Content uploaded by Mihály Földvári
Author content
All content in this area was uploaded by Mihály Földvári on Nov 24, 2024
Content may be subject to copyright.
1211
Accepted by D. Bickel: 8 Mar. 2006; published: 24 May 2006 1
ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN 1175-5334 (online edition)
Copyright © 2006 Magnolia Press
Zootaxa 1211: 1–19 (2006)
www.mapress.com/zootaxa/
A description of a new species of Diasemopsis (Diptera, Diopsidae)
from the Comoro Islands with morphological, molecular and
allometric data
MARTIN CARR1, SAMUEL COTTON2, MIHÁLY FÖLDVÁRI3 & MARION
KOTRBA 4
1Department of Biology, University of York, Heslington, York, YO10 5YW, UK, mc528@york.ac.uk
2The Galton Laboratory, Department of Biology, University College London, 4 Stephenson Way, London, NW1
2HE, UK, s.cotton@ucl.ac.uk
3Hungarian Natural History Museum, Baross u.13, H–1088 Budapest, Hungary, foldvari@nhmus.hu
4Zoologische Staatssammlung München, Münchhausenstr. 21, München D–81247, Germany.
marion.kotrba@zsm.mwn.de
Abstract
A new species of Diasemopsis (Diptera, Diopsidae) from Comoro Islands is described and
illustrated for the first time with allometric datasets. Diasemopsis comoroensis Carr & Földvári is
shown to be genetically close, but morphologically distinct from the widespread Afro-tropical
species D. meigenii (Westwood); notably a significant divergence in the degree of sexual
dimorphism within eye span has occurred between the two species. A revised molecular phylogeny
of the genus Diasemopsis is presented based on the partial sequences of four genes.
Key words: Diopsidae, Diasemopsis, Comoro Islands, new species, morphological description,
genitalia, eye span allometry, molecular phylogeny
Introduction
The diopsid stalk-eyed flies are a diverse Schizophoran family, comprising approximately
160 known species. Diopsidae is divided into the Centrioncinae, which do not possess
eyestalks and the Diopsinae, all of which do. Uniquely amongst dipterans males and
females possess eyestalks, at the end of which are located both their eyes and antennae.
Recent molecular and morphological studies have produced a robust phylogeny of the
stalk-eyed flies (Baker et al. 2001, Meier & Baker 2002, Kotrba and Balke 2006).
Emphasis within these studies has been placed on the genus Diasemopsis, a group of 51
CARR ET AL.
2 © 2006 Magnolia Press
1211
ZOOTAXA species found predominantly in sub-Saharan Africa. The composition of the genus
Diasemopsis has been the source of some debate, with the species D. meigenii and D.
minuta being placed in monotypic genera by Séguy (1955); however later authors have
determined, using both morphological and molecular data, that these species can be
confidently described as belonging to the genus Diasemopsis (Shilito 1971, Baker et al.
2001).
Here we describe a previously uncharacterised Diasemopsis species collected from the
island of Mohéli, Comoro Islands and establish its phylogenetic position within the genus
Diasemopsis. We compared aspects of its morphology with its sibling species, D. meigenii,
and highlighted a recent, rapid divergence in sexual dimorphism with respect to eye span.
Methods and Materials
All allometric and molecular work was performed on individuals taken from the same
laboratory population as the morphologically described specimens.
Morphology
Males: The specimens were studied with an Olympus (SZ 60) stereoscopic
microscope at magnifications of 10–112.5 times. The dissected genital parts (on
microscopic slides) were examined with an Olympus (BX40) light microscope attached
with a drawing tube. The drawings were made with this drawing tube (“camera lucida”).
The original pencil drawings were copied in ink on tracing paper then scanned at 50%. The
photographs were made with the SZ60 microscope equipped with an Olympus C 5050Z
digital camera.
Male genitalia were examined after treatment with 10% NaOH. In some cases the vials
containing the dissected abdomen were heated to reduce the reaction time. After different
parts were cleared sufficiently, they were put in lactic acid, 70% alcohol and glycerine.
The separated postabdomen was put into a drop of gelatine-glycerine on a microscopic
slide. This mixture is solid at room temperature and becomes fluid after careful heating.
After examination and drawing the genital parts were cleaned with NaOH or KOH and
then placed in a small plastic vial filled with glycerine and attached to the same pin which
supports the mounted specimen.
Females: The morphology of internal female genital organs was studied in freshly
killed specimens. The organs were removed under a dissecting scope and embedded on a
microscope slide in polyvinyllactophenol with an admixture of chlorazol black E. The
preparations were studied and documented in bright field as well as DIC contrast using a
Zeiss Axioskop compound microscope equipped with a Zeiss AxioCam digital camera.
© 2006 Magnolia Press 3
DIASEMOPSIS
1211
ZOOTAXA
Allometric Comparisons
D. comoroensis and D. meigenii were reared, in the laboratory, under variable larval
density to generate the variance in adult size required for the investigation of allometry
(David et al. 1998; Cotton et al. 2004). Eclosing individuals (D. comoroensis: 31 males
and 36 females; D. meigenii: 41 males and 38 females) were collected, frozen and
measured to an accuracy of 0.01 mm using a monocular microscope and the image
analysis program NIH Image (Version 1.55). Measurements were taken of eye span
(between the outermost lateral edges of the eye bulbs) and body length (from the front of
the face to the tips of the wings; Baker & Wilkinson 2001).
Absolute trait size data were non-normally distributed so differences between sexes
and species were detected using non-parametric Wilcoxon-tests. Eye span is a highly
allometric trait in stalk-eyed flies (Baker & Wilkinson 2001), so we analysed species
differences and sexual dimorphism of eye span using General Linear Models (GLMs)
containing SPECIES or SEX and BODY LENGTH as fixed factors, their interaction, and an
intercept. The significance of each effect (or interaction) was determined via F-tests on the
change in explained variance upon removal of each term from the full model. A significant
interaction implies that the relationship between eye span and body length (i.e. the
allometric slope of eye span) differs between the groups (species or sexes). An additional
model, comprising SPECIES, SEX and BODY LENGTH main effects, plus all interactions, was
also constructed. The significance of the three-way SPECIES × SEX × BODY LENGTH
interaction was used to identify a difference in sexual dimorphism between the two
species.
Gene Cloning And Phylogenetic Analysis
Genomic DNA was extracted from individual, laboratory-reared, adults ground in
TNES (50mM Tris pH 8.0, 400mM NaCl, 20 mM EDTA, 0.5% SDS), to which Proteinase
K (20mg/ml) was added and the mixture incubated at 37ºC. 5M NaCl was added and the
DNA was precipitated with EtOH. PCR was performed in 50µl volumes (5U Abgene Red
Hot DNA Polymerase, 2.5 mM MgCl2, 0.4mM dNTP). Two nuclear genes, white (w) and
wingless (wg), and two mitochondrial genes, 16S ribosomal RNA (16S) and cytochrome
oxidase subunit II (COII), were chosen for the phylogenetic analysis. The primers and
annealing temperatures used for each gene are shown in Table 1. PCR was performed over
30 cycles, with a ten minute extension time at 72°C in the final cycle.
All PCR products were ligated into the pGEM–T Easy Vector (Promega) and
transformed into Subcloning Efficiency DH5α Chemically Competent Cells (Invitrogen).
Plasmid DNA was extracted using the Qiagen Spin Miniprep kit and sequenced using T7
and SP6 primers (Macrogen Inc, Seoul, Korea). The sequences for each of the genes have
been deposited into the GenBank database (Accession numbers DQ054781 and
AY910526–AY910528). A concatenated alignment of all four genes was created using the
sequences from the other 17 available Diasemopsis species in the GenBank database as
CARR ET AL.
4 © 2006 Magnolia Press
1211
ZOOTAXA well as Teleopsis dalmanni (Wiedemann), T. quinqueguttata (Walker) and Sphyracephala
beccarii (Rondani). The alignment was produced by ClustalX (Thompson et al. 1997) and
then edited by eye.
TABLE 1. List of PCR primers.
The alignment was analysed in Modeltest 3.7 (Posada and Crandall 1998), which
determined that a GTR+I+Γ model, with a four-category gamma distribution, was the most
appropriate for phylogenetic analysis. A Maximum Likelihood (ML) tree was produced
using PAUP 4.0b10 (Swofford 2002), using the parameters determined by Modeltest 3.7.
The ML tree was bootstrapped with 1000 replicates and branches with values of less than
50% were collapsed. A Bayesian phylogeny was produced using MrBayes 3.1.1 (Ronquist
& Huelsenbeck 2003), also using a GTR+I+Γ model. The MCMC analysis ran with four
chains for 1,000,000 generations. For both tree-creating methods the Diasemopsis
sequences were rooted using the Teleopsis and Sphyracephala sequences.
Diasemopsis comoroensis Carr & Földvári, new species
(Figs 1–9, 11, 13–16)
Description
Type material: Holotype, male (Natural History Museum, London). Paratypes 4
males, 5 females (Natural History Museum, London), 5 males, 5 females (Zoologische
Staatssammlung, München), 5 males, 5 females (Hungarian Natural History Museum,
Budapest), 2 males, 2 females (Centre National de Documentation et de Recherche
Scientifique, Moroni, Comoro Islands). All type specimens (dried, double mounted,
excellent condition) taken from a laboratory culture housed at University College, London
in May 2005. Parent specimens collected at Comoro Islands, Mohéli, creek uphill from
Hoani on way toward Chalet St. Antoine, leg. M. Kotrba 21.iv.2002.
Gene Sequence Coordinate or Primer Name Annealing Temperature (°C) Reference
16S 12727–12747(S)
13270–13290(A)
50.0 Baker et al. (2001)
COII A3772
S3291
50.8 Brower (1994)
w11404–11426(S)
11975–11997(A)
58.6 Baker et al. (2001)
wg 5GTTAGAACATGTTGGATGCG3
5CGTTCAACGACAATGACCTC3
53.3 Adapted from
Baker et al. (2001)
© 2006 Magnolia Press 5
DIASEMOPSIS
1211
ZOOTAXA
FIGURES 1–2: D. comoroensis internal female genital organs. 1. D. comoroensis female
reproductive tract (ovaries omitted). Scale bar 250 µm. 2. Detail of genital papilla. Arrows indicate
the separate openings of spermathecal ducts (right) and accessory gland ducts (left).
CARR ET AL.
6 © 2006 Magnolia Press
1211
ZOOTAXA
FIGURES 3–8: 3: Multichambered ventral receptacle. Arrow indicates the tubular entrance. 4:
Ventral sclerotized ring. Arrow indicates specialized epithelium. 5: Detail of ventral receptacle.
Arrows indicate coiled spermatozoa within the cuticular chambers. 6: Spermatheca with denticles.
7: Spermatheca. Arrows indicate cuticular end apparatus of epithelial gland cells. 8: Accessory
gland. Arrows indicate cuticular end apparatus of epithelial gland cells.
© 2006 Magnolia Press 7
DIASEMOPSIS
1211
ZOOTAXA
Head. Completely black (including eye stalks) and covered with minute pale hairs.
Facial teeth smaller than half of the width of the eye stalks in the middle. Outer vertical
bristles at least as long as width of the eye stalk in the middle. Inner vertical bristles short,
but distinct, as long as 1/3rd of the width of the eye stalk in the middle.
Thorax. Uniformly grayish pollinose, except the surface of the meron. Metapleural
spine dark brown to black, at most the tip can be yellow. The spine is curved upwards
(dorsally) in anterior view. Scutellar spines 2.5–3 times as long as the scutellum.
Wing Fig. 9. Completely hyaline, except for three infuscated brownish bands. The
apical band is as broad as 1/20th of the wing length and is restricted to the space between M
and R2+3 veins (not reaching these veins). The central band is darkest around R4+5 and it is
situated between R2+3 and Cu slightly extended towards the anterior cross vein (R–M). The
proximal band is more an infuscated brownish spot below the cell cup.
Legs Fig. 11. Front coxae are shiny posteriorly and also with a lateral shiny spot at
proximal 1/3rd of the coxae. Legs are yellow–brown, brown in general; tarsi 3–5 on the
front leg are paler and yellow. Front femora incrassate (length/width approximately 4),
bearing on their ventral side two longitudinal rows of 3–5 prominent bristles each and
between these two rows of 18–22 much shorter peg-like tubercles each.
Preabdomen. Subshining black except for the following silver pollinose areas: an
uninterrupted band along posterior margin of tergite 1, lateral triangles at posterior margin
of tergite 2, distal half of tergite 3, and subsequent tergites.
Postabdomen (male) Figs 13–16. In ventral view (Fig. 13) the connection of the
hypandrium to the aedeagal apodeme is clearly visible. The membranous tip of the
hypandrium is divided into two lobes anteriorly. There are two thick hairs on the medial
inner surface of the hypandrium, the bilobed surstyli have numerous short, distinct hairs
and the gonopods bear minute hairs as well (Fig. 14). In lateral view the aedeagal apodeme
is curved (more that that of D. meigenii) and not broadening on anterior half. The ligament
connecting to the hypandrium joins in the middle of the aedeagal apodeme (Fig. 15). The
epandrium and cerci have long, dispersed hairs along their surface. Hairs on the
hypandrium are more restricted to the ventral part and are shorter than those of D.
meigenii. The distal half of the paramere is broadening towards the tip (Fig. 16).
Postabdomen (female) Fig. 1. The internal female genital organs of D. comoroensis
are most similar to those of D. meigenii as described and depicted by Kotrba (1995).
The tubular vagina is surrounded by a thick layer of muscles particularly in its anterior
region. It is anteriorly connected to the common oviduct, which descends from the paired
ovaries and lateral oviducts. The posterior end of the vagina is attached to the vulva behind
sternum 8. From the ventral anterior portion of the vagina emanates the voluminous
roundish ventral receptacle, which is composed of more than 300 tubular chambers, each
with a diameter of about 15–20
µ
m (Fig. 3). In mated females these chambers house
individual tightly coiled spermatozoa (Fig. 5). The ventral receptacle is connected with the
lumen of the vagina via a narrow, tubular duct. A dense structure dorsal of the entrance
CARR ET AL.
8 © 2006 Magnolia Press
1211
ZOOTAXA may be part of a valve mechanism. Opposite the entrance of the ventral receptacle a
pouch-like dorsal evagination of the vaginal lumen receives the spermathecal ducts and,
ventrolateral of that, the ducts of the accessory glands (Fig. 2). Posterior to this a ring-
shaped ventral sclerite is embedded in the ventral wall of the vagina. Part of the vaginal
musculature inserts on this ring-shaped sclerite, thus sparing a cushion of specialized
epithelium within its centre (Fig. 4). Two spherical spermathecae are present, one with a
diameter of about 100
µ
m, the other slightly larger. Their strongly sclerotized, dark brown
capsules are ornamented with short hollow denticles (Fig. 6). Each of these denticles is
connected to the end apparatus of an epithelial gland cell (Fig. 7). The base of the
spermathecae is not telescoped but drawn out into a tubular portion which merges
smoothly with the apical ends of the long spermathecal ducts. The bases of the
spermathecal ducts are slightly sclerotized as well next to their opening into the vagina.
Like the spermathecae, the membranous ovoid reservoirs of the accessory glands are
surrounded by epithelial gland cells with cuticular end apparatuses (Fig. 8). The delicate
ducts of the accessory glands are only about 1/3rd as long as those of the spermathecae.
A description of the entire reproductive system of another stalk eyed fly, T. whitei
(Curran) (detailed under its former name Cyrtodiopsis whitei), was given by Kotrba (1993)
including further details as well as physiological and functional aspects.
FIGURES 9 & 10: D. comoroensis (9) and D. meigenii (10) wings. The larger apical band is
clearly visible in the D. meigenii wing, spanning between the M and R2+3 veins.
FIGURES 11 & 12: D. comoroensis (11) and D. meigenii (12) first coxae.
© 2006 Magnolia Press 9
DIASEMOPSIS
1211
ZOOTAXA
FIGURES 13–16: D. comoroensis male genitalia. 13: ventral view, 14: detailed ventral view with
surstyli, gonopods, cerci (no hairs on cerci drawn), 15: detailed lateral view with gonopod and
paramere, 16: lateral view with aedeagal apodeme, hypandrium and epandrium. Scale bar 0.4 mm
for Fig 13, 16, 0.2 mm for Fig 14 and 0.1 mm for Fig 15.
CARR ET AL.
10 © 2006 Magnolia Press
1211
ZOOTAXA
FIGURES 17–20: D. meigenii male genitalia. 9: ventral view, 10: detailed ventral view with
surstyli, gonopods, cerci (no hairs on cerci drawn), 11: detailed lateral view with gonopod and
paramere, 12: lateral view with aedeagal apodeme, hypandrium and epandrium. Scale bar 0.4 mm
for Fig 17, 20, 0.2 mm for Fig 18 and 0.1 mm for Fig 19.
© 2006 Magnolia Press 11
DIASEMOPSIS
1211
ZOOTAXA
Etymology: The name refers to the type locality, Comoro Islands.
Distribution: Mohéli, Comoro Islands
Diasemopsis meigenii (Westwood, 1837)
(Figs 10, 12, 17-20)
syn. Diopsis breviseta Bezzi, 1908: 167, Type locality: Ethiopia (Eritrea)
Séguy (1955) established the new genus Chaetodiopsis for D. meigenii (Westwood, 1837,
originally in Diopsis). However, Chaetodiopsis was treated as a junior synonym of
Diasemopsis by the Catalogue of the Diptera of the Afrotropical Region (Crosskey 1980)
as well as by Meier & Baker (2002), based on sound molecular and morphological
analysis. Chaetodiopsis is embedded deeply within Diasemopsis, i.e. its recognition would
render Diasemopsis polyphyletic. Here we follow the revised classification suggested by
Meier & Baker (2002).
Material studied: 5 males, 5 females taken from a laboratory culture housed at
University College, London in May 2005. This culture was derived from a culture in the
laboratory of G. S. Wilkinson, University of Maryland at College Park, which was
originally founded from flies caught near Pietermaritzburg, South Africa in December
1994 by M. Kotrba. The dried, double-mounted specimens are deposited in the Hungarian
Natural History Museum, Budapest.
Other material: 3 males, 3 females, Lourenço marques, Mozambique, Diopsis meigeni
Westw. Det Lindner. Cameroun: 2 males, 3 females, Yaounde Obili, 20–2–1963, coll. L.
Segers: 1 male, same data, 3–1–1963 (all Zoologische Staatssammlung, München); 1
female, Nkolbisson, Dept. Nyong-Sanaga, IX. [19]68, L.G. Segers leg., van Schuytbroek
det. 19 [no year given], Chaetodiopsis meigeni (Royal Museum for Central Africa).
Republic of Congo: 1 male, Mayambé: Kiniati, 7–VI–1911, R. Mayné; 1male Ikengé, IX–
1912, R. Mayné, 5 males, Congo da Lemba, R. Mayné; dates: V–1912, V–1912, I–II–
1913, III–1913, IV–1913; 1 male, Lemfu, P. Vandereijst; 1 male, Mandungu, 25–XI–1912,
R. Mayné (all Royal Museum for Central Africa).
Head. Completely covered with minute pale hairs, generally black, but the eye stalks
are brown. Facial teeth 1.5 times longer than width of eye stalks in the middle. Outer
vertical bristles at least as long as width of the eye stalk in the middle. Inner vertical
bristles minute and weak, shorter than 1/3rd of width of the eye stalk in the middle.
Thorax. Uniformly grayish pollinose, except the surface of the meron. Metapleural
spine yellow or orange yellow and straight in anterior view. Scutellar spines 2.5–3 times as
long as the scutellum.
Wing Fig. 10. Completely hyaline, except for three infuscated brownish bands. The
apical band is as broad as 1/14th of the wing length and is restricted to the space between M
and R2+3 veins (the band reaches these veins). The central band is darkest around R4+5 and
CARR ET AL.
12 © 2006 Magnolia Press
1211
ZOOTAXA it is situated between R2+3 and Cu slightly extended towards the anterior cross vein (R–M).
The proximal band is more an infuscated brownish spot below the cell cup.
Legs Fig. 12. Front coxae are only shiny on the posterior surface. Legs are yellow-
brown in general, tibiae and first two tarsi of the front leg are black, tarsi 3–5 on the front
leg are whitish yellow (clearly contrasting the other tarsi). Front femora incrassate (length/
width approximately 4), bearing on their ventral side two longitudinal rows of 2–4
prominent bristles each and between these two rows of 24–28 much shorter peg-like
tubercles each.
Preabdomen. Subshining black except for the following silver pollinose areas: an
uninterrupted band along posterior margin of tergite 1, lateral triangles at posterior margin
of tergite 2, distal half of tergite 3, and subsequent tergites.
Postabdomen (male) Figs 17–20. The hypandrium is connected to the aedeagal
apodeme and the membranous tip of the hypandrium is continuous, not divided into two
lobes anteriorly (Fig. 17). There are 3(–4) thick hairs on the medial surface of the
hypandrium, the bilobed surstyli have numerous short, distinct hairs and the gonopods
bear minute hairs (Fig. 18). In lateral view the aedeagal apodeme is more straight (that of
D. comoroensis is curved) and also broadening towards tip. The ligament connecting to the
hypandrium joins at basal 1/3rd of the aedeagal apodeme (Fig. 20). The epandrium and
cerci have long, dispersed hairs along their surface. Hairs on the hypandrium are reaching
to the lateral part and are longer than those of D. comoroensis. The distal half of the
paramere is slightly narrowing towards the tip (Fig. 19).
Distribution: Widespread in Afro-tropical regions.
Allometric differences between D. comoroensis and D. meigenii
D. comoroensis was significantly smaller than its congener for both eye span and body
length (Table 2; Wilcoxon tests; male eye span χ2 = 40.50, male body length χ2 = 28.46,
female eye span χ2= 48.67, female body length χ2 = 34.05; all d.f. = 1, all P < 0.001). Male
D. comoroensis had a significantly shallower eye span allometry than male D. meigenii
(Fig. 21; Table 2; SPECIES × BODY LENGTH interaction terms in Table 3). However, female
D. comoroensis and D. meigenii did not differ significantly in the slope of their eye span
allometry (Fig. 21; Table 2; SPECIES × BODY LENGTH interaction term in Table 4).
Within species, females had longer body lengths than males (Table 2; Wilcoxon tests;
D. comoroensis χ2 = 3.78, d.f. = 1, P = 0.05; D. meigenii χ2 = 10.81, d.f. = 1, P = 0.001).
However, males from both species had significantly larger eye span than females before
and after controlling for body length variation (Table 2; absolute eye span Wilcoxon tests:
D. comoroensis χ2 = 18.18, d.f. = 1, P < 0.001, D. meigenii χ2 = 27.20, d.f. = 1, P < 0.001;
SEX terms in Table 5 and 6). There was a small, but significant, difference in the slope of
eye span allometry between male and female D. comoroensis, with males having a slightly
© 2006 Magnolia Press 13
DIASEMOPSIS
1211
ZOOTAXA
steeper scaling relationship (Fig. 21; Table 2; SEX × BODY LENGTH interaction term in
Table 5). Eye span allometry was significantly steeper in male D. meigenii than in their
female conspecifics (Fig. 21; Table 2; SEX × BODY LENGTH interaction term in Table 6).
However, the degree of sexual dimorphism in D. comoroensis was significantly weaker
than that observed in D. meigenii (Fig. 21; Table 2; full model SPECIES × SEX × BODY
LENGTH interaction term F = 12.22, d.f. = 1,138, P < 0.001).
TABLE 2. Mean trait size (mm ± S.D.) eye span, body length and allometric slope (± S.E.) of
males and females from each species. Allometric slope is the least-squares regression coefficient of
eye span on body length.
TABLE 3. General Linear Model of body length and species effects on eye span in male
Diasemopsis.
TABLE 4. General Linear Model of body length and species effects on eye span in female
Diasemopsis.
Species Sex Eye span Body Length Allometric Slope
D. comoroensis Male
Female
5.27 ± 0.45
4.84 ± 0.32
6.58 ± 0.34
6.70 ± 0.38
1.70 ± 0.07
0.82 ± 0.03
D. meigenii Male
Female
7.19 ± 0.96
6.10 ± 0.52
7.31 ± 0.55
7.62 ± 0.62
1.21 ± 0.08
0.80 ± 0.04
Factor SS d.f. MS F-ratio Prob > F
BODY LENGTH 97.588 1 97.588 2231.58 <0.001
SPECIES 6.098 1 6.098 139.44 <0.001
SPECIES × BODY 0.658 1 0.658 15.06 <0.001
LENGTH
ERROR 2.974 68 0.0437
Total 107.319 71
Factor SS d.f. MS F-ratio Prob > F
BODY LENGTH 39.406 1 39.406 2989.54 <0.001
SPECIES 2.699 1 2.699 204.82 <0.001
SPECIES × BODY 0.002 1 0.002 0.13 0.72
LENGTH
ERROR 0.993 70 0.0132
Total 43.031 73
CARR ET AL.
14 © 2006 Magnolia Press
1211
ZOOTAXA TABLE 5. General Linear Model of body length and sex effects on eye span in D. comoroensis.
TABLE 6. General Linear Model of body length and sex effects on eye span in D. meigenii.
FIGURE 21: Eye span allometry in D. comoroensis and D. meigenii .
= male D. comoroensis, +
= female D. comoroensis,
= male D. meigenii,
= female D. meigenii. Least-squares
regression lines are given for heuristic purposes.
Factor SS d.f. MS F-ratio Prob > F
BODY LENGTH 6.263 1 6.263 376.70 <0.001
SEX 4.967 1 4.967 298.73 <0.001
SEX × BODY 0.355 1 0.355 21.34 <0.001
LENGTH
ERROR 1.048 63 0.0166
Total 12.633 66
Factor SS d.f. MS F-ratio Prob > F
BODY LENGTH 23.292 1 23.292 613.20 <0.001
SEX 39.254 1 39.254 1033.42 <0.001
SEX × BODY 5.059 1 5.059 133.18 <0.001
LENGTH
ERROR 2.849 75 0.038
Total 90.585 78
© 2006 Magnolia Press 15
DIASEMOPSIS
1211
ZOOTAXA
Gene Sequencing and Phylogenetic Analysis
BLAST searching showed that sequences from D. meigenii had the highest similarity to
the four sequenced partial D. comoroensis genes. Nucleotide identities between the four
genes in the two species ranged from 96.7–99.8%, with both nuclear genes sharing 99.8%
identity within their coding sequences, showing that the two species must share a very
recent common ancestry.
FIGURE 22: Bayesian phylogeny of Diasemopsis species, rooted with Teleopsi s and
Sphyracephala species. Posterior probabilities are given for each branch.
CARR ET AL.
16 © 2006 Magnolia Press
1211
ZOOTAXA In both the Bayesian tree (Fig. 22) and ML tree the D. meigenii and D. comoroensis
sequences clustered together on a long branch, deeply embedded within the Diasemopsis
genus (posterior probability = 1.00 and bootstrap support = 100%). Phylogenies for each
of the individual genes all show this close relationship between the two species (data not
shown). Thus all four sequenced genes, using two different tree-creating methods, show a
very close relationship between D. comoroensis and D. meigenii. Despite the lower
number of taxa and gene sequences used, the phylogenies created were otherwise identical
to those of Baker et al. (2001) and Meier & Baker (2002), indicating that this is a robust
phylogeny of the Diasemopsis genus.
Diagnosis for the two species
External differences: D. comoroensis has a shorter eye span compared to the body length,
exhibiting only mild sexual dimorphism (large eye span, strong dimorphism in D.
meigenii). The facial teeth are shorter (long in D. meigenii) and the inner vertical bristles
are more distinct (only minute in D. meigenii). The first coxae have shiny spots on the
lateral surface at the basal 1/3 rd of the coxae (missing in D. meigenii). Metapleural spines
are dark and in anterior view upcurved (yellow and straight in D. meigenii).
Differences in the male genitalia: D. comoroensis has the aedeagal apodeme more
curved in lateral view and also the width is even along its length (straight and broadening
in D. meigenii). The linking structure from the hypandrium attaches the aedeagal apodeme
in the middle (at basal 1/3rd in D. meigenii Figs 16, 20) and the parameres are broadening
towards tip (slightly narrowing in D. meigenii Figs 15, 19). The hypandrium is bilobed
anteriorly in D. comoroensis.
Comments
Both morphological similarity and a phylogenetic approach have placed D. comoroensis
as a sister species to D. meigenii. The species are sufficiently closely related for inter-
specific mating to occur under forced laboratory conditions. However such mating
attempts do not produce any offspring (Cotton & Carr, unpublished data).
As is the case with a number of species in the current Diasemopsis molecular
phylogeny, D. meigenii and D. comoroensis are placed at the end of a long branch. It is
unknown what proportion of Diasemopsis species have been scientifically described,
however relatively few described species have been characterised from a molecular
perspective; therefore the isolated position of D. meigenii and D. comoroensis within the
tree is likely to be a result of missing taxa. Greater sampling of Diasemopsis, both on
mainland Africa and the Indian Ocean islands, may add additional undescribed taxa to the
© 2006 Magnolia Press 17
DIASEMOPSIS
1211
ZOOTAXA
known 51 Diasemopsis species. The current tree topology is ladder-like, it is likely that
further gene sequences will provide greater resolution to the Diasemopsis phylogeny and
produce a more bush-like topology. Preliminary studies of collection specimens of other
Diasemopsis species suggest there will be numerous new species to be described which
may have a strong influence on the tree.
D. comoroensis and D. meigenii exhibit very different head morphologies in terms of
the absolute difference between the two species, and within-species differences (i.e. sexual
dimorphism). D. comoroensis is mildly sexually dimorphic for eye span, both before and
after controlling for body length variation, with males having a larger trait. Male eye span
allometry is also slightly steeper than that of females in D. comoroensis. However, it is
significantly less steep (in both absolute terms and relative to females) than the eye span
allometry of male D. meigenii. It is noteworthy that the extent of dimorphism in D.
comoroensis may be partially dependent on the estimate of body size, as the use of an
alternative body size proxy (thorax length) produces an outcome of no significant sexual
dimorphism in this species (Cotton, unpublished data; sexual dimorphism remains strong
in D. meigenii with this additional measure).
The nature of selection on eye span in D. comoroensis is difficult to discern without
further study, but differences between male and female eye span may have resulted from
current or past episodes of sexual selection; all sexually-selected diopsids studied thus far
exhibit an elevated male eye span allometry (Burkhardt & de la Motte 1985; Wilkinson &
Dodson 1997; Baker & Wilkinson 2001), as predicted if eye span is an evolved signal of
body size, or correlate thereof (Green 1992; Petrie 1992, Cotton et al. 2004). For example,
in D. meigenii, there is strong sexual selection through female mate preference for large
male eye span (Cotton et al. 2006). In contrast, the slope of eye span allometries is much
shallower in, and does not differ between, female D. comoroensis and female D. meigenii,
suggesting that females from both species share similar eye span scaling optima that are
less sensitive to variation in body size.
All species from the genus Diasemopsis examined in Baker & Wilkinson (2001)
exhibited some degree of sexual dimorphism with respect to their eye span. It is
demonstrated here that D. comoroensis exhibits only very mild dimorphism, relative to its
sister species D. meigenii. As these species only recently diverged the observed
differences in eye span sexual dimorphism must have evolved rapidly. Loss of sexually
selected traits is an increasingly well-observed phenomenon (reviewed in Wiens 2001) and
has previously been recorded in an island population (Griffith et al. 1999). The colonising
of the Comoro Islands is likely to have resulted in new environmental challenges, such as
resource availability, novel predators, a reduced population size and isolation from the
ancestral population, any of which may have altered the selection pressure on increased
male eye span.
CARR ET AL.
18 © 2006 Magnolia Press
1211
ZOOTAXA Acknowledgements
Many thanks go to Andrew Pomiankowski, Kevin Fowler and Sandie Baldauf for
providing laboratory facilities during the completion of this work. Thanks also go to Marc
De Meyer (Tervuren, Belgium) for the loan of comparative material of D. meigenii. We
also thank the Ministre de la production et de lenvironnement of the Comores for kind
support and permission to collect these insects (Ref: 02–140/MPE/DGE) as well as Frank
Glaw and Michael Hiermayer for strong and reliable cooperation during the collecting trip
to the Comores in 2002.
References
Baker, R.H. & Wilkinson, G. (2001) Phylogenetic analysis of sexual dimorphism and eye-span
allometry in stalk-eyed flies (Diopsidae). Evolution, 55, 1373–1385.
Baker, R.H., Wilkinson, G. & DeSalle, R. (2001) Phylogenetic utility of different types of molecular
data used to infer evolutionary relationships among stalk-eyed flies (Diopsidae). Systematic
Biology, 50, 87–105.
Bezzi, M. (1908) Ditteri eritrei raccolti dal Dott. Andreini e dal Prof Tellini. Parte seconda. Diptera
cyclorrhapha. Bolletino della Societa Entomologica Italiana, 39(1907), 3–199.
Brower, A.V.Z. (1994) Phylogeny of Heliconius butterflies inferred from mitochondrial DNA
sequences (Lepidoptera: Nymphalidae). Molecular Phylogenetics And Evolution, 3, 159–174.
Burkhardt, D. & de la Motte, I. (1985). Selective pressures, variability, and sexual dimorphism in
stalk-eyed flies (Diopsidae). Naturwissenschaften, 72, 204–206.
Cotton, S., Fowler, K. & Pomiankowski, A. (2004). Condition-dependence of sexual ornament size
and variation in the stalk-eyed fly Cyrtodiopsis dalmanni (Diptera: Diopsidae). Evolution, 58,
1038–1046.
Cotton, S., Rogers, D.W., Small, J., Pomiankowski A. & Fowler, K. (2006) Variation in preference
for male ornaments is associated with female eyespan in the stalk-eyed fly Diasemopsis meige-
nii. Proceedings of the Royal Society. B, published online, doi:10.1098/rspb.2005.3449.
Crosskey, R.W., ed. (1980) Catalogue of the Diptera of the Afrotropical Region. British Museum
(Natural History), London. 1437 pp.
David, P., Hingle, A., Greig, D., Rutherford, A., Pomiankowski, A. & Fowler, K. (1998). Male sex-
ual ornament size but not asymmetry reflects condition in stalk-eyed flies. Proceedings of the
Royal Society. B, 265, 2211–2216.
Green, A.J. (1992). Positive allometry is likely with mate choice, competitive display and other
functions. Animal Behaviour, 43, 170–172.
Griffith, S.C., Owens, I.PF. & Burke, T. (1999) Female choice and annual reproductive success
favour less-ornamented male house sparrows. Proceedings of the Royal Society. B, 266, 765–
770.
Kotrba, M. (1993) Das Reproduktionssystem von Cyrtodiopsis whitei Curran 1936 (Diopsidae,
Diptera) unter besonderer Berücksichtigung der inneren weiblichen Geschlechtsorgane. Bon-
ner Zoologische Monographien, 33, 115 pp.
Kotrba, M. (1995) The internal female genital organs of Chaetodiopsis and Diasemopsis (Diptera:
Diopsidae) and their systematic relevance. Annals of the Natal Museum, 36, 147–159.
Kotrba, M. and Balke, M. (2006) The systematic position of Cladiopsis Séguy, 1949, and the origin
of sexual dimorphism in stalk-eyed flies (Diptera: Diopsidae) inferred from DNA sequence
data. Molecular Phylogenetics and Evolution, published online, doi: 10.1016/
© 2006 Magnolia Press 19
DIASEMOPSIS
1211
ZOOTAXA
j.ympev.2005.11.009.
Meier, R. & Baker, R.H. (2002) A cladistic analysis of Diopsidae (Diptera) based on morphologi-
cal and DNA sequence data. Insect Systematics and Evolution, 33, 325–336.
Petrie, M. (1992). Are all secondary sexual display structures positively allometric and, if so, why?
Animal Behaviour, 43, 173–175.
Posada, D. & Crandall, K.A. (1998) Modeltest: testing the model of DNA substitution. Bioinfor-
matics, 9, 817–818.
Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference under mixed
models. Bioinformatics B, 19, 1572–1574.
Séguy, E. (1955) Dipteres Diopsidae africains nouveaux ou peu connus. Bulletin de l
>
Institut
francais de l
>
Afrique noire, 17, 1102–1124.
Shilito, J.F. (1971) The genera of Diopsidae (Insecta, Diptera). Zoological Journal of the Linnean
Society, 50, 287–295.
Swofford, D.L. (2002) PAUP*. Phylogenetic analysis using parsimony (*and other methods). Ver-
sion 4. Sinauer Associates, Sunderland, Massachusetts, USA.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. & Higgins, D.G. (1997) The
CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by
quality analysis tools. Nucleic Acids Research, 25, 4876–4882.
Westwood, J. O. (1837) Description of some new species of Diopsis. Transactions of the Linnean
Society of London, 17, 543–550.
Wiens, J.J. (2001) Widespread loss of sexually selected traits: how the peacock lost its spots. Trends
in Ecology and Evolution, 16, 517–523.
Wilkinson, G.S. & Dodson G.N. (1997) Function and evolution of antlers and eye stalks in flies. In:
Choe, J. & Crespi, B. (Eds.), The Evolution of Mating Systems in Insects and Arachnids. Cam-
bridge University Press, Cambridge, pp 310–328.
