Two new species of harvestmen (Opiliones, Eupnoi, Neopilionidae) from Waitomo, New Zealand

Abstract Two new species of harvestman (Opiliones: Neopilionidae: Enantiobuninae) are described from the Waitomo region of the North Island, New Zealand, Forsteropsalis bona sp. n. and F. photophaga sp. n. Both have been collected within caves in the region, where predation on glow-worms Arachnocampa luminosa has been previously recorded for one or both species (misidentified as ‘Megalopsalis tumida’). However, both are regarded as troglophiles rather than strict troglobites due to the presence of specimens outside the cave systems, and the absence of troglobitic adaptations. Megalopsalis tumida (Forster, 1944) is identified as a junior synonym of Forsteropsalis fabulosa (Phillipps & Grimmett, 1932).


Introduction
The Waitomo region of the North Island, New Zealand, has become internationally renowned as a tourist attraction owing to its extensive cave systems. The primary reason for their fame is their large population of glow-worms Arachnocampa luminosa (Keroplatidae). These luminescent fly larvae construct silken nests on the roof of the cave from which they hang sticky threads to capture flying insects attracted to their light (Richards 1960). They are themselves predated upon by harvestmen (Opiliones), which are able to avoid entanglement by the glow-worms' threads and pull the larvae from their nests (Richards 1960;Meyer-Rochow and Liddle 1988). Richards (1960) recorded two species of harvestmen feeding on glow-worms at Waitomo. One, Hendea myersi cavernicola Forster, 1954 (Triaenonychidae), was originally described from Waitomo. The second species was identified by both Richards (1960) and Meyer-Rochow and Liddle (1988) as Megalopsalis tumida (Forster, 1944) (Neopilionidae). Richards (1960) recorded this species feeding only on mature glow-worm gnats; Meyer-Rochow and Liddle (1988) recorded it also feeding on pupae and late-instar larvae. Megalopsalis tumida is a junior synonym of Forsteropsalis fabulosa (Phillipps & Grimmett, 1932) (see below), and was originally described from near Wellington. Examination of specimens collected from caves in the Waitomo region revealed the presence of two species of Neopilionidae, both of them described as new below. Which of these was the species mentioned by Richards (1960) and Meyer-Rochow and Liddle (1988) is unknown. Forsteropsalis bona sp. n. is the more similar to F. fabulosa, but a photograph of 'Megalopsalis tumida' in Meyer-Rochow and Liddle (1988) may show F. photophaga sp. n. It is not impossible that the two were confused.
While Meyer-Rochow and Liddle (1988) regarded Hendea myersi cavernicola as a true troglobite, and did not find it outside the cave entrance, the collection of small numbers of 'M. tumida' outside the cave led them to regard it as troglophilic rather than troglobitic. Similar habits were inferred by Taylor (2013) for Megalopsalis suffugiens Taylor, 2013, described from caves in Western Australia. Further epigean specimens are recorded herein for Forsteropsalis bona, while specimens collected in the cave were only a short distance from the entrance. Forsteropsalis photophaga has not yet been conclusively recorded outside the caves, but troglophily is also suggested for this species by the absence of strong adaptations for troglobitism.

Methods
Specimens were sourced from the collection of Te Papa Tongarewa, Wellington, New Zealand (MONZ) or collected by hand by A. Probert and associates. Specimens collected by A. Probert will be deposited at the New Zealand Arthropod Collection, Landcare Research, Auckland, New Zealand (NZAC). All specimens are assigned to the area code WO (Waikato) by the system established by Crosby et al. (1998). Photographs and measurements were taken using a Nikon SMZ1500 stereo microscope and the NIS-Elements D 4.00.03 programme, and a Leica DM2500 compound microscope. Measurements are given in millimetres (mm). Coloration is described as in alcohol (live coloration is given in parentheses).
Comments. Females of this species are currently unknown. Forsteropsalis bona can be distinguished from most other Forsteropsalis species by its unarmed prosoma and enormous, sub-globose cheliceral segment II with widely bowed cheliceral fingers (Taylor 2011). In these features it strongly resembles F. fabulosa, and would key out to either F. fabulosa or F. tumida in the key to Forsteropsalis species provided by Taylor (2011). These two species are synonymised below. Forsteropsalis bona can be distinguished from F. fabulosa by the form of the pedipalpal patella: F. fabulosa has a distinct finger-like prodistal apophysis on the patella (Phillipps and Grimmett 1932: Fig. C p. 732), while the patellar apophysis is almost absent in F. bona (Fig. 1C). Forsteropsalis fabulosa also has denticles both dorsally and ventrally on the pedipalpal femur, while F. bona has denticles dorsally only. An interesting feature of Forsteropsalis bona is the presence of a strong ventrodistal tooth on the end of each of the proximal pseudosegments of the distitarsus. This tooth sits between the two spinose setae generally present on each tarsal pseudosegment in all Enantiobuninae (Fig. 1D). Such a feature has not previously been recorded for this subfamily, though it is also present in F. fabulosa (specimens from MONZ, details given in Taylor 2011). This may represent a distinct synapomorphy of these two species.
The glans of both Forsteropsalis fabulosa (Taylor 2011) and F. bona is relatively short compared to other Forsteropsalis species, and converges in shape on that of the Australian genus Megalopsalis (Taylor 2011(Taylor , 2013. Nevertheless, the remaining features of these two species support a direct relationship with other New Zealand species of Pantopsalis and Forsteropsalis, and with Forsteropsalis in particular. These features include dorsal papillae on the glans (Taylor 2011), setae on the mobile finger of the chelicera (absent in Megalopsalis except M. caeruleomontium; Taylor 2011Taylor , 2013, and an array of denticles on the medial side of the pedipalpal coxa ( Fig. 3A; Taylor 2011).

Forsteropsalis fabulosa (Phillipps & Grimmett, 1932)
Macropsalis fabulosa Phillipps & Grimmett, 1932: 731-733 Comments. Forsteropsalis fabulosa and F. tumida were distinguished in Taylor (2011) solely by the degree of dilation of the second cheliceral segment, with the latter supposedly more inflated in F. tumida than in F. fabulosa. A number of species of Forsteropsalis and its sister genus Pantopsalis are now known to vary in cheliceral dilation (Taylor 2004(Taylor , 2011(Taylor , 2013, and see Forsteropsalis photophaga below). Forsteropsalis fabulosa and F. tumida were both described from the Wellington district, and there seems to no longer be any justification for separating them as different species. Forsteropsalis tumida is therefore regarded herein as a junior synonym of F. fabulosa (syn. n.).  Etymology. From the Greek phos, light, and phagein, to eat, in reference to this species' predation of the glow-worm Arachnocampa luminosa.
Comments. Females of this species are currently unknown. The holotype of Forsteropsalis photophaga when first examined had a parasitic mite attached to the opisthosoma ( Fig. 2A). This mite is a representative of the Microtrombidiidae, a family that has not previously been recorded as parasitic on Opiliones; a more detailed description is currently being prepared by C. Taylor.
The genera Pantopsalis and Forsteropsalis have hitherto been regarded as well distinguished by the morphology of the cheliceral fingers (crescent-shaped in Pantopsalis vs bowed in Forsteropsalis), pedipalpal patellar apophysis (hypersetose and rounded in Pantopsalis, sparsely setose and triangular in Forsteropsalis) and penile bristle groups (shorter in Pantopsalis than in Forsteropsalis) (Taylor 2004(Taylor , 2011. The current species blurs this distinction: in its hypersetose and rounded pedipalpal apophysis it resembles Pantopsalis, but its elongate cheliceral fingers and penile bristle groups are more characteristic of Forsteropsalis. It also possesses an array of denticles on the medial side of the pedipalpal coxa as found in Forsteropsalis species (Fig. 3A; Taylor 2011). We therefore assign it to the latter genus herein. A hypersetose, rounded patella is also present in the female of F. grimmetti, though the male of that species possesses a more typical Forsteropsalis-type patella (Taylor 2011). It is possible that the hypersetose patella is in fact a symplesiomorphy of Pantopsalis and Forsteropsalis, with F. photophaga being a basal member of the latter genus.
Forsteropsalis photophaga can be readily distinguished from all other Neopilionidae in New Zealand by the hypertrophied denticle rows on the second cheliceral segment. The only other neopilionid with comparable chelicerae is the major male of the Tasmanian species Megalopsalis nigricans (Hickman 1957). This, however, is a much smaller species, with very different genital morphology and with small ozopores unlike those of any Forsteropsalis species (Hickman 1957, Taylor 2013.