Sediment samples were taken at several sampling stations along Urías system (north-western Mexico) (see Gómez 2020: 43, fig. 1) with an Eckman grab (sampling area of 625 cm²). Triplicate sediment cores were taken at each sampling site with acrylic corers (sampling area of 24.6 cm²) and the upper 3 cm layer of each sample was retrieved and fixed in pure ethanol. Macro- and meiofauna were separated with 500 and 38 µm sieves and meiofauna was extracted through centrifugation with Ludox HS-40 following Burgess (2001) and Rohal et al. (2016) and preserved in pure ethanol. The biological material was sorted at a magnification of 40× using an Olympus SZX12 stereomicroscope equipped with DF PLAPO 1× objective and WHS10× eyepieces, and harpacticoid copepods were stored separately in 1 ml vials with pure ethanol. Illustrations and figures were made from the whole individual and its dissected parts using a Leica DMLB microscope equipped with L PLAN 10× eyepieces, N PLAN 100× oil immersion objective, and drawing tube. The dissected parts were mounted on separate slides using lactophenol as mounting medium. Huys and Boxshall (1991) was followed for general terminology.

Figure 1. 

Cletocamptoides biushelo sp. nov., female holotype A habitus, dorsal B caudal seta V C left caudal ramus, dorsal D habitus, lateral E left caudal ramus, lateral.

BENP baseoendopod;

ENP endopod;

EXP exopod;

EXP (ENP)1 (2, 3) first (second, third) exopodal (endopodal) segment;

P1–P6 first to sixth legs.

A phylogenetic analysis was performed with 39 characters and 36 taxa (Tables 1, 2). The taxa considered in this analysis were: the genus Cletocamptus as defined below (species with poor descriptions were excluded), the genus Cletocamptoides gen. nov. as defined below, Amphibiperita neotropica Fiers & Rutledge, 1990, and some of the best described species of the subfamilies Canthocamptinae, Hemimesochrinae, and Epactophaninae (see Table 2). We used the Podogennonta as defined by Seifried (2003) as outgroup. Our phylogenetic analysis does not aim at the complete understanding of the relationships amongst all the canthocamptid subfamilies, or amongst the genera and species of Cletocamptinae subfam. nov., and is not to be regarded as definitive, but only as illustrative to support the proposal of Cletocamptinae subfam. nov. and Cletocamptoides gen. nov.

The maximum parsimony analysis was performed using Phylip ver. 3.697 (Felsenstein 2005) and Bayesian inference with MrBayes ver. 3.2.7a (Huelsenbeck and Ronquist 2001). A Wagner parsimony method with unordered multistates, randomize input order species and weights option was carried out. Missing data were coded as ‘?’, inapplicable states were coded as ‘-’. A consensus tree was obtained from the majority rule and support values were performed from 10,000 bootstrap replicates. For the Bayesian inference, all characters were of equal weights. The default prior distribution of parameters was used for MCMCMC analyses, with one cold chain and three heated chains for 1,000,000 generations and sampled every 100^th generation, discarding the first 20% of the burning. Finally, ITOL v5 was used for visualization and editing trees. We used the bioinformatics server Chihuil at Instituto de Ciencias del Mar y Limnología-Universidad Nacional Autómoma de México (ICML-UNAM).

In addition to the limited number of canthocamptid taxa used for our phylogenetic analysis, we consulted the available descriptions of 102 species distributed in 45 genera of the different canthocamptid subfamilies for our comparative analysis.

Family Canthocamptidae Brady, 1880

 Cletocamptoides biushelo sp. nov.

http://zoobank.org/658B2BB2-1122-454E-B63B-ACC05A76B601

Figs 1, 2, 3, 4, 5, 6, 7

Type locality

Urías estuary, Mazatlán, Sinaloa State, stn. 5 (23.2056°N, 106.3715°W; 0.6 m depth; organic carbon content 0.99%; organic matter content 1.71%; sand 78.61%; clay 6.72%; silt 14.67%) (see also Gómez 2020: 43, fig. 1).

Other localities

Urías estuary, Mazatlán, Sinaloa State, stn. 7 (23.2174°N, 106.3917°W; 3.7 m depth; organic carbon content 5.59%; organic matter content 9.62%; sand 10.78%; clay 37.54%; silt 51.68%) (see also Gómez 2020: 43, fig. 1).

Material examined

Female holotype dissected (ICML-EMUCOP-180119-179); 18 Jan. 2019; S. Gómez leg. One additional female from stn 7 (see above) was used for molecular analyses.

Description

Female. Total body length of holotype 411 µm measured from anterior tip of rostrum to posterior margin of caudal rami. Habitus (Fig. 1A, D) semi-cylindrical, progressively tapering posteriad, without clear demarcation between prosome and urosome; body with somitic constrictions between somites; without cuticular sensilla-bearing socles. Prosome (Fig. 1A, D) consisting of cephalothorax, P1-bearing somite fully incorporated to the latter, and three free-pedigerous somites bearing P2–P4. Rostrum (Figs 1A, D, 3A) well-developed, not fused to cephalothorax, triangular, with wide base and rounded tip, with two subdistal sensilla, with row of subdistal spinules ventrally. Cephalothorax with depressions and with sensilla as shown; posterodorsal margin serrated, lateral margin with short slender spinules. P2–P4-bearing somites with posterior sensilla as depicted; posterior margin serrated; lateral margin with short slender spinules. Urosome (Figs 1A, D, 2A) comprising fifth pedigerous somite, genital double-somite, two free abdominal somites, and anal somite. P5-bearing somite largely as in previous somite but with fewer sensilla. Second (genital somite) and third urosomites separated dorsolaterally, fused ventrally forming genital double-somite; first half with posterior margin serrated, posterior half with posterior slender short spinules; both halves with sensilla as shown; anterior half with P6 (see below); posterior half with ventrolateral small slender spinules as depicted. Fourth urosomite largely as preceding somite (second half of genital double-somite), but with more spinules lateroventrally. Fifth somite as previous one dorsolaterally, ventrally with fewer spinular rows. Anal somite slightly wider than long; with small triangular operculum ornamented with small dorsal spinules, flanked by pair of sensilla; ventrally cleft medially, with pair of medial pores, with inner small spinules along inner margin of medial cleft. Caudal rami 1.5 × as long as wide; with inner long slender spinules; with six setae as follows: seta I missing; seta II and III lateral, issuing midway outer margin of ramus, the former proximal to and slightly as long as half the length of the later; seta IV arising at outer distal corner, as long as seta III, with bulbous base; seta V longest, without breaking plane, minutely bipinnate; seta VI issuing at inner distal corner, slightly shorter than seta IV, with two proximal spinules as shown, with bulbous base; dorsal seta VII located at the center of ramus, as long as seta III, tri-articulated at its base.

Figure 2. 

Cletocamptoides biushelo sp. nov., female holotype A urosome, ventral (P5-bearing somite omitted) B P5, anterior.

Antennule (Fig. 3A) six-segmented; first segment with two inner spinular rows, remaining segments with outer spinular row as depicted; all setae smooth. Armature formula as follows: 1[1], 2[7], 3[4], 4[1+(1+ae)], 5[1], 6[8+(2+ae)].

Figure 3. 

Cletocamptoides biushelo sp. nov., female holotype A rostrum and antennule B antenna.

Antenna (Fig. 3B) with allobasis ornamented with short longitudinal row of spinules proximally, armed with two abexopodal setae (one basal, one endopodal). Exopod one-segmented, minute, with one seta. Free endopodal segment as long as allobasis, with inner spinules proximally and subdistally, and with outer subdistal frill; with two inner spines and one slender seta medially, distally with two inner spines, two medial geniculate setae, and one outer spinulose element.

Mandible (Fig. 4A) with well-developed coxa ornamented with short spinular row as shown; gnathobase well-developed, with bi- and unicuspid teeth distally, and one thick bulbous element, and one ventral pinnate seta. Palp one-segmented, with three setae (one basal, two endopodal).

Figure 4. 

Cletocamptoides biushelo sp. nov., female holotype A mandible B maxillule C maxilla D maxilliped.

Maxillule (Fig. 4B) with robust praecoxa ornamented with spinules at base of coxal endite, and medially and ventrally on arthrite, the latter with one surface seta, seven distal spines and one ventral strong unipinnate seta. Coxal endite with two setae. Endopod and exopod incorporated to basis, the latter with four, endopod with one, exopod with two setae.

Maxilla (Fig. 4C) with syncoxa ornamented with medial and distal rows of small spinules, and with few larger outer spinules; with two endites bearing three setae each. Allobasis drawn out into claw accompanied by one seta. Endopod completely incorporated to basis, represented by three setae.

Maxilliped (Fig. 4D) subchelate. Syncoxa with spinules as shown, with one subdistal seta. Basis with one anterior and one posterior spinular row, with outer spinules proximally and subdistally. Endopod drawn out into curved claw with minute accessory seta.

P1 (Fig. 5A) with elongate bare intercoxal sclerite. Praecoxa triangular, with medial subdistal row of spinules. Coxa rectangular, wider than long, with two medial transverse rows of spinules as shown, and one posterior and one anterior short row of larger spinules close to outer margin. Basis with anterior spinules medially, on outer margin, at base of inner and outer spines, and between rami. Exopod three-segmented, situated at a lower level than the endopod, reaching tip of ENP2; segments with outer and subdistal spinules as shown; EXP1 without, EXP2 with inner seta; EXP3 with two outer spines and two distal setae of which innermost geniculate. Endopod two segmented, not prehensile; segments with inner and outer long slender spinules as depicted; ENP1 unarmed; ENP2 with three setae.

Figure 5. 

Cletocamptoides biushelo sp. nov., female holotype A P1, anterior B P2, anterior.

P2–P3 (Figs 5B, 6A) with elongate bare intercoxal sclerite. Praecoxa triangular, with transverse row of spinules as shown. Coxa rectangular, wider than long; with one medial and one outer spinular row. Basis with spinules medially, at base of endopod, and between rami; with outer seta (of P2 visibly shorter than in P3). Exopod three-segmented, situated at a lower level than endopod; spinular ornamentation of segments as shown; EXP1 and EXP2 with, EXP3 without inner distal frill; EXP1 without, EXP2 with inner seta with comb tip; EXP3 with one inner element without comb tip, two distal setae, and two outer spines. Endopod two-segmented; segments with spinular ornamentation as shown; ENP1 small, as long as wide, unarmed; ENP2 elongated, with three setae.

P4 (Fig. 6B) with intercoxal sclerite, praecoxa, coxa and basis as in P3. Exopod largely as in P3 except for two apical setae, and two outer spines on EXP3. Endopod one-segmented, minute, wider than long; with two setae of which innermost reduced.

Download as

CSV

XLSX

	P1	P2	P3	P4	P5
Exopod	0,1,022	0,1,122	0,1,122	0,1,022	3
Endopod	0,210	0,111	0,111	020	5

Figure 6. 

Cletocamptoides biushelo sp. nov., female holotype A P3, anterior B P4, anterior.

P1–P5 armature formulae as follows:

P5 (Fig. 2B) with baseoendopod and exopod fused, rami separated by shallow notch; spinular ornamentation as shown. Baseoendopod with outer seta arising from short setophore; endopodal lobe with five setae. Exopod with three setae.

Genital field (Fig. 2A) with median copulatory pore on second half of genital double-somite; each P6 represented by two setae.

Male. Unknown.

Etymology

The specific epithet is an anagram for “helobius” and refers to the close relationship between the Mexican new species and C. helobius comb. nov.

Figure 7. 

Schematic representation of the evolution of the mandibular palp in Cletocamptinae subfam. nov.

The selected characters (Table 1) in the matrix (Table 2) integrate the main attributes of canthocamptids regarding the body shape (character 1), posterior ornament of the cephalothorax and prosomites (characters 2 and 3), presence of cuticular sensillum-bearing socles (character 4), shape and ornamentations of the rostrum (characters 5 and 6), segmentation of the female antennule (character 7), armature complement of the antennary allobasis and segmentation of the antennary exopod (characters 8 and 9), segmentation, shape, and armature complement of the mandibular palp (characters 10–12), armature complement of the syncoxa of the maxilliped (character 13), several features of P1–P5 (characters 14–34), and male dimorphism (characters 35–39).

The heuristic search yielded nine equally most parsimonious trees of 220 steps, and consistency index of 0.39. The extended majority rule consensus tree is shown in Figure 8. The results show the monophyly of Cletocamptinae subfam. nov. and Cletocamptoides gen. nov. which are well supported clades with high bootstrap values. The Bayesian inference analysis recognized and supported the monophyly of Cletocamptinae subfam. nov. and Cletocamptoides gen. nov. Our Bayesian results indicate that Amphibiperita is nested within Cletocamptus, rendering the relationships between the genus Cletocamptus and Amphibiperita unresolved. However, we maintained the genus Amphibiperita separated from Cletocamptus due the maximum parsimony support as well as their unique characteristics (see below).

Figure 8. 

Phylogenetic hypothesis showing the relative position of Cletocamptinae subfam. nov. and Cletocamptoides gen. nov. Bootstrap (boot) and BI posterior probabilities (pp) represented as boot/pp. * Nodes of high support, boot > 90%; pp > 0.95.

The family Canthocamptidae was created and diagnosed by Brady (1880: 47) for Canthocamptus Westwood, 1836, Mesochra Boeck, 1865, Attheyella Brady, 1880, Tetragoniceps Brady & Robertson, 1876, Diosaccus Boeck, 1872, Laophonte Philippi, 1840, Normanella Brady, 1880, Cletodes Brady, 1872, and Enhydrosoma Boeck, 1872. The subfamily was subsequently raised to family level by Sars (1906: 193). Sars (1911: 418) foresaw the separation of Pteropsyllus Scott T., 1906a, Evansula Scott T., 1906b, Tetragoniceps, and Leptastacus Scott T., 1906a from the Canthocamptidae to a different family. Several groups/subfamilies have been proposed in the past as an effort to disentangle the complicated phylogenetic relationships amongst the canthocamptid genera.

Labbé (1926) proposed a new subfamily of Canthocamptidae, the Rhyncoceratinae, and one year later, Labbé (1927) gave a more detailed description of the species within that subfamily with additional figures. Labbé’s (1926; 1927) works were plagued with taxonomical errors, wrong identifications, and descriptions, and were harshly criticized by Gurney (1927) and Lang (1948) who finally disposed of the Rhyncoceratinae. As far as we are aware, the subfamily Rhyncoceratinae was never mentioned afterwards.

In a preliminary note on the revision of the freshwater genus Canthocamptus, Chappuis (1929a) subdivided the genus into several new genera distributed in two groups based on the one- or two-segmented condition of the antennary exopod, the Biarticulata for Canthocamptus, Bryocamptus Chappuis, 1929a and Echinocamptus Chappuis, 1929a (= Bryocamptus (Echinocamptus) Chappuis, 1929a), and the Uniarticulata for Attheyella and Elaphoidella Chappuis, 1929a. That same year, Chappuis (1929b) abandoned that system. Sixteen marine, one brackish, and nine freshwater genera were known at the time of Chappuis’ (1929b) paper, but he could not inspect but freshwater material (Chappuis 1929b). However, based on the available literature, he noted that the marine and brackish genera were far different (for a list of marine and freshwater genera allocated into the Canthocamptidae by that time, see Monard 1927) and proposed and diagnosed the subfamily Canthocamptinae for freshwater canthocamptids: Canthocamptus, Paracamptus Chappuis, 1929b (= Pesceus Özdikmen, 2008), Bryocamptus, Maraenobiotus Mrázek, 1893, Hypocamptus Chappuis, 1929b, Echinocamptus (= Bryocamptus (Echinocamptus)), Ceuthonectes Chappuis, 1924, Moraria Scott T. & Scott A., 1893, Attheyella, Elaphoidella, Epactophanes Mrázek, 1893, and several species incertae sedis (Chappuis 1929b: 495–496). Pesta (1932) believed that the morphological variability of the Canthocamptidae could be better understood by grouping its genera into two subfamilies. For this, he (Pesta 1932) proposed and diagnosed the subfamily Halocanthocamptinae for those genera inhabiting coastal marine and brackish waters, as well as continental saline habitats: Mesochra Boeck, 1865, Paramesochra Scott T., 1892, Remanea Klie, 1929, Leptomesochra Sars, 1911, Evansula, and Leptastacus (Pesta 1932: 79), and the subfamily Canthocamptinae Brady, 1880 for freshwater forms: Nitocrella Chappuis, 1924, Canthocamptus, Hypocamptus, Maraenobiotus, Moraria, Epactophanes, Elaphoidella, Attheyella, Paracamptus, Echinocamptus, and Bryocamptus (Pesta 1932: 91–92). Monard (1935) expressed some doubts about the naturalness of Pesta’s (1932) scheme but believed that it was, at least, practical. Gurney (1932) did not accept the concept of the Ameiridae and subdivided the Canthocamptidae into six groups, two of which (the first and the third) were included by Lang (1936) in his list of ameirid genera. The other groups of Gurney (1932) are a) Tetragoniceps for Tetragoniceps, Phyllopodopsyllus Scott T., 1906a, Pteropsyllus, Paramesochra, Leptopsyllus Scott T., 1894, and Diagoniceps Willey, 1930, b) Evansula for Evansula, Leptastacus, and Leptopontia Scott T., 1902, c) Cletomesochra for Cletomesochra Sars, 1920 (= Heteropsyllus Scott T., 1894), Hemimesochra Sars, 1920, Nannomesochra Gurney, 1932, Leptomesochra, Pseudomesochra Scott T., 1902, and Paramesochra, and d) Canthocamptus for the freshwater canthocamptids and Mesochra.

In a preliminary note, Lang (1944) removed Paramesochra from the Canthocamptidae to his newly created family Paramesochridae Lang, 1944, and proposed a new family, Tetragonicepsidae Lang, 1944 (whose spelling was subsequently corrected to Tetragonicipitidae) for Tetragoniceps, Pteropsyllus, Diagoniceps, Phyllopodopsyllus and Paraphyllopodopsyllus Lang, 1944 (= Phyllopodopsyllus). Lang (1948) believed that the Canthocamptidae and the Cletodidae bear a sister group relationship and that the difficulties for the separation of both families are the result of poor descriptions of freshwater canthocamptids. He (Lang 1948) rejected all the subdivisions of the Canthocamptidae pending a thorough review of the family, and included into this family the following genera: Hypocamptus, Maraenobiotus, Morariopsis Borutzky, 1931, Itunella Brady, 1896, Orthopsyllus Brady & Robertson, 1873, Ceuthonectes, Spelaeocamptus Chappuis, 1933, Attheyella, Canthocamptus, Bryocamptus, Nannomesochra, Mesochra, Echinocamptus, Moraria, Paracamptus, Antarctobiotus Chappuis, 1930, Afrocamptus Chappuis, 1932, Epactophanes, and Elaphoidella.

By the early 1950’s, Borutzky (1952) re-diagnosed the Canthocamptidae and the subfamilies Halocanthocamptinae, with Mesochra and “the genera most closely related to it “ (Borutzky 1952: 124), and Canthocamptinae with Canthocamptus, Paracamptus, Bryocamptus, Arcticocamptus Chappuis, 1929b (= Bryocamptus (Arcticocamptus) Chappuis, 1929b), Echinocamptus (= Bryocamptus (Echinocamptus)), Maraenobiotus, Hypocamptus, Attheyella, Elaphoidella, and Spelaeocamptus, and added the Morariinae for Ceuthonectes, Moraria, Morariopsis and some species incertae sedis, and the Epactophaninae for Epactophanes. He (Borutzky 1952) diagnosed these four subfamilies based on the structure of the female genital field and on the structure of the male P3 ENP. Later on, Por (1986: 423) proposed and (re)diagnosed the subfamily Hemimesochrinae Por, 1986 for those genera which, despite having a non-prehensile P1 ENP, were left into the Canthocamptidae in Lang (1948); these genera are: Nannomesochra, Hemimesochra, Heteropsyllus, Mesopsyllus Por, 1960, Poria Lang, 1965 (= Hanikraia Huys, 2009), and Dahlakia Por, 1986 (= Dahlakocamptus Huys, 2009). Additionally, Por (1986) relegated to incertae sedis the genera Cletocamptus and Parepactophanes Kunz, 1935, as well as Hemimesochra rapiens Becker, 1979 (= Perucamptus rapiens (Becker, 1979)), and Heteropsyllus serratus Schriever, 1983. One year later, Hamond (1987) suggested to consider no less than 18 taxa as subgenera of Canthocamptus s. lat. “until all their respective type species, and as many others as possible from all over the world, can be studied again by modern standards “… “as part of a world revision“ (Hamond 1987: 1027; but see also Wells 2007).

With 864 species distributed in 58 valid genera (WoRMS 2021), the Canthocamptidae is currently one of the most species-rich families within Harpacticoida. The Canthocamptidae is not a natural assemblage and could be divided into several families (Huys et al. 1996; Gómez Noguera and Fiers 1997; Boxshall and Halsey 2004; Huys and Conroy-Dalton 2006). The taxonomy and the phylogenetic relationships are immersed in chaos, and the different canthocamptid subfamilies are seldom mentioned in the recent literature probably because the genera described more recently are difficult to attribute to any of the existing subfamilies, and if forced, the subfamilial diagnoses have often to be widened more to include, probably unrelated taxa, exacerbating the already complicated systematics of the group. This problem is also common at the genus level, and is evident in Hemimesochra, which is a heterogeneous grouping of distantly related taxa (see Huys and Thistle 1989). The complexity of the Canthocamptidae is also fuelled by the weak boundaries between the Canthocamptidae and the Cletodidae and their respective genera. Huys and Thistle (1989) agreed with Por (1986) in that the dismantling of the Cletodidae sensu Lang (1944, 1948) was a necessary step, but did not accept Por’s (1968) view of taking the “variability within the Cletodidae, and the notion of genus itself “… “in a much wider sense than in any other harpacticoid family“ (Por 1968: 46). Huys and Thistle’s (1989) reasoning is adopted here. Huys and Thistle (1989) succeeded in narrowing the generic diagnosis of Hemimesochra by distributing its species in Boreolimella Huys & Thistle, 1989, Perucamptus Huys & Thistle, 1989, Carolinicola Huys & Thistle, 1989, Mesopsyllus, and Pusillargillus Huys & Thistle, 1989, leaving H. clavularis Sars, 1920 as the type and only species of Hemimesochra. They (Huys and Thistle 1989) concluded that Bathycamptus Huys & Thistle, 1989, Psammocamptus Mielke, 1975 and Boreolimella are phylogenetically related, being Bathycamptus the sister taxon of Psammocamptus, that Boreolimella is probably the sister taxon of the Bathycamptus-Psammocamptus lineage, and that Mesopsyllus could be a potential relative of this genus group.

George and Schminke (2003) noted that Isthmiocaris George & Schminke, 2003 could well be related to Bathycamptus and Itunella. Later on, Huys and Kihara (2010) removed Metahuntemannia Smirnov, 1946 and Pottekia Huys, 2009 from the Nannopodidae Brady, 1880 and found that these two genera are probably related to Psammocamptus, Bathycamptus, Perucamptus, and Isthmiocaris. However, as far as we are aware, no apomorphies have been found so far for the Hemimesochrinae. The same applies for the other canthocamptid subfamilies.

Karaytuğ and Huys (2004) proposed a practical, not necessarily phylogenetic, group of marine and brackish canthocamptids, the Mesochra group, in which the male P6 is represented by unarmed membranous flaps, and the female P6 is represented by one to three setae. Following Karaytuğ and Huys (2004), this species group includes species currently allocated into the Canthocamptinae and Hemimesochrinae, and two species incertae sedis; these are: Mesochra and Amphibiperita (Canthocamptinae), and Psammocamptus, Bathycamptus, Mesopsyllus, Isthmiocaris, Hemimesochra, Poria, Perucamptus, Pusillargillus (Hemimesochrinae), and Taurocletodes Kunz, 1975, and Parepactophanes (Canthocamptidae incertae sedis). The genus Mesochra is not a natural unit and is most probably polyphyletic (Gómez Noguera and Fiers 1997; Karaytuğ and Huys 2004), but if Karaytuğ and Huys’ (2004) assumption is correct, this might reflect the artificial nature of the canthocamptid subfamilies.

More recently, George (2020) contributed with important changes to the family Cletodidae sensu Por (1986), which has been regarded in the past as the sister taxon of the Canthocamptidae. He (George 2020) proposed the Cletodoidea Bowman and Abele (1982) for some Ancorabolidae Sars, 1909 sensu George (2020) and Cletodidae for which he detected 19 autapomorphies (George 2020). George (2020) also gave an amended diagnosis for the Cletodidae, which he extended to include several genera that belonged to the Ancorabolidae, and proposed a new subfamily, the Cletodinae Scott T., 1904 for the ancorabolid Ceratonotus group sensu Conroy-Dalton (2001) and Cletodes, defined by seven autapomorphies (George 2020). For the Cletodidae, he (George 2020) detected a set of five autapomorphies: a) P1 ENP not prehensile, P1 ENP1 dwarfed; b) P1 ENP2 elongate, at least as long as ENP1; c) the medial apical geniculate seta of the Cletodoidea (seta 5-en in George 2020: 470, fig. 3C) reverted into a bipinnate or plumose seta (5-en in George 2020: 470, fig. 3D) in Cletodidae; d) the outer distal claw-like element on the P1 ENP2 in the Cletodoidea (1-en in George 2020: 470, fig. 3C) reverted into a spine (1-en in George 2020: 470, fig. 3D) in Cletodidae; and e) loss of the accompanying seta on the maxillipedal claw.

The complexity of the Canthocamptidae is evident and it is due to bad and incomplete descriptions (Hamond 1987). Hamond (1987) proposed the revision of 18 taxa as part of a world revision using modern standards. Wells (2007) commented positively on Hamond’s (1987) proposal but at the same time he foresaw that this would not find ready support due to the massive effort that this would imply. Some authors have contributed to our understanding of the intergeneric relationships within isolated subfamilies. Huys and Thistle (1989), George and Schminke (2003), and Huys and Kihara (2010) have proved, at least partially, that the subfamily Hemimesochrinae is a compact unit, not so the other canthocamptid subfamilies. In our opinion, the establishment of well-defined taxonomic taxa (subfamilies, genera, species groups, etc.) is an obligate step towards a world revision of the family. The relationships between the Canthocamptinae and other genera outside that subfamily are still uncertain, and the proposal of a new subfamily, Cletocamptinae subfam. nov., for a well-defined, seemingly monophyletic group of genera (Cletocamptus, Amphibiperita, and Cletocamptoides gen. nov.) as well as the narrowing of the diagnosis for Cletocamptus, are justified and follow the same reasoning as explained below.

The high morphological diversity of the Canthocamptidae makes difficult the detection of (aut)apomorphies to substantiate the establishment and proposal of monophyletic taxa. In this sense, the role of weighted characters for the maximum parsimony approach was essential and allowed high support. The genus Cletocamptus as known before the present study, seems to be composed of two different lineages, the merbokensis-helobius-biushelo-lineage (mhb-lineage; for C. merbokensis, C. helobius, and C. biushelo sp. nov.) and the retrogressus-lineage (for the rest of the species of Cletocamptus). The split of Cletocamptus into two genera is proposed in this study: Cletocamptoides gen. nov. for the mhb-lineage, and Cletocamptus for the retrogressus-lineage.

The reconstruction of the hypothetical ancestor of the Cletocamptinae subfam. nov. is necessary to justify the (aut)(syn)apomorphies for the three genera within the new subfamily. The hypothetical ancestor was reconstructed based on plesiomorphic and (aut)apomorphic states of a selected set of characters, as essential to explain our hypotheses and phylogenetic reasoning.

Most canthocamptids possess a body without somitic constrictions between somites, and few species, e.g., C. merbokensis comb. nov., C. helobius comb. nov., C. biushelo sp. nov., possess a body with somitic constrictions between somites conferring the body a cletodid appearance. The latter is regarded here as the derived (synapomorphic) condition for the mhb-lineage. The ancestral Cletocamptinae subfam. nov. is hypothesized to lack somitic constrictions between somites, tapering gradually posteriad. The latter is considered as the primitive condition and is present in the retrogressus-lineage.

The posterior margin of body somites of most genera of the Canthocamptidae is plain, without any ornamentation, including one species of Cletocamptinae subfam. nov., C. chappuisi, but its condition requires confirmation; it is finely or deeply serrated in some genera, e.g., Antrocamptus Chappuis, 1957, Attheyella, Canthocamptus, Elaphoidella, Moraria, Paramoriaropsis Brancelj, 1991, Spelaeocamptus, and C. dominicanus; it is ornamented with short spinules, e.g., Attheyella, and C. albuquerquensis and C. tainoi, or with long slender spinules, e.g., Cletocamptus as defined here. It is assumed that the plain condition of the posterior margin of body somites is the most primitive within the Canthocamptidae. Species of some genera developed serrate posterior margins that eventually developed into short spinules. The ancestral Cletocamptinae subfam. nov. is hypothesized to have retained the finely serrate posterior margins present in Cletocamptoides gen. nov., and displayed by other canthocamptids (see above) The presence of long slender spinules in the posterior margin of body somites of Cletocamptus as defined here, is assumed to be the derived condition.

Posterior cuticular sensillum-bearing socles have been observed only in C. merbokensis comb. nov., for which this condition is regarded as apomorphic. All other canthocamptids lack cuticular sensillum-bearing socles and is regarded here as a clear plesiomorphy and is the hypothesized condition in the ancestral Cletocamptinae subfam. nov.

The rostrum is separated from the cephalothorax in species of Bryocamptus, Gulcamptus Miura, 1969, Itunella, Mesochra, Moraria, Pordfus Özdikmen, 2008, Boreolimella, Carolinicola, Dahlakocamptus, Heteropsyllus, Perucamptus, Isthmiocaris, and Psammocamptus. It is fused and sometimes reduced in the remaining genera. The well-developed rostrum separated from the cephalothorax is regarded here as the primitive condition for the family, and is assumed to have been retained in the ancestral Cletocamptinae subfam. nov. The condition of the rostrum of C. gomezi, which has been described as fused to the cephalothorax (Suárez-Morales et al. 2013) requires confirmation.

The rostrum lacks any spinular ornamentation in the Canthocamptidae; as far as we know, ventral subdistal spinular ornamentation is present only in the Cletocamptinae subfam. nov. for which is a clear autapomorphy and justifies the new subfamily.

As far as we are aware (we could not review all the descriptions for all the species of Canthocamptidae though), the male rostrum is non-dimorphic in most canthocamptids (e.g., Canthocamptinae, Hemimesochrinae and Epactophaninae), and is regarded here as the primitive condition, which was retained in the ancestral Cletocamptinae subfam. nov. The male rostrum is sexually dimorphic in Cletocamptus and Amphibiperita for which is regarded here as synapomorphic.(it is also sexually dimorphic in Taurocletodes tumenae Karaytuğ & Huys, 2004 (the male of T. dubius (Noodt, 1958) remains unknown), but the significance of its presence in Cletocamptus and T. tumenae is uncertain). Cletocamptoides gen. nov. retained the primitive non-dimorphic male rostrum.

The canthocamptid female antennule can be eight- to five-segmented. Following Huys and Boxshall (1991) oligomerization is the main evolutionary trend in Copepoda, and the eight-segmented female antennule is the most plesiomorphic condition within the Canthocamptidae. The female antennule is six-segmented in the new subfamily, and is regarded here as the condition for the ancestral Cletocamptinae subfam. nov. Cletocamptus gravihiatus was described possessing seven-segmented female antennules, but this requires confirmation.

The canthocamptid antenna has undergone fusion of the basis and the first endopodal segment to form an allobasis. The allobasis of canthocamptids is unarmed or equipped with one or two (one basal, one endopodal) abexopodal setae. The latter is regarded here as the most primitive condition of the allobasis. Two abexopodal setae are retained in most species of Cletocamptinae subfam. nov., but some species (C. gomezi, C. feei, C. retrogressus, C. merbokensis comb. nov., C. helobius comb. nov., A. neotropica) underwent loss of one seta, most probably the endopodal element. The antennary exopod of canthocamptids is mostly one-segmented, but the two-segmented condition is present in species of Bryocamptus, Canthocamptus, Gulcamptus, Maraenobiotus, Mesochra, and Heteropsyllus; it is represented by one seta in species of Stenocaris and is absent in Dahmsopottekina Özdikmen, 2009. The antennary exopod of the Cletocamptinae subfam. nov. is elongate and one-segmented except for C. feei for which the exopod was reported as two-segmented (the latter requires confirmation though), C. chappuisi whose antennary exopod was reported as represented by one seta as in C. helobius comb. nov. and C. biushelo sp. nov., and Amphibiperita in which the antennary exopod is lost. The antennary exopod of C. dominicanus is a minute segment with a single seta. The ancestral antenna of Cletocamptinae subfam. nov. is assumed to possess an allobasis with two abexopodal setae, and a one-segmented exopod; the free endopodal lobe possessed two spines and a slender seta laterally, and three spines and two geniculate setae distally.

The mandibular palp of Podogennonta is composed of a basis with four setae; a one-segmented endopod with three proximal lateral, three subdistal lateral, and three sets of setae fused basally with three, two, and two setae, respectively; and a four-segmented exopod with armature formula 2, 1, 1, 2 (Seifried 2003). Within the Canthocamptidae, Carolinicola (Hemimesochrinae) possesses a primitive biramous palp with the basis possessing three setae, the endopod being armed with six elements, and the one-segmented exopod with four setae (Huys and Thistle 1989). The reconstruction of the mandibular palp of the ancestor of the Cletocamptinae subfam. nov. (Fig. 7A) is based on the most plesiomorphic states observed in the different taxa included in the new subfamily. The ancestral mandibular palp (Fig. 7A) of the Cletocamptinae subfam. nov. is assumed to be two-segmented; basis with two setae; endopod one-segmented, separated from the basis, and equipped with three setae, of which innermost shortest; exopod incorporated into the basis but represented by one seta. The ancestral mandibular palp underwent several changes in different lineages. The ancestral palp (Fig. 7A) lost one of the apical setae, and the basis and endopod became fused (Fig. 7B) like in C. merbokensis comb. nov. Further losses include one of the basal setae and the exopodal seta (Fig. 7C) like in C. helobius comb. nov. and C. biushelo sp. nov. On the other hand, the mandibular palp might have lost one basal seta leading the precursor of Amphibiperita-Cletocamptus (Fig. 7D). In one lineage, the mandibular palp lost the remaining basal seta and the exopodal element (Fig. 7E) resulting in an unarmed basis and a trisetose endopod like in Amphibiperita. In another lineage, the mandibular palp of the precursor of Amphibiperita-Cletocamptus underwent loss of one of the endopodal setae (Fig. 7F) leading the ancestor of Cletocamptus. In one lineage, the basis fused into the coxa with the resulting loss of the basal seta, and the exopodal seta became incorporated into the coxa (Fig. 7G) as in C. assimilis, C. axi, C. cecsurirensis, C. deborahdexterae, C. fourchensis, C. goenchim, C. koreanus, C. levis, C. nudus, C. pilosus, C. schmidti, C. sinaloensis, C. spinulosus, and C. tertius. The inner endopodal seta looks longer than the outer element in C. nudus, and this requires confirmation. One of the endopodal setae, most probably the inner short element, became lost leading a minute endopodal segment with one seta only accompanied by the accessory exopodal seta issuing from the coxa like in C. pilosus (Fig. 7H). Further loss includes the disappearance of the exopodal seta (Fig. 7I) like in C. albuquerquensis, C. gomezi, C. samariensis, C. stimpsoni, and C. tainoi. The palp of C. gomezi looks somewhat elongate in Suárez-Morales et al. (2013: 213, fig. 3D) and requires confirmation. In another lineage, the exopodal seta became lost (Fig. 7J) like in C. retrogressus. Further modifications involve the fusion of the endopod and basis (Fig. 7K) like in C. confluens, and reduction of the palp (Fig. 7L) like in C. dominicanus. A transposition of the apical short and long setae was observed in the latter species but the significance of this is uncertain.

The maxillulary basis, exopod and endopod are fused into a single unit, making difficult the homologation and correct counting of their setae. A careful and more detailed study of this appendage is pending, but it seems that the coxal endite bears two setae, and the basis, exopod and endopod are represented by three setae each, and this is considered as the ancestral state for the maxillule. On the other hand, a very strong and spinulose ventral element is present in several species of Cletocamptus, e.g., C. assimilis, C. axi, C. cecsurirensis, C. deborahdexterae, C. fourchensis, C. levis, C nudus, C. samariensis, C. schmidti, C., and C. sinaloensis, but also in C. helobius comb. nov., and C. biushelo sp. nov., and seems to have appeared independently in different lineages; the ventral seta on the praecoxal arthrite in the other species is visibly slenderer and is regarded here as the plesiomorphic condition found in the ancestral Cletocamptinae subfam. nov. The condition of C. gomezi is not conclusive and requires confirmation.

The architecture of the maxilla and maxilliped is rather constant within the new subfamily. The ancestral maxilla is assumed to bear two syncoxal endites; the endopod is assumed to be completely absorbed into the allobasis and represented by three setae. The ancestral maxilliped is subchelate; the syncoxa bears one seta; the basis is unarmed; the endopod is drawn out into a claw with one accompanying seta.

The hypothetical ancestral segmentation and armature formulae of P1–P5 is:

The inner setae of P1–P4 EXP are assumed to be whip-like in the ancestral Cletocamptinae subfam. nov., i.e., devoid of a comb tip. See Gómez et al. (2017: 350, table 5) for the species of Cletocamptus with setae with comb tip. Cletocamptoides biushelo sp. nov. bears a seta with comb tip on P2–P4 EXP2; the condition of C. helobius comb. nov. is assumed to be the same. The inner setae of P2 and P3 EXP (P4 EXP lack inner setae) of Cletocamptoides merbokensis comb. nov., and the inner setae of A. neotropica are all whip-like. The ancestral P1 ENP is assumed to be two-segmented and prehensile. For the justification of the prehensile P1 ENP as the plesiomorphic state see George (2020). The ancestral sexual dimorphism of P3 ENP is assumed to be expressed in a three-segmented ramus, with inner sinuous slender long apophysis devoid of an arrow-like tip, as seen in all the members of the Cletocamptinae subfam. nov.

The pair of P5 in the females are separated (baseoendopods not fused medially) in all species of Cletocamptinae subfam. nov. The female P5 EXP and BENP are fused in all species of Cletocamptinae subfam. nov. except for A. neotropica in which the female P5 EXP and BENP are separated. The ancestral pair of the female P5 are assumed to be separated; the exopod and the baseoendopod of each leg are assumed to be separated; the exopod is assumed to bear five setae as in most species of the new subfamily (four setae are present only in A. neotropica, C. samariensis, C. confluens, and C. merbokensis comb. nov.; three setae are present only in the new species described herein and in C. helobius comb. nov.), and the endopodal lobe is equipped with six setae as in most species of the new subfamily (seven setae have been reported for C. feei but this requires confirmation; five setae are present in C. biushelo sp. nov. and C. helobius comb. nov.). The ancestral pair of the male P5 of Cletocamptinae subfam. nov., are assumed to be fused medially as in most species of the new subfamily (the condition of C. axi, C. cecsurirensis, C. retrogressus, and C. schmidti is not conclusive), and the exopod and baseoendopod are assumed to be fused but separated by a deep notch as in most species for which the males are known (e.g., C. deborahdexterae, C. stimpsoni, and C. sinaloensis); the endopodal lobe is hypothesized to be armed with three setae as in the males of all the species of Cletocamptinae subfam. nov. for which the males are known; the ancestral exopod is assumed to bear four elements (three elements have been reported only for A. neotropica, C. confluens, C. koreanus, C. samariensis, and C. helobius comb. nov.).

The female P6 of the ancestral Cletocamptinae subfam. nov. is assumed to bear two setae; the male P6 is assumed to be represented by two articulate unarmed flaps.

The ancestral caudal rami are assumed here to be longer than broad as in some species of Cletocamptus, and equipped with seven setae.

The genus Cletocamptus was erected by Schmankevitsch (1875) for C. retrogressus. The status of several species of the genus can be found in Gómez et al. (2004), Gómez (2005), Gómez et al. (2007), Gómez and Gee (2009), Gómez et al. (2013), and Gómez et al. (2017), and the complete list of species of the genus as redefined here can be found in the generic diagnosis above.

The ancestral Cletocamptinae subfam. nov. is hypothesized to lack somitic constrictions between somites, tapering gradually posteriad. Cletocamptus retained the plesiomorphic lack of such somitic constrictions.

Cletocamptoides gen. nov. displays the ancestral shape of somites with finely serrated posterior margins (see below), and the long slender spinules displayed by most species of Cletocamptus are assumed here to be the apomorphic condition for the posterior margin of body somites. Also, Cletocamptus retained the plesiomorphic lack of posterior cuticular sensillum-bearing socles on body somites.

Amongst the species of Cletocamptus, only C. gomezi has been described with the rostrum fused to the cephalothorax (Suárez-Morales et al. 2013), but this requires further confirmation. In general, the rostrum of Cletocamptus retained its plesiomorphic condition, i.e., well-developed and separated from the cephalothorax. However, the rostrum of Cletocamptus, as in the other two genera of Cletocamptinae, Cletocamptoides gen. nov. and Amphibiperita, display the synapomorphic subdistal ventral spinules that are of diagnostic value for the new subfamily. As said above, and as far as we are aware, the male rostrum is sexually dimorphic in Cletocamptus and Amphibiperita and seems to be non-dimorphic in Cletocamptoides gen. nov. The sexually dimorphic rostrum of Cletocamptus is deemed to be derived and constitutes a synapomorphy for that genus and Amphibiperita. The significance of the dimorphic male rostrum in Cletocamptus and Taurocletodes tumenae is uncertain, but this is probably the result of convergence. The relatively low consistency index of the tree of maximum parsimony confirms homoplasy events that could be related to this convergence process.

Cletocamptus retained the six-segmented female antennule hypothesized for the ancestral Cletocamptinae subfam. nov. The seven-segmented female antennule of Cletocamptus gravihiatus awaits verification.

Cletocamptus retained the ancestral allobasis of the antenna as well as the two abexopodal setae, and only C. gomezi, C. feei, and C. retrogressus lost one of these setae (one seta only is present also in C. merbokensis comb. nov., C. helobius comb. nov., and A. neotropica). Cletocamptus also retained the plesiomorphic one-segmented elongate antennary exopod (the two-segmented condition of the antennary exopod of C. feei requires to be confirmed) and the architecture of the free endopodal segment as described above for the ancestral Cletocamptinae subfam. nov.

The architecture of the mandibular palp of Cletocamptus is variable and underwent considerable reduction as has been detailed above. Briefly, the mandibular palp of the ancestral Cletocamptus might have been two-segmented, i.e., basis and endopod separated, being the basis and endopod equipped with one and two setae respectively, and the exopod represented by one seta (Fig. 7F). The ancestral mandibular palp underwent several changes (Fig. 7G–L) across two main lineages, the retrogressus-confluens-dominicanus-lineage (rcd-lineage; Fig. 7J, K, L), and the assimilis-pilosus-albuquerquensis-lineage (apa-lineage; Fig. 7G–I) (see also above). The rcd-lineage is defined upon the synapomorphic loss of the exopodal seta; C. confluens and C. dominicanus underwent secondary fusion of the basis and endopod which is regarded as synapomorphic for them, and C. dominicanus underwent secondary apomorphic reduction of the palp. The apa-lineage is defined here upon the synapomorphic absorption of the basis into the coxa and incorporation of the exopodal seta into the latter. Cletocamptus pilosus (Fig. 7H) underwent further loss of one of the endopodal setae which is considered here apomorphic for the species. Cletocamptus albuquerquensis, C. gomezi, C. samariensis, C. stimpsoni, and C. tainoi (Fig. 7I) underwent secondary loss of the exopodal seta which is regarded here as synapomorphic for them.

The maxillule is probably the most complicated buccal appendage in Harpacticoida due to the fusion of the basis, endopod and exopod in many taxa, which makes difficult the homologation and correct counting of the setal elements. Pending a more detailed and precise study of this appendage in Cletocamptus, the coxal endite possesses two setae, and the basis, the endopod and exopod are fused into one single unit, with three setae representing each of them. On the other hand, a core of species, C. assimilis, C. axi, C. cecsurirensis, C. deborahdexterae, C. fourchensis, C. levis, C. nudus, C. samariensis, C. schmidti, C., and C. sinaloensis, share the presence of a very strong spinulose ventral seta on the praecoxal arthrite, which is regarded here as derived within the genus (the condition in C. gomezi is inconclusive; a similar seta is present in some species of Cletocamptoides gen. nov.). On the contrary, a visibly slenderer seta in the other species of the genus is regarded here as the plesiomorphic condition (a similar seta is present in C. merbokensis comb. nov.).

The maxilla and maxilliped retained the plesiomorphic architecture of the Cletocamptinae subfam. nov. as described above.

The P1 ENP of Cletocamptus is two-segmented and non-prehensile. As said above, this is the derived condition within Canthocamptidae as opposed to the plesiomorphic prehensile endopod (see George 2020). Gómez et al. (2017) gave a list of species of Cletocamptus armed with setae with comb tips on P2–P4 EXP2. The presence of such setae is deemed to be derived as opposed to the whip-like setae, but this condition seems to occur widely in the Canthocamptidae.

Sexual dimorphism in Cletocamptus is expressed in a variety of modifications. These have been outlined above. The most common modifications in the appendages of the males include the subchirocer antennule, the basis of P1 with one inner distal process, the inner spine on P2 ENP2 with several degrees of modifications, the three-segmented P3 ENP with an inner apophysis on second segment and two apical setae on the distal segment, and the fused exopod and baseoendopod of P5. Deviations occur in some species though. The shape of the outer spine on P2 ENP2 varies from slightly modified as in C. albuquerquensis, C. cecsurirensis, C. confluens, C. dominicanus, C. retrogressus, C. samariensis, and C. tainoi, to moderately modified as in C. axi, C. deborahdexterae, C. fourchensis, C. goenchim, C. gomezi, C. levis, C. schmidti, C. spinulosus, C. stimpsoni, and C. tertius, and strongly modified into a recurved spine as in C. pilosus, and C. sinaloensis. The inner spine on the male P2 ENP2 is not dimorphic in A. neotropica and Cletocamptoides gen. nov., and the sexually modified inner spine on P2 ENP2 is regarded here as an apomorphy for Cletocamptus. The male P3 ENP in most species retained the three-segmented condition with an inner apophysis on second segment which is regarded here as plesiomorphic, but the second and third segments underwent fusion leading a two-segmented ramus with the inner apophysis well-developed and situated medially on second segment as in C. dominicanus, or very short and recurved, and situated subdistally on second segment as in C. albuquerquensis, C. chappuisi, C. confluens, and C. tainoi. The former division between the second and third segments of the P3 ENP is still visible in C. dominicanus, which also possesses very long asprothekes (see Gómez et al. 2017). A similar condition of the male P3 ENP, two-segmented with inner apophysis medially on second segment, is present also in A. neotropica. The P3 ENP2 and ENP3 of C. albuquerquensis, C. chappuisi, and C. tainoi are also fused, but the inner apophysis migrated to a subapical position and underwent extreme reduction, and became recurved. A similar condition is present in C. helobius comb. nov. The condition of the male P3 ENP of C. confluens is unique in that the second segment possesses and additional outer apophysis whose female homologue is uncertain. The pair of male P5 are fused medially in most species, but the exopod and endopod are separated in C. axi, C. cecsurirensis, C. retrogressus, and C. schmidti. The significance of this is not clear, but this could be due to character reversal. The male P6 of Cletocamptus is an unarmed articulated plate. A small seta has been observed issuing from P6 in some species, but the presence of this seta is most probably due to intraspecific variability. Several sexual modifications are present also in the relative length and shape of P2 EXP and/or P3–P4 EXP as in C. retrogressus, C. confluens; C. albuquerquensis, C. samariensis, C. spinulosus, C. tainoi, and C. trichotus. Similar sexual dimorphism in P4 EXP in which the first segment is visibly more robust has been documented for Amphibiperita (see Fiers and Rutledge 1990).

The genus Amphibiperita has been diagnosed by Fiers and Rutledge (1990). They (Fiers and Rutledge 1990) gave a lengthy discussion on the relationships between Mesochra and Amphibiperita and justified the creation of the latter for M. neotropica Jakobi, 1956. Briefly, Fiers and Rutledge (1990) removed M. neotropica from Mesochra based on the sexual dimorphism of P4 EXP, lack of antennal exopod, reduced mandibular palp, number of setae on the endopods of swimming legs (lack of the outer element on ENP2), and shape of the female genital field. Fiers and Rutledge (1990) commented on the uniqueness of the shape of the sexually dimorphic male P4 EXP1, but similar dimorphism is present in some species of Cletocamptus, pointing to a probable sister-group relationship between these two genera, which can be reinforced by the synapomorphic sexually dimorphic rostrum in both genera. Fiers and Rutledge (1990) also commented on the unique lack of the outer subdistal element of P1–P4 ENP2 (Fiers and Rutledge (1990) explicitly included here the P1 ENP, but this ramus is prehensile in Amphibiperita, with ENP2 being armed with a distal outer claw, a distal inner bipinnate seta, and an inner small slender seta, and in our opinion, the inclusion of P1 ENP in Fiers and Rutledge’s (1990) comment is not correct and is most probably a slip of the pen). On the other hand, the subdistal outer element on P2–P4 ENP2 is also missing in C. merbokensis comb. nov., being the supernumerary setae in A. neotropica the main difference between both species (three, four, and three elements in total in P2–P4 ENP2 in A. neotropica, but two, two, and two setae in P2–P4 ENP2 in M. merbokensis comb. nov.). Also, the complement armature of P4 ENP2 or only segment in the Cletocamptinae subfam. nov. as defined here is consistently composed of two apical elements only in all but in A. neotropica, which possesses one inner additional seta. As noted above, sexual dimorphism in Amphibiperita is expressed in the antennule, rostrum, segmentation of the urosome, P5, P6, and more importantly in the shape of the P4 EXP1 and modification of the P3 ENP. Some comments on the dimorphic P4 EXP1 and P3 ENP, and rostrum were given above. The male P5 of Amphibiperita follows the general pattern of Cletocamptinae subfam. nov., i.e., exopod and endopod fused and both legs fused medially into an elongate plate with three endopodal and three exopodal setae. On the contrary, the baseoendopod and exopod are separated in the female P5 and is regarded here as primitive within the new subfamily.

Karaytuğ and Huys’ (2004) Mesochra group (see above) was proposed for marine and brackish canthocamptids in which the male P6 is represented by unarmed membranous flaps, and the female P6 is represented by one to three setae (Mesochra, Amphibiperita, Psammocamptus, Bathycamptus, Mesopsyllus, Isthmiocaris, Hemimesochra, Poria, Perucamptus, Pusillargillus, Taurocletodes, and Parepactophanes) as opposed to most freshwater canthocamptids in which the male sixth legs bear two or three well-developed setae (Karaytuğ and Huys 2004, and references therein). Cletocamptus and Cletocamptoides gen. nov. should be added to this list. The significance of the occasional presence of one seta in some species of Cletocamptus is still controversial (see above).

The ancestral Cletocamptinae subfam. nov. lack somitic constrictions between somites (see above). The body shape of Cletocamptoides gen. nov. is cletodid-like, i.e., body somites clearly separated by somitic constrictions, and is a probable apomorphy for the new genus. On the other hand, C. merbokensis comb. nov. is the only species with integumental sensillum-bearing socles along the posterior margin of pro- and urosomites (except anal somite). However, the presence of integumental sensillum-bearing socles are deemed to be apomorphic for C. merbokensis comb. nov., and do not seem to indicate a close relationship with the Cletodidae.

The posterior margin of body somites of most canthocamptid genera can be plain, finely or deeply serrate, or with long slender spinules (see above). The posterior margin of the body somites of Cletocamptoides gen. nov. is finely serrate and is regarded here as plesiomorphic within the Cletocamptinae subfam. nov.

The condition of the antennary exopod and the mandibular palp of C. merbokensis comb. nov. are the most primitive within Cletocamptoides gen. nov. The antennary exopod of C. merbokensis comb. nov. is one-segmented and elongate, and bears three setae (one lateral, two distal); the antennary exopod of C. helobius comb. nov. and C. biushelo sp. nov. is reduced to one minute segment bearing only one seta. Similar antennary exopods are also found in Cletodidae.

The condition of the mandibular palp is more primitive in C. merbokensis comb. nov. than in the other two species of the genus. The mandibular palp of C. merbokensis comb. nov. possesses five setae; three setae are present in C. helobius comb. nov., and C. biushelo sp. nov. Besides the supernumerary setal complement in C. merbokensis comb. nov. relative to the other two species of the new genus, the basis and exopod are still discernible in the former species, being the inner – basal – extension armed with two setae; the basis of the mandibular palp of C. helobius comb. nov. and C. biushelo sp. nov. is completely absorbed into the palp but is still discernible by the presence of one inner seta. The exopod of the mandibular palp of C. merbokensis comb. nov. is indicated by a small outer protuberance with one seta; the exopod in C. helobius comb. nov. and C. biushelo sp. nov. underwent complete reduction and is not discernible. The two apical setae on the mandibular palp of these three species are regarded here as endopodal, and judging by their relative lengths, being the inner seta visibly shorter than the outer, both setae are homologous in these three species and in those species with similar armature (see above). The one-segmented condition of the mandibular palp is the most common in the Cletodidae, but the basal, endopodal and/or exopodal setae are still discernible in some species (e.g., Paracrenhydrosoma normani Gee, 1999b); the basis and endopod are separated, and the exopod is absorbed into the basis and is represented by a seta in Pyrocletodes Coull, 1973 (e.g., P. coulli Dinet, 1976); uniramous mandibular palps (basis and endopod separated) without any trace of the exopod are present also in some species of Paracrenhydrosoma (e.g., P. cornuta Kornev & Chertoprud, 2008, P. kiai Song, Dahms, Lee, Ryu & Khim, 2014), and in the monotypic genus Nannopodella Monard, 1928.

The unisetose condition of the maxillipedal syncoxa is present in most Cletocamptinae subfam. nov., and it is unarmed only in C. merbokensis comb. nov. and C. helobius comb. nov. for which is regarded here as apomorphic. Unarmed maxillipedal syncoxae are also present in species of some other canthocamptids e.g., Attheyella, Lessinocamptus, Paramoriaropsis Brancelj, 1991, Pordfus Özdikmen, 2008, and Termomesochra Itô & Burton, 1980 (Canthocamptinae), and Stygepactophanes and Epactophanes (Epactophaninae); the maxillipedal syncoxa is unisetose in the Cletodidae (George 2020).

The P1 ENP of Cletocamptoides gen. nov. displays the derived two-segmented condition (see also George (2020) for a detailed explanation on the plesiomorphic prehensile versus the derived non-prehensile P1 ENP in Podogennonta), but the armature complements of P2 and P3 of C. helobius comb. nov. and C. biushelo sp. nov. display a more primitive trisetose condition of P2–P3 ENP2, relative to the bisetose more derived condition of P2–P3 ENP2 of C. merbokensis comb. nov. Fiers and Rutledge (1990) noticed the lack of a subdistal outer element on P3–P4 ENP2 and used that character to differentiate Amphibiperita from Mesochra. The same condition is present in C. merbokensis comb. nov.

One-segmented P4 endopods are present in species of some other canthocamptids, e.g., Antrocamptus Chappuis, 1957, Australocamptus Karanovic, 2004, Gulcamptus Miura, 1969, Hypocamptus, Lessinocamptus Stoch, 1997, Morariopsis, and Elaphoidella (Canthocamptinae), Itunella and Isthmiocaris (Hemimesochrinae), and Stygepactophanes Moeschler & Rouch, 1984 and Epactophanes (Epactophaninae), but as far as we know from the literature available, it is an elongate and well-developed ramus. Within the new subfamily, a one-segmented P4 ENP is present also in C. dominicanus, but it is elongated. The P4 ENP of Cletocamptoides gen. nov. is one-segmented but it has undergone extreme reduction and is an apomorphy, and probable autapomorphy, detected for the new genus.

Both rami of P5 are fused and separated from the supporting somite in the females and males of the genera attributed to the Cletocamptinae subfam. nov., except for the female P5 of Amphibiperita whose female P5 EXP is separated from the baseoendopod. Both P5 are fused in the males, except for some species of Cletocamptus (e.g., C. axi, C. cecsurirensis, C. retrogressus and C. schmidti), and C. helobius comb. nov. The male of C. merbokensis comb. nov. displays the more derived condition of P5, being completely absorbed into the somite, and is considered a potential apomorphy for the species. The male and female P5 of C. helobius comb. nov. and C. biushelo sp. nov. resembles those of C. dominicanus, C. albuquerquensis, C. confluens, C. tainoi, and most probably C. chappuisi, in that both rami are hardly discernible and are separated only by a superficial shallow notch. A similar P5 is present also in Carolinicola Huys & Thistle, 1989 (Hemimesochrinae). The structure of the cletodid P5 is variable; the endopodal lobe and the basis can be fused or separated, and the exopod can be fused to or separated from the baseoendopod (George 2020).

The Mesochra group as defined by Karaytuğ and Huys (2004) (see above) is characterized by the male and the female P6 being represented by unarmed membranous flaps, and by one to three setae, respectively. Besides the twelve genera attributed to this species group by Karaytuğ and Huys (2004) (see above), C. merbokensis comb. nov. also belongs to this group. The presence of such male P6 in different lineages could be the result of the artificial nature of the canthocamptid subfamilies, but it is most probably due to convergent evolution. The uni-, bi-, or trisetose condition of the female P6 is widely distributed in the family.