Full Access

Habitat differentiation vs. isolation-by-distance: the genetic population structure of Elymus athericus in European salt marshes

Corresponding Author

A.-C. Bockelmann

Laboratory of Plant Ecology,

Max-Planck-Institute for Limnology, Department of Physiological Ecology, August-Thienemann-Strasse 2, 24306 Plön, Germany

Department of Genetics, Biological Centre, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands,

Dr A.-C. Bockelmann, Department of Physiological Ecology, Max-Planck-Institute for Limnology, August-Thienemann-Strasse 2, 24306 Plön, Germany. Fax: + 49 (0)4522 763310; E-mail: [email protected]Search for more papers by this author

T. B. H. Reusch,

T. B. H. Reusch

Max-Planck-Institute for Limnology, Department of Physiological Ecology, August-Thienemann-Strasse 2, 24306 Plön, Germany

Search for more papers by this author

R. Bijlsma,

R. Bijlsma

Department of Genetics, Biological Centre, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands,

Search for more papers by this author

J. P. Bakker,

J. P. Bakker

Laboratory of Plant Ecology,

Search for more papers by this author

A.-C. Bockelmann,

Corresponding Author

A.-C. Bockelmann

Laboratory of Plant Ecology,

Max-Planck-Institute for Limnology, Department of Physiological Ecology, August-Thienemann-Strasse 2, 24306 Plön, Germany

Department of Genetics, Biological Centre, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands,

T. B. H. Reusch,

T. B. H. Reusch

Max-Planck-Institute for Limnology, Department of Physiological Ecology, August-Thienemann-Strasse 2, 24306 Plön, Germany

Search for more papers by this author

R. Bijlsma,

R. Bijlsma

Department of Genetics, Biological Centre, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands,

Search for more papers by this author

J. P. Bakker,

J. P. Bakker

Laboratory of Plant Ecology,

Search for more papers by this author

First published: 16 January 2003

https://doi.org/10.1046/j.1365-294X.2003.01706.x

Citations: 83

Share a link

Email
Wechat
Bluesky

Abstract

We investigated genetic differentiation among populations of the clonal grass Elymus athericus, a common salt-marsh species occurring along the Wadden Sea coast of Europe. While E. athericus traditionally occurs in the high salt marsh, it recently also invaded lower parts of the marsh. In one of the first analyses of the genetic population structure in salt-marsh species, we were interested in population differentiation through isolation-by-distance, and among strongly divergent habitats (low and high marsh) in this wind- and water-dispersed species. High and low marsh habitats were sampled at six sites throughout the Wadden Sea. Based on reciprocal transplantation experiments conducted earlier revealing lower survival of foreign genotypes we predicted reduced gene flow among habitats. Accordingly, an analysis with polymorphic cross-species microsatellite primers revealed significant genetic differentiation between high and low marsh habitats already on a very small scale (< 100 m), while isolation-by-distance was present only on larger scales (60–443 km). In an analysis of molecular variance we found that 14% of the genetic variance could be explained by the differentiation between habitats, as compared to only 8.9% to geographical (isolation-by-distance) effects among six sites 2.5–443 km distant from each other. This suggests that markedly different selection regimes between these habitats, in particular intraspecific competition and herbivory, result in habitat adaptation and restricted gene flow over distances as small as 80 m. Hence, the genetic population structure of plant species can only be understood when considering geographical and selection-mediated restrictions to gene flow simultaneously.

Introduction

Genetic population differentiation between sites and habitats has been the rule rather than the exception in many genetic and demographic studies on plant populations (see Hamrick & Godt 1991 for a review). Geographic differentiation results from limited dispersal in many plant species and is conveniently summarized in isolation-by-distance models, first proposed and formalized by Wright (1946). Which specific life history traits of a given species on one hand, and which attributes of the environment on the other impede or promote dispersal (and gene flow) is relatively well understood (e.g. Slatkin 1985, 1987; Hansson 1991; Bohonak 1999; Cain et al. 2000). In plants, short ‘realized dispersal’ (i.e. the successful establishment of a plant at the site), in conjunction with the limited physical distances that propagules (spores or seeds) travel, often does not allow gene flow over larger distances (Cain et al. 2000). Consequently, increasing geographical distance will lead to increasing genetic isolation of populations.

In terrestrial plants, the main obstacles for long-distance dispersal are geographical barriers such as mountain ranges, lakes, rivers, or shorelines. Conversely, in aquatic plants, the latter three of these barriers are the dispersal pathways for hydrochorous dispersal. Not much is known about the quantity of hydrochorous dispersal and realized dispersal distances (but see Koutstaal et al. 1987). Salt-marsh plants are interesting for studying dispersal because they are intermediates between terrestrial and aquatic plants. They are terrestrial in terms of their morphology, physiology and ecology, while most species are dispersed by water (although additionally also by wind or animals, Bakker et al. 1985). In our study species, Elymus athericus (Link), Gould (sensuVan der Meijden 1990), the shed seeds will float in the water for various amounts of time and may be transported over long distances by tidal currents. Ten per cent of the propagules of our study species still float after 30 days of immersion in sea water (J. P. Bakker, unpublished data). We hypothesized that a common and rapidly colonizing species such as E. athericus should have effective dispersal and would therefore show little genetic population differentiation through isolation-by-distance. Elymus athericus is one of the most common plant species in the Wadden Sea salt marshes, sometimes covering more than 70% of the area (e.g. on Schiermonnikoog, the Netherlands, 53°30′, 6°10′). Since E. athericus has increased considerably in the past three decades, leading to a concomitant decline in species diversity in many marshes (Leendertse et al. 1997; Van Wijnen et al. 1997), we were also interested to infer colonization speed and abilities from marker data.

Genetic habitat differentiation has been neglected as a potential barrier to gene flow in many population genetic studies. Adaptation of populations to locally varying conditions has mostly been approached by transplantation or common environment experiments as reviewed by Endler (1986) and Van Andel (1998). Habitat adaptation becomes apparent through significant genotype by environment interaction or home site advantage. In contrast, DNA-based molecular markers have been primarily used for studying neutral population genetic processes such as drift and gene flow, although recently, there has been increasing interest in marker-based approaches of habitat differentiation in plants (Nevo et al. 1994; Li et al. 2000b; Engelen et al. 2001). The two habitats studied here are the low and the high salt marsh, which differ in numerous biotic and abiotic factors, such as shore height, inundation and vegetation composition (Adam 1993). In a parallel study we found that biotic factors, particularly the grazing pressure of herbivores (geese, hares) were more important for genetic habitat differentiation in E. athericus than abiotic factors, such as inundation frequency (Bockelmann et al. 2002). Reciprocal transplantation experiments with different life-history stages, both seedlings and adult ramets demonstrated the heritability of habitat-specific phenotypic traits. These differences in the phenotype are indicative of fitness differences among plants originating from high and low marsh in this species (Bockelmann 2002). We therefore predict that fitness differences in plants from different origin may reduce the gene flow from one population to another (Stanton et al. 1997; Vekemans & Lefebvre 1997). Such restrictions to gene flow will be reflected by genetic differentiation at molecular marker loci such as microsatellites, although the markers themselves are selectively neutral. We hypothesized that the described (Table 2) differences in the life history between the low and the high marsh populations have repeatedly resulted in different selection regimes restricting gene flow. As a consequence we expected to find genetic differentiation in neutral DNA markers at replicated sites across the Wadden Sea area.

Table 2. Phenotypic differences of Elymus athericus at different salt-marsh sites and habitats (‘High’ and ‘Low’ refer to the shore height in the salt marsh) discriminated by anova (full factorial model) and Tukey–Kramer post hoc test with P < 0.01, d.f. = degrees of freedom

Source of variation	d.f.	Density (shoots m⁻¹)*	Shoot length†	Spikelets per spike†	Ramet number†
Habitat	2	Low < High	Low < High	Low > High	Low > High
Site	5	SNK, HS < KD, 4^th, WD, WHV	KD, 4^th, WD, WHV < SNK, HS	KD, 4^th, WD, WHV < SNK, 4^th, WD, WHV < HS	WHV > SNK, 4^th > KD, WD, HS
Habitat × Site	2	not significant	not significant	not significant	not significant
Residual d.f. of anova		108	331	331	331
n		10	30	30	30

* April 1998.
† August 1998.
KD = Kobbeduin, 4^th = 4^th-creek, WD = Willemsduin, SNK = Sönke-Nissen-Koog Vorland, WHV = Westerhever, HS = Helmsand, Low = Low marsh, High = High marsh.

The genetic population structure of most European salt marsh plants, including E. athericus, is still unknown (but see Baumel et al. 2001). In this study, we directly investigated the realized exchange of propagules between populations of E. athericus with cross-species microsatellite markers. To assess the significance of habitat-correlated genetic differentiation on a broader scale, we replicated the comparison of low and high marsh habitats at several locations as advised by Endler (1986) and Prentice et al. (1995). This is the first study where the consistency of genetic habitat differences was tested over six sites on a landscape scale (see Prentice et al. 2000 for an experimental approach of small-scale habitat differentiation).

Materials and methods

Elymus athericus

Elymus athericus[Sea Couch, synonymous taxonomic names are, e.g. Elymus pycnanthus (Godron) Melderis, Elytrigia pungens (Pers.) Tutin, or Agropyron pycnanthum (Godron) Godron & Gren] is a tall, clonal grass native to salt marshes. It occurs along the North Sea coast between southern Denmark and the southern Netherlands but can also occur further south. It is wind pollinated, as are most Poaceae. The main dispersal units are spikelets, which bear up to five seeds (A.-C. Bockelmann, personal observation). Spikelets have no obvious dispersal mechanism. Breeding experiments demonstrated that this species is hexaploid (6x = 2n = 42; Stace 1995; Hess et al. 1998). Elymus athericus is usually outcrossing (Dewey 1983), but is also self-compatible (Bockelmann 2002). Selfing does not lead to reduction in seed viability (Bockelmann 2002). Depending on site and habitat, the population morphology and phenotype of E. athericus can be very different. Higher on the marsh, we mainly find continuous meadows, whereas populations are patchy lower on the marsh (Table 1). At our study sites, there were pronounced differences in the plant phenotypic traits. Mean shoot density in the high marsh is higher and shoots are longer. The production of spikelets per spike and new ramets is higher in the low marsh (Table 2). Differences in plant phenotypes were consistent across all six sites, i.e. there was no significant statistical interaction among ‘site’ and ‘habitat’.

Table 1. Environmental conditions at the six sites and in the two habitats (‘High’ and ‘Low’ refer to the shore height in the salt marsh) with the Elymus athericus study populations

Site	Habitat	Population morphology	Shore height* (cm + MHW)	Inundation frequency year⁻¹†	Age of marsh (years)‡	Grazing by livestock§	Resident herbivores*	N-mineralization (kg ha⁻¹ yr⁻¹)*¶
Kobbeduin	high	meadow	40–98	125	125	until 1973	none	no data
	low	meadow	22–38	270	6–?	none	hare, geese	no data
4^th-creek	high	meadow	83–128	20	86	until 1958	hare	no data
4^th-creek	low	patches	50–76	115	6–?	none	hare, geese	no data
Willemsduin	high	meadow	60–102	40	25	none	hare, rabbit	3.5¶
Willemsduin	low	patches	57–86	90	13	none	hare, geese	no data
Helmsand	high	meadow	65–81	no data	no data	until 1986		26.0**
Helmsand	low	patches				none	geese	no data
Westerhever	high	meadow	50–120	120	no data	until 1984		55.0††
	low	meadow	10–20	170		none	geese	32.0††
Sönke-Nissen-	high	meadow	34–55	96	56	until 1988		30.8‡‡
Koog-Vorland	low	patches		38	5–10	none	geese	66.6‡‡

Geese are Branta bernicla bernicla and B. leucopsis, the hare is Lepus europaeus and the rabbit is Orynctolagus cuniculus.
* A.-C. Bockelmann, unpublished data.
† Bockelmann et al. (2002).
‡ Olff et al. (1993).
§ J.P. Bakker, unpublished data.
¶ Van Wijnen et al. (1999).
** R. Neuhaus, unpublished data.
†† Bockelmann & Neuhaus (1999).
‡‡ Dahl (2001).

Study sites and sampling design

Developing a sampling design to study genetic population structure and gene flow in Elymus athericus was difficult because we had no information on clone size and dispersal distance prior to the present analysis. To ensure that we included the relevant spatial scales, at which genetic differentiation is partitioned in this species, we sampled E. athericus populations on three different scales. To test for differentiation by geographical distance we chose six sites along the Wadden Sea coast in Germany and on the Dutch Island of Schiermonnikoog, which were between 2.5 and 443 km distant from one another (Fig. 1). Additional differences between sites were the age of the marsh, the land-use history and the nitrogen mineralization rate (as a measure of nitrogen availability and productivity, Table 1). At each site we recognized two subpopulations, one higher on the shore, where E. athericus forms continuous almost monospecific meadows (hereafter named ‘high’). The second subpopulation was situated lower on the marsh, where E. athericus forms patches within vegetation dominated by Puccinellia maritima and Festuca rubra (hereafter named ‘low’). The locations at different shore heights on the elevational gradient are associated with pronounced differences in inundation frequency and other ecological factors (Table 1). In each subpopulation we sampled leaf material of three individual ramets in 10 blocks to include two levels of potential clone size. The sampling distance between tillers was 1.5–3 m; the distance between blocks was 10–30 m. The total sampling size was 360. The total sampling area (approximately 5000 m² per population) was chosen according to the geomorphology or the management regime at the site to ensure potentially unrestricted gene flow. The sampled leaf material was preserved through drying in tubes with silica gel until DNA extraction.

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

The study area, indicating three sites in Germany and three on the Dutch Island of Schiermonnikoog.

Cross-species modification of microsatellites, sequencing, and segregation patterns

We used microsatellite primer pairs originally developed for two other species of Poaceae, Elymus caninus (Sun et al. 1998a) and Triticum aestivum (Röder et al. 1998). Out of 26 primer pairs tested (including 12 primer pairs of E. alaskanus; Sun et al. 1998b), five amplified and showed consistently scorable results for E. athericus (Table 3). These were used in the present study.

Table 3. Microsatellites primer pairs of Triticum aestivum (WMS) and Elymus caninus (ECGA) used to assess genetic differentiation in 12 populations of Elymus athericus

Primer name	Repeats in original species	Repeats in Elymus athericus‡	Allele size range (bp)	No. of alleles	Fluorescent label
WMS2*	(CA)₁₈	(CA)₁₀§	180–295	9	6-FAM (blue)
WMS6*	(GA)₄₀	(GA)₁₅	140–164	11	NED (yellow)
WMS44*	(GA)₂₈	no data	115–144	4	6-FAM (blue)
ECGA22†	(CT)₂₇	(CT)₁₀	128–156	4	HEX (green)
ECGA89†	(GA)₁₇	(GA)₁₂G(GA)₄	197–207	4	HEX (green)

* Röder et al. (1998).
† Sun et al. (1998a).
‡ Complete sequences can be found in GenBank Accession No. AJ517215-25.
§ Not present in all individuals.

Alleles of four of the five primers derived from other species were subsequently characterized by cloning and sequencing according to standard protocols (one primer had to be left out for logistic reasons and time constraints also prevented the development of species-specific primers). Alleles in our study are operationally defined as amplified fragments originating from one of the five primer pairs.

Sequences contained microsatellites in both E. caninus- and Triticum aestivum-derived primers (Table 3). For the latter primers in particular, changes in the nonrepeat region suggested the multiple presence of a microsatellite locus, with subsequent loss of the microsatellite at some of the loci. The occurrence of additional, nonmicrosatellite fragments has also been reported by Stachel et al. (2000). Our data also indicate that null-alleles are present. Six polymerase chain reaction (PCR) products (i.e. corresponding to a hexaploid genome) were only found occasionally in primer WMS6, whereas the average number of alleles is two for WMS44, WMS2 and ECGA89 and three for WMS6 and ECGA22, prohibiting a co-dominant analysis of data. The size range of the amplified fragments in E. athericus differed from the species for which the primers were originally designed (Table 3).

We investigated qualitatively whether the amplified microsatellite fragments would segregate in offspring, using semi-artificial pollination. For this purpose, two groups of ramets from the field (distance 20 m) were dug out and two inflorescences, one of each ramet group, were placed together in a paper bag (mesh width < 0.2 mm, n = 4). We analysed the parent genotypes and five seeds of each spike. All fragments present in the offspring were also present in the putative parents (Fig. 2). In all cases, amplified fragments present in either parent were absent in at least one offspring. We could thus infer that the primers used are inherited in a Mendelian way.

Microsatellite genotyping

For microsatellite genotyping, genomic DNA was extracted from 5–10 mg of dry leaf tissue using the Qiagen plant extraction kit (Qiagen). One microlitre (containing ∼5–10 ng) of E. athericus DNA-extract was subjected to PCR with fluorescently labelled forward primers (and unlabelled backward primers). A 10-µl reaction contained 1-µl Promega Buffer B (10×, Promega Corp.), 200 µm of each dNTP, 1.5 µm MgCl₂, 5 µm of each primer, 0.1% (wt/vol) bovine serum albumin, and 0.25 units Promega Taq polymerase. Amplification conditions were: 3 min hot start at 94 °C, followed by 10 cycles of 1 min denaturation at 94 °C, 1 min annealing at 50–60 °C, and 30 s extension at 72 °C. The following 20–23 cycles had the same conditions except a 5 °C lower annealing temperature and an extension time of 20 min for the last cycle.

For visualization on an ABI-377 automated sequencer (Applied Biosystems), the PCR products of primer pairs WMS6, WMS44 and ECGA22, and of primer pairs WMS2 and ECGA89, respectively, were pooled and size scored against an internal lane standard. Positive and negative controls were always included to check the repeatability of the results.

Data analysis and statistics

Amplified fragments were analysed using the genescan and genotyper software packages (Applied Biosystems). It follows from the DNA sequence data (see above) that the calculation of single-locus based allele frequencies in E. athericus is not possible. Consequently, we chose to analyse fragment frequencies as multilocus fingerprints in which each allele was either scored present or absent. We only analysed alleles with a frequency > 10% in at least one population.

Genotypic diversity

The genotypic or clonal diversity of a population was calculated as P_D, the proportion of detectable genotypes (Ellstrand & Roose 1997), which gives the fraction of number of genets (= clones, g) detected and ramets (r) sampled (P_D = g r⁻¹). Differences in genotypic diversity between populations from different sites were tested by contingency table test among all six populations.

Analysis of molecular variance

Between-population differentiation was partitioned into habitat-correlated and geographical effects using analysis of molecular variance (amova; Weir & Cockerham 1984; Excoffier et al. 1992; for fingerprint data). An amova was also used to calculate pairwise population differences as Φ_ST. For this purpose, we have to assume that inbreeding coefficients are constant over populations, an assumption that is reasonable given the wind-pollination and likely predominance of outcrossing in E. athericus. Prior to all analyses on population differentiation, duplicate multilocus genotypes were excluded from each population (42 samples in total). We used the computer program tfpga (Miller 1997) for nested and unnested amova, and confirmed the results independently in an unnested amova in genetix (version 3.3, Belhir et al. 1996, 98). Because of the dominant/recessive inheritance of the data obtained by partially degenerate microsatellites in a hexaploid species, heterozygosities could not be calculated.

Genetic diversity

Genetic diversity within and among populations was calculated using Shannon's diversity measure (Lewontin 1972). This measure has the advantage that it does not assume Hardy–Weinberg equilibrium as amova does. This assumption may not be met in the salt-marsh environment because of nonrandom mating, e.g. because of the patchy population structure, or within clone inbreeding. This makes it a useful tool for comparative purposes in the analysis of genetic population structure next to amova. The Shannon index was calculated according to Pielou (1975) as:

H′ = −Σ p_i ln * p_i

where p_i is the frequency of the i^th microsatellite peak (= locus). We chose the natural logarithm to calculate H′, but any log base may be adopted (Magurran 1983; p. 35). We followed Chalmers et al. (1992) to detect within-population differences of H′ depending on the primer used by averaging across primers:

where P is the primer, n is the number of primers. inline image

was calculated for each locus separately and subsequently averaged over all loci of each primer. The overall diversity of all populations (the ‘species’, sp.) inline image

can be calculated as inline image

with p̄_s being the average frequency of peaks of one primer in all samples (318 plants in our study). It follows that inline image is the component of diversity within stands of one population and that is the component of diversity between populations (Chalmers et al. 1992).

Isolation-by-distance

We explored how much differentiation between population pairs can be explained by geographical distance. Because in E. athericus, long-distance dispersal of seeds is most likely by water (Bockelmann et al. 2002), the geographical distance between the populations was measured along the coastline (excluding bays such as the Dollard).

Apart from an overall correlation, we also considered high and low populations separately to test the hypotheses that founder effects would increase the variance in differentiation among low populations. In each data set, the relation of pairwise genetic distances (assessed as Φ_ST) with geographical distance was tested with a Mantel test (1967) using genetix.

Results

Genotypic and genetic diversity

Populations in different habitats did not differ in the number of distinguishable genotypes (d.f. = 5, χ² = 0.529, P = 0.99; Table 4). Across all populations the number of identical genotypes ranged from 0 to 17%. If there were any similar multilocus genotypes they occurred within blocks (within the three plants sampled 1–3 m distant). The genetic diversity in Elymus athericus populations (expressed by the Shannon index H′) ranged from 1.41 to 4.07 (Table 4). Twenty-nine per cent of the total genetic diversity was partitioned between populations inline image . Paired t-tests revealed no differences in genetic diversity between primers, habitats, or sites.

Table 4. Within population genetic and genotypic (clonal) diversity in 12 populations of the salt-marsh species Elymus athericus (n_population= 30) from six different sites along the Wadden Sea coast and two different habitats (‘High’ and ‘Low’ refer to the shore height in the salt marsh). Replicates of a genotype within a habitat were considered as members of the same genet and excluded from the analysis (n_final= 318)

Site	Habitat	Genotypic diversity P_D	Genetic diversity
Kobbeduin	High	1.00	2.79
Kobbeduin	Low	0.93	2.56
4^th-creek	High	0.83	2.28
4^th-creek	Low	0.75	1.79
Willemsduin	High	0.88	2.54
Willemsduin	Low	0.96	3.60
Helmsand	High	0.97	3.23
Helmsand	Low	1.00	1.41
Westerhever	High	0.96	3.51
Westerhever	Low	0.90	2.65
Sönke-Nissen-	High	0.96	4.07
Koog-Vorland	Low	0.89	3.36
Average			2.82
			4.10
	within		0.72
	between		0.29

Genotypic diversity = proportion of distinguishable genotypes, P_D, Genetic diversity = Shannon index for the genetic diversity
within population = genetic diversity calculated for
each population separately,= genetic diversity calculate
across all samples,= within population genetic
diversity,= between population genetic diversity.

Population differentiation between sites and habitats

We found significant genetic differentiation of amplified fragment frequencies between sites and habitats. A one-way amova revealed that 14% of variance was distributed among all 12 subpopulations. When partitioning geographical and habitat effects using hierarchical amova we found that 8.9% (d.f. = 5; 95% CI: 4.3–14.6%) of the variance could be explained by site and 14% (d.f. = 6; 95% CI: 8.4–21.9%) by the two chosen habitats (residual d.f. = 274). The differentiation between habitats was thus 64% higher than between sites despite the small geographical distance (maximally 80 m). The amount of genetic variance explained together by site (8.9%) and habitat (14%) in amova is approximately the same amount as found with the Shannon index (29%, see above).

Of all the amplified fragments 94% occurred in at least 75% of the high marsh populations. In the low marsh 88% of the fragments were present in at least 75% of the populations. None of the fragments occurred in less than 50% of the populations.

Pairwise population differentiation and isolation-by-distance

We calculated pairwise population differentiation in Φ_ST between the 66 pairs of populations in our study (Table 5, Fig. 3). The values ranged between 0.009 and 0.315, with a significant differentiation (P < 0.05) for 59 population pairs. Between habitats within one location, all but two comparisons (KD Low vs. KD High, WHV Low vs. WHV High, Table 5) showed significant fixation indices. These populations had a patchy morphology whereas the populations without differentiation between low and high marsh were continuous populations (Table 1).

Table 5. Genetic differentiation between pairs of populations of Elymus athericus expressed by the F_ST estimator Φ_ST (Excoffier et al. 1992), bold values are not significant

	KD low	4^th high	4^th low	WD high	WD low	HS high	HS low	SNK high	SNK low	WHV high	WHV low
KD high	0.040	0.073	0.111	0.061	0.175	0.310	0.118	0.190	0.096	0.213	0.153
KD low		0.077	0.092	0.074	0.177	0.321	0.127	0.218	0.154	0.211	0.172
4^th high			0.106	0.009	0.057	0.144	0.057	0.175	0.119	0.172	0.126
4^th low				0.132	0.190	0.315	0.141	0.280	0.185	0.267	0.214
WD high					0.073	0.192	0.069	0.158	0.112	0.167	0.142
WD low						0.081	0.106	0.197	0.176	0.220	0.203
HS high							0.171	0.295	0.294	0.307	0.266
HS low								0.146	0.044	0.122	0.082
SNK high									0.117	0.021	0.016
SNK low										0.145	0.089
WHV high											0.016

KD = Kobbeduin, 4^th = 4^th-creek, WD = Willemsduin, HS = Helmsand, SNK = Sönke-Nissen-Koog Vorland, WHV = Westerhever, Low = Low marsh, High = High marsh.

We found weak but significant isolation-by-distance through pairwise comparison of population pairs among all sampled locations. Pairwise Φ_ST values were significantly correlated with geographical distances. However, the coefficient of determination was low. (R² = 0.095, P_mantel = 0.04, 5000 permutations). When testing the isolation-by-distance among populations belonging to each habitat separately correlation was almost significant for all six high populations (P_mantel = 0.063, 5000 permutations, R² = 0.60) but clearly nonsignificant for all low ones (P_mantel = 0.4, 5000 permutations, R² = 0.004). At between-sites distances < 10 km, pairs of low marsh sites showed a higher differentiation than expected under the isolation-by-distance model (Fig. 3). At larger distances (between 60 and 443 km) genetic differentiation increased with distance. The average differentiation among habitats within sites of 14% (amova, see above) corresponds to a differentiation solely through geographical distance of approximately 150 km.

Surprisingly, the Helmsand population was more similar to the populations on Schiermonnikoog, although it is situated in a clayey mainland marsh in the northeastern Wadden Sea such as Sönke-Nissen-Koog-Vorland and Westerhever (Table 1, Fig. 3).

Discussion

The use of cross-species microsatellite primers compared to other molecular markers

The cross-species microsatellite primers successfully used in this study in Elymus athericus were adapted from Triticum aestivum and Elymus caninus. Trials with other plant species have had mixed results (e.g. Van Treuren et al. 1997; White & Powell 1997; Streiff et al. 1998; Reusch 2000; Van der Velde 2001). Our results are consistent with studies where primers of E. caninus have successfully been used for other Elymus species (Sun et al. 1998a; 1998b). The success might be based on the close relationship of species within the tribe Triticeae, as suggested in the light of difficult morphological discrimination and unsolved taxonomic classification (Dewey 1983).

Cross-species microsatellites might be an alternative to the costly and time-consuming development of species-specific microsatellites. Even if insertions and deletions of parts of the repeat region prohibit co-dominant analysis as in our case, species specificity, better repeatability (Gillet 1999) and interpretation (less complex allele pattern) make it superior to random priming techniques such as random amplified polymorphic DNA or amplified fragment length polymorphism. In E. athericus the application of traditional allozyme techniques is difficult. Using this technique, we analysed the same 12 populations as in this study and found that the results are flawed by a bad repeatability and temporal variation of allozyme pattern within genotypes (Bockelmann 2002). The data derived from allozyme analysis might have overestimated genetic variation, in particular between high and low marsh. The explanation may be physiological differences in enzyme activity on a small spatial and temporal scale and even within individuals (clones) as discussed earlier by Clark & Koehn (1991). Furthermore, co-dominant interpretation is also prohibited for allozyme data by the complex banding pattern of this hexaploid species. We propose that allozyme data should, if used at all, be interpreted phenotypically rather than genetically in this species. In other Elymus species allozymes are applicable but show lower levels of polymorphism than microsatellites (Sun et al. 1999; Sun et al. 2001).

Genetic differentiation and diversity in comparison with other Elymus species

The among-population differentiation ranged between 0.089 and 0.14 (F_ST estimator Φ_ST) in E. athericus. Much stronger genetic differentiation between populations has been reported from allozyme or random amplified polymorphic DNA analysis of other Elymus species (G_ST values ranging from 0.55 to 0.95). However, in contrast to the study species E. athericus, these taxa were at least predominantly, if not exclusively, self-pollinating [Elymus alaskanus, Diaz et al. 1999; Elymus fibrosus, Diaz et al. 2000; Elymus glaucus,Knapp & Rice (1996), all 4x, 2n = (28)]. Stronger population differentiation has been predicted for self-pollinating as opposed to outcrossing species (Hamrick & Godt 1991).

Geographic isolation and dispersal

In contrast to our expectations of strong genetic exchange by means of water currents, differentiation and partitioning of genetic diversity between populations was already significant on a scale of < 100 m. Geographic distances of more than 60 km led to weak but significant isolation-by-distance in Elymus athericus, with considerable variance introduced by habitat-correlated genetic differences (see below). If we assume that E. athericus seeds or spikelets can float for 30 days on the water (J. P. Bakker, unpublished results) and further assume a maximal water velocity of 3.7 km/h of the Wadden Sea tidal current at spring tide (Dietrich et al. 1975), the potential dispersal distance might be approximately 600 km within 1 week. This is much more than the average distances from one marsh to another. However, these dispersal distances are probably overestimates. Spikelets must first of all leave the creek system and reach the Wadden Sea because velocities within creeks are much lower (0.72 km/h, Winskowsky 1998) than in open water and not always orientated towards the tidal flats, as could be shown in two pilot studies on Schiermonnikoog (A.-C. Bockelmann and T. Wels, unpublished results). Moreover, tidal currents change direction with every tidal cycle, and wind direction and wind speed may be more important in determining dispersal direction (Koutstaal et al. 1987). Nevertheless, potentially substantial gene flow between the geographical sites seems to be possible.

Rather than limits to propagule dispersal, demographic processes such as low germination rates of E. athericus can probably explain the difference between potential and realized dispersal distances. When addressing the differentiation between sites in a study on Schiermonnikoog, Bockelmann (2002) could show that germination is probably the most critical life history stage in E. athericus. Rare successful long-distance dispersal events could lead to strong founder effects in new populations. This could explain that the differentiation between sites is stronger than expected based on the potential dispersal rates. That founding events promote genetic divergence in E. athericus is also apparent from the surprisingly high genetic differentiation among young, newly founded E. athericus populations in low areas that is much stronger than would be expected through isolation-by-distance.

Habitat differentiation as barrier to gene flow

Although distances among high and low marsh habitats are short, we observed repeatedly even more differentiation between habitats within sites than between sites up to 443 km away from each other. As such, our results confirm for the first time that genetic habitat differentiation over short geographical distances (maximally 80 m) can be consistent over several sites on a landscape scale. The hierarchical sampling allowed the distinction of habitat differentiation on one hand and isolation-by-distance on the other. The results imply consistent differences in biotic or abiotic factors between the low and high habitat at all six studied sites that result in different local selection regimes for E. athericus.

Three mechanisms could explain the differentiation of microsatellite patterns between habitats through selection. First, there could be direct selection on microsatellite loci. This is very unlikely because in eukaryotes, microsatellites are thought to be situated predominantly in noncoding introns and contain no transcribed genetic information (Hancock 1999; but see Li et al. 2000a, 2000b for a discussion of possible direct selection on microsatellite loci driving habitat divergence in wild emmer wheat Triticum dicoccoides). Moreover, direct selection would lead to consistent frequency changes of specific PCR fragments of the loci under selection. While this has been demonstrated for habitat-correlated differences at allozyme loci (Prentice et al. 1995; Schmidt et al. 2000), it is probably not the case here. In E. athericus 91% of the alleles are present in at least 75% of the populations, while no PCR fragment occurs in less than 50% of the populations. The commonness of fragments can serve as an argument against the second possible mechanism, genetic linkage of microsatellite loci with functional genes.

Instead, we propose that genetic differentiation of plant traits, as a result of selection on the life-history traits of E. athericus, resulted in limited gene flow. Ecological habitat adaptations of life cycle stages, such as the establishment of seedlings, have been shown experimentally to limit realized gene flow (Bockelmann 2002). In parallel transplant experiments, we have shown that resistance to herbivores, and competitive ability towards neighbouring conspecific plants differs among genotypes from high and low habitats. We found that seedlings with parents from the high marsh were less adapted to stronger herbivory present in the low marsh, while seedlings with low marsh parents were less able to survive competition in the high marsh. This corresponds with the intensity of herbivory, respectively, the strength of competition in these habitats (Van de Koppel et al. 1996; Bockelmann 2002). Consequently, the establishment of seedlings from nonlocal populations without herbivory or competition in E. athericus is limited (Bockelmann 2002). The history of domestic livestock grazing, inundation frequency and nitrogen mineralization (Table 1) differ markedly not only between habitats but also between sites and may further contribute to genetic divergence in the phenology of E. athericus populations. It may result in differences in the timing of reproduction, the period of flowering, seed shedding or the timing of germination.

In summary, we have shown that neutral molecular markers repeatedly revealed restricted gene flow between low and high salt-marsh populations despite high potential propagule transport by water and wind. Hence, the genetic population structure of plant species can only be understood when considering geographical and selection-mediated restrictions to gene flow simultaneously.

Acknowledgements

We want to thank two anonymous reviewers for comments on the manuscript and Silke Carstensen, Catherine Schmuck and Roos Veeneklaas for their technical assistance. A.-C.B. is grateful for the support of the Max-Planck-Society through W. Lampert. A.-C.B. was funded by the Hans-Böckler-Stiftung and an Ubbo-Emmius grant of Groningen University.

References

Adam P (1993) Saltmarsh Ecology. Cambridge University Press, Melbourne.
Google Scholar
Bakker JP, Dijkstra M, Russchen PT (1985) Dispersal, germination and seed establishment of halophytes and glycophytes on a grazed and abandoned salt marsh gradient. New Phytologist, 101, 291–308.
10.1111/j.1469-8137.1985.tb02836.x
Web of Science® Google Scholar
Bakker JP, Leeuw J, Dijkema KS, Leendertse PC, Prins HHT, Rozema J (1993) Salt marshes along the coast of The Netherlands. Hydrobiology, 265, 73–95.
10.1007/BF00007263
Web of Science® Google Scholar
Baumel A, Ainouche ML, Levasseur JE (2001) Molecular investigations in populations of Spartina anglica C.E. Hubbard (Poaceae) invading coastal Brittany (France). Molecular Ecology, 10, 1689–1701.
10.1046/j.1365-294X.2001.01299.x
CAS PubMed Web of Science® Google Scholar
Belhir K, Borsa P, Goudet J, Chikhi L, Bonhomme F (1998) Genetix, logiciel sous WindowsTM pour la génétique des population. Laboratoire Génome et Population, CNRS UPR 9060, Montpellier.
Google Scholar
Bockelmann AC (2002) Ordinary and successful — the invasion of Elymus athericus in European salt marshes. PhD Thesis, University of Groningen, Groningen.
Google Scholar
Bockelmann AC, Neuhaus R (1999) Competitive exclusion of Elymus athericus from a high-stress habitat in a European salt marsh. Journal of Ecology, 87, 503–513.
10.1046/j.1365-2745.1999.00368.x
Web of Science® Google Scholar
Bockelmann AC, Bakker JP, Neuhaus R, Lage J (2002) The relation between vegetation zonation, elevation and inundation frequency in a salt marsh. Aquatic Botany, 73, 211–221.
10.1016/S0304-3770(02)00022-0
Web of Science® Google Scholar
Bohonak AJ (1999) Dispersal, gene flow, and population structure. Quarterly Review of Biology, 74, 21–45.
10.1086/392950
CAS PubMed Web of Science® Google Scholar
Cain ML, Milligan BG, Strand AE (2000) Long-distance seed dispersal in plant populations. American Journal of Botany, 87, 1218–1227.
10.2307/2656714
Web of Science® Google Scholar
Chalmers KJ, Waugh R, Sprent JI, Simons AJ, Powell W (1992) Detection of genetic variation between and within populations of Gliricidia sepium and G. maculata using RAPD markers. Heredity, 69, 465–472.
10.1038/hdy.1992.151
PubMed Web of Science® Google Scholar
Clark AG, Koehn RK (1991) Enzymes and adaptation. Genes in Ecology (eds RJ Berry, TJ Crawford, GM Hewitt), pp. 193–228. Blackwell, Oxford.
Google Scholar
Dahl M (2001) Mögliche Effekte eines Klimawandels auf die Stickstoffnettomineralisation in Vorlandsalzwiesen, PhD Thesis, Christian-Albrechts-University, Kiel.
Google Scholar
Dewey DR (1983) Historical and current taxonomic perspectives of Agropyron, Elymus, and related genera. Crop Science, 23, 637–642.
10.2135/cropsci1983.0011183X002300040009x
Web of Science® Google Scholar
Diaz O, Salomon B, Von Bothmer R (1999) Genetic variation and differentiation in Nordic populations of Elymus alaskanus (Scrib. ex Merr.) Löve (Poaceae). Theoretical and Applied Genetics, 99, 210–217.
10.1007/s001220051227
Web of Science® Google Scholar
Diaz O, Sun GL, Salomon B, Von Bothmer R (2000) Levels and distribution of allozyme and RAPD variation in populations of Elymus fibrosus (Schrenk) Tzvel. (Poaceae). Genetic Resources and Crop Evolution, 47, 11–24.
10.1023/A:1008719119756
Web of Science® Google Scholar
Dietrich G, Kalle K, Krauss W, Siedler G (1975) Allgemeine Meereskunde Gebrüder Bornträger, Berlin.
Google Scholar
Ellstrand NC, Roose ML (1997) Patterns of genotypic diversity in clonal plant species. American Journal of Botany, 74, 123–131.
Web of Science® Google Scholar
Endler JA (1986) Natural Selection in the Wild. Monographs in Population Biology (ed. RM May), Vol. 21. Princeton University Press, Princeton.
Google Scholar
Engelen AH, Olsen JL, Breeman A, Stam WT (2001) Genetic differentiation in Sargassum polyceratium (Fucales, Phaeophyseae) around the island of Curacao (Netherlands Antilles). Marine Biology, 139, 267–277.
10.1007/s002270100586
CAS Web of Science® Google Scholar
Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: am application to human mitochondrial DNA restriction data. Genetics, 131, 479–491.
CAS PubMed Web of Science® Google Scholar
Gillet EM (1999) Which DNA marker for which purpose? Final Compendium of the Research Project Development, optimisation and validation of molecular tools for assessment of biodiversity in forest trees in the European Union DGXII Biotechnology FW Research Programme Molecular Tools for Biodiversity. Http://webdoc.sub.gwdg.de/ebook/y/1999/whichmarker/index.htm.
Google Scholar
Hamrick JL, Godt MJW (1991) Allozyme diversity in plant species. In: Plant Population Genetics, Breeding, and Genetic Resources (eds HD Brown, MT Clegg, AL Kahler, BS Weir), pp. 43–63. Sinauer Associates, Sunderland.
Google Scholar
Hancock JM (1999) Microsatellites and other simple sequences: genomic context and mutational mechanisms. In: Microsatellites, Evolution and Application (eds DB Goldstein, C Schlötterer), pp. 1–9. Oxford University Press, Oxford.
Google Scholar
Hansson L (1991) Dispersal and connectivity in metapopulations. Biological Journal of the Linnean Society, 42, 89–103.
10.1111/j.1095-8312.1991.tb00553.x
Web of Science® Google Scholar
Hess HE, Landolt E, Hirzel R (1998) Flora der Schweiz und Angrenzender Gebiete, Band 1. Pteridophyta bis Caryophyllaceae. Birkhäuser. Basel.
Google Scholar
Knapp EE, Rice KJ (1996) Genetic structure and gene flow in Elymus glaucus (blue wildrye): implications for native grassland restoration. Restoration Ecology, 4, 1–10.
10.1111/j.1526-100X.1996.tb00101.x
Web of Science® Google Scholar
Koutstaal BP, Markusse MM, De Munck W (1987) Aspects of seed dispersal by tidal movements. In: Vegetation Between Land and Sea (eds AHL Huiskes, CWPM Blom, J Rozema), pp. 226–233. Dr Dr W. Junk, Dordrecht.
10.1007/978-94-009-4065-9_18
Google Scholar
Leendertse PC, Rozen AJM, Rozema J (1997) Long-term changes (1953–90) in the salt marsh vegetation at the Boschplaat on Terschelling in relation to sedimentation and flooding. Plant Ecology, 132, 49–58.
10.1023/A:1009795002076
Web of Science® Google Scholar
Lewontin RC (1972) The appointments of human diversity. Evolutionary Biology, 6, 381–398.
Google Scholar
Li Y-C, Röder MS, Fahima T et al. (2000a) Natural selection causing microsatellite divergence in wild emmer wheat at the ecologically variable microsite at Ammiad, Israel. Theoretical and Applied Genetics, 100, 985–999.
10.1007/s001220051380
Web of Science® Google Scholar
Li Y-C, Fahima T, Peng JH et al. (2000b) Edaphic microsatellite DNS divergence in wild emmer wheat Triticum dicoccoides, at a microsite: Tabigha, Israel. Theoretical and Applied Genetics, 101, 1029–1038.
10.1007/s001220051577
CAS Web of Science® Google Scholar
Magurran AE (1983) Ecological Diversity and its Measurement Croom-Helm, London.
Google Scholar
Mantel NA (1967) The detection of disease clustering and a generalized regression approach. Cancer Research, 27, 209–220.
10.1111/j.1095-8312.1998.tb01520.x
CAS PubMed Web of Science® Google Scholar
Miller MP (1997) Tools For Population Genetic Analysis, a Windows TM program for the analysis of allozyme and molecular population genetic data. Biology Department, P.O. Box 871501, Arizona State University, Tempe.
Google Scholar
Nevo E, Krugman T, Beiles A (1994) Edaphic natural selection of allozyme polymorphisms in Aegilops peregrina at a Galilee microsite in Israel. Heredity, 72, 109–112.
10.1038/hdy.1994.16
CAS Web of Science® Google Scholar
Olff H, Leeuw J, Bakker JP, Platerink RJ, Van Wijnen HJ, De Munck W (1993) Nitrogen accumulation, vegetation succession and geese herbivory during salt marsh formation on the Dutch island of Schiermonnikoog. Journal of Ecology, 81, 693–706.
10.2307/2261667
Web of Science® Google Scholar
Pielou (1975) Ecological Diversity. Wiley, New York.
Google Scholar
Prentice HC, Lönn M, Lefkovitch LP, Runyeon H (1995) Associations between allele frequencies in Festuca ovina and habitat variation in the alvar grasslands on the Baltic island of Öland. Journal of Ecology, 83, 391–402.
10.2307/2261593
Web of Science® Google Scholar
Prentice HC, Lönn M, Lager H, Rosen E, Van der Maarel E (2000) Changes in allozyme frequencies in Festuca ovina populations after a 9-year nutrient/water experiment. Journal of Ecology, 88, 331–347.
10.1046/j.1365-2745.2000.00454.x
CAS Web of Science® Google Scholar
Reusch TBH (2000) Five microsatellite loci in eelgrass Zostera marina and a test of cross-species amplification in Z. noltii and IX. japonica. Molecular Ecology, 9, 365–378.
10.1046/j.1365-294x.2000.00874-4.x
PubMed Google Scholar
Röder MS, Korzun V, Wendehake K et al. (1998) A Microsatellite map of Wheat. Genetics, 149, 2007–2023.
10.1093/genetics/149.4.2007
CAS PubMed Web of Science® Google Scholar
Schmidt PS, Bertness MD, Rand DM (2000) Environmental heterogeneity and balancing selection in the acorn barnacle Semibalanus balanoides. Proceedings of the Royal Society London B, 267, 379–384.
10.1098/rspb.2000.1012
CAS PubMed Web of Science® Google Scholar
Slatkin M (1985) Gene flow in natural populations. Annual Reviews in Ecology and Systematics, 16, 393–430.
10.1146/annurev.es.16.110185.002141
Web of Science® Google Scholar
Slatkin M (1987) Gene flow and the geographic structure of natural populations. Science, 236, 787–792.
10.1126/science.3576198
CAS PubMed Web of Science® Google Scholar
Stace CA (1995) Flora of the British Isles. Cambridge University Press, Cambridge.
Google Scholar
Stachel M, Lelley T, Grausgruber H, Vollmann J (2000) Application of microsatellites in wheat (Triticum aestivum L.) for studying genetic differentiation caused by selection for adaption and use. Theoretical and Applied Genetics, 100, 242–248.
10.1007/s001220050032
Web of Science® Google Scholar
Stanton ML, Galen C, Shore J (1997) Population structure along a steep environmental gradient: consequences of flowering time and habitat variation in the snow buttercup, Ranunculus adoneus. Evolution, 51, 79–94.
10.1111/j.1558-5646.1997.tb02390.x
CAS PubMed Web of Science® Google Scholar
Streiff R, Labbe T, Bacilieri R, Steinkellner H, Glössl J, Kremer A (1998) Within-population genetic structure in Quercus robur L. and Quercus petraea (Matt.) Liebl. assessed with isozymes and microsatellites. Molecular Ecology, 7, 317–328.
10.1046/j.1365-294X.1998.00360.x
Web of Science® Google Scholar
Sun GL, Salomon B, Von Bothmer R (1998a) Characterization and analysis of microsatellite loci in Elymus caninus. Theoretical and Applied Genetics, 96, 676–682.
10.1007/s001220050788
CAS Web of Science® Google Scholar
Sun GL, Salomon B, Von Bothmer R (1998b) Characterization of microsatellite loci from Elymus alaskanus and length polymorphism in several Elymus species (Triticeae). Genome, 41, 455–463.
10.1139/gen-41-3-455
CAS PubMed Web of Science® Google Scholar
Sun GL, Diaz O, Salomon B, Von Bothmer R (1999) Genetic diversity in Elymus caninus as revealed by isozyme, RAPD, and microsatellite markers. Genome, 42, 420–431.
CAS PubMed Web of Science® Google Scholar
Sun GL, Diaz O, Salomon B, Von Bothmer R (2001) Genetic diversity and structure in a natural Elymus caninus population from Denmark based on microsatellite and isozyme analysis. Plant Systematics and Evolution, 227, 235–244.
10.1007/s006060170051
CAS Web of Science® Google Scholar
Van Andel J (1998) Intraspecific variability in the context of ecological restoration projects. Perspectives in Plant Ecology, Evolution and Systematics, 1, 221–237.
10.1078/1433-8319-00060
Google Scholar
Van de Koppel J, Huisman J, Van der Wal R, Olff H (1996) Patterns of herbivory along a productivity gradient: an empirical and theoretical investigation. Ecology, 77, 736–745.
10.2307/2265498
Web of Science® Google Scholar
Van der Meijden R (1990) Heukel's Flora Van Nederland, 21st edn. Wolters-Noordhoff, Groningen.
Google Scholar
Van der Velde M, Bijlsma (2001) Genetic evidence for the allodiploid origin of the moss species Polytrichum longisetum. Plant Biology, 3, 379–385.
10.1055/s-2001-16453
CAS Web of Science® Google Scholar
Van Treuren R, Kuittinen H, Kärkäinen K, Baena-Gonzalez E, Savolainen O (1997) Evolution of microsatellites in Arabis petraea and Arabis lyrata, outcrossing relatives of Arabidopsis thaliana. Molecular Biology and Evolution, 14, 220–229.
10.1093/oxfordjournals.molbev.a025758
CAS PubMed Web of Science® Google Scholar
Van Wijnen HJ, Bakker JP, De Vries Y (1997) Twenty years of salt marsh succession on the coastal barrier island of Schiermonnikoog (The Netherlands). Journal of Coastal Conservation, 3, 9–18.
10.1007/BF02908174
Google Scholar
Van Wijnen HJ, Hazekamp A, Van Hooff E, Bakker JP (1999) The impact of herbivores on nitrogen mineralization rate: consequences for salt-marsh succession. Oecologia, 118, 225–231.
10.1007/s004420050722
PubMed Web of Science® Google Scholar
Vekemans X, Lefebrre C (1997) On the evolution of heavy-metal tolerant populations in Armeria maritima: evidence from allozyme variation and reproductive barriers. Journal of Evolutionary Biology, 10, 175–191.
10.1046/j.1420-9101.1997.10020175.x
Web of Science® Google Scholar
Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358–1360.
10.1111/j.1558-5646.1984.tb05657.x
CAS PubMed Web of Science® Google Scholar
White G, Powell G (1997) Cross-species amplification of SSR loci in the Meliaceae family. Molecular Ecology, 6, 1195–1197.
10.1046/j.1365-294X.1997.00297.x
CAS Web of Science® Google Scholar
Winskowsky U (1998) Strömungen im nordfriesischen Wattenmeer. In: Umweltatlas Wattenmeer, Band I: Nordfriesisches und Dithmarscher Wattenmeer (ed. Nationalpark Schleswig-Holsteinisches Wattenmeer UB), pp. 48–49. Clausen & Bosse, Leck.
Google Scholar
Wright S (1946) Isolation by distance under diverse systems of mating. Genetics, 31, 39–59.
10.1093/genetics/31.1.39
CAS PubMed Web of Science® Google Scholar

Anna-Christina Bockelmann was a PhD student at the Centre for Ecological and Evolutionary Studies (Biology Centre, University of Groningen, the Netherlands) supervised by J. P. Bakker and R. Bijlsma. The molecular work in this study was carried out in co-operation with T. B. H. Reusch from the Max-Planck-Institute for Limnology in Plön, Germany. The aim of A.-C.B.'s PhD thesis was the analysis of the invasion of Elymus athericus in European salt marshes from an ecological and evolutionary perspective.

Citing Literature

Volume12, Issue2

February 2003

Pages 505-515

Habitat differentiation vs. isolation-by-distance: the genetic population structure of Elymus athericus in European salt marshes

Abstract

Introduction