Diversity of the squid genus Leachia (Oegopsida: Cranchiidae) in the Pacific Ocean

Leachia is a cosmopolitan tropical/temperate genus of ‘glass’ squids (Cranchiidae) whose taxonomy has been plagued with instability typical of the family. Eight species are currently believed to exist worldwide, including some not yet described. This review assesses the Pacific taxa, describes a novel species, L. separata, and provides updated diagnoses for the other Pacific species. Type material and original descriptions are compared, and brief remarks provided on the Atlantic taxa. The most useful characters for identifying Leachia species appear to be the number of ocular photophores and the configuration of the ventral cartilaginous strips and associated tubercles on the mantle, with arm and sucker-ring dentition also proving useful in some taxa. Further work on the genus is needed, ideally including molecular tools.


Introduction
Cranchiids ('glass' squids) are abundant and diverse oegopsids, with representatives found in all oceans except apparently the Arctic. Many species undergo ontogenic descent (Young 1978;Evans 2018), occupying different ecological niches across their growth and development, and being preyed on by fishes, seabirds, and whales (Evans 2013). The appearance of most taxa also changes considerably across the lifespan (as shown in Evans and Bolstad 2014 for the genus Teuthowenia), and specimens are often not identified beyond genus or even family. Systematic instability, fragile morphological characters, and specimen rarity all further complicate identification and taxonomic resolution, and several undescribed species are still believed to exist (Voss 1980). Due to the scale of the family (60 + described species [Evans 2018]), its geographic extent, and the factors outlined above, recent advances in our understanding have focused on subsets of the group and/or specific regions, such as a recent review of the Pacific cranchiid taxa (Evans 2018).
Within the Cranchiidae, untangling members of the genus Leachia has provided some of the greatest taxonomic challenges. Steps toward resolving this group have been taken by a succession of authors working both globally and on local scales. The genus Leachia was first erected by Lesueur (1821) for a specimen collected from 37°S, 33°E (southeast of South Africa, although Lesueur stated that it "inhabits the Pacific Ocean"). He based the description of his new species, L. cyclura, on an illustration by Mr. Nicolas-Martin Petit (later published in Grant 1833), but did not examine the specimen himself. Lesueur placed Leachia into a new family, Loligoidea, which he proposed to separate "the Loligos" (= squids) from cuttlefish (Sepia); the species he included in Loligoidea would later be recognized as members of the oegopsid families Ommastrephidae, Onychoteuthidae, Cranchiidae, and the myopsid family Loliginidae. He characterised the genus Leachia as having an elongated cylindrical body, transparent tissue, and third arms that were longer and more robust than the other arm pairs. Some authors rejected the generic status of Leachia (e.g. d'Blainville 1823) due to its description being based on imperfect illustrations, but Grant (1833) believed it to represent the same taxon as Loligopsis Lamarck 1812 (another genus described solely from illustrations, which had also received little support from other authors). Grant resurrected Responsible Editor: R. Rosa.
Loligopsis and placed L. cyclura within it, alongside his new species L. guttata (= L. cyclura, fide Pfeffer 1912), with the type specimen of the genus being L. peronii (= nomen dubium, fide Berry 1932). In the L. guttata description, Grant mentioned lines of rough cartilaginous tubercles on the ventral surface, extending halfway down the mantle from the funnel (Grant 1833)-now recognized as one of the distinguishing features of Leachia species-but did not mention the presence of the cartilaginous line on which the tubercles are found.
A better understanding of the ecology, biology, and diversity of the genus was gained throughout the twentieth century, with authors noting the presence of luminous organs on the eyes (Joubin 1905), describing the secondary sexual characteristics of both males and females of taxa within the genus (Joubin 1931), and documenting maturation and spawning habits (Young 1975). Several new species (and genera) were described in the northern Pacific Ocean during this time including Drechselia danae Joubin 1931(= L. danae, fide Voss 1980, Zygaenopsis pacificus Issel, 1908(= L. pacifica, fide Young 1972 and Leachia dislocata Young 1972; however, few reports on the cranchiid fauna of the Southern Hemisphere were available until the late 1970s. In a much-criticised attempt to revise the New Zealand teuthofauna, Imber (1978) reported two locally occurring species of Leachia: L. cyclura and L. eschscholtzii (previously Perothis eschscholtzii Rathke 1833) and synonymized several geographically distant species to fit within those two taxa. These synonymies, which 'lumped' geographically separate and morphologically distinct species, were soundly refuted by Voss (1980).
As part of a larger work on identifying paralarval cephalopods, Voss et al. (1992) examined paralarval members of all nominal Leachia species. They concluded that six of the nominal 14 species appeared valid, but also acknowledged that the number might be revised upward to 11. They examined morphometric features of five (L. atlantica, L. danae, L. dislocata, L. lemur, and L. pacifica; L. cyclura was considered valid but was not described/compared) focussing on eye development, tubercle patterns and maturation size. Most of these species appeared to inhabit discrete, but adjacent geographic ranges, although some overlap was also noted. These observations represented a considerable advance in resolving Leachia, but the authors lacked access to specimens from several key geographic areas, such as the southern Pacific. They alluded to further forthcoming research on this genus, but this does not appear to have been published to date.
In Evans' recent review of the Pacific cranchiid fauna (2018), five Leachia species were identified (Table 1). These include two that are new to science, one of which is described below (the other, while possessing a unique suite of morphological characters, remains insufficiently known to permit formal description). For comparison, the three established Pacific species, plus L. cyclura, are rediagnosed below with a summary of key characters (with further remarks in Discussion). Tokyo National Museum (NSMT), Japan. Prior to examination, most museum specimens had been fixed in ~ 4% formalin and stored in either 70-80% ethanol or 50% isopropanol. Morphological examinations and illustrations were made using a dissecting microscope (Leica WILD M3B); an attached camera lucida was used to illustrate small specimens and characters. Morphological measures and counts follow Roper and Voss (1983). Measurements and counts for symmetrical features (those appearing on both sides of the midline, e.g. arms, eyes) were taken from the more complete side of the specimen. Ranges of indices are provided in text as X-Y-Z, where X is the lowest observed value, Y is the mean, and Z is the highest observed value. If both sides of the specimen were equally damaged, a 'minimal estimation' value was taken and noted; however, these values were not included in the mean value calculation. Mean values were not calculated for sucker counts (on either arms or tentacle clubs) as the presence of suckers, particularly on the distal portion of the arms, was variable due to damage.

Material was examined from the following institutions
Due to the unique morphology of the family Cranchiidae, with the head and mantle fused at three places, and coelomic compartments within the mantle, some specialised terms have been used to refer to specific features. The dorsal fusion between the head and the mantle is referred to as the nuchal fusion, while the two ventral fusion points (which are almost always identical to each other morphologically, symmetrical across the midline) are referred to as the funnel-mantle fusions. The cartilaginous connections at these ventral fusion points, which are thought to provide some structure for the otherwise gelatinous mantle, are generally linear and referred to as strips. These strips bear pointed tubercles, which may be simple (a single point) or complex (a cluster of points on a single tubercle). Cranchiid squid sucker dentition can be helpful in distinguishing between taxa; Fig. 1 provides examples of dentition morphology in Leachia.
Abbreviations used in text for morphological measure- Arm and tentacle club suckers were imaged using a dissecting microscope or a scanning electron microscope (SEM), after being critical-point dried and then sputtercoated in gold-palladium. When possible, beaks were extracted from preserved specimens, and soft tissues removed. Due to restrictions on destructive sampling, beaks and gladii could not be removed from most specimens housed in overseas collections.
Inked images were scanned and then digitally compiled using Adobe Photoshop.
The synonymies for individual taxa are limited to previous taxonomic descriptions and/or illustrations providing sufficient morphological detail to attribute the taxon definitively to species. When discussed in text, synonymised taxa are referred to by the original (junior) name and author, followed by the current species designation and attributing author in parentheses, e.g. "Zygaenopsis pacificus Issel 1908 (= Leachia pacifica, fide Young 1972)".

Diagnosis
Small-to medium-sized cranchiids (ML to 150 mm in examined material), body slender, with subterminal fins, together ovoid, circular, or rhombic in outline. Ventral mantle surface with a single longitudinal series of cartilaginous tubercles extending posteriorly from each ventral funnel fusion point. Eyes each with four or more circular photophores. Third arm pair more robust and over twice the length of other pairs; tentacles slender with proportionally small club.

Remarks
The most useful morphological characters for distinguishing Pacific Leachia taxa include the relative length of the ventral cartilaginous strips and the number of associated tubercles, the number of eye photophores, and the sucker dentition on the arms and tentacles (Table 1).

Geographic distribution
The distribution of this taxon is not well understood. The original description cites L. cyclura as being from the 'Pacific Ocean' (Lesueur 1821), but the collection coordinates (37°S 33°E) for the holotype are located in the Indian Ocean off eastern Africa.

Remarks
The type description of L. cyclura was based on an illustration and lacked sufficient detail to confidently attribute recent specimens to this taxon. The diagnosis above is compiled from reports by d'Orbigny (1845), Joubin (1905), Pfeffer (1912), and Nesis (1987). Some variation is apparent in these descriptions, as d'Orbigny failed to include the presence of ocular photophores, Joubin and Pfeffer reported L. cyclura having 5-6 ocular photophores, while Nesis described the species as having 8. As the type specimen is no longer extant, it is appropriate to use the earliest available information for comparison and therefore, as most material examined had more than 5-6 photophores, no specimens have been attributed to L. cyclura. Examination of Leachia material collected from the type locality would be useful in confirming diagnostic characters for L. cyclura, although multiple Leachia species may occur sympatrically (e.g. L. danae and L. dislocata).

Geographic distribution
Tropical Eastern Pacific, Baja California to Costa Rica.

Remarks
Leachia danae and L. dislocata can co-occur in the eastern Pacific, and younger individuals can be difficult to differentiate. The fins of L. danae are usually noticeably larger (FL 30-38-42% ML) than in L. dislocata (FL 20-27-35% ML). Perhaps most reliably, most specimens can easily be distinguished by the Arm III suckers: L. danae develop three distinct teeth from ~ 40 mm ML, while L. dislocata usually has 6-12 angular teeth (although several mature males examined in this study had laterally compressed suckers on Arm III with small apertures and 3 or 4 small sharp teeth). The great variation in sucker counts (26-180 in mature individuals) appears to represent sexual dimorphism, as SBMNH 464440, a mature male, had the lowest sucker count of any specimen examined despite being the second largest specimen examined.

Geographic distribution
Tropical Pacific, reported from California to Mexico, and Hawaiian Islands.

Remarks
This species most closely resembles L. pacifica (which also has a dislocated tubercle near the funnel fusions); however, these species can be distinguished by the number of ocular photophores (L. dislocata with 15 and L. pacifica with 5 or 6). The geographic range of L. dislocata overlaps with that of L. danae, but these two can also be distinguished by the number of eye photophores (20 in L. danae) or by examining arm sucker dentition (L. danae with three angular teeth on sucker ring margin, the central one noticeably enlarged). Unfortunately, small paralarval specimens (below 20 mm ML) could not be examined in this study, so further research is needed in order to complete the ontogenic series.

Geographic distribution
Western to central Pacific (most known material from Hawaiian and Japanese waters).

Remarks
Two uncatalogued paralarval specimens from Hawaiian waters were examined, courtesy of Dr. Richard Young. These specimens seem to concur with a description of P. pacificus by Sasaki (1929); however, few other historical descriptions of this species provide enough detail to distinguish it confidently from other immature Leachia. The eye photophore pattern does appear unique among similarly sized specimens of all other Leachia taxa examined in this study. The most morphologically similar species is L. dislocata, which has a sympatric distribution and also has an offset tubercle at the anterior end of the ventral cartilaginous strip. These two species can be distinguished based on eye photophore pattern (L. dislocata with 15; 5 or 6 in L. pacifica). Leachia dislocata also often has two tuberclelike protrusions at the anterior end of the gladius, which are absent in L. pacifica.

Type material
Ar ms taper ing rapidly to tip, ar m for mula III ≫ II≈IV > I: Arm I 4-7-15% ML, Arm II 7-13-20% ML, Arm III 20-30-45% ML, Arm IV 5-11-18% ML. Low protective membrane bordering suckers along entire length, webbing absent. Up to 50 suckers on arms in most specimens (two immature male specimens with 70 suckers on third arms), suckers spherical, with those at arm bases slightly largest, on slender peduncles; sucker rings with 7-10 pointed teeth (Fig. 5a); proximal margin of sucker ring with wide lip. Sex-specific arm-tip modifications: Arms III in females with fleshy membrane surrounding small suckers on distal tip of arm, lacking pigmentation (Fig. 5b), starting around 110 mm ML; Arm IVR in males with sucker series displaced to outer edge of arm on distal 25% and membranous keel along ventral margin of arm, starting around 90 mm ML.

Known distribution
New Zealand to Southern Australia; however, L. separata may have appeared in earlier accounts under other names and may ultimately prove to have a southern circumglobal distribution.

Molecular information
Fresh material of L. separata sp. nov. was fortuitously encountered in a concurrent study; tissue samples were taken, extracted, and DNA barcoded (cytochrome c oxidase subunit I [COI]; Hebert et al. 2003) for 11 individuals, reported as 'Leachia sp. nov.' (Braid and Bolstad 2019). The edited COI sequences (BIN BOLD:ADH5276) are available on the Barcode of Life Data Systems (BOLD) under the project name 'Cephalopod Fauna of the Kermadec Islands (project code KERCE).

Remarks
L. separata specimens have previously been attributed to several other names, but careful examination shows that published descriptions of these taxa do not match the present material. For instance, Leachia cyclura was originally described from an illustration of the dorsal surface of the type specimen; however, several sources (Joubin 1905;Pfeffer 1912) mentioned the presence of six eye photophores (whereas L. separata sp. nov. has eight). Both L. rhynchophorus and L. schneehagenii were described from small, damaged specimens, making detailed comparisons challenging; however, Pfeffer (1912) did describe the cartilaginous strip of L. schneehagenii as being < 25% of the mantle length and no additional cartilaginous tubercles were noted posterior to this strip. The type specimen of Pyrgopsis rynchophorus also has a short (~ 10% ML, although mantle was damaged) ventral cartilaginous strip, with no additional ventral tubercles occurring posterior to it, and examination of the eye revealed only four photophores present (although the eyes were both extremely damaged). Type material for Perothis dussumeiri Rochebrune 1884 (= Leachia sp.? fide Voss 1980) was described from a location close to that of L. cyclura, and has a ventral strip of tubercles that extend approximately 40% of the mantle length. Despite the specimen's poor condition, close examination reveals all tubercles faintly connected by the cartilaginous strip and that the strip is not linear; each strip of tubercles curves away from the ventral mid-line, forming a shallow "Z" shape. Based on existing descriptions and type material, these species all differ morphologically from L. separata.
The final possible existing name for this species was L. eschscholtzii, which was described by Rathke (1833) as having 8-12 [ventral mantle] tubercles connected by a "cartilaginous thread", ~ 50% ML, and eight eye photophoresbased on the holotype collected near Madagascar (no longer extant). Chun (1910) described a specimen he reported as a 'syntype' (ZMB 110359) of L. eschscholtzii, although it was collected 40 years later and off Borneo. This paralarval specimen differs from Rathke's description in having a cartilaginous strip ~ 40% ML with additional tubercles posterior to the strip itself. It is in fact quite similar to L. separata and could represent the same species, but given the differences from Rathke's description, is not L. eschscholtzii. Rathke (1833) both described and illustrated how the tubercles on the cartilaginous strip were connected and there is no mention of unconnected tubercles further along the ventral mantle. As Rathke provided detailed information about the outer mantle and chromatophores of the specimen, these are clearly not absent due to specimen degradation. Nesis (1987) later synonymized Rathke's missing type with L. cyclura. Comparison of historical descriptions and extant type material therefore distinguishes the taxon more commonly encountered in Aotearoa New Zealand from all known Leachia names from the Pacific and Indian Oceans. Its characters also do not align with either of the named Atlantic species, L. atlantica or L. lemur, which have cartilaginous strips 15-25% ML and lack individual tubercles posterior to the strip itself; the limited genetic information available for the genus (see "Discussion") also supports its status as distinct from these.
Leachia separata appears to retain some character states typically associated with juvenile life stages through to larger sizes than other Pacific Leachia species; among the material examined, only one specimen, a female (ML 112 mm), possessed sessile eyes (a character state normally attained by ML 40 mm in most Pacific congeners), and in several other larger specimens the eye stalks were still resorbing. Male arm modifications appeared to be developing on several mid-size specimens (ML 91, 94 mm), producing the unusual combination of stalked eyes and near-mature arm characters. Males may also mature at smaller sizes than females, as the arm-tip modifications were only just beginning to develop on the largest examined female specimen (ML 112 mm).
Leachia separata occurs sympatrically with a second apparently novel species (L. 'sp. NZ' sensu Evans 2018). While the latter remains insufficiently known to permit formal description (two juvenile specimens), preliminary characters include 10 or 11 eye photophores (eight in L. separata), and about six complex ventral tubercles that do not extend beyond the cartilaginous strip.

Etymology
Leachia separata is named for the separate cartilaginous tubercles, found posterior to the cartilaginous strip, on the ventral mantle (from the Latin 'separata' meaning 'separated'). This feature appears to be unique among known Leachia taxa.

Discussion
To clarify the Pacific Leachia fauna, physical specimens from throughout the region (and types from other regions) were examined, in conjunction with a global literature review of published Leachia records and accounts. The apparently restricted geographic ranges of most known Leachia species (Young et al. 2018) support the likelihood of the Pacific hosting a unique and characteristic species array, although detailed work is still needed on congeners from the Atlantic and Indian Oceans (including molecular data) to confirm this. A complete global review of the genus was not possible within the scope of Evans (2018), but is planned as a collaborative effort in the future, building upon groundwork laid by these observations.
Prior to this study, three Leachia species were known to inhabit the Pacific: L. danae, L. dislocata, and L. pacifica. Several additional names had been historically applied to subsets of Pacific Leachia material. The morphological characters of our new species were carefully compared with the three nominal taxa, and with accounts of the diagnostic features characterising the other three known species (L. cyclura, L. ellipsoptera, and L. lemur), as well as those associated with various historical names. All comparisons (see Table 1) led us to conclude that the New Zealand material represents at least one (likely two) undescribed Leachia species, to which none of the historical names can be applied. In the interest of improving the accuracy of Pacific biodiversity accounts, we therefore provide the present description of the novel species Leachia separata, which is well represented in NZ collections and has been sequenced (BIN BOLD:ADH5276). The two specimens we consider likely to represent a fifth Pacific species were collected farther north than the L. separata material, in an undersampled region of New Zealand's EEZ, where other more tropical oegopsid squid species are also known to occur (see Braid and Bolstad 2019). These small individuals do not yet provide enough evidence to justify a new name or full description, so we have acted conservatively. We hope that raising awareness of this possible additional taxon will enable fresh material, when encountered, to be recognised and preserved for further study.
One further name was investigated in the course of this study: L. rynchophorus Rochebrune 1884, described from the Cape of Good Hope. Despite a lack of detail in the original description, Nesis (1987, p. 270) reported this taxon as a "southern subtropical species found from Southern Africa to New Zealand," with long ventral cartilaginous strips (but no specified number of eye photophores). This bears some resemblance to our new species, but examination of the holotype of L. rynchophorus (MNHN 3.8.742, ML 20 mm, damaged) revealed clear differences from L. separata: short ventral cartilaginous strips (about 5-10% ML) and no additional posterior tubercles, which are already present at this size in L. separata. Only four circular photophores could be distinguished on the holotype's (albeit damaged) ventral eye surface. At present, we consider L. rynchophorus to be a nomen inquirendum, pending examination of additional material from the type locality, and perhaps further investigation of 'Leachia sp. A' (sensu Young et al. 2018) from the western South Atlantic.
Leachia separata sp. nov. has certainly been encountered by other authors; in one of the clearer cases, it was reported from Australia as Pyrgopsis pacificus by Allan (1945), who both noted and clearly illustrated the separated posterior tubercles. Pyrgopsis pacificus (= Leachia pacifica, fide Nesis 1987) had also been described from Japanese waters by Sasaki (1929), with illustrations depicting an animal with comparatively short cartilaginous strips (and much rounder fins), establishing the currently accepted characters of L. pacifica-in the meantime also known from Hawaii-and clearly not the same taxon as that depicted by Allan (1945). However, as is often the case with cranchiids (particularly those whose appearance changes considerably throughout ontogeny), there are also records that have proven more difficult to unravel. Chun (1910) and Pfeffer (1912) both examined Leachia material (from Borneo and Argentina, respectively) with separated tubercles on the ventral mantle, which they attributed to L. eschscholtzii (= L. cyclura, fide Nesis 1987, characterised by eight ventral eye photophores and ventral tubercle strips 20-40% ML), but which, upon further re-examination, may prove referable to L. separata, which would extend the species' known distribution.
When identifying Pacific Leachia species, eye photophore patterns and the ventral cartilaginous strips and tubercles appear to be the most useful characters. In well-preserved material, eye photophores are the most reliable character, since each species possesses a unique pattern; however, cranchiid eyes are notoriously delicate and routinely damaged during collection, reducing the utility of this character in most cases. Ventral tubercles are more robust and thus of greater utility, because the ventral mantle surface is often intact in preserved specimens; however, the differences among species may be relatively minor and require a practiced eye to distinguish. Leachia separata is the only species with tubercles extending along the ventral mantle surface posteriorly beyond the cartilaginous strip (Fig. 2), making this the most helpful feature for identifying this species. The limited material available for L. 'sp. NZ' (sensu Evans 2018) has not yet permitted a wide survey of tubercle structure in this taxon; preliminarily, both regular and antero-posteriorly compressed complex tubercles appear to be present, in fewer numbers (6 or 7) than in L. separata. The specimens are relatively small (46 and 56 mm ML), so this character should be examined in larger material when possible, although little