Três novas tornarias (Enteropneusta) das costas Sul do Brasil

The "Instituto Oceanográfico", of the University of S. Paulo, lent me several tornariae for studying purposes. These larvae were obtained from samples of plankton collected off the Island of Bom Abrigo (see map) and off Cananéia on the South Coast of Brazil, State of S. Paulo. The samples were taken from a depth of 10 mts., more or less. Fr. Müller mentioned (1898) the existence of Tornaria which probably belonged to Balanoglossus on the coast of Brazil, but did not describe the larvae. No other scientist has published anything on Brazilian tornariae since. All larvae studied from the two samples given me, were larvae without tentacles, answering the general description of the larvae belonging to Stiasny-Wijnhoff and Stiasny's type I tornaria (1931). They are generally middle sized to small. The pre-oral area has the form of an anchor in all three of them, they show the general form of a double cone, the ventral loop is normal, there are no inferior ventral lobes, lateral lobes are present, with ventral loops, and inferior dorsal lobes, the second and third cœloms are situated next to the gut, the middle-gut being large and swollen, according to Stiasny-Wijnhoff and Stiasny's description (1931, p. 185-186) of the tornariae belonging to type I. The description of the three different types of tornariae found in the samples did not correspond to the description of any of the already described tornariae belonging to type I of Stiasny-Wijnhoff and Stiasny's, and so new names were given to them - Tornaria mourei n. sp., Tornaria lilianae n. sp., and Tornaria stiasnyi n. sp. Tornaria mourei n. sp. - (Pl. I - Figs. 1 to 7) Krohn stage - 1,2 mm height x 1,1 mm wide. The anal plate is protruding, the apical plate is medially or slightly dorsally situated. The ventral belt is neither narrow nor large, taking an almost horizontal position laterally. The pre-oral area is more or less of the same size or slightly smaller than the post-oral area. It is protruding, specially in the region of the upper lip, which is very high. The mouth is highly situated and is small. The second and third cœloms are present in the Krohn stage. There is no anal ring and no secondary circum-anal ciliated band. The circum anal ciliated band is slightly ondulated. The form of the ventral loop (specific character of this tornaria) is very high and narrow, the upper and lower lips of the mouth being parallel (see figs.). The oral area is narrow. The secondary loops are not very projecting; wide at the basis and ending almost pointedly. There are two of them on the internal side of each primary loop and 2-3 on the external side of the same. On the middle loop of the pre-oral and of the post-oral areas there are 3-4 secondary loops. These are usually larger in the middle part of the primary loops and are smaller or less conspicuous at the ends. The median pre-oral and post-oral loops are taller and wider than the lateral corresponding loops (Pl. 1 - Fig. 7). The primary lower dorsal lobes are very deep. They turn a little upwards at the end and are more or less pointed. The secondary lobes, are usually deep, and sometimes narrow, opening wide. The lateral lobe is very pronounced, and situated more ventrally than laterally. It has the form of a bottle and ventrally it has a high and large loop, dorsally a smaller but also well defined loop (Pl. I - Fig. 2-7). There are apparently no branchial openings in the Krohn stage. The middle gut is cylindrical in shape and swollen. The hydrocœlom is well developed in the Krohn stage showing muscular thickenings of its walls. The proboscis vesicle or "heart" (pericardium) is present between the canal of the hydrocœlom and the blood vesicle of the proboscis which is beginning its development. The hydrocœlom is large and is pear-shaped or spindle-shaped when laterally looked at. Figs. 4 and 6 (Pl. I) show the same tornaria in the Spengel stage of retrogressive development. It is less transparent then, and shows an annular constriction in the middle of the body, this being club-shaped. This tornaria differs from the others belonging to type I of Stiasny-Wijnhoff and Stiasny (the tornaria of Balanoglossus clavigerus, Tornaria bournei, mülleri, krohni, meeki, mielcki, weldoni, ritteri, tergestina) because of differences in size, in the shape of the upper and lower lip of the oral area, non existence of the anal ring, the shape of the ventral loop, of the secondary loops, their number, the shape of the middle and hind gut the position of the apical plate, and the pre-oral area and the early development of the cœloms, the shape of the lateral lobe, the position and shape of the mouth, the general shape of the tornaria, etc. Tornaria lilianae n. sp., (Pl. II - Figs. 24-30). Krohn stage-height: 0,7 mm x 0,6 mm wide. Shape: double cone, like preceding one. Anal plate very little protruding, the ventral belt being almost vertical in position all around the body of the larva, the apical plate is medially situated, or slightly dorsally, the ventral belt is narrow, the pre-oral area is anchor shaped, and is more or less as large as the post-oral area, it is protruding, wide, occupying more or less half of the ventral region of the tornaria, the cœloma are situated next to the intestine. There is an anal ring. There is a non ondulated circum-anal ciliated band and a secondary circum-anal ciliated band. This is a small tornaria, the shape of the ventral loop being more flat than all and it is medially wide, the hollow of the mouth entrance is narrow, and situated in the middle of the body of the larva, the œsophagus therefore being longer than in the other two described tornariæ, the oral area is narrow, and small, the upper and lower lips being parallel. The secondary loops are very little projecting and show in number of two to four on the primary loops. The primary lower dorsal lobes are deep and finish near one to the other, following an horizontal course. The secondary lobes are not at all deep. The lateral lobe is large and flat, with a small loop on the ventral side and hardly a loop on the dorsal side. The second and third cœlomic cavities are present. The pore of the hydrocœlom opens a little to the left. The hydrocœlom is pear-shaped when laterally observed and round when seen ventrally. It shows no spores. The muscular band which unites it with the apical plate is long and narrow. The water canal from the hydrocœlom is not distinctly separated from it and is short. The position of the hydrocœlom is between the œsophagus and the stomach on both of which it rests. There is no proboscis vesicle or "heart" (pericardium), nor gill slits or notochord present in the Krohn stage of this tornaria. It is a tornaria very slightly transparent. The long œsophagus is well separated from the stomach which shows a median constriction, the hind-gut being also well separated from the middle gut and having the shape of an inverted short cone. The third cœlom touches with its inferior border the largest part of the inferior middle gut. Both second and third cœloms are lamelliform, and wider than the cœloms of T. stiasnyi and T. mourei. The second cœlom is very near but does not touch the stomach or middle gut. This tornaria differs from the other described tornaria of type I because of the size, which is very small, the shape of the middle gut, the horizontal lower primary dorsal lobes, the flat secondary loops and lobes, the flat lateral lobe, the existence of both cœloms in the Krohn stage, the non ondulated circum-anal ciliated band, the shape of the ventral loop, etc. Plate II - Figs. 29 and 30 show different aspects of a younger stage of probably this same tornaria, captured off the coast of Cananéia, while the Krohn stage was captured farther away off the Island of Bom Abrigo, at another time of the year. The younger larvae are in the young Metchnikoff stage of progressive development. Their identification with the older stage just described, can be done by the shapes of the apical plate, of the middle gut, of the hind gut; the long œsophagus, the vertical position of the ventral belt, the flat anal plate, the anal ring, the small transparency, the flat and large ventral loop, the parallel lips of the mouth opening. This description was based on one specimen in the Krohn stage and two specimens in the young Metchnikoff stage. Tornaria stiasnyi n. sp. - Found with greatest frequency (12 specimens from one towing net) in the plankton collected off Cananéia in September 1949. This larva was not caught in the Krohn stage, only in the young Metchnikoff stage, transition stage from Heider to Metchnikoff, older Metchnikoff stage, Spengel and Agassiz stage. The stages of progressive development (Pl. II - Figs. 14 to 21) were caught off Bom Abrigo Island in January, the stages of retrogressive development were caught off Cananéia (Pl. I-II - Figs. 8-13) in September. This larva must therefore belong to an animal with two periods of sexual maturity during one year probably. Although the Krohn stage was not caught it was possible to characterize the larva as being a new kind of type I larva (or of type V) because of its very peculiar anal area, its excentric anus, the apical plate very dorsally situated, the ventral belt being narrow and occupying a vertical position dorsally whereas ventrally in the young stages it is even slightly bent towards the outside of the larva and in elder stages it is horizontal ventrally. The oral area is pushed upwards the apical plate, excepting in the very young stages. The opening is also very elevated, even in the young stages, the whole gut being pushed anal wards. There is an anal ring which follows the general form of the anal plate, very much pushed upwards ventrally, and very short dorsally. The ventral loop is extremely elevated and narrow, the dorsal lip of the mouth entrance being parallel to it. Very characteristic are also the long cilia which develop from the circum-anal ciliated band and the small patches of rugose skin which appear all along this band, around the anus, and in the form of a large blotch at the upper and lower turns of the lateral loops in the pre-oral and post-oral bands. There are two small blotches on the anal area dorsally to the anus which can be observed in the Metchnikoff stage already. The oral area is very wide, and so is the lateral lobe, limited by a very large ventral loop and a smaller dorsal loop. The œsophagus, middle and hind gut are well separated from each other, the œsophagus being short, the middle gut cylindrical or oval in the young forms, and the hind gut being in the form of an inverted cone. The cœloms of the collar and trunk are sausage shaped and one is well separated from the other. The trunk cœlom or third cœlom has a pointed end and is larger than the collar or second cœlom. They are both present in the elder stages. These showed a great variation in size between 1,1 mm in height x 0,85 mm in width, and 0,85 mm in height x 0,59 mm in width. The Krohn stage having not been captured, it is probable that the larva may attain a larger size than the largest indicated above. The hydrocœlom is pear shaped and shows a very long muscular band of attachment to apical plate in the young forms. The elder forms show a short muscular band, well delimited from the hydrocœlom. The water pore is situated more to the left, and the canal which unites it to the hydrocœlom is very thin and well delimited from the hydrocœlom from the Metchnikoff stage onwards. The heart, proboscis vesicle or pericardium is present in the Spengel stage, as well as the bunch of cells which will form the proboscis blood vessel. Two branchial openings are present in the Spengel stage. This larva does not look like any of the other larvae of the type I of Stiasny-Wijnhoff and Stiasny. It looks very much like Tornaria hubbardi Ritter & Davis 1904, with which it has several features in common, but it differs from that tornaria by the fact that it shows a smaller number of gill slits, a well delimited apical plate, a more protruding anal area, a smaller oral area, a lateral lobe, larger pre-oral and post-oral bands. The general form of the larva is not a pointed bell shape, but a rounded bell shape. T. hubbardi as well as T. stiasnyi are being compared in the Agassiz or Spengel stage. According to Stiasny, T. hubbardi was described in the Agassiz stage of development, but Stiasny had only that stage for description. The tornaria under description, when considered apart from the younger stages, would also give the impression of an Agassiz stage of development, because of the out jutting coneshaped area, the gut pushed towards the anal region, the very large hydrocœlom, the longitudinal ciliated band being pushed towards the apical plate, but, when compared to the younger stages, it will be observed that already in the transitional stage between Heider and Metchnikoff stages, the larva shows all the above mentioned characteristics excepting the large hydrocœlom and perhaps the out jutting coneshaped anal area. This last feature does not seem so important even for Stiasny himself, because that author could not have observed it in T. hubbardi, whose anal area is flat. I should place both T. hubbardi and T. stiasnyi (Pl. I - Fig. 8) in the Spengel stage and consider Pl. II - Fig. 13 of Tornaria stiasnyi as being in the Agassiz stage. The reasons for this consideration are that in Pl. I - Fig. 8 the tornaria still shows the rests of secondary loops and lobes and is still very transparent, the whole larva having stretched a bit and diminished in size. The cœloms are larger - they have grown. Fig. 10 shows a larva typically in Spengel stage, even the circular constriction in the middle and the clavate form of the larva can be observed. The œsophagus wall is darker because of the muscular growth of its walls, in Pl. II - Fig. 13 which shows an Agassiz stage of retrogressive development. Types of these three new tornariæ are kept at the Museum of the Instituto Oceanográfico, in São Paulo. Topotypes are in the Museu Paranaense, in Curitiba. As to establishing the correspondence between these larvae and the adult animals, it is necessary to remember the distribution of the Enteropneusta in the South of Brazil (see map). In the State of Rio de Janeiro (Spengel 1893, p. 126) Glossobalanus minutus (Kowalewsky), Balanoglossus apertus (Spengel), and Schizocardium brasiliense Spengel occur. This last species also occurs in S. Sebastião in the State of S. Paulo, as well as Balanoglossus gigas Fr. Müller (Sawaya 1951, p. 730). A smaller Balanoglossus also occurs in Santos in the State of S. Paulo. In Paraná, on the Caiobá Beach we find Balanoglossus paranaensis Björnberg and a similar Balanoglossus in the Ilha do Mel. In Santa Catarina, another state further south, a small Balanoglossus and a large one B. gigas occur near to Florianópolis. Of the places mentioned above, Caiobá is the nearest to Cananéia, where the tornariæ were collected, one of them therefore probably belonging to B. paranaensis. It is the opinion of Mr. & Mrs. Stiasny (1931, p. 185) that T. hubbardi, which is very similar to T. stiasnyi should belong to a Vth type of tornaria, with different characteristics from the Ist, IInd, IIIrd and IVth types. The Stiasnys also wrote that this T. hubbardi and therefore the Vth type of larvae might belong to Schizocardium, whose larva is not known. If that should be true T. stiasnyi would probably belong to Schizocardium brasiliense. It is necessary to rear the larva and thus solve the problem.