Blackwell Publishing Ltd Gregariousness and protandry promote reproductive insurance in the invasive gastropod Crepidula fornicata: evidence from assignment of larval paternity potx

Molecular Ecology (2006) 15 , 3009–3021 doi: 10.1111/j.1365-294X.2006.02988.x © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd Blackwell Publishing Ltd Gregariousness and protandry promote reproductive insurance in the invasive gastropod Crepidula fornicata : evidence from assignment of larval paternity L. DUPONT, *‡ J. RICHARD, † Y-M. PAULET, † G. THOUZEAU † and F. VIARD * * Evolution et Génétique des Populations Marines, UMR ADMM 7144 CNRS-UPMC, Station biologique, place Georges Teissier, BP 74, 29682 Roscoff cedex, France, † Institut Universitaire Européen de la Mer — Université de Bretagne Occidentale, LEMAR UMR 6539 CNRS, Technopôle Brest-Iroise, Place Nicolas Copernic, 29280 Plouzané, France Abstract According to the size-advantage hypothesis, protandric sequential hermaphroditism is expected when the increase in reproductive success with age or size is small for males but large for females. Interestingly, some protandrous molluscs have developed gregarious strategies that might enhance male reproductive success but at the cost of intraspecific competition. The gastropod Crepidula fornicata , a European invading species, is ideal for investigating mating patterns in a sequential hermaphrodite in relation to grouping behaviour because individuals of different size (age) live in perennial stacks, fertilization is internal and embryos are brooded. Paternity analyses were undertaken in stacks sampled in three close and recently invaded sites in Brittany, France. Paternity assignment of 239 larvae, sampled from a set of 18 brooding females and carried out using five microsatellite loci, revealed that 92% of the crosses occurred between individuals located in the same stack. These stacks thus function as independent mating groups in which individuals may reproduce consecutively as male and female over a short time period, a pattern explained by sperm storage capacity. Gregariousness and sex reversal are promoting reproductive insurance in this species. In addition, females are usually fertilized by several males (78% of the broods were multiply sired) occupying any position within the stack, a result reinforcing the hypothesis of sperm competition. Our study pointed out that mating behaviours and patterns of gender allocation varied in concert across sites suggesting that multiple paternities might enhance sex reversal depending on sperm competition intensity. Keywords : larvae, microsatellites, paternity analyses, sequential hermaphrodite, social groups Received 27 January 2006; revision accepted 30 March 2006 Introduction Sex-allocation theory predicts that sequential hermaphroditism is expected when the increase in reproductive success with age or size is faster for one sex than for the other (i.e. the ‘size-advantage’ hypothesis, Ghiselin 1969; Charnov 1982). Despite theoretical advantages either over gonochorism (because of a higher lifetime reproductive potential) or simultaneous hermaphroditism (because of inbreeding avoidance), only the Gastropoda and Bivalvia have sex-changing species among the eight molluscan classes and almost all of these are protandrous (i.e. change sex from male to female; Wright 1988; Heller 1993). According to Wright (1988), the fact that most molluscan species are patchily distributed over space and/or have limited adult mobility would tend to select for protandry because males would have limited opportunity for mating (Ghiselin 1969; Hoagland 1978). Moreover, because fecundity in females generally increases with body size, Warner et al . (1975) asserted that protandry might be expected to be found in randomly mating populations, such as group spawners, where males of all body sizes have similar chances of fertilizing eggs; while protogyny would be found in the Correspondence: L. Dupont and F. Viard. Fax: +44 (0) 1752 633102; E-mail: lidu@mba.ac.uk; viard@sb-roscoff.fr. ‡Present address: Marine Biological Association, The Laboratory, Citadel Hill, Plymouth PL1 2PB, UK. 3010 L. DUPONT ET AL. © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd case of more selective matings, such as occur in pair formation. Nevertheless, it is noteworthy that several species of protandrous molluscs with internal fertilization have developed strategies of conspecific association (e.g. Hoagland 1978; Collin et al . 2005) that most notably enhance male reproductive success (Levitan 1993) but at the cost of increased intraspecific competition (Toonen & Pawlik 1994). As the direction of sex change is closely linked to the mating system, a better understanding of the evolution of sex-changing strategies requires knowing more about mating behaviours and reproductive success in sequential hermaphrodites (Wright 1988). In their review, Munday et al . (2006) have recently shown that sex-changing species use a great diversity of strategies to increase their repro- ductive success. For instance, the study by Munoz & Warner (2004) revealed that multiple paternity and sperm competition influence patterns of sex change. While sequential hermaphroditism has been extensively studied in fishes (e.g. Hoffman et al . 1985; Allsop & West 2003; Munoz & Warner 2003, 2004), rules governing sex reversal according to reproductive behaviour in molluscs have received less attention (but see Hoagland 1978; Collin 1995; Warner et al . 1996). Among gregarious protandrous mollusc species, the slipper limpet ( Crepidula fornicata ) is ideal for investigating- reproductive patterns in relation to sequential hermaph- roditism. The patterns of gender allocation of this gastropod have been investigated for almost an entire century (e.g. Coe 1936, 1938); an interest largely due to its grouping behaviour. This long-lived species (lifetime of a maximum of 10 years, Blanchard 1995) is typically found in stacks (i.e. groups of individuals attached to each other, with larger (older) individuals, usually females, at the base and smaller (younger) individuals, usually males, at the top; Coe 1936). Because fertilization is internal and these groups are perennial, they are likely to constitute independent mating groups in which sex change occurs according to various factors including sex ratio, number and size of individuals within the stack (Coe 1938; Hoagland 1978). The location of the individuals within a stack remains unchanged over time except for small individuals (male or immature), poten- tially imparting an advantage to males directly attached above a female. Coe (1938) nevertheless speculated on the importance of small mobile males that might obtain most of the fertilizations. While it is well established that sequential hermaphroditism in C. fornicata is characterized by a strong social control of sex change (Collin 1995), questions concerning mating success and sex reversal are still largely open (Gaffney & McGee 1992; Collin 1995). For instance, further studies are required to examine the timing of sex change, to measure male reproductive success, to determine the importance of small males and to investi- gate the possibility of sperm storage and competition. In marine species, effective (realized) fertilizations are difficult to observe directly in the field and are often inferred from copulatory behaviour only, thus neglecting postcopulatory events (e.g. sperm competition). Mating outcomes can nevertheless be deduced from molecular information. The use of microsatellites in parentage analyses (Jarne & Lagoda 1996) can extend to cases in which a small amount of tissue is available, as with juveniles (e.g. Viard et al . 1997) or larval (e.g. Selvamani et al . 2001) gastropods. So far, paternity analyses have been rare in marine species except in mammals (e.g. Clapham & Palsboll 1997; Coltman et al . 1998; Hoffman et al . 2003; Krutzen et al . 2004; Garrigue et al . 2005), turtles (e.g. Fitzsimmons 1998; Hoekert et al . 2002; Moore & Ball 2002; Ireland et al . 2003) and fishes (e.g. Martinez et al . 2000; Pitcher et al . 2003; Soucy & Travis 2003; Chapman et al . 2004; Naud et al . 2004; Petersen et al . 2005). In particular, very few assignments of larval paternity have been achieved in natural populations of benthopelagic invertebrates (but see Coffroth & Lasker 1998). Because larvae of C. fornicata are brooded into the pallilal cavity of the mother before being released as planktonic larvae, they can be easily retrieved for paternity analyses. To our knowledge, only one set of paternity analyses, based on an exclusion procedure, has been carried out on C. fornicata in one native (American) population from Delaware (Gaffney & McGee 1992). Although it was concluded that multiple paternity was likely to occur, the lack of power of the markers used (allozyme loci) prevented a precise assign- ment of paternity, and thus the extent of multiple paternity within a stack, reproductive success of young mobile males and the relation between mating success and sex reversal could not be assessed in this species. A second interest in studying mating strategy in C. fornicata relates to its successful colonization of Europe. Out of the 104 exotic species that have been introduced along the Channel and Atlantic coasts of France (Goulletquer et al . 2002), only a few are invasive (i.e. exotic species with significant side effects); C. fornicata is prominent among these emblematic species. Native to western North Atlantic coasts, this gastropod was repeatedly introduced into Europe during the 19th and 20th centuries (Blanchard 1997). Sex reversal and social grouping could have partici- pated to its success as a colonist, for instance by increasing effective population size as well as the probability of finding mates (Blanchard 1995; Dupont et al . 2003); yet its mating behaviour has never been directly investigated in European populations. Based on five microsatellite loci and a maximum-likelihood categorical analysis, we performed paternity assignment of larvae in three French sites invaded by C. fornicata . This study aimed at addressing the following questions and hypotheses: (i) Does paternity come mainly from the closest male to the study mother or from the largest male within the stack? (ii) What is the contribution of young LARVAL PATERNITY ASSIGNMENT IN SLIPPER LIMPETS 3011 © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd mobile males (i.e. proportion of paternity assigned outside the stack)? (iii) Is there any evidence for sperm storage and sperm competition, two important features that might influence reproductive success and sex-reversal strategies? The inferred mating patterns are discussed in light of the individual features (size, position and sex at the time of sampling) of the assigned fathers and the population characteristics at the sites (sex ratio and size at sex change). Materials and methods Collection of adults and larvae Sampling of adults was conducted by dredging during the breeding season in June 2003 in three sites (called populations in the following text; Table 1), separated by less than 10 km, in the Bay of Brest (Brittany, France), a semi-enclosed French marine ecosystem where Crepidula fornicata covered 61% of the seabed in 1995 (18 500 tonnes wet weight; Chauvaud 1998). Fifteen, 11 and 16 stacks (comprising 86, 87 and 112 adults) were sampled in the Roscanvel, Keraliou and Rozegat populations, respectively (Table 1). Six, five and seven stacks in the three populations were dismantled respectively and, except for the individual at the base (i.e. the study mother, see below), all the sampled adults were preserved in 96% ethanol for genetic analyses. From our previous surveys, the individuals located at the base of each of these 18 stacks were known to be females ( c . 100% of the stacks; L. Dupont & J. Richard, personal observation, unpublished data) and most probably with broods because of the sampling season (i.e. in the bay of Brest the maximum brooding activity occurred in May to June, with c . 80% of females carrying eggs, Richard et al . in press). These females at the base of each stack are attached to a substratum (e.g. dead shell of C. fornicata ) and incubate the capsules between the propodium and neck, with embryos packed in thin-walled capsules attached to the substratum by a peduncle (Hoagland 1986). To avoid disturbances that might alter embryo development within the capsules, the study females were kept isolated in aquaria with filtered sea water. A sand filter was used to avoid the circulation of C. fornicata larvae (400–1200 µ m, Table 1 Study populations and paternity analyses results. Location, density (L. Guérin, personal communication), sex ratio (F: M; b inomial test in parenthesis) and mean number of adults per stack are given. Results of the modal decomposition of size-frequenc y distributions are summarized by the number of age groups (i.e. cohorts; N AG ) and the averaged size and standard deviation (SD) o f each age group. Average relatedness within each population (R pop ), between each parental pair (R p ) and between each nonparental pair within stacks (R np ) are given for each population as well as the percentage of brooding females, the size at sex change (L 50 ) and the number of larvae and broods used in paternity analyses. A synthesis of the results of the paternity analysis is given by (i) the percentage of larvae with unassigned paternity; (ii) the percentage of larvae for which all individuals in the stack were excluded as potential father; (iii) the percentage of multiply sired broods; and (iv) the mean number of fathers per brood (N fathers ; including external fathers). SD, standard deviation Roscanvel Keraliou Rozegat Population characteristics Location 48°19′810”N 48°22′328”N 48°19′300”N 04°30′700”W 04°25′694”W 04°31′700”W Density (gm −2 ) 2000 100 600 N adults analysed 86 87 112 Sex ratio 0.96 : 1 (P = 0.489) 0.75 : 1 (P = 0.115) 0.69 : 1 (P < 0.001) N adults/stack ± SD 6.86 ± 2.26 8.64 ± 4.05 7.31 ± 1.86 N AG 24 4 Mode (cm) 6.58–9.32 3.40–6.36–8.86–11.25 4.52–7.02–9.12–10.76 SD (cm) 0.51–1.02 1.29–0.67–1.24–0.79 1.36–0.65–0.61–0.52 R pop −0.009 −0.010 − 0.008 R p 0.023 −0.130 0.107 R np −0.023 0.052 0.045 L 50 (cm) 9.1 9.4 9.6 Brooding females 77% 81% 57% Paternity analyses N larvae (N broods ) 77 (6) 72 (5) 90 (7) Unassigned paternity 5.2% 19.4% 23.3% External paternity 1.4% 8.6% 15.9% Multiple paternity 50% 80% 100% N fathers ± SD 2.2 ± 1.6 2.2 ± 1.1 3.1 ± 0.7 3012 L. DUPONT ET AL. © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd Pechenik et al . 2004) but allowed the passage of small particles (= 30 µ m) for feeding. After 31 days (i.e. mean time between egg laying and hatching in C. fornicata , Richard et al . in press), all the 18 females had released larvae. These females and a sample of 239 larvae (77, 72 and 90 larvae from the Roscanvel, Keraliou and Rozegat populations, respectively) were preserved in 96% ethanol for subsequent analyses. All the study individuals were sexed: following Hoagland (1978), the sexual morphs were determined morphologically, in particular according to presence or absence of penis. Besides the study mother, the presence of eggs was also recorded for the other females of the stacks. In addition, the curvilinear shell length and the location in the stack of each adult were registered. Sex and size analyses at the population level Percentage of brooding females (calculated over the total number of females) and female : male ratio were estimated for each population. Departure from a 1 : 1 sex ratio was tested using a binomial test (Wilson & Hardy 2002, p. 54). Effects of sex and population on the size of individuals were investigated using a mixed-model anova using the general linear model procedure of minitab ® release 14.1 with the sexual morphotype and the population effect incorporated into the model as fixed factors. Size at sex change ( L 50 = size at which 50% of the individ- uals are female) was calculated for each population using a logistic regression following Allsop & West (2003). To analyse the population age structure, a crude demographic structure analysis was carried out by modal decomposi- tion of size-frequency distribution using the mix 2.3 program package (MacDonald & Pitcher 1979; see Dupont et al . in press for details). DNA extraction and microsatellite typing Total genomic DNA was extracted from all adults and larvae using Nucleospin®Multi-96 Tissue Kit (MACHEREY- NAGEL). Samples were genotyped at five loci: four, namely CfCA2, CfCA4, CfCATGT and CfGT14, as defined by Dupont & Viard 2003) and one (CfH7) newly developed locus (Kruse & Viard, unpublished data; forward and reverse primer sequences: F: 5 ′ -GGTAACGTATTGCT- ACCGAAAG-3 ′ and R: 5 ′ -TCATGCGGGTTTGGTGG-3 ′ ). Loci [including CfH7 (annealing temperature of 54 ° C and 1.5 m m of MgCl 2 )] were amplified by polymerase chain reactions (PCR) according to Dupont & Viard (2003). For larvae (which yielded a small amount of extracted DNA) the only modification compared to Dupont & Viard (2003) was a pre-amplification step with the same protocol before final amplification. To avoid scoring error, each larva was genotyped twice. Paternity and relatedness analyses At the population level, number of alleles and gene diversity were estimated using genetix version 4.02 (Belkhir et al . 2004). Tests for genotypic linkage disequilibria among loci were computed with genepop version 3.3 (Raymond & Rousset 1995). The paternity analysis was performed using a maximum- likelihood-based categorical analysis following Meagher (1986) with the software cervus version 2.0 (Marshall et al . 1998). Given the genotypes of offspring, their known mothers and the candidate fathers, the paternity was assigned to the most likely father, i.e. the individual with the highest log-likelihood ratio (LOD score as defined in Meagher 1986). Computer simulations were used to assess the statistical significance of LOD scores (10 000 iterations, based on allelic frequencies of the entire population): paternity was assigned to the most likely father if the difference between the LOD score of the most likely father and that of the second most likely father was statistically significant (here with an 80% confidence level; Marshall et al . 1998). All the sampled individuals of a given population were considered as candidate fathers. Given the possib- ility for change of sex in the time interval between copulation and sampling, this list comprised all the mature individuals (i.e. males, females and individual in sexual transition), including the mother, to take into account the possibility for self-fertilization (as hypothesized by Orton 1950). The proportion of multiply sired broods was compared between populations by a Fisher test using the program struc (500 000 iterations) of the software genepop version 3.3. The relation between paternity status and size of individuals at the population level was investigated using a mixed-model anova, with the paternity status (father vs. not father) and the population effect incorporated to the model as fixed factors. Because the size of the individuals in a stack depends on the size of the individual at the base (Coe 1938), individual size was weighted by the size of the individual at the base of the same stack. To analyse the genetic relatedness between mates within maternal stacks, a pairwise microsatellite-based relatedness coefficient R xy (Queller & Goodnight 1989) was calculated using identix software (Belkhir et al. 2002). R xy values were compared among two categories: parental pairs (mother and assigned father) and nonparental pairs (mother and individuals of the maternal stack that have not been assigned as father). Results Sex ratio and comparison of size within and between populations At the level of the Bay, the three populations did not share common features in terms of gender allocation and LARVAL PATERNITY ASSIGNMENT IN SLIPPER LIMPETS 3013 © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd reproductive status (Table 1). A male-biased sex ratio is expected in protandrous species (Allsop & West 2004) including Crepidula fornicata (Hoagland 1978; Collin 1995), but in our study such a pattern was observed in Rozegat only (Table 1, binomial test, P < 0.001). This cannot be explained by the sampling procedure (dredge effects), as every population was sampled in the same way. Moreover, only intact stacks (i.e. stacks without mark left by unstuck individual) were collected. Previous studies carried out in another Breton population (Morlaix Bay) and aiming at comparing different sampling methods (scuba diving vs. dredging) also did not show differences for the mean number of individuals per stack and sex ratio according to the sampling procedure (L. Dupont & F. Viard, unpublished data). This suggests that if some individuals were lost during dredging, they should be mainly immature individuals that are small and not firmly attached to the stack. Although the sex ratio may change during the year (Richard et al. in press), the relatively high number of females in Roscanvel is thus likely to be due to a low recruitment in this population, as revealed by the demographic analysis: the smallest size-class observed in the two other populations is absent in Roscanvel (Table 1). The Rozegat population displayed the lowest number of brooding females, with only 57% of the females possess- ing egg capsules (Table 1), suggesting asynchrony in reproduction at the level of the Bay, population density effects or different environmental forcing (e.g. contaminants). As also expected in this protandrous species, a significant size difference was observed between individuals of the two sexes (d.f. = 1, F = 141.34, P = 0.000, Fig. 1) over all populations although varying according to the population (d.f. = 2, F = 18.49, P = 0.000), with a significant sex–population interaction (d.f. = 2, F = 8.43, P = 0.000). We also observed a slight variation of size at sex change (L 50 ) across populations (Table 1). Between Rozegat and Keraliou, this difference is congruent with the difference of mean size of males (5.3 ± 2.1 and 5.7 ± 2.1 cm, respectively). However, Roscanvel showed the lower size at sex change together with the higher mean size of males (7.9 ± 2.40 cm); a result due to the lack of small (young) males in the population and to the strong overlap of male and female size distributions (Fig. 1). Paternity analysis The five loci did not exhibit any linkage disequilibria. Four of them were highly polymorphic with 10–32 alleles over the 285 adults analysed, while the fifth locus CfCA2 exhibited three alleles. The three populations displayed a similar mean number of alleles (from 11.7 to 12.2) and gene diversity (from 0.768 to 0.777) over loci. This high polymorphism explained the very high value for exclusion probability estimated over the whole study (i.e. 285 adult individuals) reaching 99.5% over the five loci. As regularly observed with microsatellite loci (e.g. McCracken et al. 1999; Davis et al. 2001; Spritzer et al. 2005) and easily recognized with paternity analysis when the mother is genotyped, null alleles were noticed at two loci (CfCATGT and CfGT14). Taking advantage from the procedure implemented in cervus, we specified a non-null error rate so that a true father that mismatched at one or two loci could still be identified as the most likely parent. This procedure is reliable, provided that the exclusion power of the loci is reasonably high (Marshall et al. 1998), which is the case with our set of loci. As we were interested first in excluding as many fathers as possible, this procedure was also conservative. Note that the final assignment made with the five loci was always confirmed with the three loci that did not exhibit null alleles and for which maternal alleles always segregated in accordance with expected Mendelian proportions. Results of maximum-likelihood-based paternity analyses are summarized in Table 1 and detailed in the Appendix. None of the larvae exhibited a genotype compatible with maternal self-fertilization. Over the 239 larvae analysed, 39 could not be unambiguously assigned (i.e. with an 80% Fig. 1 Curvilinear length frequency distribution in the three stud y populations. Black, white and grey bars are featuring males, females and ‘fathers’, respectively, as assigned by paternit y analysis. 3014 L. DUPONT ET AL. © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd confidence level). They were classified as larvae with unassigned paternity and removed from subsequent analyses. Out of the 200 larvae unambiguously assigned to a father, few involved a male external to the maternal stack (Table 1). The largest proportion of external-assigned paternity was found in Rozegat (15.9%). Interestingly, multiple paternities were revealed in almost all the broods (Table 1), although the mean number of larvae analysed per female (12.8 ± 1.9, 14.4 ± 1.9 and 12.9 ± 1.7 in Roscanvel, Keraliou and Rozegat, respectively) was low compared to the approximately 25 000 larvae potentially released per female (Richard et al. in press). The maximum number of fathers assigned for a given brood was 5 (in Roscanvel, see Appendix). Differences were nevertheless observed across populations multiple paternities were observed in all the broods in Rozegat but in only half of the broods in Roscanvel. Paternity patterns thus appeared to differ according to population, although the associated Fisher test was not significant (P = 0.093). Size, position within stacks and sex of fathers Figure 2 shows the sex of the assigned fathers at the time of sampling. Surprisingly, depending on the population, 13% to 35% were females, some of them with egg capsules, suggesting not only that sex changes had occurred since the time of copulation but also that some of these individuals had subsequently reproduced as a female. Another surprising result was the location of the assigned fathers as compared to the position of the mother within the stack (Fig. 3A): individuals occupying any position in the maternal stack can be a father, including males that were not the first above the mother (Fig. 3B). Figure 3A2 showed, however, that more larvae were sired by larger males than small males in the Rozegat population. Analysis of variance showed that fathers are, on average, significantly taller than nonfathers within a stack (d.f. = 1, F = 5.21, P = 0.024). In the same way, ‘male fathers’ (i.e. fathers that were males at time of sampling) are, on average, significantly taller than ‘male nonfathers’ (i.e. males of the maternal stack that were not fathers; d.f. = 1, F = 8.76, P = 0.004). Pairwise average relatedness (R xy ) values were not different between parental and nonparental pairs when considering each population separately (Table 1) or across the overall study (H = 0.10, d.f. = 1, P = 0.749). Discussion Of 200 larvae unambiguously assigned, 183 (91.5%) were assigned to an individual belonging to the maternal stack. This pattern holds in the three Breton populations studied, confirming previous expectations based on observations of copulatory behaviour (Hoagland 1978). The mating patterns revealed by the present paternity analysis thus ruled out the hypothesis that mobile males might obtain most of the fertilizations by crawling among stacks (Coe 1936, 1938; Wilczynski 1955). In their paternity exclusion analysis on Crepidula fornicata, Gaffney & McGee (1992) suspected genetic contributions by individuals not pres- ent in the stack at the time of collection; a conclusion unfortunately hampered by the low exclusion power of the enzymatic loci used. Here, microsatellite loci afforded considerable improvement over allozyme markers: only 8.5% of the larvae in the study were estimated to be sired by individuals not present in the stack at the time of sampling. The true percentage of external fathers might be even lower as we cannot exclude the possibility of having lost some ‘candidate fathers’ from the stack since copulation (for instance during the sampling, i.e. dredging effects). Besides, the mobile males are young individuals expected to be in a side position. Seven percent of the larvae were assigned to four individuals located in a side position (out of 37 identified fathers) and three of these were old (i.e. large; one large male, one female and one individual in sexual transition). Mobile males, if any, are thus exceptions and mating between individuals within a stack is the rule. Our study highlights the importance of gregariousness in the reproductive success of C. fornicata. The species forms social groups, with a perennial assemblage of individuals of various ages and sexes that reproduce with each other and are not genetically related. Social groups are common among animals (e.g. Wilson 1975), including benthic marine invertebrates (e.g. annelids, bivalves and barnacles; Toonen & Pawlik 2001). Social interactions are known to have major influence on reproductive strategies as is well documented for harem systems in protogynous fishes (Munoz & Warner 2003). Adult aggregations, which occur in numerous gastropod species (reviewed in Baur 1998), may enhance reproductive success in species with internal fertilization and low mobility like C. fornicata by Fig. 2 Percentage of larvae, for which the assigned father was a male (M), an individual in sexual transition (T), a female (F) or a b rooding female [F(b)] at the time of sampling, in each population. LARVAL PATERNITY ASSIGNMENT IN SLIPPER LIMPETS 3015 © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd enhancing the probability for individuals to meet and realize mating (Baur 1998). The gregarious behaviour is thus increasing the long list of traits that make C. fornicata a good colonizer. As an introduced species, the combination between social grouping, repeated introductions (Blanchard 1997), dispersal ability conferred by pelagic larvae release (Viard et al. 2006) and a long season of reproduction (Richard et al. in press) is congruent with a rapid increase in population density and range expansion following primary introductions (Sakai et al. 2001). In addition, age segregation of sex allows reproduction between different age groups, a pattern favouring outbreeding as well as temporal genetic homogeneity, as described in Dupont et al. (in press). This aptitude to find a mate, in such a species with low mobility in the adult phase, is exemplified by a striking pattern observed within stacks: a significant contribution to paternity by individuals collected as transitional individuals and females in the three study populations. Two hypotheses could explain such a result: (i) bisexuality of the assigned fathers; and (ii) sperm storage by the study mothers. The former hypothesis (i.e. occurrence of transi- tory simultaneous hermaphrodites) is highly unlikely for three reasons: (i) none of the assigned fathers that were females at time of collection exhibited a penis and thus none of them could have been a functional male; (ii) although functional hermaphroditism was once suggested to occur exceptionally (Coe 1938), to our knowledge, bisexual individuals have never been documented by histology studies (see Le Gall 1980; Martin 1985 and references therein; J. Richard, unpublished data); and (iii) the sequential changes in morphology and anatomy during sex reversal in C. fornicata prevent bisexuality (Orton 1909; Coe 1938; Chipperfield 1951; Martin 1985); for example, the first step Fi g. 3 Di str ib ut i on ( percentage ) o f t h e different positions occupied within the maternal stack by the assigned fathers. (A1) Position of the father is given relative to the position of the mother (0 is the mother). ′Side position′ is used for single individuals recorded in a side position relative to the main stack. ′Secondary stacks’ refer to a situation where several individuals are forming a secondary chain branched on the primary stack. (A2) Same figure as A1, but weighted by the percentage of larvae sired. (B) Position of the father is given relative to the position of the male (position 0), which is the closest to the mother at the time of the collection within the main stack. For example, positions (−1) and (−2) refer to females situated in between the studied mother and the male closest to the mother. 3016 L. DUPONT ET AL. © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd is the cytolysis of the spermatogenic tissues; also the distal part of the gonoduct can develop into a prominent uterus with folded walls, into which a number of seminal recep- tacles opens, only when the penis degenerates, allowing the width of the gonoduct to increase and the inner walls to become folded longitudinally (Orton 1909; Chipperfield 1951). Conversely, several lines of evidence support the hypothesis of sperm storage, which was also suspected in the study of Gaffney & McGee (1992): (i) histological obser- vations (Martin 1985; J. Richard personal observation); (ii) experimental data: Hoagland (1978) asserted that ‘females can store sperm for at least one year’ according to one of her previous study (Hoagland 1975). In our experiment, one of the study mothers also produced a second set of broods after the first hatching indicating that sperm was stored for at least one month. Finally, even if a rare event of bisexuality was occurring in the study populations, this could not explain the large proportion of assigned fathers (32%, 13% and 35%, respectively, in the Roscanvel, Keraliou and Rozegat populations) that were females at time of collection. Sperm storage is thus the most likely explanation for the observed results. That some fathers are brooding females at time of collection demonstrates that an individual can effectively reproduce as both a male and a female over a relatively short time interval and contribute as both father and mother to larvae of a given cohort. Sexual trans- formation lasts 61 days (Coe 1938); a much longer time than in many protogynous fish species (Reavis & Grober 1999; Sunobe et al. 2005) but similar to other protandrous species of calyptraeids (Warner et al. 1996; Collin et al. 2005). Egg production lasts 14 days (Chipperfield 1951), whereas hatching occurs in a minimum of 21 days (Chipperfield 1951). Consequently, individuals that were both females with eggs and assigned fathers had been males and transmitted sperm at least 54 days before. Sperm storage and gregarious behaviour might thus work in concert to maximize individual reproductive success in C. fornicata. However, aggregations might also reduce individual reproductive success, especially among males, because of competition for mates (Toonen & Pawlik 1994). Another clear-cut result of our study is the remarkable level of multiple paternity, which was observed in 14 out of 18 broods. On average, two to three fathers, with a maxi- mum of five fathers, were identified in subsamples of only 11–16 larvae per brood. Although investigations of pater- nity among gastropods have been largely restricted to pulmonates (e.g. Baur 1998) with few studies within marine prosobranchs (Gaffney & McGee 1992; Paterson et al. 2001), multiple paternity has been reported in several gastropod families, suggesting that multiple copulations and fertilizations by different males are common (see Baur 1998). In terms of population effective size, multiple paternity coupled with sex reversal is an advantageous breeding tactic: the increased number of reproducing males and the sex-ratio adjustment both enhance the effective population size (Sugg & Chesser 1994; Martinez et al. 2000). Benefits of multiple paternity, sex change and social structure are combined, thus maintaining large genetic diversity as well as large effective size over time in C. fornicata populations (Dupont et al. 2003). As a result of multiple copulations, sperm from different males may compete to fertilize a single brood (Parker 1970). In addition, the occurrence of sperm storage increases the probability of biased paternity. Thus, as a result of multiple paternity and sperm storage, male–male competition might indeed occur. When sperm competition occurs, paternity is frequently determined by the relative number of compet- ing sperm present from rival males but also by sperm qual- ity (i.e. sperm size, viability and mobility; review in Snook 2005). In C. fornicata, the quantity of sperm transferred to the female could be related to the size of the male or to ease of access to the female. Interestingly, a significant size difference was observed between fathers and nonfathers in the three C. fornicata populations, and there was still a size difference when comparing only males, suggesting that the most successful males are the largest. In addition, Fig. 3A2 shows that more larvae have been sired by larger males than by small males in Rozegat population. This result means that male fertility is expected to increase with size, a pattern expected in protandrous species exhibiting a sex ratio biased towards the first sex (Charnov & Bull 1989). We indeed observed a male-biased sex ratio in Rozegat population. However, because of the grouping behaviour of C. fornicata, there is a strong interaction between age, size, sex and position in a stack. Hence the largest male was also often the closest to the mother, a position that may facilitate the copulation. This study was not designed to distinguish between the effects of size and position. In addition, only the offspring of females at the base of the stacks were examined, so that the male con- tribution to the broods of other females in the stack is unknown. Further analysis is needed to investigate precisely the components of male reproductive success. Recently, the study by Munoz & Warner (2003) of pro- togynous fish gave new insight into sex-change theory: the authors showed that social conditions indicative of sperm competition may cause sex reversal to be deferred because intense competition can substantially lower the expected reproductive success of males. Considering that the likelihood of sperm competition might influence the timing of sex change, it is likely that population character- istics such as sex ratio and optimal size at sex change vary with the intensity of sperm competition. Here, we showed that sex ratio, demographic structure and mating patterns varied across the three study populations. In particular, Rozegat displayed the highest incidence of external assigned paternity, the highest frequency of multiple paternities LARVAL PATERNITY ASSIGNMENT IN SLIPPER LIMPETS 3017 © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd together with the largest male-biased sex ratio, the largest size at sex change and the lowest proportion of brooding females. In this population, sperm competition should increase as the number or eggs available to fertilize decrease. Under the assumption that the intense male–male competition causes sex change to be delayed, it is not surprising to observe a large optimal size at sex change and a sex ratio largely male-biased. Conversely, levels of multiple pater- nity and external paternity were found to be the lowest in Roscanvel, for which the smaller size at sex change and a sex ratio close to 1:1 were noticed (Table 1). The fact that mating patterns and gender allocation patterns varied in concert across sites suggests that multiple paternities, reflecting perhaps differential sperm competition intensity, might enhance sex reversal in C. fornicata. The occurrence of male–male competition influencing sex change in protandrous gregarious species might also explain that species-forming mating groups have more variation in size at sex change within a population than solitary species do (Collin 2006). Detailed investigations of male–male competition and factors determining the timing of sex reversal in protandrous species are needed to better elucidate the evolution of sex reversal strategies especially in gregarious species. Acknowledgements This project is part of the 2001 INVABIO program of the Ministère de l’Ecologie et du Développement Durable (MEDD; project no. D4E/SRP/01115). Additional support for genotyping and iso- lation of new markers was obtained from the Network of Excellence ‘Marine Genomics Europe’ (contract n°505403). F.V. acknow- ledges support from the CNRS (ATIP program). L.D. benefited from a PhD grant from the Region Bretagne. The authors are also grateful to J.D.D. Bishop and M. Valero and three anonymous referees for critical readings, editing and improvements to the manuscript. References Allsop DJ, West SA (2003) Constant relative age and size at sex change for sequentially hermaphroditic fish. Journal of Evolu- tionary Biology, 16, 921–929. Allsop DJ, West SA (2004) Sex-ratio evolution in sex changing animals. Evolution, 58, 1019–1027. Baur B (1998) Sperm competition in molluscs. In: Sperm Competi- tion and Sexual Selection (eds Birkhead TR, Moeller AP), pp. 255– 305. Academic Press, London. Belkhir K, Castric V, Bonhomme F (2002) identix, a software to test for relatedness in a population using permutation methods. Molecular Ecology Notes, 2, 611–614. Belkhir K, Borsa P, Goudet J, Chikhi L, Bonhomme F (2004) GENETIX 4.05, logiciel sous Windows pour la génétique des populations. Lab- oratoire Génome, Population, Interactions. CNRS UMR 5000, Université Montpellier II, Montpellier, France. Blanchard M (1995) Origine et état de la population de Crepidula fornicata (Gastropoda Prosobranchia) sur le littoral français. Haliotis, 24, 75–86. Blanchard M (1997) Spread of the slipper limpet Crepidula fornicata (L. 1758) in Europe. Current state and consequences. Scientia Marina, 61, 109–118. Chapman DD, Prodohl PA, Gelsleichter J, Manire CA, Shivji MS (2004) Predominance of genetic monogamy by females in a hammerhead shark, Sphyrna tiburo: implications for shark conservation. Molecular Ecology, 13, 1965–1974. Charnov EL (1982) The Theory of Sex Allocation. Princeton University Press, Princeton, New Jersey. Charnov EL, Bull JJ (1989) Non-fisherian sex ratios with sex change and environmental sex determination. Nature, 338, 148–150. Chauvaud L (1998) La coquille St Jacques en rade de Brest; un modèle biologique d’étude des réponses de la faune benthique aux fluctuations de l’environnement. Thèse, Université de Bretagne Occidentale, France. Chipperfield PNJ (1951) The breeding of Crepidula fornicata in the river Blackwater, Essex. Journal of the Marine Biological Association of the United Kingdom, 30, 49–71. Clapham PJ, Palsboll PJ (1997) Molecular analysis of paternity shows promiscuous mating in female humpback whales (Megaptera novaeangliae, Borowski). Proceedings of the Royal Society of London. Series B, Biological Sciences 264, 95–98. Coe WR (1936) Sexual phases in Crepidula. Journal of Experimental Zoology, 72, 455–477. Coe WR (1938) Conditions influencing change of sex in molluscs of the genus Crepidula. Journal of Experimental Zoology, 77, 401– 424. Coffroth M, Lasker HR (1998) Larval paternity and male repro- ductive success of a broadcast-spawning gorgonian, Plexaura kuna. Marine Biology, 131, 329–337. Collin R (1995) Sex, size, and position: a test of models predicting size at sex change in the protandrous gastropod Crepidula forni- cata. American Naturalist, 146, 815–831. Collin R (2006) Sex ratio, life history invariants, and patterns of sex change in a family of protandrous gastropods. Evolution, 60, 735–745. Collin R, McLellan M, Gruber K, Bailey-Jourdain C (2005) Effects of conspecific associations on size at sex change in three species of calyptraeid gastropods. Marine Ecology Progress Series, 293, 89–97. Coltman DW, Bowen WD, Wright JM (1998) Male mating success in an aquatically mating pinniped, the harbour seal (Phoca vitulina), assessed by microsatellite DNA markers. Molecular Ecology, 7, 627–638. Davis LM, Glenn TC, Elsey RM, Dessauer HC, Sawyer RH (2001) Multiple paternity and mating patterns in the American alligator, Alligator mississippiensis. Molecular Ecology, 10, 1011–1024. Dupont L, Viard F (2003) Isolation and characterization of highly polymorphic microsatellite markers from the marine invasive species Crepidula fornicata (Gastropoda: Calyptraeidae). Molecular Ecology Notes, 3, 498–500. Dupont L, Jollivet D, Viard F (2003) High genetic diversity and ephemeral drift effects in a successful introduced mollusc (Crepidula fornicata: Gastropoda). Marine Ecology Progress Series, 253, 183–195. Dupont L, Bernas D, Viard F (in press) Sex and genetic structure across age groups in populations of the European marine invasive mollusc, Crepidula fornicata L. (Gastropoda). Biological Journal of the Linnean Society. Fitzsimmons NN (1998) Single paternity of clutches and sperm storage in the promiscuous green turtle (Chelonia mydas). Molecular Ecology, 7, 575–584. 3018 L. DUPONT ET AL. © 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd Gaffney PM, McGee B (1992) Multiple paternity in Crepidula fornicata (Linnaeus). Veliger, 35, 12–15. Garrigue C, Dodemont R, Steel D, Baker C (2005) Organismal and ‘gametic’ capture-recapture using microsatellite genotyping confirm low abundance and reproductive autonomy of humpback whales on the wintering grounds of New Caledonia. Marine Ecology Progress Series, 274, 251–262. Ghiselin MT (1969) The evolution of hermaphroditism among animals. Quarterly Review of Biology, 44, 189–208. Goulletquer P, Bachelet G, Sauriau PG, Noel P (2002) Open Atlantic coast of Europe — a century of introduced species into French waters. In: Invasive Aquatic Species of Europe (eds Leppäkoski E, Gollash S, Olenin S), pp. 276–290. Kluwer Academic Publishers, Dordrecht, The Netherlands. Heller J (1993) Hermaphroditism in molluscs. Biological Journal of the Linnean Society, 48, 19–42. Hoagland KE (1975) Reproductive Strategies and Evolution in the Genus Crepidula (Gastropoda: Calyptraeidae). PhD Thesis, Harvard University Press, Cambridge, Massachusetts. Hoagland KE (1978) Protandry and the evolution of environmentally- mediated sex change: a study of the mollusca. Malacologia, 17, 365–391. Hoagland KE (1986) Patterns of encapsulation and brooding in the Calyptraeidae (Prosobranchia: Mesogastropoda). American Malacological Bulletin, 4, 173–183. Hoekert WE, Neufeglise H, Schouten AD, Menken SB (2002) Multiple paternity and female-biased mutation at a microsatellite locus in the olive ridley sea turtle (Lepidochelys olivacea). Heredity, 89, 107–113. Hoffman SG, Schildhauer MP, Warner RR (1985) The costs of changing sex and the ontogeny of males under contest competition for mates. Evolution, 39, 915–927. Hoffman JI, Boyd IL, Amos W (2003) Male reproductive strategy and the importance of maternal status in the Antarctic fur seal Arctocephalus gazella. Evolution, 57, 1917–1930. Ireland JS, Broderick AC, Glen F et al. (2003) Multiple paternity assessed using microsatellite markers, in green turtles Chelonia mydas (Linnaeus, 1758) of Ascension Island, South Atlantic. Journal of Experimental Marine Biology and Ecology, 291, 149–160. Jarne P, Lagoda J-L (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution, 11, 424– 429. Krutzen M, Barre LM, Connor RC, Mann J, Sherwin WB (2004) ‘O father: where art thou?’ — Paternity assessment in an open fission-fusion society of wild bottlenose dolphins (Tursiops sp.) in Shark Bay, Western Australia. Molecular Ecology, 13, 1975– 1990. Le Gall P (1980) Etude expérimentale d’association en chaîne et de son influence sur la croissance et la sexualité chez la crépidule (Crepidula fornicata Linné 1758). PhD Thesis, University of Caen, France. Levitan DR (1993) The importance of sperm limitation to the evolution of egg size in marine invertebrates. American Naturalist, 141, 517–536. MacDonald PDM, Pitcher TJ (1979) Age-groups from size- frequency data: a versatile and efficient method of analysing distribution mixtures. Journal of the Fish Resource Board of Canada, 36, 987–1001. Marshall TC, Slate J, Kruuk LEB, Pemberton JM (1998) Statistical confidence for likelihood-based paternity inference in natural populations. Molecular Ecology, 7, 639–655. Martin M-C (1985) Etude expérimentale de l’inversion sexuelle et de la morphogénèse génitale femelle chez un mollusque hermaphrodite protandre Crepidula fornicata. PhD Thesis, University of Caen, France. Martinez JL, Moran P, Perez J et al. (2000) Multiple paternity increases effective size of southern Atlantic salmon populations. Molecular Ecology, 9, 293–298. McCracken GF, Burghardt GM, Houts SE (1999) Microsatellite markers and multiple paternity in the garter snake Thamnophis sirtalis. Molecular Ecology, 8, 1475–1479. Meagher TR (1986) Analysis of paternity within a natural population of Chamaelirium luteum. 1. Identification of most-likely male parents. American Naturalist, 128, 199–215. Moore MK, Ball RM (2002) Multiple paternity in loggerhead turtle (Caretta caretta) nests on Melbourne Beach, Florida: a microsatellite analysis. Molecular Ecology, 11, 281–288. Munday PL, Buston PM, Warner RR (2006) Diversity and flexibility of sex-change strategies in animals. Trends in Ecology & Evolution, 21, 89–95. Munoz RC, Warner RR (2003) A new version of the size-advantage hypothesis for sex-change: incorporating sperm competition and size-fecundity skew. American Naturalist, 161, 749–761. Munoz RC, Warner RR (2004) Testing a new version of the size-advantage hypohesis for sex change: sperm competition and size-skew effects in the bucktooth parrotfish, Sparisoma radians. Behavioural Ecology, 15 , 129–136. Naud M-J, Hanlon RT, Hall KC, Shaw PW, Havenhand JN (2004) Behavioural and genetic assessment of reproductive success in a spawning aggregation of the Australian giant cuttlefish, Sepia apama. Animal Behaviour, 67, 1043–1050. Orton JH (1909) On the occurence of protandric hermaphrodism in the mollusc Crepidula fornicata. Proceedings of the Royal Society of London. Series B, Biological Sciences, 81, 468–484. Orton JH (1950) Recent breeding phenomena in the American slipper-limpet, Crepidula fornicata. Nature, 165, 433–434. Parker GA (1970) Sperm competition and its evolutionary consequences in the insects. Biological Review, 45, 525–567. Paterson IG, Partridge V, Buckland-Nicks J (2001) Multiple paternity in Littorina obtusata (Gastropoda, Littorinidae) revealed by microsatellite analyses. Biological Bulletin, 200, 261–267. Pechenik JA, Blanchard M, Rotjan R (2004) Susceptibility of larval Crepidula fornicata to predation by suspension-feeding adults. Journal of Experimental Marine Biology and Ecology, 306, 75–94. Petersen CW, Mazzoldy C, Zarella KA, Hale RE (2005) Fertiliza- tion mode, sperm characteristics, mate choice and parental care patterns in Artedius spp. (Cottidae). Journal of Fish Biology, 67, 239–254. Pitcher TE, Neff BD, Rodd FH, Rowe L (2003) Multiple mating and sequential mate choice in guppies: females trade up. Proceedings of the Royal Society of London. Series B, Biological Sciences, 270, 1623–1629. Queller DC, Goodnight KF (1989) Estimating relatedness using genetic markers. Evolution, 43, 258–275. Raymond M, Rousset F (1995) genepop (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity, 86, 248–249. Reavis RH, Grober MS (1999) An integrative approach to sex change: social, behavioural and neurochemical changes in Lythrypnus dalli (Pisces). Acta Ethologica, 2, 51–60. Richard J, Huet M, Thouzeau G, Paulet Y-M (in press) Reproduc- tion of the invasive slipper limpet, Crepidula fornicata, in the Bay of Brest, France. Marine Biology . [...]... is part of L Dupont’s doctoral research on the dispersal ability and reproduction of the invasive species C fornicata along the European coasts The work was supervised by F Viard, a population geneticist with interests in studies of reproductive system and historical vs contemporary dispersal of coastal marine algae and invertebrates, including introduced species The LEMAR group is involved in studies... studies concerned with the interactions between marine invertebrate populations and environmental variability Its main goal is to understand the dynamics of invasions in coastal marine systems, by invertebrate alien species (bivalves and gastropods), together with the possible consequences on natural habitat alterations and nutrient fluxes (sedimentation/ mineralization) in a context of global change 3020... polymorphism and microsatellites in the hermaphroditic freshwater snail Bulinus truncatus Proceedings of the Royal Society of London Series B, Biological Sciences, 264, 39–44 Viard F, Ellien C, Dupont L (2006) Dispersal ability and invasion success of Crepidula fornicata in a single gulf: insights from genetic markers and larval- dispersal models Helgoländer Marine Research, 60, 144–152 © 2006 The Authors... Results of the paternity analysis for each family in each population For each stack, adults are characterized by their sex when sampled [I, immature; M, male; T, in transition; F, female; and F(b), brooding female], size, position within the stack relative to the mother and relatedness coefficient with the mother The number of assigned larvae is shown for each studied mother (NL) For each brood, the percentage... Allendorf FW, Holt JS et al (2001) The population biology of invasive species Annual Review of Ecology and Systematics, 32, 305–332 Selvamani MJP, Degnan SM, Degnan BM (2001) Microsatellite genotyping of individual abalone larvae: parentage assignment in aquaculture Marine Biotechnology, 3, 478 – 485 Snook RR (2005) Sperm in competition: not playing by the numbers Trends in Ecology & Evolution, 20, 46... 2006 Blackwell Publishing Ltd Warner RR, Robertson DR, Leigh EG Jr (1975) Sex change and sexual selection Science, 196, 633–638 Warner RW, Fitch DL, Standish JD (1996) Social control of sex change in the shelf limpet, Crepidula norrisiarum: size-specific responses to local group composition Journal of Experimental Marine Biology and Ecology, 204, 155–167 Wilczynski JZ (1955) On sex behaviour and sex... sex determination in Crepidula fornicata Biological Bulletin, 109, 353–354 Wilson EO (1975) Sociobiology: The New Synthesis Belknap Press, Cambridge, Massachusetts Wilson K, Hardy ICW (2002) Statistical analysis of sex ratios: an introduction In: Sex Ratios Concepts and Research Methods (ed Hardy ICW), pp 48–92 University Press, Cambridge, UK Wright WG (1988) Sex change in the Mollusca Trends in Ecology... Immunoreactivity in the gobiid fish Trimma okinawae during bidirectional sex change Ichthyological Research, 52, 27–32 Toonen RJ, Pawlik JR (1994) Foundations of gregariousness Nature, 370, 511–512 Toonen RJ, Pawlik JR (2001) Foundations of gregariousness: a dispersal polymorphism among the planktonic larvae of a marine invertebrate Evolution, 55, 2439 –2454 Viard F, Doums C, Jarne P (1997) Selfing, sexual... Multiple paternity and population genetic structure in natural populations of the poeciliid fish, Heterandria formosa Journal of Evolutionary Biology, 16, 1328 –1336 Spritzer D, Solomon NG, Meikle DB (2005) Influence of scramble competition for mates upon the spatial ability of male meadow voles Animal Behaviour, 69, 375 –386 Sugg DW, Chesser RK (1994) Effective population sizes with multiple paternity. .. The number of assigned larvae is shown for each studied mother (NL) For each brood, the percentage of larvae sired by individuals within the maternal stack is given Note that for some broods, the total is not 100%: this is indicative of external paternity (percentage not given in the Appendix) Population Mother Adults % larvae Sex Size (cm) Position* R Roscanvel NL = 73 fR2 NL = 13 9.6 cm R2i1 R2i2 R2i3 . 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd Blackwell Publishing Ltd Gregariousness and protandry promote reproductive insurance. Given the genotypes of offspring, their known mothers and the candidate fathers, the paternity was assigned to the most likely father, i.e. the individual