Revision of the genus Laophontodes T. Scott (Copepoda, Harpacticoida, Ancorabolidae), including the description of a new species and a key to species

Abstract The description of Laophontodes volkerlehmanskiisp. nov. (Copepoda, Harpacticoida, Laophontodinae Lang) from the deep sea of the Kairei Field, western Indian Ocean, prompted the examination of the phylogenetic status of Laophontodes T. Scott and the relationships within the genus. The allocation of L. volkerlehmanskiisp. nov. to Laophontodes based on diagnostic characters was relatively straightforward, yet phylogenetic analysis of the genus considering 39 morphological characters detected not a single autapomorphy. This indicates that Laophontodes, which seems to form a monophylum with Ancorabolina George and Bicorniphontodes George, Glatzel & Schröder, actually represents the stem-lineage, retaining the characters of the common ancestor without having developed unique derived morphological characters. Most of the 13 known species of Laophontodes can be characterised by distinct apomorphies. However, phylogenetic comparison highlights some uncertainties due to the apparent heterogeneous distribution of some derived characters across the species, the weakness of other features, and the fragmentary and inadequate description of several species, which, in combination with the unavailability of type material, prevents a detailed comparison of several phylogenetically relevant characters. Thus, the analysis presented here provides a further step towards understanding the systematic relationships of and within Laophontodes, rather than a conclusive answer. Nonetheless, a detailed character discussion and a key to species are given.

1:1 solution of Congo Red and Acid Fuchsin adapted from Michels and Büntzow (2010). Specimens were individually mounted in a drop of glycerine surrounded by a transparent, self-adhesive reinforcement ring to prevent direct contact between the specimen and coverslip and, therefore, damage to or distortion of the specimen. Images were generated using a Leica TCS SP5 consisting of a Leica DM5000 B upright microscope and three visible-light lasers. The software used was LAS AF 2.2.1 (Leica Application Suite Advanced Fluorescence). Images were taken with objective HCX PL APO CS 10.0× 0.40 DRY UV at an extinction wavelength of 561 nm with 80% acousto-optic tuneable filter. Using overlapping optical sections, passing through the whole specimen with an ideal number of sections determined by the software, a series of stacked images was generated. Table 1 lists the applied settings. To obtain a threedimensional representation from selected body parts, the data produced during the CLSM scanning was processed with the software Drishti (http://anusf.anu.edu.au/ Vizlab/drishti/). The obtained images were finalised with maximum projection and Adobe Photoshop CS6 for adjusting colour, contrast and brightness. The type material is kept in the collection of the Senckenberg Forschungsinstitut und Naturmuseum, Frankfurt am Main (Germany).
The phylogenetic analysis strictly follows Hennig (1982) and Ax (1984Ax ( , 1988Ax ( , 1995 as explained by George (2020) and without the application of any computer-based cladistic programs. Consequently, Fig. 11 is not a computer-generated cladogram; instead, it is a manually generated clear presentation of the results of the phylogenetic discussion. General terminology follows Lang (1948), Huys and Boxshall (1991), and Huys et al. (1996). Terminology referring to phylogenetic aspects follows Ax (1984); the terms "telson" and "furca" are adopted from Schminke (1976).
Abbreviations used in the text:
Male: The male differs from the female in the following characters: habitus, A1, P3 and P4 endopod, and P5.
Etymology. The epithet volkerlehmanskii is given in dedication to the 60 th birthday of LMA Lehmanski's father Volker Lehmanski (Gelsenkirchen, Germany).
Diagnostic key to the species of Laophontodes *

Discussion
Laophontodes as "survivor" of the stem-lineage The type genus of the Laophontodinae -Laophontodes -is the only representative of that subfamily which cannot be characterised by autapomorphies (cf. George and Gheerardyn 2015;George 2018George , 2020. This causes certain complications, especially as Laophontodes comprises 13 species (cf. George 2020), plus the herein described L. volkerlehmanskii sp. nov.. As discussed by George (2020), Laophontodes seems to be closely related to Ancorabolina George, 2006 and Bicorniphontodes George, Glatzel & Schröder, 2019. These three genera presumably originate from a common ancestor, which developed one derived, apomorphic, feature (character 1 in Table 3 and below) [supposed ancestral, plesiomorphic, state in square brackets]: 1. A2 exopod lost and represented by 1 tiny seta only [A2 exopod 1-segmented, at least knob-like].
For a detailed discussion of character 1, see George (2020). Since all representatives of the three named genera share that apomorphy, George (2020) interpreted it as autapomorphic for the monophylum [Ancorabolina -Bicorniphontodes -Laophontodes] (Fig. 11). Ancorabolina and Bicorniphontodes share one further apomorphy, viz. the postero-lateral cuticular processes on the cephalothorax and thus form sister-groups, but are otherwise characterised by distinct autapomorphies (George 2020). However, this is not the case for Laophontodes. Species can only be assigned to Laophontodes using diagnostic characters and this resulted in Laophontodes becoming a conglomeration of many, at least partly, unrelated species. Consequently, several authors have excluded species from Laophontodes, placing them in newly erected and phylogenetically welljustified distinct genera (e.g., Lang 1965;George 2017: Paralaophontodes Lang, 1965 Even with the addition of L. volkerlehmanskii sp. nov. as the 14 th species, we could not identify a derived feature to support the monophyletic status of Laophontodes. The apparent lack of shared morphological novelties within Laophontodes suggests that the taxon may represent the stem-lineage, retaining the derived characteristics of the common [Ancorabolina -Bicorniphontodes -Laophontodes]-ancestor, having "failed" to develop its own derived characters (Fig. 11). Whilst this is not uncommon (cf. Ax 1984;Sudhaus and Rehfeld 1992), the authors believe this might be the first evidence of a surviving stem-lineage in the Harpacticoida. It remains to be seen if future (molecular genetic) studies may support this hypothesis.

Possible relations within Laophontodes
Phylogenetic relationships within Laophontodes cannot be resolved unambiguously. This is due to several reasons. For instance, the type material of many species is no longer available, preventing re-examination and comparison of most morphological characters. Moreover, as indicated by several authors (e.g., George and Gheerardyn 2015;George 2017George , 2018George , 2020George et al. 2019;Lee and Huys 2019), species descriptions (especially, older publications) are fragmentary and of poor quality, precluding detailed comparisons between species. Nevertheless, such descriptions may be occasionally satisfactorily resolved, as shown below for character 2 (Table 3).
Laophontodes antarcticus and L. propinquus species inquirenda were excluded from the phylogenetic analysis presented herein, due to the fragmentary and imprecise descriptions by Brady (1918 and1910, respectively), and the absence of type material for re-examination.
Careful examination of the remaining 12 species revealed 38 morphological characters as phylogenetically relevant. They are listed in Table 3 (characters 2-39) and are discussed in detail below. Four out of the 38 apomorphies are considered to be convergent (4, 5, 17, 30; underlined in Fig. 11), and the remaining 34 characters as unambiguous. The result of this phylogenetic analysis is graphically summarised in Fig. 11.
The derived chirocer condition is hypothesised as synapomorphic for the nine species and this is supported by the concurrent appearance of character 3, viz. the loss of the outer element of the female P4 endopod. Although the reduction of setae/spines occurs frequently and often independently in Harpacticoida, their simultaneous loss alongside the loss of the penultimate segment in the male A1 in all nine species strongly supports its synapomorphic status. Table 3. List of 39 morphological characters used for the here presented phylogenetic analysis. In the second column, plesiomorphic states are set in square brackets. Columns 3-14: 1 = apomorphies; 0 = plesiomorphies; ? = no information available; 1 = supposed convergences. 0* = also apomorphic state present, due to intraspecific variability; explanation in the text.

No.
Character Remarks on character 2: Recent detailed descriptions of the male A1 revealed the existence of a very small fourth antennular segment between the third and the swollen fifth segment in Laophontodes (e.g., George and Gheerardyn 2015;George 2018). This tiny segment -already known for other Ancorabolidae -was first documented by Conroy-Dalton (2004) in males of other Laophontodinae (Lobopleura ambiducta Conroy-Dalton, 2004 andProbosciphontodes Fiers, 1988). The detection of a fourth antennular segment lead subsequent authors to confirm the presence of a small/tiny fourth segment in Ancorabolina George, 2006(George 2006George and Tiltack 2009;Gheerardyn and George 2010), Bicorniphontodes (George and Gheerardyn 2015;George et al. 2019), and Calypsophontodes Lee, 2012 (Gheerardyn andLee 2012). This fourth segment has been also overlooked in Laophontodes (e.g., Schizas and Shirley 1994;Arroyo et al. 2003). However, the redescription of several laophontodin species (e.g., Bicorniphontodes bicornis, Laophontodes typicus, L. whitsoni), the description of new species of Laophontodes (George and Gheerardyn 2015;George 2018), and reexamination of available material of Laophontodes mourois (George pers. obs.) proved both the existence of this segment and that it had previously gone unnoticed. Therefore, it can be assumed with some certainty that this reduced fourth segment is also present in the male A1 of L. macclintocki and L. spongiosus. Remarks on character 3: As documented by George and Gheerardyn (2015), the female P4 enp-2 in L. whitsoni apparently presents an intraspecific variability; three examined females had four setae -two apical, one outer and one inner (formula I:2:1) -, while other females lacked the outer seta (0:2:1) or even both lateral elements (0:2:0). Considering that the secondary development of a formerly deleted element is possible (cf. George 2020 and references therein) but rather improbable, we conclude that L. whitsoni originally bears all four elements in the P4 enp-2 (I:2:1). The reduction of the outer or both lateral setae is seen here as a deviation that has occurred within the species. Therefore, L. whitsoni is not grouped with those nine taxa that share the synapomorphic loss of the outer seta (character 3). Similarly, although L. gertraudae also lacks the inner seta of P4 enp-2 (character 26), a closer relationship with L. whitsoni cannot be presumed due to the rarity of this character in the latter species. Nonetheless, this intraspecific variation and potential relationships are indicated by an asterisk * in the respective fields in Table 3.
Further relationships between Laophontodes macclintocki, L. sabinegeorgeae, and L. whitsoni remain unsolved (Fig. 11). Whilst each of these species can be characterised by at least two autapomorphies (Table 3, Fig. 11), no derived characters have been found that might support any sister-group relationship.
Two inner setae on P3 exp-3 (character 4) and P4 exp-3 (character 5) are present in L. monsmaris, L. sabinegeorgeae, and L. whitsoni, while one inner seta was lost in the P3 and P4 exp-3 of the remaining Laophontodes species. This is seen as the derived state and thus as synapomorphic for the respective species. An exception is L. macclintocki, in which an inner seta is lost in the P3 and P4 exp-3. Unlike the other eight species in this group, L. macclintocki does not exhibit the synapomorphic state for characters 2 and 3, and therefore, the loss of the inner setae on the P3 and P3 exp-3 in L. macclintocki can be assumed to be convergent. The alternative would be to assume that the apomorphic character of the chirocer A1 is the result of convergent development, which is far more implausible.
Furthermore, the eight species share derived characters 6 and 7, viz. the loss of the inner seta on P2 and P3 exp-2, respectively. Although we admit that characters 6 and 7 are rather weak because the reduction of elements may occur independently (see remarks on character 3), it is assumed that, together with characters 4 and 5, they constitute a set of deviations that were developed in a common ancestor of the eight species (Table 3, Fig. 11) and are thus interpreted as synapomorphies for them.
Among the above group of eight species, Laophontodes typicus and L. mourois (as well as all more basal species) show the plesiomorphic retention of an inner seta on the third exopodal segment of P2, whereas the remaining six species share its derived loss. This is seen here as synapomorphic for L. spongiosus, L. gertraudae, L. georgei, L. volkerlehmanskii sp. nov., L. scottorum, and L. sarsi.
In the harpacticoid ground pattern, the two apical setae of P2-P4 exp-3 are longer and more flexible than the outer spines of those segments, being of almost the same size. This state is retained in most Laophontodes species except for Laophontodes sarsi, L. scottorum, L. volkerlehmanskii sp. nov., and L. georgei. These species are characterised by a clearly diminished inner apical seta of the P2-P4 exp-3, being much slenderer than the outer apical seta. This is interpreted as synapomorphic for these four species.
As with the other subgroups of the genus, relationships with those species excluded from the subgroup require further phylogenetic resolution (cf. interrogation marks in Fig. 11). The relationships of Laophontodes spongiosus and L. gertraudae with this last subgroup of four species, L. sarsi, L. scottorum, L. volkerlehmanskii sp. nov., and L. georgei, remain unclear, as no further apomorphic characters have been identified.
In addition to the derived inner apical seta in the P4 exp-3, Laophontodes sarsi, L. scottorum, and L. volkerlehmanskii sp. nov. exhibit a short, flexible seta on the P2 and P3 exp-3, whilst L. georgei retains the normal-shaped inner apical setae. This is assumed as synapomorphic for the former species.
In Harpacticoida, the inner and apical setae of P2-P4 exp-3 are usually biplumose, which must be regarded as the plesiomorphic condition. Thus, the development of unarmoured, bare setae constitutes a deviation. Accordingly, the presence of the bare, Figure 11. Cladogram summarizing the results of the phylogenetic analysis provided in the present contribution. Numbers in rectangles pointing to discussed characters listed in Table 3. Underlined numbers refer to convergent deviations. Detailed explanations are given in the text.
Remarks: The development of a bare inner apical seta in the P2 exp-3 is also present in L. typicus and L. mourois (Table 3, character 28). Nevertheless, we assume that the loss of the setal ornamentation occurred independently in these species. Shortening of the inner apical setae presumably took place before the loss of their armour in Laophontodes sarsi and L. scottorum. This assumption is further supported by the fact that these short setae remain biplumose in L. volkerlehmanskii sp. nov.. In contrast, the length of the inner apical seta in the P2 exp-3 of L. typicus and L. mourois is normal as in P3 and P4. A phylogenetic discussion of character 28 is given below.
The following characters, 15-39, listed in Table 3, characterise the different Laophontodes species. Direct comparison of characters was impossible for most species because of the lack of suitable type material. Consequently, the characterization of species by apomorphic characters is far from complete, with several species being characterised by just one potential apomorphy. Until further data are available, the current analysis provides sufficient information for an initial phylogenetic characterization of each species.

Characterization of Laophontodes species
Laophontodes sarsi, character 15 (Table 3, Fig. 11): The common shape of the harpacticoid P5 baseoendopodal setae (including Laophontodes) is a bipinnate one, with the pinnae being distinct. This is considered the plesiomorphic state. In contrast, the pinnae are strengthened and fused to the seta in L. sarsi, giving a "fish-bone" appearance (George 2018). This shape is rarely seen in Harpacticoida, and it is considered to be derived, i.e., an apomorphic state.
Laophontodes scottorum, characters 16 and 17 (Table 3, Fig. 11): George (2020) considered a cylindrical, slender body -listed here as character 1 in Table 3 -as synapomorphic for Ancorabolina, Bicorniphontodes, and Laophontodes. According to George (2020), the plesiomorphic state consists of a fusiform body that tapers posteriorly. Laophontodes scottorum deviates from character 1 in presenting a robust, rather compact body somewhat dorsoventrally compressed (character 16) (T. Scott 1907;George 2018). Moreover, the body somites are laterally extended (character 17); these lateral extensions are reminiscent of epimeres in other harpacticoid taxa (T. Scott 1907;George 2018). However, L. scottorum exhibits the synapomorphic state for characters 2-14, clearly justifying its assignment to Laophontodes. Moreover, although its body shape does not fit the synapomorphic state for Laophontodes, it does not match the plesiomorphic condition either. Instead, it can be postulated that the body shape of L. scottorum represents a secondary deviation, in addition to Character 17. Both character states are considered autapomorphic for this species.
Character 17 is also present in L. spongiosus (cf. Schizas and Shirley 1994); however, it only shares this character and apomorphies 2-8 with L. scottorum, suggesting it branched off much earlier, not sharing apomorphies 9-14. Therefore, we assume that the lateral extension of the body somites occurred convergently in these two species.
Laophontodes volkerlehmanskii sp. nov., characters 18-23 (Table 3, Fig. 11): This newly described species presents a series of morphological differences compared to the remaining species of Laophontodes. To characterise it unambiguously, the following six autapomorphies were selected: Strong serration of the posterior margin of the anal operculum (character 18), which is unique within Laophontodes -with the anal operculum of almost all other species exhibiting a row of fine spinules; subapical displacement and elongation of the furcal tube pore (character 19), compared to the usually small furcal tube pore located on the outer anterior lateral margin of the ramus in most Laophontodes species; a strongly diminished inner seta on the P1 basis (character 20), which does not reach the endopod in L. volkerlehmanskii sp. nov., similarly contrasts to the synapomorphic condition for Laophontodes in which the inner basal seta of the P1 is of moderate length, usually reaching the endopod; development of STE on the P1 outer basal seta (character 21), the P1 exp-1 outer spine (character 22), and the outer subapical seta of the male P5 exopod (character 23) are likewise exclusive derived features of L. volkerlehmanskii sp. nov., STE being rarely documented in Harpacticoida. In Laophontodes, only two species have been described possessing STE, namely L. whitsoni (characters 38, 39) (George and Gheerardyn 2015) and L. volkerlehmanskii sp. nov.. This has been confirmed by examination of various Laophontodes material (George pers. obs.). Thus, characters 18-23 are seen here as unambiguous autapomorphies of Laophontodes volkerlehmanskii sp. nov.
Laophontodes georgei, characters 24 and 25 (Table 3, Fig. 11): L. georgei was described as L. norvegicus George, 2018by George (2018 and subsequently renamed by Lee and Huys (2019), with the illustrations provided by Sars (1908) as the holotype. It may be characterised by two deviations: Firstly, L. georgei has a long, non-geniculated apical seta on the P1 enp-2 (character 24) (Sars 1908) compared to a long geniculated seta in all other Laophontodes species, as well as in Ancorabolina and Bicorniphontodes, which are considered closely related. Consequently, the geniculated seta is considered to be the plesiomorphic state. As an early harpacticoid description, being more than 100 years old (Sars 1908;as L. typicus), one might suspect this geniculation was overlooked. However, G.O. Sars was a keen observer, and in fact noted geniculated setae on the P1 exp-3 of L. georgei. Thus, there is no reason to assume that he had overlooked the geniculation in the apical seta of P1 enp-2. It is therefore concluded that in L. georgei the P1 enp-2 apical seta lost the geniculation, resulting in an autapomorphic character for that species.
Moreover, L. georgei is the only Laophontodes species that exhibits a 1-segmented P4 endopod (Sars 1908;George 2018) (character 25). This reduction of the enp-1 is interpreted as autapomorphic of the species.
Laophontodes spongiosus, characters 17 and 27 (Table 3, Fig. 11): L. spongiosus has three derived characters that are pooled as one autapomorphy (character 27): the abdominal somites except the telson are characterised by H-like cuticular dorsal structures (Schizas and Shirley 1994). Such structures are unique within Laophontodes. They are reminiscent of similar structures found in Paralaophontodes (Lang 1965;Fiers 1986;George 2017), but as shown by George (2020), no closer relationship between Paralaophontodes and L. spongiosus exists. So, although these derived features appear to be convergent for the two taxa, they are considered autapomorphic for the latter.
In addition to character 27, another deviation discussed here is interpreted as convergent (character 17, cf. L. scottorum).
Laophontodes mourois-L. typicus-group, character 28 (Table 3, Fig. 11): The transformation of pinnate/plumose setae into bare elements has been discussed above (character 12). The rather ancestral inner apical seta in the P2-P4 exp-3 of Laophontodes is biplumose, as observed in, for example, L. gertraudae, L. monsmaris, L. spongiosus, and L. whitsoni. Within the genus, however, two developmental directions were detected. The first is the reduction in length of the inner apical setae, followed by a subsequent loss of ornamentation (characters 9-11; 12-14); this is seen in L. georgei (on the P2), and in L. volkerlehmanskii sp. nov., L. scottorum, and L. sarsi (on P2-P4), and has been discussed above. A second developmental direction is seen in Laophontodes mourois and L. typicus, in which the length of the inner apical seta of the P2 exp-3 is normal, but has lost it ornamentation. This derived state is considered as synapomorphic for Laophontodes mourois and L. typicus.
Laophontodes typicus, character 30 (Table 3, Fig. 11): L. typicus does not present any exclusive morphological deviations. Compared with other Laophontodes species, L. typicus seems to retain most plesiomorphic character states. Only two deviations have been observed in the species, characters 28 and 30, and these are shared with other congeners. Of these, character 28 supports a sister-group-relationship with L. mourois (see above). In contrast, character 30 -the lack of the minute seta accompanying the maxillipedal claw -, whilst also found in L. macclintocki (see below), is thought to be the result of convergence: L. macclintocki lacks the apomorphic state of character 28, but exhibits apomorphies 4 and 5 (see below), which are not seen in L. typicus. Therefore character 30 is regarded as autapomorphic for L. typicus.
According to the description of Sars (1908), no such minute seta is present in the maxillipedal claw of L. georgei. Future examination may reveal if this is true or if the seta was overlooked by Sars (1908).
Laophontodes monsmaris, characters 31, 32 (Table 3, Fig. 11): This species exhibits two autapomorphic characters, which are unique not only within Laophontodes but also in the Laophontodinae: the maxilliped is extremely elongated and strengthened (character 31), and the penultimate abdominal somite overlaps the telson (character 32) (George 2018). Because of these autapomorphies a phylogenetic characterization of L. monsmaris is unambiguous. of such ripples is confirmed for L. gertraudae, L. monsmaris, and L. volkerlehmanskii sp. nov. (George 2018;present contribution), it is still not known if they occur in L. antarcticus, L. georgei, L. macclintocki, L. propinquus sp. inquirenda, and L. spongiosus. Consequently, it was not possible to include this character in the here presented study.

Summary and conclusion
The description of Laophontodes volkerlehmanskii sp. nov. facilitated an attempt to characterise the genus Laophontodes and to elucidate the phylogenetic relationships within the taxon. Careful comparison of 39 morphological characters led to the conclusion that Laophontodes cannot be characterised by any autapomorphies. Instead, it seems to reflect the stem-lineage of a monophylum comprised of Ancorabolina, Bicorniphontodes, and Laophontodes. While Ancorabolina and Bicorniphontodes can be characterised as monophyla and furthermore present a sister-group relationship (George 2020), Laophontodes retains the characters of the common ancestor, without having developed unique deviations that might be considered as synapomorphies of species assigned to the genus.
Similarly, discrimination of the 12 Laophontodes species examined here (L. antarcticus and L. propinquus excluded) is ambiguous. Most characters refer to the reduction of single setae or spines, which happens often and independently in harpacticoid species. Moreover, several features presumed to be derived, such as the development of fine dorsal cuticular ripples on the pedigerous somites P2-P4, or the lateral extension of the body somites, seem to be distributed quite heterogeneously amongst the species. Finally, many descriptions of Laophontodes species are incomplete or of poor quality, and the respective type material is no longer available. Those conditions have inhibited the comparison of all the morphological characters that may be otherwise of phylogenetic relevance.
Nonetheless, each of the Laophontodes species can be characterised by certain derived characters, even if convergence has to be assumed for some of them. Thus, the phylogenetic analysis undertaken provides insights into the phylogenetic relationships of and within Laophontodes and serves as the base for ongoing research.