Systematics of the Sierra Nevada endemic earwig: Eulithinus analis (Forficulidae, Dermaptera)

. Since the description of Eulithinus analis (Rambur, 1838) the taxonomic position of this Sierra Nevada endemic earwig has been controversial. It has been subdivided in different taxa, assigned to various genera or transferred to different subfamilies. With the aim of clarifying its systematics, we performed a mitochondrial phylogeographic analysis using specimens from different localities of Sierra Nevada representing the diverse phenotypes treated as differentiated taxa until now, and a phylogenetic analysis including representatives of apparently distant, but morphologically close, relatives. The phylogenetic and phylogeographic results obtained using mitochondrial (cytb, cox1) and nuclear (ITS2) markers and the study of morphological characters, indicate that the Sierra Nevada Mountain Range harbors a single species of the genus Eulithinus . Based on these molecular data, the morphological characters used to date in the internal taxonomy of this group of earwigs, especially size and shape of the cerci, lack diagnostic validity and show a large inter-and intra-populational variability. These results imply the synonymy of Eulithinus montanus (Steinmann, 1981) with Eulithinus analis (Rambur, 1838) syn. nov. and the reconsideration of Eulithinus analis outside the subfamily Allodahliinae. We established the synonymy between Eulithinus analis (Rambur, 1838) and Forficula brevis


Introduction
The high elevations of the Sierra Nevada mountains, in southern Spain, are inhabited by an endemic set of earwigs, traditionally included in the genus Eulithinus Hincks, 1935 within the family Forficulidae Latreille, 1810.The discovery of the Sierra Nevada earwig was made by Rambur (1838), along a period of intense zoological exploration of the mountains of southern Spain (Waltl 1839;Rosenhauer 1856).The zoological exploration of today's Sierra Nevada National Park has been almost continuous since then.The Sierra Nevada mountain chain, which includes the highest elevations in continental Spain (Mulhacén,3482 m), harbours an impressive list of endemic taxa (e.g., Agriades zullichi Hemming, 1933; Polyommatus golgus (Hübner, 1813); Mylabris nevadensis (Escalera, 1915); Dinodes baeticus Rambur, 1837;Baetica ustulata Rambur, 1838;Eumigus rubioi Harz, 1973) (Ruano & Tinaut 2003;Ruano et al. 2013).Rambur's (1838) description of Forficula analis Rambur, 1838 depicted unambiguously the Sierra Nevada earwig: "This species is smaller than auricularia, and more or less reddish black.The colour of the head is a slightly obscure ferruginous.The corselet is square, a little rounded posteriorly, very slightly coarse, obscure red, with the sides a little lighter; the elytra are of the same colour as the prothorax and legs; their posterior and external angle is a little dilated.The abdomen is more or less reddish black, especially on the underside; it is narrower at the base than at the tip, and a little dilated in the middle; in the male, the last terguite is depressed posteriorly in the middle, and swollen on the sides, which dilate to form an obtuse posterior angle; the penultimate half-segment almost entirely covers the last, which is deeply indented, on the underside; the claws are either fairly long and curved from the base, which is slightly enlarged with a pointed tubercle on top and a spine on the inside; or they are fairly short, with the thicker branches curved along their length; but what particularly distinguishes this species is that the plate which covers the anus extends posteriorly into a thick blade, slightly narrowed at the base, hollowed out in the middle, forming two prominent lateral and posterior angles.The posterior end of the female is almost smooth; the branches of the claw are connivent, short, a little curved and crossed towards the end.This female is very similar to brevis, but narrower and longer.As the antennae are incomplete, I do not know how many rings they have.I discovered this curious species towards the end of August in the Sierra Nevada mountains; it lives under rocks" (translated from original in French; Fig. 1).But the taxonomic position of the species was controversial since then.The taxon was included subsequently in the genera Chelidura Latreille, 1825 (type species: Forficula aptera Mégerlé, 1825) by Dubrony (1878), Lithinus Burr, 1909 (type species: Forficula analis Rambur, 1838) by Burr (1909), Eulithinus Hincks, 1935 (replacement name for Lithinus) by Hincks (1935), Pseudochelidura Verhoeff, 1902 (type species: Forficula sinuata Germar, 1825) by Burr (1908) and Steinmann (1981), and Eulithinus again by most modern authors (e.g., Ebner 1959;Steinmann 1975Steinmann , 1984;;Harz & Kaltenbach 1976;Sakai 1992Sakai , 1996;;Herrera Mesa 1999;Pérez-Valcárcel et al. 2022).The genus Lithinus, and its replacement name Eulithinus, were created to include only the Sierra Nevada earwig (Burr 1911;Hincks 1935), and although all recent authors retained the genus Eulithinus as valid, there is a long-standing controversy on its phylogenetic position.Most of the authors dealing with Sierra Nevada earwigs, included them in the subfamily Forficulinae Latreille, 1810, or Anechurinae Burr, 1907. However, Steinmann (1975), followed by Sakai (1992Sakai ( , 1996) ) and Herrera Mesa (1999) included them in the subfamily Allodahliinae Verhoeff, 1902, a subfamily previously considered to be restricted to eastern Asia (Srivastava 1993).To complicate matters, Steinmann (1981Steinmann ( , 1984Steinmann ( , 1989Steinmann ( , 1993)), after a series of taxonomic inconsistencies involving Sierra Nevada earwigs (see García-París 2017 for the problems in Anisolabididae Verhoeff, 1902) described a second taxon within the genus Eulithinus: Eulithinus montanus (Steinmann, 1981), also endemic to Sierra Nevada.Since then, the systematics of the Sierra Nevada earwig became unsettled.Some authors followed more or less strictly Steinmann's (1975Steinmann's ( , 1979Steinmann's ( , 1981Steinmann's ( , 1984) ) taxonomic proposals (Herrera Mesa 1999;Haas 2010;Pérez-Valcárcel et al. 2022), while those with local knowledge of the Sierra Nevada fauna simply disregarded part of Steinmann's criteria without further comments (Lapeira & Pascual 1980;Pascual & Barranco 2013).
We had the opportunity to examine some of the type specimens used by Steinmann (1979Steinmann ( , 1981Steinmann ( , 1984) ) for his descriptions.At the same time, we were able to collect for study, specimens representing all the phenotypes treated as separate taxa of Eulithinus in the Sierra Nevada National Park, allowing us to perform morphological and molecular analyses.
The objective of this work is to clarify the systematics of the endemic Sierra Nevada earwigs related to the Rambur' species (originally Forficula analis), in order to solve the taxonomic problems that complicate their ecological study and especially their conservation.The particular objectives are (1) to identify how many evolutionary units inhabit the high elevations of Sierra Nevada, and (2) to evaluate their phylogenetic position, including their subfamily placement.

Morphological study
We studied 12 specimens of Eulithinus at the collection of the Magyar Természettudományi Múzeum (Hungarian Natural History Museum, Budapest, Hungary: HNHM).Three of these specimens are the holotype of Pseudochelidura montana Steinmann, 1981 (= E. montanus), and respective paratypes of Eulithinus hispanicus Steinmann, 1984 (= E. montanus), and Pseudochelidura minor Steinmann, 1979.Type specimens of E. analis and F. brevis were not studied; according to Steinman (1993) the former are holded in the Natural History Museum of London.Additional 229 dry-mounted specimens of Eulithinus were studied at the Entomology collection of the Museo Nacional de Ciencias Naturales (Madrid, Spain) (MNCN_Ent); 71 ethanol preserved specimens were collected along the Sierra Nevada mountain range, and stored at -20°C at the MNCN collection.Dry-mounted and alcohol preserved specimens were used to identify variability in cerci size and shape under the stereo microscope.Extended depth-of-focus images of dry-mounted specimens were taken with a digital camera Nikon and a lens Nikon AF-S VR Micro-Nikkor 105 mm f/2.8GIF-ED, using the software Helicon Remote ver.3.9.11 and Helicon Focus ver. 7.6.4.

DNA extraction, sequencing, sequence alignment, and phylogenetic analyses
A total of 20 specimens of Eulithinus collected by biologists of the Sierra Nevada National Park Service (see Acknowledgements) and us, on three expeditions in July 2010, June 2011 and June 2010 (Ruiz & García-París 2011;García-París 2017) to Sierra Nevada, were selected for molecular analyses (Table 1).Nine of these specimens corresponded to the morphology of typical E. analis according to Rambur's (1838) original criterion, three correspond to Steinmann's (1981) criterion for E. montanus, and, four additional male specimens corresponding to intermediate cerci morphology as described by Ebner (1959).Two additional specimens were females, and two more, nymphs.
DNA extractions were performed from one leg of each specimen.DNA was extracted with a DNeasy Blood & Tissue Kits (QIAGEN NV, Hilden, Germany, Europe).Polymerase chain reaction was used to amplify fragments of two mitochondrial genes, the cytochrome c oxidase I (cox1, 658 pb) with the pair of primers LCOI1490/HCOI2198 (Folmer et al. 1994) and the cytochrome b gene (cytb, 385 pb) using the set of primers CB-J-10933 (Simon et al. 1994) and CB4 (Pons 2006); plus one nuclear region, the ribosomal internal transcribed spacer 2 (ITS2, 336 pb), amplified using the pair of primers CAS5p8sFc/ CAS28sB1d (Ji et al. 2003).PCR amplifications were performed in a final volume of 25 μl, including 3 mM MgCl 2 , 0.4 mM of each dNTP, 0.2 μM of each primer, template DNA (10-100 ng), and DNA polymerase (1 unit; Biotools).PCRs run for 40 cycles (denaturing at 94°C for 60 s, annealing at 42°C (cox1 and cytb) or 48°C (ITS2) for 60 s, and extension at 72°C for 60 s) with an initial denaturing step at 96°C for 5 min and a final extension at 72°C for 5 min.Sequencing reactions were performed using both strands in standard conditions and sequenced at Macrogen Inc. (Macrogen Spain).

Species concept
To delimit how many singular and evolutionary independent taxa can be defined within Eulithinus (sensu Hincks 1935), we adopted the evolutionary species concept (Wiley 1978) as discussed in Ruiz & García-París (2015) and Sánchez-Vialas et al. (2020).This concept considers a species as "a single lineage of ancestral descendant populations of organisms which maintains its identity from other such lineages and which has its own evolutionary tendencies and historical fate" (Wiley 1978: 18).Testing for cohesion was performed identifying levels of concordance between morphology and nuclear and mitochondrial markers (Sainz-Escudero et al. 2021;Cuesta-Segura et al. 2023).

Morphological study and phylogeographic and phylogenetic analyses
Male specimens studied from the Sierra Nevada mountain range are quite homogeneous in external characters (including pattern coloration).Only size and shape of cerci and pygidum are conspicuously variable (Figs 4-5) and may vary within populations as described in detail by Ebner (1959).The type series of P. montana consisted on the holotype and one paratype, both from Sierra Nevada (the paratype held at the Zoologiske Museum of the University at Copenhagen).We studied the holotype (Fig. 3B) and it corresponds to a male specimen of E. analis with short cerci.
The phylogeographic networks based on the mitochondrial gene cox1 marker (Fig. 6) showed six different haplotypes for Eulithinus (sensu Hincks 1935) in our sample.One haplotype was recovered from 12 male specimens collected in the Laguna de Río Seco that differed in cerci morphology: seven with large cerci (MNCN_Ent 344346,344352,344353,344356,344359,344360  ) and a male with large cerci (MNCN_Ent 344344) from Pradollano.These sequences were poorly differentiated, showing a genetic distance (uncorrected p distance ) between 0 and 0.017 (Table 2).
Nuclear ITS2 data were only obtained for two specimens from Borreguiles del Río San Juan (MNCN_Ent 344349, 344350) representing small-and large-cerci male phenotypes respectively.Both sequences were 100% identical, supporting that male cerci variation is part of a single evolutionary unit.Phylogenetic analyses of these sequences together with the available Pseudochelidura ITS2 sequences, resulted in a Fig. 4. Habitus variability in males of Eulithinus analis (Rambur, 1838) from Sierra Nevada (Spain).
A. Small specimen with short cerci, corresponding to the original description of Eulithinus montanus (Steinmann, 1981) from Laguna de las Yeguas at 2859 m of altitude (MNCN_Ent 122048).B. Large specimen with long cerci also from Laguna de las Yeguas (MNCN_Ent 121966).
well-structured clade with Elithinus analis sister to the species of Pseudochelidura, rendering Eulithinus and Pseudochelidura lineages reciprocally monophyletic (Fig. 8).The topology of the ITS2 hypothesis is not consistent with the mitochondrial hypothesis (Fig. 7) differing in the relative position of Eulithinus with respect to Pseudochelidura.
The history of the problem begins in 1979 when Steinmann described the Pyrenean species Pseudochelidura minor.Steinmann (1979) included a paratype from Sierra Nevada within the type series of the Pyrenean P. minor.We studied this paratype (Fig. 3A) and it is a male specimen of E. analis with short cerci.Steinmann thus believed that P. minor was present also in the Sierra Nevada mountains.This mistake was possible because, at a first glance, the external appearance of Pseudochelidura and Eulithinus is quite similar (size, coloration pattern, elytral development, and cerci morphology), but easy to reject upon examination of the external margin of the tegmina.Steinmann in 1981, still believing that Pseudochelidura was present in Sierra Nevada, described a new species: Pseudochelidura montana.To complicate matters, in 1984, Steinmann described one more taxon, using materials exclusively from Sierra Nevada: Eulithinus hispanicus (Fig. 3C).
Years later, Steinmann (1989) became finally aware of the morphological similarity between the genera Eulithinus and Pseudochelidura.Consequently, he synonymized E. hispanicus and P. montana, and transferred them to Eulithinus establishing the new combination: Eulithinus montanus (Steinmann, 1981).
According to the specimen labels, the holotype of P. montana Steinmann, 1981, and a paratype of P. minor Steinmann, 1979, were collected the same day by Steinmann himself at Sierra Nevada at 2700 m of altitude.The paratype of E. hispanicus Steinmann, 1984, was collected by G. Osella at the Refugio Universitario at 2600 m of altitude, together with five other male specimens identified by Steinmann as typical E. analis.These specimens represent some of the different morphs reported and described by Ebner (1959) as part of his study of E. analis population variability.
The characters used by Steinmann (1981Steinmann ( , 1984Steinmann ( , 1989Steinmann ( , 1993) ) in the description and the key to separate E. montanus and E. analis, were focused on the relative shape of male last tergite, including pygidium and forceps shape.In 1993, Steinmann indicated that male of E. analis presents "forceps strongly curved in basal section, and apically straightened; inner margins with one prominent tooth basally and pygidium ornamented with a lateral tooth" (Figs 4B, 5C).However, according to Steinmann (1993), the male of E. montanus present "forceps strongly curved basally and apically; inner margins with a small tooth medially.Pygidium ornamented with two lateral teeth" (Figs 4A, 5A).A direct examination of the morphological characters of the material of E. analis at the collection of MNCN and the type material used by Steinmann (1979Steinmann ( , 1981) ) in the description of E. montanus (= P. montana) at the collection of HNHM, shows that the external characters used by Steinmann (1979Steinmann ( , 1981Steinmann ( , 1984) ) to separate E. montanus from E. analis correspond to the intraspecific variability as already considered by Ebner (1959) (Figs 4-5, 9).
Pygidium and male forceps morphology are traits widely used as diagnostic taxonomic characters, however these are characters with high intraspecific variability in earwigs (Dohrn 1867;Diajonov 1925;Ollason 1970;Simpson & Mayer 1990;Tomkins & Simmons 1996;García-París 2017;Kirstová et al. 2020;Fontana et al. 2021;García-París et al. 2021;Jurado-Angulo et al. 2021;Cuesta-Segura et al. 2023).It has been recently demonstrated by morphological and molecular analyses (Fontana et al. 2021;Jurado-Angulo et al. 2021;Cuesta-Segura et al. 2023), that the misinterpretation of the morphological variability of male cerci in Chelidura and Pseudochelidura led to errors in the determination of the geographic distribution of the species and in the delimitation of independent evolutionary units (e.g., species).
Just as there is variability in the morphology of male cerci, there is also intraspecific variability in male genitalia within Eulithinus.Male genitalia of E. analis and E. montanus were studied in detail and illustrated by Steinmann (1981Steinmann ( , 1993)).According to Steinmann (1993) males of E. analis present "genitalia normally developed; central parameral plate somewhat narrowed basally, and expanded apically; virga within genital lobe of Forficula-type, short; external parameres comparatively long, straight, obtuse apically" (Steinmann 1993: fig. 586). Steinmann (1993) also indicated that males of E. montanus present "genitalia simple, Forficula-type; central parameral plate well developed, broad, oval, virga within genital lobe short, with basal vesicle; external parameres fully developed, more or less straight, with a very small spine at tips" (Steinmann 1993: fig. 589).These two diagnoses do not differ significantly between them, being its differences within the intraspecific levels of variability shown for example in P. cantabrica (Cuesta-Segura et al. 2023).
In addition, our phylogeographic analyses performed with specimens collected at Sierra Nevada National Park show that, at the mtDNA level, most individuals from a given population share a common cox1 haplotype with independence of cerci morphotype (Fig. 6).For example, the 12 males sequenced from the population of Laguna de Río Seco, that included specimens with cerci representing typical E. analis, typical E. montanus, and intermediate morphotypes (as described by Ebner 1959, andSteinmann 1993), share a common haplotype.All localities and specimens together are represented by 6 haplotypes in total, with a maximum divergence of 10 base pair changes between them (maximum uncorrected p distance = 0.017) (Table 2).Some specimens from Borreguiles del Río San Juan and Arroyo del Albergue de San Francisco share a common haplotype.Maximum internal haplotype divergence within a single population is represented by three base changes within the Borreguiles del Río San Juan population (uncorrected p distances = 0.005) (Table 2).Studied populations are clustered in three relatively isolated haplogroups, with apparently little exchange among them (Fig. 6): a) the high altitude specimens from the eastern-most sequenced population (Laguna de Río Seco; Figs 9E-F, 10B, D); b) the specimens from Borreguiles del Río San Juan (Fig. 9C) and Arroyo del Albergue de San Francisco, separated by a few hundred meters, and likely part of a unique population; and c) the low elevation specimens from the human modified area at the westernmost area of Pradollano (Figs 9A-B, D, 10C).
According to the morphological data (all male morphotypes are present at a single locality, and all are represented across all other populations), and the phylogeographic information (haplotypes are not segregated according to male morphotype) we suggest that all specimens studied with independence of their cerci morphology are part of single species, and therefore, E. analis and E. montanus are different names applied to a unique evolutionary unit.Thus, we propose the new synonymy: Forficula analis Rambur, 1838 = Pseudochelidura montana Steinmann, 1981 syn.nov.A few pages before the description of F. analis, Rambur described Forficula brevis Rambur, 1838 using two female specimens from Sierra Nevada.Rambur (1838) mentioned in the original description of F. brevis that the two specimens known at the time, were similar to female specimens of F. analis, but lighter in coloration, more reddish, larger (smaller than Forficula auricularia), but relatively stout with the maximum abdominal width placed a little further than the middle.Fieber (1853) suggested that F. brevis likely represented the female of F. analis.The same year, Fischer (1853: 75) proposed with doubts and only studying female specimens, the synonymy of F. brevis with Forficula decipiens Gené, 1832 (type locality: "nei contorni de Genova... sugli Apennini liguri"), probably influenced by the indications of Rambur about the absence of wings and larger size than in F. analis, and believing that, akin to other Mediterranean countries, F. decipiens was a common species in southern Spain.Since then, all recent authors treated F. brevis as a synonym of F. decipiens.However, we do not find evidences that support Fischer's (1853) synonymy.Despite detailed prospections, F. decipiens has never been found in the Sierra Nevada mountain chain (Pascual & Barranco 2013) and female specimens of F. decipiens do not show the reddish coloration or the particular body shape described by Rambur (1938).On the contrary, some particularly large females of E. analis (Fig. 11) with a general shape and coloration very similar to Rambur's (1838) original description, appear occasionally mixed with typical specimens of E. analis.These specimens not only fit precisely the description of F. brevis but also, they are found under stones in the Sierra Nevada mountains as indicated by Rambur (1838).Taking into consideration these arguments we formalize the suggestion of Fieber (1853) and propose the synonymy of Forficula brevis Rambur, 1838 with Forficula analis Rambur, 1838 syn.nov.Since F. brevis and F. analis were described in the same work, F. analis can be retained as the valid name for the species despite of being described a few pages latter (article 24.2 of International Code of Zoological Nomenclature; ICZN 1999).

Taxonomic output
Class Insecta Linnaeus, 1758 Order Dermaptera De Geer, 1773 Family Forficulidae Latreille, 1810Genus Eulithinus Hincks, 1935 Nuclear and molecular markers indicate that cerci variation in Eulithinus is part of the phenotypic plasticity within a single evolutionary unit.Under this scenario we consider that all the species level taxa described from Sierra Nevada using as diagnostic characters the shape and size of cerci and pygidium are part of a single species: E. analis (see Discussion, section The taxonomic status of Pseudochelidura montana and Eulithinus analis, p. 13).
The synonymic list of Eulithinus analis remains as follows:

Phylogenetic placement of the Sierra Nevada earwig and comments on the systematics of Forficulidae
The Sierra Nevada earwig shares with the species of Pseudochelidura a similar general morphology, including similar complex patterns of male cerci and pygidial variation and a common general habitus (Cuesta-Segura et al. 2023).Morphological resemblance between species both of Eulithinus and Pseudochelidura partially agrees with the mitochondrial phylogenetic results, which showed that the Sierra Nevada specimens are deeply nested within Pseudochelidura and sister to the Pyrenean P. sinuata (Fig. 7).However, we are well aware of the problems of using only mitochondrial DNA data for taxonomic purposes (Sainz-Escudero et al. 2021), because organismic evolution is nuclear and therefore mtDNA can only be used as a proxy when no nuclear evidence suggest otherwise (Moritz et al. 1987;Rissler et al. 2006;Singhal & Moritz 2013), because mtDNA introgression is very common across species with postzygotic isolation mechanisms (Shimizu & Ueshima 2000;Alves et al. 2003;Babik et al. 2005;Barbanera et al. 2009;Zakharov et al. 2009;Nunes et al. 2010;Mastrantonio et al. 2016), and most importantly, because its haploid nature and maternal inheritance generate very different demographic dynamics with respect to nuclear genes (Avise et al. 1987(Avise et al. , 1988;;White et al. 2008).However, it is generally considered that mitochondrial DNA data are particularly useful to depict the biogeographic and past evolutionary history of a taxon (precisely because its maternal inheritance and lack of recombination allow for its long population persistence through time, as shown since the early works of Moritz et al. 1992;García-París et al. 2003).In our case, the cytb data clearly indicates that E. analis and P. sinuata became isolated from a common ancestor much more recently than they did with respect to P. cantabrica.However, the ITS2 phylogeny clearly placed Eulithinus as the sister taxon to the Iberian Pseudochelidura, supporting, or at least not questioning their generic status.
In this situation of morphological similarity, in which, however, the presence of a marked edge in the lateral side of the tegmina of Eulithinus could be clearly interpreted as a generic diagnostic character, and with mtDNA supporting Eulithinus as deeply nested within Pseudochelidura, we cannot make a proper decision on the taxonomic status of Eulithinus.The inclusion in the analyses of the Italian species of Pseudochelidura (P.galvagnii Vigna Taglianti, 1999, and P. orsinii (Gené, 1833)) might shed some additional light on whether Eulithinus should remain as a valid genus or if, alternatively, it should be transferred to Pseudochelidura.Even Steinmann (1979Steinmann ( , 1981) ) (1975), followed by Sakai (1992Sakai ( , 1996) ) and Herrera Mesa (1999) included them in the subfamily Allodahliinae Verhoeff, 1902. Steinmann (1975) considered that the criterion "border on the sides of the elytra" given by Verhoeff (1902), which also served to describe Allodahlia Verhoeff, 1902, was a sufficiently important characteristic feature to elevate the group to the rank of subfamily.For this reason, Steinmann included in Allodahliinae all species that possess this character state including Allodahlia and Eulithinus.
Since Eulithinus is close to or nested within Pseudochelidura, it appears that the presence of a marked edge in the humeral angle of Eulithinus tegmina, absent in Pseudochelidura, is a homoplastic trait with respect to the same structure present in Allodahlia.In other words, Allodahlia and Eulithinus would present similar traits in the tegmina by mere convergence and not by presenting a close phylogenetic relationship between them.This inference renders Allodahliinae without a proper diagnostic definition.We consider that the use of subfamilies within Forficulidae should be avoided until a robust phylogenetic analysis of the entire Forficulidae would define properly internal monophyletic subunits (subfamilies) within it.
The close relationship between endemics of Sierra Nevada and Pyrenean taxa is a common theme in the biogeography of European arthropods.Some of the most emblematic insects of Sierra Nevada as Agriades zullichi or Mylabris nevadensis are sister to their Pyrenean, often extended over other European mountain systems, counterparts, and allied species, as Agriades glandon glandon (von Prunner, 1798) plus A. g. aquilo (Boisduval, 1832), and Mylabris flexuosa Olivier, 1811 (Martin et al. 2002;Wiemers et al. 2020).The isolation of the Sierra Nevada and Pyrenean (Western European) taxa have been considered a result of the extinction of the intervening populations during the interglaciar periods.Cold adapted species inhabiting the dry and thermoclimatic areas of southeastern Spain would only survive at high elevations, becoming isolated from all other mountain systems in Western Europe (Hewitt 2001;Wilson et al. 2005;Ploquin et al. 2013).Our cytb data support that Sierra Nevada populations are related to Pyrenean populations in a well-supported monophyletic group, leaving the Cantabrian populations as sister to them.

Natural history notes and conservation
Eulithinus analis is an endemic taxon to the Sierra Nevada and the adjacent Sierra de Filabres mountain chains in the Penibetic Region of southeastern Spain.The species occurs at high elevations, from 2039 (Pilar de las Yeguas, Puerto de la Ragua) to 3029 m (Laguna de Río Seco; Fig. 10A) (Ebner 1959; own data).It is not uncommon in the western area of its range, but it becomes more sporadic towards the east.
We have failed to locate the species in the Sierra de Filabres (Almería), where early 20 th century explorers located some specimens (see the historical MNCN Material examined) and where general conditions seem favourable for the species; in any case, the taxonomic status of these populations require further analyses.Eulithinus analis is relatively frequent in open areas within rocky valleys partially covered with grass, and associated with humidity, as small lakes, streams and "borreguiles" (peat bogs) (Ebner 1959) (Fig. 10), but also at the edge of pine forests (Puerto de la Ragua).Adults are usually found under stones or active at night in humid areas of beneath decaying wooden logs in pine forests during summer months (June to August).Females attend their eggs during this period, sometimes aggregated under the same stone, and often with the remains of dead males nearby (Fig. 10B).
During the explorations to localize E. analis in Sierra Nevada and Sierra de Filabres, we often found specimens of Forficula mediterranea González-Miguéns & García-París, 2020(in González-Miguéns et al. 2020), a member of the F. auricularia species group.The presence of this species at low-mid elevations of Sierra Nevada was expected, but we also observed large densities at high elevations areas in which E. analis typically occurs.In mid-August 2011, we observed more than 300 specimens of F. mediterranea and F. dentata at Puerto de la Ragua, about 50 at Laguna de Aguas Verdes at 3058 m (37°02′53″ N, 3°22′06″ W), and one female at Laguna de Río Seco.The specimen at Laguna de Río Seco (Fig. 9F) was located under a stone where females of E. analis were tending egg clutches (Fig. 10B), but we did not observe any interaction between them.At Laguna de Aguas Verdes we did not find a single specimen of E. analis, while at Puerto de la Ragua, F. mediterranea clearly dominated, but E. analis was not uncommon.Species of the F. auricularia complex appear in company of species of Pseudochelidura in the Cantabrian Mountains (Cuesta-Segura et al. 2023), but these limited observations were not designed to identify patterns of interaction between the high elevation specialists of the genus Psudochelidura and the widespread euryoic species of the F. auricularia species group.Unfortunately, we do not have previous data on the presence of F. mediterranea (even under the name of F. auricularia) at high elevations in Sierra Nevada.
Most of the area occupied by E. analis is currently under the protection of the Sierra Nevada National Park, so we do not expect immediate threats for the survival of the species other than the consequences of climate change, including a likely reduction of their presence area.It is imperative to acknowledge that the coexistence of species from the F. auricularia complex in the same areas as E. analis may constitute a potential threat to the latter species.This is attributed to the fact that species within the F. auricularia complex exhibit substantial competitive abilities, and the ongoing climatic changes are fostering their adaptation to higher altitudes (Pavón-Gozalo et al. 2011).We believe it is important to determine if the populations of E. analis from Sierra de Filabres still exist or if they have already disappeared.In this respect, we like to call attention to the record of a specimen of Chelidura, reported in Fontana et al. (2021) from Pico de Veleta, treated as doubtful by Jurado-Angulo et al. (2021), but that requires further exploring.

Fig. 2 .
Fig. 2. Geographic distribution of Eulithinus analis(Rambur, 1838).The green shading on the map represents the area of the Sierra Nevada National Park usually considered as the general area of the species.The purple dots correspond to the records of the species based on the specimens studied from the Entomology collection of the Museo Nacional de Ciencias Naturales, including dry mounted and ethanol preserved material.

Fig. 6 .
Fig. 6.Mitochondrial (cox1) network analyses of specimens of Eulithinus analis (Rambur, 1838) from Sierra Nevada.Specimens from Laguna del Río Seco are colored in green, from Pradollano in red, from Arroyo del Albergue San Francisco in yellow and from Borreguiles del Río San Juan in purple.

Fig. 7 .
Fig. 7. Bayesian phylogenetic tree based on cytb sequences of Eulithinus Hincks, 1935 and Pseudochelidura Verhoeff, 1902.Species and clade names correspond to the new classification adopted in this work.Posterior probabilities are indicated above branches.Background colours represent species and names below branches represent geographic areas.

Fig. 8 .
Fig. 8. Bayesian phylogenetic tree based on ITS2 sequences of Eulithinus Hincks, 1935 and Pseudochelidura Verhoeff, 1902.Species and clade names correspond to the new classification adopted in this work.Posterior probabilities are indicated above branches.Background colours represent species and names below branches represent geographic areas.

Fig. 9 .
Fig. 9. Habitus of live specimens of Eulithinus analis (Rambur, 1838) from Sierra Nevada (Spain). A. Male specimen with long cerci from Pradollano.B. Typical female with dark coloration and brown head from Pradollano. C. Male with intermediate sized cerci from Borreguiles del Río San Juan (MNCN_Ent433449).D. Light colored female with reddish head and pronotum from Pradollano.E. Male displaying short cerci from Laguna de Río Seco (MNCN_Ent344345).F. Male with each cercus of different size from Laguna de Río Seco.Photographs ex situ by MG-P.

Fig. 10 .
Fig. 10.General view of the habitat and natural history notes of Eulithinus analis (Rambur, 1838) in Sierra Nevada (Spain). A. Typical landscape occupied by E. analis; in front Laguna de Río Seco, where numerous female specimens were observed taking care of their eggs under stones in August.B. Two females taking care of their eggs under the same stone at Laguna de Río Seco; note the remains of a dead male on the left, and the different development degree of the two egg masses; a few minutes later, females clustered the eggs into compact masses.C. Male with intermediate long sized cerci running in alert display at Pradollano.D.Last instar nymph from Laguna de Río Seco; note the general similarity with nymphs of Pseudochelidura cantabrica (Cuesta et al. 2023: fig.5c).Photographs by MG-P.

Fig. 11 .
Fig. 11.Female specimen of Eulithinus analis(Rambur, 1838) from Borreguiles del Río San Juan (Sierra Nevada) with atypical body proportions and with coloration matching the descriptions of Forficula brevis Rambur, 1838.The presence of atypical specimens mixed with typical females of E. analis in the high elevations of Sierra Nevada supports the synonymy between F. analis and F. brevis.

Table 2 .
(Rambur, 1838))pairwise genetic distance matrix between specimens (code and morphotype cerci) used in the mitochondrial (cox1) network analyses of Eulithinus analis(Rambur, 1838).MNCN_Ent 344345 and 344358).A second haplotype is shared by four specimens, two of them nymphs (MNCN_Ent 344347, 344348) from Arroyo del Albergue de San Francisco, and two males (MNCN_Ent 344350 with large cerci and MNCN_Ent 344349 with small cerci), from the Borreguiles del Río San Juan.Four additional haplotypes were represented by single specimens, a female (MNCN_Ent 344361) from Arroyo del Albergue de San Francisco, a male with intermediate cerci (MNCN_Ent 344351) from Borreguiles del Río San Juan, and a female (field number MNCN tij11503 and field number MNCN tij11935), three with intermediate cerci(MNCN_Ent 344354, 344355 and 344357), and two Following the classification criteria used by Burr in 1907, many authors who dealt with the earwigs of Sierra Nevada, included them in the subfamily Forficulinae or Anechurinae.However, Steinmann while including Sierra Nevada specimens as paratypes of his Pyrenean Pseudochelidura minor (= P. sinuata) (Cuesta-Segura et al. 2023), or while describing his Sierra Nevada species Pseudochelidura montana, mixed Eulithinus and Pseudochelidura.