Global Species Sampling Intensity Gradients and Equalized Records for 12 Taxa

A 2026 study by Brice B. Hanberry characterizes global sampling intensity gradients using 17,200 species and 140 million georeferenced records across 12 taxa. The dataset includes a method to equalize sampling intensity, which reduced records in the greatest intensity class by 42%, and compares species distribution models from equalized, thinned, and unthinned records. The data is hosted on figshare under a CC-BY-4.0 license.

Use Cases

• Benchmarking spatial thinning methods based on global sampling intensity gradients.
• Training species distribution models using equalized sampling to reduce commission error.
• Analyzing geographic bias in biodiversity records, such as the concentration of bird samples in Europe.
• Incorporating locational distance uncertainty (e.g., 5 km) into model predictions to assess spatial concentration effects.

Strengths

• Includes data from 17,200 species across 12 different taxa.
• Processes a substantial volume of records, with the equalization method applied to 140 million initial records.
• Methodology is explicitly tested against models from range maps, showing maximized intersection and limited overprediction.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small (814.4 KB), suggesting it contains methodological documentation or summary results rather than the full raw records.

Provenance

Source

Brice B. Hanberry via figshare.

Collection Method

Kernel density estimation of georeferenced records followed by quantile-based equalization of sampling intensity.

Freshness

Last updated 2026-06-03 06:04:44; freshness should be verified.

Geography

Global, with analysis noting a concentration of bird records in Europe.