Question Rare and threatened species are a common focus of natural area protection, but selecting sites to protect them must be balanced against other conservation objectives. Using a series of wetlands as a case study, we ask: (i) will protecting sites based on species rarity capture all critical community types; (ii) do rare plant species occur in rare environments; and (iii) will safeguarding large wetlands protect taxonomic and functional richness of rare and threatened species? Location Southern New Zealand. Methods We used lists of vascular plant species from 118 wetlands (66 fens, 35 bogs and 17 marshes). The resulting species lists included 29 rare and threatened species. Nine functional attributes of the rare and threatened species were compiled. Species assemblages were ordinated using non‐metric multidimensional scaling. Permutational multivariate ANOVA tested for a difference in assemblages between wetlands with or without rare and threatened species. Wetlands were classified according to a rare environment scheme. SLOSS (single‐large‐or‐several‐small) accumulation curves determined whether species and functional richness of rare and threatened species were best captured by groups of small or large wetlands. Results Wetlands with rare and threatened species supported species assemblages, which were different from those without rare or threatened plant species. Rare and threatened species were not associated with rare environments. The presence or richness of rare and threatened species was not associated with wetland size. SLOSS analyses revealed that small wetlands were critical for capturing rare and threatened species and their functional richness. Conclusions Prioritizing wetlands with rare and threatened plant species will not meet other conservation objectives, such as the preservation of all critical community types, rare environments or large wetlands. Networks of small wetlands will be necessary to safeguard rare and threatened plant species. Complementary objectives targeted at wetlands of varying size will be necessary to protect the full range of biodiversity and ecosystem services that wetlands offer.

Context. Significant resources are spent annually in New Zealand controlling pests to mitigate impacts on native biodiversity and agricultural production, but there are few reliable estimates of the benefits. Concerns have been expressed about inconsistent monitoring methodologies, differing frequencies and intensities of control across organisations, and poor definition of desired outcomes. Aims. To conduct and report on a survey of animal and plant pest control and monitoring by regional agencies, to identify issues with current practice and to provide advice on improvements. Methods. We surveyed 15 regional agencies in New Zealand about the pest control and associated monitoring undertaken during 2005–08. We recorded the pests targeted, the control work done and its operational details, any result and/or outcome monitoring conducted, and estimated costs. Key results. About 21% of the NZ$20 million expenditure on pest control was for monitoring. Excluding compliance (62%), monitoring changes in pest populations accounted for 31% of the total monitoring expenditure, whereas only 7% was spent measuring response in the resource that was supposedly being protected. The most common monitoring design (71%) comprised a single treatment area with no non-treatment area, in which only results were monitored. Only three programs (4%) had both treatment and non-treatment areas and both results and outcome monitoring. Conclusions. Such limited outcome monitoring constrains severely the ability of regional and local authorities to provide robust justification for their pest management activities and expenditures. Implications. Improved outcome monitoring requires better design of and additional resources for monitoring programs, improved institutional/political support for long-term programs, and better definition of long-term outcomes and objectives for pest management.

The impacts of domesticated herbivores on ecosystems that did not evolve with mammalian grazing can profoundly influence community composition and trophic interactions. Also, such impacts can occur over long time frames by altering successional vegetation trajectories. Removal of domesticated herbivores to protect native biota can therefore lead to unexpected consequences at multiple trophic levels for native and non-native species. In the eastern South Island of New Zealand large areas of seral grassland-shrubland have had livestock (sheep and cattle) removed following changes in land tenure. The longterm (> 10 years) outcomes for these communities are complex and difficult to predict: land may return to a native-dominated woody plant community or be invaded by exotic plants and mammals. We quantified direct and indirect effects of livestock removal on this ecosystem by comparing plant and invasive mammal communities at sites where grazing by livestock ceased c.10-35 years ago (conservation sites) with paired sites where pastoralism has continued to the present (pastoral sites). There was higher total native plant richness and reduced richness of exotic plants on conservation sites compared with pastoral sites. Further, there were differences in the use of conservation and pastoral sites by invasive mammals: rabbits and hedgehogs favoured sites grazed by livestock whereas house mice, brushtail possums and hares favoured conservation sites. Changes in the relative abundance of invasive mammal species after removal of domesticated livestock may compromise positive outcomes for conservation in successional plant communities with no evolutionary history of mammalian grazing.

No abstract

Context. Brushtail possums (Trichosurus vulpecula) are a major pest of native biodiversity and agricultural production in New Zealand. To maximise the effectiveness of control operations, prefeeding (free-feeding) of non-toxic bait before poison is often used, but the mechanisms by which it does so, remain unclear. One possibility is that prefeeding changes foraging patterns and space use in ways that increase the likelihood of possums finding and eating a lethal dose of poison bait.Aim. To determine whether prefeeding along transects increases possum activity on the transect, and if so, how long the effect lasts.Methods. We monitored the time that radio-collared possums spent within a few metres of 350 m of aerial wire laid along a transect.Key results. Initially, possums spent only 2.6 min per night in the vicinity of the wire, but after 20 kg of bait per kilometre were placed along the wire, that time increased 20-fold on the first night and remained high for the next four nights (by which time all of the bait had been consumed). After that there was a gradual decline in time spent near the wire over a further 18 days. The increased amount of time spent near the wire was the product of both an increase in the number of visits and the duration of those visits.Conclusions. We conclude that sowing prefeed in concentrated strips is likely to greatly increase the probability of possums rapidly encountering toxic bait sown along the same strips, especially where the toxin can be sown immediately after all of the prefeed has been eaten.Implications. Possum control operations can now be designed to apply much smaller quantities of toxic bait that will potentially reduce concerns about 1080 poisoning because of a move away from broadcast sowing operations that are intuitively disliked by many, to much more localised baiting regimes.