Figures
Abstract
Small aquatic organisms harbour deep phylogeographic patterns and highly structured populations even at local scales. These patterns indicate restricted gene flow, despite these organisms' high dispersal abilities, and have been explained by a combination of (1) strong founder effects due to rapidly growing populations and very large population sizes, and (2) the development of diapausing egg banks and local adaptation, resulting in low effective gene flow, what is known as the Monopolization hypothesis. In this study, we build up on our understanding of the mitochondrial phylogeography of the halophilic rotifer Brachionus plicatilis in the Iberian Peninsula by both increasing the number of sampled ponds in areas where secondary contact is likely and doubling sample sizes. We analyzed partial mitochondrial sequences of 252 individuals. We found two deep mitochondrial DNA lineages differing in both their genetic diversity and the complexity of their phylogeographic structure. Our analyses suggest that several events of secondary contact between clades occurred after their expansion from glacial refugia. We found a pattern of isolation-by-distance, which we interpret as being the result of historical colonization events. We propose the existence of at least one glacial refugium in the SE of the Iberian Peninsula. Our findings challenge predictions of the Monopolization hypothesis, since coexistence (i.e., secondary contact) of divergent lineages in some ponds in the Iberian Peninsula is common. Our results indicate that phylogeographic structures in small organisms can be very complex and that gene flow between diverse lineages after population establishment can indeed occur.
Citation: Campillo S, Serra M, Carmona MJ, Gómez A (2011) Widespread Secondary Contact and New Glacial Refugia in the Halophilic Rotifer Brachionus plicatilis in the Iberian Peninsula. PLoS ONE 6(6): e20986. https://doi.org/10.1371/journal.pone.0020986
Editor: Corrie S. Moreau, Field Museum of Natural History, United States of America
Received: August 3, 2010; Accepted: May 17, 2011; Published: June 16, 2011
Copyright: © 2011 Campillo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by grant BOS2000-1451 from the Spanish Ministerio de Ciencia y Tecnología, grant BOS2003-0075 from the Spanish Ministerio de Educación y Ciencia and grant GRUPOS2004/20 from the Conselleria d'Empresa, Universitat i Ciència (Generalitat Valenciana) to MS and an advanced NERC fellowship NE/B501298 /1 (UK) to AG. SC was supported by a ‘V Segles’ fellowship from the Universitat de València (Spain). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Essential to understand the evolution of organic diversity, phylogeography studies the principles and processes that govern the geographical distribution of genetic lineages, mainly at the within species level [1], [2]. Although Avise gave zooplankton dwelling in ponds and lakes a fleeting treatment in his topical book [1], a consensus has built up in the last decade pointing to the existence of deep phylogeographic structures in these organisms [3]–[7]. This consensus strongly contrasts with the traditional view that zooplanktonic organisms should have high rates of gene flow due to high dispersal abilities; see for instance [8]. This contrast is known as the “dispersal-gene flow” paradox [4], since high gene flow should hamper the formation of deep phylogeographic structures.
The Monopolization hypothesis [4] was proposed to solve the paradox. The hypothesis extends the explanation of Boileau et al. [9] and states that gene flow after colonization should be strongly restricted due to a combination of factors. After the colonization of a habitat, the offspring of the population founders rapidly monopolizes the habitat, effectively hampering the success of later immigrants due to two factors: (1) their high population growth rates, building up extremely large populations relatively quickly, and (2) their rapid adaptation to local conditions. Therefore, a prediction of the hypothesis would be the occurrence of sharp contact zones between intraspecific phylogeographic lineages, with little or no overlap between them. In contrast, studies on the cladoceran Daphnia magna and on the rotifer Brachionus manjavacas have revealed signatures of secondary contact, e.g. [3], [6]. Following De Gelas and De Meester [3], under the monopolisation hypothesis model, there is a small but non-zero probability that population founders will belong to different phylogenetic groups than the regional one – due to long distance dispersal likely to be mediated by birds [10]. Nevertheless, if secondary contact is a general pattern in continental zooplankton, it could constitute a challenge for the hypothesis.
The Iberian Peninsula has been recognized as one of the major European glacial refugia during the Pleistocene glaciations [11]. Instead of forming a single, homogeneous glacial refugium, however, its complex orography and climatic diversity allowed the isolation of populations in separated areas during glacial maxima [12] resulting in patterns of phylogeographic concordance among different species which survived in the Peninsula during glaciations. For example, for terrestrial fauna at least seven putative glacial refugia were recognized while for freshwater fish glacial refugia coincided with the main river basins [12].
Iberian saline environments hold assemblages of relatively specialized fauna [13]. The distribution of saline environments in several drainage basins (see Figure 1) and a coastal chain of lagoons makes the establishment of concordance hypothesis feasible. In the last few years, new studies on the phylogeography of Iberian halophilic invertebrates allow for the evaluation of concordance patterns in these systems [5]–[7], [14], [15]. Some of these studies have been performed in the rotifer species complex Brachionus plicatilis, a group comprising approximately 12 cryptic species inhabiting coastal marine environments and salt lakes worldwide [16], [17]. Strong concordance between phylogeographic patterns of two ecologically similar species belonging to this complex, B. plicatilis and B. manjavacas, has been found in the Iberian Peninsula. Both species show two deep mitochondrial DNA lineages with a geographic orientation independent of the coastal/inland dichotomy [5], [6]. One of these lineages is more diverse and has a strong phylogeographical structure, with a remarkable similarity between the phylogeographic patterns of these species-specific lineages. In addition, B. manjavacas shows widespread secondary contact between mtDNA lineages, and indications of secondary contact have also been found in B. plicatilis [5], [18].
See pond acronyms in Table 1. Basins: 1, Duero; 2, Ebro; 3, Guadiana; 4, Júcar-Segura; 5, Guadalquivir; 6, coastal lagoons.
Obtaining reliable phylogeographic results is sensitive to the spatial sampling scope. In order to assess previous conclusions and to gain a better understanding of phylogeography in the Iberian Peninsula, we investigate B. plicatilis mtDNA variation in areas where secondary contact is suspected, and increase sample sizes in other habitats in the Iberian Peninsula to double the number of B. plicatilis sequences. Specifically, we want to ascertain the extent of secondary contact in this species, drawing conclusions on the processes involved and the implications for the Monopolization hypothesis. Finally, we review and discuss concordance patterns between the phylogeographies of B. plicatilis, and other organisms inhabiting Iberian halophilic habitats.
Materials and Methods
Rotifers of the species complex Brachionus plicatilis combine both sexual and asexual reproduction, i.e. are cyclical parthenogens, e.g. [15]. In a typical growth cycle, the active population is initiated by hatching of diapausing eggs from the sediment of the pond or lake. These hatchlings are diploid asexual females produce genetically identical daughters by parthenogenesis. After a period of clonal propagation, a sexual phase starts when, induced by environmental factors, such as population density, asexual females start produce sexual females among their offspring [19]–[23]. Sexual females produce haploid eggs. If a sexual female is not inseminated in the first hours of her life, her unfertilized eggs will develop into haploid eggs [24]. If inseminated, her fertilized, sexually produced, eggs will result in a diploid embryos, the so-called diapausing eggs. All members of the B. plicatilis species complex investigated to date show this cyclical parthenogenetic life cycle.
We isolated B. plicatilis diapausing eggs from sediment samples from six Spanish ponds using a sugar flotation technique [25]. We extracted DNA from diapausing eggs using Chelex [26], [27]. In order to select diapausing eggs belonging to B. plicatilis from the morphologically similar ones from other syntopic species of the B. plicatilis cryptic species complex, we either used a barcoding approach [28] and sequenced directly polymerase chain reaction (PCR) products (a fragment of the COI mitochondrial gene, see below) in order to use only B. plicatilis sequences, or conducted analysis by restriction fragment length polymorphism (RFLP) on PCR amplifications of the same gene [26] prior to sequencing. The latter approach was used in those lakes where B. plicatilis was not the dominant species in the diapausing egg bank.
We used specific primers COIdgF and COIdgR –see [17] for details– to amplify a fragment of the mitochondrial gene cytochrome c oxidase subunit 1 (cox1 or COI). We performed PCR in a total of 50 µl containing 3 µl of template DNA, 0.2 mM of each nucleotide, 0.6 µM of each primer, 1.5 U of Taq polymerase (BIOTOOLS) and 1× BIOTOOLS buffer (producing a final concentration of 2 mM MgCl2). A Mastercycler™ (Eppendorf) was used for PCR amplification with a cycling profile of 3 min at 94°C; 40 cycles of 30 s at 94°C, 1 min at 48°C and 1 min at 72°C; and a final extension step of 7 min at 72°C. PCR were purified (High Pure PCR Product Purification Kit, Roche) and sequenced using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer) on an ABI 3700 automated sequencer.
We computed the number of haplotypes per population and haplotype diversity (Hd) using DnaSP version 4.50.1 [29]. We performed a rarefaction analysis using RAREFAC (available from R. Petit at http://www.pierroton.inra.fr/genetics/labo/Software/Rarefac/index.html) to calculate standardized allelic richness (A) for each sampled population. We set the rarefaction size to 5. Thus, seven populations with n<5 (Poza Norte de Torreblanca, Mojón Blanco, Fuente de Piedra, El Saladar, Clot de Galvany, Atalaya de los Ojicos and Casa Nueva) were excluded from this analysis.
We calculated pairwise estimates of population differentiation (FST) using Arlequin 3.1 (30) applying the available evolutionary model closest to that recommended by ModelGenerator (see below). We performed a reduced major axis (RMA) linear regression between FST estimates and geographical distances. As the regression contained multiple pairwise comparisons, the Mantel test for significance was employed with 30,000 randomizations. All regression related analyses were carried out in IBDWS v 3.15 [31].
We constructed an alignment using the program MEGA5 [32] with all the individual sequences. Identical sequences were then collapsed using Collapse v. 1.2 [33]. We constructed a maximum likelihood phylogenetic tree of all haplotypes using the program MEGA5 [32] with 1000 pseudorreplicates to assess branch robustness, and the best fit evolutionary model to our dataset according to ModelGenerator v. 0.84 [34]. We used two outgroups: one from Canada (GenBank AF499055.1) and the other from Australia (GenBank AF524543) both from the species B. plicatilis.
We used Nested Clade Phylogeographic Analysis (NCPA) [35] to test for associations between geography and genealogy. This method allows distinguishing between contemporary and historical events (for example, gene flow from allopatric fragmentation). We followed the rules by Templeton [36] to construct the nested design on the network using the TCS v1.18 program [37]. This version of the program does not include the following nesting rule that we applied to construct the nested design: “When the nesting rules at level n create an interior, degenerate n-step clade [an n-step clade with only (n−1)-step member with actual observations, the other (n−1)-step member or members being interior nodes without any observations], such clades are nested at the (n+1) level with the n-step clade that is mutationally closest to the one (n−1)-step member of the degenerate clade with observations” (D. Posada and A.R. Templeton, personal communication). The program GeoDis 2.6 [38] was used to compute the distance measures: the clade distance (Dc), which measures the geographical range of a particular clade, and the nested-clade distance (Dn), which measures how a particular clade is geographically distributed relative to its closest evolutionary sister clade (i.e. clades in the same higher-step nesting category) [27]. GeoDis also computes the statistical significance of these distance measures by a Chi-square test after 10,000 random permutations of clades against sampling locality. This version of the program includes a Dunn-Šidák correction for multiple tests.
We then applied the inference key (version 28th April 2009) establishing the phylogeographic inferences for each clade. To detect secondary contact, we performed the analysis suggested in Templeton [39] and previously applied to the B. manjavacas Iberian phylogeography [6]. In those higher-level clades with signs of population expansion inferred by the NCPA, we performed a mismatch distribution analysis [40] in order to independently test the demographic expansion by using DNAsp version 4.50.1 [29].
Results
We sequenced a total of 154 individuals from the six sampled populations (GenBank accession numbers FJ387332–FJ387485), to which we added the 98 previously sequenced for Iberian populations (GenBank accession numbers: AF266853–AF266950) [5]. B. plicatilis was found in 20 ponds in the Iberian Peninsula, in the following basins: Duero, Ebro, Guadiana, Júcar-Segura, Guadalquivir, and in five ponds of the coastal Mediterranean chain (Figure 1). The 252 B. plicatilis cytochrome oxidase c subunit I (COI) sequences collapsed in a total of 29 haplotypes (Table 1). On average, 12.60 individuals were sequenced per population (range: 2–36; see also Figure S1). The alignment included 581 bp with no insertions, deletions, stop codons or ambiguities. There were 49 variable sites (47 synonymous and 2 non-synonymous) and 40 parsimony-informative sites. Since this alignment was slightly shorter than in Gómez et al. [5], haplotypes 18 and 16 from that work merged into one (new haplotype 16).
Twenty-five out of the 29 haplotypes were found in single ponds. The most widespread haplotype (number: 21) was found in eight ponds, seven of them from the Júcar-Segura basin, and the remaining (Clot de Galvany) from the coastal chain, but close to the former basin. There was an average of 2.2 haplotypes per pond. Among the populations showing the highest standardized allelic-richness (A), three were coastal (Estany d'en Túries, Hondo Sur and Poza Sur de Torreblanca) and one inland (Balsa de Santed 1). Standardized allelic richness for ponds with more than 5 individuals sequenced ranged from 0 to 2.831 (Table 1).
The Akaike Information Criterium II and Bayesian Information Criterium implemented in ModelGenerator v. 0.84 selected the Hasegawa-Kishino-Yano plus Gamma model (HKY+G; G = 0.15) as the best fit model of evolution for our dataset [41], with base frequencies of A = 0.20, C = 0.22, G = 0.20, and T = 0.38. The phylogenetic tree (Figure 2) showed two strongly supported Iberian lineages: lineage A, which contains predominantly northern haplotypes (inland and coastal), plus Hondo Sur, a south-eastern coastal lagoon; and lineage B, which contains predominantly southern inland haplotypes with presence also in two ponds from the coastal chain (Clot de Galvany and Hondo Sur) and some northern inland ponds (Las Eras, Balsa de Santed 1 and Balsa de Santed 2). Lineages A and B were found in 8 and 14 ponds, respectively and co-occurred in three ponds, with frequencies of lineage A in the range 3–57%. Both lineages differed in their haplotype diversity, lineage A with 23 haplotypes and lineage B with 6. The average uncorrected p-distance between these lineages is 2.7%.
Percent bootstrap support, if >50%, is shown. See Table 1 for pond acronyms.
The haplotype network consisted in 29 haplotypes, 25 1st-step clades, 10 2nd-step clades and four 3rd-step clades (Figure 3). The maximum level of parsimonious steps (P>0.95) was 10 and no connection between haplotypes exceeded this number (maximum of 8 steps between haplotypes 27 and 16).
We designated haplotypes by an H followed by a number; zeros represent missing intermediate haplotypes. We refer to clades as x-y, where x represents the nesting step (from the haplotype step to the 3rd step) and y represents the number assigned to that particular clade. Each line represents one mutation. All connections indicated by solid line between haplotypes are supported to be parsimonious at the 95% level. Locations of origin for each haplotype are given in italics.
Eight out of 17 clades with geographical and genetic variation resulted in significant permutational contingency tests, indicating non-random geographical distribution of haplotypes (Table 2). The NCPA inferred restricted gene flow / dispersal but with some long distance dispersal in two clades (clade 1–17 and the total cladogram). Clades 2-4 and 3-4 showed signs of restricted gene flow with isolation by distance. Clade 2-3 (found in coastal ponds) showed a contiguous range expansion and clade 3-2 (found in coastal and northern inland ponds) showed signs of long distance colonization, seemingly from southern coastal ponds to northern coastal ones and to those of the Ebro valley. The mismatch analysis for the clade 3-2 (Fig. 4) also supported a recent event of population expansion. Finally, clades 2-7 and 1-9 showed signs of allopatric fragmentation, between Hoya Rasa, El Saladar and Pétrola (Júcar-Segura basin) and Manjavacas (Guadiana basin) on one hand, and between Ebro and Duero basins, on the other.
Solid line represents the number of observed pairwise nucleotide differences among B. plicatilis clades within the clade. The dash line represents the expected distribution of nucleotide differences under a model of recent population expansion. Raggedness statistic r = 0.0437.
In order to test for secondary contact, we performed the analysis proposed by Templeton [39] for the total cladogram (Figure 5). As we move towards the higher-step clades, the average geographic pairwise distance remains different from zero; this happens even at the highest-clade step for six populations from five areas: Hondo Sur, Las Eras, Poza Sur, Poza Norte de Torreblanca, Balsa de Santed 1 and Estany d'en Túries. Therefore, secondary contact between two or more mtDNA clades at the highest level is the suggested inference for these ponds. In Hondo Sur, clades 3-1, 3-2 and 3-4 were found with frequencies 3.33%, 93.33% and 3.33%, respectively. In Las Eras, clades 3-2 and 3-4 overlapped with frequencies 90% and 10%, respectively. In Poza Sur, three clades (3-1, 3-2 and 3-3) overlapped with frequencies 83.33%, 13.89%, and 2.78%, respectively. In Poza Norte, clades 3-1 and 3-3 appeared with equal frequencies. In Balsa de Santed 1, clades 3-2 and 3-4 have frequencies 43.48% and 56.52%, respectively. Finally, in Estany d'en Túries, clades 3-2 and 3-3 have frequencies 14.29% and 85.71%, respectively. The distribution of clades and patterns of secondary contact between them in these populations is illustrated in Figure 6.
Ponds where the species was found are ordered on the x axis in increasing distances from the most westerly pond to the most easterly. The bars show the average pairwise geographical distance between the geographical centres of haplotypes and clades found at each of the ponds calculated from the coordinates for haplotypes and clades given in GeoDis output. For pond acronyms, see Table 1.
Note that these highest-step clades are sympatric at six ponds, a signature for secondary contact. 1: HOS; 2: HOY; 3: SA1; 4: TOS; 5: CHI; 6: SAL; 7: TUR; 8: TON; 9: GAL; 10: ERA; 11: SA2; 12: CAP; 13: MOJ; 14: FUE; 15: MAN; 16: SLD; 17: PET; 18: CLO; 19: ATA; 20: CAS. For pond acronyms, see Table 1.
B. plicatilis populations were highly subdivided (Figure 7; Table 3), with a global FST of 0.79 (P<0.001) and pairwise FST values ranging from −0.19 to 1. IBDWS detected a positive and significant correlation between genetic and geographical distance for the overall dataset (r = 0.37, P = 0.0009; genetic distance = 0.182+0.00187 geographical distance, in km).
IBDWS detected a positive and significant correlation between genetic and geographical distance (genetic distance = 0.182+0.00187 geographical distance, in km).
Discussion
The addition of new ponds in the present study has improved the understanding of phylogeographic patterns of B. plicatilis in the Iberian Peninsula. In support of previous findings [5], B. plicatilis populations were strongly subdivided and displayed a marked phylogeographic structure, but also showed evidence of secondary contacts. The two deep lineages (A and B) reported by Gómez et al. [5] were also found, although in the present study a larger level of spatial overlap was observed. Possibly, lineage A was in the area first and lineage B expanded later, in agreement with the lower genetic diversity of the latter.
Phylogeography and population structure within clades
The cladogram showed some differences with respect to the phylogram, mainly due to the nesting rules in the nested clade phylogeographic analysis. The cladogram displayed four highest-step nested clades (3rd-step) while the phylogram displayed two main lineages (A and B). Lineage B corresponded to nested clade 3-4 in the cladogram, while clades 3-1 and 3-3 corresponded to well-supported branches from lineage A in the phylogram. Nested clade 3-2 grouped basal haplotypes and less supported branches in the phylogenetic tree, corresponding to its central position on the cladogram. The cladogram found here differs slightly from that in Gómez et al. [5] in that we found an extra 3rd-step clade (3-1 in this study), and a higher complexity in clade 3-2. These changes resulted from new haplotypes found mainly in Hondo Sur lagoon. Interestingly, haplotypes from this lagoon occupy a central position in the cladogram, which results in changes in the inference of the location of a putative refugium of lineage A in B. plicatilis (see below). Accordingly to NCPA, lineage A (clade 3-2) shows an isolation-by-distance pattern (but see below), with (1) contiguous range expansion along some lagoons of the coastal chain (clade 2-3, i.e., Estany d'en Túries, Poza Sur, and Hondo Sur), and (2) possibly range expansion in the northern inland ponds (clade 2-4, i.e. Laguna de las Eras, Balsa de Santed 1, Gallocanta, and Salada de Chiprana). The lower haplotype diversity of northern inland populations (clade 2-4) and the position of their haplotypes in the network as tips are findings consistent with an expansion from outside the area, probably from the coastal chain of lagoons. This conclusion differs from what was proposed by Gómez et al. [5], since they suggested that there was a past fragmentation event between inland and coastal ponds. Nevertheless, both explanations are compatible if an initial range expansion was followed by isolation between origin and expansion areas.
Although some haplotypes were more widespread (e.g. haplotype 21), the results point out a general pattern of private haplotypes, present in single lakes. Also, we can conclude that there are separated phylogeographic groups, one distributed along the coast and northern inland ponds and the other mainly in southern inland ponds and a few ponds from the North. This finding is in agreement with expectations of the monopolisation hypothesis (but see below). There are still some unsampled areas, mainly from the Guadiana and the Guadalquivir basins, which could affect our conclusions to some extent, probably for clade 3-4 (lineage B). In addition, samples from North Africa could help us in determining the route of colonization or even, possible glacial refugia in this area.
Refugia and colonization pattern
Glacial refugia are defined as zones with suitable conditions for species to survive during the last Pleistocene glaciations. Iberian, Balkan-Greece and Italian Peninsulas have been recognized as major glacial refugia for European species [42]. In particular, the Iberian Peninsula has been acknowledged as a main glacial refugium, or more precisely as a complex of multiple glacial refugia, ‘refugia within refugia’ as expressed by Gómez and Lunt [12]. As it happens for other dispersal centres, glacial refugia are characterized by having larger allelic or haplotype richness than more recently colonized areas, even though zones of secondary contact of divergent lineages –contact zones– can show also high genetic diversity [42], [43].
In the case of Brachionus rotifers, areas within the Iberian Peninsula showing higher haplotype diversity within a clade are thus good candidates to be glacial refugia for the different lineages. Although much work is still needed in this area, there is some evidence of persistence of saline conditions in some Iberian lakes, during glacial advances [44], [45]. One of the main problems when trying to precisely establish glacial refugia for aquatic invertebrates in the Iberian Peninsula is the loss of their habitats due to human activities [7]. On one hand, many small salt ponds have been modified for agricultural purposes and, on the other, many coastal ponds are increasingly threatened by urbanization. Another important caveat in determining glacial refugia is rotifer capability for long-distance dispersal [46], which may confound patterns of regional colonization. Finally, even though we tried to retrieve samples from as many possible salt ponds in the Iberian Peninsula, the distribution of halophilic rotifers in the Iberian Peninsula as a whole is still incompletely known, and sampling can be affected by the patchy distribution of diapausing eggs in the sediment [47]. The apparent alternation between B. plicatilis and B. manjavacas often found on the superficial sediment of ponds (Montero-Pau et al., submitted) also makes the retrieval of the focus species in a particular sampling trip difficult. It should be taken into account that all these factors could be affecting the data when establishing the possible glacial refugia of B. plicatilis.
The average sequence divergence between the two main B. plicatilis lineages was 2.7%. Based on sequence divergence rates described for arthropods (1.4% of divergence per million years) [38], the split time between lineages A and B would be around 2 million years ago, predating the Pleistocene ice ages, which would mean that both lineages have survived at least several ice ages. Within lineage A, we found that coastal populations have higher haplotype diversity than inland populations. Within the coastal pond chain, the central position of Hondo Sur population in the cladogram suggests this area harboured the oldest glacial refugium for lineage A. Originally, Hondo Sur was part of a coastal lagoon, but due to silting and human activities its size was reduced through time [49]. The NCPA suggests that the species expanded from this refugium through the Iberian Mediterranean coast and entered inland regions through the Ebro basin.
Regarding lineage B, its low haplotype diversity and the close relationships of the haplotypes involved are consistent with a more recent arrival in the study area. Nonetheless, the geographic structuring of this lineage in the region and the occurrence of private haplotypes in Duero, Guadiana and Júcar-Segura basins suggest a recent glacial refugium in the Iberian Peninsula. As discussed above, more extensive sampling of the Guadiana and Guadalquivir basins and the southern coastal lagoons would be needed to conclude on the location of a putative glacial refugium for this lineage. A possibility would be that a refugium was in yet unsampled areas on the south of the Iberian Peninsula or North Africa.
From the glacial refugia, a pattern of serial founder effects resulting from successive colonizations from nearby habitats could explain the correlation between genetic and geographical distances in B. plicatilis. This idea was initially put forward for humans in their expansion throughout the world [50] and subsequently proposed to explain the correlation between genetic and geographical distances found in several continental aquatic invertebrates including B. manjavacas [6], the brine shrimp Artemia salina [7], an Australian isopod [51] and for B. plicatilis at a worldwide level [46]. Serial founder effects involve a single individual or a small group of individuals from nearby habitats colonizing the closest available habitat and this new population serving as origin for further colonizations, creating this pattern of correlation between genetic and geographical distances. Although we found a significant and positive correlation between these two distances, an important part of the variation in the genetic distance remains unexplained by geographic distance, pointing out that other factors affect this relationship. A few possible explanations could be: first, the occurrence of more than one refugia from which expansion happens; second, the patterns of secondary contact between genetically divergent lineages; third, the occurrence of long-distance dispersal; and fourth, the recent episodes of range expansions where single or highly related haplotypes have recently colonized an area. The results from the present study indicate that all of these factors would be at play in the B. plicatilis system. Despite all these factors the relationship between genetic and geographic distance was still significant.
Phylogeographic concordance in the Iberian Peninsula
The phylogeographic patterns of B. plicatilis and B. manjavacas in the Iberian Peninsula show remarkable similarities [6]. First, both species must have been sympatric in the Iberian Peninsula for a long time. Both species show two main mtDNA lineages, and, interestingly, one of these lineages in each species had much higher genetical diversity than the other. Therefore, in both species there is a major lineage that seems to have survived during several ice ages in the Iberian Peninsula, and shows strong phylogenetic structure. By contrast, in both species a lineage with a recent expansion, which does not exclude colonization from outside the Iberian Peninsula, has been found. In both species, we found processes of range expansion, allopatric fragmentation, patterns of isolation-by-distance and secondary contact between lineages. Despite the localized and isolated character of the habitats of these rotifers (salt ponds and lakes), evidence for two glacial refugia for rotifers in the Iberian Peninsula (Guadiana basin and around El Hondo Nature Reserve area) was found. Each species might have survived in different basins, possibly varying in their salinity regime [6; Montero–Pau et al., submitted].
Recent studies on the phylogeography of halophilic invertebrates are indeed supporting a pattern of high population subdivision, regional endemism and multiple refugia within the Iberian Peninsula (Figure 8). Three Iberian refugia and patterns of serial colonization have been proposed for the brine shrimp Artemia salina. The hypothesized refugia would have been located around the southeastern area (Almería), the Doñana area and possibly a third refugium from which the East coast of the Iberian Peninsula and the Balearic Islands were colonized [7]. The endangered water beetle Ochthebius glaber, which lives in Iberian hypersaline streams, has been found in three groups of populations: in the basins of the Segura, Guadalquivir and Júcar rivers [14]. The authors found signatures of range expansion, past fragmentation and restricted gene flow with isolation by distance. In the halophylic grasshopper Mioscirtus wagneri, certain similarities are apparent with the patterns found for Brachionus rotifers [15], since there are three phylogenetic groups geographically isolated among them and located in north-east, central–south-east and south-west Iberia. Genetic and geographical distances are also correlated in these grasshoppers. The authors suggested the pattern resulted from vicariance of a single ancient, more widespread population in the Iberian Peninsula, which contracted with the retraction of halophylic environments, possibly in the Early Pleistocene. The global picture arising from these studies and our results is that southern Iberian Peninsula harboured glacial refugia for different halophilic organisms (Figure 8). Nevertheless, we must compare the rotifers' phylogeographic patterns with those of the beetle and the grasshopper with caution, since long distance dispersal seems more likely in rotifers. For this reason, in the case of rotifers the presence of North African refugia is plausible, as crossing of the Gibraltar straits is a more likely scenario.
Secondary contact: Implications for the Monopolization hypothesis
One prediction of the Monopolization hypothesis [4] is that events of secondary contact are unlikely. The first migrants form a resident population which will tend to monopolize the habitat quite quickly, hampering the later effective colonization of other migrants. This monopolization is assumed to be mainly due to two facts: fast growth population rates (a numerical effect) and the rapid adaptation to the new habitat (local adaptation effect). Indeed, many aquatic organisms show little signs of secondary contact, with the different lineages being allopatric, but both Brachionus species analyzed so far do show signs of overlap in space between lineages. There are several, non-exclusive factors that could explain this finding.
First, the monopolization applies to populations of the same species and thus, one possibility is that the different lineages found were in fact different species, occupying differentiated niches. However, in our case, this is not supported by population genetic studies with nuclear loci, such as microsatellites [52], [53], which shows Hardy Weinberg equilibrium, and so recombination, in ponds where different mitochondrial lineages co-occur. Moreover, no signs of reproductive isolation between clones belonging to the two mitochondrial phylogenetic lineages A and B were found [18], supporting that these two lineages effectively belong to a single biological species.
Second, Iberian saline ponds and lakes are known for being highly variable, with strong fluctuations in water level and environmental conditions [43]. Assuming that local adaptation is common, but also that rotifer habitats have very variable ecological conditions, an immigrant arriving to a pond where conditions have just changed could have opportunities to thrive when its source pond would have similar conditions to the new conditions in the target pond, even if this pond would have already a well-established population. We hypothesize that this could increase the chances of secondary contact between lineages evolved in allopatry. Interestingly, these rotifers disperse through diapausing eggs, a resistant form that can survive for decades in the sediment [47], until environmental conditions are favourable to hatch. A previous study with rotifers in the Iberian Peninsula [53] suggested local adaptation and ecological specialization in some Iberian populations of this rotifer, but the authors stressed that environmental fluctuation seemed to hamper to some extent local adaptation to salinity and temperature in most populations.
Third, divergent lineages could have arrived to the habitat almost simultaneously, effectively establishing a population at the same time and before either of them could effectively monopolize the new habitat.
Finally, as suggested in other studies, hybrid vigour could help explain the pattern observed since inbreeding depression could affect the initial phases of colonization of a new habitat [54]. If a few migrants from the same origin colonize a new habitat, they will probably be affected by inbreeding depression [54]. Migrants belonging to other lineage and arriving later could mate with them, giving place to offspring with hybrid vigour [55].
Conclusions
Some of our results are consistent with previous studies [5]: first, B. plicatilis is made of two main phylogeographic lineages in the Iberian Peninsula, with lineage A being genetically more diverse than lineage B; second, lineages survived for several ice ages and show strong phylogeographic structure; and third, the genetically much more diverse and structured lineage A is likely the oldest one, possibly surviving several climatic cycles in the area, while lineage B expanded its range later. In contrast, additional samples included in this study have revealed that a population not sampled in previous studies –Hondo Sur– played a key role in the phylogeography of B. plicatilis. The central position of Hondo Sur population in the cladogram suggests that this area harboured the oldest glacial refugium for lineage A. Moreover, in this study we have found processes of allopatric fragmentation and range expansion with patterns of isolation-by-distance in B. plicatilis. Historical events (serial colonizations) could explain the isolation-by-distance pattern. The study also confirms that secondary contact is more frequent in these rotifers than previously thought. Finally, phylogeographic patterns shown by B. plicatilis in the Iberian Peninsula are highly concordant with those found in B. manjavacas and have certain similarities with other organisms dwelling in halophylic environments.
Supporting Information
Figure S1.
Relationship between sample size (number of individuals sequenced) and the number of haplotypes.
https://doi.org/10.1371/journal.pone.0020986.s001
(TIF)
Author Contributions
Conceived and designed the experiments: SC AG. Performed the experiments: SC AG. Analyzed the data: SC AG MS. Contributed reagents/materials/analysis tools: SC AG MS MJC. Wrote the paper: SC AG.
References
- 1.
Avise JC (2000) Phylogeography: The History and Formation of Species. Cambridge: Harvard University Press. 464 p.
- 2. Avise JC (1998) The history and purview of phylogeography: a personal reflection. Mol Ecol 371–379.
- 3. De Gelas K, De Meester L (2005) Phylogeography of Daphnia magna in Europe. Mol Ecol 14: 753–64.
- 4. De Meester L, Gómez A, Okamura B, Schwenk K (2002) The monopolization hypothesis and the dispersal-gene flow paradox in aquatic organisms. Acta Oecol 23: 121–135.
- 5. Gómez A, Carvalho GR, Lunt DH (2000) Phylogeography and regional endemism of a passively dispersing zooplankter: mitochondrial DNA variation in rotifer resting egg banks. Proc R Soc B-Biol Sci 267: 2189–2197.
- 6. Gómez A, Montero-Pau J, Lunt DH, Serra M, Campillo S (2007) Persistent genetic signatures of colonization in Brachionus manjavacas rotifers in the Iberian Peninsula. Mol Ecol 16: 3228–3240.
- 7. Muñoz J, Gómez A, Green AJ, Figuerola J, Amat F, et al. (2008) Phylogeography and local endemism of the native Mediterranean brine shrimp Artemia salina (Branchiopoda: Anostraca). Mol Ecol 17: 3160–3177.
- 8.
Mayr E (1963) Animal species and evolution. Cambridge: Belknap Press of Harvard University Press. 797 p.
- 9. Boileau MG, Hebert PDN, Schwartz SS (1992) Non-equilibrium gene frequency divergence: founder effects in natural populations persistent. J Evolution Biol 5: 25–39.
- 10. Green AJ, Figuerola J (2005) Recent advances in the study of long-distance dispersal of aquatic invertebrates via birds. Divers Distrib 11: 149–156.
- 11. Hewitt GM (2004) Genetic consequences of climatic oscillations in the Quaternary. Philos Trans R Soc B-Biol Sci 359: 183–195.
- 12.
Gómez A, Lunt DH (2007) Refugia within refugia: patterns of phylogeographic concordance in the Iberian Peninsula. In: Weiss S, Ferrand N, editors. Phylogeography of Southern European Refugia. Dordrecht: Springer. pp. 155–188.
- 13. Comín FA, Alonso M (1988) Spanish salt lakes: Their chemistry and biota. Hydrobiologia 158: 237–245.
- 14. Abellán P, Gómez-Zurita J, Millán A, Sánchez-Fernández D, Velasco J, et al. (2006) Conservation genetics in hypersaline inland waters: mitochondrial diversity and phylogeography of an endangered Iberian beetle (Coleoptera: Hydraenidae). Conserv Genet 8: 79–88.
- 15. Ortego J, Bonal R, Cordero PJ, Aparicio JM (2009) Phylogeography of the Iberian populations of Mioscirtus wagneri (Orthoptera : Acrididae), a specialized grasshopper inhabiting highly fragmented hypersaline environments. Biol J Linnean Soc 97: 623–633.
- 16. Gómez A, Serra M, Carvalho GR, Lunt DH (2002) Speciation in ancient cryptic species complexes: evidence from the molecular phylogeny of Brachionus plicatilis (Rotifera). Evolution 56: 1431–1344.
- 17. Suatoni E, Vicario S, Rice S, Snell TW, Caccone A (2006) An analysis of species boundaries and biogeographic patterns in a cryptic species complex: the rotifer Brachionus plicatilis. Mol Phylogenet Evol 41: 86–98.
- 18. Berrieman HK, Lunt DH, Gómez A (2005) Behavioural Reproductive Isolation in a Rotifer Hybrid Zone. Hydrobiologia 546: 125–134.
- 19.
Wallace RL, Snell TW, Ricci C, Nogrady T (2006) ROTIFERA: Volume 1. Biology, Ecology and Systematics. Leiden: Backhuys Publishers. 299 p.
- 20. Carmona MJ, Gómez A, Serra M (1995) Mictic patterns of the rotifer Brachionus plicatilis Müller in small ponds. Hydrobiologia 313–314: 365–371.
- 21. Carmona MJ, Serra M, Miracle MR (1993) Relationships between mixis in Brachionus plicatilis and preconditioning of culture medium by crowding. Hydrobiologia 255–256: 145–152.
- 22. Stelzer CP, Snell TW (2003) Induction of Sexual Reproduction in Brachionus plicatilis (Monogononta, Rotifera) by a Density-Dependent Chemical Cue. Limnol Oceanogr 48: 939–943.
- 23. Snell TW, Kubanek J, Carter W, Payne AB, Kim J, et al. (2006) A protein signal triggers sexual reproduction in Brachionus plicatilis (Rotifera). Mar Biol 149: 763–773.
- 24. Snell TW, Childress M (1987) Aging and loss of fertility in male and female Brachionus plicatilis (Rotifera). International Journal of Invertebrate Reproduction and Development 12: 103–110.
- 25. Gómez A, Carvalho GR (2000) Sex, parthenogenesis and genetic structure of rotifers: microsatellite analysis of contemporary and resting egg bank populations. Mol Ecol 9: 203–214.
- 26. Campillo S, García-Roger EM, Martínez-Torres D, Serra M (2005) Morphological Stasis of Two Species Belonging to the L-morphotype in the Brachionus plicatilis Species Complex. Hydrobiologia 546: 181–187.
- 27. Gómez A, Clabby C, Carvalho GR (1998) Isolation and characterization of microsatellite loci in a cyclically parthogenetic rotifer, Brachionus plicatilis. Mol Ecol 7: 1613–1621.
- 28. Hebert PDN, Cywinska A, Ball SL, Dewaard JR (2003) Biological identifications through DNA barcodes. Proc R Soc B-Biol Sci 270: 313–321.
- 29. Rozas J, Sánchez-DelBarrio JC, Messeger X, Rozas R (2003) DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496–2497.
- 30. Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary bioinformatics online 1: 47–50.
- 31. Jensen JL, Bohonak AJ, Kelley ST (2005) Isolation by distance, web service. BMC genet 6: 13–18.
- 32. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, et al. (2011) MEGA5: Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Methods. Mol Biol Evol. In press.
- 33.
Posada D (2005) Collapse: Describing haplotypes from sequence alignments. Available from http://darwin.uvigo.es/.
- 34. Keane TM, Creevey CJ, Pentony MM, Naughton TJ, Mclnerney JO (2006) Assessment of methods for amino acid matrix selection and their use on empirical data shows that ad hoc assumptions for choice of matrix are not justified. BMC Evol Biol 6: 29–45.
- 35. Templeton AR (1998) Nested clade analyses of phylogeographic data: testing hypotheses about gene flow and population history. Mol Ecol 7: 381–397.
- 36.
Templeton AR (2006) Population genetics and microevolutionary theory. Hoboken: John Wiley & Sons, Inc. 705 p.
- 37. Clement M, Posada D, Crandall KA (2000) TCS: a computer program to estimate gene genealogies. Mol Ecol 9: 1657–1659.
- 38. Posada D, Crandall KA, Templeton AR (2000) GeoDis: a program for the cladistic nested analysis of the geographical distribution of genetic haplotypes. Mol Ecol 9: 487–488.
- 39. Templeton AR (2001) Using phylogeographic analyses of gene trees to test species status and processes. Mol Ecol 10: 779–791.
- 40. Slatkin M, Hudson RR (1991) Pairwise Comparisons of Mitochondrial DNA Sequences in Stable and Exponentially Growing Populations. Genetics 129: 555–562.
- 41. Hasegawa M, Kishino H, Yano T (1985) Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. J Mol Evol 22: 160–174.
- 42. Hewitt GM (2001) Speciation, hybrid zones and phylogeography - or seeing genes in space and time. Mol Ecol 10: 537–549.
- 43. Petit RJ, Aguinagalde I, de Beaulieu J-L, Bittkau C, Brewer S, et al. (2003) Glacial refugia: hotspots but not melting pots of genetic diversity. Science 5625: 1563–1565.
- 44. Morellón M, Valero-Garcés B, Vegas-Vilarrúbia T, González-Sampériz P, Romero Ó, et al. (2009) Lateglacial and Holocene palaeohydrology in the western Mediterranean region: The Lake Estanya record (NE Spain). Quat Sci Rev 28: 2582–2599.
- 45. González-Sampériz P, Valerogarces B, Moreno A, Morellon M, Navas A, et al. (2008) Vegetation changes and hydrological fluctuations in the Central Ebro Basin (NE Spain) since the Late Glacial period: Saline lake records. Paleogeogr Paleoclimatol Paleoecol 259: 157–181.
- 46. Mills S, Lunt DH, Gómez A (2007) Global isolation by distance despite strong regional phylogeography in a small metazoan. BMC Evol Biol 7: 225–234.
- 47. García-Roger EM, Carmona MJ, Serra M (2006) Patterns in rotifer diapausing egg banks: Density and viability. J Exp Mar Biol Ecol 336: 198–210.
- 48. Knowlton N, Weigt LA (1998) New dates and new rates for divergence across the Isthmus of Panama. Proc R Soc B-Biol Sci 265: 2257–2257.
- 49. Rosselló VM (1993) Albuferas mediterráneas. Cuad de Geogr 53: 35–64.
- 50. Ramachandran S, Deshpande O, Roseman CC, Rosenberg NA, Feldman MW, et al. (2005) Support from the relationship of genetic and geographic distance in human populations for a serial founder effect originating in Africa. Proc Natl Acad Sci U S A 102: 15942–15947.
- 51. Gouws G, Stewart BA (2007) From genetic structure to wetland conservation: a freshwater isopod Paramphisopus palustris (Phreatoicidea: Amphisopidae) from the Swan Coastal Plain, Western Australia. Hydrobiologia 589: 249–263.
- 52. Gómez A, Adcock GJ, Lunt DH, Carvalho GR (2002) The interplay between colonization history and gene flow in passively dispersing zooplankton: microsatellite analysis of rotifer resting egg banks. J Evol Biol 15: 158–171.
- 53. Campillo S, García-Roger EM, Carmona MJ, Gómez A, Serra M (2009) Selection on life-history traits and genetic population divergence in rotifers. J Evol Biol 22: 2542–2553.
- 54. Tortajada AM, Carmona MJ, Serra M (2009) Does haplodiploidy purge inbreeding depression in rotifer populations? PloS ONE 4: e8195.
- 55. Haag CR, Riek M, Hottinger JW, Pajunen VI, Ebert D (2006) Founder events as determinants of within island and among island genetic structure of Daphnia metapopulations. Heredity 96: 150–158.