Issue 1, July 2010
Invasive Mammal Impacts (IMI) is a collaborative research programme looking at invasive mammal impacts on biodiversity.
In this issue:
Welcome to Newsletter No. 1 of the FRST-funded programme on invasive mammal impacts on biodiversity (IMI). IMI is a collaborative research programme involving Landcare Research, the New Zealand Department of Conservation (DOC), several New Zealand and Australian universities, regional councils, and community conservation groups throughout New Zealand. This bumper issue provides some research highlights and describes some research initiatives for the period October 2009 – June 2010. In the next issue, we’ll show you how all this work is starting to come together.
The ecological research that will be done within the IMI programme is aimed at helping conservation managers decide where and when to apply pest control. We hope this research will also provide important information to help extend biodiversity benefits beyond pest-control areas and ultimately link sites across the NZ landscape. This requires an understanding of the correct timing of pest control, as well as spatially-targeted control, within and sometimes between sites. We also investigate how best to manage multiple invasive species, including rodents (rats, mice), mustelids (ferrets, stoats), feral cats, possums, rabbits, and hedgehogs.
Our scientific approach is to use ‘management scale’ experimental manipulations of invasive mammal populations to help understand how controlling one or more pest species affects the abundance of other pests, and also affects pest impacts on native species, in and around controlled areas. By ‘management scale’ we mean that, where possible, our research is conducted at sites were DOC, community groups, and regional councils are already doing pest control.
Important aspects of the IMI programme are:
- Understanding the dynamics and trophic interactions among multiple pest species, and native fauna, in high-priority ecosystems.
- Measuring responses of mesopredators (e.g. rats and mice) to the control of top predators (e.g. stoats and feral cats) or competitors (e.g. possums).
- Monitoring movements of animals, particularly incursions and reinvasions into control areas.
- Quantifying the impacts of invasive mammals on native flora and fauna.
- Determining the spillover benefits of site-based pest mammal control, for example through increased dispersal of birds between protected sites.
- Developing computer simulation models of invasive mammal behaviours, interactions, and impacts on native biodiversity.
The programme is divided into four research strands: Forests (leader: Wendy Ruscoe), Drylands (leader: Grant Norbury), Global Change (leader: Andrea Byrom), and Theory (leader: Roger Pech). The Forests, Drylands, and Global Change strands involve major field-based projects, while the Theory strand provides the theoretical and conceptual ‘glue’ to ensure synthesis and overview of all the research projects in the programme.
Researchers in the IMI programme have linked up with DOC staff and the new Operation Kaka Project in the Tararua Ranges, north of Wellington in the lower North Island. DOC will implement a pest control operation in winter 2010 and has established a biodiversity monitoring scheme. In addition to this, Landcare Research is setting up complementary biodiversity monitoring to test whether the actual area protected by pest control is larger or smaller than the area poisoned (a ‘core effect’ or a ‘halo effect’) and whether the size of the protected area varies for different native species such as weta (an iconic invertebrate), various native birds, and native vegetation. Hannah Windley (PhD candidate with Bill Foley, Australian National University) visited the Tararuas recently with Mandy Barron and postdoctoral fellow Pen Holland, and has taken samples of kāmahi leaves. Hannah’s project aims to describe the variation in palatability of vegetation eaten by brushtail possums in Australia and New Zealand, and hopefully help identify chemical compounds within leaves of New Zealand tree species (starting with kāmahi) that alter their palatability to possums.
Wendy Ruscoe and Al Glen (postdoctoral fellow) have visited Hawke’s Bay (eastern North Island) and Northland (in the subtropical north!) to meet with regional council staff and community conservation groups to explore the potential for monitoring the benefits for native biota of wide-scale pest control being done in these areas. What do we mean by ‘wide-scale predator control’? Increasingly, many councils and community groups are adopting a ‘can do’ attitude to achieving biodiversity conservation across whole landscapes, regardless of land tenure or habitat type. This means controlling invasive mammals across really big areas – in Hawke’s Bay, for example, possum control is conducted over more than 400 000 ha of bush remnants, conservation areas, and farmland. Imagine: adding predators – rats and stoats – to the suite of species controlled could mean ‘a kiwi in everyone’s back yard’. Research will focus on the potential for wide-scale predator control to increase the survival and connectivity of native species within remnant patches of bush in what has become a complex fragmented landscape dominated by agriculture. Tom Etherington (PhD candidate with Mick Clout, University of Auckland) has submitted a proposal to work within this programme to model how the distribution of favourable habitat affects the risk of possums reinvading controlled areas (see below).
The impact of herbivory by invasive mammals on forest composition is of interest at large spatial scales (local sites, regions, and even countries). Herbivory impacts can result in multiple trees dying as a consequence of foliage consumption. Pen Holland has been developing a mechanistic model of herbivores feeding on leaves across a tree canopy, to predict browse-induced mortality of individual trees, using browsing of the introduced brushtail possum on kāmahi as a model system. The model scales up to the site level for spatially heterogeneous browse across different-sized trees with equal palatability. Pen’s model has been parameterised using independent data on the lifespan, area, and even weight of individual leaves, as well as information about individual trees such as unbrowsed foliage density and trunk diameter. The model has been validated against statistical relationships derived by Richard Duncan and colleagues from field data on kāmahi (browse and foliar cover indices, tree mortality). The main prediction of the model for pest management is the existence of a threshold consumption rate: if foliage removal from an individual tree exceeds this threshold, the tree is unable to maintain a positive foliage balance and will eventually die. However, the availability and palatability of foliage and other food sources in the local area drives consumption (and hence the ultimate fate) of individual trees, and results in a feedback loop on the fitness and subsequent population growth of the herbivore population. Already, Pen’s modelling efforts are being used to improve possum control by DOC.
Wendy Ruscoe will visit Hawai’i in August on a NZ–Hawaii conservation exchange programme. While there, Wendy will focus particularly on predatory rodents: measuring their impacts, interactions with other pests including herbivores, and quantifying the effects of rodent control on native biodiversity. She will be hosted by Tadashi Fukami from the University of Hawai’i and Stanford University with USDA, who is using 13C and 15N isotope analysis to determine food chain length, trophic niche width and functional diversity in Hawaiian forest ecosystems. Wendy will also visit several other organisations including the National Park Service and the United States Geological Survey Pacific Island Ecosystems Research Center, and she will deliver oral papers at the Hawaiian Conservation Alliance Conference and the Hawai’i Invasives Conference in Honolulu. We look forward to hearing Wendy’s report on her trip in the next issue of the IMI newsletter!
Mandy Barron, with Graeme Elliott and Josh Kemp from DOC, is developing a spatial model for rat populations. The aim is to encapsulate in the model changes in rat populations in response to fluctuations in food resources (such as beech seed mast years), and then use the model to predict the optimal timing of rat control operations to limit overall rat abundance. The model includes rat dispersal and will be used to predict invasion rates of rats back into controlled areas. The aim is to give managers an estimate of how long before rats recover to pre-control levels and thus when more control is necessary.
Up until now, we’ve known very little about mouse ecology in dryland ecosystems. As for forest ecosystems, we have found that mice are primarily driven ‘bottom-up’ by grass seed and invertebrate abundance. This is especially pronounced in tussock ecosystems during years of peak seed production (‘mast’ years) when mouse numbers reach peak densities. However, mice are sometimes driven ‘top-down’ by top-order predators (cats, ferrets, stoats). For example, in dry inland basins dominated by non-tussock grasses mice respond positively to predator control. This may explain the fact that we found no increases in the number of native lizards following predator control – the increase in mouse predation may have offset the decline in top-predator predation. In contrast, we did detect positive responses of lizards to predator control in tussock-dominated ecosystems – and here mice did not respond to predator control. These results suggest that food webs may be structured quite differently depending on the type of grassland system. This can result in very different biodiversity outcomes following control of top predators – a possible ‘mesopredator release’ effect.
Our research is progressing well with end-users, students and collaborators. Liz Rayner, a Masters student investigating the feeding behaviour of wild rabbits (co-supervised by Phil Seddon, University of Otago), has produced a draft of her thesis and it is nearly ready for submission. Her research will be used directly in a dryland ecosystem model that Dan Tompkins, Grant Norbury, Andrea Byrom and Roger Pech are developing to help managers predict the outcomes of pest management. Carlos Rouco is a postdoctoral fellow from Spain working on possum behaviour in dry grasslands. Like for mice, we know very little about possum ecology in this ecosystem. Ecologists Hanne Koster from the Netherlands and Sebastian Prinz from Germany volunteered their help to radio-track where possums den during the day. Javier Fernandez, a researcher from Spain, also visited our lab recently to help us analyse a huge dataset from the Rabbit and Land Management Programme to derive relationships between vegetation, rabbits and predators. We have been invited to write a book chapter on the ecological effects of rabbit haemorrhagic disease in New Zealand for a book entitled Relationships of biological diversity to ecological disturbance: scale, context, and nature being published by the US Wildlife Society.
We are currently planning the next phase of our research with DOC in the Macraes area (east Otago in the eastern South Island) to help them understand the wider ecological effects of predator control over a 4000-ha area designed to protect critically endangered lizards. Our proposed experimental design looks like spokes on a wheel, with transects across the predator control boundary designed to measure a ‘core effect’ or ‘halo effect’ for native species.
Unlike the other three strands of the IMI programme, the Global Change strand has a more general focus. We set up this strand firstly to develop initiatives around ‘global change’ (more below), and secondly to align some of our international work with the IMI programme. Global change drivers such as invasive species, climate and land use change are all ‘hot topics’ internationally – below we highlight some work we are developing in this strand.
People often ask the question: where in the landscape are invasive animals? How quickly are they spreading into new habitats, or reinvading habitats that have been controlled? A natural tendency is to map the distribution of invasive species, but can we go one step further and turn those static maps into dynamic, interactive maps showing the outcomes of recent pest control and predicting trends in the distribution and abundance of the species in question? Researchers James Shepherd and Mandy Barron have been doing exactly that for the introduced Australian brushtail possum in New Zealand. The ‘National Possum Model’, when it becomes fully operational, will be the first national invasive species model in the world. As well as mapping possum density in the current New Zealand landscape, the researchers aim to make predictions for future environmental conditions, including changes in land use and climate change. For example, plant growth and forest species composition may change with elevated CO2 levels, which in turn will affect the maximum possum density. We’ll keep you posted as things develop.
Andrea Byrom, Wendy Ruscoe and Guy Forrester, collaborating with Tony Sinclair (University of British Columbia) and a truly international contingent of colleagues from every continent, recently completed an analysis of 40 years of data on rodent abundance from Serengeti National Park in Tanzania (East Africa). Rodent abundance was strongly correlated to both rainfall and the SOI (Southern Oscillation Index), with the ‘short rains’ (November to January each year) being particularly important drivers of rodent outbreaks. Outbreaks of rodents can be linked to fluctuations in many other trophic levels in the Serengeti ecosystem, including small carnivores such as foxes, caracals and servals (small cats), as well as birds of prey. Rodent outbreaks are also linked to bubonic plague outbreaks in human populations in villages in the greater Serengeti ecosystem. Global change drivers may therefore strongly influence both ecosystem dynamics inside Serengeti National Park as well as the human populations surrounding the park.
In an international collaboration, Dean Anderson and colleagues recently published a study of the browsing behaviour of elk (Cervus elaphus) in Yellowstone National Park. Following the extensive 1988 fires in Yellowstone, a mosaic of high-density patches of fallen logs and regenerating lodgepole pine (Pinus contorta) saplings developed in the landscape. Such patches could potentially protect post-fire aspen (Populus tremuloides) seedlings from being browsed by elk. The researchers asked: (1) Do elk reduce their use of high density patches of coarse wood and pine saplings? and (2) Are the abundance, height, and probability of presence of aspen positively related to the density of coarse wood or pine saplings? Variation in elk pellet densities was not explained by the density of logs. The height of aspen seedlings was not related to density of logs, pine saplings or density of elk faecal pellets. The findings suggest that fire-induced coarse wood and pine saplings will not create broad-scale browsing refugia for aspen in the landscape of the Yellowstone Plateau.
Pikas as ecosystem engineers: grassland degradation on the Qinghai-Tibetan Plateau, China
Roger Pech and Andrea Byrom were recently awarded a visiting professorship for senior international scientists to the Chinese Academy of Sciences. Their host institute was the Northwest Institute of Plateau Biology in Xining, Qinghai Province, in north-western China, where they spent a month with colleague Professor Zhang Yanming in May 2010. The researchers were investigating the ecological processes likely to regulate populations of a small hamster-like lagomorph, the plateau pika (Ochotona curzoniae), a burrowing species inhabiting the Tibetan Plateau at altitudes 3200–5200 m.
The pika is considered a keystone species in alpine meadow ecosystems. Its grazing and burrowing activities influence plant species composition as well as the movement of nutrients and water through the soil profile. The plateau pika is also a primary prey for native bird and mammal predators, and its burrows are used as nest sites by small passerines. However, pikas compete with livestock for scarce forage and their burrowing activities are thought to contribute to soil erosion and, ultimately, landscape degradation. Research on the population dynamics of plateau pikas is essential both for their conservation as a keystone species and for their management as a pest species.
Interaction between climate change and predation impacts: alpine ecosystems in New Zealand
An initiative by DOC and Landcare Research scientists (Jenny Christie, Warren Chinn, Nick Singers, and Andrea Byrom) aims to understand how invasive species’ impacts on native biodiversity might be exacerbated by the effects of climate change. The aim of the new project is to examine the interaction between climate change and predation impacts using incursions of ship rats (Rattus rattus) into alpine areas as a model ecosystem. Ship rats have been recorded only occasionally in alpine ecosystems, possibly spillover from outbreaks during years of high seed and fruit production in subalpine vegetation. However, as an invasive predator, ship rats could have major impacts on groups of native biota such as weta (an iconic invertebrate), other invertebrates, lizards, and birds such as rock wrens in alpine areas. Climate change models predict more years in which mean summer temperatures climb above the putative threshold that trigger ‘masting’ events in tall tussock (Chionochloa spp.) in alpine ecosystems, and beech (Nothofagus) forests adjacent to alpine areas. Based on knowledge of invasive species impacts on native biota, the researchers aim to predict how climate change might affect long-term trends for native biota.
An exceptionally strong altitudinal gradient at Fukang Station of desert ecology in Xinjiang has potential to provide a unifying theme linking research projects on climate and land-use change. The gradient ranges from permafrost at 5000 m to alpine meadows and through forested valleys to agricultural land bordering arid dune systems at 800 m. This is an ideal 'natural laboratory' for understanding how climatic factors, interacting with ecological processes and extensive (nomadic pastoralism) and intensive (e.g. irrigated crops) land use, limit the distribution and persistence of species, communities and ecosystems.
Andrea Byrom, Al Glen, Pen Holland, Duane Peltzer, Adrian Walcroft and Roger Pech from Landcare Research will participate in an initial workshop in July 2010 at the Institute of Ecology and Geography in Xinjiang. The partner organisation involved in the project is the Institute of Computational Biology, Shanghai Institute of Biological Sciences. The Ministry of Research, Science and Technology (MoRST) and Chinese Ministry of Science and Technology (MOST) have provided financial support for the workshop as the first initiative in a New Zealand – China strategic research alliance, with the IMI Programme ‘first off the block’ in the strategic alliance. The aim is to develop a collaborative project in Xinjiang that can be matched with research across a similar cross-section of natural and modified landscapes in the IMI programme.
A major focus of the Global Change strand is to apply our knowledge of invasive mammal ecology, interactions, and impacts on biodiversity to new ecosystems. One such initiative is a PhD position, currently being advertised, on invasive species ecology in braided river ecosystems in Canterbury. The project is jointly funded by Environment Canterbury, the Bio-Protection Centre of Research Excellence at Lincoln University, and Landcare Research. Braided rivers are a unique feature of New Zealand’s South Island. They carry a substantial but highly variable water flow on the dry eastern side of the Southern Alps, and they contain many important species of native biota including vegetation, lizards, invertebrates and river birds. The project aims to investigate the role of disturbance due to natural and modified flow regimes in shaping invasive alien plant distributions and abundance in braided river ecosystems, and the cascading effects on native plants, alien mammals and ecosystem structure and function. By the time of our next newsletter we hope to have our new student up and running – watch this space!
Around the world, conservation managers are asking questions such as: How do we achieve representativeness in protecting native biodiversity? Can we protect species over their former range? Which species should we measure and monitor in order to determine conservation ‘success’? In the IMI programme, we have been asking ourselves similar questions, but with a twist: which species of native biota should we measure and monitor in order to quantify successful outcomes of invasive species control? The answer to this question is deceptively complex. It’s tempting to select a representative range of taxa – for example, to monitor birds, lizards, invertebrates and vegetation responses to invasive mammal control. But will such taxonomic representation provide information on the functional role of species in ecosystems, or might other species better help us answer our research questions? How do we conserve ‘processes’ (such as seed dispersal, invertebrate herbivory and natural vegetation succession) and not just species? Roger Pech, Clare Veltman (DOC), and postdoctoral fellow Al Glen have started developing these ideas in the Theory Strand. Watch this space!
Management of invasive species in New Zealand, and around the world, is rarely conducted at a landscape scale. More typically, pests are managed intensively in specific areas that are judged to be of high importance, with little or no management in the surrounding landscape. This may lead to persistence of some native species only in areas where pests are controlled, perhaps with periodic dispersal between patches. This is known in ecological theory as a metapopulation. If a local population becomes dangerously small or even disappears, new immigrants from neighbouring patches can provide a ‘rescue effect’. The likelihood of extinction of the metapopulation is therefore much lower than that of a single, isolated population.
How can we apply a metapopulation paradigm to improve the outcomes of pest management for native biodiversity? This is the question being addressed in a concept paper currently in preparation by postdoctoral fellow Al Glen. By synthesising the literature on metapopulation theory, conservation planning and landscape connectivity, we will help focus IMI research on understanding how to set up future linked, networks of conservation areas. The goal is to improve the timing and location of pest management to provide greater landscape connectivity for native species and hence greater overall benefits for biodiversity.
Mathematical models called friction surfaces have been used to describe how easily animals can disperse through a landscape. Also the particular routes animals choose can be described by least-cost modelling and circuit theory modelling. These recent, somewhat esoteric, mathematical techniques will be used by new PhD student, Tom Etherington, to model how possums move through agricultural landscapes containing patches of favourable habitat. The aim is to help decide where pest control will give the best results for an entire region rather than just the immediate control site. In the longer term, the models will be used to predict how pest control needs to adapt to broad-scale changes in landscapes, for example due to reforestation.
Since its release in the mid-to-late 1990s, rabbit haemorrhagic disease (RHD) has been a highly successful biocontrol method for suppressing rabbit populations in New Zealand and Australia. However, in recent years, rabbit numbers have been increasing in some area prompting new research to update pre-RHD models of rabbit population dynamics. A collaborative project has begun with Dave Forsyth from the Arthur Rylah Institute for Environmental Research, Melbourne, and Tony Arthur from CSIRO, Canberra, to develop models that can be used to help decide when and where additional ‘conventional’ control methods will be needed. As a first step, postdoctoral fellow Amy Whitehead will use a large dataset provided by John Parkes and Ben Reddiex to determine which environmental cues trigger the annual onset of breeding by rabbits.
Roger Pech and Andrea Byrom are advisors on a collaborative project funded by the Australian Research Council (ARC) entitled ‘The return of the native: reintroductions, reinvasions, and a new paradigm in restoration ecology’. The project, run by Profs Peter Banks (University of New South Wales) and Chris Dickman (University of Sydney), aims to restore bush habitats in Sydney Harbour National Park, remove invasive black rats (Rattus rattus) and ultimately reintroduce the native bush rat (Rattus fuscipes). The project is a fascinating blend of theoretical concepts and practical outcomes: because the presence of native species can often reduce an ecosystem’s susceptibility to invasion, both competition theory and invasion biology predict a role for reintroductions in the control of invasion processes. Intact assemblages are more robust to invasion by new species than disturbed ones because niches are filled and competitive processes prevent establishment. The reintroduction of common native species has the potential to make a major contribution to the control of invasive species, by preventing them from both establishing in the first place and re-establishing after control. The aim is to transform Australia’s native wildlife from victims of invasive pests to active defenders that slow or prevent the invasion process. Peter and Chris are working with Andrea and Roger to further develop some of these theoretical concepts, because they apply to both the ARC project and the IMI programme.
More information on the bush rat reintroduction project can be found at:
A one-day workshop was held jointly with Simon Fowler’s ‘Beating Weeds II’ and Bruce Warburton’s ‘Small Mammal Pest Control’ programmes on 9 June 2010 in Christchurch, which gave some of our stakeholders the opportunity to find out about the latest research Landcare Research and partners have been undertaking in order to beat both weeds and mammal pests.
This newsletter details in-house research undertaken by Landcare Research to produce timely, cost-effective and publicly acceptable vertebrate pest management. The newsletter is produced every 6 months and seeks to keep readers better informed on the progress and outcomes of research. Past issues are available online.
Abdelkrim J, Byrom AE, Gemmell N 2010. Fine-scale genetic structure of mainland invasive Rattus rattus populations: Implications for restoration of forested conservation areas in New Zealand. Conservation Genetics, in press. DOI: 10.1007/s10592-010-0085-9
Duncan RP, Holland EP, Pech RP, Barron M, Nugent G, Parkes JP 2010. The relationship between possum density and browse damage on kamahi in New Zealand forests. Austral Ecology (in press).
Forester JD, Anderson DP, Turner MG 2008. Do high-density patches of and regenerating saplings create browsing refugia for aspen (Populus tremuloides) in Yellowstone National Park (USA)? Forest Ecology and Management 253: 211–219.
Gleeson DM, Byrom A, Howitt RLJ 2010. Non-invasive methods for genotyping of stoats (Mustela erminea) in New Zealand: potential for field applications. New Zealand Journal of Ecology34 (online early).
Lettink M, Norbury G, Cree A, Seddon PJ, Duncan R, Schwarz CJ 2010. Removal of introduced predators, but not artificial refuge supplementation, increases skink survival in coastal duneland, New Zealand. Biological Conservation 143: 72–77.
Norbury G, Heyward R, Parkes J 2009. Skink and invertebrate abundance in relation to vegetation, rabbits and predators in a New Zealand dryland ecosystem. New Zealand Journal of Ecology33: 24–31.
Parkes JP, Ramsey DSL, MacDonald N, Walker K, McKnight S, Cohen BS, Morrison SA 2010. Rapid eradication of feral pigs (Sus scrofa) from Santa Cruz Island, California. Biological Conservation143: 634–641.
Pech R, Byrom A 2009. Impacts of invasive mammals on biodiversity. Kararehe Kino – Vertebrate Pest Research15: 16–17.
Pech R, Byrom, A, Anderson, D, Thomson C, Coleman M 2010. The effect of poisoned and notional vaccinated buffers on possum (Trichosurus vulpecula) movements: minimising the risk of Tb spread from forest to farmland. Wildlife Research (in press).
Ramsey DSL, Efford MG 2010. Management of bovine tuberculosis in brushtail possums in New Zealand: predictions from a spatially explicit, individual-based model. Journal of Applied Ecology (in press).
Ramsey R, Norbury G 2009. Predicting the unexpected: using a qualitative model of a New Zealand dryland ecosystem to anticipate pest management outcomes. Austral Ecology 34: 409–421.
Ruscoe W, Cave S, Sweetapple P, Pech R, Barron M, Yockney I, Perry M, Carran R, Brausch C 2010 Species interactions and consequences of pest control in forest ecosystems. Protect Autumn2010: 13–14.