Rossella podagrosa Kirkpatrick, 1907

Göcke, Christian, Janussen, Dorte, Reiswig, Henry M., Jarrell, Shannon C. & Dayton, Paul K., 2015, Rossella podagrosa Kirkpatrick, 1907 — A valid species after all, Zootaxa 4021 (1), pp. 169-177: 171-175

publication ID

publication LSID

persistent identifier

treatment provided by


scientific name

Rossella podagrosa Kirkpatrick, 1907


Rossella podagrosa Kirkpatrick, 1907 

( Fig. 1–2View FIGURE 1View FIGURE 2, Tab. 1)

Rossella podagrosa Kirkpatrick, 1907: 11  , pl. 3, figs. 2–3, pl. 5, fig. 1. Topsent 1917: 14. Tabachnick 2002: 1447. Rossella podagrosa tenuis Topsent, 1916: 4  ; 1917: 15.

Not Rossella racovitzae Burton 1929: 407  –409, fig. 1, pl. 1; 1932: 256–257; 1934: 7. Koltun 1976: 165 (pars). Barthel & Tendal 1994: 92 –95, figs. 35–36, pl. 3–4. Göcke & Janussen 2013: 116 View Cited Treatment –120, fig. 2 F, 7, tab. 6).

Material. 1 Specimen (P 1635) from 30 m depth at Cape Armitage, Ross Island, McMurdo Sound, Antarctica  ; BMNH 1908.2. 5.6 (type) from Discovery Collection, Winter Quarters Bay, Hut Point, D Net, 15.12. 1902.

Description. The sponge body is of a slender, upright shape with one round osculum at the top. It usually bears numerous buds that originate from the basal region of the mother and grow from ~ 3–5 mm to> 3 cm prior to separation. The sponges show an unusual fast growth rate increasing their length by as much as 16 cm in a decade and have been shown to increase their volume by almost 300 % in just three years. Moreover, newly separated buds themselves can grow new buds within one year ( Dayton, 1979). This mode of reproduction leads to a typical appearance of dense colonies with several specimens growing close to each other ( Fig. 1View FIGURE 1 A). Colonies are often densely covered by sediment including loads of (foreign) loose spicular material, so that in many cases only the tips of the sponges with the oscules are visible within the sediment ( Fig 1View FIGURE 1 B –D). The single specimen ( Fig. 1View FIGURE 1 E) can reach a height of approximately 20 cm. The surface bears no conules, but it can be covered by a narrow veil of protruding pentactines. Protruding diactins are not prominent. The sponge has a basal root tuft of long pentactines as it is usual for the genus. The inner cavity has a dense surface; its basal part in rare cases bears large round cavernous openings. In contrast to other Rossella  species, the texture of the sponge is soft, and it is too fragile for handling without damaging it. In fact the measurements in Dayton (1979) were made by supporting the sponges with a spoon because the sponge collapses very easily.

In here we restrict the analysis of spicules ( Tab. 1, Fig. 2View FIGURE 2) to the characters of highest taxonomical importance, which means that these are the spicules with the highest significance for distinguishing R. podagrosa  from other Rossella  spp.. The spicules of the new specimen P 1635 are in full accordance with those of the type (see comparison in Fig. 2View FIGURE 2, Tab. 1) and with the typical spicule inventory of the genus. Most characteristic spicules are the calycocomes and mesodiscohexasters. Calycocomes ( Fig. 2View FIGURE 2 A, D) are about 190 to 250 µm in diameter, they have short primary rays (in mean about 10 µm) and middle pieces (mean about 11 µm), but long secondary rays (mean about 90 µm). The number of secondary rays on each ray ranges from 2 to 6, but is most commonly 4 and very rarely higher than 5. Secondary rays are almost straight, only slightly bent outwards in the proximal region. Calycocomes are densely covered by fine spines. Mesodiscohexasters ( Fig. 2View FIGURE 2 B, E) have a very typical shape; they strongly resemble the calycocomes, but are only about half the size (mean about 129 µm) and much thinner with almost absent middle pieces. Each ray bears two to four secondary rays; there are no hemi-forms with non-split rays. Mesodiscohexasters are quite rare, but occur regularly. Microdiscohexasters ( Fig. 2View FIGURE 2 C, F) have secondary rays of two different lengths and distinct disc-shaped middle-parts. They are about 30 to 55 µm in diameter. Remarkably large oxyhexactines (mean diameter about 180 µm) and oxyhexasters (mean diameter about 174 µm), commonly as mono- or dioxyhexasters, which have only one or two primary rays split into two secondary rays, occur in about the same frequency. Dermal pentactines have smooth surfaces without hooks.

Remarks. The species Rossella podagrosa  is clearly characterized by the combination of its typical habitus with numerous buds and its rather small calycocomes with normally only 4 to 5 secondary rays on each ray. Furthermore, oxyhexactines and oxyhexasters are quite large for Rossella  (see Barthel & Tendal 1992), although Göcke & Janussen (2013) reported similar sizes from several well-known species in the eastern Wedell Sea. The appearance in situ resembles that of R. antarctica  ( Fig. 1View FIGURE 1 D) in strongly sediment-covered specimens. A clear character for differentiation in situ is that R. podagrosa  usually grows partly within the sediment, while R. antarctica  occurs on the sediment. Also, R. podagrosa  has a more elongate body form, whereas that of R. antarctica  is strongly rounded. In terms of spicules, the difference is clear: R. antarctica  has dermal pentactines with strong lateral hooks (as shown by Carter 1872) and much smaller calycocomes with different shape; primary rays and middle pieces are much more prominent and the number of secondary rays per ray is higher (compare Schulze & Kirkpatrick 1910; Barthel & Tendal 1992; Göcke & Janussen 2013). Rossella racovitzae  differs from R. podagrosa  by a larger body size, lack of buds and the presence of clear conules on the surface. Furthermore, it has much bigger calycocomes (400 µm) with a higher number of secondary rays (see Topsent 1901; Göcke & Janussen 2013). Similar sizes of calycocomes are found in R. levis  , which nonetheless differs from R. podagrosa  in some key points: it has a surface covered by numerous and very distinct conules (compare Fig. 1View FIGURE 1 C), which in the lower part of the body bear bundles of protruding diactines. Calycocomes have considerably larger middle pieces and primary rays, and secondary rays have a stronger bending than in R. podagrosa  . Microdiscohexasters have secondary rays of only one length without differentiated middle pieces (compare Kirkpatrick 1907; Göcke & Janussen 2013). Calycocomes with low numbers of secondary rays are also found in R. nuda  . Compared to those of R. podagrosa  these are almost twice as big. Also, the number of secondary rays never exceeds 4, which is lower than in R. podagrosa  (see Göcke & Janussen 2013). The habitus differs as well: R. nuda  is very smooth and appears always very clear and visible in underwater-photographs, whereas R. podagrosa  is usually sediment-covered. Rossella nuda  has been reported to grow buds of similar shape as those of R. podagrosa (Barthel & Tendal 1992)  , but they are much more rare and never form such dense aggregations of sponges as shown in Fig. 1View FIGURE 1 A for R. podagrosa  .

The long and complicated taxonomic history of the genus Rossella  is summarized in the introduction. It can be brought down to two basic attempts of synonymies, which attempted to bring some order into the high amount of often similar species ascribed to the genus: The first was done by Burton (1929) and later continued by Koltun (1976). The second attempt was done by Barthel & Tendal (1992), which was followed by Göcke & Janussen (2013). It is now obvious that both concepts contain basic systematic weaknesses that compromise the results. Burton (1929) used an attempt that was basically concentrating on the external shape and neglected most spicule information which he discounted as being of minor importance. Furthermore he was led by a strong will to reduce the number of species and therefore would rather join two species than separate them if synonymy was questionable. Koltun (1976) obviously had the same will to reduce species numbers, and was furthermore confused when he re-introduced spicule information into Burton’s spicule-less system. As Burton joined species with clearly different spicule inventory, e.g. R. levis  and R. nuda  , under one taxon name, the resulting “species” could not have a distinct characteristic spicule inventory. Koltun realized this, but did not draw the conclusion that different species were joined under one taxon name. Instead he continued synonymizing until he found a working species concept, even though this lumped together many clearly recognizably different forms/species. Barthel & Tendal (1992) realized these problems and tried to establish a new system based on spicules that included habitus information. Thus, they resurrected several former species; however, they based their work mainly on the material sampled during the EPOS-expeditions of RV Polarstern in the Weddell Sea (1989), without much consideration of the type specimens or material from other parts of the Antarctic. Also, they did not put too much emphasis on the habitus information, probably because their material was trawled and therefore delicate specimens as R. podagrosa  were not recovered in a condition that allowed for much habitus (or in situ) information. For this reason they ended up with a somewhat narrow perspective of the diversity of Rossella  which e. g. did not incorporate R. podagrosa  , which they still considered a synonym of R. racovitzae  . However, Barthel & Tendal perceptively reported a strongly bud-producing variant of R. racovitzae  that they state might be turned into a species at some point. The analysis of several Rossella  species in Göcke & Janussen (2013) was based only on Weddell Sea material, just as the study of Barthel & Tendal (1992). Here we include material from the Ross Sea showing that R. podagrosa  is a well-defined species when both habitus and spicule information are carefully considered. It is clear that a proper taxonomical analysis of Rossella  should include as much material as possible from all around the Antarctic Ocean. More importantly, all information, such as habitus (best alive in situ) spicule inventory and molecular data (if available), must be included. Rossella  once more has proven to be a complicated genus, more than previously thought, and it is still a long way to go to form a proper concept of valid Rossella  species, including a full revision of holotypes (currently underway), analysis of further material from the whole southern ocean, and detailed molecular analysis.

Budding Rossella  are not a new phenomenon; they have been reported several times, not only from the Ross Sea, but also from the Weddell Sea ( Gutt & Piepenburg 2003; Teixido et al. 2006). Unfortunately, the original material of these studies could not be directly sampled, so it was always questionable which species really does this remarkable budding. As we have now re-established R. podagrosa  and analyzed its budding habit, we come to the assumption, that in many of these cases, it might actually be the respective species, although so far not identified with its proper name due to the complicated synonymy history as shown above. Further inspection is needed to analyze its distribution and see how widespread the species really is, but it may be quite common around Antarctica  , although it probably has its main occurrence range in the Ross Sea.

The mode of life of R. podagrosa  is quite unique, as all other Rossella  spp. live on, but never within, the sediment. This difference is well visible in Figs. 1View FIGURE 1 C and D where R. podagrosa  is visible only very slightly by its oscular regions sticking out of the mud, while the more typical Rossella  appearance is visible in the adjacent large specimens of R. cf. levis  and R. antarctica  . Rossella podagrosa  nonetheless clearly covers most ground in these pictures and is certainly the most abundant species in this area of the Ross Sea. In general, sponges consist of a complicated filtering apparatus with thin channels traversing its tissue, including chambers of flagellate cells which produce a strong current from which food particles are taken in and oscules just release the filtered water (van Soest et al. 2012). It is easy to imagine that such a delicate filtering system can easily be clogged by excessive sediment. In addition, some sponges even possess the capacity of self-cleaning, and some are able to shed sediment off their surfaces ( Barthel & Wolfrath 1989; Bell, 2004; Schönberg & Suwa 2007). In her study on spongeassociated invertebrate faunas, Kunzmann (1996) showed that sponges with many protruding spicules and especially R. antarctica  with its thick veil of pentactines shelter many animals from protozoons to large polychaetes and echinoderms. They also catch certain amounts of sediment. This is visible in Fig. 1View FIGURE 1 D as well, where the specimen of R. antarctica  , although growing clearly above the ground, is heavily covered by sediment and inhabiting invertebrates such as brittle stars. It can be assumed that these associations with presumably very complicated trophic interactions function as “microbial gardens”, in which bacteria settle and form the basic nourishment for the sponges. If this is true, similar processes may be assumed for R. podagrosa  within the ground. Still, to test this hypothesis we will need much further work on microbes as well as on sponge-associated invertebrates. Another hypothesis is that the half-buried way of living might serve as a kind of hide-out for the sponge saving it from predators like Acodontaster conspicuus  as proposed by Dayton (1979). If this assumption is true, it might also serve as an explanation for the very high abundances of this species, in combination with its high growth rates.

Conclusions. It is evident that Rossella podagrosa Kirkpatrick, 1907  is a valid species. It can be clearly differentiated from other Rossella  spp. Furthermore we show that a thorough revision of the genus Rossella  will need detailed analysis of specimens from all areas of the Antarctic, and that it requires close examination of the habitus (at best based on in situ examination) as well as spicule analysis.

TABLE 1. Spicule sizes of Rossella podagrosa Kirkpatrick, 1907. Values in μm are given as follows: minimum - mean - maximum (number of spicules measured).

    BMNH 1908.2.5.6 (type)
  130-173.3-230 (30)  














Rossella podagrosa Kirkpatrick, 1907

Göcke, Christian, Janussen, Dorte, Reiswig, Henry M., Jarrell, Shannon C. & Dayton, Paul K. 2015

Rossella racovitzae

Gocke 2013: 116
Barthel 1994: 92
Koltun 1976: 165
Burton 1929: 407

Rossella podagrosa

Tabachnick 2002: 1447
Topsent 1917: 14
Topsent 1916: 4
Kirkpatrick 1907: 11