An extraordinary new fluvial bdelloid rotifer, Coronistomus impossibilis gen. nov. sp. nov., with adaptations for turbulent flow (Rotifera: Bdelloidea: Coronistomidae fam. nov.) Author Örstan, Aydin text Zootaxa 2021 2021-04-30 4966 1 16 28 journal article 6541 10.11646/zootaxa.4966.1.2 1c5c6755-0fec-4bfe-b6b7-2653e4a30a97 1175-5326 4736208 953EBF83-58FB-4BA6-BFD6-76DCE690DA30 Coronistomus impossibilis sp. nov. ( Figs. 2 -6,7A,B,8A,9-11) Material . This description was based on 25 specimens constituting the type series. The holotype ( Fig. 3 ), mounted in glycerine jelly between two cover glasses and secured in a Cobb’s slide holder, has been deposited in the Academy of Natural Sciences , Philadelphia, PA , U.S.A. ( ANSP 2134 ). Twelve paratypes (whole preserved specimens or extracted trophi) are in the collection of the author identified with the private morphovariant and locality designation A1HR2. All photographs in this description are those of the types . FIGURE 2 : Coronistomus impossibilis gen. nov., sp. nov. , composite drawing of feeding animal from the ventral side based on photographs. Only selected features of the internal anatomy are shown. FIGURE 3 : Coronistomus impossibilis gen. nov., sp. nov. , preserved holotype with open corona-mouth (shrunken to ~65% of live size). A, dorsal aspect. B, ventral aspect. Each image is a stack of three photographs. ca: caudal appendage; cn: corona; rm: rostrum; vmg: ventromedial gland. FIGURE 4 : Coronistomus impossibilis gen. nov., sp. nov. , live specimen with corona withdrawn, slightly compressed. A, entire animal crawling. B, details of anterior body. ca: caudal appendage; lp: lappet; vmg: ventromedial gland. Type locality . An approximately 850-m stretch of the Hawlings River (between the spots 39.2119 N , 77.0846 W and 39.2121 N , 77.0760 W ) within the Rachel Carson Conservation Park , Montgomery County , Maryland , USA . Most of the specimens were from the submerged surfaces of a large rock ( 39.2122 N , 77.0810 W ) in the river ( Fig. 1 ) . Diagnosis . Same as for genus with pair of red cerebral eye spots, each trophi half with three major teeth interspaced with two interproximal teeth; integument papillated; body length exceeds 550 µm ( Figs. 2 –6,8A,9,10). Etymology . The specific name of the new species, from Latin impossibilis , was suggested by the remark of my friend and fellow rotiferologist Nataliia Iakovenko that this was “an impossible creature”, which she expressed upon seeing the photographs of the species for the first time. Full description of morphology. Coronistomus impossibilis is a relatively large rotifer that exceeds 550 µm in creeping length. During creeping the combined length of the rump plus the foot is about a third of the total body length ( Fig. 4A ). The lamella on the anterior dorsal rostrum forms a pair of small, semicircular lateral lappets covering tufts of short cilia that are on the ventral side. The lappets are connected medially by a narrow strip of the lamella that occasionally appears bilobed ( Fig. 4 ). The antenna is slightly shorter than the lateral width of the neck and has two pseudosegments; the proximal pseudosegment covers ~3/4 of the total length of the antenna ( Fig. 5A ). There is a pair of red eye spots in the posterior brain. The eye spots are flat oval disks that face the anterolateral side of the brain ( Fig. 5A ) and, as a result of their orientation, appear oblique and narrow from the dorsal side ( Fig. 5B ). FIGURE 5 : Coronistomus impossibilis gen. nov., sp. nov. , eyespots and other details of live specimens. A, holotype, lateral. B, paratype, dorsal. es: eyespot; lf: longitudinal fold; lp: lappet; pp: papillae. The corona lacks pedicels and forms a weakly bilobed ciliated field on the ventral surface of the head ( Figs. 2 , 3 , 6 ). In one specimen with a total body length of 510 µm , the width of the open corona was 77 µm . There are roughly semicircular ridges near the anterior border of the corona ( Fig. 6C ). The lateral borders of the corona continue into the mouth ( Figs. 6A, B ). The open corona and the mouth are surrounded by an almost circular rim ( Fig. 7A ). The encircling structure within this circular rim is interpreted as a ring-muscle ( Fig. 7B ). The curved thickening of the integument forming the anterior dorsal border of this rim is the upper lip ( Fig. 6D ). The rostrum extends over the open corona and blocks the view of the upper lip from the dorsal side ( Figs. 2 , 7A ). FIGURE 6 : Coronistomus impossibilis gen. nov., sp. nov. , preserved holotype. A-E, focus series of open corona-mouth moving from ventral side towards dorsal side. F, diagrammatic representation of primary features of head. All to the same scale. cn: corona; ll: lower lip; mt: mouth; rm: rostrum; sr: semicircular ridge; ul: upper lip. FIGURE 7 : Coronistomus impossibilis gen. nov., sp. nov. compared with Adineta beysunae Örstan. A , live specimen showing the circular rim of integument surrounding the corona-mouth. B, same specimen preserved. C, A. beysunae , slightly compressed live specimen. All to the same scale. cr: circular rim; re: ring-muscle; rm: rostrum. The trophi are ramate and have semicircular continuous manubria. On each half (uncus) there are three widely spaced major teeth interspaced with two interproximal teeth and numerous minor teeth outside them ( Fig. 8A ). In 11 trophi that were examined (either extracted or in compressed specimens), there was no variation in the number of major and interproximal teeth. The trophi in Fig. 8A had seven anterior and 12 posterior minor teeth on one half and six anterior and 13 posterior minor teeth on the other. On two sets of trophi, including the one in Fig. 8A , one posterior minor tooth on each half appeared slightly more prominent than the others. The trophi are not protrusible. FIGURE 8 : Bdelloid trophi compared. A, Coronistomus impossibilis gen. nov., sp. nov. B, Philodina sp. C , Abrochtha sp. ipt: interproximal teeth. FIGURE 9 : Coronistomus impossibilis gen. nov., sp. nov. A, preserved specimen (stained with neutral red), dorsal aspect. B, egg of the holotype with attached debris (stack of four photographs; brightfield). C, live specimen, lateral aspect of posterior body. All to the same scale. as: anal swelling; ca: caudal appendage; dtg: dorsotransverse gland; vm: vitellarium. The integument all over the body is covered with small papillae that are most prominent around the trunk where the integument also forms narrowly spaced longitudinal folds ( Fig. 5B ). The largest organ in the neck is the dorsotransverse gland located dorsal to the anterior stomach. This gland has two lateral lobes connected by a narrower median section over the stomach ( Fig. 9A ). The most prominent gland on the ventral side is the ventromedial gland positioned slightly posterior to the anterior stomach. It consists of a circular posterior part and an elongated anterior part extending to the stomach entrance ( Fig. 4B ). The stomach has a ciliated lumen that passes into the intestine through a sphincter. There is a pair of vitellaria each with eight nuclei. Two eggs deposited by different individuals were examined: they were ellipsoid in shape and yellow-orange in color; one had smooth poles, while the other had a small nipple at one of its poles. Both eggs had bits of debris stuck on them ( Fig. 9B ). The rump and the foot are relatively thick ( Figs. 3 , 4A ). A slight swelling over the anal opening is often noticeable ( Fig. 9C ). The transverse folds of the integument of the foot are not distinct, but there appear to be two pseudosegments prior to the one carrying the toes. There are two pairs of longitudinal gland complexes in the foot ( Fig. 10A ). Each dorsolateral gland complex consists of a set of five cells aligned along each side of the foot, while each ventromedial complex has one anterior elongated syncytial mass and a large posterior cell ( Fig. 10A ). These gland complexes and the ducts that exit from their posterior ends take up most of the space inside the extended foot ( Fig. 10A, D ). FIGURE 10 : Coronistomus impossibilis gen. nov., sp. nov. , details of foot anatomy. A, preserved compressed specimen, foot. B, live specimen, toes. C, live specimen, caudal appendage with two pairs of ducts (a, a’, b, b’) and a pair of muscles (c, c’). D, live specimen, foot. an: attachment nib; ca: caudal appendage; dlc: dorsolateral gland complex; vmc: ventromedial gland complex. There is a pair of long, thick ventral toes. Each toe consists of a long proximal and a shorter distal pseudosegment ( Fig. 3B , 10B ) and appears to have two attachment nibs at its tip. There are no paired spurs present in most bdelloid species. The foot ends with a prominent caudal appendage positioned medially and extending posteriorly ( Figs. 2 , 4A , 9C , 10A, C, D ). When viewed from the dorsal side, the caudal appendage presents an outline that starts from a wide base and narrows to a blunt tip ( Fig. 4A ), but from the lateral side the outline of the appendage has a broad posterior end ( Fig 9C ). The caudal appendage has a pair of attachment nibs slightly ventral to its tip ( Fig. 10C, D ). Up to six longitudinal duct-like structures can be seen inside the caudal appendage in photographs ( Fig. 10C ). Four of these (two per attachment nib) are assumed to be ducts from the gland complexes and the remaining two are interpreted as muscle fibers inserting at the tip of the caudal appendage. Despite the presence of these muscles, the caudal appendage always remains out during crawling unlike the toes, which can be retracted into the foot as in other bdelloids. In all specimens examined or photographed alive the stomach was yellow and occasionally orange ( Figs. 4A , 5 ). When the intestine was full, the material in it usually had a dark reddish color. The vitellaria were usually yellow and sometimes had orange inclusions. Developing and deposited eggs were yellow-orange ( Figs. 4 , 9B ). In most specimens the corona was of a variable yellow color that became quite intense in a few specimens (Supplemental data). In all specimens the rostrum, brain and the foot were colorless; only in one specimen that had an intense yellow corona, did the rostrum have a light yellow tint. Dimensions . Holotype : total length, 530 µm ; ratio of length of rump plus foot to total length, 0.29; antenna length, 32 µm ; ratio of antenna length to head thickness, 0.65; caudal appendage length, 30 µm ; dimensions of one egg , 76 x 55 µm . Collective dimensions: total length (n = 15; undetermined ages): 370–566 µm ; three longest specimens: 530 µm , 533 µm and 566 µm . Mean ratio of length of rump plus foot to total length (n = 9): 0.31. Mean ratio of inter-eye distance to brain width (n = 2): 0.42. Mean antenna length (n = 4 specimens with total length> 500 µm ): 36.3 µm ; mean ratio of antenna length to head thickness: 0.75. Mean caudal appendage length (n = 4 specimens with total length> 500 µm ): 28.8 µm . Mean trophi length (n = 9): 20.5 µm . Mean egg dimensions (n = 2): 83 x 57 µm . Descriptions of habitat and behavior . All specimens of C. impossibilis were found in samples taken from submerged surfaces of rocks that were no more than about 20 cm deep in areas of the river where the water flow was fast and turbulent (Supplemental data); several samples were in fact from rock surfaces facing upstream. The samples consisted of water and fine sediment and sometimes contained parts of aquatic moss and filamentous algae. In the vicinity of the main collection location ( Fig. 1 ), the highest flow velocity of the river was 0.65 m /s in September 2019 and the mean of the measurements taken in September 2019 and August 2020 was 0.58 m /s. The collective information indicates that C. impossibilis lives in the sediment that accumulates on the submerged surfaces of rocks exposed to a relatively fast flow velocity. Although the water level in the river may fluctuate depending on the amount of precipitation, there is no season when the water flow ceases. The lowest and the highest water temperatures recorded between 30 May 2019 and 12 December 2020 were 5.7°C on 3 February 2020 and 21.8°C on 26 August 2020 , respectively. These numbers constitute the approximate annual water temperature range in the habitat of C. impossibilis . I never saw C. impossibilis swim in the glass dishes into which the samples were transferred in the laboratory or in temporary aqueous mounts used for microscopic examination. On glass surfaces C. impossibilis feeds by rapidly sweeping the surface with its corona back and forth, similar to the feeding of Adineta species (Supplemental data). It crawls at a moderate speed in the ordinary leech-like fashion of bdelloids and often stops and explores its surroundings by moving its body around in restless, jerky movements while keeping its toes and the caudal appendage fixed on the glass. During feeding and creeping, C. impossibilis uses its toes and the caudal appendage attached to the surface as a pivot and can rotate its extended body up to an angle of 90° or slightly more on either side of this pivot ( Fig. 11 ). Individuals that detach from a surface often turn sideways and start to rapidly fold and unfold their bodies; this behavior continues for several minutes. It is not known whether they would exhibit this behavior in their natural environment where they would probably be surrounded by packed sedimentary debris. Remarks on taxonomic placement and morphology . The presence of two belts of cilia surrounding the coronas of rotifers has been known at least since the 1870s. The belt that encircles the disks of the corona has been called either the trochus ( Cubitt 1872 ) or the principal wreath ( Hudson & Gosse 1889 ), while the slightly posterior belt, primarily on the ventral side of the head in bdelloids, is called either the cingulum ( Cubitt 1872 ) or the secondary wreath ( Hudson & Gosse 1889 ). Additionally, a finely ciliated field covers the ventral surfaces of the pedicels and extends into the buccal funnel ( Hartog 1896 ; Bartoš 1951 ). The entire ventral surface of the head within the corona of C. impossibilis is ciliated ( Fig. 6A ). Because of their location, the semicircular ridges near the anterior border of the corona ( Fig. 6C ) appear to be homologous with the peripheries of the disks of the typical corona. Therefore, the cilia on them constitute the trochus. In bdelloid rotifers with the typical retractable corona, the cingulum runs around the bases of the pedicels and continues into the mouth ( Bartoš 1951 ). Although there are no pedicels in C. impossibilis , the ciliated lateral borders of the corona continue posteriorly to form the buccal funnel that leads into its mouth ( Fig. 6B ). Therefore, these ciliated lateral borders of the corona of C. impossibilis appear to be homologous with the cingulum of the species with the typical corona. Hereinafter, the anatomically united corona and mouth of C. impossibilis will be referred to as the corona-mouth. If my interpretation is correct, the single belt of cilia surrounding the corona-mouth of C. impossibilis represents the union of the cingulum and the trochus. When the corona-mouth of C. impossibilis is closed and withdrawn into the head, the integument folds over itself to create a puckered structure encircling the closed mouth as in bdelloids with the typical corona on pedicels ( Örstan 2020 ). During the opening of the corona-mouth the folds of the integument move out of the way to form the circular rim surrounding the corona-mouth ( Fig. 7A ). The anterior dorsal border of this circular rim is considered to be homologous with the dorsal structure called the upper lip in species that have the corona on pedicels ( Fig. 6D ) and the ventral border posterior to the mouth may be referred to as the lower lip ( Fig. 6A ). FIGURE 11 : Coronistomus impossibilis gen. nov., sp. nov. , caudal appendage used as a pivot during crawling. A, body and and caudal appendage aligned. B, body turned 90° with respect to caudal appendage. ca: caudal appendage; rm: rostrum. For the familial placement of C. impossibilis , I will consider three bdelloid families whose members have stomachs with a lumen: the Philodinidae , Adinetidae and Philodinavidae . The characteristic philodinid corona consists of two retractable ciliated disks on pedicels. When open, the corona becomes the front end of the animal, while the mouth is located posterior to the corona on the ventral side of the head. In the Adinetidae , the corona is in the form of a flat ciliated surface covering most of the ventral head without a distinct division into halves. The adinetid corona is not retractable into the head, but it can be withdrawn into the body when the entire head is withdrawn. In comparison, the corona of C. impossibilis lacks pedicels and consists of a ciliated field on the ventral surface of the head, which is somewhat similar to the adinetid corona. A ring muscle similar to the one encircling the head of C. impossibilis ( Fig. 7B ) is also present in the head of Adineta ( Leasi & Ricci 2010 ; Fig. 7C ). However, unlike the adinetid corona, the corona of C. impossibilis is clearly, if weakly, divided into two halves and is retractable into the head ( Fig. 4 ). These dissimilarities preclude the placement of C. impossibilis either in the Philodinidae or in the Adinetidae . The only other family that may accommodate C. impossibilis is the Philodinavidae , which has so far included nine species in three genera: Abrochtha , Henoceros and Philodinavus . Among the characteristics of these species are their modified coronas. In Abrochtha small pedicels support a U-shaped trochus posterior to which lateral ciliated ridges forming the cinculum lead into the mouth ( Melone & Ricci 1995 ). In Henoceros and Philodinavus the corona consists of a very small ciliary field without pedicels. The ciliary field comprising the corona of C. impossibilis also lacks pedicels but is much larger than the coronas of Henoceros and Philodinavus . Another characteristic of the philodinavid species is their protrusible trophi on which the major teeth are crowded in the anterior half and there is no anterior connection between the manubrium and the ramus ( Fig. 8C ). In contrast, the morphology of the trophi of C. impossibilis ( Fig. 8A ) is more like that of the trophi of non-philodinavid species with more centrally located major teeth and semicircular manubria except that the gaps between the major teeth are wider and the interproximal teeth are more distinct ( Fig. 8B ). Moreover, the videos of feeding C. impossibilis do not show its trophi approaching the mouth (Supplemental data). A caudal appendage similar to that of C. impossibilis is also present in the two species of the genus Henoceros , which also lack paired spurs. In his description of the first Henoceros species (as Monoceros falcatus ), Milne (1916) referred to this appendage as a spur without an explanation and this attribution has since been accepted by all subsequent authors who discussed this species. However, in the absence of any supporting evidence (for example, embryological), the implied homology of the caudal appendage of the Henoceros species with the paired spurs of the other bdelloid species is not clear. In any case, the caudal appendage of C. impossibilis appears different from the similar-looking organ of Henoceros in that the former has two attachment nibs ( Figs. 10C, D ), while that of Henoceros falcatus has a pointed tip and presumably one attachment nib, although this has not been explicitly stated in the literature ( Milne 1916 ; Ricci & Melone 1998 ). Until the embryological development of C. impossibilis has been studied, I would prefer to refer to the subject organ simply as a caudal appendage. An additional difference between C. impossibilis and the philodinavid species is that the new species has two toes, while all of the philodinavids have four. A possible objection against this argument is that in the current taxonomy of bdelloids, the toe numbers are primarily used to separate the genera, but not the families ( Bartoš 1951 ). Strictly speaking, this is applicable only to the Philodinidae , which includes genera with toe numbers of two, three and four and one genus with an adhesive plate; whereas in the families Adinetidae , Habrotrochidae and Philodinavidae , the toe numbers of all genera are uniform as three, three and four, respectively. When the other morphological differences between some of the philodinid genera are also taken into account, the inter-generic differences in the toe numbers suggest that the Philodinidae is not monophyletic. In short, the toe numbers may indeed be appropriate, along with other traits, to separate bdelloid families. Further discussion of this matter is outside the scope of this paper. Taken together, these considerations render the Philodinavidae unsuitable for the placement of C. impossibilis . Therefore, the new family Coronistomidae is erected here to accommodate C. impossibilis . Beauchamp (1909 : Fig. XI) saw in the foot of Rotaria socialis (Kellicott) only one pair of five mononucleate glands positioned longitudinally, but could not follow their ducts as they were intertwined around each other. He did note that there was a granular substance filling the glands. Brakenhoff (1937 : Abb. 11) gave a more complete description of the two pairs of gland complexes in the foot of Embata parasitica (Giglioli) : the dorsal pair consisted of five longitudinally arranged mononucleate glands along each side of the foot, while each set of the ventral pair, positioned more medially, consisted of one trinucleate elongated gland and one posterior mononucleate gland. He was apparently able to resolve the individual ducts exiting the gland complexes and showed them to twist around each other as they extended into the spurs and the toes. The overall anatomy of the gland complexes in the foot of C. impossibilis ( Fig. 10A ) is similar to that of the gland complexes in the foot of E. parasitica ( Brakenhoff 1937 : Abb. 11). The photographs of C. impossibilis also show the presence of ducts filled with a granular substance ( Figs. 10C, D ) similar to what Beauchamp (1909) saw in the foot of R. socialis . One pair of these ducts appear to exit from the posterior ends of the dorsolateral gland complexes and cross over each other before entering the caudal appendage ( Fig. 10D ). Ducts are also present inside the toes. But because the numerous twisting ducts and the retractor muscles create a tangle too crowded to resolve easily in photographs, I have so far not been able to fully trace any of the other ducts inside the caudal appendage or the toes to any of the gland complexes. Significance of the morphological traits . An adhesive system incorporating the secretions of two different glands, one for attachment and the other for detachment, the so-called duo-gland system, has been hypothesized to operate in turbellarians, gastrotrichs, nematodes and echinoderms ( Lengerer & Ladurner 2018 ). The presence of two pairs of anatomically different gland complexes and two pairs of ducts for two attachment nibs (two ducts per nib) in the caudal appendage of C. impossibilis suggests that a duo-gland adhesive system may also be present in this species. On the other hand, the retractor muscles present in the foot of bdelloids, which are responsible for the retraction of the foot during their characteristic leech-like crawling ( Hochberg & Litvaitis 2000 ), point to a more straightforward mechanism of detachment (the “mechanical model” in Lengerer & Ladurner, 2018 ): when the retractor muscles of the foot, which reach into the caudal appendage of C. impossibilis , contract, the caudal appendage and the toes may simply be pulled off the surface without the need for a special glandular secretion to break their attachment to the substrate. But if this model is correct, it remains to be explained why there are two pairs of anatomically different gland complexes in the foot of C. impossibilis and probably in all other bdelloids. When water flows over an immersed surface a boundary layer forms on the surface. Within the boundary layer the flow velocity is zero at the surface and increases gradually to the velocity of the bulk water at a certain distance (thickness) from the surface ( Lampert & Sommer 2007 ). Many organisms live entirely within boundary layers and do not experience the flow of the bulk water. Even though at the bulk water flow velocity of the Hawlings River ( 0.58 m /s) the boundary layer may only be a few millimeters thick ( Silvester & Sleigh 1985 ), C. impossibilis , having a diameter of ~ 50 µm ( Fig. 5A ), can remain completely within the boundary layer on the surfaces of immersed rocks. However, theoretical calculations also suggest that when water is flowing at 0.58 m /s perpendicular (as opposed to parallel) to an individual of C. impossibilis (which would have a Reynolds number of ~30 for its diameter), vortexes and turbulence may develop around the rotifer ( Silvester & Sleigh 1985 ). This may not be an unrealistic scenario considering that C. impossibilis was found in samples from rock surfaces facing upstream. The Hawlings River’s flow velocity of 0.58 m /s is above Nielsen’s (1950) lower limit ( 0.5 m /s) of “swift-flowing” water. Therefore, C. impossibilis may be considered a “torrential” species. Some morphological and behavioral traits of C. impossibilis may now be interpreted as adaptations to increase the chances of these rotifers to remain in their microhabitat when they are subjected to fast and turbulent water flow. The long and the thick toes as well as the caudal appendage are expected to provide a strong attachment while serving as a pivot to allow the animal to explore its surroundings securely. The bdelloid rotifers with the typical corona on pedicels are either swimmers or stationary feeders ( Örstan 2020 ). Even though the stationary feeders may not move much, the beating of the cilia on their open coronas creates a forward thrust that makes their bodies rise above the substrate to which their feet are attached. In comparison, the corona morphology of C. impossibilis , being an adaptation for grazing within the sediment, does not make the rotifer’s body rise above the substrate and thereby reduces its chances of getting carried away by the water flow. Finally, the stickiness of the egg, as indicated by the debris attached to it ( Fig. 9B ), makes it more likely that the egg will stay where it is deposited. These proposed correlations of morphology with water flow rate should be confirmed by further study and also in other bdelloid species that live exclusively in similar habitats; otherwise, their presentation as general rules may lead to contradictory and inconsistent results ( Nielsen 1950 ; Statzner 2008 ).