1. Systematics and Distribution

Cryptopecten Dall, Bartsch and Rehder, 1938 [type-species: Cryptopecten alli Dall, Bartsch and Rehder, 1938 (=Pecten bullatus Dautzenberg and Bavay, 1912), by original designation] is a small-sized pectinid genus, now living on sublittoral and upper bathyal sandy bottoms in the warm waters of the Indo-Pacific and western Atlantic. Fossils of this genus are considerably common in the Neogene and Quaternary of Japan. The taxonomic validity of this genus is confirmed by its unique characteristics of surface sculpture, that is, the regular and simple (neither bifurcated nor inserted) radial ribs, each of which consists of a central solid ridge and a pair of lateral hollow parts covered with numerous diaphragm-like imbricated scales. The shell is generally right-convex in the adult stage, but the highly allometric growth, the mode of which is quite different in the two valves, is also worthy of notice. Corymbichlamys Iredale, 1939 [type-species: Chlamys corymbiatus Hedley, 1909, by original designation] should be regarded as a junior synonym of Cryptopecten.

The following four living and two extinct species are certainly referable to Cryptopecten.

Pecten (Chlamys) bullatus Dautzenberg and Bavay, 1912; Philippines, Indonesia, Hawaii, South China Sea, Japan and Nasca Ridge (lower sublittoral to upper bathyal); Late Pliocene to Recent.

Pecten nux Reeve, 1853; Polynesia, Melanesia, Micronesia, eastern and northern Australia, Indonesia, Philippines, Taiwan, Japan, Andaman, Oman, Seychelles, Mauritius, Kenya, Mozambique, Madagascar, eastern South Africa and Red Sea (mainly upper sublittoral); Early Miocene to Recent.

Pecten vesiculosus Dunker, 1877; Japan (mainly lower sublittoral); Middle Pliocene to Recent.

Pecten phrygium Dall, 1886; Gulf of Mexico, east coast of the United States and north of Guiana (lower sublittoral to upper bathyal); Recent.

Pecten (Aequipecten?) yanagawaensis Nomura and Zinbo, 1936; Japan; Middle Miocene.

Cryptopecten spinosus sp. nov.; Japan; Late Pleistocene.

The shell convexity is weak in C. bullatus and C. phrygium, weak to moderate in C. spinosus and C. yanagawaensis, moderate to strong in C. vesiculosus, and usually very strong in C. nux. The number of radial ribs in the examined samples is 18-23 in C. bullatus, 18-22 in C. nux, 13-18 in C. vesiculosus, 17-19 in C. phrygium, 20-25 in C. yanagawaensis and 12-15 in C. spinosus. The imbricated scales are regularly alternate in C. bullatus, C. nux and C. phrygium, but always oppositely disposed in other species. The test is very thin in C. bullatus owing to the poorly developed inner layer, but moderate or rather thick in other species.

The following subspecies are morphologically and biogeographlcally distinguishable from the nominate subspecies.

Cryptopecten nux sematensis (Oyama, 1954); Japan; Late Pleistocene.

Cryptopecten vesiculosus makiyamai subsp. nov.; Japan; Late Pliocene.

C. nux sematensis is characterized by its unusually tumid outline, and C. vesiculosus makiyamai by its large size and exceptional occurrence in fine-grained sediments.

The specimens hitherto assigned to "Cryptopecten tissotii" in various Japanese reports are almost entirely referable to C. bullatus. Cryptopecten alli Dall, Bartsch and Rehder, is evidently a junior subjective synonym of P. bullatus. Pecten hastingsii Melvill, 1888, Pecten guendolenae Melvill, 1888, Chlamys smithi Sowerby, 1908, Chlamys corymbiatus Hedley, 1909, and Pecten (Aequipecten) kikaiensis Nomura and Zinbo, 1934, are certainly synonymous with P. nux. Pecten hysginodes Melvill, 1888, may be synonymous with P. vesiculosus. Pecten inaequivalvis Sowerby, 1887, and Pecten oweni Gregorio, 1936, which were included in Cryptopecten by some authors, seem to be referable to Haumea and Comptopallium (or Decatopecten), respectively.

2. Reproduction and Growth Rings

As the result of my observation on living populations in the eastern part of Sagami Bay carried out in several different seasons, it was found that all the large-sized individuals (over 13 mm in length and height) of C. vesiculosus, regardless of the phenotype, have well-developed hermaphroditic reproductive glands in summer. The curvature of shell surface, both external and internal, becomes unusually strong in this season, and a stepwise growth ring is formed. It is, therefore, reasonable to consider that the increased internal space produced by the strong curvature corresponds well to the development of gonad. Such conspicuous growth rings occur also in several other fossil and living pectinids, and, I presume, the intervals between them indicate approximate annual growth amount after the individual reached sexual maturity.

3. Continuous Variation and Relative Growth

The intrapopulational variation of various morphometric (both univariate and bivariate) and meristic characters were examined on some selected large samples of C. vesiculosus, C. bullatus and C. nux. Because the shell size and form ratios between various lineal-measurements are growth-variant and strongly influenced by the original age composition and post-mortem sorting, some standardization is necessary. In this article the partial shell height (H1) from the origin of growth to the termination of the first growth ring, which probably signifies size at sexual maturity, was applied for the standardized shell size. The average and variability of shell shape were recognized mainly through bivariate analyses of relative growth. In almost all the samples, significant positive allometry was ascertained in the relation of convexity (T) to length (L) (only in right valve), and, on the contrary, significant negative allometry in the relation of height (H) to length (L), convexity (T) to length (L) (only in left valve) and length of two wings (D) to overall length (L). In other words, the shell outline of Cryptopecten generally changes from left-convex to right-convex with proportionally lowered disk and shortened wings through growth. Standardized form ratios, H/L, T/L and D/L, are obtained in each sample from the reduced major axes at several definite sizes.

The number of radial ribs, which is empirically almost growth-invariant except for very early growth stages, always shows a histogram resembling normal frequency distribution, giving a reliable basis for comparison between samples and between taxa.

4. Sculpture Dimorphism

All the Recent and post-Middle Pleistocene population samples of C. vesiculosus show significant dimorphism in surface sculpture; one phenotype is called Q and is characterized by the highly elevated and quadrate radial ribs and erect scales on the interspaces, while the other phenotype is called R and is characterized by the flattened and rounded radial ribs and imbricated scales covering not only the sides of ribs but also the interspaces. No individual is actually intermediate, and the relative frequency of each phenotype is clearly recognizable in all the fossil and Recent samples. No significant difference can be detected between the two phenotypes in such morphological characters as shell size, growth rate, number of ribs, allometric indices, coloration and anatomical features.

As the result of my survey on Recent samples, it was found that the two phenotypes occur in association at every locality, seeming to constitute the same interbreeding population. Such a perfect overlap of distribution with relatively stable frequency would be quite unlikely in the case of sibling species. Ecophenotypic effect and sexual dimorphism are also improbable. Though the genetic background is still unknown, various circumstantial evidences suggest that this dimorphism is controlled by a single or only a few genetic factors such as an allele or chromosomal aberration.

Somewhat similar dimorphism is found in Aequipecten commutatus from the Mediterranean and Volachlamys hirasei from Japan. The individual and geographic variations of these species should be further studied at the population level.

5. Geographic Variation

In 1973 I concluded that the geographic variation of C. vesiculosus in present-day seas is insignificant, so far as the phenotypic frequency is concerned. This conclusion was, however, based only on several Recent samples from the Pacific coast of central Honshu which do not cover the whole distribution of this species. As the result of my extensive survey on samples not only from the Pacific coast of Honshu but also from the East China Sea and the Japan Sea, some positive evidence of geographic variation was detected. The frequency of Phenotype R ranges from 0.39 to 0.46 in the large samples from the Pacific coast of central Honshu, whereas in the samples from the East China Sea and the Japan Sea it is much lower, ranging from 0.09 to 0.30. A similar tendency is also ascertained in the average number of radial ribs, which is as small as 14.5-14.8 in the samples from the Pacific coast of Honshu but commonly exceeds 15.0 in the samples from the East China Sea and the Japan Sea. The geographic variation of these characters seems to indicate the presence of some gentle clines along the Pacific coast of the Japanese Islands. Such clinal changes are not clearly detected on the samples along the Japan Sea side, though this may be due to the small sample sizes. The geographic variation of this species in the geologic past is still difficult to analyze, because the available fossil samples are strongly biassed in time and space.

The growth rate of shell and various bivariate characters are also often significantly different among Recent samples, but no conspicuous dine is recognized. This may be partly due to the deficiency of large samples, but, I presume, these characters may be more strongly influenced by local environmental factors than the phenotypic frequency and number of ribs.

6. Phyletic Evolution

The phyletic evolution of C. vesiculosus is well characterized by clear phenotypic substitution. All the Pliocene and Early Pleistocene samples are monomorphic, consisting only of the individuals of Phenotype Q. Phenotype R, so far as I am aware, first appeared in Middle Pleistocene (about 0.5 Ma) on both sides of the Japanese Islands, and the frequency seems to have increased with time. Because the amount of change evidently surpasses the range of geographic variation in the present-day seas, the phenotypic increase is mainly attributable to phyletic evolution.

The Late Pleistocene Jizôdô Formation of Boso Peninsula near Tokyo yields a number of rich fossil samples of this species. In this peninsula and its adjacent sea area the phenotypic frequency has increased from 0.15 to 0.43 during these 0.37 million years. This change could be theoretically explained by assuming a very slight difference of adaptive value (only of the order of 10-5 on average) between the two phenotypes. The causal evaluation of this change is at present no more than intuitive, because such non-adaptive causes as recurrent mutation and random genetic drift are also probable. It may be said, however, that this phenotypic change can be explained in any case by the acknowledged mechanism of microevolution.

The average number of radial ribs has also changed in excess of the range of geographic variation in the present seas. In most of the Late Pliocene samples the average exceeds 16.0, while it ranges from 14.5 to 15.3 in Recent samples. Several large Pleistocene samples from Boso Peninsula seem to indicate that this character has shifted directionally with time.

So far as the phenotypic frequency and the number of radial ribs are concerned, I would venture to say that the living populations off the Pacific coast of central Honshu are more advanced than those in the East China Sea and Japan Sea. The size of maximum individual and annual growth rate may be influenced by local conditions, but they are generally larger in Pliocene and Pleistocene samples than in Recent ones.

7. Speciation and Extinction

The three living Indo-Pacific species of Cryptopecten, i. e., C. bullatius, C. nux and C, vesiculosus, and probably also C. phrygium in the western Atlantic seem to have evolved along independent lineages at least since Late Pliocene. Although phyletic change may have occurred also in lineages other than C. vesiculosus, the clear morphological gaps among these lineages appear to indicate a punctuated pattern rather than gradual diversification (see Fig. 24).

Judging from the fundamental structure of radial ribs and geographic distribution, Cryptopecten may have already been differentiated into two species groups in Middle Miocene. One group, comprising C. bullatus, C. nux and C. phrygium, is characterized by the alternate arrangement of imbricated scales and extensive geographic distribution. C. nux is a long-ranging living species and is regarded as constituting the parental stock of this species group, because fossil records from Early Miocene onward and from Middle Pliocene onward have been known in east Africa and the northwest Pacific region, respectively. C. bullatus was probably derived from C. nux, adapting now to a somewhat deeper environment in the central and northwestern Pacific. Though there is no fossil record, C. phrygium may have arisen by geographic isolation from a peripheral population of C. bullatus.

The other species group, comprising C. vesiculosus, C. yanagawaensis and C. spinosus, is characterized by the opposite disposition of imbricated scales and restricted geographic distribution to Japan and its adjacent seas. C. yanagawaensis occurs only from the Middle Miocene of Japan and may be ancestral to the post-Early Pliocene species, C. vesiculosus, but the wide gap of fossil records makes the interspecific relation obscure. C. spinosus seems to have suddenly appeared in the Late Pleistocene coral sand facies in south Japan and to have then become extinct. It shares many common characteristics with C. vesiculosus and probably arose from the stock of this species through some geographic speciation, though the process is as yet poorly documented.

C. nux sematensis seems to represent a Late Pleistocene local population with specialized morphology near the northern periphery of distribution of C. nux. This assumption is consistent with the fact that some observed samples of C. nux nux in the peripheral areas of the present-day distribution (e. g., Queensland and the Red Sea) also show more or less specialized morphology. C. vesiculosus makiyamai from the Upper Pliocene of Japan, which exceptionally occurs in fine-grained sediments, may represent an incipient branch from the stock of C. vesiculosus, though the possibility of ecophenotypic effect is not necessarily denied.

The restored phytogeny of Cryptopecten is thus characterized by the presence of a few short-ranging species and subspecies in addition to several persistent lineages. These short-ranging taxa, I believe, represent only a fraction of the speciation and peripheral isolation that actually occurred. Speciation and geographic isolation of a population (and its absorption into the parental population) are probably ubiquitous events, and the records of a large number of other unsuccessful dead end taxa have probably simply not been discovered or have been lost.

Also worthy of notice is the fact that the geographic distribution of each species of Cryptopecten seems to be conservative through geologic times. For example, almost all the known fossils of the three living Indo-Pacific species have been found in areas adjacent to the present-day distribution, even if the northern (and southern) limit may have fluctuated in accordance with climatic changes.

8. Significance of Phenotypic Substitution in Macroevolution

In conclusion I will summarize the significance of phenotypic substitution as one of the major features of evolution. Phenotypic substitution, here defined, is the chronological change of relative phenotypic frequency in a polymorphic species. Among neontologists this is generally recognized as transient polymorphism, typically exemplified by the industrial melanism of some moths in the 19th Century (Kettlewell, 1961; Ford, 1964). Komai's (1956) study of transient polymorphism in a ladybeetle, Harmonia axyridis, over a period of 40 years suggested that some selection pressure caused by long-term climatic change is responsible for the observed definite trend. These are, of course, excellent examples of microevolution, but probably the longest observations on actual living populations.

The phyletic evolution of C. vesiculosus seems to involve phenotypic substitution over the last 500,000 years. Though the causal reasoning of this change, whether it is adaptive or non-adaptive, may be difficult to establish, this phenomenon suggests that studies of transient polymorphism could be expanded to the geological past. Polymorphic relation between two or more nominal species and sometimes phenotypic substitution have also been postulated in some fossil animals; Pleistocene-Recent bears, Ursus, by Kurten (1955), Pliocene sand dollars, Merriamaster, by Durham (1978), and Cretaceous ammonites, Gaudryceras, by Hirano (1978), for example. Although I am not necessarily in a position to confirm the adequacy of their interpretations, with careful studies of the relation between phena and taxa, phenotypic substitution could be recognized as ubiquitous phenomena in various taxonomic groups.

Phenotypic substitution seems to be one of the most intrinsic causes for the punctuated patterns of morphological change in the fossil records. Problems about its role in macroevolution may be problems of scale-how significant morphological change actually occurs by a single mutation. The morphological discontinuity between the two phenotypes of C. vesiculosus appears to be slight in one aspect but considerable in another. The two phenotypes are not significantly different in numerous unit characters, but the surface sculpture of the mutant (Phenotype R) is quite new and unique in the genus Cryptopecten and probably also in the family Pectinidae. Hollow structure of radial ribs is regarded as an important diagnostic character of Cryptopecten, but, in one and the same population of Aequipecten commutatus, hollow structure may or may not be present. As has been discussed by many evolutionists (e. g., Mayr, 1963), instantaneous speciation may be almost impossible in many animal groups in which polyploidy is rare, but, I believe, the ultimate origin of a new taxonomic character, even for a taxon higher than species, should sometimes be sought in a single gene (especially "regulatory gene") and chromosome mutation, and subsequent phenotypic substitution.

9. Further Problems

The final object of this study is to understand the evolutionary pattern and process of a particular taxonomic group, a pectinid genus in this case, from various paleontological and neontological viewpoints. As the first step, the morphological change in a part of the constituent species has been clarified to some extent at the population level, but many problems remain unsolved. With respect to C. vesiculosus, intrapopulational and geographic variations as well as chronological shift of shell morphology were detected for several characters; as the next step, developmental, cytological and genetical examinations are needed, particularly in relation to the background of the observed dimorphism. For other species of Cryptopecten, present knowledge is still far from what is required for completion of the first step, and effort should be devoted to obtaining basic information on distribution, ecology and morphology. More extensive surveys on living and fossil populations of these species are of course indispensable to the task of reconstructing their evolutionary process and interspecific relationships. The present study has brought home to me once again the profundity of the study of evolution and convinced me anew that the best understanding of evolutionary process will be accomplished by a synthesis of the results obtained through various approaches, each with their advantages and limitations.