Complex Male Mate Choice in Marine Snails

Sara Hintz Saltin

Licentiate thesis Department of Marine Ecology University of Gothenburg Till Mamma, Pappa och Hanna Abstract

The ability to recognise potential mates and choose the best possible mating-partner is of fundamental importance for most . This thesis presents studies of male mate choice within the genus Littorina. Males of this genus are sometimes observed to initiate mating with other males or with females of other species. How such suboptimal mating patterns can evolve is the theme of this thesis. In one study we investigated pre-copulatory- and copulation behaviour in L. fabalis and between this species and its sister-species L. obtusata. We found that males preferred to mount and mate with large and more fecund females rather than small females. Males also preferred to track the largest females mucus trails even though these were trails from another species (L. obtusata) although cross-matings were interrupted before completion. In a second study we found that males of three species (L. littorea, L. fabalis and L. obtusata) preferentially followed female trails. This suggests that females add a “gender cue” in the mucus. In the forth species, L. saxatilis, males followed male and female trails at random. Along with experimental evidence for high mating costs and abilities for male L. saxatilis to detect females of a related species, this suggests a sexual conflict over mating frequency. To reduce number of matings females avoid advertising their sex by disguise their mucus. The reason for the different species strategies is that L. saxatilis lives in much denser populations than the other species and therefore are the least likely to be sperm-limited. Instead, females probably get more than enough matings and disguise their trails in order to reduce the number of the costly matings, thus letting the males search blindly for mates.

Key words: Mate choice, size preference, fecundity, Littorina fabalis, Littorina obtusata, Littorina saxatilis, intraspecific mating, interspecific mating, reproductive barrier, trail- following, mating behaviour, sexual conflict

Populärvetenskaplig sammanfattning

Förmågan att känna igen och välja en lämplig partner är viktig för djur eftersom det påverkar dess möjligheter att fortplanta sig. Detta är därmed en viktig del av evolutionen av arter. Den här avhandlingen är en studie av partnerval inom familjen Littorina, som är en grupp marina snäckor (strandsnäckor). Hanar i den här familjen kan ibland inleda parning med andra hanar eller honor av en annan art. Hur sådana märkliga beteenden kan uppstå är temat för den här avhandlingen som består av två studier. I den första studien undersökte vi parningsbarriärer mellan systerarterna L. fabalis och L. obtusata genom att studera deras parningsbeteende. Vi fann att hanar föredrog att bestiga och para sig med stora och mer fertila honor hellre än små honor. Hanarna följde också gärna slemspåren från stora honor trots att de kom från honor av en annan art, L. obtusata. Därför verkar det som att hanarna inte kan avläsa, från spåret, vilken art som har lagt spåret. Däremot så var parningarna mellan arterna färre och kortare vilket tyder på att det finns någon artigenkännings mekanism som verkar vid närkontakt. I den andra studien så presenterar vi data som tyder på att det finns en sexuell konflikt mellan könen hos arten L. saxatilis. Vi fann att hanarna hos övriga tre svenska arter (L. littorea, L. fabalis och L. obtusata) kunde skilja på hon- och hanspår och att de hellre följde honspår. Det tyder på att det finns någon signal i honornas slemspår som hjälper hanarna att identifiera deras spår och därmed lättare hittar en hona att para sig med (genom att följa slemspåret). Hos den fjärde arten, L. saxatilis, följde hanarna däremot alla spår slumpmässigt, vilket kan förklaras med att honorna av denna art inte avger någon ”honsignal” i slemspåren och då kan inte hanarna skilja på hon- och hanspår. Anledningen till de olika strategierna är att L. saxatilis lever i mycket tätare populationer och risken att få för lite parningar är minimal. Tvärt om så får honorna antagligen mer än nog med parningar och genom att låta bli att skylta med att de är honor så kan de minska kostnaden genom att reducera antalet övertaliga parningarna som en följd av att hanarna söker i blindo, bland både han- och honspår, efter en partner.

Littorina saxatilis. Foto: Patrik Nilsson This thesis is based on the following papers:

Paper I: Saltin S. H, Blom E. L, Johannesson K (2010) Pre-mating behaviours in males of a marine snail; the largest female may not be the best. Manuscript.

Paper II: Johannesson K, Saltin S. H, Duranovic I, Havenhand J. N, Jonsson P. R (2010) Indiscriminate Males: Mating Behaviour of a Marine Snail Compromised by a Sexual Conflict? PLoS ONE 5(8): e12005. doi:10.1371/journal.pone.0012005

Table of contents

Introduction ...... 2 Mate choice ...... 2 The study system of Littorina ...... 3 Reproductive barriers between closely related species...... 3 Sexual conflicts...... 5 Conclusions ...... 6 References ...... 8 Acknowledgements...... 10

……………………………………………………………………………………………….

Paper I

Paper II Introduction

Evolution of reproductive strategies and species recognition are central to all sexual reproducing species. In most species mate choice is part of these processes. This thesis will focus on male mate choice in the genus of Littorina and the aim is to try to explain how the sometimes seemingly suboptimal male mating behaviours can evolve. As will be presented here male mate choice can be complicated by size-preference, which can result in maladaptive interspecific copulations between sister-species. Females that disguise their gender due to a sexual conflict can also hamper male mate search, resulting in males following and initiating copulation with other males. Mate choice within the genus of Littorina is as far as we know today exclusively male mate choice. Males choose both which female to track, which to mount and which to mate with, while the female stays more or less passive during the process. A complete lack of female choice is very unexpected and this observation will be investigated further in future studies. In this thesis I will present the main results of two studies (Paper I and II) and discuss these results in the context of conceptual models and earlier empirical findings. Starting with a brief introduction of the relevance of studying mate choice and the biological characteristics of the genus of Littorina.

Mate choice From an evolutionary perspective, mate choice can be the most important decision in an individual’s life. Some spend their whole life with one partner (Harris 1973) while most animals (not least invertebrates) meet their partner only briefly during mating. Either way choice of mate is fundamental, in particular if the entire life reproduction is invested in one or a few mating opportunities (Foellmer & Fairbairn 2003). Traditionally researchers have been focusing on female mate choice. The reason is that females invest more in each mating in terms of size of the gametes (Bateman 1948) and sometimes parental care (Trivers 1972). With higher investment follow choosiness since the loss from a bad choice will increase with increasing investment. A good mate choice, on the other hand, can increase the fitness of the offspring through enhanced survival and reproductive success for sometimes several generations. Besides investments, there can also be costs involved in mating ranging from costs of mate search to increased mortality during mating (Daly 1978). These costs are not exclusive for the female; the male also often suffer from costs of mating and both male and female tend to be more choosy if there is a high cost involved in mating (Ridley 1983). Discriminate mating (sexual selection) have given rise to many of the most remarkable characters we can see in nature with numerous examples from the avian family, e. g. the tail of the peacock (Marion et al. 1991) and the widowbird (Andersson 1982), or extensive colouration and body-shape in families of teleost fishes (Kodric-Brown 1985; Quinn & Foot 1994). During the evolution of species, sexual selection has been a major selection mechanism influencing morphology, physiology and behaviour traits in many species (Andersson 1994). Of course a fundamental and obvious strategy in mate choice must be to mate with individuals of the opposite sex and of the same species. To achieve this, mate recognition can sometimes be a complex process involving several cues and stimuli (Hankison & Morris 2003). This complexity serves to prevent interspecific matings with the risk of producing inferior hybrids (Coyne & Orr 2004). However, mate choice can sometimes be further

2 complicated by conflicting interests of preferable traits and species recognition or by sexual conflicts between the sexes, and this is the theme of my thesis.

The study system of Littorina The 19 species of the genus Littorina are widely spread in the subarctic and temperate regions of the Northern Hemisphere (Reid 1996). They all inhabit the littoral-zone, but some species live mostly submerged on seaweeds or rocks, while other species are confined to rocks above mean-water and up into the splash-zone. Snails in the upper littoral zone can have long intervals of inactivity, sometimes dried up for weeks during calm summer days (Stafford & Davies 2004). When wetted they move by muscular movements in the foot gliding on secreted mucus. This enables them to actively search for food and to perform mate search. The secreted mucus remains as a trail after the snail, and other snails may follow in the same trail. From previous studies we know that snails can gain information from mucus trails, e.g. gender identity (Johannesson et al. 2010), which enables the males to use the mucus trails during mate search. In Sweden there are four species: Littorina littorea, L. saxatilis, L. fabalis and L. obtusata. All four species have separate sexes and internal fertilisation (Reid 1996). Littorina littorea, which is the species most distantly related to the other three, is the only species in the genus that has pelagic larvae and these are released during spring at the end of the mating period (Reid 1996). Littorina saxatilis is ovoviviparous and is reproductively active year-round, while L. fabalis and L. obtusata have restricted mating seasons from spring to fall, during which they mate and lay egg-masses (Reid 1996). The two latter species are considered sister- species that diverged about 1 Mya (Kemppainen et al. 2009). They are both similar in shape with a flat spire, but adult L. obtusata are mostly larger than L. fabalis and the former also lives several years while the latter lives only one to two years (Reid 1996). The only diagnostic differences between the two species are found in the reproductive organs with males of L. fabalis having an extended tip on the penis compared to L. obtusata (Reid 1996) (Fig. 1). Both L. saxatilis and L. fabalis are polymorphic species, that is, different ecotypes are adapted to different environments. The ecotypes of L. saxatilis in Sweden comprises of the smaller and more fragile exposed morph (e-morph) living on wave-exposed cliffs and the larger and more robust sheltered morph (s-morph) living in micro-sheltered environments behind and beneath boulders. Littorina. fabalis lives on fucoid algae and inhabit sheltered waters, where the “Small-Sheltered” morph (SS-morph) lives, to moderately exposed shore where the “Large-Moderate” morph (LM-morph) occur. These morphs are different in morphology, in shell thickness and size, and to a large part these are inherited differences that seem to evolve easily due to strong potentials of local adaptation (L. saxatilis: Janson 1982; Janson & Ward 1984; Johannesson & Tatarenkov 1997, L. fabalis: Reimchen 1982; Tatarenkov & Johannesson 1998, Kemppainen et al. 2005).

Reproductive barriers between closely related species Speciation is a key area in biological research and also one of the most debated areas, ever since 1859 when Charles Darwin published “On the Origin of Species by means of natural selection” (Mayer 1942; Maynard Smith 1966; Otte & Endler 1989; Coyne & Orr 2004). To be able to understand speciation we first need to address the question of how to define species. The definitions of species are many, but the most common species concept, The Biological Species Concept states that: “Species are groups of actually or potentially interbreeding populations, which are reproductively isolated from other such groups” (Mayr 1942). Reproductive isolation is a central concept in the definition of species, making the

3 study of reproductive barriers an important part of speciation research. Reproductive barriers can be divided in three major groups (Coyne & Orr 2004): I. Pre-mating isolating barriers, which are barriers that prevents gene flow before sperm or pollen is transferred. It could be due to behavioural differences in, for example, courtship or due to mechanical isolation where the genitalia are so different in the two species that mating is not completed. It can also be due to spatial or temporal separation of the two species. II. Postmating, prezygotic isolating barriers act after sperm or pollen transfer, but before fertilization. It could be due to behavioural differences during copulation that prevents fertilization or to incompatible gametes. III. Postzygotic isolating barriers are barriers that act after the formation of a zygote, either by zygote mortality, hybrid inviability or sterility, or hybrid breakdown.

In most cases, when two species have split far back in time, there are many different reproductive barriers and it is difficult to say which barrier was the first and most important for the split. This is the reason why it is good to study speciation using closely related species, or even partially isolated ecotypes, as model organisms. Littorina is therefore an excellent genus to study to learn more about speciation. It is especially interesting to focus on L. fabalis that gives us the unique opportunity to study both reproductive barriers between the sister- species pair (L. fabalis and L. obtusata) as well as reproductive barriers within species, between ecotypes. In Paper I we studied reproductive barriers between L. fabalis and L. obtusata. In that paper mate choice had a central role since we focused on males’ choice of females and their ability to detect conspecific females both during mate search and mating. We found that there are pre-mating isolating barriers between the species acting during direct contact, although L. fabalis males seem to be unable to differ between species from mucus trails. On the contrary, males rather followed females of L. obtusata than conspecific females. This could be explained by males’ preference for large females, since L. obtusata is larger than L. fabalis. In Paper I we also tested and found that L. fabalis males indeed preferred mating with large females of its own species, which concurred with previous studies of other species in the genus showing preference for large females (L. littorea: Saur 1990; Erlandsson & Johannesson 1994) as well as observations of Spanish L. fabalis mating patterns from the field (Rolán-Alvarez et al. 1995). Even though males could not discriminate between species from mucus trails they could do so in close contact with females. That is, we found that L. fabalis males initiated mating with L. obtusata females less often compared with conspecific females - and heterospecific copulations were also interrupted after just a few minutes. Although the mechanism of species recognition remains unknown, it has been observed that males move in a stereotypic way after mounting a female (Saur 1990), which might be a way for males to evaluate their partner and prohibit interspecific copulations. Since we also found that interspecific copulations were interrupted, it is likely that there are species recognition mechanisms acting during copulation. One of the most pronounced morphological difference between the species, and the only diagnostic character, is the dissimilarity in male genitalia, where male L. fabalis have an extended tip that is lacking in male L. obtusata (Fig 1). Although it is speculative, it is possible that the dissimilarity in genital morphology is involved in species recognition.

4

Fig. 1. Male genitalia in L. obtusata (to the left) and L. fabalis (to the right). Source: Reid 1996.

In light of the results from Paper I it is intriguing that males are unable to distinguish between species early in the process; they would obviously benefit from being able to decide the species already from the mucus trail, and in this way avoid following females of the wrong species. One explanation could be that this is not a big problem in nature, e. g. even though their geographic distributions overlap, they could have a microallopatric separation in the shore caused by microscale ecological and behavioural differences. Another possible explanation could be that a species-specific cue has not yet evolved since the colonisation of Swedish waters have happened quite recently (after glaciation, about 10 000 years ago) from allopatric sites where such trait would not be selected for.

Sexual conflicts Various types of sexual conflicts are found in nature, and a common type is if the number of matings to maximize the individuals´ fitness differs between the sexes. The basis for sexual conflicts originate from the definition of the sexes; males having numerous and small gametes whereas females have few and large gametes. Hence females invest more in each offspring than do males (Bateman 1948). Moreover, males generally have the potential ability to produce many more offspring than females (Parker & Simmons 1996). Males therefore often benefit from repeated matings, while females having received enough sperm to fertilize all her eggs may not benefit from further mating. Sometimes females can obtain higher fitness by receiving additional sperm for example by increasing the genetic diversity among the offspring (Jennions & Petrie 2000). There are also other more specific cases where females receive, for example, a nutrition resource by additional matings (Boggs 1995) or where sexual contact is important for social interactions (Savage-Rumbaugh & Wilkerson 1978). But often the costs of the matings exceed the benefit of the mating at a lower mating frequency for females than for males resulting in a sexual conflict over mating frequency (Fig 2).

Fig. 2. Generally individual fitness for females start to decrease at lower mating frequency compared to males.

5 In Paper II we argue that the pattern of mating behaviour we observe in four species of the genus Littorina can be explained by an underlying sexual conflict over mating frequency between the sexes in one of the species. In three of the four Swedish species of Littorina (L. fabalis, L. obtusata and L. littorea) males have the ability to determine if a trail belongs to a male or a female. How this is done is still unknown, but the most likely explanation is that there is a gender specific chemical in the mucus that the males in the three species can detect. In the fourth species, L. saxatilis, the males seem to be unable to distinguish between conspecific males and females. It is likely that the common ancestor to all four species had this ability to distinguish between male and female mucus trails, but for some reason L. saxatilis has lost this ability. This is surprising since being able to rule out half of the trails encountered and only follow female trails seems to be a highly adaptive skill for males. The explanation to this seemingly strange pattern lies in the life-history of these species. The three species that have the ability to detect females (L. fabalis, L. obtusata and L. littorea) have higher or much higher egg production and shorter seasons of mating than L. saxatilis and they also live at lower population densities than L. saxatilis. These differences in life-history all contribute to make L. saxatilis the least sperm-limited species among the four. Indeed, females of L. saxatilis mated in the wild carries offspring sired by around 20 males (Panova et al. 2010). As the cost of mating is high (Paper II) the total cost of mating is higher in L. saxatilis than in any other of the four species, and to reduce this cost selection has favoured females of L. saxatilis that do not revile the sex in the mucus trail. For male L. saxatilis it is still optimal to mate with as many females as possible, as males have no other possibility to increase its inclusive fitness. For females, it is extremely important to survive at least as long as she is brooding her offspring under her shell. This unbalance gives rise to a sexual conflict within L. saxatilis where males try to mate as much as possible and females try to avoid male harassment by disguising their gender. Further strength for this conclusion was provided when we found that L. saxatilis males could as well differ between mucus-trails of male and female L. fabalis. The females of this species leaved a “female mark” in their trails that male L. saxatilis still recognised and followed, showing that males if they have the chance follow “female-marked” trails. This shows that the reason for L. saxatilis males following conspecific males and females at random are not that they do not care for searching for females, it is due to females disguising their trails to promote an overall lower mating frequency. Such sexual conflicts and gender mimicry are not uncommon in nature, especially not in high-density populations (Burley 1981). One example is damselflies where females mimic males in colour and behaviour to reduce number of matings and harassment during mating (Andres et al. 2002). There are many other examples like this among insects and other animals, but Paper II provides us with one of few examples in a marine invertebrate species.

Conclusions

This thesis addressed male mate choice within the genus Littorina and the aim was to try to explain the maladaptive mating-behaviours sometimes observed among males. This adds on to the many examples (summarized by Coyne 2009) showing that nature is not perfect in every sense, which is one of the arguments against the counter theory of evolution – creationism that states that all organisms was created by an intelligent being. Mate choice can influence reproductive barriers and can therefore be important for speciation. As shown in Paper I, mate choice is an important part of the reproductive barriers between the sister-species Littorina fabalis and Littorina obtusata. This finding opens for further studies of L. fabalis mate choice focusing on populations within species, namely the different

6 ecotypes (SS- and LM-morph). This gives us the opportunity to study both complete and ongoing speciation within the same species, even though the existence of reproductive barriers between the morphs is yet to be investigated. Future studies will also be focused on female mate choice since a complete lack of choice for females is highly unlikely. Even if we have no observations suggesting that females choose among the males prior to mating, cryptic female choice through sperm-selection after copulation is still one possibility. That males follow and initiate mating with both females and males (as was reported in Paper II) seems to be maladaptive from the males point of view, but this is probably a great advantage for females since it reduces the mating frequency and thereby also reduces the lifetime costs of mating. Similar sexual conflicts over mating frequency are likely to be found in species where females are unlikely to be sperm limited, for example, species that lives in high population densities with relatively low reproduction rate and long mating season, and for which there are high costs of mating. Although L. saxatilis is the only marine example this far on a sexual conflict over mating frequency, we would expect to find this in additional species with similar demographic characteristics and internal fertilization.

7 References

Andersson M. B (1982) Female choice selects for extreme tail length in a widowbird. Nature 299: 818-820 Andersson M. B (1994) Sexual Selection. Princeton University Press. Andres J. A, Sanchez-Guillen R. A, Rivera A (2002) Evolution of female colour polymorphism in damselflies: testing the hypotheses. Anim. Behav. 63: 677-685. Bateman A. J (1948) Intra-sexual selection in Drosophila. Heredity 2: 349-368 Boggs C. L (1995) Male nuptial gifts: Phenotypic consequences and evolutionary implications. Pages 215–242 in S. R. Leather and J. Hardie, editors. Insect reproduction. CRC Press, New York, New York, USA. Burley N (1981) The evolution of sexual indistinguishability. In R. D Alexander and D. V Tinkle (eds.), Natural Selection and Social Behaviour: Recent Research and New Theory. Chiron, New York / distributed outside North America by Blackwell Scientific, Oxford. Coyne J. A (2009) Why Evolution is True. OUP Oxford. Coyne J. A, Orr H. A (2004) Speciation. Sunderland mass.: Sinauer Associates, Inc. publishers. Daly M (1978) The Cost of Mating. American naturalist 112: 771-774 Erlandsson J, Johannesson K (1994) Sexual selection on female size in a marine snail, Littorina littorea (L.). J. Exp. Mar. Biol. Ecol. 181: 145-157 Foellmer M. W, Fairbairn D. J (2003) Spontaneous male death during copulation in an orb- weaving spider. Proc. R. Soc. Lond. B. 270:183-185 Hankison S. J, Morris M. R (2003) Avoiding a compromise between sexual selection and species regognition: female swordtail fish assess multiple species-specific cues. Behav. Ecol. 14: 282-287 Harris M. P (1973) The biology of the waved albatross Diomeda irrorata of Hood Island, Galapagos. Ibis 115: 483-510 Janson K (1982) Phenotypic differentiation in Littorina saxatilis (Olivi) (, Prosobranchia) in a small area on the Swedish west coast. J. moll. Stud. 48: 167- 173 Janson K, Ward R. D (1984) Microgeographic variation in allozyme and shell characters in Littorina saxatilis (Olivi) (Prosobranchia: ). Biol. J. linn. Soc. 22: 289- 307 Johannesson K, Saltin S. H, Duranovic I, Havenhand J. N, Jonsson P. R (2010) Indiscriminate Males: Mating Behaviour of a Marine Snail Compromised by a Sexual Conflict? PLoS ONE 5(8): e12005. doi:10.1371/journal.pone.0012005 Jennions M. D, Petrie M (2000) Why do females mate multiply? A rewiew of the genetic benefits. Biol. Rev. Camb. Philos. Soc. 75: 21-64 Johannesson K, Tatarenkov A (1997) Allozyme variation in a snail (Littorina saxatilis) – deconfounding the effects of microhabitat and gene flow. Evolution 51: 402-409 Kemppainen P, Nes S, Ceder C, Johannesson K (2005) Refuge function of marine algae complicates selection in an intertidal snail. Oecologia 143:402-411 Kemppainen P, Panova M, Hollander J, Johannesson K (2009) Complete lack of mitochondrial divergence between two species of NE Atlantic marine intertidal gastropods. J. Evol. Biol. 22: 2000 -2011 Kodric-Brown A (1985) Female preference and sexual selection for male coloration in the guppy (Poecilia reticulata). Behav. Ecol. Sociobiol.17: 199-205

8 Marion P, Halliday T, Sanders C (1991) Peahens prefer peacocks with elaborate trains. Anim. Behav. 41(2): 323-331 Maynard Smith, J (1966) Sympatric speciation. Am. Nat. 100: 637-650 Mayr E (1942) Systematics and the Origin of Species. Columbia University Press, New York. Otte D, Endler J. A (1989) Speciation and its Consequences. Sinauer Associates, Sunderland, Mass. Panova M, Boström J, Hofving T, Areskoug T, Eriksson A, Mehlig B, Mäkinen T, André C, Johannesson K (2010) Extreme female promiscuity in a non-social invertebrate species. PLoS ONE 5(3):e9640. doi:10.1371/journal.pone.0009640 Parker G. A, Simmons L. W (1996) Parental investment and the control of sexual selection: Predicting the direction of sexual competition. Proc. Biol. Sci. 263: 315-321 Quinn T. P, Foote C. J (1994) The effects of body size and sexual dimorphism on the reproductive behaviour of sockeye salmon, Oncorhynchus nerka. Anim. Behav. 48(4): 751-761 Reid, D. G (1996) Systematics and Evolution of Littorina. The Ray Society, Dorchester Reimchen T. E (1982) Shell size divergence in Littorina mariae and L. obtusata and predation by crabs. Can. J. Zool. 60: 687- 695 Ridley M (1983) The explanation of organic diversity. The comparative method and adaptations for mating. Clarendon Press, Oxford, UK. Rolán-Alvarez E, Zapata C, Alvarez G (1995) Multilocus heterozygosity and sexual selection in a natural population of the marine snail Littorina mariae (: Prosobranchia). Heredity 75: 17-25 Saur M (1990) Mate discrimination in Littorina littorea (L.) and L. saxatilis (Olivi) (Mollusca: Prosobranchia). Hydrobiologia 193: 261-270 Savage-Rumbaugh E. S, Wilkerson B. J (1978) Socio-sexual behavior in Pan paniscus and Pan troglodytes: A comparative study. J. Hum. Evol. 7(4): 327-344 Stafford R, Davies M. S (2004) Temperature and desiccation do not affect aggregation behaviour in high shore littorinids in north-east England. JNR EEB 1: 16-20 Tatarenkov A, Johannesson K (1998) Evidence of a reproductive barrier between two forms of the marine periwinkle Littorina fabalis (Gastropoda). Biol. J. Linnean soc. 63: 349-365 Trivers R. L (1972) Parental investment and sexual selection. In B. Campbell (Ed.) Sexual selection and the descent of man, 1871-1971 (pp 136–179). Chicago, Aldine.

9 Acknowledgements

Först vill jag tacka min superba handledare Kerstin Johannesson, som är den bästa läraren, medarbetaren och föredömet. Dessutom är hon en fantastisk fin människa och det är gaska otroligt att ha en handledare som förbihållslöst alltid sätter mig som doktorand högst på sin långa prioriteringslista. Jag skulle kunna fortsätta att skryta en halvsida till om Kerstin men jag vet att hon antagligen bara skulle påpeka att jag ska korta ner den biten eftersom det trots allt bara är rena spekulationer  Jag har ju också andra handledare. Per och Gunilla som är riktiga klippor vad det gäller statistik och experimentdesign. Det känns väldigt skönt att kunna vila på er ibland när det gäller sådana svåra men ack så viktiga bitar! Tack för all hjälp! Jag har riktigt fina kollegor också; ingen nämnd och ingen glömd brukar det ju heta… Vill ändå passa på att skicka kramar till några speciella vänner; Anna-Lisa och Christin. Sedan finns det också några starka, vackra kvinnor som inspirerar mig; Eva-Marie, Anette, Marina, Ann, Helena, Elisabeth, Anita och Sanna. Och så doktorand-gänget förståss! Robin, Angelica, Daniel, Swantje, Geno, Finn, Micke, Erika, Per och Elin. Tack även till mina studenter som jag handlett; Martin, Sara, Emmelie och speciellt Eva-Lotta som har bidragit till den här avhandlingen! Sist, men egentligen först; min härliga familj! Thomas och Sebastian som alltid står vid min sida och vars stöd jag inte står utan. Mycket kärlek tack för allt stöd också till storfamiljen; Mamma och Pappa; Hanna, Marcus och Ester; Mormor och Morfar; Ingegärd, Gunnar, Anna, Emma, Anders och Lisa.

10

Female choice in snails….?

Paper I Pre-mating behaviours in males of a marine snail; the largest

female may not be the best

Sara Hintz Saltin*, Eva-Lotta Blom and Kerstin Johannesson

Department of Marine Ecology – Tjärnö, University of Gothenburg, SE-452 96 Strömstad, Sweden *Corresponding author. Email: [email protected]

Abstract Mate choice is central to fitness of organisms and hence a key trait in the evolution of most species. As egg production is more costly than sperm production, females are usually the most choosy sex. Nevertheless, male mate choice is expected in species where males have extensive costs of mating. In snails mate search is time- and energy consuming for the males and great benefits can be obtained from being choosy early in this process, for example when choosing a mucus trail to follow. We study male mate choice in the marine snail Littorina fabalis using video recordings and live observations of pre-copulation and copulation behaviours. Upon encounter, males preferred to mate large and more fecund females over small females. When offered a choice of different sized females they most frequently tracked mucus trails of the largest females although these were of the larger sister-species Littorina obtusata. However, copulations between the two species were rare and considerably shorter than intraspecific copulations suggesting species-recognition mechanisms that prevent interspecific copulations. Hence our study shows evidence of a trade-off for males that tend to choose female traits from large females over those of small females. Although fecundity increases by female size, there is also an increased risk to spend time and energy on tracing a female of another species.

Key words: Sexual selection, mate choice, size preference, fecundity, Littorina fabalis, Littorina obtusata, intraspecific mating, interspecific mating, reproductive barrier, trail- following, copulation time, mounting time

1 Introduction Sexual selection and mate choice are of central importance in the evolution of mate recognition systems as well as in the evolution of species (Ratcliffe & Grant 1983; Wiernasz & Kingsolver 1992; Andersson 1994; Owens et al. 1999; van Doorn et al. 2009). The traditional view of mate choice is that females choose which male to mate with and males mate at every possible occasion. However, indiscriminate mating of the males should occur only if the benefit of indiscriminate mating is greater than the costs of mating and both male and female tend to be choosier about their mates in proportion to how much time and energy they invest in mating (Ridley 1983). In many different taxa males prefer to mate a large female rather than a small one. This is explained by the generally higher fecundity of large females (Parker 1970; Ridley 1983; Crespi 1989). Earlier studies have shown that in the marine snail Littorina there is sexual selection favouring large female size (L. littorea, Saur 1990; Erlandsson & Johannesson 1994; L. fabalis, Rolán-Alvarez et al. 1995b) and this is likely a consequence of a positive correlation between fecundity and female size (L. littorea, Hughes & Answer 1982; L. saxatilis, Janson 1985). Littorina fabalis and Littorina obtusata are sister-species with a large overlap in their geographic distributions, as well as a partial overlap in the microgeographic distribution on the shore; L. obtusata typically prefers either upper parts or more wave-protected parts of the shore while L. fabalis dwells in the low intertidal and in both wave-protected and more wave- exposed areas. Truly sympatric distributions are found in wave-protected parts of shore and in mid-intertidal levels (Reid 1996). The two species have also overlapping breeding seasons, which stretches from spring to autumn. It can be difficult to tell the two species apart since their colour can be similar and they have overlapping size ranges and shell morphologies, although L. obtusata are generally larger than L. fabalis and the latter has a more flat apex. The species are best discriminated from the morphology of the genital characters (Reid 1996), and in addition there are several diagnostic allozyme loci (Zaslavskaya et al. 1992; Tatarenkov 1995; Rolán-Alvarez et al. 1995a). According to genetic data the species do not currently hybridise, although they share most of the common mitochondrial haplotypes, suggesting either that hybridization has taken place after their first separation, or that the two species still share ancestral mtDNA haplotypes (incomplete lineage sorting) (Kemppainen et al. 2009). As hybrids between the sister-species never have been encountered, it is likely that there are one or several isolating mechanisms between the two species that prevent hybridisation. To depict reproductive barriers between sister-species is fundamental to understand how reproductive isolation can evolve between closely related species. Littorina fabalis is a polymorphic species and two genetically differentiated morphs or ecotypes, the LM-morph and the SS-morph are found in e.g. UK, France, Norway and Sweden (Reimchen 1982; Tatarenkov & Johannesson 1998; Kemppainen et al. 2005). In Sweden, LM-morph snails are larger and live in moderately wave-exposed habitats and the

2 smaller SS-morph snails live in sheltered habitats, often in sympatry with L. obtusata. In the zones of species overlap interspecific matings are occasionally observed (pers. obs.). Snails move by muscular movements in the foot and they glide on mucus that enables this way of locomotion, leaving a trail of mucus behind. It has been known for decades that snails often choose to follow other snails’ trails. The reasons for this behaviour have been addressed in experimental studies over the years and many explanations have been offered. One explanation could be that this behaviour has evolved as an adaptation to save energy since production of mucus can be quite costly and snails can sometimes invest as much as 23-31 % of consumed energy for mucus production (Edwards & Davies 2002). Moreover, Davies and Blackwell (2007) found that snails that follow trails can save considerable amount of their own mucus production since they only have to produce 27% of the usual amount of mucus when they move in a previously laid trail. Besides being an important part of locomotion, following in another snails mucus trail has also been suggested to provide the snails with a chance to graze microalgae (Davies & Beckwith 1999) and/or bacteria (Peduzzi & Herndl 1991) attached to the sticky mucus. Trail-following have also been proposed to be important for mate finding since males use mucus trails to track females (Erlandsson & Kostylev 1995). For example, males of three species of Littorina (L. littorea, L. fabalis and L. obtusata) follow females significantly longer then they follow males (Johannesson et al. 2010), suggesting that trail-following is linked with mate search. Most likely, trail-following have many advantages in snails, but in this study we study male Littorina fabalis tracking of female mucus trails as part of their mate-searching behaviour. Copulating time is another factor that have been used to find out if a male prefers a certain female (Hollander et al. 2005), and the rationale behind this is that much shorter copulations are observed when males mate improper mates such as another male or a juvenile or another species (Saur 1990; Erlandsson 1998). It is also possible that the time a male spend in copula position is positively correlated to the percent of eggs fertilized by that male, as is the case in another invertebrate taxon (Parker 1970). In this study we experimentally test L. fabalis male mate choice and we ask the question if in the first place there is mate choice, and secondly, if this choice is likely to improve male fitness. More specifically we were interested in if male L. fabalis would choose large (and more fecund) females over small females, and if so, how they then would avoid the problem of tracking (large) females of the partially sympatric sister-species (L. obtusata).

Materials and methods

Trail-following experiment

To test whether male Littorina fabalis would choose between females of different size classes, and between females of their own and the sister-species (L. obtusata) we performed

3 an experiment in which male choice was assessed both as trail-following and as mounting time. Snails of L. fabalis (SS-morph, the ecotype that lives in sympatry with L. obtusata) and L. obtusata were sampled at two sites along the shores of two small islands (Långholmen and Lökholmen; west of Tjärnö Marine Biological Laboratory). Sampled females of L. fabalis were measured and sorted in two size intervals (large; 9-14 mm and small; 7-9 mm) while all females of L. obtusata was of size 13-15 mm which is normal for this species and partly overlapping in size with the largest females of L. fabalis. In each of 18 replicate experimental runs, six males and six females were allowed to move freely in a square shaped arena (390*390 mm) and their movements were recorded with digital video over a period of 15 minutes. The six males were all the same species (L. fabalis) and these were of random sizes in the interval 6-12 mm. In each run the males were given equal chance to choose between females representing both different sizes and species as the setup of females were: two large L. fabalis, two small L. fabalis and two L. obtusata. All snails within a run were from the same locality to maintain familiarity among the snails. The surface of the arena was wetted with seawater and rinsed carefully between each run. Male activity (trail-following and mounting) were recorded and analysed from the video using computerised motion analysis (CellTrak for Windows, Motion Analysis Corp.). Male movements were defined as trail-following when following a female trail for more than 3 snail diameters. All other paths (including those along the edges of the arena) were excluded from the analyses. It was not possible to detect copulations with any certainty from the films due to inadequate resolution. Instead we used “mounting-time” as a proxy for the male’s time-investment in the female. Mounting time can include copulations although not all mountings lead to copulation. Mounting time was defined as the time the male stayed on a female shell after crawling up on her shell. Hence from each run we received the total distance male trackers followed female trails from the three different groups of females, as well as the total mounting time for each group of females. We pooled the results from all six males of the same run. The reason for doing this was that each trail following could not be considered as a replicate since the action of one snail could potentially affect the other snails in the same arena and hence the level of replication in our experiments was a run. We compared the result between two female groups at a time using a two-tailed binomial test. Including three groups of females in the experiment even though we only compared two groups at a time was done to provide the males with a free choice of different females, similar to what males would encounter in nature. The null-hypothesis prior to the experiment was that males would not have any preference for any group of female (large L. fabalis, small L. fabalis or L. obtusata). Since we had prior observations that L. fabalis males readily aggregated with L. obtusata females we wanted to keep the options for any preference open in this experiment by using a two-tailed test.

4 Copulation time observations

Size preference As in nature the first experiment involved both female size and species as factors influencing male mate choice. As a complement, we performed additional experiments in which we treated one factor at a time. Thus in one experiment we tested whether male Littorina fabalis showed any size preference of conspecific females. In this experiment we measured “copulating time”, which is a more direct estimate of mating effort than “mounting time” used in the previous experiment. To be able to estimate copulation time we conducted direct observations of snail matings instead of using film recordings. Prior to the experiments snails of L. fabalis were sampled, sized and sorted as described earlier. Twenty runs were performed, each including ten male and ten female L. fabalis (five large and five small, same size classes as before). All snails within a run were from the same locality. Prior to the runs the males were individually marked with nail polish so that it would be possible to follow each male. During a run, females and males were able to move freely in a circular arena (Ø 23 cm) filled with seawater. During the runs the sexual activity between males and females and the size of the female in action was noted as well as the time the pairs spent in copula position (Saur 1990). The runs lasted for one hour, but copulating pairs active when the hour had passed were followed until the male terminated the copulation. If a snail crawled out of the container it was picked up and put back in the centre of the arena. Between each run the arena was cleaned to eliminate contamination by pheromones and mucus trails. From each run we received a total copulating time for males copulating with large females and a total copulating time for males copulating with small females. Hence the level of replication in this experiment was a run and we used a two-tailed binomial test to address the hypothesis that males would prefer to mate a female of a specific size (large or small) for longer time than the other group.

Species preference: The second of the complementary experiments was designed to test whether L. fabalis males would prefer to mate females of their own species rather than females of L. obtusata. We also studied male L. obtusata mate choice. In this experiment we again measured copulation time from live observations of mating pairs. In each of six aquaria (50x40x40 cm) containing seaweed ( vesiculosus and F. serratus), fifteen snails of each sex and species (L. obtusata and L. fabalis) were placed. The aquaria were kept outdoors and supplied with running seawater. The reason for this design, compared to the design in previous experiment, was to provide a more natural environment for the interactions of the two species. Snails from different sites were kept in different aquaria to maintain familiarity among snail groups. The snails in this experiment were of random adult sizes; 6-14 mm for L. fabalis and 12-15 mm for L. obtusata. The aquaria were

5 inspected one at a time in a random order and were as gentle as possible searched through not to disturb any mating couples. As soon as a couple was found mating, inter- or intraspecific, copulation time was measured. It was only ongoing matings that were measured, if a couple started mating while the aquaria was inspected this couple was ignored. Only one aquarium at a time was inspected and as soon as all couples that were observed had finished mating the next aquarium was inspected in the same way. Thus we were not able to record the total copulation time for a mating, but the error is randomly distributed over all the observed pairs. This means that “copulation time” in this experiment is a proxy of the real copulation time, it could, nevertheless, be used as a relative measurement for comparison between interspecific and intraspecific copulations in this experiment. To test if there was a difference in copulation time within and between species a one-way analysis of variance (ANOVA) was done In this experiment all copulations were recorded, hence we also recorded the copulations performed by males of L. obtusata although they never attempted to mate with heterospecific females. The groups compared was therefore; male L. obtusata copulating with female L. obtusata, male L. fabalis copulating with female L. fabalis and male L. fabalis copulating with female L. obtusata. A posteriori test (Tukey´s test) was used to evaluate the differences between the three groups. The level of replication was aquaria and to avoid confounding results due to repeated measurements we used data from only one group per aquaria by randomly selecting 2 aquaria per group, in the statistical analysis. The hypothesis was that male L. fabalis would prefer to mate with females of its own species rather than with females of L. obtusata.

Results Male L. fabalis preferences during trail-following and mounting

In the experiment where males of Littorina fabalis were accompanied by both small and large females of their own species and females of the larger Littorina obtusata there was no significant difference between distances that L. fabalis males followed large and small conspecific females (binomial test, P = 0.5; Fig. 1). Indeed the data tentatively suggest that the males followed females of different sizes at random, although sample size of this test was smaller than in the other tests and the results should be considered as preliminary. When the males encountered a female, however, they mounted large females more frequently and/or stayed with them longer, resulting in significantly longer total mounting time in each run, compared with small females (binomial test, P = 0.001; Fig. 2). Unexpectedly, we also found that L. fabalis males not only followed females of the sister- species L. obtusata, but they followed them significantly longer distances than they followed both small and large females of their own species (binomial test, P = 0.003; Fig. 3 and P = 0.001; Fig. 4). Mounting time was however not significantly different between interspecific

6 pairs and intraspecific pairs (in binomial tests of mounting time: P = 0.09 for males mounting small L. fabalis females compared to L. obtusata females, and P = 0.4 for males mounting large L. fabalis females compared to L. obtusata females, Figs. not shown).

Male L. fabalis copulation time with females of different sizes and species

In the experiment in which males of L. fabalis were mixed with females of L. fabalis of two different size-groups, copulating time was significantly longer for males mating large compared to small females (binomial test, P = 0.002; Fig. 5). In the experiment where males and females of both species were mixed in equal numbers interspecific matings occurred (exclusively L. fabalis males that mated L. obtusata females) but was much fewer than intraspecific matings (16 % of all matings were between species). The average copulation time for an interspecific copulation was also shorter; 5 minutes compared with 20-23 minutes for intraspecific couples. Consequently, total copulation time between species was significantly smaller than that within species (1-factor ANOVA, F2, 5 = 16.52, P = 0.02). This result is illustrated in figure 6 where we can see that the total copulating time for both intraspecific couples is significantly higher than for mixed couples (Tukey´s test, P = 0.03).

250

200 (mm)

150 L. fabalis 100

50

Tracking of small 0 0 50 100 150 200 250 Tracking of large L. fabalis (mm)

Figure 1. Distances that male trackers followed small L. fabalis (y-axis) and large L. fabalis (x-axis) females. Each point represents the total distance of trail-following of 6 males of one replicate run (15 min). The diagonal indicate the expectation of equal tracking distances of both groups of females. Two-tailed P-values of binomial tests are indicated.

7 1200

(sec) 1000

800 P = 0.001 L. fabalis 600

400

200 Mounting of large 0 0 200 400 600 800 1000 1200 Mounting of small L. fabalis (sec)

Figure 2. Time (sec) that males spend mounting small L. fabalis (x-axis) and large L. fabalis (y-axis) females. Each point represents the total mounting time of 6 males of one replicate run (15 min). The diagonal indicate the expectation of equal mounting time of both groups of females. Two-tailed P-values of binomial tests are indicated.

700

600 (mm) 500 P = 0.003

400

L. obtusata 300

200

100 Tracking of

0 0 100 200 300 400 500 600 700 Tracking of small L. fabalis (mm)

Figure 3. Distances that male trackers followed small L. fabalis (x-axis) and L. obtusata (y-axis) females. Each point represents the total distance of trail-following of 6 males of one replicate run (15 min). The diagonal indicate the expectation of equal tracking distances of both groups of females. Two-tailed P-values of binomial tests are indicated.

8 700

600 (mm) 500 P = 0.001 400

L. obtusata 300

200

100 Tracking of 0 0 100 200 300 400 500 600 700 Tracking of large L. fabalis (mm)

Figure 4. Distances that male trackers followed large L. fabalis (x-axis) and L. obtusata (y-axis) females. Each point represents the total distance of trail-following of 6 males of one replicate run (15 min). The diagonal indicate the expectation of equal tracking distances of both groups of females. Two-tailed P-values of binomial tests are indicated.

80

70

L. fabalis 60

50 P = 0.002

40 (min) 30

20

10

Copulating time with large 0 0 10 20 30 40 50 60 70 80 Copulating time with small L. fabalis (min)

Figure 5. Time (min) that males spend copulating with small L. fabalis (x-axis) and large L. fabalis (y-axis) females. Each point represents the total copulating time of 10 males of one replicate run (60 min). The diagonal indicate the expectation of equal copulating time of both groups of females. Two-tailed P-values of binomial tests are indicated.

9

700

600

500

400

300

200

Total copulang me (min) 100

0 Intraspeciic copulation Intraspeciic copulation Interspeciic copulation for L. obtusata for L. fabalis for L. fabalis male and L. obtusata female

Figure 6. Total time for intra- and interspecific mating, with standard error bars. Post-hoc test (Tukey´s test) show significant differences (P = 0.03) between both intraspecific mating means compared with interspecific mating mean.

Discussion In this study we found that Littorina fabalis males prefer large females since males significantly mount and mate longer with large females compared to small females. This supports earlier observations of Spanish L. fabalis mating patterns from the field (Rolán- Alvarez et al. 1995b). Similar size preference is also known from studies of other species of the genus Littorina (L. littorea, Saur 1990; Erlandsson & Johannesson 1994). A most likely explanation is that large females of Littorina are more fecund than small females and mating a large female thus result in more offspring being sired (L. littorea, Hughes and Answer 1982; L. saxatilis, Janson 1985). Previous studies have shown that trail-following is an important way of finding a mate (Erlandsson & Kostylev 1995; Johannesson et al. 2010). In our investigations of snail trail- following behaviour we did not find any preference for size when male L. fabalis tracked different sized females of their own species. Part of the explanation could be that the size difference between the two size classes of females was too small to be appreciated by the males from the trail, although size differences were indeed detectable when mates were in physical contact prior and during mating. An alternative explanation to the lack of a choice in tracking small and large females may have been that our experiment was too small, that is, we had to few measurements of intraspecific trail-following. In contrast, L. fabalis males

10 followed the heterospecific Littorina obtustata females significantly longer than their conspecific females. The evolutionary rationale behind this observation seems puzzling, but one explanation could be that males are unable to separate conspecific females from females of the sister-species L. obtustata from their mucus trails, and that the L. obtusata female trails were only interpreted as coming from a large (and therefore preferable) female. Alternatively, males followed these large trails with other purposes than encountering a mate, for example, feeding or moving while saving own mucus production (Peduzzi & Herndl 1991; Davies & Beckwith 1999; Davies & Blackwell 2007). On the other hand, we would have expected also female L. fabalis to follow trails of L. obtusata for these purposes, but this was not the case. Preference for large females together with the lack of a species specific cue in the mucus could have the side effect of occasional fallacies of tracking females of the wrong species; L. obtusata. On the other hand, most of the time this preference for large females would probably compensate this disadvantage by increasing fitness through favourable matings with large and more fecund conspecific females. In this context it is interesting to consider why no species specific cue in the mucus trail have evolved? If the males were able to identify their own species from mucus trails, the waste of time and energy following and mounting females of another species could have been avoided. There are plenty of examples of the existence of species-specific pheromones in gastropods (Croll 1983) as well as in other taxa (copepods: Frey et al. 1998, fish: Plenderleith et al. 2005). Indeed, we recently found that males of several Littorina species including L. fabalis can discriminate between male and female trails from a gender-specific cue in mucus (Johannesson et al. 2010), showing that these type of presumably chemical cues have been evolved also in this genus. Notably, however, previous study found that L. saxatilis males followed female rather than male trails of L. fabalis indicating a lack of a strong species-recognition cue also between these species (Johannesson et al. 2010). The two species Littorina fabalis and L. obtusata have a relatively recent common ancestry about 1 million years ago (Kemppainen et al. 2009). Their current geographic distributions are to a large extent overlapping, although L. obtusata is the only species present along the North American coast (Reid 1996). However, it is unclear if they speciated in sympatry or in allopatry and possibly large parts of their postglacial distribution (including Sweden) is a result of a recent colonization about 10 000 years ago from allopatric sites in which species specific cues would not be selected for. Then it is possible that such cue has not yet evolved even F it would be a favourable adaptation. An alternative explanation could be that the geographic sympatry is nevertheless accompanied by a microallopatric distribution in the shore caused by microscale ecological and behavioural differences. In this study we also looked at copulation time and found that intraspecific couples copulated significantly longer (average 20-23 minutes) compared with interspecific couples (average 5 minutes) and the latter category was also less frequent. It seems likely that a short copulation result in considerably less or no sperm-transfer, as sperm-transfer is a time-consuming

11 process driven by ciliary movements (Buckland-Nicks pers. commun.). Thus it seems as if any post-zygotic barrier that is likely to be present (as hybrids are not found) is accompanied by a pre-copulatory reproductive barrier that comes into action after the trail-following phase. It is likely that there are one or several species recognition mechanisms working during close contact, counteracting interspecific copulation. A similar finding of heterospecific attraction due to size preference in swordtail fish (Hankison & Morris 2003) emphasises the importance of multiple cues in species recognition. In that study females rather choose large heterospecific males when accessing size and chemical cue and only choosing smaller conspecific males provided with multiple cues (chemical and skin-pattern). This study gives a view of the complex mechanisms contributing to reproductive barriers in early stages of speciation. Future studies of the mating behaviour of L. fabalis should also involve experimental tests of reproductive behaviours involving the two size morphs of L. fabalis as in light of our current results size seems to be an important component of mate choice in this species as well as in other species of Littorina.

Acknowledgements We thank Per Jonsson, Mats Lindegarth and Gunilla Toth for help with the statistical analyses. We would also like to thank the Motion Analysis Corp, Santa Rosa, CA for their support with software used for trail-following analysis. This work was performed within the Linnaeus Centre for Marine Evolutionary Biology (http://www.cemeb.science.gu.se/).

12 References Andersson M. B (1994) Sexual Selection. Princeton University Press. Crespi B. J (1989) Causes of assortative mating in arthropods. Anim. Behav. 38: 980-1000 Croll R. P (1983) Gastropod chemoreception. Biol. Rev. 58: 293–319 Davies M. S, Beckwith, P (1999) Role of mucus trails and trail-following in the behaviour and nutrition of the periwinkle Littorina littorea. Mar. Ecol. Prog. Ser. 179: 247- 257 Davies M. S, Blackwell, J (2007) Energy saving through trail following in a marine snail. Proc. R. Soc. B. doi: 10.1098/rspb.2007.0046 Edwards M, Davies M. S (2002) Functional and ecological aspects of the mucus trails of the intertidal prosobranch gastropod Littorina littorea. Mar. Ecol. Prog. Ser. 239: 129-137 Erlandsson J (1998) Evolutionary ecology of the snails Littorina littorea and L. saxatilis: mating behaviour and habitat choice. Dept. of Marine Ecology, Göteborg University Erlandsson J, Johannesson K (1994) Sexual selection on female size in a marine snail, Littorina littorea (L.). J. Exp. Mar. Biol. Ecol. 181: 145-157 Erlandsson J, Kostylev V (1995) Trail following, speed and fractal dimension of movement in a marine prosobranch, Littorina littorea, during a mating and a non-mating season. Mar. Biol. 122: 87-94

Frey M. A, Lonsdale D. J, Snell T. W (1998) The influence of contact chemical signals on mate recognition in harpacticoid copepod. Phil. Trans. R. Soc. Lond. 353: 745- 751 Hankison S. J, Morris M. R (2003) Avoiding a compromise between sexual selection and species regognition: female swordtail fish assess multiple species-specific cues. Behav. Ecol. 14: 282-287 Hollander J, Lindegarth M, Johannesson K (2005) Local adaptation but not geographical separation promotes assortative mating in snail. Anim. Behav. 70: 1209-1219 Hughes R. N, Answer P (1982) Growth, spawning and trematode infection of Littorina littorea (L.) from an exposed shore in north Wales. J. Moll. Stud. 48: 321-330 Janson K (1985) Variation in the occurrence of abnormal embryos in females of the intertidal gastropod Littorina saxatilis Olivi. J. Moll. Stud. 51: 64-68

13 Johannesson K, Saltin S. H, Duranovic I, Havenhand J. N, Jonsson P. R (2010) Indiscriminate Males: Mating Behaviour of a Marine Snail Compromised by a Sexual Conflict? PLoS ONE 5(8): e12005. doi:10.1371/journal.pone.0012005 Kemppainen P, Nes S, Ceder C, Johannesson K (2005) Refuge function of marine algae complicates selection in an intertidal snail. Oecologia 143:402-411 Kemppainen P, Panova M, Hollander J, Johannesson K (2009) Complete lack of mitochondrial divergence between two species of NE Atlantic marine intertidal gastropods. J. Evol. Biol. 22: 2000 -2011 Owens I. P. F, Bennett P. M, Harvey P. H (1999) Species richness among birds: body size, life history, sexual selection or ecology? Proc. R. Soc. Lond. B. 266: 933-939 Parker G. A (1970) Sperm competition and its evolutionary effect on copula duration in the fly Scatophaga stercoraria. J. Insect. Physiol. 16: 1301-1328 Peduzzi P, Herndl G. J (1991) Mucus trails in the rocky intertidal –a highly active microenvironment. Mar. Ecol. Prog. Ser. 75: 267-274 Plenderleith M, van Oosterhout C, Robinson R. L, Turner G. F (2005) Female preference for conspecific males based on olfactory cues in Lake Malawi cichlid fish. Biol. Lett. 1: 411-414 Ratciliffe L. M, Grant P. R (1983) Species recognition in Darwin's finches (Geospiza, Gould) I. Discrimination by morphological cues. Anim. Behav. 31(4): 1139-1153 Reid, D. G (1996) Systematics and Evolution of Littorina. The Ray Society, Dorchester Reimchen T. E (1982) Shell size divergence in Littorina mariae and L. obtusata and predation by crabs. Can. J. Zool. 60: 687- 695 Ridley M (1983) The explanation of organic diversity. The comparative method and adaptations for mating. Clarendon Press, Oxford, UK. Rolán-Alvarez E, Zapata C, Alvarez G (1995a) Distinct genetic subdivision in sympatric and sibling species of the genus Littorina (Gastropoda, Littorinidae). Heredity 74: 1-9 Rolán-Alvarez E, Zapata C, Alvarez G (1995b) Multilocus heterozygosity and sexual selection in a natural population of the marine snail Littorina mariae (Gastropoda: Prosobranchia). Heredity 75: 17-25 Saur M (1990) Mate discrimination in Littorina littorea (L.) and L. saxatilis (Olivi) (Mollusca: Prosobranchia). Hydrobiologia 193: 261-270 Tatarenkov A. N (1995) Genetic divergence between sibling species Littorina mariae Sacchi and Rastelli and Littorina obtusata (L.) (Mollusca, Gastropoda) from the White- Sea. Ophelia 40: 207-218.

Tatarenkov A, Johannesson K (1998) Evidence of a reproductive barrier between two forms of the marine periwinkle Littorina fabalis (Gastropoda). Biol. J. Linnean soc. 63: 349-365

14 Van Doorn S. G, Edelaar P, Weissing F. J (2009) On the Origin of Species by Natural and

Sexual Selection. Science 326: 1704-1707 Wiernasz D. C, Kingsolver J. G (1992) Wing melanin pattern mediates species recognition in Pieris occidentalis. Anim. Behav. 43(1): 89-94 Zaslavskaya N. I, Sergievsky S. O, Tatarenkov A. N (1992) Allozyme similarity of Atlantic and Pacific species of Littorina (Gastropoda, Littorinidae). J. Mollus. Stud. 58: 377-384

15

Paper II Indiscriminate Males: Mating Behaviour of a Marine Snail Compromised by a Sexual Conflict?

Kerstin Johannesson*, Sara H. Saltin, Iris Duranovic, Jon N. Havenhand, Per R. Jonsson Department of Marine Ecology - Tja¨rno¨, University of Gothenburg, Stro¨mstad, Sweden

Abstract

Background: In promiscuous species, male fitness is expected to increase with repeated matings in an open-ended fashion (thereby increasing number of partners or probability of paternity) whereas female fitness should level out at some optimal number of copulations when direct and indirect benefits still outweigh the costs of courtship and copulation. After this fitness peak, additional copulations would incur female fitness costs and be under opposing selection. Hence, a sexual conflict over mating frequency may evolve in species where females are forced to engage in costly matings. Under such circumstance, if females could avoid male detection, significant fitness benefits from such avoidance strategies would be predicted.

Methodology/Principal Findings: Among four Littorina species, one lives at very much higher densities and has a longer mating season than the other three species. Using video records of snail behaviour in a laboratory arena we show that males of the low-density species discriminate among male and female mucous trails, trailing females for copulations. In the high-density species, however, males fail to discriminate between male and female trails, not because males are unable to identify female trails (which we show using heterospecific females), but because females do not, as the other species, add a gender-specific cue to their trail.

Conclusions/Significance: We conclude that there is likely a sexual conflict over mating frequency in the high-density species (L. saxatilis) owing to females most likely being less sperm-limited in this species. This has favoured the evolution of females that permanently or optionally do not release a cue in the mucus to decrease excessive and costly matings resulting in unusually high frequencies of male-male copulating attempts in the wild. This is one of few examples of masking gender identity to obtain fewer matings.

Citation: Johannesson K, Saltin SH, Duranovic I, Havenhand JN, Jonsson PR (2010) Indiscriminate Males: Mating Behaviour of a Marine Snail Compromised by a Sexual Conflict? PLoS ONE 5(8): e12005. doi:10.1371/journal.pone.0012005 Editor: Neil John Gemmell, University of Otago, New Zealand Received May 23, 2010; Accepted July 14, 2010; Published August 9, 2010 Copyright: ß 2010 Johannesson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This study was performed at the Linnaeus Centre for Marine Evolutionary Biology (http://www.cemeb.science.gu/) and supported by grants from the Swedish Research Councils (VR and Formas) to KJ, JNH and PRJ. The funders had no role in study design, data collection and analysis, decision to publish,or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction the genus Littorina which have separate sexes and are typically highly promiscuous [10–12]. However, the opportunity for males For many species the number of matings that maximizes fitness to transfer sperm (mate), and female requirements for sperm, vary differs between the sexes. Male fitness is generally believed to among the four species chosen for this study, owing to differences increase with the number of matings, whereas polyandry is in length of mating season, densities of snails and overall egg favoured by selection under conditions such as depletion of sperm production ([13] and data presented herein). in a female’s reproductive tract or indirect genetic benefits via Male-female pairing in snails is preceded by mucus-trail sperm competition or cryptic female choice [1,2]. After an initial tracking, predominantly by males [12]. The accuracy of this increase in fitness with number of copulations, female fitness starts behaviour in terms of conspecific, heterosexual trailing is likely to to decline when costs of mating exceed gains from receiving be under selection, since crawling in snails requires production of additional sperm [3–6]. Thus as a consequence of this sexual substantial amounts of energetically-expensive pedal mucus conflict females may evolve measures to avoid the costs of [14,15]. Once a partner is located, a male littorinid snail mounts excessive matings while males remain selected for maximizing the female and after a counter-clockwise movement on the shell, paternity through repeated mating [7–9]. This situation may arise, he stops at the right-hand side of the shell and inserts the penis for example, in populations of promiscuous species that live at high under the shell of the partner (Fig. 1). Only rarely have females densities, store sperm over several ovarian cycles, have long been observed to actively reject a mounting male [11]. Sperm is mating seasons, but modest egg production compared to other transferred through ciliary actions in a groove along the penis species. This gives males many mating opportunities, while females (Buckland-Nicks pers. commun.), and copulations typically last for risk decreased fitness from superfluous costly matings. In this study 10–30 min [11,16]. During the copulation the female effectively we test this prediction using marine intertidal periwinkle species of ‘‘carries’’ the male, incurring a cost to the mating pair as risk of

PLoS ONE | www.plosone.org 1 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males

of the tank facilitating collection. Snails that inhabit the macroalgal canopy (in this study L. fabalis and L. obtusata) experience the area of the algal thalli as their total available surface. Hence for these two species we estimated population densities in relation to total macroalgal surface area (calculated from the surface area and wet-weight of a subsample of algae, and the total weight of the sample). In contrast, for L. saxatilis and L. littorea we estimated 3-dimensional surface area from the area of the boulders or rocky substrates (including crevices etc.). Total thallus area of 1 m2 100% cover of Ascophyllum nodosum was estimated to 24 m2, and the corresponding area for was 33 m2.

Experimental tests of female costs of mating In addition to increased predation risk [17] a main cost of mating is likely to be the increased risk of dislodgement of a Figure 1. A copulating pair of Littorina (Littorina saxatilis of copulating pair compared to the risk of a single snail, as Swedish S ecotype). The female is attached to the substratum while hydrodynamic drag will be roughly doubled for a copulating pair the male is positioned on the right-hand side of her shell, inserting his while attachment strength remains that of a single snail. In a penis under her shell. Photo: Patrik Larsson. laboratory experiment we glued male L. saxatilis shells in doi:10.1371/journal.pone.0012005.g001 copulating positions onto the females of the same species (to simulate mounting males) and compared the risk of dislodgement predation increases for mating pairs compared to single individuals with that of single females. In each run we allowed five females [17]. Dislodgement by waves is an additional selective pressure for with and five females without glued male shells attach themselves intertidal snails, as being removed from the intertidal increases risk to a PVC platform (40630 cm). The platform was then towed at of predation by crabs and fishes [18,19]. Risk of dislodgement is constant speed along the bottom of a 6 m tank. For each tow the determined by the hydrodynamic drag (proportional to the cross- number of detached females of each type was recorded. The sectional area of the snail) and attachment strength (proportional platform was towed repeatedly at increasing speeds corresponding to the foot area), and we therefore hypothesis that a copulating to the range found in breaking waves with wave heights less than pair should have considerably greater risk of dislodgement because 1 m [23]. Using results from 10 replicate tows (new snails in each), cross-sectional area is doubled with no change in attachment the difference in flow speed needed to remove half the number of strength. snails was compared between females with and without a glued It has been suggested that among gastropods in general, males male shell. This experiment was complemented by a release- identify (and are attracted to) partners prior to copulation through recapture of artificially ‘paired’ versus single females in the wild; 29 specific chemical cues added to the trail mucus by the female [20– females with and 28 without glued male shells were marked with 22]. Earlier observations suggest that female trailing by males paint and released on a natural rocky shore. Recapture took place occurs in the periwinkle species Littorina littorea [21], but appears to a week later. be remarkably random with respect to sex in male L. saxatilis [12]. This may be due to (i) male lacking the capacity to differentiate Trail-following experiments between males and females, and/or (ii) the evolution of sexual Snails of Littorina littorea, L. fabalis, L. obtusata and L. saxatilis mimicry in females, avoiding the costs of superfluous copulations. (Swedish cliff ecotype, E morph, [24]) were sampled at two sites Here, we experimentally test these two hypotheses. Furthermore, each along the shores of three small islands (Salto¨, La˚ngholmen we hypothesize that the difference in male behaviour between L. and Lo¨kholmen; in the vicinity of Tja¨rno¨ in Sweden). Littorina littorea and L. saxatilis is a consequence of a potentially more saxatilis is an exceptionally polymorphic species [13] and in pronounced sexual conflict over mating frequency in the latter addition to using the Swedish E ecotype we compared our results species, owing to its much higher density and prolonged mating- for this ecotype with the behaviour of the Swedish S ecotype (from season. To test this hypothesis we make the additional prediction two localities on Salto¨) that are morphologically and ecologically that gender-specific trail following would be in the norm in two very different and live in different microhabitats but at similarly other littorinid species, L. obtusata and L. fabalis, that share the high densities [24]. Finally, we analysed male behaviour also with restricted mating season and low density with L. littorea, although L. saxatilis from Spain sampled in exposed low-shore habitats in their yearly egg production is similar to L. saxatilis [13]. Galicia, (42u69N, 8u539W). Although considered the same species, Spanish and Swedish L. saxatilis are very distantly related and have Materials and Methods evolved separately for more than 100,000 years (Panova et al. subm.), but again densities, reproductive season and reproductive Estimates of population densities modes are very similar. Population density of each of the four species was estimated by In a first series of tests we studied male behaviour in each of the sampling 6–10 typical habitats of each species in the area of four species and the three different L. saxatilis populations investigation (the archipelago around the island Tja¨rno¨ on the separately. In one run, five males and five females were allowed Swedish west coast, 58u539N, 11u89E). In each site, 1 m2 large to move freely in an arena (,0.25 m2) and their movements were areas were investigated. First we harvested all macroalgae and recorded with digital video over a period of 15 minutes. For each these were brought back to the laboratory, and secondly we group at least 14 replicate runs (new snails for each run) testing carefully searched the rock surface and picked all visible snails trail-following behaviour were performed. The surface of the (.1–2 mm). In the laboratory, macroalgae were put in tanks with arena was wetted with seawater and rinsed carefully between each freshwater causing snails in the algal canopy to drop to the bottom run. Spatial excursions and extent of trail-following by male

PLoS ONE | www.plosone.org 2 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males conspecifics were measured from the video using computerised Table 1. Life-history traits in four species of Littorina. motion analysis (CellTrak for Windows, Motion Analysis Corp.). Male movements were defined as trail-following when following occurred for more than 3 snail diameters. All other paths and Reproductive Number of eggs paths along the edges of the arena were excluded from our Species season per year Snail density1 analyses. Hence from each run we received the total distance a L. saxatilis2 Year round3 2004 280 male tracker followed male trails and female trails, respectively. As 3 5 the level of replication in our experiments was a run, we pooled L. littorea February–June 110,000 2.3 the results from all 5 males from each run by using the total L. fabalis Spring-Autumn3 .6004 1.4 tracking distances of all 5 males. As we did not include data for L. obtusata March–September3 .5004 1.1 males tracking their own trails, we randomly removed the data from one female in each run (avoiding bias but maintaining gender Densities for L. saxatilis are indicated for the Swedish cliff ecotype (E). Data on reproductive seasons are from [13], and numbers of eggs per year are from balance during the experiment). unpublished studies by KJ except for L. littorea where data are from Buschbaum During the 15 minutes experiments, several males did not & Reise [37]. Snail density estimates are from typical habitats of each species encounter any trail to follow, and in many cases we only recorded taken into account the surface area of the substratum, see text for details. one trail-following event per run. Our main experiment was doi:10.1371/journal.pone.0012005.t001 therefore complemented by a study in which we tested male tracking behaviour in one of the groups (E ecotype L. saxatilis) over In the accompanying field experiment, we recovered signifi- a longer period of time (60 instead of 15 minutes) in order to cantly fewer females with attached male shells (3 out of 29) than remove the possibility that trial duration constrained our females without male shells (12 out of 28) (x2 = 4.6, df =1, probability of detecting sex-biased male trailing. This study was P,0.05), indicating a survival difference between the two performed in the same way as the experiments described above, experimental groups of females. with the exception that we used a larger arena (to delay snails reaching the edge as much as possible). Trail following To specifically test the hypothesis that male L. saxatilis had lost Males of all species identified the polarity of the trail and their capacity to identify female trails, we did a separate series of followed the trail in the direction of the marker snail in about 90% experiments in which we used male L. saxatilis as trackers and of the cases (Table 2). In the low-density species L. littorea, L. males and females of either L. saxatilis or L. fabalis as markers. In obtusata and L. fabalis males were also significantly more likely to random order we performed 20 replicate runs in total for the follow female mucous trails than male mucous trails (binomial test, combination with male and female L. saxatilis as markers, and 20 P = 0.006–0.029; Fig. 2A–C). Although a few males of these three with male and female L. fabalis as markers (in both using male L. species were observed to follow trails of other males (points along saxatilis as tracker). The trail-following of the trackers were the y-axis in Fig. 2), the great majority of males exclusively analysed in the same way as described above. Also in this followed female trails, or followed female trails for much longer experiments, snails were from different localities at the island of distances than male trails (points along or close to the x-axis in Salto¨. Fig. 2). In contrast, we found no evidence that male L. saxatilis In all the trail-following experiments we used a binomial test to could discriminate between the mucous trail of females and males, discriminate between two possible outcomes of the experiments; and these results were consistent for males from both the Swedish H0 – that males followed male and female trails for equal distances and the Spanish populations of this species (binomial test, P = 0.4 (or followed male distances longer than female distances); H1 – and P = 0.8; Fig. 2D–E) and for Swedish S ecotypes (P = 0.6; that males followed female mucous trails for longer distances than results not shown). In the longer-duration (60 min) runs, there was they followed other male trails. Thus, we test the directional still no deviation from random trailing with respect to sex (Fig. 2F). prediction that males followed females for longer and therefore When we used L. fabalis as marker females the overall extent of apply a one-tailed test. tracking by L. saxatilis males was slightly less pronounced; tracking in 8 of 20 runs compared to 16 of 20 for the conspecific runs, Results which may be expected for a heterospecific comparison. Snail densities Remarkably, however, in all runs where L. fabalis trails were followed, male L. saxatilis followed female trails for longer distances Combining new data for snail densities in their natural habitat than male trails (binomial test, P = 0.004) while in the same with literature data showed that densities of L. saxatilis were . experiment, L. saxatilis males, as before, did not discriminate 1006 higher than for the other species (Table 1), and therefore between male and female trails of their own species (P = 0.1; Fig. 3). expected male-female encounter rates would be two orders of magnitude greater in this species. In addition, a longer mating season further increases the total number of matings per Table 2. Polarity of trail-following in male Littorina. reproductive season in female L. saxatilis in comparison to the other species (Table 1). Hence, female L. saxatilis are far less likely to be sperm-limited than females of any of the other species. Frequency of tracks (%)

Species N Positive Negative Cost of mating Our laboratory trials showed that females with a ‘pairing male’ L. obtusata 36 89 11 required significantly lower water speed to dislodge the pair as L. fabalis 56 89 11 opposed to single females. The average water speed required to L. littorea 50 92 8 detach 50% of the females with an attached male was significantly L. saxatilis 24 100 0 lower than that for females without an attached male (0.6 m s21 vs 21 1.2 m s ; 1-factor ANOVA, two-tailed, F1,9 = 5.77, P = 0.00027). doi:10.1371/journal.pone.0012005.t002

PLoS ONE | www.plosone.org 3 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males

Figure 2. Male tracking females and other males in four species of Littorina. Distances that male trackers followed males (y-axis) and females (x-axis) in four different species of Littorina.InL. saxatilis two geographically separated populations (Sweden E-ecotype, and Spain SU-ecotype) are analysed. Each point represents the total distance of trail-following of 5 males of one replicate run (15 min, or 60 min in F). The diagonal indicate the expectation of equal tracking distances of female and male markers. One-tailed P-values of binomial tests are indicated. doi:10.1371/journal.pone.0012005.g002

PLoS ONE | www.plosone.org 4 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males

Figure 3. Littorina saxatilis males tracking males and females of their own or an other species (L. fabalis). (A) Distances that males of L. saxatilis followed males and females of their own species. (B) Distances that males of L. saxatilis followed males and females of L. fabalis. Experimental conditions and statistical tests as in Figure 1. doi:10.1371/journal.pone.0012005.g003

This shows that male L. saxatilis have the capacity to detect, and optionally removed at high densities in female L. saxatilis.AsL. show a preference for, a female-specific mucous cue, and strongly saxatilis is the most derived species in a phylogeny including the suggests that this mucous cue is absent in female L. saxatilis. four littorinids of this study [29], there are 16 possible combinations with these four species each having or not having Discussion this particular character state. Therefore, there is only one chance in 16 that loss or optional regulation of this cue in female L. saxatilis Our results show that under similar experimental densities male is a result of pure chance, while it being a novel state trail-following behaviour of the high-density L. saxatilis is different (autapomorphy) seems much more likely. from that of the related but naturally low-density periwinkle Females that do not release a cue decrease the number of L. littorea L. fabalis L. obtusata species, , and . Indeed, results obtained matings by producing a gender-neutral trail, and such a strategy in five independent experiments including geographically separat- would likely be favoured by selection if females are not sperm- ed populations and different ecotypes, show that males of L. limited and if costs of mating are substantial. Sperm-limitation saxatilis can not discriminate between female and male mucous seems most unlikely in L. saxatilis; much higher densities than in trails when tracking snails of their own species, whereas males of the other three littorinid species and longer, almost year-round, the other three species are able to identify female trails. However, mating season, give excessive opportunities to mate. Recent data male L. saxatilis can distinguish between male and female trails of show that each female simultaneously carry offspring sired by 10– another species, L. fabalis, and have thus not lost their ability to 20 males [30,31] supporting the prediction that female L. saxatilis track females using a gender-specific mucous cue. This suggests that essentially the same female identification cue is shared among are not sperm-limited. Indeed, such a high level of polyandry is species and that female L. saxatilis have either lost this cue hard to explain from the perspective of increased female fitness completely, or can optionally remove the cue when densities of and we have earlier suggested this to be a consequence of mates are high. Although we have not identified the specific cue, a convenience polyandry [31]. At the same time cost of mating is chemical cue that females add to their mucus and that attracts likely to be substantial as predation risk increases during mating, males seems quite likely, and such a cue is shown to be present in a both directly through increased susceptibility to predators [17], freshwater snail [25]. Notably, in an earlier study we observed that and indirectly through increased risk of snails being dislodged to male L. saxatilis were indeed able to discriminate between female L. the more predator-rich subtidal [this study and 18,19]. Hence saxatilis of different ecotypes, resulting in assortative mating [26]. mating costs would select for mechanisms in the female to avoid Also in this study males did not discriminate between male and both male harassment and excessive matings, as a complement to female trails, but data were only evaluated for male-female convenience mating. trackings. Costly matings may potentially be a problem also for male L. The consequence of the different trail-following behaviours are saxatilis and select for males that restrict the number of matings. As readily observed in the field: in general, numbers of males males of littorinids, however, provide no parental care, adding to mounting other males and juveniles are low (0–7%) in species of the number of life-time matings is their only way to increase life- littorinids including tropical species distantly related to L. saxatilis time fitness. Hence males would be prepared to take larger risks [11,27,28], but L. saxatilis is an obvious exception with much compared to females and strive to mate as frequently as possible higher rates of male-male or male-juvenile mating attempts (30– despite these costs. 40%) [10–12]. Collectively, these data strongly suggest that both a Is it necessary for males to track females in high-density female-specific cue added to the female mucous trail, and the populations such as L. saxatilis, when a male is likely to encounter a capacity of the males to identify this cue, are ancestral traits among high number of females anyhow? As a male that practice mucus- littorinid snails but that the gender cue is permanently lost or tracking uses both time and energy during tracking and during

PLoS ONE | www.plosone.org 5 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males mounting the partner (male-male copulation attempts are effectively reduce number of costly matings, reflecting a sexual disrupted when the mating male tries to insert the penis under conflict over mating frequency. Masking gender identity to obtain the shell of the other male [11]), a male that can avoid tracking copulations is common in animals [32,33], however evidence of other males, or juveniles, will be at an advantage over other males. females avoiding males by mimicking the male phenotype is scant. Indeed our results showed that male L. saxatilis practice female A few examples are found among species of damselflies, in which a tracking if encountering trails of males and females of another proportion of females develop into andromorph females mimick- species (L. fabalis), and this indicates that male L. saxatilis still retain ing males in colour and behaviour thereby becoming less attractive the capacity to discriminate between male and female trails if the to males and effectively reducing costs of male harassment and female cue is present. excessive matings [34,35] (but see [36] for a slightly different Why do not females produce more eggs and/or shorten the interpretation). In damselflies, however, this polymorphism is mating period to either match the high availability of sperm, or frequency-dependent with a stable proportion of the females being avoid male harassment by not being year-round receptive? In L. andromorph. This is because males learn to recognize andro- saxatilis egg production is set by the size of the female as embryos morph females if these are too common. In addition there is a cost develop to crawl-away juvenile snails inside an embryo-pouch of increased predation rate and risk of sperm limitation for under the shell of the female. To increase egg-numbers would andromorph females that must be balanced by the potential require smaller egg-sizes that may trade-off against egg survival. rewards of reduced matings (reviewed in 2). In L. saxatilis, however, Probably restricted by the size of the embryo-pouch, new-born to remove the gender-specific cue and produce neutral trails will snails are delivered at a more or less constant rate over the season. most likely evolve with no additional cost for the females, hence Consequently, juvenile females will become sexually mature at any there is no trade-off for the female snails and we predict that all of time of the year, probably setting the stage for a year-round them have adopted the ‘‘gender neutral’’ strategy. mating season. In summary, we have shown here that in three species of Acknowledgments littorinids, males actively track females by using a cue added to their mucous trail. We also showed that male L. saxatilis retain the We thank Mats Olsson, Mark Davies and two anonymous reviewers for same capacity, but fail to discriminate among trails of conspecific useful comments and improvements of the text, Johannes Bjo¨rk, Madeleine Mann, Maria Nordstro¨m and Klara Johannesson for their assistance in the males and females. From this we conclude that female L. saxatilis experimental tests, John Buckland-Nicks for his information on snail have either permanently lost this cue or have the ability to copulation physiology, and the Motion Analysis Corp, Santa Rosa, CA for optionally remove the cue under snail densities that are the norm their support with software used for trail-following analysis. This work was for this species, resulting in observed frequencies of male-male performed within the Linnaeus Centre for Marine Evolutionary Biology pairings in the wild being exceptionally high in this species. We (http://www.cemeb.science.gu.se/). have also presented direct and indirect support for costs associated with mating and we argue that female L. saxatilis are less likely to Author Contributions be sperm-limited than females of low-density species of littorinids Conceived and designed the experiments: KJ SHS PRJ. Performed the and therefore are selected for trying to decrease the number of experiments: SHS ID PRJ. Analyzed the data: KJ SHS PRJ. Contributed costly matings. Hence an obvious reason why females no longer reagents/materials/analysis tools: KJ SHS JNH PRJ. Wrote the paper: KJ have a gender cue in the mucous is that removing this cue SHS JNH PRJ.

References 1. Bateman AJ (1948) Intra-sexual selection in Drosophila. Heredity 2: 349–368. 16. Hollander J, Lindegarth M, Johannesson K (2005) Local adaptation but not 2. Arnqvist G, Rowe L (2005) Sexual conflict. Princeton: Princeton University geographic separation promotes assortative mating in a snail – support for Press. 360 p. ecological speciation. Anim Behav 70: 1209–1219. 3. Arnqvist G, Nilsson T (2000) The evolution of polyandry: multiple mating and 17. Koch N, Lynch B, Rochette R (2007) Trade-off between mating and predation female fitness in insects. Anim Behav 60: 145–164. risk in the marine snail, Littorina plena. Invertebr Biol 126: 257–267. 4. Fox CW, Rauter CM (2003) Bet-hedging and the evolution of multiple mating. 18. Johannesson B (1986) Shell morphology of Littorina saxatilis Olivi – the relative Evol Ecol Res 5: 273–286. importance of physical factors and predation. J Exp Mar Biol Ecol 102: 5. Sarhan A, Kokko H (2007) Mutiple mating in the Glanville fritillary butterfly: A 183–195. case of within-generation bet hedging? Evolution 61: 606–616. 19. Kemppainen P, Nes S, Ceder C, Johannesson K (2005) Refuge function of 6. Byrne PG, Keogh JS (2009) Extreme sequential polyandry insures against nest marine algae complicates selection in an intertidal snail. Oecologia 143: failure in a frog. Proc Roy Soc Lond B 276: 115–120. 402–411. 7. Rice WR, Holland B (1997) The enemies within: intergenomic conflicts, 20. Croll RP (1983) Gastropod chemoreception. Biol Rev 58: 293–319. interlocus contest evolution (ICE), and the intraspecific Red Queen. Behav Ecol 21. Erlandsson J, Kostylev V (1995) Trail following, speed and fractal dimension of Sociobiol 41: 1–10. movement in a marine prosobranch, Littorina littorea, during a mating and a non- 8. Holland B, Rice WR (1998) Chase-away sexual selection: antagonistic selection mating season. Mar Biol 122: 87–94. versus resistance. Evolution 52: 1–7. 22. Clifford KT, Gross L, Johnson K, Shaheen N, Harrington MA (2003) Slime-trail 9. Arnqvist G, Rowe L (2002) Correlated evolution of male and female tracking in the predatory snail, Euglandina rosea. Behav Neurosci 117: 1086–1095. morphologies in water striders. Evolution 56: 936–947. 23. Gaylord B (1999) Detailing agents of physical disturbance: wave-induced 10. Raffaelli DG (1977) Observations of the copulatory behaviour of Littorina rudis velocities and accelerations on a rocky shore. J Exp Mar Biol Ecol 239: 85– Maton and L. nigrolineata Gray (Gastropod: Prosobranchia). Veliger 20: 75–77. 124. 11. Saur M (1990) Mate discrimination in Littorina littorea (L.) and L. saxatilis (Olivi) 24. Janson K (1982) Phenotypic differentiation in Littorina saxatilis Olivi (Mollusca: (Mollusca: Prosobranchia). Hydrobiologia 193: 261–270. Prosobranchia) in a small area on the Swedish west coast. J Mollusc Stud 48: 12. Erlandsson J (2002) Do reproductive strategy and breeding season influence the 167–173. presence of mate recognition in the intertial snail Littorina? Invertebr Reprod Dev 25. Takeichi M, Hirai Y, Yusa Y (2007) A water-borne sex pheromone and trail 41: 53–60. following in the apple snail, Pomacea canaliculata. J Mollusc Stud 73: 275–278. 13. Reid DG (1996) Systematics and Evolution of Littorina. London: Ray Society. pp 26. Johannesson K, Havenhand JN, Jonsson PR, Lindegarth M, Sundin A, et al. 463. (2008) Male discrimination of female mucous trails permits assortative mating in 14. Davies MS, Jones HD, Hawkins SJ (1992) Pedal mucus production in Littorina a marine snail species. Evolution 62: 3178–3184. littorea (L.). Proceedings of the Third International Symposium on Littorinid 27. Struhsaker JW (1966) Breeding, spawning periodicity and early development in Biology Grahame J, Mill PJ, Reid DG, eds. London: The Malacological Society. the Hawaiian Littorina: L. pintado (Wood), L. picta Philippi and L. scabra (Linne´). pp 221–226. Proc Malacol Soc Lond 37: 137–166. 15. Davies MS, Blackwell J (2007) Energy saving through trail following in a marine 28. Reid DG (1986) The littorinid molluscs of mangrove forests in the Indo-Pacific snail. Proc Roy Soc Lond B 274: 1233–1236. region. London: British Museum (Natural History).

PLoS ONE | www.plosone.org 6 August 2010 | Volume 5 | Issue 8 | e12005 Indiscriminate Males

29. Winnepennickx B, Steiner G, Backeljau T, De Wachter R (1998) Details of 34. Cordero A, Carbone SS, Utzeri C (1998) Mating opportunities and mating costs gastropod phylogeny inferred from 18S rRNA sequences. Mol Phyl Evol 9: are reduced in androchrome female damselflies, Ischnura elegans (Odontata). Anim 55–63. Behav 55: 185–197. 30. Ma¨kinen T, Panova M, Andre´ C (2007) High levels of multiple paternity in 35. Andres JA, Sanchez-Guillen RA, Rivera A (2002) Evolution of female colour Littorina saxatilis. Hedging the bets? J Heredity 98: 705–711. polymorphism in damselflies: testing the hypothesis. Anim Behav 63: 677–685. 31. Panova M, Bostro¨m J, Hofving T, Areskoug T, Eriksson A, et al. (2010) Extreme 36. Sirot LK, Brockmann HJ, Marinis C, Muschett G (2003) Maintenance of a female promiscuity in a non-social invertebrate species. PlosONE 5(3): e9640. female-limited polymorphism in Ischnura ramburi (Zygoptera: Coenagrionidae). doi:10.1371/journal.pone.0009640. Anim Behav 66: 763–775. 32. Saetre GP, Slagsvold T (1996) The significance of female mimicry in male 37. Buschbaum C, Reise K (1999) Effects of barnacle epibionts on the periwinkle contests. Am Nat 147: 981–995. Littorina littorea (L.) Helgol. Mar Res 53: 56–61. 33. Hanlon RT, Naud MJ, Shaw PW, Havenhand JN (2005) Behavioural ecology – Transient sexual mimicry leads to fertilization. Nature 433: 212.

PLoS ONE | www.plosone.org 7 August 2010 | Volume 5 | Issue 8 | e12005