Spatial Capture-recapture with Memory: A New Hidden Markov Model Perspective
Lead Research Organisation:
University of St Andrews
Department Name: Mathematics and Statistics
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Organisations
| Description | Within this project we have focused on the estimation of abundance (and density) of difficult to observe species using data collected via an array of motion-sensor camera traps, typically referred to as spatial capture-recapture data. We demonstrated via simulating data that current state-of-the-art models for analysing such data can lead to over-estimates of the population size. This is because they do not account for the fact that it is more likely that an animal is next detected at a trap close to its last detection than at a trap farther away. We developed a new statistical model that is able to incorporate this "spatial memory" in the way animals move, resulting in more accurate population estimates. |
| Exploitation Route | The methods we developed suggest that their are substantial gains to be made by modelling these kinds of survey data as spatial time-to-event data, which opens up a new and promising approach that we plan to pursue. The methods and software will be useful for ecologists and conservationists assessing wildlife populations using camera trap survey data. |
| Sectors | Environment |