Available online at www.nznaturalsciences.org.nz

Is the an obligate ?

Melissa J. Griffin1, Mary Morgan-Richards and Steve A. Trewick

Ecology Group, Institute of Natural Resources, Massey University, Private Bag 11-222, Palmerston North, 1Corresponding author’s email: [email protected]

(Received 29 October 2010, revised and accepted 11 February 2011)

Abstract

Tree weta are well known of New Zealand’s forest ecosystems; however, there is limited research into their general ecology. A literature review suggests they are mainly . tree weta, Hemideina crassidens, were given the choice of possible foods that they may come into contact with in the wild. Unexpectedly, leaves of Coprosma robusta were consumed the least by the weta, which suggests a more accurate label for the diet of H. crassidens may be omnivory. There are possible implications for the mixing of foods in the diet that need further investigating.

Key words: diet – herbivory – Hemideina crassidens – omnivory – polyphagy.

Introduction diet of one species, the mountain stone weta H. maori, using faecal analysis and Weta of the family are found plant and fragments. large bodied, flightless, orthopterans that But for the other six Hemideina species, are well known insects in New Zealand. studies have primarily been directed at Nevertheless, detailed knowledge of their other questions. For example, while ecology in the wild is limited, even to investigating the possibility of seed the extent that information of their diet dispersal by tree weta, Duthie et al. (2006) is fragmentary. One general observation found seeds of Fuchsia excorticata in the is that tree weta (Hemideina spp.) are, faeces of wild H. crassidens and identified unusually for their family, predominantly a large variety of native fruits that they herbivores (Meads 1990; Green 2005; would consume in captivity. Trewick & Morgan–Richards 2005; In the orthopteran suborder , Gibbs 2009; Wehi & Hicks 2010; which includes the Anostostomatidae, but see Barrett 1991). More precisely, there is a variety of dietary traits Hemideina appear to thrive on a diet of seen. Some taxa show extreme specialisa- leaves. Although this inference has often tion, such as the stick katydids (subfamily been repeated in the literature its basis Phasmodinae) which feed exclusively on has not been extensively tested. Wilson flowers (Rentz 1996; pp. 105). Others and Jamieson (2005) examined the show generalisation in their diets, as seen

New Zealand Natural Sciences (2011) 36: 11-19. © New Zealand Natural Sciences 12 New Zealand Natural Sciences 36 (2011) Griffin et al: Tree weta and food choice 133 in the family (Rentz 1996; plant content, whereas faeces containing ambient temperature. Each container fruit (Figure 1). Some or all of the two pp. 105). tissues from an diet are softer and had a hollowed flax (Phormium tenax) moths were eaten by 87.5% of the weta. We collated records on the diet of tree may more readily break down in the envi- flower stalk as a daytime roost for the Seeds were the least preferred, with only weta observed in captivity and in the wild ronment. This difference is evident when weta (Wyman 2009). The weta were two of the 32 weta consuming some part (Table 1). From this, it is apparent that large, firm plant-based droppings of tree given seven days to acclimate to this of the seeds. The seeds were not eaten plants are indeed an important compo- and giant weta are compared to the soft environment during which time they whole, rather they were gnawed and the nent of tree weta diet. Leaves generally liquid droppings of carnivorous ground were given a maintenance diet of carrot kernel inside eaten. No weta ate only dominate the food types eaten by individ- weta and tusked weta (Gibbs 1998; pers. (Wyman et al. 2010), which was removed leaves and leaves were recorded as being uals of four species of weta (H. crassidens, obs. SAT). Method bias could easily at the start of the choice experiment. eaten less often (9/32) than other food H. femorata, H. maori and H. ricta; Table result in only the plant cuticles being Leaves and fruits used were collected fresh types provided (moths + fruit + seeds = 1). However, availability of information identified and other particles from faeces from Coprosma robusta in Palmerston 55; t test P = 0.016). The proportion of is variable and almost certainly incom- being effectively ignored, so the full diet is North, a plant known to be palatable to instances of adults and juvenile eating a plete. For instance, published records for not observed. In particular, foods that are tree weta (Table 1, Pers. Obs.). The leaves combination of these foods were 82.6% one of the most well known species, the rare but perhaps nutritionally important of Coprosma robusta were approximately and 87.8% respectively. Most (18/32) Auckland tree weta H. thoracica, consists will tend to be overlooked. 85 mm x 40 mm. The fruits of C. robusta weta ate two food items, five ate three, of seed predation, cannibalism in captiv- In light of recent speculation about are elliptical and orange, approximately one ate four items and eight ate one. ity, two instances of leaf eating and one specialised interactions between weta and 8mm x 5mm and contain two seeds of a There was no difference in the average instance of fruit eating. The majority of various plants (Duthie et al. 2006; Burns size that are too large (4.2 - 6.5mm x 2 - number of different food types eaten by published observations are of captive 2006; Morgan-Richards et al. 2008) we 3mm; Webb & Simpson 2001) for tree males (2.055) and females (2.085) (t test weta and include the foods eaten under asked the more general question: does weta to ingest whole. All fruit were ripe P = 0.43). Moths and fruit were the food experimental conditions (e.g. food choice the tree weta, Hemideina crassidens show and intact when given to the weta. The types most commonly chosen by males trials). As tree weta are generally treated a pronounced choice for particular food relatively large seeds of the fruits of this and females, adults and juveniles (Figure as herbivorous and thus usually provided types? If H. crassidens is a herbivore, we species enabled a distinction to be made 1). Although there appears to be a subtle only with leaves, captive studies may not predict that under experimental condi- between the consumption of fruit pulp difference in the food choice of adults be informative about food preference. tions, individual weta will choose leaves and predation on the seeds. compared with juveniles (Figure 1), this Field observations of natural diet include more often than other food types avail- The choice experiment was run over did not result in a change in the ranking nocturnal observations of weta eating able. In contrast, if H. crassidens is poly- two consecutive nights using wild-caught of leaves over other foods, or a change in and identification of food particles in the phagous then we expect to see each weta weta. Each weta was supplied with three the number of food types eaten per weta. faeces collected in the wild. eating a number of different food types leaves, five fruit (approximately 1g) Direct field observations provide valu- when given a choice. and two freeze-killed Wiseana moths Discussion able information but their significance is (approximately 0.5g) at the start of the difficult to quantify when there are so few Methods experiment. This volume of each food Tree weta are commonly known to eat data. Similarly, determining diet from the type meant that weta consumption was vegetation (Table 1). The capacity to remains of food components in faeces also We tested whether tree weta demonstrate not limited. There were no other foods consume leaves and develop successfully has intrinsic problems as digestion differs any partiality when offered a range in the container. Weta were checked after on a purely plant diet (Morgan-Richards among food types (Fitzgerald 1976; Tre- of food types they might encounter the first night and the total of each food 2000), appears to be a specialist trait of wick 1996). Soft tissues from fruit flesh in nature: leaves, insects, fruits and type eaten was recorded on the second this genus and the sister genus Deinacrida may not be obvious in weta droppings so seeds. 32 wild Wellington tree weta, H. morning. (Pratt et al. 2008) and leads to the quantification of fruit eating may come crassidens, were collected from southern expectation that tree weta will prefer only from the presence of seeds in the North Island. These weta included males Results plant foods when given the choice. This faeces and then only when seeds are small and females, adults and juveniles. The expectation is born out of observations enough to be ingested whole. Invertebrate weta were kept individually in two-litre The experiment recorded feeding of 32 of tree weta feeding in the wild and remains in faeces are often difficult to plastic containers with lids fitted with fine Hemideina crassidens tree weta; 11 adults the consistency and content of their distinguish and identify, even to a family mesh to allow light and air circulation (6 females and 5 males) and 21 juveniles droppings. However, in our captive level (Little 1980). Large droppings that (Wyman et al. 2010). They were kept (13 females and 8 males). The most experiment we found Wellington tree persist in the environment have a high in a quiet room with natural light and frequently eaten foods were moths and weta, H. crassidens, did not exhibit a 14 New Zealand Natural Sciences 36 (2011) Griffin et al: Tree weta and food choice 15 Scrophularaceae Hebe sp. crassidens3, maori3 Table 1. Current knowledge of the diet of tree weta, Hemideina spp. Species of weta are: Buddliea sp. crassidens3 Hemideina crassidens, H. thoracica, H. femorata, H. maori, H. ricta. Captivity is any food the weta Salicaceae Salix sp. crassidens3, maori3 ate during experimentation or while in any other captive environment. Field relates to any food Thymelaeaceae Kelleria villosa maori7 that the weta were seen eating in the wild or food particles found in faeces collected in the field. Violaceae crassidens1,4, 11 Winteraceae Pseudowintera colorata thoracica3

Food type Captivity Field Fruit – Angiosperms Agavaceae Cordyline australis crassidens2 Leaves - Gymnosperms Argophylaceae Corokia cotoneaster crassidens2 8 Pinaceae Pinus radiata crassidens * Balanophoraceae Dactylanthus taylorii crassidens8 10 10 Podocarpaceae Podocarpus nivalis maori maori Campanulaceae Pratia angulata crassidens2 Pratia physaloides crassidens2 Leaves - Angiosperms Cornaceae Griselinea littoralis crassidens2 7 Apiaceae Aniosotome imbricate maori Ericaceae Gautheria antipoda crassidens2 9 9 Apocynaceae Parsonia heterophylla ricta , femorata Lauraceae Beilschmiedia tawa crassidens2 3,4, 11 Araliaceae Pseudopanax arboreus crassidens Liliaceae Dianella nigra crassidens2 9 9 Pseudopanax colorata ricta , femorata Monimiaceae Hedycarya arborea crassidens2 9 9 Schefflera digitata ricta , femorata Myoporuaceae crassidens2 7 Asteraceae Celmisia viscose maori Onagracae Fuchsia excorticata crassidens1,2, thoracica1 crassidens2 10 10 Helichysum selago maori maori Fuchsia procumbens crassidens2 7 Raoulia hectori maori Rubiaceae Coprosma sp. crassidens2 3 9 Sonchus oleraceus crassidens , ricta , Violaceae Melicytus sp. crassidens2 femorata9 9 9 Taraxacum officinale ricta , femorata Seeds – Gymnosperms 3 Celastraceae Euonymus sp. crassidens Araucariaceae Agathis australis thoracica5 thoracica5 1 Coriaraceae Coriaria arborea crassidens Balanophoraceae Dactylanthus taylorii crassidens8 1,4 9 Cornaceae Griselinia littoralis crassidens , ricta , Podocarpaceae Dacrydium cupressinum thoracica6 femorata9,10 11 Corynocarpaceae Corynocapus laevigatus crassidens Flowers 3 Fabaceae Sophora sp. crassidens Asteraceae Brachyglottis repanda crassidens8 Trifolium repens ricta9, femorata9 9 9 Ulex europeaus ricta , femorata Other weta crassidens8, thoracica8 3 Malvaceae Hoheria sp. crassidens 3,11 Myoporaceae Myoporum laetum crassidens , crassidens3,8, maori3, maori7,10, 3 3 thoracica , maori thoracica3,8 femorata10 Myrtaceae Eucalyptus sp. crassidens1 Kunzea ericorides crassidens3, ricta9, femorata10 femorata9,10 1Wyman et al. 2010; Wyman 2009; 2Duthie et al. 2006; 3Barrett 1991; 4Fisher et al. 2007; Leptosermum scoparium maori10, femorata10 5Mirams 1957; 6Beveridge 1964; 7Wilson & Jamieson 2005; 8pers. obs., Morgan-Richards & Metrosideros sp. crassidens1 Trewick; 9Townsend 1995; 10 Little 1980; 11Rufaut 2001; 12Moller 1985. Nothofagaceae Nothofagus solandri femorata10 femorata10 Onagraceae Fuchsia excortica crassidens1 *Consumption of pine needles (P. radiata) was determined through pine cuticle found in the Piperaceae Macropiper excelsum ricta9, femorata9, crassidens12 droppings of H. crassidens collected in the field. crassidens11 Pittosporaceae Pittosporum eugenioides ricta9, femorata9 All these species of weta have also been seen to eat carrot in captivity (pers. obs., Morgan- Plantaginaceae Plantago sp. crassidens3, maori3 Richards & Trewick). Poaceae Poa colensoi femorata10 maori7 Polygonaceae Rumex obtusifolius ricta9, femorata9 Polytrichaceae Polytrichem juniperinum maori7 Rubiaceae Coprosma foetidissima crassidens11 Coprosma repens. crassidens3, 11, maori3, thoracica3 Coprosma rhamnoides femorata10 Coprosma robusta crassidens3, maori3, thoracica3 16 New Zealand Natural Sciences 36 (2011) Griffin et al: Tree weta and food choice 17

supplement plant diet by eating other optima by utilising a wide range of food insects (Brown 1983). For instance, crop types (Raubenheimer & Simpson 1993). contents of individual shield-backed ka- tydids of the genus included References plant and invertebrate remains (Gang- were 1967). Although there are many Barrett, P. (1991) Article 1. In: Keeping examples of herbivory in orthopterans weta in captivity. A series of Nine Articles (e.g., Acrididae), polyphagy is also com- for Schools and Natue-lovers (ed. G.W. mon and extends from eating a mixture Ramsey), pp. 10 - 17. Wellington of different plant tissues and species, to Zoological Gardens, New Zealand. omnivory that includes plant and animal Bernays, E.A. & Minkenberg, O.P.J.M. matter. This contrasts with the situation (1997). Insect herbivores: different of insects in general, which tend to have reasons for being a generalist. Ecology narrow dietary range (Hodkinson & 78: 1157 - 1169. Hughes 1982). Close host-plant interac- Beveridge, A.E. (1964). Dispersal and tions and coevolution often occur with destruction of seed in central North those that feed on a single plant species Island Podocarp forests. Proceedings (Ehrlich & Raven 1964). Orthopterans, of the New Zealand Ecological Society which are highly mobile as juveniles 11: 48 - 56. Figure 1. The percentage of the individual weta that ate each of the food types, grouped into and adults, may rely on food selection Brown, V.K. (1983). Naturalist’s Handbook males and females, adults and juveniles and overall the weta. to maintain homeostasis, regulate food 2: Grasshoppers. Cambridge University selection to balance nutrient/toxin intake Press, Cambridge. preference for leaves over the other food The occurrence of folivory in individu- (Raubenheimer & Simpson 2003; Jonas Burns, K.C. (2006). Weta and the types offered. In this experiment, seeds of al tree (Hemideina) and giant (Deinacrida) & Joern 2008; Hunter 2009), and it has evolution of fleshy fruits in New Coprosma robusta were the least preferred weta distinguishes them from most other been shown than diet mixing enhances Zealand. New Zealand Journal of food when given a choice of leaves, fruit, members of the Anostostomatidae, which development rates (Bernays & Minken- Ecology 30: 405 - 406. seeds and invertebrates. Two adult weta appear to be predominantly carnivorous berg 1997). Cary, P.R.L. (1983). Diet of the appeared to recognise the food potential (e.g. Cary 1983; but see New Zealand’s tree weta (Hemideina) ground weta Zealandosandrus gracilis of seeds and demonstrated an ability to Morgan-Richards et al. 2008). However, and closely related giant weta (Deina- (: Stenopelmatidae). New access the kernel of C. robusta seeds, even if not essential in the diet, carnivory crida) (Morgan-Richards & Gibbs 2001; Zealand Journal of Zoology 10: 295 - further demonstrating a capacity for appears to be important and might have Trewick & Morgan-Richards 2004) are 298. seed predation by tree weta (Table 1). In strong implications for growth rates and an ecologically prominent and diverse Crawley, M.J. (1983). Herbivory: The our experiment, tree weta that ate moths fecundity of individual weta. Carnivory group (Trewick & Morgan-Richards dynamics of animal-plant interactions. also tended to eat the fruit, but no weta may be important in maximising fitness, 2005; Trewick & Morgan-Richards (eds. D.J. Anderson, P. Greig-Smith, ate only leaves. This is inconsistent with by enabling the development of enlarged 2009). The success of this endemic line- F.A. Pitelka). Blackwell Scientific, the original inference that tree weta are heads in males that may be important in age may be related to the evolution of London. obligate herbivores, instead indicating securing mates (Kelly 2005; GW Gibbs polyphagy (folivory and carnivory), more Duthie, C., Gibbs, G. & Burns, K.C. an omnivorous or polyphagous habit, pers. comm.), and enhancing egg number akin to that of Tettigoniidae than other (2006). Seed dispersal by weta. Science but is in keeping with other observations and/or quality in females. Anostostomatidae. As there is limited di- 311: 1575. of carnivory amongst anostostomatids The feeding habits of other Orthoptera versity of Tettigoniidae in New Zealand, Eades, D.C., Otte, D., Cigliano, M.M. & (Little 1980; Barrett 1991; Wilson & are diverse, although many are herbivores; (three native species; Eades et al. 2010) Braun, H. (2010). Orthoptera species Jamieson 2005). The majority (24/32) of eating living plant tissues (Crawley 1983). the Hemideina/Deinacrida clade may have file. Retrieved 19 August 2010. http:// weta in our experiment ate two or more For example, shorthorn grasshoppers radiated into unoccupied niche space in orthoptera.speciesfile.org/HomePage. food types over just two nights, which (Acrididae) are obligate herbivores that New Zealand’s forests. We expect that aspx. further demonstrates a polyphagous specialise on grasses, while longhorn further, detailed studies of these weta will Ehrlich, P.R. & Raven, P.H. (1964). habit. grasshoppers or katydids (Tettigoniidae) reveal a capacity for targeting nutrient Butterflies and plants: a study in 18 New Zealand Natural Sciences 36 (2011) Griffin et al: Tree weta and food choice 19

coevolution. Evolution 18: 586 - 608. by a mixed-feeding insect herbivore, 108 - 112. Trewick, S.A. & Morgan-Richards, M. Fisher, P., Spur, E.B., Oglivie, S.C. & Melanoplus bivittatus (Orthoptera: Pratt, R.C., Morgan-Richards, M. & (2005). After the deluge: mitochondrial Eason, C.T. (2007). Bait consumption Acrididae). Ecological Entomology 33: Trewick, S.A. (2008). Diversification DNA indicates Miocene radiation and and residual concentrations of 546 - 554. of New Zealand weta (Orthoptera: Pliocence adaptation of tree and giant diphacinone in the Wellington Kelly, C.D. (2005). Allometry and Ensifera: Anostostomatidae) and weta (Orthoptera: Anostostomatidae). tree weta (Hemideina crassidens) sexual selection of male weaponry their relationships in Australasia. Journal of Biogeography 32: 295 - 309. (Orthoptera: Anostostomatidae). in Wellington tree weta, Hemideina Philosophical Transactions of the Royal Trewick, S.A. & Morgan-Richards, New Zealand Journal of Ecology 31: crassidens. Behavioural Ecology 16: Society B: Biological Sciences 363: 3427 M. (2009). New Zealand Biology. 104 - 110. 145 - 152. - 3437. In: Encyclopedia of Islands (eds. RG Fitzgerald, A.E. (1976). Diet of opossum Little, G.A. (1980). Food consumption Raubenheimer, D. & Simpson, S.J. 1993. Gillespie & DA Clague) pp 665 - Trichosurus vulpecula (kerr) in the and utilisation in two speces of weta The geometry of compensatory feeding 673. University of California Press, Orongorongo Valley, Wellington, (Hemideina femorata and H. maori: in the locust. Animal Behaviour 45: Berkeley. New Zealand, in relation to food-plant Stenopelmatidae). Unpublished 953 - 964. Webb, C.J. & Simpson, M.J.A. (2001). availability. New Zealand Journal of BSc (Hons) thesis. University of Raubenheimer, D., Simpson, S.J. (2003). Seeds of New Zealand gymnosperms Zoology 3:399 - 419. Canterbury, New Zealand. Nutrient balancing in grasshoppers: and dicotyledons. Manuka Press, Gangwere, S.K. (1967). The feeding Meads, M. (1990). The weta book: guide behavioural and physiological Christchurch. behaviour of Atlanticus testaceus to the identification of weta. DSIR Land correlates of dietary breadth. Journal Wehi, P.M. & Hicks B.J. (2010). Isotopic (Orthoptera: Tettigoniidae). Annuals Resources, Lower Hutt. of Experimental Biology 206: 1669 - fractionation in a large herbivorous of the Entomological Society of America Mirams, R.V. (1957). Aspects of the 1681. insect, the Auckland tree weta. Journal 60: 74 - 81. natural regeneration of the Kauri Rentz, D.C. (1996). Grasshopper country: of Insect Physiology 56: 1877 - 18882 Gibbs, G.W. (2009). The end of an (Agathis australis Salisb.). Transactions the abundant orthopteroid insects of Wilson, C.G. & Jamieson, I.G. (2005). 80-milion year experiment: a review of the Royal Society of New Zealand 84: Australia. University of New South Does melanism influence the diet of the of evidence describing the impact of 661 - 680. Wales Press: Sydney. mountain stone weta introduced rodents on New Zealand’s Moller, H. (1985). Tree wetas Rufaut, C.G. (1995). A comparative study (Orthoptera: Anostostomatidae)? “mammal-free” invertebrate fauna. (Hemideina crassicruris) (Orthoptera: of Wellington tree weta, Hemideina New Zealand Journal of Ecology 29: Biological Invasions 11: 1587 - 1593. Stenopelmatidae) of Stephans Island, crassidens (Blanchard, 1851) in the 149 - 152. Gibbs, G.W. (1998). Ruakumara Cook Strait. New Zealand Journal of presence and absence of rodents. Wyman, T.E. (2009). Tree weta as tusked weta: Discovery, ecology Zoology 12: 55 - 69. Unpublished MSc thesis. Victoria herbivores and seed dispersers and management implications. Morgan-Richards, M. (2000). University of Wellington. (Hemideina crassidens, H. thoracica). Conservation Advisory Science Notes. Robertsonian translocations and Townsend, J.A. (1995). Distribution and Unpublished BSc (Hons) Thesis. Department of Conservation, B chromosomes in the Wellington ecology of the Banks Peninsula tree Massey University, Palmerston North, Wellington. tree weta, Hemideina crassidens weta, . Unpublished New Zealand. Green, C. (2005). Using artificial refuges (Orthoptera: Anostostomatidae). MSc Thesis. Massey University, Wyman, T.E., Trewick, S.A., Morgan- to translocate and establish Auckland Hereditas 132: 49 - 52. Palmerston North, New Zealand. Richards, M. & Noble, A.D.L. (2010). tree weta Hemieina thoracica on Morgan-Richards, M. & Gibbs, G.W. Trewick, S.A. (1996). The diet of kakapo Mutualism or opportunism? Tree Korapuki island, New Zealand. (2001). A phylogenetic analysis of (Strigops habroptilus), takahe (Porphyrio fuchsia (Fuchsia excorticate) and tree Conservation Evidence 2: 94 - 95. New Zealand giant and tree weta mantelli) and pukeko (P. porphyrio weta (Hemideina) interactions. Austral Hodkinson, I.D. & Hughes, M.K. (Orthoptera: Anosostoatidae: melanotus) studied using faecal analysis. Ecology 36: 261-268. (1982). Insect herbivory. Chapman and Deinacrida and Hemideina) using Notornis 43: 79 - 84. Hall Ltd, London. morphological and genetics characters. Trewick, S.A. & Morgan-Richards, Hunter, M.D. (2009). Trophic Invertebrate 15: 1 - 12. M. (2004). Phylogenetics of New promiscuity, intraguild predation and Morgan-Richards, M., Trewick, S.A. Zealand’s tree, giant and tusked weta the problem of omnivores. Agricultural & Dunavan, S. (2008). When is it (Orthoptera: Anostostomatidae): and Forest Enotmology 11: 125 - 131. coevolution? The case of ground weta evidence from mitochondrial DNA. Jonas, J.L. & Joern, A. (2008). Host-plant and fleshy fruits in New Zealand. Journal of Orthoptera Research 13: quality alters grass/forb consumption New Zealand Journal of Ecology 32: 185 - 196.