ABSTRACT Aim Biogeographic regionalisation reliability depends on how well the data capture true species distributions. We examined how data incompleteness and disciplinary bias (taxonomic vs. ecological) influence the detection of marine biogeographic regions and transition zones. Location Southwestern Atlantic. Taxon Marine Annelids (Polychaeta). Methods Using occurrence records from NONATObase (1859–2020), we compared taxonomic, ecological and combined datasets across five marine ecoregions. We estimated diversity using Hill numbers and standardised comparisons through a sample coverage framework. Pairwise null model tests (EcoTest and BiogTest) were applied to assess regional differentiation in composition and diversity structure. Results Sampling was heavily concentrated in Eastern and Southeastern Brazil, while Amazonia and Rio Grande were severely undersampled. Ecological and taxonomic datasets showed minimal spatial overlap and distinct methodological traditions (e.g., preference for grabs vs. manual collection). Coverage‐based standardisation revealed that many perceived regional similarities were artefacts of data scarcity. Amazonia and Rio Grande emerged as distinct units only after accounting for completeness. In contrast, the Northeastern and Southeastern regions functioned as a transitional system with high species turnover but a similar diversity structure. Main Conclusions Marine regionalisation is highly sensitive to disciplinary bias and sampling completeness. Biogeographic boundaries inferred from raw data often reflect research effort rather than biological structure. Integrating disparate data sources and using completeness‐standardised metrics are essential for robust biogeographic inference, particularly for understudied marine invertebrates.
Azevedo et al. (Sun,) studied this question.