van Wyk A. J. auteur aut IRD Stuart-Smith R. D. auteur aut IRD Goetze J. S. auteur aut IRD Maire Eva auteur aut IRD Heyns-Veale E. auteur aut IRD Smit K. auteur aut IRD Langlois T. J. auteur aut IRD MacNeil M. A. auteur aut IRD Matus A. P. auteur aut IRD Lombard A. T. auteur aut IRD Carolina A. auteur aut IRD Semmens C. auteur aut IRD Clausius E. auteur aut IRD Rolim F. A. auteur aut IRD Lefcheck J. S. auteur aut IRD Monk J. auteur aut IRD Schmid J. K. auteur aut IRD Tattersall K. auteur aut IRD Ghigliotti L. auteur aut IRD Adams L. auteur aut IRD Samoilys M. auteur aut IRD Chabanet Pascale auteur aut IRD Whomersley P. auteur aut IRD Walsh P. auteur aut IRD Masuda R. auteur aut IRD Brainard R. auteur aut IRD Bernard A. auteur aut IRD text journalArticle eng text/pdf reformatted digital access Marine ecosystems are increasingly threatened by overfishing, pollution, coastal development and climate change, underscoring the need for long-term, representative information on key fish populations and habitats to inform management and policy. Underwater fish observation (UFObs) techniques, such as Underwater Visual Census (UVC), stereo-Baited Remote Underwater Video (stereo-BRUV) and Remotely Operated Vehicles (ROVs), play a key role in sustaining long-term data collection. Despite technological advancements, gaps persist in understanding research focus, geographic distribution and methodological biases inherent in these methods. We conducted a scientometric analysis of 1443 peer-reviewed publications (1953-2023), employing natural language processing and network analysis to map the research landscape. We identified 15 knowledge clusters, including marine protected areas, apex predator conservation and reef ecosystems. Our findings reveal increasing use of BRUVS and ROVs in studies of marine protected areas and subsea infrastructure, while UVC remains prevalent in shallow coral reef research. Geographic representation is skewed, with the field dominated by researchers based in Australia and the United States, and underrepresented in Africa and Southeast Asia. This imbalance highlights the need for more inclusive, globally coordinated monitoring and reporting. Our results underscore the urgency of standardising protocols within each observation method and developing interoperable reporting frameworks across techniques to maximise data comparability and foster international collaboration. Addressing these challenges will strengthen the field's capacity to inform global conservation strategies and support sustainable fisheries management. specialized epistemic mapping essential ocean variables methodological standardisation research landscape scientometric analysis underwater fish observation techniques MONDE 082 036 021 040 [Early access] [20 ] 2026 1467-2960 https://www.documentation.ird.fr/hor/fdi:010096365 10.1111/faf.70072 1467-2960 [F B010096365] https://www.documentation.ird.fr/hor/fdi:010096365 https://horizon.documentation.ird.fr/exl-doc/pleins_textes/2026-03/010096365.pdf IRD - Base Horizon / Pleins textes 2026-03-11 2026-04-07 fdi:010096365 fre