Pachastrella nodulosa Cárdenas & Rapp, 2012, sp. nov.

Pachastrella nodulosa sp. nov.

(Figures 4–10, Table 3)

Synonymy.

Pachastrella monilifera: Burdon-Jones & Tambs-Lyche 1960, p. 6; Koltun 1966, p. 30. Pachastrella sp.: Cárdenas et al. 2011, Table S1.

Not Pachastrella monilifera Schmidt, 1868: Schmidt, 1868, p. 15.

Type series. ZMBN 85242, holotype, Steinneset, Langenuen, western Norway, 59°53'N, 05°31'E, depth: 25–175 m, triangular dredge; ZMBN 85227, 85243, paratypes, Skorpeodden, Korsfjord, western Norway, 60°10'N, 05°10'E, depth: 200–400 m, triangular dredge; ZMBN 85244, paratype, Røst Reef, northern Norway, 67º30'N, 9º24'E, depth: 282–290 m, manned-submersible; ZMBN 85245, paratype, Traenadjupet, northern Norway, 66º58'N, 11º7'E, depth: 316 m, manned-submersible.

Additional material examined. NTNU-VM 54998, northern Norway.

Comparative material examined.

Pachastrella monilifera, MNHN-DT410, holotype, coast of Algiers.

Pachastrella abyssi Schmidt, 1870, MZS P0195, holotype, dry, Florida; Pachastrella cf. abyssi, HBOI 198811141008, off Savannah, Georgia, U.S.A., 31°41'N, 79°08'W, 533 m, det. P. Cárdenas; ZMAPOR 0 5301, dry, off Payne’s Bay, Barbados, 153 m (van Soest & Stentoft 1988).

Pachastrella ovisternata von Lendenfeld, 1894, MNHN-DCL4065, Portugal, det: M. Maldonado.

Pachastrella sp. 1, originally identified as P. monilifera (Lévi 1967), MNHN-DCL680L, 681, 691, slides, South Africa, 47–130 m, det. P. Cárdenas.

Pachastrella sp. 2, originally identified as P. monilifera (Lévi & Lévi 1989), MNHN-DCL3372, slides, off Lubang Island, Philippines, MUSORSTOM 1, 14°01'N, 120°16'E, 183–185 m, det. P. Cárdenas.

Outer morphology (Fig. 4). Massive irregular or cup-shaped (ZMBN 85242, ZMBN 85243). Large specimens (ca 20 cm high and 25 cm wide) observed with the manned-submersible were usually cup (Fig. 4 G–H) or half-cup shaped with a deep or shallow central depression. Surface and choanosome color is whitish or yellowish-cream to dark brown (alive and in ethanol). Specimen is slightly compressible. Surface is very rugose, as sandpaper. Choanosome is dense and also rugose (due to the abundance of calthrops). Upper and side surfaces around the central depression are knobby (each knob with a diameter of 0.5–1 cm) while the lower surface is more irregular. Small oscules are grouped in the oval to circular depressions on the upper surface (Fig. 4 F), where knobs are absent. Each oscule has a diameter of 0.5–1.5 mm. Pores are not visible. Cortex is not conspicuous. Macroepibionts include bivalves, polychaetes, ascidians, hydroids, sea anemones, pycnogonids, and other sponges (e.g. Cyamon spinispinosum (Topsent, 1904), Hexadella dedritifera).

Skeleton (ZMBN 85242) (Fig. 4 E). Thin cortex of microstrongyles (36–73 µm thick). Calthrops (with no particular orientation), microstrongyles and abundant amphiasters fill the choanosome. Microstrongyles are also found lining the canals. Foreign material and spicules are abundant (e.g. sigmas, discorhabds and tylotes).

Spicules (ZMBN 85242) (Fig. 5). (a) calthrops, few dichotomous actines, actine length: 47- 244.3 -768 µm; actine width: 4- 28.7 -73 µm. (b) microxeas, rare, smooth length: 44.8- 91.1 -190 µm; width: 0.7- 2.0 -2.5 µm (N=26). The longest of these microxeas (109–190 µm) seemed thinner (0.7–0.9 µm), almost whip-like and may represent a seperate category of microxeas, or be foreign, only six were found. (c) microstrongyles, in high numbers, microspiny, straight, centrotylote or not, length: 13.6- 15.7 -26 µm; width: 3.1- 4.7 -5.8 µm. (d) microrhabdose streptasters, rare, straight or slightly bent, sometimes centrotylote, length: 23- 30.3 -34.7 µm; width: 1.4- 2.1 -2.6 µm (N=7). (e) amphiasters, ca. 5-11 actines (usually 5 actines on each end of the naked shaft, actines often have a tip with two spines), microspiny, length: 10.8- 14.8 -22.6 µm; width: 7- 11.5 -17.5 µm.

Distribution (Fig. 6). Norwegian Sea and Norwegian coast (Burdon-Jones & Tambs-Lyche, 1960; Koltun, 1966; this study).

Depth. 25– 400 m.

Discussion. P. nodulosa is a fairly common species in Norway. During a three-hour dive in Traenadjupet (northern Norway) with the manned-submersible JAGO, we spotted more than a dozen large specimens (Fig. 4 G– H). Pachastrella species can be difficult to discriminate and since P. monilifera was the first species of this genus to be described, many records from the Atlantic and Pacific have excessively used that name. Kirkpatrick (1902) and von Lendenfeld (1907) were among the first authors to discriminate new species from the P. monilifera complex. According to the latest revision (Maldonado 2002), there are three Pachastrella species in the NEA: P. monilifera, P. o v i s t e r n a t a and P. chuni. Upon collection, the Norwegian specimens were initially identified as P. monilifera, following previous Norwegian records (Burdon-Jones & Tambs-Lyche 1960; Koltun 1966). However, P. nodulosa differs from the P. monilifera holotype (Fig. 7 A–B) by (i) its knobby surface, (ii) the absence of large oxeas (noticed by Koltun (1966)), (iii) its amphiasters have on average less actines and these are less spiny, and finally (iv) differs morphology and rareness of the microxeas. Fairly common in P. nodulosa, microrhabdose streptasters were not found in the holotype of P. monilifera. This was not considered to be a difference since they can be quite rare, and they have been observed in other Mediterranean specimens (Table 3). P. nodulosa is also different from P. ovisternata (Fig. 7 C–D) and P. chuni for the same reasons as for P. monilifera, and also because of the absence of small dichotriaenes (P. o v i s t e r n a t a and P. chuni) and mesodichotriaenes (P. o v i s t e r n a t a) in the choanosome. We however note that a knobby surface has also been observed in P. o vi s te r n at a (M. Maldonado, pers. com.) so this character may not be unique to P. nodulosa. Interestingly, the morphologies of the largest category of calthrops were somewhat characteristic in each NEA species: often plagiotriaene-like with stout regular actines in P. monilifera (Fig. 10 A); plagiotriaene-like or not, with thinner regular actines in P. nodulosa (Fig. 10 B); and often irregular actines in P. ovisternata (Fig. 10 C) or P. chuni. The morphology of the large calthrops seems to be consistent: large calthrops from drawings of Mediterranean P. monilifera are very similar to what we observed in the type (Topsent 1894; Babiç 1922). The last species of the Atlanto-Mediterranean area, Pachastrella echinorhabda Pulitzer-Finali, 1972 from the Mediterranean, has characteristic acanthorhabds, absent in all other Atlantic species, including P. nodulosa.

... Continued on next page Pachastrella sp. 183- Cup 2000-3500/ yes 80-1500/ - - 15/ - 15 Philippines 185 50-100 40-200 7 MNHN-DCL3372

(Lévi & Lévi 1989)

spicules measured for this study.

spicules from ZMAPOR 0 5301 measured for this study

By identifying P. nodulosa, our study further confirms the existence of a P. monilifera species complex and thereby shows how necessary a comprehensive revision of P. monilifera and its many synonyms is. This is beyond this study, but our observations have nonetheless led to a few taxonomical conclusions that could initiate this sizeable work. As shown above, skeleton organization and large calthrops morphology has been shown to be useful to discriminate species, we recommend their usage to reassess the status of Pachastrella species.

P. abyssi from Florida has been synonymized with P. monilifera (Topsent 1894), but occurrence of the same sponge species on both sides of the Atlantic was considered suspicious, so we re-examined the holotype (Figs. 7 E- F; 9). It has a plate-like shape (Fig. 7 E), similar to the shape of P. monilifera (Bo et al. 2012). But the microscleres observed in the holotype of P. abyssi (Fig. 9 B-I) are quite different from the ones of P. monilifera and P. nodulosa. First of all, we observed a few meso/dichotriaenes in our spicule preparations of the holotype (rhabdome: 32- 50.5 - 75 x 4 - 7.7 -10 µm (N=4); protoclade + deuteroclade: 20- 31.7 -55 µm + 15- 27.3 -55 µm (N=17)), similar in size and shape to the ones found in P. o v i s t e r n a t a. These meso/dichotriaenes are absent in P. monilifera. Furthermore, compared to P. monilifera, the microstrongyles in P. abyssi are more “chubby” and consequently less centrotylote; some can also be fairly elongated, but they are not common (Fig. 9 G). The amphiasters have on average more actines, some of which can even be placed on the shaft (which is usually naked in P. monilifera and P. nodulosa); the shaft can occasionally be bent and therefore some of these strepasters could be called spirasters (Fig. 9 D-E). We also noticed some very small sanidaster-like amphiasters with numerous actines (Fig. 9 F); these are not common but they were never observed in the other Pachastrella species examined. Moreover, in the holotype of P. abyssi (Fig. 9 B-C) microxeas are clearly thicker (Table 3) than in P. monilifera (Fig. 8 B–C). We could also distinguish two size classes of microxeas in P. monilifera, against only one in P. abyssi, P. chuni and P. nodulosa. These microxeas are common in P. abyssi, P. c h un i and P. monilifera but rare (or even foreign?) in P. nodulosa. Microrhabdose streptasters are common in P. abyssi and P. nodulosa, they are rare in P. monilifera. Finally, the large calthrops are regular, rarely bent, and very stout which makes them different from those of the other species (Fig. 10). It is therefore quite clear that the holotype of P. abyssi is different from P. monilifera, P. chuni and P. nodulosa: it should be considered as a valid species. However, P. abyssi remains a poorly known species and after examining two other Pachastrella cf. abyssi specimens from the North-western Atlantic we are still unsure about the relevance and variability of some characters observed in the holotype of P. abyssi. The Barbados specimen of P. cf. abyssi (ZMAPOR 0531) matched the holotype, except for four observations. The microxeas, which had been overlooked by van Soest & Stentoft (1988), are slightly thinner than in the holotype (Table 3). The microstrongyles are smaller (Table 3), not “chubby” and consequently more centrotylote. We could not find meso/dichotriaenes, but even in P. ovisternata, these can be absent from certain parts of the sponge. We also noticed that ZMAPOR 0 531 had only large styles and strongyles instead of oxeas; but this may be just intra-specific variation since the large calthrops often had blunted actines (often shortened) and since van Soest & Stentoft (1988) observed oxeas in other Barbados specimens. The Georgia specimen (HBOI 198811141008) shared most characters with P. abyssi including the presence of spirasters and sanidaster-like streptasters. But it also differed from the holotype regarding the microxeas: it had smaller microxeas that are centrotylote and rough (vs. larger smooth non-centrotylote in holotype). It seems like an important difference but we have examined too few specimens to conclude whether HBOI 198811141008 could be a different species at this point. We could not find meso/dichotriaenes in HBOI 198811141008 either.

We have examined additional material identified as P. monilifera, from South Africa (MNHN-DCL680, 681, 691). Once again, we found it slightly different from all the other Atlantic Pachastrella. It is characterized by i) large calthrops with frequent bent actine tips, ii) smaller and more compact microstrongyles. This material should be carefully compared with the holotype of Pachastrella caliculata Kirkpatrick, 1902 (a synonym of P. monilifera) in order to eventually propose the resurrection of this South African species. We have also re-examined P. monilifera from the Philippines (MNHN-DCL3372, 3373). In our opinion, i) the very large oxeas and calthrops (with irregular branching actines) and ii) the small dichotriaenes (as in P. c h u n i and P. ovisternata) clearly discriminate it from P. monilifera. Further studies are needed to say if it represents a new species.

Many species described in the Pachastrella genus actually lack the characteristic microstrongyles and amphiasters, which make us believe they should be moved to the genus Poecillastra: Pachastrella cribrum Lebwohl, 1914, Pachastrella fusca Lebwohl, 1914, Pachastrella incrustata Bergquist, 1968, Pachastrella scrobiculosa Lebwohl, 1914. A re-examination of the type of Pachastrella multipora Dickinson, 1945 will be necessary to take a decision, notably because this species has lost its monaxonic microscleres. Southern Atlantic and Antarctic specimens, until now also identified as P. monilifera, are also in need of a revision to confirm their status (Table 3). All in all, we consider seven species of Pachastrella to be valid (P. abyssi, P. chuni, P. echinorhabda, P. monilifera, P. nodulosa sp. nov., P. ovisternata, and P. loricata), while P. multipora and several Pachastrella sp. await confirmation of their status (Table 3). These results illustrate once more how thorough reexamination of a cosmopolitan species (P. monilifera) can reveal a species complex. Following this, further identifications and synonyms of P. monilifera (and P. ovisternata) from remote geographical locations (especially from the Indian and Pacific Ocean) should be considered with caution.

Etymology. From the latin word ‘ nodulosa ’ meaning small knot or small nodule. It refers to the knobby surface of this species.