Occurrence

Both Tolypelepis undulata and T. mielnikensis come from the youngest, upper Silurian (Pridoli) sediments of the Paleobaltic shelf (Fig. 2). Older or equal in age (Ludlow or Pridoli; Loeffler and Jones, Reference Loeffler and Jones1977) are rocks from Somerset Island, Arctic Canada, where Tolypelepis leopoldensis was collected. Other Silurian tolypelepidids, such as Asketaspis interstincta or Tolypelepis lenzi are known from Laurentian epicontinental sediments (Mackenzie Mountains), but their exact dating is uncertain (Dineley and Loeffler, Reference Dineley and Loeffler1976). Only some Tolypelepis sp. remains from the same area of Laurentia are well dated and referred to the early Silurian (Wenlock; Soehn and Wilson, 1991). The early Silurian (early Wenlock) agnathan fish, Athenaegis, currently considered as a possible sister taxon of the “higher Heterostraci” (Janvier, Reference Janvier1996), comes also from this part of Laurentia. Its suggested basal position, stratigraphic age and geographic province are suggestive of a Laurentian origin of the whole heterostracan clade, but this is very uncertain at this time.

Figure 2

Paleogeographical (Silurian map modified from Scotese, Reference Scotese2001) distribution of tolypelepidids in the stratigrapic context with location of described material.

Remarks

Two genera are included in Tolypelepididae: Tolypelepis Pander, Reference Pander1856 and Asketaspis Dineley and Loeffler, Reference Dineley and Loeffler1976. Given the range of taxonomic characters applied herein, Asketaspis should be included within the genus Tolypelepis, which would result in monotypy of Tolypelepididae and reduction of its diagnosis to that of the genus. However, Asketaspis, a monotypic genus with a A. interstincta represented by only a single specimen which has not been available for study, and thus is left as valid. The present ranking of the “tolypelepidids” is here considered tentative until new material will clarify the situation.

Genus Tolypelepis Pander, Reference Pander1856

Diagnosis

Dorsal shield broad and the preorbital length rather short (orbital ratio 0.14), subdivided into epitega. On the central epitegum and postrostral field the dentine ridges short and in most grouped into scale-like areas, with narrower, lower ridges arranged around a coarser, higher central ridge (after Denison, Reference Denison1964).

Tolypelepis mielnikensis new species

Figure 3.1, 3.2

1977 Tolypelepis undulata (Pander); Märss, pl. 1, fig. 3; fig. 2D.

Diagnosis

Rostral region constricted at orbital level with orbital notches deep. Dorsal shield of moderate dimensions (width ratio 0.62), with uniform width. Postbranchial lobes of dorsal shield not evident. Pineal macula distinct. Ornamentation of central epitegum of short dentine ridges of low relief. Ridges in anterior part of epitegum generally uniform in height, but difference between median and adjacent ridges increasing posteriad, the units becoming scale-like.

Description

The dorsal shield is complete, but the middle right side of the specimen is slightly collapsed. Nevertheless, the left part of the dorsal shield and rostral region are perfectly preserved except a small scratch of the left side of the specimen that does not affect the general shape of this part. Therefore, the natural shape of the shield can be restored. The shield length is 37 mm and width 23 mm. It is oval and of rather uniform width, and slightly arched dorsally. The anterior border is smoothly rounded and lacks the median rostral process. The width of the dorsal shield is reduced in front of the deep orbital notches (both sides of this part are symmetrical); the preorbital process is well developed. The lateral edges are almost straight except anteriorly and posteriorly where they are lightly curved. In lateral view, the edge is also straight. The branchial notch and post-branchial lobes are not developed. The posterior margin is zigzag and concave. The epitega are distinct; the post-rostral field can be distinguished from the central epitegum. Boundaries between epitega are clearly designated by sutures and by the pattern of dentine ridges.

The ornamentation of the dorsal shield comprises non-uniform dentine ridges of low relief, with scalloped edges. Ridges (5 per mm) on the rostral epitegum are uniform in height, broad, and arranged mostly transversely in the anterior part of the epitegum. Posteriorly on the epitegum, they are shorter and form small denticles. The most posterior ridges are shorter. They are orientated diagonally, some of them grouped into units. The ridges (5 per mm) on the lateral epitega are almost uniform in height. They are narrow and long, generally arranged parallel to the margin of the epitegum. In the middle of the lateral epitegum there are several singular ridges with higher relief and oriented diagonally. In the orbital region, ridges (4 per mm) are curved around the orbital notch. On the central epitegum the ridges are variable in height, and the average density is 4 per mm. On the anterior half of the shield the median ridge of each unit has the same height or is subtly higher than adjacent ones, the difference increasing posteriad. In the posterior half of the epitegum the ornamentation looks like imbricated, fused scales. The typical size of each unit is 1.8 mm long (±0.3 mm) and 1 mm (±0.2 mm) wide. Such scale-like units are absent from the postrostral field of the epitegum. Only one ridge group resembles such a unit. The ridges have uniform height and largely diagonal arrangement. The pineal macula is clearly visible on both specimens. Its central region is elevated and surrounded by encircling ridges.

The lateral line system is conspicuous over the entire specimen. It consists of pores in the superficial layer arranged in two pairs of longitudinal canals and four transverse commissures, perpendicular to the longitudinal ones (Fig. 4.1).

Etymology

Named after the type locality, Mielnik village.

Material

Dorsal shield, ZPAL Ag–II/1 (holotype); incomplete dorsal shield, ZPAL Ag–II/2.

Occurrence

Mielnik IG–1 drill core, depth 602.3–606.8 m, upper Silurian (Pridoli).

Remarks

The following character states of the dorsal shield are to be considered in a comparative analysis of the armor: the pattern of superficial ornamentation, dimensions and proportions of the dorsal shield, outline of the shield (uniform or variable width associated with lateral margin straight or slightly convex, posterior edge convex or concave), and pore canal system.

Eight tolypelepidid specimens of six species are considered here (Table 1). The length of the dorsal shield ranges from 27 to 38 mm. The Mielnik specimen matches this variability range, and so does Asketapis interstincta.

Table 1

Measurements and ratios of well–preserved specimens of the Tolypelepididae.

Regarding the overall proportions of the shield, expressed by width ratio, T. mielnikensis (width ratio 0.62), the longest among the tolypelepidids, is the slimmest of all, whereas the range for species of Tolypelepis is 0.67–0.77. If not ontogenetic (which is unclear as yet), this difference may be a taxonomically useful feature at a generic or specific level. Morphometric characters, other than the width ratio, do not significantly differ from those of T. undulata (Table 1).

More readily discriminated from other tolypelepidids is Tolypelepis leopoldensis. With the length (35 mm) comparable with the Mielnik specimen (37 mm), it is slightly broader (width ratio 0.77) than other Tolypelepis representatives (0.67–0.74), thus displaying a different combination of morphometric characters. It also differs in orbital width ratio (0.31), which is lower than in other tolypelepids (0.44–0.52), and in a very high pineal length ratio (0.57), which is dramatically different from that of all other tolypelepidids (0.22–0.28). In other words, the pineal macula is placed much more posteriorly in T. leopoldensis. These data show the amount of variability within the genus Tolypelepis, and, combined with other characters, these differences are here considered specific (see below).

Another alleged tolypelepidid, Asketaspis, seems not to differ in morphometric terms from Tolypelepis at the generic level (Table 1), but some other characters (e.g., presence of inter-epitegal tubercles) may suggest the opposite (see below).

Apart from overall proportions, the shape of the shield depends on the degree of tapering, which is expressed by having either straighter or more convex sides. Tolypelepis mielnikensis is uniform in width. In this regard it recalls both Asketaspis interstincta (Fig. 4.5; Dineley and Loeffler, Reference Dineley and Loeffler1976:fig. 22, pl. 6, fig. 1), and one specimen of Tolypelepis undulata figured by Märss (Reference Märss1977:fig. 2 D, pl. fig. 3), the lateral margins of these specimens being straight in dorsal view. All three specimens contrast with other representatives of Tolypelepis, where the lateral margin is convex (Rohon, Reference Rohon1893:pl. I, fig. 45; Kiaer, Reference Kiaer1932:l. X; Stensio, Reference Stensiö1958:fig. 169B; Loeffler and Jones, Reference Loeffler and Jones1977:fig. 3, pl. 80, figs. 3, 5). This unique character may suggest that all three specimens are congeneric (or conspecific if belonging to Tolypelepis), and belong to Asketaspis, but other characters (e.g., presence of interepitegal tubercles, do not support this view (see below)).

The Mielnik specimen is unique in that the rostral region is constricted anterior to the orbits. The specimen Pi 5003/GIT 168-4 figured by Märss (Reference Märss1977:fig. 2D) shows a similar rostral shape. In contrast, Asketaspis and other representatives of Tolypelepis (Fig. 4.3–4.7) lack such narrowing, the tapering of the anterior part of the shield being gradual and smooth beginning from the orbital region to the anterior tip of the rostral margin (Märss, Reference Märss1977:fig. 3, fig. 11 A–B; Denison and Loeffler, 1976:fig. 21–22, pl. 6, figs. 1, 3; Loeffler and Jones, Reference Loeffler and Jones1977:fig. 3, pl. 80, figs. 3, 5). Rostral region narrowing in front of orbits is known in Poraspis (Denison, Reference Denison1964; Loeffler and Jones, Reference Loeffler and Jones1977; Elliott and Petriello, 2011) so this feature concerns the generic rather than specific level, which precludes the inclusion of the Mielnik specimen in Asketaspis. Moreover, the Mielnik specimen differs from Asketaspis in the degree of delimitation of the orbital epitegum, which is not distinct in the former (similar to Tolypelepis), while being distinct in the latter (Dineley and Loeffler, Reference Dineley and Loeffler1976:fig. 22, pl. 6, fig. 1).

Complete articulated skeletons, of tolypelepidid-like agnathan fish, are only known for one species, Athenaegis chattertoni, recorded from the early Wenlock (early Silurian) of the Mackenzie Mountains, and originally considered to be a tolypelepidid (Soehn and Wilson, Reference Soehn and Wilson1990). More than 14 specimens were assigned to A. chattertoni. In terms of dorsal shield proportions and outline, Athenaegis differs from tolypelepidids at a generic level, but it is excluded from the Tolypelepididae on the basis of branchial structure (Janvier, Reference Janvier1996). If so, its strikingly tolypelepidid ornamentation (i.e., grouping of dentine ridges into scale-like units, each with narrow and low ridges arranged around a wider central ridge) is to be considered plesiomorphic. In taxonomic terms, the width of the ridges is important, defined as the number of ridges per millimeter (Denison, Reference Denison1964), their length, orientation, and variation in height and thickness. Tolypelepis shares with Asketaspis transversely aligned ridges, but differs in having diagonally oriented, short ridges, sometimes grouped into units (T. undulata, T. lenzi, or T. mielnikensis).

On the lateral epitegum of T. mielnikensis, ridges are arranged longitudinally and are parallel to the lateral margin (Fig. 3.1, 3.4), which is the pattern shared by all tolypelepidids and by Athenaegis as well, associated with the long and narrow shape of the lateral epitegum. The width of ridges in the lateral epitegum of Tolypelepis undulata is 8 per mm on average (Table 1). Tolypelepis mielnikensis has slightly coarser ridges (7 per mm). The information about Asketaspis (6.5 ridges per mm), given by Dineley and Loeffler (Reference Dineley and Loeffler1976) is not precise regarding the topography of the dorsal shield. There is no information about T. leopoldensis and T. lenzi.

Figure 3

Tolypelepis mielnikensis, n. sp. (1) holotype Ag–II/1, dorsal shield; (2) fragment of dorsal shield Ag–II/2; (3) drawing of left lateral view of holotype Ag–II/1; (4–5) drawing of the dentine ridge pattern of the dorsal shield.

The ornamentation of the central epitegum in T. mielnikensis consists of short dentine ridges of low relief and is similar to that of Asketaspis interstincta, and particularly so in the changing relations of the ridge heights. In the anterior part of the epitegum of both these taxa, the ridges are generally uniform in height, but the difference between the median ridge and adjacent ones increases posteriad. This changing pattern of ornamentation is shared also by the specimen Pi 5003/GIT 168-4 of Tolypelepis undulata (Märss, Reference Märss1977:pl. fig. 3), which stresses the similarity of these three specimens, as suggested by the shield outline. In other Tolypelepis species, the median ridge is higher than the adjacent ones on the whole central epitegum. However, a significant difference does exist between both the Mielnik specimen and Märss’s specimen (1977:fig. 2D) on one hand and Asketaspis on the other. The former lacks an inter-epitegal band of tubercles, characteristic of Asketaspis, and so a generic difference between those taxa is supported. However, the character may be subject to infrageneric variability as demonstrated by the presence of the tubercles in Tolypelepis leopoldensis.

Most posterior units of the Mielnik specimen look like imbricated scales (Fig. 3.1, 3.4) as they do in some other Tolypelepis species, T. lenzi (Dineley and Loeffler, Reference Dineley and Loeffler1976:fig. 22, pl. 6, fig. 1) and specimens of T. undulata described by Rohon (Reference Rohon1893:pl. I, fig. 45) and Kiaer (Reference Kiaer1932:pl. X). The main difference between the taxa is in the relative size of units, which in T. mielnikensis are proportionally smaller than those in other Tolypelepis species, but larger than those in Asketaspis. In this regard, the Mielnik specimen is morphologically intermediate between Tolypelepis and Asketaspis, but the mosaic pattern of similarities within the Tolypelepis group suggests undetermined variability at the generic or specific level.

Comparisons of other characters are inconclusive. The posterior edge of the shield is concave in the specimen from Mielnik (if this is not an artifact of preservation), while convex in Tolypelepis. In Asketaspis the posterior part of the dorsal margin is not preserved.

In Tolypelepis, the construction of the branchial region is unclear, as it is in Tolypelepis mielnikensis. However, a fragment of Tolypelepis branchial plate (Märss, Reference Märss1977:fig. 6B) reveals a branchial notch, and, posterior to it, an expanded dorsal lobe of the branchial plate. Asketaspis displays also a branchial notch and moderately well-developed postbranchial lobe (Dineley and Loeffler, Reference Dineley and Loeffler1976). This structure is probably more or less typical for the early cyathaspidiforms, tolypelepidids included, which thus recalls the morphology of the cyathaspidid genus Nahanniaspis (Dineley and Loeffler, Reference Dineley and Loeffler1976:figs. 33–34). Presumably, the overall branchial structure was similar (branchial notch mostly created by the incised branchial plate). The situation is different in Asketaspis which displays a branchial notch and postbranchial lobe, both moderately well developed. The dorsal shield of T. mielnikensis apears straight in lateral view (Fig. 3.3), as it does in Athenaegis, thus suggesting that the postbranchial lobe is absent. Whether these minor differences are properly understood and taxonomically valid for low level taxonomy cannot currently be estimated.

The sensory line canal system of the Mielnik specimen is typical for cyathaspidiforms, with two pairs of longitudinal canals and at least three transverse commissures on the dorsal shield (Fig. 4.1). The pineal branches of the supraorbital canals do not join each other, similar to the other Cyathaspidiformes, and in contrast to Pteraspidiformes where the canals join and make a V-shaped pineal canal.

Figure 4

Diversity of dorsal shield outlines in the Tolypelepididae. (1–2) Tolypelepis mielnikensis, n. sp.; (3–4) Tolypelepis undulata; (5) Asketaspis interstincta; (6) Tolypelepis lenzi; (7) Tolypelepis leopoldensis. (2) after Märss, Reference Märss1977; (3) after Rohon, Reference Rohon1893 and Märss, Reference Märss1977; (4) after Kiaer, Reference Kiaer1932 and Märss, Reference Märss1977; (5–6) after Dineley and Loeffler, Reference Dineley and Loeffler1976; (7) after Loeffler and Jones, Reference Loeffler and Jones1977. Abbreviations: cd1–cd4, dorsal transverse commissures; cep, central epitegum; ifc, pores of infraorbital canal; ldl, pores of lateral dorsal canal; lep, lateral epitegum; mdl, pores of medial dorsal canal; rep, rostral epitegum; soc, pores of supraorbital canal.