Projects per year
Abstract
Conventional distance sampling adopts a mixed approach, using model-based methods for the detection process, and design-based methods to estimate animal abundance in the study region, given estimated probabilities of detection. In recent years, there has been increasing interest in fully model-based methods. Model-based methods are less robust for estimating animal abundance than conventional methods, but offer several advantages: they allow the analyst to explore how animal density varies by habitat or topography; abundance can be estimated for any sub-region of interest; they provide tools for analysing data from designed distance sampling experiments, to assess treatment effects. We develop a common framework for model-based distance sampling, and show how the various model-based methods that have been proposed fit within this framework.
Original language | English |
---|---|
Pages (from-to) | 58-75 |
Number of pages | 18 |
Journal | Journal of Agricultural, Biological and Environmental Statistics |
Volume | 21 |
Issue number | 1 |
Early online date | 3 Sept 2015 |
DOIs | |
Publication status | Published - Mar 2016 |
Keywords
- Distance sampling
- Line transect sampling
- Model-based inference
- Point transect sampling
Fingerprint
Dive into the research topics of 'Model-based distance sampling'. Together they form a unique fingerprint.Projects
- 1 Finished
-
National Centre for Statistical Ecology: National Centre for Statistical Ecology
Buckland, S. T. (PI)
1/10/10 → 30/09/15
Project: Standard
Profiles
-
Stephen Terrence Buckland
- Statistics - Emeritus Professor
- School of Mathematics and Statistics - Emeritus Professor
- Centre for Research into Ecological & Environmental Modelling
- Marine Alliance for Science & Technology Scotland
Person: Emeritus Professor