Integrating citizen science data sets to estimate bird population dynamics
Philipp Boersch-Supan (British Trust for Ornithology)
Friday 31st January, 2020 15:00-16:00 Maths 311B
Mobile apps and online data entry portals have facilitated the reporting of opportunistic biodiversity observations. There is interest in using these data to calculate population indices for regions and species where formal monitoring is limited or absent. However, drawing inferences from opportunistic records is challenging because of preferential sampling and large heterogeneity in recording effort. Trend estimation is further complicated by computational challenges arising from the large size and high degree of sparseness of opportunistic citizen science datasets.
I will give an overview of different citizen science schemes for bird recording run by the British Trust for Ornithology and the challenges and opportunities they provide for understanding avian population dynamics. The focus will be on two schemes collecting data from unmarked individuals on a UK wide scale, the structured Breeding Bird Survey and the opportunistic BirdTrack scheme. I will compare and contrast current modelling approaches to estimate population trends based on either scheme, and present ongoing work aimed at combining information from multiple schemes using state-space models to integrate detection–nondetection and count data into a single analysis while accounting for individual detection probabilities during sampling.