The Oldest Representative of the Rove Beetle Tribe Pinophilini (Coleoptera: Staphylinidae: Paederinae), from Upper Cretaceous Burmese Amber

The recently reviewed subtribe Procirrina comprises eight extant genera with a predominately (sub)tropical distribution. Previous phylogenies consistently recover the tribe Pinophilini of the subfamily Paederinae monophyletic. No fossils of the tribe have been described, although compression fossils are known from the Cenozoic Green River Formation (50.3–46.2 Ma) as well as inclusions from the Miocene Dominican (20.43–13.65 Ma) and Mexican (20–15 Ma) ambers. Here we describe †Cretoprocirrus trichotos Jenkins Shaw and Żyła gen. et sp. n., the oldest fossil representative of the tribe Pinophilini, from Upper Cretaceous Burmese amber (ca. 99 Ma). Phylogenetic analyses of morphological data allow its unambiguous placement in the subtribe Procirrina. †Cretoprocirrus trichotos is the second genus of Paederinae described from Burmese amber and provides an important insight into the evolution of the subfamily.


Introduction
The family Staphylinidae contains over 64,000 described species in 32 subfamilies, making it the most speciose animal family known [1]. Among all subfamilies of rove beetles, the Paederinae are one of the more diverse, with 7584 described extant species and 34 fossil species [1][2][3]. Recent molecular [2], total-evidence [3] and morphological [4] phylogenetic analyses of Paederinae have contributed a wealth of data, helped to improve the classification [2] and provided the first morphological matrices for the entire subfamily [3,4]. InŻyła et al. [3], the authors also described the first representatives of the subfamily Paederinae from Burmese amber and summarized the fossil record of the subfamily. Before that, some molecular data for Paederinae was produced as a side result of larger studies focused on Coleoptera in general or Staphyliniformia and related families. Zhang and Zhou [5] included four genera of Paederinae and McKenna et al. [6] contained 11 paederine genera, two of them from the tribe Pinophilini. The latter study, however, was based on only two gene fragments [6].
The tribe Pinophilini is one of four tribes that constitute the subfamily Paederinae [2]. Pinophilini is further divided into two subtribes, namely Pinophilina and Procirrina. Pinophilini has been recovered as monophyletic with strong support based on molecular [2,6], total-evidence [3] and

Specimen, Deposition, Photography and Measurements
The beetle specimen is embedded in a medium size oval piece (32 × 26 × 5 mm) of Burmese amber from the Hukawng Valley, Kachin State, northern Myanmar and has the specimen number NIGP172214. The age of the amber is considered as Late Cretaceous (earliest Cenomanian, ca. 99 Ma) based on UePb zircon dating [14]. The specimen is quite well preserved, but observations are only possible in dorsal and ventral views. The head of the beetle is directed slightly downwards, meaning ventral head characters are not visible. The right side of the specimen (particularly from the elytra to the anterior portion of the abdomen) is partly obscured by organic matter. The holotype (NIGP172214) is deposited in the Nanjing Institute of Geology and Palaeontology (NIGP), Chinese Academy of Sciences (CAS). The specimen was studied using an Olympus SZ61 (Tokyo, Japan) stereomicroscope. Photographs were taken using a Canon EOS 5D MkII (Tokyo, Japan) with a Canon MP-E 65 mm macro lens (Tokyo, Japan) or a Zeiss AX10 microscope and Zeiss Aziocam 105 color. Photographs were stacked using Helicon Focus. X-ray micro-computed tomography (micro-CT) was attempted using an Xradia MicroXCT-400, but reconstruction was unsuccessful, probably because of lack of density contrast.
The following measurements are given (all in millimeters): head width (HW), measured from the outer margin of one eye across to the outer margin of the other; head length (HL), measured from the anterior margin of the frontoclypeus to the anterior margin of the nuchal constriction; neck width (NW), measured across the narrowest part of the nuchal constriction; pronotal length (PL), measured mid-longitudinally from the anterior to the posterior margin; pronotal width (PW), measured across the widest place, usually at about the anterior third; elytral length (EL), measured from the posterior edge of the scutellum to a line across the posterior most portion of the posterior margin of the elytra; elytral width (EW), measured across the widest place. Measurements were taken using ImageJ [15].

Phylogenetic Analyses
Initial examination of the fossil suggested that it was likely to be either a member of, or closely related to, the subtribe Procirrina (tribe Pinophilini) of the subfamily Paederinae. To ascertain the systematic position of the fossil, the matrix of Herman [9] was compiled in Mesquite [16] and modified according to our taxon sampling. This matrix was published as an erratum in Herman [9] but refers to the study by Herman [7] and is available at http://hdl.handle.net/2246/6421. We found it necessary to re-word some characters and their states without changing their meaning.
We added 23 more characters, some used before and some used here for the subfamily for the first time. A list of the characters studied and their respective states is provided below. To make the phylogenetic analysis more representative of the entire subfamily, seven additional taxa representing a broad spectrum of Paederinae and two additional outgroups, Pseudopsis Newman, 1834 (Pseudopsinae) and Quedius molochinus (Gravenhorst, 1806) (Staphylininae), were added. Unknown character states were coded with "?" and inapplicable states with "-". The final matrix consisted of 57 characters scored for 21 taxa and is available as the nexus file in Supplementary Materials and in MorphoBank (project no. 3619) under this permalink http://morphobank.org/permalink/?P3619.
Bayesian inference and maximum parsimony methods were used for phylogenetic analyses to test for topological congruence.
For the Bayesian inference, MrBayes 3.2.6 [17] was used running on CIPRES Science Gateway v3.3. (phylo.org). The data were analyzed using the Mkv model [18] and default settings for priors. The analyses used four chains (one cold and three heated) and two runs of 10 million generations; they were conducted using a gamma distribution. The convergence of the two runs was visualized in Tracer v1.6 [19] and by examining the potential scale reduction factor (PSRF) values and the average standard deviation of split frequencies in the MrBayes output. Topological convergence was confirmed using the R package, RWTY (R We There Yet) [20].
The maximum parsimony (MP) analyses were conducted in TNT 1.5 [21] using the "traditional search" option to find the most parsimonious trees (MPTs) under the follow parameters: memory set to hold 999,999 trees; tree bisection-reconnection (TBR) branch-swapping algorithm with 1000 replications saving 10 trees per replicate; and zero-length branches collapsed after the search. All character states were treated as unordered and equally weighted. Node support was assessed with 1000 bootstrap (BS) pseudoreplicates.

Morphological Characters
A single asterisk (*) indicates characters from Herman [7], two asterisk (**) characters were included inŻyła et al. [3] and Bogri et al. [4] and three asterisk (***) characters are used for the subfamily for the first time.

Phylogenetic Analyses
The Bayesian inference (BI) analysis reached convergence with an average standard deviation of split frequencies well below 0.01 after 2 million generations. Most PSRF values were 1.000 (maximum 1.002). All effective sample size (ESS) values in Tracer [19] were above 200. The cumulative split frequency of all clades using RWTY [20] indicated that all chains reached convergence. We also visualized the average standard deviation of split frequencies (ASDSF) across chains as a function of chain length, which confirmed that the chains converged in their estimate of topology and support values. The maximum parsimony (MP) analysis under equal weights resulted in 68 most parsimonious trees (MPTs) with 195 steps, consistency index (CI) = 0.44 and retention index (RI) = 0.68. Overall, both BI and MP produced trees of similar topology. The placement of the fossil was the same in both. The supported topological congruence between methods is summarized on the BI tree ( Figure 1). The MP strict consensus tree is provided in Supplementary Materials Figure S1. The
Material. Holotype, female, specimen number NIGP172214 (Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences). Most of the specimen is covered in a thin layer of air bubbles. In the dorsal view, there is a single pseudoscorpion (Arachnida: Pseudoscorpiones) to the right of the beetle.
Diagnosis. As for the genus.
Material. Holotype, female, specimen number NIGP172214 (Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences). Most of the specimen is covered in a thin layer of air bubbles. In the dorsal view, there is a single pseudoscorpion (Arachnida: Pseudoscorpiones) to the right of the beetle.
Pronotum elongate, distinctly longer than wide, widest at anterior third behind front angles ( Figure 3A); sides with dense erect setae and each with single long setae, sides meeting straight base of pronotum at obtuse angle. Dorsal surface of pronotum covered in small setae. Basisternum with sharp longitudinal ridge; prosternal transverse carina acutely pointed medially.
Mesoscutellum and base of elytra deeply depressed. Elytra elongate (Figures 2A and 3B); weakly tapering posteriad; laterally with dense setae; humeral angles distinct. Elytra with random punctation and area around elytral suture somewhat raised. Posterior margin of conjoined elytra emarginate, with row of setae ( Figure 3C). Apicolateral setae of elytra absent. Metaventrite with several (probably four) erect macro setae in addition to usual smaller setae.
Abdomen densely covered in erect setae (some setae almost as long as length of single tergite) (Figure 2A,B); widest at segment VII. Paratergites absent, abdomen cylindrical (Figure 2A,B). Sternite III with intercoxal carina. Tergite IX with putative incomplete left lateroapical process.

Discussion
As with many diverse organism groups, rove beetles (Coleoptera: Staphylinidae) have enormous potential for addressing questions in evolutionary biology; yet their sheer diversity acts as an impediment to their taxonomic inventory and phylogenetic comprehension. The subtribe Procirrina is just another group waiting to be explored and better understood. A quick glance at the current knowledge of the group points towards some interesting evolutionary patterns. Their predominant restriction to tropical environments, presence in arid Australia [7] and Madagascan caves [8] and the existence of two genera (Oedichirus and Palaminus) that exhibit disjunct distributions between the Old and New World are all phenomena requiring inquiry. Ideally, exploration of these should be done through fossil-calibrated dating of a molecular phylogeny or total-evidence dating. However, species level revisions and molecular data are far too incomplete to permit such research on Procirrina at present. Despite this, the description of extinct stem lineages is critical for their inclusion in future works, especially those that would address the phylogeny of Procirrina or Paederinae in a temporal context.
Burmese amber has proven to be a rich and diverse source of arthropod inclusions. As of 2018, 1192 species have been described from the deposit [23]. Among all the organisms that occur in Burmese amber, insects are the most diverse and abundant group [23]. It is unsurprising that numerous rove beetles (Staphylinidae) have been described from this deposit; with over 40 species reported so far, the number is increasing almost weekly.
Until now, within the subfamily Paederinae, just a single genus, DiminudonŻyła and Jenkins Shaw, 2019 with two described species, is known from Burmese amber. The genus was placed as subtribe incertae sedis within the tribe Lathrobiini based on a total-evidence phylogenetic analysis [3]. In the present paper we do not just describe the second genus of Paederinae from Upper Cretaceous Burmese amber, but the first fossil representative of Pinophilini and at the same time the oldest record of that group. The finding has important implications for our understanding of paederine evolution by showing that Procirrina (and therefore Pinophilini) rove beetles existed at least as far back as the Late Cretaceous.
The new species is placed in Paederinae based on the following characters, which are considered diagnostic of the subfamily [3,24]: antennal insertions concealed under "shelf" and therefore not visible from above; and hypomeron of prothorax with well-developed postcoxal process. Placement in the subtribe Procirrina is supported by the following diagnostic characters: securiform maxillary palpomere 4; protarsomeres 1-4 inflated; and abdominal segments IV to VII without paratergites [2,7]. Additional characters that support the placement in Procirrina include the long procoxae (about as long as profemora) and posterior margin of the combined elytra emarginate [7]. Cretoprocirrus is unusual among procirrine genera in having the metatibia of almost equal width along its entire length, not apically expanded, another character mentioned by Herman [7] that helps to define the subtribe. Cretoprocirrus shares with Oedichirus the presence of a spiniform pencil of setae on antennomere 11. Herman [7] suggested that it may be used to wick a secretion from the surface, although there is so far no evidence for this. Our phylogenetic analysis confidently places the fossil species within the subtribe Procirrina; however, its exact position remains unresolved. With the addition of our new genus, Procirrina now contains nine genera (Procirrus, Neoprocirrus, Oedichirus, Oedodactylus, Palaminus, Paraprocirrus, Pseudoprocirrus, Stylokyrtus and Cretoprocirrus; Figure 3G,H).
Burmese amber is thought to be the product of a coniferous tree belonging to the Cupressaceae or Araucariaceae, which existed in a moist tropical environment [25]. The precise paleoenvironmental conditions of the Burmese amber forest have not yet been fully established, but palynological study is indicative of a humid, warm, temperate climate [26]. Further evidence of Burmese amber formation in a tropical environment is provided by the presence of distinctly tropical taxa [25]. As mentioned by Herman [7], Procirrina are typically diverse in the tropical and subtropical regions of the world. Therefore, the presence of an extinct lineage of Procirrina from the Upper Cretaceous Burmese amber is not surprising.
With the description of Cretoprocirrus trichotos here, it is apparent that Paederinae rove beetles had diversified (or at least had begun to), into their major groups by the late Cretaceous. In addition to the taxon described here and Diminudon [3], we are aware of other, potentially distantly related paederines, in Burmese amber, which await description and phylogenetic placement.

Conclusions
Cretoprocirrus trichotos Jenkins Shaw andŻyła gen. et sp. n. is the first formally described fossil species belonging to the rove beetle tribe Pinophilini of the subfamily Paederinae. Based on phylogenetic analyses of a morphological dataset, the new genus is confidently placed in the subtribe Procirrina. Its discovery in the Upper Cretaceous amber along with the earlier mentioned Cenozoic and Miocene fossils [7,12,13] indicate a relatively long evolutionary history of this group dating back to the Mesozoic, with a high likelihood of more fossils waiting to be discovered.