MALE DISCRIMINATION OF FEMALE MUCOUS TRAILS PERMITS ASSORTATIVE MATING IN A MARINE SNAIL SPECIES

Transcription

1 doi: /j x MALE DISCRIMINATION OF FEMALE MUCOUS TRAILS PERMITS ASSORTATIVE MATING IN A MARINE SNAIL SPECIES Kerstin Johannesson, 1,2 Jon N. Havenhand, 1 Per R. Jonsson, 1 Mats Lindegarth, 1 Annika Sundin, 1 and Johan Hollander 1,3 1 Department of Marine Ecology, Göteborg University, Tjärnö Marine Biological Laboratory, S Strömstad, Sweden 2 Kerstin.Johannesson@marecol.gu.se Received May 14, 2008 Accepted August 5, 2008 Recent research has shown the potential for nonallopatric speciation, but we lack an adequate understanding of the mechanisms of prezygotic barriers and how these evolve in the presence of gene flow. The marine snail Littorina saxatilis has distinct ecotypes in different shore microhabitats. Ecotypes hybridize in contact zones, but gene flow is impeded by assortative mating. Earlier studies have shown that males and females of the same ecotype copulate for longer than mates of different ecotype. Here we report a new mechanism that further contributes to reproductive isolation between ecotypes in the presence of gene flow. This mechanism is linked to the ability of males to track potential partners by following their mucous trail. We show that cliff ecotype males follow the trails of females of the same ecotype for longer than females of the alternate (boulder) ecotype. In addition, cliff males are more likely to follow the mucous trail in the correct direction if the trail is laid by a cliff-female. The capacity to discriminate the ecotype of female mucous trails combined with differential copulation times creates a strong prezygotic reproductive barrier between ecotypes of L. saxatilis that reduces gene flow from cliff to boulder ecotypes by 80%. KEY WORDS: Assortative mating, male mate-choice, marine gastropod, mechanism of reproductive isolation, sexual isolation, sexual selection. Models of nonallopatric (sympatric and parapatric) speciation challenge the traditional view that speciation arises from longterm physical isolation of populations, and results emerging from empirical work support nonallopatric speciation as a possible alternative to allopatric speciation in nature (e.g., Doebeli and Dieckmann 2003; Barluenga et al. 2006; Savolainen et al. 2006; Gavrilets et al. 2007; Gavrilets and Vose 2007). Although in many cases it seems that reproductive isolation has evolved in the presence of gene flow, we still lack detailed explanations of how such barriers evolve. The marine snail Littorina saxatilis provides an interesting case for studies of incipient reproductive isolation. Habitat- 3 Current address: Department of Animal and Plant Sciences, University of Sheffield, Sheffield, N10 2TN, United Kingdom specific ecotypes of this species exhibit an essentially onedimensional mosaic pattern along rocky coasts of the NE Atlantic. Adjacent (parapatric) populations of contrasting ecotypes are connected by gene flow, but this gene flow is impeded by nonrandom mating of ecotypes in zones of overlap (Johannesson et al. 1995; Pickles and Grahame 1999; Rolán-Alvarez et al. 2004; Grahame et al. 2006; Panova et al. 2006; Johannesson 2008). Studies of mate-choice and sexual selection in both laboratory and field situations have led to the suggestion that different mating behaviors can result in assortative mating (e.g., nonrandom microdistribution of ecotypes and longer copulation times with like mates; Rolán-Alvarez et al. 1999; Hollander et al. 2005). The present study investigates a completely new mechanism of reproductive isolation: the ability of males to identify female ecotype, and localize them, by following their mucous trails C 2008 The Author(s). Journal compilation C 2008 The Society for the Study of Evolution. Evolution 62-12:

2 STUDY SNAIL POPULATION In Sweden Littorina saxatilis forms distinct ecotypes on rocky coasts; the E ecotype is small (4 7 mm), fragile, and dwells in crevices and cracks of exposed (E) cliff surfaces, whereas the S ecotype is large (6 12 mm), robust, and inhabits wave-sheltered (S) boulder habitats rich in crabs (Janson 1982, 1983). At the interface of cliffs and boulders both ecotypes occur and overlap in distribution over a zone typically 5 50 m wide (Janson and Sundberg 1983; Panova et al. 2006). Snails of both sexes are promiscuous and mating is frequent throughout the year except during days in the winter when temperature falls below zero, or during warm summer days when emerged snails are dried up and hibernating inside the shell. The male is the most active partner during mating and upon locating a potential mate mounts the partner s shell, positions himself on the right-hand side, and initiates copulation by entering the penis under the shell of the partner. Observations suggest that females rarely try to reject a mating partner at this stage of the copulation (Saur 1990; Pickles and Grahame 1999), however matings vary in length from a few minutes to almost an hour, and short matings are likely to reflect failed mating attempts with no or little transfer of sperm (Hollander et al. 2005). Similar short copulation attempts are observed when a male mistakenly mounts another male or even another species (Saur 1990), indicating that males have a problem in identifying potential mates upon initial contact (some 35% of copulating pairs in the field are, for example, male-male pairs, Saur 1990; Erlandsson 2002). Littorina saxatilis lacks pelagic larvae, and the genetic structure of Swedish populations is correspondingly characterized by strong isolation-by-distance at larger scales (10 4 to 10 5 m) (Janson 1987). At a smaller scale (10 2 to 10 3 m) local geography is the main structuring factor: snails from the same island are typically closely related in neutral markers independent of snail ecotype (Johannesson and Tatarenkov 1997; Johannesson et al. 2004), but detailed investigations of the genetic structure within islands nevertheless show that gene flow between adjacent subpopulations of contrasting ecotypes is impeded by a partial barrier (Panova et al. 2006). Mating experiments in the laboratory support this finding; males mate assortatively when females of both ecotypes are equally available (average copulation times are much longer with like females; Hollander et al. 2005). SNAIL MUCOUS TRAILS All snails produce costly pedal mucus for movement. Snails may follow each other s mucous trails for various reasons, one of which is the potential saving in own mucus production (Davies and Blackwell 2007). However, the observation from several species that males are more likely than females to follow the mucous trail of other individuals, has lead to the suggestion that males may use mucous trails to track potential mates (Erlandsson and Kostylev 1995; Erlandsson 2002). To test this hypothesis in L. saxatilis we video-recorded male and female movements in an experimental arena in which E males and females of both E and S ecotypes were present, and compared tracking distances of males following mucus trails laid by females of each ecotype. E and S ecotype snails differ most in size: an average E snail is only half the size of an average S snail. Size can be an important component of mate choice (Erlandsson and Rolán-Alvarez 1998; Hollander et al. 2005), and consequently we compared tracking distances of males tracking females of normal sizes (E females half the size of S females), and males tracking females of similar sizes (large E females small S females). This allowed us to identify effects of ecotype-specific traits that were independent of size. If trail following is indeed a mate-searching behavior there should be a dominance of positive polarity, that is, males predominantly tracking toward the female. We thus assessed polarity of all males that tracked females, and compared the extent of correct polarity among males tracking females of different size and ecotype. Materials and Methods SAMPLING AND TRAIL-FOLLOWING SETUP Snails were sampled from two genetically independent populations in separate bays (Janson and Ward 1984) on the island of Saltö (58 53 N, 11 8 E), close to Tjärnö Marine Biological Laboratory on the Swedish west coast. We used a round wetted arena (Ø 36 cm) in which trail following was recorded for 1 h using a digital video camera. In each experiment we used five individuals of each of four different size/ecotype categories of females; normal E, large E, small S, and normal S (large E and small S were similar size, normal S was twice the size of normal E). We used 20 E males as trackers (all normal E). (S males are much more sensitive to disturbance in the laboratory environment and therefore much less active than E males, and we therefore decided to use only E males to optimize the amount of data we could get with available experimental resources). Each male tracker was defined a priori as a tracker for a specific category of female, so that results among female groups were not confounded. Each experiment was replicated (with new snails) eight times for each bay (16 experiments in total). Snail densities during experiments were m 2, which is within the range of densities found in nature. Length of trail following was estimated from the number of 1-cm 2 steps shared by marker and tracker snails. Paths sharing only one square were excluded from the analysis and paths adjacent to the edge of the arena were dismissed completely. Furthermore, trails involving more than one marker snail were not considered. We also recorded the polarity of trail following; EVOLUTION DECEMBER

3 positive polarity being trackers following the trail in the same direction that it was laid. An analysis of variance (ANOVA) was used to test the hypothesis that males could discriminate between trails laid down by females of different size/ecotype. The experimental design included the fixed factor Female (female size/ecotype) and the random factors Bay (origin of snails) and Experiment (nested under Bay). Data were square-root transformed to remove heteroscedasticity (Cochran s C). Pooling of sources of variation in the ANOVA was used to improve statistical power of the test for an effect of the factor Female (see below under Pooling Procedures). Specific a priori hypotheses regarding the ability of males to discriminate between trails of the two ecotypes, and the effects of female size were tested with orthogonal contrasts. MODEL OF PATH FOLLOWING To examine the central question of whether snails actively followed mucous trails, or whether trail-following resulted from random movement we compared the observed patterns of trail following with a null-model of snail motility. The objective was to estimate the probability that trails of tracker snails would coincide with those of marker snails to the extent observed in the experiments described above. The model considered the circular spatial domain of the experimental arena (Ø 36 cm), within which paths were modeled as a correlated random walk where the angle between successive steps was sampled from a normal distribution (mean = 0,SD= 22.5 ). This produced model paths similar in form to observed L. saxatilis paths. Paths were reflected at the boundaries, but path segments closer than 1 cm to the boundary perimeter were not included in the sequence analysis. In each run of the model, we simulated a full hour of movement by two independent snails, at a mean velocity of cm s 1 and a time step of 1 s (parameters estimated from observations of snail movements). A sequence analysis was then performed to search for overlapping trail segments and determine their length. Paths were first transformed to a series of integer coordinates corresponding to the spatial resolution (1 cm) of the experimental recordings of L. saxatilis movement. A matching algorithm then searched for overlapping path segments of successively longer length. This analysis was performed for 160 simulated pairs of paths. We estimated the probability of obtaining a sequence overlap of a specified length from the frequency distribution of overlapping path lengths. The model was coded and run in MatLab (MathWorks Inc., Natick, MA). Results MOVEMENT SIMULATION Model simulations showed that male trail following was not simply a consequence of stochastic movement. In simulations, snails followed trails for more than four consecutive spatial steps in 5% of the cases. The corresponding figure from our experiments was 42%. Hence we concluded that the trail-following we observed in our experiments was an active behavioral trait and not simply an emergent consequence of random snail movement. TRAIL FOLLOWING Male tracking behavior varied significantly with female size/ecotype category (Table 1, pooled ANOVA). Specifically, males (E ecotype) tracked normal E females for longer distances than normal S females (Contrast test, Table 1; Fig. 1). Normal Table 1. ANOVA and a priori test (Orthogonal Contrasts) of the effect of female category (size/ecotype) on the distance over which males tracked the female trail. Distance was tested against three factors: Female (four levels, E normal and large, S small and normal), Bay (two levels), and Experiment (16 levels). Data were square-root transformed. Results of full and pooled tests are shown. ANOVA Source df Mean square F-value P-value Pooled terms Pooled MS New F New P Female (A) Bay (B) Female Bay (C) C+E+F Expt (Bay) (D) Female Expt (Bay) (E E+F 6.48 Residual (F) Contrasts (a priori test) Comparison Full design P-value Pooled design P-value Normal E vs. normal S LargeEvs.smallS Dependent: Total distance of male tracking (square root transformed) EVOLUTION DECEMBER 2008

4 Figure 1. Tracking distance for males tracking females of different ecotype and size (normal E, large E, small S, and normal S). Light bars show total distance, and dark bars net distance in the direction of the female (positive minus negative distance), per tracking male (N = 80) during 1 h with standard error of means. Level of significance (from Tables 1 and 2) is indicated for the comparisons: Normal E versus normal S, and large E versus small S (these are equal sizes). S females are twice as large as normal E females. On the other hand, we found no significant difference in tracking distance for males following similar-sized S and E females (small S and large E) (Contrast test of Table 1 and Fig. 1). This indicated that male s tracking behavior was influenced by in particularly size of female tracker. Analysis of the polarity of male tracking presented a slightly different picture (light vs. dark bars, Fig. 1, Table 2): males demonstrated a clear capacity to detect the polarity of female mucous trails, typically tracking about three times further in the direction of positive polarity (P < for all categories of females). More intriguing was that males tracked females of their own ecotype in the correct direction (positive polarity) more frequently than they tracked S females positively. Consequently, net distances (positive polarity distance minus negative polarity distance) for males tracking normal E females were higher than when tracking normal S females (Table 2, Fig. 1). Interestingly, this difference persisted when males tracked E and S females of similar size (large E and small S, Table 2, Fig. 1). Thus extent of tracking polarity by males was largely independent of female size, but clearly dependent on female ecotype. Pooling procedures The factor Experiment (nested within Bay) was highly significant in our analyses (Tables 1 and 2) owing to a large variation in general activity among experiments (total tracking distance of each experiment ranged between 356 and 1599 steps, with a mean of 827 and a standard deviation of 352). The reason we experienced different activities during different experiments is unknown to us and could be related to different activities at different times of the day or during different days. The significant factor Experiment is not further interpreted because it was not part of any a priori hypotheses. However, this variation did not affect our interpretation of the main hypothesis as the interaction Female Experiment (Bay) remained nonsignificant (element E in Tables 1 and 2) indicating that the variation in activity among experiments was similar for different groups of females. We pooled the nonsignificant interactions Female Bay and Female Experiment (Bay) with the Residual variation to increase the power of the remaining tests. (The results of both full and pooled tests are presented in Tables 1 and 2). Discussion Male L. saxatilis localizes females (for mating) by following their mucus trails. Our results show that the success of this behavior is dependent upon two factors: (1) the size similarity between the tracking male and the female marker that lays the trail; and (2) the female s ecotype. Cliff ecotype males (E, the subject of these investigations) showed a clear preference for trails laid by snails EVOLUTION DECEMBER

5 Table 2. ANOVA and a priori test (Contrasts) of the effect of female category on the net distance over which males tracked the female trail. Net distance, positive polarity distance minus negative polarity distance. Male net distance was tested against three factors: Female (four levels, E normal and large, S small and normal), Bay (two levels), and Experiment (16 levels). Data were square-root transformed. Results of full and pooled tests are shown. ANOVA Source df Mean square F-value P-value Pooled terms Pooled MS New F New P Female (A) Bay (B) Female Bay (C) C+E+F Expt (Bay) (D) Female Expt (Bay) (E E + F 2.46 Residual (F) Contrasts (a priori test) Comparison Full design P-value Pooled design P-value Normal E vs. normal S Large E vs. small S Dependent: Net distance of male tracking (square root transformed). within the size-range of females of the same ecotype (normal E, large E, small S; light bars, Fig. 1). Secondly, males improved tracking success by determining the direction (polarity) of the trail and following the trail toward the female more often than expected by chance. Intriguingly, the accuracy of this mechanism was greater when tracking females of his own ecotype (normal and large E; dark bars, Fig 1). This provides clear evidence that trail-following behavior can mediate assortative mating between L. saxatilis ecotypes, and according to the experimental results of our study, assortative trail-following effectively halves the likelihood of E males encountering S females when E and S females are equally represented. This reduction is magnified by the fact that copulation time with dislike mates is shorter (Hollander et al. 2005), such that the combination of these prezygotic barriers substantially reduces mating success of E males with S females in comparison to own ecotype females (Fig. 2). From detailed microsatellite analyses of wild populations we know that gene flow between E and S ecotypes is reduced to 18% of gene flow within each ecotype over small distances (20 m) (Panova et al. 2006). Interestingly, this corroborates the quantitative estimates of prezygotic isolation (7 20%, Fig. 2), and suggests that trail-following and copulation time are the essential mechanisms of reproductive barriers in the study populations. Some earlier investigations have obtained results that seem contradictory to our findings, however close examination of the respective experimental designs reveals plausible explanations for this. For example, Erlandsson et al. (1999) found no evidence for trail-discrimination in Spanish ecotypes of L. saxatilis. In that study, however, single males were given the opportunity to follow (or not) only one female trail at any given time. This choice between a trail and no trail, does not permit evaluation of whether males choose between trails of different females. In a second study on mate-searching behavior of Spanish L. saxatilis, Rolán- Alvarez et al. (1999) found that females of different ecotypes were located at random by males, and concluded that there was no precopulatory mate choice. In their experiment they removed size differences between the two Spanish ecotypes (which also differ in average size by a factor of 2, Johannesson et al. 1993) by using equally sized snails. Consequently their results are similar to what we found; when sizes of ecotypes are equal males tend to follow females of different ecotypes equally long distances (large E and small S; light bars, Fig. 1). Identifying the mechanisms that underpin this reproductive barrier is an important first step to understanding how reproductive isolation has evolved in this system. Nevertheless, the explanation for why the assortative mating evolved remains an important question. Evolution of assortative mating is often linked to inferior hybrids, as prezygotic barriers are expected to be reinforced by selection in such situations (Butlin 1995). However, hybrids between E and S ecotypes have higher survival rate in the hybrid zone than both parental forms (Janson 1983) and produce intermediate numbers of eggs and embryos of approximately similar quality to the parental ecotypes (Janson 1985). Consequently, a direct selection pressure for assortative mating between E and S ecotypes seems to be missing. Still the deviation from random mating that we observe is substantial with a > 80% reduction in gene flow (Panova et al. 2006; Fig. 2) EVOLUTION DECEMBER 2008

6 Figure 2. Simplified model comparing the mating success of E males mating own or alien ecotype females. Mating success of crossecotype matings is standardized against success of mating own females. The comparison of normal E females versus normal S females is underlined, and the comparison of large E females and small S females (these females being of the same size) is in italics. Bold figures indicate relative gene flow within and between ecotypes. 1 Experimental data from Hollander et al. (2005). 2 Nm (migration) values based on microsatellite difference (F ST ) estimated by Panova et al. (2006). An alternative explanation is that this reproductive barrier is a secondary consequence of ecological differentiation and specialization of the two ecotypes to their different ecological niches (cliffs and boulders) with large inherited differences in size, shape, and behaviors (Johannesson and Johannesson 1996). Size, for example, is the single most important morphological factor and it is under strong divergent selection. In cliff habitats small individuals are favored as they can hide in crevices to protect from heavy wave action, whereas large individuals resist crab attacks much better in boulder habitats (e.g., Johannesson 1986). Too large differences in size between mates might impose mechanical constraints on mating and this, rather than reinforcement selection, might be the mechanism that impede random mating where cliff and boulder ecotype snails overlap in distribution. Such an isolation mechanism based on ecotype size differences can explain some of our observations. Nonetheless, we found also size-independent discrimination of mates that cannot be logically explained in this way. Indeed, size-independent mechanisms are indicated at two different stages of the mating procedure: polarity of trail-tracking (this study), and variation in copulation times (see fig 7 in Hollander et al. 2005, and table 5 in Rolán-Alvarez et al for a similar result in Spanish ecotypes). Both these observations are intriguing and cannot be explained by a mechanism related to general size differences between ecotype populations. Possibly future investigations that unveil the details of how males recognize the polarity of a trail might be useful to shed light on the size-independent components of assortative mating between the two Swedish ecotypes of L. saxatilis. ACKNOWLEDGMENTS This research was funded through grants from the Swedish Research Council to KJ (contract ), PJ (contract ), and JH (contract ). LITERATURE CITED Barluenga, M., K. N. Stolting, W. Salzburger, M. Muschick, and A. Meyer Sympatric speciation in Nicaraguan crater lake cichlid fish. Nature 439: Butlin, R. K Reinforcement an idea evolving. Trends Ecol. Evol. 10: Davies, M. S., and J. Blackwell Energy saving through trail following in a marine snail. Proc. R. Soc. Lond. B 274: Doebeli, M., and U. Dieckmann Speciation along environmental gradients. Nature 421: Erlandsson, J Do reproductive strategy and breeding season influence the presence of mate recognition in the intertidal snail Littorina? Invert. Reprod. Develop. 41: Erlandsson, J., and V. Kostylev 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: Erlandsson, J. and E. Rolán-Alvarez Sexual selection and assortative mating by size and their roles in the maintenance of a polymorphism in Swedish Littorina saxatilis populations. Hydrobiologia 378: Erlandsson, J., V. Kostylev, and E. Rolán-Alvarez Mate search and aggregation behaviour in the Galician hybrid zone of Littorina saxatilis. J. Evo. Biol. 12: Gavrilets, S., and A. Vose Case studies and mathematical models of ecological speciation. 2. Palms on an oceanic island. Mol. Ecol. 16: Gavrilets, S., A. Vose, M. Barluenga, W. Salzbuger, and A. Meyer Case studies and mathematical models of ecological speciation. 1. Cichlids in a crater lake. Mol. Ecol. 16: Grahame, J., C. S. Wilding, and R. K. Butlin Adaptation to a steep environmental gradient and an associated barrier to gene exchange in Littorina saxatilis. Evolution 60: Hollander, J., M. Lindegarth, and K. Johannesson Local adaptation but not geographic separation promotes assortative mating in a snail support for ecological speciation. Anim. Behav. 70: Janson, K Phenotypic differentiation in Littorina saxatilis Olivi (Mollusca Prosobranchia) in a small area on the Swedish west coast. J. Moll. Stud. 48: Selection and migration in two distinct phenotypes of Littorina saxatilis in Sweden. Oecologia 59: Variation in the occurrence of abnormal embryos in females of the intertidal gastropod Littorina saxatilis Olivi. J. Moll. Stud. 51: EVOLUTION DECEMBER

7 Allozyme and shell variation in two marine snails (Littorina, Prosobranchia) with different dispersal abilities. Biol. J. Linn. Soc. 30: Janson, K., and P. Sundberg Multivariate morphometric analysis of two varieties of Littorina saxatilis from the Swedish west coast. Mar. Biol. 74: Janson, K., and R. D. Ward Microgeographic variation in allozyme and shell characters in Littorina saxatilis Olivi (Prosobranchia: Littorinidae). Biol. J. Linn. Soc. 22: Johannesson, B Shell morphology of Littorina saxatilis Olivi the relative importance of physical factors and predation. J. Exp. Mar. Biol. Ecol. 102: Johannesson, B., and K. Johannesson Population differences in behaviour and morphology in Littorina saxatilis: Phenotypic plasticity or genetic differentiation? J. Zool. 240: Johannesson, K Inverting the null-hypothesis of speciation: a marine snail perspective. Evol. Ecol. [Epub ahead of print] doi: /s Johannesson, K., and A. Tatarenkov Allozyme variation in a snail (Littorina saxatilis) deconfounding the effects of microhabitat and gene flow. Evolution 51: Johannesson, K., B. Johannesson, and E. Rolán-Alvarez Morphological differentiation and genetic cohesiveness over a microenvironmental gradient in the marine snail Littorina saxatilis. Evolution 47: Johannesson, K., E. Rolán-Alvarez, and A. Ekendahl Incipient reproductive isolation between two sympatric morphs of the intertidal snail Littorina saxatilis. Evolution 49: Johannesson, K, J. Lundberg, C. André, and P. Nilsson Island isolation and habitat heterogeneity correlate with DNA variation in a marine snail (Littorina saxatilis). Biol. J. Linn. Soc. 82: Panova, M., J. Hollander, and K. Johannesson Site-specific genetic divergence in parallel hybrid zones suggests nonallopatric evolution of reproductive barriers. Mol. Ecol. 15: Pickles A. R., and J. Grahame Mate choice in divergent morphs of the gastropod mollusc Littorina saxatilis (Olivi): speciation in action? Anim. Behav. 58: Rolán-Alvarez, E., J. Erlandsson, K. Johannesson, and R. Cruz Mechanisms of incomplete prezygotic reproductive isolation in an intertidal snail; testing behavioural models in wild populations. J. Evol. Biol. 12: Rolán-Alvarez, E., M. Carballo, J. Galindo, P. Moran, B. Fernández, A. Caballero, R. Cruz, E. G. Boulding, and K. Johannesson Nonallopatric and parallel origin of local reproductive barriers between two snail ecotypes. Mol. Ecol. 13: Saur, M Mate discrimination in Littorina littorea (L.) and L. saxatilis (Olivi) (Mollusca:Prosobranchia). Hydrobiologia 193: Savolainen, V., M. C. Anstett, C. Lexer, I. Hutton, J. J. Clarkson, M. V. Norup, M. P. Powell, D. Springate, N. Salamin, and W. J. Baker Sympatric speciation in palms on an oceanic island. Nature 441: Associate Editor: N. Wedell 3184 EVOLUTION DECEMBER 2008