Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean

Deep-sea ecosystems are among the least explored on Earth, and traditional sampling methods often underestimate fragile megabenthic species such as gorgonians. As a result, the use of underwater vehicles for visual surveys has increased considerably. This study combines underwater video surveys and...

ver descrição completa

Detalhes bibliográficos
Autores: Carmona-Rodríguez, A., Gómez-Donoso, Francisco, Cazorla, Miguel, Cobo-Viveros, Alba Marin, Aguilar, Ricardo, Ramos-Esplá, Alfonso A., Guijarro-García, Elena
Tipo de documento: artigo
Estado:Versão publicada
Data de publicação:2026
País:España
Recursos:Consejo Superior de Investigaciones Científicas (CSIC)
Repositório:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/419149
Acesso em linha:http://hdl.handle.net/10261/419149
https://api.elsevier.com/content/abstract/scopus_id/105027446823
Access Level:Acceso aberto
Palavra-chave:Bathyal
Deep-learning
Gorgonian forests
Soft bottoms
VME
id ES_293e52d6bf94dca5360acf2cb338df6c
oai_identifier_str oai:digital.csic.es:10261/419149
network_acronym_str ES
network_name_str España
repository_id_str
spelling Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western MediterraneanCarmona-Rodríguez, A.Gómez-Donoso, FranciscoCazorla, MiguelCobo-Viveros, Alba MarinAguilar, RicardoRamos-Esplá, Alfonso A.Guijarro-García, ElenaBathyalDeep-learningGorgonian forestsSoft bottomsVMEDeep-sea ecosystems are among the least explored on Earth, and traditional sampling methods often underestimate fragile megabenthic species such as gorgonians. As a result, the use of underwater vehicles for visual surveys has increased considerably. This study combines underwater video surveys and artificial intelligence (AI) to characterize the distribution and density of the bamboo coral Isidella elongata, a habitat forming species in the southeastern Iberian margin. Ten video transects obtained with a remotely operated towed vehicle (ROTV) between 300 and 725 m depth were analyzed using a YOLOv8-based detection algorithm trained on 983 manually annotated frames. Manual counting identified 2237 colonies (2347 ind/ha on average), revealing dense aggregations in the Cartagena Canyon and Seco de Palos seamount. AI-based detection achieved overall satisfactory performance, reproducing spatial patterns and colony size structure. However, it tended to overestimate large colonies and underestimate small ones, depending on image quality, specimen size, and environmental complexity. Although this methodology is not perfect, it provides a robust exploration tool for rapid localization and quantification of I. elongata meadows, greatly reducing video processing time. The integration of automated detection algorithms and standardized image databases is expected to enhance future deep-sea monitoring efforts. This work provides the first quantitative assessment of I. elongata populations in the southeastern Spanish margin, highlighting the potential of AI for the study and conservation of Vulnerable Marine Ecosystems (VMEs) in the Mediterranean Sea.The first author was supported by a grant from Ministry of Science, Innovation and Universities (Grant ID: FPU22/02966). This research was performed in the scope of the LIFE + IP INTEMARES project (LIFE15 IPE ES 012), coordinated by the Biodiversity Foundation of the Ministry for the Ecological Transition and the Demographic Challenge.Peer reviewedElsevierAgencia Estatal de Investigación (España)Ministerio de Ciencia, Innovación y Universidades (España)Ministerio para la Transición Ecológica y el Reto Demográfico (España)Carmona-Rodríguez, A. [0000-0001-9068-444X]Guijarro-García, Elena [0000-0001-7398-2381]Consejo Superior de Investigaciones Científicas [https://ror.org/02gfc7t72]202620262026info:eu-repo/semantics/articlehttp://purl.org/coar/resource_type/c_6501Publisher's versioninfo:eu-repo/semantics/publishedVersionapplication/pdfhttp://hdl.handle.net/10261/419149https://api.elsevier.com/content/abstract/scopus_id/105027446823reponame:DIGITAL.CSIC. Repositorio Institucional del CSICinstname:Consejo Superior de Investigaciones Científicas (CSIC)InglésCentro Oceanográfico de Murcia, (COMU)The underlying dataset has been published as supplementary material of the article in the publisher platform at DOI https://doi.org/10.1016/j.marenvres.2025.107830https://doi.org/10.1016/j.marenvres.2025.107830Síinfo:eu-repo/semantics/openAccessoai:digital.csic.es:10261/4191492026-05-22T06:33:51Z
dc.title.none.fl_str_mv Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
title Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
spellingShingle Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
Carmona-Rodríguez, A.
Bathyal
Deep-learning
Gorgonian forests
Soft bottoms
VME
title_short Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
title_full Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
title_fullStr Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
title_full_unstemmed Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
title_sort Artificial intelligence as a tool for bionomic transects: the case of Isidella elongata (Esper, 1788) forests in the Western Mediterranean
dc.creator.none.fl_str_mv Carmona-Rodríguez, A.
Gómez-Donoso, Francisco
Cazorla, Miguel
Cobo-Viveros, Alba Marin
Aguilar, Ricardo
Ramos-Esplá, Alfonso A.
Guijarro-García, Elena
author Carmona-Rodríguez, A.
author_facet Carmona-Rodríguez, A.
Gómez-Donoso, Francisco
Cazorla, Miguel
Cobo-Viveros, Alba Marin
Aguilar, Ricardo
Ramos-Esplá, Alfonso A.
Guijarro-García, Elena
author_role author
author2 Gómez-Donoso, Francisco
Cazorla, Miguel
Cobo-Viveros, Alba Marin
Aguilar, Ricardo
Ramos-Esplá, Alfonso A.
Guijarro-García, Elena
author2_role author
author
author
author
author
author
dc.contributor.none.fl_str_mv Agencia Estatal de Investigación (España)
Ministerio de Ciencia, Innovación y Universidades (España)
Ministerio para la Transición Ecológica y el Reto Demográfico (España)
Carmona-Rodríguez, A. [0000-0001-9068-444X]
Guijarro-García, Elena [0000-0001-7398-2381]
Consejo Superior de Investigaciones Científicas [https://ror.org/02gfc7t72]
dc.subject.none.fl_str_mv Bathyal
Deep-learning
Gorgonian forests
Soft bottoms
VME
topic Bathyal
Deep-learning
Gorgonian forests
Soft bottoms
VME
description Deep-sea ecosystems are among the least explored on Earth, and traditional sampling methods often underestimate fragile megabenthic species such as gorgonians. As a result, the use of underwater vehicles for visual surveys has increased considerably. This study combines underwater video surveys and artificial intelligence (AI) to characterize the distribution and density of the bamboo coral Isidella elongata, a habitat forming species in the southeastern Iberian margin. Ten video transects obtained with a remotely operated towed vehicle (ROTV) between 300 and 725 m depth were analyzed using a YOLOv8-based detection algorithm trained on 983 manually annotated frames. Manual counting identified 2237 colonies (2347 ind/ha on average), revealing dense aggregations in the Cartagena Canyon and Seco de Palos seamount. AI-based detection achieved overall satisfactory performance, reproducing spatial patterns and colony size structure. However, it tended to overestimate large colonies and underestimate small ones, depending on image quality, specimen size, and environmental complexity. Although this methodology is not perfect, it provides a robust exploration tool for rapid localization and quantification of I. elongata meadows, greatly reducing video processing time. The integration of automated detection algorithms and standardized image databases is expected to enhance future deep-sea monitoring efforts. This work provides the first quantitative assessment of I. elongata populations in the southeastern Spanish margin, highlighting the potential of AI for the study and conservation of Vulnerable Marine Ecosystems (VMEs) in the Mediterranean Sea.
publishDate 2026
dc.date.none.fl_str_mv 2026
2026
2026
dc.type.none.fl_str_mv info:eu-repo/semantics/article
http://purl.org/coar/resource_type/c_6501
Publisher's version
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/10261/419149
https://api.elsevier.com/content/abstract/scopus_id/105027446823
url http://hdl.handle.net/10261/419149
https://api.elsevier.com/content/abstract/scopus_id/105027446823
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv Centro Oceanográfico de Murcia, (COMU)
The underlying dataset has been published as supplementary material of the article in the publisher platform at DOI https://doi.org/10.1016/j.marenvres.2025.107830
https://doi.org/10.1016/j.marenvres.2025.107830
Sí
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:DIGITAL.CSIC. Repositorio Institucional del CSIC
instname:Consejo Superior de Investigaciones Científicas (CSIC)
instname_str Consejo Superior de Investigaciones Científicas (CSIC)
reponame_str DIGITAL.CSIC. Repositorio Institucional del CSIC
collection DIGITAL.CSIC. Repositorio Institucional del CSIC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869404992442466304
score 15,812429

Registros relacionados