Impact of deep learning and post-processing algorithms performances on biodiversity metrics assessed on videos

PLOS One | https://doi.org/10.1371/journal.pone.0327577 August 11, 2025 13 / 14 confidence scores to minimize misidentifications. The Jaccard similarity index was used to measure the difference in com- position of species present on the videos and the composition estimated after automated processing using algorithms. (TIF) Acknowledgments The authors thank the two anonymous reviewers who helped improve and clarify this manuscript. Author contributions Conceptualization: Valentine Fleuré, Sébastien Villéger. Methodology: Valentine Fleuré. Visualization: Valentine Fleuré. Writing – original draft: Valentine Fleuré. Writing – review & editing: Valentine Fleuré, Kévin Planolles, Thomas Claverie, Baptiste Mulot, Sébastien Villéger. References 1. Bush A, Sollmann R, Wilting A, Bohmann K, Cole B, Balzter H, et al. Connecting Earth observation to high-throughput biodiversity data. Nat Ecol Evol. 2017;1(7):1–9. 2. Sparrow BD, Edwards W, Munroe SEM, Wardle GM, Guerin GR, Bastin J-F, et al. Effective ecosystem monitoring requires a multi-scaled approach. Biol Rev Camb Philos Soc. 2020;95(6):1706–19. https://doi.org/10.1111/brv.12636 PMID: 32648358 3. Spalding M. World atlas of coral reefs. 2001. 4. Trenkel V, Vaz S, Albouy C, Brind’Amour A, Erwan D, Laffargue P, et al. We can reduce the impact of scientific trawling on marine ecosystems. Marine Ecol Progress Series. 2018;609. 5. Cristescu ME, Hebert PDN. Uses and misuses of environmental DNA in biodiversity science and conservation. Ann Rev Ecol Evol Syst. 2018;49:209–30.

6. Rostami B, Nansen C. Application of active acoustic transducers in monitoring and assessment of terrestrial ecosystem health—A review. Methods Ecol Evol. 2022;13(12):2682–91. 7. Wetz JJ, Ajemian MJ, Shipley B, Stunz GW. An assessment of two visual survey methods for documenting fish community structure on artificial platform reefs in the Gulf of Mexico. Fisheries Res. 2020;225:105492. 8. Besson M, Alison J, Bjerge K, Gorochowski TE, Høye TT, Jucker T, et al. Towards the fully automated monitoring of ecological communities. Ecol Lett. 2022;25(12):2753–75. https://doi.org/10.1111/ele.14123 PMID: 36264848 9. Bilodeau SM, Schwartz AWH, Xu B, Paúl Pauca V, Silman MR. A low-cost, long-term underwater camera trap network coupled with deep residual learning image analysis. PLoS One. 2022;17(2):e0263377. https://doi.org/10.1371/journal.pone.0263377 PMID: 35108340 10. Shortis M, Harvey ES, Abdo DA. A review of underwater stereo-image measurement for marine biology and ecology applications. Oceanogr Marine Biol. 2009;47:257–92. 11. Griffin RA, Robinson GJ, West A, Gloyne-Phillips IT, Unsworth RKF. Assessing fish and motile fauna around offshore windfarms using stereo baited video. PLoS One. 2016;11(3):e0149701. https://doi.org/10.1371/journal.pone.0149701 PMID: 26934587 12. Magneville C, Leréec Le Bricquir ML, Dailianis T, Skouradakis G, Claverie T, Villéger S. Long-duration remote underwater videos reveal that graz- ing by fishes is highly variable through time and dominated by non-indigenous species. Remote Sens Ecol Conserv. 2023;9(3):311–22.

13. Dunkley K, Dunkley A, Drewnicki J, Keith I, Herbert-Read JE. A low-cost, long-running, open-source stereo camera for tracking aquatic species and their behaviours. Methods Ecol Evol. 2023;14(10):2549–56. 14. Maslin M, Louis S, Godary Dejean K, Lapierre L, Villéger S, Claverie T. Underwater robots provide similar fish biodiversity assessments as divers on coral reefs. Remote Sens Ecol Conserv. 2021;7(4):567–78. https://doi.org/10.1002/rse2.209 15. Desgarnier L, Mouillot D, Vigliola L, Chaumont M, Mannocci L. Putting eagle rays on the map by coupling aerial video-surveys and deep learning. Biol Conserv. 2022;267:109494. 16. Schiele M, Rowcliffe JM, Clark B, Lepper P, Letessier TB. Using water‐landing, fixed‐wing UAVs and computer vision to assess seabird nutrient subsidy effects on sharks and rays. Remote Sens Ecol Conserv. 2023;10(3):416–30. https://doi.org/10.1002/rse2.378 17. Gilbert NA, Clare JDJ, Stenglein JL, Zuckerberg B. Abundance estimation of unmarked animals based on camera-trap data. Conserv Biol. 2021;35(1):88–100. https://doi.org/10.1111/cobi.13517 PMID: 32297655