INTRODUCTION

The echinoid fauna of Japanese waters is remarkably rich and about 160 species have been recorded by Shigei (1986). With such extensive species diversity, taxonomy and phylogeny of the echinoids have been extensively studied from the morphological and/or palaeontological standpoints (Nisiyama, 1966, 1968; Shigei, 1974, 1986). However, many unresolved problems still remain for the phylogenetic and evolutionary relationships of the echinoids. For an elucidation of these problems, it would be desirable to actively introduce the molecular approaches which are more analytic and quantitative than the usual morphological methods into the field of echinoid systematics.

In the previous papers, we have reported on the biochemical systematics of echinoids of four different families (Toxopneustidae, Strongylocentrotidae, Echinometridae, and Temnopleuridae) of the order Echinoida (Matsuoka, 1980, 1985, 1986, 1987; Matsuoka and Suzuki, 1989; Matsuoka and Hatanaka, 1991; Matsuoka and Inamori, 1996), two families (Phymosomatidae and Stomopneustidae) of the order Arbacioida (Matsuoka and Nakamura, 1991), and family Diadematidae of the order Diadematoida (Matsuoka, 1989). Through these serial biochemical systematic studies, we could provide reliable and, in some cases, critical information to the echinoid systematics.

Another evolutionarily attractive echinoid group whose phylogeny has not yet been studied at molecular level is the order Cidaroida. The echinoid taxonomists claim that the order Cidaroida is a primitive echinoid group which diverged earlier, from the morphological standpoint (Shigei, 1974, 1986). The echinoids have a unique morphology when compared with the order Echinoida which is generally thought to be the recent echinoids. For example, they possess a few but thick rod-like primary spines. Furthermore, coronal plate of the test is large, the ambulacral plate is of simple type, the interambulacral plate bears single large tubercle, and the gill slits are absent. These morphological characters have been regarded as primitive by echinoid taxonomists (Shigei, 1974, 1986).

Since the order Cidaroida includes many rare species and deep-sea species, it is much more difficult to collect specimens. In addition, the number of individuals which can be sampled is decreasing rapidly now because of rapid marine pollution. Therefore, the molecular phylogenetic studies of these echinoids have not been carried out for the long time. Fortunately, we recently collected four species of the order: Stereocidaris japonica (Döderlein), Eucidaris metularis (Lamarck), Prionocidaris baculosa (Lamarck), and Phyllacanthus dubius Brandt. These echinoids are representative species of the order Cidaroida from Japanese waters. Taxonomically, they belong to the family Cidaridae of the order Cidaroida. The order Cidaroida consists of three families including fossil species alone and two families including extant species. All the living species of the order Cidaroida are classified into the family Cidaridae except for one species, and thus the family Cidaridae can be considered a representative group of the order Cidaroida (Shigei, 1974, 1986). However, as mentioned above, the number of species of the Cidaridae that are commonly found in Japanese waters are exceedingly scanty, when compared with many other echinoid groups.

More recently, Littlewood and Smith (1995) examined the sequences of 18S ribosomal RNA genes of various echinoids. The aim of their study was to clarify the relationships among different orders of the higher taxonomic rank. In the study, they provided no useful information about the phylogenetic relationships within the order Cidaroida, because they used only two European species, Cidaris cidaris and Eucidaris tribuloides, from the cidarids. Therefore, until now, the phylogenetic relationships among members of the cidarids have not yet been studied by molecular approaches, and thus the phylogeny within the cidarids still remain uncertain.

In this study, we have attempted an allozyme study to clarify the phylogenetic relationships among four species of the family Cidaridae (Cidaroida) from Japanese waters. This is the first report of biochemical systematics of the cidarids using allozyme analysis.

MATERIALS AND METHODS

RESULTS AND DISCUSSION

Eighteen genetic loci were inferred from the allozyme variation observed in the 11 different enzymes. Table 1 shows the allele frequencies at 18 genetic loci. Nine loci (Xdh, Pgm, Sod-2, Alk-1, Alk-2, Est-2, Est-4, Amy-1, and Amy-2) were monomorphic and exhibited a single active band which varied inter-specifically. Est-3 showed single- and double-banded phenotypes between species, and they were assumed as the products of two different alleles. The remaining eight loci (Mdh, Odh, Pgi, Sod-1, Po, Alk-3, Est-1, and Lap) were polymorphic in at least one species. In general, the enzymes such as SOD, EST and LAP are highly polymorphic in echinoids or asteroids.

Table 1

Allele frequencies at 18 genetic loci in the four echinoids of the family Cidaridae

Relationship between the estimation of average heterozygosity (H) and sample size has been reported by some workers in the electrophoretic surveys of genetic variation within population (Nei and Roychoudhury, 1974; Nei, 1978; Gorman and Renzi, 1979). Their general conclusion was that even a few individuals are sufficient for estimating H value if the number of loci examined is large. Furthermore, Nei (1987) reported that 20–30 loci have often been used in allozyme studies and the number of loci is adequate for estimating H values. Bearing their findings in mind, we calculated expected average heterozygosity per locus (H) in the four species. The H values obtained were 7.4% for S. japonica, 0% for E. metularia, 14.0% for P. baculosa, and 3.3% for Ph. dubius. We have previously reported the amount of genetic variation within populations of various echinoderm species (Matsuoka et al., 1993). According to it, the H value (14.0%) of P. baculosa was comparable to those of many other deep-sea echinoderms, while the H values of other three species (E. metularia and Ph. dubius are shallow water echinoids) were comparable to those of many other shallow water echinoderms.

To quantify the degree of genetic differentiation between the four species, we calculated genetic identity (I) and genetic distance (D) between two of each species by the method of Nei (1972) from the allele frequencies data in Table 1. Table 2 shows the matrices of I and D values between all pairs of four species. The I and D values between four cidarids were in the range of 0.141–0.417, with a mean of 0.289, and in the range of 0.875–1.959, with a mean of 1.315, respectively.

Table 2

Genetic identities (above diagonal) and genetic distances (below diagonal) between four species of the family Cidaridae

Although the number of individuals used in this study is small, it has been shown that the number of loci used has far more effect on the D estimates than the number of individuals sampled (Nei and Roychoudhury, 1974; Nei, 1978; Gorman and Renzi, 1979). Indeed, it has been demonstrated in the study of Anolis lizards that the D estimates from a sample size of only one or two individuals deviate only slightly from those from a much larger sample size (Gorman and Renzi, 1979). Furthermore, our serial allozyme studies on the phylogeny of echinoderms suggested that in the phylogenetic studies with allozyme analysis at higher taxonomic rank of different genera the number of individuals has hardly any effect to the estimation of the genetic distance and the phylogenetic relationships (Matsuoka, unpublished data).

Table 3 summarizes the I and D values observed between two confamilial genera of the orders Echinoida and Diadematoida reported previously (Matsuoka, 1987, 1989; Matsuoka and Suzuki, 1989; Matsuoka and Inamori, 1996). Furthermore, Table 4 shows the I and D values observed between two echinoids of different families. All the data were compiled from our serial biochemical systematics of echinoids by using allozyme analysis (Matsuoka, 1987; Matsuoka and Nakamura, 1991), and thus they can be compared at the equivalent level. As evident from Table 3, the mean I and D values between 28 pairs of confamilial genera were 0.501 and 0.706, respectively. The present mean I and D values between 6 pairs of the family Cidaridae were 0.289 and 1.315, respectively, and these values were not comparable to those between confamilial genera in other echinoids. On the other hand, the mean I and D values between 13 pairs belonging to different families were 0.295 and 1.245, respectively, and these values were comparable to the present values. Namely, the genetic distances between four species of the family Cidaridae are considerably higher (about two times) than those between genera in many other echinoids, and rather equivalent with those between families.

Table 3

Genetic identity (I) and genetic distance (D) between two species of confamilial genera in various echinoids

Table 4

Genetic identities (I) and genetic distances (D) between two echinoids belonging to different families

Nei (1975) reported that the genetic distance (D) estimated electrophoretically corresponds well with the divergence time (T) from the common ancestor, and that T of two taxa can be calculated by T=5×10⁶D (years). Judging from the findings, it is considered that these four cidarid species diverged at earlier time: the present allozyme study indicated that four cidarids studied here are much more genetically differentiated from one another than members of families of the orders Echinoida and Diadematoida.

Although the higher genetic distances more than D =1 were observed in this study, the similar values have also obtained in the allozyme study among seven echinoids belonging to the two different families, Toxopneustidae and Strongylocentrotidae, of the order Echinoida, as mentioned above (Matsuoka, 1987). In the biochemical dendrogram constructed, the seven species of the two families examined were clearly divided into two large clusters corresponding to the two families. Further, the allozyme results including the relationships among taxa at species and genus level were strongly supported by the other biochemical and non-molecular studies including the morphological studies. It indicates that even the higher genetic distances more than D =1 can provide reliable information to the estimation of the phylogenetic relationships among taxa (Matsuoka, 1987; Matsuoka et al., 1994). Thus, it is considered that the higher genetic distances among the cidarids obtained in this study have hardly any effect to the estimation of the phylogeny of the cidarids.

To clarify their phylogenetic relationships, the molecular phylogenetic tree for four species was constructed from Nei's genetic distance matrix of Table 2 by using the UPGMA clustering method of Sneath and Sokal (1973). The phylogenetic tree (Fig. 1) indicated the following: (1) S. japonica and Pr. baculosa are most closely related to each other among the four species and diverged later. (2) E. metularia is more closely related to the cluster of S. japonica and Pr. baculosa than Ph. dubius. (3) Ph. dubius is the most distant among the four species and diverged first.

Fig. 1

Molecular phylogenetic tree for the four echinoids of the family Cidaridae (Cidaroida) which was constructed from the Nei's genetic distances (Nei, 1972) by using the UPGMA clustering method. The divergence time estimated from the Nei' equation (Nei, 1975) using the genetic distance is also given in the phylogenetic tree.

The molecular phylogenetic tree (Fig. 1) shows not only their relationships, but also the sequence of their evolutionary divergences. From the equation of Nei (1975), their divergence times were calculated. According to them, Ph. dubius diverged firstly about 8 million years ago (MY). Next, E. metularia diverged from the common ancestor of S. japonica and Pr. baculosa about 6 MY. Lastly, the divergence of S. japonica and Pr. baculosa occurred about 4 MY. These divergence times are considerably older than those of echinoids belonging to the orders Echinoida and Diadematoida (Matsuoka, 1987, 1989; Matsuoka and Suzuki, 1989; Matsuoka and Inamori, 1996), and it shows their older evolutionary origin.

With respect to the taxonomic relationships among the four cidarids, Shigei (1974) classified the four species into three subfamilies based on the morphology of globiferous pedicellaria: the subfamily Stereocidarinae for S. japonica, the Cidarinae for E. metularia, and the Rhabdocidarinae for Pr. baculosa and Ph. dubius. In this system, he suggested the close affinity between Pr. baculosa and Ph. dubius. Thereafter, Shigei (1986) revised his previous subfamily system and proposed the new subfamily system based on the morphology of ambulacral plates and spines besides globiferous pedicellaria. Under the revised system, the present four cidarids are classified into the following three subfamilies: the subfamily Stereocidarinae for S. japonica, the Stylocidarinae for Pr. baculosa and E. metularia, and the Phabdocidarinae for Ph. dubius. The present molecular study clearly indicated the close affinity between Pr. baculosa and S. japonica. Shigei (1976, 1986) claimed in his two subfamily systems that S. japonica is an isolated species from other members of the family Cidaridae. However, the present allozyme study does not support his view concerning the systematic position of S. japonica. On the other hand, Shigei (1986) suggested that Ph. dubius may be a phylogenetically distant species from other cidarids, based on the unique morphological structures of peristomial ambulacral plates, secondary tubercles and so on. The present allozyme result strongly supported his view for the systematic position of Ph. dubius.

There exists a morphologically very similar echinoid to Ph. dubius. The species is Phyllacanthus imperialis. In contrast to the fact that Ph. dubius is endemic to Ogasawara Islands, Ph. imperialis is widely distributed in the Indo-West Pacific Ocean. More recently, Shigei (1994) suggested that these two echinoids may be conspecific from the morphological standpoint. The previous allozyme study demonstrated that the tropical echinoid, Echinometra mathaei consists of four distinct species (Matsuoka and Hatanaka, 1991). Therefore, the problem whether Ph. dubius and Ph. imperialis are the same single species or two distinct species may be clarified by allozyme study. In the near future, if Ph. imperialis is collected, we are planning to investigate their taxonomic and genetic relationship by allozyme analysis.