Retrodisplacement of the oral and anal openings in dendrasterid sand dollars.
Received August 19, 1994. Accepted September 6, 1994.
Echinoid tests are ideal subjects for morphogenetic studies, for two reasons that may seem contradictory. On one hand, the basic organization of the test is similar in nearly all post-Paleozoic forms, and thus it is relatively easy to establish the homology of skeletal elements. But at the same time, the gross morphology of the test is extraordinarily variable. For example, an echinoid test may display near-perfect pentameral symmetry, or it may have well-defined bilateral symmetry. Its overall shape may be spheroidal, cylindroidal, or discoidal. The mouth and anal openings may be central, or they may lie at the edges of the test. Because all modern tests share the same underlying structure, comparative studies can identify the developmental changes responsible for the striking differences in morphology.
An echinoid test is composed of sutured calcite plates, which may number in the thousands. The coronal plates, which form the bulk of the test, are arranged in 20 radial columns. Each column begins at the apical system, a group of small specialized plates on the aboral (upper) surface. The columns continue downward to the peristome, or mouth opening, on the underlying oral surface. The 20 columns are grouped into ten distinct pairs. Five of these pairs form the ambulacra, and they alternate with the five remaining pairs, which form the interambulacra. Each ambulacrum and interambulacrum has an "adapical" end (at the apical system) and an "adoral" end (at the peristome).
Regular echinoids exemplify the primitive condition for the class. In regular echinoids, each ambulacrum (or interambulacrum) closely resembles the other four, and thus the test displays pentamerous radial symmetry. The test outline is circular or pentagonal, with the apical system and the peristome located at the centers of their respective surfaces. The periproct, or anal opening, is also located centrally, within the plates of the apical system.
The morphology of irregular echinoids is more complex. In irregular echinoids, the periproct migrates from the apical system into the posterior interambulacrum. The periproct defines an anterior-posterior (or "longitudinal") axis, and the test displays bilateral symmetry about this axis. The ambulacra (or interambulacra) may differ in morphology, and thus the outline of the test can be quite variable. The apical system and the peristome are commonly displaced toward the anterior or posterior ends of the test.
Many irregular echinoids are anatomically polarized: the peristome lies near the anterior end of the test, whereas the periproct lies at the posterior end (fig. 1A). This morphology, which is particularly well developed in spatangoids and holasteroids, has often been linked to the development of unidirectional locomotion (Durham 1966; Kier 1974; Smith 1984). From a functional standpoint, it is probably advantageous to place the oral and anal openings at opposite ends of the body. In fact, this arrangement is so characteristic of bilateral metazoans that it is commonly taken for granted.
However, other irregular echinoids are not polarized. In many holectypoids and clypeasteroids, the peristome and the periproct are both found on the oral surface of the test: the peristome typically lies near the center, whereas the periproct lies between the peristome and the posterior margin (fig. 1B). The oral and anal openings of such unpolarized tests may occur in remarkably close proximity. In some mellitid sand dollars, for example, these structures are separated only by a single small interambulacral plate.
From a functional standpoint, the unpolarized morphology is rather problematical. It is not surprising to find the peristome at the center of the oral surface, because this is the primitive condition for the class. But why should the periproct be located close by? If it is advantageous to separate these structures, then the periproct should lie on the aboral surface, or at the posterior margin of the test. In some holectypoids and clypeasteroids, the periproct does occupy one of these positions. But in other cases, the periproct has shifted in the "wrong" direction: it has moved anteriorly, away from the posterior margin and toward the peristome, thereby reducing the separation between the oral and anal openings.
This phenomenon may be termed "retrodisplacement," because the direction of periproct movement is opposite to the expected direction of periproct movement in irregular echinoids. The term "retrodisplacement" has been borrowed from medical terminology, where it refers to "displacement backwards of an organ of the body" (Gove 1976, p. 1940).
The dendrasterid sand dollars provide one of the most intriguing examples of retrodisplacement (fig. 1C). The dendrasterids are a small clypeasteroid group, of Late Miocene to Recent age, found along the Pacific coast of North America. During the course of dendrasterid evolution, the periproct has shifted anteriorly in the manner described above, from the posterior margin onto the oral surface. Furthermore, the dendrasterid peristome has undergone retrodisplacement as well: it has moved posteriorly, away from the anterior margin and toward the periproct. In dendrasterids, the mouth opening is located on the posterior half of the test, which is an extremely rare condition among echinoids.
A new model for oral surface development in the dendrasterid sand dollars is presented here. An earlier study, Beadle (1989), focused on the development of the aboral surface, and on the position of the apical system. Here, the same approach is applied to the oral surface. Together, these two papers provide a comprehensive model for the evolution of the characteristic features of the dendrasterid test, including the peristome, periproct, apical system, ambulacra, and interambulacra.
The retrodisplacement of the peristome and periproct are two of the effects predicted by the developmental model. The model proposes that the positions of the peristome and periproct (on the oral surface) and the position of the apical system (on the aboral surface) are affected by the same developmental processes. If this interpretation is correct, then selective pressure for apical system displacement could cause the peristome and periproct to shift as well. In other words, the retrodisplacement of these two structures could be related to developmental constraints.
The dendrasterid sand dollars, and the materials and methods used to measure their morphological characteristics, are reviewed here. Two different patterns of oral development, the "Echinarachnius pattern," which occurs in other northern Pacific sand dollars, and the "Dendraster pattern," which occurs in dendrasterids, are then compared. The model presented here is based upon two hypotheses: first, that a heterochronic change in the Echinarachnius pattern of development could simultaneously affect many aspects of test morphology; and second, that the predicted effects are characteristic features of the Dendraster pattern.
THE DENDRASTERID SAND DOLLARS
The sand dollar Dendraster is a familiar member of temperate-water faunas along the Pacific coast of North America, from southern Alaska to Baja California. Three living species are commonly recognized: D. excentricus, D. laevis, and D. vizcainoensis (Durham et al. 1980). Dendrasterids also have an extensive record in the provincial Neogene (Kew 1920; Grant and Hertlein 1938; Wagner 1970; Durham 1978). The three extant Dendraster species are all known as fossils, and some 20 additional species are known exclusively as fossils. Eight more fossil species have been assigned to the extinct genus Merriamaster, which is often associated with Dendraster in Pliocene deposits.
Modern taxonomists generally recognize the Dendrasteridae as a distinct family. However, the composition of this family remains unsettled. The only genera that are universally accepted as dendrasterids are Dendraster and Merriamaster. Here, as in Beadle (1991), the term "dendrasterid" refers exclusively to these two forms.
Dendrasterid tests are characterized by their peculiar aboral morphology. In most sand dollars, the apical system is located at the approximate center of the aboral surface. In dendrasterids, however, the apical system is displaced toward the posterior end of the test (fig. 2A). This morphology is produced by an asymmetrical growth pattern that favors the development of the anterior areas (Beadle 1989). This growth pattern also affects the five petals, which hold specialized respiratory tubefeet. In mature dendrasterids, the unequal development of the petals is readily apparent (fig. 2A).
The unusual morphology of the aboral surface is linked to an equally unusual behavior. Most clypeasteroids are deposit feeders, and thus they lie horizontally on the substrate. Living Dendraster can feed in this manner, but they can also assume an inclined, vertical position, with only the anterior third of the test in contact with the substrate. The remainder of the test projects upward into the water column, such that suspended particles can be collected there (Timko 1976). The asymmetry of the aboral surface keeps the apical system and the petals above the substrate when the test is in the inclined position.
In the first description of a dendrasterid species, Eschscholtz (1831, p. 19) reported that both the apical system and the peristome were displaced toward the posterior end of the test, and he named the new species Scutella excentrica (now D. excentricus). The term "eccentric" has been widely applied to dendrasterids ever since. Whereas eccentricity usually refers to the apical system, it has also been used to describe the peristome (e.g., Woodring et al. 1940, p. 82; Raup 1956, p. 693). Here apical eccentricity refers to the displacement of the apical system on the aboral surface, whereas peristomial eccentricity refers to the displacement of the peristome on the oral surface. Both forms of eccentricity can be defined quantitatively, as described below.
Apical Eccentricity.--The most conspicuous feature of a dendrasterid test is the eccentricity of the apical system. Many workers have published measurements of apical eccentricity, using a variety of definitions. However, the most detailed studies defined apical eccentricity by the following ratio: the distance from the apical system to the posterior margin, divided by one-half of the test length (fig. 2A). This definition was proposed by Raup (1956, pp. 685-686) because "eccentricity is usually expressed in terms of the position of the center of the apical system relative to the midpoint of the antero-posterior diameter of the test." Raup's definition was later used by Stanton et al. (1979), and it was also adopted here.
The numerical value for apical eccentricity may be designated as [e.sub.a] Possible values of [e.sub.a] range from 0 to 2. When the apical system is centered, then [e.sub.a] = 1. If the apical system is displaced posteriorly, then [e.sub.a] < 1. If the apical system is displaced anteriorly, then [e.sub.a] > 1.
Raup (1956, p. 686) defined an arbitrary measuring point for the apical system: "the midpoint of an imaginary line connecting the two posterior genital pores." This point is located near the center of the apical system (in fact, it is slightly posterior to the center). The genital pores are located at the adapical ends of the posterior paired interambulacra. These pores are absent in small juvenile tests, but the adapical ends of the interambulacra can still be identified, because they form narrow gaps between the conspicuous ambulacral petals.
Peristomial Eccentricity.--The displacement of the peristome in dendrasterids is not as obvious as the conspicuous displacement of the apical system. Only a few measurements of peristomial eccentricity have been published (Woodring et al. 1940, p. 82; Durham 1949, p. 51). Nonetheless, the displacement of the peristome can be well developed, particularly among fossil species (fig. 3).
Previous definitions of peristomial eccentricity are unsatisfactory; the definition used by Durham (1949) does not include an exact measuring point for the peristome, whereas the definition used by Woodring et al. (1940) is only valid when comparing tests of similar size. The latter definition includes a measurement from the anterior end of the test to the posterior end of the peristome. This measurement is affected by the relative size of the peristome, which is in turn dependent on the overall size of the test.
Here peristomial eccentricity was measured by the following ratio: the distance from the posterior margin of the test to the posterior edge of the peristome, divided by the distance from the anterior margin of the test to the anterior edge of the peristome (fig. 2B). This definition is independent of peristome size or test size.
The numerical value for peristomial eccentricity may be designated as [e.sub.p]. Possible values of [e.sub.p] range from zero to infinity. Measured [e.sub.p] values can be interpreted in the same manner as [e.sub.a] values. Thus, if [e.sub.p] = 1, the peristome is centered at the midpoint of the longitudinal axis. If [e.sub.p] < 1, the peristome is displaced posteriorly, and if [e.sub.p] > 1, the peristome is displaced anteriorly.
It should be noted that [e.sub.a] and [e.sub.p] are defined differently (fig. 2) and have different ranges. An [e.sub.a] value of 0.75 is not "equivalent" to an [e.sub.p] value of 0.75, although both values do signify posterior displacement. In theory, [e.sub.p] could be measured using the adoral ends of the posterior paired interambulacra, in the same general manner as [e.sub.a]. This procedure would be impractical, however, because the oral surface has no conspicuous features (such as petals or genital pores) to delineate the ends of the interambulacra. Special preparation is often required to identify the boundaries of the interambulacral plates.
MATERIALS AND METHODS
Eccentricity measurements were obtained from three well-known dendrasterid species. The first species, Dendraster gibbsii, is a Mio-Pliocene form with highly developed eccentricity. The second species, D. laevis, is a Recent form with poorly developed eccentricity. The final species, D. excentricus, is the most abundant Recent dendrasterid; it is known for its morphological variability. Previous quantitative studies suggest that these three species encompass most of the total variation in dendrasterid apical eccentricity (Kew 1920; Woodring et al. 1940; Raup 1956; Wagner 1970). For comparison, another northern Pacific sand dollar, Echhnarachnius parma, was measured as well. This Recent species is not a dendrasterid, and it is conventionally regarded as noneccentric.
The specimens were examined at the Los Angeles County Museum of Natural History, in the Sections of Invertebrate Zoology and Invertebrate Paleontology. The measured tests were not chosen randomly; they were selected to maximize intraspecific variation. Particular efforts were made to find specimens with unusually high or low eccentricities, although tests with obvious injuries or malformations were not included. Apical and peristomial eccentricities were measured using Vernier calipers, following the definitions outlined above (fig. 2). The measurements were rounded to the nearest 0.1 mm.
The fossil D. gibbsii tests were collected at a single upper Miocene locality, LACMIP 22696, in the lower Etchegoin (or Jacalitos) Formation near Coalinga, California. It is reasonable to assume that these specimens represent a single species. The Recent D. laevis, D. excentricus, and E. parma tests came from a wide range of localities. The samples included 40 D. gibbsii, 28 D. laevis, 76 D. excentricus, and 36 E. parma specimens.
This study also included measurements of ambulacral plate size in D. excentricus. Thirty tests were obtained from the University of California (Berkeley) Department of Paleontology. The tests came from lot A-4943, from Crescent Beach, near Eastsound, on Orcas Island, Washington. The outlines of the plate boundaries were copied from the tests onto transparent plastic sheets, and the plate areas were then calculated with a digitizer. The same methods were used previously (Beadle 1989) to measure plate size in E. parma.
THE ECHINARACHNIUS PATTERN OF DEVELOPMENT
The living sand dollars of the northern Pacific are commonly assigned to two families: the Dendrasteridae and the Echinarachniidae. The exact composition of these families remains a matter of debate. However, it is clear that the type genera, Dendraster and Echinarachnius, are characterized by very different patterns of oral and aboral growth.
The best-known representatives of these genera are the type species, D. excentricus and E. parma. Here the growth pattern displayed by the former species will be termed the Dendraster pattern, whereas the growth pattern displayed by the latter species will be termed the Echinarachnius pattern. These patterns also occur in other northern Pacific sand dollars, but D. excentricus and E. parma are the best-documented examples.
In adult E. parma tests, the anterior and posterior areas are about equally well developed, and the apical system and the peristome are approximately central. However, the apparent regularity of the adult test is misleading. The test actually undergoes two opposing phases of strongly asymmetric growth during the course of its development.
First Phase of Unequal Growth.--The oral plates of E. parma are not distributed symmetrically: the posterior areas consistently have one or two fewer plates than the anterior areas (Beadle 1989, figs. 3-5). This phenomenon may be termed the "oral plate disparity." The oral plate disparity was first documented in juvenile E. parma tests by Gordon (1929, table 1) and later in adult tests by Durham (1955, table 2).
The oral plate disparity is caused by unequal growth early in ontogeny, before metamorphosis. At this time, the developing test of E. parma is markedly asymmetrical: the anterior areas have more plates, whereas the posterior areas have fewer plates (Gordon 1929). The accelerated development of the anterior areas early in ontogeny has been termed "the first phase of unequal growth" (Beadle 1989, p. 210).
The test becomes flattened soon after metamorphosis, which divides the test into two separate surfaces. Thereafter, no plates are added to the oral surface; new plates can only be generated aborally, and the aboral plates do not migrate around the margin to the oral surface. The oral surface retains its original complement of plates throughout life, and thus the asymmetrical distribution of these plates is permanent.
Second Phase of Unequal Growth.--Whereas juvenile E. parma tests are markedly asymmetrical, the adult tests are not. Gordon (1929, p. 319) found that "the corona with large anterior and small posterior areas is characteristic only of the imago and of very small urchins. During development this condition is gradually altered such that in the adult all the ambulacra and all the interambulacra are almost equally well developed."
This change in morphology is caused by another phase of asymmetrical growth. Soon after metamorphosis, the premetamorphic growth pattern is reversed; the posterior areas develop at a faster rate than the anterior areas. The accelerated development of the posterior areas late in ontogeny has been termed "the second phase of unequal growth" (Beadle ] 989, p. 212).
The first phase of unequal growth (which favors anterior development) and the second phase (which favors posterior development) are normally balanced, and at maturity, the test appears to be symmetrical. This unusual mode of growth affects many of the characters on the oral surface, as outlined below.
Position of the Peristome.--In E. parma, the peristome is typically very close to the center of the oral surface, such that 0.95 < [e.sub.p] < 1.05 (fig. 4). Eccentricities within this range are so slight that they are unlikely to be noticed by a casual observer. Some E. parma tests do have greater or lesser eccentricities, as shown in figure 4, but such specimens are rare. It should be noted that the figured sample was selected nonrandomly, to maximize variation. This sample shows more variability than would be expected in any natural population.
The position of the peristome does not appear to be closely linked to the position of the apical system. There was little or no correlation between [e.sub.a] and [e.sub.p] in the 36 E. parma specimens ([r.sup.2] = 0.276).
Morphology of the Ambulacra.--The peristome of E. parma is typically centered along the longitudinal axis, which suggests that the size of the posterior oral areas is approximately equal to that of the anterior oral areas. However, the posterior areas normally contain one or two fewer plates, resulting from the oral plate disparity described above. It follows that the plates in the posterior areas should be somewhat larger, on the average, than the plates in the anterior areas.
Variation in oral plate size can be studied by measuring plate areas with a digitizer. In E. parma, the second plates in each column (counting from the peristome) are usually the largest (fig. 5A). These plates are known as the "first post-basicoronal plates" (because the initial plates in each column, adjacent to the peristome, are known as the "basicoronal plates"). The first post-basicoronal plates are well suited to quantitative analysis, because they are relatively large and easy to identify. Each of the five ambulacra contains a pair of these plates (fig. 5A). It is simplest to measure the combined area of both plates.
Beadle (1989) measured the relative size of l SO ambulacral plate pairs in 30 Recent E. parma tests. The "relative size" of a plate pair was defined as the area of the plate pair, divided by the area of the plate pair in ambulacrum III of the same specimen. The measured relative sizes varied widely (fig. 6A). However, the plates in the anterior ambulacrum (area III) were consistently the smallest (fig. 6B). The plates in the anterior paired ambulacra (areas II and IV) averaged 25.5% larger, whereas the plates in the posterior paired ambulacra (areas I and V) averaged 46.5% larger (fig. 6B). These changes are statistically significant (t-test, 99.9% level). Thus, the size of the oral ambulacral plates of E. parma varies considerably, and the variation is strongly correlated with the longitudinal position of the ambulacra. The plates in the posterior areas are significantly larger than the plates in the anterior areas, as predicted above.
In summary, the oral ambulacra of E. parma vary in morphology, and this variation is linked to their position along the longitudinal axis (fig. 5A). The anterior ambulacra have more plates. The greater number of plates in the anterior areas is a consequence of the first phase of unequal growth. However, the posterior ambulacra have larger plates. The greater relative size of the plates in the posterior areas is due to the second phase of unequal growth.
Morphology of the Interambulacra.--In juvenile sand dollars, each oral interambulacrum is "continuous"; the interambulacral plates form an unbroken series from the peristome to the margin. Over time, however, the interambulacral plates may be displaced by the adjacent ambulacral plates. The basicoronal plate, which adjoins the peristome, may lose contact with the other plates in the series, such that the interambulacrum becomes "discontinuous." Many (although not all) sand dollars have discontinuous interambulacra as adults.
Mature E. parma tests usually have at least one discontinuous interambulacrum (fig. 5C). Durham (1955, table 3) discovered that the posterior interambulacra of E. parma were the most likely to become discontinuous. Lohavanijaya and Swan (1965) later found that the discontinuous condition develops in a predictable sequence across the oral surface. The posterior interambulacrum (area 5) is almost always the first area to become discontinuous. The posterior paired interambulacra (areas 1 and 4) are usually next, and the anterior paired interambulacra (areas 2 and 3) are usually last (Beadle 1989, fig. 5).
Thus, the longitudinal axis affects the interambulacra of E. parma, as well as the ambulacra. The interambulacra become discontinuous in a gradual sequence, from posterior to anterior (fig. 5C). This pattern reflects the accelerated development of the posterior oral areas. The sequential appearance of discontinuous interambulacra is one of the most conspicuous effects of the second phase of unequal growth.
Periproct Position.--In irregular echinoids, the periproct lies between the two plate columns of the posterior interambulacrum (area 5). In E. parma, the periproct is typically located in a "marginal" position, on the posterior edge of the test. The periproct remains marginal throughout life. It shows little or no tendency to migrate onto the oral surface.
Development of Scaphechinus.--The distinctive oral plate characters that occur in E. parma can also be recognized in other northern Pacific sand dollars. It is reasonable to assume that these forms share the Echinarachnius pattern of development. The living Scaphechinus mirabilis (which is the type species of the genus Scaphechinus) is one such example.
First, S. mirabilis appears to be affected by the oral plate disparity: the anterior oral areas tend to have one or two more plates than their posterior counterparts (Nisiyama 1968, p. 118). This disparity is particularly evident among the interambulacra.
Second, the relative size of the oral ambulacral plates varies: the posterior areas have larger plates than the anterior areas. The first post-basicoronal plates were described by Nisiyama (1968, p. 117); he measured the length of the suture between adjacent plates in different areas. The suture between the plates in ambulacra I and V (the I-V suture) is typically the longest. The I-II suture and the IV-V suture are only about half as long as the I-V suture. The II-III suture and the III-IV suture are only about one-third as long as the IV suture. These findings imply that the plates in the posterior paired ambulacra (areas I and V) are the largest, that the plates in the anterior paired ambulacra (areas II and IV) are smaller, and that the plates in the anterior ambulacrum (area III) are the smallest.
Third, the interambulacra become discontinuous sequentially, from posterior to anterior. At 5 mm, all five areas are continuous. The posterior interambulacrum (area 5) is discontinuous at a size of 9 mm, the posterior paired interambulacra (areas 1 and 4) are discontinuous at 18 mm, and the anterior paired interambulacra (areas 2 and 3) are discontinuous at 32 mm (Nisiyama 1968, fig. 49).
In Recent S. mirabilis, all five areas normally become discontinuous. However, Hayasaka and Morishita (1947) described a fossil population in which many adults retained one or more continuous interambulacra. The continuous condition was never observed in area 5, occurred rarely in areas 1 and 4 (7 of 42 cases) and was most common in areas 2 and 3 (35 of 42 cases). These findings demonstrate that the posteriorto-anterior gradient in interambulacral development can even be detected in fossil tests.
The fossil sand dollar fauna of the northern Pacific is diverse, but incompletely known. The available evidence suggests that the distinctive oral characters found in living forms, such as Scaphechinus and Echinarachnius, can also be recognized in extinct Mio-Pliocene forms, such as Astrodapsis, Remondella, Vaquerosella, and Kewia. Some of these fossil forms occur in close geographic and stratigraphic proximity to the earliest dendrasterids (Beadle ]991). Several authors have suggested that the dendrasterids were derived from extinct Miocene echinarachniids (Kew 1920; Durham 1949, 1955; Wagner 1970).
THE DENDRASTER PATTERN OF DEVELOPMENT
In dendrasterids, the aboral surface is typically eccentric (fig. 2A): the anterior areas are much larger than the posterior areas. It is clear that anterior development is favored during dendrasterid ontogeny. This asymmetrical growth pattern is not unique, however, because it also occurs in juvenile E. parma. The development of the anterior areas is favored early in the ontogeny of E. parma, during the first phase of unequal growth described above.
The Echinarachnius pattern of development also includes a second phase of unequal growth, which favors the posterior areas. The second phase counterbalances the effects of the first phase, and at maturity, the test appears to be regular. However, if a heterochronic change in development suppressed the second phase of unequal growth, the morphology of the test would change dramatically. The asymmetry found in juvenile tests would persist throughout life: the anterior areas would remain larger than the posterior areas. On the aboral surface, the petals would be unequally developed, and the apical system would be displaced toward the posterior end of the test. Beadle (1989) proposed that the aboral morphology of dendrasterids arose in this manner.
A similar approach can be applied to oral morphology. If the second phase of unequal growth were weakened or eliminated, the effects on oral development would be profound. It is possible to predict these effects, and to compare them with the actual oral characters that occur in adult dendrasterids.
Position of the Peristome.--In E. parma, the anterior oral areas have more plates than the posterior areas, because of the first phase of unequal growth. The peristome remains central in spite of this disparity, because the second phase of unequal growth increases the relative size of the plates in the posterior areas (fig. 5A). If the second phase were suppressed, this compensating factor would disappear. The anterior areas, with more plates, would be disproportionately large, and thus the peristome would be eccentric posteriorly.
Dendrasterid morphology is fully consistent with the predictions of the developmental model. First, dendrasterids are affected by the oral plate disparity. Durham (1949, 1955) studied the arrangement of oral plates in Dendraster and Merriamaster, and found that the posterior plate columns typically had one or two fewer plates than the corresponding anterior columns (fig. 5B). The disparity is particularly apparent in juvenile tests (Durham 1955, fig. 11). Second, the dendrasterid peristome is almost always eccentric posteriorly (fig. 4). In nearly all of the 144 Dendraster tests measured here, [e.sub.p] < 1, although diligent searching uncovered a few tests with 1.00 < [e.sub.p] < 1.05. The [e.sub.p] values were highest for D. Iaevis: typically 0.80 < [e.sub.p] < 1.00. The [e.sub.p] values were much lower for D. gibbsii, with 0.55 < [e.sub.p] < 0.75. The [e.sub.p] values for D. excentricus occupied an intermediate range: usually 0.70 < [e.sub.p] < 0.95.
In nearly all of the tests examined here, both [e.sub.p] < 1 and [e,sub.a] < 1 (fig. 4), although rare exceptions did exist. The [e,sub.a] values were highest for D. laevis: usually 0.80 < [e,sub.a] < 1.00. At the opposite extreme, in the D. gibbsii tests, 0.40 < [e,sub.a] < 0.60. The [e,sub.a] values for D. excentricus were typically intermediate, with 0.60 < [e,sub.a] < 0.80. These results are generally consistent with those obtained by Raup (1956) and Stanton et al. 1979). In one of Raup's D. excentricus samples (locality 7), the mean[e,sub.a] value was 0.930, but the specimens in this sample were reassigned to D. Iaevis by Wagner (1970, p. 170).
Raup (1956, p. 693) suggested that "it would be interesting to see if eccentricity of the peristome parallels that of the apical system." In fact, there does appear to be a linear relationship between these characters (fig. 4). The correlation between[e,sub.a] and [e.sub.p] in the 144 measured dendrasterid specimens was strong ([r.sup.2] = 0.815).
Morphology of the Ambulacra.--In E. parma, the anterior oral ambulacra have more plates, but the posterior oral ambulacra have larger plates (fig. 5A). The relative size of the ambulacral plates increases from anterior to posterior (fig. 6B). If the second phase of unequal grownth were suppressed, this distinctive pattern would never develop. Instead, all of the ambulacra would have plates of comparable size.
The ambulacral plates of a D. excentricus sample were measured, using the same methods employed for the E. parma sample. The plates of these species differed in two important respects. First, the D. excentricus plates were much less variable, in terms of relative size, than the E. parma plates (fig. 6A). The standard deviation for the entire D. excentricus sample was less than half the value for the entire E. parma sample (0.095 vs. 0.237, n = 150 for both samples). Second, in E. parma, relative plate size increased steadily from anterior to posterior, whereas in D. excentricus, no such pattern existed (fig. 6B).
In the D. excentricus sample, the anterior ambulacrum (area 111) usually had the smallest plates. The plates in the anterior paired ambulacra (areas II and IV) averaged 9.6% larger. However, the plates in the posterior paired ambulacra (areas I and V) were almost the same size as those in area III; they averaged just 2.0% larger (fig. 6B). The size differences between the plates in areas III, I, and V were not statistically significant.
Thus, the oral arrrbulacral plates of D. excentricus follow the general predictions of the developmental model. There is little variation in relative plate size, and the variation that does exist is not strongly correlated with the longitudinal position of the ambulacra.
Morphology of the Interambulacra.--In E. parma, the interambulacra become discontinuous in a gradual sequence, from posterior to anterior. Adult tests are often "partially discontinuous," with a mixture of continuous and discontinuous areas (fig. SC). If the second phase of unequal growth were suppressed, this pattern would never develop. The five interambulacra would become discontinuous simultaneously.
In Dendraster and Merriamaster, the discontinuous condition appears at about the same time in each interambulacrum (fig. 5D). No living or fossil dendrasterid is known to display the partially discontinuous condition. In living D. excentricus, the transition from the fully continuous condition to the fully discontinuous condition occurs rapidly, when the test length is 10-15 mm ([Durham 1955, fig. 11). The simultaneous development of discontinuous interarnbulacra in dendrasterids is consistent with the prediction above.
Position of the Periproct.--In both E. parma and D. excentricus, the periproct is normally adjacent to the third pair of post-basicoronal plates (Gordon 1929; Durham 1955). However, the periproct occupies different positions on the tests of these two species. In E. parma, the periproct is marginal: it lies on the posterior edge of the test (fig. 7A). But in D. excentricus, the periproct is "submarginal"; it lies entirely on the oral surface of the test, between the posterior edge and the peristome (fig. 7B).
The dendrasterid periproct undergoes retrodisplacement late in ontogeny. In small D. excentricus tests, less than 8 mm, the periproct is marginal. As the test enlarges, the periproct becomes "admarginal"; its posterior edge remains in contact with the margin, but its anterior edge extends onto the oral surface. The periproct gradually continues to shift anteriorly, and it becomes fully submarginal when the test measures about 15-20 mm (Durham 1955, fig. 11). Durham (1955, p. 102) examined "several thousand" Recent D. excentricus tests, and found that every adult specimen had a submarginal periproct.
Fossil dendrasterids are considerably more variable. One of the oldest known species, the Upper Miocene D. elsmerensis, has a true marginal periproct (Durham 1949). Another Upper Miocene species, D. sullivani, has an admarginal periproct (Durham end Morgan 1978). Even Pliocene dendrasterids occasionally have marginal or admarginal periprocts: Woodring et al. (1940) described D. coalingaensis tests (pi. 10; figs. 10, 12) and D. gibbsii tests (pi. 42; fig. 4) with these morphologies. However, Quaternary dendrasterids consistently have submarginal periprocts. Thus, the submarginal condition has apparently been favored over the course of dendrasterid evolution.
As the dendrasterid periproct moves onto the oral surface, it is accompanied by the adjacent plates in the posterior interambulacrum. The third post-basicoronal plates are of particular interest. In E. parma, the third plate pair is essentially marginal; the two plates barely extend onto the oral surface (fig. SC). The same condition occurs in juvenile D. excentricus (Durham 1955, fig. 11). In adult D. excentricus, however, the third pair is largely or entirely on the oral surface (fig. 5D).
The migration of these plates can be interpreted as a response to the oral plate disparity. In northern Pacific sand dollars, the posterior oral areas contain fewer plates than the anterior oral areas. In theory, a plate disparity of this sort could be counterbalanced in two ways, either by enlarging the plates in the posterior areas, or else by adding plates to the posterior areas. In E. parma, the first strategy is used: posterior development is favored during the second phase of unequal growth. In D. excentricus, the second strategy is used: posterior plates migrate from the margin onto the oral surface. The retrodisplacement of the periproct serves as a conspicuous marker of this plate migration.
Retrodisplacement of the Peristome.--In dendrasterid sand dollars, the peristome is consistently displaced toward the posterior end of the test (fig. 1C). The retrodisplacement of the peristome could be related to the unusual dendrasterid feeding behavior. For example, the eccentricity of the oral surface could help to keep the peristome above the substrate when the test assumes the inclined, suspension-feeding position. In addition, the displacement of the peristome could expedite the transport of food particles across the oral surface. When Dendraster suspension feeds, most food particles are collected on the posterior half of the test, and thus the posterior food grooves are much more highly developed than the anterior grooves (fig. 2B). The posterior displacement of the peristome would serve to reduce the length of the posterior food grooves, which could speed the overall movement of food to the mouth.
Retrodisplacement of the Periproct.--Mio-Pliocene dendrasterids often have marginal or admarginal periprocts. But in modern forms, the periproct always moves from the posterior margin onto the oral surface, into an area with a high density of food grooves (fig. 1 C). Certain other irregular echinoids, particularly clypeasteroids and holectypoids, also have submarginal periprocts (fig. 1B). There is no functional explanation for this morphology.
Apical versus Peristomial Eccentricity.--The apparent link between apical end peristomial eccentricity in dendrasterids (fig. 4) is not dictated by functional constraints. There is probably a functional correlation between the position of the peristome and the position of the "apex" (the highest point of the test). Both of these features are closely associated with the internal lantern; the peristome is centered beneath the lantern, whereas the apex is centered above it. However, this correlation does not extend to the apical system. In dendrasterids, unlike most other echinoids, the apical system is not located at the apex of the test: it typically occupies a separate position, between the apex and the posterior margin. Because the position of the apical system is not linked to the position of the apex, there is no reason to assume that it should be linked to the position of the peristome.
In irregular echinoids, the peristome commonly shifts toward the anterior end of the test (fig. 1A). But in dendrasterid sand dollars, the peristome undergoes retrodisplacement: it consistently moves in the opposite direction, toward the posterior end (fig. 1C). In dendrasterids, the peristome is consistently located on the posterior half of the test, which is a very rare condition among echinoids.
The dendrasterid periproct also undergoes retrodisplacement. In irregular echinoids, the periproct commonly migrates from the aboral surface to the posterior margin (fig. I A). But in dendrasterids, the periproct shifts anteriorly, rather than posteriorly (fig. 1C). It moves from the posterior margin onto the oral surface, into an area densely covered by food grooves (fig. 2B).
Why does retrodisplacement occur in dendrasterids? The posterior shift of the peristome could be related to the unusual dendrasterid feeding behavior. However, other aspects of retrodisplacement are harder to interpret. First, there is a strong correlation between the displacement of the peristome and the displacement of the apical system (fig. 4). This correlation does not appear to be dictated by functional constraints. Second, there is no obvious explanation for the retrodisplacement of the periproct. In fact, it is easier to see the adaptive costs of this feature than any adaptive benefits. Third, retrodisplacement is associated with a major change in oral development. In other northern Pacific sand dollars, the anterior and posterior ambulacra (and interambulacra) develop at different rates. In dendrasterids, the anterior and posterior oral areas develop at similar rates.
The developmental model presented here provides a means of linking these phenomena. The model proposes that the Dendraster pattern of development was derived from the Echinarachnius pattern. It assumes that a heterochronic change in the Echinarachnius pattern could simultaneously affect many aspects of test morphology, including the apical system, peristome, periproct, ambulacra, and interambulacra.
The Echinarachnius pattern of development includes two phases of asymmetric growth. The first phase, which favors anterior growth, occurs early in ontogeny. This growth phase permanently affects the distribution of plates on the oral surface of the test. Even in adults, the anterior areas have more plates than the posterior areas. The difference is not large (usually one or two plates per area), but it is real.
The second phase, which favors posterior growth, begins soon after metamorphosis. The second phase of unequal growth also has conspicuous effects on oral morphology. In adult tests, a distinct posterior-to-anterior gradient appears among the oral ambulacra: the posterior areas have larger plates than the anterior areas (figs. 5A, 6). Another posteriorto-anterior gradient affects the oral interambulacra: the posterior areas become discontinuous before the anterior areas (fig. SC).
The two phases of unequal growth are normally balanced. At maturity, the anterior and posterior areas are approximately equal, and thus the test appears to be symmetrical. However, if the second phase of unequal growth were weakened or eliminated, the effects on oral morphology would be profound.
First, the gradient in ambulacral morphology would disappear. The anterior and posterior ambulacra would have plates of roughly similar size. In fact, this feature is characteristic of dendrasterids (figs. 5B, 6).
Second, the gradient in interambulacral morphology would disappear. The anterior and posterior interambulacra would become discontinuous at about the same time. This is also a typical dendrasterid characteristic (fig. SD).
Third, the peristome would be displaced toward the posterior end of the test. The posterior oral areas have fewer plates than the anterior areas; if the second phase of unequal growth was suppressed, the posterior areas would remain disproportionately small. As a result, the effective distance between the peristome and the posterior margin would shrink. Posterior displacement of the peristome is one of two possible responses to this shrinkage ([D.sub.ps] of fig. 7). Nearly all dendrasterids show posterior displacement of the peristome (fig. 4). This character is particularly well developed in fossil species (fig. 3).
Fourth, the plates at the posterior margin, along with the associated periproct, would be displaced onto the oral surface. Anterior displacement of the marginal plates is the second possible response to the shrinkage of the posterior oral areas ([D.sub.pp] of fig. 7). This character occurs in nearly all dendrasterids, except for one early fossil species.
In summary, the Dendraster pattern of oral development can be derived directly from the Echinarachnius pattern, through suppression of the second phase of unequal growth. This model successfully accounts for the unusual features of the dendrasterid oral surface, including the retrodisplacement of the peristome and periproct. However, one important question remains: what adaptive benefits are conferred by this developmental change?
The best-known dendrasterid characteristic is the eccentricity of the apical system on the aboral surface (fig. 2A). In theory, apical eccentricity can be derived in the same manner as the oral characters described above, by suppressing the second phase of unequal growth. In addition, there is a strong correlation between the displacement of the apical system on the aboral surface and the displacement of the peristome on the oral surface (fig. 4). These theoretical and empirical considerations suggest that oral and aboral morphology may be linked. In other words, the processes that affect the development of the aboral surface may simultaneously affect the oral surface as well.
Apical eccentricity is regarded as a valuable adaptation, because it facilitates a unique form of suspension feeding. It is likely that apical eccentricity was derived from a heterochronic change in test development. But if this developmental change affected both surfaces of the test, then new oral characters would have been produced as well. These characters could be considered "by-products" of the process that produced apical eccentricity.
This model implies that the unusual oral characters of dendrasterids may have little or no inherent value: their existence may be due solely to developmental constraints. It is, of course, impossible to prove that a given character has no adaptive value. However, it is also difficult to provide plausible functional explanations for the retrodisplacement of the dendrasterid periproct, for the changes in ambulacral and interambulacral development, and for the strong correlation between apical and peristomial eccentricity. The alternative model presented here does account for these phenomena.
In conclusion, I propose that many seemingly unrelated characters can appear simultaneously, through a single, critical change in development. In such cases, the characters are effectively linked, and thus selective pressure for any one would inevitably produce the others as well. In dendrasterids, selective pressure for apical eccentricity on the aboral surface has simultaneously produced unusual new characters on the oral surface, including retrodisplacement of the peristome and periproct.
This study was based in part on a 1990 doctoral dissertation at the Johns Hopkins University, Department of Earth and Planetary Sciences. The author was supported at Johns Hopkins by a National Science Foundation Graduate Fellowship. The author gratefully acknowledges the assistance of G. Hendler, G. Kennedy, S. Suter, and C. Groves at the Los Angeles County Museum of Natural History. This paper was improved by comments from S. Stanley (the Johns Hopkins University), H. Lescinsky (University of California at Davis), R. Mooi (California Academy of Sciences), and two anonymous reviewers.
[Figures 1 to 7 ILLUSTRATION OMITTED]
Beadle, S. C. 1989. Ontogenetic regulatory mechanisms, heterochrony, and eccentricity in dendrasterid sand dollars. Paleobiology 15:205-222.
--. 1991. The biogeography of origin and radiation: dendrasterid sand dollars in the northeastern Pacific. Paleobiology 17: 325-339.
Durham, J. W. 1949. Dendraster elsmerensis Durham, n. sp. American Journal of Science 247:49-62.
--. 1955. Classification of clypeasteroid echinoids. University of California, Publications in Geological Sciences 31:73-198.
--. 1966. Phylogeny and evolution. Pp. U266-U269 in R. C. Moore, ed. Treatise on invertebrate paleontology, part U, Echinodermata 3. Geological Society of America and University of Kansas, Boulder, Colo. and Lawrence, Kans.
--. 1978. Polymorphism in the Pliocene sand dollar Merriamaster (Echinoidea). Journal of Paleontology 52:275-286.
Durham, J. W., and S. R. Morgan. 1978. New sand dollars (Echinoidea) of the genera Merriamaster and Dendraster from the Purisima Formation, California. California Academy of Sciences, Proceedings, series 4. 41:297-305.
Durham, J. W., C. D. Wagner, and D. P. Abbott. 1980. Echinoidea: the sea urchins. Pp. 160-176 in R. H. Morris, D. P. Abbott, and E. C. Haderlie, eds. Intertidal invertebrates of California. Stanford University Press, Stanford, Calif.
Eschscholtz, F. 1831. Zoologischer Atlas, enthaltend Abbildungen und Beschreibungen neuer Thierarten, wahrend des Flottcapitains von Kotzebue zweiter Reise um die Welt, auf der Russisch-Kaiserlichen Kriegeschlupp Predpriaetie in den Jahren 18231826. Viertes Heft. G. Reimer, Berlin.
Gordon, I. 1929. Skeletal development in Arbacia, Echinarachnius, and Leptasterias. Philosophical Transactions of the Royal Society of London B 217:289-334.
Gove, P. B., ed. 1976. Webster's third new international dictionary of the English language, unabridged. Merriam, Springfield, Mass.
Grant, U. S., IV, and L. G. Hertlein. 1938. The west American Cenozoic Echinoidea. Publications in Mathematical and Physical Sciences 2. University of California, Los Angeles.
Hayasaka, I., and A. Morishita. 1947. Notes on some fossil echinoids of Taiwan, II. Acta Geologica Taiwanica 1:93-110.
Kew, W. S. W. 1920. Cretaceous and Cenozoic Echinoidea of the Pacific coast region of North America. University of California Publications in Geology 12:Z3-236.
Kier, P. M. 1974. Evolutionary trends and their functional significance in the post-Paleozoic echinoids. Paleontological Society Memoir 5, Journal of Paleontology 48(3)1-95.
Lohavanijaya, P., and E. E Swan. 1965. The separation of postbasicoronal areas from the basicoronal plates in the interambulacra of the sand dollar, Echinarachnius parma (Lamarck). Biolagical Bulletin 129:167-180.
Nisiyama, S. 1968. The echinoid fauna from Japan and adjacent regions, II. Paleontological Society of Japan, Special Paper 13.
Raup, D. M. 1956. Dendraster: a problem in echinoid taxonomy. Journal of Paleontology 30:685-694.
Smith, A. 1984. Echinoid palaeobiology. Allen and Unwin, London.
Stanton, R. J., Jr., J. R. Dodd, and R. R. Alexander. 1979. Eccentricity in the clypeasteroid echinoid Dendraster: environmental significance and application in Pliocene paleoecology. Lethaia 12:75-87.
Timko, P.L. 1976. Sand dollars as suspension-feeders: a new description of feeding in Dendraster excentricus. Biological Bulletin 151:247-259.
Wagner, C. D. 1970. Evolution among some clypeasteroid echinoids. Ph.D. diss. Department of Paleontology, University of California, Berkeley.
Woodring, W. P., R. Stewart, and R. W. Richards. 1940. Geology of the Kettleman Hills oil field, California. U.S. Geological Survey, Professional Paper 195. U.S. Government Printing Office Washington, D.C.
|Printer friendly Cite/link Email Feedback|
|Author:||Beadle, Steven C.|
|Date:||Dec 1, 1995|
|Previous Article:||Variation in species diversity and shell shape in Hawaiian land snails: in situ speciation and ecological relationships.|
|Next Article:||Origin and recent endemic divergence of a Caspian Mysis species flock with affinities to the "glacial relict" crustaceans in boreal lakes.|