J. Mar. Biol. Ass. U.K. (2005), 85, 835^841 Printed in the United Kingdom
Taxonomic distinction of Ophelia barquii and O. bicornis (Annelida: Polychaeta) in the Mediterranean as revealed by ISSR markers and the number of nephridiopores Ferruccio Maltagliati*$, Marco CasuO, Tiziana LaiO, Daniela Iraci SareriP, Daniela Casu*O, Marco Curini GallettiO, Grazia CantoneP and Alberto Castelli* *Dipartimento di Scienze dell’Uomo e dell’Ambiente, Universita' di Pisa, Via A. Volta 6, 56126 Pisa, Italy. Dipartimento di Zoologia e Antropologia Biologica, Universita' di Sassari, Via F. Muroni 25, 07100 Sassari, Italy. P Dipartimento di Biologia Animale ‘M. La Greca’, Universita' di Catania, Via Androne 81, 95124 Catania, Italy. $ Corresponding author, e-mail:
[email protected]
O
Ophelia bicornis sensu lato is a polychaete living in intertidal sandy habitats of Mediterranean and European Atlantic coasts, whose systematics have been strongly debated in the past few decades. In the present work the count of nephridiopores was coupled with genetic analysis carried out with DNA markers (inter simple sequence repeats) for a total of 30 individuals collected at six Italian beaches. Exact test, analysis of molecular variance, non-metric multidimensional scaling and assignment tests clearly separated individuals with ¢ve nephridiopore pairs from those with six pairs. This ¢nding validated results of a recent allozyme study in which O. bicornis sensu lato was split into O. bicornis sensu stricto (six nephridiopore pairs) and O. barquii (¢ve nephridiopore pairs). This paper represents a further contribution to the estimation of biodiversity within marine invertebrates.
INTRODUCTION In recent years a number of molecular studies on marine invertebrates have greatly contributed to the assessment of local biodiversity through the resolution of complex taxonomic cases. This led some authors to recommend the deposition of molecular data such as DNA samples of one syntype specimen and a list of diagnostic DNA banding patterns for formal taxonomic nomenclatural purposes (Westheide & Schmidt, 2003). However, a more powerful taxonomic approach requires the integration of molecular data with additional types of information such as morphology, behaviour, physiology, etc. (Lee, 2004). This cross-validation is needed to obtain a sound interpretation of the observed di¡erences in a taxonomic perspective. In the marine environment there is a growing suspicion that the present perception of biodiversity is based on nonrepresentative data. The problem is experienced even in conspicuous invertebrate macrofaunal groups, in which the number of described species is often underestimated. In particular, polychaetes are characterized by a large number of apparently widely distributed or cosmopolitan species. Detection of diagnostic di¡erences at the molecular level has revealed that many of these formerly alleged species are actually complexes of morphologically identical or slightly di¡erent cryptic species (i.e. Maltagliati et al., 2000, 2001, 2004; Scaps et al., 2000; Schmidt & Westheide, 2000). The polychaete Ophelia bicornis sensu lato Savigny 1818 is distributed along the North Sea, Mediterranean, Black Sea, North African and European Atlantic coasts, occurring intertidally in high energy, ¢ne to medium sediments. Journal of the Marine Biological Association of the United Kingdom (2005)
According to Wilson (1948), species belonging to the genus Ophelia Savigny, 1818 release their gametes on the sand where they will be swept away by the tidal or wave current into the water-column. Their larvae are potentially capable of long-distance dispersal, adopting a benthic lifestyle after several weeks of planktonic life. The debate on the taxonomic status of a number of species within the genus Ophelia has a long history owing to the limited reliability of morphological diagnostic traits usually employed for species identi¢cation in the group (Maltagliati et al., 2004 and references therein). Ophelia bicornis Savigny 1818 is the type species of the genus as well as the Family Opheliidae Savigny 1818. Fauvel (1927) distinguished in European Ophelia the species O. radiata and O. bicornis and the variety O. bicornis barquii on the basis of the number of gill pairs. However, the use of this character in taxonomy induced frequent misidenti¢cations due to its high variability, its overlapping distribution across species and the common occurrence of many cases of asymmetries and anomalies. This led Bellan (1964), Cantone & Costa (1975) and Amoreux (1977) to aggregate the above taxa into a single taxon identi¢ed as Ophelia bicornis sensu lato, characterized by remarkable morphological variability across Atlantic and Mediterranean localities. Successively, Pilato et al. (1978) proposed that individuals from an eastern Sicilian beach could be assigned to two separate species on the basis of the number of nephridiopore pairs: O. barquii with ¢ve nephridiopore pairs and O. bicornis with six pairs. The ¢rst attempt to clarify the taxonomy of the genus Ophelia by means of genetic markers was made by Britton-Davidian & Amoreux (1982). These authors tried to relate the observed allozyme patterns to the number of gill pairs,
836
F. Maltagliati et al.
ISSR distinguished Ophelia barquii from O. bicornis Table 1. Primer sequences used in the inter simple sequence repeats (ISSR) analysis, number of polymorphic bands per primer and range of molecular weight in base pairs (bp) ampli¢ed by polymerase chain reaction^ISSR for 30 individuals of Ophelia.
Primer
Sequence (5’^3’)
No. of polymorphic bands
Size-range of polymorphic bands (bp)
IT2 IT3 SAS3 PT1
(CA)8AC (CA)8AG (GAG)4C (GT)8C
12 12 11 9
580^1560 600^1500 600^1600 350^1700
proteic genetic marker on individuals with di¡erent numbers of nephridiopore pairs. To accomplish this, we employed inter simple sequence repeats (ISSRs) (Zietckiewicz et al., 1994) on individuals with ¢ve and six nephridiopore pairs collected at six Mediterranean localities. The ISSR technique relies on the ampli¢cation of the DNA region between closely-spaced, inversely oriented microsatellites (or simple sequence repeats, SSRs) by means of a single primer composed of a short microsatellite sequence with one to four degenerate nucleotides anchored at the 5’ or 3’ end.
MATERIALS AND METHODS Sampling, nephridiopore count and DNA extraction Figure 1. Location of the six sandy beaches where samples of Ophelia were collected. Sardinian samples: PL, Platamona; PF, Porto Ferro and MG, Mugoni. Sicilian samples: CF, Cefalu'; CT, Catania and CB, Calabernardo.
identifying two ‘forms’ along Atlantic and Mediterranean French coasts. The two Atlantic forms were assigned to O. bicornis (15 gill pairs) and O. radiata (14 gill pairs). However, these authors did not solve the problems of: (i) the occurrence of asymmetrical individuals; and (ii) the ¢nding of controversial results for the Mediterranean population, which shared alleles with both the Atlantic O. bicornis and O. radiata. Recently, a study employing allozyme and morphological markers gave support to the existence of two valid species of Ophelia in the western Mediterranean (O. bicornis and O. barquii) (Maltagliati et al., 2004). Unreliability of many morphological traits was found with the exception of the number of nephridiopore pairs, that were consistent with allozyme patterns in separating O. barqui from O. bicornis, validating Pilato et al.’s (1978) observations. Interestingly, in some localities the two species were found in sympatry with di¡erent abundances, O. bicornis being more abundant than O. barquii (Maltagliati et al., 2004). Taxonomic resolution power of allozymes has been criticized by a number of authors because of possible nonneutrality; thus, the aim of the present work was to verify the correctness of the taxonomic separation of O. barquii from O. bicornis in the Mediterranean by using a nonJournal of the Marine Biological Association of the United Kingdom (2005)
A total of 30 specimens of Ophelia bicornis sensu lato was collected from the upper intertidal zone at three sandy beaches in Sardinia and three in Sicily in June 2004 (Figure 1). Five specimens were collected at each of the following Sardinian localities, Platamona (PL: 40851’N 08835’E), Porto Ferro (PF: 40841’N 08812’E) and Mugoni (MG: 40837’N 08813’E), and Sicilian localities, Cefalu' (CF: 38802’N 13859’E), Catania (CT: 37830’N 15805’E) and Calabernardo (CB: 36852’N 15808’E) (Figure 1). The array of ¢ve individuals collected at each locality will be referred as ‘sample’ hereafter. Nephridiopores of live specimens were counted under a stereomicroscope, then individuals were sacri¢ced in absolute ethanol and stored at 48C until genetic analysis was conducted. Genomic DNA was extracted from portions of individuals corresponding approximately to 20 mg using QIAGEN1 DNeasy Tissue kit (QIAGEN Inc., Valencia, California) according to the manufacturer’s instructions. Once extracted, DNA was stored in solution at 48C until ISSR ^ polymerase chain reaction (PCR) ampli¢cations. PCR ampli¢cations and visualization of ISSRs
The ISSR ^ PCR products correspond to DNA sequences delimited by two inverted microsatellites. ISSRs are generated by protocols very similar to those of RAPDs (random ampli¢ed polymorphic DNA), except that ISSR primer sequences are planned for microsatellite regions. Moreover, ISSRs are quick and easy to handle, and they have reproducibility because the longer lengths
1000
900
800
750
700
650
600
580
1500
1350
1300
1100
1050
950
820
800
780
750
700
600
1600
1450
1400
1350
1300
1050
1000
900
780
680
600
1700
1350
1200
830
720
620
550
380
350
Mugoni
0 0 0 1 0
0 0 0 0 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 1 1 0 0
1 1 1 0 0
0 0 0 0 0
1 0 0 1 1
1 1 1 0 0
0 0 1 1 1
1 1 0 0 0
1 1 1 0 0
0 0 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 0 1 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
1 1 0 0 0
1 0 0 0 0
0 0 0 0 0
1 1 1 0 1
0 0 0 0 0
1 0 1 1 1
0 0 0 0 0
1 1 1 0 1
0 1 1 1 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 0 0 0 0
0
5+5 Calabernardo
1 1 1 1 1
0 0 0
1 0 0 0 0
0 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 1 1 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 0 0 1
0 0 0 0 0
0 0 0 0 0
1 1 0 1 1
1 1 1 1 1
0 0 0 0 0
1 1 1 0 0
1 1 0 0 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 0 1
0 0 0 0 0
1 1 0 1 1
1 1 1 0 1
0 0 0 0 0
Catania
1 1 1 1 1
0 0 0 0 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 1 0 0 0
0 0 0 0 0
0 1 1 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 0 0 0
1 1 1 1 1
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 0 0 0 0
1 0 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 0 0 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 1 1 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 0 1 1 1
0 0 0 0 0
1 1 1 1 1
0 1 1 1 1
0 0 0 0 0
Cefalu'
0 0 0 0 0
0 0 0 0 0
1 0 1 0 0
1 1 0 1 0
1 1 1 1 1
1 1 1 1 1
0 1 0 0 0
0 0 0 0 0
0 0 1 1 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 0 0 0 0
0 1 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 1 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 1 1 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 1 0 1 1
1 1 1 1 1
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 1 0 0
0 1 0 0 1
Porto Ferro
0 0 0 0 0
0 0 0 0 0
1 0 1 1 1
0 0 0 0 0
1 1 1 1 1
1 1 0 0 0
1 1 0 0 0
1 1 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 0 0 0 0
1 1 0 0 0
0 0 1 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 0 0 0
0 0 0 0 0
1 1 1 1 1
1 1 0 1 0
1 1 1 1 1
1 1 1 1 1
0 1 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 1 1 1
0 0 0 0 0
1 1 0 1 1
0 0 0 0 0
0 0 0 0 0
1 1 0 1 1
0 0 0 0 0
0 0 0 0 0
0 1 1 0 1
0 1 1 0 1
1 1 1 1 1
0 1 1 0 0
1 1 0 1 0
1 0 0 1 1
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
0 0 0 1 0
0 1 0 0 1
0 1 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 1 0
0 0 0 0 0
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
0 0 0 0 0
0 0 0 0 0
0 0 0 0 1
1 1 0 0 0
0 0 0 0 0
1 1 1 1 0
0 0 0 0 0
0 0 0 0 0
1 1 1 1 1
bp
Morphotype
(O. barquii)
Morphotype 6+6
(O. bicornis)
Platamona
0
F. Maltagliati et al. 837
1200
PT1
1300
SAS3
1400
IT3
1550
IT2
ISSR distinguished Ophelia barquii from O. bicornis
Journal of the Marine Biological Association of the United Kingdom (2005)
Table 2. Inter simple sequence repeats primers, approximate band sizes (bp) and pro¢les obtained for 30 individuals of Ophelia from six Italian localities. Morphotype(species)-private bands are in bold.
838
F. Maltagliati et al.
ISSR distinguished Ophelia barquii from O. bicornis
Table 3. Matrix of combined probabilities for each pairwise comparison between the six samples of Ophelia. Probabilities of signi¢cantly di¡erent samples are in bold.
Porto Ferro Mugoni Cefalu Catania Calabernardo
Platamona
Porto Ferro
Mugoni
Cefalu Catania
1.0000 0.0051 1.0000 0.0026 0.0000
0.0839 0.9999 0.0052 0.0001
0.0008 0.9482 0.7943
0.0037 0.0000
bromide solution for 15 min. The ISSR banding patterns on gels were visualized using a photo-UV transilluminator system and recorded by digital photography. We obtained bands corresponding to dimensions of ampli¢ed DNA fragments of approximately 350 to 1700 base pairs (bp) (Table 1). One hundred bp ladders (DNA Molecular Weight Marker XIV, Roche1) were run for reference with each primer. Statistical analyses
1.0000
and higher annealing temperatures of their primers decrease the amount of template primer mismatch artefacts typical of RAPD ampli¢cations. We used four primers provided by Proligo1 Primers and Probes, Proligo France SAS (Table 1). The PCR reaction mixture of 25 ml volume contained 0.5 units of Taq DNA polymerase (Pharmacia1), 1reaction bu¡er (Pharmacia1), 2.5 mM MgCl2, 0.2 mM primer, 200 mM of each dNTP (Roche1), and up to 30 ng of genomic DNA. The PCR ampli¢cation was performed in a MJ PTC-100 Thermal Cycler (MJ Research1) programmed for 1 cycle of 3 min at 948C, 45 cycles of 40 s at 948C, 45 s at 558C and 1min and 40 s at 728C to complete partial ampli¢cation. At the end a post-treatment for 5 min at 728C and a ¢nal cooling at 48C were performed. For each primer, negative controls and replicates were included in the ampli¢cations in order to exclude PCR artefacts and verify repeatability of results. The PCR products were analysed by electrophoresis using a 2% agarose gel (1520 cm) in 1TAE bu¡er (0.04 M Tris-acetate, and 0.001M EDTA). Gels were run at 80 V for 3 hand stainedbysoaking ina1 ml/10 mlethidium
Like RAPDs, ISSRs are treated as dominant diallelic markers where the dominant allele A determines the presence of the band; hence, AA and Aa individuals produce bands (phenotype 1), while aa individuals do not (phenotype 0). Sample genetic di¡erentiation was assessed with the program TFPGA (web site http://bioweb. usu.edu/mpmbio/index.htm) by using an exact test. This analysis uses a contingency table approach to determine if signi¢cant di¡erences in band frequencies exist among groups of individuals. Hierarchical relationships were estimated by analysis of molecular variance (AMOVA). Total genetic variation was partitioned into: (i) between morphotypes; (ii) among samples within morphotype; and (iii) within sample using AMOVAPREP (web site http://bioweb.usu.edu/mpmbio/index.htm) and WINAMOVA (Exco⁄er et al., 1992). Signi¢cance levels were calculated using randomization tests. A null distribution was obtained by allocating each individual to randomly chosen populations and the variance components estimated from 10,000 permutations. This procedure eliminated the normality assumption required for the analysis of variance but which is inappropriate for molecular data (Exco⁄er et al., 1992). For FST (within samples), the variance component was tested by randomizations across all samples; for FSC (among samples within morphotype), it was assumed that the morphotypes are
Table 4. Results of hierarchical analysis of molecular variance (AMOVA) derived from the cluster analysis computed from the distance matrix constructed using Exco⁄er et al.’s (1992) formula. P-values, calculated from a random permutation test (10,000 replicates), and F-statistics represent the probability of obtaining by chance alone a more extreme variance than the observed values (Exco⁄er et al., 1992).
Source of variation
df
MS
Variance component
Percentage of variance
F-statistics
P
Between morphotypes Among samples within morphotype Within samples
1 4 24
79.833 13.667 3.233
4.411 2.087 3.233
45.3 21.4 33.3
FCT ¼0.453 FSC ¼0.392 FST ¼0.668
50.001 50.001 50.001
Table 5. Results of the assignment tests between the six Ophelia samples analysed. The values in the table matrix represent the number of individuals from the source location (rows) that were assigned by genotype to a particular location (columns).
Platamona Porto Ferro Mugoni Cefalu Catania Calabernardo
Platamona
Porto Ferro
Mugoni
Cefalu
Catania
Calabernardo
5 0 0 0 0 0
0 5 0 0 0 0
0 0 5 0 0 0
0 0 0 5 0 0
0 0 0 0 4 1
0 0 0 0 1 4
Journal of the Marine Biological Association of the United Kingdom (2005)
ISSR distinguished Ophelia barquii from O. bicornis
F. Maltagliati et al. 839
Figure 2. The UPGMA consensus dendrogram of Nei’s (1978) genetic distances between samples. Bootstrap values were obtained after 10,000 replicates.
Figure 3. Non-metric multidimensional scaling of 30 Ophelia individuals based on genetic dissimilarity matrix [17Nei & Li’s (1985) similarity] calculated from inter simple sequence repeats data. Circles and squares correspond to individuals with ¢ve (O. barquii) and six nephridiopore pairs (O. bicornis), respectively. Locality abbreviations are given in Figure 1.
real but samples are not, so that randomizations occurred within morphotypes; and for FCT (between morphotypes), it was assumed that the samples were real and the morphotypes were arti¢cial, so that randomizations of samples were made across morphotypes. The UPGMA cluster analysis on pairwise Nei’s (1978) genetic distances between samples was carried out to construct a dendrogram using the program TFPGA. Nodes of the dendrogram were tested using bootstrapping with 10,000 replicates. The ISSR patterns were further analysed by the program RAPDPLOT 3.0 (web site ftp:// lamar.colostate.edu/pub/wcb4/) to generate a matrix of 17Nei & Li’s (1985) similarity index between individuals. Similarity index was de¢ned as S¼2NAB/(NA+NB), where NAB is the number of bands that individuals A Journal of the Marine Biological Association of the United Kingdom (2005)
and B share in common and NA is the number of bands in individual A and NB is the number of bands in individual B. Non-metric multidimensional scaling (nMDS) was applied to the dissimilarity matrix in order to reveal possible groupings of the individuals using the program STATISTICA 5.1 (web site http://www.stasoft.com). This analysis uses a function minimization algorithm that evaluates di¡erent con¢gurations with the goal of maximizing the goodness-of-¢t. Stress index measures reliability of the nMDS plot: the smaller the stress index, the better the ¢t of the reproduced distance matrix to the relative distances on the plot. Further, an assignment test was performed on individuals of the six localities using the calculator available online at http://www2.biology.ualberta.ca/jbrzusto/ Doh.php. This calculator takes genotypes of individuals
840
F. Maltagliati et al.
ISSR distinguished Ophelia barquii from O. bicornis
from each population and determines from which population each individual is most likely to have come, by using the highest probability of an individual’s genotype in any of the populations. Probabilities were obtained by permuting 10,000 times genotypes within populations.
RESULTS AND DISCUSSION Integration and cross-validation of di¡erent types of data are desirable tools for good taxonomic practice. This is particularly true with polychaetes, a taxon often comprising widely distributed species which lack reliable morphological characters. Indeed, a number of studies reported that molecular techniques coupled with morphological analyses allowed di¡erentiation of taxonomic di¡erences at a very ¢ne scale (e.g. Schmidt & Westheide, 1999; Scaps et al., 2000; Maltagliati et al., 2001, 2004). In this study specimens collected at Mugoni, Catania and Calabernardo exhibited ¢ve nephridiopore pairs, while those collected at Platamona, Porto Ferro and Cefalu' six pairs. In general, band presences varied in several ways across samples (Table 2). The number of nephridiopore pairs and ISSR patterns yielded perfectly congruent results in distinguishing two groups of individuals, denoting the absence of hybridization and introgression between the two morphotypes. This also suggested complete reproductive isolation between individuals of di¡erent morphotypes, resolving the controversy regarding the taxonomic status of Ophelia bicornis sensu lato in the Mediterranean. Indeed, in this region, Ophelia barquii (Fauvel, 1927), with ¢ve nephridiopore pairs, is a valid species, separated from Ophelia bicornis Savigny, 1818, with six nephridiopore pairs. This result represents a sound con¢rmation based on non-proteic genetic markers of the recent morphological and allozyme survey on western Mediterranean populations (Maltagliati et al., 2004). The DNA fragment IT3-800 was produced by all 15 individuals of O. bicornis analysed and no bands were exhibited by the 15 individuals of O. barquii; conversely, the band IT3-820 was produced only by all individuals of O. barquii analysed (Table 2). Therefore, IT3-800 and IT3820 can be considered fully diagnostic loci and the oligonucleotide IT3 a ‘diagnostic primer’. Previous allozyme attempts by Britton-Davidian & Amoreux (1982) and Maltagliati et al. (2004) failed in detecting fully diagnostic loci for Ophelia species. In particular, in the latter investigation the two morphotypes based on the number of nephridiopore pairs exhibited only nearly ¢xed di¡erences at the loci GPI* and IDH-2*. These loci could not therefore be considered fully diagnostic. Moreover, the present study revealed a large number of ISSR loci (29 out of 44¼66%) that can be considered as nearly diagnostic, namely loci that exhibit species-private bands but absence of bands is common to individuals of both species (Table 2). The separation of O. bicornis and O. barquii was ¢rmly supported by statistical evidence. With the exception of the comparison between samples from Porto Ferro and Mugoni, exact tests produced signi¢cant values for all pairwise combinations between di¡erent morphotypes, whereas within-morphotype pairwise comparisons did not (Table 3). The analysis of molecular variance (Table 4) indicated that the greatest portion of genetic variation was Journal of the Marine Biological Association of the United Kingdom (2005)
found between individuals with di¡erent numbers of nephridiopore pairs (45.3%). The variances among samples within morphotype (21.4%) and within samples (33.3%) accounted for the remaining molecular variance (Table 4). The F-statistics highlighted signi¢cant genetic di¡erentiation between the two morphotypes (FCT ¼ 0.453, P5 0.001) (Table 4), con¢rming the separation in di¡erent species. However, FSC and FST were signi¢cantly di¡erent from zero as well, suggesting: (i) population structuring within each of the two morphotypes; and (ii) within-population genetic heterogeneity, respectively. Although the number of individuals per population herein analysed was not su⁄cient to reliably obtain within-population estimates of genetic variability, the signi¢cant within-sample genetic heterogeneity detected could be due to the intrinsic hypervariability of ISSR markers. Interestingly, similar within-sample genetic heterogeneity was obtained in an ISSR survey of the north-western Atlantic brooding bivalve Gemma gemma, where FST relative to samples collected within 10 m patches was signi¢cantly high, whereas FCT between two localities hundreds of km distant does not (Casu et al., 2005). The UPGMA analysis of Nei’s (1978) genetic distances grouped samples in two distinct clusters, each comprising individuals of one morphotype, regardless of geographical distance between samples (Figure 2). All nodes were provided with high bootstrap support (470), with the exception of the node linking Porto Ferro and Platamona (bootstrap value¼51). Reliability of nMDS plot based on 17Nei & Li’s (1985) similarity index (Figure 3) was ensured by the low value of stress index (s¼0.080), indicating that distances among individuals on the plot accurately represented their genetic distances. The ¢rst dimension of the nMDS plot separated clearly individuals with ¢ve nephridiopore pairs from their six pairs counterparts (Figure 3). Moreover, in the cloud of points distributed on the left part of the plot two groups were distinguishable. The ¢rst group corresponded to the two Sicilian samples from Catania and Calabernardo (intermixed in the upper left part of the plot) and the second to the Sardinian sample from Mugoni (in the lower left part of the plot) (Figure 3). This suggested the presence of subspeci¢c phylogeographical structuring in O. barquii, given that the Sardinian sample and the two Sicilian ones are located in the western and eastern Mediterranean basins, respectively. Results of the assignment tests showed that all individuals were assigned to the respective morphotypes and localities, with the exception of one individual from Catania that was assigned to Calabernardo and one individual from Calabernardo that was assigned to Catania (Table 5), suggesting a degree of connectivity between these two geographically close populations of O. barquii. However, the analysis of a larger number of both localities and individuals per locality is needed to give sound insight on species genetic structure and between-population connectivity (gene £ow). This study indicated a need for an overall taxonomic revision of the genus Ophelia employing a morphological/ genetic approach. The ¢rst step should be the clari¢cation of the taxonomic status of O. bicornis sensu lato across its extra-Mediterranean range. Within this framework, the
ISSR distinguished Ophelia barquii from O. bicornis diagnostic ISSR primer IT3 used in this study coupled with the count of nephridiopore pairs could be tested on extra-Mediterranean individuals of O. bicornis sensu lato and other Ophelia species. Cross-validation employing other classes of markers such as ecological, behavioural or physiological traits are needed in order to fully understand the mechanisms of sympatric speciation hypothesized by Maltagliati et al. (2004). Furthermore, the exact distribution of O. bicornis and O. barquii and their seasonal variation in abundances and reproductive characteristics, as for instance the shifting of the reproductive period of one species in respect to the other, remain to be described. This research was ¢nancially supported by the European Community INTERREG III, involving research groups from Sardinia, Tuscany, and Corsica and by the ‘Centro di Eccellenza’ of the University of Sassari, Italy. We wish to thank Anna Gelosa, who kindly facilitated the collection of the samples from Mugoni and Porto Ferro and Valentina Bacciu, who assisted during laboratory work.
REFERENCES Amoreux, L., 1977. Ophelia bicornis Savigny, 1818, Ophelia radiata (Delle Chiaje, 1828), two phenotypical forms of the same species. In Essays on polychaetous annelids in memory to Dr Olga Hartman. Allan Hancock Foundation Publications, pp. 267^ 278. Los Angeles: University of Southern California Press. Bellan, G., 1964. Contribution a' l’e¤tude syste¤matique, bionomique et e¤cologique des Anne¤lides Polyche'tes de la Me¤diterrane¤e. Recueil des Travaux de la Station Marine d’Endoume, 33, 1^372. Britton-Davidian, J. & Amoreaux, L., 1982. Biochemical systematics of two sibling species of polychaete annelids: Ophelia bicornis and O. radiata. Biochemical Systematics and Ecology, 10, 351^354. Cantone, G. & Costa, G., 1975. Variabilita' nel numero di branchie, delle papille perianali e delle papille rettali in una popolazione di Ophelia bicornis Savigny delle coste orientali della Sicilia (Annelida Polychaeta). Pubblicazioni della Stazione Zoologica di Napoli, 39, Supplement, 22^36. Casu, M., Maltagliati, F., Cossu, P., Lai, T., Curini Galletti, M., Castelli, A. & Commito, J.A., 2005. Fine-grained spatial genetic structure in the bivalve Gemma gemma from Maine and Virginia (U.S.A.), as revealed by Inter-Simple Sequence Repeat markers. Journal of Experimental Marine Biology and Ecology, in press. Exco⁄er, L., Smouse, P.E. & Quattro, J.M., 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA data. Genetics, 131, 479^491. Fauvel, P., 1927. Polyche'tes se¤dentaire. In Faune de France. Vol. 16. pp. 1^494. Paris: Lechevalier.
Journal of the Marine Biological Association of the United Kingdom (2005)
F. Maltagliati et al. 841
Lee, M.S.Y., 2004. The molecularization of taxonomy. Invertebrate Systematics, 18, 1^6. Maltagliati, F., Camilli, L., Lardicci, C. & Castelli A., 2001. Evidence for morphological and genetic divergence in Perinereis cultrifera (Polychaeta: Nereididae) from two habitat types at Elba Island (Italy). Journal of the Marine Biological Association of the United Kingdom, 81, 411^414. Maltagliati, F., Casu, M. & Castelli, A., 2004. Morphological and genetic evidence supports the existence of two species in the genus Ophelia (Annelida, Polychaeta) from the Western Mediterranean. Biological Journal of the Linnean Society, 83, 101^ 113. Maltagliati, F., Peru, A.P., Casu, M., Rossi, F., Lardicci, C., Curini-Galletti, M. & Castelli, A., 2000. Is Syllis gracilis (Polychaeta: Syllidae) a species complex ? An allozyme perspective. Marine Biology, 136, 871^879. Nei, M., 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics, 89, 583^ 590. Nei, M. & Li, W.H., 1985. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences of the United States of America, 76, 5269^5273. Pilato, G., Belcastro, G. & Cassibba, R., 1978. Il valore speci¢co di Ophelia barquii Fauvel, 1927 (Annelida, Polychaeta). Animalia, 5, 395^403. Scaps, P., Rouabah, A. & Lepretre, A., 2000. Morphological and biochemical evidence that Perinereis cultrifera (Polychaeta: Nereididae) is a complex of species. Journal of the Marine Biological Association of the United Kingdom, 80, 735^736. Schmidt, H. & Westheide, W., 1999. Genetic relationships (RAPD ^ PCR) between geographically separated populations of the ‘‘cosmopolitan’’ interstitial polychaete Hesionides gohari (Hesionidae) and the evolutionary origin of the freshwater species Hesionides riegerorum. Biological Bulletin. Marine Biological Laboratory, Woods Hole, 196, 216^226. Schmidt, H. & Westheide, W., 2000. Are the meiofaunal polychaetes Hesionides arenaria and Stygocapitella subterranea true cosmopolitan species? results of RAPD^ PCR investigations. Zoologica Scripta, 29, 17^27. Westheide, W. & Schmidt, H., 2003. Cosmopolitan versus cryptic meiofaunal polychaete species: an approach to a molecular taxonomy. Helgoland Marine Research, 57, 1^6. Wilson, D.P., 1948. The larval development of Ophelia bicornis Savigny. Journal of the Marine Biological Association of the United Kingdom, 27, 540^553. Zietckiewicz, E., Rafalsky, A. & Labuda, D., 1994. Genome ¢ngerprinting by Inter-Simple Sequence Repeat (ISSR)anchored polymerase chain reaction ampli¢cation. Genomics, 20, 176^183.
Submitted 14 January 2005. Accepted 20 May 2005.
