Landscape influences on possum dispersal
Dispersal of possums into new territory is a key issue for possum management, resulting in the spread of bovine tuberculosis (TB) and rapid recolonisation of controlled areas. Landscape features act to limit possum dispersal but just how is not fully understood. Knowing how these features influence dispersal is important for large-scale possum control programmes, especially since they affect potential possum dispersal back into areas under management.
Tom Etherington, a PhD student at The University of Auckland and Landcare Research Joint Graduate School in Biodiversity and Biosecurity, is currently trying to identify the main landscape features that influence possum dispersal at large regional scales.
Collecting sufficient information on individual possum dispersal across large regional areas would be very costly. Instead Tom is using a ‘landscape genetics’ approach to provide an indirect measure of connectivity between possum populations across a landscape (Fig. 1). The assumption is that possum populations that are more closely related are likely to be more geographically connected. The Figure shows that such genetic connectivity is clearly not a simple function of how close or far apart possum populations are in the landscape as the ‘crow flies’, and that other variables must also be acting to influence possum dispersal.
To identify what landscape features may be limiting possum dispersal, Tom is analysing the landscape genetics data using a geographic information system (GIS) approach called ‘least-cost modelling’. Cost refers to the energy needed to get around particular topographic features (e.g. steep hills), behavioural preferences (e.g. forest remnants) and mortality risk (e.g. crossing rivers or roads). Using this information, least-cost modelling can find pathways between pairs of locations that represent the most efficient route of dispersal by balancing the distance travelled with the cost to traverse that landscape. The total cost associated with each least-cost pathway is then taken as a measure of connectivity.
A large number of scenarios, in which different combinations of landscape features have different costs, are being analysed and compared with the genetic data to identify what landscape features are important in limiting possum dispersal. Preliminary results suggest that large improvements in understanding the connectivity of possum populations across landscapes can be achieved by accounting for the costs of dispersal associated with certain landscape features (Fig. 2). The analyses indicate that the main factors restricting possum dispersal are the size of major rivers, followed by the absence of tree and scrub landcover.
Once the analyses are completed, GIS maps that represent the landscape in terms of costs to possum dispersal will be produced. These ‘cost maps’ will then be used as inputs to further analyses that will enable large-scale suppression programmes to tailor management of possums based on whether parts of a landscape are more or less isolated, or are more or less likely to act as dispersal pathways.
The collection of genetic data was undertaken under contract to the Animal Health Board (Project R-10625). The analysis was funded by the New Zealand Government through an International Doctoral Research Scholarship.
Thomas Etherington, George Perry, Phil Cowan & Mick Clout
Contact Thomas: teth001@aucklanduni.ac.nz
Fig. 1(a) The location of possum populations for which genetic data have been collected, with neighbouring possum populations connected to form a network. Thicker lines indicate more closely related possum populations with lower genetic distances.
Fig. 1(b) A correlation between genetic distance and straight-line distance, where each point represents a neighbouring possum population pair. If the landscape does not affect dispersal then possum populations that are closer together should be more related, and possum populations that are further apart should be less related. The poor correlation suggests that landscape features must be affecting possum dispersal.
Fig. 2(a) Preliminary results from the least-cost modelling have identifi ed pathways that are a more realistic representation of connectivity between neighbouring possum populations.
Fig. 2(b) A correlation between genetic distance and least-cost-path total cost, where each point represents a neighbouring possum population pair. If the landscape does affect dispersal then possum populations that are more connected should be more related, and possum populations that are less connected should be less related. The reasonable correlation suggests that landscape features do affect possum dispersal.
