Journal of Heredity 2003:94(4)
© 2003 The American Genetic Association 94:310-314
Microsatellite Variation and Differentiation in North Atlantic Eels
From the Department of Genetics, University of Georgia, Athens, GA 30602.
Address correspondence to Judith E. Mank at the address above, or e-mail: jemank{at}uga.edu.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
We screened 11 populations of American, European, and Icelandic eels (Anguillidae) for allelic variation and genetic divergence at six polymorphic microsatellite loci. Within either of the two recognized Anguilla species in the North Atlantic (rostrata in the Americas, anguilla in Europe), population genetic structure was statistically significant but weak; fully 95% of the total genetic variation was present within geographic locales rather than distributed among them. The two Anguilla species also overlap greatly in allelic frequencies, so the available data proved ineffective for addressing hypotheses about the possible hybrid origins of some Icelandic eels. The overlapping microsatellite profiles contrast with nearly diagnostic species differences documented previously in allozymes and mtDNA. This and similar empirical findings in several other species support theoretical concerns that homoplasy (convergent evolution) in allelic states can compromise the utility of rapidly mutating microsatellite loci for certain types of microevolutionary questions regarding gene flow and species differences.
Anguillid eels of the North Atlantic traditionally have been divided into two taxonomic species: Anguilla rostrata (American eels) from the Western Hemisphere, and A. anguilla (European eels) from Europe, Iceland, and north Africa. However, only one quasi-diagnostic morphological feature is known (Tesch 1977); individuals of A. rostrata exhibit 103–110 (mean 107.1) vertebrae (or myomeres in larvae), whereas the count in A. anguilla ranges from 110–119 (mean 114.7). Anguillid eels are catadromous, meaning that maturing individuals spend most of their lives in estuaries or freshwater but migrate to the ocean to reproduce. Based on the observed geographic distributions of newly hatched larvae, both American and European eels spawn in the tropical western Atlantic Ocean (the Sargasso Sea region). Thus both the morphological and distributional evidence raise questions as to whether these two "species" are indeed genetically distinct, and if so by how much, and what might account for their reproductive isolation.
In recent years, molecular genetic data have been gathered to address these and related issues [review in Avise (2003)]. More than a dozen allozyme loci were surveyed first [Williams and Koehn (1984) and references therein]. One of these proved to be nearly diagnostic for the two taxonomic species, with allele Mdh-2a present in a frequency of about 0.96 in all American eel populations, but only 0.10 in European samples. Allele frequencies at the other allozyme loci differed little or not at all between A. rostrata and A. anguilla. Later analyses of mitochondrial DNA (mtDNA) restriction sites (Avise et al. 1986) uncovered the first definitively diagnostic genetic characters for the two species: all specimens grouped into two continent-specific mtDNA clades separated by about 3% net sequence divergence (after correction for mean within-taxon variation). These matrilineal distinctions between American and European eels subsequently were confirmed by direct mtDNA sequence analyses [see references in Avise (2003)].
Another long-standing question stemming from the eels' catadromous life history is whether conspecific populations from different locales within a continent have come from a panmictic spawn. Available allozyme data barely dispute this possibility; apart from mild clinal variation at a few loci, sample allele frequencies otherwise were indistinguishable across locales. Williams et al. (1984) provisionally attributed the clines to spatial variation in intra-generational selection pressures rather than departures from random mating within a generation. Recently, in assays of nuclear microsatellite loci, Wirth and Bernatchez (2001) reported statistically significant (albeit low) levels of geographic population structure in A. anguilla sampled across this species' vast European range, a result that they attributed to nonrandom mating and restricted gene flow.
Finally, questions have arisen about the genetic status of Icelandic eels. Although generally considered A. anguilla, populations in Iceland have a frequency distribution of vertebral counts that is shifted somewhat toward A. rostrata and slightly lower in mean than populations from continental Europe. This observation, coupled with nonrandom associations between vertebral counts and allozyme and mitochondrial genetic markers (Avise et al. 1990; Williams et al. 1984), raised the possibility that some Icelandic eels may be of mixed (hybrid) ancestry between A. anguilla and A. rostrata. Additional diagnostic markers from the nuclear genome are needed to critically test this hypothesis.
Here we compare multiple population samples of both A. rostrata and A. anguilla with respect to microsatellite markers. These data could shed further light on genetic relationships between these taxa, and perhaps also help to address several of the other questions raised here.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The nuclear DNA samples employed (Table 1) were stored at -20°C for more than a decade. They had been extracted from eels collected in earlier mtDNA surveys of American, European, and Icelandic populations (Avise et al. 1986, 1990).
|
Six microsatellite loci developed by Wirth and Bernatchez (2001) were amplified in the Anguilla samples with fluorescently labeled primers. Polymerase chain reactions (PCRs) of 10 µl consisted of 1 µl DNA, 1 µM each of forward and reverse primers, 7.5 µM of each dNTP (Roche), 5 U Taq DNA polymerase (Promega), and MgCl2 concentrations as shown in Table 2, in 1x MgCl2-free PCR buffer (Promega). Cycling conditions were a 5 min denaturation at 95°C, 30 cycles of 1 min at 95°C, 1 min at annealing temperature (shown in Table 2), and 1 min elongation at 72°C, followed by an 8 min final elongation at 72°C. For fluorescent DNA fragment analysis, the reactions were pooled and electrophoresed on polyacrylamide gels using an ABI 377 automated DNA sequencer with a 500 size standard (Applied Biosystems). Fragment data were extracted with GeneScan and analyzed using Genotyper (Applied Biosystems).
|
The genotypic data were entered into Arlequin (Schneider et al. 2000) to estimate genetic distance values and conduct Hardy–Weinberg tests. Values of GST (a measure of population differentiation for multiallelic loci; Nei and Kumar 2000) and their standard errors from 1,000 bootstrap repetitions were calculated using Poptree (Takezaki 1998). Another genetic distance measure, Da (Nei et al. 1983), calculated using Dispan (Ohta 1993), served as the basis for generating a neighbor-joining (NJ) tree (Saitou and Nei 1987) to show relationships among population samples. The tree was midpoint rooted, and levels of confidence in tree nodes were calculated from 1,000 bootstrap replications.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Species' allelic frequencies at the six amplified loci are shown in Figure 1. The average total number of alleles per locus was 41 (range 30–57) and the mean observed heterozygosity per population was 0.82 (range 0.79–0.86), as shown in Table 2. Most of the loci conformed to Hardy–Weinberg expectations (HWE) at both the local-population and species levels. The only exceptions involved 16 of 78 tests (20%) that displayed a mild (P <.05) departure from HWE, but without a discernable pattern.
|
Table 3 summarizes the estimates of GST for population samples of A. rostrata and A. anguilla considered both separately and pooled. Values in all cases were low, but statistically significant. Figure 2 plots the observed ranges of genetic distances as measured by the Da metric in various subsets of populations within and between the two nominal Anguilla taxa. The largest average pairwise distance was between American and European groups. Less distance was observed between pairwise comparisons of American and Icelandic groups. Figure 3 shows a NJ tree for the 11 surveyed eel populations, as generated from a matrix of these genetic distance values. When midpoint-rooted, this tree marginally separates all populations of American eels from those sampled in Iceland and Europe.
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Geographic Variation Among Conspecific Populations
Allelic variation at all six microsatellite loci was high, but only weakly partitioned among the geographic locales of each species. Thus, although both A. anguilla and A. rostrata show statistically significant population-genetic differentiation, the GST values were remarkably low considering the vast geographic ranges surveyed; fully 94–97% of the total genetic variation occurred within rather than among population samples. Furthermore, for such highly polymorphic loci, any apparent population structure based on finite samples must be interpreted cautiously due to stochastic sampling errors in the allele frequency estimates and the strong likelihood that some rare alleles will be missed.
Our results are consistent with the low geographic differentiation reported in previous microsatellite assays of European eels (Wirth and Bernatchez 2001), and they extend such findings to the American species. Our study also generally agrees with the negligible population-genetic divergence reported in intraspecific allozyme surveys of North Atlantic eels, suggesting that any departures from apparent eel panmixia are modest at best (Williams and Koehn 1984). The relative paucity of spatial differentiation within either A. rostrata or A. anguilla on continent-wide scales is undoubtedly related to these animals' catadromous life history, which likely promotes a "concerted" genetic evolution of widely separated freshwater populations via the ongoing opportunity for interpopulational matings and range-wide juvenile dispersal in nearly every generation. These unusual peculiarities of eel natural history in the North Atlantic region appear to have greatly inhibited the accumulation of population-genetic differences over time, certainly in comparison to most other widely distributed species of freshwater and marine fishes that for reasons of geography and life history are far more spatially isolated or dispersal restricted (Avise 2000).
Genetic Differences Between American and European Eels
Allelic frequency histograms for the two species overlap extensively (Figure 1). Thus no fixed allelic differences were observed between A. anguilla and A. rostrata, and indeed, private alleles (those present in only one species, in this case) invariably were rare in the available samples. The only partial exception involved locus Ang101, where one allele (20) was present in a frequency of 0.16 in A. anguilla but was not observed in A. rostrata. Other private alleles, albeit of low frequency, may exist in these species, but would not be detectable without vastly larger sample sizes.
An accumulation of such subtle allele frequency differences across loci probably accounts for the fact that many of the composite pairwise genetic distances between A. anguilla and A. rostrata slightly exceed those among regional populations within either species (Figure 2). Likewise, such reasoning probably explains why conspecific populations of the two recognized species tend to group separately (albeit only marginally so by bootstrap criteria) in the NJ tree (Figure 3).
This NJ tree also places the Icelandic eels in an intermediate position between those from the American and European continents. Although this genetic depiction might be interpreted as consistent with the hypothesis that some Icelandic eels are of hybrid ancestry, the available microsatellite data are inadequate to critically test this possibility. Needed instead are multilocus batteries of nuclear-gene markers that clearly distinguish A. rostrata from continental A. anguilla. No microsatellite locus assayed here remotely approaches this condition.
In principle, the paucity of allelic differentiation between American and European eels at microsatellite loci could be due to any of several phenomena. First, substantial gene flow between A. rostrata and A. anguilla would cause their allelic profiles to coincide. However, such genetic exchange seems highly unlikely because these two taxa are distinct in mtDNA composition and nearly so at the nuclear Mdh locus (Avise et al. 1986; Williams and Koehn 1984). Second, the extensive overlap in microsatellite allelic profiles could be due to the retention of ancestral alleles in both sister species. However, this too seems improbable given the small effective evolutionary population sizes estimated for these species (Avise et al. 1986, 1988) and their clear genetic differences at mitochondrial and at least some other nuclear loci (Avise 2003).
Homoplasy as Noise
We hypothesize that a primary contributor to the microsatellite allelic overlap between these two species is extensive homoplasy (evolutionary convergence) associated with mutation-driven saturation effects. High microsatellite mutation rates (typically 10-3 in many taxa; Goldstein and Schlotterer 1999) and the stepwise nature of mismatch repair mutations (Ohta and Kimura 1973), coupled with the likelihood of physical or selective constraints at the upper end of the spectrum of allelic sizes and physical constraints at the lower end (Goldstein and Schlötterer 1999), imply that mutations to some allelic states probably occur independently and routinely in different populations (Taylor et al. 1999). If so, similar allelic frequency profiles (such as those in Figure 1) may not evidence either extensive recent genetic exchange or the retention of ancestral polymorphisms. Instead, they probably reflect, at least in part, homoplasy with regard to allelic size. This suggestion also raises a cautionary note that could apply to microsatellite allelic profiles that overlap among geographic populations within a species.
Thus, at microsatellite (or any other) loci with high mutation rates and numbers of qualitative allelic states constrained within narrow limits, many mutational alterations will merely shuffle alleles among the finite number of available slots between the boundaries. Such cryptic homoplasy can lead to significant underestimates of genetic divergence (Chapman et al. 1979; Taylor et al. 1999). The probability of homoplasy increases with divergence time (Estoup and Cornuet 1999), all else being equal, but the phenomenon also may be pervasive in large populations or those that have expanded historically. Furthermore, for populations or species that have gone through periodic or recent bottlenecks, the surviving alleles might be diagnostic for the populations in question, but nonetheless largely uninformative or even misinformative with regard to exact historical relationships.
Previous studies have identified homoplasy as an important factor contributing to overlaps in microsatellite allele frequencies among species. Judging from the long period of reproductive isolation between chimpanzees and humans, and the high mutation rates at microsatellite loci, Garza and Freimer (1996) deduced that many of the microsatellite alleles shared by these two primate species must not be identical by descent, but rather are the products of convergent evolution to common allelic states. In several species of Lake Malawi cichlids that began diverging about 700,000 years ago, detailed molecular-genetic appraisals of microsatellite alleles provided more direct evidence that homoplasy with respect to the number of repeat units was common (van Oppen et al. 2000; Zardoya et al. 1996).
At the intraspecific level also, cryptic homoplasy with regard to allelic size at the microsatellite can compromise historical analyses. Convergent evolution of this sort has been documented convincingly in, for example, horseshoe crabs (Orti et al. 1997), European rabbits (Queney et al. 2001), and humans (Garza and Freimer 1996). Such studies illustrate a growing realization that rapidly evolving loci with constrained numbers of alleles will not always be ideal for microevolutionary questions regarding historical population structure, gene flow, or the diagnosis of sibling species (Garza et al. 1995; Nauta and Weissing 1996).
|
Acknowledgments |
|---|
This work was supported by a University-Wide Fellowship from the University of Georgia (to J.E.M.) and by funds from the Pew Foundation (to J.C.A.). We thank Elizabeth Dakin, Mark Mackeiwicz, Brady Porter, and DeEtte Walker for useful comments on the manuscript.
|
Footnotes |
|---|
Corresponding Editor: Martin Tracey
Received November 8, 2002
Accepted March 23, 2003
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
-
Avise JC, 2000. Phylogeography: the history and formation of species. Cambridge, MA: Harvard University Press.
Avise JC, 2003. Catadromous eels of the North Atlantic: a review of molecular genetic findings relevant to natural history, population structure, speciation, and phylogeny. In: Eel biology (Aida K, ed). New York: Springer-Verlag.
Avise JC, Ball RM, Arnold J, 1988. Current versus historical population sizes in vertebrate species with high gene flow: a comparison based on mitochondrial DNA lineages and inbreeding theory for neutral mutations. Mol Biol Evol. 5:331-344.[Abstract]
Avise JC, Helfman GS, Saunders NC, Hales LS, 1986. Mitochondrial DNA differentiation in North Atlantic eels: population genetic consequences of an unusual life history pattern. Proc Natl Acad Sci USA. 83:4350-4354.
Avise JC, Nelson WS, Arnold J, Koehn RK, Williams GC, Thorsteinsson V, 1990. The evolutionary genetic status of Icelandic eels. Evolution. 44:1254-1262.[CrossRef]
Chapman RW, Avise JC, Asmussen MA, 1979. Character state restrictions and boundary conditions in the evolution of quantitative multistate characters. J Theor Biol. 80:51-64.[CrossRef][Web of Science][Medline]
Estoup AC, Cornuet J-M, 1999. Microsatellite evolution: inferences form population data. In: Microsatellites: evolution and applications (Goldstein DB and Schlotterer C, eds). Oxford: Oxford University Press; 49–65.
Garza J, Freimer N, 1996. Homoplasy for size at microsatellite loci in humans and chimpanzees. Genome Res. 6:211-217.
Garza J, Slatkin M, Freimer N, 1995. Microsatellite allele frequencies in humans and chimpanzees with implications for constraints on allele size. Mol Biol Evol. 12:594-603.[Abstract]
Goldstein DB, Schlötterer C, (eds), 1999. Microsatellites: evolution and applications. Oxford: Oxford University Press.
Nauta M, Weising F, 1996. Constraints on allele size at microsatellite loci: implications for genetic differentiation. Genetics. 143:1021-1032.[Abstract]
Nei M, Kumar S, 2000. Molecular evolution and phylogenetics. Oxford: Oxford University Press.
Nei M, Tajima F, Tateno Y, 1983. Accuracy of estimated phylogenetic trees from molecular data. J Mol Evol. 19:153-170.[CrossRef][Web of Science][Medline]
Ohta T, 1993. DISPAN: genetic distance and phylogenetic analysis. http://www.bio.psu.edu/People/Faculty/Nei/Lab/software.htm. Last modified August 2000.
Ohta T, Kimura M, 1973. The model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a genetic population. Genet Res. 22:201-204.[Web of Science][Medline]
Orti G, Pearse DE, Avise JC, 1997. Phylogenetic assessment of length variation at a microsatellite locus. Proc Natl Acad Sci USA. 94:10745-10749.
Queney G, Ferrand N, Weiss S, Mougel F, Monnerot M, 2001. Stationary distribution of microsatellite loci between divergent population groups of the European rabbit (Oryctolagus cuniculus). Mol Biol Evol. 18:2169-2178.
Saitou N, Nei M, 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 4:406-425.[Abstract]
Schneider S, Roessli D, Excofier L, 2000. Arlequin: a software for population genetics data analysis, version 2.000. Geneva, Switzerland: University of Geneva.
Takezaki N, 1998. POPTREE: Population tree construction. http://www.bio.psu.edu/People/Faculty/Nei/Lab/software.htm. Last modified August 2000.
Taylor JS, Sanny JSP, Breden F, 1999. Microsatellite allele size homoplasy in the guppy (Poecilia reticulata). J Mol Evol. 48:245-247.[Web of Science][Medline]
Tesch FW, 1977. The eel: biology and management of Anguillid eels. London: Chapman & Hall.
van Oppen MJH, Rico C, Turner GF, Hewitt GM, 2000. Extensive homoplasy, nonstepwise mutations, and shared ancestral polymorphisms at a complex microsatellite locus in Lake Malawi cichlids. Mol Biol Evol. 17:489-498.
Williams GC, Koehn RK, 1984. Population genetics of North Atlantic catadromous eels (Anguilla). In: Evolutionary genetics of fishes (Turner BJ, ed). New York: Plenum Press; 529–560.
Williams GC, Koehn RK, Thorsteinsson V, 1984. Icelandic eels: evidence for a single species of Anguilla in the North Atlantic. Copeia 1984:221–223.
Wirth T, Bernatchez L, 2001. Genetic evidence against panmixia in the European eels. Nature. 409:1037-1040.[CrossRef][Medline]
Zardoya R, Vollmer DM, Craddock C, Streelman T, Karl S, Meyer A, 1996. Evolutionary conservation of microsatellite flanking regions and their use in resolving the phylogeny of cichlid fishes (Pisces: Perciformes). Proc R Acad Lond B. 263:1589-1598.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. E. Maes and F. A. M. Volckaert Challenges for genetic research in European eel management ICES J. Mar. Sci., October 1, 2007; 64(7): 1463 - 1471. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||