We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Artificial Neural Networks (ANN) to Model Microplastic Contents in Commercial Fish Species at Jakarta Bay
Summary
Researchers applied artificial neural networks (ANN) to model microplastic content in two commercial fish species from Jakarta Bay, using fish weight and length as proxies, and found that the weight-plus-length model best predicted pellet-type microplastics in both anchovy and mackerel stomachs. The study demonstrates ANN as an effective tool for projecting microplastic contamination levels in commercially important fish populations.
Jakarta Bay is known as one of the marine ecosystems that have been contaminated by microplastics. Despite massive loads of microplasticcontamination, Jakarta Bay is also habitat to potential commercial fish species, including anchovy Stolephorus commersonnii and mackerel Rastrelliger kanagurta. While information on the microplastic contents and their determining factors is still limited, the goal of this study was touse artificial neural networks (ANN) as a novel and useful tool to model the determinants of microplastic content in fish in Jakarta Bay, using fish weight and length as proxies. Inside the stomachs of S. commersonnii and R. kanagurta, the order of microplastics from the highest to thelowest was fiber > film > fragment > pellet. Based on the RMSE values of 3.199 for S. commersonnii and 2.738 for R. kanagurta, the ANNmodel of fish’s weight + length ~ pellet was found to be the best fitted model to explain the correlation of fish weight and length with microplastic content in the stomach. The results indicate that ANN is suitable for solving large, complex problems in determining and projecting microplastic contents and provides better estimates that can be used to manage R. kanagurta and S. commersonnii along with microplastic contamination threats.