ACANTHODESMIIDAE Haeckel, 1862

Suzuki, Noritoshi, Caulet, Jean-Pierre & Dumitrica, Paulian, 2021, A new integrated morpho- and molecular systematic classification of Cenozoic radiolarians (Class Polycystinea) - suprageneric taxonomy and logical nomenclatorial acts, Geodiversitas 43 (15), pp. 405-573 : 484-487

Family ACANTHODESMIIDAE Haeckel, 1862

Acanthodesmida Haeckel, 1862: 237, 265-266 [as both family and tribe]; 1882: 445 [as a tribe]; 1887: 970, 973 [as a subfamily of Coronida]. — Zittel 1876-1880: 123 [rank unknown]. — Mivart 1878: 179 [as a subsection]. — Stöhr 1880: 86 [as an order].

Acanthodesmidae – Claus 1876: 158 [in suborder Thalassicollea ].

Acanthodesmiden – Hertwig 1879: 196-200 [as a family].

Perispyrida Haeckel, 1882: 443 [as a subfamily]; 1887: 1092, 1095 [as a subfamily].

Triostephanida Haeckel, 1882: 445 [as a subfamily].

Circospyrida Haeckel, 1882: 443 [nomen dubium, as a tribe]; 1887: 1024, 1072.

Eucoronida Haeckel, 1882: 445 [as a tribe]; 1887: 970, 976 [as a subfamily].

Trissocyclida Haeckel, 1882: 446 [as a tribe]; 1887: 970, 982 [as a subfamily].

Monostephida Haeckel, 1882: 447 [nomen dubium, as a subfamily].

Tympanida Haeckel, 1887: 937, 987-991 [as a family]. — Bütschli 1889: 1978 [as a family]. — nec Rüst 1892: 177 [as a family]. — Anderson 1983: 29 [as a family].

Semantida Haeckel, 1887: 937, 953-956 [as a family]. — Bütschli 1889: 1977 [as a family]. — Anderson 1983: 29 [as a family].

Coronida Haeckel, 1887: 937, 967-970 [as a family]. — Bütschli 1889: 1977 [as a family]. — nec Rüst 1892: 176. — Anderson 1983: 29 [as a family].

Lithocircida Haeckel, 1887: 940 [as a subfamily].

Cortiniscida Haeckel, 1887: 956 [as a subfamily].

Protympanida Haeckel, 1887: 990, 991 [nomen dubium, as a subfamily].

Semantidae – Popofsky 1908: 267; Popofsky 1913: 297. — Schröder 1914: 87-88. — Clark & Campbell 1945: 29. — Chediya 1959: 169. — Cachon & Cachon 1985: 292.

Tympanidiidae [sic] – Poche 1913: 219 (= Tympaniidae).

Coronidiidae – Poche 1913: 219.

Semantididae [sic] – Poche 1913: 219 (= Semantidae ). — Campbell 1954: D106. — Tan & Tchang 1976: 270. — Chen & Tan 1996: 152. — Tan & Chen 1999: 271. — Tan & Su 2003: 85.

Coronidae [sic] – Popofsky 1913: 300 (= Coronidiidae). — Schröder 1914: 87. — Chediya 1959: 171. — Tan & Tchang 1976: 270. — Cachon & Cachon 1985: 292.

Tympanidae – Popofsky 1913: 301. — Schröder 1914: 87. — Chediya 1959: 173. — Cachon & Cachon 1985: 292-293.

Acanthodesmiidae – Campbell 1954: D106. — Riedel 1967b: 296; 1971: 656. — Riedel & Sanfilippo 1970: 523;1971: 1590. — Petrushevskaya 1971a: 260; 1971b: 990; 1981: 353-354. — Dumitrica 1973a: 840; 1979: 35. — Petrushevskaya & Kozlova 1972: 532- 533. — Sanfilippo & Riedel 1973: 526. — Nakaseko et al. 1975: 173. — Nishimura 1990: 116, 118 ( sensu emend. ). — Takahashi 1991: 101. — van de Paverd 1995: 200-201. — Anderson et al. 2002: 1017. — De Wever et al. 2001: 230, 232. — Afanasieva et al. 2005: S305. — Afanasieva & Amon 2006: 155.

Lithocircinae – Campbell 1954: D106.

Semantidinae [sic] – Campbell 1954: D106 (= Semantinae).

Cortiniscinae – Campbell 1954: D106. — Chediya 1959: 170.

Acanthodesmiinae – Campbell 1954: D106-107. — Petrushevskaya 1981: 356-357. — Afanasieva et al. 2005: S305-306. — Afanasieva & Amon 2006: 156-157.

Eucoronidinae – Campbell 1954: D108.

Trissocyclinae – Campbell 1954: D108. — Chediya 1959: 172.

Protympaniinae – Campbell 1954: D108 [nomen dubium].

Perispyridinae – Campbell 1954: D116 (not from the Mesozoic Perispyridium ). — Petrushevskaya 1981: 354. — Afanasieva et al. 2005: S305. — Afanasieva & Amon 2006: 155.

Circospyridinae – Campbell 1954: D114 [nomen dubium]. — Petrushevskaya 1981: 364-366. — Afanasieva et al. 2005: S305. — Afanasieva & Amon 2006: 155.

Lithocyrtinae [sic] – Chediya 1959: 168 (= Lithocircinae).

Acanthodesminae [sic] – Chediya 1959: 171 (= Acanthodesmiinae).

Eucoroninae [sic] – Chediya 1959: 172 (= Eucoronidinae).

Protympaninae [sic] – Chediya 1959: 173 [nomen dubium] (= Protympaniinae).

Circospyrinae [sic] – Chediya 1959: 181 [nomen dubium] (= Circospyridinae).

Perispyrinae [sic] – Chediya 1959: 184 (= Perispyridinae). — Tan & Su 1982: 166.

Trissocyclidae – Goll 1968: 1416-1417 ( sensu emend. ). — Hollis 1997: 83.

Spyridae [sic] – Boltovskoy 1998: 33 [nomen nudum] (= Spyrididae).

TYPE GENUS. — Acanthodesmia Müller, 1856: 485 [type species by subsequent designation ( Campbell 1954: D107): Lithocircus vinculatus Müller, 1856: 484 ].

INCLUDED GENERA. — Acanthodesmia Müller, 1856: 485 (= Acanthostephanus n. syn., Octotympanum n. syn., Tristephaniscus n. syn., Tristephanium n. syn., Triostephus n. syn., Tympanura n. syn., Zygostephus n. syn., Zygostephanus n. syn.; Tympanium synonymized by Nigrini & Lombari 1984: N75; Lithocoronis synonymized by Petrushevskaya 1971a: 274). — Dictyospyris Ehrenberg, 1846: 385 (= Dictyospyrissa synonymized by Petrushevskaya 1971a: 267;? Dictyospyrantha n. syn.; Dictyospyrella synonymized by Kozlova 1999: 164). — Eucoronis Haeckel, 1882: 445 (= Acrocoronis with the same type species; Acrocubus , Apocubus , synonymized by Petrushevskaya 1971a: 267; Coronidium synonymized by Petrushevskaya 1981: 358). — Lithocircus Müller, 1856: 484 (= Archicircus , Archistephus , synonymized byPetrushevskaya 1971a: 269). — Lithocubus Haeckel, 1882: 446 . — Lithotympanium Haeckel, 1882: 447 . — Semantis Haeckel, 1887: 956 (= Cortiniscus n. syn.). — Tricolospyris Haeckel, 1882: 443 (= Perispyris n. syn.). — Trissocyclus Haeckel, 1882: 446 (= Tricyclarium with the same type species; Tricirconium n. syn., Tricyclonium n. syn.; Tricircarium , Trissocircus, Zygostephanium , synonymized by Petrushevskaya & Kozlova 1972: 533). — Tympanomma Haeckel, 1887: 1004 .

NOMINA DUBIA. — Circospyris , Dendrocircus , Dictyospyromma , Dipocoronis , Dipocubus , Hexacoronis , Monostephus , Plectocoronis , Podocoronis, Prismatium , Stephaniscus, Stephanolithis , Stylocoronis , Tetracoronis , Tetracubus , Tripocoronis , Tripocubus , Tympaniscus , Zygostephaniscus .

INVALID NAME. — Lithotympanum , Tympanidium .

DIAGNOSIS. — Acanthodesmiidae formed by a sagittal ring with twin cupola or twin set of body frames. The Lo-axis is parallel to Lg-axis while the Sh-axis is parallel to the Sg-axis. No significant skeleton developed below the basal ring. The basal ring is construct- ed of two to six basal pores. The endoplasm is situated within the space encapsulated by the sagittal ring. The space of the cupola is occupied by algal symbionts.

STRATIGRAPHIC OCCURRENCE. — Middle Paleocene-Living.

REMARKS

This family is probably an artificial group. The very young form of Lithocircus closely resemble Zygocircus (Stephaniidae) but the former is considered to be a young stage of some acanthodesmioid genera. This family is distinguishable from the Cephalospyrididae by the presence of a significant skeleton below the basal ring and from the Paradictyidae in the Lo-axis parallel to the Sg-axis but not to Lg-axis. Living appearance and cytological ultrafine structure were illustrated in Acanthodesmia (Anderson 1983: fig. 1.2.C; Cachon & Cachon 1985: fig. 53.b; Matsuoka 1993a; fig. 2.9; 2017: fig. 16; Sugiyama & Anderson 1998b; Suzuki & Aita 2011: fig. 5.K; Suzuki & Not 2015; fig. 8.11.2; Matsuoka et al. 2017: Appendix B; Zhang et al. 2018: 10, figs 2.40-2.45, pl. 15, figs 4.22, 4.26) and Lithocircus ( Probert et al. 2014: S2, VEPO-10). Algal symbionts of Acanthodesmia were identified as Gymnoxanthella radiolariae , the same dinoflagellate species as those of Dictyocoryne elegans ( Euchitoniidae , Spumellaria ) and Dictyopodium (originally Pterocanium , Lithochytrididae , Nassellaria in Yuasa et al. 2016). Meanwhile, algal symbionts of Lithocircus were identified as Brandtodinium nutricula by Probert et al. (2014).

VALIDITY OF GENERA

Acanthodesmia

The following combination has the same type species, respectively: Tristephanium and Triostephus ; Tympanium , Tympanidium and Tympanura ; and Zygostephanus and Zygostephus . The genera synonymized in this paper were classified into two families by Campbell (1954): “Acanthodesmidae” whose skeleton is formed by one sagittal ring (D-ring), a horizontal basal ring, and a well-developed vertical meridian ring ( Campbell 1954: D106) and “Paratympanidae” whose skeleton is composed of one sagittal ring (D-ring), a horizontal basal ring, and two parallel vertical meridian rings ( Campbell 1954: D108). The former is “twin q -ring” and the latter is “twin z -ring” parallel to “twin q- ring” in the sense of Petrushevskaya (1969: fig. 1.7). Both “family” specimens are actually found at least in the northeastern Indian Ocean (confirmed by Zhang Lanlan & Suzuki Noritoshi in the same slides). These specimens cannot be differentiated even at specific level by any other characters except by the number of vertical meridian rings. In consideration of this situation, the number of parallel vertical meridian rings is not applicable at not only a family level but also at genus level. These families are subdivided into several subfamilies by the number of gates (opening) in Campbell (1954). The definition of a “gate” is unclear because it is not defined by geometric rules. For example, Acanthodesmia belongs to the “Acanthodesmiinae” which are defined as having “ five large gates, or openings, between rings ” ( Campbell 1954: D106), but as for the number of openings following geometric rules, Acanthodesmia has eight openings (not explained in detail here). According to Campbell (1954: D107-108), Acanthodesmia is characterized by partly latticed gates and Lithocoronis by armed rings with arborescent spines. The description of “partly latticed gates” in Acanthodesmia is wrong because the type-illustration has no latticed parts ( Müller 1859b: pl. 1, fig. 7). A “Partly latticed gate” is visible in well-preserved fully-grown specimens. The genus name Acanthostephanus seems to appear only in the first description of Haeckel (1879: 705) and its type species is marked by thorny rings. There are many intermediate forms between Acanthodesmia through Acanthostephanus to Lithocoronis in same samples, suggesting ontogenetic variations. High variety in Acanthodesmia has already been commented by Petrushevskaya (1971a: 274). Triostephus is characterized by a sagittal ring (D-ring) and frontal rings (twin q -ring) of different sizes and forms whereas Tristephaniscus by a D-ring and twin q -rings alike ( Campbell 1954: D108). Different shape and size between D-ring and twin q -ring depends highly on species, so it cannot be applied for genus criteria. The type-illustration of Tristephus ( Haeckel 1887: pl. 93, fig. 9) is similar to that of Acanthostephanus ( Haeckel 1879: pl. 16, fig. 7) with thorny rings but their differences are the presence of a twin a -f ring which is parallel to the basal ring and several feet on Tristephus. As shown in the support image for Acanthostephanus, the development of feet are intra- or infra-species variations. Several specimens display incomplete a-f ring and, thus, some species can present Acanthostephanus -forms to Tristephus -forms. This means that it is not necessary to separate Tristephus from Acanthostephanus. Zygostephanus is marked by a vertical ring without a sagittal constriction ( Campbell 1954: D108). The meaning of “without sagittal constriction” is not understandable because no sagittal constrictions on vertical rings (in this case, D-ring and twin q -ring) are observed in Acanthodesmia (support image for Acanthodesmia in the Atlas ). The typeillustration of Zygostephanus muelleri , the type species of Zygostephanus ( Haeckel 1862: pl. 12, fig. 2), looks to lack a basal ring. As like the support image for Acanthodesmia in the Atlas , the basal ring is easily overlooked without special care. Octotympanum and Tympanium are characterized by the presence of parallel twin q- rings and twin z -rings forementioned because they were classified in the “Paratympanidae”. According to Campbell (1954: D108), Octotympanum is marked by incomplete equatorial rings but this characteristic is meaningless for any identification even at species level due to differences in the ontogenetic growth. The type-illustrations of Tympanium as well as Octotympanum are nearly identical to the type-illustration of Lithocoronis within species or at species level, except for the presence/absence of twin z -rings. The basal ring has two or four polygonal pores depending on the development stage of doble l-rod ( Goll 1972a: pl. 63, fig. 2 for two basal pores-type; Goll 1969: pl. 60, fig. 3 for four basal pores-type). The two basal pores type has two unified Ca - and Cerv -pores and the four basal pore type has very large twin Ca -pores and small twin Cerv -pores. J -pore is unknown. All the genera listed here are synonymized here like this way. The oldest available name Acanthodesmia is validated.

Dictyospyris

All the genera synonymized here belonged to the “Circospyridinae” sensu Campbell (1954 : D114) but this diagnosis is too incomplete to permit to precisely identify them. Real type specimens for Dictyospyris , Dictyospyrella and Dictyospyrissa were re-discovered in the Ehrenberg collection ( Ogane et al. 2009b: pl. 9, figs 2a, 2b for Dictyospyris [as a topotype], pl. 75, figs 3b, 3c for Dictyospyrella , and pl. 38, figs 1a, 1b for Dictyospyrissa ). Lack of basal feet is a distinguishing character for the “Circospyridinae” according to Campbell (1954), but the most important common structure in them are the presence of a latticed cephalic wall with small pores and the absence of any spines derived from the initial spicule system. The basal ring of Dictyospyrella in the lectotype ( Ogane et al. 2009b: pl. 75, figs 3b, 3c) comprises three large pores (twin Ca -pores and a unified Cerv -pore) and two tiny pores (twin J -pores). The drawing of Dictyospyris fenestra by Ehrenberg (1876: pl. 19, fig. 11), the type species of Dictyospyrissa , looks as having four large basal pores. The real specimen of the Ehrenberg’s drawing shown in Ogane et al. (2009b: pl. 38, figs 1a, 1b) first confirmed that this specimen is obliquely oriented in the microscopic slide. Referred to the lectotype photo, these four large pores correspond to twin J- pores for the upper pores and twin Ca -pores for the lower pores. The Cerv -pores are invisible in the lectotype but the support image for Dictyospyrissa displays a unified Cerv pore in the lower side of two photos. The topotype of Dictyospyris trilobata ( Ogane et al. 2009b: pl. 9, figs 2a, 2b), the type species of Dictyospyris , looks to have 3 large basal pores which correspond to twin Ca -pores and a unified Cerv -pore. Probable tiny twin J -pores are visible in the lower side of the specimen in the lower photo of the support image for Dictyospyris . These observations permit to conclude that Dictyospyris has five basal pores composed of large twin Ca -pores, a small to large unified Cerv -pore and small to large twin J -pores. If the unified Cerv -pore is large, the double J -pores are small. By contrast, if the unified Cerv - pore is small, the twin J- pore is large. Due to this pattern, the number of large basal pores changed as three or four among them. Large variation is only recognized in basal rings.

The type-illustration of Dictyospyris stalactites ( Haeckel 1887: pl. 89, fig. 7), the type species of Dictyospyrantha , surely fits to the description of this genus as well as the diagnosis of “Circospyridinae” sensu Campbell (1954) . A probable Dictyospyrantha specimen is illustrated by Goll (1968: pl. 173, figs 21-24; 1972b: pls 73-74). Four basal pores are present with relatively large twin Ca -pores besides MB and relatively smaller unified Cerv -pores besides the V-rod. A unified J -pore is large and placed on the dorsal side of the test. In consideration of this basal pore pattern Dictyospyris and Dictyospyrantha may be different genera, but there is no time possibility to fix a much better position of Dictyospyrantha in this Atlas . According to Campbell (1954: D114), Dictyospyris , Dictyospyrissa and Dictyospyrella are respectively characterized by a basal ring with four-heart shaped basal pores, four large basal pores and three large basal pores. Under the current taxonomic system for Nassellaria ( De Wever et al. 2001) , it is impossible to synonymize genera with different numbers of basal pores on the basal ring. On the other hand, if this variation is plausible, Dictyospyris , Dictyospyrissa and Dictyospyrella can be synonymized as a single genus. The oldest available name is Dictyospyris among them including Dictyospyrantha .

Eucoronis

The combination of Eucoronis and Acrocoronis and that of Acrocubus and Apocubus have respectively the same type species. As Acanthodesmia , the genera synonymized here are classified into the “ Acanthodesmiidae ” ( Coronidium , Eucoronis ) and the “Paratympanidae” (Acrocubus) sensu Campbell (1954 : D107-108). The “Paratympanidae” are defined by two parallel rings but the reliability of the type-illustration for Acrocubus is suspected. Referred to real specimens and following the terminology in the remarks of the Acanthodesmioidea , the type-illustration of Eucoronis is a view from the lateral plane (Lt-plane) ( Haeckel 1887: pl. 82, fig. 6; the support image for Eucoronis in the Atlas ) and the type-illustration of Coronidium is a view from the supra side of the equatorial plane (Eq-plane). The support image for Coronidium in the Atlas is a view from the inferior side of the Eq-plane. The referable images shown in Goll (1968: pl. 175, figs 4, 5, 8, 9, pl. 176, figs 8, 10, 12; 1972a: pl. 69, fig. 3) display four basal pores in the basal ring. Four basal pores comprise relatively small twin Cerv -pores and very large twin Ca -pores. A tiny twin J -pore is placed on the lateral side of the shell. The principle of these basal pores is common for both Acanthodesmia and Dictyospyris . Campbell (1954) characterized Coronidium by four open lateral gates, Eucoronis by six large gates, absence of large basal feet, simple gates and armed rings with short thorns, and Acrocubus by lack of an equatorial ring and basal ring without feet. All these diagnoses, whoever, are helpless to characterize this genus. Eucoronis , Acrocoronis and Acrocubus were simultaneously published in Haeckel (1882: 445, 445 and 446 in ascending order). In respect to the first revision by Petrushevskaya (1971a: 267), Eucoronis is validated here. The relationships among Eucoronis , Trissocyclus and Tympanomma need more studies.

Lithocircus

Archistephus has the same type species as Archicircus . Archicircus has already been synonymized with Lithocircus by Petrushevskaya (1971a: 269), but this genus is mixed with juvenile forms of Acanthodesmia, Semantis, Tricolospyris (Acanthodesmiidae) , Zygocircus (Stephaniidae) and many genera of the Cephalospyrididae . It is practically impossible to differentiate a true Lithocircus from a young form of some Acanthodesmioidea .

Semantis

Both Semantis and Cortiniscus were classified in the “ Semantididae ” whose skeleton is composed of a vertical sagittal and a horizontal basal ring ( Campbell 1954: D106). They are mixed with not only the true Semantis and Cortiniscus but also with young forms of some Acanthodesmioidea . This definition is not based on basal pore patterns and construction of the initial spicular system. As a strict differentiation at genus level based on these characters will need more time, we simply synonymized both these genera for a practical usage. These two genera were simultaneously published in Haeckel (1887: 956 for Semantis and 963 for Cortiniscus ). Semantis is validated among them in consideration of realistic type specimen images ( Haeckel 1887: pl. 92, fig. 2).

Tricolospyris

Campbell (1954: D116) characterized Tricolospyris as “lattice complete on all sides, otherwise like Perispyris ”. The “complete lattice” is obviously the supplemental skeletal part by secondary growth mode defined inOgane et al. (2009c). The basal pore parallel to the equatorial plane is two basal pores ( Goll 1972b: pl. 1, figs 4, 5, pl. 3, fig. 1, pl. 4, fig. 3, pl. 6, fig. 4, pl. 7, fig. 4, pl. 9, fig. 12). These two basal pores are very large: the Japanese rice spatula-shaped twin Ca -pore. The large twin Cerv -pore is also visible at an oblique angle from the inferior view (the basal view). The large twin J -pore is also present at an oblique angle from the basal view. The presence of robust double l- and L -rods to form twin Ca -pores is common with Ceratospyris , suggesting phylogenetic relationships. Tricolospyris and Perispyris were simultaneously published in Haeckel (1882: 443 for both genera). Tricolospyris is selected here as a valid name because real specimens are recognized within this genus.

Trissocyclus

The combinations of Trissocyclus and Tricyclarium and that of Trissocircus and Tricircarium have respectively the same type species. The genera synonymized here are classified into the “Zygostephaninae” (Zygostephanium) with four lateral gates and the “Trissocyclinae” ( Trissocyclus , Tricyclonium , Trissocircus , Tricirconium ) with eight large gates in Campbell (1954: D108). The differences among the four genera in “Trissocyclinae” are the relative size differences of the sagittal rings (D-ring), the simplicity of the “gates” and the latticed conditions. According to Campbell (1954), Trissocyclus and Tricyclonium are different from Trissocircus and Tricirconium , the latter having simple gates. However, no obvious differences are recognizable in the typeillustrations of these four genera. Referred to the relative size differences of the sagittal ring, the combination of Trissocyclus and Trissocircus and that of Tricyclonium and Tricirconium are indicate their respective synonymy. Real specimens of Trissocyclus are commonly found but any real specimens identifiable as Tricyclonium or Tricirconium have not been encountered so far. As in previous genera, the difference between Trissocyclus and Tricyclonium being only the relative difference in size of their sagittal ring, we synonymize all these four genera until real Tricyclonium or Tricirconium representatives can be illustrated. The basal ring illustrated by Goll (1968: pl. 175, figs 1-5, 7-9) shows four pores which comprise larger rectangle twin Ca -pores and small elliptical twin Cerv -pores. J -pores are unknown. Zygostephanium is considered to have four lateral gates but not eight. We suspect an incorrect recognition of the number of gates so that this genus is also synonymized with the remaining genera until a new study can be conducted.Two oldest available names were simultaneously published in Haeckel (1882: 446 for Trissocyclus and Trissocircus ). As real specimens are found for Trissocyclus stauroporus , Trissocyclus is validated.