Reducing variability in habitat condition assessments is critical if nature‐market metrics are to represent true biodiversity change

The need for standardized metrics for measuring losses and gains in biodiversity has resulted in many countries and private sector initiatives looking to adapt England's Statutory Biodiversity Metric for mandatory Biodiversity Net Gain (BNG). BNG requires a minimum 10% uplift in biodiversity units out of infrastructure development, and the number of biodiversity units depends in part on field‐based habitat condition assessments. We carried out simulations to explore the influence of uncertainty in condition assessments and found that there is inherent variability that introduces bias, even when a habitat's true condition remains unchanged. To explore likely real‐world variability, we used an online test targeting ecologists ( n = 155) involved in BNG assessments and follow‐up interviews with a sub‐set of respondents ( n = 21). In our online test, BNG ecologists were more likely to correctly assess habitats in poor condition than in good condition. Depending on the scenario and habitat, assessment variability could be significant enough to lead to a reported 10% uplift in biodiversity units even if there is no uplift. We acknowledge that our online test might not capture all true variability, but our findings underpin the focus on training and support for ecologists to reduce variability and more detailed industry guidance to ensure that BNG legislation in England delivers for nature. It is key to recognize the role of habitat condition assessment for the metric to measure true change and when adapted elsewhere.