Bijdragen tot de Dierkunde, 63 (3) 129-148 (¡993)

SPB Academie Publishing bv, The Hague

The taxonomic significance of the short-shafted mesotriaene reviewed by

parsimony analysis: validation of Pachastrella ovisternata Von Lendenfeld

(Demospongiae: Astrophorida)

Manuel Maldonado

Centre d'Estudis Avançais de Blanes, Camide Santa Barbara s/n, 17300-Blanes, Girona, Spain

Keywords: Astrophorida, Pachastrellidae, parsimony analysis, ,

Abstract P. ovisternata válida nuevo ejemplar prueba que es una especie

rabdoma corto y corrobora la existencia de mesotrienas de en es-

pecies del género Pachastrella. The spicule complement of a specimen collected Los análisis de parsimonia muestran que la especie Yodomia from the Atlantic coast of Spain was noted to be similar to that tradicionalmente al Yodomia la perfecta, asignada género por of Pachastrella monilifera, but particularly characterized by the

presencia de mesotrienas, debería ser reclasificada en el género presence of short-shafted mesotriaenes. After detailed examina- Pachastrella , cerca de P. ovisternata. Los análisis indican tam- tion, it was concluded that the specimen undoubtedly belonged

bién que las mesotrienas que aparecen en Calthropellidae y to the species Pachastrella ovisternata Von Lendenfeld, 1894.

Pachastrellidae son el resultado deuna evolución convergente en For a long time, this species has been regarded a synonym of la forma de las espículas tetraxonas de estos dos grupos. El estu- Pachastrella moniliferaSchmidt, 1868. Examination of this new

dio a nivel específico indica que las mesotrienas han surgido más specimen showed that Pachastrella ovisternata was a valid spe- de una vez en el seno de la familia Pachastrellidae. Por tanto, Pachastrella. cies of the genus Moreover, it denoted the real exis-

la presencia de mesotrienas no puede ser considerada, en sí mis- tence of mesotriaenes in this genus.

ma, como un carácter diagnóstico a nivel de género. Sin embar- The parsimony analyses indicated that the species Yodomia

go, se ha mostrado como un carácter muy fiable para distinguir perfecta, traditionally assigned to the genus Yodomia by the

taxones a nivel específico. presence of mesotriaenes, ought to be reclassified in the genus

Pachastrella, close toP. ovisternata. These analyses also showed that mesotriaenes are a product of a convergent evolution in Introduction shape of tetraxon spicules in Calthropellidaeand Pachastrelli-

dae. At the species level it is suggested that short-shafted mesotriaenes have arisen more than once in the family Pacha- with short-shafted mesotriaenes

strellidae. Thus, the presence of mesotriaenes itself cannot be (mesocalthrops or mesodichotriaenes) are extreme- regarded as a diagnosticcharacteristic at generic level. However, ly rare. According to the literature, these spicules it was noted to be a reliable feature in order to distinguishtaxa are present in species belonging to the family at specific level.

Calthropellidae, such as Pachastrissa pathologica

(Schmidt, 1868) (sensu Von Lendenfeld, 1903;

Resumen Levi, 1973) and Pachataxa enigmatica Lévi & Lévi,

1983. They also occur in some pachastrellids, such

as the species of the Yodomia Lebwohl, La espiculación de un ejemplar de demosponjaprocedentede las genera

costas 1914 and atlánticas españolas se apreció muy similar a la de Pacha- (sensu Dendy, 1916) Triptolemus Solías, strella monilifera, caracterizada la presencia de pero por 1888. Additionally, a peculiar case of short-shafted mesotrienas de rabdoma corto. Tras un estudio detallado se ha mesotriaenes becoming desmas has been reported concluido que dicho ejemplar pertenece a la especiePachastrella Lévi Lévi in the by & (1989) enigmatic genus ovisternata VonLendenfeld, 1894. Esta especie, sin embargo, ha Brachiaster 1925. sido Wilson, considerada, desde hace mucho tiempo, un sinónimo de

Pachastrella monilifera Schmidt, 1868. La descripción de este Many of these descriptions of species with

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 130 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

mesotriaenes are very ancient. Furthermore, most (MNCN: 36, according to the old code), Sphinctrella gracilis

Solías, 1888 (BMNH: 94:11:16:144 to 146), Sphinctrellahorrida species have not been recorded since their initial Schmidt, 1870 (BMNH: 70:5:3:47), Sphinctrellacribrifera Sol- in In such the description, except a few cases. cases, las, 1888 (BMNH: 89:1:1:39), and Calthropellasimplex Solías, short lists of records indicate that these species are 1888 (BMNH: ?, schizotype stored as substratum of S. cri-

limitedto a small geographical area. Conse- usually brifera). In addition, some specimens collected in the Alboran

of these quently, most species are presently regard- Sea belonging to the species cited below, were also examined:

Pachastrella monilifera Schmidt, 1868, Poecillastra compressa ed as taxonomieoddities. The genus Yodomia, for (Bowerbank, 1866), Poecillastra amygdaloides (Carter, 1876), instance, has virtually been forgotten by present- Spinctrella verrucolosa Pulitzer-Finali, 1983, Stoeba plicatus it invalidated. day taxonomists, although was never (Schmidt, 1868), Dercitus bucklandi (Bowerbank, 1861) and Consequently, the existence of diverse types of Pachastrissa pathologica (Schmidt, 1868). Other data required

has unnoticed in mesotriaenes unfortunately gone were obtained from descriptions in scientific literature.

The skeletal study was accomplished using the standard interesting discussions on the evolution and rela-

methodology(Riitzler, 1978). Micrographs were obtained using tionships of the tetraxon spicules (e.g., Reid, 1970;

a HITACHI-S-2300 SEM. Levi, 1991). The taxonomie history of the genus The parsimony analyses were performed using "Paup 3.0 g"

which was reclassified from Brachiaster, recently (Swofford, 1989),according to the assumptionsexplained in the

Lithistida "incertae sedis" (Wilson, 1925) to As- text. trophorida, Pachastrellidae(Lévi & Lévi, 1989) and subsequently relocated in Lithistida Tetracladina Systematic description (Lévi, 1991), is also indicative of the systematic in- stability affecting most of the species concerned in

Genus Pachastrella 1868 this study. Schmidt,

The collection of a new pachastrellid specimen with Pachastrella ovisternata Von Lendenfeld, 1894 mesotriaenes prompted the present investiga- Pis. tion on the relationships between taxa that have (Fig. 1; I-III) short-shafted mesotriaenes, using parsimony anal- ysis. Synonymy. - Pachastrella ovisternata Von Lendenfeld, 1894:

439.

Pachastrella monilifera(Schmidt); Topsent, 1902: 14; 1904: 93,

Materials and methods 94 (in part).

Non: Pachastrella ovisternata sensu Ferrer-Hernández, 1914: 7

( =P. monilifera). The specimen described here was collected duringan expedition carried out in 1982 by the Spanish Institute of Oceanography.

The location data were: muddy bottom at 300 m deep,near Cape Description. - Massive specimen, irregular in

Finisterre (northeasternAtlantic coast ofSpain); latitude/longi- and in size The shape 20 x 20 x 14 cm (Pl. I). sur- tude are unknown. The specimen was preserved dry and deposit- face is uneven, wrinkled in some areas and usually ed in the Museo Nacional de Ciencias Nacionales in Madrid covered with slime. Ostia are circumscribed to a (MNCN-01.01/6). The holotype of Pachastrella ovisternata, large concave area resulting from a convoluted borrowed from the Naturhistorisches Museum Wien (NMW- conical in Inv.Nr. 525 and 526) was checked against the collected specimen fold. Non-aquiferous papillae, 5-8 mm and material described by Ferrer-Hernández under the (1914) height, along with hispidating tracts, protruding name of Pachastrella ovisternata (MNCN-202 & 203, according 3-4 mm, present on the inhalant surface. Oscules to the old code). are 1 -2 mm in diameter, scattered on theremaining Type material and abundant other material stored in the Brit- surface. Colour white. ish Museum of Natural History in London (BMNH), the Museo greyish Stony consistency

Civico di Storia Naturale "G. Doria" in Genoa (MSNG) and when dry.

MNCN were examined in order to adequately record charac- teristics at generic level. Type material examined was asfollows: Spicules: Characella pachastrelloides (Carter, 1876) (BMNH: 10:1:1680 Oxeas: curved or slightly flexuous, blunt-pointed and 1681), Characella tripodaria (Schmidt, 1868) (BMNH:

la, 3600-7500 m X 16-45 in size. 68:3:2:36), Poecillastra rudiastra Pulitzer-Finali, 1983 (MSNG: (Figs, b),

C:E:47161), Sphinctrella linaresi Ferrer-Hernández, 1914 Calthrops and pseudocalthrops: clads are usually

Downloaded from Brill.com10/11/2021 08:28:45AM via free access Bijdragen tot de Dierkunde, 63 (3) - 1993 131

1. of Fig. Spicules Pachastrella ovisternata: (a) oxeas, (b) blunt end of oxea, (c) calthrops, (d) dichotriaene, (e)mesotriaene, (f) metaster,

(g) amphiasters, (h) plesiasters, (i) oval microstrongyles, (j) microrhabdoid streptaster.

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 132/I M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

Plate I. Overall view of the specimen.

straight in the smaller forms, but malformations Oval microstrongyles: usually elliptical in shape, are occasionally observed in the greater forms (Fig. fully covered with minute spines (Fig. li, PI. Hid)

size. lc, PL lib). Clads display a wide range in size, and 12-19 /¿m x 5-9 in

Microrhabdose forms measuring 122-1100 ¡¿m x 8-55 /¿m. streptasters: thin, very long

Dichotriaenes: with 1-2 short-shafted, occasionally one (25-45/¿m x fim), exhibiting strong spines or or two undivided clads only (Figs. Id, e, Pl. lia). microspiny turbercles (PI. IIIc; Fig. lj). They are

The rhabdome measures 65-85 ¿im X 8-12 /¿m, scarce. The thick central axis along with the groups the minute actines protoclads 25-40/nm x 8-12 /im, and the deu- of degenerate (resembling spines) teroclads 46-95 The is made these resemble micro- x 7-10 /im. rhabdome spicules polytylote usually shorter than the clads. strongyles when examined under the light micro-

Mesodichotriaenes: similar to short-shafted dicho- scope. triaenes in which shaft the (rhabdome sensu lato) is Similar spicules have been reported in some spe-

both sides of the cladome of prolonged on forming a cies Triptolemus Solías, 1888, Pachastrella, and rhabdome sensu stricto and an epirhabdome. Yodomia. They were initially described as mi-

Epirhabdome usually shorter than rhabdome sensu croxeas or derived spicules (e.g., Carter, 1876; Sol- stricto (Fig. le; Pl. Ile, d). Clads are always di- las, 1888; Kirkpatrick, 1903), sometimes under the

and chotomous no malformations were observed. ambiguous name of "microrhabds" (e.g., Kirk-

Dimensions are similar to those of the dicho- patrick, 1902; Pulitzer-Finali, 1971). However, I triaenes. with authors Epirhabdome measures 40-85 /im x agree such as Dendy (1916: 234),

7-12 )j.m. Sarà (1959) and Lévi & Lévi (1983: 153), who have

Downloaded from Brill.com10/11/2021 08:28:45AM via free access Bijdragen tot de Dierkunde, 63 (3) - 1993 133/II

Plate II. Megascleres: (a) dichotriaene, (b) calthrops and dichomesotriaene, (c) & (d) different views of dichomesotriaenes.

the thick interpreted these spicules as reduced streptasters. consistency to . A crust (1-1.5 ¿¿m)

Amphiasters: with short shaft and microspiny ac- of microstrongyles reinforces the ectosome. The tines. Microspines are only evident under SEM ex- choanosomal skeleton displays a diffuse architec- amination (Fig. lg, PI. Ilia). The spicules are ture. Tracts of oxeas show a vague radial pattern

from the inner choanosome to the 17-22 /xm in their overall length, with actines mea- up ectosome,

6—10 in abundant. in tufts at the inhalant suring /xm length. They are very protruding long hispidating

Amphiasters are very rarely transformed into zones. Calthrops are abundant everywhere in the

but sanidasters or metasters (Fig. If). choanosome, no special spatial organization is

Plesiasters: with usually 3-5 actines covered by apparent. microspines (Fig. lh; PL Illb). Actines are 10-15 Dichotriaenes and mesodichotriaenes are more

abundant in the inner choanosome, in /xm in length. They are very scarce. contrast to

the general pattern of the Astrophorida (Solías,

Skeleton: 1888). Streptasters are densely scattered everywhere

Spicules are densely arranged, providing a hard in the choanosome.

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 134/III M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

Plate III. Microscleres: (a) amphiaster, (b) plesiaster, (c) reduced streptaster, (d) oval microstrongyle.

Distribution. - Known only from the coasts of however, this interesting species has been disguised

and for time due several Portugal betweenFaro Sines (Von Lendenfeld, a long to consecutive systemat-

1894) and the coasts of Galicia (present record), at ic interpretations: (1) Von Lendenfeld (1903) ex- a depth of 300 m. Specimens with mesodichotri- plicitly defined the genus Pachastrella as Pacha- aenes mentionedby Topsent(1902) under the name strellidae without mesotriaenes (p. 73), although of Pachastrella abyssi might actually correspond to he paradoxically included in the same paper(p. 75)

Pachastrellaovisternata, but data about their loca- a description of P. ovisternata, underlining the tion were omitted by the author. presence of characteristic mesodichotriaenes. (2)

Later, P. ovisternata was claimed a junior synonym

of Pachastrella monilifera by Topsent (1902: 14; Taxonomic discussion 1904: 93, 94). The systematic value of the mesotri-

The level considered low collected specimen completely fits the holo- aenes at specific was by Top- type of Pachastrella ovisternata. Unfortunately, sent, who assumed that these spicules occasionally

Downloaded from Brill.com10/11/2021 08:28:45AM via free access - Bijdragen tot de Dierkunde, 63 (3) 1993 135

appeared in Pachastrella from sporadic modifica- + P. abyssi) as proposed by Topsent (1902,

tions in dichotriaenes which originated, in their 1904). Consequently, Pachastrella ovisternata Von turn, from occasional modifications in calthrops Lendenfeld, 1894 is herein claimed as a valid spe-

different from (Topsent, 1902, 1904). (3) The disregard of the sys- cies, clearly P. monilifera. tematic significance of the mesotriaenes which It is also worth noting that P. ovisternata resem- emerged from Topsent's interpretations brought bles the Indian species Yodomia perfecta Dendy,

Ferrer-Hérnandez (1914) to incorrectly record the 1916, which has a spicule complement of long

mi- name of P. ovisternata for specimens lacking oxeas, mesotriaenes, calthrops, amphiasters,

characteristic crorhabdoid mesotriaenes, dichotriaenesand long streptasters ("reduced streptasters" oxeas. After re-examination, I came to the con- according to Dendy, 1916), and oval micro- clusion that the material described by Ferrer- strongyles. Y. perfecta is only differentiated from

Hernández (1914) actually belongs to the species P. ovisternata in minor detailsof its skeleton, such

Pachastrella monilifera. as the absence of dichotriaenes and the occurence

The illegitimacy of P. ovisternata has been sup- of tetracladose mesotriaenes in addition to the ported for a long time mainly because of a vague dichotomous ones.

Yodomia 1914, erected for report on specimens bearing dichotriaenes and The genus Lebwohl, mesodichotriaenes. These specimens were at first Yodomia ijimai, was modified by Dendy (1916) in regarded by Topsent (1902: 14, 17) as "peculiar" order to include a second species, Yodomiaperfec-

since materialbelonging to P. abyssi Schmidt, 1870, but, ta. Both species appear to be very rare, they

himself claimed have been recorded of Leb- two years later, Topsent (1904) only once (Sea Japan: synonymy of P. abyssi and P. monilifera. Meso- wohl, 1914) and twice (Indian Ocean -Saya de

dichotriaenes were envisaged by Topsent (1902, Malha-: Dendy, 1916; Burton, 1959), respectively.

of such the 1904) as spicules of sporadic occurrence, i.e. occa- Some skeletal features Y. ijimai, as sional results of haphazard modifications in cal- presence of long-shafted triaenes (rhabdome mea- throps and thus of no systematic value. suring up to 2200 /¿m and undeveloped clads),

undivided absence In contrast, most evidence indicates that meso- mesotriaenes with clads, of dichotriaenes of P. ovisternata are a well estab- microrhabdoid streptasters, and presence of curved lished category of spicules: (1) They are always per- microstrongyles with occasional oxeote ends (pre- fect in shape (PI. lie, d), while the occurrence of sumably arising from microxeas and non-homolo-

that some malformationswouldbe expected under Top- gous with the oval microstrongyles), suggest sent's assumption. (2) In addition, a great variabili- this species is not closely related to Y. perfecta. In ty in size occurs in the calthrops whereas the size is contrast, Y. perfecta and P. ovisternata are noted

in of tradi- especially uniform in the mesotriaenes. (3) The ar- to be very closely related, spite having

of mesotriaenes in the is differ- different rangement sponge tionally belonged to genera. ent to that of the calthrops. Mesodichotriaenes are In a discussion on therelationships concerning P.

Brachiaster particularly confined to the inner choanosome, ovisternata, the monospecific genus

characterized the of whereas calthrops indiscriminately occur every- Wilson, 1925, by presence where in the sponge. (4) Finally, there are no other mesotriaenes transforming to desmas, can not be reports on specimens with mesotriaenes eitherin P. obviated. This genus was formerly regarded as a

lithistid "incertae sedis" monilifera or in P. abyssi (cf. Schmidt, 1870, 1880; (Wilson, 1925), later

Carter, 1876; Solías, 1888; Van Soest & Stentoft, redescribed in Pachastrellidae (Lévi & Lévi, 1989)

1988). Calthrops and mesodichotriaene mega- and recently reclassified into Lithistida Tetracladi-

in scleres are further compared in PI. lib. na (Levi, 1991). The presence Brachiaster of

Therefore, the skeletal differences noted, mainly some skeletal characters also present in typical the presence of especially long oxeas and meso- pachastrellids, such as diversely branched meso- dichotriaenes, isolate and identify P. ovisternata triaenes (although becoming desmas), oval micro- from the of moni- and careful considera- remaining complex species lifera(P. strongyles metasters, require

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 136 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

into solve the tion. In my opinion, and taking account the Poecillastra-Sphinctrella), to relation-

Pachastrella fact that species such as Yodomia perfecta and ships among Brachiaster simplex,

Pachastrella ovisternata were overlooked by Lévi monilifera, P. echinorhabdaPulitzer-Finali, 1973,

(1991) in his discussion about the relationships of P. ovisternata, and Yodomia perfecta. This multi-

Brachiaster, it would be advisable to reconsider the ple outgroupconsisted of all four species belonging

of this in the to with Stoeba at possible systematic relocation genus Triptolemus along spp. (defined

Pachastrellidae. generic level). Such a heterogeneous set is allowed,

since "the outgroup comparisons need not be con-

strained Linnean hier- by nomenclatural rank or

Phylogenetic analysis archical structure" (Watrous & Wheeler, 1981).

The parsimony analysis was seen to be the best way Taxonomic background to tackle the obscure relationships between the

The carried in above-mentioned species. study was All taxa included the analyses as well as some dis- out in three phases. The first phase analyzed the carded taxa are briefly diagnosed and discussed relationships among the pachastrellid genera, in- below. cluding Brachiaster. The second phase explored the possible generic placement of the problematic spe- Ingroup genera:

such Yodomia Yodomia Poecillastra 1888: cies, as ijimai, perfecta, Solías, Pachastrellidae having and Pachastrella ovisternata. The third and phase ex- oxeas, calthrops, or pseudocalthrops as amined the relationships between several species megascleres. Microscleres consist of microxeas in a traditionally assigned to three different genera single category and several types of streptasters

(Pachastrella, Brachiaster, and Yodomia), based with long and thin actines. Tetraxon spicules occur on the results of the preliminary parsimony ana- everywhere in the choanosome. lyses. Characella Solías, 1888: Pachastrellidae with a

spicule complement similar to Poecillastra, but

where Outgroup selection tetraxons are restricted to subectosomal loca-

utilized in all tions and of microxeas A multiple outgroup was analyses, two categories always occur. since it is the most stringent test of the ingroup Some authors (e.g., Pulitzer-Finali, 1983; Van monophyly (Bergquist & Kelly-Borges, 1991). Soest & Stentoft, 1988) doubtthe validity of this ge-

According to the traditional family diagnoses nus, but no complete study has been made in this

(e.g., Lévi, 1973), members of Theneidae could matter. I have examined holotypes and material be- be thought to be the most appropriate outgroup longing to Characella pachastrelloides and Ch. for tackling the relationships among pachastrel- tripodaria, noting that the location of tetraxons lid genera. However, calthropellid and stellet- was always restricted to peripheral choanosome. tid genera were used to elaboratethe outgroup. The Sphinctrella Schmidt, 1870: Pachastrellidae whose occurrence of short-shafted mesotriaenes along oxeas are divided into two categories according to with calthrops in some calthropellids, as well as the their choanosomal or ectosomal location. Cal-

of in stellettids restricted the occurrence streptasters some (e.g., throps or pseudocalthrops are to genus Stryphnus), were considered very useful fea- peripheral choanosome. Microscleres consist of tures in ascertaining the homoplasous or homolo- one or several kinds of streptasters and one or two gous nature of such characteristics in Astropho- categories of microxeas. Oscules are grouped in ex- rida. halant areas surrounded by palisades of ectosomal

The same reasoning was applied in the analysis oxeas. at specific level and, thus, the subgroup consist- Stoeba Solías, 1888 (sensu Dendy, 1905): Pacha- ing of Triptolemus-Stoeba was chosen as out- strellidae with calthrops or pseudocalthrops par-

the of group (rather than subgroup consisting tially or completely replaced by short-shafted

Downloaded from Brill.com10/11/2021 08:28:45AM via free access - Bijdragen tot deDierkunde, 63 (3) 1993 137

dichotriaenes (as, for instance, in Stoeba extensa feld and Lévi, typical species remaining in this

natalensis Oxeas in the Atlantic-Mediterranean are Dendy, 1905 or St. Burton, 1926). genus region

are absent. Microscleres are exclusively a single Calthropella stelligera (Schmidt, 1868), C. recondi-

type of sanidaster with degenerate actines, located ta Pulitzer-Finali, 1973, and C. simplex.

habit 1936: in an ectosomal layer. Endolithic or parasitic Pachataxa De Laubenfels, Calthropellidae

of this with euasters and characteristic characterizes species genus. calthrops, regular

Triptolemus Solías, 1888: Parasitic Pachastrellidae aberrant euasters displaying microrhabdoid shape

whose tetraxons are exclusively short-shafted (sometimes named ataxasters). Oxeas are absent.

Short-shafted mesotriaenes in mesotriaenes with clads diversely branched. Oxeas occur some species

be absent in all such as Pachataxa Lévi & may species (according to Sarà, enigmatica Lévi, 1983,

1959), since those formerly described in Triptole- but they are absent in P. lithistina(Schmidt, 1880)

mus intextus (Carter, 1876) ( = T. parasiticus Car- and P. lutea Pulitzer-Finali, 1986.

cladosus 1888: Stellettidae whose ter, 1876, sensu Levi & Lévi, 1983) and T. Stryphnus Solías, mega-

Solías, 1888 presumably belonged to the host scleres are long-shafted triaenes and oxeas. Micro-

well sponges. Microscleres are streptasters, including scleres consist of euasters as as streptasters.

degenerate forms in species such as T. cladosus,

(according to Lévi & Levi, 1983: 153), T. simplex Discarded genera:

Sarà, 1959, and T. intextus. Microxeas are present Indeed, it would be interesting to include in the

in the species T. cladosus and T. incertus Kirk- analysis of the pachastrellid relationships some

patrick, 1903. controversial genera (mentioned below), but this

Pachastrella Schmidt, 1868: Pachastrellidae whose task would take too long and would complicate this

discussed be- megascleres are oxeas and calthrops or pseu- study enormously. These genera are

of the docalthrops. Microscleres consist of oval micro- low to provide a better understanding

strongyles and several kinds of streptasters in- problematic taxonomie background.

cluding degenerate forms. The presence of meso- Dercitus Gray, 1867: This genus with toxas and

different triaenes in some species, presumably belonging to massive habit, is, in my personal opinion,

The of the microrhabdoid this genus, is examined further below. from Stoeba. shape

Brachiaster Wilson, 1925: Monospecific genus microscleres is the main support for the traditional

whose megascleres are styloxeas, short-shafted assumed identity between Dercitus and Stoeba.

mesotriaenes with clads diversely branchedand des- However, the microrhabdoid shape may be de-

mas with mesotriaene crepis (named mesotriders, ceptive in establishing relationships, since it is a

according to Reid, 1970). Short-shafted dichocal- morphological convergence which may be seen

throps, transforming to desmas, may be present or in microxeas (e.g., Characella pachastrelloides),

Lévi Micro- Pachastrel- absent (Wilson, 1925; & Levi, 1989). streptasters (e.g., Triptolemus simplex,

and and in in scleres are oval microstrongyles metasters. la echinorhabda) even euasters (e.g., spe-

cies of Pachataxa: Topsent, 1897; De Laubenfels,

of is Outgroup genera: 1936). As far as thepresence toxas concerned,

Pachastrissa Von Lendenfeld, 1903: Calthropelli- it must be concededthatthis is an enigmatic feature

euasters whose with the microsclere dae with oxeas, calthrops, and regular (ac- relationship remaining

cording to Von Lendenfeld, 1903; Levi, 1973). types in Astrophorida is unclear; it may be impor-

Short-shafted mesotriaenes with unbranched clads tant from an evolutionary point of view, but it is

uninformative in alter- are present in some species, such as Pachastrissa a parsimony analysis. The pathologica (Schmidt, 1868) and P. inopinata native taxonomie opinion, viz. both genera should

(Pulitzer-Finali, 1983). be separated, has been postulated previously by

Calthropella Solías, 1888: Calthropellidae with authors such as Dendy (1905), Burton (1926), and

and absent. Vacelet whereas calthrops regular euasters. Oxeas are & Vasseur (1971), most contem-

After the application of the criteria of Von Lenden- porary authors do not make such a generic distinc-

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 138 M. Maldonado - Taxonomicsignificance of the mesotriaene in Astrophorida

Table I. General data matrix. Lowercase letters indicate character states, positive symbols (+) indicate taxa included in each analysis, negative symbols (-) correspond to taxa and/or characters not used in the analyses. Characters 1-11 were used in the analyses at the generic level and characters 2, 4, 5, 9, 10, 12-16 in the analyses at the specific level. For explanation of the characters, see text.

OTUS CHARACTERS TREES

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Tl T2 T3 T4

STRYPHNUS abbaac bbaaa - — — — — + +

PACHASTRISSA a b a b aba a b a a d - — — — — + +

PACHA TAXA a a a ababa a c a a d - — —

CALTHROPELLA aaabaaabaad — —

POECILLASTRA ababaabbabac - — — — —

CHARACELLA a b b b a abb a be a c - — —

SPINCTRELLA - — - — a c b b a a b a be a c

STOEBA baabaacaaaacaabc + + + +

TRIPTOLEMUS b aba a b a bda aba ac - — — — — + + + —

PACHASTRELLA - 1 ababaadaabb

PACHASTRELLA - 2 a b a b aba d a a b b

BRACHIASTER simplex abaabc abaabbac baa + + + +

YODOMIA ijimai ababbebdeae

Yodomia perfecta ababbadaabbabc bba - + — +

Pachastrella ovisternata ababbadaabbabbba - + - +

Pachastrella monilifera -b - b a - -- a b - a a b b a — — - +

Pachastrella echinorhabda — b - ba — - - a b — babe a — — — +

Triptolemussimplex -a — a b - — — a a — c c c b b --- +

- — Triptolemus intextus ? a b — — — a a -ccbbb--- +

Triptoleus incertus — a - ab - -- ba - ccbaa - — — +

Triptolemus cladosus - ? — a b - — - b a - c c bba — — — +

tion. Both opinions are debatableand it is obvious cies Acanthotriaenacripta Vacelet, Vasseur & Lévi, that this from 1976. The of systematic matter is far being com- spicule complement consists acan-

solved. In the elimination of this pletely any case, thodichotriaenes, spirasters, rhaphides and per- genusfrom thepresent analysis was thought advisa- haps, small oxeas. In spite of being formerly ble so as to avoid further complications, as it hardly described in the Pachastrellidae, the relationships affects either the aims or the conclusions of this betweenthis genusand theremaining pachastrellids study. analyzed here are very obscure.

1968: Artificial erected to Lamellomorpha Bergquist, Monospecific ge- NetheaSolías, 1888: genus con- nus erected for Lamellomorpha strongylata Berg- tain species having calthrops with underdeveloped

quist, 1968 and located by the original author in the actines. It was rejected in Lévi's classification dustbin order Epipolasida. The megasclere comple- (1973) and questioned by other authors (e.g., ment lacks tetraxon and consists exclusive- spicules Pulitzer-Finali, 1983). In my opinion, this genus ly of flexuous strongyles. Microscleres are two ought to be formally abandoned and its species kinds of degenerate (microrhabdoid) streptasters reclassified in other pachastrellid genera: Nethea

with along amphiaster-metaster forms having well- nana (Carter, 1880) and N. amygdaloides (Carter,

actines. Such microscleres developed could cor- 1876) clearly belong to the genus Poecillastra; respond to a very modified pachastrellid, but for Nethea dissimilis Sarà, 1959 probably belongs to the present the genus is excluded from the analysis the genus Stoeba. due to its uncertain status.

Acanthotriaena Vacelet, Vasseur & Lévi, 1976: Character analysis

erected for Monospecific genus the enigmatic spe- A total of sixteen characters were used in the dif-

Downloaded from Brill.com10/11/2021 08:28:45AM via free access Bijdragen tot de Dierkunde, 63 (3) - 1993 139

the found the inter- ferent analyses (Table I). Characters 1-11 were possible, according to states at utilized for the analyses at generic level, whereas nal nodes in the generic cladogram. Assumed an-

12-16 in the included in the searches. characters 2, 4, 5, 9, 10, were utilized cestors were never analyses at specific level. Most of these characters

Characters concern skeletal features, since other characters, were as follows:

tactile surface Habit: Endolithic habits char- such as color, consistence, and so on, 1. and parasitic (lb) have unknown evolutionary importance. All acterize the generaStoeba and Triptolemus, where-

in the lack this trend in habit characters were unweighted order to avoid sub- as remaining taxa (la).

Uninformative characters 2. Oxeas: in in jective evaluations. were Present a single category most

when charac- Oxeas divided into rejected, or reinterpreted possible, as genera (2b). are two categories ter states. A question mark indicates a doubtful (ectosomal and choanosomal) in Sphinctrella (2c).

in Taxonomie in Stoeba and the character state an OTU (Operational Oxeas are absent (2a) outgroup

Unit). genera Pachataxa and Calthropella (sensu Von

Multiple character states were utilized, when Lendenfeld, 1903; Lévi, 1973). They are also absent

to introduce into the the varia- in incertus and T. but its necessary, analysis Triptolemus simplex,

inherent in taxa in this is doubtfulin cladosus and T. intextus bility supraspecific (genera, presence T. case). Multiple states were first treated under the (e.g., Solías, 1888; Sarà, 1959). The presence of a

hypothesis of "uncertainty" and then as "poly- single category of oxeas was assumed to be the an-

morphisms" ( = variability, not Dollo characters). cestral conditionin the analysis at the specific level.

The states assigned to the terminal taxa in the clado- 3. Tetraxon location: Tetraxon spicules may occur

gram, obtained under the hypothesis of uncer- either dispersed in the choanosome (3a) or restrict- tainty, are expected to be states of diagnostic value ed to choanosomal peripheral zones (3b), as is not-

at generic level since they minimize the tree length ed in species of Characella, Sphinctrella, and

(monomorphic ancestral state). Conversely, those Stryphnus. states relegated to the condition of possible assign- 4. Calthrops and calthropoid spicules: Calthrops ments (less parsimonious) will lack diagnostic value or calthropoid spicules (including short-shafted

at generic level, since they would correspond to dichotriaenes, e.g., in Characella or Stoeba) are apomorphies exclusively arising in only a few spe- present in most taxa (4b). They are always absent in

cies of the genus. Triptolemus (4a). However, their presence is sub-

The use of multiple states facilitates the addition ject to intraspecific variation in Brachiaster simplex of new or controversial species (in this case Yodo- (cf. Wilson, 1925; Levi & Levi, 1989). They display mia ijimai, Yodomia perfecta and Pachastrella interspecific variation in Pachataxa, i.e., they are , ovisternata) to a supraspecific system, even without absent in Pachataxa enigmatica and present in adding new characters. This procedure offers an ex- P. lithistina and P. lutea. The state "4b" was ploratory approach to the generic relationships of assumed to be the ancestral condition for this these species. Moreover, the addition of new taxa character in the analysis at the specific level.

constitutes an intuitive method to check the robust- 5. Short-shafted mesotriaenes: Mesotriaenes are ness of the previously inferred classification (Sokal present (5b) in all species of Triptolemus, Yodomia et al., 1992). ijimai, Yodomiaperfecta, Pachastrellaovisternata,

in No character was polarized the study at the and some species of Pachastrissa, such as P. generic level, since the relationships between the pathologica (Pl. IV a-d) and P. inopinata. outgroup and the ingroup were assumed "a priori" Mesotriaenes, transforming to mesotrider desmas,

be eitherunsolved known. It would be in Mesotriaenes to or poorly occur Brachiaster (5c). are lacking

difficult actually to logically support any given in the remaining taxa (5a). The state "5a" was as- polarity. At the specific level in the study, the ances- sumed to be the ancestral conditionin the analysis tral conditionof the characters was assumed, when at the specific level.

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 140/IV M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

Plate IV. Tetraxon spicules of Pachastrissa pathologica:(a) mesotriaene with underdevelopedactine (mesocalthrops), (b) calthrops and mesotriaene with well-developed actines, (c) well-developed mesotriaene (mesocalthrops), (d) well-developed mesocaltrops and under-

developed form (left lower corner).

Downloaded from Brill.com10/11/2021 08:28:45AM via free access Bijdragen tot de Dierkunde, 63 (3) - 1993 141

6. Long-shafted triaenes: According to the tradi- strongyles in the literature, ought to be terminologi-

tionally accepted diagnosis of the family Pacha- cally differentiated from both above-mentioned

of strellidae (Lévi, 1973), these spicule types ought to types. The shape the microrhabd is clearly a con-

be absent (6a). However, anatriaenes have been vergent morphology for several lines of micro-

described in some pachastrellid species (6b), such scleres (Wiedenmayer, 1977) and, thus, the use of

as Characella tripodaria ( =Sphinctrella linaresi), the ambiguous term "microrhabd" ought to be

Poecillastra armata Hanitsch, 1895 and Poecillas- avoided.

tra rudiastra. The presence of long-shafted ortho- The results of the generic cladograms indicated

plagiotriaenes (6c) characterizes the genusStryphus that presence/absence of oval microstrongyles is a

and they also occur in Yodomia ijimai. very consistent character (Consistency Index = 1,

7. Absent in all Retention Index = take Streptasters: calthropellid genera 1). This fact allows us to

(7a). They are regular forms with long, thin actines, for granted the presence of microstrongyles as an ir-

and occur in Poecillastra, Characella, Sphinctrella, reversible apomorphy for the ingroup (analysis at

Brachiaster, some species of Triptolemus, and the specific level).

felt: reinforced for outgroup Stryphnus (7b). They are exclusively 11. Ectosomal The ectosome is

forms with degenerate (reduced) actines (7c) in differentkinds of microscleres. Streptasters are the

Stoeba. Both morphologies simultaneously occur principal spicules in the felt of Stoeba, Triptole-

in Pachastrella and some species of Triptolemus mus intextus, T. simplex, and the outgroup genus

Oval make the (7d). Stryphnus (11a). microstrongyles up

8. Euasters: These spicules are absent inall ingroup ectosomal layer in Pachastrella, Brachiaster, Y.

taxa (8a). Euasters with a regular shape occur in all perfecta, and P. ovisternata (lib). Microxeas (or

in where outgroup genera (8b), except Pachataxa, derived forms) are the principal spicules in the felt

are with forms of Characella, Poecillastra, some they present together degenerate Sphinctrella, spe-

(sometimes named ataxasters) (8c). Degenerated cies of Triptolemus (T. cladosus, T. incertus) and

euasters (under the name of acanthosphaera) occur Yodomia ijimai (11c). Euasters reinforce the ecto-

in Yodomia of all three of the ijimai (8d). some calthropellid genera out-

9. Microxeas: Absent in most genera (9a). They oc- group (lid).

cur as a single category (9b) in Poecillastra, Yodo- 12. Habit (at specific level): This character is a re-

mia ijimai, Triptolemus incertus and T. cladosus. make of the character "1", but used exclusively at

Most species of Characella and Sphinctrella display the specific level. The endolithic-parasitic state

two categories of microxeas (9c), although some- (12c) has been retained as a characteristic feature of

times is found all of only one category because of the in- species Stoeba and Triptolemus. However,

traspecific and/or interspecific variability. Absence the non-endolithic-parasitic habit was split into

of microxeas was assumed to be the ancestral condi- massive (12a) and encrusting (12b) shapes. The an-

tion at the specific level. cestral condition is unknown, since theresult of the

10. Oval microstrongyles: Present in all species of generic cladogram does not allow us to deduceif the

Brachiaster, Pachastrella, and in Yodomiaperfecta ancestral condition was massive or encrusting.

(10b). The remaining species included in the anal- 13. Maximal branching in mesotriaenes: Mesotri-

lack this The "microrhabds" of aenes of Brachiaster those of and yses spicule (10a). , Triptolemus,

Characella pachastrelloides or Yodomia ijimai Yodomia perfecta have clads reaching tri- or

(sometimes namedmicrostrongyles) are assumed to tetracladose states (13c), whereas they are dichoto-

be non-homologous spicules of the oval micro- mous in Pachastrella ovisternata (13b). To assign a

strongyles (cf. Dendy, 1916), since they are longer, state of branching to species previously coded with

curved, and occasionally display oxeote ends in- the state "5a" ( = absence of mesotriaenes) is im-

dicating their origin from microxeas. In the same possible. So then, a third hypothetical state, "13a"

of way, degenerate streptasters Stoeba and Pacha- (absence of cladome) was assigned to those taxa

referred to as microrhabds or micro- with the state "5a". In this this strella, usually particular case, way

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 142 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

a single category (15b), although there are two mor-

phological categories (under the names of elongat-

ed microrhabds and cylindrical acanthorhabds) in

Pachastrella echinorhabda (15c). The ancestral

condition could not be deduced from the results of

the generic cladograms.

16. Streptaster categories in the ectosomal felt: A

felt containing both degenerate and regular strept-

asters (16b) occurs in Triptolemus intextus and T.

simplex. Stoeba has exclusively degenerate strept-

asters (16c). The remaining species lack strept-

asters in the felt (16a). The ancestral condition is

unknown.

Parsimony methods

The "Paup 3.0 g" program analyses the unordered,

ordered and irreversible characters under Fitch,

Kluge & Farris, and Camin-Sokal optimizations,

respectively (Swofford, 1989). The branch-and-

bound method was used to yield the most par-

simonious trees.

method The bootstrap (Felsenstein, 1985) was

applied to obtain confidence limits on the phyloge-

ny inferred at the specific level. Bootstrapping was

first applied to the complete data matrix.

Fig. 2. Tree 1, displayingthe inferred relationships according to the traditional concepts of the genus Pachastrella. The black Parsimony analysis results squares indicate apomorphy, the open squares indicate homo- The relationships inferred for the pachastrellid plasy, numbers in italics indicate terminal polymorphism. L(U) genera are presented in tree 1 (Fig. 2). The output = tree length under the hypothesis of uncertainty for multiple included another most parsimonious but it states. L(P) = tree length under polymorphism hypothesis for tree,

of multiple states. CI(U) = consistency index under uncertainty displayed a politomy for all three subgroups hypothesis. CI(P) = consistency index under polymorphism pachastrellids (Poecillastra-Characella-Sphinctrel- hypothesis. la, Pachastrella-Brachiaster, and Triptolemus-

Stoeba). A consensus tree leaves the relationships to treat incompatible characters is considered to be among these three subgroups unsolved. The am-

the more realistic than the use of question marks. The biguity in the relationships among subgroups state "13a" was assumed to be the ancestral condi- stems from the fact that it was not possible to order tion, obviously in congruence with the polarity of the character "11" unless speculative assumptions character 5. were taken. Note that in the input data (Table I),

14. still Regular streptaster: Absent in Stoeba (14a). In the presence of mesotriaenes was not included

of Pachastrella most species, they are exclusively forms with long in the generic concept (OTU = Pa- thin actines (14b), but in Triptolemus simplex they chastrella-1). occur along with forms having short conicalactines An exploratory analysis of the generic assigna-

(14c). State 14b was assumed to be the ancestral tion for Yodomia ijimai, Y. perfecta, and Pacha- condition. strella ovisternata is shown in tree 2 (Fig. 3). The

15. Degenerate streptasters: Absent in Brachiaster parsimony analysis yielded three equally parsi- and in monious All three T. incertus (15a). In most species they occur trees. above-mentioned pacha-

Downloaded from Brill.com10/11/2021 08:28:45AM via free access - Bijdragen tot de Dierkunde, 63 (3) 1993 143

Fig. 3. Tree 2, displaying anapproach to genericassignation for Fig. 4. Tree 3, displaying the inferred relationships after read-

traditionally overlooked species with mesotriaenes. Note that justing the generic concept of Pachastrella and Yodomia.

is in evidence for the polyphyly genus Yodomia (sensu Dendy). Legend for symbols as in Fig. 2.

Legend for symbols as in Fig. 2.

parsimonious trees. A consensus tree leaves the

strellid subgroups were reconfirmed in this analy- relationships between all three above-mentioned

sis, but a consensus tree again leaves the relation- pachastrellid subgroups unresolved.

ships among them unsolved. Note in the input The consistency index (CI) for all of the clado-

matrix that Y. and ovisternata is moderate (Table I) perfecta P. grams or low, especially under the are indistinguishable at the generic level (that is, us- hypothesis of polymorphisms for the multiple ing characters to diagnose genera) and closely relat- character states. This fact, however, does not indi- ed Pachastrella of mesotriaenes. bad inference to regardless The cate a of the generic relationships. It species Y. ijimai, on theother hand, is placed in the must be conceded that demonstrationof homopla-

far from Y. in of outgroup, perfecta, spite having sy by parsimony analysis is inherent in obtaining a been included traditionally in the same genus. low CI in the cladograms. A presumably homopla-

The definitive relationships inferred for this set sous character has to be coded identically in all of genera are shown in tree 3 (Fig. 4). For this anal- OTUs concerned in order to demonstrate that it has

the of mesotriaenes ysis, presence in some species of been acquired separately for them. This fact has a the Pachastrella considered genus was (OTU = Pa- negative repercussion in the CI value of the clado- chastrella-2) and besides, Yodomia was regarded as gram. The decrease in CI under the hypothesis of a monospecific exclusively Y. genus containing polymorphisms is a logical consequence of the ijimai. Parsimony analysis yielded three equally intra-OTU variability inherent to the characters

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 144 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

= = = I = AP Table II. Evaluation of the cladograms. NC = number of character; T type: U unordered, O ordered, irreversible;

states the lowercase letters to character = plesiomorphic states assumed, IP = plesiomorphic inferred from cladogram; correspond

of for unknown character = index under the uncertainty multiple states, questionmarks (?) indicate states; CI(U) consistency hypothesis

states; CI(P) = consistency index under polymorphism hypothesis.

GENERIC LEVEL SPECIFIC LEVEL

TREE 1 TREE 3 TREE 4

NC T AP IP CI(U) CI(P) CI(U) CI(P) T AP IP CI(U) CI(P)

- - 1 U ? a 1.00 1.00 1.00 1.00 _ _ -

2 u ? b 0.66 0.50 0.66 0.50 U a b 1.00 1.00

3 u 7 a 0.50 0.50 0.50 0.50 - - - - -

4 u J b 0.50 0.25 0.50 0.25 u b b 1.00 0.50

5 u J a 1.00 0.50 0.50 0.33 u a a 0.66 0.66

- - 6 u 7 a 1.00 0.66 1.00 0.66 - - -

7 u 7 b 1.00 0.75 1.00 0.75 - - - - -

- - - 8 o 7 7 1.00 1.00 0.75 0.75 - -

9 u J a 1.00 0.50 1.00 0.40 u a a 1.00 1.00

10 u 7 a 1.00 1.00 1.00 1.00 1 a a 1.00 1.00

- - 11 u 7 a 1.00 0.75 0.75 0.60 - - -

- - - - 7 b 12 - - - 0 1.00 0.66

13 ------0 a a 0.50 0.50

14 ------o b b 1.00 1.00

- - - - - b 15 - - u 7 0.66 0.66

16 ------u 7 a 1.00 1.00

with multiple states. The advantage of using multi- contribution to clarify the obscure relationships

ple states under the hypothesis of polymorphisms is which have been so far shrouding the enigmatic ge-

that this procedure permitted the detection of hid- nus Triptolemus.

den homoplasies and reversions in a study at the It is worth noting that no bootstrap test detected

generic level, which otherwise would be omitted by statistically significant monophyly in the ingroup

exclusively affecting some isolated species in each set, although a common linkage for all four species

of times. Such genus. If this variability is overlooked, the estima- of Triptolemus was found 90% the tion less than the 95% of the character consistency might result in un- a percentage, although suggested

realistically high values (Table II). by Felsenstein (1985), is large enough to be signifi-

Characters 1 (habit) and 10 (oval microstron- cant. In any case, the results of the bootstrap

gyles) were the most consistent at the generic level method ought to be regarded as approximative for

(Table II). Characters 4 (calthrops) and 5 (mesotri- two reasons: (1) tests were applied to data in which

the least the of character aenes) were noted to be consistent because homoplasous nature some states

of their intra-OTU variability and their homopla- was intentionally overlooked, and (2) characters 5

sous nature, respectively. and 13 are not fully independent of each other, and

The relationships inferred for the set of species neither are characters 14, 15 and 16, although they

the characterized by presence of oval microstron- are treated as stochastically independent in the

shown in tree 4 The gyles are (Fig. 5). parsimony resampling processes. analysis yielded one single most parsimonious tree. Table II shows characters 4 (calthrops) and 13

A close relationship seems evident on the clado- (maximal branching of the clads in mesotriaenes) as gram between all species included in the ingroup. the least consistent at the specific level. Note that

The relationship detected in all analyses between cladogram 4 (Fig. 5) indicates that mesotriaenes

and Stoeba in of Triptolemus provides a very interesting (sb) have arisen more than once this set spe-

Downloaded from Brill.com10/11/2021 08:28:45AM via free access Bijdragen tot de Dierkunde, 63 (3) - 1993 145

ijimai, and the species Yodomia perfecta ought to

be relocated in the genus Pachastrella, near Pacha-

strella ovisternata. The genus Yodomia (sensu Leb-

wohl, 1914) is a borderline genus between Pacha-

strellidae and Stellettidae, whose definitivelocation

can not be ascertained from the analyses performed

in this paper. In my view, it is closer to a stellettid

genus (similar to Stryphnus, for instance) than to

pachastrellid genera, although a new analyses most-

ly including stellettid taxa would be necessary to be

absolutely certain on this matter. It is worth men-

tioning that the parsimony analysis always yielded

trees in which Yodomia ijimai was related to

Stryphnus, even if the acanthosphaeras were not

assumed to be a type of reduced euasters.

The species set consisting of Brachiastersimplex,

Pachastrella perfecta (formerly Yodomia), P. ovi-

sternata, P. monilifera, and P. echinorhabda might

be assumed to be a single taxonomie unity, but it

does not have statistically significant monophyly. It

must be conceded that the parsimony analysis lacks

a sensitive method for distributing species in supra-

specific ranks, and the usual procedure consists of

assigning a new rank (clade) after each branching

point on the tree. It is generally more useful to take

Fig. 5. Tree 4, displaying the specific relationships inferred for an evolutionary attitude, constructing supraspecif- species of Pachastrella and Brachiaster. Note that the relation- ic taxa according to the degree of divergence be- ships are also well solved in the multiple outgroup. Legend for tween OTUs. So then, as far as the number of apo- symbols as in Fig. 2. morphic states between nodes is concerned, three

in this subgroups can be distinguished analysis (P. cies. Sarà's echinorhabda- Parsimony analyses support opinion P. monilifera, P. ovisternata-P. per-

(1959) that oxeas ought to be regarded as absent in fecta, and Brachiaster simplex). The state "5b" all of species Triptolemus. Oxeas described by (presence of mesotriaenes) is one of the two

Carter in = the first and the (1876) Triptolemus intextus ( T. para- apomorphies separating second siticus), probably belong to the host sponge Coral- subgroup and its consistency is very low at generic listes. If oxeas are assumed to be actually present in level. Thus, it would be advisable to keep both such have be character species, they to regarded as specific subgroups under the single generic name of reversals. Pachastrella.

As for the generic location of Brachiaster sim-

plex, cladograms indicate that this OTU is separat-

General discussion ed from therelated at least for species two apomor-

phies: "5c" (presence of mesotriders) and "15a"

Parsimony analyses clearly indicate that the genus (absence of degenerate streptasters). In this case,

in the the number of Yodomia, way diagnosed by Dendy (1916), is apomorphies can not really prove a polyphyletic taxon. Consequently, this genus whether or not this taxon indeed belongs to Pacha-

to be used in the strict sense of Lebwohl strella or should be considered as a different ought only genus

related (1914), containing only the type species Yodomia (although closely to Pachastrella). It is

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 146 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

worth noting, however, that the relationship be- (character 13), it was found to be one of the least tween the mesotriaenes of Pachastrella (5b) and the consistent characters in the analysis at the specific mesotriders of Brachiaster (5c) is unclear. If the level (Table II). At first sight, the geographical state "5b" (instead of "5c") is assigned to separation between both species might be good

Brachiaster, the analysis indicates that mesotri- evidence to warrant their specific separation.

in it is worth that other aenes are a shared homologue (synapomorphy) However, mentioning pacha- both taxa. This fact strengthens the affinity be- strellid and calthropellid species, such as Sphinc- tween both taxa and it even introduces the possibili- trella annulata (Carter, 1880) (e.g. Topsent, 1897, ty of a generic synonymy. However, there are other 1904, 1928; Carter, 1880) or Pachastrissa geodioi- aspects involved in this matter, which merit careful des (cf. Carter, 1876; Topsent, 1897, 1904; Des- examination: (1) only two specimens of Brachiaster queyroux-Faúndez, 1981), are simultaneously pres-

known Lévi in Lusitanian are (cf. Wilson, 1925; & Lévi, 1989) ent both and Indonesian-Philippine and one of them, the holotype, notably macerated; locations.

(2) dichotriaenes becoming non-mesotrider des- The parsimony analysis clearly indicated that the mas, described in the holotype, are absent in the presence of mesotriaenes in calthropellids and recent specimen; (3) oxeas and microscleres, de- pachastrellids is a homoplasy. Moreover, mesotri- scribed in the recent specimen, were not found in aenes seem to have arisen separately in Triptolemus the holotype. The lack of some small skeletal ele- and Pachastrella-Brachiaster, in spite of being ments (microscleres) in the holotype may be tenta- closely related. Notwithstanding, it is possible that, tively attributedto its maceration state (cf. Wilson, in this particular case, the homoplasy might have

1925), but, in my view, this assumption is not con- resulted from genetic polymorphism (rather than

the of from that arises and vincing to explain complete absence long ox- convergent evolution) persists eas. In addition, the absence of dichotriaenestrans- through speciation events, segregating the same forming to non-mesotrider desmas also remains un- variants at differentpoints on the tree (Felsenstein, explained in the recent specimen (Lévi & Lévi, 1983).

the skeletal differences Short-shafted mesotriaenes 1989). In a word, true be- correspond to a spe- tween both specimens remain enigmatic and, in my cial morphology of tetraxon spicules, which has opinion, the assumed conspecific nature of both been independently acquired from calthrops (in specimens should be regarded as an uncertainty. Calthropellidae), short-shafted plagiotriaenes (e.g.

Therefore it seems advisable to keep the genus Yodomia ijimai), and short-shafted orthodichotri-

Brachiaster, as such, till further material produces aenes (e.g. Pachastrella, Triptolemus). Details of more useful information. the shape displayed in mesotriaenes of different

In the same way, the description of new speci- groups allow the establishment of morphological mens would be necessary to clarify the matter con- types corroborating the homoplasous nature de- cerning a possible identity between Pachastrella tected in the parsimony analyses: (1) Spicules with ovisternata and Pachastrellaperfecta. Both species clads undivided, equally long andarranged at equal are only distinguished by two characters: (1) angles (mesocalthrops) (PI. IVb, c, d); presence of presence/absence of dichotriaenes in their respec- a fifthclad usually underdeveloped (Pl. IV a, d) is tive spicule complements; (2) occurrence of mul- characteristic; they occur in calthropellid species. tibranched clads in mesotriaenes of P. perfecta ver- (2) Short-shafted plagiomesotriaenes with un- sus dichotomous clads in P. ovisternata. Both divided clads and a well-developed mesoactine presumably distinctive characters are not very relia- (epirhabdome). As far as known, they occur exclu- ble taxonomie of view. from a point Presence- sively in Yodomia ijimai. (3) Short-shafted meso- absence of dichotriaenes is subject to intraspecific triaenes characterized by a well-developed epirhab-

in variability pachastrellid and calthropellid spe- dome and a diversely branched cladome arranged cies, such as Characella pachastrelloides or Pacha- in upright position. Spicules with undivided clads strissa geodioides (Carter, 1876). As for the maxi- can occur, but they are regarded as either imma- mal state in branching of clads in mesotriaenes ture or underdeveloped ontogenetic states. These

Downloaded from Brill.com10/11/2021 08:28:45AM via free access - Bijdragen tot de Dierkunde, 63 (3) 1993 147

in of Pachastrella spicules occur some species This research was supported by a Generalitat of Catalunya

and by the funds of "Fauna Ibérica-II" Project (CICYT (PI. lie, d) and in all species of Triptolemus. (4) grant PB89-0081). Short-shafted orthotriaenes, diversely branched,

in becoming mesotriders. They occur monotypic

its relation with the above- genus Brachiaster, but Literature cited mentioned type (3) is unclear.

There is no satisfactory explanation for the Bergquist, P.R., 1968. The marine fauna of New Zealand: shortage of records on pachastrellids and calthro- Porifera, Demospongiae, Part 1 (Tetractinomorpha and pellids with short-shafted mesotrianes. However, N.Z. = Lithistida). Mem. oceanogr. Inst., 37 ( Bull. N.Z. there are several possibly connected items that are Dep. scient, ind. Res., 188): 1-105.

worth considering: (1) No paleontological evidence Bergquist, P.R. & M. Kelly-Borges, 1991. An evaluation of the

with in the fossil record indicates the existence of meso- genus Tethya (Porifera: Demospongiae: Hadromerida)

descriptions of new species ofthe Southwest Pacific. The Bea- triaenes prior to calthrops. (2) Short-shafted meso- gle, Rec. Northern Territ. Mus. Arts Sci., 8 (1): 37-72, pis. triaenes are only present in genera and families I—II.

characterized by the of calthrops or In: presence Bowerbank, J.S., 1861. List of the British sponges. [R.]

"pseudocalthrops". (3) Fossil desmas derived from McAndrew's "List of marine invertebrate fauna". Rep. Br.

Advmt. London: 36. mesotriaenes (mesotriders) do not exist, while Ass. Sci., 30, 830-836, pi. Bowerbank, J.S., 1866. A monograph of the British Spongia- numerous fossil desmas originating from calthrops dae, 2: 1-388 (Ray Society, London). do occur (e.g., De Laubenfels, 1955; Reid, 1970). Burton, M., 1926. Description of South African sponges collect- (4) Surprisingly, desmas having a mesotriaene cré- ed in the South African Marine Survey. Part I. Myxospongia

in the pis occur exclusively recent monospecific ge- and Astrotetraxonida. Rep. Fish. mar. biol. Surv. Un. S.

Afr., 4: 1-29, pis. 1-6. nus Brachiaster Wilson, 1925 (Levi & Lévi, 1989; Burton, M., 1959. Sponges. Scient. Rep. John Murray Exped., Lévi, 1991). (5) The epirhabdome or fifth actine 10(5): 151-281. was noted to be notably or slightly less developed Carter, H. J., 1876. Descriptions and figures of deep-sea sponges than the rhabdome sensu stricto or the equivalent and their spicules, from the Atlantic Ocean, dredged up on

actine in all short-shafted mesotriaenes studied. board H.M.S. 'Porcupine' chiefly in 1869 (concluded). Ann.

nat. Hist., 18: 388-410, pis. XIV-XVI. In conclusion, most evidence suggests that the Mag. (4)

Carter, H.J., 1880. Report on specimens dredged up from the presence of the fifth actine or epirhabdome might Gulf of Manaar and presented to the Liverpool Free Museum be either a recent or relatively late evolutionary ac- by Capt. W.H. Cawne Warren. Ann. Mag. nat. Hist., (5) 5: for the quisition calthropoid spicules, although a 437-457, pis. XVIII-XIX; (5) 6: 35-61, pis. IV-VI; (5) 6:

detailed study involving abundant fossil material 129-156, pis. VII-VIII.

Dendy, A., 1905. Report on the collected by Professor would be necessary to be certain on this assertion. sponges

Herdman at Ceylon in 1902. Report to the Government of

Ceylon on the pearl oyster fisheries of the Gulf of Manaar, 3

Acknowledgements (Suppl. 18): 57-246, pis. I—XVI (Royal Society, London).

Dendy, A., 1916. Report on the Homosclerophora and Astro-

Dr. I am indebted to Jesús Benito from the ComplutenseUniver- tetraxonida collected by H.M.S. "Sealark" in the Indian

sity of Madrid for the opportunityto study this interesting speci- Ocean. Trans. Linn. Soc. Lond., (2, Zool.), 17 (2): 225-271,

44-48. men, and to Drs. S. Stone (BMNH, London), E. Kritscher pis.

(NMW, Wien), O. Soriano (MNCN, Madrid) and V. Rainieri Desqueyroux-Faúndez, R., 1981. Révision de la collection

(MSNG, Genoa) for the loan of material. I also thank Drs. d'épongés d'Amboine (Moluques, Indonésie) constituée par

R.W.M Van Soest (Zoölogisch Museum, Amsterdam), S.M. Bedot et Pictet et conservée au Muséum d'histoire naturelle de

Stone (BMNH, London), Q. Ferreiro (Museu e Laboratório Genève. Revue suisse Zool., 88 (3): 723-764.

Zoológico, Coimbra), and D. Kühlmann (Zoologisches Muse- Felsenstein, J., 1983. Parsimony in systematics: biological and

I A. Rev. 14: um, Berlin) for their help in the search for the type material. statistical issues. Ecol. Syst., 313-333. also thank Dr. D. Byrne for correcting the English. The Techni- Felsenstein, J., 1985. Confidence limits on phylogenies: an ap- cal staffof the Microscopy Service (Central University of Barce- proach using the bootstrap. Evolution, 39 (4): 783-791.

del lona) are thanked for their friendly and efficient assistance. Spe- Ferrer-Hernández, F., 1914. Esponjas Cantábrico. Parte 2a. cial thanks are due to two anonymous reviewers, whose con- III. Myxospongida. IV. Tetraxonida. V. Triaxonida. Trab.

Mus. Nat. structive comments have notably improved the manuscript. nac. Cieñe. Madrid (Zool.), 17: 3-44.

Downloaded from Brill.com10/11/2021 08:28:45AM via free access 148 M. Maldonado - Taxonomic significance of the mesotriaene in Astrophorida

research methods: Gray, J.E., 1867. Notes on the arrangementof sponges, with the R.E. Johannes (eds.), Coral reefs:

Proc. 299-313 description of some new genera. Zool. Soc. London, (Unesco, Paris).

1867: 492-558, pis. XXVII-XXVIII. Sarà, M., 1959. Specienuove di Demospongieprovenientidaac-

1895. Notes collection of from the del Golfo di Annuar. 1st. Mus. Zool. Hanitsch, R., on a sponges que superficiali Napoli.

West coast of Portugal. Proc. Trans. Lpool. biol. Soc., 9: Univ. Napoli, 11 (7): 1-22, pl. I.

205 -219, pis. XII-XIII. Schmidt, O., 1868. Die Spongien der Kiiste von Algier. Mit

den des Adriatischen Kirkpatrick, R., 1902. Descriptions of South African sponges. Nachtragen zu Spongien Meeres, Suppl.

Mar. Invest. S. Afr., 1: 219-232, pis. I—III. 3: i—iv, 1-44, pis. I-V (W. Engelmann, Leipzig).

1903. of South African 1870. einer des Atlan- Kirkpatrick, R., Descriptions sponges. Schmidt, O., Grundzüge Spongien-Fauna

Part II. Mar. Invest. S. Afr., 2: 171-180, pl. IV. tischen Gebietes: i-iv, 1-88, pis. I-VI (W. Engelmann,

M.W. 1936. A discussion of the fauna Laubenfels, de, sponge Leipzig).

of the Dry Tortugas in particular and the West Indies in gener- Schmidt, O., 1880. Die Spongien des Meerbusen von Mexico

V-X al, with material for a revision of the families and orders of (und des Caraibischen Meeres), 2: 33-90, pis. (G.

the Porifera. Pubis. Carnegie Inst. Wash., 467 ( = Pap. Tor- Fischer, Jena).

Barbados tugas Lab., 30): i-iii, 1-225, pis. 1-22, 1 map. Soest, R.W.M. van& N. Stentoft, 1988. deep-water

Laubenfels, M.W. de, 1955. Porifera. In: R.C. Moore (ed.), sponges. Stud. Fauna Curaçao, 70 (215): 1-175.

Treatise on invertebrate paleontology: E21-EI22 (Univ. Sokal, R.R., J. Kim & F.J. Rohlf, 1992. Character and OTU

Kansas Press, Lawrence). stability in five taxonomie groups. J. Classification, 9:

Lebwohl, F., 1914. Japanische Tetraxonida. I. Sigmatophora 117-140.

W.J. and II. Astrophora metastrosa. J. Coll. Sci. imp. Univ. Solías, 1888. Report on the collected by

H.M.S. the 1873-76. scient. Tokyo, 35 (2): 1-116, pis. I-IX. "Challenger" during years Rep.

Lendenfeld, R. von, 1894. Eine neue Pachastrella. Sber. Akad. Results Voy. Challenger, (Zool.) 25 (63): i-clxvi, 1-458, pis.

Wiss. Wien (Mat.-nat. Classe, Abth. I), 103 (1): 439-441, I-XLIV, 1 map.

pl. I. Swofford, D.L., 1989. PAUP 3.0 g. Phylogeneticanalysis using

R. Lendenfeld, von, 1903. Tetraxonia. Das Tierreich, 19: i-xv, parsimony (user's manual): i-iii, 1-43 (IllinoisNatural His-

1-168. tory Survey, Champaign).

Levi, C., 1973. Systématique de la classe des Demospongiaria Topsent, E., 1897. Spongiaires de la Baie d'Amboine. Voyage

(Démosponges). In: R-P. Grassé (éd.), Traité de zoologie, 3 de MM. M. Bedot et C. Pictet dans l'Archipel malais. Revue

(1): 557-631 (Masson et Cie, Paris). suisse Zool., 4 (3): 421-487, pis. XVIII-XX1.

scient, méd. Lévi, C., 1991. Lithistid sponges from the Norfolk Rise. Recent Topsent, E., 1902. Les Asterostreptidae. Bull. Soc.

and Mesozoicgenera. In: J. Reitner & H. Keupp (eds.), Fossil Ouest, 11 (2): 1-18.

and recent sponges: 72-82 (Springer Verlag, Berlin, Topsent, E., 1904. Spongiaires des Açores. Résuit. Camp,

Heidelberg). scient. Prince Albert I, 25: [i-iii], 1-280, pis. I—XVIII,

Lévi, C. & P. Lévi, 1983. Éponges Tétractinellides et Lithistides [281-316],

bathyales de Nouvelle-Calédonie. Bull. Mus. natn. Hist, nat., Topsent, E., 1928. Spongiaires de l'Atlantiqueet de la Méditer-

Paris, (Sect. A, Sér. 4), 5 (1): 101-168, pis. I— XIII. ranée, provenant des croisières du Prince Albert 1er de Mona-

Lévi, C. & P. Lévi, 1989. Spongiaires (MUSORSTOM 1 & 2). co. Résuit. Camp, scient. Prince Albert I, 74: [i—v], 1-376,

Forest In: J. (éd.), Résultats des campagnes MUSORSTOM, [377-399], pis. I-XI.

coralliens 4. Mém. Mus. natn. Hist, nat., (A) 143: 25-103. Vacelet, J. & P. Vasseur, 1971. Eponges des récifs de

1973. collection of Tuléar 51 — I-IV. Pulitzer-Finali, G., Report on a sponges (Madagascar). Téthys, (Suppl. 1): 126, pis.

from the Bay ofNaples. I. Sclerospongiae,Lithistida, Tetrac- Vacelet, J., P. Vasseur & C. Lévi, 1976. Spongiaires de la pente

tinellida, Epipolasida. Pubbl. Staz. zool. Napoli, 38: "1970": externe des récifs coralliens de Tuléar (sud-ouest de Madagas-

328-354. car). Mém. Mus. natn. Hist, nat., (A, Zool.) 99: 1-116, pis.

Pulitzer-Finali, G., 1983. A collection of Mediterranean Demo- I-X.

spongiae (Porifera) with, in appendix, a list ofthe Demospon- Watrous, L.E. & Q.D. Wheeler, 1981. The out-group compari-

giae hitherto recorded from the Mediterranean Sea. Annali son method of character analysis. Syst. Zool., 30: 1—11.

Mus. civ. Stor. nat. Giacomo Doria, 84: 445-621. Wiedenmayer, F., 1977. Shallow-water sponges of the western

Pulitzer-Finali, G., 1986. A collection of West Indian Demo- Bahamas (Experientia Supplementum, 28): 1-287 (Birk-

spongiae (Porifera). In appendix, a list of the Demospongiae haüser Verlag, Basel & Stuttgart).

hitherto recorded from the West Indies. Annali Mus. civ. Wilson, H.V., 1925. Siliceous and horny sponges collected by

Stor. nat. Giacomo Doria, 86: 65-216. the U.S. Fisheries steamer "Albatross" duringthe Philippine

Reid, R.E.H., 1970.,Tetraxons and demospongephylogeny. In: Expedition, 1907-10. Bull. U.S. natn. Mus., 100 (2,4): i-vii,

W.G. Fry (ed.), The biology of the Porifera. Symp. zool. Soc. 273-532, pis. 37-52.

London, 25: 63-89 (Academic Press, London/New York).

Riitzler, K., 1978. Sponges in coral reefs. In: D.R. Stoddart & Final draft received: 19 March 1993

Downloaded from Brill.com10/11/2021 08:28:45AM via free access