A new suction feeder and miniature teleosteomorph, Marcopoloichthys mirigioliensis , from the lower Besano Formation (late Anisian) of Monte San Giorgio

A new species ( Marcopoloichthys mirigioliensis ) of the stem teleosteomorph genus Marcopoloichthys is described from the lower Besano Formation (late Anisian at Monte San Giorgio, southern Switzerland), making this new species distinct from Marcopoloichthys furreri from the Prosanto Formation (early Ladinian at Ducanfurgga, southeastern Switzerland). Marcopoloichthys mirigioliensis n. sp. is smaller (ca. 32 mm standard length) than M. furreri (ca. 40 mm standard length), and in addition, the two species have some important differences in the caudal endoskeleton and fin, e.g., number of epaxial and hypaxial basal fulcra, uroneural structure, size of hypurals, and presence versus absence of urodermals. Marcopoloichthys mirigioliensis n. sp. is the smallest member of Marcopoloichthyidae which is currently known from at least five species living in the Triassic of China (one species), Italy (two and others that remain undescribed), and Switzerland and according to current information, with its ca. 32 mm standard length is candidate to be considered a miniature fish. Additionally, this size makes it the smallest known stem teleost. As in other marcopoloichthyids, the buccal and suspensorium anatomy of M. mirigiolensis n. sp. corresponds to that of suction-feeder fishes.


Introduction
Marcopoloichthyids are small stem teleosteomorphs (Arratia, 2022) of about five centimeters maximum length, which are easy to identify because of their characteristic morphology, including a special mouth configuration as suction feeders with the upper and lower jaws able to displace anteriorly during feeding, scaleless and torpedo-like body, and a vertebral column with a persistent, functional notochord and well-developed arcocentral vertebral elements.They are currently known from one genus (Marcopoloichthys Tintori et al., 2007) with a Eurasian distribution (Arratia, 2022;Tintori et al., 2007) and four species: Marcopoloichthys ani Tintori et al., 2007 in the Anisian of China, M. andreettii Tintori et al., 2007, andM. faccii Tintori et al., 2007 (previously described as Pholidophorus faccii by Gortani in 1907) in the Ladinian and Carnian of Italy, and M. furreri Arratia, 2022 in the Ladinian of southeastern Switzerland.A new, smaller marcopoloichthyid, collected in eight different dolomite beds in the lowermost part of the Besano Formation (Late Anisian), southern Switzerland, is described herein and is compared with other marcopoloichthyids.Considering that this fish is one of two oldest known suction-feeders (along with Marcopoloichthys ani of China), information on the history of the site and its geology is given below.
The Middle Triassic succession of Monte San Giorgio is part of the Southern Alps in Northern Italy (province of Varese) and southern Switzerland (canton Ticino; Fig. 1).It starts with fluvio-deltaic deposits (Bellano Formation, Illyrian; Sommaruga et al., 1997).The late Anisian sediments indicate the progressive transgression of a shallow, epicontinental sea and expansion of a carbonate platform (San Salvatore Dolomite; Zorn, 1971) north of an emerged land area, which is now covered by the Po Plain (Brusca et al., 1981;Picotti et al., 2007).During the latest Anisian and Ladinian, although shallow water sedimentation continued in the north, an intraplatform basin opened in the area of the Monte San Giorgio, which led to the deposition of the Besano Formation, San Giorgio Dolomite, and Meride Limestone (Bernasconi, 1994;Furrer, 1995;Rieber, 1973;Röhl et al., 2001;Stockar et al., 2012).
The Besano Formation (''Grenzbitumenzone'' of Frauenfelder, 1916;Rieber, 1973) consists of an alternation of dolomites and black shales of up to 16 m in thickness, including the Anisian/Ladinian boundary in its uppermost part (Fig. 2).A volcanic-ash layer six meters below this boundary yielded a U-Pb minimum age of 242.1 ± 0.6 Myr (Mundil et al., 2010).The lower Besano Formation, with laminated to stromatolitic dolomites, very thin, organic-rich, black shales, and a few vertebrate fossils, was deposited in a shallow, restricted, carbonate platform environment (Röhl et al., 2001).The middle part, with prominent black shales in between dolomitic mud-to packstones and a highly diverse vertebrate fauna, documents a slightly deeper intraplatform basin with euxinic and anoxic conditions.The upper Besano Formation, with dominant laminated dolomites, thin black shales, and well preserved, but less diverse vertebrates, was deposited again in a shallow, restricted, carbonate platform environment (Röhl et al., 2001).
The new Marcopoloichthys material from the Besano Formation of Monte San Giorgio, southern Switzerland (late Anisian, Illyrian, R. reitzi Ammonoid Zone) is a bit younger than Marcopoloichthys ani from the Guanling Formation (middle Anisian, Pelsonian) of southern China (Yunnan and Guizhou Provinces; Tintori et al., 2007), but older than M. andreettii, from the base of the Cunardo Formation (early Ladinian) of Northern Italy (Province of Varese; Tintori et al., 2007) and M. faccii from the early Carnian of northeastern Italy (Province of Udine; Tintori et al., 2007).It is also older than M. furreri from the Prosanto Formation (early Ladinan) of southeastern Switzerland (Arratia, 2022; see also Fig. 1).However, all species lived in the Triassic Tethys Ocean, Marcopoloichthys ani in the Eastern Tethyan realm and all the others in the Western Tethyan realm.

Material and methods
The studied fossil material was collected in 1968, during the last year of the systematic excavation at Point 902/ Mirigioli (Meride, canton Ticino, southern Switzerland) under the direction of Emil Kuhn-Schnyder (Paleontological Institute and Museum University of Zurich, PIMUZ).The best-preserved specimens were mechanically prepared with the aid of sharpened steel needles.The material studied here consists of 28 specimens, which are catalogued in the collection of the Paleontological Institut and Museum, University of Zurich, Switzerland (PIMUZ).The fishes were photographed with a Keyence digital microscope VHX7000N with VXH-7100 camera unit at the PIMUZ.Illustrations were executed under a microscope equipped with a camera lucida attachment.

Anatomical terminology
The terminology of the skull roof bones follows Schultze (2008 and literature cited therein) that has been recently confirmed using other evidence (Teng et al., 2019).To avoid confusion, the first time that the parietal and postparietal bones are cited in the text, as well as in all figures, the traditional terminology is shown in square brackets, e.g., parietal bone [= frontal]: pa [= fr].The terminology of the vertebral column follows Arratia et al. (2001) and Arratia (2015), whereas that of the caudal endoskeletal elements follows Schultze and Arratia (1988, 1989, 2013), Arratia and Schultze (1992) and Schultze and Arratia (2013).The method of counting vertebrae follows Tintori et al. (2007), an approach followed by Arratia (2022), to ensure that the results are comparable.The terms fin rays, scutes, fulcra and its different types, procurrent rays, and principal rays follow definitions provided by Arratia (2008Arratia ( , 2009)).

Description
The new fish, Marcopoloichthys mirigioliensis sp.nov., is small, about 30 to 35 mm SL (with 32 mm being the most frequent value), with an elongate, torpedo-like body, a moderately large head that occupies about 36% of the SL, and a narrower caudal peduncle that is followed by a deeper caudal fin.The orbit is moderately large; its diameter is about 30% of the head length, and the preorbital length is moderately long, about 33% of the head length.
The insertion of the dorsal fin is located slightly in front of the middle of the body length and is opposed ventrally to the origin of the pelvic fins; the insertion of the anal fin is placed opposite to the posterior part of the dorsal fin.The insertion of the anal fin is just in front of the last pterygiophore of the dorsal fin or slightly posterior; thus, the insertion of the moderately short anal fin is closer to that of the pelvic fins than to the caudal fin.Consequently, the new fish, as in other marcopoloichthyids, has a relatively long caudal peduncle (about 38% of standard length in the holotype PIMUZ T 3030).The small pectoral fins, poorly preserved, have a low position, closer to the ventral margin of the body than to the middle region of the flank.The caudal fin has both lobes of nearly the same size with its posterior margin deeply forked.The distal portions of all fin rays are commonly crushed or not preserved.

Skull roof and braincase
The skull roof is incompletely preserved in all specimens and does not allow a detailed description of its shape, nor of specific bones.
Remains of the parietal [= frontal] bones are exposed laterally, i.e., above the orbits, and they have a gentle incurvation when the fish is preserved with closed jaws.
The region where the nasals would be placed is poorly preserved, but an elongate rectangular nasal bone and a shorter accessory or additional nasal bone are preserved in PIMUZ T 3030 (Figs. 3,4).An incomplete, elongate bone, the mesethmoid, is preserved at the most anterior tip of the snout.It has a similar T-shape, as in Marcopoloichthys furreri, but commonly one of its lateral processes is destroyed in the available specimens; the preserved process is distinct, well developed, and laterally extended.
The postparietal [= parietal] and supratemporotabular [= dermopterotic] are poorly preserved so that a proper description is not possible, but the supratemporotabular would be forming the lateral border of the skull roof in PIMUZ T 3030 (Fig. 4); but its suture with the postparietal is not preserved, while the autosphenotic and its sutures with the supratemporotabular and parietal show that the bone is unfused in the holotype.Because of poor preservation, it is unclear whether the skull roof bones are independent or whether they are partially fused in available specimens.A fragment of a possible extrascapula is preserved in PIMUZ T 3030.The external surface of the skull roof bones is covered by ornamentation, consisting of small, irregular ridges and tubercles.Information on the cephalic sensory canals and pit-lines is not available because of the ornamentation covering the skull roof bones; however, traces of sensory canals are not observed in places where the ornamentation is missing.
The orbitosphenoid is not preserved in the available material.The lateral ethmoid is well ossified and slightly bent.Little information is available about the parasphenoid and vomer, although the parasphenoid is partially exposed in the holotype (Fig. 4) and is preserved as an imprint in PIMUZ T 4411.

Orbit and circumorbital bones
The new species has a moderately large orbit (Figs. 3,4) of about 30% of head length.All specimens studied herein were not feeding at the moment of their death, and their orbit is preserved almost rounded; in contrast, in other marcopoloichthyids preserved during feeding, the orbit and the mouth are protracted anteriorly, and the orbit is oval-shaped (Arratia, 2022: Figs. 4, 7B).
The circumorbital series is incompletely preserved in the holotype and partially destroyed or not preserved at all in other specimens so that it is not possible to describe the anatomical characteristics of all bones.Additionally, the ornamentation obscures most of the boundaries between bones in the holotype.As in other marcopoloichthyids, the circumorbital series is open dorsally due to the absence of supraorbital bones.Anteroventrally, a possible antorbital and infraorbitals 3-5 (Fig. 4) are observed, while a possible small, squarish dermosphenotic is preserved at the posterodorsal corner of the orbit in PIMUZ T 3030.A faint remnant of an anterior sclerotic bone is preserved.Posterior to the dermosphenotic and posterodorsal infraorbitals, one or two suborbitals are preserved, but the ornamentation obscures details on these bones, as well as on infraorbital 3 and the trajectory of the infraorbital canal.
Upper jaw The upper jaw (Figs. 4, 5) consists of a moderately long premaxilla and a long maxilla, which are poorly preserved.A supramaxilla has not been observed and is considered absent as in other marcopoloichthyids.The premaxilla and maxilla either are broken or not preserved in the available specimens.The premaxilla is about half of the length of the maxilla, and when the mouth is closed, the premaxilla is placed anteroventrally to the maxilla.The premaxilla is elongate, and its anterior thin process is incompletely preserved in the available material and slightly gently curved and edentulous (Fig. 6B).The longer and larger maxilla (Fig. 5) is narrow anteriorly and widens caudally into a plate-like posterior part.Its posterior margin is slightly oblique.Both bones, premaxilla and maxilla, lack teeth.If ornamentation was present, no remains are visible in any specimen.
Lower jaw The lower jaw (Figs. 4, 5) is not well preserved in the available material, and only the anterior margin and ventral part are visible.As in other marcopoloichthyids, the jaw is prominent, relatively short, deep, and somewhat triangular-shaped, with the quadrate-mandibular articulation placed at a point roughly below the midorbit (Fig. 4) when the mouth is closed.The oral margin lacks teeth as in other marcopoloichthyids.The ventral margin is almost straight and has a parallel running ridge below which a series of small pores indicate the course of the mandibular sensory canal (see specimens PIMUZ T 2976, 4411 and 4428).The lower jaw narrows anteriorly at the symphyseal region, which is anteroventrally placed and widens dorsally, producing a very high coronoid process that is covered by lateral bones in the available material (Fig. 5).Thus, the elements forming the coronoid process are unknown in this species, but it is expected to be formed by the dentary and surangular, as in other marcopoloichthyids (see Arratia, 2022: Figs. 4B, 8).The anteroventral margin of the jaw has the characteristic shape described for Marcopoloichthys furreri, due to the presence of a slightly rounded anteroventral process (Fig. 4).The posterior part of the right lower jaw is preserved medially in PIMUZ T 2976, and because only one ossification is observed, this is interpreted here as the result of fusion among the articular, angular and retroarticular.The postarticular process of the angular bone (Fig. 4) is rudimentary or non-existent.
The jaw ornamentation is very characteristic in Marcopoloichthys furreri; however, the lateral surface of the jaw is covered by other bones in this new fish material, and the exposed portion is not very informative, so that the condition of the ornamentation remains unclear for Marcopoloichthys mirigioliensis sp.nov.and right elements are superposed in PIMUZ T 4411), and the hypohyal region is not preserved.Remnants of branchial arch elements are preserved in a single specimen (PIMUZ T 4432; Fig. 6), but a description is not possible.Specimen PIMUZ T 4411 is an interesting specimen that provides information on the suspensorium, but there is a problem in that the left and right quadrates and hyomandibulae (Fig. 6B) are preserved and overlapping each other, making the interpretation of the cranial bones difficult.The suspensorium is only partially preserved.The left hyomandibula is almost vertically oriented, with a slight anteroventral inclination on  4411).Scale bar: 2 mm.Abbreviations: a.prra, anal proximal radials; c.dra, compound dorsal radial; dor.r, dorsal fin rays; ns, neural spines; pel.r, pelvic rays; sc, scutes its shaft.Because of the position of the bone, its cranial facet seems to articulate largely with the supratemporotabular region [= dermopterotic]; because of the poor preservation or no preservation of the braincase (which probably was not ossified in life), other components of the cranial facet are unknown.There is a significant distance between the ventral ossified part of the hyomandibula and the quadrate and it is expected that this space was filled with the symplectic or hyosymplectic cartilage (Fig. 6B), because no ossified section is observed between the lower part of the hyomandibula and the ossified lower part of the symplectic (which is not preserved in the specimen illustrated in Fig. 6, probably due to its cartilaginous condition).The quadrate is a triangular bone, with a well-defined rounded condyle for articulation with the lower jaw; it is unclear how large the metapterygoid was.An elongate, triangular-shaped symplectic is observed in PIMUZ T 3030 (Fig. 4).A Opercular and branchiostegal series, and gular plate The preopercle is described here, although it is a bone associated with the suspensorium.The opercular and branchiostegal series are poorly preserved or incomplete so that their descriptions are also incomplete.The preopercle is a large and L-shaped bone, which is slightly expanded postero-ventrad at the confluence of both of its arms.No specimen preserves the complete bone; its dorsal arm is slightly longer than the ventral one when the mouth is closed and almost reaches the lateral side of the supratemporotabular [= dermopterotic] dorsally (Fig. 3C,  D), and the ventral arm seems to be mainly carrying the preopercular canal.This canal is positioned along a slightly broad groove and blind-ending tubules are clearly visible in PIMUZ T 2976, T 4415, and especially in T 4419, where the preopercles are partially preserved.
A description of the opercle, subopercle and interopercle is not possible, because the bones are damaged or not preserved in the available material (see Figs. 3,4,5).A similar situation is observed in relation to the branchiostegal series, whose number of rays remains unknown, as in other marcopoloichthyids.A gular plate has not been observed and is interpreted as absent, as in other marcopoloichthyids.
Vertebral column, intermuscular bones, and ribs A few specimens provide partial information on the vertebral column.The caudal vertebrae are better preserved than the abdominal ones, including the holotype PIMUZ T 3030 (Fig. 3) and paratypes PIMUZ T 4410 (Fig. 7B) and PIMUZ T 4411 (Fig. 6A).
The vertebral column is aspondylous (see Arratia et al., 2001 for the different types), with well-developed dorsal and ventral arcocentral elements forming the centra, and a persistent, functional notochord in adults.There are about 35-38 vertebrae, including five hypural segments, making the count comparable with those of Tintori et al. (2007) and Arratia (2022).There are about 19-21 caudal vertebral segments (Fig. 3); small interdorsal and interventral arcocentral elements alternating irregularly with the well-developed basidorsal and basiventral arcocentral elements have been observed, as in Marcopoloichthys furreri.No remains of centra are present in the ural region (see description of caudal fin).
The neural and haemal spines (Figs. 3,7,9,11) of the caudal region are narrow, except for those of the preural centra (see below).The neural and haemal spines are moderately inclined toward the body axis, increasing slightly their caudal inclination posteriad.The first haemal spines are short, not extending between the anal pterygiophores or just reaching them.Due to the poor preservation of the abdominal or precaudal vertebrae, the presence and structure of the "supradorsal carrier" found in Marcopoloichthys furreri (Arratia, 2022: Fig. 9) could not be confirmed.The total number of parapophyses (Fig. 3) is unclear due to poor preservation of the precaudal region.The first ones would be covered by the opercle and the dorsal bones of the pectoral girdle.The parapophyses are lightly squarish in shape, and each bears a small articular cavity close to its ventral margin.No ribs are preserved in the studied specimens.They were not reported or illustrated in Marcopoloichthys ani (Tintori et al., 2007: Fig. 4) or in Marcopoloichthys furreri (Arratia, 2022: Fig. 9), and it is accepted here that marcopoloichthyids do not have ossified ribs.
There is no reliable information on the neural arches of the abdominal vertebrae, which are distorted, disarticulated, or not preserved at all in the available specimens.Consequently, it is not possible to confirm the presence of a stout and short epineural process emerging at the posterolateral margin of the arch, as in Marcopoloichthys furreri.
The series of supraneural bones is commonly either not preserved, distorted, or displaced and covered by other structures.The total number of supraneural bones is unknown.The series extends up to the expanded, platelike, compound first dorsal proximal radial, and it does not extend between the most anterior proximal radials, as in Marcopoloichthys furreri.

Pectoral girdle and fins
The pectoral girdle is poorly preserved or not preserved in the available specimens.The dermal bones forming the teleostean girdle are the posttemporal (linking the girdle with the posterior region of the cranium), supracleithrum, cleithrum, clavicle, and postcleithra.The posttemporal, supracleithrum, and clavicle are incompletely preserved or not preserved in the available material (Figs. 3,4,6).Chondral bones, including the scapula, coracoid, and proximal and distal radials, are not preserved in the available material.
The sigmoidal-shaped cleithrum (Figs. 4, 6) is an elongate, narrow bone forming most of the pectoral girdle and is a heavily ossified bone, with a long and narrow dorsal limb and a slightly expanded and curved ventral limb.Due to the poor preservation of the cleithrum it is unclear whether a serrated appendage is on its anteromedial surface.Remains of the scapula and coracoid are preserved in the available material, but they are not informative.
The pectoral fin (Figs. 3,4,6,7) is positioned near the ventral margin of the body.The total number of pectoral rays is unknown, because the fins are incomplete or disarticulated, but one thick, first pectoral ray and nine to 11 rays are preserved in PIMUZ T 3030.All rays have long, delicate bases and are scarcely branched and segmented distally.Basal fulcra or fringing fulcra have not been observed.

Pelvic girdles and fins
The pelvic girdles are exposed in a few specimens (e.g., Figs. 3,7,8 and 9).A large, elongate plate-like basipterygium (or pelvic plate) is slightly curved medially; its posterior part is slightly broader than the anterior one and presents an elongate posteromedial process that it is longer than the one present in Marcopoloichthys furreri.The rays per fin are difficult to count due to their poor preservation, but seven or eight rays are preserved in specimen PIMUZ T 4409 and nine in PIMUZ T 4432.This last specimen is preserved with the pelvic fin expanded so that it is possible to observe that the two most inner rays are thinner than the preceding rays.

Dorsal fin and radials
The insertion of the dorsal fin is almost at the same level of the insertion of the pelvic fins and the fin does not overlap the anal fin (Fig. 9).The dorsal fin (Figs. 3,6A,7,9) and its radials are not well preserved, with all bones partially displaced or damaged so that a precise total number of dorsal fin rays and radials cannot be provided.The holotype PIMUZ T 3030 has ca.15 rays preserved, but the radials are destroyed; in contrast, PIMUZ T 4411 has nine proximal radials, but only 10 incomplete or broken rays are preserved.Medial and distal radials are not preserved in the available specimens.In all specimens, except one, the first dorsal proximal radial is destroyed or incomplete; this plate-like proximal radial is nicely shown in Fig. 9.It is an ax-shaped element, with an anteriorly expanded, slightly rectangular bony plate that is partially separated proximally of a posterior bar-like radial; this morphology suggests that this compound element may be the result of fusion and modification of at least three proximal radials.All other proximal radials are slightly expanded distally, with the last one more so, having a broader distal articular surface.However, it is unknown how many rays are supported by this last pterygiophore.

Anal fin and radials
The anal fin and its pterygiophores are not well preserved in the available material, making a description difficult.Additionally, there is variation in number of proximal radials and amount of fusion involved between proximal radials.The most complete series of proximal anal radials, or the most informative, is that present in PIMUZ T 4424 (Fig. 10).In this specimen, the first proximal radial is broader and apparently formed only by a single expanded proximal radial.This first proximal radial is long and curves anterodorsally giving the bone a characteristic shape in marcopoloichthyids (Arratia, 2022;Tintori et al., 2007).This first proximal radial is followed by four elongate, but shorter and narrower radials in this specimen.The last radial is an elongate element, bearing a narrow, thin anterior process that extends dorsally between the distal tips of the haemal spines and has a broad distal portion for articulation with about four lepidotrichia (Fig. 10).In specimen PIMUZ T 4410a (Fig. 9), the first proximal radial is expanded ventrally, giving the impression that this is a result of fusion of two radials; the bone is very elongate and curves markedly anterodorsally.This element is followed by three shorter and simple proximal radials, and the last one is a shorter proximal radial that expands posteroventrally in an elongate, somewhat triangular-shaped plate, which supports about four rays that are poorly preserved in this specimen.The holotype has relatively well preserved anteriormost radials, but the last, expanded one is partially destroyed.In total, the anal series of proximal radials includes five separate elements and ca.nine rays.

Caudal fin and endoskeleton
The caudal fin and endoskeleton are preserved in several specimens.The homocercal caudal fin (Figs. 3, 7A) is deeply forked, with a few shorter and thinner middle principal rays compared to the long, marginal segmented and branched principal rays (Fig. 11).As in Marcopoloichthys furreri, many rays still preserve a thin layer of ganoine on their surfaces.
Two or three preural vertebrae support the anterior basal fulcra.The preural vertebrae, as well as the ural ones, are supported by a functional notochord.Consequently, except by the arcocentra, no centra are formed, and the region is monospondylous, in contrast to the irregular diplospondylous vertebral segments present in some of the mid-caudal vertebrae (Fig. 11).The two preural segments (corresponding to preural centra 1 and 2) are characterized by the presence of well-developed ventral arcocentra with moderately broad and flat haemal spines, which distally support the last principal rays, one procurrent ray and the series of hypaxial basal fulcra (Fig. 11).Dorsally, the neural arches or arcocentra and spines of these two preural vertebrae are missing, and their space is partially occupied by the uroneural plate, a condition different to that in Marcopoloichthys furreri (Arratia, 2022: Figs. 13, 14).The neural spines of the last caudal and preural vertebrae 3 and 4 are inclined posteriorly, closer to the body axis, and they do not support the most anterior basal fulcra.
The preservation of the neural spines of preural vertebrae 3-4 suggests that they have a central core of cartilage surrounded by a thin, perichondral ossification.An anterior process at the base of neural spines is apparently absent.The haemal spines of preural centra 1-2 are broader than most anterior spines.The haemal spine of preural vertebra 4 and more anterior ones are narrower.The haemal spines of the most preural vertebrae are thinly perichondrally ossified.The haemal spines of preural vertebrae 1-3 (Fig. 11) bear a short and narrow process dorsally at their limit with the expanded ventral arcocentra, which is often broken.A hypurapophysis on the lateral wall of the ventral arcocentrum or haemal arch of preural centrum1 is absent.
Posterior to the neural spine of preural centrum 3, a large chondral neural plate is positioned (Fig. 11).In many specimens, this region is damaged or the elements are displaced, except for PIMUZ T 3030, which is illustrated in Fig. 11.Because of its position as part of the ural region and the lack of neural arches or arcocentra, the bony plate is considered here as a uroneural plate or "a uroneural of special kind" following Arratia and Schultze (2013; see section of Discussion).Certainly, due to its position and shape, such an element in Marcopoloichthys mirigioliensis n. sp., increased the stiffness of the tail during locomotion, which is a function of the uroneurals.
No epurals are present, and there is no space left for them between the distal tips of the enlarged uroneural plate and the bases of the epaxial basal fulcra.Five hypurals (Fig. 11) are present, all of them close together so that a hypural diastema between hypurals 2 and 3 is absent.Hypural 1 is expanded at its proximal region and seems to have preserved part of the ventral arcocentrum as in M. furreri.Hypural 1 is the longest and broadest element of the series, whose size diminishes posteriorly.Hypural 1 (Fig. 11) supports the lowest principal rays, while hypurals 2 and 3 support the thin and short middle principal rays.Several thin and narrow bases of the principal rays articulate directly with one hypural without producing a special angle.
There are ca.eight epaxial basal fulcra, which are followed by an incomplete series of fringing fulcra, which only reach to the mid-region of the dorsal margin of the first unsegmented principal ray.The most anterior basal fulcra are bifurcated at their bases (Fig. 11).There are 20 or 21 principal rays with narrow bases that are segmented and branched distally as in Marcopoloichthys furreri (Arratia, 2022).Their most distal segments are commonly not preserved, except for the shorter rays of the middle region of the fin, which have thin and narrow bases branching once or twice distally.The articulation between segments of the principal rays is straight.
Ventrally, there are about 12 hypaxial basal fulcra.There is one short procurrent ray that is followed by a short series of hypaxial fringing fulcra (Fig. 11).In addition, an accessory fulcrum is present between the procurrent ray and the hypaxial basal fulcra (Fig. 11).
One incomplete dorsal scute and another incomplete ventral scute precede the epaxial and hypaxial lobes, respectively.A series of four slightly displaced ovoid and circular urodermals is present (Fig. 11).A new suction feeder and miniature teleosteomorph Scales The body is devoid of scales, except for two large oval cloacal scales or scute-like elements (Figs. 7, 8, 9 and 10) placed around or close to the urogenital region as in other marcopoloichthyids.

Stomach content
In the region between the anterior most anal pterygiophore, pelvic fins and enlarged scales or scutes of specimen PIMUZ T 4410, there is a series of small structures that are herein interpreted as stomach content remains (Fig. 7B).These remains are represented by tiny phosphatic elements (length about 0.2-0.3mm) that probably are hooks of a coleoid cephalopod (see Rieber, 1970).These hooks are smaller than the ones found in the holotype and only specimen of Phragmoteuthis?ticinensis Rieber, 1970 recovered in bed 17 and are currently under study to identify them taxonomically.

Analysis of certain anatomical features
The torpedo-like body of Marcopoloichthys mirigioliensis n. sp.(Fig. 3) is one of the features distinguishing marcopoloichthyids from all other Mesozoic teleosteomorphs, most of which have oblong or fusiform body shapes (Arratia & Schultze, 2024: Fig. 13).Another feature that characterizes marcopoloichthyids is the possession of a scaleless body, a feature also presents in the Triassic teleosteomorph Prohalecites (Tintori, 1990) and rarely found in actinopterygians (e.g., in Birgeria and Gymnoichthys).According to the available information, Marcopoloichthys and Prohalecites are the only known Triassic teleosteomorphs with "naked" bodies; however, a scaleless, torpedo-like body is a unique combination of characters only found in marcopoloichthyids among teleosteomorphs (Arratia, 2022;Arratia & Schultze, 2024).Marcopoloichthyids have two (occasionally three) enlarged oval scales around the urogenital region.Modified scales or so-called cloacal scales are frequently present surrounding the urogenital region of extant teleosts, and they are distinctly distributed and shaped in some Jurassic 'pholidophoriforms' (GA, pers.observation).Modified scales or 'pre-anal scutes' are also known from other stem teleost groups (e.g., pachycormids), but detailed descriptions of these elements are not available for many of the stem teleosts so that comparisons are not possible at the present time.They are not present in the other scaleless Triassic teleosteomorph Prohalecites (Tintori, 1990), and the presence of commonly enlarged scales in the urogenital region is interpreted as a unique feature of marcopoloichthyids (Arratia, 2022).
Another noteworthy feature is the structure of the buccal apparatus with an elongate triangular maxilla bearing a narrow anterior articular process and lacking teeth on its oral margin, and an elongate premaxilla, with a long and narrow proximal region lacking the ascending process characteristic of more advanced teleosts (Figs. 4, 5;Tintori et al., 2007: Fig. 3;Arratia, 2022: Figs. 4, 5, 7, 8).In contrast to the apparent weakness of the upper jaw, the lower jaw is a massive structure with an enlarged anterior portion of the dentary and also lacking teeth (Figs. 4, 5;Tintori et al., 2007: Fig. 3;Arratia, 2022: Figs. 5, 7, 8).The posterior part of the jaw of Marcopoloichthys mirigioliensis n. sp.exposed medially in one specimen (see Fig. 5) confirms the fusion of the three bones at the posterior part of the jaw into an angulo-articulo-retroarticular bone, as previously described and illustrated for M. furreri (Arratia, 2022: Fig. 8).Unfortunately, the presence or absence of a coronoid bone could not be clarified in the available specimens of M. mirigioliensis n. sp.Both jaws are edentulous in marcopoloichthyids, which is a unique feature among Triassic stem teleosts (Arratia & Schultze, 2024), but also it is a rare feature among neopterygians, especially teleosts in general.
The preserved quadrate (Fig. 6B) of the new species is a small triangular bone with a comparatively robust articular condyle for the articular facet of the lower jaw, similar to the conditions described for M. ani and M. furreri.The articulation of the quadrate-lower jaw lies below the middle orbital region in Marcopoloichthys mirigioliensis n. sp., and the quadrate relates to the dorsal part of the suspensorium (hyomandibula) through an elongate (assumed) cartilaginous hyosymplectic region (Fig. 6B).In contrast, the quadratelower jaw articulation is positioned below the anterior half of the orbit in M. ani (Tintori et al., 2007: Fig. 3) and some specimens of M. furreri, a condition that would imply that the lower jaw was comparatively shorter than in other marcopoloichthyids and the hyosymplectic region was even longer than that in M. mirigioliensis n. sp.
The vertebral column of Marcopoloichthys mirigioliensis n. sp. is characterized by the presence of a persistent, functional notochord, as in other marcopoloichthyids.
Unfortunately, its abdominal or precaudal region is poorly preserved and often partially disarticulated; thus, the presence of short epineural processes (as in M. furreri) could not be confirmed in this species, as well as the structure and number of supraneurals dorsally placed to the neural spines.The total number of supraneural bones is unknown in M. mirigioliensis n. sp.The series extends up to the expanded, plate-like, compound first dorsal proximal radial, and it does not extend between the most anterior proximal radials, as in Marcopoloichthys furreri.However, it extends between the anteriormost proximal dorsal radials in Marcopoloichthys ani (Tintori et al. 2007: Fig. 4).
The first dorsal pterygiophore of Marcopoloichthys mirigioliensis n. sp. is a compound element that resulted from the fusion of ca.four proximal radials.It is axeshaped.The same element in M. furreri has a similar shape (see Table 1), and it is also formed by four proximal radials (Arratia, 2022: Figs. 9, 11, 12).The shape of this compound element may change among marcopoloichthyids (Table 1).An axe-shaped first proximal radial has also been described for Marcopoloichthys ani (Tintori et al., 2007).
Fin-rays are commonly poorly preserved in marcopoloichthyids and their distal tips are destroyed or absent due to their fragile structure.Nine to 11 pelvic rays were reported for Marcopoloichthys furreri (Arratia, 2022) and 11 rays for Marcopoloichthys ani, but the number of rays remains unknown for M. andreetti and M. faccii (Tintori et al., 2007).For counts of fin rays in marcopoloichthyids, see Table 1.The fin rays of paired fins and of the dorsal and anal fins are thinly ossified in marcopoloichthyids.They are segmented and distally branched, including generally only one branching, as observed in Marcopoloichthys furreri, with better preserved rays (Arratia, 2022).
The tail of marcopoloichthyids (including the new species; Figs. 3, 7A, 11) is homocercal, with both lobes of similar shape and size, and five hypurals somehow producing a fan-shaped figure.This pattern differs from other Triassic teleosteomorphs (e.g., Prohalecites and pholidophorids), which have a hemiheterocercal tail (Arratia, 2013(Arratia, , 2015;;Arratia & Tintori, 1999;Tintori, 1990).Contrary to other stem teleosteomorphs where the caudal endoskeleton is known, the Swiss marcopoloichthyids do not possess epural bones (Fig. 11; Arratia, 2022: Figs. 13, 14); however, six dorsal epurals were identified in M. ani, but these elements have an uncommon position above the neural spines of preural centra 3-5 and arcocentral elements in front of hypurals 1-3, which are followed by an unnamed element that is fragmented distally (Tintori et al., 2007: Fig. 5).In contrast, the region above the hypurals (ural region, but only represented by the notochord, without ural centra) present about three broad uroneurals in M. furreri, but only one expanded plate-like uroneural fills the epaxial region in front of the bases of the hypurals in Marcopoloichthys mirigioliensis n. sp.(Fig. 11).These chondral expanded elements or just one plate would increase the stiffness of the tail during locomotion, as the independent uroneurals do in other teleosts.The interpretation here is that these elements, as in the "uroneural-like" ones in pachycormids and other Jurassic stem teleosts, represent different morphologies of the teleostean uroneurals or modified ural neural arches (e.g., Arratia & Lambers, 1996;Arratia & Schultze, 2013;Schultze & Arratia, 2013).
The urodermals seem to be variably present in marcopoloichthyids or their absence may be the result of poor preservation, but urodermals have been mentioned for Marcopoloichthys ani (Tintori et al., 2007: Fig. 5) and the new species here described (Fig. 11; Table 1).They also are present in another scaleless fish, Prohalecites, where they are numerous (Arratia & Tintori, 1999).

Feeding in marcopoloichthyids
The head of marcopoloichthyids looks very different in specimens that died when they were feeding (illustrating highly protractile jaws) and those not feeding (with the mouth closed), and these differences have been nicely illustrated for Marcopoloichthys furreri (see Arratia, 2022: Fig. 7).In contrast, the studied specimens of Marcopoloichthys mirigioliensis n. sp.usually are preserved with the mouth closed (Fig. 3).Interpreting the morphological differences between both jaw positions, including positional changes of some cranial bones, the function of additional bones, and the massive ossification of some bones was discussed in Arratia (2022), and some of these are summarized here.One of the major changes is in the shape of the head, which is somewhat triangular when the mouth was closed, whereas there is an antero-posterior elongation of the cranium that is accompanied with a dorso-ventral compression during suction, which can be referred to as "functional integration" (Klingenberg, 2014).For example, changes in the shape of the upper and lower jaws and expansion of the skull in M. furreri were significant enough to generate an intraoral pressure to draw water and prey into the mouth (as has been demostrated for extant fishes (e.g., Day et al., 2015;Lauder, 1985;Wainwright et al., 2015).Such action involved multiple integral components, as for instance, the T-shaped mesethmoid, nasals, accessory nasals, elongate upper jaws, the enlarged suspensorium, the cleithrum-strongly expanded antero-ventrally-as well as the clavicle, and support of the pectoral fins by the scapula and coracoid, whose integrated kinetic movements maximized suction forces and the chance to engulf prey.The integration of these mechanisms during prey capture also involves lower jaw length and the length of the ascending process of the premaxilla in extant teleosts (Kane et al., 2019).Interestingly, the lower jaw of Marcopoloichthys has a moderate length, and its articular region with the quadrate is placed at about the level of mid-orbit, but when the fish was feeding, the lower jaw displaced anteriorly to below the anterior half of the orbit, closer to the anterior orbital margin.Marcopoloichthys did not have an ascending process in the premaxilla (which is present with varying degrees of development in extant teleosts).The anterior articular regions of the premaxilla and maxilla were weakly developed in marcopoloichthyids (Figs. 5,6), and when the fish was feeding, both bones displaced anteriorly, forming the lateral walls of the buccal tube, together with a strongly ossified mesethmoid, which was loosely articulated postero-laterally with the nasal bones.The characteristics of the bones forming the buccal tube and suspensorium of the lower jaw in marcopoloichthyids and their comparison to extant teleosts support the interpretation that they were suction feeders.
The jaws of Triassic teleosteomorphs are formed consistently by the same bones: an upper jaw comprising a small, mobile premaxilla, a maxilla, and two supramaxillae (which may be absent in some taxa, marcopoloichthyids), and a lower jaw formed laterally by a dentary or dentalosplenial, angular, and surangular.Although not to the degree of marcopoloichthyids, other Triassic teleosteomorph jaws (e.g., pholidophorids) show some protractibility and were interpreted also as suction feeders by Arratia and Schultze (2024).There is no information concerning the diet of most Triassic stem teleosts, except Prohalecites and the pholidophorid Pholidorhynchodon malzannii that were interpreted as carnivorous fishes by Arratia and Schultze (2024).The remains of food in the stomach of Marcopoloichthys mirigioliensis n. sp., as described above, indicates a carnivorous diet.

Size and miniature stem teleosteomorphs
Among marcopoloichthyids, M. mirigioliensis is the smallest species, reaching a standard length between 30 and 35 mm, with an average of ca.32 mm, whereas M. furreri appears to reach the largest size of ca.40 mm (see Table 1).These values make marcopoloichthyids a small body-sized species group (see Arratia & Schultze, 2024) and M. mirigioliensis n. sp. a candidate to be interpreted as a miniature stem teleosteomorph.As Schultze et al. (2022) and Arratia & Schultze (2024: p. 18-19) noted, most Triassic stem teleosts are small-bodied in comparison to contemporaneous non-teleost fishes.Myers (1958) was the first one to address what today is understood as "miniaturization".He noticed a high frequency of reductions and paedomorphies in the anatomy of certain fish groups, including reduction of scales, number of fin rays, and simplification of the sensory canal system in very small fishes.Later, miniaturization, was formally defined for extant fishes by Weitzmann and Vari (1988), to involve individuals reaching sexual maturity at 20 mm SL or less and not growing longer than 26 mm SL, and usually exhibiting paedomorphic characters.However, such views have being challenged by Burns and Bragança (2024) who proposed a broader concept of miniaturization in which miniature fishes are the ones that have undergone significant reduction in adult body size when compared to their sister group and that exhibit some degree of morphological simplifications (i.e., reductions and losses).This new concept considers the importance of including phylogenetic relationships in the study of miniatures.Thus, following Myers' (1958) ideas, the absences of scales and teeth in the jaws in marcopoloichthyids can be interpreted as paedomorphic features.Another possible paedomorphic feature is the reduction or absence of the supraorbital canal in the skull roof bones; this canal has not been observed in marcopoloichthyids.The phylogenetic position of marcopoloichthyids among teleosteomorphs is unclear; Marcopoloichthys furreri (the most complete-known marcopoloichthyid) stands in a trichotomy with (large) Aspidorhynchiformes and more advanced teleosts of different sizes (Fig. 12) in the phylogenetic hypothesis of Arratia (2022: Fig. 16); consequently, the possible sister group of Marcopoloichthyidae is still unknown.

Taxonomic comments
According to the available information, with 28 specimens, Marcopoloichthys mirigioliensis n. sp. is based on the largest sample of marcopoloichthyids from one specific deposit.However, the quality of preservation is not as outstanding as that of M. furreri from the Prosanto Formation of southeastern Switzerland (Arratia, 2022), making the description difficult for some features, and a few others are unknown (e.g., the state of fusion in the skull roof bones or the condition of the abdominal or precaudal vertebral region).The new fish has the distinctive synapomorphies of Marcopoloichthyidae and although the information is incomplete for M. andreettii and M. faccii, Marcopoloichthys mirigioliensis differs from all other species known in numerous characters (see Diagnosis and Table 1), supporting its status as a new species.

Fig. 1
Fig.1Locality and stratigraphic position of the described fish material from the UNESCO World Heritage site Monte San Giorgio (Southern Alps of Canton Ticino, southern Switzerland), compared to the section at Ducanfurgga near Davos (Upper Austroalpine Silvretta Nappe of Canton Graubünden, southeastern Switzerland; modified fromScheyer et al., 2017)

Fig. 12
Fig. 12 Phylogenetic relationships of stem teleosts illustrating the position of Marcopoloichthys standing in a trichotomy at Node D (abbreviated from Arratia, 2022: Fig. 16)

Table 1
Arratia (2022) (2007) vertebrae, fin rays and other features in species of Marcopoloichthys and their geographic provenance and age Data of Marcopoloichthys ani, M. andreetti and M. faccii are fromTintori et al. (2007)and those of M. furreri are fromArratia (2022)