Developing a new biodiversity risk indicator based on finance metrics

Abstract:

As human activities continue to impact wildlife and the natural environment, it has become critical to quantify the risk of biodiversity loss in order to assess these impacts and prioritize mitigation strategies. Existing indicators of biodiversity status (e.g., the Living Planet Index, Shannon diversity index) combine aspects of species abundance, evenness, and diversity to measure how biodiversity is changing over time. Here we explore the application of risk metrics from financial portfolio theory, specifically portfolio volatility and marginal contribution to risk (MCTR), to evaluate their effectiveness for assessing the risk of biodiversity loss. We calculate these metrics using the most comprehensive existing dataset on species abundance, the USGS North American Breeding Bird Survey, which provides abundance estimates for >400 bird species from 1970-2017. We examine whether MCTR values can be predicted based on species traits, and compare trends in MCTR at multiple spatial scales. We show that portfolio volatility and MCTR provide complementary insights with existing indicators, by incorporating fluctuations in abundance, which is not represented in existing indicators but known to be correlated with species extinction risk. We also find demographic traits are better predictors of MCTRs than habitat- and niche-based traits, and the species contributing the most to portfolio risk differ across spatial scales. These newly identified biodiversity indicators will be combined with critical risk indicators from other domains using machine learning methods and high-dimensional network modeling to explore the complex interconnectedness among risks in human-natural systems.