178-10 Body-size Bias in Occurrence-based Biodiversity Datasets

Session: Paleontology, Diversity, Extinction, Origination

Presenting Author:

Leah Kahn

Authors:

Kahn, Leah X.¹, Heim, Noel A.², McClain, Craig R.³, Monarrez, Pedro M.⁴, Trindade-Santos, Isaac⁵, Webb, Thomas J.⁶, Payne, Jonathan L.⁷

(1) Stanford University, Stanford, CA, USA, (2) Tufts University, Earth & Climate Sciences, Medford, MA, USA, (3) University of Louisiana at Lafayette, Lafayette, LA, USA, (4) Virginia Tech, Blacksburg, VA, USA, (5) University of Louisiana at Lafayette, Lafayette, LA, USA, (6) University of Sheffield, Sheffield, United Kingdom, (7) Stanford University, Stanford, CA, USA,

Abstract:

Analyses of large, occurrence-based biodiversity datasets have greatly increased our understanding of biodiversity and evolutionary patterns. However, biodiversity knowledge is incomplete, with known biases toward organisms that are large, charismatic, and/or useful to humans. Characterizing and quantifying biases in biodiversity datasets is necessary, as unrecognized bias can limit our ability to accurately interpret evolutionary and ecological processes from biodiversity patterns. Here, we use body-size data from extensive databases developed for fossil and extant marine animals to assess body-size bias in three commonly used occurrence-based natural history databases: the Paleobiology Database (PBDB), the Ocean Biodiversity Information System (OBIS), and the GenBank sequence database. By referencing occurrences from these occurrence databases against comprehensive taxonomic lists from the Sepkoski compendium of fossil marine animal genera and the World Register of Marine Species (WoRMS), we determine the presence or absence of taxa in each database and model the probability of sampling as a function of body size. In total, our analyses include occurrence data for 23,311 fossil genera and 62,733 extant species, and sequence data for 70,344 species. We consistently find that taxa with small body sizes have a lower probability of being sampled, and this bias is present across all three occurrence databases. Among fossil taxa, bivalves, gastropods, ostracods, rhynchonellate brachiopods, and ray-finned fishes all show significant associations between small body size and reduced probability of sampling, while cephalopods, trilobites, and strophomenate brachiopods show non-significant body size-sampling relationships. For extant taxa, bivalves, gastropods, cephalopods, and ray-finned fishes show significant bias against sampling of small-bodied species in both OBIS and GenBank. The presence of size bias across different data types and taxonomic groups suggests that the bias against observing and documenting small-bodied taxa is not attributable to a single factor such as sampling method or technological limitations. As body size is associated with many organismal and species-level characteristics, our findings underscore the importance of considering body-size bias when using occurrence data to study biodiversity, ecology, and evolution, both past and present.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-10400

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.