The male sexual apparatus in the order Scorpiones

48 Pages·2017·14.69 MB·English

by  Lionel Monod

Monodetal.FrontiersinZoology (2017) 14:51
DOI10.1186/s12983-017-0231-z
RESEARCH Open Access
The male sexual apparatus in the order
Scorpiones (Arachnida): a comparative
study of functional morphology as a tool to
define hypotheses of homology
Lionel Monod1*, Lucie Cauwet1, Edmundo González-Santillán2 and Siegfried Huber1
Abstract
Background: Inseminationinscorpionsiscarriedoutbymeansofapartlysclerotizedstructure,thespermatophore,
whichiscomposedoftwoseparatehalves,thehemispermatophores.Inmostgenerathesereproductivestructures
canbeusedtodifferentiatespecies.However,manytaxasuchasthegenusEuscorpiusandthefamilyDiplocentridae
lackthemorphologicaldiversityobservedinthecopulatoryorgansofmanyotherarthropods,renderingthemuseless
forspeciesleveltaxonomy.Suchstructuralstasis,however,suggeststhathemispermatophoreshaveevolvedrelatively
slowlyandmaythusprovideastrongerphylogeneticsignalforrecognizingsupra-genericranksthanpreviously
thought.Basedonthepostulatethatthephenotypicstabilityobservedinsomegroupsistheconsequenceof
functionalconstraint,themostcomprehensivecomparativestudyofthemalesexualapparatustodatewas
conductedforacompletereassessmentofthemorphology,phylogeneticvalueandhypothesesofhomologyof
thesestructures.
Results:Hemispermatophores,pre-andpost-inseminationspermatophores,aswellastheinherentmechanismsof
insemination,werestudiedacrossthewholeorder,allowingtherecognitionanddescriptionofaseriesoffivebasic
bauplansforthecapsularregion.Forthemostpart,thesepatternsappeartobeconsistentwithineachmajor
taxonomicgroup,butseveralcasesofincongruencebetweenspermatophoremorphologyandtaxonomyraises
questionsaboutthemonophylyofsomeclades.TheBothriuridaearetraditionallyregardedasabasalscorpionoid
family.However,exceptforthegenusLisposoma,bothriuridhemispermatophoresandspermatophoresare
morphologicallymoresimilartothoseoftheChactoideathantothoseofscorpionoids.Ontheotherhand,themale
copulatorystructuresofthehormuridclade(Hormiops(Hormurus+Liocheles))aremoreakintothoseof
DiplocentridaeandHeteroscorpionidaethantothoseofotherhormurids.
Conclusions:Spermatophorecapsularpatternsappearstobecongruentwitharecentphylogenyoftheorder
ScorpionesbasedonphylogenomicdatathatplacedBothriuridaeoutsideofScorpionoideaandLiochelesoutsideof
Hormuridae,incontradictonwithearlierphylogeneticreconstructionsbasedonmorphology.Thisraisesquestions
aboutthepotentialuseoffunctionallyconstrainedtraitstoassessthereliabilityofcontradictingphylogenetic
hypothesesandemphasizestheneedforathoroughreassessmentofthescorpionphylogeneticrelationships.
Keywords:Hemispermatophore,Spermatophore,Insemination,Bauplan,Naturalselection,Sexualselection,
Functionalconstraints,Phylogeneticvalue
*Correspondence:[email protected]
1Départementdesarthropodesetd’entomologieI,Muséumd’histoire
naturelle,RoutedeMalagnou1,1208Genève,Switzerland
Fulllistofauthorinformationisavailableattheendofthearticle
©TheAuthor(s).2017OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.0
InternationalLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinkto
theCreativeCommonslicense,andindicateifchangesweremade.TheCreativeCommonsPublicDomainDedicationwaiver
(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.
Monodetal.FrontiersinZoology (2017) 14:51 Page2of48
Background and the stalk and guides semen from the spermato-
Morphologyofthescorpionreproductiveapparatus phore cavity to the female genital tract.
In scorpions, sperm is not transferred directly by a sex- This last part, the capsule, is responsible for the sperm
ual organ, but using a spermatophore, an external, partly transfer per se. It is composed of two sclerotized ridges/
sclerotized structure containing the spermatozoa, which carinae (distal and basal) delimiting a soft membrane, the
is produced and deposited on the ground by the male “spermduct”(sensuStockwell[4]).Thecapsuleshowsthe
[1]. This spermatophore is composed of two halves, the greateststructuralvariationinthespermatophore,ranging
hemispermatophores, which are secreted by the paraxial fromsimpleopeningstocomplexeversiblestructures.
organs [2] and joined together when they are expelled
from thebody duringcourtship. Hemispermatophoresastaxonomiccharacters
The spermatophore is composed of four distinct Vachon [5] was the first to recognize the taxonomic
parts (Fig. 1). Although their shapes and proportions valueof hemispermatophoreandtouseitasadiagnostic
can be extremely variable, each part has the same character in the families Buthidae Koch, 1837 and Scor-
function in all taxa: (1) the pedicel is used to glue pionidae Latreille, 1802. In the sixties and seventies, San
the spermatophore to the substrate, (2) the stem is Martin [6–9], Maury [10–17] and Cekalovich [18–22]
composed of soft membranes and contains the sperm- emphasized its importance for the systematics of
atozoa, (3) the stalk (= flagellum and distal lamina Bothriuridae Simon, 1880. It was only in the late seven-
sensu Lamoral [3]) is usually sclerotized and acts as a ties/eighties that hemispermatophores started to be
lever to trigger the compression of the stem in most treated more widely as taxonomically informative charac-
taxa, and (4) the capsule is located between the stem ters in others families: initially in Hormuridae Laurie,
a b c d
Chaerilidae Buthidae Scorpiopidae Hormuridae
distal carina
GEND dstepisremtarmiln c adal urmicnetambrane GEND tseucbtuexm
E E terminal membrane
L basal carina L
basa carina
Fig.1Hemispermatophoreintoto,lateralaspect,groundplan.Thefourpartsofthestructure(stalk,stem,pedicelandcapsule)areindicated
foreachofthefourdistantlyrelatedtaxa,emphasizingtheconservedpatterndespitethevariableproportions.a,ChaerilidaePocock,1893.
b,ButhidaeKoch,1837.c,ChactidaePocock,1893.d,HormuridaeLaurie,1896.Scales,0.5mm(a),1mm(c),2mm(d),3mm(b)
Monodetal.FrontiersinZoology (2017) 14:51 Page3of48
1896 [3, 4, 23–28] and Urodacidae Pocock, 1893 [4, 23], diagnose species, but is also considered to have evolved
theninCaraboctonidaeKraepelin,1905[4,29],Chactidae too rapidly to be phylogenetically informative [51, 52].
Pocock, 1893 [4, 29, 30], Iuridae Thorell, 1876 [29], However, numerous phylogenetic studies of various
Vaejovidae Thorell, 1876 [4, 29, 31–33], Superstitioniidae arthropod lineages have successfully included male geni-
Stahnke, 1940 [34], Typhlochactidae Mitchell, 1971 [34], talic characters, demonstrating that reproductive struc-
Heteroscorpionidae Kraepelin, 1905 [26, 35, 36] and tures may also be useful in determining supra-specific
Diplocentridae Karsch, 1880 [4, 37, 38]. The morphology phylogenetic relationships[53].
of hemispermatophores from several families became Scorpion hemispermatophores can be divided into dif-
knownonlyrecently:in1989forChaerilidaePocock,1893 ferent ‘subunits’ that are arguably subject to different se-
[4], Euscorpiidae Laurie, 1896 [3, 4, 39] and Scorpiopidae lective pressures and thus show different patterns of
Kraepelin, 1905 [4]; in 2001 for Troglotayosicidae [40]; in evolution and therefore various degrees of phylogenetic
2005 for Hemiscorpiidae Pocock, 1893 [41]; in 2006 for value [54–56]. While some characters follow the general
Pseudochatidae Gromov, 1998 [42]. Nevertheless, despite trend of rapid and divergent evolution driven by strong
thegeneralized useof hemispermatophoreas adiagnostic sexual selection observed in most male genitalia, others
character in taxonomic descriptions since the beginning show only limited variation among closely related taxa,
of the 21rst Century, our knowledge for a few groups, i.e. making them more suitable for the reconstruction of the
Chactidae,Pseudochactidae,Typhlochactidae,Troglotayo- deeperphylogeneticrelationships[56].
sicidaeandScorpiopidae,stillremainsfragmentary. Mattoni et al. [56] pointed out that the traits of sper-
Although hemispermatophores are usually distinct for matophores, which are mechanically constrained to per-
closely related species, their morphology can also be ex- form a particular task, would be less likely to evolve
tremely uniform insome genera([4,43,44];Monod, un- rapidly and randomly than, for instance, features subject
published data) or even in some families [4], conveying to sexual selection. Functional constraint limits the mor-
little to no taxonomic information at the species level. phological variability, and when a structure is involved
Such structural stasis suggests that hemispermatophores in performing a precise mechanistic task, its shape can-
have evolved relatively slowly at least in some groups not be quickly and radically modified without causing
and may thus potentially provide a strong phylogenetic severefunctionaldisruption[57–60].
signal athigher taxonomic levels. Based on this reasoning, slowly evolving mechanically
constrained characters are potentially the most inform-
Hypothesisofhomologyinhemispermatophores ative for reconstructing deeper phylogenetic nodes. It is
Hemispermatophores have very rarely been considered therefore necessary to consider not only the morphology
in a more global evolutionary perspective because their of hemispermatophores but also their functional aspect
study is usually purely descriptive. Lamoral [3] was the in order to correctly assess homology. In this respect,
first to propose structural homologies among the hemi- the study of the mechanisms of insemination by com-
spermatophores of Buthidae and Scorpionidae. He con- parison of pre- and post-mating spermatophores over a
sidered that the flagellum of buthids was an extension of widerangeoftaxaisofparamountimportanceinunder-
the capsule and was therefore not homologous to the standing how the different features that compose the
distal lamina of scorpionids. Stockwell [4] disagreed with male reproductive apparatus relate to each other in
this hypothesis, pointing out the very limited sampling distant taxa.
of intermediate taxa examined by his predecessor and
presented a more congruent hypothesis of homology Post-CopulatorySpermatophoresinSystematics
based on comparison of a wider range of taxonomic Despite their putative usefulness for systematics [54],
groups. His work is thus far the only available study giv- pre- and post-copulatory spermatophores have been
ing a comparative assessment of the morphology of treated relatively marginally in the scorpion literature
hemispermatophores across the whole order. Several [1, 2, 4, 61–94], and only few studies present an ana-
phylogenetic analyses [45–50] subsequently used Stock- lysis of functional morphology and the mechanisms
well’s work to define characters and character states of involved. To our knowledge, spermatophore morph-
hemispermatophores for their respective matrices, but ology has never been compared in detail across dis-
each of these studies focused on infraordinal taxonomic tantly related taxa and used in a phylogenetic study
groupsratherthan theorderasawhole. in the way that it has been for the order Amblypygi
Thorell, 1883 [95, 96]. The limited data available on
Phylogeneticvaluesandfunctionalmorphologyof spermatophores is probably a consequence of the dif-
Hemispermatophores ficulty in obtaining these structures, necessitating as
Due to the high phenotypic dissimilarity seen in many it does the availability of living adult animals of both
organisms, the male sexual apparatus is widely used to sexes and the observation of successful matings. This
Monodetal.FrontiersinZoology (2017) 14:51 Page4of48
gap in our knowledge of the reproductive structures in sheet of cork bark on the bottom. In order to create a
male scorpions has probably contributed to the difficulty humidity gradient the cork bark was moistened on one
ofestablishingreliablehypothesesofhomology. side with a plant mister. When mating occurred, sper-
matophores were retrieved from the enclosure immedi-
Achievementsofthepresentstudy ately after the pair separated and placed into 75%
We present here the most comprehensive comparative ethanol. In order to obtain pre-insemination spermato-
study of the scorpion male reproductive apparatus, not phores, the pairs were separated after the male had de-
limited to morphology of hemispermatophores but also posited the spermatophore on the substrate and the
including comparison of spermatophore morphology, an unusedspermatophorewasthenplacedin75%ethanol.
analysis of the mechanisms of insemination and of the
function of the various structures. The largest database Dissectionandexamination
todatewasgatheredbytheexamination ofmuseumma- Spermatophores and hemispermatophores were exam-
terial but also by searching the literature for all pub- ined with a ZEISS Stemi SV8 stereomicroscope. Mature
lished illustrations of hemi- and spermatophores. male specimens were dissected using microsurgical scis-
Moreover, we doubled the data available for pre- and sors and forceps for extracting both of their hemisper-
post-copulatory spermatophores and present the first matophores. Paraxial organ tissue was removed either
comparative study of these structures over a range of manually with forceps or chemically with a solution of
distantly related taxa. This allow us (1) to elaborate a Proteinase K (concentration: 10 mg/mL; Qiagen, Venlo,
more consensual terminology applicable to the whole The Netherlands). Chemical extraction was performed
order, (2) to identify characters most suitable to provide by immersing hemispermatophores in the solution and
strong phylogenetic signal at supra-specific phylogenetic then placing them in an oven at 45–50 °C for 15 min to
levels, (3) to reassess homology hypotheses for hemi- an hour depending on the size and sclerotization of the
spermatophores, with an emphasison the capsule. Based structure. Once the soft tissue of the paraxial organ was
on these new paradigms, we identify and describe five sufficiently digested, the hemispermatophores were re-
general patterns (bauplans) for the capsular region and trieved from the solution and thoroughly rinsed with
map them onto available phylogenetic reconstructions in water.
order to uncover inherent evolutionary trends. The roles Depending on the taxa, some parts of the hemisper-
of natural and sexual selection in the evolution of scor- matophore capsular region are composed of soft mem-
pion copulatory structures are then evaluated. Finally, branes that can be easily damaged during dissection
conflicts between morphological and molecular phyloge- whether chemical or manual. The thinner membranes
nies and the reliabilty of each hypothesis are also dis- can be degraded by long exposure to Proteinase K, or
cussed based on congruence between these hypotheses torn apart by strong traction of the forceps, especially
andthereproductive morphology. after prolonged storage in alcohol which can stiffen
them. In fresh specimens, these membranes remain flex-
Methods ible and are less likely to be damaged during dissection.
Acquisitionofspermatophores Therefore, whenever possible the dissection was carried
Live specimens were purchased from the pet trade, do- outimmediatelyafterthespecimenwaseuthanized.
nated by colleagues or collected in the field in compli- However, most of the time fresh specimens were not
ance with the legal requirements of each country. available. In those cases, the size and state of preserva-
Spermatophores were obtainedfrom breeding specimens tion of the specimen were evaluated in order to deter-
during field expeditions or subsequently in the labora- mine the best methods of extraction to minimize the
tory. Specimens were kept in captivity at temperatures risk of damaging the structure. Manual extraction was
between 25 and 30 °C, without seasonal changes. Each performed with relatively good success on large taxa,
specimen was housed individually in a plastic container even those preserved for a long time, while chemical ex-
provided with a mixture of bark mulch and peat as sub- traction was preferred for smaller taxa. On several occa-
strate and pieces of cork bark for hiding. Each specimen sions, several specimens of the same species were
was fed 2–4 crickets (Acheta domestica (Linnaeus, dissected in order to obtain a hemispermatophore show-
1758), Gryllomorpha dalmatina (Ocskay, 1832), Gryllus ingthe completeset ofinformative characters.
assimilis (Fabricius, 1775) or Gryllus bimaculatus De
Geer, 1773) or 1–2 cockroaches (Blaptica dubia Serville, Evaluationofinseminationmechanism
1838) every two weeks. Each enclosure was watered The position of a spermatophore during mating was
once a week or every two weeks depending on the mois- inferred by placing the spermatophore on a preserved
tureneedofeachtaxon.Formating,afemaleandamale female of the corresponding species, following the
were placed in a larger plastic container with a thin method used by Jacob et al. [90]. For taxa in which
Monodetal.FrontiersinZoology (2017) 14:51 Page5of48
the spermatophore remains attached to the female prepared in order to allow a direct comparison of the
gonopore for some time after copulation, i.e. the genus two structures, which we considered necessary to under-
Chiromachus, and the sub-genera Monodopistha- stand theprocessofcapsulareversion.
canthus Lourenço, 2001 and Nepabellus Francke, 1974
(see corresponding paragraph for details), the position Scanningelectronmicroscopy
of the structure could be ascertained in situ directly Spermatophores of Hormiops davidovi Fage, 1933
after the mating occurred. preservedin 75% ethanol weredehydratedin a gradedal-
Assessment of the insemination processes also in- coholseries,criticalpointdriedinaSPI-DRYcriticalpoint
volved a precise localization of the capsular foramen, the dryer (SPI supplies, West Chester, PA, U.S.A.), mounted
opening through which the semen is expelled from the on standard aluminium stubs (diameter 12.5 mm, height
spermatophore. It is clearly visible under a microscope 6 mm; Agar Scientific, Essex, U.K.), and sputter-coated
andalwaysassociated withthepresenceofsperm. with gold in a Cressington Sputtercoater 108 Auto. The
samplewasexaminedwithaZeissDSM940ASEM.
Photographsandillustrations
Line drawings were produced using a drawing tube Terminologyofpositionalaspectofhemispermatophores
mounted on the SV8 stereomicroscope. Pencil sketches Lamoral [3] proposed a terminology to identify the dif-
were subsequently inked and scanned for further pro- ferent orientational aspects of the hemispermatophore
cessing and editing. High-resolution images were taken based on the anatomical position of the structure inside
with a custom-built stacking system and with a DSLR the body prior to dissection. Stockwell [4], on the other
camera equiped with a VariMag II DSLR microscope hand, used the position of the functional spermatophore
adapter system (CNC Supply, cape Coral, FL, U.S.A.) in relation to the position of the male during copulation
and mounted on a stereomicrocope. Zerene Stacker as the reference system. Although most researchers
(Zerene Systems, Richland, WA, U.S.A.) was used to subsequently used Lamoral’s terminology to describe
assemble images taken at different focal planes into a hemispermatophores, Stockwell’s alternative is more
single image with greater depth of field. Illustrations and appropriate, as pointed out by Cauwet [97]. Depending
photographs were edited (background removal and on the taxonomic group, hemispermatophore capsules
contrast adjustment) in Adobe Photoshop CS5 and may not have the same orientation inside the male body,
plates prepared with Adobe illustrator CS5 (both from and its orientation is, moreover, difficult to assess with
Adobesystems,San Jose,CA,U.S.A.). accuracy without completely removing the tergites. On
Line drawings are preferred over photographs and the other hand, the orientation of deposited spermato-
SEM to illustrate hemispermatophores for the following phores is always the same across the order, with the cap-
reasons: (1) the small size and transparency of the struc- sule and sperm duct always facing away from the male
ture sometimes results in photographs uninterpretable towards the female gonopore. Furthermore, contrary to
in three dimensions, whereas drawing with a drawing Lamoral’s hemispermatophore terminology, Stockwell’s
tube enables a precise interpretation of the complex can be applied to both hemispermatophores and sper-
shape of the capsule; (2) drying a hemispermatophore matophores. A second terminology for spermatophores
for SEM often causes shrivelling or deformation, even is therefore not needed and direct comparison between
when using critical point drying. For spermatophores, the two structures is straightforward. The positional
photographs were deemed appropriate because the terminology of Stockwell [4] adapted by Cauwet [97]
transparency is less problematic than for hemispermato- (Fig. 2)isthususedinthepresent contribution.
phores (spermatophores are filled with semen), and they
are preferred over SEM because of the paucity of mater- Nomenclatureofhemispermatophores
ialavailable. The tendency among scorpion taxonomists to use differ-
Rather than illustrating all hemispermatophores exam- ent nomenclatural systems, adapted to the particular
ined, major patterns were identified and representative group studied but not necessarily to other taxa, has
taxa were selected. As explained above, it is difficult and arguably hampered the establishment of a consensual
time-consuming to obtain spermatophores, and each terminology based on reliable interpretations of hom-
specimen provides crucial information that can poten- ology [97]. The terminology used in the present paper is
tially facilitate future work. Illustrations of spermato- thus adapted from several nomenclatural systems, i.e.
phores from most of the genera for which the structure Vachon [5], Alexander [62, 63], San Martin [7], Lamoral
was previously unknown are therefore presented, even [3], Francke [83], Stockwell [4], Peretti [85], Monod &
when they belong to closely related taxa. When a hemi- Volschenk [98], Monod & Lourenço [41] and Monod
spermatophore and a spermatophore were available fora [49]. Whenever possible, we have adopted the first term
given taxon, a plate representing both structures was proposed. In some cases, however, older terms were
Monodetal.FrontiersinZoology (2017) 14:51 Page6of48
a
LEGEND
distal carina
tectum
subex
terminal membrane
basal carina
b c d
Fig.2Positionalnomenclatureofthemalereproductivestructure.a,diagrammaticrepresentationofthepost-inseminationspermatophoreof
Liochelescf.australasiaeFabricius,1775(Thailand,RoiEtProvince,MHNG),lateralaspect,indicatingitspositionrelativetomaleandfemaleduring
copulation.b,diagrammaticrepresentationofthepre-inseminationspermatophoreHeterometrusmysorensisKovařík,2004(India,MHNG),lateral
aspect.c-d,diagrammaticrepresentationofthelefthemispermatophoreofHeterometrusindus(DeGeer,1778)(SriLanka,MHNG),contra-lateral
(c)andanterior(d)aspects.Abbreviations:A(anterior),B(basal),D(distal),L(lateral),CL(contra-lateral)andP(posterior)
considered less appropriate than more recent ones, lamelliformhemispermatophores.Distallaminais,however,
either because they define the character concerned less stillusedinthepresentpublicationtodesignatethepartof
accurately, or because a more recent name has gained thestalk, distal to thetransverse ridge,which usuallybears
common acceptance among researchers. Furthermore, a laminar hook, in lamelliform hemispermatophore. Lame
several new names are introduced, some to designate distale [5]; Flagelle (pars recta + pars reflexa) (Buthidae)
newly recognized structures or parts, and others to des- [5];Posteriorprocess+blade(Buthidae)[99];Stalk[62,63];
ignate features alreadydescribed but forwhichall names Flagellum (Buthidae) [4, 68, 83]; Blade [69]; Flagella [64];
available were considered inappropriate. For instance, Láminadistal[6,7,9,18,76,77,100];Distallamina[3,49,
numerous parts of hemispermatophore capsule have 90]; Flagellum (Pars recta + pars reflecta + parsbireflecta)
been referred to as lobes. This vague wording does not [3];Lamina(Lamellaintext)[83];Distalpart[101];Distal
properly describe either morphology or function and lamella[4];Flagelo (Buthidae) [84];Lámina[85];Lamina
wasthussystematically replaced. [30,86,102];Flagellum(Pseudochactidae)[42].
All the terms used in this publication and their syn-
onymies with earlier nomenclatural systems are listed Latero-distalcrest(Bothriuridae):Cresta[7,9,76,77,
below. The synonyms are listed in chronological order 85];Crest[86];Distalcrestoflamina[102].
with their references and a definition for each new name Antero-distal crestofdistallamina:Lateralcrest
is provided. The list is divided into three sections corre- [3];distal crestofdistallamina[49].
sponding to the main parts of the hemispermatophore Laminar hook:Lobeinterne[5];Chitinous hook [63];
(Fig. 1),i.e.the stalk,the capsule andthestem. Lóbulointerno[18];Protuberancia espiniforme
Stalk: The term proposed by Alexander [62] is retained [76];Lóbulodistal [9,77];Lobulacióndorsales
here because it can be applied to both flagelliform and esclerificadasdellobulióinterno[100];Hook
Monodetal.FrontiersinZoology (2017) 14:51 Page7of48
[3,41,49,98];Distallobe[101];Dorsalfold[86]; Subex+tectum:Lobemédian (Buthidae)[5];Inferior
Basallobe+Outerdistallobe+Innerdistallobe outerprocess(Buthidae)[99];Medial lobe
[90];Dorsalapophysis[30]. (Buthidae)[68];Laterallobe[69];Membrane
Transverseridge:Distalcrestofmedianlobe+Ectal [90];Trough[30].
crestofmedianlobe[3];Suturatronco-laminar Terminalmembraneofsperm duct:Membranethat
[85];Laminar-trunksuture[86];Transverseridge surroundstheforamen,probablyalways eversible
[41,49,98]. and intromittentwhenpresent,preventssperm
backflowduringinsemination.Pórcionbasal[9];
Capsule: Part of the spermatophore between stalk and Membrane[103];Hojacapsular interna[85];
stem responsible for sperm transfer and composed of Internalcapsular sheet[86];Spermduct[90];
two sclerotized ridges/carinae (distal and basal) delimit- Median lobe[30].
ing a soft membrane, the “sperm duct”. Capsule [62, 64, Physema(Greekwordmeaningsomething inflated,
83]; Capsule portion [63]; Cápsula [77]; Sperm tube puffedup,bubble):Externallyinflated
(Buthidae)[83];Cápsula [85];Capsule [42,86,90]. membranouspouch observedonpost-
insemination spermatophores of bothriurids
Spermduct:Membranouspart ofthe capsule, and of several chactoid taxa, and formed by the
eversibleinmanytaxa.Protuberance[69]; eversion of the capsular terminal membranes.
Median transverse trough (part) [3]; Outer Capsular concavity: Lóbulo externo [6]; Lóbulo
lobe (Buthidae) [3]; Sperm duct [4, 83]; externo [9]; Concavidad capsular [85];
External lobe (Buthidae/Chaerilidae) [4]; Capsular concavity [86, 102].
Sperm duct (part) [49]. Hemisolenos (from the Greek words hemi and
Capsulardistalcarina:Lobe interne(Buthidae)[5]; solen which respectively mean half and pipe/
Innerprocess(Buthidae)[99];Internallobe channel): Lobe basal [5]; Valve [62, 63];
(Buthidae)[68,69];Lateral process(Buthidae) Inner lobe [3]; Internal lobe [101];
[64]; Lobulación dorso-externa posterior del internobasal reflection of sperm duct [4];
lobulió interno [100]; Inner lobe (Buthidae) [3]; Lamella [41, 49, 98].
Median lobe (Buthidae/Chaeriidae) [4]; Lóbulo Holosolenos(fromthe Greekwordsholos andsolen
interno (Buthidae) [84]. whichrespectivelymeanwhole/entireandpipe/
Capsularbasalcarinae:Lobeexterne(Buthidae)[5]; channel):Pipe-likestructureonthespermatophore
Superiorouterprocess(Buthidae) [99];External capsulecomposedofthetwohemisolenosand
lobe(Buthidae)[68,69];Medianprocess throughwhichthesemenistransferredintothe
(Buthidae)[64];Medianlobe(Buthidae)[3];Basal femalegenitaltract.
lobe(Buthidae)[3:Fig.99];laterallobe[90]. Accessoryapophysis(of hemisolenos):Accessory lobe
Basalhook(Buthidae):Lobebasal[5];Obliquevertical [41];Lamellaraccessorylobe[49].
process[99];Basallobe[4,68,69];Obliqueprocess Accessoryhook(of hemisolenos):Lamellaraccessory
[64];Basallobe(Buthidae)[3];Medianlobe hook [49].
(Buthidae)[3:Fig.99];Lóbulobasal[84];Process Clasper:Sclerotizedintromittentapophysisthat widens
(Pseudochatidae)[42]. thefemale genitaltract andprovidesasecure
Capsularforamen(spermatophore):Openingofthe anchoringforthe spermatophore. Lobeexterne
spermduct through which thesemen isexpelled. [5];Sacculus[63];Lóbulobasal[6,7,9,18,76,
Paired spermexits (Buthidae)[64];Foseta[77]; 77];Lobulaciónbasal[100];Basallobe[3,101];
Foramenoductoespermático(Buthidae) [84]; Lobebasal[103];Lóbulo capsular[85];Capsular
Foramen[85,86]. lobe[86];Crown-likestructure[90];Distal lobe
Subex(Latin wordmeaningbasallayer,support): Basal [41,49,98];Basallobeofcapsule [102];Ental
surface ofthespermduct.Escodatura dellóbulo lobe[30].
interno[9];floorofthespermduct[4];Basallobe Matingplug[4].
[98];Basal lobe[41(Figs.7,34,35),49]. Distalbarbofthe matingplug[4].
Tectum (Latin wordmeaningroof):Distalsurfaceof Basalplate ofthematingplug [4].
thespermduct.Lobemédian [5];Lóbulo interno
[6,7,9,76,77,100];Median lobe[3];Lobe Stem: Basal part composed of soft membranes and
interne[103];accessorydistal lobe[101]; containing the spermatozoa. Stem [63, 64]; Basal tube
Hojacapsular externa [85];External capsular [68]; Porción basal [6, 7, 9, 18, 76, 77, 100]; Basal
sheet[86];Posterior lobe[41,49,98]; portion [3, 102]; Trunk [4, 30, 42, 83, 86, 90]; Basal
Internallobe[102]. part [101]; Tronco [84, 85]; Basal trunk [49].
Monodetal.FrontiersinZoology (2017) 14:51 Page8of48
Truncalflexure:Sillon articulaire[5];Replieguebasal invagination ofthe capsularbasaledgerepresentstheul-
[7,18,76];Escotadurabasal[9,77,100]; Median timate step of this process eventally resulting in the es-
transversecleavage[3];Truncalflexure[4,49,83, tablishment of the more complex pattern. For these
86,90];Articular suture [101]; Flexión capsular reasons, both characters were assigned linear transform-
[85]; Basal fold [102]. ation series and treated as additive (ordered) [108]. The
Pedicel:Enlargedstickybasethatfixesthe consistency index and retention index of each character
spermatophoreto thesubstrate.Anterior process were calculated on each of the cladograms in order to
[99];Wings[62];Pedalwings[63];Anchor piece evaluatetheirlevelsof homoplasy onthedifferenttrees.
[68,69];Basalplate [64];Lengüeta/pieovoidal
largo [7];Pie[9,18,76,77,100];Foot +Stalk [3]; Results
Pedicel[49,83,86];Cylindrical gland[101]; Sampling
Pedicelo[84,85];Foot [42,90]. Hemispermatophores of 122 species (5% of known spe-
cies)belongingto71genera(35%ofdescribedgenera)and
Ancestralstatereconstructions 17families(85%ofknownfamilies)werestudied(Fig.3a–
The five bauplans identified during the course of this c), incorporating material from various museum collec-
study were optimized onto the currently available phylo- tions [see Additional file 6 for a complete list of speci-
genetic trees of the order Scorpiones proposed by mens]. The relevant literature was thoroughly checked to
Stockwell[4],Coddingtonetal.[104],Soleglad&Fet[46],/ inventory all published illustrations of hemispermato-
Solegladetal.[105],andSharmaetal.[106]underthepar- phores and spermatophores. Illustrations from published
simony criterion with Mesquite version 2.75 [107]. The taxonomic descriptions and morphological studies were
presence/absence of invagination on the basal edge of usedtocomplementspecimenexamination,allowingusto
capsule was also optimized onto the phylogenies of confirm the constancy of defined patterns within taxo-
Stockwell [4], Prendini [45], and Sharma et al. [106] using nomicgroupsforwhichonlyalimitednumberoftaxahad
thesamemethod. previously been examined. Although we tried to be as
Character matrices were produced for each of the exhaustive as possible, some illustrations may have been
cladograms (Additional files 1, 2, 3, 4, 5). The different missed. Photographs or drawings of hemispermatophores
bauplans seems to reflect a gradual evolutionary transi- and/or spermatophores of 578 species (24% of known
tion from a simple plain sperm duct towards more com- species), belonging to 119 genera (58%) and 19 families
plex evertible capsules by successive foldings. The (100% of known families when omitting the family
a FAMILIES b GENERA c SPECIES
S
E 5% 5%
OR 10%
H
OP 35% 35% 21%
T
A
M
R
PE 85% 74%
S
MI 30%
E
H
d e f 1% 2%
S 7%
RE 9%
O
H 35%
P
O 45%
T
A
M
PER 20% 84% 97%
S
Additional taxa
Taxa examined Taxa not examined
illustrated in literature
Fig.3Piechartsshowingdatacoverageofthepresentstudyforhemispermatophores(a–c)andspermatophores(d–f).a,dPiechartsshowing
theproportionoffamiliesforwhichdatawereobtainedfromexaminationofspecimensorfromtheliteratureduringthisstudy,andthe
proportionoftaxaforwhichdatawasmissing.b,ePiechartsshowingtheproportionofgeneraforwhichdatawereobtainedfromexamination
ofspecimensorfromtheliteratureduringthisstudy,andtheproportionoftaxaforwhichdatawasmissing.c,fPiechartsshowingthe
proportionofspeciesforwhichdatawereobtainedfromexaminationofspecimensorfromtheliteratureduringthisstudy,andtheproportion
oftaxaforwhichdatawasmissing
Monodetal.FrontiersinZoology (2017) 14:51 Page9of48
AkravidaeLevy,2007whichisonlyknownfromhollowed- involve two sets of characters: (1) characters pertaining
out exoskeletons and probably extinct [109, 110]) were to the overall architecture, which are responsible for the
found in the literature [see Additional file 7 for the ejection ofsemen from the spermatophore,and (2)char-
completelistandAdditionalfile8forreferences]. actersof the capsule, which ensure the transfer ofsemen
The combined dataset (data from examined material into the female genital tract. In all scorpion taxa, semen
and from the literature) represent 619 species (26% of expulsion is carried out by the same mechanism, i.e. an
known species), 132 genera (65%) and 19 families increase of pressure in the internal cavity of the sperm-
(100%) (Fig. 3a–c). The hemispermatophore bauplan for atophore induced by a bending of the whole structure
each of these taxa was coded [see Additional file 9 for [2, 63, 64, 84, 90]. As a result of this conserved mechan-
the complete matrix]. In the case of hemispermato- ism, the general groundplan of the spermatophore with
phores for which no specimen was available, the vast four distinct parts remains the same across the order
majority of published illustrations are good enough to despite aconsiderablemorphologicaldiversity(Fig. 1).
assess similarity with the material examined, and, in The capsule, that ensures the insemination per se,
most cases, to unambiguously assign hemispermato- shows an even greater structural variation, but its archi-
phores to one of the recognized patterns, even if the tecturefollowsalinearevolutionarypathway(seediscus-
capsularregionwasnot accuratelydepicted. sion) probably due to underlying mechanical constraints.
For some taxa no material or only incomplete As for the overall groundplan, while proportions and
hemispermatophores were examined, and in a few sizes of the various capsular features may differ signifi-
case the bauplan remains difficult to determine with cantly, the variation of the structural pattern is actually
certainty based on published illustrations alone, i.e. quite limited, mainly consisting of a gradual complexifi-
the genus Lisposoma Lawrence, 1928, and the fam- cation of the invaginations and foldings of the sperm
ilies Pseudochactidae and Superstitioniidae. Hemi- duct. Five basic bauplans (Fig. 4), accounting for the
spermatphores of these taxa were nonetheless structural changes of the capsule observed in the exam-
assigned to a recognised pattern, and the reasons for ined material, were identified. Each of these bauplans
each decision are given (see relevant paragraphs for roughly corresponds to an additional fold of the sperm
detail), pending the study of specimens to confirm duct. Thesearedescribedbelowfrom thesimplesttothe
the proposed hypotheses. most complicated architecture. The term ‘bauplan’ is
In additiontothe hemispermatophores,119 spermato- used here to designate an overall architectural pattern or
phores from 36 species (1% of known species) belonging organizationinvariantamongawiderangeoftaxa[60].
to 14 genera (7%) and 7 families (35%) (Fig. 3d–f) [See In most illustrations of the capsule presented, the dif-
Additional file 10 for the complete material list] were ferent parts are color-coded as follows: (1) the distal car-
obtained and studied in the course of this study. This ina is in green, (2) the basal carina in blue, (3) the sperm
structure was previously unknown for 30 of these duct is given in yellow in the first bauplan, (4) the tec-
species, for 9 of these genera, for two of the families and tumisinyellowandthesubexinorangeinthefollowing
for one of the sub-families [see Additional file 10]. Our bauplans, and (5) the terminal membrane is in pink.
dataset does not generally overlap with material from Corresponding parts in hemispermatophores and sper-
earlier publications and thus represents major progress matophores, and homologous parts in hemispermato-
in our knowledge of scorpion spermatophores, present- phores/spermatophores of different taxa are thus easily
ing information about more taxa than in all previous identifiable.
publications combined (spermatophores of 34 species
belonging to 13 genera and 9 families). The combined I.‘No-fold’bauplan
data represent 66 species (3% of known species) belong- This is the simplest pattern observed, with a non-folded
ing to 33 genera (16%) and 11 families (55%) (Fig. 3d–f). sperm duct (Fig. 4a). This is present in the families
Moreover, 22 of the spermatophores obtained are from Buthidae, Chaerilidae (Figs. 5–6), and putatively also in
Hormuridae and Scorpionidae, two families for which thePseudochactidae.
very little data was previously available in the literature. ThemonogenericfamilyChaerilidaepossesshemisper-
Thesespermatophores are amongthe mostcomplex and matophores with the simplest capsule, consisting of two
they provide essential data for a thorough comparison carinae surrounding an unfolded sperm duct ([4];
with the better-known spermatophores of the families Fig. 5a–c yellow). Although there is no visible foramen,
Bothriuridae,ButhidaeandEuscorpiidae. the semen can nevertheless be expelled because the
membranes of the two hemispermatophore are not fixed
Architectureofthecapsule together in the spermatophore between the capsular
The primary function of spermatophores is to ensure in- ridges (arrow in Fig. 5d). The same basic “no-fold” pat-
semination. The mechanisms by which this is achieved tern is observed in the family Buthidae ([3, 4]; Fig. 6).
Monodetal.FrontiersinZoology (2017) 14:51 Page10of48
a b
‘No-fold’ bauplan, lateral aspect ‘One-fold’ bauplan, lateral aspect
Caraboctoninae
Buthidae
Iurinae
Chaerilidae
Superstitioniidae?
Pseudochactidae? ttteeeccctttuuu Typhlochactidae
mmm
sssuuubbbeeexxx
c d
‘Two-folds’ bauplan, anterior aspect ‘Three-folds’ bauplan, anterior aspect
Bothriuridae
Chactidae
Uroctonus (Chactidae)
Euscorpiidae
Vaejovidae
Hadrudinae
Scorpiopidae
Troglotayosicidae
e f
‘Four-folds’ bauplan, anterior aspect ‘Four-folds’ bauplan, lateral aspect
Diplocentridae
Hemiscorpiidae
Heteroscorpionidae
Hormuridae
Lisposoma (Bothriuridae)?
Scorpionidae
Urodacidae
2
11
Fig.4Bauplansandfoldingprocessesofthehemispermatophorecapsule,diagrammaticrepresentationscomplementedbycamera-lucida
drawingswithindicationofthecorrespondingarea.a,‘No-fold’pattern,lateralaspect.b,‘One-fold’pattern,arrowsindicatethedirectionofthe
foldingdelimitingthesubexandtectum,lateralaspect.c,‘Two-folds’pattern,anterioraspect.d,‘Three-folds’patternobservedinUroctonusmor-
daxThorell,1876,anterioraspect,arrowsshowtheinvaginationofthetectum.e–f,‘Four-folds’pattern,anterior(e)andlateral(f)aspects,arrow1
showstheinvaginationofthebasalcapsularedgeandarrow2indicatestheextensionofthehemisolenos
However, in contrast to chaerilid hemispermatophores, a bent backward during mating, medially in chaerilids
foramen (cf in Fig. 6c and g) is clearly visible in buthid (dotted arrow in Fig. 5d) and more basally in the elon-
hemispermatophores, because the duct membrane is gated buthidspermatophores.
shifted away from the longitudinal axis as a result of According to illustrations in Prendini et al. [42], the
wider capsular carinae. In some other buthid taxa, like hemispermatophore of Pseudochactidae appears to be
observed here in Hottentotta jayakari (Pocock, 1895), morphologically intermediate between those of buthids
the duct membrane even extends above the basal carina and chaerilids. The stem is short as in chaerilids, and
(sdm in Fig. 6g–h yellow). The insemination process is the stalk is flagelliform but much thicker than that ob-
similar inthe Chaerilidae(this publication)andButhidae served in buthids. The capsule is not shown in detail in
[64, 65, 68–72, 75, 78, 80–84, 88], without alteration of Prendini et al. [42], therefore the bauplan could not be
the structure of the capsule. Furthermore, unlike other determined and potential similarities with buthid and
spermatophores, buthid and chaerilid spermatophores chaerilid capsules could not be assessed accurately. An
do not have a fixed truncal flexure; the stem is simply apophysis at the base the stalk was interpreted by