3.1.1. Pseudooneotodus mitratus (Moskalenko Reference Moskalenko1973)

The type horizon of this species, the morphological simplicity of which is misleading, is the late Sandbian Baksan formation of Siberia (Moskalenko Reference Moskalenko1973). Its probable ancestor P. nostras (Moskalenko Reference Moskalenko1973) comes from the underlying early Sandbian Chertovo formation. It differs from P. mitratus in a rounded T-outline, one of the lobes having incipient denticulation (Moskalenko Reference Moskalenko1973, pl. 17:12–15). Moskalenko (Reference Moskalenko1973, pl. 17:16) also illustrated and described in open nomenclature a possibly congeneric specimen from the late Darriwilian Kirensk formation with all lobes prominently tuberculated. This suggests proximity to Polonodus as probable relative.

P. mitratus appeared in Baltica in the Baltoniodus navis Zone (boulder E-360, supplementary Table 41; Fig. 2a) and continued its occurrence to the Amorphognathus ordovicicus Zone (Fig. 5), usually being represented by a single element per sample. An exception is the borehole Lesieniec IG 1 with a sample of E. robustus Zone (Les-1371/3; supplementary Table 56) that yielded seven elements and similar occurrences in the A. superbus Zone (supplementary Table 66). This does not allow reasonable restoration of the apparatus but it is likely that it consisted of only one pair of elements. The Silurian species of the genus are represented by elements with a virtually circular outline (e.g., Sansom Reference Sansom1996).

3.1.2. Lundodus gladiatus (Lindström Reference Lindström1955a)

The species occurs in Baltica in the Prioniodus elegans and Oepikodus evae Zones. Bagnoli & Stouge (Reference Bagnoli and Stouge1997) proposed its apparatus reconstruction with denticulated S elements slightly resembling those of Stolodus, but this is based on a small material and requires verification. Five P elements in the Ottenby section sample Ot-10 (supplementary Table 2) in my material do not add anything to its knowledge.

3.1.3. Nericodus capillamentum Lindström Reference Lindström1955a

The ancestry of Polonodus was proposed by me to be in a species illustrated from Ottenby as Nericodus (?) sp. (Dzik Reference Dzik1983) but apparently conspecific with Polonodus? lofgreni Stouge & Bagnoli Reference Stouge and Bagnoli1988. Löfgren (Reference Löfgren1985, fig. 4aab; also Löfgren Reference Löfgren1990, fig. 4p) illustrated its S elements and noticed (Löfgren Reference Löfgren1990, p. 256) that ‘these ramiforms are much less Baltoniodus-like than the ramiforms’ attributed by her to Polonodus. Moreover, there was a large gap in the fossil record between these chronospecies. It has to be pointed out that they are very rare in the Baltic region and appeared there as immigrants from elsewhere, being abundant in Newfoundland and South America. Possibly the lineage can be traced back to the earliest Floian and latest Tremadocian in the succession of species of Kallidontus (Pyle & Barnes Reference Pyle and Barnes2002) from British Columbia. In the course of evolution of this lineage, P elements developed ramification similar to that in Polonodus and S elements became gracile and denticulated like homologous elements of Pygodus. ‘Polonodus’ corbatoi (Serpagli Reference Serpagli1974) from Argentina may belong to the same lineage (Lehnert Reference Lehnert1993; Albanesi Reference Albanesi1998). Probably Nericodus Lindström Reference Lindström1955a, with N. capillamentum Lindström Reference Lindström1955a from the late Volkhov age Lower Limbata Limestone of Yxhult as its type species (Miller Reference Miller1980), has nomenclatorial priority over Kallidontus. Fragmentary elements from sample E-360, as well as some other samples, may be conspecific with it (Fig. 2b, c).

3.1.4. Polonodus clivosus (Viira Reference Viira1974)

The holotype of Ambalodus clivosus Viira Reference Viira1974 from the Ohesaare borehole depth of 510.35 m of the late Kunda age was originally interpreted as a sinistral element with bifurcated processes (probably P₁). It co-occurs in the same sample as another specimen (probably P₂) interpreted as a dextral element named Ambalodus? sp. n. The illustrated paratype of P. clivosus comes from coeval strata at the Ab'ya borehole. Another late Kunda age sample from the Ohesaare borehole (depth 510.12 m) yielded specimens closely similar to the holotype of P. clivosus, described as Amorphognathus? sp. n. This set of elements provides enough evidence to substantiate attribution of specimens from the Histiodella kristinae Zone of Newfoundland to the same species (Stouge Reference Stouge1984). It co-occurs with the more derived Polonodus tablepointensis Stouge Reference Stouge1984, in respect to bifurcation of processes. I suggest that these are extreme morphotypes within the same population and that the Newfoundland assemblage of Polonodus represents an ontogenetic series of gradual ramification of processes in P₁ elements (Fig. 8). Juvenile specimens (Stouge Reference Stouge1984, pl. 13:10) referred to P. newfoundlandensis from older strata (with H. tableheadensis) show four processes with bifurcation that develops at some distance from the cusp in the P₁ element. This refers also to the Chinese Dzikodus hunanensis Zhang Reference Zhang1998b that seems closely related to P. clivosus. However, the pattern of ramification of its ventral(?) process in relatively robust P₁ elements may not be homologous to that in P. tablepointensis. Zhang (Reference Zhang1998b) noticed that P₂ elements do not form mirror-image pairs in P. tablepointensis and D. hunanensis. She proposed to use this difference to erect a separate genus.

The brittle thin-walled Baltic specimens of Polonodus are never complete enough to exhibit a different ramification from that of the holotype of P. tablepointensis. All specimens illustrated by Viira (Reference Viira1974; also Dzik Reference Dzik1976, fig. 29c, d) represent P₁ elements, but the specimen from sample E-080 photographed in SEM by me (Dzik Reference Dzik1976, pl. 43:1) is a P₂ element. I doubt if more than one species of Polonodus occurred in Baltica.

An extrapolation of the trend towards ramification of processes and thinning of the crown in P₁ elements may suggest an alternative for the ancestry for the Polonodus lineage. Either this is a balognathid, with the apparatus rooted in that of Trapezognathus, as interpreted by Zhang (Reference Zhang1998b) or rather a fryxellodontid rooted in Kallidontus lofgreni as proposed by me (Dzik Reference Dzik1983). The nicely documented origin of Pygodus from Polonodus (Zhang & Sturkell Reference Zhang and Sturkell1998; Zhang Reference Zhang1998a) implies that conodonts of these genera had a similar architecture of their apparatuses. The S elements of Pygodus are of bizarre shapes and denticulation (inter alia, McCracken Reference McCracken1991; Zhen et al. Reference Zhen, Wang, Zhang, Bergström, Percival and Chen2011b). The balognathid elements interpreted by Löfgren (Reference Löfgren1990) and Zhen et al. (Reference Zhen, Cooper, Simes and Percival2011a) as S and M elements of Polonodus may represent an extreme morphotypes of Baltoniodus.

3.1.5. Pygodus trimontis Hamar Reference Hamar1966

Pygodus lunnensis Zhang Reference Zhang1998b and P. anitae Bergström Reference Bergström1983 form a connecting link between Polonodus and Pygodus (Zhang & Sturkell Reference Zhang and Sturkell1998). A complete succession of Pygodus species with a set of S elements was assembled by Zhang (Reference Zhang1998a).

Specimens of Pygodus characterised by the wider space between the ventral denticle rows on the P₁ elements do not seem to show extremes of population variability of P. serra but rather represent a separate biological species partially coeval with it (Fig. 8). In the borehole Lesieniec IG 1 it is represented by 74 specimens (sample Les-137 1/3; supplementary Table 56). The name proposed by Hamar (Reference Hamar1966; and perhaps P. xinjiangensis Wang & Qi Reference Wang and Qi2001) is available for it. Its occurrence is inserted between ranges of P. serra and P. anserinus both in the Baltic and Małopolska Massif material. In other regions of the world, it continued to occur well within the range of typical P. anserinus (Hamar Reference Hamar1966; Zhang Reference Zhang1998a).

3.1.6. Pygodus serra (Hadding Reference Hadding1913)

The topotype P₁ specimen from Fågelsång shows a wide space between the dorsal denticle rows (that houses an additional fourth row in P. anserinus), the morphology represented in the Eoplacognathus lindstroemi and latest E. robustus Zones (Bergström Reference Bergström2007). It is thus conspecific with P. protoanserinus Zhang Reference Zhang1998a, which seems to be a part of the continuous transition to P. anserinus (Fig. 8; Dzik Reference Dzik1994, fig. 26). In the Lesieniec IG 1borehole (sample Les-137prz), specimens transitional to P. anserinus occur; that is, with denticles forming incipient four rows, but the actual transition towards the fourth carina probably took place a little later.

3.1.7. Pygodus anserinus Lamont & Lindström Reference Lamont and Lindström1957

The fourth row of tubercles initially developed at maturity and gradually expanded to earlier and earlier stages in the evolution of this chronospecies. The holotype has this row of tubercles weakly developed, indicating its position in the early stage of evolution, which is well represented in the fossil record (Bergström Reference Bergström1971). A few S elements co-occur with P ones in boulder E-351. Paiste et al. (Reference Paiste, Männik and Meidla2022) attributed a geniculate M element to Pygodus anserinus but its incipient platform and sharp denticulation makes it different from associated unquestionable ramiform elements of Pygodus.

3.2.1. Acanthodus uncinatus Furnish Reference Furnish1938

Löfgren (Reference Löfgren1997a) interpreted specimens of this species occurring at the Swedish localities Brattefors and Oreholmen as reworked from older strata. The specimen from Brattefors with dorsally gaping basal cone (Löfgren Reference Löfgren1997a, fig. 5b) fits the probably homologous element of Drepanodus expansus Chen & Gong, 1986 sensu Pyle & Barnes (Reference Pyle and Barnes2002, pl. 6:7). Its affinity to Drepanoistodus, proposed by these authors, seems unlikely because geniculate M elements of this genus are missing at Brattefors. Agematsu et al. (Reference Agematsu, Sashida, Salyapongse and Sardsud2008) described specimens from Thailand similar to those from Brattefors.

3.2.2. Variabiliconus variabilis (Lindström Reference Lindström1955a)

Löfgren et al. (Reference Löfgren, Repetski and Ethington1999) presented the apparatus reconstruction and proposed its close relationship to the North American Midcontinent species, V. bassleri (Furnish Reference Furnish1938). Rare specimens of the species occur throughout the Ottenby section (supplementary Table 2).

3.2.3. Strachanognathus parvus Rhodes Reference Rhodes1955

The species is relatively rare in Baltica, but in large samples discrete types of its elements can be arranged into symmetry transition series, including the bilaterally symmetric S₀ element, strongly asymmetric probable M element and relatively robust P elements (Fig. 7; Dzik Reference Dzik1994). The specimen of Parapaltodus simplicissimus Stouge Reference Stouge1984 illustrated by Zhen et al. (Reference Zhen, Percival, Cooper, Simes and Wright2009a, fig. 2n) differs from Strachanognathus only in the lack of ventral denticle and may be related to its ancestor. P. simplicissimus preceded S. parvus in Baltica but is very rare in samples (Stouge & Bagnoli Reference Stouge and Bagnoli1990; Rasmussen Reference Rasmussen2001; Löfgren Reference Löfgren2003; Stouge & Nielsen Reference Stouge and Nielsen2003).

3.3.1. Scandodus rhomboideus (Pander Reference Pander1856)

According to Bergström (Reference Bergström1988), S. furnishi Lindström, 1955 is its younger synonym. The holotype of S. furnishi (Lindström Reference Lindström1955a, pl. 5:3) from the early Latorpian (Hunneberg) Lower Planilimbata Limestone of Skultorp has its dorsal (‘anterior’) tip of the base broken, but it seems to be gaping there. Lindström (Reference Lindström1971), van Wamel (Reference van Wamel1974; also Johnston & Barnes Reference Johnston and Barnes2000) included S₀ elements of Drepanoistodus morphology in the apparatus and transferred it to this genus under the name D. conulatus (Lindström Reference Lindström1955a). In the present samples from Ottenby the set of large hyaline elements, including those similar to the holotype of S. furnishi, resembles rather those of Drepanodus (Fig. 9). Admittedly, I am not sure of its taxonomic determination.

3.3.2. Tropodus comptus (Branson & Mehl Reference Branson and Mehl1933)

Bagnoli et al. (Reference Bagnoli, Stouge and Tongiorgi1988) and Tolmacheva et al. (Reference Tolmacheva, Fedorov and Egerquist2003) identified the main element types of the apparatus. The most characteristic is its S₀ element with five ridges forming incipient processes. Tropodus sweeti (Serpagli Reference Serpagli1974) as interpreted by Lehnert et al. (Reference Lehnert, Keller and Bordonaro1998) may be its relative.

3.3.3. Scolopodus sublaevis Pander Reference Pander1856

This is the type species of Scolopodus occurring in Baltica from the late Prioniodus elegans to Microzarkodina ozarkodella Zone (Fåhraeus Reference Fåhraeus1982). Tolmacheva (Reference Tolmacheva2006) restored the apparatus (Fig. 7) and proposed to use the name S. striatus Pander Reference Pander1856 for this species because the S. sublaevis morphotype had not been identified by herself in the section studied by Pander (Reference Pander1856), but the latter remains the type species of the genus chosen by Lindström (Reference Lindström1955a).

3.3.4. Decoriconus peselephantis (Lindström Reference Lindström1955a)

Löfgren (Reference Löfgren1998) attributed elements of the apparatus to six ozarkodinid locations. I failed to see them in my material, probably because of not having enough experience with ‘simple cones’ (Fig. 7). There is apparently an evolutionary continuity of the Decoriconus lineage from the Tremadocian to Llandovery. I apply the chronospecies D. mercurius and D. pesequus introduced by Löfgren (Reference Löfgren1998) referring to their geological age rather than to morphology (Fig. 2m–t).

A finely striated specimen from the basalmost sample Ot-1 from the Ottenby section may be transitional between Scolopodus and Decoriconus but the evidence is hardly conclusive. Specimens from Thailand identified by Agematsu et al. (Reference Agematsu, Sashida, Salyapongse and Sardsud2008) as Parapanderodus striatus (Graves & Ellison Reference Graves and Ellison1941) lack the medial fissure but have a smooth base and the cusp ornamented with numerous riblets resembling in this respect Decoriconus.

3.4.1. Diaphanodus latus (van Wamel Reference van Wamel1974)

D. latus is probably conspecific with the Furongian to Tremadocian type species of Diaphanodus (Bagnoli et al. Reference Bagnoli, Barnes and Stevens1987, Reference Bagnoli, Faúndez and Stouge2015; Johnston & Barnes Reference Johnston and Barnes2000). Its apparatus is known owing to clusters described by Andres (Reference Andres1988). In Baltica it is known from the Oepikodus evae Zone at Lava, Russia (Tolmacheva et al. Reference Tolmacheva, Koren, Holmer, Popov and Raevskaya2001b). According to Szaniawski (Reference Szaniawski2015) it is a protoconodont (that is, a chaetognath). The crucial difference between the chaetognath grasping spines and conodont elements is their mode of growth: centripetal versus centrifugal – that is, secreted by the soft tissue located inside the basal cavity or by an epithelium covering the element from outside. In the former the growth increments are visible on the external surface of the grasping spine; in the latter growth lines are within the basal cavity, the element exterior remaining smooth. Some of the Diaphanodus elements show transverse striation on their surface (Szaniawski Reference Szaniawski2015, fig. 2f) but this may be an impression of internal surface of the crown tissue. The evidence for its centripetal growth is rather weak. Moreover, in some conodonts this tissue layer may be very thin and wrinkled (e.g., Dzik Reference Dzik2009; Dzik & Moskalenko Reference Dzik and Moskalenko2016, fig. 4).

Rasmussen (Reference Rasmussen2001) suggested the presence of a related species in the Oepikodus evae Zone at the Norwegian locality of Herram.

3.4.2. Stolodus stola (Lindström Reference Lindström1955a)

Most elements of the apparatus of S. stola are similar to those of Diaphanodus and Coelocerodontus in having a very deep basal cavity and sharp costae (Fig. 6), but distal portions of their cusps are straight, making them similar to S elements of prioniodontids and unlike the chaetognath spines (inter alia, Löfgren Reference Löfgren1993). In sample E-368 (54 elements, supplementary Table 37) two element types with a short cusp co-occur, one tricostate, the other bicostate. Possibly they correspond to P locations. Already Lindström (Reference Lindström1964, fig. 29) noticed that there are two kinds of gracile elements with four costae. Among those with costae developed at two sides there seems to be a bilaterally symmetric one.

3.4.3. Coelocerodontus trigonius Ethington Reference Ethington1959

The type horizon of this species is the Late Ordovician Stewartville Member of the Galena Formation in Iowa (Ethington Reference Ethington1959). Nowlan et al. (Reference Nowlan, McCracken and Chatterton1988), Zhang (Reference Zhang1998b) and Zhen et al. (Reference Zhen, Percival and Webby2003) presented composition of its apparatus. It includes not only sharply costate S elements but also robust elements with curved tips at least remotely resembling those in Diaphanodus and bicostate elements virtually identical with that illustrated by Rasmussen (Reference Rasmussen2001). Rare specimens in my samples lack ridges on their posterior or anterior sides but share an extremely deep basal cavity with those of C. trigonius.

3.5.1. Scalpellodus latus (van Wamel Reference van Wamel1974)

M and S₃ elements of this species have a somewhat wider base (Figs 2d–k, 5) than in its successor S. gracilis. The holotype of van Wamel's (Reference van Wamel1974) Protopanderodus latus comes from about 1.40 m above the base of the Horns Udde Formation near Horns Udde, Öland and co-occurs with B. navis (that is, Volkhov in age).

3.5.2. Scalpellodus gracilis (Sergeeva Reference Sergeeva1974)

Löfgren (Reference Löfgren1978) listed the differences in the M elements (her short-based drepanodiform) that have a sharper ventral keel on the base, more symmetrical in S. gracilis than in S. latus (Fig. 5). The angular bending was proposed to be higher above the base than is typical for the holotype S. latus S element of Kunda age of the Popovka River, Ingria (Sergeeva Reference Sergeeva1974). Even if these subtle differences between chronospecies are real, they are obliterated by the high population and within-apparatus variation, acknowledged already by Löfgren (Reference Löfgren1978) and illustrated by Stouge & Bagnoli (Reference Stouge and Bagnoli1990). The only obvious difference in respect to S. latus that I am able to see is the geological age. Scalpellodus viruensis Löfgren Reference Löfgren1978 may be conspecific with S. gracilis.

3.5.3. Walliserodus costatus Dzik Reference Dzik1976

The type population of this species comes from the basal, early Kunda age, part of the Mójcza Limestone in the Holy Cross Mountains, Poland. It differs from W. ethingtoni in multicostate S₀ and S₁ elements and in their more robust appearance (Fig. 7; Dzik Reference Dzik1994; Albanesi Reference Albanesi1998). Löfgren (Reference Löfgren1978, p. 116) referred early Darriwilian (Kunda age) populations of Walliserodus to ‘Paltodus’ iniquus Viira Reference Viira1974 but the holotype of the latter species is of late Darriwilian (Lasnamägi) age and morphologically closer to W. ethingtoni (Viira Reference Viira1974, text-fig. 124a, b).

Rasmussen (Reference Rasmussen2001, p. 60) claimed that the Silurian Acodus curvatus Branson & Branson Reference Branson and Branson1947 (=Paltodus debolti Rexroad Reference Rexroad1967), the type species of Walliserodus Serpagli Reference Serpagli1967, differs from the Ordovician species attributed to this genus in having a less complex apparatus, deeper basal cavities and higher contribution of symmetrical elements. Although there is undoubtedly some difference between them, a continuity of the lineage across the Ordovician–Silurian boundary was traced by Zhang & Barnes (Reference Zhang and Barnes2002). No geniculate M element proposed by Rasmussen (Reference Rasmussen2001) to characterise the genus Costiconus is associated with elements of Walliserodus morphology in my material.

3.5.4. Walliserodus ethingtoni (Fåhraeus Reference Fåhraeus1966)

S elements of this species differ from W. nakholmensis in bearing only few ribs and having a flat or even concave ventral surface (Fig. 7; Albanesi Reference Albanesi1998). The cross-section of the S₀ element is triangular. The holotype of ‘P’. iniquus is roughly coeval with the holotype of W. ethingtoni and I do not see any significant difference between these forms.

3.5.5. Walliserodus nakholmensis (Hamar Reference Hamar1966)

The ventral surface of S elements of this species is convex and their cross-section is more oval than in the previous species (Fig. 7). Some samples show transitional morphology, with varying convexity of elements, but to prove the evolutionary transition would require more numerous samples than those available to me at the moment.

3.5.6. Panderodus sulcatus Fåhraeus Reference Fåhraeus1966

The contents of the apparatus were illustrated by Viira et al. (Reference Viira, Mens and Nemliher2006b). This includes a symmetric S ₀ element with slits on both sides of the cusp (Fig. 6). It is noteworthy that elements with two slits occur also in some Llandovery species (Zhang & Barnes Reference Zhang and Barnes2002).

Although the apparatus structure of Panderodus points to Scalpellodus as the ancestor, no transition has been recognised in the Baltic region, as pointed out by Löfgren (Reference Löfgren1978). The oldest occurrence of Panderodus in the erratic boulders studied here is together with Baltoniodus norrlandicus.

Elsewhere, probably the earliest member of this evolutionary lineage is ‘Protopanderodus’ nogamii (Lee Reference Lee1975) from the Oepikodus evae Zone. Although Zhen et al. (Reference Zhen, Percival and Webby2003, p. 207) claim that the slit in its elements ‘is not a true panderodontid furrow, and disappears just before reaching the basal margin’, I am not able to discern any difference between it and that of Panderodus. The main difference between ‘P.’ nogamii and P. sulcatus is the alate appearance of the S₀ element in the former.

3.5.7. Panderodus gracilis (Branson & Mehl Reference Branson and Mehl1933)

The presence of only one slit in all elements of the apparatus (Fig. 6) seems to be the main difference of this species from P. sulcatus. This may mean that the medial symmetrical element was lacking in the apparatus. The virtually bilaterally symmetrical elements with prominent costae on both sides (but only one panderodont slit) reported by Trotter & Webby (Reference Trotter and Webby1994) may represent the S₁ location.

The slit in the panderodontid elements is narrower than the size of secretory cells known from imprints on various conodont elements, and could hardly hide any canal to transfer venom as proposed by Szaniawski (Reference Szaniawski2009); this is further contradicted by the mode of secretion of conodont elements, permanently covered with secretory tissue. Probably a better analogue would be the belemnite alveolar slit; that is, an attachment for a kind of ligament.

3.5.8. Panderodus panderi (Stauffer Reference Stauffer1935)

The specimens of a rather robust appearance that may belong to this species co-occur with Amorphognathus superbus in erratic boulder E-305 from Mochty, Poland (Dzik Reference Dzik1983).

3.5.9. ‘Belodella’ serrata Dzik Reference Dzik1976

This is clearly a close relative of Panderodus, as indicated by the presence of the ‘panderodont furrow’ (Fig. 7) but, unlike the Ordovician and Silurian Panderodus serratus Rexroad Reference Rexroad1967 (Mellgren Reference Mellgren2010), its fine denticles are inclined towards the element base and there is a sharp posterior ridge along the ventral margin of the element (Dzik Reference Dzik1976). Zhang et al. (Reference Zhang, Tarrant and Barnes2011, fig. 14.18) identified denticulated specimens also in P. gracilis.

3.5.10. Belodina compressa (Branson & Mehl Reference Branson and Mehl1933)

Leslie (Reference Leslie1997) proposed a complete reconstruction of the Belodina apparatus, composed of 14 asymmetric elements, and thus similar to that of the Silurian Panderodus as restored by Dzik & Drygant (Reference Dzik and Drygant1986). This may mean that the lineage of Belodina emerged from Panderodus after losing the medial S₀ element. Transitional forms are known from Laurentia (e.g., Nowlan et al. Reference Nowlan, McCracken and Chatterton1988). The species has been reported from the Åland Islands, Finland by Merrill (Reference Merrill1980).

3.5.11. Belodina confluens Sweet Reference Sweet1979

The late Ordovician B. confluens was a descendant of B. compressa (Branson & Mehl Reference Branson and Mehl1933), from which it differs in that the external margin of the largest element in the apparatus is regularly curved throughout its length, rather than distinctly flattened for a short distance above its basal corner. Associated elements with oblique orientation of sharp denticles belonged to the same apparatus, as shown by the fused cluster illustrated in Nowlan (Reference Nowlan1979, fig. 35.2) and reinterpreted by Leslie (Reference Leslie1997; also Bergström in Trotter & Webby Reference Trotter and Webby1994, p. 485).

3.5.12. Dapsilodus viruensis (Fåhraeus Reference Fåhraeus1966)

This is the oldest chronospecies of the Dapsilodus lineage (Fig. 7). Its apparatus composition was restored already by Löfgren (Reference Löfgren1978). The element size increased in its evolution and their appearance became more gracile (Dzik Reference Dzik1994).

3.5.13. Dapsilodus mutatus (Branson & Mehl Reference Branson and Mehl1933)

The oblique striae near the external margin of the relatively large elements make the species easy to distinguish (Fig. 7; Dzik Reference Dzik1994).

3.5.14. Scabbardella altipes (Henningsmoen Reference Henningsmoen1948)

This is a species abundant in subpolar Gondwanan conodont communities, but subordinate in Baltica. Its evolutionary origin remains to be determined. The typical panderodont furrow (Fig. 7) makes it similar to Dapsilodus, which frequently co-occurs but apparently had less deep- or cold-water preferences.

3.5.15. Parapanderodus striatus (Graves & Ellison Reference Graves and Ellison1941)

The distinction between the Laurentian populations of this species and its Baltic relative P. quietus Bagnoli & Stouge Reference Bagnoli and Stouge1997 (Fig. 7) remains unclear. Smith (Reference Smith1991) described fused clusters of Parapanderodus composed of nine and five elements, which supports the reconstruction of the apparatus based on isolated elements. They show gradation in their size, although the gradient seems opposite to that in the Silurian Panderodus from Podolia (Dzik & Drygant Reference Dzik and Drygant1986). Toxotodus? gabriellae Löfgren Reference Löfgren1998 from the O. evae Zone of Sweden may be related to the Parapanderodus lineage.

3.6.1. Cornuodus longibasis (Lindström Reference Lindström1955a)

This is an almost cosmopolitan and long-ranging species, easily identified owing to the expanded base of the P₁ element (Fig. 2v–aa; e.g., Zhen et al. Reference Zhen, Percival, Liu and Zhang2009b, Reference Zhen, Percival and Molloy2015a). It appeared for the first time in Baltica in the Paroistodus proteus Zone at Ottenby, found in sample Ot-5 (supplementary Table 2). According to Löfgren (Reference Löfgren1999b) and Johnston & Barnes (Reference Johnston and Barnes2000) there was only a single lineage of Cornuodus in the Baltic realm. This is contradicted by recurrent, although restricted, occurrence of elements with an obliquely cut base.

3.6.2. Cornuodus bergstroemi Serpagli Reference Serpagli1967

The obliquely cut base of some S elements, identical with those described by Serpagli (Reference Serpagli1967), as well as twisted cusp of the probable M element (Fig. 7), distinguishes this species from C. longibasis (Dzik Reference Dzik1994). Its oldest record in the present material is in the Eoplacognathus reclinatus Zone in the borehole Lesieniec IG 1 (sample Les-141/30; supplementary Table 53); as a rare element it occurs up to the Amorphognathus ordovicicus Zone. It may also be present elsewhere in older strata but is variously understood by other authorities.

3.6.3. cf. Teridontus gallicus Serpagli et al. Reference Serpagli, Ferretti, Nicoll and Serventi2008

A single specimen from sample E-389, associated with B. norrlandicus, closely resembles P₁ elements of T. gallicus from the P. deltifer Zone of Montagne Noire, France (Serpagli et al. Reference Serpagli, Ferretti, Nicoll and Serventi2008).

3.6.4. Semiacontiodus cornuformis (Sergeeva Reference Sergeeva1963a)

The holotype comes from the Kunda age stratum at Popovka (Sergeeva Reference Sergeeva1963a). According to Löfgren (Reference Löfgren1999a) this species immigrated to the Baltic region in the M. parva Zone together with the evolutionary emergence of Baltoniodus norrlandicus. In younger samples (e.g., E-271) the ridge on the cusp of S elements extends almost to the base. Perhaps these late populations deserve a separate chronospecies status. Leslie (Reference Leslie2000) proposed to resurrect the name Scandodus polonicus Spassov & Teller Reference Spassov and Teller1963 and apply it to a species having symmetrical elements with a shallow medial groove, relatively prominent lateral ribs and asymmetrical elements with prominent flange on one side. Such elements characterise the late S. cornuformis in my material. It remains unknown which bed in the Mójcza section was sampled by Spassov & Teller (Reference Spassov and Teller1963), but elements of Baltoniodus illustrated by them seem to represent B. norrlandicus that at Mójcza is associated with early S. cornuformis.

Leslie (Reference Leslie2000) suggested that the similarity of this species to the early Tremadocian species of Semiacontiodus is convergent and that there is no continuity between them. He included the Darriwilian species in Staufferella. I do not share this idea.

3.6.5. Semiacontiodus longicostatus (Drygant Reference Drygant1974)

Unlike S. cornuformis, S₀ elements of this species are roughly rectangular at the cusp base. Other S elements had a flat ridge that extends up to its base, and is not restricted to the distal part of the cusp. The holotype comes from the Uhaku age strata in the borehole Pishcha-16, depth 410 m; that is, from the Pygodus anserinus Zone (Drygant Reference Drygant1974). A morphology of S₀ transitional between S. cornuformis and S. longicostatus is represented in sample Ha-4 at Hälludden, Öland (supplementary Table 19b). A continuous gradation connects also S₀ elements with long cusps and those of S. bulbosus (Löfgren Reference Löfgren1978) morphology (for instance in sample E-355), which makes their status as separate species uncertain.

3.6.6. Semiacontiodus davidi Löfgren Reference Löfgren1999a

The S₀ element lacks lateral ribs (Fig. 6) and is medially sharp (Löfgren Reference Löfgren1999a). In the latter aspect it resembles the geologically younger, and possibly related, S. carinatus.

3.6.7. Semiacontiodus carinatus Dzik Reference Dzik1976

The prominent lateral ribs (Fig. 6) may speak in favour of the origin of S. carinatus from S. cornuformis instead of S. davidi. The contents of the apparatus were illustrated by Viira et al. (Reference Viira, Mens and Nemliher2006b).

3.6.8. Protopanderodus rectus (Lindström Reference Lindström1955a)

This is the oldest of the Baltic species of its lineage (Mellgren & Eriksson Reference Mellgren and Eriksson2006). The most characteristic aspect of most Protopanderodus species is the morphology of the M element (Fig. 7). In P. rectus its posterior surface is gently convex with a wide furrow along the dorsal margin (Fig. 2ah–am). It is not clear which species was counted by Bagnoli & Stouge (Reference Bagnoli and Stouge1997) as Protopanderodus sp. in sample 652 from Horns Udde, Sweden.

3.6.9. Protopanderodus cooperi (Sweet & Bergström Reference Sweet and Bergström1962)

Protopanderodus cooperi probably originated from P. rectus in the late Lenodus variabilis Zone or somewhat later (supplementary Table 15c) and differs from it in the gentle concavity of the external furrow on the posterior surface of the M element, the central convexity being bordered by such furrows on both sides (Fig. 7). This species is frequently referred to as P. robustus (Hadding Reference Hadding1913) but the holotype of Drepanodus robustus Hadding Reference Hadding1913 seems to me to be truly a Drepanodus (Albanesi Reference Albanesi1998). Its S element shows ‘a marked tendency to develop additional costae’ (Lindström Reference Lindström1955b, p. 108), which does not happen in Protopanderodus. Therefore, I follow Zhang (Reference Zhang1998b) in using the name P. cooperi for it. Protopanderodus formosus (Fåhraeus Reference Fåhraeus1966) is its junior synonym.

3.6.10. Protopanderodus parvibasis (Löfgren Reference Löfgren1978)

The posterior surface of the M element cusp is acutely convex and bordered by relatively flat surfaces on both sides in specimens from the Lenodus variabilis Zone (supplementary Table 44c). In this aspect it is transitional between P. rectus and P. graeai, being restricted to strata preceding those, in which P. graeai appears (its reports from the older strata should be confirmed).

3.6.11. Protopanderodus graeai (Hamar Reference Hamar1966)

The furrow in M elements along the inner margin is deep and delimits a sharp ridge along the middle of the cusp (Fig. 7). The margin of the base of all elements runs almost transversely, unlike the sinuous one in P. rectus.

3.6.12. Protopanderodus sulcatus (Lindström, 1955)

Together with P. floridus and P. calceatus this species belongs to the lineage characterised by the presence of a deep furrow along the posterior side of the cusp in some S elements (both sides in S₀ element; Fig. 5). Their M element has a narrow furrow along the posterior surface of the cusp (Fig. 2an–aq). In the boulder E-314a some S elements have the furrow very shallow and are closely similar to those of P. rectus. However, there is no discontinuity in the elements’ variation, especially in respect of M ones, that would substantiate distinction of two sympatric species. A single element M shows a sharp ridge resembling that in P. graeai, but its large size suggests that this is the end-member of population variability.

3.6.13. Protopanderodus floridus Bagnoli & Stouge Reference Bagnoli and Stouge1997

According to Bagnoli & Stouge (Reference Bagnoli and Stouge1997) it differs from P. sulcatus in that the central convexity on the posterior surface of the cusp of the M elements is angular and bordered by an external furrow.

3.6.14. Protopanderodus calceatus Bagnoli & Stouge Reference Bagnoli and Stouge1997

The external narrow furrow in the relatively flat cusp of M element is deeper than in P. floridus. The S elements developed a shallow additional furrow near the external margin (Fig. 5).

3.6.15. Protopanderodus varicostatus (Sweet & Bergström Reference Sweet and Bergström1962)

Zhen et al. (Reference Zhen, Cooper, Simes and Percival2011a) illustrated the topotype material from the Pratt Ferry Formation of Alabama. The external furrow on S elements is much deeper than in P. calceatus (Fig. 5). I referred such specimens to P. gradatus (Serpagli Reference Serpagli1974) that may be conspecific rather with P. sulcatus (Dzik Reference Dzik1994). Armstrong (Reference Armstrong1997) described under this name a form from Scotland transitional to P. liripipus. Bergström (Reference Bergström1971, p. 100) dated the transition between these chronospecies as the early Baltoniodus gerdae Zone.

3.6.16. Protopanderodus liripipus Kennedy Barnes & Uyeno Reference Kennedy, Barnes and Uyeno1979

This is a member of the lineage (Fig. 5) that terminated in P. insculptus (Branson & Mehl Reference Branson and Mehl1933), all elements of which bear a prominent denticle at the base. Transitional forms are known and both species have a high correlative potential (Nowlan et al. Reference Nowlan, McCracken and McLeod1997, pl. 2:10; Zhen et al. Reference Zhen, Percival and Molloy2015a).

3.6.17. Drepanodus concavus (Branson & Mehl Reference Branson and Mehl1933)

Owing to a wide population variability, the early populations of Drepanodus are difficult to classify (Fig. 9). The M elements from Ottenby (sample Ot-5) may represent this Midcontinent species as interpreted by Kennedy (Reference Kennedy1980) but the Drepanoistodus-like M element included in the apparatus by Ji & Barnes (Reference Ji and Barnes1990) may contradict this.

3.6.18. Drepanodus arcuatus Pander Reference Pander1856

The apparatus (Figs. 2ab–ag, 9) has been restored by Dzik (Reference Dzik1994) and Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2003) although with opposite order in elements notation. They interpreted the element proposed by me to be P₂ as the M one, referring to the similarity of the Drepanodus apparatus to that of Cordylodus. The counterargument can be put forward that there are two kinds of elements of Drepanodus with short cusps and a robust appearance, whereas the morphological disparate element with long cusp (pipaform) has a notch that resembles incipient geniculation. Until a complete fused cluster or natural assemblage is found, this problem will remain unsettled. Few large elements in samples from the O. evae Zone developed ridges along the dorsal side of the cusp, unlike later populations of D. reclinatus, in which most elements are of such morphology. According to Bergström (Reference Bergström1988), D. planus (Pander Reference Pander1856), referred to by Bagnoli & Stouge (Reference Bagnoli and Stouge1997, table 1, text-fig. 5), is synonymous with D. arcuatus. Zhen et al. (Reference Zhen, Percival and Molloy2015a, Reference Zhen, Zhang, Tang, Percival and Yu2015b) and Agematsu et al. (Reference Agematsu, Sashida, Salyapongse and Sardsud2008) reported this species from as ancient strata as the early Tremadocian Cordylodus lindstroemi Zone.

3.6.19. Drepanodus reclinatus (Lindström Reference Lindström1955a)

Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2003; also Zhen et al. Reference Zhen, Cooper, Simes and Percival2011a, Reference Zhen, Wang, Zhang, Bergström, Percival and Chen2011b) proposed to use this species name instead of D. robustus for the costate form of Drepanodus (Fig. 9). This is probably a reasonable way to avoid confusion resulting from different interpretation of Hadding's (Reference Hadding1913) holotype that ‘possesses several small but distinct lateral costae’ (Bergström Reference Bergström2007, p. 81) that is typical for Drepanodus, not Protopanderodus, as pointed out by Dzik (Reference Dzik2020).

3.6.20. Drepanodus parformis Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2003

This is a species of Drepanodus with very wide bases of S elements. Possibly specimens labelled Drepanoistodus? sp. in Dzik (Reference Dzik2020, fig. 8f–j) belong to it.

3.6.21. Drepanodus aff. santacrucensis Dzik Reference Dzik1994

The flat M elements associated with rather indifferent Drepanodus elements in a sample from the Gałajny IG 2 borehole (Gał-18sp; supplementary Table 59) resemble D. santacrucensis and may represent an initial stage towards its development from D. arcuatus.

3.7.1. Paltodus pristinus (Viira Reference Viira1970)

The apparatus of this species (Fig. 9) was partially restored by Szaniawski (Reference Szaniawski1980) and completed by Löfgren (Reference Löfgren1997a). These authors applied subspecies rank to it but this is in conflict with the biological understanding of subspecies as a geographic race. Closely related subspecies should not co-occur in the same locality or its understanding as a segment of evolutionary lineage (chronosubspecies) has to be indicated, as proposed by Dzik & Trammer (Reference Dzik and Trammer1980). P. pristinus differs from its successor P. deltifer in the lack of geniculation in the M element and lack of angulation at the anterior surface of P elements. It was reported to co-occur with its probable successors in the Uuga section in Estonia (Löfgren et al. Reference Löfgren, Viira and Mens2005) but it remains unclear whether this is a result of reworking or application of the vertical concept of chronospecies.

3.7.2. Paltodus deltifer (Lindström Reference Lindström1955a)

The apparatus (Fig. 9) was adequately illustrated in Bagnoli et al. (Reference Bagnoli, Stouge and Tongiorgi1988) and Löfgren (Reference Löfgren1997a). It differs from the preceding stage in evolution in the M element in having geniculate, anteriorly angular P elements, and in the S₀ element having lateral expansions of the base. Paltodus peracutus (Lindström Reference Lindström1955a) sensu Bagnoli et al. (Reference Bagnoli, Stouge and Tongiorgi1988) seems to be a morphotype within the population variability of P. deltifer.

3.7.3. Paltodus subaequalis Pander Reference Pander1856

The apparatus of this species (Fig. 9) differs from its predecessor, P. deltifer, in most of its elements bearing costae along the base of the cusp (Bagnoli et al. Reference Bagnoli, Stouge and Tongiorgi1988; Löfgren Reference Löfgren1997a).

3.7.4. Besselodus semisymmetricus (Hamar Reference Hamar1966)

The contents of the apparatus (Fig. 7) were illustrated by Leslie (Reference Leslie2000) and Viira et al. (Reference Viira, Mens and Nemliher2006b).

3.7.5. Paroistodus numarcuatus (Lindström Reference Lindström1955a)

This is a cosmopolitan species (Agematsu et al. Reference Agematsu, Sashida, Salyapongse and Sardsud2008). Its M element is of the shape typical for the genus but, unlike its younger species, only rarely do elements of the apparatus show the inversion of basal cavity (Löfgren Reference Löfgren1997b).

3.7.6. Paroistodus proteus (Lindström Reference Lindström1955a)

All elements of the apparatus developed an inverted basal cavity in their external parts that is not as much expanded as is in P. originalis (Löfgren Reference Löfgren1997b).

3.7.7. Paroistodus parallelus (Pander Reference Pander1856)

P and S elements of this species bear sharp ridges on both sides of the cusp (Fig. 9). According to Johnston & Barnes (Reference Johnston and Barnes2000), both costate and acostate elements co-occur in the lower part of the stratigraphic range of this species, suggesting ancestry in P. proteus (Löfgren Reference Löfgren1997b). However, there is no such transition between P. parallelus and P. originalis and their co-occurrence as separate species is accepted (Fig. 3h).

3.7.8. Paroistodus originalis (Sergeeva Reference Sergeeva1963a)

This species is similar to P. proteus (Figs 3i–m, 9), from which it apparently derived allopatrically, reappearing in the Baltic region after some time of almost exclusive occurrence of P. parallelus (Löfgren Reference Löfgren1997b).

3.7.9. Paroistodus horridus Barnes & Poplawski Reference Barnes and Poplawski1973

Löfgren (Reference Löfgren1997b) pointed out that denticulation of P and S elements developed in Paroistodus several times. Albanesi & Barnes (Reference Albanesi and Barnes2000) applied vertical concept of chronospecies to the apparently anagenetic evolution from P. originalis to P. horridus, which resulted in artificial overlap ranges of allegedly sympatric species.

3.7.10. Drepanoistodus basiovalis (Sergeeva Reference Sergeeva1963a)

This seems to be a cosmopolitan species (e.g., Albanesi Reference Albanesi1998; Agematsu et al. Reference Agematsu, Sashida, Salyapongse and Sardsud2008). The holotype comes from the Volkhov age strata at Volkhov (Sergeeva Reference Sergeeva1963a). Among a total of 1,029 elements of Drepanoistodus in the erratic boulder E-368, 13 M elements represent D. basiovalis, 266 of D. forceps, and there are 19 intermediate forms (Figs. 3q–v, 9).

D. stougei Rasmussen Reference Rasmussen1991, characterised by a rounded ventral part of the M element, may be within the population variability range of D. basiovalis. It was referred to as D. basiovalis by Stouge & Bagnoli (Reference Stouge and Bagnoli1990), D. aff. D. contractus by Bagnoli & Stouge (Reference Bagnoli and Stouge1997) and D. contractus by Dzik (Reference Dzik2020). The inverted basal cavity characterises some specimens but I am not convinced that this is not a trait of advanced ontogenetic age of the individuals.

3.7.11. Drepanoistodus contractus (Lindström 1955)

‘Oistodus’ contractus is characterised, according to Lindström (Reference Lindström1955a, Reference Lindström1955b), by a very short base (and extremely long cusp). M elements of such morphology in the present sample are not sufficient to decide whether they truly represent a separate species or are rather extremes of the population variability of D. basiovalis.

3.7.12. Drepanoistodus forceps (Lindström Reference Lindström1955a)

Apparatus elements of the distacodontid conodonts are of a rather simple morphology and very variable (Figs. 3n–p, 9), providing little information of taxonomic value. This especially refers to sympatrically co-occurring Dapingian and Darriwilian species of Drepanoistodus, population variability of which may exceed the difference between species. Dzik (Reference Dzik1983) performed a biometric study to delimit the most commonly occurring Baltic species. The length of the base, length of the cusp and height of the external margin of M elements were measured in 101 specimens from boulder E-116 of the Baltoniodus navis Zone and from boulder E-314 of the Oepikodus evae Zone (supplementary Tables 5, 37). As expected, the ranges of their variation strongly overlap, but at least in the case of the combined height of the external margin and the ratio of basis to cusp length, the standard deviations from the mean values do not overlap (Dzik Reference Dzik1983, fig. 8b). This indicates that they significantly differ from each other, which may be interpreted as an evolutionary change towards a longer and longer cusp and higher external margin of the M elements between the time horizons represented by the boulders.

Alternatively, there are at least two separate biological species (D. basiovalis and D. forceps) represented in the younger sample E-116, but only one of them (D. forceps) is present in the older sample E-314B. M elements in this sample bear a longitudinal ridge on the cusp, but all associated S and P elements have both sides smooth. The latter lineage, characterised by almost equal length of the base and cusp, probably evolved towards an acute external end and a prominent costa along the cusp, typical for even younger species of Venoistodus. It is noteworthy that the M specimens with acute ventral end are not the most common in samples with D. forceps.

The same goal to delimit Drepanoistodus species was undertaken by Rasmussen et al. (Reference Rasmussen, Eriksson and Lindskog2021) with application of basically different methodology. They applied principal component analysis for 58 specimens selected from more than 20 samples of different age and geographic origin. I find such an approach hardly meaningful in terms of evolutionary population biology.

An enigmatic aspect of early populations of probable D. forceps is the co-occurrence of rare S elements with prominent anterior costa (Fig. 3n, o). Even the very large number of elements in sample E-360 (supplementary Table 41) has not helped in solving the taxonomic status of these elements. They are tentatively interpreted as a plesiomorphic trait consistent with the stratigraphic position of the sample.

3.7.13. Venoistodus balticus Löfgren Reference Löfgren2006

The holotype of this species comes from the Eoplacognathus pseudoplanus Zone (Fig. 7); the type population is thus very close to its probable ancestor, D. forceps. The S elements identified by Löfgren (Reference Löfgren2006) have their sides almost smooth, with only indistinct ridge on the cusp.

3.7.14. Venoistodus venustus (Stauffer Reference Stauffer1935)

The holotype of V. venustus comes from the Sandbian Glenwood Formation of Minnesota (Stauffer Reference Stauffer1935) and the apparatus was described by Nowlan et al. (Reference Nowlan, McCracken and Chatterton1988). M elements of this morphology are associated with prominently costate S elements in the Holy Cross Mountains, Poland and Baltic strata of similar age (Dzik Reference Dzik1994). The oldest Venoistodus occurs in the erratic boulder E-334 of Eoplacognathus robustus Zone age (supplementary Table 58).

3.8.1. Prioniodus robustus Lindström Reference Lindström1955a

The species preceded P. elegans and may be the oldest denticulated member of the lineage in the Baltic region (Löfgren Reference Löfgren1985, Reference Löfgren1994; Rasmussen Reference Rasmussen2001; Tolmacheva et al. Reference Tolmacheva, Koren, Holmer, Popov and Raevskaya2001b, Reference Tolmacheva, Fedorov and Egerquist2003). Perhaps undenticulated ‘Prioniodus cf. gilberti’ of Löfgren et al. (Reference Löfgren, Viira and Mens2005) belongs to it, but an alternative relationship to Diaphorodus is likely as well. Zhen et al. (Reference Zhen, Zhang, Chen and Wang2023) proposed its Chinese ancestry.

3.8.2. Prioniodus elegans Pander Reference Pander1856

This is a species that was cosmopolitan for a brief environmental change event in the earliest Floian (e.g., Löfgren Reference Löfgren1978; Bergström Reference Bergström1988; Stouge & Bagnoli Reference Stouge and Bagnoli1988; Bagnoli & Stouge Reference Bagnoli and Stouge1997; Zhen et al. Reference Zhen, Zhang, Tang, Percival and Yu2015b).

3.8.3. Stiptognathus microdentatus (van Wamel Reference van Wamel1974)

This species preceded not only O. evae but also P. elegans (van Wamel Reference van Wamel1974, fig. 15) and probably has its relatives in the Canning Basin Emanuel Formation, Australia (Zhen & Nicoll Reference Zhen and Nicoll2009).

3.8.4. Oelandodus elongatus (Lindström 1955a)

Johnston & Barnes (Reference Johnston and Barnes2000) reported a cluster of at least four pairs of S elements, similar to each other, and a pair of M elements. The apparatus reconstruction by van Wamel (Reference van Wamel1974) of O. elongates erroneously includes Protoprioniodus elements. A poorly preserved juvenile M element from Ottenby (sample Ot-6) may represent this species.

3.8.5. Oelandodus costatus van Wamel Reference van Wamel1974

Elements of this species have an incipient platform at the base (e.g., Bergström Reference Bergström1988). Bagnoli & Stouge (Reference Bagnoli and Stouge1997) mentioned the occurrence at Horns Udde of ‘Cooperognathus aranda’, a highly derived representative of this Australian lineage (Zhen et al. Reference Zhen, Percival and Webby2003), but did not illustrate or describe the specimens.

3.8.6. Protoprioniodus papiliosus (van Wamel Reference van Wamel1974)

This is another derived species of Australian affinity (Zhen Reference Zhen2023), reported also from Argentina by Lehnert (Reference Lehnert1993) and from Newfoundland by Johnston & Barnes (Reference Johnston and Barnes2000).

3.8.7. Paracordylodus gracilis Lindström, 1955

Tolmacheva & Löfgren (Reference Tolmacheva and Löfgren2000; also Tolmacheva & Purnell Reference Tolmacheva and Purnell2002 and Tolmacheva Reference Tolmacheva2014) presented the apparatus structure of this deep-water species and identified biometric distinctions between populations from different regions of the world.

3.8.8. Oepikodus evae (Lindström Reference Lindström1955a)

Oepikodus communis (Ethington & Clark Reference Ethington and Clark1964), having a tetraramous S₀ element diagnostic for the genus, preceded O. evae (inter alia, Lehnert Reference Lehnert1995; Zhen et al. Reference Zhen, Liu and Percival2007). According to Pyle & Barnes (Reference Pyle and Barnes2002) it is a descendant of Acodus neodeltatus Pyle & Barnes Reference Pyle and Barnes2002, adenticulate but with M elements bearing an elongated dorsal process. This is unlikely and roots of the lineage are rather in Australia, where another adenticulate forms with tetraramous S ₀ elements have been found (Nicoll & Ethington Reference Nicoll and Ethington2004; Zhen & Nicoll Reference Zhen and Nicoll2009). Smith (Reference Smith1991) described a cluster of elements of O. communis, lacking M elements and with two sinistral P elements displaced to the centre of the cluster, which suggests its coprolithic (or regurgitated) origin. This suggests that early prioniodontids had apparatuses of composition similar to that of the ozarkodinids.

The S elements of O. evae bear a very long dorsal process (Fig. 9) that rarely is preserved in loose material (Landing Reference Landing1976, pl. 4:1). In the Argentinian San Juan Formation, O. evae is succeeded by O. intermedius Serpagli Reference Serpagli1974 that differs from its predecessor in lacking denticulation on the anterior process of P elements and apparently being the next step in the evolution (Serpagli Reference Serpagli1974). In a sense, this is a case of evolutionary reversal to the morphology of O. communis that differs from O. intermedius only in being of a more gracile appearance (Lehnert Reference Lehnert1995).

3.9.1. Diaphorodus delicatus (Branson & Mehl Reference Branson and Mehl1933)

Kennedy (Reference Kennedy1980), while proposing the genus Diaphorodus with D. delicatus as its type species, pointed out that the apparatus of the type species of Acodus, A. erectus Pander Reference Pander1856, remains unknown but A. deltatus is generally assumed to be representative for its genus (Zhen & Nicoll Reference Zhen and Nicoll2009). Diaphorodus delicatus differs from Acodus in having a more prominent cusp and longer processes. Perhaps together with Triangulodus amabilis Rasmussen Reference Rasmussen2001 from the O. evae Zone, it is an early chronospecies of the Triangulodus lineage.

3.9.2. Triangulodus brevibasis (Sergeeva Reference Sergeeva1963a).

The holotype comes from Volkhov strata at Popovka, Ingria (Sergeeva Reference Sergeeva1963a). The apparatus (Figs. 3a–g, 6) was restored by Lindström (Reference Lindström1971). Tolmacheva (Reference Tolmacheva2001, pl. 3:1–9 and 4:1–10) reported in open nomenclature two species of Tripodus from the Lava section in Ingria. The one from the upper part of the Oepikodus evae Zone differs from that occurring in the lower part in having longer and more widely disposed processes of S₀ elements. Perhaps they both represent parts of the Triangulodus lineage.

3.9.3. Triangulodus alatus Dzik Reference Dzik1976

The fossil record of this species is not in a continuity with T. brevibasis but, being different only in more prominent processes (Fig. 6), such a relationship seems likely.

3.9.4. Acodus deltatus Lindström Reference Lindström1955a

The apparatus of this species (Fig. 9) is well known (inter alia, Bagnoli et al. Reference Bagnoli, Stouge and Tongiorgi1988; Viira et al. Reference Viira, Mens and Nemliher2006b). As indicated above, it is closely related to D. delicatus and probably gave ancestry to Baltoniodus by developing denticulation on processes, but the transition remains untraced (Stouge & Bagnoli Reference Stouge and Bagnoli1990).

3.9.5. Baltoniodus triangularis (Lindström Reference Lindström1955a)

Stouge & Bagnoli (Reference Stouge and Bagnoli1990; Bagnoli & Stouge Reference Bagnoli and Stouge1997) considered this species as representing the Trapezognathus lineage. It was suggested by the morphology of its tetraramous S₁ element but is contradicted by the M elements shape (Fig. 7). As identified by Lindström (Reference Lindström1971), its dorsal process is not perpendicular to the cusp (Wu et al. Reference Wu, Calner and Lehnert2017). However, Bergström & Löfgren (Reference Bergström and Löfgren2009) allowed a very wide range of variability of M elements included in this species, ranging from the morphology typical of Baltoniodus (Bergström & Löfgren Reference Bergström and Löfgren2009, fig. 4t), through those with the undenticulated dorsal process perpendicular to the cusp, typical of early Trapezognathus (Bergström & Löfgren Reference Bergström and Löfgren2009, fig. 4c, r, z, ac), up to prominently denticulated specimens of Trapezognathus pectinatus morphology (Bergström & Löfgren Reference Bergström and Löfgren2009, fig. 4d, s). The same refers to S elements that probably represent co-occurring species of Baltoniodus and Trapezognathus.

3.9.6. Baltoniodus navis (Lindström Reference Lindström1955a)

According to Stouge & Bagnoli (Reference Stouge and Bagnoli1990, p. 10) this is the oldest recognised species of its lineage, but B. triangularis as understood by Lindström (Reference Lindström1971) and partially by Bergström & Löfgren (Reference Bergström and Löfgren2009, also Wang et al. Reference Wang, Stouge, Chen, Li, Wang, Finney, Zeng, Zhou, Chen and Erdtmann2009) better fits such place in the evolutionary sequence. Baltoniodus navis (Fig. 7) differs from its successor, B. norrlandicus, in a more robust appearance of all elements, as well as in shorter processes of P₁ and M elements (Fig. 4a–h). Large samples are necessary to trace the transition. Unfortunately, this is not offered by the Les-154ko to Les-148prz series of samples from the Lesieniec IG 1 core (supplementary Tables 39, 41). Attribution of Baltoniodus elements sets from the middle of the sequence to either chronospecies has to remain arbitrary.

3.9.7. Baltoniodus norrlandicus (Löfgren Reference Löfgren1978)

Stouge & Bagnoli (Reference Stouge and Bagnoli1990) understood this species in terms of the vertical chronospecies concept and they even introduced B. clavatus, corresponding to morphotype 8 on fig. 13 in Dzik (Reference Dzik1994). Potentially, it would be possible to interpret it in population concept terms and apply to the segment of the lineage with such modal value for S₃ elements, but this does not seem practical. Baltoniodus norrlandicus (Fig. 7) may be conspecific with Falodus parvidentatus Sergeeva, 1963.

3.9.8. Baltoniodus medius (Dzik Reference Dzik1976)

The species (Fig. 7) is characterised by a short but still distinct anterior process in S₂ (S₃ according to Dzik's Reference Dzik2015 interpretation).

3.9.9. Baltoniodus prevariabilis (Fåhraeus Reference Fåhraeus1966)

In this species the anterior process in S₂ is completely reduced (Fig. 7) and almost undistinguishable from that in S₃ elements.

3.9.10. Baltoniodus variabilis (Bergström Reference Bergström1962)

The contents of the apparatus (Fig. 7) were illustrated by Viira et al. (Reference Viira, Mens and Nemliher2006b). The posterior part of the basal cone in the P₁ element developed a platform with triangular outline.

3.9.11. Baltoniodus alobatus (Bergström Reference Bergström1971)

The platform in the P₁ element is robust but relatively narrow, of irregular outline (Fig. 7). A gradual transition connects B. variabilis with B. alobatus in the Holy Cross Mountains (Dzik Reference Dzik1994). This may mean that the latter species is an immigrant to Baltica that replaced B. gerdae.

3.9.12. Baltoniodus gerdae (Bergström Reference Bergström1971)

The range of this species, which is characterised by the denticulated process (Fig. 7) that replaced the platform in P₁, was traced within a 0.25 m thick interval in the Dalby Limestone in Sweden. It persisted longer in the Appalachians than in the Baltic region (Bergström Reference Bergström1971). In Estonia it is missing in some localities as a result of sedimentary hiatus (Paiste et al. Reference Paiste, Männik, Nõlvak and Meidla2020). A shallowing and erosional event is evidenced also by the co-occurrence of reworked Kunda age conodonts with B. gerdae in the boulder E-358. This presumably warm-water species is not present in the Holy Cross Mountains. (Dzik Reference Dzik1994).

3.9.13. Barrandegnathus bohemicus (Dzik Reference Dzik1983)

Stouge (Reference Stouge2005) reported this species, originally known only from the locality Mýto near Prague, from the late Volkhov B. norrlandicus to L. antivariabilis Zones at localities Skelbro in Bornholm, Denmark, as well as Limensgade, Fågelsång and Albjära in Skania, Sweden. According to Stouge & Nielsen (Reference Stouge and Nielsen2003), the Volkhov conodont communities at Fågelsång show a Gondwanan cold-water affinity whereas those of the Kunda age contain Laurentian elements.

3.9.14. Trapezognathus pectinatus (Dzik Reference Dzik2020)

The diagnostic aspect of the species is the prominently denticulated dorsal process of the M elements (Fig. 5). Zentagnathus gertrudisae Albanesi, 2023 from the early Darriwilian of Argentina (Albanesi et al. Reference Albanesi, Monaldi, Barnes, Zeballo and Ortega2023) seems conspecific. I suggest that some specimens attributed to B. triangularis by Bergström & Löfgren (Reference Bergström and Löfgren2009) belong to Z. pectinatus. Its presence in Morocco suggests the Gondwanan source area for its immigration to the Holy Cross Mountains and marginal areas of Baltica.

3.9.15. Trapezognathus diprion (Lindström Reference Lindström1955a)

P and S elements in this species are adenticulated (Fig. 8) or weakly denticulated (Bagnoli & Stouge Reference Bagnoli and Stouge1997). As pointed out above, the M elements differ from those of Baltoniodus in orientation of the dorsal process that is perpendicular to the cusp (Fig. 4i–h).

3.9.16. Trapezognathus quadrangulum Lindström Reference Lindström1955a

The P elements are weakly denticulated (Fig. 8; Stouge & Bagnoli Reference Stouge and Bagnoli1990) and with a sharp almost transversely oriented posterior lobe on the dorsal process (Viira et al. Reference Viira, Löfgren, Mägi and Wickström2001). Perhaps elements determined as Baltoniodus triangularis Carlorosi et al. (Reference Carlorosi, Heredia and Aceñolaza2013) belong here. Specimens from the Lesieniec IG 1 borehole (samples Les-154ko, 154tyl and 154śr) have incipiently denticulated posterior lobes and may represent advanced T. quadrangulum or T. falodiformis Sergeeva Reference Sergeeva1963b.

The holotype of Lenodus falodiformis is an M element of Volkhov age from the Popovka River, Ingria (Sergeeva Reference Sergeeva1963b). Its dorsal and ventral processes are finely denticulated but no ridge, diagnostic for Lenodus, is developed on the basal cone of the holotype. This makes it rather a member of the Trapezognathus grade, although not necessarily synonymous with T. quadrangulum as proposed by Stouge & Bagnoli (Reference Stouge and Bagnoli1990).

Lindström (Reference Lindström and Ziegler1977) applied the name Amorphognathus falodiformis for the transitional form between Baltoniodus and Amorphognathus but Bergström (Reference Bergström1983) pointed out that its apparatus was too poorly known to be of use. A few specimens from Hälludden, Öland were attributed to this species and have been illustrated by me (Dzik Reference Dzik1983, fig. 7a–e), but Rasmusssen (Reference Rasmussen1991) suggested that this is a morphotype within the range of variability of L. variabilis. The nomenclatorial problem remains to be settled (Stouge Reference Stouge1989).

3.9.17. Lenodus antivariabilis (An Reference An1981)

The species as understood by Bagnoli & Stouge (Reference Bagnoli and Stouge1997), with undivided posterior process of its P₁ (Pb in Bagnoli & Stouge Reference Bagnoli and Stouge1997) element and the posterior lobe of the dorsal process almost parallel to it (Fig. 8), is the first member of the Lenodus lineage, possibly next to T. falodiformis. Its Baltic successor, L. variabilis (but also L. antivariabilis as understood by Wu et al. Reference Wu, Calner, Lehnert, Lindskog, Erlström and Joachimski2018), has this process bifurcated. Zhang (Reference Zhang1998b) included in the Chinese L. antivariabilis forms even more advanced, having the ventral branch of the process longer that the dorsal one. This is a morphology close to that of Eoplacognathus zgierzensis.

The oldest occurrence of Lenodus M elements is in an erratic boulder from Mokrzeszów in the Sudetes, Poland, transported there in the most extensive Pleistocene glaciation (E-374; supplementary Table 17d). Their ventral margin is virtually smooth in juveniles but developed distinct denticulation in adults. The ridge on its basal cone extends into a weakly denticulated process of at least double the length of the cusp. The posterior process of the P₁ elements is variable, in juveniles lacking bifurcation that apparently developed later in the ontogeny. The holotype of the type species of Lenodus, L. clarus is an M element from the Kunda age strata at the Popovka River, Ingria (Sergeeva Reference Sergeeva1963b). Its ventral process is prominently denticulated; the posterior process is also finely denticulated. This makes its synonymy with L. antivariabilis likely.

3.9.18. Lenodus variabilis (Sergeeva Reference Sergeeva1963a)

The diagnostic trait of this species is a P₁ element with bifurcated posterior process (Fig. 8), its rami being of not more than equal length (Dzik Reference Dzik1994; Löfgren Reference Löfgren2003). Such an element has been found in sample Ha-4 at Hälludden, Öland (supplementary Table 19b).

3.9.19. Lenodus crassus (Chen & Zhang in Wang Reference Wang1993)

Dzik (Reference Dzik1976) suggested that the beginning of the Eoplacognathus lineage is marked by elongation of the dorsal branch of the posterior process in the P₁ element. In this respect L. crassus is at the Lenodus grade. Bifurcation of the posterior process with its branches of similar length points to early Lenodus variabilis as its probable ancestor (Zhang Reference Zhang1997). Although typical for the Yangtse Platform (Zhang Reference Zhang1998b), it had its Baltic acme of early Kunda age (Männik & Viira Reference Männik and Viira2012; S. Stouge, pers. comm. 2023).

3.9.20. Amorphognathus tvaerensis (Bergström Reference Bergström1962)

The contents of the apparatus (Fig. 6) were illustrated by Viira et al. (Reference Viira, Mens and Nemliher2006b). Dextral P₁ elements with relatively straight dorsal process have been found in the boulder E-173 and in the Lesieniec IG 1 borehole sample Les-134śr (supplementary Tables 60, 62).

3.9.21. Amorphognathus viirae Paiste et al. Reference Paiste, Männik and Meidla2023

Populations of Amorphognathus from the Baltoniodus alobatus Zone differ from those preceding in a sinuous dorsal process of the dextral P₁ element (Fig. 6; Paiste et al. Reference Paiste, Männik and Meidla2023).

3.9.22. Amorphognathus ventilatus Ferretti & Barnes Reference Ferretti and Barnes1997

This species precedes A. superbus Rhodes Reference Rhodes1955 in the Holy Cross Mountains, Poland (Dzik Reference Dzik1999) and Estonia, being a member of the Oandu age fauna (Nōlvak et al. Reference Nōlvak, Hints and Männik2006; Männik & Viira Reference Männik and Viira2012). Its origin from the advanced A. tvaerensis seems well substantiated by the M element morphology, with a fan-like arrangement of denticles (Fig. 6). Its type horizon, the Kalkbank horizon in Thuringia (Ferretti & Barnes Reference Ferretti and Barnes1997), seems to mark a global expansion of warm-water conodonts preceding the Late Ordovician cooling (Dzik Reference Dzik1999).

3.9.23. Amorphognathus superbus (Rhodes Reference Rhodes1953)

The diagnostic character of this species, distinguishing it from the ancestral A. tvaerensis, is the absence of postero-dorsal lobe in the dextral P₁ element (Fig. 6). M elements are very variable in populations of the late Amorphognathus and usually rare in conodont samples because of their small size (inter alia, Dzik Reference Dzik1994, Reference Dzik1999; Männik Reference Männik2017). This is the main cause of controversies in determining particular species. Ferretti et al. (Reference Ferretti, Bergström and Barnes2014a), in their taxonomically important study on the Late Ordovician conodonts from Wales, proposed co-occurrence of at least four species of Amorphognathus in the same sample and allowed up to nine sympatric species of this genus in their material. I find this biologically unrealistic. The within-sample variation of the diagnostic character for the species is unimodal throughout the Mójcza Limestone in the Holy Cross Mountains (Dzik Reference Dzik1990, fig. 17; Reference Dzik1999). Without considering frequencies of particular morphotypes within their Gaussian distribution, it is hardly possible to determine position of particular chronospecies in an evolutionary continuum.

3.9.24. Amorphognathus sp. n. Dzik Reference Dzik1999

Merrill (Reference Merrill1980) illustrated the M element from the Åland Islands, Finland, associated with other element types he attributed to A. complicatus Rhodes Reference Rhodes1955 and A. superbus (Bergström Reference Bergström1971), but it seems to represent rather this unnamed species.

3.9.25. Amorphognathus ordovicicus Branson & Mehl Reference Branson and Mehl1933

The only M element of the Katian Amorphognathus, with a dominant cusp diagnostic for this species (Fig. 6; Bergström & Leslie Reference Bergström and Leslie2010), has been found in erratic boulder E-207 (supplementary Table 69). Zhang et al. (Reference Zhang, Mirza and Barnes2016) illustrated its almost intact specimen. Other element types are undistinguishable from those of related species but presumably erratics of the Ostseekalk lithology are of similar age and Amorphognathus specimens from them represent A. ordovicicus.

3.9.26. Amorphognathus duftonus Rhodes Reference Rhodes1955

A characteristic dorsal spur on the M elements (Fig. 6) is the diagnostic trait of this species (e.g., Ferretti et al. Reference Ferretti, Bergström and Barnes2014a, Reference Ferretti, Messori and Bergström2014b), well known from Podolia (Dzik Reference Dzik1999). Despite the relatively sparse Late Ordovician conodont fossil record in Estonia (Männik & Viira Reference Männik and Viira2012), Kaljo et al. (Reference Kaljo, Hints, Männik and Nõlvak2008, fig. 8f, g) collected the diagnostic element of A. duftonus from the earliest Porkuni age bed immediately preceding the immigration of Noixodontus girardeauensis (Satterfield Reference Satterfield1971). They identified the species as A. ordovicicus and it is unclear if such determined conodont from the Hirnantian of Latvia (Hints et al. Reference Hints, Hints, Kaljo, Kiipli, Männik, Nõlvak and Pärnaste2010) is conspecific.

3.9.27. Rhodesognathus elegans (Rhodes Reference Rhodes1953)

Specimens from the borehole Lesieniec IG 1 (Les 124 1/3; supplementary Table 67), although rather poorly preserved, provide a complete set of element types of this species (Bergström & Sweet Reference Bergström and Sweet1966). Possibly conspecific elements of somewhat larger size have been found in the erratic boulder E-213 (supplementary Table 67).

The Rhodesognathus lineage is closely related to Amorphognathus (Fig. 5) but its exact ancestry remains unknown. I suggested (Dzik Reference Dzik1994) that the main trend in its evolution is the reduction of ramification in the platform of P₁ elements. If true, R. elegans would be the end-member of the lineage.

3.9.28. Sagittodontina sp.

The mid Darriwilian (Kunda age) early populations of the lineage, co-occurring with Eoplacognathus zgierzensis, differ from those of the late Darriwilian type species in a relatively wide angle between dorsal and ventral processes of P₂ elements. Also, the M element, probably erroneously referred to Eoplacognathus by Dzik (Reference Dzik1994, pl. 20:15) differs from that of the type species population in having a less posteriorward-oriented prominently denticulated postero-dorsal process (Fig. 8). M elements are invariably broken in the Baltic material but two relatively complete elements from the borehole Lesieniec IG 1 (sample Les-145) bear an extremely long transversely oriented posterior process, more than two times longer than the cusp and with 13 rather prominent denticles.

3.9.29. Sagittodontina kielcensis (Dzik Reference Dzik1976)

The species (Fig. 8) appears in Estonia in the P. anserinus Zone (Hints et al. Reference Hints, Viira and Nõlvak2012). After a long gap in occurrence, the last member of this lineage, Noixodontus girardeauensis, enters the Baltic region again close to the end of the Ordovician (Männik & Viira Reference Männik and Viira2012). It marks the global Hirnantian cooling (McCracken & Barnes Reference McCracken and Barnes1982; Bergström et al. Reference Bergström, Saltzman and Schmitz2006). Its co-occurrence with Kockelella and Ozarkodina (Männik & Viira Reference Männik and Viira2012) indicates surprisingly early immigration of the Silurian fauna from an unknown refuge (unless the Noixodontus specimens are reworked).

3.9.30. Eoplacognathus zgierzensis Dzik Reference Dzik1976

Erratic boulders E-211 and E-266 (supplementary Table 47) yielded P₁ elements with branches of the bifurcated posterior process of proportions diagnostic for E. zgierzensis. This is the most ancient member of the Baltic Eoplacognathus lineage (Fig. 5), characterised by the elongated ventral branch of posterior process in the P₁ element (the homologous bifurcated process was originally short in Lenodus and remained such in Sagittodontina, Amorphognathus and Rhodesognathus). The Chinese L. antivariabilis sensu Zhang (Reference Zhang1998b) had the ventral branch of the process slightly longer than the dorsal one, which may mean that the E. zgierzensis lineage is rooted there. Saadre et al. (Reference Saadre, Einasto, Nõlvak and Stouge2004) preferred to classify E. zgierzensis (as well as E. pseudoplanus) in Lenodus, although both lineages seem to co-occurred in the mid Kunda.

The M element with prominently denticulated postero-dorsal process, referred to E. zgierzensis by Dzik (Reference Dzik1994, pl. 20:15), belonged to a co-occurring unnamed early species of Sagittodontina. However, there is a gradation within the same sample ranging to morphologies typical of Lenodus, presumably characterising also E. zgierzensis.

3.9.31. Eoplacognathus pseudoplanus (Viira Reference Viira1974)

The holotype of Ambalodus pseudoplanus Viira Reference Viira1974 is a P₂ element that comes from the Kunda age strata in the borehole Ohesaare (depth 508.87 m; Viira Reference Viira1974, p. 54, fig. 43). It does not differ from the homologous element of E. zgierzensis found at 510.44 m. The distinguishing and diagnostic aspect of these species is the length of the dorsal branch of the anterior process of the P₁ element (Dzik Reference Dzik1976). Also, the posterior process in the P₁ element of this species was much more elongated that in the preceding E. zgierzensis (Fig. 5).

Stouge & Bagnoli (Reference Stouge and Bagnoli1990) described non-P elements in E. pseudoplanus, contradicting Dzik (Reference Dzik1976) in that non-P elements were lost in the course of evolution from Lenodus to Eoplacognathus. Dzik (Reference Dzik1994) already corrected this in respect to E. zgierzensis.

3.9.32. Eoplacognathus suecicus Bergström Reference Bergström1971

The holotype chosen by Bergström (Reference Bergström1971, pl. 1:6) is a juvenile dextral P₂ element. Specimens of such ontogenetic age do not differ much from those of stratigraphically preceding species and this has resulted in their different interpretation by some authors. The associated sinistral element, as well as sinistral elements from the type locality (Zhang Reference Zhang1999), shows the ventral process (Zhang & Sturkell Reference Zhang and Sturkell1998) significantly longer than that of E. pseudoplanus, from which E. suecicus evolved (Fig. 5). Diagnostic for the species is also the short and bifurcated posterior process of the P₁ element (Zhang Reference Zhang1999, fig. 1), in Estonia characterising the late Aseri age. Juveniles recapitulate the morphology of ancestral E. pseudoplanus (Zhang Reference Zhang1999, fig. 3). Specimens attributed to these species described by Mestre & Heredia (Reference Mestre and Heredia2019) from Argentina under the generic name of Lenodus truly represent species of this genus.

3.9.33. Eoplacognathus reclinatus (Fåhraeus Reference Fåhraeus1966)

The diagnostic P₂ elements of this species changed during their ontogeny, with the ventral process becoming longer. Mature specimens are closely similar to those of E. robustus and some of the differences between these chronospecies may be due to a change in population dynamics (distribution of mortality) rather than evolution, which was traced in Sweden by Bergström (Reference Bergström1971). There is an evolutionary continuity from this chronospecies (the type for Baltoplacognathus Zhang, Reference Zhang1998b) to E. lindstroemi (the type for Eoplacognathus Hamar Reference Hamar1966) eventually resulting in a decrease of asymmetry of P₂ elements, and their separation into different genera seems redundant (Fig. 5).

3.9.34. Eoplacognathus robustus Bergström Reference Bergström1971

The ontogenetic change of P₂ elements is expressed in the length of the dorsal process also in this species. It deserves its name owing to a generally large size of elements. Still, even juveniles are clearly different from the preceding and apparently ancestral E. reclinatus in respect to the dorsal process length (Fig. 5).

The erratic boulder E-294 yielded two S₃ elements of E. robustus. Heredia & Mestre (Reference Heredia and Mestre2019) identified some S and M elements of this apparatus, although their elements are much more gracile than elements under study here, and puzzlingly similar to those of Baltoniodus.

3.9.35. Eoplacognathus lindstroemi (Hamar, 1964)

Heredia & Mestre (Reference Heredia and Mestre2019), Paiste et al. (Reference Paiste, Männik and Meidla2022) and Ferretti & Bergström (Reference Ferretti and Bergström2022) proposed a controversial partial reconstruction of the apparatus including Baltoniodus-like ramiform elements (Fig. 5).

3.9.36. Eoplacognathus elongatus (Bergström Reference Bergström1962)

The characteristic star-like P₁ elements of this species (Fig. 5) occur from the latest Pygodus anserinus to early Baltoniodus gerdae Zones (Bergström Reference Bergström1971, p. 100).

3.9.37. Yangtzeplacognathus foliaceus (Fåhraeus Reference Fåhraeus1966)

According to Zhang (Reference Zhang1999), the first populations of Y. foliaceus co-occur with E. suecicus, which is probably their ancestor, but the transition has not been traced in the Baltic region (Fig. 5). In the Goldwyer Formation, Canning Basin, Australia, Y. foliaceus co-occurs with Lenodus, as shown by a short bifurcation of the anterior process in P₁ element, identified as E. suecicus by Watson (Reference Watson1988, pl. 6:33), that seems to agree with reinterpretation of its age by Zhen (Reference Zhen2020).

3.9.38. Yangtzeplacognathus protoramosus Chen et al. Reference Chen, Chen and Zhang1983

The Yangtse lineage of Yangtzeplacognathus reappeared in Estonia (Männik & Viira Reference Männik and Viira2012), the Holy Cross Mountains (Dzik Reference Dzik1994) and the Lesieniec IG 1 borehole (sample Les-138) in the late E. robustus Zone.

3.9.39. Cahabagnathus sweeti (Bergström Reference Bergström1971)

The holotype of C. sweeti comes from the base of the Effna Formation in Virginia and this chronospecies is a continuation of the North American Midcontinent lineage initiated by C. friendsvillensis (Bergström Reference Bergström1971; Leslie & Lehnert Reference Leslie and Lehnert2005). Drygant (Reference Drygant1974) based his Petalognathus bergstroemi on two P₂ specimens from the depth of 405 m in the Pishcha-16 borehole in Volhynia, Ukraine, of morphology virtually identical with the Appalachian type specimen. Its age is roughly determined by the occurrence of Eoplacognathus elongatus at the depth of 407 m. Another species exotic to Baltica but of the Yangtse affinity, Complexodus pugionifer, co-occurs in the same interval of the core (405–407 m) as well as in the Fågelsång section in Sweden, where B. variabilis has also been found (Bergström Reference Bergström2007), but in Estonia it ranges only up to the Pygodus anserinus Zone (Männik & Viira Reference Männik and Viira2012).

3.10.1. Phragmodus costulatus (Lindström Reference Lindström1955a)

The apparatus (Fig. 7) was restored by Bagnoli & Stouge (Reference Bagnoli and Stouge1997). This is probably the ancestral chronospecies of the Phragmodus lineage, an early offshoot of the balognathids that appeared in the Baltic region in the Oepikodus evae Zone (Dzik Reference Dzik1994, Reference Dzik2015; Bagnoli & Stouge Reference Bagnoli and Stouge1997). Gutiérrez-Marco et al. (Reference Gutiérrez-Marco, Albanesi, Sarmiento and Carlotto2008) claimed that it had its ancestry in the Andean part of Gondwana.

3.10.2. Phragmodus polonicus Dzik Reference Dzik1978

This chronospecies differs from P. costulatus in the sinuosity of the dorsal process of S_0–2 elements (Fig. 7; Dzik Reference Dzik1994). It occurs in the Holy Cross Mountains, Poland in the Lenodus variabilis Zone, and appeared somewhat later in Norway (as Nordiora torpensis Rasmussen Reference Rasmussen2001) and the Canning Basin, Australia (as Phragmodus polystrophos Watson Reference Watson1988; Zhen Reference Zhen2019). In Estonia, it occurs in the E. robustus to B. variabilis Zones (Viira et al. Reference Viira, Aldridge and Curtis2006a; Hints et al. Reference Hints, Viira and Nõlvak2012). Probably the Late Ordovician P. undatus Branson & Mehl Reference Branson and Mehl1933 is the last member of the lineage. It was typical for Laurentia but appeared also in the Baltic Oandu age together with Belodina confluens (Paiste et al. Reference Paiste, Männik and Meidla2022).

3.10.3. Complexodus originalis Chen & Zhang Reference Chen and Zhang1984

This species is of Yangtse origin but known also from Turkey (Kozlu et al. Reference Kozlu, Göncüoğlu, Sarmiento and Gül2002). Biometrical data from the Mójcza Limestone in the Holy Cross Mountains, Poland show that soon after its immigration to the region, C. originalis gradually developed sinuosity of the ventral branch of the anterior process in the course of evolution towards C. pugionifer (Dzik Reference Dzik1994). Thereafter a return to the original morphology is observed, with some gap in the record. An even more complex succession of immigration events was reported in Estonia by Männik & Viira (Reference Männik and Viira2012). A species of Complexodus is recorded there already in the E. suecicus and Y. foliaceus Zones, while C. pugionifer appeared after a gap in the E. lindstroemi Zone to continue its presence until the B. variabilis Zone, replaced by C. originalis in the middle of its range.

3.10.4. Complexodus pugionifer (Drygant Reference Drygant1974)

The species was originally identified in the Pishcha-16 borehole in Volhynia, Ukraine (Drygant Reference Drygant1974) and its complete apparatus has been restored by me (Dzik Reference Dzik2015), based on materials from the Mójcza Limestone.

3.11.1. Icriodella superba Rhodes Reference Rhodes1953

The Icriodella lineage had its root in Avalonia (Dzik Reference Dzik1990). The species co-occurs with A. superbus in the erratic boulder E-305 from Mochty, Poland (supplementary Table 67; Dzik Reference Dzik1983). Two kinds of M elements can be distinguished in a loose material (Fig. 6). A complete apparatus structure of a Late Ordovician close relative is known, owing to natural assemblages from the Soom Shale, South Africa (Aldridge et al. Reference Aldridge, Murdock, Gabbott and Theron2013).

3.12.1. Ansella jemtlandica (Löfgren Reference Löfgren1978)

The apparatus of this species was illustrated and described in detail by Stouge (Reference Stouge1984), Zhang (Reference Zhang1998b), and Zhen et al. (Reference Zhen, Cooper, Simes and Percival2011a, Reference Zhen, Wang, Zhang, Bergström, Percival and Chen2011b). Zhang (Reference Zhang1998b) determined its position in a Chinese lineage showing a gradual increase in the basal cavity depth. Fåhraeus & Hunter (Reference Fåhraeus and Hunter1985) identified an Ansella species with denticulation extended to the P elements, which is apparently a derived trait. This may mean that the ancestral apparatus for the lineage was undenticulated, with relatively shallow basal cavity. This places Ansella in proximity to Walliserodus, having a similar ground-plan of S₀ elements. It is preceded in the Table Head Formation of Newfoundland by A. sinuosa (Stouge Reference Stouge1984), characterised by a rather generalised appearance of M elements (Stouge Reference Stouge1984), and may be ancestral to it.

3.12.2. Ansella sp.

Hints et al. (Reference Hints, Viira and Nõlvak2012) determined Ansella species with obliquely oriented denticles as A. serrata (Dzik Reference Dzik1976) but ‘Belodella’ serrata Dzik Reference Dzik1976 has fine denticles inclined towards the element base, opposite to the Ansella pattern. Moreover, ‘B.’ serrata bears the panderodont furrow, which means that a separate panderodontid genus should be erected for it.

3.12.3. Hamarodus brevirameus (Walliser Reference Walliser1964)

Together with Dapsilodus and Scabbardella, all characterised by thin element walls (Fig. 7), this is a member of the cold- or deep-water Late Ordovician fauna (e.g., Stouge & Rasmussen Reference Stouge and Rasmussen1996; Ferretti & Barnes Reference Ferretti and Barnes1997; Ferretti Reference Ferretti1998) but is known also from the Appalachians (Nowlan et al. Reference Nowlan, McCracken and McLeod1997). In the Nanjing region of China this species seems to have a longer range than elsewhere (Chen & Zhang Reference Chen and Zhang1984). Only a single element possibly belonging to Hamarodus has been found in the Baltoniodus alobatus Zone in the borehole Niwa (sample 1093sp; supplementary Table 66), which would be one of the earliest known occurrences of this lineage.

3.13.1. Oistodus lanceolatus Pander Reference Pander1856

Element S₀ of this species (Fig. 6) is much smaller than other parts of the apparatus (e.g., Löfgren Reference Löfgren1994, fig. 6:39). The Laurentian species ‘Acodus’ bransoni (Ethington & Clark Reference Ethington and Clark1981) from the Paroistodus proteus to Prioniodus elegans Zones (Ji & Barnes Reference Ji and Barnes1990, Reference Ji and Barnes1994) may be ancestral to it.

3.13.2. Jumudontus gananda Cooper Reference Cooper1981

Nicoll (Reference Nicoll1992) restored its apparatus with an alate S₀ element suggesting affinity to Histiodella, but the geniculate M element he attributed to this species does not fit the rest of the apparatus.

3.13.3. Histiodella holodentata Ethington & Clark Reference Ethington and Clark1982

The composition of the apparatus of Histiodella and its early evolution was traced by McHargue (Reference McHargue1982) and Zhen et al. (Reference Zhen, Bauer and Bergström2022) based on the Oklahoma material. The lineage is probably of Laurentian ancestry and only sporadically immigrated to Baltica.

3.13.4. Histiodella kristinae Stouge Reference Stouge1984

The denticles of the ventral process are of the same height as the cusp (Fig. 7). It is a successor, of high correlative value, of H. tableheadensis Stouge Reference Stouge1984, characterised by the cusp in P₁ element being higher than the denticles of the ventral process (Stouge Reference Stouge1984; Zhang Reference Zhang1998b).

3.14.1. Erraticodon alternans (Hadding Reference Hadding1913)

Polygnathus alternans (Hadding Reference Hadding1913) from the E. robustus Zone at Fågelsång, Sweden is a senior synonym of E. balticus Dzik Reference Dzik1978 (Bergström Reference Bergström2007). This cosmopolitan warm-water species (Fig. 8) appeared in Baltica as early as the M. parva Zone (Löfgren Reference Löfgren1985) and continued to reappear sparsely until the E. lindstroemi Zone (Männik & Viira Reference Männik and Viira2012).

3.14.2. Spinodus spinatus (Hadding Reference Hadding1913)

This is a lineage of extremely long duration extending from the O. evae to A. superbus Zone but still poorly known because of rarity (Zhen et al. Reference Zhen, Cooper, Simes and Percival2011a; Stouge Reference Stouge2012). The apparatus of an early species of Spinodus from the Floian Lower Head Formation of Newfoundland was reconstructed by Johnston & Barnes (Reference Johnston and Barnes2000) who pointed out its similarity to Periodon. It differs from S. spinatus (Fig. 7; Albanesi Reference Albanesi1998) in closely spaced denticles of S elements. The M element bears two denticles on its ventral process, which makes it different from Erraticodon.

3.15.1. Periodon primus Stouge & Bagnoli Reference Stouge and Bagnoli1988

Periodon selenopsis Serpagli Reference Serpagli1974 was listed in Bagnoli & Stouge (Reference Bagnoli and Stouge1997) as co-occurring with P. primus in samples from north of Horns Udde but not commented on in the text. Presumably, these are undenticulated and denticulated specimens within population variability of a single biological species (Fig. 9).

3.15.2. Periodon flabellum (Lindström Reference Lindström1955a)

The P and S elements of Periodon are rather variable (Figs. 4aa–af, 9) and P. macrodentatus (Graves & Ellison Reference Graves and Ellison1941) as understood by Rasmussen (Reference Rasmussen2001; also Stouge et al. Reference Stouge, Harper, Sevastopulo, O'Mahony and Murray2016 and Zhen et al. Reference Zhen, Percival, Cooper, Simes and Wright2009a) seems to represent the transitional stage of evolutionary transformation to P. aculeatus in respect to the diagnostic incipient denticulation of M elements.

3.15.3. Periodon aculeatus Hadding Reference Hadding1913

Johnston & Barnes (Reference Johnston and Barnes2000, text-fig. 4) and Stouge (Reference Stouge2012) observed gradual development of denticulation of M elements in the Periodon lineage in the Floian Lower Cow Head Formation. Periodon species (Fig. 9) were typical for deeper-water environments than their relative Microzarkodina (Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008, p. 44).

3.15.4. Microzarkodina russica Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008

This is the oldest Baltic species of Microzarkodina from the latest Oepikodus evae Zone (Fig. 5). Its P elements have a few small denticles on the ventral process and the S₁ (Sb2 in Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008) elements have a long denticulated medial process. In these aspects it is somewhat closer to Periodon than the majority of Microzarkodina species. Somewhat surprisingly, a similar denticulation on the medial process developed late in the evolution of the lineage in M. hagetiana and, even more prominent, in its successor M. ozarkodella, but in S₂ elements. This means that a rather profound reorganisation of the S series took place.

Microzarkodina buggischi Lehnert Reference Lehnert1995 that occurs in the San Juan Formation in Argentina below M. flabellum, has the S₀ element with prominently denticulated medial process (Lehnert Reference Lehnert1995), which makes it similar to early Periodon (elements S₁ and S₂ remain unknown). However, in contrast to M. buggischi, lacking ventral denticulation, the P elements of M. russica may develop more than one ventral denticle. Albanesi et al. (Reference Albanesi, Carrera, Cañas and Saltzman2006) and Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2008) proposed to transfer M. buggischi to Texania and would therefore be related to T. heligma Pohler, 1994 that co-occurs with Oepikodus evae and above the first occurrence of Periodon primus and P. selenopsis in the San Juan Formation (Albanesi Reference Albanesi1998; Mango & Albanesi Reference Mango and Albanesi2020). Anyway, the apparatus structure of both M. russica and M. buggischi points to at least distant relationship between Microzarkodina and Periodon.

3.15.5. Microzarkodina flabellum (Lindström Reference Lindström1955a)

Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2008) identified a population transitional in morphology and age (early B. triangularis Zone) between M. russica and M. flabellum in Kinnekulle, Sweden. This is the next in order chronospecies of the Microzarkodina lineage that continued to have the S₁ elements bearing only a single denticle or lacking denticulation of the medial process (Figs 4t–z, 5). The P elements only rarely bear more than one ventral denticle. All the Microzarkodina species presented by Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2008, fig. 1) in their meticulous study have partially overlapping stratigraphic ranges. This means that they were defined in vertical (typologic) terms, the approach followed by most authors dealing with this genus. This unavoidably implies presence of pairs of artificially sympatric species at each evolutionary transition.

3.15.6. Microzarkodina parva Lindström Reference Lindström1971

The S₀ elements that developed rudimentary fourth process and adenticulate S₁ elements make this species different from the segments of the lineage preceding and following it (Fig. 5; Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008, fig. 14).

3.15.7. Microzarkodina hagetiana Stouge & Bagnoli Reference Stouge and Bagnoli1990

This species is characterised by the medial process in S₂ elements (Sb1 in Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008) with a few denticles, the S₁ elements remaining adenticulate, which points to M. ozarkodella as its phylogenetic successor (Fig. 5). The stratigraphic range of M. hagetiana strongly overlaps with M. bella Löfgren Reference Löfgren2000, restricted in its occurrence to Öland and Finngrundet in Sweden, where it is represented by small, possibly juvenile specimens (Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008, fig. 1). It is not clear to me whether these are series of sympatric populations of distinct biological species or results of differences in population dynamics (distribution of mortality). Conodont elements grew by adding denticles (Dzik & Trammer Reference Dzik and Trammer1980, fig. 3) and the appearance of elements classified as M. hagetiana and M. bella (to me difficult to discern and separate from population variability so nicely presented by Löfgren & Tolmacheva Reference Löfgren and Tolmacheva2008) may be due to their ontogenetic age at the moment of death.

According to Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2008, p. 36, fig. 9x–z) ‘some levels of the upper stratigraphical range [of M. hagetiana] yield elements with up to three’ ventral denticles. This suggests a gradual transition to M. ozarkodella, as this is the most characteristic diagnostic character of the species. A sample (Les-151śr) taken from the core of the borehole Lesieniec IG 1 somewhat below the first occurrence of M. ozarkodella (Les-149tył; supplementary Table 40), yielded P elements with only one ventral denticle. Co-occuring elements of Baltoniodus show a morphology transitional between B. navis and B. norrlandicus.

3.15.8. Microzarkodina ozarkodella Lindström 1971

Löfgren & Tolmacheva (Reference Löfgren and Tolmacheva2008, fig. 12) quantitatively documented the gradual increase of the number of ventral denticles in the P elements from numerical domination of specimens with two denticles to those with domination of three denticles and rare specimens with five denticles (Fig. 5). In the large sample from the erratic E-348 (supplementary Table 46) specimens of M. ozarkodella are not especially variable and bear three to four denticles, but in samples from the borehole Lesieniec IG 1 of comparable size (Les-145) from one to four.

Elements of this species occur sporadically in as young strata as the B. gerdae Zone, but this is probably due to reworking.

3.16.1. Yaoxianognathus rhodesi (Lindström Reference Lindström1970)

Leslie's (Reference Leslie2000) restoration of the apparatus of Ozarkodina? abrupta Branson & Mehl Reference Branson and Mehl1933 shows a close similarity to this species. He proposed to classify it in Yaoxianognathus and I adhere to such a taxonomic decision (Fig. 5) despite derived morphology of the type species of this genus. An almost complete set of elements co-occurring with those of Amorphognathus superbus was found in erratic boulder E-305 from Mochty, Poland (supplementary Table 67; Dzik Reference Dzik1983). In Estonia it occurs in the early Katian Amorphognathus ventilatus Zone (Oandu age) but also in the late Katian early A. ordovicicus Zone (Vormsi age) (Viira Reference Viira1974). In Podolia, it co-occurs with A. duftonus, which in England characterises the latest Katian (Orchard Reference Orchard1980).

Zhen et al. (Reference Zhen, Burrett, Percival and Lin2010) proposed that Tasmanognathus careyi Burrett Reference Burrett1979 from the latest Sandbian or earliest Katian of Tasmania is ancestral to the Laurentian Y. abruptus (Branson & Mehl Reference Branson and Mehl1933) as interpreted by Leslie (Reference Leslie2000). A problematic aspect of such interpretation is that, in the apparatus of Y. abruptus, the symmetrical element bears a strong medial process and very short lateral processes, unlike T. careyi that lacks any medial denticulated process in these elements, of a unique morphology as for ozarkodinid conodonts. Its morphology in the type species of Yaoxianognathus remains unknown.

In Y. rhodesi the S₀ element is of morphology typical for the post-Ordovician ozarkodinids (Orchard Reference Orchard1980) and it may be near their roots. If the non-M (that is, the non-holotype) elements of Gallinatodus elegantissimus Albanesi in Albanesi et al. (Reference Albanesi, Monaldi, Barnes, Zeballo and Ortega2023) from Argentina (Albanesi et al. Reference Albanesi, Monaldi, Barnes, Zeballo and Ortega2023) are joined with the M element attributed to Pyramidens cactus Albanesi in Albanesi et al. (Reference Albanesi, Monaldi, Barnes, Zeballo and Ortega2023) (unfortunately a prioniodontid P₁ element was chosen as its holotype), this may be an early Darriwilian species ancestral for the Yaoxianognathus lineage.

4.4.1 Bothnian Bay

The most detailed quantitative sampling of the Ordovician at the main route of the glacier along the Bothnian Bay is the Finngrundet borehole studied by Löfgren (Reference Löfgren1985). The core covers the late Floian, Dapingian and earliest Darriwilian conodont succession (Fig. 11). The late Floian is marked by the sudden alternation of Paltodus, Acodus, Paracordylodus and Oepikodus. Apparently, this is an expression of a global environment instability that resulted in short migration events (or rather changes in areas inhabited by these species). An intriguing aspect of the late Floian to early Dapingian transition is the gradual increase and then decrease in contribution of the Drepanoistodus forceps lineage and corresponding decrease and then recovery of the Paroistodus lineage. Perhaps this is a case of replacement by competing species, mentioned above in respect to the Putilovo mud mound (Tolmacheva et al. Reference Tolmacheva, Fedorov and Egerquist2003). Like all other Baltic sections, Baltoniodus gradually and slowly increased its contribution during the Dapingian until it became the dominant element of the assemblage. It was followed, in a more chaotic pattern, by Microzarkodina. A gradual increase from the latest Floian to a decrease in the Darriwilian is also shown by Protopanderodus rectus, which was replaced by Semiacontiodus cornuformis. The Finngrunded record of the conodont succession is especially instructive because it shows the actual changes in frequencies of conodont species almost not distorted by uneven sampling.

Figure 11 The Ordovician conodonts succession recorded in the Finngrundet borehole drilled in the Bothnian Bay, Sweden (data from Löfgren Reference Löfgren1985; supplementary Table 31).

4.4.2. Ingria

Almost equally detailed is the succession in Ingria, ranging from the latest Tremadocian to early Darriwilian (Fig 10; Tolmacheva et al. Reference Tolmacheva, Koren, Holmer, Popov and Raevskaya2001b, Reference Tolmacheva, Fedorov and Egerquist2003; supplementary Tables 4, 6, 11). The exclusive relationship between Paroistodus and Drepanoistodus forceps is recorded there in the Oepikodus evae Zone and near the Tremadocian–Floian boundary, but not later. The discrepancy between the record of these species’ succession at the Putilovo mud mound and the coeval Lava section is intriguing. In more general terms, the Ingria succession shows a relative stability of conodont faunas until the Oepikodus evae immigration. Actually, the appearance of this exotic elements did not cause any profound reorganisation of the conodont community.

4.4.3. Bornholm and Scania

Faunal influences from opposite directions in the late Dapingian and early Darriwilian are recorded in Bornholm and Scania (Fig. 12; supplementary Table 19; Stouge & Rasmussen Reference Stouge and Rasmussen1996). The source of Barrandegnathus, identified by Stouge (Reference Stouge2005) in the Komstad Limestone, is the Czech part of Gondwana, whereas Histiodella (Stouge & Nielsen Reference Stouge and Nielsen2003) is of Laurentian origin. Warm climate elements include Erraticodon, Spinodus and Cahabagnathus in the Almelund Shale at the classic locality Fågelsång in Scania (Hadding Reference Hadding1913; Lindström Reference Lindström1955b; Bergström Reference Bergström2007). Spinodus spinatus was a species preferring deeper-water environments. It is extremely rare in other Baltic localities. Complexodus is probably of the South Chinese Yangtse continent affinity. Such disparity in source areas for immigrants may have resulted from the location of Scania exposed on the margin of Baltica.

Figure 12 The Ordovician conodonts succession recorded at Fågelsång in Scania, Sweden (data from Stouge & Nielsen Reference Stouge and Nielsen2003 and Bergström Reference Bergström2007; supplementary Table 19).

4.4.4. Island Öland

A closer approximation to the beginning of colonisation by conodonts is offered by exposures in the Swedish island Öland (Fig. 9; Stouge & Bagnoli Reference Stouge and Bagnoli1990; Bagnoli & Stouge Reference Bagnoli and Stouge1997; Bagnoli et al. Reference Bagnoli, Stouge and Tongiorgi1988; supplementary Tables 2, 8, 10, 18, 20, 27). The quantitatively described succession starts there from the latest Paltodus deltifer Zone and ends with Eoplacognathus pseudoplanus Zone. Unlike for Ingria, the oldest conodont assemblage sampled at Ottenby is dominated by Paltodus, not Paroistodus. Drepanodus is second in importance; other species are rare, except for Acodus. Assemblages from younger strata are similar to those in eastern Baltica.

4.4.5. Västergötland

One would expect that the Ordovician conodonts from Swedish localities located westward of Scania, thus exposed to the Laurentian influences, would contain Midcontinent faunal elements. This is not necessarily the case and may be restricted to the initial stage in the formation of the Baltic conodont faunal realm. Probably the most complete record of the latest Tremadocian succession is that from Brattefors (Fig. 13; Löfgren Reference Löfgren1997a; supplementary Table 1). It starts from the Laurentian lineage of Variabiliconus (Löfgren et al. Reference Löfgren, Repetski and Ethington1999; supplementary Table 1). Darriwilian assemblages recorded in a few other localities are clearly Baltic (Löfgren Reference Löfgren1995, Reference Löfgren2003; supplementary Tables 32, 34, 35, 45, 49, 50). Paroistodus and Paltodus are dominant there, followed by Drepanodus. Some reworked material is also present.

Figure 13 The Ordovician conodonts succession recorded at the Gullhögen Quarry (data from Löfgren Reference Löfgren1995, Reference Löfgren2003; supplementary Tables 35, 45, 49, 50), Lanna and Orreholmen (Löfgren Reference Löfgren1995; supplementary Tables 32, 34), and Brattefors, (Löfgren Reference Löfgren1997a; supplementary Table 1) in Västergötland, Sweden.

4.4.6. Dalarna

A different pattern to that in Västergötland can be observed in central Sweden (Fig. 14; Löfgren Reference Löfgren1985, Reference Löfgren1994, Reference Löfgren1995; supplementary Tables 4, 28, 33). The Floian conodont community is virtually the same as at Brattefors (Fig. 13). However, as long as the limited data allow one to infer, the Darriwilian is characterised by a relatively low contribution from Baltoniodus (Löfgren Reference Löfgren1995). This places it relatively close to the Caledonides.

Figure 14 The Ordovician conodonts succession recorded at Leskusänget (data from Löfgren Reference Löfgren1995; supplementary Table 33), Sjurberg (Löfgren Reference Löfgren1985; supplementary Table 28), and Fjäcka (Löfgren Reference Löfgren1994; supplementary Table 4) in Dalarna, Sweden.

4.4.7. Jämtland

Northward of Dalarna, the faunal succession shows gradual increase in contribution from the Baltoniodus lineage typical for the Baltic region (Fig. 15; supplementary Tables 14, 22, 24–26, 30). It slowly became more and more important after its appearance as B. triangularis and, after a period of domination, retreats near the end of the local conodont record in the Eoplacognathus robustus Zone (Löfgren Reference Löfgren1978; Zhang & Sturkell Reference Zhang and Sturkell1998; Zhang Reference Zhang1999).

Figure 15 The Ordovician conodonts succession recorded at Gammalbodberget, Gusta, Gärde, Kloxåsen, Sommartjärnen and Lunne (data from Löfgren Reference Löfgren1978; Zhang & Sturkell Reference Zhang and Sturkell1998; supplementary Tables 14, 22, 24–26, 30) in Jämtland, Sweden.