Blaisdell (1906) described Eschatoporis nunenmacheri, a new genus and species of blind tenebrionid collected from under a rock next to a spring. He compared this species to Eulabis Eschscholtz, 1829 and Cerenopus LeConte, 1851, at that time placed in the Scaurini. Blaisdell suggested that the tribe be expanded to include Eschatoporis or that it be placed in a new tribe which he named Eschatoporini. Lacordaire (1859) placed Eulabis, Centrioptera Mannerheim, 1843, and Cryptoglossa Solier, 1836 in his tribe Scaurides Billberg, 1820. Eulabis was placed in “Groupe III: Nyctoporides” (Lacordaire 1859: 131) while Centrioptera and Cryptoglossa were placed in “Groupe IV: Centrioptérides” along with Cerenopus (Lacordaire 1859: 135). This is perhaps why Gebien in both his catalogs (1910, 1937) placed Eschatoporis in the Cryptoglossini, which now includes the genera Cryptoglossa (= Centrioptera) and Asbolus LeConte, 1851 (= Cryptoglossa) see Aalbu (1985, 2005). Despite LeConte’s removal of Cerenopus and Eulabis from the Cryptoglossini (LeConte 1862), subsequent catalogs listed Eschatoporis in the Cryptoglossini.

Aalbu (1985: 50) moved Eschatoporis from the Cryptoglossini (subfamily Pimeliinae) to the subfamily Lagriinae, based on morphological data (see discussion below), but placed it as incertae sedis within the lagriine tribal classification due to the lack of specimens to dissect at that time. Doyen (1994: 445–446) later tentatively placed Eschatoporis in the tribe Goniaderini (Lagriinae). This placement was accepted by Aalbu et al. (2002). Later, Aalbu (2005) placed Eschatoporis in the Laenini based on the lack of defensive gland reservoirs and the presence of multiple non-glandular spermathecal tubules. The placement of Eschatoporis in the Goniaderini by Aalbu and Smith (2014), as pointed out by Kanda (2016), was an accidental error. At that time, Eschatoporis should have remained in the Laenini as pointed out Matthews et al. (2010: 577). These errors, as well as the shuffling of Eschatoporis between various tribes, were recently summarized (Kanda 2016), and helped emphasize that a reevaluation of the placement of this genus was overdue.

Over the past decade material belonging to a new species of Eschatoporis has been collected from a cave in Napa County, California; thus allowing for a representative of the genus to be sequenced and analyzed within the context of a large pre-existing molecular dataset for the Lagriinae (Kanda et al. 2015). This new species, Eschatoporis styx, is described below.

DNA was extracted from a specimen of Eschatoporis styx sp. n. collected from the type locality (Clay Cave), using a Qiagen DNeasy Blood and Tissue kit. Four gene fragments were amplified: 28S nuclear ribosomal DNA (28S), arginine kinase (ArgK), carbamoyl phosphate synthetase domain of the rudimentary gene (CAD), and wingless (wg). These gene fragments were previously sequenced for the Lagriinae sampled in Kanda et al. (2015). Polymerase chain reactions (PCRs) were performed on either an Eppendorf Mastercycler ProS or Mastercycler gradient Thermal Cycler using Ex Taq DNA polymerase (TaKaRa) and basic protocols recommended by the manufacturers. Primer pairs and cycler profiles are described in Kanda et al. (2015). PCR products were cleaned, quantified, and sequenced at the University of Arizona’s Genomic and Technology Core Facility using a 3730 XL Applied Biosystems automatic sequencer. Assembly of multiple chromatograms of each gene fragment and initial base calls were made with Phred v. 0.020425.c (Green and Ewing 2002) and Phrap v. 0.990319 (Green 1999) as orchestrated by Mesquite’s Chromaseq v. 1.12 package (Maddison and Maddison 2014a, 2014b) with subsequent modifications by Chromaseq and manual inspection. Final sequences are available on GenBank (accessions MF370333‒MF370336).

Sequences were incorporated into matrices from Kanda et al. (2015). The final matrix (http://insectbiodiversitylab.org/data/) includes 31 Lagriinae spanning all currently recognized tribes and five outgroup taxa from other subfamilies of Tenebrionidae.

Ribosomal 28S gene fragments were aligned using MAFFT v. 7.130b (Katoh and Standley 2013) and the L-INS-i algorithm. ArgK sequences were manually aligned, asthere were no indels among our sampled taxa. CAD and wg were first translated to amino acid sequences, which were aligned using MAFFT v. 7.130b (Katoh and Standley 2013) and the L-INS-i algorithm with default parameter values. The nucleotide sequences were then mapped onto the amino acid alignment using Mesquite (Maddison and Maddison 2014b). The 28S, CAD, and wg alignments contained regions with numerous indels. These poorly aligned regions were identified using the server version of Gblocks (Castresana 2000; Talavera and Castresana 2007) with all options for less stringent block selection chosen. For CAD and wg, the “Codon” option was selected to maintain the triplet codons in the alignment.

Phylogenetic analyses were performed on a concatenated dataset of all four genes using maximum likelihood (ML), Bayesian (MB), and parsimony (MP) methods. For ML and MB analyses, optimal dataset partitions and substitution models were identified using the BIC implemented in PartitionFinder v.1.1.1 (Lanfear et al. 2012) from initial schemes based on genes and codon position. Two analyses were conducted, first restricting examined models to only those available in RAxML (ML) and then restricting models to only those available in MrBayes (MB). The inferred optimal data partition for ML analyses grouped first and second codon positions of all genes in the first partition, codon position three of ArgK and wg in the second partition, codon position three of CAD in the third partition, and 28S in the fourth partition. GTR+I+G was identified as the optimal substitution model for all partitions. The optimal partitioning scheme for MB analyses was the same, but SYM+I+G was identified as the optimal substitution model for the fourth partition.

Maximum Likelihood (ML) analyses were performed using RAxML v. 8.2.9 (Stamatakis 2014) implemented through the Zephyr v. 1.1 package (Maddison and Maddison 2015) in Mesquite (Maddison and Maddison 2014b). Five hundred independent searches for the maximum likelihood tree and 1000 bootstrap replicates were run on all datasets. Bayesian analyses were conducted using MrBayes v. 3.2.2 (Ronquist et al. 2012) on servers maintained by the CIPRES Scientific Gateway (Miller et al. 2010). Analyses were run for 36.8 million generations using default search parameters (two independent runs each with one cold chain and three hot chains). The two runs were considered to have converged when the standard deviation of split frequencies fell below 0.01 and the estimated sample size (ESS) for all parameters was greater than 200, suggesting adequate mixing between the two independent runs. ESS was calculated using Tracer v. 1.6 (Rambaut et al. 2014).

Clay Cave is located in oak woodland in the California wine country adjacent to the northern margin of San Francisco Bay, California. Known since the 1870s, the cave formed as a soil pipe cave in an ash flow of the Miocene Sonoma Volcanics, a continental packet of rhyolitic to andesitic volcanoclastic sediments and tephras. The cave consists of 229 m of linear passage with several small rooms floored with a seasonal stream (see Elliott et al., in press: fig.17). It appears that this cave originated along root casts in the bedded volcanic sediments that are mostly altered to smectite clay locally stained with iron oxides. Subsequent invasion by seasonal streams has integrated the initial fist-sized soil pipes into vadose canyon passages. The cave has at least two seeping springs. Clay Cave also has a rich biota, including some unusual terrestrial invertebrates, and is ranked fourth in the most bio-diverse caves of California (Elliott et al. in press).

Repeated attempts to find larvae in the cave or acquire larvae from adults in the lab yielded no results.

Species of the tribe Eschatoporiini seem to be associated with deep interstitial layers in rocky soils or underground water flows. Specimens of the tribe are either collected in deep rocky soil layers or in caves, both near springs. In Clay Cave, most Eschatoporis styx were collected under rocks rather than walking freely. Sometimes specimens have been found dead in standing small pools water from spring seepage in the cave. Specimens of Eschatoporiini remain very rare in collections. For instance, as far as we know Eschatoporis nunenmacheri has not been recollected since 1958.

Eschatoporis species are very similar in appearance and biology to the laenine genus Hypolaenopsis (Masumoto, 2001), which was originally placed in Adeliini but subsequently transferred to Laenini (see Schawaller 2008). Some Hypolaenopsis species are superficially very similar to Eschatoporis, differing only in size. The species H. nanpingica (originally described in Laena by Schawaller 2001), which is blind with only an eye scar remaining (as in E. nunenmacheri), could be mistaken for a large species of Eschatoporis based on external morphology. Other Hypolaenopsis species also have reduced eyes. Even the species-rich genus Laena contains taxa with reduced eyes; such as L. subcoeca Kaszab, 1973 and L. sherpa Schawaller, 2002 (both from forest litter in Nepal), and L. deplanata Weise, 1878 from Turkey in which the eyes are reduced to single facets (Schwaller, personal communication).

Masumoto (2001) mentioned that all specimens of Hypolaenopsis nanpingica were taken from the upper hypogean zone by digging soil mingled with gravel beneath large stones to the depth of 20–30 cm, about five or six meters above a stream.