Samples were collected from two localities in the municipality of Jarabacoa (Dominican Republic), in the island of Hispaniola. Jarabacoa is located in La Vega province, in a valley of the Cordillera Central (central ranges) with a mean altitude of 530 m a.s.l. The area is characterized by a siliceous substrate, and wet tropical climate, with an average annual temperature of 20 °C and 1723 mm of mean annual precipitation (Climate-Data.org 2023). The Cordillera Central is included in a single biogeographical area, the Central‐Eastern district, which has one of the highest richness of plant genera and endemic species of Hispaniola (Cano-Ortiz et al. 2017).

Ostracod samples were collected in the frame of a wider survey and various projects on the aquatic invertebrate biodiversity of Hispaniola, which sampled varied habitats, focusing particularly on potential predators of mosquito larvae (Rodríguez Sosa et al. 2019; Olmo et al. 2024). Invertebrate samples were collected by suction from the water stored in between the base of bromeliad leaves, using either a plastic Pasteur pipette, or a 60 mL syringe coupled to a 40 cm flexible hose following Júnior et al. (2017). Most of the Bromeliaceae plants were located in private gardens or nearby, and were tentatively determined as belonging to the genus Neoregelia. In the laboratory, the samples were filtered through a 350 μm mesh size filter and fixed in 70% ethanol.

The dissection of ostracod specimens for optical microscopy inspection was done following the protocol described in Namiotko et al. (2011). Soft parts were embedded in HydroMatrix^® for the preparation of permanent slides. Shells were stored dry in micropaleontological slides. Drawings were done using a camera lucida on a Leica microscope. Some pictures were taken using a Nikon^® Eclipse E800 epifluorescence microscope, either with white light or with UV light (340–380 nm) plus a blue filter (435–485 nm). Some individuals were critical-point dried in toto or without the valves. These individuals, plus separated valves of other individuals, were coated with a thin layer of Au-Pd for SEM observation in a Hitachi S-4800 or a SCIOS-2 at the University of València.

In this work we follow Mesquita-Joanes et al. (2024) in accepting the suggestion of Tanaka et al. (2021) to raise the subfamily Timiriaseviinae to the family rank, and provide a diagnosis of the family. This diagnosis is established after the differences indicated by Martens (1995) and Danielopol et al. (2018) between Timiriaseviinae and Limnocytherinae. However, most ostracodologists have traditionally considered the Timiriaseviinae as a subfamily within the Limnocytheridae, ever since the review by Colin and Danielopol (1978) (e.g., Savatenalinton et al. 2008; Karanovic and Humphreys 2014; Danielopol et al. 2018; Meisch et al. 2019).

The selection of critical characters to build the identification key was based on those used by Pinto and Jocqué (2013), plus those stressed by Danielopol et al. (2014), supported in some cases by some of the characters included in the phylogenetic tree of Pereira et al. (2022).

Abbreviations used in the text and figures include the following:

Cp carapace;

CL carapace length;

H height of valves;

L length of valves;

LV left valve;

RV right valve;

W width of shell;

A1 antennula;

A2 antenna;

Md mandibula;

Md-palp mandibular palp;

Mx maxillula;

T1 first thoracopod;

T2 second thoracopod;

T3 third thoracopod;

CR caudal ramus;

Hp hemipenis;

DL distal lobe;

CoP copulatory process;

LR lower ramus.

Chaetotaxy nomenclature follows mainly Broodbakker and Danielopol (1982), Martens (1987), Meisch (2000), and Pereira et al. (2023). We follow mostly Sames (2011a, 2011b) and Danielopol et al. (2014) for carapace traits terminology. However, terms used for the description of the general shape in dorsal or ventral view of the carapace follow those commonly used for leaves summarized by Hickey (1973), by applying those terms for the tip of leaves to the shape of the anterior part of the carapace, and those for the base of leaves to the shape of the posterior part of the carapace. Note that these terms differ in some cases from those used in the ostracod literature, and in Elpidium descriptions in particular, but are more widely used in general in Biology for morphological descriptions.

Ethanol-fixed ostracods were individually transferred to 1.5 mL microtubes using a thin brush. Single specimens from the type locality (e.g., P459=MUVHNZY0040) were digested at 55 °C overnight using 180 µL T1 buffer and 20 µL proteinase K, and DNA was extracted with the Nucleospin DNA extraction kit (Macherey-Nagel™) following the manufacturer’s instructions. The large ribosomal subunit (18S) gene region was amplified using primers 18S_5F 5’-GCG AAA GCA TTT GCC AAG AA-3’ and 18S_9R 5’-GAT CCT TCC GCA GGT TCA CCT AC-3’ (Carranza et al. 1996). Amplifications were carried out using ~ 10 ng of genomic DNA in a reaction containing 1 U of Taq polymerase (Amersham), 1× buffer (Amersham), 0.2 mM of each primer and 0.12 mM dNTPs. The polymerase chain reaction (PCR) thermal profile included an initial denaturation step at 94 °C for 4 min, followed by 30 cycles of 94 °C for 30 s, 50 °C for 30 s, 72 °C for 30 s, and a final extension at 72 °C for 20 min. Sequences were obtained using the Big-Dye Ready Reaction kit v. 3.1 (Applied Biosystems) on an ABI Prism 3770 automated sequencer at the MACROGEN sequencing facilities. Chromatograms for each DNA sequence were checked with BioEdit v. 7.2.5 (Hall 1999) and sequence alignment was conducted with Muscle v. 3.6 (Edgar 2004). Model selection was carried out for the sequence alignment using the Bayesian Information Criterion (BIC) as implemented in ModelTest-NG v. 0.1.7 (Darriba et al. 2020). Maximum likelihood phylogenetic reconstruction was then completed with the corresponding DNA substitution model with ultrafast bootstrap (1000 replicates) as implemented in IQ-TREE v. 2.0 (Minh et al. 2020).

Class Ostracoda Latreille, 1802

Subclass Podocopa G.O. Sars, 1866

Order Podocopida G.O. Sars, 1866

Suborder Cytherocopina Baird, 1850

Superfamily Cytheroidea Baird, 1850

We have obtained new 18S rDNA sequences for Elpidium alarconi sp. nov. and Cyprideis torosa, with GenBank accession numbers PP648174 and PP648175, respectively. The 18S rDNA sequence alignment had 739 bp in length and followed the GTR substitution model according to BIC model selection. The phylogenetic tree obtained (Fig. 8) placed Elpidium alarconi sp. nov. in the same clade as Metacypris, Cytheridella, and Gomphodella, all of them belonging to the Timiriaseviidae (formerly subfamily Timiriaseviinae). This clade becomes clearly separated from the genus Limnocythere, and therefore the Limnocytheridae s.s. Interestingly, the clade formed by the Timiriaseviidae genera, holds a more basal position within the Cytheroidea, splitting earlier than Limnocytheridae, but also than other families, including Xestoleberididae, Loxoconchidae, Cytheridae and Cytherideidae, among others.

Elpidium alarconi sp. nov. has a shell morphology that does not differ widely from those of other Elpidium species with symmetric smooth valves, shell closure with left valve embracing right valve, and truncate posterior margin in dorsal view, such as E. higutiae, E. purium, E. litoreum, E. pintoi or even the type species E. bromeliarum. Yet, some of these species are either larger, as E. bromeliarum, or more elongated (E. higutiae, E. litoreum). The remaining two species, E. purium and E. pintoi, are very similar in dorsal view and their carapace sizes overlap with that of E. alarconi sp. nov. However, both lack a basal digitiform expansion in the distal lobe of the hemipenis, which is present, and very long, in the new species. This relatively straightforward distinction between species could be established thanks to a previous review of the variability of morphological traits in the genus Elpidium by Danielopol et al. (2014). These authors highlighted the importance of valve surface (smooth or ornamented), shell size and closure (left or right valve overlapping the other one), and its outline in dorsal view, including symmetry or asymmetry of valves, shape of posterior margin (pointed, rounded, truncate, invaginated), and length/width relationship. These traits are very useful for morphological characterization of Elpidium species, and therefore for identification keys, so we also used them in the new key provided, which now includes 20 described species. However, besides the indication of a pointed shape, we used the more precise term “obtuse” for an angle > 90°, and “acute” for an angle < 90°, and rather than “invaginated”, we used the term “cordate”, following Hickey (1973). We call for a more general use of this terminology, well established in the literature for leaf shape, but which can be also applied to ostracod shape in dorsal or ventral view.

In some cases, carapace morphology alone is not enough to easily distinguish between similar species, and other characters may be needed. Indeed, the most diversified morphological trait in Elpidium ostracods is the shape of the hemipenis and, in particular, that of its distal lobe, copulatory process and lower ramus (Danielopol 1975; Danielopol et al. 2014; Pereira et al. 2022). Hemipenis morphology has long been considered an essential character in ostracod phylogeny, allowing species determination in lineages with similar shell structure (Danielopol 1969; Hart and Hart 1974; Bisquert-Ribes et al. 2023), and this seems to be also the case in the genus Elpidium. Besides E. alarconi sp. nov., there are other species that also have a basal digitiform expansion in the distal lobe of the hemipenis: E. chacoense, E. cordiforme, E. picinguabaense, E. merendonense, E. maricaoense, and E. higutiae. But out of these, this expansion is as long or longer than the basal seta of the distal lobe only in E. alarconi sp. nov., E. cordiforme, and E. merendonense. It is nevertheless distinctly shorter in E. merendonense than in the other two species, and this species is furthermore distinguished because of an asymmetric carapace shape in dorsal view, and a lower ramus of the hemipenis with a blunt tip. Despite the similarity of the digital expansion of E. cordiforme with that of the new species, its distal lobe has a blunt tip (pointed in the new species). In addition, E. cordiforme has a cordate posterior margin in dorsal view (hence its name), whereas E. alarconi sp. nov. has a truncate posterior margin, although a slight invagination (i.e., quasi-cordate shape) can be appreciated in some shells. Together with the distal lobe shape, the morphology of the lower ramus is also remarkable in the new species, as it is thinner than in any other member of the genus, and L-shaped, somehow resembling a piolet or a very thin sock with an acuminate tip. The lower rami are also pointed and even almost L-shaped in other species of Elpidium, but always thicker at the basal part, as for instance in E. higutiae, E. maricaoense, E. oxumae, or E. cordiforme. Taking these hemipenis characters into account, E. cordiforme is one of the species closer to E. alarconi sp. nov., although the former has a twisted copulatory process, unlike any other species of the genus. Furthermore, the distal lobe of E. merendonense and the lower ramus of E. maricaoense are the most similar hemipenis structures to those of E. alarconi sp. nov.

Another interesting morphological trait apparently differing between species of the genus Elpidium, according to the literature, is the strength of the separation between segments 4a and 4b of the antennula. Most species have these segments only partially or weakly separated, as described for E. maricaoense, E. littlei, E. wolfi, E. litoreum, E. cordiforme, E. laesslei, E. merendonense, E. heberti, E. oxumae, E. picinguabaense, E. eriocaularum, and E. higutiae (Tressler 1941; Pinto and Jocqué 2013; Pereira et al. 2019, 2022, 2023), while others, including E. bromeliarum, E. martensi, and E. purium are described as having a single, undivided, fourth segment (Pinto and Purper 1970; Danielopol et al. 2014; Pereira et al. 2023). Consequently, most authors considered a five-segmented antennula as a diagnostic character of the genus (Pinto and Jocqué 2013; Danielopol et al. 2014; Pereira et al. 2019). However, E. alarconi sp. nov. shows a distinctly clear separation between segments 4a and 4b in most specimens (weaker in others) under standard microscopic observation in transmitted light, so that its antennula appears as having six segments, rather than five. Also six segments are apparent in the graphic description of E. heberti; although the authors indicate that the fourth segment is “partially subdivided” when describing the species in the text, it is drawn as divided with a continuous line in their figure (Pereira et al. 2019: fig. 9a), while other species described in the same publication show a dashed line. In the original description of the type species, and in a subsequent revision and establishment of neotypes, Müller (1881) and Pinto and Purper (1970) stated that the antennula usually has five segments, but that it can exceptionally have six. Pinto and Jocqué (2013) suggested this separation might not be fully functional. Later, Pereira et al. (2022), when performing a phylogenetic analysis of the genus using a list of coded characters (detailed in the Supplementary information of their publication), characterized all species for which they found information on this trait, as having a partially fused fourth segment of the antennula. They concluded, after re-examining preparations of most species, that this morphological feature was shared for all Elpidium species analyzed, and considered that, according to microscopic observations, the segmentation was most probably only occurring on one side of the segment, but not in the other (Pereira, pers. comm.). We tested this possibility in the case of E. alarconi sp. nov., and could confirm it; even if the separation was quite clear under standard transmitted light, the use of fluorescence and scanning microscopy confirmed that it was partial, and only present in the inner side of the segment for each antennula. It remains to be confirmed whether this feature is shared with all other species of the genus, and which is its functional and evolutionary significance.

The species described in this work represents the first member of the genus Elpidium identified to species level for the island of Hispaniola; it must be noticed that Acosta-Mercado et al. (2012) previously recorded two undetermined species collected from liverworts. Its presence in this island does not come as a surprise, considering that several Elpidium species had been found in the neighboring islands of Cuba, Jamaica, and Puerto Rico, in addition to those found in the mainland (Fig. 9). At present, Jamaica can be considered the area with the highest density of Elpidium species worldwide, most of them endemic to the island (Little and Hebert 1996; Danielopol et al. 2014; Pereira et al. 2019). The high diversity and endemicity of the genus Elpidium was previously noted by Little and Hebert (1996), based on allozyme and mitochondrial data and on partial description of hemipenis morphologies, although they did not formally describe any species. They also highlighted the role of isolation and restricted dispersal in phytotelma ostracods for speciation, although more recent works have shown how they may disperse via phoresis using mostly amphibians and snakes (Lopez et al. 1999, 2005; Sabagh and Rocha 2014; Cunha et al. 2023). The high diversity and endemicity of Elpidium has been further supported by studies in Brazil and Argentina during the past few years (Pereira et al. 2022, 2023; Díaz et al. 2024), and corroborated by the present survey. It seems therefore that the species diversity of the genus Elpidium may be much higher than previously expected. The small number of samples collected hitherto from phytotelmata in tropical countries probably caused that only 20 species of Elpidium are known to date, but we expect many more to be discovered in the future, considering the large areas in the Neotropics that have remained unexplored for this habitat (Jocque et al. 2013; Pereira et al. 2022) (Fig. 9).

The high endemicity of Elpidium species is challenged by the wide distribution of the type species E. bromeliarum, recorded from Southern Brazil to Central America and Jamaica (Fig. 9). However, we must be cautious, as probably most of the records out of Brazil might be erroneous (Pereira et al. 2023). Indeed, even if some authors cited E. bromeliarum from Costa Rica (Pinto and Jocqué 2013; Pereira et al. 2023) based on early work by Picado (1913), this author did not confirm that the Elpidium species he found was E. bromeliarum, but a similar species: “Metacypris (Elpidium) sp. (fig. 42, B). La Mica, 1500 mètres. Ce crustacé est, d’après, G. W. Müller une espèce très voisine d’Elpidium bromeliarum. Quand le Crustacé est vivant, il présente cependant une pigmentation différente de celle de l’espèce décrite par Fritz Müller...” (Picado 1913: 336). [Transl: “Metacypris (Elpidium) sp. (fig. 42, B). La Mica, 1500 metres. This crustacean is, according to G. W. Müller, a species closely related to Elpidium bromeliarum. When the crustacean is alive, however, it presents a different pigmentation from that of the species described by Fritz Müller...”]. Picado (1913) included Elpidium bromeliarum in his list of “Animaux bromelicoles actuellement connus”, but he specifically wrote that this species lived in Brazilian epiphytic bromeliads, not in Costa Rica. Later on, it was Tressler (1956) who recorded E. bromeliarum from Jamaica (Fig. 9), although he did not discuss or show diagnostic characters of the hemipenis, so it may be considered an unreliable record (Pereira et al. 2023). The presence of E. bromeliarum in Guatemala (Pérez et al. 2012) should also be considered doubtful, because the authors only provided valve pictures, and it would be necessary to check the morphology of the copulatory apparatus to confirm this determination. Furthermore, Pinto and Jocqué (2013) described E. merendonense one year later from Honduras; it would therefore be interesting to check whether or not the species determined as E. bromeliarum from Guatemala may actually belong to a different species, perhaps E. merendonense. Finally, E. bromeliarum has also been recorded from French Guiana (GBIF.Org 2023), but we could not find further information on morphological aspects of this record, which is quite far from other geographic locations of the species, so we consider it should be taken with caution. Actually, the confusion on the identification and distribution of E. bromeliarum can be traced back to its discovery; in his description of male copulatory organs, Müller (1881) included at least three different morphologies of the hemipenis distal lobe, suggesting it was very variable. However, these different morphotypes most probably belong to different species of Elpidium. This confusion was continued in the review of Pinto and Purper (1970), as they also showed some clearly different hemipenes as belonging to the same species, although they may actually correspond to different ones (Pereira et al. 2017, 2023).

Another potential issue for understanding the biogeography of Elpidium is the presence of E. maricaoense in Florida (Tressler 1956). Even though this record was noted by the same author who described the species earlier from Puerto Rico (Tressler 1941), and considering the high diversity of species in the Caribbean and the lack of morphological information for the Florida specimens, the presence of E. maricaoense in mainland America needs to be corroborated by further sampling in Florida. In addition, undetermined species of Elpidium have been recorded from other locations besides Costa Rica (Picado 1913), including Brazil, Florida and Mexico (Mercado-Salas et al. 2021; GBIF.Org 2023), so we would expect the genus to be widespread in the Neotropical region, and many more species to be described in the future. Consequently, the early suggestion by Müller (1881) that E. bromeliarum should be widely distributed in Brazil, is not corroborated by recent data, although it has been shown that the genus Elpidium has probably colonised most of the Neotropical region.

The new finding of E. alarconi sp. nov. in Hispaniola should initially be considered as an endemism for the island. However, considering that it was collected from bromeliads in managed gardens or nearby secondary forests, it would not be surprising that future research may record it in other regions, considering also its morphological proximity to several mainland species, and the worldwide proliferation of exotic ostracods driven by human movements (McKenzie and Moroni 1986; Valls et al. 2014). This might be one of the reasons for the lack of congruence between the geographic distribution of Elpidium species and their phylogenetic relationship using morphological data (Pereira et al. 2022). These authors only found a clear relationship between a clade of Elpidium species and their restricted distribution in Jamaica. They suggested that the lack of a phylogeographic pattern for most of the species relies on the scarcity of studies and/or the lack of critical morphological information for some species described long ago. We agree that these are the main reasons for the unresolved Elpidium biogeography, although we would not discard human-mediated movement of Elpidium species through bromeliad trade for gardening, as shown for other ostracod species in relation to the trade of aquatic plants for cultivation, gardening or aquaculture (McKenzie and Moroni 1986; Matzke-Karasz et al. 2014; Valls et al. 2014; Smith et al. 2024). Pereira et al. (2023) proposed using the genus Elpidium as a model group to study biogeographic areas of endemism, but considering the issue of expanding exotic ostracods, this kind of studies should be focused on sampling bromeliads mostly in undisturbed environments, far from human-impacted sites.

Our molecular phylogeny analysis placed E. alarconi in the same clade as Metacypris, Cytheridella and Gomphodella, and far from the branch where Limnocythere was positioned in the phylogenetic tree. Assuming that the Limnocythere specimen whose DNA sequence is deposited in the repository has been accurately identified, and that it is representative of the Limnocytherinae, these results provide further support for the suggestion that the former subfamilies Timiriaseviinae and Limnocytherinae should be promoted to family level (Tanaka et al. 2021). The Timiriaseviinae subfamily was established to accommodate a fossil species of the genus Timiriasevia by Mandelstam (in Kashevarova et al. 1960). Shortly after, the subfamily Metacyprinae was established by Danielopol (1965), initially as a tribe (Metacyprini) of the subfamily Limnocytherinae, to include the genera Metacypris, Elpidium, Afrocythere and Cordocythere. Later on, the tribe Metacyprini was promoted to subfamily, and considered a junior synonym of the Timiriaseviinae (Colin and Danielopol 1978; Danielopol et al. 2018).

Although most recent authors consider the Timiriaseviinae a subfamily included in the Limnocytheridae Sars, 1928 together with the subfamily Limnocytherinae, after considering their major differences in shell and soft parts anatomy (Martens 1995; Danielopol et al. 2018) and the long genetic distance between them (Tanaka et al. 2021), we decided to accept the proposal of these last authors to promote the Timiriaseviinae to family Timiriaseviidae, as also did earlier Mesquita-Joanes et al. (2024), although the molecular basis for this promotion needs to be further tested with more sequences of species belonging to the Limnocytheridae s.s. Nevertheless, despite some morphological similarities between both families, which might be considered large enough as to hamper the proposed change of taxonomical levels suggested by Tanaka et al. (2021) and adopted here, we consider that the differences between them are even stronger, supporting a separation in two distinct families. Regarding similarities, there are three characters that are shared between both groups (Danielopol et al. 2018), but which can be considered relatively weak or even plesiomorphic, and therefore not well founded for their use in sustaining their monophyly: i) the distal antennular aesthetasc, fused with a distal seta, shows a much longer fused zone in the Limnocytheridae s.s. than in the Timiriaseviidae, in which this fusion is very short (as in Gomphodella or Gomphocythere) or even not observed (in Elpidium, Intrepidocythere, or Metacypris); ii) the presence of three claws in the last segment of the antenna may be regarded as an important trait, but it might be considered plesiomorphic, as it appears also in the primitive Bythocytheridae, and Entocytheridae; and iii) the presence of a minute seta in the last podomere of thoracopods, this segment fused with the final claw, may be a remnant of the posterior seta that some other Cytherocopina hold in the last segment of thoracopods (when it is not fused with the claw). For instance, it is the only posterior seta present in the thoracopod endopods of some Cytherocopina (e.g., in Terrestricythere, Bythocypris, Bairdoppilata) or in Darwinuloidea (e.g., in Vestalenula). It is interesting to notice how the first thoracopod of adult males of Terrestricythere hold a small posterior seta in their modified claw, probably resulting from the fusion of the last segment with the claw, as the male endopod has only two segments, while there are three in the female (Horne et al. 2004). Furthermore, we can see a similar shape of a fused segment-claw with a tiny seta in Amnicythere prespensis (in Petkovski and Keyser 1992), a species in the family Leptocytheridae, therefore also outside the Limnocytheridae s.l. Conversely, there are some species of Limnocytheridae s.l. for which that minute seta has not been observed or illustrated, as it occurs in several Limnocythere (Martens, 1990) or in Intrepidocythere (Pinto et al. 2008), although it may have been missed by the authors when illustrating them. Consequently, it does not seem appropriate to keep Timiriaseviinae and Limnocytherinae together in the same family on the basis of such a loose character of a minute seta, considering that it is not present in all species, and that it is also present in other species outside the family Limnocytheridae s.l., suggesting it is a plesiomorphic trait.

Regarding morphological differences between Limnocytheridae s.s. and Timiriaseviidae, we consider these are more consistent and strong enough as to support their separation as two distinct families: (i) unlike the Timiriaseviidae, females of the Limnocytheridae s.s. do not have a brooding chamber in their valves. This is an important morphological trait, related to reproduction and readily observed in the female carapace of most Timiriaseviidae; (ii) another very important trait, in our view, is the segmentation of the maxillular palp. It has only one segment in the Timiriaseviidae (Elpidium, Cytheridella, Intrepidocythere, Metacypris, Gomphodella, Gomphocythere) but two in the Limnocytheridae s.s. (e.g., in Limnocythere, Korannacythere and Leucocythere); (iii) still another important trait differing between the two families is the ventral seta on the second antennular segment, which is situated in a medial or proximal position in the Timiriaseviidae, but in the distal margin in the Limnocytheridae; in addition, (iv) the antennula is five-segmented in the Limnocytheridae, but in the Timiriaseviidae it can be five-segmented (as in Cytheridella), six-segmented (as in Metacypris and Gomphocythere) or with a partial segmentation of the 4^th segment, i.e. apparently six-segmented but not completely (as in Elpidium, Gomphodella or Intrepidocythere) and (v) the distal lobe of the hemipenis is articulated in the Timiriaseviidae, but not in the Limnocytheridae. This can be considered an important trait as well, because of its potential functional role in reproduction. Furthermore, (vi) a recent review of the sieve-type pore canals (StPC) in the Limnocytheridae s.l. by Danielopol et al. (2018) concluded that these pore canals, when present, have a seta inside them in the Limnocytheridae s.s. (type C StPC) but not in the Timiriaseviidae (type B StPC). Finally, if this taxonomic scheme with two separate families (Limnocytheridae s.s. and Timiriaseviidae) is accepted, a derived conclusion should be to promote their constitutive tribes to subfamilies: Timiriaseviini Mandelstam, 1960, Cytheridellini Danielopol & Martens, 1989 and Gomphodellini Danielopol et al., 2018 would therefore change to Timiriaseviinae Mandelstam, 1960, Cytheridellinae Danielopol & Martens, 1989 and Gomphodellinae Danielopol et al., 2018, all belonging to the family Timiriaseviidae; and the Limnocytheridae s.s. would be composed by the subfamilies Leucocytherinae Danielopol & Martens, 1989 and Limnocytherinae Klie, 1938 (previously as tribes Leucocytherini Danielopol & Martens, 1989 and Limnocytherini Klie, 1938, belonging to the subfamily Limnocytherinae Sars, 1928).

Within the Timiriaseviidae, previous phylogenies using morphological traits, positioned the genus Elpidium either alone in a branch separated from another that included Gomphodella, Metacypris and Cytheridella (Karanovic 2009) or together with Metacypris in a branch separated from Gomphodella or Cytheridella (Karanovic and Humphreys 2014). In contrast, our 18S phylogenetic tree suggests Elpidium might be closer to Gomphodella than to Metacypris or Cytheridella. Such different pattern has consequences for the interpretation of the biogeographic origin of Elpidium; as Gomphodella is exclusive to the Australian region, the phylogenetic association between these two genera suggests an ancient vicariant origin from the breakage of Gondwana, when Australia became separated from Antarctica and South America, similar to the findings of Sigvardt et al. (2021) for Lynceus (Laevicaudata). Therefore, our findings do not support the alternative process of a dispersal event from Eurasia or Africa to South America, as previously proposed by Karanovic (2009) in relation to the morphological similarities between Elpidium and Metacypris, and to the rich fossil record of the latter. Yet some morphological traits point to other relationships. For instance Elpidium lack StPC, while Gomphodella or Cytheridella present this type of pores on its valves (Danielopol et al. 2018). Another interesting trait is the row of posteroventral type-A2 pores (with rim and seta) on the peripheral marginal infold of valves of Elpidium. A similar row of pores is observed in Cytheridella (Danielopol et al. 2023: fig. 11) and in Intrepidocythere ibipora (Pinto et al. 2008), although in a more external position in this case. Still another particular trait shared by Elpidium, Intrepidocythere and Cytheridella is the presence of a seta at the base of the articulated distal lobe of hemipenes, not described in other species of the family. The inconsistencies between morphological and molecular phylogenetic relationships inside the family calls for further molecular analysis of other genera of Timiriaseviidae, and a more detailed morphological work, which combined would help understanding the phylogeny and early biogeography of this interesting family of non-marine ostracods.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This research was partially supported by Ministerio de Economía, Industria y Competitividad (Spanish Government), through project CRUSTRESS (code PID2020-112959GB-I00) awarded to FP and FMJ. FP acknowledges the project “CIDEGENT/2019/028 - BIOdiversity PAtterns of Crustacea from Karstic Systems (BIOPACKS): molecular, morphological, and functional adaptations” funded by the Conselleria d’Innovació, Universitats, Ciència i Societat Digital. This work was also partially funded by the project: “Búsqueda, caracterización y evaluación de agentes ecológicamente amigables para el control de mosquitos (Diptera: Culicidae) de importancia médica en República Dominicana”, supported by the Fondo Nacional de Innovación y Desarrollo Científico y Tecnológico (FONDOCyT), Ministerio de Educación Superior, Ciencia y Tecnología (MESCyT) of the Dominican Republic (Project No. 2018–19–2B2–043) awarded to Pedro María Alarcón Elbal and JR. Pedro Alarcón and María Altagracia Rodríguez Sosa are thanked for their collaboration in sampling and project coordination.

Conceptualization: FMJ, JR. Data curation: FP, FMJ, JR. Formal analysis: FP, FMJ. Funding acquisition: FP, FMJ, JR. Investigation: FP, ÁG, JR, FMJ. Methodology: ÁG, FP, FMJ, JR. Project administration: FP, JR. Resources: JR, FMJ, FP. Supervision: FMJ. Validation: JR, FMJ. Visualization: FP, ÁG, FMJ. Writing - original draft: FMJ. Writing - review and editing: FP, ÁG, JR, FMJ.

Francesc Mesquita-Joanes https://orcid.org/0000-0001-7168-1980

Ángel Gálvez https://orcid.org/0000-0002-5562-7316

Ferran Palero https://orcid.org/0000-0002-0343-8329

Juan Rueda https://orcid.org/0000-0002-7629-8881

All of the data that support the findings of this study are available in the main text.