Feaella (Tetrafeaella) obscura sp. nov. – a new pseudoscorpion species from the Maldives (Arachnida, Pseudoscorpiones), and an updated identification key to the subgenus Feaella (Tetrafeaella)

The Feaellidae Ellingsen, 1906 is a small but ancient family of pseudoscorpions with 20 extant species across the Southern Hemisphere, and fossils from the Lower Cretaceous of Myanmar and the Eocoene of Europe. Here, we describe and illustrate Feaella (Tetrafeaella) obscura sp. nov. as a new species from the Maldives archipelago in the Indian Ocean. This is the first record of Feaella from a young oceanic island and may indicate a potential for long-distance dispersal in this lineage. We also elevate Feaella (T.) capensis nana Beier, 1966 to full species rank as F. (T.) nana Beier, 1966 and provide an identification key to the members of the subgenus Feaella (Tetrafeaella), thereby facilitating the identification of species.


Introduction
Pseudoscorpions belonging to the family Feaellidae Ellingsen, 1906 are amongst the most unusual arachnids and have a spectacular morphology that includes a carapace with multiple frontal lobes, a dorsoventrally flattened and heavily sclerotised body, and pedipalps that are stout, heavily armed and without a venom gland in both fingers (Harvey et al. 2016b;Judson 2017). The family has a Pangaean distribution and amber fossils are known from the Lower Cretaceous of Myanmar (Henderickx and Boone 2016) and Baltic amber in Europe (Hendrickx and Boone 2014; Harms and Dunlop 2017), whereas twenty extant species are found in tropical and subtropical regions in central and southern Africa, north-western Australia, tropical India and Sri Lanka, the Seychelles, Madagascar, and Brazil ( Fig. 1) (Harvey 2013(Harvey , 2018Harvey et al. 2016a, b;Judson 2017).
Their strange morphology aside, feaellids are also of fundamental importance in understanding the evolution of character systems in pseudoscorpions (Benavides et al. 2019). According to the most recent transcriptomic analysis (Benavides et al. 2019), Feaellidae and its sister-family Pseudogarypidae Chamberlin, 1923 are the sister group to all pseudoscorpions with venom glands in the pedipalp fingers and placed within their own suborder Atoposphyronida Harvey, 2019. The present-day distribution of the family and the fossil record (Harms and Dunlop 2017) suggests that this is a remnant lineage that may have survived in relictual habitats across the Southern Hemisphere.
In this paper, we are describing Feaella (Tetrafeaella) obscura sp. nov. from the Maldives, which are comprised predominantly of young reef-islands and atolls in the Indian Ocean (Kench et al. 2005). The description is based on old museum specimens rather than recently collected material but our species does not match any of the described species and documents the first possible case of oceanic dispersal in this lineage. To diagnose this species accurately and assist the identification of Feaella species across the world, we also provide a key to the species of Feaella (Tetrafeaella). Finally, we also discuss the taxonomic status of Feaella (Tetrafeaella) capensis nana Beier, 1966 from southern Africa and elevate this population to species rank.

Material and methods
The specimens were found during an inventory in the collections of the Hungarian Natural History Museum (HNHM) and stored in 70% ethanol. They were cleared in a 3:1 mixture of lactic acid and gelatine to be examined with a Zeiss Stemi 2000-c stereomicroscope and a Zeiss Axioskop 2 light compound microscope. Drawings and measurements were made with the aid of the Zeiss Axioskop 2 microscope. Measurements were taken using the Olympus Soft Imaging analySIS work 5.0 software. Digital images were taken with a custom-made BK Plus Lab System by Dun, Inc. with integrated Canon EOS 7D Mark II, microscopic lens (5× and 10× magnification) and the Zerene stacker version 1.04 software. Scanned electron images were taken from temporarily dried specimens mounted on copper wire, using a Hitachi TM4000 Plus scanning electron micrograph (SEM). Mensuration follows the reference points in Chamberlin (1931) as does the terminology, except for modifications for the pedipalps and chelal trichobothria (Harvey 1992), the chelicera (Judson 2007), and the faces of the appendages (Harvey et al. 2012).

Superfamily Feaelloidae Ellingsen, 1906 Family Feaellidae Ellingsen, 1906 Subfamily Feaellinae Ellingsen, 1906
Remarks. The genus Feaella is divided into three subgenera: Feaella (Feaella), Feaella (Difeaella), and Feaella (Tetrafeaella) based on the number of frontal lobes on the carapace (Mahnert 1982). Our new species has four lobes and is assigned to Tetrafeaella, hence the diagnosis is against similar species with four protuberances. The species can also be diagnosed against the east Asian genus Cybella Judson, 2017 which lacks pleural sclerites on the abdomen which are obvious in our species. Etymology. The name of the species refers to the obscure evolutionary and geographic origins of the species that is unlikely to have evolved on the young island that is the locus typicus.

Diagnosis.
A typical Feaella (Tetrafeaella) habitus (Figs 2A, B, 3A) that differs from all the other species of the subgenus by the following combination of characters: with the anteromedian lobes of carapace being closer to each other than to the anterolateral ones; anterolateral pair approximately as broad as anteromedian pair, anteromedian lobes longer. Palpal trochanter with prolateral triangular protuberance, femur very robust (1.70-1.76×) with two pronounced triangular process on prolateral corner near base. Chelae are approximately as long as the palpal femora. Chelae with dorsal protuberance at basal third of finger, and a large medial tooth at the base of each fingers. Chelal trichobothriotaxy: esb and est situated on basal third of retrolateral face; ib, isb and ist situated basally; eb and it situated subdistally; dt situated distally; esb midway between est and isb; ist equidistant from est and esb; isb is closer to ib than to ist; on the movable chelal finger t and sb are situated midway between b and st. Pedal coxa I with medioposterior depression (coxal pit), a single primary coxal spine posteriorly; coxa II with 9 secondary spines in males and 8-11 in females. Pleural membrane with a dorsal row of 15 and a ventral row of 14 sclerotised pleural platelets. See detailed differentiation from the most similar species in the Differential diagno-sis, and differentiation from all the other species of the subgenus in the key.
Carapace: anterior margin with four lobes, the anteromedian lobes are closer to each other than to the anterolateral ones; anterolateral pair approximately as broad as anteromedian pair, anteromedian lobes longer (Figs 2D,3E). Four lobes between eyes. Lateral margins of carapace nearly parallel, slightly widened medially; carapace in females 1.27-1.28×, in male 1.24× longer than broad; with two pairs of eyes situated on tubercles away from anterior carapace margin; all eyes with tapetum; with numerous inconspicuous setae; with a pair of postero-lateral processes; shallow anterior and posterior furrows present. Three slit-like lyrifissures at the level of posterior one-fourth of carapace on each side, and two near posterior base. (Figs 2E,3B): pedipalpal coxa with strong basal lateral processes; with numerous small setae and 3 acuminate apical setae. Coxa I with medioposterior depression (coxal pit), a single primary coxal spine posteriorly; coxa II with 9 secondary spines in male and 8-11 in females (Fig. 3C). Pedipalpal coxa somewhat longer as combined length of leg coxae I-IV.

Coxal region
Legs (Fig. 5D, E): claws simple, arolium shorter than claws. Metatarsi and tarsi fused. Each patella with a shallow dorsal depression. Tarsi without tactile setae, vestitural setae short and acuminate. Patellae I and II slightly shorter than femora I and II; patellae III and IV nearly twice as long as femora III and IV.
Abdomen: longer than broad, somewhat ovoid; tergites II-IX and sternites IV-X with distinct median suture lines; tergite XI and sternite XI fused; tergite XII and sternite XII (anal sclerites) strongly scle-rotised; most segments with numerous setae; tergite XII and sternite XII with two setae; anal region with raised circular rim (Fig. 3D). Setae of sternites II and III longer than vestitural setae. Pleural membrane with a dorsal row of 15 and a ventral row of 14 sclerotised pleural platelets.       Feaella (T.) obscura is most similar in general habitus of carapace and pedipalps and in most characters (see in the key below) to the following species from Africa, India and the Seychelles.
It differs from F. (T.) capensis in having a pronounced and triangular dorsomedial process on the prolateral corner of palpal femur near its base (absent in F. (T.) capensis), conical shape of the frontal carapace lobes (long and triangular), pleural membrane with a dorsal row of 15 and ventral row of 14 platelets (15 and 15 in F. (T.) capensis) (Beier 1955).
Feaella (T.) obscura differs from F. (T.) nana in having a pronounced and triangular dorsomedial process on the prolateral corner of the palpal femur near its base (absent in F. (T.) nana). Furthermore, in having pleural membrane with a dorsal row of 15 and a ventral row of 14 sclerotised pleural platelets (in F. (T.) nana the number of these are 14 and 14). In Feaella (T.) obscura on fixed chelal finger isb much closer to ib than to ist and these three trichobothria do not form a straight row, and trichobothrium eb situated clearly closer to the fingertip than to ist (in F. (T.) nana isb, ib and ist are equally placed and form a straight row, and eb is situated midway between ist and the finger tip) (Beier 1966).
The new species differs from F. (T.) indica in having a pronounced and triangular dorsomedial process on the prolateral corner of palpal femur near its base (in F. (T.) indica this process is absent). The chelae of F. (T.) obscura are approximately as long as the palpal femora (in F. (T.) indica these are distinctly shorter than the palpal femora). Furthermore, the palpal femur length/width ratio of F. (T.) obscura is 1.70-1.76× (this value is 1.90× in F. (T.) indica) (Chamberlin 1931;Beier 1932).
In the case of F. (T.) affinis Hirst, 1911 trichobotria sb and st of movable chelal finger are closer to b than t (in the new species sb and st are situated midway between b and t). Furthermore, in F. (T.) affinis trichobothrium ist of fixed chelal finger is at the level of esb and est is placed distal to them (in F. (T.) obscura, ist is equidistant from est and esb) (Mahnert 1978).

Remarks on locality
The specimens were found in soil samples of the Hungarian Natural History Museum and collected by Mr. Győző Horváth who was a friend of Dr. Sándor Mahunka (former director of HNHM, Department of Zoology) and occasionally collected soil samples for the museum. The collecting label clearly points to the Maldives, however, the island name Kabudu is suspected to be a misspelled form of Kaudu [0°17'N, 73°1'E] or Kandudu islands [2°19'N, 72°55'E]. Mr. Horváth passed away some years ago and no further data are available.

Biogeography
The present-day distribution of Feaellidae has been interpreted as a prime example of continental vicariance (Harvey 1989) although this hypothesis is based on dis-tributional data only and requires further testing. One observation that may lend support to this hypothesis is that all species records come from old and stable landmasses (continents or ancient islands such as Madagascar and the Seychelles), with the exception of the Bonaparte Archipelago in north-western Australia which, however, is geologically part of the mainland (Henderson and Johnson 2016). This is a distributional pattern to be expected for poor dispersers with a long evolutionary history.
The present finding is the first record of the Feaellidae from an oceanic island other than the Seychelles which, in contrast to the Maldives, are geologically old and known to support ancient Gondwanan fauna (Gerlach and Marusik 2010). The atolls of the Maldives emerged as recently as 5500-4500 years ago as lagoonal sediments were exposed by shoreline changes (Kench et al. 2005), although volcanic activity in the Maldive Ridge has also shaped this region for as long as 60 million years and many of the atolls have underlying volcanic basalt (Droxler 1992). We cannot exclude the hypotheses that the presence of Tetrafeaella here is a result of vicariance and long-term persistence in an island landscape that changed dynamically over time but dispersal is equally likely, in particular since many islands drowned and re-emerged as sea levels changed. We are not sure where F. (T.) obscura originates but we have checked the new species carefully against congeners from India and Sri Lanka, which are the most promising candidates for a source population. Our new species is clearly different from F. indica but also from several undescribed species from Madagascar. So, if F. (T.) obscura originated elsewhere it must be from a region where the genus is presently undescribed. Dispersal modes are also not clear and passive oceanic drift is as likely as anthropogenic dispersal. The Maldives is a popular tourist destination and timber as well as soil samples are imported regularly from other regions to support the local industry. The evolutionary origins of F. (T.) obscura remain enigmatic until fresh samples are being collected for genetic studies.
The status of Feaella (T.) capensis nana Beier, 1966 While our focus is the description of a new species, the study of relevant literature also reveals inconsistencies in the description of F. (T.) capensis from South Africa. The species was originally described from Cape Point in the Western Cape but Beier (1966) also recorded this species from Kruger National Park in Limpopo Province, almost 1700 km apart. There are several differences in morphology between both populations, including body size (females from Cape Point: 2.6 mm in body length; 1.7 mm from Kruger) and chelal length (females from Cape Point: 0.68 mm; 0.40-0.50 mm from Kruger) but Beier preferred to treat the Kruger populations as a subspecies, F. (T.) capensis nana, next to the nominate form Feaella (T.) capensis capensis. Our perception of range sizes in Feaella has changed drastically since then and recently molec-ular studies have shown that most species have extremely small ranges and low dispersal capacities (Harvey et al. 2016a). We also emphasise the morphological differences such as chelal trichobothriotaxy: on the fixed chelal finger of F. (T.) capensis capensis isb is much closer to ib than to ist and these three trichobothria do not form a straight row, and trichobothrium eb is situated clearly closer to the fingertip than to ist, whereas in F. (T.) capensis nana isb, ib and ist are equally placed and form a straight row, and eb is situated midway between ist and the fingertip. Fur-thermore, the number of dorsal and ventral rows of sclerotised pleural plates of the pleural membrane are 15 + 15 in F. (T.) capensis capensis, and 14 + 14 in F. (T.) capensis nana (Beier 1955(Beier , 1966. Hence, we elevate F. (T.) capensis nana to full species rank, forming the new combination F. (T.) nana Beier, 1966 comb. nov. We also emphasise that F. capensis was recorded from several locations in Zimbabwe (Beier 1964) and, again, these populations are much smaller than the nominate form and likely represent distinct species pending detailed study.
Identification key to the species of the subgenus Feaella (Tetrafeaella) Since Chamberlin's first key to the family (Chamberlin 1931) which included four species only, several keys were published for the African fauna (Beier 1955, Mahnert 1982. To better diagnose F. (T.) obscura from species with a similar morphology, we provide the following iden-tification key which is constructed based on our new data and the literature (Tullgren 1907;Hirst 1911;Chamberlin 1931;Beier 1932Beier , 1947Beier , 1955Beier , 1959Beier , 1966Mahnert 1978Mahnert , 1982Harvey 1989;Harvey et al. 2016a). The proportions are given for both sexes whenever possible.  Pronounced and triangular dorsomedial process on prolateral corner of palpal femur near its base is absent. All four anterior lobes of the carapace long and triangular, the anteromedian ones somewhat longer than the anterolaterals.