© 2000 by The Society for Integrative and Comparative Biology
What a New Model of Skeletal Homologies Tells Us About Asteroid Evolution1
1 Department of Invertebrate Zoology and Geology, California Academy of Sciences, Golden Gate Park, San Francisco, CA 94118-4599
2 UMR CNRS 5561, Université de Bourgogne 6, bd. Gabriel F-21000, Dijon, France
 |
SYNOPSIS |
|---|
TOP
 SYNOPSIS INTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
The Extraxial-Axial Theory (EAT) is applied to the body wall homologies of asteroids. Attempts to characterize major plate systems of asteroids as axial or extraxial, particularly those that are highly organized into series, can be problematic. However, the Optical Plate Rule (OPR) is instrumental in establishing that ambulacrals and terminals are axial. It is equally clear that the region aboral to the marginal frame is a part of the perforate extraxial body wall (with the possible exception of the centrodorsal, which is likely imperforate extraxial). Previously established EAT criteria, particularly those strongly rooted in the embryologically expressed boundary between axial and extraxial body wall in larvae, suggest that marginals, and perhaps adambulacrals, are extraxial in origin. We also explore the extraxial nature and phylogenetic significance of the odontophore. Our data from both juveniles and adults show that plate and tube foot addition sequences occur according to the OPR, and shed light on poorly known homologies of the asteroid mouth frame. These data indicate that the mouth angle ossicle must at least contain the first ambulacral, although we cannot rule out the possibility that the first adambulacral also contributes to the construction of this ossicle. The interpretations provided by the EAT for all ossicles suggest a synapomorphy scheme for somasteroids, ophiuroids, and asteroids.
|
INTRODUCTION |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Recent works on asteroid phylogeny are only now focusing on the cladistic data that can help achieve an understanding of the homologies of stelleroids (Blake, 1987
; Gale, 1987; Smith and Jell, 1990; Lafay et al., 1995). However, few of these studies agree on even the most basic of scenarios or tree topologies. Some of the problems stem from a lack of appropriate data for certain taxa, and from a diversity of approaches and proclivities of individual workers. But a significant amount of difficulty lies in the lack of a framework for homologies rooted in morphology, ontogeny, and paleontology. As a result, many interpretations of stelleroid morphology exist for the origins and homologies of even the most distinctive of stelleroid features. These include adambulacral and marginal plates, and ossicles involved in the mouth frame (see Fig. 1; Spencer and Wright, 1966; Blake, 1989; Hotchkiss, 1993). Ontogeny can help us trace homologies that become obscured by subsequent modifications of adult morphology. Paleontology, like ontogeny, can be helpful in providing fossils that do not yet exhibit some of the evolutionary events (autapomorphies) that might have "written over" and masked synapomorphies among basal forms. Therefore, in many cases, ontogeny and paleontology can remove some of the burden of autapomorphy that obscures the earliest events in the history of organisms. However helpful this removal can be, there needs also to be a cohesive model of homologies that can draw upon the integration of morphology, ontogeny, and paleontology.
We have developed such a model, the Extraxial-Axial Theory (EAT), for the Echinodermata, using empirical data on the configuration of certain plate systems, their ontogeny, and distribution among major clades (Mooi et al., 1994; David and Mooi, 1998; Mooi and David, 1997, 1999). The EAT embodies basic ontogenetic differences in two main types of echinoderm skeleton: extraxial and axial (Mooi et al., 1994). We have argued that the model operates similarly to the system of homologies established for the vertebrates (Mooi and David 1997, 1999). Like the vertebrate scheme, the EAT can provide a basis for character analysis that removes some of the "noise" in approaches scoring superficial similarities as if they were homologous.
Axial skeletal elements are added according the Ocular Plate Rule (OPR) which summarizes several principles common to all members of the phylum. 1) New axial plates are added in a blastema at the end of growing ambulacral series usually associated with terminal plate (called the ocular in echinoids). 2) Each of the 5 terminals marks the end of a growing radial component of the hydrocoel (that is, the radial canal of the water vascular system). 3) The newest plates in the series are at the tip of the ambulacrum, and the oldest are adjacent to the mouth. 4) In virtually all echinoderms, the new plates form a biserial sequence of ambulacrals, and each ambulacral is associated with a tube foot. 5) In all echinoderms investigated so far, the tube foot appears after the plate begins to form, so that the tube foot is separated from the mouth by at least part of the most proximal ambulacral (referred to as the basicoronal, or mouth frame plate in some forms). 6) In most echinoderms the paired columns of each axial series are staggered, so that there is a zig-zag suture separating the 2 columns in the biserial sequence. In asteroids and ophiuroids, the ambulacrals in these columns have secondarily lined up to form paired ambulacrals.
In extraxial skeleton, new elements can be added anywhere. Although certain extraxial elements in some taxa can show secondary organization, there are no ontogenetic principles applicable to all echinoderms that govern the pattern in which extraxial elements are added during growth. Extraxial skeleton can be further subdivided. Perforate extraxial skeleton contains several types of orifices not found in any other type of skeleton, including epispires, hydropores, gonopores, and the anus. Plate systems falling under the general category of perforate extraxial skeleton include thecal plates such as the radials and basals of some pelmatozoans, and plates associated with the aboral surfaces of arms such as the brachials of crinoids. An expanded list can be found in Mooi and David (1997: Table 1). Imperforate extraxial skeleton is virtually absent in the living taxa except crinoids. In crinoids, it appears to be restricted to stems and holdfasts, but is probably represented by the centrodorsal in stelleroids.
The morphological criteria for recognizing components of the EAT are based on well-known embryological patterns correlated with these body wall regions (David and Mooi, 1996, 1998; summarized in Mooi and David, 1997, 1999). Axial skeleton is strictly associated with the rudiment in larval echinoderms. The first skeletal elements of the axial system appear adjacent to the primordial lobes. New plates in the developing ambulacral system are laid down according to the OPR in the pattern described above. No extraxial elements are known to appear in association with hydrocoel, or according to the OPR. In forms such as echinoids, there is a well-demarcated boundary between the region in which axial elements appear, and that in which extraxial elements originate. We use this to make a distinction between the rudiment and non-rudiment parts of the echinoderm larva. The axial and extraxial body wall divisions are derived from the rudiment and the non-rudiment, respectively, indicating that the two skeletal types are established early in development (David and Mooi, 1996). During metamorphosis of taxa such as echinoids, much of the non-rudiment portion of the body is lost or histolyzed. However, significant portions of the non-rudiment region is retained into adulthood in some clades such as asteroids. Therefore, establishing the position of the boundary between rudiment and non-rudiment is of great importance in distinguishing between axial and extraxial body wall of stelleroids.
Using the EAT, Mooi and David (1997) and David and Mooi (1998) have questioned the homology of the marginal rings in stelleroids and edrioasteroids, suggested homologies between arms of crinozoans and stelleroids, and undermined the concept of the cryptosyringids (Smith, 1984). However, more comprehensive application of the EAT to asteroids in particular is needed. This paper is intended to use the model to address homologies in a variety of asteroid plate systems, with a focus on the vexing topic of the asteroid mouth frame.
|
ASTEROIDS AND THE EXTRAXIAL/AXIAL BOUNDARY |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Spencer and Wright (1966)
originally developed the terms axial, adaxial, and extraxial as categories for the plate systems they saw in stelleroids. Our uses of the terms axial (1966, p. U9) and extraxial (1966, p. U17) are partially congruent with theirs. Adaxial structures such as adambulacrals and marginals (Figs. 1A, B) form the focus of much of this work because adaxial systems do not occur in other echinoderms to which the EAT has been applied.
|
In echinoids, the boundary between axial and extraxial body wall types becomes more prominent as the rudiment develops. The distinction of this demarcation is related to a peculiar aspect of echinoid morphology: the adult echinoid's test is almost entirely axial (Mooi et al., 1994
). The drastic reduction of extraxial body wall is an important feature of echinoid metamorphosis, and is attended by the post-metamorphic restriction of the extraxial skeleton to the apical system. This maintains a clear boundary between the two body wall types.
Post-metamorphic development is very different in echinoids and stelleroids. Echinoids cease adding extraxial elements at this point, but asteroids continue to produce extraxial ossicles throughout ontogeny. The retention of a relatively large amount of extraxial skeleton into adulthood results in a less dramatic post-metamorphic reduction of the extraxial region than in echinoids. This difference means that it is more difficult to discern the axial/extraxial boundary in adult asteroids. However, in early post-metamorphic development, a clear differentiation between aboral extraxial elements (notably the centrodorsal [centrale], primary and subsequent interradials, and primordial carinals) can be seen in developing asteroids (Fewkes, 1888a: Pl. I, Fig. 6; Komatsu, 1975: Figs. 25–34a; Komatsu and Nojima, 1985: Fig. 2).
These observations indicate that the terminals are formed in association with the hydrocoel in the rudiment, as in other echinoderms. In contrast, the "primary circlet" of centrodorsal and primary interradials form more aborally, close to, or embedded in the non-rudiment part of the body wall. Hyman (1955: Fig. 126) suggested homology between interradial plate pairs formed on the aboral surface of juvenile Asterina and the first interambulacrals of echinoids. We do not agree because these plates are extraxial in asteroids, whereas the interambulacrals are axial in echinoids. Therefore, in these early stages, the boundary can be drawn between these aboral plates and those formed adoral to the terminals. As new elements are added to both axial and extraxial systems, the boundary becomes more difficult to determine because the two systems form closely spaced series along the arms, and tightly integrate to form the adult morphology. We will return to this problem below.
Few authors have explored the implications of Spencer and Wright's (1966) uses of the terms axial, adaxial, and extraxial. However, Hotchkiss (1993) provided a listing of stelleroid and crinozoan plate systems. In his scheme, extraxial skeleton (principally plates of the aboral surface, including the primary circlet) were listed as elements "derived at metamorphosis from the right side of the larva" (Hotchkiss, 1993: Table 2). Those derived from the left side of the larva (the rudiment) included axial, adaxial, admarginal, and marginal systems (Hotchkiss, 1993: Table 3).
The EAT is largely in agreement with Hotchkiss' assignments as far as Spencer and Wright's (1966) extraxial components are concerned, but discussion is warranted concerning the derivation of adaxial (adambulacral and marginal) systems from the left side of the larva. The assignment of these systems to the EAT body wall regions rests on where the boundary is placed in larval asteroids. In most asteroids, by the time these series are forming, the animal has already achieved a post-metamorphic level of development (imago). The distinction between rudiment and non-rudiment portions of developing asteroids is much more difficult to make at this later stage. Nevertheless, Hotchkiss (1993) has neatly highlighted a major problem in asteroid homology, which we address below.
|
AXIAL ELEMENTS: AMBULACRALS AND TERMINALS |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
With respect to the homologies of asteroids discussed here, the single most important aspect of the EAT is the OPR. The OPR can be used to establish the pattern of plate formation in the axial skeleton. In all other echinoderms, axial skeleton includes body wall elements formed according to the OPR, such as terminals (oculars), flooring plates, cover plates, and basicoronal (= mouth frame) plates (Mooi and David, 1997
: Table 1). In asteroids, it is easy to identify the terminal and ambulacral plates as part of the axial system, and establish their relationship to the hydrocoel components (ring and radial canals) normally associated with the axial skeleton. Blake and Guensburg (1988) reviewed this association, and described not only variations upon this theme among the asteroids, but also the strongly conserved basic pattern emphasized below.
That axial elements in asteroids follow ontogenetic trajectories characteristic of the OPR is evident from descriptions of skeletal development for forms as diverse as forcipulates (Fewkes, 1888a; Gemmill, 1914 [Asterias]; Gordon, 1929 [Leptasterias]), paxillosidans (Hörstadius, 1939; Komatsu, 1975; Komatsu and Nojima, 1985 [Astropecten]), velatids (Gemmill, 1912 [Solaster], 1920 [Crossaster]); and spinulosidans (Siddall, 1979 [Echinaster]). Hyman (1955) and Blake (1994) have provided excellent summaries of the literature dealing with ontogenetic events in the development of skeletal elements to be discussed below. Coupled with our own observations of very early post-metamorphic specimens we have tentatively identified as Pteraster (Fig. 2), this literature helps us to form a generalized picture of the operation of the OPR in asteroid rays (Fig. 3).
|
|
Ambulacrals develop along the growing radial canals of the water vascular system. In asteroids, new ambulacrals are added in a blastema associated with the terminal plate at the end of growing ambulacral series. Therefore, the newest ambulacrals are adjacent to the terminal, and the oldest are proximal to the mouth (Fig. 3). As in non-stelleroid echinoderms, asteroid ambulacrals form a biserial column. However, asteroids are unusual among echinoderms in that the ambulacrals are directly opposite each other across a straight perradial suture, rather than staggered to create a zig-zag suture (Figs. 1, 2A).
Most importantly, as in other echinoderms, each ambulacral is associated with a tube foot, and there is a single tube foot per plate for the entire ambulacrum. The new tube foot appears just distal to the newly formed ambulacral. Therefore, moving distally from the mouth, the pattern is: ambulacral, tube foot, ambulacral, tube foot, and so on (Figs. 2, 3). This appears to be plesiomorphic and ubiquitous in echinoderms (Mooi and David 1999). Therefore, there can be little doubt that the ambulacrals, terminals, and ring and radial water vessels of asteroids contribute to the typical axial system as seen in other echinoderms.
|
UNAMBIGUOUSLY EXTRAXIAL ELEMENTS |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
As Spencer and Wright (1966)
pointed out, the areas aboral to the marginals are extraxial, and so are the members of the primary circlet. The position of these plates on the non-rudiment side of the larva (see above) suggests that ossicles occurring on the aboral surface are extraxial. In many asteroids, these elements are anisotropically arranged, space-filling ossicles—a hallmark of extraxial skeleton. We fully recognize that many of the plates found in this region can be organized into series, and that the arrangement of plates by itself might not signify that they are extraxial. However, it is equally apparent that this ordering of plate systems is secondary. The plesiomorphic condition for stelleroids, as shown by basal ophiuroids and Archegonaster (Smith and Jell, 1990) is for the aboral surface to be composed of many, small, disorganized ossicles similar to the perforate extraxial skeleton of basal echinoderms such as edrioasteroids (Mooi and David, 1999).
Virtually all extraxial body wall in asteroids is of the perforate variety, since it includes body openings such as papular pores, hydropores in the madreporite, gonopores, and the anus. Imperforate extraxial body wall is greatly reduced in asteroids, with the possible exception of the centrodorsal. If the centrodorsal is homologous with the distal attachment plate in the stem of juvenile comatulid crinoids (Hotchkiss, 1993), then it is likely a vestigial imperforate extraxial element. The dramatic reduction of the imperforate extraxial region is a major event in the diversification of the Echinodermata. We have interpreted it as a synapomorphy for asteroids, ophiuroids, echinoids, and holothuroids, to the exclusion of the crinoids (Mooi et al., 1994; Mooi and David, 1997; David and Mooi, 1998).
Many asteroids possess a subset of aboral plates known as carinals running along the top of the arm (Fig. 1). Hotchkiss (1993) reviewed the homologies and phylogenetic significance of upper arm plates in ophiuroids, concluding that they were homologous with asteroid carinals. If true, then the presence of carinals in asteroids could be interpreted as a plesiomorphic character. Lovén (1874: Pl. 3) illustrated considerable variation in carinal pattern among rays in the same specimen, and what appear to be new carinals forming several plates proximal from the terminal. This reinforces the inference that carinals are extraxial.
|
PROBLEMATIC EXTRAXIAL ELEMENTS |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Adambulacrals
In asteroids, the ambulacrals arch to form a deep ambulacral groove (Fig. 1B). The perradial parts of the ambulacrals are flexed upward, with their abradial ends sitting on the adambulacral series (Fell, 1963
; Blake, 1998) (Fig. 1A, B). The two series are offset by one-half of a plate, so that the ambulacral precedes the adambulacral as you move away from the mouth. However, for almost the entire length of the arm (with the possible exception of the mouth angle ossicles), there is still a single adambulacral for each ambulacral (Blake and Guensburg, 1988). In somasteroids and ophiuroids, which appear to show the plesiomorphic condition, the adambulacrals are placed laterally, opposite the abradial ends of the ambulacrals (Fell, 1963; Smith and Jell, 1990; Blake, 1998) (Fig. 4A, B).
|
It is the use of the term "adaxial" that forms the greatest point of departure between our interpretation and that of Spencer and Wright (1966)
. Spencer and Wright considered adambulacrals to be part of the adaxial skeleton, but this does not address the question of whether they originate on the axial or extraxial side of the larval boundary between these body wall types. The OPR seems to be operating in the case of adambulacrals, because the oldest adambulacrals are near the mouth, and the youngest near the terminal. In addition, both the adambulacral and marginal series originate on the oral side of the terminals (Figs. 2, 3). This reasoning has suggested to some workers (Mooi et al., 1994; Mooi and David, 1997; David and Mooi, 1999) that adambulacrals are part of the axial system.
However, studies of basal stelleroids suggest that the adambulacrals are not part of the axial body wall. Hyman (1955) suggested that the first virgals of somasteroids and the adambulacrals of asteroids were homologous. Paul and Smith (1984, p. 469) stated that the "primary cover plates of stromatocystitids became transformed into the adambulacral ossicles of asteroids and the lateral plates of ophiuroids." However, we are unable to support the suggestion that cover plates and adambulacrals are homologous, largely because no mechanism is proposed whereby an external appendage such as a cover plate can become an element of the body wall. In the somasteroid Archegonaster, Smith and Jell (1990) identified an adambulacral supporting the abradial members of the rest of each virgal series. Blake and Guensburg (1993) inferred that these were homologous to the "first virgal" of other somasteroids. The OPR does not operate in the formation of virgals, and we would not hesitate in assigning virgals to the extraxial body wall. The implication is that adambulacrals would be extraxial as well.
The terminal of asteroids is unlike that of other echinoderms in being secondarily enlarged, leading to "sheltering" of newly formed adambulacrals. However, in ophiuroids, the newly formed adambulacrals form outside of the terminal, off to either side of it (Murakami, 1937: Fig. 7). In contrast, the newly formed ambulacrals are clearly within the proximal arc of the terminal. This condition is also evident in somasteroids (which have a very small terminal), suggesting that this is the plesiomorphic condition for stelleroids. The first adambulacrals also appear to be a part of the interradial skeleton confluent with the aboral surface, rather than a part of the rudiment (Fewkes, 1887: Pl. III). This might constitute embryological evidence that in stelleroids, the boundary between elements produced in the rudiment and those produced outside the rudiment is between the ambulacral and adambulacral series.
If adambulacrals are extraxial, some of the odd displacements of the adambulacral series that we have observed in the "adoral carina" (a series of tightly associated proximal adambulacrals found among forcipulate taxa [Fig. 1A]) could be accounted for by the fact that unlike the ambulacrals, adambulacrals are not formed according to the strictures of the OPR. Also, some of the variations seen in Paleozoic stelleroids (sublaterals, and the possible absence of adambulacrals in the distal regions of the arms in some forms) can be attributed to the evolutionary lability of the extraxial skeleton (Mooi and David 1997).
Marginals
Like adambulacrals, marginals (supra- and inframarginals [Fig. 1A, B]) display superficial obedience of the OPR. Blake (1978) showed quite convincingly that marginals, even when they are represented by multiple rows of serially homologous plates, arise adjacent to the terminal. In contrast to adambulacrals and ambulacrals, which originate oral to the terminal, marginals (as well as carinals) originate on the aboral side (Gordon 1929: Fig. 25A), outside the proximal convexity of the terminal. Also, marginals do not show the one-to-one correspondence with ambulacrals that adambulacrals display. This suggests distinct modes of formation for adambulacrals and marginals, and raises doubts as to whether the OPR operates for marginals in particular.
The ossicles between the marginals and the adambulacrals on the oral surface (Fig. 1A), are clearly extraxial. If marginals are axial, then we would have a region of extraxial skeleton between two axial systems on the oral surface. Therefore, a body wall system known to be contiguous in all other echinoderms would be rather oddly split by extraxial skeleton (David and Mooi, 1999).
Even if it is technically possible for the OPR to operate in the formation of marginals, we feel that they are best considered as highly ordered extraxial ossicles. It is not unusual for extraxial skeleton to follow ontogenetic patterns that converge on the OPR. Even imperforate extraxial skeleton, such as in crinozoan stems, can exhibit regularities in the way new ossicles are added (Roux 1997; Breimer 1978). In attenuated but growing areas of echinoderm body wall such as arms, there are few ways or places to add new elements to expanding body wall areas. The addition of new axial ambulacrals necessitates the addition of new extraxial plates at the arm tip to allow the rest of the body wall to keep up with axial plate addition. This can be likened to the construction of a new highway. The road itself lengthens when the pavers put down new asphalt at its terminus. The guard rails and power lines go up in the same place, at the same time, in pace with the lengthening end of the highway. From a distance, it all looks like part of the same process. However, closer inspection reveals differences between the mechanism putting asphalt down, and that putting up the rails and poles.
So it is perhaps unsurprising to see a convergent, OPR-like pattern in serially homologous plate systems like marginals. One way to test for extraxial attributes of these systems is to search for new marginal ossicles added in places other than near the terminal. Verrill (1914: Fig. 1) illustrated how otherwise extremely similar rows of ossicles running along the sides of the arm can vary in the distance between their distalmost member and the terminal.
Whether marginal frames are axial or extraxial, they are not homologous in edrioasteroids and stelleroids. Blake (1994), Mooi and David (1997, 1999) and David and Mooi (1999) reviewed this hypothesis in the context of recent attempts to derive asteroids from edrioasteroids (Smith and Jell, 1990). Marginal rings of stelleroids and edrioasteroids are fundamentally different. In the former, they are embedded within perforate extraxial skeleton, whereas in the latter, they separate imperforate from perforate extraxial regions.
|
ODONTOPHORES |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
In asteroids, the odontophore is a small, often butterfly- or shield-shaped ossicle situated interradially above the distal ends of the mouth angle ossicles (Fig. 1C). Spencer and Wright (1966
: p. U14) suggested that the odontophore was axial: "an inframarginal which in the course of phylogeny became occluded from the margin and adapted as part of the ‘jaw’ system." McKnight (1975) separated somasteroids from asteroids at least in part because of the possession of an odontophore in the latter. Blake (1990, pp. 354–355) suggested that an unpaired, axillary ossicle like that shown in juvenile Promopalaeaster "is well positioned to be the homologue of the modern odontophore." Although the axillary in Paleozoic palasteriscids is external, it is just distal to the mouth angle ossicles, like the odontophore in crown asteroids. The axillary is also shield-shaped (Blake, 1994) and similar enough to the odontophore that it gets labeled as such in these Paleozoic forms (Spencer and Wright, 1966: Fig. 41).
The odontophore does not form as part of what is now recognized as axial skeleton, and seems to represent an internalized member of the marginal series (Blake, 1990). If this is the case, then whether or not it is homologous to the axillary, the odontophore is best interpreted as part of the perforate extraxial body wall. The possession of an odontophore internal to the mouth angle ossicles can only be interpreted as an apomorphy of crown asteroids.
|
MOUTH ANGLE OSSICLES |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Previous interpretations
Spencer and Wright (1966
, p. U13) stated: "There has been much argument about the origins and homologies of the mouth frame." Hyman (1955, p. 260) suggested that in the mouth frame, "there are two adambulacral and four ambulacral pieces for each arm, as the ambulacral pieces are composed of the first two ambulacrals of each side of the ray." Hyman then indicated that in some asteroids, the adambulacral elements were most prominent, making an "adambulacral peristome". In the Asteriidae, the ambulacral pieces were larger, creating an "ambulacral peristome". Spencer and Wright (1966) also recognized these two types of mouth frame. Fell (1963) showed that the mouth angle ossicles (MAOs) in somasteroids are made of the first ambulacrals (Fig. 4A). The derived nature of the mouth frame, along with the aforementioned modifications, makes determining the homology of the MAOs a major problem in asteroid phylogeny.
Blake (1994) summarized the literature on mouth frame homologies, particularly that on the ontogeny of MAOs. He concluded that more research was needed, particularly because some evidence suggested that MAOs are part of the ambulacral series, but that other fossil evidence suggested an adambulacral origin. Smith and Jell (1990) homologized the MAOs of asteroids with ophiuroid jaw plates, largely on the basis of their work with the Paleozoic somasteroid, Archegonaster (Fig. 4A). Hendler (1978) showed that the MAOs in ophiuroids, like those in somasteroids, originate as the first ambulacrals (Fig. 4B). This fits the OPR because the tube foot is just distal to its corresponding ambulacral. If somasteroids and ophiuroids build their MAOs from the first ambulacrals, it becomes important to determine if the asteroids build theirs in the same way.
Asteroid MAOs line up directly with the adambulacral series, and bear sculpturing and spination very similar to that found on adambulacrals (Fig. 5). Ambulacral plates, on the other hand, do not bear spines at any point in their ontogeny. The highly derived musculature that joins adambulacrals in the more distal part of the series continues onto the MAO. Therefore, unlike the situation in somasteroids and ophiuroids, there is a strong superficial resemblance between the outer parts of the MAO and the proximal adambulacral ossicles. In addition, if the MAO is considered the first adambulacral (Fig. 4C), then the first ambulacral would also be associated with an adambulacral ossicle, just as in the rest of the arm.
|
Difficulties arise when the OPR is applied to this interpretation, and when comparisons are made with other echinoderms. If the MAO is adambulacral, then the first tube foot is proximal to the first ambulacral. This situation would be a unique departure from the pattern seen everywhere else in the Echinodermata. The ambulacral mouth frame of other echinoderms also supports the ring canal, which is the situation seen in asteroid MAOs.
Fewkes (1888a, p. 23) suggested that a given MAO did not form as a separate ossicle, but as a greatly enlarged abradial part of the first ambulacral. In spite of this, he called the MAOs "adambulacral circumorals," probably in reference to the position of the ossicles. It is not clear whether Fewkes thought the MAOs represented a composite element, although he ultimately seems to conclude that the MAOs are made up of a joining of first ambulacrals and first adambulacrals (Fig. 4D).
In a paper that appeared slightly later the same year, Fewkes (1888b) seems to have decided that the MAOs were not composite plates. However, he was still unable to say definitely "whether these oral plates are homologous with the other ambulacrals or adambulacrals of the arm" (Fewkes 1888b, p. 100). In spite of this, he showed quite conclusively that the developmental pattern of the MAOs does not justify the division of asteroids into "those with an adambulacral (non-forcipulates) and those with a combination of ambulacral and adambulacral (forcipulates) mouth plates" (Fewkes, 1888b, p. 101), a finding that remained overlooked by Hyman (1955) and Spencer and Wright (1966).
Gemmill (1912) turned Fewkes' (1888a) conclusion on its head by deriving the first ambulacrals from perradial extensions of the first adambulacrals (Gemmill, 1920) (Fig. 4C). Again, it is not clear that they were meant to represent composite entities.
Gordon (1929, p. 329) did not support either of these ideas because in "Leptasterias as in Asterias no separate adambulacral plate is formed in the neighborhood of the oral ambulacral plate. Instead, the outer extremity of each oral ambulacral plate is greatly enlarged and thickened." Hörstadius (1939: Figs. 36, 37) showed much the same configuration, with each MAO "closing" what would be the proximal edge of the first tube foot pore. Siddall's (1979) electron micrographs support this idea, although it is difficult to see if there is fusion between the MAO and the next ambulacral (Blake 1994), or if they are one and the same plate. If Gordon (1929) is correct, then the MAO is the first ambulacral, with a hypertrophied abradial part of that ossicle coming to lie in an interradial position, in line with the adambulacral series (Fig. 4E).
Mouth angle ossicles in adults
It is important to know if this pattern can be detected in adults, so that a survey of other asteroid groups for which ontogenetic data is lacking can be made. Using material housed at the California Academy of Sciences, we prepared mouth frames from each of the orders listed in Blake (1989)(species data available from the first author): Brisingida (Brisingidae: Astrolirus), Forcipulatida (Asteriidae: Pisaster, Asterias, Rathbunaster; Heliasteridae: Heliaster; Zoroasteridae: Zoroaster; Myxoderma), Valvatida (Asterinidae: Asterina; Goniasteridae: Mediaster), Notomyotida (Benthopectinidae: Pectinaster), Paxillosida (Astropectinidae: Astropecten; Luidiidae: Luidia; Ctenodiscidae: Ctenodiscus, Porcellanasteridae: Eremicaster), Velatida (Solasteridae: Solaster, Pterasteridae: Pteraster), Spinulosida (Echinasteridae: Echinaster, Henricia).
In every case, the MAOs were represented by a single ossicle from each half-ray, without vestigial fusion planes that might indicate they were composite entities. In addition, the aboral, more internal part of each MAO supported a portion of the ring canal, and joined the "first" ambulacral (sensu Fewkes, 1888a) on either side of the pore through which the first tube foot passes (Figs. 1C, 2, 5). The MAO is then followed by a tube foot, which is in turn followed by an unequivocal ambulacral (Figs. 4E, 5).
New interpretations
Hypotheses regarding the affinities of the MAOs cannot be based on observations of the condition in adults alone, and it is for this reason that previous work has resulted in the diversity of arguments that Spencer and Wright (1966) noted. It is not possible to falsify or support the suggestion that the MAO is ambulacral in origin without careful consideration of developmental and paleontological evidence.
Gordon (1929) was the first to explicitly recognize the ontogenetic derivation of the MAO from the first ambulacral. If she was correct, then the so-called "first ambulacral" of many studies (e.g., Fewkes, 1888a) is in fact the second ambulacral (Fig. 4E, 5). This is similar to the configuration in ophiuroids, in which the partially fused first and second ambulacrals of adjacent rays join interradially to form the jaw (Hendler, 1978) (Fig. 4B). In asteroids, the first ambulacrals do not meet across the perradius, but are twisted away from the rest of the ambulacral series to meet interradially, as in somasteroids (Fig. 4A). The result is an abradial bend in each half-ambulacrum as it approaches the mouth, forming a small buccal slit like that in Paleozoic stelleroids (Fig. 4A).
The presence of spines on the MAO has caused some to consider the MAO to be at least partially adambulacral in origin, because spines have been construed as a "marker" that distinguishes ambulacrals from adambulacrals (Hyman, 1955). We are unable to see why superficial resemblance induced by what are probably adaptive features of the MAO should overturn the evidence supporting the purely ambulacral nature of the MAO: 1) the absence of sutural vestiges that could indicate fusion with any other plates (see below); 2) the development of the MAO directly from the first ambulacral in early ontogeny; 3) the ambulacral, tube foot, ambulacral pattern signifying the OPR pattern seen in all other echinoderms; 4) the presence of the ring canal on the MAO.
We cannot completely rule out the possibility that the MAO represents a first ambulacral to which the first adambulacral has been fused (Fig. 4D). If so, then the adambulacral that should be associated with the first ambulacral has carried this association to an extreme by fusing with the first ambulacral. The remarkable similarity, spination, and alignment of the external part of the MAO to the adambulacrals would thereby be explained. A major problem is that there is no evidence of fusion at any point in the ontogeny of the asteroid MAO. Fusion, although uncommon in echinoderms, usually leaves evidence that there was a suture between the fused elements at some point. This can show up as faint suture lines between, or in the early ontogeny of the ossicles in question, as in ophiuroid jaws (Hendler, 1978) or vertebral ossicles (Ludwig, 1882; Fewkes, 1887).
Hendler (1978, p. 90) stated that "there seems to be no unequivocal homologue of the adambulacral-1 [the adambulacral that would be immediately distal to the MAO in asteroids] in the oral frame of recent ophiuroids." Therefore, it would appear that ophiuroids lack the adambulacral that would otherwise be in association with the first ambulacral (Figs. 4B, E). In other words, the situation in ophiuroids undermines the idea that every ambulacral must have a corresponding adambulacral. This is consistent with the idea that the MAOs of asteroids are made only of ambulacral elements, because the loss of the adambulacral ossicle that should be associated with the first ambulacral would represent a synapomorphy for asteroids, ophiuroid-like somasteroids, and ophiuroids (Fig. 6), rather than provide evidence against the configuration in Figure 4E.
|
Therefore, the least complicated way to interpret these observations is to suppose that the MAO is the first ambulacral. As the mouth frame of both asteriid and non-asteriid taxa matures, the relationships of the MAO to the rest of the axial series becomes more difficult to trace. During ontogeny, the MAO becomes highly modified, with a hypertrophied external region that is turned abradially from the ambulacral position and then downward to place it in alignment with the adambulacral series (Figs. 2, 3, 5).
In summary, the available evidence from both ontogeny and adult morphology suggests that the first ambulacral is at least involved in the construction of the MAO. It is possible that the first adambulacral is also part of the MAO, but there is some good evidence against this idea. If the MAO is purely first ambulacral, then the asteroids have essentially the same condition seen in other stelleroids.
|
IMPLICATIONS FOR PHYLOGENY OF THE STELLEROIDS |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Earlier application of the EAT (Mooi and David, 1997
; David and Mooi, 1998, 1999), as well as the work of other authors (Blake, 1994) has resulted in a topology that makes the cryptosyringid grouping (Smith, 1984, 1988, inter alia) untenable. Recent work has mirrored Fell (1963) in suggesting that arms of crinoids and stelleroids are of the same general construction (Mooi et al., 1994; David and Mooi, 1999), and that the derivation of stelleroids from an edrioasteroid ancestor is not well supported by application of EAT principles. The emerging picture is that shown in David and Mooi (1998: Figs. 3, 4), with a monophyletic Asterozoa possessing arms homologous with those of crinozoans (David and Mooi, 1999).
Blake (1998) lists several features that should be studied in any phylogenetic analysis of the Asterozoa. Because Blake did not explicitly place these onto a tree topology, we have used his list, in conjunction with the findings described above to suggest synapomorphies for the major stelleroid clades (Fig. 6). For example, based on our analysis of MAO homologies, we suggest that the enlarged first ambulacral (character 23) is a synapomorphy for the asteroids, and a further modification of character 7 that supports monophyly of the ophiuroids and asteroids. Reexamination of somasteroid morphology in light of the EAT is imperative, as several characters (5–10) indicate that exploration of the possibility that somasteroids are not monophyletic is warranted (Fig. 6). Some, such as Chinianaster and Villebrunaster, appear to be basal to the stelleroids, but others, such as Ophioxenikos, appear to form a stem lineage to the ophiuroids (Fig. 6). There are still large gaps in our knowledge of forms pivotal to our understanding of the phylogeny of stelleroids, some of which are already being filled by work such as that of Dean (1998).
Successful reinterpretation of major features of stelleroids underscores the importance of character analysis. Ultimately, we seek an integration of morphology, ontogeny, and paleontology to produce a reliable phylogeny that is based not only on many characters, but on independent characters in whose homologies we can have a high level of confidence.
|
ACKNOWLEDGMENTS |
|---|
We wish to thank D. Blake, D. Janies, C. Mah and L. Villier for their patient and helpful exchanges. C. Mah also helped in the collections and with preparations. J. Pearse provided a helpful review that greatly improved the manuscript. Support to Mooi was provided by California Academy of Sciences In-house Funds and a CNRS Associate Researchship. This paper is a contribution of the theme "Signal morphologique de l'Évolution" of the UMR CNRS 5561 "Biogéosciences."
|
FOOTNOTES |
|---|
1 From the Symposium Evolution of Starfishes: Morphology, Molecules, Development, and Paleobiology presented at the Annual Meeting of the Society for Integrative and Comparative Biology, 6–10 January 1999, at Denver, Colorado.
2 E-mail: rmooi{at}cas.calacademy.org
|
REFERENCES |
|---|
|
TOP
SYNOPSISINTRODUCTIONASTEROIDS AND THE...AXIAL ELEMENTS: AMBULACRALS AND...UNAMBIGUOUSLY EXTRAXIAL ELEMENTSPROBLEMATIC EXTRAXIAL ELEMENTSODONTOPHORESMOUTH ANGLE OSSICLESIMPLICATIONS FOR PHYLOGENY OF...REFERENCES |
|---|
Blake, D.B. 1978. The taxonomic position of the modern sea star Cistina Gray, 1840. Proc. Biol. Soc. Wash, 91:234-241.
Blake, D.B. 1987. A classification and phylogeny of post-Paleozoic sea stars (Asteroidea: Echinodermata). J. Nat. Hist, 21:481-528.[CrossRef]
Blake, D.B. 1989. Asteroidea: Functional morphology, classification and phylogeny. In M. Jangoux and J.M. Lawrence (eds.)Echinoderm studies 3, pp179-223Balkema, Rotterdam.
Blake, D.B. 1990. Paleobiological implications of some Upper Ordovician juvenile asteroids (Echinodermata). Lethaia, 23:347-357.
Blake, D.B. 1994. Re-evaluation of the Palasteriscidae Gregory, 1900, and the early phylogeny of the Asteroidea (Echinodermata). J. Paleont, 68:123-134.[Abstract]
Blake, D.B. 1998. Morphological characters of early asteroids and ophiuroids. In R. Mooi and M. Telford (eds.)Echinoderms: San Francisco, pp5-7Balkema, Rotterdam.
Blake, D.B., and T.E. Guensburg. 1988. The water vascular system and functional morphology of Paleozoic asteroids. Lethaia, 21:189-206.[Web of Science]
Blake, D.B., and T.E. Guensburg. 1993. New Lower and Middle Ordovician stelleroids (Echinodermata) and their bearing on the origins and early history of the stelleroid echinoderms. J. Paleont, 67:103-113.[Abstract]
Breimer, A. 1978. General morphology. Recent crinoids, In R.C. Moore and C. Teichert (eds.)Treatise on invertebrate paleontology, Part T, Echinodermata 2,. pp9-58TGeol. Soc. Amer. and Univ. of Kansas Press, Lawrence.
David, B., and R. Mooi. 1996. Embryology supports a new theory of skeletal homologies for the phylum Echinodermata. Comptes Rendus Acad. Sci. Paris, 319:577-584.
David, B., and R. Mooi. 1998. Major events in the evolution of echinoderms viewed by the light of embryology. In R. Mooi and M. Telford (eds.)Echinoderms: San Francisco, pp21-28Balkema, Rotterdam.
David, B., and R. Mooi. 1999. Comprendre les échinodermes: la contribution du modèle extraxial-axial. Bull. Soc. Géol. Fr, 170:in press.
Dean, J. 1998. What makes an ophiuroid? A morphological study of the problematic Ordovician stelleroid Stenaster and the palaeobiology of the earliest asteroids and ophiuroids. Zool. J. Linn. Soc, 126:225-250.[CrossRef]
Fell, H.B. 1963. The evolution of the echinoderms. Smithson. Inst. Ann. Rep, 1962:457-490.
Fewkes, J.W. 1887. On the development of the calcareous plates of Amphiura. Bull. Mus. Comp. Zool., Harvard, 13:107-150.
Fewkes, J.W. 1888. aOn the development of the calcareous plates of Asterias. Bull. Mus. Comp. Zool., Harvard, 17:1-56.
Fewkes, J.W. 1888. bOn the serial relationship of the ambulacral and adambulacral calcareous plates of the starfishes. Proc. Boston Soc. Nat. Hist, 24:96-117.
Gale, A.S. 1987. Phylogeny and classification of the Asteroidea (Echinodermata). Zool. J. Linn. Soc. Lond, 89:107-132.
Gemmill, J.F. 1912. The development of the starfish Solaster endeca Forbes. Trans. Zool. Soc. Lond, 20:1-71.
Gemmill, J.F. 1914. The development and certain points in the adult structure of the starfish Asterias rubens, L. Phil. Trans. Roy. Soc. Lond. B, 205:213-294.
Gemmill, J.F. 1920. The development of the starfish Crossaster papposus, Mueller and Troschel. Quart. J. Microscop. Sci, 64:155-190.
Gordon, I. 1929. Skeletal development in Arbacia, Echinarachnius and Leptasterias. Phil. Trans. Roy. Soc. Lond. B, 217:289-334.
Hendler, G. 1978. Development of Amphioplus abditus (Verrill) (Echinodermata: Ophiuroidea). II. Description and discussion of ophiuroid skeletal ontogeny and homologies. Biol. Bull, 154:79-95.
Hörstadius, S. 1939. Über die Entwicklung von Astropecten aranciacus L. Pubbl. Staz. Zool. Napoli, 17:222-312.
Hotchkiss, F.H.C. 1993. A new Devonian ophiuroid (Echinodermata: Oegophiurida) from New York State and its bearing on the origin of ophiuroid upper arm plates. Proc. Biol. Soc. Wash, 106:63-84.
Hyman, L. 1955. The invertebrates,. Vol4:Echinodermata, McGraw-Hill, New York.
Komatsu, M. 1975. On the development of the sea-star, Astropecten latespinosus Meissner. Biol. Bull, 148:49-59.
Komatsu, M., and S. Nojima. 1985. Development of the seastar, Astropecten gisselbrechti Döderlein. Pac. Sci, 39:274-282.
Lafay, B., A.B. Smith, and R. Christen. 1995. A combined morphological and molecular approach to the phylogeny of asteroids (Asteroidea: Echinodermata). Syst. Biol, 44:190-208.[CrossRef]
Lovén, S. 1874. Études sur les échinoïdées. Kongl. Svensk. Vetensk.-Akad. Handl, 11:1-91.
Ludwig, H. 1882. Zur Entwicklungsgeschichte des Ophiurenskelettes. Z. Wiss. Zool, 36:181-200.
McKnight, D.G. 1975. Classification of somasteroids and asteroids (Asterozoa: Echinodermata). J. Roy. Soc. N. Z, 5:13-19.
Mooi, R., and B. David. 1997. Skeletal homologies of echinoderms. Paleont. Soc. Papers, 3:305-335.
Mooi, R., and B. David. 1999. Evolution within a bizarre phylum: Homologies of the first echinoderms. Amer. Zool, 38:965-974.
Mooi, R., B. David, and D. Marchand. 1994. Echinoderm skeletal homologies: Classical morphology meets modern phylogenetics. In B. David, A. Guille, J.P. Féral, and M. Roux (eds.)Echinoderms through time (Echinoderms Dijon),. pp87-95Balkema, Rotterdam.
Murakami, S. 1937. On the development of the calcareous plates in an ophiurid larva, Ophiopluteus serratus. Annot. Zool. Japon, 16:135-147.
Paul, C.R.C., and A.B. Smith. 1984. The early radiation and phylogeny of echinoderms. Biol. Rev, 59:443-481.
Roux, M. 1997. Classification et ontogenèse chez les crinoïdes: une révision nécessaire de la hiérarchie des caractères. Bull. Soc. Zool. Fr, 122:371-378.
Siddall, S.E. 1979. Development of ossicles in juveniles of the sea star Echinaster sentus. Bull. Mar. Sci, 29:278-282.
Smith, A.B. 1984. Classification of the Echinodermata. Palæontology, 27:431-459.
Smith, A.B. 1988. Fossil evidence for the relationships of extant echinoderm classes and their times of divergence. In C.R.C. Paul and A.B. Smith (eds.)Echinoderm phylogeny and evolutionary biology, pp85-97Oxford Univ. Press, Oxford.
Smith, A.B., and P.A. Jell. 1990. Cambrian edrioasteroids from Australia and the origins of the starfishes. Mem. Queensland Mus, 28:715-778.
Spencer, W.K., and C.W. Wright. 1966. Asterozoans. In R. C. Moore (ed.)Treatise on invertebrate paleontology, Part U, Echinodermata 3,. pp4-107UGeol. Soc. Amer. and Univ. of Kansas Press, Lawrence.
Verrill, A.E. 1914. Monograph of the shallow-water starfishes of the North Pacific coast from the Arctic Ocean to California. Smithson. Inst. Harriman Alaska Ser, 14:1-408.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||