Elsevier

Biological Conservation

Volume 191, November 2015, Pages 640-649
Biological Conservation

Using pest monitoring data to inform the location and intensity of invasive-species control in New Zealand

https://doi.org/10.1016/j.biocon.2015.08.022Get rights and content

Highlights

  • Rat and possum abundances were related to pest control and environmental variables.

  • High-intensity control substantially reduced rat and possum abundances.

  • Low-intensity control dominates pest management but had no measurable effect.

  • Environmental variables can help to prioritise pest control locations.

  • Pest monitoring may ensure that scarce resources are not wasted on ineffective control.

Abstract

Improving decision-making about where and how invasive species management is conducted requires a better understanding of spatial variation in pest abundance and the extent to which different pest control regimes reduce those abundances. We measured how the relative abundance of invasive rats and possums, indexed at 147 forest sites in northern New Zealand, varied in response to environmental variables, and to the category of pest control at the site: no control (NC), periodic possum (PP), low-intensity rat and possum (LRP), and high-intensity rat and possum (HRP). We found that climate and topography strongly influenced the rat index, while vegetation characteristics strongly influenced both indices. These variables may therefore be useful for predicting pest impacts and prioritising locations for pest control. HRP control substantially reduced pest abundance indices, with model-predicted values for the rat index that were 99% lower than in areas of no control, and 91% lower for the possum index. In contrast, indices did not differ significantly among NC, PP, and LRP. PP and LRP regimes dominate pest control in New Zealand, but may be of limited conservation value, at least in terms of the ‘average’ operation in our study region. Globally, conservation agencies with limited budgets frequently avoid monitoring pest populations, since resources spent on monitoring are no longer available for management. However, our results highlight the value of collecting and analysing pest monitoring data, both for informing the location of future management and for ensuring that scarce resources are not wasted on ineffective control.

Introduction

Invasive species are one of the major threats to biodiversity worldwide (Sala et al., 2000), but their management (‘pest control’) is rarely straightforward. Pest control is expensive, and there are typically only sufficient resources available to control a fraction of the area in which threats from invasive species occur (Wilson et al., 2007). Moreover, there are often multiple methods available for controlling a given invasive species, and these may vary in their efficacy, resource-intensiveness, humaneness, and risk to non-target species (Shea et al., 2002, Parkes and Murphy, 2003, Govindarajulu et al., 2005, Reddiex et al., 2006, Beausoleil et al., 2010, Massei et al., 2011). The challenge for conservation managers working with limited resources is to select locations for pest control where conservation gains will be the greatest, then to use control methods that will predictably and efficiently reduce pest abundances to the levels necessary to meet conservation objectives.

The collection and analysis of pest monitoring data may play an important role in meeting this challenge. First, analysis of the distribution and abundance of pests may reveal that easily-measured environmental variables can be used to predict pest occurrence or impacts. This information has typically been used to predict where future invasions are likely to occur (e.g. Ficetola et al., 2007, Zhu et al., 2007, Roura-Pascual et al., 2009), but can also be used to prioritise locations for control within the existing range of an invasive species (e.g. Porphyre et al., 2014). Second, monitoring how pests respond to control operations improves understanding of whether a particular control regime is likely to achieve conservation objectives, which is important both for designing effective pest management programmes and for increasing the ethical defensibility of pest control that kills sentient animals or poses risks to non-target species (Reddiex and Forsyth, 2006, Warburton and Norton, 2009). However, conservation agencies often do not monitor pest populations, or collect monitoring data but do not analyse them (Shea et al., 2002, Reddiex et al., 2006, Possingham, 2012, Walsh et al., 2012). There is a need for increased monitoring and analysis of pest populations, to improve understanding of where limited pest control resources should be spent and to validate assumptions about the type of control required to achieve conservation objectives (Shea et al., 2002, Reddiex et al., 2006, Warburton and Norton, 2009, Walsh et al., 2012).

As in many parts of the world, the control of invasive mammals is a central part of conservation management in New Zealand (Clout, 2001). On the mainland (North Island and South Island), this control primarily targets the ship rat (Rattus rattus), a globally invasive species which threatens native biodiversity on islands worldwide (e.g. MacFarland et al., 1974, Monteiro et al., 1996, Martin et al., 2000, Jouventin et al., 2003, Harris and Macdonald, 2007), and the brushtail possum (Trichosurus vulpecula), which is invasive throughout New Zealand (Wright, 2011). These species consume a wide range of plant and animal taxa, and are believed to be the primary threat to many native species (Craig et al., 2000, Innes et al., 2010a, Wright, 2011). An enormous amount of time and effort is spent attempting to reduce their abundance, with control covering more than a million hectares and costing of tens of millions of New Zealand dollars each year (Wright, 2011).

The particular methods used for this control vary widely, involving a range of control devices (mostly utilising various traps and poisons) which are employed at widely varying intensities (i.e. variable spacing of devices and frequency of control) (Gillies, 2002, Parkes and Murphy, 2003, Beausoleil et al., 2010). However, although some of these pest control regimes demonstrably achieve conservation objectives (e.g. Innes et al., 1999, Gillies, 2002, Armstrong et al., 2006, Warburton and Norton, 2009), there is considerable uncertainty about the efficacy of others. For example, possum control is often repeated on a 1–8 year cycle, but return times longer than a few years may not provide benefits to native species (Parkes and Murphy, 2003, Brown and Urlich, 2005). Similarly, even relatively intensive rat control operations may not be able to reduce rat abundance to levels required to protect rat-sensitive species (Gillies, 2002). Lastly, there is increasing recognition that rats, possums, and other invasive mammals may negatively affect each other, so that control of one species may cause an increase in the abundance or impacts of others (Rayner et al., 2007, Sweetapple and Nugent, 2007, Ruscoe et al., 2011). In the face of this uncertainty, there is a need for increased monitoring and analysis of pest control operations, so that the efficacy of particular regimes can be better understood.

Despite the considerable resources spent controlling rats and possums over large parts of New Zealand, this control can still only cover a fraction of the area in which threats from these species are likely to occur (Craig et al., 2000, Wright, 2011). To aid in prioritising locations for pest control, several studies have examined determinants of spatial variation in rat or possum abundance in New Zealand. Fraser et al. (2003) examined how possum capture rates varied at a national scale, and found that this variation was primarily influenced by climatic variables. Similarly, Porphyre et al. (2014) found that climate, together with the presence of forest cover, strongly influenced possum capture rates in a montane grassland environment. Several studies have examined how ship rat capture rates vary within forests in New Zealand, and these suggest that rat abundance is influenced by a variety of factors, including distance from the forest edge, vegetation, topography, and altitude (King et al., 1996, Harper et al., 2005, Christie et al., 2006, Christie et al., 2009, Ruffell et al., 2014). However, little is known about how possum and rat abundances vary within lower elevation environments at the regional scale. This information would be helpful for prioritising areas for pest control, because these environments make up the majority of New Zealand's land area (MfE, 2002) and include a disproportionate number of New Zealand's threatened species (Walker et al., 2008), and because conservation management agencies frequently operate at a regional scale (regional councils and Department of Conservation regional conservancies, for example).

In this study, we measured rat and possum relative abundances at sampling stations systematically located in forest habitats throughout the Auckland region of New Zealand, as part of a region-wide biodiversity monitoring programme. We also quantified local vegetation characteristics (structure and composition), topography and climate, measures of forest fragmentation and urbanisation, and the presence and type of pest control at each sampling station. Our aim was to use this dataset to improve understanding of where and how pest control should be conducted in the region, by examining (1) which variables could be used to predict spatial variation in rat and possum abundance at a regional scale, and (2) how effectively the different pest control regimes in the region could reduce rat and possum abundance indices.

Section snippets

Study region

We conducted our study in the Auckland region of northern New Zealand. The region is c. 5000 km2 in area, and comprises lowlands and hill country on a relatively narrow isthmus of land (< 70 km at the widest point), together with a number of near-shore islands. The maximum mainland altitude is 722 m (McClure, 2012a). The climate is temperate–subtropical (McClure, 2012b), with a mean annual temperature ranging from c. 11–16 °C across the region as measured from Land Environments of New Zealand

Results

The 147 sampling stations captured a wide gradient of variation in our continuous predictor variables, and covered most levels of our categorical predictor variables with reasonable sampling intensity (Table 1). An exception was the ‘kahikatea’ category of our vegetation class variable, which occurred at only two sampling stations (Table 1). Plots of model residuals suggested that assumptions were met for rat and possum GLMMs.

From our 63 candidate models, there was no single model that

Discussion

The management of invasive species is a key part of biodiversity conservation globally, but resources are typically only available to manage a fraction of the area in which threats from these species exist (Craig et al., 2000, Wilson et al., 2007). Moreover, pest control operations can have variable and unpredictable effects on pest abundances (Reddiex and Forsyth, 2006, Warburton and Norton, 2009, Walsh et al., 2012). In the face of these scarce resources and variable outcomes, effective

Conclusions

Monitoring the relative abundance of invasive rats and possums throughout the Auckland region revealed that these species show considerable geographic variation in response to local vegetation characteristics, climate and topography, and HRP pest control. Local vegetation and climate and topography may therefore help to prioritise locations for pest control when resources are scarce, while HRP control appears able to predictably and substantially reduce the abundances of these species where

Acknowledgements

We gratefully acknowledge the many pest control operators who helped us to map the locations of pest control throughout the region, and the large number of staff at Auckland Council who provided fieldwork assistance. JR was supported by a Hackett Postgraduate Research Scholarship throughout the research. RKD was supported during the writing of the manuscript by an Australian Research Council Future Fellowship FT100100040. JI acknowledges core funding for Crown Research Institutes from the

References (78)

  • I. Atkinson

    Derivation of vegetation mapping units for an ecological survey of Tongariro National Park, North Island, New Zealand

    N. Z. J. Bot.

    (1985)
  • K. Barton

    MuMIn: Multi-model Inference. Version 1.9.13

  • D. Bates et al.

    lme4: Linear Mixed-effects Models Using S4 Classes. Version 0.999999-0

    (2011)
  • N. Beausoleil et al.

    How humane are our pest control tools? Part 1

    Vertebrate Toxic Agents and Kill Traps in Mammal Species

    (2010)
  • O. Bjørnstad

    ncf: Spatial Nonparametric Covariance Functions. Version 1.1-5

    (2013)
  • R. Brockie

    Effect of commercial hunters on the numbers of possums, Trichosurus vulpecula, in Orongorongo Valley, Wellington

    N. Z. J. Ecol.

    (1982)
  • K.P. Brown et al.

    Aerial 1080 operations to maximise biodiversity preservation

    (2005)
  • K.P. Burnham et al.

    Model Selection and Multi-model Inference: A Practical Information-theoretic Approach

    (2002)
  • J. Christie et al.

    Measuring stoat (Mustela erminea) and ship rat (Rattus rattus) capture success against micro-habitat factors

    N. Z. J. Ecol.

    (2006)
  • J. Christie et al.

    Environmental predictors of stoat (Mustela erminea) and ship rat (Rattus rattus) capture success

    (2009)
  • R. Core Team

    R: A Language and Environment for Statistical Computing

    (2015)
  • J. Craig

    Rodents of the greater Auckland region

    Tane

    (1983)
  • J. Craig et al.

    Conservation issues in New Zealand

    Annu. Rev. Ecol. Syst.

    (2000)
  • R.K. Didham et al.

    Rethinking the conceptual foundations of habitat fragmentation research

    Oikos

    (2012)
  • G. Ficetola et al.

    Prediction and validation of the potential global distribution of a problematic alien invasive species — the American bullfrog

    Divers. Distrib.

    (2007)
  • J. Flux

    Seventeen years of predation by one suburban cat in New Zealand

    N. Z. J. Zool.

    (2010)
  • K. Fraser et al.

    Predicting spatial patterns of animal pest abundance: a case study of the brushtail possum (Trichosurus vulpecula)

    Science for Conservation

    (2003)
  • C. Gillies

    Managing Rodents on the New Zealand Mainland — What Options are Currently Available? Summary of a Workshop Session at the Department of Conservation ‘Mainland Island’ Hui, Omapere, 20–23 August 2001

    (2002)
  • C. Gillies et al.

    The prey of domestic cats (Felis catus) in two suburbs of Auckland City, New Zealand

    J. Zool. Soc. Lond.

    (2003)
  • C. Gillies et al.

    Six years of intensive pest mammal control at Trounson Kauri Park, a Department of Conservation “mainland island”, June 1996–July 2002

    N. Z. J. Zool.

    (2003)
  • J. Gordon et al.

    Belled collars reduce catch of domestic cats in New Zealand by half

    Wildl. Res.

    (2010)
  • P. Govindarajulu et al.

    Matrix model investigation of invasive species control: bullfrogs on Vancouver Island

    Ecol. Appl.

    (2005)
  • R. Gregory et al.

    Structured Decision Making: A Practical Guide to Environmental Management Choices

    (2012)
  • C. Grueber et al.

    Multimodel inference in ecology and evolution: challenges and solutions

    J. Evol. Biol.

    (2011)
  • G. Harper et al.

    Habitat use by three rat species (Rattus spp.) on Stewart Island/Rakiura, New Zealand

    N. Z. J. Ecol.

    (2005)
  • D.B. Harris et al.

    Interference competition between introduced black rats and endemic Galápagos rice rats

    Ecology

    (2007)
  • T. Hothorn et al.

    multcomp: Simultaneous Inference In General Parametric Models. Version 1.3.1

    (2013)
  • J. Innes

    Advances in New Zealand mammalogy 1990–2000: European rats

    J. R. Soc. N. Z.

    (2001)
  • J. Innes

    Ship rat

  • Cited by (13)

    View all citing articles on Scopus
    View full text