The new fossil genus and species Balticeler kerneggeri appears to be a member of a paraphyletic collection of lineages sometimes called “middle-grade” Carabidae. The evidence for this comes from the presence of synapomorphies in Balticeler that group it with so-called higher carabids, i.e., the subfamily Harpalinae, combined with the lack of a derived characteristic that would place it within crown-Harpalinae. Four character states present in Balticeler that are derived within Carabidae are:

1. (i) antennal cleaner anisochaetous Grade B (Hlavac 1971);
2. (ii) mesocoxae conjunct (Bell 1967; Beutel 1992);
3. (iii) procoxal cavities externally closed (Bell 1967; Beutel 1992);
4. (iv) external lobe of metepimeron well developed (Beutel 1992).

These derived states indicate that Balticeler is more closely related to Harpalinae than are those lineages that are sometimes called “basal-grade” carabids, i.e., those with plesiomorphic states for these four characters.

The character state plesiomorphic within Carabidae that indicates Balticeler is not a member of crown Harpalinae is:

1. (v) seta in mandibular scrobe present (Maddison et al. 1999).

Within the middle-grade carabids, our evidence suggests that Balticeler is a member of the subfamily Trechinae sensu Maddison et al. (2019) (= Trechinae + Patrobinae in sense of previous authors, but without Lissopogonini). We propose the following as derived character states that help place Balticeler within Trechinae:

1. (vi) dorsal surface of aedeagal median lobe open, unsclerotized [a proposed autapomorphy of Trechinae]
2. (vii) clypeus quadrisetose [a proposed synapomorphy of Bembidarenini + Trechini]
3. (viii) antennomere 2 pubescent [a proposed autapomorphy of Trechitae]
4. (ix) pedicellus about as long as scapus [a proposed autapomorphy of Trechitae]
5. (x) umbilicate series of setae arranged in a 4+2+2 pattern [a proposed autapomorphy of Trechitae]

These characters are discussed in turn, below.

One of the primary characteristics that supports the placement of Balticeler in Trechinae is the surface of the aedeagal median lobe, which is not sclerotized dorsally and thus completely open between the basal and apical ostia and, thus, the lobes of median lobe basal bulb are free (Fig. 23). Within carabid groups characterized by character states i-v listed above, a completely open dorsal surface of the aedeagal median lobe is only present in Patrobini, Bembidarenini and two basal clades of Trechini (in the paraphyletic Trechodina). Based on a multigene analysis of Trechinae, Patrobini was identified as the sister group to the supertribe Trechitae, which includes all other Trechinae (Maddison et al. 2019). Within Trechitae, Bembidarenini was identified as the sister group of Trechini. The phylogenetic tree inferred from molecular data suggests that a dorsally open aedeagal median lobe is an autapomorphic character of Trechinae, and that it was secondarily closed at least twice in the evolution of Trechitae: once in the ancestor of the Trechina stem group, and a second time in the ancestor of the Clade B1 stem group of Maddison et al. (2019) which includes Anillini, Bembidiini, Pogonini, Tachyini, Sinozolini, and Zolini. If so, Balticeler might be considered a crown or stem group member of the subfamily Trechinae in the sense of Maddison et al. (2019).

This view is further supported by an additional character state observed in the fossil specimens which makes it more likely that Balticeler represents a member of Trechitae: the clypeus is quadrisetose, with two primary setae on each side. This character state is also developed in Bembidarenini and Trechini (absent only in Omalodera Blanchard; Arnaud Faille, pers. comm. 2021), but absent in all other Trechinae. If a quadrisetose clypeus represents a synapomorphic character state of Bembidarenini and Trechini, as seems probable, the likewise quadrisetose Balticeler could therefore be considered a member of the Bembidarenini + Trechini clade. We present as a working hypothesis that the quadrisetose clypeus represents a derived character state within both Trechinae and Trechitae, and that it evolved in the stem lineage of a clade comprising Balticeler + Bembidarenini + Trechini.

A quadrisetose clypeus is likewise developed in two other groups of middle-grade Carabidae, the genus Psydrus LeConte (Psydrini) and Gehringiini, neither of which are closely related to Trechitae (Maddison et al. 2019). Of the three genera within Psydrini, only Psydrus has a quadrisetose clypeus (Nomius Castelnau and Laccocenus Sloane have a bisetose clypeus); the quadrisetose clypeus is presumably derived within Psydrini. If a quadrisetose clypeus was an indication of a relationship with psydrines, it would suggest a sister group relationship to Psydrus itself. A quadrisetose clypeus is present in both Gehringia Darlington (based upon Darlington 1933; Deuve 2005, and the specimens we have examined) and Afrogehringia Baehr, Schüle, and Lorenz (based upon the two specimens we have examined). The third genus of gehringiine, Helenaea Schatzmayr and Koch (1934) appears to have only two setae on the anterior margin of the clypeus based upon available descriptions (e.g., Deuve 2005), but given the sister group relationship between Afrogehringia and Helenaea (Baehr et al. 2009), and the modified clypeus of Helenaea, this is likely a secondary loss.

Psydrines and gehringiines, however, differ from Balticeler in a number of aspects that speak against the potential relationship suggested by the setation of the clypeus. Both groups differ from Balticeler in having a very small or absent outer lobe of the metepimeron and by a dorsally closed aedeagal median lobe base. In addition, if Balticeler were the sister group of Psydrus, one would predict that it would have eight setae along the apical margin of the labrum (a synapomorphy of Psydrus and Nomius). Psydrini also lack an elytral plica, as well as having a unique position of the posterior spur of the antennal cleaner, which is distinctly distad of the clip setae, characteristics one would expect Balticeler to share if it were the sister to Psydrus. In addition to the metepimeral shape and aedeagal structure, gehringiines differ from Balticeler by the underlapping of elytral edges, apically truncate elytra, and by the metacoxa which laterally extends to elytral epipleuron, although these are all likely derived characters of gehringiines and don’t exclude the possibility that Balticeler is the sister group to the tribe.

The proposed phylogenetic position of Balticeler within Trechitae and the Bembidarenini + Trechini clade is difficult to verify due to a striking autapomorphy of the fossil taxon: the basal protarsomeres of Balticeler are not sexually dimorphic. Because of the absence of any modifications on the male protarsomeres, they lack the tooth-like prolongation on the outer apical margin. Uniquely dentate basal protarsomeres of males is currently the only known morphological synapomorphy of adults of the supertribe (Maddison et al. 2019). However, a female-like shape of male protarsomeres is also present in at least some Trechina (e.g., Aphaenops Bonvouloir, Jeannel 1928). The phylogenetic position of Aphaenops deeply nested within subtribe Trechina is strongly supported by additional morphological and molecular data (Jeannel 1928; Faille et al. 2010), and in that group the lack of sexual dimorphism and subsequent loss of dentate male protarsomeres is surely a derived state within Trechitae. We propose that the female-like protarsomeres in male Balticeler represent a similar secondary loss.

Balticeler shares with Trechitae the derived state of a pubescent antennal base. Pubescence begins from the second (Balticeler, Bembidiini, Tachyini, Zolini, the genus Chaltenia Roig-Juñent and Cicchino within Sinozolini) or first antennomere (all other Trechitae except Pogonini). Pubescence begins from the third or fourth antennomere in Pogonini, Patrobini and in all other members of the middle-grade carabids examined (including 15 genera of Moriomorphini, Lissopogonus Andrewes, and 11 genera of Broscini), except for those taxa that are either generally pubescent (e.g., Cymbionotum Baudi, Apotomus Illiger, brachinines) or have additional setae in many places of the body (all three genera of Psydrini). The smooth antennal base in Pogonini could represent a reversal since this group clusters within Clade B1 Trechitae of Maddison et al. (2019).

The pedicellus is markedly slender and about as long as the scapus in Balticeler and all other Trechitae except Bembidiini, Tachyini, Zolini, some Pogonini, and very few Trechini (e.g., Aphaenops queffeleci Cabidoche, some Pachydesus Motschulsky, Sporades Fauvel, Trechisibius Motschulsky, Trechosia Jeannel, Arnaud Faille, pers. comm. 2021). We view this as a derived character state within middle-grade carabids. Outside of Trechitae, all of the middle-grade carabids we have examined have a pedicellus no longer than 0.7 of the length of the scapus, which is similar in length to those trechites in Clade B1 of Maddison et al. (2019) that have a shorter pedicellus. We propose that the shorter pedicellus in some but not all members of Clade B1 is due to reversal.

Reduction of the number of setae in the umbilicate series and consequent separation into three or two groups was repeatedly evolved in many different groups of Carabidae. The 4+2+2 pattern, however, is observed only in Trechitae, and we consider it an autopomorphy of the group. Beside the fossil Balticeler, this pattern occurs in Bembidarenini, Trechini, Bembidiini, Tachyini, some Sinozolini and Anillini. In Lovriciina, the number of umbilicate setae is reduced further. One or more additional setae in the umbilicate series are observed in other species of Trechitae with larger adults. Pogonini possess an almost-continuous umbilicate series similarly to that found in the Trechitae sister group Patrobini. It is therefore likely that the 4+2+2 pattern evolved in the Trechitae stem group and was subsequently modified in some terminal groups.

Within Trechitae, we propose that Balticeler is a member of the stem group of Bembidarenini + Trechini, and thus the sister group of the two recent lineages of this clade. Evidence for it being a member of the stem group rather than the crown group comes from lack of an apomorphy that Balticeler should possess if it were within crown Bembidarenini + Trechini. All four genera of Bembidarenini, as well as all genera of Trechini we have examined (including members of all Trechodina subgroups, including the Trechobembix assemblage sensu Maddison et al. 2019, and various Trechina) possess a single long suborbital seta in approximately the same position on each side. Some other groups of Trechinae possess a scattering of short setae suborbitally (including, for example, some Bembidion Latreille, some Sinechostictus Motschulsky, Orzolina Machado, Phrypeus Casey, some Anillini), but lack a bilaterally symmetrical pair of long setae in that region. Similar scattered short setae in the suborbital region are also present in several non-Trechitae tribes, including the three genera of Psydrini and brachinines. The only other species of Trechinae we have seen with a single long suborbital seta are within two groups: many species of the subtribe Tachyina (including species of Meotachys Erwin, Nothoderis Boyd and Erwin, Paratachys Casey, Pericompsus LeConte, Tachyta Kirby, and Tachyura Motschulsky), and a few members of Pogonini [e.g., Sirdenus grayii (Wollaston)]. We have confirmed the lack of a long suborbital seta in Zolini (based on an examination of eight genera, including Pterocyrtus Sloane), Sinozolini (all three genera), four genera of Pogonini, Anillini (six genera examined, including Nesamblyops Jeannel), Xystosomina (five genera examined), all eight genera of Bembidiini, and Patrobini (15 genera examined, including Platidiolus Chaudoir). Outside of Trechinae a single long suborbital seta is lacking in the 15 genera of Moriomorphini examined, all three genera of Psydrini, as well as Gehringia, Lissopogonus, Broscini, Melaenini, and in all other tribes outside of Harpalinae we have examined (with Nototylini, Mantichorini, and Protopaussini being the only tribes outside of Harpalinae we have not examined). The only lineage of carabids outside of Trechinae that possesses a single long suborbital seta at a fixed position that we know of is the subtribe Pericalina, a lebiine group deeply nested within the large subfamily Harpalinae. Thus, it appears that the suborbital seta is derived within Trechitae. Although it has arisen more than once in Trechitae, the lack of a suborbital seta in Balticeler provides evidence against it being a member of crown Bembidarenini + Trechini. In addition, Balticeler lacks two apomorphies that place it outside of the phylogenetically diverse tribe Trechini: (1) Balticeler lacks Trechini-like supraorbital furrows, and (2) it lacks a recurrent stria on the elytron.

In addition, Balticeler can be reasonably excluded from the crown groups of other described tribes of Trechitae. From crown Anillini, Bembidiini, or Tachyini, it can be excluded because of the lack of subulate apical palpomeres. There is evidence that it is not a member of crown Sinozolini, as it lacks a bifid tooth on the mentum. Similarly, it appears to not be a member of crown Pogonini, as it lacks the larger number of umbilical setae present in that group.

There are two alternative placements of Balticeler which should be considered, based upon some similarities it shares with two other groups of Carabidae: within Promecognathini or its stem group, and within Scaritinae or its stem group.

Balticeler shares two distinctive apomorphies with Promecognathus Chaudoir: (i) long, thin, straight mandibles; (ii) lack of external sexual dimorphism, with unmodified male protarsomeres and with the last visible abdominal sternite with two pairs of setae (rather than one pair typical of males). Similarly long and thin mandibles are present in some other carabid groups (e.g., Lovriciina and some trechines), and thus this character state is homoplastic. Males having the last visible abdominal sternites with two pairs of setae, as is typical in females, also occur in the intertidal Bembidion subgenus Desarmatocillenus Netolitzky (Sasakawa 2007). However, Promecognathus has plesiomorphic states in two thoracic characters that indicate it is not part of the middle-grade of carabids, in contrast to Balticeler: Promecognathus (i) has disjunct mesocoxae, and (ii) lacks an externally visible lobed metepimeron. In addition, Promecognathus lacks all of the apomorphies mentioned above that Balticeler shares with Trechitae.

As with Promecognathus, Balticeler shares with most scaritines unmodified male protarsomeres, and with the scaritine tribe Dyschiriini, as well as many species of Clivinini and Scaritini, the last visible abdominal sternite with two pairs of setae in males. Long and thin mandibles are present in some lineages of Clivinini, e.g., Camptidius Putzeys, Camptodontus Dejean, Climax Putzeys, and Stratiotes Putzeys. In external shape, with its subcylindrical pronotum and hind body, Balticeler is particularly similar to species of Dyschiriini. In addition, several species of Dyschirius Bonelli (incl. Dyschiriodes Jeannel) share the reduced pronotal marginal border and elytral basal border with Balticeler (but not the long and thin mandibles). However, scaritines are characterized by at least two notable synapomorphies which Balticeler lacks: (i) the protibia is markedly modified as an adaptation to the fossorial way of life, with large tooth-like expansions laterally and apically; (ii) the insertion of the antenna, if viewed from above, is overlapped by a lateral extension of the frontal plate. Because both these character states are lacking in Balticeler the latter is unlikely to belong in crown scaritines. In addition, as mentioned for Promecognathus, scaritines have plesiomorphic states in two thoracic characters that indicate they are not part of the middle-grade of carabids: (i) disjunct mesocoxae, and (ii) lack of an externally visible lobed metepimeron.

The placement of Balticeler within Trechitae as a member of the stem lineage of Bembidarenini + Trechini is only moderately well supported. All of the character states that suggest this placement are homoplastic, with the derived state having evolved several times within Carabidae, or lost within some group of trechites. Placement of Balticeler as a member of stem Bembidarenini + Trechini and the evidence against Balticeler being within any other living tribe either within or outside of Trechitae would require, should we feel compelled to place Balticeler within a tribe, the creation of a new monobasic tribe to house it. However, for the moment we prefer to leave Balticeler incertae sedis within the classification of Trechitae, outside of any existing tribe. Creating a new tribe for Balticeler now would set a precedent that could lead to the creation of many tribal-level taxa should additional lineages of various stem groups of Trechitae be discovered in amber deposits, especially Burmese amber. We prefer a more cautious approach, in which new family group names are created when the evidence for their distinctiveness is sufficiently strong to make them moderately stable as new stem lineages are discovered. Once the amber fauna is better known, or the morphological studies of Trechinae provide stronger evidence for the placement of Balticeler, a more stable and complete tribal classification can be devised.