The order Palpigradi Thorell, 1888 remains one of the least studied arachnid groups, likely due to their small size (0.7–2.0 mm), the general low abundance of most species, and their strict habitat requirements. Currently, 138 species of palpigrades have been described worldwide (Harvey 2003; World Palpigradi Catalog 2022), and the majority are known from only one or a few specimens. The first species, Eukoenenia mirabilis (Grassi & Calandruccio, 1885), was described from Europe in the late nineteenth century. Since then, 35 species have been recorded from Europe, all belonging to the family Eukoeneniidae Petrunkevitch, 1955. Of these, 34 species are assigned to the genus Eukoenenia Börner, 1901, while a single species belongs to the genus Leptokoenenia Condé, 1965, Leptokoenenia gallii (Christian, 2009).

To date, 11 species of palpigrades have been reported from the Iberian Peninsula and the Balearic Islands (Peyerimhoff 1906, 1908; Condé 1951; Mayoral and Barranco 2002; Barranco and Mayoral 2014; Mayoral and Barranco 2017a, 2017b). Their diagnostic morphological characters suggest that these taxa belong to several distinct phylogenetic lineages, reflecting a complex evolutionary history in the region. Ecologically, Iberian palpigrades were found in a wide range of habitats. Some species are endogean, having been collected under rocks or in leaf litter, whereas others are strictly subterranean and occur exclusively in caves. They also display varying degrees of morphological adaptations associated with life underground. Among these 11 species, four belong to the E. mirabilis-berlesei species group: E. mirabilis (Grassi & Calandruccio, 1885), Eukoenenia amatei Mayoral & Barranco, 2017, E. gadorensis Mayoral & Barranco, 2002, and E. indalica Mayoral & Barranco, 2017. There are only eight species worldwide currently included in the E. mirabilis-berlesei group: E. mirabilis, E. berlesei (Silvestri, 1903), E. ferratilis Souza & Ferreira, 2011, E. tetraplumata Montaño, 2007, E. madeirae Strinati & Condé, 1995, E. gadorensis, E. indalica and E. amatei. Notably, only some of those occurring in the Iberian Peninsula appear to include taxa showing different levels of adaptations to subterranean environments.

In this study, we describe a new species collected in the cave “Sima de Pla de les Llomes” (Serra, Valencia, Spain) and belonging to the E. mirabilis-berlesei group. The new species displays pronounced morphological traits consistent with subterranean adaptation, expanding the morphological and biological diversity known in this group. In addition, we provide a comparative analysis and discussion of all known species of the E. mirabilis-berlesei group, including both cave-adapted and non–cave-adapted taxa, with the aim of identifying which morphological traits are most likely to be modified in response to life underground in this group of arachnids.

The palpigrade was collected using a small brush and stored in a small vial containing 70% ethanol. It was cleared in the laboratory using 50% lactic acid and mounted on slides using Hoyer’s medium. Illustrations were made using a Nikon Optiphot-2 microscope with a lucid camera attached. All measurements are given in micrometers. Setal terminology and abbreviations follow Condé (1994), Christian and Christophoryová (2013), Mayoral and Barranco (2013, 2017a, 2017b), and Souza and Ferreira (2018). The type specimen is deposited in the Scientific Collections Centre of the University of Almeria (CECOUAL). Pictures and fluorescence images were taken using a Leica DM LB2 microscope with a Leica DFC450C camera, using a DAPI filter (350/50 excitation and 460/50 emission).

Morphometric data for statistical analysis for each palpigrade species was compiled from the literature. All specimens data for each character studied were extracted and used from each species description, except for those from the immature stages. The data for E. gadorensis was compiled from Mayoral and Barranco (2002), E. tetraplumata from Montaño (2007), E. ferratilis from Souza and Ferreira (2011), E. amatei from Mayoral and Barranco (2017a), E. indalica from Mayoral and Barranco (2017b), E. madeirae from Strinati and Condé (1995), E. mirabilis from Börner (1901), Condé (1989) and Strinati and Condé (1995), and E. berlesei from Condé (1984) and Mayoral and Hernández-Borroto (2023). Differences between endogean and hypogean groups were assessed using the built-in T.TEST function, applying two-tailed Welch’s t-tests to account for unequal variances and sample sizes. For each morphological character, the mean, median, standard deviation (SD), minimum, maximum, and the 5^th–95^th percentile interval of the observations were calculated separately for the endogenous and hypogean groups. The percentage of elongation for each character was calculated using the group means (% elo/A) and the group medians (% elo/M) of the endogenous and hypogean groups. Percentage elongation was calculated as (hypogean−endogean)/endogean ×100. All statistical analyses were conducted in Microsoft Excel (Microsoft Corp.).

Eukoenenia serrana sp. nov. carries one blade in the lateral organ, seven setae on the basitarsus IV (grt, gla, r, 2 esp and 2 esd), 10 + 10 setae on the propeltidium, and 13 tritosternal setae. The combination of these characters relates this species with those in the E. mirabilis-berlesei group, sensuChristian (2009). There are only eight species in this reduced group, most of them are fairly small and endogean, however in recent years several cave-adapted dwelling species were described in this group (Mayoral and Barranco 2017a; Mayoral and Barranco 2017b). Remarkably, a sexual dimorphism of the post-sternal glandular warts are reported for E. mirabilis and E. berlesei. They exhibit glandular structures and their setae on sternites IV and VI, whereas males lack these structures on sternite IV and retain them only on sternite VI. Condé (1991, 1996) interpreted this character as neotenic in males, because none of the immature stages (A, B, or C) possess glandular setae on sternite IV; this juvenile trait is thus retained in the adult males of the group. In contrast, the number of setae on sternite IV increases throughout development: stage A individuals bear four setae, while stages B and C of E. mirabilis carry six (four in E. berlesei), indicating a progressive addition of setae through successive molts leading to the adult female condition (Condé 1976, 1987b).

The Iberian palpigrades in the E. mirabilis-berlesei group were previously divided into two subgroups (Mayoral and Barranco 2017a). If we apply this classification to the whole E. mirabilis-berlesei group, the first subgroup is characterized by the presence of an external glandular complex on sternites IV and VI in females; the species included are E. mirabilis reported worldwide (Harvey et al. 2006) and E. berlesei reported from Italy, France, Corsica, Algeria, Virgin Islands and Cuba (Rémy 1949; Condé 1951, 1984, 1987a, 1987b; Mayoral and Hernández-Borroto 2023). The second subgroup includes species with an external glandular complex on sternites VI but not on IV: E. madeirae from the island of Madeira, E. gadorensis, and E. indalica from southern Spain. A third subgroup could be made to accommodate species that lack an external glandular complex on the IV or VI sternites: E. ferratilis from a cave in Minas Gerais, Brazil, and E. amatei from southern Spain. Eukoenenia tetraplumata from Mexico does not fit in any of these groups since the female only carries an external glandular complex on sternite IV, and the male only carries it on the sternite VI (Montaño 2007).

Eukoenenia serrana sp. nov. belongs to the third group, it lacks a visible external glandular complex on either of the IV/VI sternites. This relates this new species from caves in the east of Spain more closely to E. ferratilis in Brazil and specially E. amatei in the south of Spain (Souza and Ferreira 2011; Mayoral and Barranco 2017b). Eukoenenia ferratilis is a smaller species, non-cave-adapted and it is easy to differentiate from Eukoenenia serrana sp. nov. by its overall small size; for example, the total length of the body (910–1240 vs 1596), length of the prosoma (250–305 vs 370), length of palp tibia, (93–120 vs 247), length of leg I tibia (90–120) or length of leg IV basitarsus (88–115 vs 256). On the other hand, E. amatei is a bigger species that showed certain traits characteristic of cave adapted species (like elongation of the basitarsus on leg IV). Both species carry two a setae (a₁, a₂), and two s slender setae (s₁, s₂) on each side of the sternites IV–VI, therefore their ventral chaetotaxy is identical. Eukoenenia serrana sp. nov. can be differentiated from E. amatei in the morphology of the d setae of the chelicerae: d₁ (truncated vs slender), d₂ (truncated in both species) and d₃ (slender vs truncated), as well as the length of the t setae on the metapeltidium t₃ (139) > t₂ (109) > t₁ (17) vs t₂ (98) > t₃ (65) > t₁ (25). The leg coxae chaetotaxy has been used to separate species and groups of species in palpigrades (Christian and Christophoryová 2013), and it also shows differences between these two species. In particular, the number of thickened setae, presumably excretory, in coxae I–IV (6–6–4 vs 3–3–2). Eukoenenia serrana sp. nov. appears to have a larger body size and with proportionally longer palps and legs than E. amatei. These differences are evident in total body length (1596 vs 1176), prosoma length (370 vs 330), palp length (1359 vs 1002), total length of leg I (2090 vs 1614) and the total length of leg IV (1541 vs 1207).

Eukoenenia serrana sp. nov. is the twelfth palpigrade species reported from the Iberian Peninsula, and the fourth in the E. mirabilis-berlesei group displaying troglomorphic characters.

Morphological differentiation between endogean (soil species) and hypogean (cave–dwelling/cavernicolous species) palpigrades was not systematically addressed in palpigrades, and most of the palpigrade reports are limited to taxonomic descriptions. Condé (1996) did the only formal comparison available until now with 9 species (4 endogeans and 5 cavernicolous) using 9 characters, for which many species were missing data (Condé 1996, Table 1). This table has been the reference point to determine whether palpigrades fall in one group or the other; this analysis included the total length of the body (L), the length of the prosomal shield (B), the length of the flagellum (fl), the length of the basitarsus leg IV (IVbta), the number of the blades in the lateral organ (O.lat), and the ratios bta/ti, B/bta IV and L/fl. Condé (1996) noted a “syndrome of cave adaptation” in Eukoenenia where there is an increase in the overall body length, prosomal shield, basitarsus IV and flagellum, as well as an increment in the number of elements in the lateral organs. It is remarkable the clear separation of the two groups of palpigrades regarding their average IVbta length, being 66 μm for the endogenous and 245 μm for the cave adapted; similarly, the most derived cavernicolous species may carry 8–13 elements in the lateral organ on each side, whereas endogeans generally have 1–3, occasionally 4–6, and interpreted as preadaptation to hypogean habitats (Condé 1996). In addition, endogean species show nearly double B/btaIV ratio of cavernicolous forms (3.51 vs 1.69), reflecting that the basitarsus IV of troglobitic species becomes disproportionately elongated relative to the prosomal shield. As a result, greater elongation of basitarsus IV in subterranean taxa drives the B/btaIV ratio downward, providing a clear morphological distinction between endogean and cavernicolous species.

Mammola et al. (2021) conducted an alternative analysis based on environmental variables to determine how they affect the distribution and morphological adaptations of subterranean and cave-dwelling palpigrades. They found that i) climate, and ii) nutrient disposition and habitat have a significant influence on the distribution of palpigrades adapted to caves and soil.

Comparative morphometric analyses of the E. mirabilis-berlesei group revealed a pronounced trend of larger body size and elongated appendages in hypogean species compared with their endogean counterparts. Percentage-based comparisons indicate that elongation is particularly marked in appendicular measurements, including basitarsus IV length, tibial segments, and overall length of leg I. These differences are consistently observed when calculated using both group means and medians, suggesting that the pattern is robust to distributional asymmetries. Our results suggest that the appendage elongation is strongest in leg I, followed by leg IV, and weakest in the palp; the tarsus–tibia and tibia measurements show similar results. Leg I serves as the dominant tactile appendage during locomotion and environmental surveying, whereas leg IV seems to contribute more to support and walking. By contrast, characters describing the prosomal shield or ratios expressing its relative size tend to show reduced values in hypogean species, reflecting a shift toward relatively longer appendages rather than uniform body enlargement. An increase in the number of blades has been interpreted as a hallmark of advanced cave adaptation in palpigrades, the persistence of a single element in E. serrana sp. nov. or E. indalica indicates that there may be exceptions and that an increase in the number of blades in the lateral sensory complex is not a pre- or corequisite for appendage elongation. This differential evolution of characters implies that different troglomorphic characters may be subject to distinct functional constraints, pressures and/or evolutionary pathways.