Reprint of “Is the Genetic Structure of Mediterranean Ruppia Shaped By

Aquatic Botany 115 (2014) 45–53

Contents lists available at ScienceDirect

Aquatic Botany

jou rnal homepage: www.elsevier.com/locate/aquabot

Reprint of “Is the genetic structure of Mediterranean Ruppia shaped

ଝ

by bird-mediated dispersal or sea currents?”

∗

Ludwig Triest , Tim Sierens

Research Group ‘Plant Biology and Nature Management’, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: In the European part of the Mediterranean at least 15 cpDNA haplotypes of Ruppia can be distinguished

Received 6 July 2010

and characterized the West basin as a diversity hotspot. Ruppia cirrhosa shows a West–East differentia-

Received in revised form 16 June 2011

tion and clear isolation-by-distance between each basin. We investigated whether the maternal cpDNA

Accepted 23 September 2011

differentiation between and within subbasins of the Mediterranean could shed light on distribution and

Available online 24 February 2014

dispersal phenomena of a morphological variable species complex. Complementary nuclear ITS markers

showed three variants and allowed to detect hybrids with Ruppia maritima. Haplotypes differed sig-

Keywords:

niﬁcantly in leaf and fruit features for Ruppia drepanensis. Haplotypes A, D and E had numerous seeds

cpDNA microsatellites

trnH-psbA whereas haplotypes B and C were mostly vegetative. The scattered distribution of rare haplotypes argued

rbcL for occasional dispersal at long distances. However, birds as vectors of maternal cpDNA markers did not

ITS homogenize the genetic structure but it showed the presence of scattered isolated haplotypes reﬂect-

Ruppia ing a thin tail of long distance dispersal events. We observed a strong maternal isolation-by-distance

Seagrass between subbasins of the West basin and within the Balearic subbasin. It was found paradoxal that the

Genetic diversity

most continuous widespread haplotype B also had lowest number of fruits. Sea currents are discussed

Dispersal

as a potential dispersal vector at broad geographic scale for the most marine haplotype B variants of R.

cirrhosa, hereby resembling other seagrasses.

1. Introduction Ruppia occurs in a wide variety of coastal lagoon and conti-

nental brackish to saltwater habitats. The morphological plasticity,

The genus Ruppia has a cosmopolitan, but discontinuous dis- adaptation to temporal or permanent habitats and especially tradi-

tribution and is found on all continents, including many isolated tions in national ﬂora publications or regional wetland inventories

islands from tropical to subarctic regions (Green and Short, 2003). still create much confusion in naming Ruppia species. As a conse-

Ruppia maritima L. is the most widely distributed species of this quence no reliable distribution range maps could be provided as

truly global seagrass genus (Short et al., 2007). In the Mediter- was done for all other seagrass species (Short et al., 2007). The eco-

ranean region three taxa are recognised, namely R. maritima, typic and genotypic variation at population level partly remains not

Ruppia cirrhosa (Petagna) Grande and Ruppia drepanensis Tineo, understood (Den Hartog and Kuo, 2006) although this type of infor-

the latter as an inland ecotype of the SW Mediterranean (also mation is essential for seagrass conservation genetics (Waycott

as variety R. cirrhosa (Petagna) Grande var. drepanensis (Tineo) et al., 2006). Therefore, a direct comparison of the putative diag-

Symoens). Morphological studies (Aedo and Fernandez-Casado, nostic features between distinct cpDNA haplotypes using the same

1988; Cirujano and Garcia-Murillo, 1990), cytotaxonomical inves- populations will give more clarity on the morphological variability,

tigations (Cirujano, 1986; Talavera et al., 1993; Van Vierssen et al., distinctiveness and distribution of Ruppia taxa.

1981), isozyme polymorphisms (Triest and Symoens, 1991) and Chloroplast sequences generally revealed very low variability

chloroplast DNA sequence analyses (Triest and Sierens, 2010) con- in most aquatic plant and seagrass populations (Mader et al., 1998;

ﬁrmed the Mediterranean Ruppia diversity. Talbot et al., 2004; Triest et al., 2007; Koga et al., 2008; Provan

et al., 2008; Tan et al., 2008) unlike the polymorphic Ruppia (Triest

and Sierens, 2009, 2010; Ito et al., 2010). In a previous study of 53

water bodies across the European part of the Mediterranean, 15

DOI of original article: http://dx.doi.org/10.1016/j.aquabot.2011.09.009.

haplotypes were revealed and showed a much higher nucleotide

ଝ

This article is a reprint of a previously published article. For citation purposes,

diversity of cpDNA in the Western than in the Eastern basin. This

please use the original publication details “Aquatic Botany” 104 (2013) 176–184.

∗ hotspot of diversity caused an overall gradient and isolation-by-

Corresponding author. Tel.: +32 02 629 34 21; fax: +32 02 629 34 13.

E-mail address: [email protected] (L. Triest). distance (IBD) at basin level and was stronger between both basins

46 L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53

than within them (Triest and Sierens, 2010). A signiﬁcant IBD was Accession numbers are listed for ITS1 (JN113280–JN113282) and

observed within the West basin but not in the East basin. The ITS2 (JN113283–JN113285), whereas AJ012292 and FJ495523 were

latter had too low variability and thus could not be considered previously listed for Ruppia.

in further IBD tests. Genetic patterns of Mediterranean marine

plant and animal populations often display a West–East differen- 2.2. DNA extraction, ampliﬁcation and sequencing

tiation because the connection between both basins was narrower

between Sicily and Tunisia during the periods of Pleistocene and Genomic DNA extractions and ampliﬁcation for three cpSSR

Quaternary glacial maxima and because of the sea currents circula- primer pairs (Ccmp 2, Ccmp3 and Ccmp 10), a non-coding region

tion patterns. A barrier to gene ﬂow between the East and West (trnH-psbA) and a partially coding (rbcL) region were as in Triest

Mediterranean was suggested for Posidonia oceanica (L.) Delile and Sierens (2010).

(Arnaud-Haond et al., 2007; Serra et al., 2010) and explained as PCR and direct amplicon sequencing of nuclear ITS1 and

vicariance. The Ruppia cpDNA haplotypes of Eastern basin popu- ITS2 spacers (White et al., 1990) was done with primers

lations represent a very small subset of those from the Western for the ITS1 spacer (ITS1: TCCGTAGGTGAACCTGCGG and ITS2:

basin, thereby suggesting a historical eastward dispersal of a single GCTGCGTTCTTCATCGATGC) and the ITS2 spacer (ITS3: GCATCGAT-

R. cirrhosa haplotype over long distances (Triest and Sierens, 2010). GAAGAACGCAGC and ITS4: TCCTCCGCTTATTGATATGC) using a PCR

There is ample evidence that birds are the main vec- reaction of 25 ␮l containing 1× PCR buffer, 200 ␮M of each dNTP,

tor for dispersal of R. maritima seeds at short distances 3 mM MgCl2, 200 nM of each primer, 0.5 ␮l of BSA (10 ␮g/␮l) and

◦

(Figuerola and Green, 2002; Figuerola et al., 2002; Charalambidou 1 unit of Taq polymerase. Reaction: 95 C for 4 min followed by 35

◦ ◦ ◦

and Santamaria, 2005). Long-distance dispersal (LDD) is far more cycles of 95 C for 1 min, 54 C for 1 min, 72 C for 2 min and a ﬁnal

difﬁcult to estimate as it is a rare event in a thin-tailed dis- extension step of 5 min.

persal kernel (Nathan et al., 2008). The role of birds in LDD was

argued (Figuerola and Green, 2002; Figuerola et al., 2002), crit- 2.3. Morphology and habitat

ically reviewed (Clausen et al., 2002) and partially answered in

favour of birds as effective vectors (Charalambidou and Santamaria, For each haplotype group (A, B, C, D, E) we measured follow-

2005; Rodriguez-Perez and Green, 2006; Brochet et al., 2010). ing features on reference herbarium material (deposited at BRVU)

Nevertheless it is not clear to what extent such dispersal events from each population: leaf width, ﬂower peduncle length, podog-

have inﬂuenced the species genetic structuring at broader geo- yne length, achene length, achene width, number of inﬂorescences

graphic scales. R. maritima is a taxon with a single cpDNA haplotype per plant and number of fruits per plant. The number of measure-

whereas R. cirrhosa is more polymorph with unique allele variants ments for each haplotype ranged from 10 to 95 for leaf width;

(Triest and Sierens, 2010). Maternal genetic markers can be relevant 26–62 for podogyne and achene sizes; 12–48 for inﬂorescences and

at different geographic levels to infer dispersal patterns. However peduncle sizes and 4–58 for individual shoots. The habitat type was

if chloroplast capture occurred after introgressive hybridization, scored on a scale from 1 to 5 reﬂecting an increased marine inﬂu-

then the information obtained solely from cpDNA might blur the ence: (1) inland waters, (2) coastal brackish, (3) coastal temporary

interpretation of distribution ranges. Therefore nuclear markers are saltmarsh, (4) coastal permanent lagoon, (5) coastal lagoon with

needed to reveal introgression events. Zostera and seaweeds.

We investigated the distribution and morphological traits of

Ruppia chloroplast haplotypes in the European part of the Mediter- 2.4. Data treatment

ranean. The objective was to estimate whether maternal cpDNA

differentiation between and within subbasins of the Mediterranean DNA sequences were aligned with CLUSTAL W (Thompson et al.,

followed an isolation-by-distance model using ﬂight distances or 1994). The 2300 bp long haplotypes were deﬁned on basis of

sea current distances. The distribution pattern of each cpDNA vari- transitions, transversions, indels and mononucleotide repeats (Cor-

ant, their nuclear DNA identity, morphology, ﬁtness traits and rigendum: haplotype B5 as mentioned in Triest and Sierens, 2010

habitat type will be compared to discuss on the potential role of his- has to be replaced by haplotype E3). One haplotype (D) referred to

torical dispersal through either birds or sea currents. Additionally R. maritima whereas 14 haplotypes (groups A, B, C, E) referred to

we will comment on ‘how marine’ the different Ruppia haplotypes an unresolved complex including R. drepanensis (A) and R. cirrhosa

are in the Mediterranean. (B, C, E). A minimum spanning network using NETWORK 4.5.1.0

(Fluxus Engineering) served as basis for the haplotype deﬁnition.

2. Materials and methods Morphological measurements were tested for signiﬁcant

differences with one-way ANOVA (Kruskal–Wallis) and pair-

2.1. Study sites and plant materials wise Mann–Whitney U test. Genetic differentiation between

pairs of regions (˚ST), Slatkin’s ˚ST/(1 − ˚ST) were calculated

Ruppia plants were collected in 2006, 2007, 2008 and 2009 in with ARLEQUIN (Excofﬁer et al., 2005) considering the pairwise

56 water bodies from 38 wetland areas in the European part of the differences between haplotypes. Slatkin’s ˚ST/(1 − ˚ST) was

Mediterranean (Table 1, Fig. 1). In each site we collected up to 30 used for testing isolation-by-distance between pairs of regions

individual shoots (ramets) along a 30 m transect, thereby largely with geographical distances obtained as the average distance

avoiding identical genets because visible clumps of a ramet were between populations. Straight ﬂight distances and distances

smaller than 1 m diameter. Leaves were dried on silica gel and a following major sea currents (both log transformed) were used.

reference herbarium for each population was deposited at BRVU The tests were done at the within basin level (East or West),

(herbarium of the Vrije Universiteit Brussel). A total of 1546 individ- between subbasins of each basin and within subbasins. We

ual shoots was investigated for cpDNA sequence variability in ﬁve considered subbasins corresponding to relevant Mediterranean

genes (ccmp-2, ccmp-3, ccmp-10, trnH-psbA, rbcL) as published in biogeographical subdivisions, taking into account the major

Triest and Sierens (2010). Genbank Accession numbers are listed currents (http://www.ifremer.fr/lobtln/OTHER/Terminology.html

for ccmp-2 (JN1013249–JN113255), ccmp-3 (JN113257–113259), and http://www.mediterranean-yachting.com/winds.htm) and

ccmp-10 (JN113260–JN113263), trnH-psbA (JN113266–JN113271) currentology of the Mediterranean Sea (Blondel et al., 2010). These

and rbcL (JN113275–JN113278). Here we add new information are the coastlines of the Alboran, Balearic (including Lyon Gulf),

from two nuclear spacers (ITS1 and ITS2) of which Genbank Tyrrhenian, Adriatic, Ionian and Aegean subbasins. The populations

L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53 47 study)

DNA neighbouring (This

ABC

ITS-B ITS-A ITS-B ITS-A ITS-B ITS-A ITS-B ITS-A Nuclear ITS-C ITS-A ITS-C ITS-B ITS-B ITS-A ITS-A ITS-B ITS-B ITS-B ITS-B ITS-A ITS-A ITS-A ITS-B ITS-B ITS-B ITS-A ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B ITS-B marker and

number

) 2010

wetland

the

haplotype

Sierens, C4 C3 C3

C1 C1 C1 C1 C1 C2, C2, C2, C3 and

Greece,

A2 A2 C1, C1 C1, C1, B3, B3, C1 B2, B2, B2 B2 B2 B2, E2 B2 C2 C2,

(cp-capture) (cp-capture) (cp-capture)

GR: Triest

E3 Chloroplast A1, A1, B1 E1 B1 E1 B1, B1 B1 D1 C1, C1 C1 B4, B1 D1 C1, B1, B1, D1 B1, B1, B1, B1, B1, B2 B1, C1 E3 C1 B1, B1, B1, B1 B1 B1, B1 B1 B1 B2 B1, B1 ( and

Slovenia

1.58 ◦ complex). 08.77 04.04 06.31 03.33 51.68 16.33 49.61 38.76 36.81 18.94 15.50 44.78 11.57 15.58 12.25 59.96 48.48 48.48 48.28 36.29 27.39 52.37 52.03 37.03 29.07 56.15 14.50 55.32 28.36 14.31 36.26 36.22 55.32 25.52 23.15 17.55 SLO:

0 ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ 01.25 0 0 0 0 3 4 2 6 2 2 3 4 4 3 4 4 4 8 8 8 8 9 ◦ 10 20 22 26 12 11 12 13 12 13 13 21 21 21 Longitude 9 − − − − − − − Italy,

I: cirrhosa

France;

00.55 00.47 04.28 20.77 01.89 09.28 24.69 42.57 42.78 11.07 36.98 11.11 38.24 15.56 56.47 59.09 54.03 33.15 21.57 21.57 22.75 52.07 58.09 53.43 53.22 26.06 51.58 46.43 25.43 23.05 42.50 34.46 29.28 31.40 19.57 15.54 46.33 10.53 ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦ belong

39 38 37 39 40 40 42 38 40 39 36 36 38 39 42 39 39 39 43 43 43 43 39 39 38 38 39 37 42 42 41 45 44 45 45 38 38 Latitude 39 Spain;

SP: others

all

indicate

Saler

whereas

Motte

Ampurdà

Tsoukalio de

salinas

del

in Foglianu capture and

Badiola

Lago

Grande

Albufera

(Abbreviations ¸

and

NP,

Tamarit

l’oie and Puc

Salina cpDNA

lagoons

Victoria and Pescaia, Istai

Torreblanca

Giusta Mongofre Salina Aiguamolls

Koronissiu

Capouillet Pallisade

Grappa lagoon

Griells lagoon

with Prokopos Mar

Cavanata Cogoni

wetlands

Lago

Moros, della Valencia Torre lagoon

Trou Le La

Manjavacas

els lagoon de Secovlje Santa Stagno

Carnon Kotychi

saltmarsh Western

Saline

Comacchia Salinas Lake de

Cabanes,

Borgo

Valle

dels

Buda

Monolimni

Strunjan Aliki

Pola,

Corallo

Monte

Logarou,

hybrids

Bou Lake

Giudeu Grau,

are

Localities Laguna Valverde Almerimar, Roquetas Santa Albufera Marjal Prat Illa L’Estartit, Empuriabrava, Es Addaia, Son Between Camargue, Camargue, Camargue, Oristano, Oristano, Cagliari, Chia, Su Porto Trapani Castiglione Orbetello Circeo, Grado, Valle Portoroz, Piran, Arta, Messolonghi Ilia, Achaia, Aigio, Evros, #

, Mediterranean

maritima

1 R.

30 30 90 30 30 60 30 60 30 30 30 30 30 30 30 60 30 60 30 30 60 60 60 92 58 15 25 29 23 29 31 28 27 26 46 36 N 120 Ruppia

are

* of

drepanensis

identiﬁcation §

sites) §

are

Wetlands

DNA §

F16 GR33ABCD (56 I21 SP3 SP6 I29 38 SP1 SP2A SP5ABC SP9 F15 I25ABC I28AB GR38ABC SP10AB I24 SP4AB I27 SP7# SP8* SP11 SP13# SP14* F17 F18* I20 I22 I23 I30 SLO31AB SLO32 GR34AB GR36AB GR37 SP12AB I19 I26AB GR35AB plants; nuclear

and

number

subbasin =

N Spain

subbasins delta

subbasin subbasin

subbasin NP delta Brava haplotypes

subbasin

delta Gulf subbasin Almeria Alicante Valencia Trieste

and

basin

Mancha basin

1 W-Greece SE-Sardinia Evros Gulf Gulf Gulf Gulf Ebro Costa Menorca Lion Rhône W-Sardinia W-Sicily N-Tyrrhenian S-Tyrrhenian Peleponesos

La Adriatic Alboran Donana Balearic Tyrrhenian Ionian Aegean Basins SW-Inland West East Chloroplast Table waterbodies;

48 L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53

Fig. 1. Distribution of cpDNA haplotypes of Ruppia populations in the European part of the Mediterranean. Upper map indicates haplotypes A (R. drepanensis), D (R. maritima),

E (part of R. cirrhosa complex) and hybrids. Lower map indicates haplotypes B and C (major part of R. cirrhosa complex). Populations can have mixed haplotype composition

of B and E and are shown on upper and lower map respectively. Symbol sizes are proportional to the number of observations. Numbers refer to 38 wetlands (as in Table 1

but without country codes).

along coastlines of islands of Menorca (Balearic), West Sardinia distribution (Fig. 1). Three ITS1 amplicon (305–306 bp) and three

(Balearic), South East Sardinia (Tyrrhenian) and Sicily (Tyrrhenian) ITS2 amplicon (443 bp) sequences were fully linked. A total of 38

were considered according to their subbasin. Pairwise distances substitutions and four deletions characterized those three linked

between each population were estimated from the coordinates ITS variants. ITS1 amplicon was somewhat more variable (19/306

using straight paths (Euclidean direct ﬂight distances) and using variable sites) than ITS2 (19/443 variable sites) giving an overall

paths constructed along with major sea currents (shortest pos- identity of 97.1% and 98.6%, respectively.

sible routes). Distances were obtained from drawing paths in These variants differed mainly in substitutions and were named

Google Earth. All statistical tests were obtained with STATISTICA ITS-A, ITS-B and ITS-C. The nuclear ITS-A variant was observed

software. in the most divergent haplotypes D and group E whereas the

nuclear variant ITS-B was shared by the most common haplotypes

B and C (Table 1). In a few populations the nuclear ITS-A variant

3. Results

was observed in individuals with haplotype B1, thereby indicating

hybridization. This suggests chloroplast capture of haplotype B1

3.1. Haplotype and ITS deﬁnition

(from egg cell of haplotype B) through hybridization with pollen of

a taxon characterized by nuclear ITS-A sequences and was detected

Five groups of chloroplast haplotypes (named A, B, C, D, E) dis-

in three populations from Spain (Table 1, Fig. 1). These populations

tinguished R. maritima (haplotype D1) from R. drepanensis (A1, A2),

were from Valverde (Donana NP), Marjal dels Moros (Puc¸ ol) and

the latter not fully resolved from a diverse R. cirrhosa complex

Addaia (Menorca). The nuclear ITS-C variant coincided with hap-

(haplotypes B1, B2, B3, B4, C1, C2, C3, C4, E1, E2, E3, E4). The hap-

lotype group A. The sequence homology of nuclear ITS1 and ITS2

lotype groups A, B and C each differed only in a few substitutions,

was lowest for comparisons between taxa with haplotypes D (or

a microsatellite repeat or an indel whereas E was more divergent

E) and other haplotypes (94–96%). The ITS sequence homology was

(Triest and Sierens, 2010). A total of 15 haplotypes from 53 water

largest for the comparison between taxa with haplotypes A and B

bodies formed the basis for further analysis of their morphology and

L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53 49

(or C), reaching 98.5–98.7%. Only one hybrid population (Valverde,

Donana NP) showed intra-individual ITS variability.

3.2. Morphology

For each of the ﬁve haplotype groups (A, B, C, D, E) we com-

pared leaf and fruiting inﬂorescence features (Table 2). Leaf width

ranged from 0.1 to 1.1 mm and differed signiﬁcantly between hap-

lotypes A, D/E and B/C (ANOVA F4,204 = 92; p < 0.0000). Leaves were

nearly capillary and about 0.1 mm wide for haplotype A, with 25–75

percentile values between 0.2–0.3 mm for D/E and 0.5–0.9 mm for

B/C.

The inﬂorescences were clearly different between haplotype

Fig. 2. Pairwise cpDNA haplotype differentiation (haplotype groups B, C and E)

groups. In haplotype D the peduncle was less than 25 mm whereas

of Ruppia cirrhosa populations against their geographical distance between sub-

in the other haplotypes it reached 30–120 mm in our samples. Hap- basins. Dark symbols refer to comparisons using sea current distances (regression

equation); open symbols refer to comparisons using ﬂight distances.

lotype D had straight or one-curled peduncles whereas B, C and E

were spirally (sometimes only one curl), often reddish. Many-coiled

and dense, often whitish, peduncles were found in haplotype A. measured according to sea currents gave signiﬁcant IBD unlike

The podogyne length was very variable, but smaller for haplotype straight ﬂight distances (Fig. 2). At subbasin level an overall IBD was

B (F4,199 = 7.85; p = 0.0001). detected as well as within the Balearic subbasin only, indicating a

The achenes were shorter for haplotype A (1.8–2.0 mm as 25–75 restricted historical dispersal (Table 3, Fig. 3).

percentile values), D (1.9–2.1 mm) and E (2.0–2.2 mm) than for The slope of an IBD model was higher between basins (West

B/C (2.1–2.6 mm) with most groups signiﬁcantly different, except versus East) than between subbasins (e.g., Alboran, Balearic,

D from E and B from C (F4,199 = 10.56; p < 0.0000). Achenes were Tyrrhenian) and lowest for within subbasin comparisons. The IBD

narrowest for haplotype A (1.0–1.1 mm) than for all other groups model also was stronger for distances that follow sea currents than

(1.2–1.5 mm). The number of ripe fruits per inflorescence reached for direct flight distances. We obtained slope values of 1.19 (flight)

the largest numbers (6–8 on average) for haplotype A, whereas to 1.23 (flow) between basins, 0.47 (flight – but a non significant

all other haplotypes had less fruits (on average 4–6), though with relationship) to 0.74 (ﬂow) between subbasins (Western Mediter-

many outliers (F4,153 = 3.87; p = 0.005). The reproductive ﬁtness ranean) and 0.04 (Eastern Mediterranean – but a non signiﬁcant

(fruits/plant) clearly differentiated the many-seeded haplotype D relationship). Lowest slope values were obtained for comparisons

(on average about 100) from haplotype A/E (about 30) and B/C within subbasins of the Western Mediterranean (0.29 for ﬂight and

(on average less than 5 and often zero) (F4,122 = 152; p < 0.0000). 0.24 for sea current). Overall the R values were very low at large

Despite a large variability, all haplotypes were signiﬁcantly dif- geographic level because identical haplotypes occurred at large dis-

ferent in reproductive ﬁtness (Mann-U, all with p < 0.001) except tances. At a smaller geographic level, within the Balearic subbasin,

between haplotypes A and E. the R explained up to 32% of the IBD variation (Table 3).

The habitat type scored on a scale from 1 to 5 (from no to a The maximum geographic range between populations of similar

strong sea inﬂuence) was averaged for each haplotype (Table 2). cpDNA variants was used as a proxy for accumulated histor-

This indicated that the habitat of haplotype B was more marine ical seed or shoot dispersal events. Haplotype B1 occurred at

than haplotypes A (Mann-U; p = 0.003) and D (Mann-U; p = 0.003). large ranges across subbasins within the East basin, from the

Haplotype B (when in combination with ITS-B and not as a hybrid Adriatic over the Ionian towards to Northern Aegean Sea encom-

with ITS-A) thus should be considered as the most marine Ruppia passing about 2000 km (when measured along coastlines and

genotype of the Mediterranean. sea currents) or 1180 km (straight ﬂight distances across Balkan

peninsula). In the West basin haplotype B1 ranged over more

3.3. Distribution patterns, distances and IBD at subbasin level than 1900 km (sea currents) or 1500 km (straight ﬂight) between

Southern Iberian and Italian peninsula. The maximum observed

The geographic distribution of similar haplotypes that refers geographic range between populations of similar haplotypes was

to maternal dispersal events showed contrasting patterns among high for A1 (>500 km inland), B2 (1450 km, but could be homoplasy

the variants in the Mediterranean (Fig. 1). The most continuous

and widespread haplotype B1 occurred on all three peninsula and

islands studied. The second common haplotype C1 was mainly con-

ﬁned to the Western Mediterranean (only one individual detected

in samples from the Adriatic subbasin). All other Ruppia haplotypes

showed a scattered distribution (A1, B2, C2, D1, E1, E3) or were –

up to now – detected from a single area (A2, B3, B4, C3, C4, E2, E4,

E5).

At the level of two Mediterranean basins a signiﬁcant IBD was

observed within the West basin but not in the East basin (Table 3).

The latter had too low variability, indicating historical dispersal

of only haplotype B1 (and B2 if no homoplasy). The East basin

therefore could not be considered in further IBD tests. In the West

basin a strong IBD was found for pairwise distances between three

subbasins (Alboran, Balearic and Tyrrhenian) when considering dis- Fig. 3. Pairwise cpDNA haplotype differentiation (haplotype groups B, C and E) of

Ruppia cirrhosa populations against their geographical distance within subbasins.

tances following major sea currents (Table 3, Fig. 2). This indicates

Dark symbols refer to comparisons within West subbasins (circles = ﬂight distances;

restricted historical dispersal between those subbasins and was

triangles = sea current distances with regression equation); open symbols refer to

mainly explained by the scattered haplotypes E and local unique

comparisons within East subbasins (circles = ﬂight distances; squares = sea current

variants at short distances. Pairwise distances between subbasins distances).

50 L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53

Table 2

Measurements of morphological characters for ﬁve cpDNA haplotype groups (comprising three nuclear ITS-variants) in 56 Mediterranean Ruppia populations (mean, minimum

and maximum values are given). Signiﬁcant differences (Mann–Whitney-U, p < 0.05) are indicated as a,b,c,d.

Diagnostic feature Sample size of Haplotype A Haplotype B Haplotype C Haplotype D Haplotype E

haplotype group ITS-C ITS-B ITS-B ITS-A ITS-A

(A/B/C/D/E) R. drepanensis R. cirrhosa complex R. cirrhosa complex R. maritima R. cirrhosa complex

a b b c,d d

Leaf width (mm) 10/95/63/31/10 =or <0.1 0.75 (0.4–1.0) 0.71 (0.4–1.1) 0.29 (0.2–0.5) 0.25 (0.2–0.3)

a b a,c a,c a

Podogyne length (mm) 26/40/62/48/28 14.0 (5–23) 10.0 (4–21) 13.5 (2–23) 13.7 (2–34) 17.3 (8–25)

a b b c c

Achene length (mm) 26/40/62/48/28 1.8 (1.5–2.0) 2.3 (1.4–3.0) 2.4 (1.7–3.2) 2.0 (1.5–2.2) 1.1 (1.2–2.8)

a b b b,c b,c

Achene width (mm) 26/40/62/48/28 1.1 (0.9–1.2) 1.3 (1.0–1.7) 1.4 (1.0–2.0) 1.3 (1.0–1.5) 1.3 (0.8–1.8)

a a a b b

Achene L/W 26/40/62/48/28 1.7 (1.5–2.0) 1.8 (1.3–2.6) 1.8 (1.1–2.5) 1.6 (1.3–2.0) 1.6 (1.2–1.9)

a b,c b,c c c

Number of fruits/inﬂorescence 17/39/48/42/12 6.9 (5–10) 4.8 (2–8) 5.6 (1–12) 5.6 (3–10) 4.5 (2–8)

a b c d a

Number of fruits/plant 4/58/44/12/9 31.5 (25–37) 3.8 (0–56) 10.1 (0–50) 97.7 (72–128) 33.6 (20–54)

Peduncle 17/39/48/42/12 Many dense coils Spirally Spirally Straight/1 curve Straight/1 curve

Long, whitish Long, reddish Long, reddish Short, <25 mm Short, <25 mm

a b b,c c b,c

Sea inﬂuence (scale 1–5) (3/41/20/6/6) 1.0 3.7 (2–5) 3.4 (2–5) 2.7 (1–4) 3.2 (2–4)

of a mononucleotide T-repeat), C2 (790 km, but could be homo- bias in resulting relationships and further interpretation as a R. mar-

plasy of a mononucleotide T-repeat), D1 (1150 km) and E3 (380 km itima complex. Likewise, shortened sequences of MatK were used.

between Italy and Sardinia), except for B3 (10 km), C3 (70 km), E1 Obviously, the parsimony information for R. cirrhosa was underesti-

(15 km). mated in their study and unfortunately more confusion was added

to the already confusing taxonomy of Ruppia.

4. Discussion

We revealed haplotype groups using both exons (rbcL) and 4.1. Dispersal by birds

introns (ccmp, trnH-psbH) of the chloroplast genome and com-

bined these with nuclear spacer (ITS1 and 2) information and Birds are considered as the most likely vector of LDD

morphological measurements. Haplotype group A with nuclear through seeds. Waterfowl carry seeds on their feet, feathers

ITS-C corresponded to a morphologically distinct R. drepanensis and in the gut. Seeds might germinate when birds deposit

characterized by capillary leaves and long, many-coiled peduncles. them in suitable habitats (Fishman and Orth, 1996). The role

The unique and most distinct haplotype D combined with ITS-A of birds in LDD was argued on basis of feeding experiments,

corresponded to a many-seeded R. maritima with short peduncles. simulation of ﬂight distances versus retention time of seeds

Haplotype groups B and C that shared ITS-B were morphologically in the gut and germination capacity of R. maritima seeds

distinct in having wider leaves, longer achenes and spirally pedun- (Figuerola and Green, 2002; Figuerola et al., 2002). The role of birds

cles. This polymorphic group was assigned to a R. cirrhosa complex. in dispersal is not at all questioned, but it is more difﬁcult to obtain

Haplotype group E was more related to B and C than to any other experimental evidence for their role in LDD or in homogenizing a

haplotype group and was considered as part of the haplotypic R. genetic structure. Several putative steps going from fruiting period

cirrhosa complex as a result of chloroplast capture after hybridiza- until germination and establishment in a distant habitat were crit-

tion. The latter group contained the nuclear ITS-A and approached ically reviewed (Clausen et al., 2002) but partially answered in

morphological features of R. maritima, except for their lower num- favour of birds as effective vectors even if the season of seed produc-

ber of fruits per plant. The latter complex has caused much of the tion is decoupled from bird migration periods (Charalambidou and

taxonomic uncertainty in Mediterranean Ruppia. Santamaria, 2005; Rodriguez-Perez and Green, 2006; Brochet et al.,

Several chloroplast exons and a nuclear coding gene of Ruppia 2010). To be effectively dispersed though birds, the various Ruppia

taxa at a global scale revealed hybridization and polyploidization haplotypes should be highly fertile. We observed that haplotype D

as important phylogenetic events (Ito et al., 2010). The Asian and (R. maritima, including the introgressed hybrids with nuclear ITS-

Oceanic taxa R. megacarpa S. Mason, R. tuberosa Davis and Tomlin- A and capture of haplotype B) was many-seeded with on average

son and R. polycarpa S. Mason could be clearly resolved. However, 3 times more fruits per plant than haplotypes A and E and even

Ito et al. (2010) could not clearly distinguish R. maritima from R. cir- 10–30 times more than haplotypes B and C, respectively. The prob-

rhosa (R. drepanensis was not studied) and proposed a world-wide ability of bird-mediated seed dispersal of B/C populations thus will

R. maritima complex without considering the evidence of a distinct be reduced if these plants ﬂower and fruit less freely. Therefore we

R. maritima taxon and a polymorphic R. cirrhosa complex in Europe hypothesize that the many-seeded haplotypes A, D, and E have a

(Triest and Sierens, 2009, 2010). Analysis of methodological aspects higher probability for dispersal by birds than B and C. This did not

reveal that Ito et al. (2010) used a small part of the rbcL (542 bp) result in a lowered genetic structuring throughout the Mediter-

instead of the full sequence (>1300 bp). They designed primers for ranean distribution range because haplotype groups A, D and E

a shorter rbcL on basis of a R. maritima sequence and hence caused a show a scattered distribution and were mainly observed in the

Table 3

Overview of signiﬁcant isolation by distance models (IBD) at different geographic levels (basin and subbasin) using genetic differentiation ˚ST/(1 − ˚ST) against straight ﬂight

(left) and sea current (right) distances (both log km). N = number of population pairs; r = Pearson correlation, p = signiﬁcance level, R = coefﬁcient of determination.

2 2

Geographic level N IBD (Flight distances) r p R IBD (Sea current r p R

distances)

Between basins 210 y = −2.88 + 1.19x 0.37 0.0001 0.14 y = −3.40 + 1.23x 0.25 0.0002 0.06

Within basins (West or East) 255 y = −0.24 + 0.37x 0.23 0.0003 0.05 y = −0.49 + 0.44x 0.29 0.0001 0.08

Within West basin 210 y = −0.46 + 0.50x 0.27 0.0001 0.07 y = −0.67 + 0.54x 0.32 0.0001 0.10

Between West subbasins 144 n.s. – – – y = −1.24 + 0.74x 0.26 0.0020 0.07

Within West subbasins 66 y = −0.10 + 0.29x 0.34 0.0053 0.12 y = −0.02 + 0.24x 0.29 0.0169 0.09

Within Balearic subbasin 36 y = −0.53 + 0.49x 0.55 0.0005 0.31 y = −0.55 + 0.48x 0.57 0.0003 0.32

L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53 51

Western part whereas haplotypes B1 and C1 were most common genetic differentiation when compared to non-visited populations.

throughout the European part of the Mediterranean. In a related RAPD study, King et al. (2002) compared two coastlines

Dispersal vectors with low displacement over long distances in the Baltic Sea for genetic differentiation in P. pectinatus and

and short seed passage time are expected to contribute hardly to concluded that the SW Baltic coastline populations known for high

LDD, even when the seed load is very high (Nathan et al., 2008). waterfowl concentration, showed a reduced isolation. On basis

We found no evidence for an extensive bird-mediated dispersal of cumulative differentiation values it was suggested that water-

lowering the IBD within the West basin because there was a sig- fowl can increase the neighbourhoods up to about 150–200 km.

niﬁcant genetic structure between the subbasins, the latter when Their interpretation is based on a priori grouping of populations

considering sea currents, but not when considering ﬂight distances considered to have the greatest bird trafﬁc coming and going to

among the subbasins. This can be explained partly from the iden- arctic Russia, and those that are less visited. A non signiﬁcant

tical haplotypes at longer distances but also from the occasionally trend towards lower differentiation was explained by bird-trafﬁc

large differentiation at short distances. At short distances (about regions. In our opinion, migrating birds are a potential explana-

<20 km) waterfowl was shown to have an effect on R. maritima tion, however, a physical barrier such as the highly fragmented

pond populations and their seed bank densities throughout the archipelago region could be an alternative explanation for the

year in SW Spain and not only at times of major concentrations of stronger IBD of the Northern most populations.

migratory fowl in autumn (Rodriguez-Perez and Green, 2006). Win- Ito et al. (2010) detected similar sequences in exons of one

ter observations on teal in the Camargue showed that dispersal of nuclear and four chloroplast genes for R. megacarpa from North

propagules indeed can be decoupled from the time of seed produc- Eastern Asia and Southern Oceania. This disjunct distribution was

tion of aquatic plants (Brochet et al., 2010). Over larger distances, explained as a bird-mediated long distance dispersal. Given the

ﬂamingos were proposed as an important factor in wetland migra- abovementioned constraints of using partial, shortened sequences

tions across the Mediterranean (Rodriguez-Perez and Green, 2006). of conservative genes, these materials need further analysis using

Greater Flamingos are adapted to shallow wetlands and individ- full rbcL sequences and less conservative genes before inferring long

uals can travel long distances (Amat et al., 2005). Flamingos were distance dispersal events.

reported to connect Western Mediterranean wetlands with East Albeit, it will be very difﬁcult to detect traces of bird-mediated

basin wetlands in Turkey (Balkiz et al., 2007). From our observa- dispersal events in large and continuous populations with genetic

tions during the ﬁeld work on Ruppia and from bird lists of protected methods. Distribution patterns of unique alleles might be more

areas, Ruppia is often found in the same areas from which ﬂamingo informative than overall gene diversity and differentiation meas-

populations were recorded. ures. It is a paradox that for Ruppia the least fertile haplotypes B and

In the Mediterranean, Ruppia fruits are produced at different C showed the largest and most continuous distribution whereas in

times of the year according to the taxon and the habitat type being comparison the most fertile haplotypes A, D and E showed a scat-

ephemeral or permanent (Malea et al., 2004). In large lagoons, R. tered distribution. The most marine chloroplast genotype can, after

cirrhosa is ﬂowering and fruiting from June to July, whereas in cpDNA capture through hybridization with R. maritima, be hitch-

ephemeral sites this can be a month earlier (Gesti et al., 2005). R. hiked to other, even inland habitats. The three cases of introgression

maritima set fruits in May–June, whereas R. drepanensis even might we found showed that pollen of R. maritima fertilized the R. cir-

fruit earlier, depending on the ephemeral habitat (personal obser- rhosa ovules. This also ﬁts with the observation that R. maritima

vations). Haplotypes A (R. drepanensis), D (R. maritima) and E (part is more fertile and thus has a higher probability to cross-pollinate

of R. cirrhosa complex) produce a large number of seeds and such with other taxa.

a high seed rain most likely results in a rich seed bank through-

out the year or even years. These haplotypes are mostly known 4.2. Dispersal through sea currents

from isolated, shallow, sometimes ephemeral habitats depending

on freshwater for germination. R. maritima seeds germinated best at R. cirrhosa can occur in saltmarshes, salinas, estuaries and

low chlorinities in strong contrast to much lower germination rates coastal lakes, nowadays mostly protected from direct sea inﬂu-

of R. cirrhosa (Van Vierssen et al., 1984). R. drepanensis and R. mar- ence by dunes and dikes. R. cirrhosa populations can co-occur

itima both occur in inland saline lakes in Spain whereas R. maritima with other seagrasses such as Zostera. We observed R. cirrhosa

also established far inland in few isolated saline habitats of West- together with Zostera in several lagoon areas of Italy (Oristano,

ern and central Europe. Therefore, it is evident that haplotypes A (R. Circeo, Orbetello, Grado) and Greece (Amvrakikos gulf and Mes-

drepanensis), D (R. maritima) and most likely also haplotype E (part solonghi lagoons). Their establishment might have originated from

of R. cirrhosa complex) are dispersed by birds. Together with the drifting propagules transported by sea currents. Vegetative propa-

scattered distribution pattern of these haplotypes, it makes sense gation of seagrasses can only allow the plants to colonize adjacent

to state that LDD by birds are rare events and therefore not at all non vegetated area or get mixed but not far away. Seed production

lowering the genetic differentiation between population pairs at on ﬂowering and drifting shoots might increase the dispersal dis-

various distances. tance. The seeds of seagrasses generally are too heavy to travel a

There are only few reports on the putative homogenizing effect long distance by water currents alone (Orth et al., 1994) although

of bird-mediated dispersal over large geographic distances of fruits of P. oceanica ﬂoat and are potentially able to disperse for

submerged waterplants. These studies were conducted with dom- several weeks under the inﬂuence of surface currents (Serra et al.,

inant markers and are not entirely convincing that populations 2010). Dispersal distances of detached seeds of Zostera marina L.

connected through bird migration routes indeed had an overall appear to be very limited (Orth et al., 1994; Ruckelshaus, 1996).

lowered differentiation. Using RAPD markers, Mader et al. (1998) Orth et al. (1994) frequently observed ﬂoating reproductive shoots

showed that the slope of geographic against genetic distance at the sea surface. They suggested that long-distance dispersal and

was lower for Potamogeton pectinatus L. populations that were colonization of distant habitats may be achieved via these ﬂoat-

frequently visited by swans in comparison to those that were ing plants. Fully ripened Ruppia fruits detach and sink readily to

not visited by swans. However, in our opinion this interpretation the bottom. However, larger shoot fragments with rhizomes and

should be revised because the lower slope for bird-visited pop- ﬂowering/fruiting parts regularly detach and survive when ﬂoat-

ulation differentiation was caused by higher pair-wise genetic ing (Personal observation in Oristano, Stagno Istai, Porto Corallo

differentiation at small geographic distance and not, as should be lagoon and Circeo in June 2008). The most probable means of sea-

in case of homogenization over longer distances, due to lowered grass colonizing nearby shores is through transportation of seeds

52 L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53

on ﬂoating ﬂowering or fruiting shoots. These shoots may have latter however contained a single mononucleotide repeat that

become detached through stochastic disturbances such as stormy could be the result of independent events. We summarize that IBD

weather or the feeding activities of water fowl burrowing in the gave better ﬁt when using sea current distances than direct ﬂights.

sediment. An important question is whether Ruppia shoots can root This IBD relationship was low within basins because genetic dif-

again after a period of ﬂoating. ferentiation can occur at short distances and identical haplotypes

Rafting of seeds on ﬂoating reproductive shoots is an impor- might exist at large distances. Birds are considered as important

tant dispersal strategy for eelgrass, allowing for colonization of vectors of dispersal at short distances and occasional LDD events

new areas (Källström et al., 2008). Eelgrass disperses by ﬂoating of the most fertile Ruppia populations of haplotypes A (R. drepa-

on ocean currents sometimes for great distances and with the nensis), D (R. maritima) and E (part of R. cirrhosa complex). Our

help of birds, but with low seed survival (Phillips and Menez,˜ results highlight a paradox, namely a large Mediterranean distri-

1988). Buoyancy for at least 26 days was observed for Z. marina. bution of the least fertile haplotypes and therefore we propose

Wind-driven surface-transport velocity of detached post-ﬂowering historical sea currents as a potential vector for accumulated dis-

inﬂorescences gave a dispersal potential of up to 150 km in one persal of the most marine variants of R. cirrhosa. This might explain

season along the Western Swedish coast (Källström et al., 2008). an overall continuous distribution and a lower diversity in the East

This estimate of about 150 km corresponds somehow with esti- basin. The alternative explanation that waterfowl would be the only

mates (200 km) of dispersal potential from genetic studies (Reusch responsible vector for LDD from West to East basin is less likely.

et al., 1999; King et al., 2002). P. oceanica showed moderate gene Nevertheless, good candidates of additional vectors along coast-

ﬂow between populations of the Tyrrhenian subbasin (Procaccini lines are ﬂamingos that connect wetlands over large distances. Such

et al., 2001). Floatation of buoyant diaspores (Enhalus, Thalassia, hypothesis should be tested in more detail by studying R. cirrhosa

Posidonia) and detached shoots with reproductive material can populations in ﬂamingo dominated habitats.

have dispersal distances of 100 m to 10 km (Harwell and Orth,

2002; Lacap et al., 2002). Z. marina reproductive fragments were

Acknowledgements

estimated to have a ﬂoatation potential according to the speed of

water ﬂow of about 23 km/6 h but strong winds even may enhance

This project was ﬁnanced by the Fund for Scientiﬁc Research

(Harwell and Orth, 2002). Other estimates for Thalassia testudinum

Flanders (KN 1.5.124.03, research projects G.0056.03, G.0076.05,

K.D. Koenig were up to 15 km (Kaldy and Dunton, 1999) or even up

Sabbatical leave contract for L. Triest), the Vrije Universiteit Brus-

to 350 km using ﬂoating potential and 360 km using microsatellite

sel (OZR 1172, OZR1189BOF). Jordi Figuerola (Estación Biológica

based IBD (Van Dijk et al., 2009), 63.5 km for Enhalus acoroides (L.f.)

Donana,˜ Sevilla), Xavier Quintana (Univ. Girona), Patrick Grillas

Royle, up to 73.5 km for Thalassia hemprichii (Ehrenb. ex Solms)

(Tour du Valat), Eva Papastergiadou, Chrysoula Christa, Kostas Ste-

Asch. (Lacap et al., 2002) and over 300–400 km during stormy

fanidis (Univ. Patras), Luka Kastelic (Secovlje Salina NP, Slovenia),

weather (Kendall et al., 2004). Rafting provides a typical dispersal

Alenka Popic (Slovenian Inst. Nature Conservation, Piran), Athana-

range within tens of kilometres (Thiel and Gutow, 2005). Genetic

sios Mogias (Univ. Thrace) and Pere Fraga i Arguimbau (Menorca)

structure of local seagrass populations suggests that gene ﬂow is

kindly helped with the indication of suitable sites or with collecting

limited and depends largely on vegetative dispersal (Procaccini

plant shoots. Tess Driessens helped with DNA extraction.

and Mazzella, 1998; Reusch et al., 1999). Sea currents also were

reported as dispersal vector for coastal land plants such as Beta vul-

garis L. subsp. maritima (L.) Arcang. (Fievet et al., 2006) and Cakile

References

maritima Scop. (Gandour et al., 2008). Thus even coastal terrestrial

plant parts can be dispersed through sea currents.

Aedo, C., Fernandez-Casado, M., 1988. The taxonomic position of Ruppia populations

Ocean currents are relatively slow but their vector load in prin- along the Cantabrian coast. Aquat. Bot. 32, 187–192.

Amat, J., Rendon, M., Rendon-Martos, M., Garrido, A., Ramirez, J., 2005. Ranging

ciple is without limit over ecological timescales and especially so

behaviour of greater ﬂamingos during the breeding and post-breeding periods:

over longer timescales (Nathan et al., 2008). The Mediterranean

linking connectivity to biological processes. Biol. Conserv. 125, 183–192.

Sea is a unique system and cannot be considered as a miniature Arnaud-Haond, S., Migliaccio, M., Diaz-Almela, E., Teixeira, S., Van de Vliet, M.S.,

ocean because of its microtidal system and heterogeneity in salin- Alberto, F., Procaccini, G., Duarte, C.M., Serrao, E., 2007. Vicariance patterns in

the Mediterranean Sea: west-east cleavage and low dispersal in the endemic

ity, currents and temperatures (Blondel et al., 2010). If sea currents

seagrass Posidonia oceanica. J. Biogeogr. 34, 963–976.

represent an effective vector for accumulated dispersal of partic-

Balkiz, Ö., Özesmi, U., Pradel, R., Germain, C., Siki, M., Amat, J., Rendon-Martos, M.,

ular taxa, then their populations should be located in the sea or Baccetti, N., Béchet, A., 2007. Range of the Greater Flamingo, Phoenicopterus

roseus, metapopulation in the Mediterranean: new insights from Turkey. J.

at least in lagoons with an actual or historical connection to the

Ornithol. 148, 347–355.

sea. Such genotypic variants are also expected to co-occur in habi-

Blondel, J., Aronson, J., Bodiou, J.-Y., Boeuf, G., 2010. The Mediterranean Region.

tats with seagrasses or seaweeds and eventually possess adaptive Biological Diversity in Space and Time. Oxford University Press, Oxford.

Brochet, A., Guillemain, M., Fritz, H., Gauthier-Clerc, M., Green, A., 2010. Plant dis-

traits for dispersal by water. We observed that in the habitats with

persal by teal (Anas crecca) in the Camargue: duck guts are more important than

most marine inﬂuences, only haplotype B or C occurred whereas

their feet. Freshwater Biol. 55, 1262–1273.

in the more sheltered, shallow lakes, ponds and ditches all haplo- Charalambidou, I., Santamaria, L., 2005. Field evidence for the potential of waterbirds

as dispersers of aquatic organisms. Wetlands 25, 252–258.

types could be observed. The paradox about why the least fertile

Cirujano, S., 1986. El género Ruppia L. (Potamogetonaceae) en la Mancha (Espana).

haplotypes being the most widely distributed across the Mediter-

Bolet. da Soc. Brot., Série 2 59, 293–303.

ranean should be investigated with both genetic methods using Cirujano, S., Garcia-Murillo, P., 1990. Ruppiaceae. Fontqueria 28, 159–165.

Clausen, P., Nolet, B.A., Fox, A.D., Klaassen, M., 2002. Long-distance endozoochorous

microsatellites and ecological studies on the ﬂoating potential of R.

dispersal of submerged macrophyte seeds by migratory waterbirds in north-

cirrhosa rhizomes and shoots. A additional question raised here is

ern Europe – a critical review of possibilities and limitations. Acta Oecol. 23,

whether lagoonal R. cirrhosa are adapted to vegetative survival and 191–203.

dispersal on long term by water currents, hereby resembling other Den Hartog, C., Kuo, J., 2006. Taxonomy and biogeography of seagrasses. In: Larkum,

A.W.D., Orth, R.J., Duarte, C. (Eds.), Seagrasses: Biology, Ecology and Conserva-

seagrasses or if the marine habitat merely is a secondary habitat

tion. Springer, Dordrecht, pp. 1–23.

and too harsh for allocating in fruits.

Excofﬁer, L., Laval, G., Schneider, S., 2005. Arlequin ver. 3. 0: an integrated software

On the basis of our observations, we assume that haplotype package for population genetics data analysis. Evol. Bioinformatics Online 1,

47–50.

B1 is the most continuous and widespread distributed maternal

Fievet, V., Touzet, P., Arnaud, J.-F., Cuguen, J., 2006. Spatial analysis of nuclear and

genotype of the Mediterranean Ruppia. The variants B3 and B4

cytoplasmic DNA diversity in wild sea beet (Beta vulgaris ssp. maritima) popu-

were local variants whereas B2 occurred in several subbasins. The lations: do marine currents shape the genetic structure. Mol. Evol.

L. Triest, T. Sierens / Aquatic Botany 115 (2014) 45–53 53

Figuerola, J., Green, A., 2002. Dispersal of aquatic organisms by waterbirds: a review Reusch, T., Stam, W., Olsen, J., 1999. Size and estimated age of genets in eel-

of past research and priorities for further studies. Freshwater Biol. 47, 483–494. grass, Zostera marina, assessed with microsatellite markers. Mar. Biol. 133,

Figuerola, J., Green, A., Santamaria, L., 2002. Comparative dispersal effectiveness of 519–525.

wigeongrass seeds by waterfowl wintering in south-west Spain: quantitative Rodriguez-Perez, H., Green, A., 2006. Waterbird impacts on widgeongrass Ruppia

and qualitative aspects. J. Ecol. 90, 989–1001. maritima in a Mediterranean wetland: comparing bird groups and seasonal

Fishman, J., Orth, R.J., 1996. Effects of predation on Zostera marina seed abundance. effects. Oikos 112, 525–534.

J. Exp. Marine Biol. Ecol. 198, 11–26. Ruckelshaus, M., 1996. Estimation of genetic neighbourhood parameters from pollen

Gandour, M., Hessini, K., Abdelly, C., 2008. Understanding the population genetic and seed dispersal in the marine angiosperm Zostera marina L. Evolution 50,

strucutre of coastal species (Cakile maritima): seed dispersal and the role of sea 856–864.

currents in determining population structure. Genet. Res. 90, 167–178. Serra, I., Innocenti, A., Di Maida, G., Calvo, S., Migliaccio, M., Zambianchi, E., Pizzi-

Gesti, J., Badosa, A., Quintana, X., 2005. Reproductive potential in Ruppia cirrhosa galli, C., Arnaud-Haond, S., Duarte, C., Serrao, E., Procaccini, G., 2010. Genetic

(Petagna) Grande in repsonse to water permanence. Aquat. Bot. 81, 191–198. structure in the Mediterranean seagrass Posidonia oceanica: disentangling past

Green, P., Short, F.T., 2003. World Atlas of Seagrasses. UNEP-WCMC University of vicariance events from contemporary patterns of gene ﬂow. Mol. Ecol. 19,

California Press, Berkeley and Los Angeles, p. 298. 557–568.

Harwell, M., Orth, R.J., 2002. Seed bank patterns in Chesapeake Bay eelgrass (Zostera Short, F., Carruthers, T., Dennison, W., Waycott, M., 2007. Global seagrass distribution

marina L.): a baywide perspective. Estuaries 25, 1196–1204. and diversity: a bioregional model. J. Exp. Mar. Biol. Ecol. 350, 3–20.

Ito, Y., Ohi-Toma, T., Murata, J., Tanaka, M., 2010. Hybridization and polyploidy of Talavera, S., Garcia-Murillo, P., Herrera, J., 1993. Chromosome numbers and a

an aquatic plant, Ruppia (Ruppiaceae), inferred from plastid and nuclear DNA new model for karyotype evolution in Ruppia L. (Ruppiaceae). Aquat. Bot. 45,

phylogenies. Am. J. Bot. 97, 1156–1167. 1–13.

Kaldy, J., Dunton, K., 1999. Ontogenetic photosynthetic changes, dispersal and sur- Talbot, S.L., Wyllie-Echeverria, S., Ward, D.H., Rearick, J.R., Sage, G.K., Chesnay, B.,

vival of Thalassia testudinum seedlings in a subtropical lagoon. J. Exp. Marine Phillips, R.C., 2004. Genetic characterization of Zostera asiatica on the Paciﬁc

Biol. Ecol. 240, 193–212. Coast of North America. Aquat. Bot. 85, 169–176.

Källström, B., Nyqvist, A., Aberg, P., Bodin, M., Andrea, C., 2008. Seed rafting as Tan, B., Liu, K., Yue, X.-L., Liu, F., Chen, J.-M., Wang, Q.-F., 2008. Chloroplast DNA varia-

dispersal strategy for eelgrass (Zostera marina). Aquat. Bot. 88, 148–153. tion and phylogeographic patterns in the Chinese endemic marsh herb Sagittaria

Kendall, M., Battista, T., Hillis-Starr, Z., 2004. Long term expansion of a deep potamogetifolia. Aquat. Bot. 89, 372–378.

Syringodium ﬁliforme meadow in St. Croix, US Virgin Islands: the potential role Thiel, M., Gutow, L., 2005. The ecology of rafting in the marine environment. I. The

of hurricanes in the dispersal of seeds. Aquat. Bot. 78, 15–25. ﬂoating substrata. In: Gibson, R., Atkinson, R., Gordon, J. (Eds.), Oceanography

King, R., Gornall, R., Preston, C., Croft, J., 2002. Population differentiation of Pota- and Marine Biology: An Annual Review, vol. 42, pp. 181–264.

mogeton pectinatus in the Baltic Sea with reference to waterfowl dispersal. Mol. Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensi-

Ecol. 11, 1947–1956. tivity of progressive multiple sequence alignment through sequence weighting,

Koga, K., Kadono, Y., Setoguchi, H., 2008. Phylogeography of Japanese water crawfoot position-speciﬁc gap penalties and weight matrix choice. Nucleic Acids Res. 22,

based on chloroplast DNA haplotypes. Aquat. Bot. 89, 1–8. 4673–4680.

Lacap, C., Vermaat, J., Rollon, R., Nacorda, H., 2002. Propagule dispersal of the SE Asian Triest, L., Sierens, T., 2009. High diversity of Ruppia meadows in saline ponds and

seagrasses Enhalus acoroides and Thalassia hemprichii. Mar. Ecol. Progr. Ser. 235, lakes of the western Mediterranean. Hydrobiologia 634, 97–105.

75–80. Triest, L., Sierens, T., 2010. Chloroplast sequences reveal a diversity gradient in the

Mader, E., Van Vierssen, W., Schwenk, K., 1998. Clonal diversity in the submerged Mediterranean Ruppia cirrhosa complex. Aquat. Bot. 93, 68–74.

macrophyte Potamogeton pectinatus L. inferred from nuclear and cytoplasmic Triest, L., Symoens, J.J., 1991. Isozyme variation in populations of the submerged

variation. Aquat. Bot. 62, 147–160. halophyte Ruppia (Ruppiaceae). Op. Bot. Belg. 4, 115–132.

Malea, P., Kevrekidis, T., Mogias, A., 2004. Annual versus perennial growth cycle in Triest, L., Tran Thi, V., Sierens, T., 2007. Chloroplast microsatellite markers reveal

Ruppia maritima L.: temporal variation in population characteristics in Mediter- Zannichellia haplotypes across Europe using herbarium DNA. Belg. J. Bot. 140,

ranean lagoons (Monolimni and Drana lagoons Northern Aegean Sea). Bot. Mar. 109–120.

47, 357–366. Van Dijk, J.K., Van Tussenbroek, B.I., Jiménez-Durán, K., Márquez-Guzmán, G.J.,

Nathan, R., Schurr, F., Spiegel, O., Steinitz, O., Trakhtenbrot, A., Tsoar, A., 2008. Mech- Ouborg, J., 2009. High levels of gene ﬂow and low population genetic struc-

anisms of long-distance seed dispersal. Trends Ecol. Evol. 23, 638–647. ture related to high dispersal potential of a tropical marine angiosperm. Mar.

Orth, R.J., Luckenbach, M., Moore, K., 1994. Seed dispersal in a marine macrophyte: Ecol. Prog. Ser. 390, 67–77.

implications for colonization and restoration. Ecology 75, 1927–1939. Van Vierssen, W., Van Wijk, R., Van der Zee, J., 1981. Some additional notes on the

Phillips, R., Menez,˜ E., 1988. Seagrasses. Smithsonian Contributions to the Marine cytotaxonomy of Ruppia taxa in western Europe. Aquat. Bot. 11, 297–301.

Sciences. No. 34. Smithsonian Institution Press, Washington, D.C. Van Vierssen, W., Van Kessel, C., Van Der Zee, J., 1984. On the germination of Ruppia

Procaccini, G., Mazzella, L., 1998. Population genetic structure and gene ﬂow in the taxa in western Europe. Aquat. Bot. 19, 381–393.

seagrass Posidonia oceanica assessed using microsatellite analysis. Mar. Ecol. Waycott, M., Procaccini, G., Les, D., Reusch, T., 2006. Seagrass evolution, ecology and

Progr. Ser. 169, 133–141. conservation: a genetic perspective. In: Larkum, A.W.D., Orth, R.J., Duarte, C.

Procaccini, G., Orsini, L., Ruggiero, M., Scardi, M., 2001. Spatial patterns of genetic (Eds.), Seagrasses: Biology, Ecology and Conservation. Springer, Dordrecht, pp.

diversity in Posidonia oceanica, an endemic Mediterranean seagrass. Mol. Ecol. 25–50.

10, 1413–1421. White, T., Bruns, T., Lee, S., Taylor, J., 1990. Ampliﬁcation and direct sequencing of

Provan, J., Wilson, S., Portig, A.A., Maggs, C.A., 2008. The importance of reproductive fungal ribosomal RNA genes for phylogenetics. In: Innis, M., Gelfand, D., Snin-

strategies in population genetic approaches to conservation: an example from sky, J., White, T. (Eds.), PCR Protocols: A Guide to Methods and Applications.

the marine angiosperm genus Zostera. Cons. Gen. 9, 271–280. Academic Press, New York, pp. 15–322.