A Highly Troglomorphic New Genus of Sminthuridae (Collembola, Symphypleona) from the Brazilian Semiarid Region

Simple Summary A new genus and species of cave springtail are herein described. The new species has a remarkable morphology, suggesting it evolved within caves, such as the appendages and chaetae elongation, loss of body pigments, and 5 + 5 eyes reduced in size. The new species may also be occasionally cannibalistic, as one of the analyzed females devoured another male from her species. Abstract Here, we describe the highly troglomorphic Troglobentosminthurus gen. nov. from Água Clara cave system, Caatinga domain, Bahia, Brazil. Troglobentosminthurus luridus gen. nov. sp. nov. has remarkably long antennae, legs and furca, and lacks body pigments, except for small orange eye patches which also show a reduction in the number of eyes (5 + 5) and lens sizes. The overall morphology of the genus, with long and highly sub-segmented antennae, resembles other Sminthurinae of the Temeritas-group, especially Temeritas Richards and Galeriella Ćurčić and Lučić. However, it is unique, especially in the combination of the number of antennae IV subsegments and eyes, frontal head chaetotaxy and empodial complex morphology. Two type specimens have remnants of a mite and another specimen from the new species in their gut contents, supporting the species may be occasional predators and even cannibals. We also provide identification keys and comparative tables to the subfamilies of Sminthuridae and the Temeritas-group of genera.


Introduction
The knowledge on the Brazilian Symphypleona strongly increased during the past two decades. In 2003, there were only 40 species recorded in the entire country, while in 2022, there are 73 nominal species recognized in Brazil [1,2]. Similarly to the Poduromorpha and Entomobryomorpha, this remarkable rise in the taxonomic understanding of the Symphypleona was mainly due to the description of new species rather than the register of previously described taxa from other countries [2]. In this context, the Brazilian fauna of Symphypleona is unique, with 57 endemic species (about 78% of its total), as well as some genera only found in Brazil, like Arborianna Bretfeld, 2002 [3] (Katiannidae), Keratosminthurus Zeppelini 2020 in [4], and Varelasminthurus Silva, Palacios-Vargas and Bellini, 2015 [5] (both Sminthuridae).

Materials and Methods
Specimens were collected manually with entomological brushes and ethanol in the dark zones of Água Clara cave system (more details in topic 3.3, habitat and threats), preserved in 70% ethanol at 6 • C, cleared in Nesbitt's solution, washed in Arlé's liquid and mounted on glass slides in Hoyer medium, mixing [30,31] procedures. Ethanolfixed specimens were photographed in a Nikon SMZ1500 stereomicroscope with a DS-Ri1 camera, under NIS-Elements AR v.4.51 software. Morphological studies and first drawings were made under Leica DM750 phase-contrast microscope, with an attached drawing tube. Final drawings were vectorized in Adobe Illustrator 2020 software. The type series was deposited at the Collection of Subterranean Invertebrates of Lavras (ISLA), linked to the Center for Studies in Subterranean Biology from the Federal University of Lavras ( biologiasubterranea.com.br (accessed on 10 June 2020)).
The terminology used in the text and figures follows [32] to labial palp papillae and guard-chaetae, [33] to the labral chaetotaxy, [34] to the head and anterior large abdomen chaetotaxy, Vargovitsh system [35][36][37] with adaptations for the posterior large abdomen chaetotaxy, [38] for the small abdomen chaetotaxy, and [39] to part of the tibiotarsal chaetotaxy. Regarding the dens we considered the dorsal, dorso-internal and dorso-lateral rows summed as the dorsal chaetotaxy. Drawings and observations were made based on the entire type series.
The abbreviations used in the description and figures are: Abd.-abdominal segment(s); Ant.-antennal segment(s); Th.-thoracic segment(s). In the drawings, structures 1. Ungues with cavity and/or a filament-like tunica, tibiotarsal distal whorl with six or fewer chaetae .

Troglobentosminthurus gen. nov. Bellini, Medeiros and Souza
Diagnosis of the genus. Appendages remarkably elongated. Chaetae smooth and long, especially on antennae, tibiotarsi and posterior large abdomen, spine-like chaetae absent on head frontal area and dorsal large abdomen. Most chaetae alveoli slightly elongated, somewhat neosminthuroid-like, on the body and appendages. Ant. IV Temeritas-like, longer than the body combined with Ant. I-III, with 44 subsegments, Ant. II apically with one ventro-lateral bothriotrichum-like sensillum. Antennae and head chaetotaxy not sexually dimorphic. Mouthparts normal, not elongated, labial palp guard-chaeta a1 laterally displaced and curved. Large abdomen without dorsal vesicles, small abdomen formed by Abd. V-VI. Bothriotricha A-D present and elongated, especially D, longer than the dens. Parafurcal area without sexually dimorphic or neosminthuroid chaetae. Trochanter III with a large spine in a rounded alveolus, tibiotarsi I-III distal whorls with 8/9/9 chaetae, respectively, pretarsi I-III with only a small anterior chaeta, ungues without cavity and tunica, but with underdeveloped pseudonychia, unguiculi without the apical filament. Ventral tube sacs warty, tenaculum with 1 + 1 distal chaetae. Dens ventral formula from the apex to the basis as 3, 2, 2, 1 . . . 1, mucro without chaeta, with the inner lamella serrated and the outer smooth, mucronal apex symmetric.
Etymology. The new genus honors our dear friend, the emergent skillful biospeleologist Dr. Diego de Medeiros Bento, nicknamed "Troglobento". Diego works at the National Center for Cave Research and Conservation (ICMBio/CECAV), tracking, mapping and studying caves and cave fauna in Brazil, mainly in the semiarid region of Rio Grande do Norte state, Brazil. His efforts, as well as those of many other professionals working for CE-CAV and other entities, helped to reveal and preserve several Brazilian caves. Contrary and irresponsibly, the Brazilian president signed in 2022 the federal decree 10,935, which enables the destruction of even the most relevant caves in Brazil for commercial purposes, putting at risk such efforts. Thus, the new genus honors not only Dr. Diego Bento but also many other Brazilian professionals committed to the conservation of underground environments.
Remarks. The overall morphology of the new genus puts it within the Temeritas-group sensu Medeiros et al. [29], which gathers Temeritas Richards, 1963 in [51], Richardsitas Betsch, 1975 [52], GaleriellaĆurčić and Lučić, 2007 in [53], and Keratosminthurus. This group of genera and Troglobentosminthurus gen. nov. share: antennae mostly longer than the body, Ant. IV with 18-46 subsegments, post-antennal chaeta absent and metatrochanteral spine present. The Ant. IV with 44 subsegments, presence of eyes, head without sexually dimorphic features, large abdomen lacking spines but with slender chaetae, parafurcal area without neosminthuroid chaetae and tibiotarsi without capitate tenent-hairs put the new genus closer to Temeritas stricto sensu Medeiros and Bellini [54]. However, Temeritas differs from Troglobentosminthurus gen. nov. especially in the frontal area of the head with spines (vs. absent), unguiculi with the apical filament (vs. without), dens ventrally with 13 chaetae (vs. 9), specimens pigmented, mostly with complex color patterns (vs. lacking pigments), eyes 8 + 8 and normally developed (vs. 5 + 5 reduced eyes) [54]. The highly troglomorphic profile of Troglobentosminthurus gen. nov. strongly resembles that of Galeriella, a monotypic cave genus only recorded from Bosnia and Herzegovina [53,55], but Galeriella differs from Troglobentosminthurus gen. nov. in Ant. IV with 32 subsegments (vs. 44), eyes absent (vs. 5 + 5), ungues without teeth and pseudonychia (vs. present), tenaculum without chaetae (vs. 1 + 1) and mucro with both lamellae serrated (vs. only one). The distribution supports these genera emerged independently as well, since Galeriella was only found in the Balkan Peninsula, Palaearctic Region, and the new genus in the southern region of Caatinga Domain, Neotropical Region. Galeriella also presents a remarkably underdeveloped chaetotaxy, different from most Sminthurinae, including the Temeritas-group, which suggests that the two type specimens, listed as males, may be juveniles, or at least that some chaetae at the proximal antennal articles, head, tibiotarsi and ventral dens may have been overlooked in the original description [53]. Such doubts prevent us from better comparing the genus with Troglobentosminthurus gen. nov. Further comparisons between the Temeritas-group genera are presented in Table 2 and the identification key below.
Although the antennal elongation of Troglobentosminthurus gen. nov. is supposedly linked to the cave life, it may also be the result, at least partially, of a shared ancestral with Temeritas, which also have long and subsegmented Ant. IV, but live in epigeic ecosystems [54]. Morphology, as previously discussed, and distribution support this hypothesis. There are currently 12 species of Neotropical Temeritas, with seven of them registered from Brazil, of which three were also found in the Caatinga domain [2,54]. Nevertheless, only a phylogenetic study could test if Troglobentosminthurus gen. nov. and Temeritas are closely related.
The elongation of the tibiotarsi in the new genus splits the chaetae on each whorl following Nayrolles [39], suggesting at first that the distal whorl is similar to the Songhaicinae with only five or six chaetae. However, a more detailed study of the homology of the tibiotarsal chaetotaxy shows they fit the diagnosis of Sminthurinae, with 8/9/9 chaetae on tibiotarsi I-III, respectively, and three chaetae occur on the upper region of the distal tibiotarsi (see more details in the description of the new species). Other peculiar details on the morphology of the new genus are: the labial palp papillae similar to that of Sminthurus Latreille, 1802 [56] and Disparrhopalites naasaveqw, with the a1 guard-chaeta large, apically curved and laterally displaced, almost or clearly external to the papilla C [27,32], a trait not reported to any Temeritas, Richardsitas or Keratosminthurus species (unknown to Galeriella) [4,29,54]; the absence of the posterior pretarsal chaeta, only recorded before within Sminthuridae to the Songhaicinae genera Varelasminthurus and Disparrhopalites Stach, 1956 [5,27,57]; and the Ant. II apically with one ventro-lateral bothriotrichum-like sensillum, a feature previously only observed in the Bourletiellidae [26].  It is important to state we did not consider any chaetae on the parafurcal area as neosminthuroid because they all have similar slightly elongated alveoli, in accordance with most chaetae seen on the appendages and body. Further details on the chaetae which present different alveoli are detailed in the description of the new species.
Abdominal appendages (Figure 9). Ventral tube with 1 + 1 chaetae on the corpus and 1 + 1 on the lateral flaps, sacs long and warty. Manubrium with nine, and dens with 27 dorsal chaetae, respectively ( Figure 9A), dens ventrally (anteriorly) with nine chaetae inside rounded alveoli, following the formula from the apex to the basis: 3, 2, 2, 1 . . . 1 ( Figure 9B). Mucro with the inner lamella slightly serrated with about 70 small teeth, outer lamella smooth; mucro sexually dimorphic, with large crenulations on males in the serrated edge; mucronal chaeta absent ( Figure 9B-D). Ratio mucro:dens:manubrium of the holotype as 1:4.6:1.8. Tenaculum with three distal teeth plus one proximal basal tubercle on each ramus, corpus with 1 + 1 apical chaetae ( Figure 9E).   Etymology. The new species was named luridus (from Latin: lurid or shockingly macabre) due to its pale, ghostly habitus, and due to one of the type-series females has remnants of a devoured male on its guts, suggesting the species can be cannibalistic or necrophile.
Habitat and threats. Specimens of Troglobentosminthurus luridus gen. nov. sp. nov. were found in two caves from the Água Clara cave system (ACCS), located in the karst region of Serra do Ramalho, Carinhanha municipality, Bahia state, Brazil ( Figure 1A). The ACCS is composed of four limestone caves (Gruna da Água Clara, Gruna dos Índios, Lapa dos Peixes I and Lapa dos Peixes II caves) which are trespassed by an intermittent stream, active during the austral summer (October until March) ( Figure 1B). The whole system presents around 24 km, but there are still unsurveyed galleries. According to Köppen's climate classification system, the local climate is "Aw", with dry winter and an average annual rainfall of 640 mm [59]. The Serra do Ramalho region is inserted in the Caatinga domain (an exclusive Brazilian semiarid biome), with transitional areas to the Cerrado (Brazilian Savanna) [60].
Specimens of Troglobentosminthurus luridus gen. nov. sp. nov. were only found in deep and moistened areas of the dark zone of the system, which have only horizontal caves. It is interesting noting that the two caves located at the intermediate portion of the ACCS (Gruna dos Índios and Lapa dos Peixes caves) have the main conduit quite dry (even with the presence of some ponds) due to the wind, which trespasses both caves. Such caves present entrances on both sides of their main conduit, thus favoring the airflow. Hence, individuals of Troglobentosminthurus luridus gen. nov. sp. nov. were not observed in this intermediate portion of the ACCS, probably due to the unstable climatic conditions of the caves.
The individuals were found both associated with small bat guano pellets (especially in the Lapa dos Peixes II cave, Figure 2A,B) but also freely walking on the muddy sediment in the deep areas of the Gruna da Água Clara cave ( Figure 2C,D). In the latter, specimens were often found in areas with no visible organic debris, so it is likely that they might be feeding on some microbial content associated with the cave sediments. It is interesting to highlight that the caves of ACCS are seasonally flooded, as strong water flows trespass the caves during the rainy period, coming from several entrances along the ACCS ( Figure 1D). Thus, individuals of Troglobentosminthurus luridus gen. nov. sp. nov. somehow overcome such floodings, probably by climbing up the cave's walls, which can be considerably high ( Figure 1E), or entering rock cracks, but this subject certainly merits further research.
It is also noteworthy that in the deep regions where specimens are found, there are almost only troglobitic species [61], thus indicating that only highly specialized species can thrive under the restrictive conditions occurring in such areas (especially strong oligotrophy). Among the 30 cave-restricted species observed in the ACCS, only 8 (~26%) are formally described [61]. Thus, it is of paramount importance to describe the remaining troglomorphic species in order to preserve the ACCS, since non-described species are often ignored in conservation actions [62].
The Serra do Ramalho region has been the target of increasing deforestation in the last decades, mainly due to charcoal production and the establishment of monocultures ( Figure 1C). Forest removal in the areas that directly influence the caves can alter the organic sediments supply to the caves and negatively impacts bat communities as well, thus affecting the food resources available for the hypogean fauna [61]. Furthermore, species associated with systems with low availability of guano and carcasses, such as the ACCS, may be more dependent on allochthonous food resources [61].
Considering the recent setbacks in the Brazilian laws regarding cave protection [63], it is urgent to describe cave-restricted species from an important hotspot of subterranean biodiversity as the Agua Clara cave system. Since the caves are under serious threat, it is more urgent than ever to formally describe such endemic species, as they might become the only effective protection for the habitats they rely on.
Biology. The overall morphology of Troglobentosminthurus luridus gen. nov. sp. nov. with extraordinarily long appendages and chaetae, loss of pigments and part of the eyes, as well as the reduction of the eye lenses, support the species is a troglobite, especially compared to other species of the Temeritas-group. As discussed before, the limited distribution of the species within the ACSS, only found in deeper areas, reinforces this hypothesis.
The type specimens have soil particles in the guts mixed with some dark matter, suggesting the specimens may feed on bat guano, cave sediments with microbes or particles of organic matter ( Figure 10A). However, the gut contents of two specimens were more unusual. One of the studied males has two mite legs in its contents ( Figure 10A), while the holotype, a female, has the remains of one male of the same species in its digestive system ( Figure 10B,C). In the case of the analyzed male paratype, we could not track any other clear part of the mite other than the legs, but in the case of the female, we could confirm the morphology of the six empodial complexes of the digested specimen as the ones described to Troglobentosminthurus luridus gen. nov. sp. nov. Although this devoured specimen was fragmented, we could also identify its testicles, both labial palps and many body chaetae on the remains found inside the holotype, supporting it is a male of her own species (Figure 10B,C). In this scenario Troglobentosminthurus luridus gen. nov. sp. nov. specimens may be occasional predators/cannibals, or even necrophiles and feed on dead arthropods. Such uncommon feeding habitats for springtails must be better studied to confirm which hypothesis fits best the biology of the new species. However, under extreme oligotrophic conditions, such as those observed in the areas where the specimens are found inside the ACCS, all raised behaviors (predation/cannibalism/necrophagy) are quite plausible. Figure 10. Troglobentosminthurus luridus gen. nov. sp. nov., gut contents: (A) a male's contents, with a mite leg at the center and dark matter mixed with soil particles at the right side; (B) holotype (female) contents, yellow circle marks an empodial complex, red marks the testicles, and blue marks one labial palp; (C) holotype (female) contents, black circles mark four devoured empodial complexes.
Remarks. Since the genus is monotypic, the species and genus diagnoses should be considered complementary. They also may have conflicting characters, as specific traits marked as diagnostic of the genus and vice-versa, which should be better understood with the description of further species of Troglobentosminthurus gen. nov. in the future, if any. For taxonomical comparisons, see the remarks of the new genus and Table 2.

Discussion
The description of Troglobentosminthurus gen. nov., as well as Keratosminthurus and the Songhaicinae, test the boundaries of the internal systematics of the Sminthuroidea. Until now, no representative molecular analysis aiming to resolve the internal relationships of the Symphypleona was published. Recent studies using different markers and phylogenetic methods only gathered a few taxa of the order, unable to clearly investigate the validity and affinities of its internal groups, as in [64][65][66][67][68][69]. In this scenario, arguably, the most reliable phylogenetic data available are based on morphology, as in [25,41,42,70]. However, it is important to state that, even with limited sampling, molecular analyses do not clearly support important subdivisions of the Symphypleona. For instance, Schneider et al. [65] tree does not support the Dicyrtomidae as a family, Yu et al. [66] do not support the Sminthurididae, and Yu et al. [66], Sun et al. [68] and Cucini et al. [69] do not support the suborder Appendiciphora. These are all well-delimited taxa based on morphology [26], and poor and uneven sample size and choice of markers could be blamed for such results. Nevertheless, more detailed phylogenetic analyses of the Symphypleona using modern tools are lacking, making its internal organization sometimes dubious and weakly supported in the light of the discovery of new taxa, as in the case of the Sminthuroidea. We hope in a near future more data on the systematics of the Symphypleona, as well as the Collembola as a whole, emerge to clarify its organization and to test different hypotheses concerning their evolution.

Conclusions
In this paper we described the highly troglomorphic Troglobentosminthurus gen. nov. from a cave system in the Brazilian semiarid region. The new genus shows a remarkable morphology, with troglomorphisms not recorded before in any Brazilian species of springtails and intermediate traits which made us revise the diagnosis of the Sminthurinae and discuss the exclusive morphology of the Bourletiellidae. Troglobentosminthurus luridus gen. nov. sp. nov. is discovered at the same time the Brazilian laws regarding cave protection become more flexible to favor commercial expansion, while the cave system where the species inhabits is developed in a karst already threatened by deforestation. Urgent conservation strategies are clearly in need to preserve this new taxon.  Institutional Review Board Statement: Ethical review and approval were waived for this study, due to Brazilian laws which do not require permission from an institutional ethics committee on the use of animals for taxonomical studies with microartropods.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data is contained within the article. All biological material is deposited at ISLA (# 79662 and 79664) as previously stated.