Among many others, huge thanks go to Carla Mere Roncal, Arianna Basto, Zuzana Burivalova, and my advisor Christopher Searcy for seeing me through to the end of this wild project. Can't wait to share what we're able to do with this workflow in the future…

🧵(8/8)

We think this hybrid approach could be a flexible and repeatable means to answer questions about interior forest birds in the tropics that wouldn't be possible using a single data source

🧵(7/8)

The pooled models outperformed eBird-only models for predicting fine grain species-habitat associations in intact forest, but also retained information from eBird about species occurrence patterns in non-vegetated areas away from intact forest, including those subject to human disturbance

🧵(6/8)

For a five interior forest suboscine species in Madre de Dios, southeastern Peru, we combined long duration acoustic monitoring survey data and a filtered subset of eBird data available for the region with a high resolution land cover dataset of the region’s key ecological gradients

🧵(5/8)

Acoustic monitoring data is also becoming more available in the tropics… it is only cost and time-effective over small areas, but is still a great way to survey important ecological gradients for vocal organisms. CS tech for rapidly processing audio data, such as BirdNET, greatly helps

🧵(4/8)

There is less biodiversity data available for the tropics than the temperate zone, but citizen science data has been accumulating over time. eBird data are free to use and cover a large spatial extent, but sometimes have problems with survey evenness and quality

🧵(3/8)

Ornithologists in the temperate zone have combined structured and unstructured survey data to improve species distribution models for birds (see here for a good example), so is it possible to do this in the tropics as well?

onlinelibrary.wiley.com/doi/full/10....

🧵(2/8)