Hivanua tekao Maddison, sp. nov.

Hivanua flavipes (Berland, 1933), comb. nov., transferred from Havaika.

Hivanua nigrescens (Berland, 1933), comb. nov., transferred from Plexippus.

Hivanua nigrolineata (Berland, 1933), comb. nov., transferred from Havaika.

Hivanua rufescens (Berland, 1934), comb. nov., transferred from Habronattus.

Hivanua tekao Maddison, sp. nov.

Hivanua triangulifera (Berland, 1933), comb. nov., transferred from Havaika.

An arbitrary combination of letters, containing a reference to the largest two islands of their range, Hiva Oa and Nuku Hiva. Grammatical gender: feminine.

Reflective scales and relatively sparse setation on the legs distinguish Hivanua and Havaika from other genera of the pellenine subgroup of harmochirines, which have fuller and more varied setation. Hivanua is distinct from Havaika by a more posterior placement of the epigynal coupling pocket (ECP). In Hivanua, the crescent-shaped atrial ridges shielding the openings reach posteriorly only as far as the midpoint of the ECP; in Havaika, the atrial ridges merge with the posterior end of the ECP (Prószyński 2002, 2008). Male first leg of Hivanua unusually long; for example, the holotype of H. tekao sp. nov. has a body length of 6.9 mm but a first leg length (femur to tarsus) of 12.5 mm. Third patella+tibia about the same length as fourth (distinctly longer in Pellenes and Habronattus). Palp with bulb smaller relative to cymbium and tibia compared to Habronattus and Pellenes. TmA sometimes present, unlike Havaika. First leg tibia usually or often with four anterior ventral macrosetae (other harmochirines with fewer). (Four macrosetae present in all H. tekao/H. cf. tekao, H. flavipes, H. nigrolineata, and about half of the H. rufescens specimens, mostly juveniles).

Six species are placed in Hivanua, five of which were described by Berland (1933, 1934). A new species, H. tekao sp. nov., is described below, and one of Berland’s, H. rufescens, is partially redescribed. The other four Berland species are not redescribed here except via illustrations of their female holotypes (in BPBM, examined; Figs 3–10). The holotype of N. flavipes is from “Hiva Oa, Mont Temetiu, 1300 m. d’alt.”, that of H. nigrescens is from “Tahuata: sommet du Haaoipu, 900 m”, that of H. nigrolineata from “Nukuhiva: Ooumu”, and that of H. triangulifera from Tahuata. There is variation among species in body form, with H. nigrolineata narrow and linearly marked, and H. nigrescens robust and with a rough texture. H. nigrescens was inexplicably synonymized with Plexippus paykulli (Adouin, 1826) by Berland himself. The holotype of H. nigrescens is clearly a Hivanua (Figs 5, 6), similar to H. tekao sp. nov. One wonders if Berland confused it with his Sandalodes magnus Berland, 1933, which is indeed a synonym of P. paykulli, and whose figures appeared in the same plate as Sandalodes nigrescens.

Figures 3–10. 

Berland’s holotypes of four Hivanua species, each showing habitatus and ventral view of epigyne 3, 4 holotype of Sandalodes flavipes Berland, 1933 5, 6 holotype of Sandalodes nigrescens Berland, 1933 7, 8 holotype of Sandalodes nigrolineatus Berland, 1933 9, 10 holotype of Sandalodes triangulifera Berland, 1933. Abbreviations: CO, copulatory opening; ECP, epigynal coupling pocket. Scale bars: 1.0 mm for bodies; 0.1 mm for epigynes.

Several of Berland’s Marquesan harmochirines are placed in Hivanua only tentatively. Hivanua rufescens can be placed with the type species H. tekao sp. nov. with confidence based on the molecular evidence. These two species, along with H. nigrescens and H. triangulifera, are large bodied, distinctly larger than most of the Hawaiian Havaika. The remaining two species, H. nigrolineata and H. flavipes, are considerably smaller-bodied and more delicate, and could easily be mistaken for Havaika. They share with the larger Hivanua one distinction from Havaika, the more anterior placement of epigynal atria. For this, and for geographical parsimony, I will here place them into Hivanua, but this should be considered provisional until more material can be found and studied.

Species taxonomy of Hivanua is made difficult by the simplicity of the markings and genitalia, by the paucity of specimens, and by the fact that Berland’s type specimens are mostly female, harder to distinguish than males. Berland considered specimens from different islands as conspecific without good explanation. Adding to these difficulties is confusion over the geographic provenance of some specimens, mentioned under H. rufescens below.

Similar to H. tekao sp. nov., large bodied, with long appendages, especially first legs in male, and light to medium brown throughout, except for indistinct markings. Distinguished from H. tekao sp. nov. by lack of TmA (Fig. 11).

Male (based on specimen IDWM.22076). Carapace length 3.95, width 2.86; abdomen length 3.8. Carapace (Fig. 14): slightly swollen at the cheeks, as if the cheliceral muscles are strong. Medium brown, with two longitudinal thoracic bands. Clypeus medium orange-brown, with a few white setae. Chelicerae vertical, orange-brown, with only sparse setae (Fig. 13). One simple retromarginal and two promarginal teeth. Palp, like legs, uniformly brown with few setae. Patella and tibia unusually long compared to other pellenine harmochirines. Embolus thin, originating at about 7:30 (Fig. 11). Lacks TmA. Legs brown, front legs darker (medium rusty brown), back legs paler (light honey-brown). First legs especially long. First tibia with three anterior and three posterior ventral macrosetae. Abdomen indistinctly marked, with a trace of a central chevron.

Female. See Prószyński (2002).

The habitat is a “mountain ridge cloud forest” (Gillespie 2003). Although no specific notes were taken regarding the collecting methods for H. rufescens, the material listed below was bycatch of fieldwork seeking Tetragnatha and was most likely collected at night and from foliage (R. Gillespie pers. comm.).

1 male (IDWM.22076), 11 juveniles (including IDWM.22080, 23002) in EMEC with data French Polynesia: Marquesas Islands: Hiva Oa, Temetiu Ridge, 1170 m elev., 28-VI-2000, leg. R. Gillespie, G. Roderick. Gillespie (2003) reported this locality at 9.81°S, 139.08°W.

Prószyński (2002) did not provide any explanation for placing this species in Habronattus. A large and elbowed terminal apophysis (TmA) has been considered a synapomorphy of Habronattus (Maddison and Hedin 2003), but H. rufescens has no TmA (Arnedo and Gillespie 2006; Fig. 11). Nonetheless, Prószyński may have noticed some shared trait, because indeed H. rufescens is more closely related to Habronattus than to Havaika, and its congener H. tekao sp. nov. does have an elbowed TmA. The gap between the embolus and tegulum is larger in H. rufescens (Fig. 11) than in other Harmochirina lacking a TmA, as if leaving room for a TmA that was lost only recently.

There is some confusion about the geographic provenance of H. rufescens and perhaps also H. tekao sp. nov. Berland’s original description list the types of H. rufescens as from Nuku Hiva, but, as reported by Prószyński (2002), the labels with the type specimens in NHMUK (examined by D. Sherwood pers. comm.) indicate a collecting locality of Hiva Oa, 133 km to the southeast. Those specimens, studied by Prószyński (2002), do indeed appear to be the types, not only because his drawings match well Berland’s original drawings, but also because their labels seem clearly to be of the types. They read “Sandalodes rufescens Berland Type F et M”, and “1926.1.27.297-304; Hiva Oa, Marquesas Is.; C.L. Collenette 31.12.24; 3000–4000 ft.; S. Y. ‘St George’; S.E.R.A.” (D. Sherwood pers. comm.). This indicates the specimens were collected by the Scientific Expeditionary Research Association, from the ship “St George”, and formally accessioned by the NHMUK in 1926. While there could be an error in this label, it is more reasonable to respect the physical material and instead assume that Berland made an error in the publication.

Figures 11–14. 

Hivanua rufescens (Berland, 1934), specimen IDWM.22076 from Hiva Oa. 11 palp, ventral view 12 same, prolateral 13 face 14 habitus. Scale bars: 0.1 mm for palp; 1.0 mm for body.

The material collected more recently by Gillespie and others from Nuku Hiva and Hiva Oa (Arnedo and Gillespie 2006) could help resolve Berland’s confusion if it confirmed H. rufescens on Hiva Oa, except that there is unfortunately a similar contradiction between vial labels and published information for these more recent specimens. For a specimen they list as “Habronattus rufescens, Marquesas, Nuku Hiva”, Arnedo and Gillespie (2006) gave a palp photo (their figure 2Q) that can be matched to the specimen in EMEC here labelled IDWM.22076 (as indicated by the unusual dark line on the tegulum; Fig. 11). Its palp is a good match to the paratype of H. rufescens illustrated by Prószyński (2002). Thus, the specimen they reported as H. rufescens appears to be properly identified. However, that specimen and the accompanying juveniles are in vials whose labels indicate they were collected from Hiva Oa, not Nuku Hiva. Conversely, the other Marquesas male they discussed, “Habronattus sp. Marquesas, Hiva Oa” can be identified by details of setal placement (in their figure 2P) as specimen IDWM.22075 in a vial labelled as from Nuku Hiva. The DNA sequences they reported are likewise attributed to the correct species, but to the wrong islands. Those reported for H. rufescens (DQ531803 and DQ532084) are close matches to those obtained here from juveniles accompanying their H. rufescens male, while those sequences reported (DQ531801 and DQ532082) for the specimen that is here called H. tekao IDWM.22075 (their figure 2P) are closely similar to those obtained here from other specimens of H. tekao sp. nov. from Nuku Hiva. All of this is consistent with Arnedo and Gillespie attributing the palp and DNA to the correct specimens but recording their localities incorrectly. It is possible that Berland’s misreported locality for H. rufescens misled Arnedo and Gillespie to doubt and mistake the locality of their matching specimen.

I provisionally interpret the labeling of the vials to be correct for both the Berland and Gillespie specimens. The known specimens of H. rufescens are, therefore, from Hiva Oa. The male of H. rufescens that Arnedo and Gillespie showed in Fig. 2Q and whose DNA was reported as DQ531803, etc., is now labelled as specimen IDWM.22076. The known specimens of H. tekao sp. nov. are interpreted as from Nuku Hiva. The male shown in Arnedo and Gillespie’s figure 2P and whose DNA was reported as DQ531801, etc., is now labelled as specimen IDWM.22075.

Male holotype (W. Maddison voucher code d560, in BPBM), with data French Polynesia: Marquesas Islands: Nuku Hiva, Mt Tekao, 1200 m elev. 23-VI-2000, leg. R. Gillespie. Female paratype (W. Maddison voucher code IDWM.23001, in EMEC), with data French Polynesia: Marquesas Islands: Nuku Hiva, Mt Tekao, 1185 m elev., 23-VI-2000, leg. R. Gillespie, L. Shapiro. Gillespie (2003) reported the 1185 m elevation locality at 8.86°S, 140.17°W. (See comments on provenance under H. rufescens.)

Derived from the name of the type locality; treated as a noun in apposition.

Embolus accompanied by a terminal apophysis (TmA), lacking in other species of Havaika and Hivanua. The TmA is long, thin, and elbowed, and thus resembles that of Habronattus (Griswold 1987). Otherwise, similar to H. rufescens, H. tekao sp. nov. is large bodied and with long appendages, especially the first legs in the male, and light to medium brown throughout except for indistinct markings. Females differ from those of H. nigrolineata, H. flavipes, and H. triangulifera in being more robust, with abdominal markings indistinct. Females are paler than the holotype of H. nigrescens. However, given the lack of clarity of which females belong to H. tekao sp. nov., any attempt to identify them is difficult at present.

Male (based on holotype, specimen d560). Carapace length 3.96, width 2.92; abdomen length 3.70. Carapace (Fig. 17): medium to dark brown, with two longitudinal thoracic bands of paler integument and thin covering of white scales, and marginal band of sparse white scales. Remainder of carapace thinly covered in setae, some in ocular quadrangle with bronze sheen. Clypeus medium to dark brown, sparsely covered with setae, with some long pale setae overhanging chelicerae (Fig. 18). Chelicerae vertical, orange-brown, with patch of white scales basally. One simple retromarginal and two promarginal teeth. Palp, like legs, uniform brown with few setae. Patella and tibia unusually long compared to other pellenine harmochirines. Embolus thin, originating at about 7:00 (Fig. 15). TmA present, narrowing to a point, angled toward 10:30 initially, then bending (and thus elbowed) as it nears the embolus. Legs light brown, the front legs slightly darker. First legs especially long. First tibia with four anterior and three posterior ventral macrosetae. Length of femur I 3.65, II 2.29, III 2.76, IV 2.60; patella + tibia I 5.31, II 2.76, III 2.76, IV 2.71; metatarsus + tarsus I 3.54, II 2.14, III 2.66, IV 2.71. Abdomen indistinctly marked, with a trace of a central chevron.

Figures 15–26. 

Hivanua tekao sp. nov. and a specimen that may be distinct, all from Nuku Hiva 15–18 H. tekao sp. nov. holotype, specimen d560 15 palp, ventral view 16 same, prolateral 17 habitus 18 face 19–22 H. tekao sp. nov. male, specimen IDWM.22075 19 palp, ventral view 20 same, prolateral 21 habitus 22 face. 23–26 H. cf. tekao, specimen d561 23 palp, ventral view 24 same, prolateral 25 habitus 26 face. Abbreviations: TmA, terminal apophysis; e, embolus. Scale bars: 0.1 mm for palps; 1.0 mm for bodies.

Female (based on specimen IDWM.23001). Carapace length 4.01, width 3.02; abdomen length 4.90. Carapace (Fig. 27): Structure, colour as in male. Clypeus brown covered with white scales, which overhang chelicerae. Chelicerae orange-brown, with a few white scales near the base. One simple retromarginal and two promarginal teeth. Legs light brown with a few scattered white scales. First tibia with four anterior and three posterior ventral macrosetae. Length of femur I 2.45, II 2.19, III2.60, IV 2.60; patella + tibia I 3.28, II 2.60, III 2.76, IV 2.76; metatarsus + tarsus I 2.08, II 1.77, III 2.40, IV 2.60. Abdomen medium brown except that posterior third is covered with a prominent triangular white patch, with white scales. Epigyne with basic simple Habronattus or Bianor-like form, with a triangular ECP placed centrally, flanked by two crescent-shaped atria (Fig. 29). The vulva shows the spermatheca forming a compact coil, much like those of Habronattus (Fig. 30).

Three specimens can be reasonably securely considered to be H. tekao sp. nov. The male chosen as holotype, d560, closely resembles another male, IDWM.22075 in markings and palp; the female described, IDWM.23001, is placed next to the holotype in the phylogeny based on 1195 gene loci. The short branch lengths and discordance between mitochondrial and nuclear results (Fig. 1 vs. Fig. 2) would be consistent with those and all the other Mount Tekao specimens belong being a single species. However, on Mount Tekao there are two forms of males distinct enough that they might have been suspected as separate species. Male specimen d561 has longitudinal white stripes along the side of the abdomen and the embolus arising at 6:00, while males d560 and IDWM.22075 lack the stripes and have the embolus arising at about 7:00. Subadult male IDWM.22077 appears to match d561, with white stripes and 6:00 embolus origin (developed enough to see through the subadult integument), while subadult male IDWM.22078 lacks stripes and appears to have a 7:00 embolus origin (though this is unclear). The stripes could easily be polymorphic, but a difference in bulb rotation like that seen between Figs 15, 19 versus Fig. 24 would typically mark a different species, based on patterns in other groups. However, those differences do not form a clear pattern on the molecular phylogeny. From the UCE data (Fig. 1), one could suspect an unstriped less-rotated species (d560, IDWM.23001) and a striped more-rotated species (d561, IDWM.22077, 78, 79), but that would require doubting the appearance of the subadult IDWM.22078. It would also be contradicted by the mitochondrial genome, which places female IDWM.22079 instead of IDWM.23001 with the unstriped male d560. These two females appear morphologically the same, except that IDWM.23001 might have an extra coil in the spermatheca. That, however, would be against the expectations of IDWM.23001 belonging to a male with a shorter embolus. And, despite the difference in palp rotation, the striped male d561 has only one nucleotide difference with the unstriped male IDWM.22075 (Arnedo and Gillespie’s sequenced male from Nuku Hiva) in 16SND1.

Against all this confusion, I have decided to refer three specimens to H. tekao sp. nov. (male d560, male IDWM.22075, female IDWM.23001) and treat them as type material, and the remainder as possibly conspecific, naming them “H. cf. tekao”. Applying formal species delimitation methods to the UCE data might be able to help resolve it, but with so few specimens, it is prudent to wait until more specimens are available to determine if there is a second species.

The specimens from “above Toovii” are listed as “beated from ohia”. The others from Mount Tekao, including the type specimens, are not associated with specific habitat data. However, the specimens were likely on foliage. The type locality is a “high montane wet forest” (Gillespie 2003). The specimens were collected as bycatch of fieldwork seeking Tetragnatha, which was primarily at night and involved looking on foliage (R. Gillespie pers. comm.).

Figures 27–38. 

Females sequenced of Hivanua from Nuku Hiva, all H. tekao sp. nov. or a closely related species. Each shows habitatus, face, ventral view of epigyne, and dorsal view of cleared vulva 27–30 specimen IDWM.23001 31–34 specimen IDWM.22079 35–38 specimen d558. Scale bars: 1.0 mm for bodies; 0.1 mm for epigynes.

These are all identified only tentatively, as H. cf. tekao. The following are all in the EMEC, from French Polynesia: Marquesas Islands: Nuku Hiva. One male (voucher code d561), one subadult male (IDWM.22078) and 4 juveniles from Mt. Tekao, 1200 m elev., 23-VI-2000, leg. R. Gillespie. One male (IDWM.22075) and one subadult males (IDWM.22077) from Mt. Tekao, 1185 m elev., 23-VI-2000, leg. R. Gillespie, L. Shapiro. Two females (one is IDWM.22079) and one juvenile from Mt. Tekao, 1100 m elev., 24-VI-2000, leg. R. Gillespie. One female (d558) from Mt. Tekao, 1200 m elev., 25-VI-2000, leg. R. Gillespie. Two females and two juveniles from above Toovii, ~2800 ft., beaten from ohia, 18-vii-2001 Claridge.

The possibility that the specimens here described could be conspecific with one of Berland’s female holotypes should be addressed. The other holotype from Nuku Hiva, that of H. nigrolineata, is quite different, delicate bodied and with lineate markings. The male that Berland placed with H. triangulifera (not examined; location of specimen unknown) is from Nuku Hiva and could match that of H. cf. tekao shown in Figs 23–26, but there is no evidence to associate either of them with the female holotype of H. triangulifera, which is from another island, and differs from the females here considered to be H. tekao sp. nov. in having a simple clear chevron marking (the triangles of its specific epithet). The most obvious candidate for a match of H. tekao sp. nov. with a Berland species is with H. nigrescens, which, like H. tekao sp. nov., is large and robust. However, H. nigrescens is from a different and distant island, closer to Hiva Oa. The epigynes are too simple and poorly known to help. Because of the geographical distance, and to have a traceable name to which to attach the DNA data, the specimens here studied from Nuku Hiva are described as a new species.