Risk Analysis of Invasive Species and Pests Using Citizen Science Data: Evaluation of Existing Analysis Techniques Using Layered Simulation

• Activity Number: 364
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Risk Analysis
• Author(s): Marijke Welvaert* and Peter Caley
• Companies: CSIRO and CSIRO
• Keywords: biosecurity risk ; citizen science ; model evaluation ; simulation

Abstract

Citizen science data have recently drawn attention as a means to infer species' distribution models (SDM) and could also have the potential of monitoring invasive species and pests. However, statistical species' distribution modelling techniques are typically not designed to cope with the added noise levels and distorted observation process in citizen science data. Therefore, it is unclear whether SDM's derived from citizen data are useful for risk assessments of the ecological and/or economical impact of the invasive species/pest. To evaluate the modelling techniques, we developed a simulator that couples the realistic simulation of species distribution with the citizen science observation process. This tool enables the statistical validation of SDM techniques and delineates the parameter space under which citizen surveillance is useful for making inference on the distribution of invasive species/pests. In this paper, we detail the different layers in our simulator and evaluate the statistical properties of applying popular SDM techniques on citizen science data. Actual examples of citizen science data are presented.