The pervasive influence of sampling and methodological artefacts on a macroecological pattern: the abundance–occupancy relationship
Article first published online: 10 APR 2008
© 2008 The Author. Journal compilation © 2008 Blackwell Publishing Ltd
Global Ecology and Biogeography
Volume 17, Issue 4, pages 457–464, July 2008
How to Cite
Wilson, P. D. (2008), The pervasive influence of sampling and methodological artefacts on a macroecological pattern: the abundance–occupancy relationship. Global Ecology and Biogeography, 17: 457–464. doi: 10.1111/j.1466-8238.2008.00385.x
- Issue published online: 10 JUN 2008
- Article first published online: 10 APR 2008
- Abundance–occupancy relationship;
- classification and regression tree;
Aim To investigate the influence of sampling and methodological artefacts on the correlation between abundance and occupancy.
Location Global scope.
Methods A fixed effects weighted regression model was fitted to standardized effect size for 175 examples of correlations between abundance and occupancy. A regression tree model with standard effect size as the dependent variable was also fitted to the data.
Results Standard effect size, and therefore the correlation between abundance and occupancy, was found to be strongly influenced by the type of abundance measure used to characterize the abundance–occupancy relationship. Local mean abundance (also referred to as ecological mean abundance) was primarily responsible for negative correlations. Negative correlations also resulted from a mismatch in the sampling extents of abundance and occupancy measures.
Main conclusions The combination of abundance and occupancy measures selected to characterize the abundance–occupancy relationship for a given set of data has a profound impact on the sign of the correlation between the selected measures. Previous attempts to understand the processes giving rise to the pattern represented by the abundance–occupancy relationship have confounded sampling artefacts (e.g. spatial extent of abundance and occupancy information) and methodological artefacts (e.g. combining a truncated abundance measure such as local mean abundance with an untruncated occupancy measure such as proportion of occupied samples). Thus, a revision of the approach currently used to define and evaluate competing explanatory models of the abundance–occupancy relationship appears to be necessary.