Which is the optimal sampling strategy for habitat suitability modelling

摘要: Abstract Designing an efficient sampling strategy is of crucial importance for habitat suitability modelling. This paper compares four such strategies, namely, ‘random’, ‘regular’, ‘proportional-stratified’ and ‘equal-stratified’—to investigate (1) how they affect prediction accuracy (2) sensitive are to sample size. In order compare them, a virtual species approach (Ecol. Model. 145 (2001) 111) in real landscape, based on reliable data, was chosen. The distribution the sampled 300 times using each strategies sizes. data were then fed into GLM make two types prediction: presence/absence. Comparing predictions known allows model be assessed. Habitat assessed by Pearson's correlation coefficient presence/absence Cohen's κ agreement coefficient. results show ‘regular’ ‘equal-stratified’ most accurate robust. We propose following characteristics improve design: increase size, prefer systematic random (3) include environmental information design.