Evidence for direct oviposition into substrates by the New Zealand stick insect Spinotectarchus acornutus

ABSTRACT Stick insects (Phasmatodea) have many different oviposition strategies, reflecting a range of adaptive behaviours and morphologies to best place and secure eggs in their environments. Oviposition strategies in Aotearoa New Zealand phasmids are not well documented, but the literature so far suggests that they drop individual eggs to the ground from their position in the foliage. Here, we present evidence for an oviposition strategy unique among the Rō stick insects of Aotearoa New Zealand. Individual female Spinotectarchus acornutus were observed inserting their eggs in a range of substrates, in particular the bark of trees and in spaces within the textured surfaces of tree fern trunks. We also highlight how the specific morphology of their eggs may be an adaptation to assist in substrate attachment, while their elongated secondary ovipositor could aid in egg insertion into substrates.


Introduction
Oviposition is a crucial phase in the life histories of insects, as the location and timing of egg-laying greatly influences the survival and successful development of offspring (Refsnider and Janzen 2010;Lee and Lee 2020).The decision of where to lay is likely under strong selection, which has led to diversification of oviposition behaviours (Cury et al. 2019).Egg-laying in insects is often mediated by choice of oviposition substrate, with the choice being made based on multiple context-depended sensory cues (Almohamad et al. 2009).Once a suitable substrate has been found, female insects deposit their eggs via their ovipositors.This complex abdominal structure has diversified dramatically across insect orders, often shaped by the type of substrate upon which the egg is laid.Some insects use their ovipositors to drill (e.g.woodwasps) (Madden 1974), dig (e.g.grasshoppers) (Thompson 1986) or pierce (e.g.parasitoid wasp) (Arakawa 1987) into the chosen substrate (Cury et al. 2019).
The leaf and stick insects (Phasmatodea) are represented by around 3000 described species worldwide (Bradler and Buckley 2018).These insects are renowned for their incredible camouflage with both juveniles and adults resembling parts of plants (Shi et al. 2019), while eggs sometimes also bear a striking resemblance to plant seeds (Nakata 1961;O'Hanlon et al. 2020).Across species however, phasmids display great variation in their morphology and behaviour, and their egg morphology and oviposition strategy are similarly diverse (Carlberg 1983;Sellick 1998;Goldberg et al. 2015;Bradler and Buckley 2018;Robertson et al. 2018;O'Hanlon et al. 2020).As such, egg morphology and oviposition strategy have often historically been used as basic characters in phasmid classification (Sellick 1997a(Sellick , 1997b(Sellick , 1997c;;Bradler et al. 2014).A notable feature is the worldwide association of ants with stick insect eggs.In such cases, ants are attracted to the capitulum present on the eggs, which they then pick up and move around, acting as an agent of dispersal (Stanton et al. 2015).Detailed description of egg morphology is often given when describing a new species, and knowledge of how these eggs are oviposited is an important characteristic (Sellick 1997a(Sellick , 1997b(Sellick , 1997c)).
Oviposition in phasmids can be divided into five different strategies (based on Robertson et al. 2018): 1: Drop or flick eggs, 2: Bury/insert eggs into soil or crevices, 3: Glue eggs to substrate, 4: Pierce eggs into leaves, 5: Produce ootheca.Bradler (2009) showed that dropping eggs is the plesiomorphic oviposition trait, while inserting/burying or gluing eggs were inferred to be derived character states that have evolved convergently and repeatedly across Phasmatodea.Previous studies on the evolution of oviposition techniques have described multiple transitions from dropping eggs to burying/inserting eggs into crevices (Robertson et al. 2018).Robertson et al. (2018) estimated that a minimum of 16 shifts to inserting eggs occurred during phasmid diversification.Interestingly, an example of one of these extreme shifts occurs in a clade which includes the New Zealand taxa as well as the New Caledonian genera Trapezaspis, Microcanachus and Canachus (Robertson et al. 2018).In fact, several Lanceocercata species, the clade to which the New Zealand and most New Caledonian species belong (Buckley et al. 2010), insert their eggs into soil/crevices.For example, the Lord Howe stick insect (Dryococelus australis) lays eggs subterraneously using their elongated secondary ovipositor (Honan 2008).The secondary ovipositor is composed of the abdominal tergum 10 and operculum (Bradler 2003).Similarly, the New Caledonian species Canachus alligator also oviposits by inserting eggs into the soil (Jourdan and Delfosse 2010).
The Aotearoa New Zealand stick insect fauna, or rō or whē in the Māori language, currently contains nine genera and 23 described species (Salmon 1991;Jewell and Brock 2002;Buckley and Bradler 2010;Buckley et al. 2014).All of these genera are endemic to New Zealand, and are found from the coast to alpine regions (Dennis et al. 2014).Phylogenetic work suggests that the separation between the genus Spinotectarchus and all other species represents the deepest phylogenetic split in the New Zealand stick insect fauna (Trewick et al. 2008;Buckley et al. 2010;Dunning et al. 2013).Spinotectarchus acornutus (Hutton 1899) is a species of stick insect found in the northern parts of the North Island of New Zealand, as well as some offshore islands.Like many other New Zealand stick insects, they are commonly found on a range of native plants, including rātā (Metrosideros robusta, M. perforata), kiekie (Freycinetia banksii), mānuka (Leptospermum scoparium), kānuka (Kunzea spp), Dracophyllum species, Astelia species, Gahnia species, native Rubus species, kamahi (Weinmannia silvicola), Cyathodes species, Lophomyrtus bullata and mingimingi (Leucopogon fasciculatus) (Salmon 1991;Buckley et al. 2010;personal observation Merien 2022;New Zealand Arthropod Collection records 2022).Spinotectarchus acornutus is small and robust, with the female body measuring between 4-4.8 cm while males measure between 3.5-3.7 cm (Salmon 1991;Jewell and Brock 2002).Sexual dimorphism is apparent, with the males being thinner and shorter than their female counterparts (Salmon 1991).
In most New Zealand species, the oviposition technique is to simply let single eggs drop to the forest floor, such as occurs in Clitarchus hookeri (Stringer 1970;Salmon 1991).Salmon (1991) spent a considerable amount of time describing the eggs of New Zealand stick insect species but did not document specifically how they were laid.However, it can be inferred from his writing and the way he kept the stick insects in cages that he believed the eggs were simply dropped for all species.Salmon (1991) bred stick insect species for many years to study their biology, but only ever mentioned collecting eggs from the bottom of cages and laying them out on substrate to hatch.However, Asteliaphasma jucunda has been noted to carry its eggs for several days before depositing them (Trewick and Morgan-Richards 2014).It is unknown whether A. jucunda drop their eggs or inserts them into substrates after carrying them.
Recently, while rearing S. acornutus in captivity, we noticed that eggs were sometimes lodged in the seams or the zip of the mesh cage (Figure 1A).For another study, live adult female S. acornutus were put on bark backgrounds to photograph.During these periods, most individuals would wiggle their abdomen back and forth, inserting their ovipositor/ operculum into the crevices of the bark.It was previously assumed that S. acornutus lay their eggs by simply dropping them, like other New Zealand species.However, after these observations, we decided to investigate what type of oviposition strategy S. acornutus uses by providing various substrates for them to lay in and recording them with an infrared camera overnight.We also used scanning electron microscopy to investigate the surface morphology of the eggs and operculum in order to assess if these morphological features supported our observations of oviposition behaviour.

Methods
To investigate oviposition in S. acornutus, we observed individuals collected from the Upper Nihotupu Dam Track, Waitakere Ranges, Auckland, New Zealand (36°56'11.7"S174°33'32.5"E).We introduced adult female S. acornutus (n = 18) to mesh cage enclosures (40 cm x 40 cm x 60 cm) for a period of 10 days.After a period of acclimation to the cage, tree fern trunk bark (Cyathea smithii, Dicksonia squarrosa) was used to line one vertical side of the cage.This was done by gluing pieces of tree fern trunk to a large wooden board using non-toxic glue.The tree fern trunk substrate was allowed to completely dry before being inserted into the mesh cage.The wooden board was leaned against one side of the mesh cage, where it remained upright (Figure 1B).We used tree fern trunk as the substrate when filming the cage overnight as that is what this species are often found on in the wild and it is a common substrate for species of climbing rātā (Metrosideros robusta, M. perforata), the main host plant for S. acornutus.The population was then monitored by recording the cage using an infrared camera, XA40 Canon Camcorder between 7pm and 7am for four consecutive nights.The bark and enclosure were thoroughly checked each morning for any eggs inserted into crevices or found at the bottom of the cage.Every morning, the number of eggs laid was counted and removed from the cage.The bark substrate was also removed from the mesh cage.We define oviposition behaviour as females gyrating their abdomen while repeatedly inserting their ovipositor into substrate.
To investigate the surface appearance of S. acornutus operculum and eggs, we used an FEI Quanta 200-F field emission environmental scanning electron microscope (ESEM; FEI, Oregon, USA) operated at an electron accelerating voltage of 10 kV in the presence of water vapour in the analysis chamber (average pressure 0.6 Torr).Female S. acornutus specimens were preserved in a −20 °C freezer following natural mortality.These individuals came from the same population used for oviposition observations.Prior to analysis, the specimens were allowed to dry naturally for a few days and then the abdomen and eggs were mounted on stubs with black carbon adhesive tape.The drying technique used was natural dehydration at ambient conditions for 72 hrs.The specimens were then double platinum sputter coated using a Quorum Q150RS (Quorom Technologies, East Sussex, England) for 60 s to alleviate sample charging effects.

Observations of oviposition
Almost immediately after placing the bark substrate in the cage, female S. acornutus climbed onto it and started moving their abdomen in a circular motion while inserting their operculae into a crevice of the bark (videos available in supplementary data) (Figure 1C).This continued for a few minutes until they ceased movement, with their abdomen still inserted into the crevice.After another minute, the females straightened up and left the site of oviposition.A careful check revealed an egg in the crevice they had inserted their abdomen into.The oviposition event took around 8 min in total.Throughout the video, individuals were often seen moving from the mesh side of the cage onto the bark substrate and ovipositing into it.Whether all eggs were successfully inserted into the crevices is hard to distinguish on video, however multiple attempts were observed (Table 1).In the table, the number of bark visitations refers to females simply walking over the bark substrate.Ovipositing on bark refers to the number of times it could be clearly discerned that females were attempting oviposition through the characteristic gyrating of the abdomen and inserting their operculae into the substrate.

Location of oviposition
Eggs were found on both the vertical tree fern trunk (Figure 1D) and the bottom of the cage.It is unknown whether the eggs found at the bottom of the cage were dropped deliberately, or simply fell off the substrate due to ineffectual insertion.The substrate was visited and the operculae were inserted into it several times, significantly more than the number of eggs laid (Table 1).

Scanning electron microscopy (SEM)
Images of S. acornutus obtained using the ESEM shows the presence of mechano-sensory hairs on the surface of the operculum.The surface of the eggs are also covered in hair-like processes.

Discussion
Observations of oviposition by captive female S. acornutus demonstrate that they frequently insert eggs into crevices in the surface of fern trunks and potentially into the bark of trees or other plants.This has remained undocumented until now and highlights that this oviposition behaviour contrasts strongly with that of other rō or stick insects in Aotearoa New Zealand.The most common egg-laying strategy for stick and leaf insects both globally and in New Zealand is for females to remain in the foliage and simply drop or flick the individual eggs from their ovipositor onto the ground (Robertson et al. 2018).However, in some other phasmid species outside of New Zealand, females have been observed inserting their eggs into crevices such as bark, or even directly into soil (Sellick 1997b).For example, members of the Heteropteryginae insert their eggs into soil, with some species possessing a secondary ovipositor, an enlarged epiproct (supraanal plate) and an elongated, tapered operculum (abdominal sternum VIII) (Bradler and Buckley 2018).It appears other stick insect species such as the Malaysian moss mimic Orthostheneboea exotica also move in a similar manner to insert their eggs into moss and crevices (Youtube video link: https://www.youtube.com/watch?v=gkopGjmpAP4).
Orthostheneboea exotica drill their 'bullet shaped' eggs into the mossy substrate, one at a time and likely lay one to two eggs a day (Seow-Choen 2017).Other stick insect species such as Sipyloidea and Orxines (Necrosciinae) will also insert their eggs into bark and plant crevices (Bedford 1978;Gunning 1987;Sellick 1998).
While we found a large number of eggs found inserted in tree fern crevices, a large number were also found on the ground, and more observations of ovipositional behaviour were made than the number of eggs found.It is possible that the substrate chosen (bare tree fern bark) in this study is not the most optimum, or not the only one used for egg insertion in nature.Due to our observations of egg insertion presented here, and due to the high frequency with which S. acornutus are found on tree fern trunks, we do however assert that tree fern bark is likely a common substrate for oviposition.Potentially, the dense covering of climbing rātā that this species feeds on and which grows up the trunks of tree ferns provides more locations for egg insertion in nature, or perhaps other substrates exist.The disparity between the number of oviposition behaviour events and the total number of observed eggs (both found in the bark and at the bottom of the cage) could be to several possibilities.It is possible that these gyrating movements were used by females to probe the substrate to determine its suitability for egg insertion.As such, the behaviour itself may have only sometimes led to actual oviposition, which was not possible to determine from video footage.Females of many insect species show sophisticated mechanisms of oviposition site assessment and choice via the sensory capabilities of the ovipositor (Renwick and Chew 1994;Carrasco et al. 2015), and such probing occurs repeatedly until the best site for egg insertion is found.It is also possible that females may have attempted to oviposit but were not successful, and subsequently moved out of the camera's view and then returned later for another attempt.This would have inflated the number of oviposition behaviours witnessed.Another possibility is that S. acornutus uses both oviposition strategies of inserting eggs into substrate and dropping them passively to the ground.For example, Eurycantha calcarata will lays eggs by dropping them to the ground when not provided with suitable oviposition substrate.When provided with either soil or sand, they will oviposit by inserting their eggs into that substrate (Hsiung 1897).Further research is needed to tease these possibilities apart in S. acornutus.
The insertion of eggs in substrate may offer specific functional benefits such as access to preferred host plants and/or protection from parasites and predators (O'Hanlon et al. 2020).For example, the stick insect species Spinosipyloidea doddi adheres its eggs to the plant Alstonia muelleriana.This milkwood plant produces a white noxious chemical when exposed and may serve as a deterrent against predators (Brock and Hasenpusch 2009;O'Hanlon et al. 2020).Non-random oviposition site choice can have fitness consequences (Refsnider and Janzen 2010) and may be especially important for organisms that have limited dispersal abilities, such as robust or flightless species, and organisms in which parental care is absent (García-González and Gomendio 2003;Kiflawi et al. 2003).While currently unknown, an interesting next step would be to determine the cues female S. acornutus use to choose the location for placement of their eggs, and whether they discriminate towards specific substrate types.
The passive dropping of eggs has been hypothesised to have coevolved with stick insects' camouflage, as it allows females to oviposit without the need to move to new locations or engage in additional behaviour when ovipositing therefore maintaining their crypsis (Robertson et al. 2018).Alternatively, inserting eggs into a substrate or strategic placement on plants could be considered a form of concealment.Concealment of eggs may be selected for as an anti-predator defence for the eggs themselves and could be considered an important extension of the stick insect's own suite of camouflage strategies.Some stick insect eggs have already been noted to strongly resemble plant seeds, consistent with stick insect plant mimicry (Bedford 1978;Compton and Ware 1991;Hughes and Westoby 1992;Goldberg et al. 2015).Better knowledge of how eggs are predated on in the wild could help us to understand how this oviposition strategy has evolved, and its possible advantages.
The genus Spinotectarchus forms a separate lineage all other remaining New Zealand species and the related the New Caledonian species (Buckley et al. 2010;Dunning et al. 2013).Furthermore, egg substrate insertion is a common oviposition strategy in the New Caledonian genera.Records of the oviposition techniques of New Caledonian stick insects such as Asprenas impennis, Microcanachus matileorum, Carlius fecundus, Labidiophasma rouxi, Canachus spp and Trapezaspis spp, show that they all bury or insert their eggs into soil/crevices (Jourdan and Delfosse 2010;Delfosse 2013;Robertson et al. 2018).To accommodate this strategy, many New Caledonian species also have elongated secondary ovipositors likely adapted for inserting eggs into substrates (Buckley et al. 2009;2010).In S. acornutus, the operculum extends beyond the tip of the anal segment unlike all other New Zealand species (Jewell and Brock 2002).We hypothesise that egg insertion into substrates is a plesiomorphic character state as Spinotectarchus is the sister group to the other New Zealand stick insect genera and this character state is common in New Caledonian genera which are sister taxa to the New Zealand clade (Buckley et al. 2010;Dunning et al. 2013).We hypothesise that the egg dropping strategy re-evolved later among the New Zealand species.
The egg morphology of S. Acornutus was investigated using scanning electron microscopy (SEM), which showed that the eggs have numerous hair-like processes covering the entire surface (Figure 2B) (Salmon 1991;Jewell and Brock 2002).Hair-like spiny processes have also been described on eggs of many other stick insect species (Sellick 1997c) including egg inserting species such as Sipyloidea sipylus (Necroscinae) and the New Caledonian species Carlius fecundus (Mazzini et al. 1993;O'Hanlon et al. 2020).These 'hairs' have also been described as 'shell surface projections', 'spines' and 'setae', and exhibit variation in length and the presence of flat or hooked end of hairs (Sellick 1997c).The presence of these hairs on the eggs (either completely covering it, or only on certain parts such as the capitular/capsule collar) appears to be strongly associated with stick insect species who bury, insert or glue their eggs as an oviposition strategy (Sellick 1997c).However, 'hairy' eggs have also been described from species which are known egg droppers such as Epidares nolimetangere.This peculiar Bornean species has very hairy eggs, which the female catapults over her head, catching it with her antennae, and then burying it into the ground using her legs (Abercrombie 1992).In the case of this species, the presence of hairs on the egg capsule help capture by the antennae.However, in other phasmid species, O'Hanlon et al. (2020) proposed that these hairs function in a similar manner to plant seeds dispersed by epizoochory, whereby they adhere to feathers and fur of passing animals.In the case of S. acornutus eggs, it is possible that the presence of Velcro-like hairs helps the egg to better adhere to bark and remain inside crevices.They could also function by aiding the egg to better resemble a plant seed or conceal more effectively in the habitat, as suggested above in other phasmid species.

Conclusion
Through observations we have determined that S. acornutus females can insert their eggs into crevices as part of its oviposition strategy.More research needs to be done to determine if this is the sole oviposition strategy or if they also drop eggs.Further research should also be conducted to determine which type of substrate they prefer for inserting when ovipositing in nature, and how choice of site may play a part in their oviposition behaviour.The behaviours described here are unlike other New Zealand species of stick insects, for which we lack any evidence for egg insertion and are most likely to be egg droppers.This finding leads us to reinterpret the morphology of S. acornutus eggs as displaying possible adaptations for egg insertion.

Figure 1 .
Figure 1.A, Photo of mesh cage with eggs of S. acornutus inserted into the seams; B, Mesh cage containing 18 female S. acornutus, with the bark background on one side, and feeding plants in the middle; C, Female S. acornutus inserting its operculum into the crevice of the tree fern bark to lay her egg; D, Photos of S. acornutus eggs inserted into tree fern trunk bark.

Figure 2 .
Figure 2. A, SEM image of the female terminalia of S. acornutus; B, SEM image of a S. acornutus egg, showing the hair-like processes.

Table 1 .
Results of night observations of 18 Spinotectarchus acornutus in cage with bark substrate.