Evolution of intra-ejaculate sperm interactions: do sperm cooperate?

Biol. Rev. (2011), 86, pp. 249–270. 249doi: 10.1111/j.1469-185X.2010.00147.x

Evolution of intra-ejaculate sperminteractions: do sperm cooperate?

Dawn M. Higginson∗ and Scott PitnickDepartment of Biology, Syracuse University, 107 College Place, Syracuse, New York 13244, USA

(Received 11 January 2010; revised 31 May 2010; accepted 2 June 2010)

ABSTRACT

Sperm are often considered to be individuals, in part because of their unique genetic identities produced as a resultof synapsis during meiosis, and in part due to their unique ecology, being ejected away from the soma to continuetheir existence in a foreign environment. Selection at the level of individual sperm has been suggested to explain theevolution of two enigmatic sperm phenotypes: sperm heteromorphism, where more than one type of sperm is producedby a male, and sperm conjugation, where multiple sperm join together for motility and transport through the femalereproductive tract before dissociation prior to fertilization. In sperm heteromorphic species, only one of the spermmorphs typically participates in fertilization, with the non-fertilizing ‘‘parasperm’’ being interpreted as reproductivealtruists. Likewise, in species with sperm conjugation, high levels of sperm mortality have been suggested to be requiredfor conjugate break-up and this has been considered evidence of kin-selected altruism. However, it is unclear if spermpossess the heritable variation in fitness (i.e. are individuals) required for the evolution of cooperation. We investigatethe question of sperm individuality by focusing on how sperm morphology is determined and how sperm conjugatesare formed. Concentrating on sperm conjugation, we discuss functional hypotheses for the evolutionary maintenanceof this remarkable trait. Additionally, we speculate on the potential origins of sperm heteromorphism and conjugation,and explore the diversification and losses of these traits once they have arisen in a lineage. We find current evidenceinsufficient to support the concept of sperm control over their form or function. Thus, without additional evidence ofhaploid selection (i.e. sperm phenotypes that reflect their haploid genome and result in heritable differences in fitness),sperm heteromorphism and conjugation should be interpreted not as cooperation but rather as traits selected at thelevel of the male, much like other ejaculatory traits such as accessory gland proteins and ejaculate size.

Key words: sperm heteromorphism, sperm conjugation, haploid expression, sperm morphological evolution, motility,sperm competition.

CONTENTS

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250II. A brief overview of spermatogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

III. Haploid expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251IV. Types of putative sperm cooperation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252

(1) Sperm heteromorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252(a) Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252(b) Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252

(2) Sperm conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253(a) Types of conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

( i ) Primary conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254( ii ) Secondary conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

(b) Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257( i ) Primary conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257( ii ) Secondary conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

* Address for correspondence: Tel: 315-443-7248; Fax: 315-443-2012; E-mail: [email protected]

Biological Reviews 86 (2011) 249–270 © 2010 The Authors. Biological Reviews © 2010 Cambridge Philosophical Society

250 Dawn M. Higginson and Scott Pitnick

(c) Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259( i ) Motility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259( ii ) Alternative hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

V. Evolutionary patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261(1) Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261

(a) Sperm heteromorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261(b) Conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

(2) Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262(a) Conjugate size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262(b) Division of labour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263

(3) Losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263VI. Evidence for sperm cooperation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264

(1) Are sperm individuals? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264(2) Is there haploid selection? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264(3) Are sperm heteromorphism and conjugation cooperation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

VII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265VIII. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266

IX. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266

I. INTRODUCTION

Sperm are exceedingly unusual cells due to their dual nature(Sivinski, 1984). On one hand, they are highly differentiated,and seemingly simple, haploid cells of a male’s body thattransmit his genes from one generation to the next. On theother hand, after ejaculation sperm are free-living organismswith unique genetic identities that perform a host of functionsindependent of the male that produced them, including (i)maturation, transformation or capacitation after ejaculation,(ii) location of an egg in a water column or navigation of afemale reproductive tract, (iii) prolonged storage within thefemale, (iv) a variety of biochemical and cellular interactionsbetween sperm, seminal plasma and the female, and often,(v) competing with the sperm of other males to fertilize afemale’s eggs (Pitnick, Hosken & Birkhead, 2009b; Pitnick,Wolfner & Suarez, 2009c; Pizzari & Parker, 2009). Thus,it is little wonder that sperm are frequently interpreted asindividuals, replete with their own evolutionary interests (i.e.fitness costs and benefits that might differ from the male) andreproductive fitness (e.g. Moore et al., 2002; Pizzari & Foster,2008; Sivinski, 1984).

The concept of sperm individuality has been broadlyapplied to explain the evolution of two unusual spermatozoicforms: heteromorphism and conjugation. Sperm heteromor-phism occurs when more than one type of sperm is producedin a single testis, with typically only one morph participatingin fertilization (Till-Bottraud et al., 2005). Sperm conjugationinvolves the joining together of two or more sperm for motil-ity or transport before dissociating at the site of fertilization orstorage (Pitnick et al., 2009b). Remarkably, these seeminglydisparate phenomena sometimes co-occur with two distinctsperm types joining together to form complex conjugates(e.g. Afzelius & Dallai, 1983; Ferraguti, Grassi & Erseus,1989; Healy & Jamieson, 1993), thereby suggesting that theselective function of these two phenotypes may be comple-mentary. Both sperm heteromorphism and conjugation have

been interpreted as sperm cooperation, including reproduc-tive altruism, to enhance fertilization success in competitiveenvironments (Baker & Bellis, 1987; Buckland-Nicks, 1998;Holman & Snook, 2006; Immler, 2008; Kura & Nakashima,2000; Pizzari & Foster, 2008; Pizzari & Parker, 2009; Silber-glied, Shepherd & Dickinson, 1984; Sivinski, 1984), akin tothe sterile worker castes of eusocial insects. This conceptualinterpretation implies that these sperm morphologies evolvedthrough selection acting at the level of individual sperm. Tra-ditionally, however, it is the diploid genome of the male thathas been thought to control the phenotypes of the sperm theyproduce (Eddy, 2002; Joseph & Kirkpatrick, 2004). Thus,any variation in the reproductive fitness of sperm would beattributable to males and not to the genotypes of individualsperm. It is unclear if spermatozoa themselves possess theheritable variation in fitness required for the evolution ofcooperation, or any other trait.

Discussion of sperm heteromorphism and conjugationhas often been confused by unstated assumptions aboutthe meaning of cooperation. Evolutionary biologists usethe term cooperation to describe social interactions, wherethe actions of one individual benefit another individual,that have evolved through fitness benefits accrued via kinselection (Hamilton, 1964), reciprocal interactions betweengroup members (Trivers, 1971), and higher-order selectionamong interacting groups (e.g. for reviews see Michod, 1999;Okasha, 2006). In general usage, however, cooperation sim-ply means to work together, irrespective of evolutionaryprocess (we term this ‘‘functional cooperation’’). Furthercomplicating matters, cooperation can shift from ‘‘evolu-tionary’’ to ‘‘functional’’ during transitions in the unit ofselection. For example, evolutionary cooperation amonginteracting unicellular organisms becomes functional coop-eration if the organisms merge to form a multicellular entity(Michod, 1999). In a recent review, Immler (2008) definedsperm cooperation as ‘‘the partitioning of function and/orthe mutual interaction between sperm of one male (i.e. sibling


Do sperm cooperate? 251

sperm) to increase a male’s fertilisation success’’. Unfortu-nately, Immler’s (2008) definition of sperm cooperation isnot explicit about origins and could include both functionaland evolutionary cooperation. For clarity, we recommendthat the term sperm cooperation be applied only to instanceswhere the traits of interest have evolved by sperm-levelselection.

To all appearances, heteromorphic sperm or members ofa sperm conjugate function in concert with other sperm, sat-isfying the criteria functional cooperation, but the selectiveenvironments that have resulted in the evolution of theseunusual phenomena are unknown. If males control spermphenotype, then sperm may be viewed as captives of maleevolutionary interests, and variation in sperm phenotypeshould be interpreted similarly to variation in other ejacu-latory traits such as seminal proteins or ejaculate volume.However, if sperm can influence their fate (i.e. haploid geneexpression underlies relevant variation in sperm form), thenit is both appropriate and necessary to consider the evo-lutionary interests of sperm and potential conflict with themale that produced them (Immler, 2008; Pizzari & Foster,2008). To establish if sperm heteromorphism or conjugationrepresents evolutionary or merely functional cooperation itmust be determined if selection acts at the level of sperm,males or both. Simply put: are sperm individuals?

The biological processes that control sperm phenotypeand conjugate formation have been largely overlooked indiscussions of sperm cooperation but provide clues to thequestion of sperm individuality. Here, we review the sper-matogenic mechanisms determining sperm form and discussthe functional hypotheses for the maintenance of sperm het-eromorphism and conjugation. Additionally, we speculate onthe potential origins and diversification of these remarkabletraits. Lastly, we critically examine the evidence for haploidselection of sperm and thus, the potential for evolutionarycooperation among sperm.

II. A BRIEF OVERVIEW OF SPERMATOGENESIS

Spermatogenesis in the majority of metazoa involves, asan early step, the proliferation of the germ cell populationby a species-specific number of amplifications (i.e. mitoticdivisions; Kurokawa & Hihara, 1976; Schiff, Flemming &Quicke, 2001; Scharer, Da Lage & Joly, 2008; Virkki, 1969,1973). Development is syncytial with each of the resultinggroups of ‘‘sister’’ primary spermatocytes remaining con-nected to one another by relatively large intercellular bridgesthrough which materials are shared (e.g. Braun et al., 1989).Each spermatocyte then goes through two meiotic divisionsto become four haploid spermatids. It is not until late inspermiogenesis (i.e. following elongation of the flagellum inthe case of flagellated sperm) that spermatids become fullycellularized. One of the final events in the production ofspermatozoa, even after ‘‘individualization’’ of the flagella,is for the sperm heads to separate from one another, thesebeing embedded up to this point in somatic support cells or

an extracellular matrix (for reviews of animal spermatogen-esis see, e.g. Fuller, 1993; L’Hernault, 2006; Rouse, 2006;Scheltinga & Jamieson, 2003; White-Cooper, Doggett &Ellis, 2009).

III. HAPLOID EXPRESSION

Sperm possess unique haploid genomes resulting from chro-mosomal segregation and crossing over during meiosis thatwould seemingly provide the necessary heritable varianceon which selection might act. However, the extent to whichwithin-ejaculate variation in sperm form and function isa consequence of haploid gene expression, and the tar-get of sperm-level selection, is an open question (Joseph& Kirkpatrick, 2004). Sperm phenotypes are predomi-nantly determined by testicular gene expression and hencethe diploid genome of the male (Eddy, 2002). Althoughthe general consensus was that mature spermatozoa weretranscriptionally inert (Hecht, 1998), it is now clear thatpost-meiotic gene expression occurs (reviewed by Dadoune,Siffroi & Alfonsi, 2004; Erickson, 1990; Joseph & Kirkpatrick,2004). The precise timing of expression has not, however,been determined for most of the genes required for sper-matogenesis and for sperm function. Such determination iscomplicated by the fact that, due to DNA condensation andrepackaging during spermatogenesis, some genes requiredfor spermatocyte or spermatid function will be transcribed inprimary spermatocytes, with RNA persisting to be translatedlater in sperm development (reviewed by Dorus & Karr,2009; Hecht, 1998; Kleene, 2001).

Even with post-meiotic gene expression, it is altogetherunclear how much of the phenotypic variation among spermwithin an ejaculate is attributable to allelic variation amonghaplotypes, and hence amenable to sperm-level selection(Joseph & Kirkpatrick, 2004). As described above, syncytialdevelopment with sister spermatids connected by large inter-cellular bridges through which cytoplasm can be shared iswidespread, occurring even in diploblastic basal metazoanlineages (Gaino et al., 1984). This arrangement has beenexperimentally demonstrated with mice to result in pheno-typically diploid spermatids, despite haploid expression ofgenetic differences between developing sperm cells (Braunet al., 1989; Erickson, 1990). We are aware of only twocharacteristics of sperm form or function that have beendemonstrated to exhibit natural variation resulting fromhaploid gene expression. First, the sperm adhesion molecule(Spam1) in mice is an antigen that enables sperm to penetratethe cumulus and is involved in sperm-egg binding. The geneis expressed post-meiotically and the mRNA is compartmen-talized within a developing spermatid. The resulting proteindoes not diffuse in the cytoplasm but rather is immediatelyinserted into the plasma membrane (Zheng, Deng & Martin-DeLeon, 2001a). Reduced expression of Spam1 results inreduced ability of sperm to penetrate the cumulus andfemales mated to males with Spam1-deficient sperm producesmaller than average litters (Zheng et al., 2001b). The second



example involves the well-studied t-locus meiotic drive sys-tem in mice. Sperm from males hemizygous for the t-locus(t/+) have poor motility and egg penetration, presumably ofwild-type sperm, resulting in non-Mendelian inheritance ofthe t-locus, with the t-haplotype being transmitted to up to99% of their offspring (reviewed in Lyon, 2003). The t complex

responder (Tcr) gene and its wild-type counterpart sperm motil-

ity kinase-1 (Smok1) are expressed post-meiotically. The geneproducts are retained within haploid sperm cells, resulting inphenotypic differences between t and wild-type sperm (Veronet al., 2009). It is unlikely coincidental that both examplesof heritable variation in fitness attributable to sperm haplo-type occur in mice, well-studied model organisms for geneexpression. Increasingly refined molecular techniques mightreveal additional haploid expressed and compartmentalizedgene products in other taxa.

IV. TYPES OF PUTATIVE SPERMCOOPERATION

(1) Sperm heteromorphism

There is always some, and often considerable, within-malevariation in sperm morphology (reviewed by Calhim, Immler& Birkhead, 2007; Cohen, 1973, 1975; Immler, Calhim &Birkhead, 2008; Morrow & Gage, 2001; Pitnick et al., 2009b;Ward, 1998). Baker & Bellis (1987) proposed that such vari-ation might be an adaptation to sperm competition, withdifferent sperm morphologies fulfilling different roles in anejaculate (the ‘‘kamikaze sperm’’ hypothesis). This hypothe-sis, however, has not been supported by either comparative(Harcourt, 1988, 1991; Møller, 1989) or experimental analy-ses (Moore, Martin & Birkhead, 1999), and such variation isgenerally attributed to poor quality control by the male anderrors in spermatogenesis (e.g. Cohen, 1967; Hunter & Birk-head, 2002). Nevertheless, it is clear that there is a separate,distinct phenomenon of sperm heteromorphism, in whichdifferent types of sperm perform discrete functions. Here werefer to cases where there are distinct morphological classesof sperm whose production is tightly regulated (Friedlander,1997; Schrader, 1960b). Because this phenomenon has beenthe subject of numerous reviews (Baccetti, 1972; Baccetti& Afzelius, 1976; Buckland-Nicks, 1998; Dallai, Lupetti& Mencarelli, 2006; Fain-Maurel, 1966; Friedlander, 1997;Hayakawa, 2007; Hodgson, 1997; Jamieson, 1987; Jamieson,Dallai & Afzelius, 1999; Silberglied et al., 1984; Swallow &Wilkinson, 2002; Till-Bottraud et al., 2005), we restrict our-selves here to addressing briefly a few of the most salientfeatures and conceptual implications.

Sperm heteromorphism has numerous, independent ori-gins throughout Metazoa. First discovered by von Siebold(1836) in a gastropod mollusc, sperm heteromorphismhas since been described for various species of rotifers,gastrotrichs, platyhelminths, nematodes, pogonophorans,molluscs, annelids, tardigrades, centipedes, spiders, insects,echinoderms, priapulids, hemichordates, cephalocordates,

urochordates and chordates (Till-Bottraud et al., 2005). Innearly all known cases, two distinct sperm morphologies areproduced (Swallow & Wilkinson, 2002; Till-Bottraud et al.,2005; but see e.g. Buckland-Nicks et al., 1982; Chawanji,Hodgson & Villet, 2005). The two sperm types may appearvery different from one another, or they may be of gener-ally similar appearance, differing only in total length or theproportional dimensions of their parts (e.g. head:tail ratio).Although not confirmed in all cases, it is probably true thatonly one of the sperm types (often referred to as ‘‘eusperm’’or ‘‘eupyrene’’ sperm) is ever genetically functional in eggfertilization (Till-Bottraud et al., 2005; but see Au, Reunov &Wu, 1998). The non-fertilizing sperm type (or ‘‘parasperm’’)may lack DNA (‘‘apyrene’’ sperm) or contain only a partialchromosome complement (‘‘oligopyrene’’ sperm; reviewedby Buckland-Nicks, 1998; Hayakawa, 2007). In other cases,the chromosomal complement of parasperm may be normal.Nevertheless, these sperm may not be functional in fertiliza-tion, as demonstrated for numerous sperm-heteromorphicDrosophila species (Snook & Karr, 1998; Snook, Markow &Karr, 1994).

(a) Mechanisms

Irrespective of the degree of morphological divergencebetween heteromorphic sperm, all are made in the sametestis and derived from similar spermatogonia or spermato-cytes (Healy & Jamieson, 1981; Lucas, 1971). Physiologicalaspects of the responsible spermatogenic mechanisms, how-ever, may differ substantially among taxa. Production of thedifferent sperm types may be spatially separated into differ-ent compartments within each testis, such as occurs with the‘‘harlequin lobes’’ of hemipteran insects (Schrader, 1960b).Production may be temporally separated within the sameregion of the testis, as occurs in Lepidoptera, with eupyrenesperm produced essentially prior to pupation and apyrenesperm produced thereafter, yet both sperm types derivedfrom the same bipotential spermatocytes (Friedlander, 1997).Both sperm types may be produced contemporaneously andside-by-side, albeit in separate spermatogenic cysts, as occursin some Drosophila species (Beatty & Burgoyne, 1971; Beatty& Sidhu, 1969). Finally, as with the sculpin (fish) Hemilepi-

dotus gilberti, paraspermatids and euspermatids may co-occurwithin the same cyst with intercellular bridges between them(Hayakawa, 2007; Hayakawa, Komaru & Munehara, 2002).Experimental work with moths has demonstrated an acti-vational role of hormones, with glucose and ecdysteroidinteracting in the induction of apyrene sperm production(e.g., Jans, Benz & Friedlander, 1984; Kawamura, Sahara &Fugo, 2003).

(b) Function

Decades of research has resulted in six non-mutually exclu-sive functional hypotheses for sperm heteromorphism includ-ing (1) non-adaptive errors in spermatogenesis, (2) transportor capacitation of eusperm aided by parasperm, (3) enhancedfertilization success in the presence of sperm competition,



Table 1. Summary of the different types of conjugation. See text for references and more detailed explanations of terms

Type of conjugation Examples Description Taxonomic distribution

Primary spermatodesm(-a, -ta) sperm produced by a singlespermatogonium remain groupedtogether, with the headsembedded in gelatinous material

fish, gymnolaemate bryozoan, insects,monotremes

synspermia encapsulated, syncytial sperm spidersSecondary pairs spermatozoa form pairs in the

seminal vesiclesAmerican marsupials, insects, marine

snails, millipedesrouleaux spermatozoa form orderly stacks guinea pigs, insects, naked-tail

armadillos, solifugesbundles mature spermatozoa are bound

together by extracellular materialinsects

trains sperm cling or adhere to oneanother in a disorganized manner

muroid rodents

heterospermatozeugmata heteromorphic sperm formconjugates

annelids, bivalves, gastropods, insects,polychaete and oligochaete worms,prosobranch snails

coenspermia mature, coiled spermatozoa areencapsulated

spiders

(4) protection from spermicidal environments, (5) nutrientprovisioning of females or other sperm via senescence ofparasperm, and (6) control of sex ratio. However, most of thecurrent functional hypotheses lack convincing support in anyspecies (for reviews see Dallai et al., 2006; Friedlander, 1997;Hodgson, 1997; Holman & Snook, 2006; Jamieson, Dallai& Afzelius, 1999; Swallow & Wilkinson, 2002; Till-Bottraudet al., 2005). Given the diversity of forms and numerous inde-pendent origins, it is likely that parasperm have evolved fordifferent (or multiple) reasons in different lineages. Perhapsthe one clear conclusion to be drawn from sperm hetero-morphism is that sperm may serve a variety of functions inaddition to, or before they are able to participate in, eggfertilization. Selection can either shape a single, complexsperm phenotype to perform all of these functions, or therecan be a division of labour, with discrete sperm types eachspecialized to perform different reproductive functions.

(2) Sperm conjugation

Sperm conjugation occurs when two or more sperm phys-ically unite for motility or transport through the femalereproductive tract. The sperm typically disassociate fromone another only after reaching the site of sperm storage orfertilization (but see Buckland-Nicks et al., 1999; Hayashi &Kamimura, 2002a,b for a description of sperm conjugatesdisassociating in the female’s bursa prior to movement tothe spermatheca). Similar to sperm heteromorphism, conju-gation has numerous evolutionary origins among a diversityof taxa, including annelid and polychaete worms, gastropodmolluscs, myriapods, spiders, insects and both marsupial andeutherian mammals (Pitnick et al., 2009b).

(a) Types of conjugation

Here, we briefly review the usage of terminology in theliterature to describe sperm conjugates and we recommend

stricter criteria for the application of terms. Our goal is tofacilitate future evolutionary investigations of sperm conjuga-tion, including evolutionary origins, transitions, trajectories,constraints and morphogenetic convergence. Synthesis ofsperm conjugation has been inhibited by a lack of cohesionin the literature due to the phenomenon being referred to bymany different, often taxon-specific, terms [i.e. pairs, cou-ples, conjugates, rouleaux, bundles, trains, aggregates (withspermatostyles), spermatodesm (-a, -ta), spermatozeugma(-ta) and spermatophores (Table 1)]. Apart from ‘‘pairs’’ and‘‘rouleaux’’ (meaning a stack of disc-shaped objects), theterms that have been applied are too general to ascribeto variation in biological phenomenology. The Greek rootword ‘‘zeugma’’ means yoke, and ‘‘desma’’ means bond,fetter or head-band. Hence, ‘‘spermatozeugma’’ and ‘‘sper-matodesma’’ are synonymous with one another and with allother terms used to describe sperm conjugates. For example,‘‘spermatozeugma’’ has been used to refer to the complexconjugates formed when numerous, tiny eusperm attach to agiant, highly modified parasperm in some prosobranch snails(see Section IV.2a.ii), as well as to the conjugates of someinternally fertilizing fish that involve highly variable numbersof monomorphic sperm embedded within an extracellularmatrix (Downing & Burns, 1995).

When considering developmental mechanisms underlyingdifferent forms of sperm conjugation, it is useful to dis-criminate conjugates derived from spermatogenic processes(‘‘primary conjugation’’) from those forms arising throughpost-spermatogenic mechanisms (‘‘secondary conjugation’’),as this dichotomy has important implications for variationwithin species (and within ejaculates) and as well as forevolutionary diversification among species in conjugate size(see Section V.2a). Primary conjugation is achieved throughpostponement of the disassociation of the sperm heads ofeach cyst from the material they are embedded in. Insteadof individualizing at the end of spermatogenesis, sister sperm



remain attached to one another until they reach the site ofsperm storage or fertilization within the female reproduc-tive tract. Such heterochronic evolution is not unusual withsperm. For example, numerous taxa have evolved to delaysperm maturation, transformation (i.e. capacitation) and/oractivation until they are within the female (reviewed by Pit-nick et al., 2009c). With secondary conjugation, sister spermwithin cysts undergo typical disassociation from one anotherduring spermatogenesis and then later, ‘‘downstream’’ of thetestes (e.g. within the seminal vesicles), conjugate with notnecessarily sister spermatozoa.

( i ) Primary conjugation. The term ‘‘spermatodesm(-a, -ta)’’has most frequently been used to describe conjugates formedby the products of a single spermatogonium remaininggrouped together following spermiogenesis, with their headsembedded in a cyst cell or cap of gelatinous material andthe tails free (Fig. 1B, G). Conjugation is primary, and weencourage future use of this term only when referring specif-ically to this mechanism. Such spermatodesms have beendescribed from many species of insects, including orthopter-ans (Fig. 1C, D; Baccetti, 1986), a grylloblattid (Dallai et al.,2005), a mantophasmid (Dallai et al., 2003), some hemipter-ans (Chawanji et al., 2005; Jamieson et al., 1999; Nur, 1962),several species of beetles (Fig. 1A, E, F, G; Breland & Lino-Neto et al., 2008; Simmons, 1970; Sasakawa, 2007; Takami& Sota, 2007), members of the hymenopteran suborderSymphyta (Fig. 1B; Quicke et al., 1992; Schiff et al., 2001)and several species of dragonflies (Abro, 1998). Althoughreferred to as spermatozeugmata, the conjugates of thegymnolaemate bryozoan Membranipora membranacea are alsoprobably spermatodesms, as the syncytial sperm occur inaggregates of 32 or 64 (Temkin & Bortolami, 2004), andthus are likely generated by a spermatogenic mechanism (seeSection V.2a). What appear to be spermatodesms have alsorecently been described in a monotreme: the short-beakedechidna Tachyglossus acelatus, although it has not been estab-lished that the sperm comprising each bundle are the productof a single cyst (Johnston et al., 2007).

( ii ) Secondary conjugation. All other forms of spermconjugation are secondary. Given the many independentevolutionary origins of secondary conjugation and dra-matic subsequent diversification, it is no surprise that themechanisms by which sperm become conjoined are diverse.Any categorization of this variation will be somewhat con-trived. Nevertheless, some broad categories do emerge basedon variation in mechanisms, which may include variousattributes of sperm per se and/or extracellular materials, andin the size and organization of conjugates, both of whichco-vary with the mechanisms.

Sperm ‘‘pairing’’ or ‘‘coupling’’ has referred to cases wherespermatozoa form pairs in the seminal vesicles. Sperm ofsuch species typically exhibit asymmetrical, flat-sided headsthat facilitate pairing and result in a shape apparentlyconducive to locomotion of the bi-flagellate product (e.g.Fig. 2B, C, I). In the relic silverfish, Tricholepidion gertschi, cellmembranes of mature, individualized sperm fuse to form acommon syncytium, creating a bi-nucleate and bi-flagellate

spermatozoan (Alberti, 2000; Alberti & Weinmann, 1985;Dallai et al., 2001, 2004). Sperm pairing in the firebratThermobia domestica uniquely involves the intertwining andcell adhesion of the anterior third of the length of thepaired sperm (Bawa, 1964). Other insects with sperm pairinginclude some species of burrowing water beetles (D.M.Higginson, unpublished data) and diving beetles (Auerbach,1893; Ballowitz, 1895; Dallai & Afzelius, 1985; Mackie &Walker, 1974; Werner, 1983). Pairing has been reported inmarine snails of the genus Turritella (Afzelius & Dallai, 1983),millipedes (Reger & Cooper, 1968), and the new worldopossums (Biggers & Creed, 1962; Moore, 1996; Phillips,1970; Selenka, 1887).

The term ‘‘rouleaux’’ has been used to describe the orderlystacks of sperm that form in the lumen of the cauda epi-didymides of guinea pigs and naked-tail armadillos (Fawcett& Hollenberg, 1963; Heath, Meritt & Jeyendran, 1987;Shepherd & Martan, 1979). These conjugates are organizedwith the convex surface of one paddle-shaped sperm headfitting into the concave surface of another, followed by adhe-sion of their plasma membranes (Fawcett, 1975; Fawcett &Hollenberg, 1963; Shepherd & Martan, 1979). We suggestthat this term should be extended to other taxa for whichsperm bear adaptations for orderly stacking, including somesolifuges (Fig. 2L; Alberti, 2000) and dytiscid beetles (Fig. 2G,K; D.M. Higginson & S. Pitnick, unpublished data). In thecase of dytiscids with rouleaux, sperm heads are cone-shapedwith a concave, hooded base. The rouleaux form within theseminal vesicles when the apical point of one sperm headslips into the hood of another sperm’s head, in the mannerof stacking cups (Fig. 2G).

Sperm ‘‘bundles’’ are conjugates in which head bindingis largely achieved by extracellular material (reviewed byHayashi, 1997) and does not necessarily involve any mor-phological attributes of sperm. The sperm bundles of somemegalopteran fishflies involve both mechanisms; the hookedtips of sperm heads embed in an extracellular hyaline mate-rial (Fig. 2A; Hayashi, 1996, 1997). In the majority of cases,the heads of sperm within bundles are tightly clustered withbundle shape slender and elongate in some species and nearlyspherical in others (Hayashi, 1996, 1998). The bundles ofsome homopteran cicadas, leaf hoppers and spittlebugs areuniquely organized with the sperm heads aligned along arope- or rod-like structure composed of hyaline material(Fig. 2H, J; Hayashi & Kamimura, 2002a) that are simi-lar in appearance to the spermatostyles of whirligig beetlespermatodesms (Fig. 1E).

Sperm ‘‘trains’’ form when sperm cling or adhere to oneanother in an imprecise location by a ‘‘grappling’’ structure(and may not involve any extracellular material), resultingin conjugates that are highly variable in size and exhibitinga relatively disordered arrangement (apart from all headspointing in roughly the same direction). Sperm trains areonly known to occur in several species of muroid rodents(Fisher & Hoekstra, 2010; Immler et al., 2007; Moore et al.,2002), where the associations between sperm form by aconspicuous apical hook (Fig. 2D, E).



Fig. 1. Primary conjugation. (A) Spermatodesm of the leaf beetle Xanthogaleruca sp. Merged interference contrast and fluorescenceimage of Hoechst-stained nuclei provided by R. Dallai. (B) Scanning electron micrograph (SEM) of a spermatodesm from thesawfly Arge pagana. After Lino-Neto et al. (2008). (C) Transmission electron micrograph (TEM) and (D) darkfield microscopy ofspermatodesm of the katydid Platycleis intermedia. From Viscuso et al. (1998). (E) Interference contrast and (F) SEM of spermatostylesof the whirligig beetle Dineutus assimilis. Images by D. M. Higginson and B. A. Byrnes. (G) Interference contrast image of thespermatodesm of the crawling water beetle Haliplus immacucollis. Image by D. M. Higginson. Note: images are not to scale.a = acrosome; f = flagella; m = median axis; n = nuclei.



Fig. 2. Secondary conjugation. (A) Sperm bundles of the fishfly Parachauliodes japonicus. Note the extreme variation in conjugate size.Image by F. Hayashi. (B) Phase contrast image of the paired spermatozoa of the opossum Monodelphis domestica. Image providedby H. D. M. Moore. A longitudinal transmisson electron micrograph (TEM) section through the paired heads (C) highlights theasymmetry of heads and the angle of flagellum attachment. After Moore (1996). (D) Scanning electron micrograph (SEM) of thesperm head of the wood mouse Apodemus sylvaticus showing the large apical hook. Upon ejaculation, the apical hook opens facilitatingformation of large sperm trains (E). After Moore et al. (2002). (F) Interference contrast image showing tight association of spermheads a conjugate of the diving beetle Rhantus consimilis. Image by D. M. Higginson. (G) Fluorescence image of Hoechst-stainednuclei of a large rouleaux in the diving beetle Hydroporus sp. Each annulation represents the transition between one sperm head andthe next. Sperm tails are faintly visible in the background. Image by D. M. Higginson. (H) Interference contrast image of spermattached to a hyaline rope in the leafhopper Ledra auditura. Image by F. Hayashi. (I) Darkfield microscopy of the paired sperm ofthe diving beetle Graphoderus liberus. Image by D. M. Higginson. (J) Interference contrast image of sperm attached to a hyaline rodin the spittlebug Aphrophora major. Image by F. Hayashi. (K) Darkfield microscopy of two rouleaux in the diving beetle Neoporusdimidiatus. Each conjugate is composed of eight sperm. Image by D. M. Higginson. (L) TEM longitudinal section through a rouleauxin the solifugid Eusimonia sp. nov. Sperm are plate-shaped and arranged as stacked pairs that can be found in both the testes andvas deferens. While the presence of rouleaux in the testes casts doubt on whether conjugation is secondary, we have classified itas such based on three lines of evidence. First, no stages of spermatogenesis have been observed in adults, suggesting that spermmaturation occurs before males reach reproductive maturity. Second, the pairing of sperm within the stack and third, the highlyvariable number of sperm conjugates is consistent with secondary conjugation. Image by A. E. Klann. Note: images are not to scale.



Sperm heteromorphism and conjugation sometimes co-occur. In the simplest case, Turritella spp. snails produce singleapyrene sperm and paired eusperm (Afzelius & Dallai, 1983).Similarly, the ground beetle Scarites terricola produces shortsperm that form large ‘‘sperm bundles’’ and long sperm thatremain single (Sasakawa, 2009). In other species, euspermand parasperm interact to form complex conjugates, typicallyreferred to as ‘‘spermatozeugma(-ta).’’ Often, the paraspermare specialized to aid in the transport of the eusperm (Healy &Jamieson, 1981). In some gastropods, eusperm attach to thesurface of parasperm that effectively act as transport vessels(Fig. 3C). These parasperm may be gigantic vermiform cellsor possess multiple tails (ranging from seven to hundreds; e.g.Fretter, 1953; Healy & Jamieson, 1993). Perhaps even moreunusual are the conjugates of some annelids, where hundredsof apyrene sperm join together to form hollow cylindersthat fill with loose eusperm (Fig. 3D; Braidotti, Ferraguti &Fleming, 1980; Ferraguti et al., 1989). ‘‘Spermatozeugma’’has also been applied to the monomorphic conjugates of someinternally fertilizing fish where the heads of large numbersof sperm are embedded in an extracellular matrix (Downing& Burns, 1995; Fishelson et al., 2007). The extracellularmatrix appears to be secreted by the Sertoli cells andcoheres the sperm of a single cyst (Fishelson et al., 2007,but see Grier, 1984); thus, we suggest referring to theseconjugates as spermatodesms. To mitigate ongoing confusionbetween researchers working on different taxa, Pitnick et al.(2009b) proposed a new term, ‘‘heterospermatozeugma’’to describe conjugates composed of heteromorphicsperm or sperm which adopt discrete, alternative roles.Heterospermatozeugma have been described in prosobranchsnails (Fretter, 1953; Hanson, Randall & Bayley, 1952;Nishiwaki, 1964; Woodard, 1940), bivalves (Jespersen,Kosuge & Lutzen, 2001), some polychaete and oligochaeteworms (Braidotti et al., 1980; Ferraguti, 2000; Ferragutiet al., 1989; von Nordheim, 1989) and some dytiscid beetles(Fig. 3A, B, E, F; D.M. Higginson, unpublished data).

Finally, an unusual form of sperm conjugation is observedin some arachnids. A sheath, probably secreted by epithelialcells of the male vas deferentia, surrounds mature, coiledspermatozoa such that sperm take the form of small, immotilespheres (Alberti, 1990, 2000). Sperm may be individuallyencapsulated (‘‘cleistospermia’’- most Araneomorphae:Alberti, 1990; Alberti & Colye, 1991; Alberti & Weinmann,1985; Boissin, 1973; Wu, Song & Chen, 1997) or they maybe conjugated into various-sized groups (‘‘coenospermia’’-Theraphosidae and Filistatidae: Alberti, 1990; Alberti,Afzelius & Lucas, 1986; Alberti & Weinmann, 1985; Bertkau,1877). For example, individual coenospermia of Liphistius cf.phuketensis can include more than 30 sperm (Michalik, 2007).Coenospermia may also include parasperm in addition tofertilizing eusperm (Alberti, 2005). In some taxa, at theend of spermiogenesis, spermatids fuse to form syncytialspermatozoa that are then encapsulated (‘‘synspermia’’-Scytodiade, Sicariidae, Segestriidae, Dysderidae: Alberti,2000; Alberti & Weinmann, 1985; Michalik et al., 2004).Synspermia are thus a kind of spermatodesm. Ensheathed

spermatodesms are not unique to spiders, but also havebeen described for certain insects: mealy bugs (Nur, 1962)and silverfish in the family Ateluridae (Dallai et al., 2002;Wygodzinsk, 1958).

Although the term ‘‘spermatophore’’ has been used insome cases when referring to sperm conjugates (e.g. sperma-todesms of the grasshopper Conocpehalus saltator; Cruz-Landim& Ferreira, 1977), we consider such usage incorrect and asource of confusion worth avoiding. Conjugates differ fromspermatophores, the latter being chitinous or cellular capsulessurrounding sperm aggregates and seminal fluid (typically anentire ejaculate; Davey, 1960; Jamieson, 1987). The capsulesare derived in part from male accessory gland secretions(Chapman, 1998) and, in the case of internally inseminatingspecies, may be formed into characteristic shapes within thefemale reproductive tract. Whereas it is true that ensheathedspermatodesms of some species may similarly enclose secre-tions from the male genital tract (Alberti, 2000; Dallai et al.,2002), ensheathed or encapsulated sperm are often furtherpackaged within a larger spermatophore (Alberti, 2000), asmay be other forms of conjugated sperm (e.g. Hayashi, 1996).We therefore see no justification for equating the phenomenaof sperm conjugation and spermatophore production. Thereare myriad species where males produce spermatophores butdo not exhibit sperm conjugation.

(b) Mechanisms

( i ) Primary conjugation. Spermatodesms are character-ized by a mass of extracellular material that binds togetherthe products of a single cyst. In vestimentiferan worms,spermatodesms are formed by transversal rows of microfila-ments that encircle the spermatozoa and hold them together(Marotta et al., 2005). In orthopteran insects, conjugates areformed by a ‘‘muff’’ of polysaccharides secreted by the glan-dular walls of the deferent ducts (Jamieson et al., 1999). Sper-matodesms may be further stabilized by protein interactions.In Mantophasmatodea, adherens junctions and a thin layerof extracellular material (Dallai et al., 2003) act in concertto bind the spermatozoa together. However, adherens junc-tions were not observed in the very similar spermatodesms ofGalloisiana yuasai (Grylloblattodea), although cross striationsin the intercellular space between sperm suggest that septatejunctions may be present (Dallai et al., 2005). As describedabove, the bound sperm of spermatodesms may additionallybe surrounded by a sheath arising in the testes (Nur, 1962) orfrom the vasa deferentia (Alberti, 2000; Dallai et al., 2002).

We have described spermatodesms as the product ofnormal spermatogenic mechanisms, with sperm bundlesessentially delaying the last step of individualization. Sperma-todesms may however bear additional conjugation-specificcellular adaptations in response to selection for coordinatedgroup motility (although we are not aware of any studies thatcompare the ultrastructure of the bound heads of sperma-todesms in one species with the ultrastructure of sperm headswithin a sperm cyst just prior to individualization in a closelyrelated species without sperm conjugation). For example inwhirligig beetles of the genus Dineutus, upon leaving the testes



Fig. 3. Heterospermatozeugmata. (A, B) Fluorescence image of Hoechst-stained sperm nuclei of the diving beetle Hygrotus sayi.(A) Mature, individualized sperm as they exit the testis. Two sperm morphs are clearly visible, one with filamentous-shaped nucleiand a second with cone-shaped nuclei. (B) After passage through the deferent ducts, both sperm morphs interact to form a largeconjugate composed of thousands of sperm. The tip of a cone-shaped sperm head slips into the concave portion of anothercreating the ‘‘backbone’’ of the conjugate to which the filamentous-headed sperm attach. No flagella are visible. Images by D. M.Higginson. (C) Interference contrast image of a heterospermatozeugmata of the marine snail Fusitriton oregonensis. Approximately fiftyeusperm attach to a worm-shaped parasperm. This species also produces a second, lancet-shaped parasperm (not shown). Neitherof the parasperm morphs contain chromatin. After Buckland-Nicks et al. (1982). (D) Scanning electron micrograph (SEM) of theheterospermatozeugmata of the oligochaete worm Tubifex tubifex. Parasperm are helically arranged to form a hollow cylinder thatis filled with loose eusperm. The pictured conjugate has been broken to reveal the inner cortex of eusperm. After Ferraguti et al.(1988). (E) Fluorescence image of Hoechst-stained nuclei of a partially dissociated conjugate recovered from the sperm storage organof a female diving beetle, Hygrotus picatus. (F) Intact conjugate of H. picatus. Highly motile short sperm conjugate with a second,weakly motile and much longer sperm type. Images by D. M. Higginson. Note: images not to scale. eu = eusperm; p = parasperm;s = short morph sperm; ∗ = long morph sperm.

the sperm bundles pass through a long series of slender ducts,within which the gelatinous material into which the headsare embedded becomes lengthened and hardened into along, stiff rod (a ‘‘spermatostyle’’) along which the heads areattached (Fig. 1E, F; Breland & Simmons, 1970). In othercases, such as the spermatodesms of some katydids, the spermdisplay a very precise, crystalline arrangement (see Fig. 1C).Spermatodesms may also undergo profound reorganizationafter ejaculation. For example, the spermatodesms of severalkatydid species lose their polysaccharide cap but the spermdo not dissociate. Instead, the spermatozoa become tightlylinked by their acrosomes (Viscuso, Brundo & Sottile, 2002).

( ii ) Secondary conjugation. With secondary conjugation aswell, there is the appearance of adaptations to selection fordesign conducive to efficient, coordinated motility. Therehave been detailed morphological studies of sperm pairs for

several species of opossum (Moore, 1996; Phillips, 1970;Temple-Smith & Bedford, 1980), for example, and alldescribe intimate association of the plasma membrane thatoverlies the acrosome and peripheral regions that result inwhat Moore (1996, p 606) describes as ‘‘coordinated align-ment of exquisite precision that enables spermatozoa tobehave as a single biflagellate unit’’.

Secondarily-conjugated sperm often have numerousintermembranous particles at the site of conjugation. In thecomplex heterospermatozeugmata formed by Tubifex tubifex,

large numbers of parasperm are joined together along theirflagella by adherens-like and septate-like junctions to forma cortex around a core of loose eupyrene sperm (Braidottiet al., 1980; Ferraguti et al., 1989). Intermembranous particlesmay form specialized regions (Bawa, 1964; Dallai & Afzelius,1985; Dallai et al., 2004) or be more generally distributed



(Afzelius & Dallai, 1983; Dallai & Afzelius, 1987). Oftenelectron-dense or granular material is associated with theregions of adhesion (Bawa, 1964; Dallai & Afzelius, 1987;Mackie & Walker, 1974; Werner, 1976, 1983) and theintermembrane particles may represent anchor sites for aspecialized glycocalyx (Dallai & Afzelius, 1985).

Additionally, extracellular material may be associated withsecondary conjugation. For example, sperm of the silverfishTricholepisma aurea are loosely bundled into small groupsby a granular substance (Dallai et al., 2004). Similarly, thesperm of the dytiscid beetle subfamily Colymbetinae arecemented together by an electron-dense substance that iscomposed of carbohydrate and protein (Dallai & Afzelius,1985, 1987; Mackie & Walker, 1974; Werner, 1983). Theunusual intertwined sperm pairs of the firebrat T. domestica

also show an electron-opaque substance filling the gapsbetween the distinct opposing cell membranes. Additionally,there appear to be some regions in which the membranesurrounding both spermatozoa is continuous (Bawa, 1964).

Conjugation achieved through purely biomechanicalinteractions between sperm appear to be a relatively rarephenomenon, with muroid rodents presenting the mostobvious example. In the wood mouse, Apodemus sylvaticus,the apical hook of epididymal spermatozoa is ‘‘closed’’ andattached to a peri-nuclear process (Fig. 2D; Moore et al.,2002). Upon release into the fertilization medium, the apicalhook undergoes morphological remodelling such that theapical hook ‘‘opens’’ and is released from the peri-nuclearprocess. The hooks then become entangled with other spermand the inner surface of the hooks adheres to sperm theycontact (Fig. 2E; Moore et al., 2002). In other muroid rodents,however, cell-cell adhesion appears to be absent and theapical hooks seem only to stabilize sperm groups (Immleret al., 2007). Biomechanical interactions between sperm arealso evident in the formation of rouleaux in some dytiscidbeetles (Fig. 2G, K; D.M. Higginson, unpublished data).However, given that these rouleaux can comprise thousandsof sperm, it is likely that some cell-cell adhesion also is at play.

(c) Function

Similar to sperm heteromorphism, conjugation takes manyforms, has multiple independent origins and has likelyevolved for more than one reason. Several of the (non-mutually exclusive) proposed functions of sperm conjugationoverlap with sperm heteromorphism; sperm conjugation hasbeen hypothesized to (1) facilitate sperm transport (Afzelius& Dallai, 1987; Breland & Simmons, 1970; Dallai &Afzelius, 1985), (2) enhance competitive fertilization success(Moore et al., 2002), and (3) provide physical protectionfrom spermicidal environments (Phillips, 1970). Additionally,sperm conjugation has been proposed to (4) permit molecularexchange between spermatozoa (Auerbach, 1893; Bawa,1975) and to (5) enhance egg penetration (Mackie & Walker,1974). We review the limited information pertaining to thefunction of sperm conjugation. However, in no case is thecritical biology understood sufficiently to evaluate alternative

hypotheses, resulting in a universal lack of knowledgeregarding the adaptive significance of sperm conjugation.

( i ) Motility. It is generally regarded that sperm conju-gation enhances sperm motility, an expectation grounded onsound hydrodynamic and biomechanic principles. In brief,maximum sperm velocity is proportional to the wavelengthgenerated, and wavelength depends in part on the length ofthe flagellum (Lighthill, 1976). Longer sperm are predictedto swim faster because they generate increased force withproportionately less drag (Dresdner & Katz, 1981; Dusen-bery, 2000; Hoekstra, 1984; Turner, 2006). A comparativeanalysis of 29 species of African cichlids (Fitzpatrick et al.,2009) and a quantitative genetic analysis of intraspecificvariation (Mossman et al., 2009) provide strong support forthis predicted relationship. Nevertheless, our understandingof the relationship between sperm size and motility is farfrom complete (Humphries, Evans & Simmons, 2008). Moststudies of variation in sperm size within populations havefailed to find any relationship with motility (e.g. Birkheadet al., 2005; Gage et al., 2002; reviewed by Montgomerie& Fitzpatrick, 2008; Pizzari & Parker, 2009). Also, recentmanipulative experiments have found that social factors (i.e.relative dominance of males) relevant to the competitivefertilization environment can significantly impact motilitywithout affecting sperm morphology (Cornwallis & Birk-head, 2007; Rudolfsen et al., 2006). Hence, within species,non-morphological aspects of sperm energetics [e.g. levelsof adenosine triphosphate (ATP)] appear to predict motilitybetter than does sperm size (e.g. Burness et al., 2004). More-over, sperm performance will in large part be determined byexternal conditions, and the selective environment in whichsperm migrate and compete to fertilize eggs tends to becomplex and in general is poorly understood in all specieswith internal fertilization (Pitnick et al., 2009c).

Some forms of sperm conjugation will lengthen eachmotile entity (e.g. sperm trains of mice, the spermatodesmsof whirligig beetles and the rouleaux of some diving beetles;see Figs 1E, F and 2E, G, K), whereas with some other formsall of the tails will essentially be side by side (e.g. Figs 1A, B, Gand 2B, C, F, I). Either way, the number of sperm per conju-gate may correlate with swimming velocity due to enhancedforce generation without substantially greater drag. This the-oretical expectation should hold best if the beating of spermflagella within a conjugate is synchronous, thus minimizinginterference between flagella. Indeed, studies of diverse taxahave noted that the multiple flagella within conjugates dobeat in a coordinated manner (e.g. Ferraguti et al., 1988;Hayashi, 1996, 1998; Moore & Taggart, 1995; D.M. Hig-ginson, personal observations). Such coordinated motility,in the case of spermatodesms of the bryozoan Membrani-

pora membranacea, has been observed to include spontaneousshifts between small-amplitude, large-amplitude and reversewaveforms (Temkin & Bortolami, 2004).

Although not well understood, synchronization betweenflagella in contact with one another may be a natural hydro-dynamic property of cilia and flagella, as consequence ofreduced drag (Gray, 1930; Machin, 1963; Taylor, 1951).



Recent investigations of the behaviour of bull sperm in vitro

have confirmed that, when two sperm contact one another,they synchronize their beat phase and frequency (Woolleyet al., 2009). The velocity of synchronized sperm ‘‘pairs’’was found to be significantly greater than that of eithersingle sperm prior to or subsequent to their contact andsynchronization. This effect was attributed to an increase inwave velocity of the flagella—presumed to be a consequenceof combined force generation with less than additive drag(which is the equivalent of sperm swimming in lower viscosityfluid; Woolley et al., 2009).

Unfortunately, relatively few studies have quantified motil-ity of conjugated sperm, and those that have done socollectively provide inconclusive or mixed results. Sperm ofthe firebrat, T. domestica are only motile when in conjugatedpairs, not as single spermatozoa (Bawa, 1964). Swimmingspeed of paired sperm of the marine snail Turritella communis

was examined under a range of viscosity conditions, with nodifferences from that of single sperm detected (Ishijima, Ishi-jima & Afzelius, 1999). By contrast, sperm pairs were foundto swim faster than unpaired sperm in the opossum M. domes-

tica, with the difference in motility increasing with increasingviscosity of the medium (note: fluid within the isthmic regionof the oviduct of M. domestica is highly viscous; Moore, 1996;Moore & Taggart, 1995). However, because opossum spermare morphologically adapted for pairing with asymmetricalheads (Fig. 2B, C), the significance of this comparison is notaltogether clear. Unpaired sperm tend to swim in circles andthe thrust produced by the flagellum is largely dissipated bylateral movements of the head. Head displacement is damp-ened in paired sperm, presumably due to the acquired headsymmetry, the angle of insertion of the flagella (Fig. 2C),and the reported (but not quantified) observation that ‘‘theflagellum of each paired spermatozoon always beats in equalbut opposite synchrony with its partner’’ (Moore & Taggart,1995, p. 951). If true, the mechanism underlying such coor-dinated, 180◦ out-of-phase synchrony remains a mystery.Sperm trains of European wood mice (Fig. 2E; Moore et al.,2002), deer mice Peromyscus maniculatus (Fisher & Hoekstra,2010) and Norway rats Rattus norvegicus (Immler et al., 2007)swim faster than their respective single sperm, but spermtrains of the house mouse Mus musculus do not (Immler et al.,2007). Hayashi (1998) found a significant positive relation-ship between the size of sperm bundles and swimming speedin the fishfly Parachauliodes japonicus.

( ii ) Alternative hypotheses. Sperm transfer —Sperm conju-gation may facilitate the transfer of large numbers ofspermatozoa to females and has been proposed to be aprecursor to more efficient mechanisms of transfer, such asspermatophores (Afzelius & Dallai, 1987; Breland & Sim-mons, 1970; Dallai & Afzelius, 1985). Reducing sperm lossmay be important in Bryozoa where conjugates are releasedinto the water column before entering the intertentacularorgan of a maternal individual, dissociating and migratingto the surface of the ovary (Temkin, 1994). However, thisis unlikely to be a common function of conjugation; we areunaware of conjugation occurring in externally fertilizing

species and, in contrast to Bryozoa, the vast majority ofspecies deposit conjugates in the female reproductive tract.Moreover, while the hypothesis has not been subject toevolutionary analysis, the concomitance of conjugation andspermatophores as mentioned above (Alberti, 2000; Hayashi,1998; Sasakawa, 2007; Takami & Sota, 2007) suggests thatconjugation serves a function beyond merely grouping sperm.

Sperm competition—The female reproductive tract providesthe arena of competition and intimately links femalepreference with sperm competition (Eberhard, 1996; Pitnicket al., 2009c). In addition to the potential role of conjugatemotility in sperm competitiveness, we propose that spermconjugation may produce sperm morphologies that are‘‘preferred’’ by females. Sperm size is positively correlatedto competitive fertilization success in the snail Viviparus ater

(Oppliger, Ribi & Hosken, 2003), the nematode Caenorhabditis

elegans (LaMunyon & Ward, 1999), the bulb mite Rhizoglyphus

robini (Radwan, 1996), and the fly Drosophila melanogaster

(Miller & Pitnick, 2002; Pattarini et al., 2006). In the onlystudy to examine conjugate size and female reproductivetract morphology, spermatodesm length was found to bepositively correlated with spermathecal length in 30 speciesof the ground beetle tribe Pterostichini (Sasakawa, 2007),but remains to be tested using rigorous statistical methodsaccounting for phylogeny. Sperm conjugation may be amechanism for increasing size in an energetically inexpensiveway. Sperm size has been proposed to trade-off against spermnumbers (Parker, 1982), although empirical studies haveoften failed to find a relationship (reviewed by Parker et al.,2010; Pizzari & Parker, 2009; Snook, 2005; but see Pitnick,1996). Sperm conjugation may avoid this potential trade-offand maintain high sperm numbers without the increasedmetabolic costs of producing large sperm [e.g. large testes(Pitnick, 1996) or delayed maturation (Pitnick, Markow &Spicer, 1995)].

Protection from spermicidal environments—Conjugation mayalso preserve sperm viability, to the extent that the femalereproductive tract presents an environment that is damagingto sperm (Birkhead, Møller & Sutherland, 1993; Holman &Snook, 2006). The rouleaux of guinea pigs nominally par-titions the acrosomes of all but the uppermost spermatozoain the stack from the environment (Fawcett, 1975; Friend &Fawcett, 1974; Phillips, 1970) and may thereby protect theacrosomes from degradation. A similar protective role hasbeen proposed for sperm pairing in American marsupials,where sperm are tightly apposed along the acrosomal surface(Bedford, Rodger & Breed, 1984). In fertilization medium,paired sperm of opossums maintained motility for longerperiods than did single sperm (Moore & Taggart, 1995).While intriguing, this result does not specifically addressacrosomal integrity and should be interpreted with cau-tion as the sperm heads of opossums are morphologicallyadapted for conjugation and sperm that fail to pair may bedefective in some way. In many cases, however, conjuga-tion leaves the acrosomes exposed (e.g. some diving beetles:Dallai & Afzelius, 1985; rodents: Breed, 2004; Immler et al.,2007). Moreover, acrosomes are a nearly universal feature



of animal spermatozoa, whereas conjugation is rare (Pitnicket al., 2009b). In species with sperm conjugation, there is noevidence that environmental degradation of acrosomes iscommonplace and the role of conjugation in maintenance ofacrosomal integrity has not been subjected to experimentalinvestigation. It is also plausible for some conjugate formsthat the chemical environment in the interior of conjugatesmay be significantly different from the surrounding environ-ment. In the complex heterospermatozeugma of oligochaeteworms, small molecules can pass through the cortex ofparasperm into the central, eusperm-containing core and hasbeen proposed to increase sperm survival during the longstorage period before fertilization (Ferraguti et al., 1988).However, it remains to be demonstrated that the femalereproductive tract contains nutrient molecules and that theseare of the appropriate size to permeate the conjugates.

Conversely, dissociation of conjugates has been proposedto result in the death or damage of some of the participatingsperm. In the wood mouse, sperm-train break-up is suggestedto be concomitant with premature acrosome reaction byapproximately half of the participating spermatozoa (Mooreet al., 2002). Likewise, it has been argued that the dissociationof sperm pairs in the opossum tends to result in loss of motilityof one of the participants, possibly due to disruption of the cellmembrane (Moore & Moore, 2002). Where cell-cell adhesionis important for conjugate formation and stability it seemsinevitable that sperm degradation or death will result inconjugate break-up. Additionally, theoretical models suggestthat hydrodynamically synchronized sperm clusters fall apartwhen there is substantial disparity in sperm beat frequenciesas would be the case with weakly motile, dying sperm (Yang,Elgeti & Gompper, 2008). Nonetheless, it does not necessarilyfollow that conjugate break-up results in sperm death orthat sperm mortality is a normal mechanism of conjugatedissociation. In the wood mouse, for example, it is difficult toassess the significance of a high rate of acrosome reaction inconjugated sperm without also knowing the ‘normal’ rate ofpremature acrosome reaction for single sperm in this speciesunder the experimental conditions examined. Unfortunately,in most cases, the appropriate experimental tests have notbeen undertaken to substantiate the potential associationbetween conjugate break-up and sperm mortality. At least insome species, however, conjugate break-up is not associatedwith sperm mortality; the paired sperm of the divingbeetle Graphoderus liberus undergo natural or mechanicallyinduced separation without associated sperm mortality (D.M.Higginson, K.R. Henn & S. Pitnick, in preparation).

Molecular exchange—Molecular exchange to limit variabil-ity between spermatozoa was the earliest proposed functionof conjugation (Auerbach, 1893; Bawa, 1975). As describedabove, cytoplasmic bridges between developing spermatidseffectively reduce phenotypic variation between spermatozoa(e.g. Braun et al., 1989). Thus, molecular exchange amongmembers of a conjugate would only be important if there waspost-individualization haploid expression. To date, there isno evidence of exo- or endocytosis in conjugated sperm (e.g.Dallai & Afzelius, 1987; Dallai et al., 2003; Friend & Fawcett,

1974; but see Werner, 1976). Additionally, any hypotheticalmolecular exchange between spermatozoa would be limitedto cases of intimate association between sperm heads, suchas is seen in most, but not all, forms of secondary conju-gation; the distance between primarily conjugated sperm istypically occupied by gel-like or solid extracellular materialthat would present a barrier to efficient diffusion of molecules(Amsden, 1998).

Egg penetration—Lastly, sperm conjugation has been pro-posed as a mechanism to improve sperm penetration of theegg membrane by the joint action of the multiple acrosomespresent in a conjugate or by increasing thrust force (Mackie& Walker, 1974). This hypothesis minimally requires thatsperm remain conjugated until contacting an egg. We canfind no evidence supporting a role of sperm conjugation infertilization. Where the timing of conjugate dissociation isknown, it invariably occurs before contact with an egg (opos-sum: Taggart et al., 1993; guinea pigs: Martan & Shepherd,1973; fishflies: Hayashi, 1997; snails: Buckland-Nicks et al.,1999). Furthermore, polyspermy has not been reported inany species with sperm conjugation.

V. EVOLUTIONARY PATTERNS

Discussion of the evolution of sperm heteromorphism andconjugation has often focused on theoretical analysis ofthe role of haploid expression and the origins of cooperation(Haig & Bergstrom, 1995; Immler, 2008; Kura & Nakashima,2000; Parker & Begon, 1993; Pizzari & Foster, 2008; Swal-low & Wilkinson, 2002). Given the current lack of empiricalsupport for haploid control (see Section VI.1), we advocate amore pragmatic approach combining phylogenetic hypoth-esis testing to identify selective environments (e.g. matingsystem, female reproductive tract morphology) associatedwith the origin, maintenance or loss of sperm heteromor-phism or conjugation. We further advocate renewed effortto identify the physiological and molecular mechanisms thatproduce these remarkable variations in sperm morphology.Below, we consider possible physiological origins, modifica-tions and losses of sperm heteromorphism and conjugation.

(1) Origins

(a) Sperm heteromorphism

It seems probable that heteromorphic sperm with aberrantchromosome complements originate from mutations disrupt-ing the normal course of spermatogenesis resulting in nondis-junction, nuclear elimination, or asymmetric cytokinesis. Inthe sculpin Hemilepidotus gilberti, eusperm and paraspermdevelop synchronously within a common cyst (Hayakawaet al., 2002). Spermatogenesis proceeds typically until the sec-ond meiotic division, when asymmetrical cytokinesis results intwo haploid eusperm and one diploid parasperm (Hayakawa,2007). In some cases, the asymmetrical cytokinesis that givesrise to parasperm is reminiscent of that seen during oogen-esis (Fain-Maurel, 1966). Spatial or temporal changes in the



chemical environment in which sperm develop might alsoresult in abnormal chromosome complements. The testes ofhemipteran insects are partitioned into several lobes, withone morphologically and presumably chemically distinct‘‘harlequin’’ lobe (Schrader, 1960a,b; Schrader & Leucht-enberger, 1951). During meiosis in the harlequin lobes,autosomes typically form a chain or become an indistin-guishable mass that is laterally displaced from the normalmitotic axis and fails to segregate appropriately during celldivision (Schrader, 1945, 1946a, b, 1960a). Likewise, in Lep-idoptera, changes in circulating hormones with the onset ofpupation result in a shortened prophase and the eliminationof nuclei from developing spermatids (Friedlander, 1997;Kawamura et al., 2003; Sahara & Kawamura, 2004).

The origins of size dimorphism in contemporaneouslydeveloping spermatids with normal chromosomal comple-ments are enigmatic; spermatogenesis is similar to closelyrelated sperm monomorphic species and cysts seeminglyshare a common developmental environment. Additionally,the paucity of studies of spermatogenesis in sperm-length-heteromorphic species provides few clues to the develop-mental mechanisms that result in alternative morphs. In thehandful of studies addressing spermatogenesis, there wereno clear differences in development between cysts that couldexplain the resultant long and short sperm in cicadas or fruitflies (Chawanji et al., 2005, 2006, 2007; Hauschteck-Jungen &Maurer, 1976; Kubo-Irie et al., 2003). Comparative genomicanalysis between sperm hetero- and monomorphic Drosophila

species targeting spermatogenesis-related genes and regula-tory sequences might reveal prospective variants associatedwith sperm heteromorphism. Candidate genes could be sub-jected to functional analysis to characterize their role inspermatogenesis.

(b) Conjugation

By definition, primary conjugation is the result of delayedsperm individualization and thus has most likely resultedfrom mutations delaying or suppressing expression of genesinvolved in the final stages of spermatogenesis. By contrast,secondary conjugation is achieved by several, distinctly dif-ferent means (i.e. biomechanical interactions, cell-surfaceinteractions, membrane fusion, or a combination thereof).Accordingly, it seems likely that there may be more evolu-tionary avenues that result in secondary sperm conjugationthan primary conjugation.

Some sperm structures that have evolved in the con-text of selection for enhanced fertility or competitivenessmight preadapt sperm for conjugation. The sperm headsof most species of muroid rodents have large apical hooksprimarily composed of acrosomal material (Breed, 2004,2005; Roldan, Gomendio & Vitullo, 1992), suggesting thatthe hooks evolved in the context of sperm-egg interactions.Alternatively, sperm-female interactions might have selectedfor sperm morphologies that improve sperm retention andmovement through the female reproductive tract; apicalhooks act like barbs, allowing sperm to attach to the femaleepithelium periodically and to maintain their position within

the tract, potentially reducing energy expenditure (Firman& Simmons, 2009). Nonetheless, the hooks of some speciesinteract to form sperm conjugates (see Section IV.2a.ii; Imm-ler et al., 2007; Moore et al., 2002). Although it is not clearif all sperm with apical hooks conjugate, in rodents, hooksappear to be a prerequisite for conjugation. Likewise, hoodedsperm heads in diving beetles (Jamieson et al., 1999) mightfacilitate the formation of rouleaux in the deferent ducts ofmales where sperm are densely packed, oriented in a singledirection, and motile. Hooded sperm heads are found in atleast three subfamilies, collectively accounting for more thanhalf of all diving beetle species (Dallai & Afzelius, 1985, 1987;Mackie & Walker, 1974; Werner, 1976, 1983). Currently,phylogenetic relationships within Dytiscidae are insufficientlyresolved (Miller, 2001; Ribera, Vogler & Balke, 2008) to inferif morphological features of sperm heads, such as the depthof the hood, are correlated with conjugate evolution.

Sperm pairing might have originated by cooption ofcell surface proteins involved in other cell-cell interactions.‘‘Green-beard’’ selection has been proposed to explain spermpairing in opossums (Moore & Moore, 2002). Green-beardselection for cooperation occurs when a gene, or groupof linked genes, produces a recognizable phenotype (e.g. agreen beard) and cooperates with other individuals that sharethat phenotype (Dawkins, 1976; Hamilton, 1964). Green-beard-mediated cooperation is generally considered to beevolutionarily unstable because of the breakdown of linkagedisequilibrium between the genes encoding the phenotypeand those controlling the cooperative behaviour (Lehmann& Keller, 2006). However, green-beard effects may beimportant in the evolution of secondary sperm conjugation.Moore & Moore (2002) propose that cell-adhesion moleculesmediate sperm conjugation in the opossum. In this case,a single gene might control both the green-beard andcooperation. Here, linkage disequilibrium would be absoluteand could not decay over time. As pointed out by Keller(2002), more information is required to determine if green-beard selection was important in the evolution of spermpairing. First, the molecular mechanism of conjugationshould be confirmed and it be demonstrated that onlysperm with the green-beard conjugate. Second, the fitnessbenefits (if any) must be quantified. Neither the molecularmechanism nor reproductive fitness benefits have beenstudied in opossums or any other species with sperm pairing.

(2) Modifications

(a) Conjugate size

The mechanisms by which sperm conjugate have importantselective consequences, as they may determine the extentof among-male variation and otherwise constrain the evolv-ability of conjugate size. With sperm pairing, for example,the morphological and cellular mechanisms of conjugation(see Section IV.2b.ii) are not conducive to the formation ofconjugates of more than two sperm. In the grey short-tailedopossum Monodelphis domestica, for instance, presumably allmales in the population have paired sperm, and within males



80–90% of spermatozoa in the caudal region of the epi-didymis were found to be (correctly) paired, with remainingsperm being either single or members of misaligned pairs(Moore, 1996). Because single sperm are either immotile orexhibit impaired or limited motility (see below), they tendnot to be among the sperm population competing to fertilizeeggs (e.g., Moore 1996); hence, there is effectively no additivegenetic variation in conjugate size. Where sperm pairing hasarisen, it has presumably gone to fixation or been lost.

There will similarly be limited variation in conjugate sizewhen conjugates form by primary, spermatogenic mech-anisms (i.e. spermatodesms). The number of sperm perconjugate (N ) will be determined by the formula, N = 2n × 4,with n equal to the number of pre-meiotic, mitotic divisions(also referred to as amplifications). That is, each germ cell giv-ing rise to a sperm cyst first divides binomially n times, withthe resulting primary spermatocytes then entering meio-sis to each give rise to four spermatids. [Note that somespecies are known to show intermediate or slight deviationsfrom the N = 2n × 4 formula (e.g. Nur, 1962; reviewed byScharer et al., 2008)]. The number of amplifications tends tobe species-specific and to exhibit considerable phylogeneticinertia (Kurokawa & Hihara, 1976; Schiff et al., 2001; Schareret al., 2008; Virkki, 1969, 1973). As a consequence, irrespec-tive of the species-specific number of sperm per conjugate(e.g. 16, 32, 64, 128, 256, etc.), minimal variation within andbetween males in conjugate size is expected (e.g. Nur, 1962).There is known to be some variation in the number of spermper cyst, and hence theoretically in the number of spermper spermatodesm, although the extent of such variation andits consequences have not been well explored (see Schareret al., 2008). With primary conjugation, therefore, any evo-lutionary response to selection for increased or decreasedconjugation size requires the gain or loss of an amplificationstep (and the respective doubling or halving of the number ofsperm per conjugate), which does not appear to be an evolu-tionarily labile trait. Moreover, given that a sperm conjugatehas a complex functional phenotype (e.g. architecture, flag-ellum length, beat frequency) that is presumably subject tomultivariate selection, it may be that hypothetical alternativediscrete character states (e.g. an increase from 256 to 512sperm per conjugate) will not reside on adaptive peaks.

Other forms of conjugation are more conducive to exten-sive and continuous variation in conjugate size (e.g. seeFig. 2F, G). However, no thorough qualitative, nor quantita-tive genetic, studies of variation in sperm conjugate size havebeen conducted with any species. The rouleaux of guinea pigsare reported typically to include from 2 to 14 sperm (Fawcett& Hollenberg, 1963), and those of the solifugid Eusimonia

mirabilis (see Fig. 2L for a closely related species) vary from 4to more than 50 sperm (A. Klann & G. Alberti, personal com-munication). Among rouleaux-producing species of dytiscidbeetles, some produce rouleaux of relatively invariant size(6–8 in Neoporus dimidiatus), whereas others, such as speciesof Hydroporus, produce rouleaux that vary in size from onlya few to thousands of sperm (D.M. Higginson, unpublisheddata). Similarly, coenospermia of the spider Liphistius cf.

phuketensis were observed to contain anywhere from only afew to over 30 spermatozoa (Michalik, 2007). The mech-anism of sperm train formation also inherently results inconjugates of variable size. Sperm from the vas deferens andcaudal epididymis were found to be in groups of 5–50 in theNorway rat and from 3 to 30 in the house mouse (Immleret al., 2007). In vitro, trains can grow to include thousandsof sperm (Moore et al., 2002). Studies of conjugate size her-itability, or even repeatability across ejaculates within maleswould prove valuable, as clearly would investigations of therelationship between among-male variation in conjugate sizeand (non-competitive and competitive) fertilization success.

(b) Division of labour

Understanding of the evolution of heterospermatozeugmatamay be enriched by theory developed to explain the originsof multicellularity. Heterospermatozeugmata are typically(perhaps always) composed of fertile and non-fertile sperm(see Section IV.2a.ii) and can be viewed as analogous to dif-ferentiated cell clusters with reproductive and somatic cells.Formation of simple cell aggregates is postulated to be the firststep towards multicellularity, with specialization of cell func-tion occurring secondarily (Mable & Otto, 1998; Michod,2007; Michod & Roze, 2001; Pfeiffer & Bonhoeffer, 2003);thus, one might expect conjugation to evolve before spermheteromorphism in lineages leading to heterospermatozeug-mata. However, models of the evolution of multicellularityare based on the assumption that cells behave as individ-uals and the applicability of these models to the evolutionof heterospermatozeugmata will depend on the extent thatsperm demonstrate evolutionary interests independent of themale that produced them. As both conjugation and spermheteromorphism have evolved independently in several lin-eages (see Section IV.2a.ii), it seems equally probable thatparasperm might evolve prior to conjugation in lineages withheterospermatozeugmata. Currently, incomplete knowledgeof sperm characters in basal taxa does not permit inferenceabout the sequence of evolution of heterospermatozeugmata.

(3) Losses

The developmental process of sperm morphogenesis willinfluence the evolution of sperm traits. Sperm heteromor-phism and primary conjugation that likely arose throughspermatogenic mechanisms may be more resistant to lossthan secondary conjugation. The overwhelming majority ofmutations disrupting the normal course of spermatogenesisare expected to have severe, negative fitness consequences(e.g. Cooke & Saunders, 2002; Hackstein et al., 1990; Waki-moto, Lindsley & Herrera, 2004).

Once sperm heteromorphism arises, it typically persists in alineage. Sperm heteromorphism appears to have originatedin the basal lineages of Lepidoptera and is ubiquitous inthe derived clades (Jamieson et al., 1999). Interestingly, bothmonomorphic and heteromorphic sperm have been reportedin members of Micropterygidae, the sister group to the restof Lepidoptera (Jamieson et al., 1999). It has been suggested



that sperm heteromorphism might have two independentorigins in Lepidoptera (Sonnenschein & Hauser, 1990;Swallow & Wilkinson, 2002). However, the hypothesis ofmultiple origins of sperm heteromorphism in Lepidopterahas not been tested and may not be supported by currentphylogenetic hypotheses (Weigmann, Regier & Mitter, 2002).In flies, sperm heteromorphism occurs in some Drosophila spp.(Drosophilidae) and in stalk-eyed flies (Diopsidae). Spermheteromorphism in Drosophila appears to have a single originin the obscura subgroup and is maintained in all examinedmembers of the clade (N = 17 of 41 species; Gao et al., 2007;Snook et al., 1994; Swallow & Wilkinson, 2002). Outside ofthe obscura subgroup, sperm polymorphism has only beenreported in some (but not all) populations of D. tessieri (Joly,Cariou & Lachaise, 1991). While sperm heteromorphismappears to be an evolutionarily stable trait, stalk-eyed fliesprovide an example of loss of sperm heteromorphism. Spermdimorphism is the ancestral condition of Diopsidae, butonly monomorphic sperm are found in the derived genusDiasemopsis (Presgraves, Baker & Wilkinson, 1999).

Conversely, traits that mediate secondary conjugationsuch as sperm gross morphology (Pitnick et al., 2009b) andcell surface proteins (Vacquier, 1998) are remarkably labilemaking this type of conjugation susceptible to loss or modifi-cation. Phylogenetic analyses examining sperm morphologyin rodents suggest that apical hooks that participate in con-jugate formation (see Section IV.2b.ii) are the ancestralcondition of the superfamily Muroidea and have been secon-darily lost numerous times resulting in paddle-shaped spermheads (Breed, 2004, 2005). (Note: earlier analyses using pre-vious phylogenetic hypotheses for generic and subfamilyrelationships proposed paddle-shaped sperm as the ancestraltrait; Roldan et al., 1992). Although it has not been verified,we can infer that the loss of apical hooks also results in aloss of conjugation. Unfortunately, no studies have examinedthe evolutionary trajectory of sperm conjugation for any lin-eage. Future studies examining the evolution of conjugationthrough a lineage would be informative for identifying mat-ing systems and selective environments that result in the lossor maintenance of conjugation.

VI. EVIDENCE FOR SPERM COOPERATION

Two conditions are required for the evolution of sperm coop-eration (Nowak, 2005). First, sperm must possess heritablevariation in reproductive fitness (i.e. be individuals). Second,individual sperm must be subject to natural selection. Inlight of these conditions, we consider if sperm heteromor-phism or conjugation fulfill the requirements for evolution ofcooperation.

(1) Are sperm individuals?

Sperm have unique haploid genomes resulting from segrega-tion and crossing over during meiosis that would seeminglyprovide the necessary heritable variance. The weight of

empirical evidence, however, suggests that sperm genomicdiversity is rarely reflected in sperm phenotypes due tomale control of sperm form and sharing of post-meioticallyexpressed gene products (see Section IV.1 and 2). This con-clusion may in part be due to a general lack of knowledgeof haploid gene expression and technical hurdles required todemonstrate compartmentalization of transcripts and geneproducts within developing spermatids. Nonetheless, untilevidence to the contrary exists, we contend that in generalsperm are not individuals and sperm phenotype should beconsidered an expression of the male phenotype, similar toother ejaculatory traits such as seminal proteins.

(2) Is there haploid selection?

The meiotic drive genes and alleles are notable examplesof haploid selection (reviewed in Immler, 2008; Lyon,2003; Lyttle, 1991; Taylor & Ingvarsson, 2003). However,outside of meiotic drive, there is little evidence of sperm-level selection (Clark, Dermitzakis & Civetta, 2000; Pitnick,Dobler & Hosken, 2009a), the second requirement forthe evolution of sperm cooperation. Sperm morphologygenerally exhibits rapid and dramatic evolutionarydiversification (reviewed by Pitnick et al., 2009b). Selectionon sperm form and function is known to be intense, largely asa consequence of postcopulatory sexual selection (i.e. spermcompetition and cryptic female choice; reviewed by Birkhead& Møller, 1998; Eberhard, 1996; Keller & Reeve, 1995;Pitnick et al., 2009b,c; Pizzari & Parker, 2009). The mostwidely investigated sperm attribute known to be subject topostcopulatory sexual selection is flagellum length (reviewedby Snook, 2005; Pitnick et al., 2009b; Pizzari & Parker,2009). Using crosses between discrete laboratory populationsthat had been subjected to divergent artificial selection forsperm length (i.e. long-sperm and short-sperm populations:Dobler & Hosken, 2009; Miller & Pitnick, 2002), Pitnicket al. (2009a) demonstrated with both Drosophila melanogaster

and Scathophaga stercoraria that haploid gene expression doesnot contribute to sperm length for these species. Moregenerally, by analyzing segregation ratios of offspring ofheterozygous males derived from chromosome-extractedlines of D. melanogaster representing a range of sperm-precedence phenotypes, Clark et al. (2000) demonstratedthat sperm competition success depends on the diploid malegenome and is not a property of the haploid sperm.

Moore et al. (2002) postulated that the pronounced apicalhook on the sperm head of mice, which appears to have diver-sified in response to postcopulatory sexual selection (Immleret al., 2007; but see Firman & Simmons, 2009) and is critical tothe formation of sperm trains (Moore et al., 2002), is a result ofhaploid gene expression. In support of this claim, some genesinvolved in the patterning of the hooked heads of mice havebeen shown to exhibit post-meiotic expression (Kim et al.,1989; Kleene, 2001; Xu et al., 1999). However, to establish ifthis trait might be subject to haploid selection it is necessaryto demonstrate that post-meiotic gene products responsiblefor apical hooks are not shared between syncytial spermatidsand that hook morphology influences fertilization success.



(3) Are sperm heteromorphism and conjugationcooperation?

Three lines of evidence indicate that, in the majority of cases,sperm heteromorphism is male controlled. First, the influenceof haploid gene expression on parasperm morphogenesis isprecluded in those species in which parasperm lack nuclearDNA (e.g. apyrene sperm of Lepidoptera). Second, spermheteromorphism occurs in the haploid males of Dahlbominus

fuscipennis (Hymenoptera; Lee & Wilkes, 1965) where spermare genetically identical to one another. Third, if spermmorphology were controlled by the haploid genotype of thesperm, morphological variation would be expected to occurwithin cysts rather than among cysts or testicular lobes.Instead, most of the observed intraspecific variation in spermform is observed among males with relatively little within-male variation (reviewed by Pitnick et al., 2009b). Productionof alternative sperm morphs (i.e. parasperm) is controlled bycirculating hormones (Lepidoptera, Friedlander, 1997; Janset al., 1984; Kawamura et al., 2003) or is spatially isolated fromthat of eusperm (i.e. harlequin lobes in Hemiptera: Schrader,1960b; separate cysts in Diptera: Beatty & Burgoyne, 1971).

Evidence similarly supports male control of primaryconjugation. Males produce the extracellular material orsheaths that bind sperm together (Alberti, 2000; Dallaiet al., 2002; Nur, 1962). Thus it seems likely that theseremarkable traits have evolved through male-level selectionand do not represent evolutionary cooperation, but ratherthe theoretically unproblematic, functional cooperation seenin multicellular organisms below the level of the individual.Sub-individual cooperation among cells is maintained byselection for organismal integrity via germ line sequestration(Buss, 1987), high levels of relatedness among cells (MaynardSmith & Szathmary, 1995) and policing of selfish tendencies(Boyd & Richerson, 1992; Frank, 2003).

It is less clear whether the male or the sperm controlsecondary conjugation. It is conceivable that mature spermmight exert considerable influence over the ‘‘decision’’ toconjugate and the duration of the association. Whereas thecementing materials that bind conjugates together in thediving beetle Dytiscus marginalis are derived from cells sur-rounding the developing spermatids, in the related genusRhantus the cementing material appears to be derived fromthe sperm themselves (Werner, 1976, 1983). It is unknown,however, whether production of the material is controlledby the sperm haploid genome or if it results from sharedor male-derived transcripts. Ejaculates from different malesfrequently overlap in female reproductive tracts of a promis-cuous deer mouse, Peromyscus maniculatus, where copulationswith different males can occur less than one minute apart(Dewsbury, 1985). Sperm conjugate after ejaculation and, in

vitro, conjugated more frequently with sibling sperm than withthe sperm of full-sibling littermates or of heterospecific males(Fisher & Hoekstra, 2010). This capacity for discriminationwas not found in the monogamous sister species P. polionotus;their sperm conjugated indiscriminately with regard to thedegree of relatedness (Fisher & Hoekstra, 2010). The mecha-nism that permits discrimination among sperm is unknown,

but it seems probable that it is mediated through macro-molecules on the sperm surface. If this is the case, the ‘selfrecognition’ may be sperm-specific or male-specific. Knowl-edge of the timing of gene expression and whether or not geneproducts are shared between developing spermatids woulddetermine the relative influence of the diploid male genotypeor haploid sperm genotype on conjugation in P. maniculatus.

Sperm influence over secondary conjugation could resultin conflict between male and sperm evolutionary interests(Immler, 2008; Pizzari & Foster, 2008). We expect that thelevel of male-sperm conflict, if it exists, will be inverselyrelated to the intensity of sperm competition (Parker &Begon, 1993); when sperm competition is high, both maleand individual sperm fitness may be maximized by cooper-ation. However from the sperm-level view, when there arefew or no sperm from rival males, competition between sib-ling sperm becomes an increasingly important selective forceand cooperation may become unfavourable. Competitionbetween sibling sperm may reduce male reproductive fitness,even in monogamous systems, by reducing the number offertilization-competent sperm per ejaculate (e.g. killing ofY-chromosome-bearing sperm in the case of sex chromo-some meiotic drive) or by displacing sibling sperm from thesite of storage or fertilization. Male-level selection for adap-tations that reduce intra-ejaculate competition in favour ofimproved whole-ejaculate success aligns the interests of malesand the sperm they produce. When competition between sib-ling sperm is restricted or prevented, individual sperm fitnesscan only be maximized by enhancing inter-ejaculate compet-itive success (Frank, 2003). We refer interested readers to tworecent reviews for a more detailed discussion of sperm-levelselection and male-sperm conflict (Immler, 2008; Pizzari &Foster, 2008).

Nonetheless, to date, no studies have investigated whetherthe haploid genotype of sperm influences the likelihood, orany other aspect, of conjugation. It is possible that ever-increasingly sophisticated molecular techniques will reveal arole of individual sperm in controlling conjugation. However,in accordance with the preponderance of current evidencesuggesting that sperm do not possess heritable variation infitness, it is prudent to interpret secondary conjugation asa trait that has evolved via male-level selection and not anexample of evolutionary cooperation.

VII. CONCLUSIONS

(1) Although rare, sperm heteromorphism and conju-gation are both taxonomically widespread and havemultiple, independent origins throughout Bilateria.Sperm heteromorphism and conjugation are largelyrestricted to internally fertilizing species, with sculpinfish and echinoderms providing exceptions. Addi-tionally, heterospermatozeugmata - complex spermconjugates composed of more than one sperm type- have evolved in multiple, phylogenetically distinct



lineages, suggesting that sperm conjugation and het-eromorphism might have complementary functions.

(2) Sperm conjugation may arise from spermatogenicmechanisms (primary) or from post-spermatogenicmechanisms (secondary). Secondary conjugation isexpected to have greater evolutionary lability andpermits formation of more complex conjugates (i.e.heterospermatozeugmata).

(3) A multitude of ultrastructure studies have provideddetailed understanding of conjugate morphology, yetour understanding of the functional or evolutionarysignificance of this morphology is rudimentary. Manyhypotheses for sperm conjugation have been advancedbut few have been tested and none have convincingempirical support.

(4) Sperm heteromorphism and primary conjugationappear to be male controlled and, in theabsence of evidence of sperm individuality, spermheteromorphism and primary conjugation should notbe interpreted as evolutionary cooperation but ratheras male-selected traits. Secondary conjugation mightbe influenced by the sperm haploid genome, but nostudies have examined haploid contribution to theconjugation phenotype. Barring evidence of sperm-level selection for conjugation, we suggest the useof the moniker ‘sperm cooperation’ to be restrictedto discussions of functional interactions among spermand be avoided when discussing the evolution of spermphenotype.

VIII. ACKNOWLEDGEMENTS

We would like to thank Mollie Manier, Tom Starmer, AlUy and two anonymous reviewers for helpful comments onthis paper and David Woolley for stimulating discussionsof synchronized swimming by sperm. Additionally, weare indebted to John Buckland-Nicks, Bridget A. Byrnes,Romano Dallai, Marco Ferraguti, Fumio Hayashi, AnjaE. Klann, Harry D. M. Moore, and Renata Viscuso forpermitting us to reproduce or use unpublished photographs.This work was supported by the National Science Foundation(DDIG-0910049 to D.M.H and S.P.; DEB-0814732 to S.P.)and the Natural Sciences and Engineering Research Council(PGS-D 331458 to D.M.H.).

IX. REFERENCES

Abro, A. (1998). Structure and development of sperm bundles in the dragonfly Aeshna

juncea L. (Odonata). Journal of Morphology 235, 239–247.Afzelius, B. A. & Dallai, R. (1983). The paired spermatozoa of the marine snail,

Turritella communis Lamark (Molluscoa, Mesogastropoda). Journal of Ultrastructure

research 85, 311–319.Afzelius, B. A. & Dallai, R. (1987). Conjugated spermatozoa. In New Horizons in

Sperm Cell Research (ed. H. Mohri), pp. 349–355. Japan Science Society, Tokyo.Alberti, G. (1990). Comparative spermatology of Araneae. Acta Zoologica Fenn. 190,

17–34.

Alberti, G. (2000). Chelicerata. In Reproductive biology of invertebrates, vol. 9, Part B.Progress in male gamete ultrastructure (ed. B. G. M. Jamieson), pp. 311–388. JohnWiley and Sons, Ltd., New York.

Alberti, G. (2005). Double spermatogenesis in Chelicerata. Journal of Morphology 266,281–297.

Alberti, G., Afzelius, B. A. & Lucas, S. M. (1986). Ultrastructure of spermatozoaand spermatogenesis in bird spiders (Theraphosidae, Mygalomorphae, Araneae).Journal of Submicroscopic Cytology 18.

Alberti, G. & Colye, R. A. (1991). Ultrastructure of the primary male genital system,spermatozoa and spermiogenesis of Hypochilus pococki (Araneae, Hypochilidae).Journal of Acarchnology 19, 136–149.

Alberti, G. & Weinmann, C. (1985). Fine structure of spermatozoa of somelabidognath spiders (Filistatidae, Segestriidae, Dysderidae, Oonopidae, Scytodidae,Pholcidae; araneae’ Arachnida) with remarks on spermiogenesis. Journal of Morphology

185, 1–35.Amsden, B. (1998). Solute diffusion within hydrogels. Mechanisms and models.

Macromolecules 31, 8382–8395.Au, D. W.-t., Reunov, A. A. & Wu, R. S.-s. (1998). Four lines of spermatid

development and dimorphic spermatozoa in the sea urchin Anthocidaris crassispina

(Echinodermata, Echinoida). Zoomorphology 118, 159–168.Auerbach, L. (1893). Remarkable behaviour of the spermatozoa of Dytiscus marginalis.

Journal of the Royal Microscopical Society, Series 2 13, 622.Baccetti, B. (1972). Insect sperm cells. Advances in Insect Physiology 9, 315–397.Baccetti, B. (1986). Spermatozoa and phylogeny in orthopteroid insects. In Biology of

Orthopteroid Insects (ed. B. Baccetti), pp. 12–112. John Wiley and Sons, New York.Baccetti, B. & Afzelius, B. A. (1976). The biology of the sperm cell. In Monographs

in Developmental Biology, vol. 10 (ed. A. Wolsky). Karger, Basel.Baker, R. R. & Bellis, M. A. (1987). ‘Kamikaze’ sperm in mammals? Animal Behaviour

36, 936–938.Ballowitz, E. (1895). Die dopplespermatozoen der Dyticiden. Zeitschrift fur

Wissenschaftliche Zoologie 60, 458–499.Bawa, S. R. (1964). Electron microscope study of spermiogenesis in a fire-brat insect,

Thermobia domestica Pack. The Journal of Cell Biology 23, 431–445.Bawa, S. R. (1975). Joined spermatozoa. In The Functional Anatomy of the Spermatozoon

(ed. B. A. Afzelius), pp. 259–266. Pergamon Press, Oxford.Beatty, R. A. & Burgoyne, P. S. (1971). Size classes of the head and flagellum of

Drosophila spermatozoa. Cytogenetics 10, 1770189.Beatty, R. A. & Sidhu, N. S. (1969). Polymegaly of spermatozoan length and its

genetic control in Drosophila species. Proceedings of the Royal Society of Edinburgh B 71b,14–29.

Bedford, J. M., Rodger, J. C. & Breed, W. G. (1984). Why so many mammalianspermatozoa - a clue from Marsupials? Proceedings of the Royal Society B: Biological

Sciences 221, 221–233.Bertkau, P. (1877). Uber die ubertragungsorgane und die spermatozoen der spinnen.

Verhandlungen des Naturhistorischen Vereins preussisher Rheinlande 34, 28–32.Biggers, J. D. & Creed, R. F. S. (1962). Conjugate spermatozoa of the North

American opossum. Nature 196, 1112–1113.Birkhead, T. R. & Møller, A. P. (1998). Sperm Competition and Sexual Selection.

Academic Press, London.Birkhead, T. R., Møller, A. P. & Sutherland, W. J. (1993). Why do females

make it so difficult for males to fertilize their eggs? Journal of Theoretical Biology 161,51–60.

Birkhead, T. R., Pellatt, E. J., Brekke, P., Yeates, R. & Castillo-Juarez, H.(2005). Genetic effects on sperm design in the zebra finch. Nature 434, 383–387.

Boissin, L. (1973). Etude ultrastructurale de la spermiogenes de Meta bourneti Simon(Arachinides, Araneides, Metinae). Comptes rendus de la 2e runion d’archnologie d’expression

fancaise, Montpellier, 7–22.Boyd, R. & Richerson, P. J. (1992). Punishment allows the evolution of cooperation

(or anything else) in sizable groups. Ethology and Sociobiology 13, 171–195.Braidotti, P., Ferraguti, M. & Fleming, T. P. (1980). Cell junctions between

spermatozoa flagella within the spermatozeugmata of Tubifex tubifex (Annelida:Oligochaeta). Journal of Ultrastructure Research 73, 299–309.

Braun, R. E., Behringer, R. R., Peschon, J. J., Brinster, R. L. & Palmiter,R. D. (1989). Genentically haploid spermatids are phenotypically diploid. Nature

337, 373–376.Breed, W. G. (2004). The spermatozoon of Eurasian murine rodents: its

morphological diversity and evolution. Journal of Morphology 261, 52–69.Breed, W. G. (2005). Evolution of the spermatozoon in muroid rodents. Journal of

Morphology 265, 271–290.Breland, O. P. & Simmons, E. (1970). Preliminary studies of the spermatozoa and

the male reproductive system of some whirligig beetles (Coleoptera: Gyrinidae).Entomological News 81, 101–110.

Buckland-Nicks, J. (1998). Prosobranch parasperm: sterile cells germ cells thatpromote paternity? Micron 29, 267–280.

Buckland-Nicks, J., Bryson, I., Hart, L. & Partridge, L. (1999). Sex and asnail’s sperm: on the transport, storage and fate of dimorphic sperm in Littorinidae.Invertebrate Reproduction and Development 36, 145–152.



Buckland-Nicks, J.,Williams, D., Chia, F.-S. & Fontaine, A. (1982). The finestructure of the polymorphic spermatozoa Fusitriton oregonensis (Mollusca: Gastropoda)with notes on the cytochemistry of the internal secretions. Cell and Tissue Research

227, 235–255.Burness, G. P., Casselman, S. J., Schulte-Hostedde, A. I., Moyes, C. D. &

Montgomerie, R. (2004). Sperm swimming speed and energetics vary with spermcompetition risk in bluegill (Lepomis macrochirus). Behavioral Ecology and Sociobiology 56,65–70.

Buss, L. W. (1987).The evolution of individuality. Princeton University Press, Princeton,New Jersey.

Calhim, S., Immler, S. & Birkhead, T. R. (2007). Postcopulatory sexual selectionis associated with reduced variation in sperm morphology. PloS One 2, e413.

Chapman, R. F. (1998).The insects: structure and function. Cambridge University Press,Cambridge.

Chawanji, A. S., Hodgson, A. N. & Villet, M. H. (2005). Sperm morphology infour species of African platypleurine cicadas (Hemiptera: Cicadomorpha: Cicadidae).Tissue & Cell 37, 257–267.

Chawanji, A. S., Hodgson, A. N. & Villet, M. H. (2006). Sperm morphology infive species of cicadettine cicadas (Hemiptera: Cicadomorpha: Cicadidae). Tissue &

Cell 38, 373–388.Chawanji, A. S., Hodgson, A. N., Villet, M. H., Sanborn, A. F. & Phillips,

P. K. (2007). Spermiogenesis in three species of cicadas (Hemiptera: Cicadidae). Acta

Zoologica 88, 337–348.Clark, A. G., Dermitzakis, E. T. & Civetta, A. (2000). Nontransitivity of sperm

precedence in Drosophila. Evolution 54, 1030–1035.Cohen, J. (1967). Correlation between chiasma frequency and sperm redundancy.

Nature 215, 862–863.Cohen, J. (1973). Cross-overs, sperm redundancy and their close association. Heredity

31, 408–413.Cohen, J. (1975).Gamete redundancy - wastage or selection? In Gamete Competition in

Plants and Animals (ed. D. L. Mulcahy), pp. 99–112, Amsterdam.Cooke, H. J. & Saunders, P. T. K. (2002). Mouse models of male infertility. Nature

Reviews Genetics 3, 790–801.Cornwallis, C. K. & Birkhead, T. R. (2007). Changes in sperm quality and

numbers in response to experimental manipulation of male social status andfemale attractiveness. The American Naturalist 170, 758–770.

Cruz-Landim, C. D. & Ferreira, A. (1977). Spermatophore formation inConocephalus saltator (Saussure). (Orthoptera: Conocephalidae). International Journal

of Insect Morphology and Embryology 6, 97–104.Dadoune, J.-P., Siffroi, J.-P. & Alfonsi, M.-F. (2004). Transcription in haploid

male germ-cells. International Journal of Cytology 237, 1–55.Dallai, R. & Afzelius, B. A. (1985). Membrane specializations in the paired

spermatozoa of dytiscid water beetles. Tissue & Cell 17, 561–572.Dallai, R. & Afzelius, B. A. (1987). Sperm ultrastructure in the water beetles

(Insecta, Coleoptera). Bollettino di Zoologia 54, 301–306.Dallai, R., Carapelli, A., Nardi, F., Fanciulli, P. P., Lupetti, P., Afzelius,

B. A. & Frati, F. (2004). Sperm structure and spermiogenesis in Coletinia sp.(Nicoletiidae, Zygentoma, Insecta) with a comparative analysis of sperm structurein Zygentoma. Tissue & Cell 36, 233–244.

Dallai, R., Frati, F., Lupetti, P. & Adis, J. (2003). Sperm ultrastructure ofMantophasma zephyra (Insecta, Mantophasmatodea). Zoomorphology 122, 67–76.

Dallai, R., Lupetti, P., Carapelli, A., Frati, F. & Afzelius, B. A. (2002). Spermstructure and spermiogenesis in Atelura formicaria Heyden (Zygentoma, Insecta). Acta

Zoologica 83, 245–262.Dallai, R., Lupetti, P., Frati, F., Nardi, F. & Afzelius, B. A. (2001). Binucleate

and biflagellate spermatozoa in Tricholepidion gertschi Wygodzynsky (Insecta,Zygentoma). Tissue & Cell 33, 606–613.

Dallai, R., Lupetti, P. & Mencarelli, C. (2006). Unusual axonemes of hexapodspermatozoa. Int. Rev. Cytol. 254, 45–99.

Dallai, R., Ryuichiro, M., Uchifune, T., Lupetti, P. & Frati, F. (2005). Thesperm structure of Galloisiana yuasai (Insecta, Grylloblattodea) and implications forthe phylogenetic position of Gyrlloblattodea. Zoomorphology 124, 205–212.

Davey, K. G. (1960).The evolution of spermatophores in insects. Proceedings of the Royal

Entomological Society of London 35, 107–113.Dawkins, R. (1976). The selfish gene. Oxford University Press, Oxford.Dewsbury, D. A. (1985). Interactions between males and their sperm during multi-

male copulatory episodes of deer mice (Peromyscus maniculatus). Animal Behaviour 33,1266–1274.

Dobler, R. & Hosken, D. J. (2009). Response to selection and realized heritabilityof sperm length in the yellow dung fly (Scatophaga stercoraria). Heredity.

Dorus, S. & Karr, T. L. (2009).Sperm proteomics and genomics. In Sperm biology,

an evolutionary perspective (ed. T. R. Birkhead, D. J. Hosken and S. Pitnick),pp. 433–459. Elsevier, London.

Downing, A. L. & Burns, J. R. (1995). Testis morphology and spermatozeugmaformation in 3 genera of viviparous halfbeaks - Nomorhamphus, Dermogenys andHemirhamphodon (Teleostei, Hemiramphidae). Journal of Morphology 225, 329–343.

Dresdner, R. D. & Katz, D. F. (1981). Relationships of mammalian sperm motilityand morphology to hydrodynamic aspects of cell function. Biology of Reproduction 25,920–930.

Dusenbery, D. B. (2000). Selection for high gamete encounter rates explains thesuccess of male and female mating types. Journal of Theoretical Biology 202, 1–10.

Eberhard, W. G. (1996). Female control: sexual selection by cryptic female choice.Princeton University Press, Princeton, NJ.

Eddy, E. M. (2002). Male germ cell gene expression. Recent Progress in Hormone Research

57, 103–128.Erickson, R. P. (1990). Post-meiotic gene expression. Trends in Genetics 6, 264–269.Fain-Maurel, F.-A. (1966). Aquisitions recentes sur les spermatogeneses atypiques.

Annee Biologique, ser 4–5, 513–564.Fawcett, D. W. (1975). The mammalian spermatozoon. Developmental Biology 44,

394–436.Fawcett, D. W. & Hollenberg, R. (1963). Changes in the acrosome of guinea

pig spermatozoa during passage through the epididymis. Zeitschrift fur Mikroskopische

Anatomie 60, 276–292.Ferraguti, M. (2000). Euclitellata. In Reproductive biology of invertebrates, vol. IX. Part

B. Progress in male gamete ultrastructure and phylogeny (ed. B. G. M. Jamieson),pp. 125–182. John Wiley & Sons, Ltd, New York.

Ferraguti, M., Bernardini, G., Melone, G. & Dallai, R. (1988). Structure andfunction of the metachronal wave in Tubifex tubifex spermatozeugmata (Annelida,Oligochaeta). Journal of Ultrastructure and Molecular Structure Research 99, 79–95.

Ferraguti, M., Grassi, G. & Erseus, C. (1989). Different models of tubificidspermatozeugmata. Hydrobiologia 180, 73–82.

Firman, R. C. & Simmons, L. W. (2009). Sperm competition and the evolution ofthe sperm hook in house mice. Journal of Evolutionary Biology 22, 2505–2511.

Fishelson, L., Gon, O., Holdengreber, V. & Delarea, Y. (2007). Comparativespermatogenesis, spermatocytogenesis, and spermatozeugmata formation in malesof viviparous species of Clinid fishes (Teleostei: Clinidae, Blennioidei). The Anatomical

Record 290, 311–323.Fisher, H. S., & Hoekstra, H. E. (2010). Competition drives cooperation among

closely related sperm of deer mice. Nature 463, 801–803.Fitzpatrick, J. L., Montgomerie, R., Desjardins, J. K., Stiver, K. A.,

Kolm, N. & Balshine, S. (2009). Female promiscuity promotes the evolutionof faster sperm in cichlid fishes. Proceedings of the National Academy of Science 106,1128–1132.

Frank, S. A. (2003). Repression of competition and the evolution of cooperation.Evolution 57, 693–705.

Fretter, V. (1953).The transference of sperm from male to female prosobranch, withreference also, to the pyramidellids. Proceedings of the Linnean Society of London 164,217–224.

Friedlander, M. (1997). Control of the eupyrene-apyrene sperm dimorphism inLepidoptera. Journal of Insect Physiology 43, 1085–1092.

Friend, D. S. & Fawcett, D. W. (1974). Membrane differentiations in freeze-fractured mammalian sperm. Journal of Cell Biology 63, 641–664.

Fuller, M. T. (1993). Spermatogenesis. In The development of Drosophila melanogaster, vol.Vol. I (ed. M. Bate and A. M. Arias), pp. 71–147. Cold Spring Harbor LaboratoryPress.

Gage, M. J. G., Macfarlane, C. P., Yeates, S., Schackleton, R. & Parker,G. A. (2002). Relationships between sperm morphometry and sperm motility inAtlantic salmon. Journal of Fish Biology 61, 1528–1539.

Gaino, E., Burlando, B., Zunino, L., Pansini, M. & Buffa, P. (1984). Origin ofmale gametes in Spongia officinalis (Porifera, Demospongiae). International Journal of

Invertebrate Reproduction 7, 83–93.Gao, J.-j., Watabe, H.-a., Aotsuka, t., Pang, J.-f. & Zhang,Y.-p. (2007).

Molecular phylogeny of the Drosophila obscura species group, with emphasis onthe Old World species. BMC Evolutionary Biology 7, 87.

Gray, J. (1930). The mechanism of ciliary movement. VI. Photographic andstroboscopic analysis of ciliary movement. Proceedings of the Royal Society of London

B 107, 313–332.Grier, H. J. (1984). Testis structure and formation of spermatophores in the

atherinimorph teleost Heraichtyhys setnai. Copeia, 833–839.Hackstein, J. H. P., Beck, H., Hochstenbach, R., Kremer, H. & Zacharias, H.

(1990). Spermatogensis in Drosophila hydei: a genetic survey. Roux’s Archives of develop-

mental biology 199, 251–280.Haig, D. & Bergstrom, C. T. (1995). Multiple mating, sperm competition and

meiotic drive. Journal of Evolutionary Biology 8, 265–282.Hamilton, W. D. (1964). The genetical evolution of social behaviour. I. Journal of

Theoretical Biology 7, 1–16.Hanson, J., Randall, J. T. & Bayley, S. T. (1952). The microstructure of the

spermatozoa of the snail Viviparus. Experimental Cell Research 3, 65–78.Harcourt, A. H. (1988). Deformed sperm are probably not adaptive. Animal Behaviour

37, 863–864.Harcourt, A. H. (1991). Do mammals produce adaptive non-fertilizing sperm? In

Comparative Spermatology 20 Years After, vol. 75 (ed. B. Baccetti), pp. 765–770. RavenPress, New York.



Hauschteck-Jungen, E. & Maurer, B. (1976). Sperm dysfunction in sex ratiomales of Drosophila subobscura. Genetica 46, 459–477.

Hayakawa, Y. (2007). Parasperm: morphological and functional studies on nonfertilesperm. Icthyological Research 54, 111–130.

Hayakawa, Y., Komaru, A. & Munehara, H. (2002). Ultrastructural observationsof eu- and paraspermiogenesis in the cottid fish Hemilepidotus gilberti (Teleostei:Scorpaeniformes: Cottidae). Journal of Morphology 253, 243–254.

Hayashi, F. (1996). Insemination through an externally attached spermatophore:Bundled sperm and post-copulatory mate guarding by male fishflies (Megaloptera:Corydalidae). Journal of Insect Physiology 42, 859–866.

Hayashi, F. (1997). A trypsin-degradable protein agglutinates fishfly sperm-bundles.International Journal of Insect Morphology and Embryology 26, 63–66.

Hayashi, F. (1998). Sperm co-operation in the fishfly, Parachauliodes japonicus. Functional

Ecology 12, 347–350.Hayashi, F. & Kamimura, Y. (2002a). Sperm-bundles of Cicadomorpha insects

(Homoptera). Natural Environmental Science Research 15, 9–15.Hayashi, F. & Kamimura, Y. (2002b). The potential for incorporation of male

derived proteins into developing eggs in the leafhopper Bothrogonia ferruginea. Journal

of Insect Physiology 48, 153–159.Healy, J. M. & Jamieson, B. G. M. (1981). An ultrastructural examination of

developing and mature paraspermatozoa in Pyrazus ebeninus (Mollusca, Gastropoda,Potamididae). Zoomorphology 98, 101–119.

Healy, J. M. & Jamieson, B. G. M. (1993). Euspermatozoa, paraspermatozoaand spermatozeugmata of Littoraria (Palustorina) articulata (Prosobranchia:Caenogastropoda) with special reference to the Pseudotrich. Acta Zoologica 74,321–330.

Heath, E., Meritt, D. A. & Jeyendran, R. S. (1987). Rouleaux formation byspermatozoa in the naked-tail armadillo, Cabassosus unicinctus. Journal of Reproductive

Fertility 79, 153–159.Hecht, N. B. (1998). Molecular mechanisms of male germ cell differentiation.

BioEssays 20, 555–561.Hodgson, A. N. (1997). Paraspermatogenesis in gastropod molluscs. Invertebrate

Reproduction and Development 31, 31–38.Hoekstra, R. F. (1984). Evolution of gamete motility differences. II. Interaction with

the evolution of anisogamy. Journal of Theoretical Biology 107, 71–83.Holman, L. & Snook, R. R. (2006). Spermicide, cryptic female choice and the

evolution of sperm form and function. Journal of Evolutionary Biology 19, 1660–1670.Humphries, S., Evans, J. P. & Simmons, L. W. (2008). Sperm competition: linking

form to function. BMC Evolutionary Biology 8, 319.Hunter, F. M. & Birkhead, T. R. (2002). Sperm viability and sperm competition in

insects. Current Biology 12, 121–123.Immler, S. (2008). Sperm competition and sperm cooperation: the potential role of

diploid and haploid expression. Reproduction 135, 275–283.Immler, S., Calhim, S. & Birkhead, T. R. (2008). Increased postcopulatory sexual

selection reduces the intramale variation in sperm design. Evolution 62, 1538–1543.Immler, S., Moore, H. D. M., Breed, W. G. & Birkhead, T. R. (2007). By hook

or by crook? Morphometry, competition and cooperation in rodent sperm. PloS One

2, e170.Ishijima, S., Ishijima, S. A. & Afzelius, B. A. (1999). Movement of Turritella

spermatozoa: direction of propagation and chirality of flagellar bends. Cell Motility

and the Cytoskeleton 44, 85–95.Jamieson, B. G. M. (1987). The ultrastructure and phylogeny of insect spermatozoa. Cambridge

University Press, Cambridge.Jamieson, B. G. M., Dallai, R. & Afzelius, B. A. (1999). Insects: their spermatozoa and

phylogeny. Science Publishers, Inc., Enfield.Jans, P., Benz, G. & Friedlander, M. (1984). Apyrene-spermatogenesis-inducing

factor is present in the haemolymph of male and female pupae of the codling moth.Journal of Insect Physiology 30, 495–497.

Jespersen, A ., Kosuge, T. & Lutzen, J. (2001). Sperm dimorphism andspermatozeugmata in the commensal bivavle Pseudopythina macrophthalmensis

(Galeommatoidea, Kelliidae). Zoomorphology 120, 177–189.Johnston, S. D., Smith, B., Pyne, M., Stenzel, D. & Holt, W. V. (2007). One-

sided ejaculation of echidna sperm bundles. The American Naturalist 170, E162–E164.Joly, D., Cariou, M. L. & Lachaise, D. (1991). Can sperm competition explain

sperm polymorphism in Drosophila teissieri? Evolucion Biologica 5, 25–44.Joseph, S. B. & Kirkpatrick, M. (2004). Haploid selection in animals. Trends in

Ecology and Evolution 19, 592–597.Kawamura, N., Sahara, K. & Fugo, H. (2003). Glucose and ecdysteroid increase

apyrene sperm production in in vitro cultivation of spermatocysts of Bombyx mori.Journal of Insect Physiology 49, 25–30.

Keller, L. (2002). Marsupial sperm pairing: a case of ‘sticky’ green beards? Trends in

Ecology and Evolution 17, 113.Keller, L & Reeve, H. K. (1995). Why do females mate with multiple males? The

sexually selected sperm hypothesis. Advances in the Study of Behavior 24, 291–315.Kim, E., Waters, S. H., Hake, L. E. & Hecht, N. B. (1989). Identification and

developmental expression of smooth muscle gamma-actin in post-meiotic malegerm cells of mice. Molecular and Cellular Biology 9, 1875–1881.

Kleene, K. C. (2001). A possible meiotic function of the peculiar patterns of geneexpression in mammalian spermatogenic cells. Mechanisms of Development 106, 3–23.

Kubo-Irie, M., Irie, M., Nakazawa, T. & Mohri, H. (2003). Ultrastructure andfunction of long and short sperm in Cicadidae (Hemiptera). Journal of Insect Physiology

49, 983–991.Kura, T. & Nakashima, Y. (2000). Conditions for the evolution of soldier sperm

classes. Evolution 54, 72–80.Kurokawa, H. & Hihara, F. (1976). Number of first spermatocytes in relation

to phylogeny of Drosophila (Diptera: Drosophilidae). International Journal of Insect

Morphology and Embryology 5, 51–63.LaMunyon, C. W. & Ward, S. (1999). Evolution of sperm size in nematodes: sperm

competition favours larger sperm. Proceedings of the Royal Society B: Biological Sciences

266, 263–267.Lee, P. E. & Wilkes, A. (1965). Polymorphic spermatozoa in the hymenopterous

wasp Dahlbominus. Science 147, 1445–1446.Lehmann, L. & Keller, L. (2006). The evolution of cooperation and altruism - a

general framework and a classification of models. Journal of Evolutionary Biology 19,1365–1376.

L’Hernault, S. (2006). Spermatogenesis. In Wormbook (ed. T. C. E. R. Community),pp. 1–14.

Lighthill, J. (1976). Flagellar hydrodynamics. SIAM Review 18, 161–230.Lino-Neto, J., Dolder, H., Mancini, K., Mercati, D. & Dallai, R. (2008). The

short spermatodesm of Arge pagana (Hymenoptera: Symphyta). Tissue & Cell 40,185–193.

Lucas, A. (1971). Les gametes des mollusques. Haliotis 1, 185–214.Lyon, M. F. (2003). Transmission ratio distortion in mice. Annual Review of Genetics 37,

393–408.Lyttle, T. W. (1991). Segregation distorters. Annual Review of Genetics 25, 511–557.Mable, B. K. & Otto, S. P. (1998). The evolution of life cycles with haploid and

diploid phases. BioEssays 20, 453–462.Machin, K. E. (1963). The control and synchronization of flagellar movement.

Proceedings of the Royal Society of London B 158, 88–104.Mackie, J. B. & Walker, M. H. (1974). A study of the conjugate sperm of the dytiscid

water beetles Dytiscus marginalis and Colymbetes fuscus. Cell Tissue Research 148, 505–519.Marotta, R., Melone, G., Bright, M. & Ferraguti, M. (2005). Spermatozoa

and sperm aggregates in the Vestimentiferan Lamellibrachia luymesi compared withthose Riftia pachyptila (Polychaeta: Siboglinidae: Vestimentifera). Biological Bulletin

209, 215–266.Martan, J. & Shepherd, B. A. (1973). Spermatozoa in rouleaux in the female guinea

pig genital tract. The Anatomical Record 175, 625–630.Maynard Smith, J. & Szathmary, E. (1995).The major transitions in evolution. W. H.

Freeman and Company Limited, New York, NY.Michalik, P. (2007). Spermatozoa and spermiogenesis of Liphistius cf. phuketensis

(Mesothelae, Araneae, Arachnida) with notes on phylogenetic implications. Arthropod

Structure & Development 36, 327–335.Michalik, P., Dallai, R., Giusti, F. & Alberti, G. (2004). The ultrastructure of

the peculia synspermia of some Dysderidae (Araneae, Arachnida). Tissue & Cell 36,447–460.

Michod, R. E. (1999). Darwinian dynamics: evolutionary transitions in fitness and individuality.

Princeton University Press, Princeton, New Jersey.Michod, R. E. (2007). Evolution of individuality during transition from unicellular to

multicellular life. Proceedings of the National Academy of Science 104, 8613–8618.Michod, R. E. & Roze, D. (2001). Cooperation and conflict in the evolution of

multicellularity. Heredity 81, 1–7.Miller, G. T. & Pitnick, S. (2002). Sperm-female coevolution in Drosophila. Science

298, 1230–1233.Miller, K. B. (2001). On the phylogeny of the Dytiscidae (Insecta: Coleoptera) with

emphasis on the morphology of the female reproductive system. Insect Systematics &

Evolution 31, 45–92.Møller, A. P. (1989). Ejaculate quality, testes size and sperm production in mammals.

Functional Ecology 3, 91–96.Montgomerie, R. & Fitzpatrick, J. L. (2008). Testes, sperm, and sperm

competition. In Reproductive biology and phylogeny of fishes (ed. B. G. M. Jamieson).Science Publishers, Enfield, NH.

Moore, H. D. (1996). Gamete biology of the New World marsupial, the grey short-tailed opossum, Monodelphis domestica. Reproduction, Fertility and Development 8, 605–615.

Moore, H. D., Dvorakova, K., Jenkins, N. & Breed, W. (2002). Exceptionalsperm cooperation in the wood mouse. Nature 418, 174–177.

Moore, H. D., Martin, M. & Birkhead, T. R. (1999). No evidence for killer spermor other selective interactions between human spermatozoa in ejaculates of differentmales in vitro. Proceedings of the Royal Society B: Biological Sciences 266, 2343–2350.

Moore, T. & Moore, H. D. (2002). Marsupial sperm pairing: A case of ‘sticky’ greenbeards? Trends in Ecology and Evolution 17, 112–113.

Moore, H. D. & Taggart, D. A. (1995). Sperm pairing in the opossum increases theefficiency of sperm movement in a viscous environment. Biology of Reproduction 52,947–953.



Morrow, E. H. & Gage, M. J. G. (2001). Consistent significant variation betweenindividual males in spermatozoal morphometry. Journal of Zoology, London 254,147–153.

Mossman, J., Slate, J., Humphries, S. & Birkhead, T. R. (2009). Spermmorphology and velocity are genetically codetermined in the zebra finch. Evolution

63, 2730–2737.Nishiwaki, S. (1964). Phylogenetic study on the type of the dimorphic spermatozoa

in Prosobranchia. Sci. Rep. Tokyo Kyouiku Daigaku Sect. B 11, 237–275.Nowak, M. A. (2005). Five rules for the evolution of cooperation. Science 314,

1560–1563.Nur, U. (1962). Sperms, sperm bundles and fertilization in a mealy bug, Pseudococcus

obscurus Essig. (Homoptera: Coccidea). Journal of Morphology 111, 173–199.Okasha, S. (2006). Evolution and the levels of selection. Oxford University Press Inc.,

New York.Oppliger, A., Ribi, N.-G. & Hosken, D. J. (2003). Sperm length influences

fertilization success during sperm competition in the snail Viviparus ater. Molecular

Ecology 12, 485–492.Parker, G. A. (1982). Why are there so many tiny sperm? Sperm competition and

the maintenance of two sexes. Journal of Theoretical Biology 96, 281–294.Parker, G. A. & Begon, M. E. (1993). Sperm competition games: sperm size and

number under gametic control. Proceedings of the Royal Society B: Biological Sciences 253,255–262.

Parker, G. A., Immler, S., Pitnick, S. & Birkhead, T. R. (2010). Spermcompetition games: sperm size (mass) and number under raffle and displacement,and the evolution of P2. Journal of Theoretical Biology, in press.

Pattarini, J. M., Starmer, W. T., Bjork, A. & Pitnick, S. (2006). Mechanismsunderlying the sperm quality advantage in Drosophila melanogaster. Evolution 60,2064–2080.

Pfeiffer, T. & Bonhoeffer, S. (2003). An evolutionary scenario for the transitionto undifferentiated multicellularity. Proceedings of the National Academy of Science 100,1095–1098.

Phillips, D. M. (1970). Ultrastructure of spermatozoa of the wooly opossum, Caluromys

philander. Journal of Ultrastructure Research 33, 381–397.Pitnick, S. (1996). Investment in testes and the cost of making long sperm in Drosophila.

The American Naturalist 148, 57–90.Pitnick, S., Dobler, R. & Hosken, D. J. (2009a). Sperm length is not influenced by

haploid gene expression in the flies Drosophila melanogaster and Scathophaga stercoraria.Proceedings of the Royal Society of London B 276, 4029–4034.

Pitnick, S., Hosken, D. J. & Birkhead, T. R. (2009b). Sperm MorphologicalDiversity. In Sperm biology, an evolutionary perspective (ed. T. R. Birkhead,D. J. Hosken and S. Pitnick), pp. 69–149. Elsevier, London.

Pitnick, S., Markow, T. A. & Spicer, G. S. (1995). Delayed male maturity is a costof producing large sperm in Drosophila. Proceedings of the National Academy of Science 92,10614–10618.

Pitnick, S., Wolfner, M. F. & Suarez, S. S. (2009c). Ejaculate- and sperm-female interactions. In Sperm biology, an evolutionary perspective (ed. T. R. Birkhead,D. J. Hosken and S. Pitnick), pp. 245–302. Elsevier, London.

Pizzari, T. & Foster, K. R. (2008). Sperm sociality: Cooperation, altruism, andspite. PLoS Biology 6, e130.

Pizzari, T. & Parker, G. A. (2009). Sperm competition and sperm phenotype.In Sperm biology, an evolutionary perspective (ed. T. R. Birkhead, D. J. Hosken andS. Pitnick), pp. 205–244. Elsevier, London.

Presgraves, D. C., Baker, R. H. & Wilkinson, G. S. (1999). Coevolution of spermand female reproductive tract morphology in stalk-eyed flies. Proceedings of the Royal

Society B: Biological Sciences 266, 1041–1047.Quicke, D. L. J., Ingram, S. N., Baillie, H. S. & Gaitens, P. V. (1992). Sperm

structure and ultrastructure in the Hymenoptera (Insecta). Zoologica Scripta 21,381–402.

Radwan, J. (1996). Intraspecific variation in sperm competition success in the bulbmite: a role of sperm size. Proceedings of the Royal Society B: Biological Sciences 263,855–859.

Reger, J. F. & Cooper, D. P. (1968). Studies on the fine structure of spermatids andspermatozoa from the millipede Polydesmus sp. Journal of Ultrastructure Research 23,60–70.

Ribera, I., Vogler, A. P. & Balke, M. (2008). Phylogeny and diversification ofdiving beetles (Coleoptera: Dytiscidae). Cladistics 23, 1–28.

Roldan, E. R. S., Gomendio, M. & Vitullo, A. D. (1992). The evolution ofeutherian spermatozoa and underlying selective forces: female selection and spermcompetition. Biological Reviews 67, 551–593.

Rouse, G. W. (2006). Annelid sperm and spermiogenesis. In Reproductive Biology

and Phylogeny of Annelida (ed. G. W. Rouse and F. Pleijel), pp. 45–76. SciencePublishers, Inc., Enfield.

Rudolfsen, G., Figenschou, L., Folstad, I., Tveiten, H. & Figenschou, M.(2006). Rapid adjustments of sperm characteristics in relation to social status.Proceedings of the Royal Society of London B 273, 325–332.

Sahara, K. & Kawamura, N. (2004). Roles of actin networks in peristaltic squeezingof sperm bundles in Bombyx mori. Journal of Morphology 259, 1–6.

Sasakawa, K. (2007). Sperm bundle and reproductive organs of carabid beetles tribePterostichini (Coleoptera: Carabidae). Naturwissenschaften 94, 384–391.

Sasakawa, K. (2009). Marked sperm dimorphism in the ground beetle Scarites terricola:a novel type of insect sperm polymorphism. Physiological Entomology 34, 387–390.

Scharer, L., Da Lage, J.-L. & Joly, D. (2008). Evolution of testicular architecturein Drosophilidae: a role for sperm length. Evolutionary Biology 8, 143.

Scheltinga, D. M. & Jamieson, B. G. M. (2003). Spermatogenesis and the maturespermatozoon: Form, function and phylogenetic implications. In Reproductive Biology

and Phylogeney of Anura (ed. B. G. M. Jamieson), pp. 119–251. Science Publishers,Inc., Enfield.

Schiff, N., Flemming, A. J. & Quicke, D. L. J. (2001). Spermatodesmata of thesawflies (Hymenoptera: Symphyta): evidence for multiple increases in sperm bundlesize. Journal of Hymenoptera Research 10, 119–125.

Schrader, F. (1945). Regular occurrence of heteroploidy in a group of Pentatomidae(Hemiptera). Biological Bulletin 88, 63–70.

Schrader, F. (1946a). Autosomal elimination and preferential segregation in theharlequin lobes of certain Discocephalini (Hemiptera). Biological Bulletin 90,265–290.

Schrader, F. (1946b). The elimination of chromosomes in the meiotic divisions ofBrachystethus rubromaculatus Dallas. Biological Bulletin 90, 19–31.

Schrader, F. (1960a). Cytological and evolutionary implications of aberrantchromosome behavior in the harlequin lobe of some Pentatomidae (Heteroptera).Chromosoma 11, 103–128.

Schrader, F. (1960b). Evolutionary aspects of aberrant meiosis in some Pentatominae(Heteroptera). Evolution 14, 498–508.

Schrader, F. & Leuchtenberger, C. (1951). The cytology and chemical nature ofsome constituents of the developing sperm. Chromosoma 4, 404–428.

Selenka, E. (1887). Studien uber Entwicklelungsgeschichte. Ver Heft das Opossum (Didelphys

virginiana). C. W. Kriedels, Verlag, Wiesbaden.Shepherd, B. A. & Martan, J. (1979). Morphology and fertility of guinea-pig

spermatozoa aged in vivo. Archives of Andrology 2, 53–58.Silberglied, R. E., Shepherd, J. G. & Dickinson, J. L. (1984). Eunuchs: the role

of apyrene sperm in Lepidoptera. The American Naturalist 123, 255–265.Sivinski, J. (1984). Sperm in Competition. In Sperm Competition and Evolution of Animal

Mating Systems (ed. R. L. Smith), pp. 85–115. Academic Press, London.Snook, R. R. (2005). Sperm in competition: not playing by the numbers. Trends in

Ecology and Evolution 20, 46–53.Snook, R. R. & Karr, T. L. (1998). Only long sperm are fertilization-competent in

six sperm-heteromorphic Drosophila species. Current Biology 8, 291–294.Snook, R. R., Markow, T. A. & Karr, T. L. (1994). Functional nonequivalence

of sperm in Drosophila pseudoobscura. Proceedings of the National Academy of Science 91,11222–11226.

Sonnenschein, M. & Hauser, C. L. (1990). Presence of only eupyrene spermatozoain adult males of the genus Micropterix hubner and its phylogenetic significance(Lepidoptera: Zeugloptera, Micropterigidae). International Journal of Insect Morphology

and Embryology 19, 269–276.Swallow, J. G. & Wilkinson, G. S. (2002). The long and short of sperm

polymorphisms in insects. Biological Reviews 77, 153–182.Taggart, D. A., Johnson, J. L., O’Brien, H. P. & Moore, H. D. M. (1993). Why

do spermatozoa of American Marsupials form pairs? A clue from the analysis ofsperm-pairing in the epididymis of the Grey Short-Tailed Opossum, Monodelphis

domestica. The Anatomical Record 236, 465–478.Takami, Y. & Sota, T. (2007). Sperm competition promotes diversity of sperm

bundles in Ohomopterus ground beetles. Naturwissenschaften 94, 543–550.Taylor, D. R. & Ingvarsson, P. K. (2003). Common features of segregation

distortion in plants and animals. Genetica 117, 27–35.Taylor, G. (1951). Analysis of the swimming of microscopic organisms. Proceedings of

the Royal Society A: Mathematical and Physical Sciences 209, 447–461.Temkin, M. H. (1994). Gamete spawning and fertilization in the gymnolaemate

bryozoan Membranipora membranacea. Biological Bulletin 187, 143–155.Temkin, M. H. & Bortolami, S. B. (2004). Waveform dynamics of spermatozeug-

mata during transfer from paternal to maternal individuals of Membraniporamembranacea. Biological Bulletin 206, 35–45.

Temple-Smith, P. D., & Bedford, J. M. (1980). Sperm maturation and theformation of sperm pairs in the epididiymis of the opossum, Didelphis virginiana.The Journal of Experimental Zoology 214, 161–171.

Till-Bottraud, I., Joly, D., LaChaise, D. & Snook, R. R. (2005). Pollen andsperm heteromorphism: convergence across kingdoms? Journal of Evolutionary Biology

18, 1–18.Trivers, R. L. (1971). The evolution of reciprocal altruism. The Quarterly Review of

Biology 46, 35–57.Turner, R. M. (2006). Moving to the beat: a review of mammalian sperm motility

regulation. Reproduction, Fertility and Development 18, 25–38.Vacquier, V. D. (1998). Evolution of gamete recognition proteins. Science 281,

1995–1998.Veron, N., Bauer, H., Weiße, A. Y., Luder, G., Werber, M., & Herrmann, B. G.

(2009). Retention of gene products in syncytial spermatids promotes non-Mendelian



inheritance as revealed by the t complex responder. Genes & Development 23,2705–2710.

Virkki, N. (1969). Sperm bundles and phylogenies. Zeitschrift fur Zellforschung und

Mikroskopische Anatomie 101, 13–27.Virkki, N. (1973). Evolution of sperm cell number per bundle in insects. Anales de la

Escuela Nacional de Ciencias Bioloigcas (Mexico) 20, 23–34.Viscuso, R., Brundo, M. V. & Sottile, L. (2002). Mode of transfer of spermatozoa

in Orthoptera Tettigoniidae. Tissue & Cell 34, 337–348.Viscuso, R., Narcisi, L., Sottile, L. & Barone, N. (1998). Structure of

spermatodesms of Orthoptera Tettigonioidea. Tissue & Cell 30, 453–463.von Nordheim, H. (1989). Vergleichende ultrastukturuntersuchungenader eu- und

paraspermien von 13 Protodrilus arten (Polychaeta, Annelida) und ihre taxonomischeund phylogenetische Bedeutung. Helgolander wissenschaftliche Meeresuntersuchungen 43,113–156.

von Siebold, C. T. (1836). Fernere beobachtungen uber die spermatozoen der tiere.2. Die spermatozoen der Paludina vivipara. Arch. Anat. Physiol. Wiss. Med 1, 232–255.

Ward, P. I. (1998). Intraspecific variation in sperm size characters. Heredity 80,655–659.

Wakimoto, B. T., Lindsley, D. L. & Herrera, C. (2004). Towards a compre-hensive genetic analysis of male fertility in Drosophila melanogaster. Genetics 167,207–216.

Weigmann, B. M., Regier, J. C. & Mitter, C. (2002). Combined molecular andmorphological evidence on the phylogeny of the lepidopteran lineages. Zoologica

Scripta 31.Werner, G. (1976). Entwicklung und bau der doppelspermien bei den Dytisciden

Acilius sulcatus L., Dytiscus marginalis L. und Hydaticus transversalis Pont. (Coleoptera).Zoomorphologie 83, 49–87.

Werner, G. (1983). The joint seprmatozoa of Dytiscidae. 4th International Symposium on

Spermatology, 1982, 167–170.White-Cooper, H., Doggett, K. & Ellis, R. E. (2009). The evolution of

spermatogenesis. In Sperm biology, an evolutionary perspective (ed. T. R. Birkhead,D. J. Hosken and S. Pitnick), pp. 151–182. Elsevier, London.

Woolley, D. M., Crockett, R. F., Groom, W. D. I. & Revell, S. G. (2009). Astudy of synchronisation between the flagella of bull spermatozoa, with relatedobservations. The Journal of Experimental Biology 212, 2215–2223.

Woodard, T. M. (1940). The function of the apyrene spermatozoa of Goniobasis

laqueata (Say). Journal of Experimental Zoology 85, 103–123.Wu, X.-M., Song, D.-X. & Chen, D.-Y. (1997). Ultrastructure of spermatozoa and

spermatogenesis in wolf spider Pardosa astrigera (Araneae: Lycosidae). Acta Zoologica

Sinica.Wygodzinsk, P. (1958). Beobachtungen an spermatolophiden und spermatophoren

bei Nicoletiidae (Thysanura, Insecta). Zoologischer Anzeiger 161, 280–287.Xu, X., Toselli, P. A., Russell, L. D. & Seldin, D. C. (1999). Globozoopermia in

mice lacking the subunit casein α’ kinase II catalytic subunit. Nature Genetics 23,118–121.

Yang, Y., Elgeti, J. & Gompper, G. (2008). Cooperation of sperm in two dimensions:synchronization, attraction, and aggregation through hydrodynamic interactions.Physical Review E 78, 061903.

Zheng, Y., Deng, X. & Martin-DeLeon, P. A. (2001a). Lack of sharing of Spam1(Ph-20) among mouse spermatids and transmission ratio distortion. Biology of

Reproduction 64, 1740–1738.Zheng, Y., Deng, X., Zhao, Y., Zhang, H. & Martin-DeLeon, P. A. (2001b).

Spam1 (PH-20) mutations and sperm dysfunction in mice with the Rb(6.16) orRb(6.15) translocation. Mammalian Genome 12, 822–829.


Documents

Evolution of intra-ejaculate sperm interactions: do sperm cooperate?