Skip to content
DraconiticResearch
Search Sources API
197,761 studies31 sourcesUpdated 1 min ago
Peer-reviewedCC BY 4.0Royal Society Open Science

Using machine learning effectively in marine benthic ecology: a question-driven toolkit

  • Laurence H. De Clippele1
  • Joao Balsa2
  • Lara M. Beckmann3,4
  • Corie M.B. Boolukos5
  • Itziar Burgues6
  • Anna De la Torriente Diez7
  • Blair A.A. Easton8
  • Teresa C. Ferreira9
  • Alice Guzzi10,11,12
  • Lucy Harris13
  • Marie E. Kaufmann14,15
  • Ariadna Martínez-Dios16
8 more authors
  • Jose C. Mendoza17
  • Tanya G. Riley18
  • Vicente I. Villalobos14,15
  • H. Poppy Clark19
  • Carlos Dominguez Carrió2,16
  • Johanne Vad20
  • Tiffany Vlaar21
  • Abraham G. Smith22
22 affiliations
  1. 1School of Biodiversity, One Health and Veterinary Medicine, University of Glasgow 1 , Glasgow, UK
  2. 2Instituto de Investigação em Ciências do Mar – Okeanos, University of the Azores 2 , Ponta Delgada, Azores, Portugal
  3. 3Tjärnö Marine Laboratory, University of Gothenburg 3 , Strömstad, Västra Götaland County, Sweden
  4. 4Gothenburg Global Biodiversity Centre, University of Gothenburg 4 , Gothenburg, Västra Götaland County, Sweden
  5. 5Department of Geography, School of the Human Environment, University College Cork 5 , Cork, County Cork, Ireland
  6. 6Centro de Estudios Avanzados de Blanes, Spanish National Research Council 6 , Blanes, Spain
  7. 7Instituto Español de Oceanografía, IEO-CSIC, Spanish National Research Council 7 , Santander, Spain
  8. 8The Lyell Centre, Heriot-Watt University 8 , Edinburgh, UK
  9. 9Centre for Environmental and Marine Studies (CESAM), Department of Biology, University of Aveiro 9 , Aveiro, Aveiro District, Portugal
  10. 10Department of Earth, Environmental and Life Sciences (DISTAV), University of Genoa 10 , Genoa, Liguria, Italy
  11. 11Italian National Antarctic Museum (MNA, Section of Genoa), University of Genoa 11 , Genoa, Liguria, Italy
  12. 12National Biodiversity Future Center (NBFC) 12 , Palmero, Italy
  13. 13Ocean and Earth Science, University of Southampton 13 , Southampton, UK
  14. 14Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research 14 , Bremerhaven, HB, Germany
  15. 15Department of Biology/Chemistry, University of Bremen 15 , Bremen, HB, Germany
  16. 16Consejo Superior de Investigaciones Científicas (ICM-CSIC), Institut de Ciències del Mar 16 , Barcelona, Spain
  17. 17IEO-CSIC, Spanish National Research Council, Instituto Español de Oceanografía 17 , Madrid, Community of Madrid, Spain
  18. 18Scalloway Campus, Shetland College UHI 18 , Lerwick, UK
  19. 19University of Aberdeen 19 , Aberdeen, UK
  20. 20School of GeoSciences, The University of Edinburgh 20 , Edinburgh, UK
  21. 21School of Mathematics & Statistics, University of Glasgow 21 , Glasgow, UK
  22. 22Department of Computer Science, University of Copenhagen 22 , Copenhagen, Capital Region of Denmark, Denmark
Open at sourcePDFDOICite

Abstract

Manual annotation of marine image datasets is time-consuming, driving interest in machine learning (ML) to automate image analysis. However, benthic ecologists often lack the expertise needed to determine when and how ML can be applied effectively to their datasets. To explore the practical use of ML in benthic ecology, user perceptions and model training experiences were collected during a workshop. RootPainter, an interactive, corrective-annotation ML software, was used to introduce ML concepts and workflows. Participants developed models to segment benthic organisms from time-lapse and remotely operated vehicle imagery. Real-time feedback during training improved efficiency and understanding of model behaviour. Pre- and post-project qualitative surveys identified ecological and technical factors influencing model performance. From these findings, we propose five questions to guide efficient, robust annotation of marine benthic images for ML model development: (i) ‘do you have expert taxonomic and machine learning knowledge?’; (ii) ‘what is your ecological research question?’; (iii) ‘is the species sufficiently abundant?’; (iv) ‘are you consistent in your selection of training data?’; and (v) ‘have you anticipated the strengths and limitations of your imaging tool?’. Responses to these questions can provide practical guidance to help marine ecologists make informed choices, accelerate model training and improve performance.

Related

  1. Strategic science is needed to mitigate warming-induced emissions and bioclimate feedbacks effectively and efficiently
    Brian Buma, Pascal H Badiou, Sheel Bansal +7
    Environmental Research LettersPeer-reviewedCC BY 4.0Full textEnvironmental science & pollutionIOP/AASCrossref
  2. Anthropogenically induced shifts in sediment vegetation impact coastal benthic microbiomes
    Borja Aldeguer-Riquelme, Meagan A Beatty, Luis H Orellana +5
    HALPreprintPublisher termsFull textEcologyMicrobiologyHAL
  3. Machine-learning prediction of freezing temperature in Pb-contaminated soils remediated with biochar
    Laila Safi
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textEnvironmental science & pollutionCrossref
  4. Sol-Gel-Encapsulated Biofilms on Biochar Effectively Degrade PCBs and Extend Biodegradation Activity under Environmentally Relevant Conditions.
    Qin Dong, Weilun Zhao, Andres Martinez +6
    Environmental science & technologyPeer-reviewedCC BYFull textEnvironmental science & pollutionEurope PMC
  5. Atmospheric Predictability beyond 30 Days with Machine Learning
    P. Trent Vonich, Gregory J. Hakim
    Artificial Intelligence for the Earth SystemsPeer-reviewedLicence not statedFull text13 citedAtmospheric science & meteorologyOpenAlex

Cite

BibTeXRIS
BibTeX
@article{clippele2026using,
  title = {{Using machine learning effectively in marine benthic ecology: a question-driven toolkit}},
  author = {Laurence H. De Clippele and Joao Balsa and Lara M. Beckmann and Corie M.B. Boolukos and Itziar Burgues and Anna De la Torriente Diez and Blair A.A. Easton and Teresa C. Ferreira and Alice Guzzi and Lucy Harris and Marie E. Kaufmann and Ariadna Martínez-Dios and Jose C. Mendoza and Tanya G. Riley and Vicente I. Villalobos and H. Poppy Clark and Carlos Dominguez Carrió and Johanne Vad and Tiffany Vlaar and Abraham G. Smith},
  journal = {Royal Society Open Science},
  year = {2026},
  month = sep,
  doi = {10.1098/rsos.251833},
  url = {https://royalsocietypublishing.org/rsos/article/13/9/251833/483440/Using-machine-learning-effectively-in-marine},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Using machine learning effectively in marine benthic ecology: a question-driven toolkit
AU  - Laurence H. De Clippele
AU  - Joao Balsa
AU  - Lara M. Beckmann
AU  - Corie M.B. Boolukos
AU  - Itziar Burgues
AU  - Anna De la Torriente Diez
AU  - Blair A.A. Easton
AU  - Teresa C. Ferreira
AU  - Alice Guzzi
AU  - Lucy Harris
AU  - Marie E. Kaufmann
AU  - Ariadna Martínez-Dios
AU  - Jose C. Mendoza
AU  - Tanya G. Riley
AU  - Vicente I. Villalobos
AU  - H. Poppy Clark
AU  - Carlos Dominguez Carrió
AU  - Johanne Vad
AU  - Tiffany Vlaar
AU  - Abraham G. Smith
PY  - 2026
DA  - 2026/09/30
JO  - Royal Society Open Science
DO  - 10.1098/rsos.251833
UR  - https://royalsocietypublishing.org/rsos/article/13/9/251833/483440/Using-machine-learning-effectively-in-marine
AB  - Manual annotation of marine image datasets is time-consuming, driving interest in machine learning (ML) to automate image analysis. However, benthic ecologists often lack the expertise needed to determine when and how ML can be applied effectively to their datasets. To explore the practical use of ML in benthic ecology, user perceptions and model training experiences were collected during a workshop. RootPainter, an interactive, corrective-annotation ML software, was used to introduce ML concepts and workflows. Participants developed models to segment benthic organisms from time-lapse and remotely operated vehicle imagery. Real-time feedback during training improved efficiency and understanding of model behaviour. Pre- and post-project qualitative surveys identified ecological and technical factors influencing model performance. From these findings, we propose five questions to guide efficient, robust annotation of marine benthic images for ML model development: (i) ‘do you have expert taxonomic and machine learning knowledge?’; (ii) ‘what is your ecological research question?’; (iii) ‘is the species sufficiently abundant?’; (iv) ‘are you consistent in your selection of training data?’; and (v) ‘have you anticipated the strengths and limitations of your imaging tool?’. Responses to these questions can provide practical guidance to help marine ecologists make informed choices, accelerate model training and improve performance.
SN  - 2054-5703
ER  -