No abstract

Technological advancements in remote sensing and GIS have improved natural resource managers’ abilities to monitor large-scale disturbances. In a time where many processes are heading towards automation, this study has regressed to simple techniques to bridge a gap found in the advancement of technology. The near-daily monitoring of dredge plume extent is common practice using Moderate Resolution Imaging Spectroradiometer (MODIS) imagery and associated algorithms to predict the total suspended solids (TSS) concentration in the surface waters originating from floods and dredge plumes. Unfortunately, these methods cannot determine the difference between dredge plume and benthic features in shallow, clear water. This case study at Barrow Island, Western Australia, uses hand digitising to demonstrate the ability of human interpretation to determine this difference with a level of confidence and compares the method to contemporary TSS methods. Hand digitising was quick, cheap and required very little training of staff to complete. Results of ANOSIM R statistics show remote sensing derived TSS provided similar spatial results if they were thresholded to at least 3 mg L−1. However, remote sensing derived TSS consistently provided false-positive readings of shallow benthic features as Plume with a threshold up to TSS of 6 mg L−1, and began providing false-negatives (excluding actual plume) at a threshold as low as 4 mg L−1. Semi-automated processes that estimate plume concentration and distinguish between plumes and shallow benthic features without the arbitrary nature of human interpretation would be preferred as a plume monitoring method. However, at this stage, the hand digitising method is very useful and is more accurate at determining plume boundaries over shallow benthic features and is accessible to all levels of management with basic training.

Machine-assisted object detection and classification of fish species from Baited Remote Underwater Video Station (BRUVS) surveys using deep learning algorithms presents an opportunity for optimising analysis time and rapid reporting of marine ecosystem statuses. Training object detection algorithms for BRUVS analysis presents significant challenges: the model requires training datasets with bounding boxes already applied identifying the location of all fish individuals in a scene, and it requires training datasets identifying species with labels. In both cases, substantial volumes of data are required and this is currently a manual, labour-intensive process, resulting in a paucity of the labelled data currently required for training object detection models for species detection. Here, we present a “machine-assisted” approach for i) a generalised model to automate the application of bounding boxes to any underwater environment containing fish and ii) fish detection and classification to species identification level, up to 12 target species. A catch-all “fish” classification is applied to fish individuals that remain unidentified due to a lack of available training and validation data. Machine-assisted bounding box annotation was shown to detect and label fish on out-of-sample datasets with a recall between 0.70 and 0.89 and automated labelling of 12 targeted species with an F1 score of 0.79. On average, 12% of fish were given a bounding box with species labels and 88% of fish were located and given a fish label and identified for manual labelling. Taking a combined, machine-assisted approach presents a significant advancement towards the applied use of deep learning for fish species detection in fish analysis and workflows and has potential for future fish ecologist uptake if integrated into video analysis software. Manual labelling and classification effort is still required, and a community effort to address the limitation presented by a severe paucity of training data would improve automation accuracy and encourage increased uptake.

No abstract

Increasing sea surface temperature and extreme heat events pose the greatest threat to coral reefs globally, with trends exceeding previous norms. The resultant mass bleaching events, such as those evidenced on the Great Barrier Reef in 2016, 2017, and 2020 have substantial ecological costs in addition to economic and social costs. Advancing remote (nanosatellites, rapid revisit traditional satellites) and in-field (drones) technological capabilities, cloud data processing, and analysis, coupled with existing infrastructure and in-field monitoring programs, have the potential to provide costeffective and timely information to managers allowing them to better understand changes on reefs and apply effective remediation. Within a risk management framework for monitoring coral bleaching, we present an overview of how remote sensing can be used throughout the whole risk management cycle and highlight the role technological advancement has in earth observations of coral reefs for bleaching events.