Allozymic variation in the clam genus Eurhomalea ( Bivalvia : Veneriidae ) along southern South American coast Variación alozímica en el género de almejas Eurhomalea ( Bivalvia : Veneriidae ) a lo largo de la costa sur de

The correspondence between allozymic variation and specific differentiation was studied in four populations corresponding to the three nominal, allopatric species of clam genus Eurhomalea (E. rufa, E. lenticularis, E. exalbida) described for southern South America. Allozyme variation scored in 12 loci was high as indicated by heterozygosity levels (15.8-20.7 %) and by the presence of only three monomorphic loci (Hk-2, Icd-2, and Xdh-1). These high estimates of allelic variability were influenced by the low levels of interspecific genetic similarity (I = 0.64) and for the high conspecific values of genetic identity observed. The high estimates of substructuring found at the species level (FST = 0.39) contrasted with the low differentiation (FST = 0.027) and high migration rate (Nm = 9) existing among conspecific samples. Diagnostic allele fixation coinciding with specific recognition was recorded at locus Hk-1 whereas nearly-fixed differences at loci (Adh, α-Gpd, Icd-1, Ldh, Odh, Pgm-3) differed sharply in frequency among species. The Wagner procedure and the neighborjoining algorithm produced a similar tree topology highly related to the geographic distance and to their taxonomic recognition. The lack of coincidence between patterns of allozymic variation and the two distinctive shell morphs (flat and globoid) occurring in E. exalbida from Ushuaia bay do not support their taxonomic recognition.

Traditionally, three allopatric species of Eurhomalea (E.rufa, E. lenticularis, and E. exalbida) have been recognized in southern South America.The distributional range of E. rufa extends from Panama to northern Chile whereas E. lenticularis has a restricted distribution limited to the coast of central Chile.Eurhomalea exalbida has the southernmost and extended distribution, from Chiloé island to Buenos Aires (Carcelles 1950, Osorio et al. 1979, Reid & Osorio 2000).In spite of being an exploited resource (anonymous 1998), no comparative study has yet been published on the genus.In fact, the few and scattered research on ecological variables, growth and production, reproduction, recruitment rate, survival, and energetic variation have been conducted only at the population level (Urban & Tesch 1996, Jaramillo et al. 1998, Sousa et al. 1998, Campos et al. 1999, Lomovasky et al. 2001, 2002a, 2002b, Morriconi et al. 2002).Morphological characters in Eurhomalea have shown to be somewhat confusing for systematic purposes (Soot-Ryen 1959, Osorio et al. 1979, Ramírez 1993, Reid & Osorio 2000).Recent paleontological data suggest that E. exalbida and E. lenticularis should be grouped in a new genus, Retrotapes (Del Río 1997).In addition, two highly distinctive shell morphs are suspect to represent different age groups in E. exalbida from Ushuaia bay, Argentina (Lomovasky personal communication).
Here, the allozymic relationships and the degree of population structuring of Eurhomalea species are examined by electrophoretic markers.Having in mind the morphological differences observed in E. exalbida, we aimed to test whether there was a correspondence between shell morphology and the electromorphic patterns recorded for each one of them.

RESULTS
Although high standard measures of genetic variation were found in Eurhomalea, our estimates are within the limits reported for other bivalves (Ayala et al. 1973, Buroker et al. 1979a, 1979b, Gallardo et al. 1998).Heterozygosity level and percent polymorphism (x = 49.3 %) was high as reflected by the twelve variable loci having up to six different alleles across populations (Table 1).Direct-count heterozygosity fluctuated between 15.8 and 20.7 % and significant heterozygote deficiency (ranging from three to four loci) was found across the sampled localities (Table 1).Diagnostic alleles differences were found among the three species at locus Hk-1.Nearlyfixed allelic differences between rufa-like, lenticularis-like and exalbida-like samples ocurred at six loci (Adh, α-Gpd, Icd-1, Ldh, Odh, and Pgm-3).Loci α-Gpd, and Icd-1 were diagnostic for the discrimination of E. rufa from the other two species.Locus Odh in E. exalbida was fixed or nearly-fixed for allele B, whereas allele A is fixed in E. lenticularis.
Nei's genetic identity was very high among conspecific samples of E. exalbida (I = 0.989) but it was much lower at the interspecific level (I = 0.724 between E. exalbida and E. lenticularis).Genetic identity was 67.7 % between E. exalbida and E. rufa and it was 53.8 % between E. lenticularis and E. rufa.Lack of discrimination and high identity values (I = 0.995) were obtained between the two sympatric morphotypes of E. exalbida.
Three clusters highly consistent with both the geographic location and the specific status of the samples were produced by the Wagner tree.The first one contained the northern sample of E. rufa (Punta Choros); the second was formed by E. lenticularis from Algarrobo, and the third was formed by E. exalbida from San Juan Bay and Ushuaia Bay (not shown).An identical tree topology with high bootstrap support was also obtained with FITCH (Fig. 1).
Low estimates of population substructuring were found within the three samples of E. exalbida (F ST = 0.027).The migration rate derived from this F ST value corresponds to an exchange of nine individuals per generation.Nevertheless, if the five samples are treated as conspecifics, the standardized variance of gene frequencies (F ST = 0.39) suggest an exchange rate of 0.4 individuals per generation.

DISCUSSION
A high variability in allozymic patterns has been reported in marine organisms (Ayala et al. 1973, Campbell et al. 1975, Nevo 1990, Dufresne et al. 2002, Gardner & Thompson 2002).Mean polymorphism (49 %) and the average heterozygosity level (18.3 %) is high in Eurhomalea but within the range of other marine invertebrates (Buroker et al. 1979a, 1979b, Nevo et al. 1984, Day & Bayne 1988, Backeljau et al. 1994, Gallardo & Carrasco 1996).The observation of an overall deficiency of heterozygotes is a feature common in other studies on more than 50 species of mollusks such as mytilids, pectinids, ostreids and cardids (Nevo et al. 1984).
The electrophoretic data recorded for the total Eurhomalea populations provides further evidence for a generalized heterozygote deficiency at enzymatic loci in bivalves.A number of possible explanations for this deficiency include the presence of null alleles, mis-scoring of gels, inbreeding, aneuplody, the Wahlund effect, non random mating, and selection against heterozygotes (Zouros & Foltz 1984, Foltz 1986, Zouros et al. 1988, Gaffney et al. 1990, Gardner & Thompson 1999).Our analyses do not permit an assessment in the detection of the factors responsible for the residual heterozygote deficiencies.Deviation of genotypic proportions may reflect an incorrect procedure for obtaining genotype data, or alternatively, that these populations depart from Hardy Weinberg equilibrium.
A consistent association between geographic distribution and allozymic differentiation was generated by both clustering procedures.Resulting from the alternative fixation or nearlyfixation of alternative alleles, these topologies reflect also a clear segregation into discrete genetic units.In contrast, low intraspecific F ST estimates and Nei's distance values reflect regional cohesiveness and essential lack of population substructuring in E. exalbida.The high intraspecific estimates of gene flow (Nm = 9) suggest that either for a two-dimensional stepping stone or an island model, random differentiation through genetic drift is prevented (Slatkin 1985).Previous electrophoretic surveys have established that Nei's mean genetic identity is ≤ 0.85 in about 97 % of interspecific comparisons, whereas in 98 % of intraspecific comparisons is more than 0.85 (Ayala 1975, Skibinski et al. 1980, Sarver et al. 1992, Backeljau et al. 1994).Our results for the conspecific populations of E. exalbida was I = 0.989 whereas the same parameter estimated at the interspecific level fluctuated between 0.538 and 0.724.These results are totally consistent with the morphological and distributional criteria used for species recognition in Eurhomalea as supported by its correspondence with the demarcation limits of allozymic data above mentioned.
Based on the evolutionary species concept that emphasizes its genetic integrity over hybridization (Wiley 1981), allozymic studies have recognized sibling species by the presence of fixed allelic differences and large genetic distances (Chambers 1978, Woodruff 1988).Our data also show that Eurhomalea species are allozymically different as demostrated in other molluscan taxa (Liu et al. 1991).The intrapopulational morphs of shell shape (flat and globoid) found in E. exalbida from Ushuaia, assumed to represent reproductively isolated units has no genetic basis to substantiate their taxonomic recognition.Most probably, these morphs represent age groups having distinctive growth rates (Lomovasky, unpublished data).Accordingly, this trait should be considered as a polymorphic one.
Since fishery management plans require basic biological knowledge to determine the conservation status of any exploited resource, the existence of three Eurhomalea species and a detailed knowledge of its conservation status should be considered in future governmental actions.Árbol de consenso mayoritario derivado de una matriz de distancia (neighbor joining) de cinco poblaciones de Eurhomalea.Los valores de bootstrap mostrados en los nodos son una medida de repetitividad basada en 1.000 iteraciones.

Fig. 1 :
Fig.1: Majority-rule consensus tree derived from a distance matrix (neighbor-joining) of five populations of Eurhomalea.Bootstrap values shown at nodes are a measure of repeatability based of 1,000 bootstrap iterations.