Leveraging participatory science data to detect and quantify elevational range shifts in birds

Montague Neate-Clegg, Morgan Tingley, I-Ching Chen, Natalia Ocampo-Peñuela

One of the primary expectations—and observed consequences—of persistent warming temperatures is range shifts to higher elevations. However, the ability to detect and attribute climate-driven range shifts has historically been limited to rare locations with sufficient time-series data, which are especially rare outside temperate regions of the Northern Hemisphere. Leveraging data from participatory science programs may offer the key to studying elevational range shifts at broader spatial scales and at higher temporal resolution. With an elevational gradient of almost 4000 m and a high density of checklists from the platform eBird, we focused on the island of Taiwan and demonstrated the estimation of elevational range shifts for 69 species of land bird from 2011–2025. Specifically, we used generalized additive models to estimate species’ peak elevations (where occurrence probability is highest), lower and upper elevational limits, and elevational range breadths in every year, and then modeled those estimates over time (while propagating uncertainty) to quantify shift rates. We found that bird species’ peak elevations have shifted upslope by 13 m/yr, with lower and upper elevational limits shifting upslope by 8 and 21 m/yr, respectively, leading to range expansions averaging 15 m/yr. We also conducted a power analysis with six hypothetical species to determine the number of presence points necessary to reliably estimate elevational metrics with participatory science data using our methods. We found that, for a single species in a single year, even 100 presences was sufficient to result in 95% coverage of elevational metric estimates 95% of the time. However, accuracy generally decreased for the peak elevations of species that live close to topographic boundaries (sea level or mountaintops). After demonstrating how participatory science data can be used to quantify elevational range shifts, we provide a set of best practices to enable future analyses of elevational ranges and elevational shifts around the world where sufficient eBird data exist.