Wehi et al.: Predictors of tree weta occupancy Predictors of relative abundance of tree weta (Hemideina thoracica) in an urban forest remnant Priscilla M. Wehi 1,3 * , Murray Jorgensen 2 and Dai K.J. Morgan 1,4 1 Department of Biological Sciences, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand 2 Department of Statistics, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand 3 Present address: Landcare Research, Private Bag 1930, Dunedin 9054, New Zealand 4 Present address: NorthTec, Private Bag 9019, Whāngarei 0148, New Zealand *Author for correspondence (Email: wehip@landcareresearch.co.nz) Published online: 15 April 2015 Abstract: Restoration of urban forest remnants is an increasing activity worldwide, but the effects of restoration efforts on local wildlife in urban remnants remain poorly understood. Understanding the beneits of restoration can also be confounded because of dificulties in monitoring the abundance of representative species, or understanding their ecological requirements. We studied tree weta (Hemideina thoracica) in an urban kahikatea (Dacrycarpus dacrydioides) forest remnant in Hamilton City: we estimated relative abundance of tree weta, and examined the relationship of weta occupancy of artiicial refuges to tree size, bark depth, distance from a centrally lit path, and distance from the edge of the remnant. Estimates of tree weta abundance were comparable with those at other mainland sites. A range of tree species supported tree weta activity, but more weta were encountered on large trees, and fewer on kahikatea relative to other tree species. Occupancy records from 40 artiicial refuges on kahikatea trees over 28 consecutive days revealed that tree stem diameter and proximity to the central path predicted occupancy on the dominant kahikatea trees. The data indicate constructed features of urban remnants, such as paths, can negatively affect habitat quality in urban forest remnants. Maintaining large trees in urban parks can provide critical habitat through the provision of natural cavities for weta. Erecting artiicial cavities on these trees to gain estimates of tree weta density can also provide indications of ecosystem recovery, for example, after pest removal. Keywords: forest fragment; indicator species; New Zealand; Orthoptera; restoration Introduction Restoration and conservation of small forest remnants is important for urban communities, despite debate about their biodiversity value (McDonnell 2007). Urban remnants can act as ecological reservoirs for species, provide habitat sites so that metacommunities are linked by dispersal (Leibold et al. 2004), and provide opportunities for people to interact with nature (Crane & Kinzig 2005; Miller 2005). However, many urban remnants are modiied or degraded, for example by large-tree felling (e.g. Whaley et al. 1997), and have characteristics such as paths and artiicial lighting that can have implications for species’ reproductive biology, foraging and behavioural responses (e.g. in nocturnal species; Navara & Nelson 2007). There is now a large body of literature indicating roads inluence dispersal and mortality patterns for a range of species (Trombulak & Frissell 2000; Laurance et al. 2004; Fahrig & Rytwinski 2009). However, the effect of paths on the distribution of invertebrates, in particular, within small remnants is largely unknown, despite the potential to mitigate effects with effective urban design during restoration. Degraded urban remnants are often the focus of restoration by community groups, whose choice of plant species for revegetation will ultimately inluence the composition of both lora and fauna, ecosystem functioning, and the future trajectory of the ecosystem (Hobbs & Norton 1996). However, restoration efforts can proceed with little monitoring of objectives, or knowledge of local species abundance and habitat (Aronson et al. 1995). Artiicial nest boxes, or refuges, are increasingly attached to trees in managed ecosystems as a conservation tool (Jackson et al. 2013), to monitor species and provide increased habitat for cavity dwelling species. Despite this, it is often unclear which plant species provide important habitat for local fauna, particularly for cryptic invertebrates. Tree wētā (Hemideina spp., Anostostomatidae) have been used as indicators to monitor the effects of habitat restoration in New Zealand (Spurr & Berben 2004; Fisher et al. 2007; Watts et al. 2011), and are part of an iconic group of endemic Orthoptera that are of high conservation value (Watts et al. 2008). Within New Zealand, the anostostomatids are represented by tree wētā Hemideina spp., giant wētā Deinacrida spp., ground wētā Hemiandrus spp., and two species of tusked wētā (Anisoura nicobarica and Motuweta isolata).The Auckland tree wētā (Hemideina thoracica) is abundant throughout most of the northern North Island. It is a lightless, large-bodied (adults 3–7 g) generalist herbivore (Wehi & Hicks 2010; Wehi et al. 2013b) that rests during the day in tree cavities before emerging to feed at night. However, despite the use of tree wētā when monitoring restoration efforts, population estimates are dificult to achieve as most tree wētā are cryptic, and it is extremely dificult to sight tree wētā in cavities during the day without destructive sampling of trees. This limitation has led to the frequent use of artiicial refuges in ecological studies of these wētā, with tree wētā quickly colonising artiicial refuges that are placed close to natural refuges (Kelly 2006a; Wehi et al. 2013a). We investigated patterns of natural refuge occupancy in the Auckland tree wētā in an urban forest remnant dominated New Zealand Journal of Ecology (2015) 39(2): 0-0 © New Zealand Ecological Society. Available online at: http://www.newzealandecology.org/nzje/