The following section includes the previous project I led or was involved in. Click on the tabs to see detailed information.
Topics
01
Species Distribution Modelling
Habitat Characterization
Comparing habitat characterizations around bird- and observer-locations. We found model accuracy improved 22.9% when habitat characterization was centered on birds.
02
Species Richness
Breeding Bird Survey Taiwan vs. eBird
Comparing species richness at standardized duration between the two largest avian datasets in Taiwan.
03
Benchmarking Bird Community
Neotropical Birds
We analyzed long-term monitored sites to assess the change in bird communities in Central and South America.
Species Distribution Modelling
Habitat characterization
Habitat may look distinct when zoom-in the scale
Birds are not always located within the exact habitat where observers conduct surveys. Accurately capturing the habitat that birds actually occupy is critical when building species distribution models. We compared model predictive performance using satellite imagery centered around observer locations versus bird locations for 105 breeding bird species in Oregon. Our results showed that models were more predictive when habitat was centered on the birds’ locations. See the predicted map snapshot below.
Northern Harrier
Predicted maps
Habitat characterization on observer’s location
Habitat characterization on birds’ location
Want to learn more about the study?
This work is now published in Ecological Indicators.
Species Richness
It is common to see studies overlook the importance of comparing biodiversity measurements at the same sampling effort, which results in biased data interpretation.
In this project, we compared species richness from the Breeding Bird Survey (BBS) and eBird Taiwan datasets at a 60-minute duration. Plus, we applied a species richness estimator (Chao1) on both datasets to calibrate imperfect detection and variable sampling efforts.
We found that BBS outperforms eBird in species richness. Even though we calibrated species richness in eBird, BBS still had higher species richness. We speculate observers in the BBS program possess skilled identification, aiming to detect more species. In addition, BBS sites are distributed to cover diverse habitats within a 2×2 km area, which may contribute to a higher diversity of birds. Our insight highlights the importance of incorporating variable sampling effort when comparing different datasets.
Learn the whole story?
This project is now available to read in Scientific Reports.
Discover the bird diversity
Benchmarking Bird Community
Need more collaboration with other projects
Big Bird Plots
We analyzed more than 500 research articles focused on bird communities in Central and South America. Birds in Panama show higher abundance and species richness than plots from Amazonia and Guianan plateau. We propose future continuation collaboration with ForestGEO long-term monitoring program.
The figure on the right shows common keywords used in literature published and their relationship among each study focus.
500+
Research Article Analyzed
5
Field Stations
Understand more?
This project is now available to read in Frontiers in Ecology and Evolution.