doi:10.1111/j.1420-9101.2004.00789.x

MINI REVIEW

Pollen and sperm heteromorphism: convergence across kingdoms? I. TILL-BOTTRAUD,* D. JOLY,  D. LACHAISE  & R. R. SNOOKà *Laboratoire d‘Ecologie Alpine, CNRS, Grenoble, France  Laboratoire Populations, Ge´ne´tique et Evolution, CNRS, Gif sur Yvette Cedex, France àDepartment of Animal and Plant Sciences, University of Sheffield, Sheffield, UK

Keywords:

Abstract

angiosperms; animals; heteromorphism; invertebrates; plants; pollen; sexual selection; sperm; sperm competition; vertebrates.

Sperm competition theory predicts that males should produce many, similar sperm. However, in some species of animals and plants, males exhibit a heteromorphism that results in the production of at least two different types of sperm or pollen grains. In animals, sperm heteromorphism typically corresponds to the production of one fertile morph and one (or more) sterile morph(s), whereas in plants two or more pollen morphs (one of which can be either sterile or fertile) are produced in all flowers but sometimes in different anthers. Heteromorphism has arisen independently several times across phyla and at different phylogenetic levels. Here, we compare and contrast sperm and pollen heteromorphism and discuss the evolutionary hypotheses suggested to explain heteromorphism in these taxa. These hypotheses include facilitation, nutritive contribution, blocking, cheap filler, sperm flushing or killing for animals; outcrossing and precise cross-pollen transfer or bet-hedging strategy for plants; cryptic female choice for both. We conclude that heteromorphism in the two phyla is most likely linked to a general evolutionary response to sexual selection, either to increase one male’s sperm or pollen success in competition with other males, or mediate male/female interactions. Therefore, although sperm and pollen are not homologous, we suggest that heteromorphism represents an example of convergence across kingdoms.

Introduction To solve similar problems, unrelated organisms can either evolve different traits or converge on similar solutions. Sperm and pollen are obviously important for fertilization but also have been shown to be targets of sexual selection (Birkhead & Møller, 1998; Simmons, 2001; Skogsmyr & Lankinen, 2002). Here, we review the phenomenon of sperm and pollen heteromorphism (i.e. the production of multiple sperm or pollen morphologies within an individual) to assess the extent to which its occurrence in both kingdoms is a response to similar selective pressures. In both plants and animals, fertilization involves three steps: (1) transport of the male gamete from the male to Correspondence: Rhonda R. Snook, Department of Animal and Plant Sciences, University of Sheffield, Western Bank, Sheffield S10 2TN, UK. Tel.: 44 114 222 0126; fax: 44 114 222 0002; e-mail: [email protected]

the female, (2) transport within female organs, and (3) fertilization sensu stricto, albeit not necessarily immediately after gamete transport. In animals and nonseed plants, these steps involve sperm whereas in seed plants (spermatophytes, e.g. angiosperms and gymnosperms) the first two steps involve the gametophyte in the form of pollen grains and it is only the last step that involves sperm (Fig. 1a–d). However, fertilization in both animals and plants becomes more complicated in that females rarely have only one concurrent mate, resulting in sperm or pollen competition (see Willson & Burley, 1983; Birkhead & Møller, 1998; Simmons, 2001; Skogsmyr & Lankinen, 2002 for reviews). Instead of a simple liaison between one male’s gametes and the female partner, sperm or pollen from multiple males compete for fertilization of a set of eggs/ovules. In animals, after spawning or insemination, ejaculates may overlap, resulting in the direct competition between sperm for fertilization (Parker, 1970; Smith, 1984; Birkhead &

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

1

2

I. TILL-BOTTRAUD ET AL.

Møller, 1998). In angiosperms, pollen is transported either singly or in ‘packets’ of variable number (pollen loads) that are the equivalent of an ejaculate, and pollen from several males can be deposited on a stigma via abiotic or biotic pollinating agents leading to pollen competition (‘mixed-donor pollination’; see for example, Waser & Price, 1984; Thomson et al., 1986; Mulcahy & Mulcahy, 1987). Competition can occur both within and among males, however, because selection is thought to be stronger between males in animals (Parker, 1998), most research has focused on ejaculate traits that increase a male’s competitive success against a rival’s. Distinct from animal sperm competition, angiosperm sperm from two pollen grains, whether from the same or different males, never compete with one another. Despite these differences between animals and plants, one trait that appears to have responded to sperm/pollen competition similarly is the number of sperm/pollen produced; increased number increases an individual’s reproductive success (Parker, 1982; Simmons, 2001).

(a)

The plant and animal kingdoms have evolved separately for possibly 1 billion years (e.g. Nei et al., 2001). As a consequence, pollen grains (or anthers) and animal spermatozoa (or testes) are not homologous because their life cycle positions are different: the relationship between meiosis, fertilization, and mitosis is altered. This alteration is because of the multicellularity of the haploid generation in plants and the lack of a differentiated germ line. Additionally, their structure is divergent (Fig. 1a–d). The relevant difference between plants and animals is the portion of the life cycle in which pollen and sperm are produced and the manner in which they are produced, leading to variation in relative gene expression between pollen and sperm. Whereas both sperm (Beatty, 1972; Parker, 1993; Haig & Bergstrom, 1995) and pollen (Ressayre et al., 1998) morphology are likely under diploid control (but see Flynn & Rowley, 1971; Parker & Begon, 1993; Guzzo et al., 1994; Kreunen & Osborn, 1999), gene expression in pollen is much greater than in sperm likely because they are multicellular (Walbot & Evans, 2003). Therefore, selection will act both among

(b) Plant

Animal Zygote/seed Stem cells

diploid phase

Zygote

Spermatocyte

Double fertilization

Meiosis

Ovocyte Ovule

1 Embryo sac

Central cell

4 Microspores

Meiosis

Fertilization

haploid phase

haploid phase Egg

2 Sperm cells

Sperm

(c)

x MITOSES

Stem cell (2n)

Spore mother cells

diploid phase

MEIOSIS

(d)

MEIOSIS

Pollen mother cell (2n)

2x spermatocytes (2n)

(e)

4 Pollen grains

Membrane Acrosome

1 MITOSIS

4 microspores (n)

4x 2x sperm (n) 4 pollen grains 2 cells each (n)

(f)

Exine

Aperture

Intine

Head Nucleus

Vegetative nucleus

Mitochondria

Reproductive cell

Mid piece

Reproductive nucleus

Microtubule Flagellum

5 – 200 µ

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

and within pollen loads in plants, whereas in animals selection acts primarily among ejaculates from different males. Natural and sexual selection act on the motility, morphology and/or ornamentation of male gametes/ gametophytes to facilitate fertilization (Faegri & Van der Pijl, 1966; Sivinski, 1984). The effect of these selection pressures on sperm and pollen morphology is that the morphology is supposed to be fixed within species (Parker, 1982; Gage & Morrow, 2003), but can be variable among families, genera and species. The vast taxonomic variation in sperm/pollen shape, together with conservation of sperm/pollen morphology within species, allows these cells to be used as taxonomic indicators in both animals and plants (Faegri & Iversen, 1964; Afzelius, 1979; Jamieson, 1987, 1999, 2000a,b; Jamieson et al., 1995). However, the same between-clade variation leads to questions about the driving force behind rapid divergence in sperm/pollen morphology (Erdtman, 1966; Afzelius, 1975; Sivinski, 1984; Joly et al., 1989; Pitnick et al., 1995). Moreover, this intraspecific conservation in morphology is not always true, such as in the case of sperm (Fig. 2a,b) and pollen heteromorphism (Fig. 2c,d). Sperm/pollen heteromorphism is not a curiosity limited to a few minor taxa (Tables 1 and 2). Sperm heteromorphism has arisen independently several times and occurs in both invertebrate and vertebrate phyla,

3

including molluscs (both prosobranch gastropods and bivalves), arthropods (spiders, insects), centipedes, annelids, echinoderms, platyhelminthes, nematods (Table 1), and fish. It occurs in half of the invertebrate taxonomical groups investigated so far, and over 10% of invertebrate families in which sperm structure has been studied (Table 1). Similarly, pollen heteromorphism is quite common and broadly distributed in angiosperms, occurring in wind- and insect-pollinated species and in eudicots and magnolids. One-third of angiosperm families are estimated to have heteromorphic genera or species (Table 2). Pollen heteromorphism has arisen multiple times at different levels in the phylogenetic trees (Dajoz et al., 1991; Mignot et al., 1994) and even within single genera (Viola: Nadot et al., 2000; Medicago: Bena, 1998; Epilobium: Mignot, 1995). Both in animals and plants, the presence of heteromorphism has not been systematically assessed, and hence, its prevalence may be higher than current estimates.

Expression of heteromorphism Animals In some organisms, the presence of multiple types of sperm is easily identifiable because the sperm types are morphologically distinct (Fig. 2a,b). For example, in

Fig. 1 Male gametogenesis and fertilization in plants and animals. (a) In animals, no mitoses occur in the haploid stage (except for some parthenogenetic species where males are haploid organisms); sperm are the direct products of meiosis that develop in the testes during spermatogenesis. (b) In plants, the products of meiosis are microspores that develop into haploid organisms, the gametophytes. Each male gametophyte (pollen grains) will in turn produce two genetically identical sperm cells through mitosis. One sperm cell will fuse with the ovule, and the second with the central cell (that has two nuclei) to produce a triploid endosperm. (c) In animals, spermatogenesis is relatively similar from one species to the other. It starts with a variable number of mitosis (that range from 4 to 14 in invertebrates, Callard & Callard, 1998, · in the figure), to produce spermatocytes. Each spermatocyte produces four spermatids through meiosis that differentiate into spermatozoa. Mature sperm are stored in different places in the male reproductive tract, depending on the organism. In most cases, sperm acquire motility when released from either the testis or storage. Fertilization is either internal or external. In internal fertilization, sperm must swim or be transported to the female and then enter the female reproductive tract and potentially stored in special sperm storage organs. Usually, only a minority of the sperm succeed in reaching areas of the female reproductive tract where fertilization may occur. (d) Angiosperm pollen grains develop in the anthers. The pollen mother cell produces four haploid microspores through meiosis (plants are also called sporophytes). Each microspore will develop into a pollen grain following one or two mitotic divisions. One of the nuclei forms the vegetative cell. Its genes are expressed and it metabolically supports the first stage of pollen tube growth. Pollen grains are the male gametophytes of angiosperms (homologous to the prothallus of ferns), whereas the female gametophytes are the embryo sacs, which are embedded in sporophytic tissues. Once on a stigma, pollen tubes will exit the pollen through one aperture and grow through the style (female tissues) to transport the two sperm cells to the embryo sac. Angiosperm pollen grains have either two or three nuclei when mature. In binucleate pollen grains, the second nucleus divides during pollen tube elongation to give two sperm cells. In trinucleate pollen grains, this division takes place during pollen maturation in the anther. (e) Animal sperm generaly consist of four components: (1) a head containing the haploid genetic material in a nucleus and (2) an acrosome that releases cytolytic chemicals to dissolve egg membranes for fertilization, (3) a midpiece that contains mitochondria used to generate energy for motility, and (4) the flagellum that is composed of an axoneme. However, in many taxa, any component may dominate or be absent. Animal spermatozoa are always in an aqueous environment. They have very short lifespans outside the male or female reproductive tracts, but within the female reproductive tract, particularly in species where females have sperm storage organs, sperm lifespan can approach a few decades (Verma, 1974; Porter & Jorgensen, 1988; Wheeler & Crichton, 1990; Pamilo, 1991; Baur, 1998). (f) Pollen grains are composed of two (vegetative and reproductive) or three (vegetative and two sperm, not shown here) genetically identical haploid cells, cytoplasmic reserves, and are surrounded by a thick multilayer wall (intine + exine). Thinner areas in this wall are called apertures, such as pores or furrows, and serve as exits for the initiation of pollen tube growth and mediate some pollen–stigma interactions. Pollen grains are not motile and plants must rely on a third party to be transported to the female organs. Pollen grains are very sensitive to environmental conditions and usually exhibit a short lifespan.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

4

I. TILL-BOTTRAUD ET AL.

Fig. 2 Sperm and pollen heteromorphism. In animals, sperm heteromorphism is differentially expressed with some species, for example, producing short, nonfertile sperm as in Drosophila pseudoobscura (a) or long, nonfertile sperm as in the mollusc, Viviparus ater (b). In plants, pollen heteromorphism is often expressed as variation in aperture number with plants either producing either four or five apertures as in Viola tricolor (c) or either three or four apertures as in Nicotiana tabacum (d).

Table 1 Distribution of sperm heteromorphism in invertebrates. Taxa

SHF

SHG

SHS

Myriapoda* Mollusca  Annelidaà Nematoda§ Crustacea– Platyhelminthes** Chelicerata   Echinodermataàà Priapulida§§ Tardigrada–– Lophophorata*** Hemichordata    Urochordataààà Cephalocordata§§§ Gastrotricha–––

5 (19) 40 (117) 6 (74) 6 (46) 4 (75) 1 (122) 1 (106) 2 (56) (2) (8) (15) (6) (13) (1) (7)

7 (36) 55 (201) 16 (186) 8 (79) 4 (180) 1 (190) 1 (157) 2 (103) (5) (12) (15) (6) (33) (1) (12)

25 (63) 59 (286) 31 (267) 9 (108) 5 (247) 2 (234) 2 (192) 2 (148) (7) (20) (16) (6) (53) (4) (16)

Taxa are ordered according to the importance of sperm heteromorphism at the species level. For the insect see the recent review of Swallow & Wilkinson (2002). The number of sperm heteromorphic families (SHF), genera (SHG) and species (SHS) are given with the total number (in brakets) of families, genera and species investigated for sperm morphology or ultrastructure. *Mazzini et al. (2000).  Buckland-Nicks (1995), Healy (1995, 2000). àFerraguti (2000), Rouse (2000). §Justine & Jamieson (2000), Yushin et al. (2002), Yushin & Zograf (2002). –Medina (1995), Jamieson & Tudge (2000), Tudge (2003). **Baˆ & Marchand (1995), Justine (1995), Watson & Rohde (1995).   Alberti (2000). ààAu et al. (1998), Jamieson (2000b). §§Storch et al. (2000). ––Rebecci et al. (2000). ***Jamieson (2000b).    Jamieson (2000b). àààBurighel & Martinucci (2000). §§§Jamieson (2000b). –––Ferraguti & Balsamo (1995).

insects and some fish, sperm development occurs in cysts (Fig. 1c) and all sperm within a cyst are at the same developmental stage. In sperm heteromorphic invertebrates, double (or multiple) spermatogenesis occurs in which, at any one time, each cyst produces only one sperm morph that is clearly recognized and unambiguous from each other at some point following meiosis (Buckland-Nicks, 1998; Ferraguti, 2000; Ferraguti et al., 2002). The production of different sperm types may be located in different parts of the testis (Schrader, 1960) and within a testis, the cysts may switch production between types (Friedla¨nder, 1997). Some organisms, such as mammals, produce various sperm morphs to greater or lesser extents (as classified by the World Health Organization, 1999). One difficulty with considering these sperm types as heteromorphic is that there is much overlap between the various morphs, making unambiguous categorization of each sperm difficult (Amann, 1989), thus assessing their respective function and potential adaptive significance problematical. Because of this, we do not consider species in which the sperm morphs are not clearly distinct. The expression of sperm heteromorphism in animals varies (Table 3). Two general patterns occur; either all sperm morphs appear fertilization competent either because they have the appropriate haploid DNA content and/or are transferred to females (Ansley, 1954; Au et al., 1998; Kubo-Irie et al., 1998; Mazzini et al., 2000), or sperm morphs are not functionally equivalent with one morph participating in fertilization and the other(s) being fertilization incompetent, with sterility a consequence of lacking the appropriate DNA composition (Meves, 1902; Friedla¨nder & Hauschteck-Jungen, 1982), structural incapability of fertilization (Healy & Jamieson, 1981), or unknown factors (Snook et al., 1994; Snook & Karr, 1998). The distinguishing feature between these patterns is fertilization ability, but classifying taxa into these

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

Table 2 Distribution of pollen heteromorphism in different subclasses of eudicots angiosperms. Subclass

PHO

PHF

PHG

PHS

Ranunculiidae Hamamelididae Caryophyllidae Dilleniidae Rosidae Asteridae

2 1 2 4 3 7

2 2 2 5 5 8

5 3 2 6 5 13

19 3 2 17 6 30

(2) (1) (3) (6) (4) (8)

(3) (2) (3) (8) (6) (13)

(10) (3) (3) (9) (11) (25)

(44) (3) (3) (31) (42) (75)

The number of pollen heteromorphic orders (PHO), families (PHF), genera (PHG) and species (PHS) are given with the total number of orders families, genera and species observed in brackets. Monocots and Magnoliidae were excluded from the survey because pollen from these taxa have mostly only one aperture (but see Ressayre et al., 2002). Caryophyllidae, on the other hand, are under-represented because pollen from this subclass is polyaperturate (i.e. more than 10 apertures), and any variation in aperture number probably has a lower impact on physiological characteristics (see text). Adapted from Mignot (1995).

groups can be difficult because of the lack of data on whether all sperm types are truly capable of fertilization. Most evidence of sperm heteromorphism comes from ultrastructural studies (Table 1) that, by their very nature, prevent determining the fertilization capacity of each morph. For the majority of animals in which the fertilization competency of each sperm type has been deduced through either DNA content or experimentally tested, only one morph participates in fertilization and the other morph(s) does not. The sperm participating in fertilization is referred to as the eupyrene type (or eusperm containing the normal haploid chromatin content) and, most generally, the nonfertilization type as parasperm (Hodgson, 1998). Parasperm can be classified according to chromatin content variation (e.g. sperm with little chromatin are oligopyrene, those with an excess of chromatin are hyperpyrene, and those without chromatin are apyrene). Most parasperm are either oligopyrene or apyrene. Plants In angiosperms, the different morphs are fertile and produced in similar proportions by all anthers of all flowers of a plant (Till-Bottraud et al., 1994) but they have different physiological characteristics that influence germination and survival rates (Table 3; Dajoz et al., 1991, 1993; Till-Bottraud et al., 1999). The different morphs are sometimes produced within the same tetrad (i.e. from the same germinal cell or the plant equivalent of a sperm cyst; Mignot et al., 1995; Ressayre et al., 1998, 2002), and are determined by variations of the meiosis parameters, such as cytokinesis, tetrad conformation and cell wall deposition (Ressayre et al., 2002). This definition is restrictive compared with that of invertebrates and eliminates the production of

5

different pollen morphs by different anthers of a flower (anther heteromorphism). Given the parallel of anther heteromorphism with double spermatogenesis in animals (i.e. the stamens differentiate early in development), anther heteromorphisms will also be considered. One type of anther heteromorphism occurs in heterostylous species where all plants have two (or three in some species) anther or style heights (Barrett, 2002; Table 3). Each anther height produces a specific pollen morphology that can only germinate and grow in sameheight stigmas. Thus, pollen with different morphologies very rarely compete on a stigma. Anther heteromorphism may also occur when two, or rarely three, different anther types are localized in different areas of the flower (heterandry; Faegri & Van der Pijl, 1966). The anthers from which pollen is systematically transported to the stigma are called pollinating anthers and produce fertile pollen. The anthers from which pollen is never or only rarely transported to the stigma are called feeding anthers and act as an attractant to insect visitors. Pollen from feeding anthers can be absent, sterile or fertile (Bowers, 1975). This type of anther heteromorphism is the equivalent of double spermatogenesis in invertebrates.

The adaptive significance of animal and angiosperm heteromorphisms Sperm and pollen heteromorphism may represent a neutral trait that does not affect fitness. In plants, this idea is unlikely because the different morphs have different physiologies that affect their survival and germination rate. In mammals, sperm heteromorphism may be a consequence of errors during meiosis (Harcourt, 1991) but this is doubtful in invertebrates because heteromorphic sperm production is highly developmentally regulated. Each sperm cyst produces only one type at a time and each type is morphologically distinct. Moreover, every organism would then be predicted to exhibit sperm heteromorphism as meiotic errors are ubiquitous. Sperm heteromorphism is also unlikely to be neutral because in many species, parasperm represent at least 50% of the ejaculate (Gage, 1994; Snook & Markow, 2001). Sperm heteromorphism could respond to selection; males vary in sperm morph size (Joly & Lachaise, 1994; Snook, 1997, 2001), male ejaculate composition is repeatable and sperm heteromorphism responds to selection (R. Snook, unpublished data), all suggesting a heritable component. Likewise, pollen heteromorphism appears heritable and individuals within a pollen heteromorphic species can have different proportions of the different types (Viola: Dajoz et al., 1993; Nicotiana: TillBottraud et al., 1995), indicating at least the potential for selection to act on the trait. These reasons and both the frequency with which animals (Table 1) and plants (Table 2) exhibit male gametic-related heteromorphism,

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

6

I. TILL-BOTTRAUD ET AL.

Table 3 The various heteromorphic patterns in animals and plants. Taxa Invertebrates Insecta: Diptera

Drosophila obscura group spp.

Expression of heteromorphism

References

Parasperm type shorter in length than the eusperm type but each type with a nucleated sperm head Sperm types produced simultaneously Only long sperm participates in fertilization

Joly et al. (1991) Snook (1997) Bressac & Hauschteck-Jungen (1996) Snook et al. (1994), Snook & Karr (1998) Pasini et al. (1996)

No difference in DNA content or specificity for DNA fluorochromes, thus parasperm cannot be classified according to chromatin content Drosophila teissieri, Drosophila yakuba

Unknown whether both sperm types are fertilization competent

D. Joly, unpublished data

Parasperm type anucleated (apyrene) and shorter in length than eusperm; parasperm sterile Both sperm types derive from the same bipotential spermatocytes but production is asynchronous; eupyrene spermatogenesis starts during larval instars and stops after pupation whereas apyrene spermatogenesis begins just before or after pupation and persists in imagoes

Friedla¨nder & Hauschteck-Jungen (1982)

Different sperm sizes/chromatin content located in different parts of the testis (harlequin lobes) Abnormal chromatin content likely sterile; otherwise unknown whether all sperm types are fertile In Cicadidae, the two sperm types have different nuclear lengths (short and long); only long sperm are fertile

Schrader (1960), Tuzet (1977)

Insecta: Hymenoptera (Dahlbominus fuscipennis)

Five sperm types with three not entering female sperm storage organs and therefore functionally sterile. The two remaining types have either dextral or sinistral helices from head to tail Unknown whether both sperm types are fertile

Lee & Wilkes (1965)

Myriapoda: Chilopoda

Macro- and microsperm differ in volume of primary spermatocytes by a factor 140 and have different specificities for fluorochromes Unknown whether all types are fertile

Ansley (1954), Prunescu (1992), Carcupino et al. (1999)

Macro- and microsperm with differences in the DNA volume (but not content) and histone content of the nuclei; located in different parts of the testis Production of sperm types discordant Unknown whether all types are fertile

Ansley (1954), Prunescu (1992)

Crustacea: Decapoda

Synchronous production of sperm with variable number of spikes and arms from the nucleus Unknown whether all types are fertile

Jamieson & Tudge (2000)

Annelida

Paraspermatozoa with or without normal or reduced acrosomes or Paraspermatozoa with smaller nuclei (and lower amount of DNA) with incompletely condensed chromatin or Paraspermatozoa with large mitochondria and tails with swollen membranes All parasperm are sterile

Ferraguti (2000)

Insecta: Lepidoptera

Insecta: Hemiptera

Scutigera forceps

Friedla¨nder (1997)

Kubo-Irie et al. (2003)

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

7

Table 3 Continued Taxa

Expression of heteromorphism

References

Mollusca

Long anucleate/oligonucleate parasperm and short eusperm

Nematoda

Species with two testes in tandem position; large (anterior testis) and small (posterior testis) aflagellate eupyrene sperm; only the small sperm enter the female Species with only one testis: small and large eupyrene sperm which enter female receptacle; fertility of both types is not determined

Fain-Maurel (1966), Buckland-Nicks (1998) Riemann (1986)

Echinodermata

Eupyrene sperm with symmetrical or asymmetrical midpieces Fertility of both types is not determined

Au et al. (1998)

An erythrocyte-shaped unflagellated binuclear parasperm and flagellate eusperm both produced in the same cyst and developmentally regulated

Hayakawa et al. (2002a,b)

One type of sperm has a larger head than the other sperm type

Watanabe et al. (2000)

Pollen differ for size, ornamentation or aperture number

Till et al. (1989), Dajoz et al. (1991), Mignot et al. (1994) Ressayre et al. (1998, 2002)

Vertebrates Osteichthyes: Scorpaeniformes (Hemilepidotus gilberti)

Actinopterygii: Syngnathiformes (Syngnathus schlegeli) Plants Angiosperms (eudicots and magnolids)

Determined by meiotic processes (cytokinesis, tetrad conformation, cell wall deposition) All pollen types potentially fertile and the different morphs have different physiological characteristics that influence germination and survival rate

Dajoz et al. (1991, 1993), Till-Bottraud et al. (1999)

Asteridae: Nicotiana

Two main aperture numbers (three or four) Variable proportions of the two morphs with temperature

Till-Bottraud et al. (1995)

Dilleniidae: Viola

Aperture number varies from two to six (different main types in different species)

Dajoz et al. (1993), Till-Bottraud et al. (1999), Nadot et al. (2000)

Rosidae: Medicago

Aperture number varies from two to six (different main types in different species)

Bena (1998)

Rosidae: Epilobium

Aperture number varies from two to four (different main types in different species)

Mignot (1995)

Equivalent to the double spermatogenic lines in animals Two or three levels of anther/style length. Different pollen types produced in different anthers heights Right/left anther and stigma positioning. Pollinating and feeding anthers

Faegri & Van der Pijl (1966) Barrett (2002)

Anther heteromorphism 1-heterostylous species (Oxalis, Primula, Armeria, Lythrum) 2-enantiostylous species (Solanum rostratum, Cassia, Heterandra)

in concert with the distribution across both closely and distantly related taxa, suggests that this phenomenon is adaptative. Most hypotheses to explain animal sperm heteromorphism consider that the morphs are not functionally equivalent (as experimentally demonstrated), and hence, attempt explaining the evolution of a sterile caste of gametes. In contrast, hypotheses for pollen heteromorphism consider that all the morphs within a plant are fertile and hence, need to account for this coexistence. Below we discuss these hypotheses and review current evidence for the adaptive signifi-

Bowers (1975)

cance of sperm and pollen heteromorphism with the aim of understanding the extent to which these heteromorphisms of animals and angiosperms are evolutionarily comparable. Facilitation The facilitation hypothesis suggests that parasperm function to assist the transfer or storage of fertilizing sperm (Katsuno, 1977; Osanai et al., 1987). Two decades of work on lepidopterans suggest that apyrene sperm aid

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

8

I. TILL-BOTTRAUD ET AL.

the transport of eusperm (Sahara & Takemura, 2002, 2003), however, most evolutionary literature ignores facilitation. Perhaps this is because facilitation is generally considered a proximate explanation for sperm heteromorphism whereas the other hypotheses are considered ultimate. Yet, if parasperm assist eusperm and thereby increase the reproductive success of the male, then facilitation is an ultimate explanation for the production of sterile gametes. The most obvious prediction from the facilitation hypothesis is that in the absence of parasperm, eusperm would not either be transferred by males or stored by females and therefore fertilization would not occur. This prediction is difficult to test because, until recently, separating the sperm types has been problematic and artificial insemination difficult. New work in the lepidopteran, Bombyx mori, overcame some of these methodological difficulties and found that neither sperm type alone could fertilize eggs, but when mixed, the fertilization frequency was 80–95%, demonstrating that apyrene sperm are necessary for eupyrene sperm fertility (Sahara & Takemura, 2002, 2003). In insects, this result is the strongest evidence to date that parasperm assist eusperm in some way, perhaps by facilitating transfer of sperm-tosperm storage organs or somehow making eusperm fertilization competent. Facilitation has been a favoured hypothesis to explain sperm heteromorphism in molluscs. For example, in aphallic mollusc species, parasperm deliver eusperm for fertilization (for review, see Buckland-Nicks, 1998). In Tonnoidea, males produce two parasperm types, a lancet morph whose function is unclear and a carrier morph that delivers cohorts of eusperm to females or the site of fertilization (Buckland-Nicks, 1998). Additionally, in some species, spermatozeugmata composed of a mixture of eu- and parasperm may swim faster than independent eusperm (Buckland-Nicks, 1998), hence the parasperm may facilitate the delivery of eusperm. In the gastropod snail, Viviparus ater, a male’s fertilization success is dependent on the length of the longer nonfunctional parasperm, and these longer parasperm could help eusperm reach the site of fertilization faster (Oppliger et al., 2003). Parasperm may also function in facilitation by preventing the diffusion of eupyrene sperm during sperm transfer. In the marine sculpin fish, Hemilepidotus gilberti, parasperm decreases the lateral dispersion width of eusperm (Hayakawa et al., 2002a). In the mollusc, Goniobasis, parasperm entangle eusperm and are thought to prevent premature eusperm dispersal (Woodward, 1940) but how important this is to fertilization success is unknown. In some molluscs (Woodward, 1940; Buckland-Nicks, 1998) and annelids (Ferraguti, 2000), both sperm types are grouped after mating to form rod-like spermatozeugmata in which the parasperm surrounding the eusperm protect and hold the eusperm together, carrying them either towards the spermathecal opening

or assisting in the penetration through the female’s body wall during dermal impregnation. An obvious weakness of the facilitation hypothesis is that many species do not require additional sperm types for transport to or within the female for fertilization. What mating system and fertilization conditions lead to the evolution of a sterile sperm morph whose function is to deliver fertile sperm to the site of fertilization or to somehow make eusperm fertile is unclear. Aphallic species may be one candidate group in which facilitation may be necessary for sperm transfer but this does not explain sperm heteromorphism in phallic species. In some taxa, it may be easier to co-opt sperm production machinery to create a cell type (e.g. parasperm) that delivers eusperm since locomotory function is already part of the system, rather than develop additional special morphological structures that deliver sperm. Developing sperm separation and artificial insemination techniques are needed to elucidate whether sterile sperm function in facilitation (or other hypotheses; see below), as well as unravelling the reproductive biology and selective agents that promote the evolution of sperm heteromorphism. In angiosperms, one type of facilitation is known as the ‘population’ effect (Brewbacker & Majumder, 1961) in which proteins stored in the outer envelope of the pollen grain diffuse on the stigma and stimulate pollen tube germination (Chen et al., 2000). Pollen heteromorphism with a sterile and a fertile morph could be selected for facilitation when both morphs are transferred simultaneously and sterile pollen enhances the germination of fertile pollen by increasing the protein concentration on the stigma, but pollen from a male donor could similarly enhance germination of pollen from a rival male. Instead, the feeding anthers in anther heteromorphic plants are probably a better equivalent of facilitation in animals. Feeding anthers attract the pollinator and are positioned such that pollen from the pollinating anthers are maximally transferred on to the part of the pollinator that will touch the stigmas in the next visit to a flower (Barrett, 2002). This mechanism, therefore, increases a male’s reproductive success when several plants are visited in sequence by a pollinator, i.e. under mixed-donor pollination. A test of this hypothesis should compare pollination efficiency of plants both with and without feeding anthers from the same species, but also without the special positioning of the style related to feeding anther position. No such plants exist naturally, and thus, this test has never been performed. In plants, the production of an additional pollen type does not require any morphogenetic innovation as heteromorphism is due to small variations in meiosis parameters (Ressayre et al., 1998, 2002). Indeed, pollen homomorphism is probably more costly as it requires very strict control of microsporogenesis parameters.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

Nutritive contribution In many animals, the parasperm do not persist in female sperm storage areas, suggesting that these parasperm could break down and serve as a nutritional gift to the fertilizing sperm, the female and/or to offspring (Schrader, 1960; Richards, 1963; Riemann & Gassner, 1973; Sivinski, 1980; Buckland-Nicks, 1998). The likelihood of insect parasperm serving this function is low primarily because most sperm heteromorphic insects transfer sperm in packages containing energetically expensive seminal fluids that probably provide much more nutrition than the simple parasperm could (Swallow & Wilkinson, 2002). Only one experiment, performed on Drosophila pseudoobscura, has tested whether parasperm are incorporated into female somatic and ovarian tissues; no relationship between the disappearance of radiolabelled parasperm and radiolabel uptake in the female was found (Snook & Markow, 1996). In the marine sculpin fish, the parasperm are unlikely to be nutritive because the fish are oviparous so that nutrition for embryos depends on yolk laid down prior to mating and the length of time in which eusperm are motile appears unrelated to the presence of parasperm, eliminating the sperm-feeding hypothesis (Hayakawa et al., 2002a,b). The mollusc literature strongly conjectures that parasperm may be a nutrient contribution (Buckland-Nicks, 1998), although this idea has not been tested. In contrast to the simple structure of insect parasperm (Fig. 2a), mollusc parasperm (Fig. 2b) can have copious secretory vesicles that are rich in polysaccharides and glycoproteins. If parasperm provide nutrients, then they probably benefit the female rather than eusperm, as gastropod eusperm contain independent stores of glycogen (Buckland-Nicks, 1998) and only eusperm are transferred to the seminal receptacle. Eusperm can be stored in females for over 3 months, embedded in invaginations of epithelial cells suggesting little need for additional nutrients, therefore, it is unlikely that parasperm serve to benefit the fertile sperm. The potential for parasperm to serve as a nutritive contribution to offspring may be exemplified in the mollusc, Aporrhais pespelicanis, in which both eu- and parasperm can enter eggs but only eusperm pass into the entrance cones for fertilization whereas parasperm partially break down before being eliminated from the egg (Healy & Jamieson, 1981). Whether this partial donation improves survivorship of the offspring and whether the eusperm and parasperm that enter the eggs are from the same male or different males is unknown. Similar to the facilitation hypothesis, evidence for the nutrient contribution hypothesis is weak, primarily because of the lack of experimental data. Further, why these nutrient contributions would be transferred in relatively small sperm compared with a larger volume of seminal fluids, as in most nuptial gifts, is difficult to

9

imagine, particularly if these nutrients are critical for increased survival and/or fecundity of the partner or offspring. Moreover, in multiply mating species, these sperm donations would be prone to cuckholdry (Markow & Ankney, 1984). Experiments that increase the number of parasperm transferred to females, assessing whether either female survivorship or fecundity increased, or whether eupyrene sperm lived longer, would provide support for this hypothesis. Current methods are not available to manipulate the number of each sperm type transferred to females. Anther or pollen heteromorphisms are also unlikely to serve a nutritive role for the receptive plant or the seed. By analogy to animals, the sterile pollen produced by feeding anthers could serve as a nutrient function. Although this is their role towards pollinators, there are several reasons why this would be improbable at the level of the plant. First, sterile pollen do not usually reach the stigmas because they are not positioned on the pollinator where the animal could transfer it to the relevant female tissues (Bowers, 1975; Barrett, 2002). Secondly, being sterile, they usually have little or no cytoplasm and do not germinate, limiting the contact with the style to the pollen outer envelopes. Thirdly, the trophic relationship between male and female functions is opposite to animals; the female stigmatic tissues have been shown to actively provide nutrients for pollen tube growth (Lord, 2003), making contributions by any extra pollen superfluous. Blocking Parasperm may block fertilization sites from rival sperm by either forming sperm plugs or congregating at the site of sperm storage entrances (Woodward, 1940; Sivinski, 1980; Baker & Bellis, 1995). In some insects, parasperm are generally not present in the female reproductive tract when females remate and so are unlikely to function as blockers (Snook, 1998; Cook & Wedell, 1999). Some species of molluscs produce two types of parasperm: carrier sperm that serve in facilitation (see above) and lancet sperm. Lysosomes in these lancet parasperm could create a hostile prefertilization environment for rival sperm and these lancet sperm form a plug in the female bursa that binds eusperm (Buckland-Nicks, 1998). However, the effectiveness of lancet sperm in preventing either female remating or the storage of rival sperm is untested. Parasperm of the marine sculpin fish both facilitate the delivery of the fertilizing sperm type to the site of fertilization (see above), but also may hinder later rival male sperm from reaching eggs. Both eu- and parasperm types are simultaneously released into the water but parasperm form clumps at the seawater/ovarian fluid boundary (Hayakawa et al., 2002a,b). These clumps engulf eusperm, suggesting that sneaker rival male eusperm may be prevented from penetrating through

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

10

I. TILL-BOTTRAUD ET AL.

the ovarian fluid surrounding a female’s released eggs. However, the efficacy of this phenomenon serving to block rival sperm is unclear because parasperm engulf both self- and nonself-sperm (Hayakawa et al., 2002a) and how quickly and how many self eusperm reach the eggs prior to clump formation has not been determined. The blocking hypothesis suffers from lack of experimental data. Additionally, the effectiveness of sperm plugs in preventing matings and/or sperm transfer by second males is unclear. In some species, sperm plugs do appear to reduce sperm competition (Polak et al., 2001) or evolve under sperm competitive situations (Dixson & Anderson, 2002) whereas in others, blocking the female reproductive tract may not decrease the opportunity for sperm competition (Jia et al., 2002; Parga, 2003). Understanding why some species may use sterile sperm for this function and other species fertile sperm or seminal fluids will require a more in-depth study of the genetics and physiologies of these different strategies. Angiosperm heteromorphisms are unlikely to function in blocking. Because pollen from feeding anthers is unlikely to be transferred to the stigma (see above; Barrett, 2002), they cannot block reception of further pollen loads. Large loads of fertile pollen on a stigma block new loads from landing or germinating (Marshall & Ellstrand, 1988). However, mixtures of pollen from different donor plants occur on the pollinator prior to deposition on a stigma, therefore, there can be only very few pollen grains from one individual competing against pollen from other donors. A blocking strategy by the production of sterile pollen would not be selectively favoured because if the few pollen grains from one individual in a pollen load were sterile, then all fertilizing chances would be lost. In some species, additional cells are released together with pollen (Iqbal & Wijesekara, 2002), and, whereas untested, these cells have been suggested to block rival males’ pollen from reaching stigmas. The additional cells, however, are diploid cells derived from anther tissues and not a specific pollen morph. Cheap filler In species with internal fertilization and in which females store sperm, females may remate when stored sperm loads go below some threshold (Thibout, 1975; Sugawara, 1979). Parasperm have been suggested to serve as ‘cheap filler’ that increase sperm loads and deceive the female into delaying remating, thereby increasing the amount of time that a male’s sperm will be used before his partner remates (Silberglied et al., 1984). This type of sperm competition mechanism (along with blocking) is referred to as ‘defensive’ as it results in a male defending his ejaculate from a potential rival male. The probability that parasperm function as cheap filler is limited to those taxa with internal fertilization, sperm

storage, and female ability to assess those stores and thus, generally this hypothesis has only been tested in insects. Results are mixed. Comparative analyses in lepidopterans conclude that parasperm may serve as cheap filler (Gage, 1994; Morrow & Gage, 2000) and in an experimental study on Pieris napi, remating females had fewer parasperm than nonremating females (Cook & Wedell, 1999). In contrast, experimental evidence in the dipteran, D. pseudoobscura, found no difference between remating and nonremating females in the number of short sperm in storage (Snook, 1998). The relationship between the production of sterile sperm and the cheap filler hypothesis in noninsect taxa is undocumented because female remating behaviour in these groups is not well-studied. The cheap filler hypothesis assumes that sterile sperm are cheaper to produce. In insects, sterile sperm are smaller than fertile sperm and can lack expensive DNA, suggesting that parasperm are likely to be cheap compared with long, nucleated sperm. However, the cost of producing sterile and fertile sperm has not been empirically determined. Whereas the energetic resources to produce one short sperm are likely to be less than one long sperm, it is the net cost of the ejaculate that matters. In sperm heteromorphic Drosophila, whether sterile sperm are cheaper to produce compared with long, fertile sperm will likely depend on the total length of each sperm type and the total number of each sperm type in the ejaculate. For example, the ejaculate of D. pseudoobscura has about 25 000 sperm, half of which are short sperm measuring 90 lm and half that are long measuring 360 lm (Snook et al., 1994). If one was to take length and number as an estimate of cost, then in this species long sperm ‘cost’ about four times that of short sperm. However, in D. athabasca, males produce 9000 short sperm measuring 120 lm compared with 1000 long sperm that are 1527 lm (Snook, 1997; unpublished data). In this species, long sperm ‘cost’ only about 1.4 times as much as short sperm. Do these estimates bear any relationship to whether D. pseudoobscura is a better candidate for short sperm functioning as cheap filler (which it does not appear to do; Snook, 1998) than D. athabasca? Or are these between-species estimates unrelated to whether cheap filler operates? Whereas this naı¨ve calculation may work in dipterans because both sperm types have a nucleus, in lepidopterans, parasperm have a nucleus that is eventually discarded during development; whether this is cheaper than retaining a nucleus is speculative. In molluscs, the parasperm are larger than the eusperm and can be produced in relatively large numbers (Buckland-Nicks, 1998). Hence, it is unlikely that mollusc parasperm can be considered ‘cheap’ relative to the production of eusperm and thus could not function as cheap filler. The concept of cheap filler in plants is problematic because there is no pollen or sperm storage and ‘remating’ of a plant usually involves different flowers. Flower receptivity is indeed diminished by

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

pollen deposition, but plant receptivity is not, and plants do not control receiving a pollen load. Moreover, because an external agent transports the pollen, pollen deposition on a stigma is often limiting for seed set (T. L. Ashman, unpublished data) and in several species flowers actually wait to be pollinated and then wilt very rapidly after pollen deposition (Ashman & Schoen, 1994). These life history and reproductive traits suggest that pollen heteromorphism is not a response to female remating and does not function as cheap filler. Sperm flushing or killing Sperm flushing suggests that parasperm physically displace a previous male’s sperm from the female sperm stores (Baker & Bellis, 1988; Buckland-Nicks, 1998). The controversial ‘kamikaze’ sperm hypothesis suggests that parasperm kill a rival male’s sperm (Baker & Bellis, 1988, 1995). These sperm competition mechanisms are referred to as ‘offensive’ as it results in the last male to mate trying to gain fertilization opportunities against an existing male’s sperm. Whereas game theory has predicted the conditions under which kamikaze sperm could evolve (Kura & Nakashima, 2000), no tests have been performed to assess whether parasperm function to flush resident sperm from storage or kill rival sperm. Perhaps the most likely candidate taxonomic group for such a function would be the Tonnoidae molluscs, in which the lancet sperm have lysosomes that could create a hostile prefertilization environment for rival sperm (BucklandNicks, 1998), but the action and specificity of these lysosomes has not been characterized. The lack of data for this hypothesis is probably due to the difficulties in distinguishing sperm from different males; without some kind of identifying marker to distinguish rival male’s sperm from each other, determining whether one is killing/flushing the other’s sperm is currently impractical. Moreover, it is unclear how sperm could kill a rival’s sperm and avoid killing self-sperm, as there is no evidence for the ability to distinguish between self- and nonself-sperm (Gilchrist & Partridge, 1995). This problem could be circumvented if the parasperm entered the sperm storage organs first; they could kill rival sperm before the male’s eusperm arrived in storage. In Drosophila, the short sperm disappear from storage before fertilization, a pattern consistent with the kamikaze sperm hypothesis, but short sperm enter sperm storage organs at the same time as long sperm (Snook & Markow, 2001), which would require recognition between rival and self-sperm. Furthermore, long sperm may be better candidates for flushing resident sperm compared with short sperm as they may take up more volume, although this would depend on the proportion of each sperm type in the ejaculate. As with other hypotheses, why a separate sperm class would be necessary for such a function, unless the costs of producing the parasperm were much less than the costs of eusperm (which are

11

unquantified) is unknown. In addition, if sperm competition is frequent in a species, killer sperm could act as a pathogenic agent, inducing resistance to killing in competing males. This resistance may be highly favoured and could invade the population, ameliorating the effects of parasperm. This response may then in turn increase the killing ability of parasperm or could result in selection for a different trait to impact sperm competition. In plants, pollen deposited on the stigmatic surface triggers pollen tube growth through these female tissues, precluding rival pollen from displacing these growing pollen tubes. Indirect evidence of interference between pollen tubes from different donors has been reported in some species (Landi & Frascaroli, 1988; Marshall et al., 1996) possibly in relation to pollen tube growth rate (Lankinen & Skogsmyr, 2002), but has never been linked to differences in pollen morphology or to heteromorphism. In vitro interaction between pollen from different male donors (both in heteromorphic or homomorphic species) was not seen (I. Till-Bottraud, unpublished data). In unfavourable environmental conditions, pollen tube tips can swell or become convoluted (Lush, 1999), but no such morphologies were observed either. Cryptic female choice In the gastropod snail, V. ater, a male’s fertilization success is related to the length of the longer parasperm (Oppliger et al., 2003). This may be because longer parasperm deliver the eusperm faster or can carry more sperm to the site of fertilization. However, females could also be biasing paternity in favour of males with longer parasperm (Oppliger et al., 2003). Parasperm have been suggested as indicator traits, signalling to females either that males are producing healthy eusperm or that males are of superior genetic quality (Swallow & Wilkinson, 2002). Parasperm could also be a form of copulatory courtship, functioning to influence females to use that male’s sperm (Eberhard, 1996). The field of cryptic female choice (CFC) is relatively new and experiments that can distinguish male and female effects on paternity success are difficult to design. However, more studies are finding that paternity success in animals depends on interactions between male’s ejaculates and females (e.g. Miller & Pitnick, 2002; Evans et al., 2003; Snook & Hosken, 2004). There is no a priori reason to dismiss sperm heteromorphism potentially functioning in CFC, however, there is no compelling evidence to suspect parasperm function for this mechanism. The action of parasperm functioning in CFC is probably more likely in organisms with internal fertilization as females may have more control over fertilization compared with species having external fertilization (Eberhard, 1996). As with all hypotheses related to the adaptive significance of sperm heteromorphism, this idea clearly requires experimental evidence and begs

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

12

I. TILL-BOTTRAUD ET AL.

the question of why and what species would evolve parasperm for CFC compared with sperm monomorphic species. Pollen heteromorphism may serve in CFC. There is abundant gene expression in pollen, with approximately 60% of the genes expressed in pollen also being expressed in the plant (Mascarenhas, 1989), and stigmas have an active part in directing or even pulling down pollen tubes through the style (Lord, 2003). Pollen tube growth could be a very good cue for the female to assess individual pollen quality. Indeed, the pollen grain’s genotype influences its fertilization success (Ottaviano et al., 1980), proving a male · female interaction (and thus a possible role in CFC in fertilization success). However, similarly to animals, it is difficult to disentangle female choice from male–male competition, and studies showing some advantage of one pollen type (i.e. bigger pollen grains have faster pollen tubes; Stanton & Preston, 1986; Mulcahy et al., 1988) can be interpreted as either CFC or male–male competition. Outcrossing and precise cross-pollen transfer The hypotheses of outcrossing and cross-pollen transfer are specific to explaining pollen and anther heteromorphisms. Many plants species are bisexual, and this may lead to selfing. Furthermore, in insect-pollinated species, there may be interference between the male and female plant reproductive tissues with respect to pollen transfer. These difficulties require mechanisms to promote precise pollen transfer among individuals. Two such mechanisms are heterostyly (see above) and enantiostyly (Barrett, 2002). Enantiostyly, i.e. left/right positioning of fertile anthers and stigmas, is often coupled with the presence of sterile anthers located in the middle of the flower to position the pollinator. In enantiostyly, the two flower morphologies can be on different plants or on the same plant. Both systems are selected for because pollen from one type of plant or flower is located on the pollinator so as to be deposited on the stigma of the other type of plant or flower, thus providing efficient pollen transfer. In animals, many mechanisms reduce successful interspecific sperm/egg interactions and promote intraspecific interactions. For example, sperm coat proteins evolve rapidly, are species-specific, and determine fertilization ability with an egg (Swanson & Vacquier, 2002). In species with female sperm storage organs, organ length and sperm length can be positively correlated (Briskie & Montgomerie, 1993; Joly & Bressac, 1994; Pitnick et al., 1999; Presgraves et al., 1999; Morrow & Gage, 2000) and influence sperm competitive success (Miller & Pitnick, 2002). An analogous situation to that found in plants, however, would be mechanisms limiting inbreeding that are associated with heteromorphism. Whereas there is some evidence that reproductive proteins (e.g. Riginos & McDonald, 2003) and sperm

(Morrow & Gage, 2001; Snook, 2001; Pitnick et al., 2003) vary intraspecifically, and that this variation may result in reproductive isolation (Pitnick et al., 2003), there is currently no study that addresses the association between these solutions in animals and the presence of sperm heteromorphism. Bet-hedging strategy Pollen heteromorphism is primarily viewed as a bethedging strategy. In Viola, there is a trade-off between rapid germination (pollen with many apertures) and survival (pollen with fewer apertures; Dajoz et al., 1991, 1993; Till-Bottraud et al., 1999). Rapid germination is often seen as the result of pollen competition (Snow & Spira, 1991). If pollen competition occurs on a stigma (i.e. pollen grains from two or more plants arrive simultaneously), then the many aperturate pollen will be the first to germinate and therefore usually reaches the ovules first. However, angiosperm reproduction relies on a third party for pollen transfer that can prove quite unreliable in some conditions. For example, when the weather is too cold, windy or rainy, insects do not fly. The many aperturate pollen grains will have reduced survival in these conditions whereas the fewer aperturate, more slowly germinating, types will survive for subsequent fertilization (Dajoz et al., 1991). The proportions of the different morphs appear to be optimized in different pollination conditions (Till-Bottraud et al., 1994, 1999, 2001). Therefore, pollen heteromorphism appears to be a bet-hedging strategy in response to germination rate and pollination unreliability. The possibility that sperm heteromorphism in animals is a bet-hedging strategy depends on the occurrence of sperm morphologies that are fertilization competent. Whereas species with dimorphic ‘typical’ sperm (i.e. with no apparent structural or fertilization abnormalities) are regularly observed in the animal kingdom, whether DNA content and fertilization capacity varies between types is generally undocumented. Typical does not necessary equate to fertilization competency (Snook et al., 1994; Snook & Karr, 1998). Furthermore, the reproductive biology of many taxa remains unstudied and so the likelihood of bet-hedging strategies operating is unknown. Clearly, the evolutionary significance of dichotomous sperm lines requires their respective fertilization capabilities to be investigated carefully, in concert with the reproductive biology of the organism. If the various intraspecific sperm morphs are demonstrated to be fertilization competent and have different efficiencies in different conditions, then sperm heteromorphism may be analogous to pollen heteromorphism. Subsequently, the question of the survival and capacity to reach the oocyte and the environmental variation of each sperm type would require investigation.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

Conclusion In plants and animals, pollen/sperm heteromorphism arose independently several times across widely distributed taxa and at different phylogenetic levels. Here, we have reviewed the hypotheses suggested to explain the occurrence of these phenomena across both kingdoms. One cannot rule out the hypothesis that sperm/pollen heteromorphism is selectively neutral, i.e. that the genes coding for heteromorphism are not favoured or counterselected by the resulting phenotype (Selosse et al., 2001) but there is substantial circumstantial evidence in both plants and animals that this phenomenon is not neutral. Sperm/pollen heteromorphism appears to be adaptive but its functional significance requires explanations that have proved difficult to determine. There is no a priori reason to predict that either sperm or pollen heteromorphism have a similar adaptive significance or that these hypotheses are mutually exclusive and indeed, several of these hypotheses have been supported in the same taxa (e.g. facilitation and cheap filler in lepidopterans). After 25 years of speculation and some experimentation, no hypothesis is strongly supported, mostly because only limited experimental evidence that tests these hypotheses. More studies on heteromorphism have been performed in animals than in plants, and pollen and sperm heteromorphisms have been studied from different angles. The difference in focus is likely due to disparities in the development and functional equivalence of each sperm/pollen type (e.g. usually all pollen types are fertile and produced in the same mother cell, whereas generally only one of the animal sperm morphs is fertile and each morph is produced in different cysts). Moreover, in animals, the functional role of alternative sperm types involves interactions between sperm of either the same male (e.g. facilitation) or different males (e.g. blocking, cheap filler) whereas in plants each pollen type behaves individually within and between males (e.g. pollen do not interact according to their morphologies). These differences are probably a consequence of variation in haploid gene expression between sperm and pollen, and this variation has likely influenced how sperm and pollen competition is studied; among males for sperm and both within and among males for pollen competition. Hence, whereas aspects of sperm and pollen competition are functionally analogous, the types of relevant data that have been collected differ thus making comparisons regarding the evolutionary significance of sperm/pollen heteromorphism between the kingdoms challenging. Several research avenues need to be pursued to gain a comprehensive picture of heteromorphism across taxa and to assess the degree to which animal and plant heteromorphisms represent either parallel trends or idiosyncrasies. First, hypotheses to explain the adaptive significance of sperm heteromorphism need to be adequately tested. Currently, some hypotheses have

13

intriguing partial evidence of support, but either crucial data on reproductive biology (i.e. female remating frequency and the action of sperm competition) is missing or the critical experiments generally have not been done. These experiments include, but are not limited to: testing whether parasperm are cheaper to produce and delay female remating behaviour, determining whether parasperm is an indicator trait, and examining whether parasperm delay eusperm dispersal thereby increasing male sperm competitive success. To argue that sperm heteromorphism is adaptive, the trait must be empirically linked to increased male fitness compared with when the trait is absent or reduced. No such study has been performed. Secondly, extremely little is known about sperm heteromorphism in nonseed plants in which the three fertilization steps involve sperm. Could evidence from the animal kingdom provide any predictions about its existence? For example, with respect to gamete dispersal, there are ‘plant-like’ animals (Pemberton et al., 2003) but are there nonseed plants that have multiple morphologies of sperm and reproductive biologies that facilitate sexual selection and sperm competition? Thirdly, there is currently no true equivalent of pollen heteromorphism in animals; i.e. different sperm types produced by the same germinal cells, with all sperm fertile, and functioning as a bet-hedging strategy with regard to either the relevant environment or to the female reproductive tract. The difficulty in identifying the equivalent in animals lies in determining the fertilization capacities of various morphologies of sperm. Most studies have not examined whether all sperm types are fertilization competent. Fourthly, all heteromorphisms described here are based on observable morphological differences (e.g. size, shape, presence or absence of a nucleus, aperture number) that translate into variation in fertilization capacity. In animals, one morph is usually fertile and the other(s) sterile. In animals in which both sperm types have the same DNA content and otherwise appear normal, the reason why the sterile type is infertile is unknown. In contrast, the functional differences between pollen morphs are more subtle and correspond to differences in physiological characteristics. In most cases, however, the physiological differences were only identified as a proximal explanation of the morphological differences. With more studies identifying the action of CFC and interactions between sperm and the female reproductive tract determining sperm competition success, the existence of sperm morphs with different physiological capabilities, but with similar morphologies, which could vary between females and/or environments is not impossible. We predict more subtle cases of sperm/ pollen heteromorphism will likely be identified as our technological capacity to examine finer cellular, physiological, and biochemical details increases.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

14

I. TILL-BOTTRAUD ET AL.

Finally, following from some of the previous points, we encourage researchers to examine whether their study organism exhibits putative sperm/pollen hetermorphism, and, if so, to experimentally determine the fertilization capacities of each type. In this context, possible incompatibility systems that have been thoroughly studied in plants (Matton et al., 1994; Williams et al., 1994), but much less in animals (Tregenza & Wedell, 2000), will have to be taken into account to determine fertilization capacity of the different morphs as it is dependent on the female context in which fertility/sterility is studied. There might therefore be situations in which one sperm morphology corresponds to eusperm and another to parasperm in some females, and the reverse in others. We suggest that sperm and pollen heteromorphism arose as a response to the same evolutionary force. The conditions favouring both heteromorphisms appear to be most plausibly linked to sexual selection, that is the risk of gametic competition (e.g. pollen competition on a stigma, or sperm competition in the female reproductive tract) and unpredictability at some stage in the reproductive process (e.g. age of the pollen at deposition or selection by female choice), or male/female interactions (e.g. nuptial gift). Therefore, similar selective forces may have resulted in functionally analogous adaptations associated with sperm and pollen and could provide an example of convergence across kingdoms.

Acknowledgments This paper came about from IT-B and RRS participating in a conference ‘Evolutionary ecology of the prezygotic stage in animals and plants’, held in February 16–21, 2003 at Monte Verita`, Ascona, organized by Giorgina Bernasconi, Barbara Hellriegel, Bernard Schmidt and Io Skosmyr, and funded by Centro Stefano Franscini, ETH Zurich, Swiss Academy of the Natural Sciences, and Swiss National Science Foundation. The authors wish to thank the organizers of the conference for providing the opportunity to embark on this endeavour, organizing this contribution, and patience during its development. Also the authors would like to acknowledge the following people for their help in collecting taxonomic data and kindly sharing their knowledge on the various taxonomic groups: Jean-Marie Demange (Myriapoda), Marco Ferraguti (Annelida and Pogonophora), Jean-Lou Justine (Nematoda), Christopher Tudge (Crustacea) and Vladimir Yushin (Nematoda). The photograph of Vivaparus ater sperm is courtesy of Sonja Sbilordo, University of Zurich. The paper has benefited from critical readings from P.-H. Gouyon, and two anonymous reviewers. This research was supported by an ACC SV3 ‘Ge´ne´tique et Environnement: de´terminants ge´ne´tiques de l’adaptation a` l’environnement’ from the French Research Ministary to DJ and IT-B and the US National Science Foundation to RRS. Irene Till-Bottraud, Dominique Joly and Rhonda R. Snook contributed equally to the work.

References Afzelius, B.A. 1975. The Functional Anatomy of the Spermatozoon. Pergamon Press, Oxford. Afzelius, B.A. 1979. Sperm structure in relation to phylogeny in lower metazoa. In: The Spermatozoon (D. W. Fawcett & J. M. Bedford, eds), Urban and Schwarzenberg, Inc., Baltimore, Munich. Alberti, G. 2000. Chelicerata. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds), Vol. IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B (B. G. M. Jamieson, ed.), pp. 311–387. John Wiley & Sons, Chichester. Amann, R.P. 1989. Can the fertility potential of a seminal sample be predicted accurately. J. Androl. 10: 89–98. Ansley, H.R. 1954. A cytological and cytophotometric study of alternative pathways of meiosis in the house centipede (Scutigera forceps, Rafinesque). Chromosoma 6: 656–695. Ashman, T.L. & Schoen, D.J. 1994. How long should flowers live? Nature 371: 788–791. 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. Baˆ, C.T. & Marchand, B. 1995. Spermiogenesis, spermatozoa and phyletic affinities in the Cestoda. In: Advances in Spermatozoan Phylogeny and Taxonomy (B. G. M. Jamieson, J. Ausio & J.-L. Justine, eds), pp. 87–96. Me´m. Mus. Natl Hist. Nat., Paris. Baker, R.R. & Bellis, M.A. 1988. ‘Kamikaze’ sperm in mammals? Anim. Behav. 36: 936–939. Baker, R.R. & Bellis, M.A. 1995. Human Sperm Competition. Chapman and Hall, London. Barrett, S.C.H. 2002. The evolution of plant sexual diversity. Nat. Rev. Genet. 3: 274–284. Baur, B. 1998. Sperm competition in molluscs. In: Sperm Competition and Sexual Selection (T. R. Birkhead & A. P. Moller, eds), Academic Press, San Diego, USA. Beatty, R.A. 1972. The genetics of size and shape of spermatozoan organelles. In: The Genetics of the Spermatozoan (R. A. Beatty & S. Gluecksohn-Waelsch, eds), pp. 97–115, Edinburgh University Press, Edinburgh. Bena, G. 1998. De la reconstruction phyloge´ne´tique aux hypothe`ses e´volutives: une histoire du genre Medicago L. Institut National Agronomique, Paris-Grignon. Birkhead, T.R. & Møller, A.P. 1998. Sperm Competition and Sexual Selection. Academic Press, San Diego. Bowers, K.A.W. 1975. The pollination ecology of Solanum rostratum (Solanaceae). Am. J. Bot. 62: 633–638. Bressac, C. & Hauschteck-Jungen, E. 1996. Drosophila subobscura preferentially selects long sperm for storage and use. J. Insect Physiol 42: 323–328. Brewbacker, J.L. & Majumder, S.K. 1961. Cultural studies of the pollen population effect on the self-incompatibility inhibition. Am. J. Bot. 48: 457–464. Briskie, J.V. & Montgomerie, R. 1993. Patterns of sperm storage in relation to sperm competition in passerine birds. Condor 95: 442–454. Buckland-Nicks, J. 1995. Ultrastructure of sperm and sperm-egg interaction in Aculifera: implications for molluscan phylogeny. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson, J. Ausio & J.-L. Justine, eds), pp. 129–153. Me´m. Mus. Natl Hist. Nat., Paris.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

Buckland-Nicks, J. 1998. Prosobranch parasperm: sterile germ cells that promote paternity? Micron 29: 267–280. Burighel, P. & Martinucci, G.B. 2000. Urochordata. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Vol. IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part C (B. G. M. Jamieson, ed.), pp. 261–298. John Wiley & Sons, Chichester. Callard, G.V. & Callard, I.P. 1998. Spermatogenesis, in nonmammals. In: Encylopedia of Reproduction (E. Knobil & J. D. Neill, eds), pp. 563–570. Academic Press, San Diego, London. Carcupino, M., Baldacci, A., Fausto, A.M., Scapigliati, G. & Mazzini, M. 1999. Sperm dimorphism in Chilopoda: comparison of Scolopendromorpha and Geophilomopha. Invert. Reprod. Dev. 35: 45–53. Chen, Y.F., Matsubayashi, Y. & Sakagami, Y. 2000. Peptide growth factor phytosulfokine-alpha contributes to the pollen population effect. Planta 211: 752–755. Cook, P.A. & Wedell, N. 1999. Non-fertile sperm delay female remating. Nature 397: 486. Dajoz, I., Till-Bottraud, I. & Gouyon, P.H. 1991. Evolution of pollen morphology. Science 253: 66–68. Dajoz, I., Till-Bottraud, I. & Gouyon, P.-H. 1993. Pollen aperture polymorphism and gametophyte performance in Viola diversifolia. Evolution 47: 1080–1093. Dixson, A.F. & Anderson, M.J. 2002. Sexual selection, seminal coagulation and copulatory plug formation in primates. Folia Primatol. (Basel) 73: 63–69. Eberhard, W.G. 1996. Sexual Selection by Cryptic Female Choice. Princeton University Press, Princeton, NJ. Erdtman, G. 1966. Pollen Morphology and Plant Taxonomy. Hafner, New York. Evans, J.P., Zane, L., Francescato, S. & Pilastro, A. 2003. Directional postcopulatory sexual selection revealed by artificial insemination. Nature 421: 360–363. Faegri, K. & Iversen, J. 1964. Textbook of Pollen Analysis. Munksgaard, Copenhagen. Faegri, K. & Van der Pijl, L. 1966. The Principles of Pollination Ecology. Pergamon Press, New York. Fain-Maurel, M.A. 1966. Acquisitions re´centes sur les spermatogene`ses atypiques. Anne´e Biologique 11–12: 513–564. Ferraguti, M. 2000. Euclitellata. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds.) Vol. IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B (B. G. M. Jamieson, ed.) pp. 125–182. John Wiley & Sons, Chichester. Ferraguti, M. & Balsamo, M. 1995. Comparative spermatology of Gastrotricha. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson & J. Ausio, eds), pp. 105–118. Me´m. Mus. Natl Hist. Nat., Paris. Ferraguti, M., Marotta, R. & Martin, P. 2002. The double sperm line in Isochaetides (Annelida, Clitellata, Tubificidae). Tissue Cell 34: 305–314. Flynn, J.J. & Rowley, J.R. 1971. The primexine of Nelumbo nucifera. Experientia 27: 227–228. Friedla¨nder, M. 1997. Control of the eupyrene-apyrene sperm dimorphism in Lepidoptera. J. Insect Physiol. 43: 1085– 1092. Friedla¨nder, M. & Hauschteck-Jungen, E. 1982. Differential basic nucleoprotein kinetics in the two kinds of Lepidoptera spermatids: nucleate (eupyrene) and anucleate (apyrene). Chromosoma 85: 387–398.

15

Gage, M.J.G. 1994. Associations between body size, mating pattern, testis size and sperm lengths across butterflies. Proc. Roy. Soc. Lond. B 258: 247–254. Gage, M.J.G. & Morrow, E. 2003. Experimental evidence for the evolution of numerous, tiny sperm via sperm competition. Curr. Biol. 13: 754–757. Gilchrist, A.A. & Partridge, L. 1995. Male identity and sperm displacement in Drosophila melanogaster. J. Insect Physiol. 41: 1087–1092. Guzzo, F., Balban, B., Bracco, F. & Mariani, P. 1994. Pollen development in Liriodendron tulipifera: some unusual features. Can. J. Bot. 72: 352–358. Haig, D. & Bergstrom, C.T. 1995. Multiple mating, sperm competition and meiotic drive. J. Evol. Biol. 8: 265–282. Harcourt, A.H. 1991. Sperm competition and the evolution of nonfertilizing sperm in mammals. Evolution 45: 314–328. Hayakawa, Y., Munehara, H. & Komaru, A. 2002a. Obstructive role of the dimorphic sperm in a non-copulatory marine sculpin, Hemilepidotus gilberti, to prevent other males’ eusperm from fertilization. Environ. Biol. Fishes 64: 419– 427. Hayakawa, Y., Akiyama, R., Munehara, H. & Komaru, A. 2002b. Dimorphic sperm influence semen distribution in a noncopulatory sculpin Hemilepidotus gilberti. Environ. Biol. Fishes 65: 311–317. Healy, J.M. 1995. Comparative spermatozoal ultrastructure and its taxonomic and phylogenetic significance in the bivalve order Veneroida. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson & J. Ausio, eds), pp. 155–166. Me´m. Mus. Natl Hist. Nat., Paris. Healy, J.M. 2000. Mollusca: relict taxa. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Vol. IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B (B. G. M. Jamieson, ed.), pp. 21–80. John Wiley & Sons, Chichester. 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. Hodgson, A.N. 1998. Paraspermatozoa. In: Encylopedia of Reproduction (E. Knobil & J. D. Neill, eds), pp. 656–668. Academic Press, San Diego, London. Iqbal, M.C.M. & Wijesekara, K.B. 2002. Cells of the connective tissue differentiate and migrate into pollen sacs. Naturwissenschaften 89: 39–42. Jamieson, B.G.M. 1987. The Ultrastructure and Phylogeny of Insect Spermatozoa. Cambridge University Press, Cambridge. Jamieson, B.G.M. 1999. Spermatozoal phylogeny of the Vertebrata. In: Chapter 29. The Male Gamete: From Basic Science to Clinical Applications (C. Gagnon, eds), pp. 303–331. Cache River Press, Vienna (USA). Jamieson, B.G.M. 2000a. Reproductive Biology of Invertebrates. Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B. John Wiley & Sons, Chichester. Jamieson, B.G.M. 2000b. Reproductive Biology of Invertebrates. Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part C. John Wiley & Sons, Chichester. Jamieson, B.G.M. & Tudge, C.C. 2000. Crustacea-Decapoda. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part C (B. G. M. Jamieson, ed.), pp. 1–96. John Wiley & Sons, Chichester.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

16

I. TILL-BOTTRAUD ET AL.

Jamieson, B.G.M., Ausio, J. & Justine, J.-L. 1995. Advances in Spermatozoal Phylogeny and Taxonomy. Me´m. Mus. Natl Hist. Nat., Paris. Jia, Z.Y., Duan, E.K., Jiang, Z.G. & Wang, Z.W. 2002. Copulatory plugs in masked palm civets: prevention of semen leakage, sperm storage, or chastity enhancement? J. Mammalogy 83: 1035–1038. Joly, D. & Bressac, C. 1994. Sperm length in Drosophilidae (Diptera): estimations by testis length and receptacle lengths. Intl J. Insect Morphol. Embryol. 23: 85–92. Joly, D. & Lachaise, D. 1994. Polymorphism in the sperm heteromorphic species of the Drosophila obscura group. J. Insect. Physiol. 40: 933–938. Joly, D., Cariou, M.L., Lachaise, D. & David, J.R. 1989. Variation of sperm length and heteromorphism in Drosophilid species. Gen. Sel. Evol. 21: 283–293. Joly, D., Cariou, M.-L. & Lachaise, D. 1991. Can sperm competition explain sperm polymorphism in Drosophila teissieri? Evolucion Biologica 5: 25–44. Justine, J.-L. 1995. Spermatozoal ultrastructure and phylogeny in the parasitic Platyhelminthes. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson & J. Ausio, eds), pp. 55–86. Me´m. Mus. Natl Hist. Nat., Paris. Justine, J.-L. & Jamieson, B.G.M. 2000. Nematoda. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B. (B. G. M. Jamieson, ed.), pp. 183–266. John Wiley & Sons, Chichester. Katsuno, S. 1977. Studies on eupyrene and apyrene spermatozoa in the silkworm, Bombyx mori, L. (Lepidoptera: Bombycidae): I. The intratesticular behavior of the spermatozoa at various stages from the 5th instar to the adult. Appl. Entomol. Zool. 12: 142–153. Kreunen, S.S. & Osborn, J.M. 1999. Pollen and anther development in Nelumbo (Nelumbonaceae). Am. J. Bot. 86: 1662–1676. Kubo-Irie, M., Nakazawa, T. & Mohri, H. 1998. Ultrastructural observations of sperm in Cicadidae (Homoptera). In: 8th International Symposium on Spermatology, Montreal, August 17–22. Kubo-Irie, M., Irie, M., Nakazawa, T. & Mohri, H. 2003. Ultrastructure and function of long and short sperm in Cicadidae (Hemiptera). J. Insect Physiol. 49: 983–991. Kura, T. & Nakashima, Y. 2000. Conditions for the evolution of soldier sperm classes. Evolution 54: 72–80. Landi, P. & Frascaroli, E. 1988. Pollen-style interactions in Zea mays. In: Sexual Reproduction in Higher Plants (M. Cresti, P. Gori & E. Pacini, eds), pp. 315–320. Springer, Berlin. Lankinen, A. & Skogsmyr, I. 2002. Pollen competitive ability: the effect of proportion in two-donor crosses. Evol. Ecol. Res. 4: 687–700. Lee, P.E. & Wilkes, A. 1965. Polymorphic spermatozoa in the hymenopterous wasp Dahlbominus. Science 147: 1445–1446. Lord, E.M. 2003. Adhesion and guidance in compatible pollination. J. Exp. Bot. 54: 47–54. Lush, W.M. 1999. Wither chemotropism and pollen tube guidance? Trends Plant Sci. 4: 413–418. Markow, T.A. & Ankney, P.F. 1984. Drosophila males contribute to oogenesis in a multiple mating species. Science 224: 302– 303. Marshall, D.L. & Ellstrand, N.C. 1988. Effective mate choice in wild radish: evidence for selective seed abortion and its mechanism. Am. Nat. 131: 739–756.

Marshall, D.L., Folsom, M.W., Hatfield, C. & Bennett, T. 1996. Does interference competition among pollen grains occur in wild radish, Raphanus raphanistrum? Evolution 50: 1842–1848. Mascarenhas, J.P. 1989. The male gametophyte of flowering plants. Plant Cell 1: 657–664. Matton, D.P., Nass, N., Clarcke, A.E. & Newbigin, E. 1994. Selfincompatibility: how plants avoid illegitimate offspring. Proc. Natl Acad. Sci. U S A 91: 1992–1997. Mazzini, M., Carcupino, M. & Fausto, A.M. 2000. Myriapoda. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part C (B. G. M. Jamieson, ed.), pp. 117–142. John Wiley & Sons, Chichester. Medina, A. 1995. Spermatozoal ultrastrucutre in Dendrobranchiata (Crustacea, Decapoda): taxonomic and phylogenetic considerations. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson & J. Ausio, eds), pp. 231–242. Me´m. Mus. Natl Hist. Nat., Paris. Meves, F. 1902. Ueber oligopyrene und apyrene spermien und uber ihre entstechung nach beobachtungen an Peludina und Pygaera. Archive fur Mikroskopische Anatomie 61: 1–84. Mignot, A. 1995. Contraintes et se´lection dans l’e´volution : le cas du pollen. PhD thesis, Universite´ Franc¸ois Rabelais, Tours. Mignot, A., Hoss, C., Dajoz, I., Leuret, C., Henry, J.-P., Dreuillaux, J.-M., Heberle-Bors, E. & Till-Bottraud, I. 1994. Pollen aperture polymorphism in the Angiosperms: importance, possible causes and consequences. Acta Botanica Gallica 141: 109–122. Mignot, A., Dajoz, I. & Till-Bottraud, I. 1995. Heteromorphism of pollen aperture number and heterogeneity at the tetrad stage in two Viola species and Nicotiana tabacum. C. R. Acad. Sci. 318: 453–457. Miller, G.T. & Pitnick, S. 2002. Sperm-female coevolution in Drosophila. Science 298: 1230–1233. Morrow, E.H. & Gage, M.J.G. 2000. The evolution of sperm length in moths. Proc. Roy. Soc. Lond. B 267: 2333–2337. Morrow, E.H. & Gage, M.J.G. 2001. Consistent significant variation between individual males in spermatozoal morphometry. J. Zool. 254: 147–153. Mulcahy, D.L. & Mulcahy, G.B. 1987. The effects of pollen competition. Am. Scientist 75: 44–50. Mulcahy, D.L., Mulcahy, G.B., Popp, R., Fong, N., Pallais, N., Kalinowski, A. & Marien, J.N. 1988. Pollen selection for stress tolerance or the advantage of selecting before pollination. In: Sexual Reproduction in Higher Plants (M. Cresti, P. Gori & E. Pacini, eds), pp. 43–50. Springer, Berlin. Nadot, S., Ballard, J.H.E., Creach, J.B. & Dajoz, I. 2000. The evolution of pollen heteromorphism in Viola: a phylogenetic approach. Plant Syst. Evol. 223: 155–172. Nei, M., Xu, P. & Glazko, G. 2001. Estimation of divergence times from multiprotein sequences for a few mammalian species and several distantly related organisms. Proc. Natl Acad. Sci. U S A 98: 2497–2502. Oppliger, A., Naciri-Graven, Y., Ribi, G. & Hosken, D. 2003. Sperm length influences fertilization success during sperm competition in the snail Viviparus ater. Mol. Ecol. 12: 485–492. Osanai, M., Kasuga, H. & Aigaki, T. 1987. Physiological role of apyrene spermatozoa of Bombyx mori. Experientia 43: 593– 596. Ottaviano, E., Sari-Gorla, M. & Villa, M. 1980. Pollen tube growth rates in Zea mays: implications for the genetic improvement of crops. Science 210: 437–438.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Pollen and sperm heteromorphism

Pamilo, P. 1991. Life span of queens in the ant Formica excecta. Insects Soc. 38: 111–119. Parga, J.A. 2003. Copulatory plug displacement evidences sperm competition in Lemur catta. Int. J. Primatol. 24: 889–899. Parker, G.A. 1970. Sperm competition and its evolutionary consequences in the insects. Biol. Rev. 45: 525–567. Parker, G.A. 1982. Why are there so many tiny sperm? Sperm competition and the maintenance of two sexes. J. Theor. Biol. 96: 281–294. Parker, G.A. 1993. Sperm competition games: sperm size and number under adult control. Proc. Roy. Soc. Lond. B 253: 245– 254. Parker, G.A. 1998. Sperm competition and the evolution of ejaculates. In: Sperm Competition and Sexual Selection (T. R. Birkhead & A. P. Møller, eds) pp. 3–54, Academic Press, San Diego. Parker, G.A. & Begon, M. 1993. Sperm competition games: sperm size and number under gametic control. Proc. Roy. Soc. Lond. B 253: 255–262. Pasini, M.E., Redi, C.A., Cavaglia, O. & Perotti, M.E. 1996. Ultrastructural and cytochemical analysis of sperm dimorphism in Drosophila subobscura. Tissue Cell 28: 165–175. Pemberton, A.J., Hughes, R.N., Manriquez, P.H. & Bishop, J.D.D. 2003. Efficient utilization of very dilute aquatic sperm: sperm competition may be more likely than sperm limitation when eggs are retained. Proc. Roy. Soc. Lond. B 270: S223–S226. Pitnick, S., Spicer, G.S. & Markow, T.A. 1995. How long is a giant sperm? Nature 375: 109. Pitnick, S., Markow, T.A. & Spicer, G.S. 1999. Evolution of multiple kinds of female sperm-storage organs in Drosophila. Evolution 53: 1804–1822. Pitnick, S., Miller, G.T., Schneider, K. & Markow, T.A. 2003. Ejaculate-female coevolution in Drosophila mojavensis. Proc. Roy. Soc. Lond. B 270: 1507–1512. Polak, M., Wolf, L.L., Starmer, W.T. & Barker, J.S.F. 2001. Function of the mating plug in Drosophila hibisci Bock. Behav. Ecol. Sociobiol. 49: 196–205. Porter, S.D. & Jorgensen, C.D. 1988. Longevity of harvester ant colonies in southern Idaho. J. Range Manage. 41: 104–107. Presgraves, D.C., Baker, R.H. & Wilkinson, G.S. 1999. Coevolution of sperm and female reproductive tract morphology in stalk-eyed flies. Proc. Roy. Soc. Lond. B 266: 1041– 1047. Prunescu, C.C. 1992. The beginning of double spermatogenesis in Scutigera coleoptrata (Chilopoda). Berichte des Natursissenschaftlich-Medizinischen Vereins in Innsbruck Suppl. 10: 93–97. Rebecci, L., Guidi, A. & Bertolani, R. 2000. Tardigrada. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B (B. G. M. Jamieson, ed.), pp. 267–292. John Wiley & Sons, Chichester. Ressayre, A., Godelle, B., Mignot, A. & Gouyon, P.-H. 1998. A morphogenetic model accounting for pollen aperture pattern in flowering plants. J. Theor. Biol. 193: 321–334. Ressayre, A., Godelle, B., Raquin, C. & Gouyon, P.H. 2002. Aperture pattern ontogeny in Angiosperms. J. Exp. Zool. 294: 122–135. Richards, A.G. 1963. Giant sperm in the cockroach, Periplaneta americana. Entomol. News 74: 57–60. Riemann, F. 1986. Nicascolaimus punctatus gen. et sp. n. (Nematoda, Axonolaimoidea), with notes on sperm dimorphism in free-living marine nematodes. Zoologica Scripta 15: 119–124.

17

Riemann, J.G. & Gassner, G. 1973. Ultrastructure of the lepidopteran sperm within the spermathecae. Ann. Entomol. Soc. Am. 66: 154–159. Riginos, C. & McDonald, J.H. 2003. Positive selection on acrosomal sperm protein, M7 lysin, in three species of the mussel genus Mytilus. Mol. Biol. Evol. 20: 200–207. Rouse, G.W. 2000. Polychaeta, including Pogonophora and Myzostomida. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part B (B. G. M. Jamieson, ed.), pp. 81–124. John Wiley & Sons, Chichester. Sahara, K. & Takemura, N. 2002. Double copulation of a female with sterile diploid and polyploid males recovers fertility in Bombyx mori. Zygote 10: 23–29. Sahara, K. & Takemura, N. 2003. Application of artificial insemination technique to eupyrene and/or apyrene sperm in Bombyx mori. J. Exp. Zool 297A: 196–200. Schrader, F. 1960. Evolutionary aspects of aberrant meiosis in some Pentatominae (Heteroptera). Evolution 14: 498–508. Selosse, M.A., Albert, B. & Godelle, B. 2001. Reducing the genome size of organelles favours gene transfer to the nucleus. Trends Ecol. Evol. 16: 135–141. Silberglied, R.R., Shepherd, J.G. & Dickinson, J.L. 1984. Eunuchs: the role of apyrene sperm in Lepidoptera? Am. Nat. 123: 255–265. Simmons, L.W. 2001. Sperm Competition and its Evolutionary Consequences in the Insects. Princeton University Press, Princeton and Oxford. Sivinski, J. 1980. Sexual selection and insect sperm. Fl. Entomol. 63: 99–111. Sivinski, J. 1984. Sperm in competition. In: Sperm Competition and the Evolution of Animal Mating Systems (R. L. Smith, ed.), pp. 86–116. Academic Press, San Diego. Skogsmyr, I. & Lankinen, A. 2002. Sexual selection: an evolutionary force in plants? Biol. Rev. 77: 537–562. Smith, R.L. 1984. Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, San Diego. Snook, R.R. 1997. Is the production of multiple sperm types adaptive? Evolution 51: 797–808. Snook, R.R. 1998. The risk of sperm competition and the evolution of sperm heteromorphism. Anim. Behav. 56: 1497– 1507. Snook, R.R. 2001. Absence of latitudinal clines in sperm characters in North American populations of Drosophila subobscura (Diptera: Drosophilidae). Pan-Pacific Entomol. 77: 261–271. Snook, R.R. & Hosken, D.J. 2004. Sperm death and dumping in Drosophila. Nature 428: 939–941. Snook, R.R. & Karr, T.L. 1998. Only long sperm are fertilizationcompetent in six sperm-heteromorphic Drosophila species. Curr. Biol. 8: 291–294. Snook, R.R. & Markow, T.A. 1996. Possible role of nonfertilizing sperm as a nutrient source for female Drosophila pseudoobscura Frovola (Diptera: Drosophilidae). Pan-Pacific Entomol. 72: 121– 129. Snook, R.R. & Markow, T.A. 2001. Mating system evolution in sperm-heteromorphic Drosophila. J. Insect Physiol. 47: 957–964. Snook, R.R., Markow, T.A. & Karr, T.L. 1994. Functional nonequivalence of sperm in Drosophila pseudoobscura. Proc. Natl Acad. Sci. U S A 91: 11222–11226. Snow, A.A. & Spira, T.P. 1991. Pollen vigor and the potential for sexual selection in plants. Nature 352: 796–797.

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

18

I. TILL-BOTTRAUD ET AL.

Stanton, M.L. & Preston, R.E. 1986. Pollen allocation in wild radish: variation in pollen grain size and number. In: Biotechnology and Ecology of Pollen (D. L. Mulcahy, G. B. Mulcahy & E. Ottaviano, eds), pp. 461–466. Springer, New-York. Storch, V., Kempendorf, C., Higgins, R.P., Shirley, T.C. & Jamieson, B.G.M. 2000. Priapulida. In: Reproductive Biology of Invertebrates (K. G. Adiyodi & R. G. Adiyodi, series eds). Volume IX. Progress in Male Gamete Ultrastructure and Phylogeny. Part C (B. G. M. Jamieson, ed.), pp. 1–19. John Wiley & Sons, Chichester. Sugawara, T. 1979. Stretch reception in the bursa copulatrix of the butterfly, Pieris rapae crucivora, and its role in behavior. J. Comp. Physiol. A 130: 191–199. Swallow, J.G. & Wilkinson, G.S. 2002. The long and short of sperm polymorphisms in insects. Biol. Rev. 77: 153–182. Swanson, W.J. & Vacquier, V.D. 2002. Reproductive protein evolution. Ann. Rev. Ecol. Sys. 33: 161–179. Thibout, E. 1975. Analyse des causes de l’inhibition de la re´ceptivite´ sexuelle et de l’influence d’une e´ventuelle seconde copulation sur la reproduction chez la teigne du poireau, Acrolepia assectella (Lepidoptera: Pluttelidae). Entomol. Exp. Appl. 18: 105–116. Thomson, J.D., Price, M.V., Waser, N.M. & Stratton, D.A. 1986. Comparative studies of pollen and fluorescent dye transport by bumble bees visiting Erythronium grandiflorum. Oecologia (Berlin) 69: 561–566. Till, I., Valdeyron, G. & Gouyon, P.-H. 1989. Polymorphisme pollinique et polymorphisme ge´ne´tique. Can. J. Bot. 67: 538– 543. Till-Bottraud, I., Venable, D.L., Dajoz, I. & Gouyon, P.-H. 1994. Evolution of pollen morphology: a game theory model. Am. Nat. 144: 395–411. Till-Bottraud, I., De Paepe, R., Mignot, A. & Dajoz, I. 1995. Pollen heteromorphism in Nicotiana tabacum (Solanaceae). Am. J. Bot. 82: 1040–1048. Till-Bottraud, I., Vincent, M., Dajoz, I. & Mignot, A. 1999. Pollen aperture heteromorphism: variation in pollen type proportions along altitudinal transects in Viola calcarata (Violaceae). C. R. Acad. Sci. Paris 322: 579–589. Till-Bottraud, I., Gouyon, P.H., Venable, D.L. & Godelle, B. 2001. The number of competitors providing pollen on a stigma strongly influences intraspecific pollen aperture number variation. Evol. Ecol. Res. 3: 231–253. Tregenza, T. & Wedell, N. 2000. Genetic compatibility, mate choice and patterns of parentage: invited review. Mol. Ecol. 9: 1013–1027.

Tudge, C.C. 2003. Endemic and enigmatic: the reproductive biology of Aegla (Crustacea: Anomura: Aeglidae) with observations on sperm structure. Memoirs Museum Victoria 60: 63– 70. Tuzet, O. 1977. La spermatogene`se. In: Traite´ de Zoologie. Anatomie, Syste´matique, Biologie (P. P. Grasset, ed.), pp. 139– 276. Masson, Paris. Verma, L.R. 1974. Honeybee spermatozoa and their survival in the queen’s spermatheca. Beeworld 55: 53–61. Walbot, V. & Evans, M.M.S. 2003. Unique features of the plant life cycle and their consequences. Nat. Rev. Genet. 4: 369–379. Waser, N.M. & Price, M.V. 1984. Experimental studies of pollen carry-over: effects of floral variability of Ipomosis agregata. Oecologia 62: 262–268. Watanabe, S., Hara, M. & Watanabe, Y. 2000. Male internal fertilization and introsperm-like sperm of the seaweed pipefish (Syngnathus schlegeli). Zool. Sci. 17: 759–767. Watson, N.A. & Rohde, K. 1995. Sperm and spermiogenesis of the ‘Turbellaria’ and implications for the phylogeny of the phylum Platyhelminthes. In: Advances in Spermatozoal Phylogeny and Taxonomy (B. G. M. Jamieson & J. Ausio, eds), pp. 37– 54. Me´m. Mus. Natl Hist. Nat., Paris. Wheeler, D.E. & Crichton, E.G. 1990. Comparative ultrastructure of ant spermatozoa (Formicidae: Hymenoptera). J. Morphol. 206: 343–350. Williams, E.G., Clarcke, A.E. & Knox, R.B. 1994. Genetic Control of Self-incompatibility and Reproductive Development in Flowering Plants. Kluwer Academic Publishers, Dordrecht. Willson, M.F. & Burley, N. 1983. Mate Choice in Plants: Tactics, Mechanisms and Consequences. Princeton University Press, Princeton, NJ. Woodward, T.M. 1940. The function of the apyrene spermatozoa of Goniobasis laqueata (Say). J. Exp. Zool. 85: 103–185. World Health Organization 1999. Laboratory Manual for the Examination of Human Semen and Sperm-cervical Mucus Interaction. Cambridge University Press, Cambridge, England. Yushin, V.V. & Zograf, Y.K. 2002. Electron microscopic study of the spermatogenesis in a free-living marine nematode Neochromadora poecilosoma (Chromadorida, Chromadoridae). Biologiya Morya, Vladivostok 28: 47–52. Yushin, V.V., Coomans, A. & Malakhov, V.V. 2002. Ultrastructure of spermatogenesis in the free-living marine nematode Pontonema vulgare (Enoplida, Oncholaimidae). Can. J. Zool. 80: 1371–1382. Received 7 January 2004; revised 4 June 2004; accepted 7 June 2004

J. EVOL. BIOL. 18 (2005) 1–18 ª 2004 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY