Attempted tests of the surrogacy value of the ED environmental diversity measures highlight the need for corroboration assessment of surrogacy hypotheses

A recent paper in this journal (Hortal et al., 2009) claimed to have evaluated the ED biodiversity surrogates methods of Faith and Walker (1994, 1996), and to have provided evidence for poor performance of the continuous ED method. In fact, their study neither used nor evaluated the continuous ED method. Here, I document their misrepresentation. I then discuss some constructive lessons emerging from their study and other recent studies that have attempted tests of ED surrogacy value. The need to consider the actual degree of support that observed evidence provides for a hypothesis about surrogacy raises general issues for evaluations of indicators' performance, and suggests a greater role for corroboration assessments. Guidelines for achieving this cover three aspects of surrogates testing: experimental design of tests, ongoing corroboration assessment of evidence produced by tests, and accumulation of lessons learned from multiple test studies over time.