Wendy A. Ruscoe

Unexpected consequences of control: competitive vs. predator release in a four-species assemblage of invasive mammals

Invasive species are frequently the target of eradication or control programmes to mitigate their impacts. However, manipulating single species in isolation can lead to unexpected consequences for other species, with outcomes such as mesopredator release demonstrated both theoretically and empirically in vertebrate assemblages with at least two trophic levels. Less is known about the consequences of species removal in more complex assemblages where a greater number of interacting invaders increases the potential for selective species removal to result in unexpected changes in community structure. Using a replicated Before-After Control-Impact field experiment with a four-species assemblage of invasive mammals we show that species interactions in the community are dominated by competition rather than predation. There was no measurable response of two mesopredators (rats and mice) following control of the top predator (stoats), but there was competitive release of rats following removal of a herbivore (possums), and competitive release of mice following removal of rats.

CLIMATE AND NET CARBON AVAILABILITY DETERMINE TEMPORAL PATTERNS OF SEED PRODUCTION BY NOTHOFAGUS

We analyzed seed production of mountain beech (Nothofagus solandri var. cliffortioides) forest along an elevational gradient in New Zealand from 1020 to 1370 m (treeline) for the years 1973-2002. We used seed production data from nine elevations and a site- and species-specific net carbon (C) availability model from two elevations (1050 m and 1340 m) to examine how three variables (temperature, soil moisture, and net C avail- ability) during three key periods (resource priming, flowering primordia development, and flowering) influence seed production. These three strongly interrelated variables have all been considered determinants of seed production but have not previously been tested to- gether in a single analysis. Seed production increased over the 30-year period, with the greatest increases at high elevations; this increase was driven by a greater frequency of intermediate-sized seeding years. We then determined how temperature, soil moisture, and net C availability determined seeding, and examined whether temporal trends in the seeding data could be linked to similar temporal trends in temperature, soil moisture, or net C availability. High seed production was related to cool summers with high soil moisture during resource priming, warm summers during flower primordia development, and low net C availability during flowering. Positive temporal trends in temperatures during the period of flower primordia development accounted for the increase in seed production, suggesting that increasing temperatures are promoting more frequent seed production at high elevations.

A comparison of population estimates and abundance indices for house mice inhabiting beech forests in New Zealand

Populations of house mice were sampled on nine grids in Fiordland National Park between May and November 1999, using live-capture and footprint-tracking tunnel methods. Trapped mice were removed from three grids (approximately 3.24 ha each), and marked and released on the other six. Density estimates were obtained using recapture data from the grids where mice were released (Mh (jackknife) model from program CAPTURE), and rate-of-capture data from grids where mice were removed (Mbh (removal) CAPTURE model). Density estimates were used to evaluate the performance of 4 indices of mouse abundance by contrasting R2-values of their regression on estimated mouse density. Indices evaluated were: minimum number of individuals known to be alive (MNA) (total number of individual mice caught over the course of a trapping session), one-night trap catch (number of mice caught on first night of each trapping session expressed as captures per 100 trap-nights), three-night trap catch (same index estimated from number of mice caught over the first three nights), and tracking-tunnel index (proportion of nine tunnels that had mouse tracks). While MNA, one-night trap catch, and three-night trap catch were all significantly correlated with estimated density, MNA was most strongly correlated, with R2 varying between 0.67 to 0.87 depending on whether 3, 4 or 5 nights’ capture data were used. Variation in tracking-tunnel index was unrelated to mouse density on our grids.

Effect of prefeeding, sowing rate and sowing pattern on efficacy of aerial 1080 poisoning of small-mammal pests in New Zealand

Context Aerial poisoning using sodium fluoroacetate (1080) is an important but controversial technique used for large-scale control of brushtail possums (Trichosurus vulpecula) and other pests in New Zealand. The technique reliably produces near total kills of possums and rats, provided that many tens of baits (and therefore many tens of individually lethal doses) are sown for each target animal present. Aim The aim of this study was to further refine aerial 1080 poisoning by determining the effect of prefeeding, sowing rate, and sowing pattern on effectiveness. Methods Eighteen experimental treatments comprising all possible combinations of three sowing rates (1, 2, and 5 kg/ha of bait), three frequencies of non-toxic prefeed (0, 1, and 2) and two sowing patterns (parallel and cross-hatched) were applied to each of two forested areas. Treatment effectiveness was assessed from changes in the rate of interference recorded on baited cards for three species: possum, ship rat (Rattus rattus) and mouse (Mus musculus). Key results Outcomes were highly variable, ranging from increases in pest activity to near total reductions. Possum reductions were highest where one or two prefeeds were used, and at the higher sowing rates (2 or 5 kg/ha), but with some interactions between these factors. For rats, two prefeeds resulted in the highest reductions but sowing rate had no effect. For mice, post-poisoning indices were often high, indicating low effectiveness. Conclusions Some treatments were highly effective so poor kills were unlikely to have resulted from pests not encountering bait, or the bait being unpalatable. Rather they appeared to reflect sub-lethal poisoning either as a result of low acceptance (as a result of a lack of familiarity and/or satiation) or bait fragmentation. We infer that for possum and rats prefeeding helps reduce this risk of sub-lethal poisoning not only by increasing familiarity, but also (in conjunction with high sowing rates) by increasing the bait encounter rate, particularly for possums. Implications There is scope to further reduce the amount of toxic bait sown and the cost of poisoning, without compromising efficacy, by fine-tuning the balance between prefeeding and sowing rate based on which species are being targeted and, for possums, reducing bait fragmentation.

Advances in New Zealand mammalogy 1990–2000: House mouse

The house mouse in New Zealand has always been assumed to be Mus musculus L. (Murphy & Pickard 1990), because no genetic analysis had ever been done on its taxonomic status. J.A.Mahoney determined that the house mouse in Australia is M. domesticus, and that the correct authority should be Sage & Sage, 1943 (Bannister et al. 1988), because the earlier name M. domesticus Rutty 1772 is unavailable. Australian researchers have adopted this decision, e.g. Singleton & Redhead (1990). Studies of New Zealand material now in progress confirm that most house mice here are M. domesticus (or M. musculus domesticus) (P. Jamieson, unpubl. data), and when this analysis is completed and published, future references to New Zealand mice will presumably adopt the change in name here too.

Thresholds in plant–herbivore interactions: predicting plant mortality due to herbivore browse damage

Patterns of herbivore browse at small scales, such as the rate of leaf consumption or plant preferences, drive the impact of herbivores on whole-plant processes, such as growth or survival, and subsequent changes in plant population structure. However, herbivore impacts are often non-linear, highly variable and context-dependent. Understanding the effect of herbivores on plant populations therefore requires a detailed understanding of the relationships that drive small-scale processes, and how these interact to generate dynamics at larger scales. We derive a mathematical model to predict annual rates of browse-induced tree mortality. We model individual plant mortality as a result of rates of foliage production, turnover and herbivore intake, and extend the model to the population scale by allowing for between-tree variation in levels of herbivore browse. The model is configurable for any broadleaved tree species subject to vertebrate or invertebrate browse, and is designed to be parameterized from field data typically collected as part of browse damage assessments. We parameterized and tested the model using data on foliage cover and browse damage recorded on kamahi trees (Weinmannia racemosa) browsed by possums (Trichosurus vulpecula) in New Zealand forests. The model replicated observed patterns of tree mortality at 12 independent validation sites with a wide range of herbivore densities and browse damage. The model reveals two key thresholds; in plant foliar cover, indicating when individual trees may be at high risk from browse-induced mortality, and in herbivore intake, leading to high rates of mortality across the whole population.

Century-scale effects of invasive deer and rodents on the dynamics of forests growing on soils of contrasting fertility

Understanding the long-term impacts of invasive mammalian browsers and granivores in mixed forests is difficult due to the many processes potentially affecting the demography of long-lived trees. We constructed individual-based spatially explicit simulation models of two mixed conifer–angiosperm forests, growing on soils of contrasting phosphorus (P) availability, to investigate how browsing by invasive red deer (Cervus elaphus scoticus) and granivory by invasive rodents (primarily house mouse Mus musculus) might alter forest dynamics. Models were parameterized with field data. Seedling growth and survival rates were estimated inside and outside deer exclosures. Seed predation rates were estimated at high and low rodent densities in mast and non-mast seeding years. For the alluvial terrace forest, which grew on P-rich soil, our model contained 15 tree species dominated by angiosperms; our model of the P-poor marine terrace forest contained seven species dominated by conifers. The two forest models were us...

Trophic Interactions Among Native and Introduced Animal Species

In the absence of exotic mammals, the beech forest system is strongly driven bottom-up. The sporadic heavy seeding of beech trees results in a cascade of population increases in the native fauna, without any known reciprocal effects. In dryland ecosystems, the nutrient pulses occur annually during spring flushes of herbaceous plants. In both the little modified beech forest and the highly human-impacted dryland ecosystems, mammalian introductions have been made at both the herbivore and predator trophic levels. These exotic additions have created strong top-down effects on indigenous fauna because predator abundance (stoat, ferret, and cat) is driven mainly by exotic prey species (mice and rabbits). Predator numbers can reach levels not normally possible without the introduced prey, and this can potentially lead to extinction of the native fauna. The worst scenario for native prey occurs when mice and rabbit numbers fluctuate widely. This leads to acute bouts of predation caused by the increases in predator numbers (in the case of stoats), or as ferrets and cats switch to native species following sudden declines in rabbit abundance. We now know enough about some processes in beech forest and dryland ecosystems to build prototype models that will help to predict the wider effects of controlling introduced species, identify critical knowledge gaps, and ultimately guide management decisions to achieve desired biodiversity outcomes.

A new location record for kiore (Rattus exulans) on New Zealand's South Island

Abstract Thirteen kiore (Rattus exulans), also known as the Pacific or Polynesian rat, were trapped in Waitutu Forest, Southland, New Zealand in November 2002 and February 2003. This is a new mainland location for kiore, approximately 75 km south of the closest recorded extant population in the Borland Valley. Kiore remained undetected at Waitutu Forest during the first 18 months of trapping, until the population responded to tree seeding. Morphological characters of the trapped animals are compared with those of previously reported kiore and ship rats (R. rattus), to aid in the correct identification of future captures. Kiore may be more widespread in mainland New Zealand than presently known for reasons of misidentification or trapping when numbers are very low.

Bait shyness in possums induced by sublethal doses of cyanide paste bait

Cyanide paste covered in a flour-based lure is commonly used to control populations of the Australian brushtail possum, New Zealands foremost vertebrate pest. When possums were given sublethal amounts of a standard paste bait, most ate it initially, but 75% refused it 7 days later, having become bait-shy. Most possums were able to discriminate between the toxic paste and the harmless flour-based lure. A less concentrated formof the paste was less effective in killing possums, in contrast toearlier data on the lethality of cyanide. A lower sublethal dose led to a weaker and less persistent bait shyness. Further refinement is therefore required to obtain the optimum concentration that balances reduced shyness and maximum kill. Surviving possums were offered cyanide paste again after 90 and 300 days and only 13 and 17% respectively succumbed, indicating a longlived shyness, as has previously been demonstrated after sublethal poisoning with 1080 baits. The shyness did not generalize towards 1080 paste or pellet baits, or cholecalciferol gel baits. Sustained control of possums is therefore unlikely to be achieved by repeated use of cyanide paste, but changing to other types of toxic bait should provide more effective long-term control.