Mode of use of sexually dimorphic glands in a Neotropical harvestman (Arachnida: Opiliones) with paternal care

Males of Iporangaia pustulosa (Arachnida: Opiliones) have a sexually dimorphic metatarsus IV, which is thicker and with more glandular pores in males. Here we tested the hypothesis that this glandular area is used by males to leave chemicals in the environment, predicting that the animals would rub the metatarsus IV against the substrate. We have made recordings both in the field and in the laboratory, in several distinct contexts during the day and at night, comprising 67 hours of observations. We also experimentally tested the reaction of both sexes to a filter paper rubbed on the metatarsus gland, with adequate controls. We report and describe for the first time that the metatarsal gland of I. pustulosa is used to leave chemicals on the substrate by rubbing or touching it against the substrate. We also provide evidence that males can control the release of secretions of the metatarsal gland IV.

Chemicals released by such glands may have several roles, such as attracting the opposite sex and stimulating courtship (Suter and Renkes 1982;Olsson et al. 2006;Gaskett 2007;Brum et al. 2012), providing information on the size of the individual (Carazo et al. 2011), social status (Martín et al. 2007), physiological state (Ibáñez et al. 2012) and territory marking (Gosling and Roberts 2001;Khannoon et al. 2011). The behavioural reaction of males and females towards conspecific chemicals also varies: when males find chemicals of conspecific males, the responses are avoiding the site (Melville et al. 2003;Miller and Formanowicz 2011), running away or becoming aggressive (Khannoon et al. 2011). When females detect chemicals of conspecific males, they often spend more time on the spot (Wabnitz et al. 1999;Pearl et al. 2000;Steiner and Ruther 2009). Such behaviours are obviously only possible in animals that have an accurate sense of smell or taste, and that is true for some arachnids including species in the order Opiliones ).
The use of chemicals is particularly important for harvestmen. Chemoreception is used for finding back shelters (Donaldson and Grether 2007;Teng et al. 2012;Santos et al. 2013), food and predator detection Chelini et al. 2009;Costa and Willemart 2013), detection of alarm pheromones (Machado et al. 2002) and conspecific recognition (Willemart and Hebets 2012).
In the harvestman Iporangaia pustulosa Mello-Leitão 1935 (Gonyleptidae, Progonyleptoidellinae), males take care of eggs (Requena et al. 2009). A male can copulate with several females, which will add eggs to the batch the male is guarding on the inferior part of leaves (Machado et al. 2004;Requena et al. 2009). This species has a sexually dimorphic metatarsus IV that is thicker and darker in males, which bear more of such glandular pores than females. The small pores are widely spread on that specific region of the metatarsus ). The mode of use of these glands is unknown. There is no mention of use of these glandular openings during copulation (Requena and Machado 2014). It is not known if males are capable of attracting the female to themselves or the egg batch. Because some harvestmen species have been observed to drag body parts against the substrate possibly leaving chemicals (Willemart and Hebets 2012;Willemart et al. , 2009Fernandes and Willemart 2014), we tested the hypothesis that this glandular area is used by males to leave chemicals in the environment, predicting that the animals would rub the metatarsus IV against the substrate. We also tested how both sexes react to a filter paper rubbed on the metatarsus gland. Finally, we report a putative use of a gland on the tarsus I that is also absent in females.

Collection and maintenance in the laboratory
We collected the animals in Parque Estadual de Intervales, in Ribeirão Grande, South of the State of São Paulo (24°14′S, 48°04′W). We conducted the fieldwork from February to April 2012, which is the reproductive season for this species (Requena et al. 2009). For both the preliminary trials and the actual experiments we collected a total of 68 males and 72 females and brought them to the laboratory. All the animals were maintained in individual plastic containers (20 cm diameter × 8 cm height) with a retreat, a wet cotton ball and soil on the bottom. The animals were maintained at room temperature, with the natural day-night cycle, and fed once a week with moistened dog food.

Data collection
We conducted the experiments and behavioural observations in the morning, afternoon and at night. Including preliminary experiments, the total sampling times of the descriptive parts were 51 h in the laboratory and 16 h in the field. These aimed at identifying how the metatarsal gland of legs IV is used. We used a Sony Handycam HDR-XR550 (Sony Corporation, Tokyo, Japan) with 'nightshot' when necessary.

Field recordings
We recorded individuals of I. pustulosa in the morning (06.00-12.00), afternoon (12.00-16.30) and at night (19.00-05.30), between 21 February and 3 April 2012. We used focal animal sampling with continuous recording (Martin and Bateson 2011) for 10 minutes, with 84 males randomly selected. This is in addition to other recordings that did not last 10 minutes because the animals were not observable after some minutes (n = 25). Typically, all the animals were wandering or stationary on the vegetation.

Laboratory recordings
We recorded the animals in the morning (09.00-10.15), afternoon (13.30-16.00) and at night (21.00-02.00) between 21 March and 2 April 2012. A male and a female (n = 18) were introduced in the same terrarium (31 cm × 27 cm; height 35 cm) with humid soil on the bottom and a 40 cm plant in the centre. Five minutes after introducing the two individuals in the terrarium, we started recording continuously, for 30 min per pair.

Behavioural experiment
In this experiment we tested the reaction of conspecifics to a filter paper rubbed against the metatarsal gland IV (n = 9 males and 14 females). We ran the trials in the morning (07.30-10.30), afternoon (12.30-18.00) and at night (18.00-22.00), between 23 February and 24 March 2012.
The tested animal was introduced in an arena (15 × 15 × 15 cm) with paper towel on the bottom and a wet cotton ball to provide humidity. While we kept the specimen in a vial where it acclimated for five minutes, we rubbed a 1 × 1 cm filter paper against both metatarsi IV of one male. We repeated the procedure with a second piece of filter paper rubbed against the exact same region of a female. We then introduced both filter papers and a third blank one side by side, in a row, separated by approximately one cm. They were placed in front of the tested animal, at one cm from the vial. The position of the three filter papers was systematically pre-determined so that the sequence would vary between trials. We released the animal and recorded it for 15 minutes, one individual at a time.
Harvestmen usually attempt to touch a chemical stimulus source they have detected ). We predicted that more males and females would touch the filter paper rubbed against males. Individuals that did not touch the filter papers were retested until they touched at least one of the three filter papers. Each tested individual had two secretion donors (a male and a female) and each donor only gave secretions for one tested individual. Not all donors were used as subjects in the experiment. The combinations were made in a way that most individuals (males and females) were donors and subjects. Chi square was used to test for differences between treatments in the trials.

Field recordings
We have observed two ways of using the metatarsal gland IV: when gently touching the substrate, the leg would make contact with the substrate only once and then be taken off the substrate; when the leg was rubbed on the substrate, the animal would clearly drag it against the substrate. Repeatedly touching the metatarsus IV on the substrate was observed only in one male, which touched the glandular region on the margins of a leaf for 44 s and then again on the same leaf for 24 s (Figure 1). Tibia, calcaneus and tarsus also touched the leaf. Rubbing the metatarsus IV dorsoventrally was observed in two males: the first one rubbed the metatarsus IV against a twig while it was walking, three times in 1.51 s. Another male did it against a leaf margin, once in 1.8 s.

Laboratory recordings
Two males were observed dragging the metatarsus IV: while walking, the male would lean downwards to one side, extending the leg IV of the same side and dragging the metatarsal gland IV (Figure 2). Chelicerae and legs II occasionally touched the substrate. Legs I either supported the body or were not touching the substrate. One male did it five times in an interval of 4 min 18 s (mean = 3.7 s ± 2.30 s standard deviation [SD]). Another male did it nine times in an interval of 13 min 11 s (mean: 2.6 s ± 1.29 s SD). Another behavioural category observed in the laboratory was similar to one observed in the field, in which the male legs III and IV of one side would support the body while the metatarsus from the other side would gently touch the substrate. Legs I and II did not touch the substrate. This behaviour was observed once and its duration was 1 s.

Experiment: reaction of conspecifics to secretions of the metatarsus IV
None of the individuals displayed any stereotyped behaviour before or after touching the filter paper, irrespective of the stimulus on it. They just wandered in the arena   Figure 3. Number of males and females of the harvestman Iporangaia pustulosa, tested separately, that touched three 1 × 1 cm pieces of filter paper available simultaneously. The filter papers were rubbed against the sexually dimorphic proximal portion of the metatarsus IV, where males bear lots of pore glands. One piece of filter paper was rubbed against the two metatarsi of a male, the other one on the same regions of a female and the last one was blank.

Other observations
In captivity, one male was observed rubbing the tarsus IV: when walking, the male twisted the right tarsus IV from left to right, keeping it in contact with the substrate and gently dragging it at the end of the behaviour (Figure 4). Males of I. pustulosa also bear glandular openings on the tarsi I (Willemart et al. 2010), and its mode of use is unknown. In the field, we recorded a male on the inferior side of a leaf displaying rapid dorsoventral whipping movements with a leg I. Occasionally, the male displayed leg-threading, passing this same tarsus between the chelicerae and pedipalps, and then restarted the whipping movements.

Discussion
We have described how I. pustulosa uses the metatarsal gland to leave chemicals on the substrate. We also provide evidence that males can control the release of secretions of the metatarsal gland IV and report a possible use of the gland in tarsus I of males, which is also sexually dimorphic. This is the first field observation on how a sexually dimorphic gland is used in the order Opiliones. The low sample size of our observations reflects the fact that the behaviours described herein are very hard to notice and last a few seconds. This species has been studied for about 10 years in the field (Machado et al. 2004;Requena et al. 2009Requena et al. , 2012, we have recorded and analysed videos of 136 males for 27 h and these herein are the only reports of such behaviours. As these behaviours are hard to spot and because of the implications discussed below, we consider their description very relevant. In non-opilionid taxa, rubbing has also been observed, either directly on conspecifics or on the substrate. Other animals leave faeces and urine for marking. In every case, marking is somehow related to communication (Ralls 1971;Peters and Mech 1975;Atkeson and Marchinton 1982;Erlinge et al. 1982;Hutchings and Figure 4. Iporangaia pustulosa male twisting the right tarsus IV, rubbing it against the substrate (seta). White 2000; Barja et al. 2004;Jarau et al. 2004). Rubbing, pressing and dragging body parts has been observed in invertebrates (insects: Jarau et al. 2004;Bhadra et al. 2007;Roux et al. 2010) and vertebrates (amphibians: Teodecki et al. 1998;mammals: Kruuk et al. 1984;Pereira and Oliveira 2010;Mertl-Millhollen 2007). Previous papers have reported rubbing body parts against the substrate in other harvestmen species:  observed males and females of I. pustulosa and Neosadocus sp., rubbing legs III and IV on the substrate while they walk. Willemart and Hebets (2012) reported Leiobunum vittatum rubbing the lateral, ventral and dorsal regions of the body and Donaldson and Grether (2007) described Prionostemma sp. rubbing the dorsal region of the body against the substrate. Adding to these previous descriptions, here we described the twisting of tarsus IV, which can be a way of rubbing the non-sexually dimorphic tarsal aggregated pores Gainett et al. 2014), openings of glands of unknown function. However, none of these papers had reported sexually dimorphic glands being rubbed except for the very recent paper of Fernandes and Willemart (2014). The implication of rubbing sexually dimorphic glands is that harvestmen seem to leave chemicals not only for putative orientation in the environment or aggregation formation but also for sexual purposes (see Andersson 1994).
Males of some harvestmen species take care of eggs and may leave the batch unprotected in some periods of the day (Mora 1990;Hara et al. 2003;Machado et al. 2004;Proud et al. 2011). Three hypotheses have been raised to explain this behaviour: (I) males leave the batch to forage; (II) males patrol the batch at a distance as a way of repelling predators or other males; and (III) males are looking for females to copulate with (Machado et al. 2004). Based on our findings, we suggest a fourth hypothesis, which is indirectly related to the third one: males leave the batches to mark the surroundings and therefore maximize female attraction. If females can match males' scent to the scent of their secretion or if males also mark the leaf where they are (or the eggs recently laid by females if they are already guarding eggs), these could be clues to where males are located (see discussion in Wyatt 2003). Moreover, it is known that I. pustulosa females prefer males guarding eggs recently deposited by other females (Requena et al. 2012; see also Tallamy 2001). Our data allow us to suggest that female evaluation of males could also happen in the absence of the male or the eggs, since chemicals could provide information about the physiological state of the male (Rentala et al. 2002;Martín and López 2011). Finally, chemicals left by males may also repel male competitors, therefore avoiding direct fights with conspecifics for females (Martín et al. 2007;Ibáñez et al. 2012).
The non-significant results we obtained in the experiment suggest that the production of secretion by the metatarsal gland IV may be interrupted even during the reproductive season (when we collected our data). Such interruption is not related to the hour of the day since we have made observations in the morning, afternoon and at night. Males may produce chemicals only when they are going to use the secretion or they might have interrupted it because of our manipulation. Because harvestmen have contact chemoreceptors, because they have been reported to use contact chemoreception in several contexts, and because sexually dimorphic glands are typically used in intraspecific communication (Wyatt 2003;Willemart et al. 2009), we do not believe I. pustulosa would have failed to detect any chemicals if they were on the filter papers.
The behaviour of whipping legs I (where another sexually dimorphic gland is present) after passing it through the chelicerae could potentially be another form of communicating to conspecifics. Males may be diluting the secretions from the glands with water coming out of their mouth, just like they do with their defensive secretions (Gnaspini and Hara 2007). By quickly whipping the leg, it could be that secretions are dropped down to the soil, in the surroundings of the plants where the male is standing. This would be similar to what aphids do (Del-Claro and Oliveira 1996) to attract ants.
The proximal data we provide herein not only complement studies with ultimate approaches but may also help to answer ultimate questions such as why males leave their batches. We therefore hope to have helped in the understanding of this interesting system, which has contributed to the comprehension of the evolution of mating systems and parental care (Requena et al. 2009(Requena et al. , 2012.