The Biodiversity Intactness Index: a model-based indicator of ecological integrity

First proposed in 2005, the Biodiversity Intactness Index (BII) is a measure of ecological integrity. BII estimates the overlap between sites’ present-day ecological assemblages and their condition prior to marked anthropogenic change. It ranges from 100% (the assemblage is essentially intact) down to 0% (any species present are not native) and can be averaged across any region of interest. The PREDICTS project estimates BII from statistical models that relate biodiversity to land-use and related pressures, using our compilation of biodiversity survey data from over 50,000 sites worldwide that includes a taxonomically representative set of 75,000 plant, invertebrate and vertebrate species. As a model-based indicator, BII can be projected at high spatial and temporal resolution, so has a wide range of scientific, policy and commercial uses. Our first global estimate (in 2016) showed that BII had already fallen below the proposed planetary boundary of 90%, even though several limitations meant it underestimated losses at that time. We have been improving our input data, analytical methods and software engineering throughout the ten years since then. In this talk, I show these have overcome many of the problems with our early estimates of BII, and allowed us to now estimate both it and PDF (the potentially disappeared fraction of species) at kilometre resolution for each year since 2000.

Since 2023, our work has received support from Bloomberg via a data licence with the Natural History Museum. I will describe how Bloomberg combine their database of physical assets with global BII data produced by the Natural History Museum to provide summaries of biodiversity-related risks for nearly 50,000 companies. These insights form part of the sustainability data available to over 350,000 Bloomberg Terminal subscribers, helping investment professionals to make more nature-positive investment. Lastly, I will outline ongoing and planned developments, such as incorporating a broader range of drivers – including climate change – into our projections.