New Zealand has two endemic widow spiders, Latrodectus katipo Powell, 1871 and L. atritus Urquhart, 1890. Both species face many conservation threats and are actively managed. The species status of the Latrodectus spiders of New Zealand was assessed using molecular (COI, ITS1, ITS2) and morphological methods and with cross-breeding experiments. Latrodectus katipo and L. atritus were not found to be reciprocally monophyletic for any of the gene regions or morphological traits. Other than colour, which is variable, there were no morphological characters that separated the two species, which cross-bred in the laboratory and produced fertile eggsacs. Colour variation is clinal over latitude and correlates significantly with mean annual temperature. We conclude that L. atritus is a junior synonym of L. katipo. An example of introgression from the Australian species L. hasseltii Thorell, 1870 was also detected and its conservation implications are discussed.

Populations of the Australian redback spider, Latrodectus hasseltii Thorell 1870, were first recorded in New Zealand in the early 1980s and in Osaka, Japan in 1995. Reliable records suggest that naturalised populations of L. hasseltii in New Zealand are present only in Central Otago and New Plymouth. In Central Otago, L. hasseltii feeds on endangered invertebrates, such as Prodontria modesta (Broun 1909). Latrodectus hasseltii is also a hazard to the New Zealand endemic L. katipo through interbreeding and competitive displacement. CLIMEX TM was used to model the potential global distribution of L. hasseltii based on current climate, and using ArcGIS Ò 9.2, areas of suitable climate in New Zealand were overlaid with favourable habitats to identify areas most suitable for L. hasseltii establishment. In addition, shelter that urban areas offer L. hasseltii were modelled in CLIMEX and incorporated into ArcGIS to produce maps indicating cities and built up areas where the species could establish. The presence of L. hasseltii in New Zealand and Japan, and its possible spread to other areas, is of human health significance, and the species may also impact on native biodiversity.

The spider family Oxyop~~dae is revised for New Zealand. The sole species, Oxyopes gracilipes (White, 1849), occurring nationwide, is redescribed and distribution records are given. It is hypothesised that 0, gracilipes is Australian in origin.

This is the first published invertebrate survey focusing on a low-altitude shrubland community in New Zealand. Invertebrates were collected from a remnant native shrubland (450 m

The Opiliones family Neopilionidae is restricted to the terranes of the former temperate Gondwana: South America, Africa, Australia, New Caledonia and New Zealand. Despite decades of morphological study of this unique fauna, it has been difficult reconciling the classic species of the group (some described over a century ago) with recent cladistic morphological work and previous molecular work. Here we attempted to investigate the pattern and timing of diversification of Neopilionidae by sampling across the distribution range of the family and sequencing three markers commonly used in Sanger-based approaches (18S rRNA, 28S rRNA and cytochrome-c oxidase subunit I). We recovered a well-supported and stable clade including Ballarra (an Australian ballarrine) and the Enantiobuninae from South America, Australia, New Caledonia and New Zealand, but excluding Vibone (a ballarrine from South Africa). We further found a division between West and East Gondwana, with the South American Thrasychirus/Thrasychiroides always being sister group to an Australian–Zealandian (i.e. Australia + New Zealand + New Caledonia) clade. Resolution of the Australian–Zealandian taxa was analysis-dependent, but some analyses found Martensopsalis, from New Caledonia, as the sister group to an Australian–New Zealand clade. Likewise, the species from New Zealand formed a clade in some analyses, but Mangatangi often came out as a separate lineage from the remaining species. However, the Australian taxa never constituted a monophyletic group, with Ballarra always segregating from the remaining Australian species, which in turn constituted 1–3 clades, depending on the analysis. Our results identify several generic inconsistencies, including the possibility of Thrasychiroides nested within Thrasychirus, Forsteropsalis being paraphyletic with respect to Pantopsalis, and multiple lineages of Megalopsalis in Australia. In addition, the New Zealand Megalopsalis need generic reassignment: Megalopsalis triascuta will require its own genus and M. turneri is here transferred to Forsteropsalis, as Forsteropsalis turneri (Marples, 1944), comb. nov.