Towards More Rigorous Assessment of Biodiversity

Abstract

Biodiversity is often ill-defined and subjectively surveyed, resulting in inefficient and ambiguous estimates. Strengths and deficiencies of prevailing survey techniques are appraised through a review of selected literature. Analogies with forest inventory are used to suggest options for more efficient and rigorous biodiversity assessment. Techniques such as variable-probability and model-based sampling, especially when used in conjunction with generalized linear modelling, offer efficient alternatives to more traditional assessments based on quadrants and nested plots. Bayesian methods offer scope to combine expert and local knowledge with formal samples, and warrant further investigation. Suggestions for further research are given.