Printer Friendly

The influence of self-fertilization and population dynamics on the genetic structure of subdivided populations: a case study using microsatellite markers in the freshwater snail Bulinus truncatus.

The distribution of neutral genetic variation in a group of populations is generally viewed as a balance between genetic drift and gene flow (Wright 1978; Slatkin 1985; Barton and Whitlock 1996). Several models, such as the well-known island models, have been built to analyze this balance. They usually assume that the local populations are at equilibrium between migration and genetic drift. However, the long-term persistence of populations is not realistic in numerous ecological situations (e.g., habitat instability). Extinction and recolonization of local populations therefore appear as additional major processes. This gave rise to the metapopulation concept (sensu Levins 1968). The genetic consequences of these processes depend on several factors, such as the origin and number of colonists and the migration rate or the extinction rate (Slatkin 1985; Wade and McCauley 1988). These aspects have recently received both theoretical and empirical interest (reviews in Slatkin 1985; Olivieri et al. 1990; McCauley 1993; Barton and Whitlock 1996; Harrison and Hastings 1996). Less well-understood is the role of the mating system. Maruyama and Tachida (1992) used the island model to show that the distribution of genetic variability varies with the selfing rate. This is because both the effective size of local populations, and the effective migration rate when migration takes place at the diploid stage (Jarne 1995) are lowered in selfing populations. The mating system may also affect the colonizing potential, and therefore the variability (see Schoen et al. 1996). The neutral variability is also expected to be affected by selection at other loci, through processes such as background selection, which both decreases the variability within populations (Charlesworth et al. 1993) and increases the differentiation among populations (Charlesworth et al., in press). These effects are especially strong in selfing populations, because of limited recombination.

In practice, analyses of geographic patterns of neutral gene distribution used F-statistics (Wright 1978) and allozymes (references in Avise 1994). However, some species are so depauperate in allozyme polymorphism that these markers cannot be used, even at a large geographic scale (Murphy et al. 1996). Microsatellite markers may be especially useful in such studies, because they have been shown to be extremely variable, even in species deprived of allozyme polymorphism (Queller et al. 1993; Viard et al. 1996). They also exhibit other desired qualities, that is, neutrality, Mendelian inheritance, and codominance. However, microsatellites probably evolve under peculiar models of mutation, and exhibit high mutation rates (Charlesworth et al. 1994; Jarne and Lagoda 1996). Whether this may influence our perception of population functioning has been only partially analyzed (Goldstein et al. 1995; Slatkin 1995b; Rousset 1996).

As the metapopulation concept deals with discrete, patchily distributed populations, freshwater organisms occupying discretely distributed habitats, are a priori good examples of its potential applications. Tropical freshwater snails present some further advantages. First, annual cycles of drought and flood, especially in the Sahelian area, limit the persistence of habitats (Brown 1994). Populations are therefore subject to periodic size variation on a short time scale, and are expected to experience genetic drift. At a larger time scale, genetic drift is also expected, in connection with very severe reductions in effective size (population crashes) and even extinction, if the latter is followed by recolonization of a few individuals (founder event). Second, these snails are hermaphroditic, and most species retain the capacity for self-fertilization (Jarne and Stadler 1995). The influence of the mating system on genetic variability must therefore be considered.

We focus here on the population genetics of the tropical freshwater snail Bulinus truncatus analyzed from the within-pool scale (few hundreds of meters) to the whole species distribution area (several thousands of kilometers; from the Sahelian area to Mediterranean islands). We analyze the variability of four polymorphic microsatellite loci in 38 natural populations. This largely extends previous studies (Viard et al. 1996, 1997b), which were restricted to some populations from Niger. The extensive polymorphism uncovered allow us to analyze the forces acting within and among populations (self-fertilization and strong variation in population size, migration, and colonization), taking advantage of the fact that the natural populations considered experience various environmental conditions (different types of habitats and regions). We also show through a slight extension of the theory that high selfing rates depress the variability within populations under stepwise models of mutation. Differences in population differentiation across loci are discussed in terms of the relationship between homoplasy and the mutation rate. We finally consider the relevance of our results for the evolution of phally polymorphism.


Species and Populations Studied

The freshwater snail B. truncatus (Gastropoda, Pulmonata) is one of the major vectors of various species of Schistosoma, agents of bilharziasis in humans and cattle in Africa (Brown 1994). Bulinus truncatus is an allotetraploid species distributed over most of Africa, several Mediterranean islands, and part of the Middle East. Its habitats are submitted to annual variation in water availability, which imposes wide fluctuations in snail density (Madsen et al. 1988; Brown 1994; Vera et al. 1995) and reduce the genetic variability through genetic drift (Viard et al. 1996). This is especially true for natural pools in the Sahelian area, characterized by annual cycles of drought and flood, but also for habitats in North Africa and the Mediterranean islands, where B. truncatus generally occupies irrigation systems or oueds (temporary rivers).

This species has been intensively studied with regard to its reproductive system. It is a simultaneous hermaphrodite characterized by the occurrence of two sexual morphs: regular hermaphroditic individuals, referred to as euphallic individuals, coexist with aphallic individuals, which are deprived of the male copulatory organ. Euphallic snails can self-fertilize and cross-fertilize as both male and female, while aphallic snails can only self-fertilize and cross-fertilize as female (Larambergue 1939). The aphally ratio (frequency of aphallic individuals) is highly variable among natural populations (Schrag et al. 1994a,b; Doums et al. 1996b). Aphally influences the selfing rate, since selfing is obligatory in strictly aphallic populations. Laboratory-crossing experiments (Doums et al. 1996c), progeny-array analyzes (Viard et al. 1997b), and inbreeding coefficient computations (Viard et al. 1996) indeed strongly suggested that B. truncatus is a predominantly selfing species. Aphally is determined by both environmental and genetic factors (Larambergue 1939; Schrag et al. 1994b; Doums et al. 1996a).

Thirty-eight populations were sampled between 1992 and 1995 over a large part of the distribution area of B. truncatus ([ILLUSTRATION FOR FIGURE 1 OMITTED], Table 1). The habitat was classified according to water availability as permanent (water always available), semipermanent (water available more than six months per year), or temporary (water available up to six months). In most samples, snails were checked for their sexual morph and the aphally ratio was estimated. Besides covering different aphally ratios and types of habitat (Table 1), our sampling also allowed us to contrast variation at different geographic levels, that is, among different, characteristic, geographic areas (Mediterranean islands, north and south of the Sahara), among the different samples within these areas, and within pools (large pools in Niger covering several square kilometers; see Viard et al. 1996, 1997b) or rivers.

Analysis of Microsatellite Loci

DNA extraction was performed according to Jarne et al. (1992). Individual genotypes were assessed using four microsatellite loci (BT1, BT6, BT12, and BT13; see Viard et al. 1996). They were chosen because of their high polymorphism and unambiguous amplification patterns. BT1 and BT6 on one side, and BT12 and BT13 on the other side, are dinucleotide and tetranucleotide repeat loci, respectively. The Mendelian inheritance of alleles has previously been demonstrated [TABULAR DATA FOR TABLE 1 OMITTED] and the four loci were analyzed according to the protocol described in Viard et al. (1996), and slightly modified in Viard et al. (1997b).

Statistical Analyzes

For each population, we estimated the allelic frequencies, the mean number of alleles ([n.sub.all]), the observed heterozygosity ([H.sub.o]) and the gene diversity ([H.sub.e]). Tests for deviation from Hardy-Weinberg expectations at each locus and for genotypic linkage disequilibria among polymorphic loci were computed within each population with the software Genepop 2.0 (Raymond and Rousset 1995b). The estimator [Mathematical Expression Omitted] of [] was computed according to Weir and Cockerham (1984) using Genypop 2.0. We also estimated the selfing rate S using the classical relationship [] = S/(2 - S), since this relationship does not depend on the mutation model of the markers used (Rousset 1996).

We analyzed the relationships among the aphally ratio, the genetic parameters and the geographic coordinates using nonparametric tests of comparison for qualitative characters (type of habitat) and regression analysis for proportion data (arcsine transformed before analysis).

The genetic structure among populations was analyzed by testing for differentiation among populations (Raymond and Rousset 1995a) and computing the estimator [Mathematical Expression Omitted] of [F.sub.ST] according to Weir and Cockerham (1984) using Genepop 2.0. Microsatellite loci probably evolve under a stepwise mutation model (SMM), although an infinite alleles model (IAM) cannot always be rejected (review in Jarne and Lagoda 1996). The estimator [Mathematical Expression Omitted] of [F.sub.ST] is an intraclass correlation coefficient for the probability of identity by descent under the IAM, and for the probability of identity in state under the SMM (Rousset 1996). We also tested for isolation by distance (Slatkin 1993) at various geographical levels analyzing the independence between geographic and genetic distances. Geographic distances among populations were calculated according to the formula

[Mathematical Expression Omitted], (1)

where R is the radius of the earth, [[Phi].sub.A] and [[Phi].sub.B] are the latitudes (in radians) and [[Theta].sub.A] and [[Theta].sub.B] the longitudes (in radians; positive and negative for the eastern and the western longitudes, respectively) of two populations A and B in the northern hemisphere. Genetic distances were pairwise [Mathematical Expression Omitted]-values. The null hypothesis of independence between geographic and genetic distances was tested against the hypothesis of a positive correlation expected under isolation by distance, estimated as Spearman's rank correlation coefficient. The observed correlation coefficient was compared to the distribution of correlation coefficients obtained from Mantel-like permutations (5000 permutations) of the genetic and geographic distance matrices (Genepop 2.0). This was performed over all populations and within each geographic area, namely the Mediterranean islands (Sardinia and Corsica), North Africa (Morocco and Algeria), West Africa (Senegal), and the Sahelian area (Niger) separately. The populations from Ivory Coast and Burkina Faso were not analyzed here as their sample size was too low, and individuals from Burkina Faso were analyzed at three loci only.

Hierarchical analyzes of population structure were also conducted in two ways on a subset of 22 populations. We considered seven populations only in Niger (namely populations 12, 15, 17, 19, 23, 26, 30; cf. [ILLUSTRATION FOR FIGURE 1 OMITTED] and Table 1), both to have the same range of geographic distances across the different geographic areas (the populations sampled in Niger are the only ones so closely spaced) and because the program used (see below) was limited in the number of alleles. First, the hierarchical analysis was performed by partitioning [Mathematical Expression Omitted] into [Mathematical Expression Omitted] and [Mathematical Expression Omitted], the correlation for pairs of alleles among individuals within populations and among populations within regions (Weir and Cockerham 1984). Analyzes were performed on five different areas, namely the Mediterranean islands, North Africa, north of the Sahara (Mediterranean islands and North Africa), south of the Sahara (Niger and Senegal), and north versus south of the Sahara. The calculations were made using a program written in Turbo Pascal, kindly provided by Y. Michalakis. Second, we analyzed the variation of [Mathematical Expression Omitted] and [Mathematical Expression Omitted] while sequentially pooling different samples using the graphical method of Goudet et al. (1994) using Fstat 1.2 (Goudet 1995). Four levels were analyzed. At level 0, all 22 populations were considered separately. At level 1, populations from each of the following areas: Corsica, Sardinia, Algeria, Morocco, Niger, and Senegal, were considered together. We then contrasted at level 2 larger geographic areas corresponding to large physical and/or historical barriers, namely the Mediterranean islands, North Africa, and south of the Sahara. At level 3, populations from the Mediterranean islands and North Africa were pooled together and contrasted with populations from south of the Sahara.

Our analysis often implied replicated independent tests, some of which may be significant by chance alone. We thus used Fisher's method for combining independent results (Sokal and Rohlf 1995). A significant combined probability means that the null hypothesis is violated in at least one of the tests performed.


Microsatellites and Within-Population Polymorphism

Unambiguous patterns of PCR amplification were generally observed despite the allotetraploidy of B. truncatus. An exception is the population of Dyoro in which four bands were observed for each individual at locus BT13, and clear segregation patterns could not be determined. These data were therefore discarded from the present analyzes.

A large number of alleles was observed over the 38 populations analyzed (1210 individuals) with 6, 15, 48, and 55 alleles at loci BT1, BT6, BT12, and BT13, respectively. The mean number of alleles per population (size range: size in base pairs of the shortest and largest alleles) was 1.58 (176188), 2.16 (116-167), 6.24 (206-406), and 6.81 (244-436) at loci BT1, BT6, BT12, and BT13 respectively. The mean gene diversity over populations were 0.15, 0.25, 0.51, and 0.54 (loci ordered as above).

The within-population polymorphism was highly variable among populations. Indeed, four populations, namely KEK, Arbatache, NPK, and Natio, were monomorphic at all loci. In contrast, some populations exhibited a huge number of alleles (e.g., 19 and 17 alleles were observed in Lampsat at loci BT12 and BTI 3, respectively). The mean number of alleles per population is given in Table 2. Gene diversity varied from 0 to 0.76 across populations, but was generally high (above 0.5). Observed heterozygosity was always low, varying from 0 to 0.24 (Table 2). The number of alleles and gene diversity were positively correlated (r = 0.89, n = 36, P [less than] [10.sup.-3]).

The four loci usually showed highly significant deviations (P [less than] [10.sup.-5]) from genotypic proportions expected under Hardy-Weinberg equilibrium. Nonsignificant deviations were observed only in Boundoum (locus BT1, P = 0.27), RD4RD (BT12, P = 1) and Fint (BT13, P = 0.13). Combined probability [TABULAR DATA FOR TABLE 2 OMITTED] tests over all loci showed highly significant departure (P [less than] [10.sup.-5]) for all populations, except Fint (P = 0.13). [Mathematical Expression Omitted]-values over all loci, and selfing rates estimated from [Mathematical Expression Omitted]-values, were usually very large (Table 2), except in Fire and Sidi Chikh, for which one locus only was polymorphic.

Exact tests for genotypic linkage disequilibria resulted in a large number of significant values (P [less than] 0.05) for each pair of loci per population (44 of 114). No pairs of loci appeared in linkage disequilibrium in the populations Djanet, RD4RD, Budoni, Lotzorai, Orbo, BoyzeI, Doubalma, Bala, Kobouri, Niumpalma, and Namaga PM. In BoyzeII, Kokourou, and Kotaki, the independence was rejected for each pair of loci when the test was possible (i.e., polymorphic loci). The combined P-values for each pair of loci across all populations were extremely low (P [less than] [10.sup.-5]).

Geographic Origin, Type of Habitat and Aphally

No correlation was found between the type of habitat and either the aphally ratio (Kruskal-Wallis test, H = 3.73, n = 36, P = 0.15), the observed heterozygosity (H = 4.27, n = 36, P = 0.12), or and the [Mathematical Expression Omitted]-values (H = 1.43, n = 34, P = 0.49). A slightly significant correlation was observed with both the number of alleles (H = 7.56, n = 36, P = 0.02) and the gene diversity (H = 6.53, n = 36, P = 0.04). More alleles and greater gene diversity values were observed in semipermanent and permanent habitats than in temporary habitats. Similar results were obtained on a locus per locus basis. No correlation was detected between the aphally ratio and either the number of alleles (regression analysis, r = 0.11, n = 36, P = 0.52), the observed heterozygosity (r = 0.05, n = 36, P = 0.76), the gene diversity (r = 0.07, n = 36, P = 0.67), and the [Mathematical Expression Omitted]-values (r = 0.04, n = 32, P = 0.82). These results also hold when analyzing each locus separately. Significant correlations were observed between the latitude and either the number of alleles (r = -0.40, n = 38, P = 0.01), the observed heterozygosity (r = -0.41, n = 38, P [less than] [10.sup.-3]), or the gene diversity (r = -0.38, n = 38, P = 0.02), though not with the [Mathematical Expression Omitted]-values (r = 0.10, n = 34, P = 0.58). Similar results were obtained when analysing each locus separately, except for the observed heterozygosity at loci BT1 [TABULAR DATA FOR TABLE 3 OMITTED] and BT6. Significant correlations were also obtained between the longitude and either the number of alleles (r = -0.39, n = 38, P = 0.02), or the [Mathematical Expression Omitted]-values (r = 0.40, n = 34, P = 0.02), though not with the observed heterozygosity (r = 0.27, n = 38, P = 0.10) or the gene diversity (r = 0.14, n = 38, P = 0.41). Similar correlations were obtained when analysing BT12 and BT13 separately, though not always for BT1 or BT6.

Among-Population Structure

The overall differentiation among populations was highly significant at the four loci,' as well as for the combined test over loci (P [less than] [10.sup.-5]), and corresponded to a mean [Mathematical Expression Omitted]-value of 0.54 (Table 3). In contrast to the [Mathematical Expression Omitted] and [Mathematical Expression Omitted]-values, the [Mathematical Expression Omitted]-values varied among loci (Table 3). The two tetranucleotide loci (BT12 and BT13) showed very similar [Mathematical Expression Omitted]-values, but these were much lower than those obtained at the two dinucleotide loci. We also analyzed the structure between each pair of populations using exact tests, and found nonsignificant P-values (i.e., no structure) in a few cases only (55 of 565, 7 of 688, 4 of 701, and 1 of 628 at the 1% level at loci BT1, BT6, BT12, and BT13, respectively). The associated [Mathematical Expression Omitted]-values across loci ranged from 0 (e.g., between Mari Sud and Mari Nord) to 1 (e.g., between NPK and Arbatache).

The Mantel-like test for independence between geographic and genetic distances over all populations and loci was significant [TABULAR DATA FOR TABLE 4 OMITTED] (P = 0.001). Significant values were also obtained when the two dinucleotide loci (P = 0.006) and the two tetranucleotide loci (P = 0.011) were taken into account separately. No pattern of isolation by distance was found when considering the populations from the Mediterranean islands only (P = 0.10), nor for the North African populations (P = 0.54). A slight though significant pattern was found within Niger (P = 0.05) and Senegal (P = 0.04).

The hierarchical analysis of the distribution of variability showed that the populations of B. truncatus are subdivided at all levels (Table 4). Any barrier among groups of populations (e.g., sea, desert) has a significant effect, as the among-population within-region component is always far greater than the among-region component. The results of the hierarchical analysis of genetic subdivision by sequentially grouping the samples are given in Figure 2. Similar curves were observed at each locus and over all loci. Exact tests of population differentiation at each locus and combined tests over all loci were always significant (P [less than] [10.sup.-5]), whatever the level of grouping.


Three important results will be discussed in turn: (1) the high polymorphism observed within highly selfing populations experiencing population size variation, population crashes and/or colonizing events; (2) the relative influence on population structure of genetic drift and gene flow at various hierarchical levels and in different regions; and (3) the effect of the mutation rate and homoplasy on population structure.

Microsatellite Polymorphism, Selfing, and Population Dynamics

We observed a high polymorphism over the whole dataset. This is especially true for the two tetranucleotide loci BT12 and BT13, in line with previous results (Viard et al. 1996). Our results contrast with previous analyzes describing the natural populations of B. truncatus as poorly polymorphic, based on allozymes. Njiokou et al. (1993) studying 17 populations from Niger and Ivory Coast found a mean number of alleles per locus between 1 and 1.1, and observed no heterozygotes. Similar results were obtained for the three populations from Algeria studied here (Hamza 1992). That microsatellite markers are far more polymorphic than allozyme markers is due to mutation rates two or three orders of magnitude higher than those of allozymes (Amos et al. 1996; Primmer et al. 1996 and references therein). However, microsatellite markers usually exhibit higher gene diversity values than observed here, even when populations are expected to be deprived of genetic variability, as for example in some social insects ([H.sub.e] = 0.9, Evans 1993) or in fragmented habitats ([H.sub.e] = 0.8, Spencer et al. 1995).

Almost all populations were characterized by a low observed heterozygosity. The highest value was 0.24 in Boundoum and eight populations did not exhibit any heterozygote. Numerous genotypic disequilibria were also observed. Only 11 of 38 populations did not exhibit disequilibria between any pair of loci. Both the heterozygote deficiencies and genotypic disequilibria may be considered as consequences of selfing. Selfing decreases the proportion of heterozygotes, the rate of decay of gametic disequilibrium, and the effective recombination rate (Hedrick 1980), homogenizing genotypes within lines and creating familial structures (Allard 1975). That selfing rate estimates were always higher than 80% (except in two populations) suggests that B. truncatus is a preferential selfer, confirming previous results (Viard et al. 1996, 1997b). Selfing must therefore be considered as a major force molding the genetic structure of this species, as previously shown in other hermaphroditic organisms (Schoen and Brown 1991; Jarne 1995; Jarne and Stadler 1995). We also showed that despite the high mutation rates at microsatellite loci, the gene diversity values were sometimes low (e.g., KEK, Djanet). This also may be due to the mating system, since selfing populations are expected to maintain less neutral genetic polymorphism than randomly mating populations for three main reasons, including a lower effective population size (Pollak 1987), genetic hitchhiking (Hedrick 1980), and background selection against deleterious alleles (Charlesworth et al. 1993). These predictions have been derived under an infinite alleles (sites) model (IAM) of molecular evolution. However, microsatellites are thought to fit the SMM or the TPM rather than the IAM (review in Freimer and Slatkin 1996; Jarne and Lagoda 1996). 'We show in the Appendix how selfing depresses the genetic variability of a population under the SMM or the TPM, as a consequence of reduced effective population size, genetic hitchhiking and background selection. However there is currently no straightforward method available to distinguish between the three processes in selfing populations, and the comparative approach (see Schoen and Brown 1991; Jarne 1995) cannot be used because very few microsatellite data are available.

High selfing rates cannot be the only source of the loss of genetic polymorphism. All populations indeed exhibited very high selfing rates, though with very different levels of variability. This may be a consequence of the population dynamics. Three processes may explain these results: (1) low, and/or cyclic variation in, population size; (2) population crashes and/or extinction, followed by recolonization by a few individuals in a metapopulation model; and (3) founder events following the recent colonization of previously unoccupied areas. These three points will be examined in turn.

Small populations maintain by definition less variability than larger ones (Crow and Kimura 1970). Low densities and scattered populations were indeed observed in Corsica, Sardinia, and Morocco (pers. obs.) compared to high densities in the ponds from Niger studied here (Doums et al. 1997). However these observations provide only a rough estimate of actual population sizes, and cannot be related to effective population sizes. Cyclic variation in water availability in the Sahelian area throughout the year should also be considered, as it has been related to variation in population size in a large number of studies (see Brown 1994), especially in the temporary habitats occupied by B. truncatus in Niger (Vera et al. 1995). This cyclic variation in population size decreases the effective population size (Crow and Kimura 1970), and subsequently leads to a reduction in neutral allele diversity. We showed elsewhere (Viard et al. 1997b) that the annual variation in surface of the pools in Niger correlates well with the genetic pattern observed on a short (one-year) time scale. In the present study, some genetic parameters were significantly correlated with the type of habitat (e.g., number of alleles), although the associated probability values were not extremely low. These equivocal results may be due to contrasted patterns within a given type of habitat. For example, some permanent habitats exhibited a substantial genetic variability (e.g., Senegal) whereas others appeared poorly polymorphic (e.g., Boyze). It is likely that our measure (water availability) of the periodic variation in the prevailing environmental conditions is not accurate. Studies of population dynamics would be required to ascertain these points.

The contrasted patterns may result from population crashes and extinction/recolonisation processes. Severe reductions in population size decrease both the number of alleles and gene diversity, as mainly a function of the size of the population during the bottleneck and its growth rate after the bottleneck (Nei et al. 1975; Cornuet and Luikart 1996 and references therein). We have no evidence of a recent major crash in any of the populations studied. However the water level of the environment considered here, especially in the Sahelian area, fluctuates in time, provoking local extinctions. Dried-out zones are recolonized from adjacent areas during the rainy season, as suggested by empirical studies in freshwater snails species (Woolhouse 1988), especially B. truncatus (Betterton 1984), and by isolation by distance in Niger (see next section). A temporal survey in this country did not indicate a severe loss of variability associated with this process (Viard et al. 1997b), even if the time scale considered (a year) was short. These and the present results suggest that recolonization, subsequent to population crashes or extinction, acts as a source of gene flow maintaining a high level of genetic diversity and counteracting the influence of genetic drift and selfing (Slatkin 1985; McCauley 1993). At a larger scale, a pond (or even an oued) may completely disappear in years of exceptional drought. Ecological observations in Algeria fit this scenario, with a recent appearance of B. truncatus in KEK and a low number of founding individuals (Hamza 1992). This is consistent with the low genetic diversity of this population.

The differences may be interpreted as a consequence of the recent colonization of new areas by low numbers of colonizers. Populations from south of the Sahara (Niger, Burkina Faso, and Senegal) appear far more polymorphic than those from north of the Sahara (North Africa and Mediterranean Islands). These correlations might be indicative of successive colonizations from the south to the north of the actual distribution area. Bulinus truncatus probably originates from East or West Africa, though this is poorly documented (Brown 1994). As the Sahara constitutes a strong barrier, the colonization of North Africa and the Mediterranean islands may be due to few individuals, and include successive founding events. This is consistent with the pattern of differentiation (see next section).

Scenarios (2) and (3) include recent bottlenecks. Testing for their occurrence is in principle possible through a comparison of observed and expected polymorphism under various mutation models (Tajima 1989; Cornuet and Luikart 1996). However these tests are designed for panmictic populations, and no test is currently available for partially selfing populations.

Among-Population Structure

A large differentiation was observed over the whole study and at each level of sampling (within local habitats, countries, or regions), except for some semipermanent ponds in Niger (Mari and Namaga). That [Mathematical Expression Omitted] increased and [Mathematical Expression Omitted] decreased when pooling populations (from level 0 to level 3; cf. [ILLUSTRATION FOR FIGURE 2 OMITTED]) suggests that each level represents a different breeding unit and that isolation by distance occurs (Goudet et al. 1994).

We will first analyze the situation at a short geographic scale. Both a weak genetic structure and isolation by distance were detected in Senegal. A linear stepping-stone model may be best suited to describe this situation, although exchanges between neighboring populations are probably not symmetric but rather follow the water current. A weak genetic structure was also observed in natural temporary and semipermanent ponds in Niger, but may be best described using an island model. These ponds are indeed characterized by annual cycles of contraction and expansion. Colonization of the upper parts of the ponds at the beginning of the rainy season occurs from the lower parts with an annual redistribution of variability among local populations (Viard et al. 1996, 1997b).

Over a larger scale (among habitats within the different areas), a marked structure was detected north of the Sahara (especially in Corsica, Sardinia, and Algeria) in contrast to the pattern observed in Niger (Viard et al. 1996). This indicates that exchanges between local populations are limited. At an even larger scale, the hierarchical analysis of the genetic variability suggests that strong barriers prevent migration, even at short distances (e.g., between Corsica and Sardinia). A marked genetic structure between Sardinian and Corsican populations has been observed in other organisms (e.g., mosquitoes; Chevillon et al. 1995). The Sahara, just as the Mediterranean Sea, appears as another important natural barriers to gene flow over the distribution area of B. truncatus. This is indicated by marked differences in genetic characteristics for populations from north versus south of the desert (e.g. number of alleles, gene diversity).

The large differentiation observed at all levels of our study may also be related to the high selfing rates. This is not unexpected, as shown by the comparative analyzes conducted in plants (Hamrick and Godt 1989; Schoen and Brown 1991) and animals (Jarne 1995). The reasons why the estimators of population differentiation are higher in selfers than in outcrossers include a lower effective population size (Maruyama and Tachida 1992) and migration rate (Jarne 1995). Various forms of selection (e.g., background selection or balancing selection) are also efficient in increasing population differentiation at neutral markers. The effect is expected to be stronger when recombination is low, as in selfers (Charlesworth et al., in press).

Estimators of Population Differentiation, Mutation Rate, and Homoplasy

A striking result of our study is the important difference between [F.sub.ST] estimates from dinucleotide (0.71) and tetra-nucleotide (0.48) loci (Table 3). This may be due to the mutation process itself. The SMM and TPM generate some homoplasy, while the IAM does not by definition (see Freimer and Slatkin 1996; Jarne and Lagoda 1996). Adding constraints on allele size decreases the number of possible allelic states for both the SMM and the TPM, and therefore increases homoplasy (Goldstein et al. 1995; Garza and Freimer 1996). This may have consequences on the use of microsatellites for the study of genetic differentiation. When comparing two loci with different mutation rates and equal constraints on allele size in a population, nearly all possible allelic states may be represented at the locus with the highest mutation rate, whereas a subset only is represented at the other locus: the higher the mutation rate, the higher the homoplasy and the weaker the genetic structure. The magnitude of this effect depends on the effective size of the populations considered: mutation homogenizes the genetic variability among populations, while genetic drift acts as diversifying factor (Nauta and Weissing 1996). This may explain why the dinucleotide loci exhibited a much larger differentiation than the tetranucleotide loci. A condition is that the mutation rates are smaller at the dinucleotide than at the tetranucleotide loci. Absolute mutation rates are not available in B. truncatus. However we previously showed that the mutation rate at dinucleotide loci is about seven times lower than at tetranucleotide loci in this species (Viard et al. 1996). Whatever the relative difference in [F.sub.ST] estimates among loci, we observed a marked genetic structure and isolation by distance. Thus, even if homoplasy has to be considered when using microsatellite loci, especially among species (e.g., Garza and Freimer 1996), it does not prevent the analysis of the distribution of genetic variability within species, even over a large geographic scale.

The Evolution of Phally Polymorphism

Selfing is obligatory in purely aphallic populations. A correlation between aphally and selfing is thus expected, and the evolution of phally polymorphism might be driven by selective factors playing on the evolution of the selfing rate (see Jarne and Charlesworth 1993; Uyenoyama et al. 1993). However we did not detect a correlation between aphally and selfing over the whole dataset, suggesting that phally polymorphism may be considered as more or less neutral with regard to the selfing rate. Environmental factors, such as the type of habitat, are also not correlated with the aphally ratio. This does not completely falsify the hypothesis that the current environmental conditions have no influence on the evolution of aphally, since our test is rather crude (for a more refined approach, see Schrag et al. 1994a), neither can we reject the role of past environmental conditions. However other factors may have a more marked influence on the evolution of aphally, such as differences in life-history traits between the two sexual morphs (Doums and Jarne 1996), or the genetic stochasticity deriving from the population dynamics. This idea might be tested by comparing the polymorphism at microsatellite loci and for the genetic component determining aphally. That selfing rates are high may also be interpreted as a result of the marked genetic structure observed (Ronfort and Couvet 1995). Both the distribution of selfing and aphally may therefore be influenced by stochastic forces, due to the instability of the habitat and the population history.


Our study indicates that selfing and reduction in population size are major forces enhancing local genetic drift, and subsequently the genetic structure of B. truncatus whatever the geographic scale considered. On the other hand, gene flow through extinction and recolonization processes and founding events are of great importance. The natural populations of B. truncatus may therefore function as a metapopulation the features of which depend on the spatial and temporal scales. At a large scale (continent), marked subdivisions are observed, so that genetic drift may overcome the effects of migration. On a much smaller scale (within habitats), migration and colonization, acting as a source of gene flow, have a more marked effect resulting in a limited genetic structure. The high genetic diversity and the significant genetic structure over the whole area studied suggest that the species has a potential for evolutionary change despite high selfing rates.


The authors are grateful to the whole staff at the O.C.C.G.E. laboratory in Niamey for facilitating their stay in Niger; to P. Bremond, T. Dan Kountche, P. David, B. Delay, C. Doums, H. Escaffre, D. Ibrahim, and A. Islamane for help in collecting snails in Niger; to P. Bremond, J. M. Duplantier and J.-C. Ernoult, F. Hamza, and K. Khallaayoune for providing us molluscs and information from Ivory Coast and Burkina Faso, Senegal, Algeria and Morocco, respectively; to Y. Michalakis for providing a program for hierarchical analysis; to D. Charlesworth for access to an unpublished manuscript; to P. David, C. Doums, and F. Rousset for discussions; to B. Delay for constant support; and to J. Britton-Davidian, P. David, D. Waller, and an anonymous referee for critical reading of the manuscript. This work was supported by CNRS-UMR 5554 (Universite Montpellier II), GREG (94/84) and the Ministere de l'Environnement (EGPN 94019). FV was supported by a fellowship from CNRS. This is contribution No. 97 090 of Institut des Sciences de l'Evolution.


ALLARD, R. W. 1975. The mating system and microevolution. Genetics 79:115-126.

AMOS, W., S. J. SAWCER, R. W. FEAKES, AND D. C. RUBINSZTEIN 1996. Microsatellites show mutational bias and heterozygote instability. Nature Genet. 13:390-391.

AVISE, J. C. 1994. Molecular markers, natural history and evolution. Chapman and Hall, London.

BARTON, N. H., AND M. C. WHITLOCK. 1996. The evolution of metapopulations. Pp. 183-210 in I. Hanski and M. Gilpin, eds. Metapopulation dynamics: ecology, genetics and evolution. Academic Press, New York.

BETTERTON, C. 1984. Spatiotemporal distributional patterns of Bulinus rohlfsi (Clessin), Bulinus forskali (Ehrenberg) and Bulinus senegalensis Muller in newly irrigated areas in northern Nigeria. J. Mollusc. Stud. 50:137-152.

BROWN, D. S. 1994. Freshwater snails of Africa and their medical importance. 2d ed. Taylor and Francis Ltd., London.

CHARLESWORTH, B., M. T. MORGAN, AND D. CHARLESWORTH. 1993. The effect of deleterious mutations on neutral molecular variation. Genetics 134:1289-1303.

CHARLESWORTH, B., P. SNIEGOWSKI, AND W. STEPHAN. 1994. The evolutionary dynamics of repetitive DNA in eukaryotes. Nature 371:215-220.

CHARLESWORTH, B., M. NORDBORG, AND D. CHARLESWORTH. 1997. The effects of local selection, balanced polymorphism and background selection on equilibrium patterns of genetic diversity in subdivided populations. Genet. Res. 68:In press.

CHEVILLON, C., G. ADDIS, M. RAYMOND, AND A. MARCHI. 1995. Population structure in Mediterranean islands and risk of genetic invasion in Culex pipiens L. (Diptera: Culicidae). Biol. J. Linn. Soc. 55:329-343.

CORNUET, J.-M., AND G. LUIKART. 1996. Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:2001-2014.

CROW, J. E, AND M. KIMURA. 1970. An introduction to population genetics theory. Burgess Publ. Co., Minneapolis, MN.

DIRIENZO, A., A. C. PETERSON, J. C. GARZA, A.M. VALDES, M. SLATKIN, AND N. B. FREIMER. 1994. Mutational processes of simple-sequence repeat loci in human populations. Proc. Nat. Acad. Sci. USA 91:3166-3170.

DOUMS, C., AND P. JARNE. 1996. The evolution of phally polymorphism in Bulinus truncatus (Gastropoda, Planorbidae): the cost of male function analyzed through life-history traits and sex allocation. Oecologia 106:464-469.

DOUMS, C., P. BREMOND, B. DELAY, AND P. JARNE. 1996a. The genetical and environmental determination of phally polymorphism in the freshwater snail Bulinus truncatus. Genetics 142: 217-225.

DOUMS, C., R. LABBO, AND P. JARNE. 1996b. Stability and genetic basis of phally polymorphism in natural populations of the self-fertile freshwater snail Bulinus truncatus. Genet. Res. 68:22-33.

DOUMS, C., F. VIARD, A.-F. PERNOT, B. DELAY, AND P. JARNE. 1996C. Inbreeding depression, neutral polymorphism and copulatory behavior in freshwater snails: a self-fertilization syndrome. Evolution 50:1908-1918.

DOUMS, C., F. VIARD, P. DAVID, AND P. JARNE. 1997. Phally status and size in Niger populations of Bulinus truncatus (Gastropoda: Planorbidae). J. Mollusc. Stud. 63:111-115.

EVANS, J. D. 1993. Parentage analyzes in ant colonies using simple sequence repeat loci. Molec. Ecol. 2:393-397.

FREIMER, N. B., AND M. SLATKIN. 1996. Microsatellites: evolution and mutational processes. Pp. 51-72 in D. Chadwick and G. Cardew, eds. Variation in the human genome. Ciba Foundation Symposium 197, Wiley, Chichester, UK.

GARZA, C. J., AND N. B. FREIMER. 1996. Homoplasy for size at microsatellite loci in humans and chimpanzees. Genome Res. 6: 211-217.

GOLDSTEIN, D. B., A. R. LINARES, L. L. CAVALLI-SFORZA, AND M. W. FELDMAN. 1995. An evaluation of genetic distances for use with microsatellite loci. Genetics 139:463-471.

GOUDET, J. 1995. Fstat: a computer program to calculate F-statistics. Vers. 1.2. J. Hered. 86:485-486.

GOUDET, J., T. DE MEEUS, A. J. DAY, AND C. J. GLIDDON. 1994. The different levels of population structuring of the dogwhelk, Nucella lapillus, along the South Devon coast. Pp. 81-95 in A. R. Beaumont, ed. Genetics and evolution of aquatic organisms. Chapman and Hall, London.

HAMRICK, J. L., AND M. J. GODT. 1989. Allozyme diversity in plant species. Pp. 43-63 in A. H. D. Brown, M. T. Clegg, A. L. Kahler, and B. S. Weir, eds. Plant population genetics, breeding, and genetic resources. Sinauer, Sunderland, MA.

HAMZA, F. 1992. Bulinus truncatus (Audouin, 1827) d'Algerie-Recensement de ses biotopes dans le Nord-Est algerois-Structure genetique de ces populations mediterraneennes. Comparaison avec deux populations sahariennes. Universite des Sciences et de la Technologie Houari Boumedienne, Alger, Algeria.

HARRISON, S., AND A. HASTINGS. 1996. Genetic and evolutionary consequences of metapopulation structure. Trends Ecol. Evol. 11:180-183.

HEDRICK, P. W. 1980. Hitchhiking: a comparison of linkage and partial selfing. Genetics 94:791-808.

JARNE, P. 1995. Mating system, bottlenecks and genetic polymorphism in hermaphroditic animals. Genet. Res. 65:193-207.

JARNE, P., AND D. CHARLESWORTH. 1993. The evolution of the selfing rate in functionally hermaphrodite plants and animals. Annu. Rev. Ecol. Syst. 24:441-466.

JARNE, P., AND P. J. L. LAGODA. 1996. Microsatellites, from molecules to populations and back. Trends Ecol. Evol. 11:424-429.

JARNE, P., AND T. STADLER. 1995. Population genetic structure and mating system evolution in freshwater pulmonates. Experientia 51:482-497.

JARNE, P., B. DELAY, C. BELLEC, G. ROIZES, AND G. CUNY. 1992. Analysis of mating systems in the schistosome-vector Bulinus globosus by DNA fingerprinting. Heredity 68:141-146.

LARAMBERGUE, M. DE. 1939. Etude de l'autofecondation chez les gasteropodes pulmones: recherches sur l'aphallie et la fecondation chez Bulinus (Isadora) contortus Michaud. Bull. Biol. France Belgique 73:19-231.

LEVINS, R. 1968. Evolution in changing environments. Princeton Univ. Press, Princeton, NJ.

MADSEN, H., A. A. DAFFALA, K. O. KAROUM, AND F. FRANDSEN. 1988. Distribution of freshwater snails in irrigation schemes in the Sudan. J. Appl. Ecol. 25:853-866.

MARUYAMA, K., AND H. TACHIDA. 1992. Genetic variability and geographical structure in partially selfing populations. Jpn. J. Genet. 67:39-51.

MCCAULEY, D. E. 1993. Evolution in a metapopulations with frequent local extinction and recolonization. Oxf. Surv. Evol. Biol. 9:109-134.

MURPHY, R. W., J. W. SITES JR., D. G. BUTH, AND C. H. HAUFLER. 1996. Proteins: isozyme electrophoresis. Pp. 51-120 in D. M. Hillis, C. Moritz, and B. K. Mable, eds. Molecular systematics. Sinauer, Sunderland, MA.

NAUTA, M. J., AND F. J. WEISSING. 1996. Constraints on allele size at microsatellite loci: implications for genetic differentiation. Genetics 143:1021-1032.

NEI, M., T. MARUYAMA, AND R. CHAKRABORTY. 1975. The bottleneck effect and genetic variability in populations. Evolution 29: 1-10.

NJIOKOU, F., C. BELLEC, P. BERREBI, B. DELAY, AND P. JARNE. 1993. Do self-fertilization and genetic drift promote a very low genetic variability in the allotetraploid Bulinus truncatus (Gastropoda: Planorbidae) populations? Genet. Res. 62:89-100.

OHTA, T., AND M. KIMURA. 1973. A model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a finite population. Genet. Res. 22:201-204.

OLIVIERI, I., D. COUVET, AND P.-H. GOUYON. 1990. The genetics of transient populations: research at the metapopulation level. Trends Ecol. Evol. 5:207-210.

POLLAK, E. 1987. On the theory of partially inbreeding finite populations. I. Partial selfing. Genetics 117:353-360.

PRIMMER, C. R., H. ELLEGREN, N. SAINO, AND P. A. MOLLER. 1996. Directional evolution in germline microsatellite mutations. Nature Genet. 13:391-393.

QUELLER, D. C., J. E. STRASSMANN, AND C.R. HUGHES. 1993. Microsatellites and kinship. Trends Ecol. Evol. 8:285-288.

RAYMOND, M., AND F. ROUSSET. 1995a. An exact test for population differentiation. Evolution 49:1280-1283.

-----. 1995b. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J. Hered. 86:248-249.

RONFORT, J., AND D. COUVET. 1995. A stochastic model of selection on selfing rates in structured populations. Genet. Res. 65:209-222.

ROUSSET, F. 1996. Equilibrium values of measures of population subdivision for stepwise mutation processes. Genetics 142: 1357-1362.

SCHOEN, D. J., AND A. H. D. BROWN. 1991. Intraspecific variation in population gene diversity and effective population size correlates with the mating system in plants. Proc. Nat. Acad. Sci. USA 88:4494-4497.

SCHOEN, D. J., M. T. MORGAN, AND T. BATAILLON. 1996. HOW does self-pollination evolve? Inferences from floral ecology and molecular genetic variation. Phil. Trans. Roy. Soc. Lond., B 351: 1281-1290.

SCHRAG, S. J., A. O. MOOERS, G. T. NDIFON, AND A. F. READ. 1994a. Ecological correlates of male outcrossing ability in a simultaneous hermaphrodite snail. Am. Nat. 143:636-655.

SCHRAG, S. J., G. T. NDIFON, AND A. F. READ. 1994b. Temperature-determined outcrossing ability in wild populations of a simultaneous hermaphrodite snail. Ecology 75:2066-2077.

SLATKIN, M. 1985. Gene flow in natural populations. Annu. Rev. Ecol. Syst. 16:393-430.

-----. 1993. Isolation by distance in equilibrium and non-equilibrium populations. Evolution 47:264-279.

-----. 1995a. Hitchhiking and associative overdominance at a microsatellite locus. Mol. Biol. Evol. 12:473-480.

-----. 1995b. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139:457-462.

SOKAL, R. R., AND F. J. ROHLF. 1995. Biometry. 3d ed. Freeman, New York.

SPENCER, P. B. S., D. M. ODORICO, S. J. JONES, H. D. MARSH, AND D. J. MILLER. 1995. Highly variable microsatellites in isolated colonies of the rock-wallaby (Petrogale assimilis). Mol. Ecol. 4: 523-525.

TAJIMA, F. 1989. The effect of change in population size on DNA polymorphism. Genetics 123:597-601.

UYENOYAMA, M. K., K. E. HOLSINGER, AND D. M. WALLER. 1993. Ecological and genetics factors directing the evolution of self-fertilization. Oxf. Surv. Evol. Biol. 9:327-381.

VALDES, A.M., M. SLATKIN, AND N. B. FREIMER. 1993. Allele frequencies at microsatellite loci: the stepwise mutation model revisited. Genetics 133:737-749.

VERA, C., P. BREMOND, R. LABBO, F. MOUCHET, E. SELLIN, D. BOULANGER, J.-P. POINTIER, B. DELAY, AND B. SELLIN. 1995. Seasonal fluctuations in population densities of Bulinus senegalensis and B. truncatus (Planorbidae) in temporary pools in a focus of Schistosoma haematobium in Niger: implications for control. J. Mollusc. Stud. 61:79-88.

VIARD, F., P. BREMOND, R. LABBO, F. JUSTY, B. DELAY, AND P. JARNE. 1996. Microsatellites and the genetics of highly selfing populations in the freshwater snail Bulinus truncatus. Genetics 142: 1237-1247.

VIARD, F., C. DOUMS, AND P. JARNE. 1997a. Selfing, sexual polymorphism and microsatellites in the hermaphroditic freshwater snail Bulinus truncatus. Proc. Roy. Soc. Lond. B Biol Sci. 264: 39-44.

VIARD, F., F. JUSTY, AND P. JARNE. 1997b. Population dynamics inferred from temporal variation at microsatellite loci in the selfing snail Bulinus truncatus. Genetics 146:973-982.

WADE, M. J., AND D. E. MCCAULEY. 1988. Extinction and recolonization: their effects on the genetic differentiation of local populations. Evolution 42:995-1005.

WEIR, B. S., AND C. C. COCKERHAM. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38:1358-1370.

WOOLHOUSE, M. E. 1988. Passive dispersal of Bulinus globosus. Ann. Trop. Med. Parasitol. 82:315-317.

WRIGHT, S. 1978. Evolution and the genetics of populations. Vol.4, Variability within and among natural populations. Univ. of Chicago Press, Chicago.


Self-fertilization depresses the genetic variability within populations as a consequence of reduced effective population size, genetic hitchhiking, and background selection. This can be shown as follows. In an inbreeding population the effective population size is expected to be [N.sub.out]/(1 + F) with [N.sub.out] the inbreeding effective size of the corresponding randomly mating population and F the inbreeding coefficient at equilibrium (assuming a binomial distribution of successful gametes produced; Pollak 1987). We mentioned in the text that, under a mixed-mating model at genetic equilibrium, F = S/(2 - S) with S the selfing rate. In purely selfing populations (S = 1 and F = 1), the effective size is therefore expected to be divided by two. It can then easily be shown that the gene diversity (formula in Ohta and Kimura 1973) is expected to be halved in selfing populations provided that [N.sub.e]u [much less than] 1, where [N.sub.e] is the effective population size and u the nutation rate. With increasing values of [N.sub.e]u, gene diversity is less than halved. An alternative perspective on the same point is provided by Valdes et al. (1993) and DiRienzo et al. (1994). These authors, based on the coalescent theory, showed that under the SMM and TPM, at equilibrium, the variance of an allelic distribution is:

[Mathematical Expression Omitted] (A1)

where [a.sub.1] and [a.sub.2] are the repeat numbers of two alleles sampled from a population and [[[Sigma].sup.2].sub.m] the variance of the change in allele size as a result of each mutational event. Using this formula, in a selfing population, we obtain:

[Mathematical Expression Omitted]. (A2)

A lower variance of allele size is therefore expected in mixed-mating populations, under the SMM, down to a halving in purely selfing populations.

A second reason for the loss of polymorphism in selfing populations is that selectively advantageous alleles can drive linked neutral alleles to very high frequencies (genetic hitchhiking: Hedrick 1980; Schoen et al. 1996), because of reduced effective recombination. Estimating quantitatively the influence of hitchhiking is difficult. However, its role can be understood qualitatively following Slatkin (1995a). This author considered the extreme case of strong selection and complete linkage, that is a hitchhiking event after which "only the allele at the microsatellite locus that was initially linked to the advantageous allele at the selected locus remains in the population." The variance in allele size under a TPM (see formula above) will then evolve under the conflicting influences of mutation and genetic drift, such that the variance at generation t after the hitchhiking event is the variance at equilibrium multiplied by 1 - [1 - [(1/2[N.sub.e]).sup.t]]. Variability at microsatellite loci is expected to be restored much more quickly after a hitchhiking event that at loci with lower mutation rates (Slatkin 1995a), though the time required to fully recover the equilibrium variability does not depend on the mutation rate.

Polymorphism may also be lost in selfing populations because of background selection against deleterious mutations (Charlesworth et al. 1993). Background selection plays because it reduces the coalescence time of those neutral alleles going to fixation, proportionally to the frequency of gametes bearing the lowest number of mutations at the selected loci. Slatkin (1995a) suggested that it should work with microsatellites, when selected and neutral alleles are physically linked. We further suggest that it should also work under high selfing rates because selfing reduces the opportunity of recombination between chromosomes bearing different genetic loads.
COPYRIGHT 1997 Society for the Study of Evolution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Viard, F.; Justy F.; Jarne, P.
Date:Oct 1, 1997
Previous Article:Speciation and population genetic structure in tropical Pacific sea urchins.
Next Article:The effect of temperature on body size and fecundity in female Drosophila melanogaster: evidence for adaptive plasticity.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters