Boosting biodiversity monitoring using smartphone-driven, rapidly accumulating community-sourced data

Abstract

Kunming-Montreal Global Biodiversity Framework increased the demand for biodiversity distribution data. To gather species observation from the public, we introduced a mobile application called 'Biome' in Japan. By employing species identification algorithms and gamification elements, Biome has gathered >5M observations since its launch in 2019. However, cloud-sourced data often exhibit spatial and taxonomic biases. Species distribution models (SDMs) enable infer species distribution while accommodating such bias. We investigated Biome data's quality and how incorporating Biome data influences the performance of SDMs. Species identification accuracy of Biome data exceeds 95% for birds, reptiles, mammals, and amphibians, but seed plants, molluscs, and fishes scored below 90%. The distributions of 132 terrestrial plants and animals across Japan were modeled, and their accuracy was improved by incorporating Biome data into traditional survey data. For endangered species, traditional survey data required >2,000 records to build accurate models (Boyce index >0.9), though only ca.300 records were required when Biome was blended. The unique data distributions may explain this improvement: Biome data covers urban-natural gradients uniformly, while traditional data is biased towards natural areas. Combining multiple data sources offers insights into species distributions across Japan, aiding protected area designation and ecosystem service assessment.

Harnessing Community Science and Open Research-based Data to Track Distributions of Invasive Species in Japan

“…People take part in community science projects because they recognize their value for the environment and/or want to gain knowledge. Gamification is an additional feature that can boost data collection (Fraisl et al 2022;Atsumi et al 2024). Finally, we want to emphasize that some important taxa that require special knowledge were badly underrepresented in both CSD and ORD.…”

“…We used the data that had been collected by February 2023. Their species accuracy was 91.7% and 95.5% for common seed plants and insects, respectively (Atsumi et al 2024). Before the analysis, we filtered out records deemed to be invalid based on location metadata and users' reactions to the record (see Atsumi et al (2024) for detailed procedures).…”

“…iNaturalist has accumulated 190 million occurrence records worldwide uploaded by 3 million observers (https://www.inaturalist.org, accessed in June 2024). In Japan, Biome, a mobile application developed by Biome Inc. in 2019, has accumulated >7 million records (Atsumi et al 2024).…”

“…Nevertheless, multiple factors limit the information that can be extracted from massive CSD or ORD (Beck et al 2014;Amano et al 2016). CSD records are often concentrated in regions with high human population densities and are easily accessible (Geurts et al 2023;but Atsumi et al 2024). Besides, CSD is biased toward large charismatic species (e.g., mammals, birds, butterflies), and records of inconspicuous or unpopular groups (e.g., spiders) tend to be rare (Binley et al 2023).…”