ARTIFICIAL INTELLIGENCE FOR ENVIRONMENTAL MONITORING: AUTOMATED IDENTIFICATION AND COUNTING OF BENTHIC MACROFAUNA USING CONVOLUTIONAL NEURAL NETWORKS
DOI:
https://doi.org/10.56238/revgeov17n6-099Keywords:
Environmental Bioindicators, Convolutional Neural Networks (CNNs), Decision Making, Climate ChangeAbstract
Benthic macrofauna; including mollusks, crustaceans, and polychaetes; plays a crucial role in regulating biogeochemical cycles, stabilizing sediments, and maintaining aquatic ecosystem resilience. Owing to their high sensitivity to environmental perturbations, these organisms are extensively used as bioindicators in ecological monitoring and impact assessments. Nevertheless, their traditional identification and enumeration rely on manual, labor-intensive procedures that demand taxonomic expertise and are susceptible to subjectivity and human error. To overcome these limitations, this study presents a robust and fully automated system based on Convolutional Neural Networks (CNNs) for the detection and quantification of benthic macrofauna from digital microscopy images. The proposed workflow encompasses standardized image acquisition, taxonomic annotation, advanced pre-processing, data augmentation, and supervised deep learning model training implemented in Python. Evaluation of the system revealed a mean F1-score of 0.93, mAP50 of 0.939, and inference times below 10 milliseconds per image, while maintaining stable validation precision, recall, and mAP values throughout training. Moreover, precision and recall metrics remained consistently high across diverse taxa, and the mean Average Precision (mAP50 and mAP50–95) improved steadily during training. These outcomes confirm the system's high generalization capacity and computational efficiency, enabling rapid, reliable, and reproducible ecological data generation. Beyond technical performance, the solution supports scalable biodiversity assessments, strategic environmental decision-making, and the optimization of conservation efforts, particularly in the face of climate change. This integration of artificial intelligence with benthic ecology represents a significant advancement in digital environmental monitoring, offering a cost-effective, time-efficient, and scientifically rigorous alternative to conventional analytical methods.
Downloads
References
Avelino, D.F.G., Silva, A.M.C., Avelino, P.G., Sá, M.M.S., Soares, A.X., 2023. Macroinvertebrados bentônicos como bioindicadores da qualidade ambiental dos recifes de arenito da praia de Porto de Galinhas (Pernambuco). Revista Brasileira de Meio Ambiente 11(1), 182–201.
Barbier, E.B., Hacker, S.D., Kennedy, C., Koch, E.W., Stier, A.C., Silliman, B.R., 2011. The value of estuarine and coastal ecosystem services. Ecological Monographs 81(2), 169–193. https://doi.org/10.1890/10-1510.1.
Borja, A., Dauer, D.M., Grémare, A., Elliott, M., 2010. The importance of setting targets and reference conditions in assessing marine ecosystem quality. Ecological Indicators 10(1), 1–9. https://doi.org/10.1016/j.ecolind.2009.07.008.
Bradshaw, C., Iburg, S., Morys, C., Sköld, M., Pusceddu, A., Ennas, C., Jonsson, P., Nascimento, F.J.A., 2024. Effects of bottom trawling and environmental factors on benthic bacteria, meiofauna and macrofauna communities and benthic ecosystem processes. Science of the Total Environment 921, 171076.
Duarte, C.M., Hendriks, I.E., Moore, T.S., Olsen, Y.S., Steckbauer, A., Ramajo, L., Carstensen, J., Trotter, J.A., McCulloch, M., 2013. Impact of ocean acidification on the structure and function of benthic ecosystems. Biogeosciences 10, 2761–2774. https://doi.org/10.5194/bg-10-2761-2013.
Ferreira, B., Carvalho, R., Silva, E., 2022. Deep learning for coral reef classification: A survey. Ecological Informatics 70, 101703.
Ghimire, D., Lee, J., 2021. Deep learning-based aquatic organism classification: A review. Aquaculture Reports 19, 100568. https://doi.org/10.1016/j.aqrep.2021.100568.
Hale, R., Bigham, K.T., Rowden, A.A., Halliday, J., Nodder, S.D., Orpin, A.R., Frontin-Rollet, G., Maier, K.L., Mountjoy, J.J., Pinkerton, M.H., 2024. Bioturbation and faunal-mediated ecosystem functioning in a deep-sea benthic community recovering from a severe seabed disturbance. Deep-Sea Research Part I 204, 104235.
Han, C., Xu, Z., Liu, X., 2021. Characteristics of macrofaunal assemblages and their relationships with environmental factors in a semi-enclosed bay. Marine Pollution Bulletin 167, 112348. https://doi.org/10.1016/j.marpolbul.2021.112348.
Howard, J., Sutton-Grier, A., Herr, D., Kleypas, J., McLeod, E., Pidgeon, E., Simpson, S., 2017. Coastal blue carbon: methods for assessing carbon stocks and emissions factors in mangroves, tidal salt marshes, and seagrass meadows. Conservation International, Arlington, VA.
Intergovernmental Panel on Climate Change (IPCC), 2023. Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Geneva. Available at: https://www.ipcc.ch/report/sixth-assessment-report-synthesis-report.
Krizhevsky, A., Sutskever, I., Hinton, G.E., 2012. ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems 25, 1097–1105.
Kuhlmann, M.L., 2022. Revisão dos índices da comunidade de macroinvertebrados bentônicos aplicados aos rios e reservatórios do estado de São Paulo. CETESB, São Paulo.
LeCun, Y., Bengio, Y., Hinton, G., 2015. Deep learning. Nature 521(7553), 436–444.
Li, P., Liu, J., Bai, J., Tong, Y., Meng, Y., Diao, X., Pan, K., Zhu, X., Lin, G., 2022. Community structure of benthic macrofauna and the ecological quality of mangrove wetlands in Hainan, China. Frontiers in Marine Science 9, 861718. https://doi.org/10.3389/fmars.2022.861718.
McLeod, E., Chmura, G.L., Bouillon, S., Salm, R., Björk, M., Duarte, C.M., Lovelock, C.E., Schlesinger, W.H., Silliman, B.R., 2011. A blueprint for blue carbon: toward an improved understanding of the role of vegetated coastal habitats in sequestering CO₂. Frontiers in Ecology and the Environment 9(10), 552–560. https://doi.org/10.1890/110004.
Mizuno, K., Terayama, K., Ishida, S., Godbold, J.A., Solan, M., 2024. Combining three-dimensional acoustic coring and a convolutional neural network to quantify species contributions to benthic ecosystems. Royal Society Open Science 11, 240042.
Moura, R.B., Dalto, A.G., Salloranzo, I.A., Moreira, D.L., Lavrado, H.P., 2023. Community structure of the benthic macrofauna along the continental slope of Santos Basin and São Paulo Plateau, SW Atlantic. Ocean and Coastal Research 71, e23032.
Neves, R.A.F., Valentin, J.L., 2011. Revisão bibliográfica sobre macrofauna bentônica de fundos não-consolidados em áreas costeiras prioritárias para conservação no Brasil. Arquivos de Ciências do Mar 44(3), 59–80. https://doi.org/10.32360/acmar.v44i3.153.
Pardo, J.C.F., Poste, A.E., Frigstad, H., Quintana, C.O., Trannum, H.C., 2023. The interplay between terrestrial organic matter and benthic macrofauna: framework, synthesis, and perspectives. Ecosphere 14, e4508. https://doi.org/10.1002/ecs2.4508.
Sousa, L.K.S., Cutrim, M.V.J., Nogueira Júnior, M., Oliveira, V.M., 2023. Does dredging activity exert an influence on benthic macrofauna in tropical estuaries? Case study on the northern coast of Brazil. Iheringia, Série Zoologia 113, e2023009.
Sun, X., Dong, J., Hu, C., Zhang, Y., Chen, Y., Zhang, X., 2021. Use of macrofaunal assemblage indices and biological trait analysis to assess the ecological impacts of coastal bivalve aquaculture. Ecological Indicators 127, 107713. https://doi.org/10.1016/j.ecolind.2021.107713.
United Nations Environment Programme (UNEP), 2021. Making Peace with Nature: A Scientific Blueprint to Tackle the Climate, Biodiversity and Pollution Emergencies. Nairobi. Available at: https://www.unep.org/resources/making-peace-nature.
Villon, S., Mouillot, D., Chaumont, M., Claverie, T., Villéger, S., 2018. Automatic underwater fish species classification with limited data using few-shot learning. Ecological Informatics 48, 26–34. https://doi.org/10.1016/j.ecoinf.2018.07.006.
Xu, M., Wang, J., Li, F., Chen, X., 2020. Automatic underwater image classification using deep convolutional neural networks. Pattern Recognition Letters 130, 259–265. https://doi.org/10.1016/j.patrec.2019.12.003.
Zhang, S., Qiao, H., Li, H., 2019. Plankton image classification using deep convolutional neural networks. Neurocomputing 335, 215–223.
Zhu, Z., Wood, C.M., Zhang, Y., Ma, Z., Wu, J., Wang, C., Jiang, Z., Yu, H., 2021. Deep learning for ecological monitoring in marine environments: A review. Ecological Informatics 61, 101240. https://doi.org/10.1016/j.ecoinf.2021.101240.
