed D. willistoni and provided several illustrations of the male terminalia of this species. Eberhard and Ramirez (2004) presented several Scanning Electron Microscopy (SEM) images of male and female terminalia of D. willistoni. Rohde et al. (2010) provided photos of the hypandria of D. willistoni, D. equinoxialis, D. tropicalis and D. paulistorum without indicating the semispecies while suggesting the importance of this structure for their identification. Recently, Souza et al. (2014) provided SEM images of D. willistoni, which was used as outgroup of the phylogenetic reconstruction of the D. saltans group. Civetta and Gaudreau (2015) shown photos of external male genitalia and aedeagus of D. willistoni and the subspecies D. willistoni quechua. Also, there are SEM images of male terminalia of four sibling species (D. willistoni, D. tropicalis, D. equinoxialis and D. paulistorum) available in Emilio Goeldi Museum database (marte.museu-goeldi.br).

Burla et al. (1949) only found slight morphological differences that were insufficient for the identification of single individuals, while Spassky (1957) noted differences in the male genitalia that did permit such identification. In other way, Malogolowkin (1952) described the general morphology of the male genitalia of four sibling species (D. willistoni, D. equinoxialis, D. tropicalis and D. paulistorum) as very similar but completely different from non-sibling species. In addition, this author found no differences in the penises of these four sibling species. Spassky (1957), however, observed that the shapes of the penises and their gonapophyses differed in the sibling species.

With respect to D. insularis, only a few illustrations were presented by Spassky (1957). The previous descriptions and illustrations of D. paulistorum were based on Andean-Brazilian specimens once the specimens were collected in areas where only this semispecies is found (Mogi das Cruzes, São Paulo, Brazil in Burla et al. (1949) and Malogolowkin (1952); Tamandaré, Pernambuco, Brazil in Rohde et al. (2010)). The remaining species, D. pavlovskiana, has not been collected and is no longer available in Stock Centers. There is no description in the literature of the male genitalia of this species.

In this scenario, our objective was to characterize and compare the male terminalia of the species of the D. willistoni subgroup, including the D. paulistorum semispecies complex, using Scanning Electron Microscopy and Light Microscopy. We attempt to describe this morphological trait within species and find features that differentiate the sibling species and semispecies of the willistoni subgroup and verify if these species can be recognized based on characters of the male terminalia.

Material and methods

Fly stocks

The stocks were reared in a cornmeal medium (Marques et al., 1966) at a constant temperature and humidity (17 ± 1 °C; 60% rh). Some individuals were preserved in 70% ethanol, and mounted stubs and slides will be deposited in Fundação Oswaldo Cruz (Fiocruz) Collection. All strains used in this study are listed in Table 1.

Table 1
Species and strains used in this study. The numbered strains are represented in the figures.

Species	Semispecies	Population sample	Localities	Source
D. equinoxialis [1]		Mexico	Apazapán/Mexico	Lee Ehrman and Yong Kyu Kim
D. equinoxialis		FNT12	Belterra/Brazil	Mário Josias Müller
D. equinoxialis		FNT19	Belterra/Brazil	Mário Josias Müller
D. equinoxialis		FNT32	Belterra/Brazil	Mário Josias Müller
D. equinoxialis		Tape03	Santarém/Brazil	Mário Josias Müller
D. equinoxialis		Tape31	Santarém/Brazil	Mário Josias Müller
D. equinoxialis		Tape32	Santarém/Brazil	Mário Josias Müller
D. equinoxialis		Tape63	Santarém/Brazil	Mário Josias Müller
D. equinoxialis		Tape89	Santarém/Brazil	Mário Josias Müller
D. equinoxialis		207018	Belém do Pará/Brazil	Marlucia Martins
D. insularis [2]		SL03	Saint Lucia/Lesser Antilles	Jeffrey R. Powell
D. insularis		SL15	Saint Lucia/Lesser Antilles	Jeffrey R. Powell
D. tropicalis [3]		0801	San Salvador/El Salvador	Tucson Stock Center
D. willistoni [4]		WIP4	Salvador/Brazil	Antônio R. Cordeiro and Helga Winge
D. willistoni		Gdh4-1	Guadalupe/Caribe	Tucson Stock Center
D. willistoni		Ribeirão Preto	Ribeirão Preto/Brazil	Cláudia Rohde
D. willistoni		Mexico	Apazapán/Mexico	Margaret Kidwell
D. willistoni		DW-Pool Ita	Itapuã/Brazil	André Schnorr
D. willistoni		DW-Pool Faz	Serra do Cipó/Brazil	Carlos Vilela
D. willistoni		CIP	Serra do Cipó/Brasil	Carlos Vilela
D. willistoni		COR	Coronilla/Uruguay	Beatriz Goñi
D. willistoni		ISC	Santa Catarina/Brazil	Daniela de Toni
D. willistoni		DLA	Dois Lajeados/Brazil	Cláudia Rohde
D. willistoni		MAS	Morro Santana/Brazil	Ana Lauer Garcia
D. willistoni		Joinville	Joinville/Brazil	Jonas Döge
D. willistoni		17A2	Eldorado do Sul/Brazil	Vera L.S. Valente
D. paulistorum [5]	Amazonian	Pará q3	Belém do Pará/Brazil	Lee Ehrmann and Yong Kiu Kim
D. paulistorum		Tape 60	Santarém/Brazil	Mário Josias Müller
D. paulistorum [6]	Andean-Brazilian	MLC	Santa Catarina/Brazil	Marco Gottschalk
D. paulistorum		RIB	Ribeirão Preto/Brazil	Vera L.S. Valente
D. paulistorum		RMT	Porto Alegre/Brazil	Ana Lauer Garcia
D. paulistorum		Pool-Bur	Brasília/Brazil	Rosana Tidon
D. paulistorum		Játon-Sacha	Jaton Sacha/Equador	Margaret Kidwell
D. paulistorum [7]	Centroamerican	Lancetilla	Lancetilla/Honduras	Lee Ehrman and Yong Kyu Kim
D. paulistorum [8]	Interior	INT	Llanos/Colombia	Lee Ehrman and Yong Kyu Kim
D. paulistorum [9]	Orinocan	ORI	Georgetown/Guyana	Lee Ehrman and Yong Kyu Kim
D. paulistorum [10]	Transitional	Santa Marta	Santa Marta/Colombia	Lee Ehrman and Yong Kyu Kim

Species recognition

Species were confirmed with the Acph-1 (Acid Phosphatase) electrophoresis protocol (Garcia et al., 2006). Also, DNA sequences were generated and compared with sequences available in GenBank – mitochondrial gene fragments COI (Cytochrome oxidase I) (deposited by Gleason et al., 1998), COII (Cytochrome oxidase II) (deposited by Robe et al., 2010), and nuclear genes fragment Adh (Alcohol dehydrogenase) (deposited by Gleason et al., 1998 and Robe et al., 2010). The sequences obtained will be presented in a further study.

Scanning electron microscopy (SEM) preparation and observation

Male terminalia were treated with 10% KOH (Wheeler and Kambysellis, 1966 modified by Bächli et al., 2004) and dissected in glycerol. Terminalia were dehydrated for 20–30 s with acetone washes in the following concentrations: 30%, 50%, 75% and 100%. The entire terminalia and separated parts were mounted in stubs with carbon tape and metalized with gold in a Balzers SCD050 sputter coater. Visualization and image capture were performed in JSM6060 Scanning Electron Microscope in Centro de Microscopia da Universidade Federal do Rio Grande do Sul. We observed approximately 50 seven-day-old specimens of each species and semispecies.

Light microscopy preparation and observation

Male terminalia were prepared as previously described. The terminalia, aedeagi and hypandria were mounted in a non-permanent glycerin jelly medium (Klaus et al., 2003). The slides were observed and photographed with a Carl-Zeiss Standard phase contrast microscope. We analyzed the genital structures of 7-day-old males of each species and semispecies (750 individuals).

Terminology and references

The morphological terminology used in this study follow McAlpine (1981), Grimaldi (1990), Vilela and Bächli (1990), Bächli et al. (2004) and Souza et al. (2014). Figures of external male genitalia of D. willistoni subgroup, under light microscopy, are shown in Supplementary material 1 (S1).

The morphological terminology used in this study follow McAlpine (1981), Grimaldi (1990), Vilela and Bächli (1990), Bächli et al. (2004) and Souza et al. (2014). Figures of external male genitalia of D. willistoni subgroup, under light microscopy, are shown in Supplementary material 1 (S1).

Results

Epandrium, surstylus and prensisetae

The cerci are not fused to the epandrium (Figs. 3 and 4; Fig. S1) in all the species analyzed. The surstylus is elongated into a hook at the bottom, is not micropubescent, has up to 12 prensiseta (also called primary teeth) in D. paulistorum Orinocan and D. paulistorum Interior, and up to 18 prensiseta in D. equinoxialis (Fig. 5) in addition to one large prensisetae and one or two seta on the ventral hook.

Fig. 3
General example of external male genitalia of the D. willistoni subgroup (picture of D. paulistorum Amazonian [5]). Scale bar 50 μm. EP, epandrium; CE, cerci; SU, surstylus; PR, prensisetae; AE, aedeagus; VL, ventral lobe; HY, hypandrium.

Fig. 4
Scanning electron microscopy of D. willistoni species subgroup external male genitalia. Scale bar 50 μm. The strains are in brackets and listed in Table 1. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropiacalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

Fig. 5
Scanning electron microscopy of D. willistoni species subgroup surstyli and prensiseta. Scale bar 20 μm. The strains are in brackets and listed in Table 1. SU, Surstylus; PR, Prensisetaes; VH, Ventral hook. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

The cerci are not fused to the epandrium (Figs. 3 and 4; Fig. S1) in all the species analyzed. The surstylus is elongated into a hook at the bottom, is not micropubescent, has up to 12 prensiseta (also called primary teeth) in D. paulistorum Orinocan and D. paulistorum Interior, and up to 18 prensiseta in D. equinoxialis (Fig. 5) in addition to one large prensisetae and one or two seta on the ventral hook.

Drosophila equinoxialis presents 18 prensiseta, nine smaller and nine larger, and one setae on the ventral hook (Figs. 4A and 5A). D. insularis has approximately 17 prensiseta of equal size and one setae in the ventral hook (Figs. 4B and 5B). D. tropicalis has 13–14 crescent size prensiseta and one setae in the ventral hook (Figs. 4C and 5C). D. willistoni also has 13 crescent sized prensiseta and one setae in the ventral hook (Figs. 4D and 5D).

Among the D. paulistorum semispecies, we found that Drosophila paulistorum Amazonian has a surstylus with 15 prensiseta, nine longer and six shorter, and one setae in the ventral hook (Figs. 4E and 5E). D. paulistorum Andean-Brazilian has a surstylus with 15 prensiseta, eight longer and seven shorter, and two seta in the ventral hook (Figs. 4F and 5F). D. paulistorum Centroamerican has a surstylus with 15 prensiseta, eight larger and seven smaller, and two seta in the ventral hook (Figs. 4G and 5G). D. paulistorum Interior has a surstylus with 12 crescent size prensisetae and two seta in the ventral hook (Figs. 4H and 5H). D. paulistorum Orinocan has a surstylus with 12 prensiseta of approximately the same size and two seta in the ventral hook (Figs. 4I and 5I). D. paulistorum Transitional has a surstylus with 12 prensisetae of approximately the same size, with one larger prensisetae in the middle and two seta in the ventral hook (Figs. 4J and 5J). The distance between the sides of surstylus and hooks could be an artifact of the SEM preparation and is not considered a diagnostic character for species identification.

Hypandrium

The hypandrium is smaller than the epandrium; it is approximately 1/4 to 1/3 of the size of the epandrium. The hypandria in all of the species analyzed have, in the apical region, one pair of heavily sclerotized median teeth, lobes with paramedian seta on the apex and lateral extensions (Figs. 6 and 7). The relative size and thickness of the median teeth, as well as the size and shape of the lobes, vary within the species (Figs. 6 and 7).

Fig. 6
Scanning Electron Microscopy of the D. willistoni species subgroup hypandria. Scale bar 50 μm. The strains are in brackets and listed in Table 1. LO, lobes; TE, teeth; LE, lateral extensions; SE, seta. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

Fig. 7
Hypandria of D. willistoni species group. Scale bar 0.1 mm. The strains are in brackets and listed in Table 1. LO, lobes; TE, teeth; LE, lateral extensions; SE, seta. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

In D. equinoxialis, the hypandrium is triangular, with well-developed trapezoidal lobes and lobe seta convergent to teeth. The teeth are large and thick, twice the height of the lobes, aligned with the lobes, and do not touch each other. The lateral extensions are very prominent toward the top (Figs. 6A and 7A). D. insularis has a triangular hypandrium, with almost absent lobes. This species presents one or two paramedian seta in the lobes and presents the smaller and more separated teeth that occur in the willistoni subgroup (Figs. 6B and 7B). D. tropicalis presents a triangular hypandrium, with very large round lobes and seta convergent to teeth and almost touching them. It also has large thick teeth, twice the height of the lobes, inserted slightly below the lobe line. The lateral extensions are located under the lobes (Figs. 6C and 7C). D. willistoni also has a triangular hypandrium, with subtle round lobes. Its very close large, thick teeth are thrice the height of the lobes and are inserted far below the lobe line. The lateral extensions are similar to those of D. tropicalis, but are adjacent to the lobes (Figs. 6D and 7D).

In the D. paulistorum complex, D. paulistorum Amazonian presents a square-shaped hypandrium, dome-shaped lobes, and slightly convergent seta. The lobes almost touch each other and the teeth are large and slightly separated, inserted in the lobe line. The lateral extensions are prominent toward the top, but less than in D. equinoxialis (Figs. 6E and 7E). D. paulistorum Andean-Brazilian presents a square-shaped hypandrium, slightly square-shaped lobes, convergent seta and very close, medium-sized, thin teeth that are twice the height of the lobes and inserted in the lobe line. There is a visible gap between the lobes and teeth. The lateral extensions are almost continuous with the lobes (Figs. 6F and 7F). D. paulistorum Centroamerican presents a rectangular hypandrium, which is the most elongated in D. willistoni subgroup, irregular-shaped lobes, convergent seta and very close, medium-sized, thin teeth that are twice the height of the lobes and inserted below the lobe line. The lateral extensions are similar to D. willistoni but nearer to the lobes (Figs. 6G and 7G). D. paulistorum Interior is very similar to D. paulistorum Andean-Brazilian, but the lateral extensions are not continuous with the lobes and there is no gap between the teeth and lobes (Figs. 6H and 7H). D. paulistorum Orinocan hypandrium is the smallest of the subgroup, is square-shaped with small round lobes, convergent seta and medium-sized, thin teeth that are thrice the height of the lobes and inserted a little below the lobe line. Lateral extensions are expanded to external sides of the lobes (Figs. 6I and 7I). D. paulistorum Transitional presents a square-shaped hypandrium, small round lobes very close to the teeth, convergent seta and large thick teeth that are almost thrice the height of the lobes and inserted in the lobe line. Lateral extensions are prominent toward the top, higher than the lobes (Figs. 6J and 7J).

Aedeagus, aedeagal apodeme, paramere and lateral projections

In all of the species of the D. willistoni subgroup that have been analyzed, the aedeagus is dorsally membranous and ventrally directed downwards, as a bird beak-like protusion at the distal end (distiphallus), with two lateral projections at the anterior half, covered with some tiny spines that are not always visible in preparations. The aedeagal apodeme is as long as the aedeagus, is bar shaped and is linked to the aedeagus by a membranous tissue. The parameres are smooth and are also linked to the apodeme by a membranous tissue. The paramere anchors the aedeagus to the hypandrium through the lateral expansions of hypandrium.

The most noticeable difference within the willistoni subgroup is the distal portion of the aedeagus (distiphallus). This is very prominent and curved in Drosophila tropicalis (Figs. 8C, 9C and 10C), long and straight in D. willistoni (Figs. 8D, 9D and 10D), and straight and shorter than D. willistoni in D. equinoxialis (Figs. 8A, 9A and 10A). In D. insularis, this structure is the shortest among the siblings (Figs. 8B, 9B and 10B). There are small variations in size within the D. paulistorum semispecies, but the shape of the distiphallus is unique in each incipient species (Figs. 8E-J, 9E-J and 10E-J).

Fig. 8
Scanning electron microscopy of D. willistoni species subgroup aedeagi. Scale bar 20 μm. The strains are in brackets and listed in Table 1. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

Fig. 9
Aedeagi, aedeagal apodeme and parameres of the D. willistoni species subgroup; lateral view. Scale bar 0.1 mm. The strains are in brackets and listed in Table 1. AA, Aedeagal apodeme; DP, Distiphallus; LP, Lateral projections; PA, Paramere. (A) D. equinoxialis [1]; (B) D. insularis [2]; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

Fig. 10
SEM of aedeagi, aedeagal apodeme and parameres of the D. willistoni species subgroup; lateral view. Scale bar 50 μm. The strains are in brackets and listed in Table 1. AA, Aedeagal apodeme; DP, Distiphallus; LP, Lateral projections; PA, Paramere. (A) D. equinoxialis; (B) D. insularis; (C) D. tropicalis [3]; (D) D. willistoni [4]; (E) D. paulistorum Amazonian [5]; (F) D. paulistorum Andean-Brasilian [6]; (G) D. paulistorum Centroamerican [7]; (H) D. paulistorum Interior [8]; (I) D. paulistorum Orinocan [9] and (J) D. paulistorum Transitional [10].

The lateral projections exhibit some differences within the subgroup, which are more notable in D. willistoni, D. tropicalis, D. equinoxialis and D. insularis. In D. equinoxialis and D. paulistorum Orinocan the distal portion of lateral expansions is rounded; in D. insularis, D. tropicalis and D. willistoni it is pointy; and in the remaining D. paulistorum semispecies, it is slightly pointy.

The aedeagal apodeme also shows variation; however, it is not species specific. In D. equinoxialis, D. insularis, D. tropicalis and D. willistoni, the aedeagal apodeme is rod shaped, without ornamentation in the distal portion (Figs. 9A-D and 10A-D). In the Amazonian, Andean-Brazilian, Interior and Transitional semispecies, the aedeagal apodeme is also rod shaped but with a small rounded expansion in the distal portion (Figs. 9E,F,H,J and 10E,F,H,J), and in the Centroamerican and Orinocan semispecies, there is a fan-like expansion in the distal area (Figs. 9G,I and 10G,I).

Discussion

This study sheds new light on the identification of the D. willistoni sibling species subgroup, especially regarding the D. paulistorum complex. We presented here for the first time images of the male terminalia of the six semispecies of the D. paulistorum. Now, the male identification of the cryptic species of this subgroup could be easier and quicker than enzymatic, molecular and chromosomal approaches.

We found major differences especially in D. willistoni, D. tropicalis, D. equinoxialis and D. insularis. While there is a strong sexual isolation within these species, it has been reported that they occasionally interbreed (Dobzhansky et al., 1957; Winge and Cordeiro, 1963; Winge, 1965; Cordeiro and Winge, 1995), and several degrees of reproductive affinity are present between the sibling species group (reviewed in Cordeiro and Winge, 1995).

Our findings are consistent with the results presented in Spassky (1957) regarding D. equinoxialis, D. insularis, D. tropicalis, D. willistoni and D. paulistorum Andean-Brazilian. However, a notable aspect of this comparison is that intraspecific variation was not observed in our results. We analyzed the male genitalia of D. equinoxialis, D. insularis, D. willistoni and D. paulistorum Andean-Brazilian from several localities (Table 1), including some recently collected strains, and no remarkable character variation was found (data not shown). This fact does not imply that there are no intraspecific variation in these species and in other species and semispecies of the willistoni subgroup. Also, laboratory strains may have less character variation than found in nature. Burla et al. (1949), however, concluded that “the variability is great enough to make identification of single individuals hazardous”.

Burla et al. (1949) found differences in the hypandria of four of the sibling species – D. willistoni, D. paulistorum, D. tropicalis and D. equinoxialis – and described the D. tropicalis hypandrium as similar to that of D. paulistorum in shape, although larger. These authors also stated that the D. willistoni hypandrium is the most distinctive. In contrast with the findings of Burla et al. (1949) we observed that the hypandria of D. insularis and D. tropicalis are the most distinctive with respect to the remaining species. D. willistoni seems to be more similar to the D. paulistorum semispecies than to the other species. Some features could only be observed in SEM: D. willistoni and D. insularis presented two seta in the apex of the lobes, only in one side (Fig. 6B); Despite the low frequency of this modification (1:50 in D. willistoni and 2:50 in D. insularis), it is an interesting feature. The specimens with this characteristic did not present any other peculiarity.

Pasteur (1970) considered the teeth of the claspers (the prensisetae in the surstylus) to be the single character of the male genitalia that differentiates the D. paulistorum incipient species. In our results (Fig. 5), the number of prensisetae is consistent with the range of values previously observed by Pasteur (1970). We observed two groups regarding the number of prensisetae – the semispecies Amazonian, Andean-Brazilian and Centroamerican with 15 prensisetaes each and Centroamerican, Interior and Transitional semispecies with 12 prensisetaes each, although the size and arrangement of the prensisetae are not the same for each one. Pasteur (1970) found variation in the number of prensisetaes in each semispecies from different localities. In this study, we observed D. paulistorum Andean-Brazilian from several localities and the number of prensisetae was constant, even in the recently collected strains.

In the present report, we show that there are several diagnostic characters of the male genitalia useful for differentiating the species and even the semispecies of the D. willistoni species subgroup. However, the aedeagus is not the most important trait for identifying the subgroup's cryptic species, as is common in other Drosophila species groups. In the studied species, the hypandrium seems to be the main character for species identification. In addition to this, we suggest the visualization of the aedeagus under light microscopy for identification purposes, since this is a membranous structure and may be deformed in SEM.

Based on visual observations of hypandrium D. insularis seems to be the most distinctive species, related to the other sibling species. Within the D. paulistorum semispecies, the most dissimilar is the transitional, especially with respect to the hypandrium shape, surstylus and prensisetae. Such variation is in accord with the assumptions of Spassky et al. (1971) and Dobzhansky (1970), stating that the diversification of the semispecies is apparently still in progress.

Concerning the evolutionary relationship among the willistoni subgroup, an attempt to phylogenetic reconstruction was made using the characters of the male terminalia observed and described in this study. The reconstruction, however, were inconclusive, since the generated tree presented some polytomies and low support values for the characters (data not shown). Despite of the differences observed, it is possible that these characters did not accumulated enough differences to represent the evolutionary history of the subgroup. Nevertheless, these characters could be useful in a combined phylogeny and will be presented in a further study.

Although complete reproductive isolation is not present within the species and semispecies of the D. willistoni subgroup (reviews in Ehrmann and Powell (1982) and Cordeiro and Winge (1995)), the number of differences among the male genitalia found in our observations is relevant, especially between the D. paulistorum semispecies. In insects, the rapid divergence in male genitalia is so pronounced that even recently diverged sibling species show a high degree of variation in the male genitalia (Richards, 1927; Liu et al., 1996; Song, 2009), as we have observed in the D. willistoni species subgroup. Finally, in response to our own question, sibling species of the D. willistoni subgroup can be recognized through combined microscopy techniques.

Acknowledgements

We thank to CNPq, CAPES and FAPERGS (process 10/0028-7) for financial support. We thank to Dr. Hermes José Schmitz, Dr. Marco Gottschalk and MsC. Jean Lucas Poppe for precious suggestions. We thank to Dr. Yong Kyu Kim, Dr. Lee Ehrmann, Dr. Jeffrey R. Powell and all the other collectors for the provided strains. We also thank the anonymous reviewers for the suggestions.

Appendix A Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi: 10.1016/j.rbe.2015.09.006.

References

• Ayala, F.J., Tracey, M.L., 1973. Enzyme variability in the Drosophila willistoni group. VIII. Genetic differentiation and reproductive isolation between two subspecies. J. Hered. 120-124.
• Ayala, F.J., Tracey, M.L., Barr, L.G., Ehrenfeld, J.G., 1974. Genetic and reproductive differentiation of the subspecies D. equinoxialis caribbensis Evolution. 28, 24-41.
• Bachli, G., 2015. Taxodros – The Database on Taxonomy of Drosophilidae, Available at: http://www.taxodros.uzh.ch (accessed in 02/2015).
  » http://www.taxodros.uzh.ch
• Bächli, G., Vilela, C.R., Escher, S.A., Saura, A., 2004. The Drosophilidae (Diptera) of Fennoscandia and Denmark. Fauna Entomol. Scand. 39, 1-364.
• Burla, H., DaCunha, A.B., Cordeiro, A.R., Dobzhansky, T., Malogolowkin, C., Pavan, C., 1949. The willistoni group of sibling species of Drosophila Evolution. 3, 300-314.
• Civetta, A., Gaudreau, C., 2015. Hybrid male sterility between Drosophila willistoni species is caused by male failure to transfer sperm during copulation. BMC Evol. Biol. 15, 75-83.
• Cordeiro, A.R., Winge, H., 1995. Levels of evolutionary divergence of Drosophila willistoni sibling species. In: Levine, L. (Ed.), Genetics of Natural Populations ofTheodosius Dobzhansky. Columbia University Press, New York, pp. 241–261.
• Dobzhansky, T., 1970. Genetics of the Evolutionary Process. Columbia, New York.
• Dobzhansky, T., Powell, J.R., 1975. The willistoni group of sibling species of Drosophila In: King, R.C. (Ed.), Handbook of Genetics. Plenum Press, New York, pp. 589–622.
• Dobzhansky, T., Spassky, B., 1959. Drosophila paulistorum a cluster of species in statu nascendi Proc. Natl. Acad. Sci. U. S. A. 45, 419-428.
• Dobzhansky, T., Ehrman, L., Pavlovsky, O., 1957. Drosophila insularis, a New Sibling Species of the willistoni Group, vol. 5721. University of Texas Publications, pp. 39–47.
• Dobzhansky, T., Ehrman, L., Pavlovsky Spassky, B., 1964. The superespecies D. paulistorum Proc. Natl. Acad. Sci. U. S. A. 51, 3-9.
• Eberhard, W.G., Ramirez, N., 2004. Functional morphology of the male genitalia of four species of Drosophila: failure to confirm both lock and key and male female conflict. Ann. Entomol. Soc. Am. 97, 1007-1017.
• Ehrman, L., 1961. The genetics of sexual isolation in Drosophila paulistorum Genetics. 46, 1025-1038.
• Ehrman, L., 1965. Direct observation of sexual isolation between allopatric and between sympatric strains of the different Drosophila paulistorum races. Evolution. 19, 459-464.
• Ehrmann, L., Powell, J.R., 1982. The Drosophila willistoni species group. In: Ashburner, M., Carson, H.L., Thompson Jr., J.N. (Eds.), The Genetics and Biology of Drosophila, vol. 3b. Academic Press Inc., New York, pp. 193–220.
• Garcia, A.C.L., Rohde, C., Audino, G.F., Valente, V.L.S., Valiati, V.H., 2006. Identification of the sibling species of the Drosophila willistoni subgroup through the electrophoretical mobility of acid phosphatase-1. J. Zool. Syst. Evol. Res. 44, 212-216.
• Gleason, J.M., Griffith, E.C., Powell, J.R., 1998. A molecular phylogeny of the Drosophila willistoni group: conflict between species concepts?. Evolution. 52, 1093-1103.
• Grimaldi, D.A., 1990. A phylogenetic revised classification of genera in the Drosophilidae (Diptera). Bull. Am. Mus. Nat. Hist. 197, 1-139.
• Hsu, T.C., 1949. XI. The external genital apparatus of male Drosophilidae in relation to systematics. In: Studies in Genetics of Drosophila The University of Texas Publications, pp. 80–142.
• Kastritsis, C.D., 1967. A comparative study of the chromosomal polymorphs in the incipient species of the Drosophila paulistorum complex. Chromosoma 19, 208–222.
• Klaus, A.V., Kulasekera, V.L., Schawaroch, V., 2003. Three-dimensional visualization of insect morphology using confocal laser scanning microscopy. J. Microsc. 212, 107–121.
• Liu, J., Hercer, J.M., Stam, L.F., Gibson, G.C., Zeng, Z.B., Laurie, C.C., 1996. Genetic analysis of a morphological shape difference in the male genitalia of Drosophila simulans and mauritiana Genetics 142, 1129–1145.
• Malogolowkin, C., 1952. Sobre a genitália dos Drosophilidae (Diptera). III. Grupo willistoni do gênero Drosophila Rev. Bras. Biol. 12, 79–96.
• Marques, E.K., Napp, M., Winge, H., Cordeiro, A.R., 1966. A cornmeal, soybean flour, wheat germ medium for Drosophila Drosoph. Inf. Serv. 41, 187.
• McAlpine, J.F., 1981. Morphology and terminology. Adults. In: McAlpine, J.F., Peterson, B.V., Shewell, G.E., Teskey, H.T., Vockeroth, J.R., Wood, D.M. (Eds.), Manual of Nearctic Diptera, vol. 1. Biosystematics Research Institute, Ottawa, pp. 9–63.
• Pasteur, G., 1970. Biometrical data on the semispecies of the Drosophila paulistorum complex. Evolution. 24, 156-168.
• Perez-Salas, S., Ehrmann, L., 1971. Mechanisms of male sterility in hybrids of the Drosophila paulistorum group. Genetics. 69, 63-70.
• Perez-Salas, S., Richmond, R.C., Pavlovsky, O.A., Kastritsis, C.D., Ehrman, L., Dobzhansky, T., 1970. The interior semispecies of Drosophila paulistorum Evolution. 24, 519-527.
• Richards, O., 1927. The specific characters of the British humblebees (Hymenoptera). Trans. Entomol. Soc. Lond. 75, 233-268.
• Robe, L.J., Cordeiro, J., Loreto, E.L.S., Valente, V.L.S., 2010. Taxonomic boundaries, phylogenetic relationships and biogeography of the Drosophila willistoni subgroup (Diptera: Drosphilidae). Genetica. 138, 601-617.
• Rohde, C., Garcia, A.C.L., Valiati, V.H., Valente, V.L.S., 2006. Chromosomal evolution of sibling species of the Drosophila willistoni group. I. Chromosomal arm IIR (Muller's element B). Genetica. 126, 77-88.
• Rohde, C., Monteiro, A.G.F., Cabral, W.B.M., Silva, D.M.I.O., Oliveira, G.F., Montes, M.A., Garcia, A.C.L., 2010. The importance of the identification of the willistoni subgroup of Drosophila at the species level: the first evidence of D. equinoxialis in the Northeast region of Brazil. Drosoph. Inf. Serv. 93, 118–122.
• Santos, R.A., Valente, V.L.S., 1990. On the occurence of Drosophila paulistorim Dobzhansky and Pavan (Diptera, Drosophilidae) in an urban environment: ecological and cytogenetic observations. Evol. Biol. 4, 253-268.
• Song, H., 2009. Species-specificity of male genitalia characterized by shape, size and complexity. Insect Syst. Evol. 40, 159.
• Souza, T.A.J., Noll, F.B., Bicudo, H.E.M.C., Madi-Ravazzi, L., 2014. Scanning electron microscopy of male terminalia and its application to species regocnition and phylogenetic reconstruction in the Drosophila saltans group. PLOS ONE. 9, 1-12.
• Spassky, B., 1957. Morphological Differences Between Sibling Species of Drosophila, vol. 5721. University of Texas Publications, pp. 48–61.
• Spassky, B., Richmond, R.C., Pérez-Salas, S., Pavlovsky, O., Mourão, C.A., Hunter, A.S., Hoentgsberg, H., Dobzhansky, T., Ayala, F.J., 1971. Geography of the sibling species related to Drosophila willistoni, and of the semispecies of the Drosophila paulistorum complex. Evolution. 25, 129-143.
• Townsend, J.I., 1954. Cryptic subspeciation in Drosophila belonging to the subgenus Sophophora Am. Nat. 842, 339-351.
• Valiati, V.H., Valente, V.L.S., 1996. Observations on ecological parameters of urban populations of Drosophila paulistorum Dobzhansky & Pavan (Diptera, Drosophilidae). Rev. Bras. Entomol. 40, 225-231.
• Vilela, C.R., Bächli, G., 1990. Taxonomic studies on Neotropical species of seven genera of Drosophilidae (Diptera). Mitteilungen der Schweizerischen Entomologischen Gesellschaft. 63, 1-332.
• Wheeler, M.R., Kambysellis, M., 1966. Notes on Drosophilidae (Diptera) of Samoa, vol. 6615. University of Texas Publications, pp. 533–563.
• Winge, H., 1965. Interspecific hybridization between the six species of Drosophila willistoni group. Heredity. 20, 9-19.
• Winge, H., PhD thesis. 1971. Níveis de divergência evolutiva no grupo críptico da Drosophila willistoni Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil.
• Winge, H., Cordeiro, A.R., 1963. Experimental hybrids between Drosophila willistoni Sturtevant and Drosophila paulistorum Dobzhansky and Pavan from southern marginal populations. Heredity. 18, 215-222.

Publication Dates

• Publication in this collection
  Oct-Dec 2015

History

• Received
  6 Apr 2015
• Accepted
  4 Aug 2015