Skip to main content
  • UQ Home
  • Contacts
  • Study
  • Maps
  • News
  • Events
  • Library
  • my.UQ
The University of Queensland

School of Veterinary ScienceSchool of Veterinary Science

Site search
Homepage Site menu
  • Home
  • About
    • Welcome
    • Accreditation
    • History
    • News
    • Events
    • Job opportunities
    • Our strategic plan- vision, mission and values
    • Intranet
  • Study
    • Undergraduate
    • Postgraduate Coursework
    • Higher Degree by Research (HDR)
  • Research
    • Research strengths & impacts
    • Research Collaborations
    • Facilities
    • Research Centres and Units
  • Services & Hospitals
  • Community
    • Continuing Education and Professional Development
    • Pets-for-Life Adoption
    • Giving
    • Alumni and Community
      • Feature Events
      • Newsletters
      • Student Life
      • Vet School Activities
    • Providing Student Placements
    • Clinical trials
  • Our people
    • Academic
    • Administrative Professional
    • Scientific Professional
    • Clinical Professional
    • Adjunct and Honorary
  • Student Support
    • Scholarships & prizes
    • Engaging with Students
    • Pre-Clinical and Clinical Placements
    • Student Clinical Skills Hub
    • Support for international BVSc students
  • Contact
    • Staff Directory

Ravi Dissanayake, PhD candidate, Thesis Review Seminar , "Using long-term passive surveillance data to describe and model spatio-temporal patterns of koala sightings and risk factors for koala mortalities in South-East Queensland"

30 July 2019
12:00pm
Building 8106 Room 211 Gatton Campus
Ravi’s advisors are A/ Prof Joerg Henning, A/ Prof Rachel Allavena, and Prof Mark Stevenson

 

The koala has faced threats since European settlements but the nature of these threats have changed over time. It has been estimated that an average koala population decline by 24% during last 15-21 and estimated 53% losses over the next 15 to 21 years in Queensland. The South East Queensland (SEQLD) population is under severe threat due to loss of habitat and fragmentation as a result of rapid urbanisation, and deaths of koalas from diseases and vehicular accidents. Hence the replacement of population by natural breeding is not sufficient to main the population. Therefore, it is important to take evidence based conservation strategies to protect the declining population. Towards this end, understanding of population size, its distribution and risk factors for deaths are important. For population estimates or for monitoring population, conducting systematic surveys are not feasible for large geographical areas for long time periods. Thus an alternative data sources we explored and a long term citizen science database was used to retrieve incidental koala sightings and location of koala deaths for this study.    Using citizen science data have many advantages as volunteers have contributed to collect incidental koala sightings continuously from 1996-2013 in a large geographical area. Spatial temporal analysis of incidental sightings revealed no decline in sightings although the population is declining in SEQLD suggesting the active participation and enthusiastic reporting of sightings by Queenslanders. Temporal trends in sightings mirrored the breeding season of koalas. Sightings were high in residential areas followed by agricultural, and parkland. The density of reported koala sightings decreased as distance from primary and secondary roads increased, indicative of a higher search effort near roads. Incidental sightings are biased and lacks search effort information which is generally available with systematically collected survey data. The analysis shows that sighting are biased towards primary and secondary roads and less sightings being reported from western part of the study area further away from the coastal areas. Therefore, bias correction techniques were applied and search effort was estimated to model and map koala distribution. Three modelling approaches were used under the Poisson point process frame work each generating biased and bias corrected distribution of population. Spatial bias correction using distance to roads covariates improved the predicted koala density in areas where sightings were not reported. Search effort used to correct for uneven effort was by using Boosted Regression Tree modelling technique (BRT). The koala distribution models developed using incidental sightings and distribution maps revealed that koala population is high in eastern coastal areas and low in western parts of SEQLD suggesting that more active collection of incidental sightings is needed to improve the representativeness in sampling. These population distribution maps can be used to improve collection of incidental sightings in low density areas to improve the representativeness of sampling. The model which was developed using long term koala sighting data shows a high density in the Moreton bay area compared to all others locations suggesting a high activity of reporting of incidental sighting. The modelling and mapping of risk of koala deaths due to vehicle hits using Bayesian approach shows varying risk along different road types and hot spot areas but lower risk in western parts of the study area. Overall, this study has identified biases in incidental sightings, developed koala population density distribution maps using bias correction techniques and mapped risk across the study area which can be used as planning tools to take mitigation actions to protect both koala habitat and unexpected deaths of koalas due to accidents.

 

Attending staff and students please submit your audience feedback here

© The University of Queensland
Enquiries: +61 7 3365 1111   |   Contact directory
ABN: 63 942 912 684   |   CRICOS Provider No: 00025B
Emergency
Phone: 3365 3333
Privacy & Terms of use   |   Feedback   |   Updated: 26 Jul 2019
Login