Memoirs of the Museum ol" Victoria 56(l):235-244 (1997) 28 February 1997 https://doi.org/10.24199/j.mmv.1997.56.09 SPERMATOLOGICAL EVIDENCE SUPPORTS THE TAXONOMIC PLACEMENT OF THE AUSTRALIAN ENDEMIC HAIRY STONE CRAB, LOMIS HIRTA (DECAPODA: ANOMURA: LOMIDAE)

Christopher C. Tudge

Zoology Department. University of Queensland, Brisbane, Qld 4072, Australia Present address: Crustacea. Museum of Victoria, 71 Victoria Crescent. Abbotsford. Vic. 3067. Australia

Abstract

Tudge, C.C., 1997. Spermatological evidence supports the taxonomic placement of the Australian endemic hairy stone crab, Lomis hirta (Decapoda: Anomura: Lomidac). Memoirs of the Museum of Victoria 56(1): 235-244. An ultrastructural investigation of the spermatozoal morphology of the Australian endemic hairy stone crab, Lomis hirta, has revealed a unique sperm morphology supporting placement within its own family and superfamily. The spermatophore and spermatozoal morphology agree with morphological evidence that place this crab within the infraordcr Anomura. the relatively unmodified spermatozoon suggests a basal origin for this taxon (with respect to other anomurans) and may be similar to the ancestral sperm type for many families in this infraordcr. An independent lineage from the base of the Anomura has been previously suggested for Lomis on the basis of many morphological synapomorphies being shared with representatives from disparate families of the Anomura.

Introduction tive of its sharing many characters with the paguroids, galatheoids and thalassinoids. Previous classifications of the decapod McLaughlin ( 1 983a) similarly placed Lomis in a infraorder Anomura (Borradaile, 1907; basal position, with respect to the Anomura, but Glaessner, 1969), recognised the four superfam- had it evolving independently as a distinct evol- ilies, Thalassinoidea. Paguroidea, Galatheoidea utionary line from a common ancestral stock. and Hippoidea. More recently the thalassinoids Lomis continues to be problematic in regard to have been excluded from the Anomura and the its relationship to the remainder ofthe Anomura constituent anomuran superfamilies redefined (McLaughlin, 1983b; Martin and Abele, 1986; as the Paguroidea, Lomoidea. Galatheoidea and Richterand Scholtz, 1994). Hippoidea (McLaughlin, 1983b; McLaughlin The use of spermatozoal morphology for sys- and Holthuis, 1985; Poore, 1994). McLaughlin tematic, taxonomic and phylogenetic studies, and Holthuis stated that the term 'Anomala' was initially pioneered by workers such as Koltzoft" the more appropriate name for these constituent ( 1 906) and Retzius ( 1 909), has been successfully superfamilies (minus the thalassinoids), but the applied to the Crustacea (Jamieson, 1 99 1) and term 'Anomura' is more commonly used and has in the Anomura and Brachvura in particular been adopted herein. (Tudge, 1992, 1995a, b; Jamieson, 1994, 1995; The superfamily Lomoidea contains the and references therein). In this paper, the ultra- monospecific genus Lomis in the family Lomi- structure of the spermatozoa of Lomis hirta is dae. This hairy stone crab is endemic to the described for the first time and the phylogenetic rocky southern coasts of Australia and like its implications are discussed. porcellanid relatives is very crab-like in appear- ance. Lomis hirta was first described as a hairy Methods porcellanid by Lamarck (1818), then Bouvier

( 1 894, 1 895) considered it a symmetrical hermit The single male specimen of Lomis hirta crab, but this enigmatic crab has since been elev- (Lamarck, 1818) was collected and fixed by Dr ated to its own family and superfamily Gary Poore, Curator of Crustacea, Museum of (McLaughlin, 1983a). Pilgrim (1965) reassessed Victoria, Australia from Flinders Reef. Victoria the literature and examined fresh specimens and in September 1993. The male reproductive came to the conclusion that Lomis was a basal material (both testes including the ducts of the offshoot of the paguroids (along with the Pylo- vasa deferentia) was removed from the crab and cheiidae) and that this basal position was indica- immediately fixed in cold glutaraldehyde for a

235 236 C. C. TUDGE

minimum of 2 hours at 4°C then posted to Bris- spermatophorc, is shown at the electron micro- bane at ambient temperature where the remain- scope level to be composed of a single homo- der of the fixation and embedding process was geneously granular electron-dense layer (Fig. carried out. 2A) and to be of variable thickness.

Light microscopy Spermatozoa/ morphology For light microscopy, glutaraldchydc-fixcd At the light microscope level the spermatozoa sperm were viewed and photographed under an of Lomis hirta are irregular in outline, but basi- Olympus BH2 Nomarski interference contrast cally globular in form with one to three small microscope. Micrographs were taken with an vertices which may be extended into microtubu- attached Olympus OM-2 camera. lar arms (Figs 1B-E). Up to two microtubular Transmission electron microscopy arms have been observed in a single spermato- The standard fixation procedure (outlined zoon but sometimes three vertices are apparent below) for transmission electron microscopy (Fig. IB). A ring-shaped acrosome vesicle is was carried out in a Lynx -el. Microscopy Tissue obvious on one surface of the sperm cell. The Processor (Australian Biomedical Corporation, entire sperm cell can be up to 6um in diameter Ltd, Mount Waverley. Victoria, Australia), after (refer to Figs 2C and 4 throughout). In all figures the initial glutaraldehyde fixation and first phos- the acrosomal or apical end of the spermatozoon phate buffer wash. is considered the anterior pole while the

Portions of the testis (approximately I mm') opposite, basal or nuclear end is posterior. were fixed in 3% glutaraldehyde in 0.2 M phos- The acrosome vesicle is an irregular, inverted phate butter (pH 7.2), with 3% sucrose added, cup-shaped structure, approximately 2.5 umi for a minimum of 1 h at 4°C. They were washed wide, and is broadly penetrated posteriorly by in phosphate butter (3 washes in I 5 min), post- the perforatorial chamber (Figs 2B, C; 3A, B, D). fixed in phosphate buffered 1% osmium tctrox- The acrosome vesicle wall is relatively thin and ide for 80 min; similarly washed in buffer and is composed of three distinctive zones. The most dehydrated through ascending concent rat ions of external region of the vesicle is made up of an cthanol (40-100%). After being infiltrated and interrupted layer of extremely electron-dense embedded in Spurr's epoxy resin (Spurr, 1969), material embedded in a granular zone of lesser thin sections (500-800 A thick) were cut on a electron density. The extremely electron-dense LKB 2128 UM IV microtome with a diamond material is interpreted to be the equivalent of the knife. Sections were placed on carbon-stabilised dense operculum, which normally caps the acro- colloidin-coated 200 urn mesh copper grids and some vesicle in anomurans, but its extension stained (according to Daddow, 1986) in Rey- around the entire acrosome vesicle is unusual nold's lead citrate (Reynolds, 1963) for 30 s, (Figs 2B, C; 3A, B, D, F). As mentioned this rinsed in distilled water, then 6% aqueous uranyl opercular layer is not a continuous, evenly dis- acetate for 1 min. Reynold's lead citrate again tributed structure but forms an interrupted for 30 s and a final rinse in distilled water. dense layer. The electron-dense, granular zone, Micrographs were taken on an Hitachi H-300 here interpreted as the outer acrosome zone, sur- transmission electron microscope at 80 kV. rounds the operculum and composes the majority of the thin acrosome vesicle. Occurring between this combined dense outer acrosome Results layer and the perforatorial chamber wall is a Spcrmaiophorc morphology zone of similar granularity, the inner acrosome The spermatophores of Lomis hirta were only zone (Figs 2C; 3A. B, D-F). This latter zone viewed in thick sections for transmission elec- is more electron-pale than the outer acrosome tron microscopy and the dimensions and mor- zone. phology can only be estimated. The spermato- The perforatorial chamber invaginates into phores are pedunculate, having the sperm-filled the acrosome vesicle from the posterior end and ampulla attached to a long thin stalk which in extends almost to the anterior most tip. It has a turn is connected to a basal plate or pedestal. broad posterior region which tapers anteriorly to The entire spermatophores have been estimated a blunt point (Figs 2C; 3A). The open, posterior at approximately 300 um in length and the end is slightly constricted when compared to the ampullar dimensions at 100 um long X 50 um greatest width of the perforatorial chamber. The wide (Fig. 1A). The spermatophorc wall (Fig. perforatorial chamber contents can be divided

I E), which constitutes the ampullar region of the into two distinct regions. Internally there is a TAXONOMIC POSITION OF THE HAIRY STONE CRAB 237

stained thick resin section showing partially sectioned Figure I. A-E Lomis lurta (Lomidae). A. Toluidine spcrmatophore and contained spermatozoa. B-D. Light micrographs of spermatozoa showing one or two micro- the vas deferens. Note: the thick spermato- tubular arms. E. Light micrograph of spermatozoa extruded from phore wall. Scale bars as indicated (original). - pedestal: s - spermatozoon: st = stalk; sw = a = ampulla; av = acrosome vesicle; ma = microtubular arm: ped spcrmatophore wall. 238 C. C. TUDGE

Figure 2. A-C I .onus hula (Lomidac). Transmission electron micrographs of spermatozoa. A. Detail of'spcr- matophorc wall surround ing spermatozoa. B. Low power shot oft hree spermatozoa. C. Longitudinal section (LS) of a spermatozoon. Arrowhead indicates probable actin filaments. Scale bar - I urn. except where indicated (original). av - acrosoine vesicle; cy = cytoplasm; ia = inner acrosome zone; m = mitochondrion; ms = membrane system: mt = microtubules: n = nucleus; npm = nucleo-plasma membrane: o = operculum; oa = outer acrosome zone; p = perforatorial chamber: sw = spermatophorc wall. TAXONOMIC POSITION OF THE HAIRY STONE CRAB 239 core of longitudinally arranged striations, simi- Ailogalathea. the hippid, Hippa, and several lar in appearance to actin filaments, while exter- porcellanid species (Tudge, 1995a). nal to this is a homogeneous, finely granular The spermatozoon of Lomis hirta (Figs 2, 3, 4) region of moderate electron density (Figs 2C; possesses spermatozoa! characters, such as 3A, B; 4). The posterior contents of the perfor- microtubular arms (possible three?) and a con- atorial chamber appear continuous with the centrically arranged acrosome vesicle poster- cytoplasm beneath the acrosome vesicle. iorly penetrated by a pcrforatorial chamber, The cytoplasm is very extensive and com- which justify its position in the Anomura but its pletely surrounds the acrosome vesicle, with the spcrmatozoal morphology is distinct enough to widest areas lateral to it (Figs 2C; 3D, F; 4). warrant its own family and superfamily. Some of Anterior and posterior to the acrosome vesicle the characteristics of the spermatozoon of the cytoplasm is reduced to a thin layer with no Lomis are considered unique. The variation in obvious organelles present, except for a pair of sperm cell shape shown within an individual is centrioles which occur beneath the pcrforatorial unusual and the amorphous form of the sperma-

chamber (Fig. 3C). Lateral to the acrosome ves- tozoa (Figs 1 B-E; 2B), and to a lesser extent, the icle the cytoplasm contains many organelles acrosome vesicle, distinguishes Lomis from including an anterior region of loosely aggre- other investigated anomurans (Hinsch. 1980, gated membranes and associated with these, 1991; Tudge, 1992, 1995a, b). The small acro- many small spherical mitochondria, of which some vesicle completely embedded in the cyto- some appear cristate and others more degener- plasm has not been recorded for any other ate (acristate) (Figs 2C; 3D, F). Scattered anomuran spermatozoon and is only approxi- amongst these organelles are short and long mated by that seen in the thalassinoids, Axius bundles of microtubules which represent and Callianassa (Tudge. 1995a, b). The micro- internal sections of the microtubular arms. tubular bundles, which appear to be concen- Some bundles of microtubules are seen to extend trated in the cytoplasm, can also extend into the into the nuclear material (Figs 3D, E). nucleus (Figs 2C; 3D-F). It is not known if these The nucleus of the spermatozoa of Lomis hirta microtubular bundles, which are the bases of the

is irregular in form but retains a basically globu- microtubular arms, are cytoplasmic in origin (as lar appearance (Figs 1B-D). It is composed Off in other investigated anomurans) or nuclear in homogeneous, coarsely granular material of origin (as seen in the spermatozoa of astaci- moderate electron density and is surrounded deans, palinurans and some thalassinoids and externally by a thickened nucleo-plasma mem- brachyurans). Although the sperm cells appear brane (Figs 2B, C; 3D; 4). No distinct membrane to have three vertices, only one or two microtu- separates the cytoplasm from the nucleus. As bular arms have been seen on a single sperma- previously mentioned small sections of the tozoon (Figs 1B-D) and it is not known if three internal microtubular bundles are apparent in microtubular arms is standard for Lomis. Three the nucleus (Fig. 2C). microtubular arms have been recorded for all investigated paguroids and galatheoids, with the exception of the Porcellanidac; which, with the Discussion hippids and thalassinoids. have four or more The paucity of information concerning the mor- microtubular arms (Tudge, 1995a). Another phology of the spermatophores of Lomis hirta unique feature of the spermatozoa of Lomis is the discontinuous electron-dense operculum, make it difficult to do more than speculate about surrounds the entire acrosome vesicle their structure. It can be ascertained that the which apical spermatophores are pedunculate (Fig. I), and (Figs 2C; 2A, B, D-F) and not just the pole, investigated anomurans. therefore typical of the Anomura studied to date as in all other An concerning this structure is (Tudge, 1991, 1995a). Until further morphologi- alternative theory it is not an operculum at all, but simply cal and ultrastructural studies are carried out on that dense zone of the acrosome vesicle. This the spcrmatophore of this crab it will not be another possibility assumes that the operculum is there- known if it has a spcrmatophore morphology fore lacking which would similarly be an auta- that substantiates its morphological and spcr- matozoal uniqueness. The homogeneous, granu- pomorphy. in The Lomidae have been assigned to a basal lar appearance of the spcrmatophore wall position in the Anomura, based on adult somatic Lomis (F\g. 2A) is similar to that recorded in the McLaughlin, spermatophores of several pagurid hermit crabs, characters (Pilgrim, 1965; 1983a, Martin and Abele, 1 986; Richter and Scholtz, the parapagurid, Sympagurus, the galatheid. b; 240 C. C. TUIXJE

Figure 3. A-F. Lomh tuna (Lomidae). Transmission electron micrographs of spermatozoa. A. Detail of longi- luilinal section (LS) of acrosome vesicle. Arrowhead indicates probable actin filaments. B. Transverse section (TS) through acrosome vesicle. C. Detail of pair of ccntriolcs. D. Oblique section through a spermatozoon showing long internal bundle of microtubules. K. Detail of microtubular bundle. F. Detail of cytoplasmic orga- nelles. Scale bars - lum, except where indicated (original). ce = ecntnole: cy = cytoplasm; ia = inner acrosome zone; m = mitochondrion: ms = membrane system; mt = microtubules: n = nucleus: o = operculum; oa = outer acrosome zone; p « perforatorial chamber. TAXONOMIC POSITION OF THE HAIRY STONE CRAB 241

operculum membrane systems some zone

ter acrosome zone perforatorial chamber cytoplasm

tochondrion

crotubules

actin filaments?

nucleo-plasma membrane

longitudinal section of a spermatozoon, based on a tracing Figure 4. Lonvs hinu (Lomidae). Sem.diagrammatic (original). of a micrograph. Scale bar = I urn 242 C. C. TUDGE

1994). but there are differences between their was not one of the anomuran taxa used. The several conclusions. Not all authors dealt with all ano- spermatozoa of Lomis appear to share in with two previously muran taxa. Pilgrim ( 1 965) thought the family to characters common be an early offshoot of the Anomura (within investigated thalassinoids {Axius and Callia- which he included the Thalassinidea). nassa) but a detailed comparison of spermato- McLaughlin (1983a: 435) stated "the Lomidae zoal morphology between the Lomoidea and represents a distinct evolutionary lineage within Thalassinoidea will have to wait pending a more the Anomala. sharing the basic characters of the thorough investigation into the spermatozoal infraorder, but having evolved independently form and diversity within the latter super- from the ancestral stock common to all of its family. major taxa", and decided it was a specialised The ancestor of some or all of the anomuran and highly evolved taxon within the infraorder. families may have had a spermatozoal mor- Martin and Abele (1986) thought the Lomidae phology very similar to that of Lomis and the to be the sister family to the Lithodidae, together diversity (admittedly restricted by a conserva- sister taxa to the Paguroidea. Richter and tive ground-plan) of sperm morphology encoun- Anomura could Scholtz ( 1 994) placed the Lomidae as the sister tered in extant families of the group of the Paguroidea (including Lithodidae have emerged through modification of this in the latter superfamily) based on uropod mor- ancestral sperm form. It is conceivable that the phology. anomuran spermatozoal type (in the Paguroidea The fact that the phylogenetic relationship of in particular) developed from a lomid-like mor-

Lotnis has sparked such debate and still remains phology by 1 ) increasing the size and complexity unclear is reflected in its lack of recognisable of the acrosome vesicle, 2) restricting the oper- morphological apomorphies and its sharing of culum to the apical pole, 3) shifting the acro- numerous synapomorphic characters with more some vesicle to a superior position on the remotely related anomuran taxa. Like many cytoplasm and 4) establishing three, equidistant other families in the Anomura (the Hippidae, microtubulararms. The lomid lineage may have Lithodidae, Pylochelidae and Aeglidae, for independently evolved from this ancestor but example) the Lomidae is only represented as fos- retained the spermatozoal morphology which sils from the Recent (Glaessner, 1 969) and there- was ancestral to the remainder of the Anomura. fore palaeontological evidence of their evolution The lack of distinct morphological apomor- is limited. phies, the sharing of morphological characters The spermatozoal data supports the views with as many as nine other anomuran families presented by Pilgrim (1965) and McLaughlin and the newly presented spermatological evi-

( 1 983a) that Lomis is basal to, or a separate lin- dence, indicating a unique and particularly eage from, the ancestor of the Anomura + Thal- unmodified sperm morphology, would appear assinidea (Pilgrim, 1965) or Anomala (anomu- to vindicate a basal or ancestral position of rans excluding thalassinideans) (McLaughlin, Lomis in relation to the remainder of the Ano-

1 983a). This same spermatozoal data is less sup- mura. portive of the view expressed by Richter and Scholtz (1994) that the Lomidae is the sister Acknowledgments taxon of the Paguroidea and does not support The author wishes to thank Dr Gary Poore the close relationship between Lomis and litho- (Museum of Victoria) for collecting and under- dids suggested by Martin and Abele (1986). taking the initial fixation of the specimen of Although the spermatozoal morphology in the Lomis hirta. Mrs Lina Daddow (Zoology Lithodidae is only described from light micro- Department, The University of Queensland) is scope drawings of Lilhodes maja (see Retzius, thanked for technical assistance with electron 1 909), it can be clearly ascertained that the over- microscopy. The support and careful reading of all spermatozoal morphology is unlike that of the manuscript by Prof. Barrie Jamieson Lomis and more similar to investigated pagu- (Zoology Department, The University of roids (Tudge, 1992, 1995a, b). A close link Queensland) is gratefully acknowledged. between Lomis and lithodids seems unlikely on the present spermatological evidence. Poore References ( 1 994) clearly showed a sister group relationship between the Thalassinidea and the Anomura Borradailc, L.A., 1907 On the classification of the based on morphological characters but the pos- decapod crustaceans. Annals of the Maga:ine of ition of Lomis in this analysis is not known as it Natural History 19: 457-486. .

TAXONOMIC POSITION OF THE HAIRY STONE CRAB 243

Bouvier, E.L.. 1894. Sur les caracteres et revolution Martin, J.W. and Abele. L.G., 1986. Phylogenetic dcs Lomisines. nouveau groupc dc Crustaces ano- relationships of the genus Aegla (Decapoda: Ano- moures. Complex Rendus Hebdomadaires de mura: Aeglidae), with comments on anomuran Seances dc I'Academic dcs Sciences, Paris 118: phylogeny, Journal ofCrustacean Biology 6: 576- 1353-1355. 616. Bouvier. E.L.. 1895. Recherche sur les affinites dcs McLaughlin, P. A., 1 983a. A review of the phylogenetic Lithodcs et des Lomis avec les Pagurides. Annates position of the Lomidae (Crustacea: Decapoda: dcs Science Naturclles, Zoologie (Serie 7) 18: 157— Anomala). Journal of Crustacean Biology 3: 431- 213. 437.

Daddow, L.Y.M., 1 986. An abbreviated method of the McLaughlin, P.A., 1983b. Hermit crabs — are they double lead stain technique. Journal of Submi- really poly phy\elk'? Journal ofCrustacean Biology

croscopic Cytoiogy 18: 21 1-224. 3: 608-62 1

Glaessner. M.F., 1969. Decapoda. Pp. R399-R533 in: McLaughlin, P.A. and Holthius, L.M., 1 985. Anomura Moore. R.C. (ed.). Treatise on invertebrate paleon- versus Anomala. Crustaceana 49: 204-209. tology, Arthropoda 4. Part R. vol. 2. Geological Pilgrim, R.L.C.. 1965. Some features in the mor- Society of America and University of Kansas phology of Lomis hirla (Lamarck) (Crustacea: Press: Lawrence. Decapoda) and a discussion of its systematic pos- Hinsch. G.W., 1980. Spermiogenesis in Coenobitacly- ition and phvlogeny. Australian Journal of

pealus, I. Sperm structure. International Journal Zoology 13: 545-557. of Invertebrate Reproduction 2: 189-198. Poore, G.C.B.. 1994. A phylogeny of the families of Hinsch. G.W., 1991. Ultrastructure of the sperm and Thalassinidea (Crustacea: Decapoda) with keys to spermatophores of the anomuran crab Pleuron- families and genera. Memoirs of the Museum of

codes planipes. Journal of Crustacean Biology 1 1: Victoria 54: 79-120.

17-22. Retzius, G., 1 909. Die spermien der Crustaceen. Biol- Jamieson, B.G.M.. 1991. Ultrastructure and phylo- ogische Untersuchungen 14: 1-54. geny of crustacean spermatozoa. Memoirs of the Reynolds, E.S., 1963. The use of lead citrate at high pH Queensland Museum 31: 109-142. as an electron opaque stain in electron Jamieson, B.G.M., 1994. Phylogeny of the Brachyura microscopy. Journal of Cell Biology 17: 208- with particular reference to the Podotremata: evi- 212. dence from a review of spermatozoal ultrastruc- Richter, S. and Scholtz, G., 1994. Morphological evi- ture (Crustacea, Decapoda). Philosophical Trans- dence for a hermit crab ancestry of lithodids actions of the Royal Society of London B 345: (Crustacea, Decapoda, Anomala, Paguroidea).

373-393. Zoologische Anzeiger 233: 1 87-2 1 0. Jamieson, B.G.M., 1995. Phylogeny of the Brachyura: Spurr, A.R., 1969. A low viscosity epoxy-resin embed- evidence from spermatozoal ultrastructure (Crus- ding medium for electron microscopy. Journal of tacea, Decapoda). In: Jamieson, B.G.M., Ausio, J. Ultrastructure Research 26: 31-43. and Justine, J.-L. (eds.). Advances in spermatozoal Tudge, C.C., 1991. Spermatophore diversity within phylogeny and taxonomy. Memoires du Museum and among the hermit crab families, Coenobiti- National d'Histoire Naturelle (Paris) 166: 265- dae, Diogenidae and Paguridae (Paguroidea. 283. Anomura. Decapoda). Biological Bulletin 181: Koltzoff, N.K., 1906. Studien uberdiegestalt derzellc. 238-247. ultrastructure her- I. Untersuchungen iiber die spermien der deca- Tudge, C.C.. 1992. Comparative of poden, als einleitung in das problem der zellen- mit crab spermatozoa (Decapoda: Anomura: gestalt. Archiv fur Mikroskopisehe Anatomie und Paguroidea). Journal of Crustacean Biology 12: Entwicklungs-Geschichte 67: 364-572. 397-409. Lamarck, J., 1818. Histoire naturelle des animaux Tudge, C.C., 1995a. The ultrastructure and phylogeny sans vertebres, presentant les caracteres generaux of anomuran crab spermatozoa. Ph.D. Thesis, et particuliers de ces animaux, leur distribution, Zoology Department, University of Queensland, leurs classes, leurs families, leurs genres, et la Brisbane, Australia. 346 pp. citation des principales especes qui s'y rapportent; Tudge, C.C., 1995b. Ultrastructure and phylogeny of precedee d'une introduction offrant la determi- the spermatozoa of the infraorders Thalassinidea nation des caracteres essentiels de I'animal, sa and Anomura (Decapoda, Crustacea). In: Jamie- Justine, J.-L. (eds.). distinction du vegetal et des autres corps naturels, son, B.G.M., Ausio, J. and enfin, I'exposition des principes fondamentaux Advances in spermatozoal phylogeny and tax- Memoires du Museum National d'His- de la zoologie. 5: 1-612. onomy. toire naturelle (Paris) 166: 251-263.