Crustose coralline algae can contribute more than corals to coral reef carbonate production

Communications Earth & Environment, Journal Year: 2023, Volume and Issue: 4(1)

Published: April 6, 2023

Abstract Understanding the drivers of net coral reef calcium carbonate production is increasingly important as ocean warming, acidification, and other anthropogenic stressors threaten maintenance structures services these ecosystems provide. Despite intense research effort on production, inclusion a key forming/accreting calcifying group, crustose coralline algae, remains challenging both from theoretical practical standpoint. While corals are typically primary builders contemporary reefs, algae can contribute equally. Here, we combine several sets data with numerical modelling to demonstrate that match or even exceed contribution production. their importance, often inaccurately recorded in benthic surveys entirely missing budgets. We outline recommendations improve into such budgets under ongoing climate crisis.

Language: Английский

---

Scalable semantic 3D mapping of coral reefs with deep learning

Methods in Ecology and Evolution, Journal Year: 2024, Volume and Issue: 15(5), P. 916 - 934

Published: March 14, 2024

Abstract Coral reefs are among the most diverse ecosystems on our planet, and essential to livelihood of hundreds millions people who depend them for food security, income from tourism coastal protection. Unfortunately, coral existentially threatened by global climate change local anthropogenic pressures. To better understand dynamics underlying deterioration reefs, monitoring at high spatial temporal resolution is key. However, conventional methods quantifying cover species abundance limited in scale due extensive manual labor required. Although computer vision tools have been employed aid this process, particular structure‐from‐motion (SfM) photogrammetry 3D mapping deep neural networks image segmentation, analysis data products creates a bottleneck, effectively limiting their scalability. This paper presents new paradigm underwater environments ego‐motion video, unifying systems that use machine learning adapt challenging conditions under water, combined with modern approach semantic segmentation images. The method exemplified northern Gulf Aqaba, Red Sea, demonstrating high‐precision unprecedented significantly reduced required costs: given trained model, 100 m video transect acquired within 5 min diving cheap consumer‐grade camera can be fully automatically transformed into point cloud min. We demonstrate accuracy performance (of least 80% total accuracy), publish large dataset videos along frames annotated dense benthic classes. Our scales up reef taking leap towards automatic transects. advances transects reducing labor, equipment, logistics, computing cost. help inform conservation policies more efficiently. computational learning‐based Structure‐from‐Motion has broad implications fast low‐cost other than reefs.

Language: Английский

---

Exploring coral reef communities in Puerto Rico using Bayesian networks

Ecological Informatics, Journal Year: 2024, Volume and Issue: 82, P. 102665 - 102665

Published: June 5, 2024

Most coral reef studies focus on scleractinian (stony) corals to indicate condition, but there are other prominent assemblages that play a role in ecosystem structure and function. In Puerto Rico these include fish, gorgonians, sponges. The U.S. Environmental Protection Agency conducted unique surveys of communities across the southern coast included simultaneous measurement all four assemblages. Evaluating results from community perspective demands endpoints for assemblages, so patterns were explored by probabilistic clustering measured variables with Bayesian networks. found have stronger associations within than between taxa, unsupervised learning identified three cross-taxa relationships potential ecological significance. Clusters each assemblage constructed using an expectation-maximization algorithm created factor node jointly characterizing density, size, diversity individuals taxon. clusters characterized variables, taxa examined, such as stony fish variables. Each nodes then used create set meta-factor further summarized aggregate monitoring taxa. Once identified, taxon-specific meta-clusters represent can be examined regional or site-specific basis better understand risk assessment, management delivery services.

Language: Английский

---

Assessment of the utility of underwater hyperspectral imaging for surveying and monitoring coral reef ecosystems

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Nov. 30, 2023

Abstract Technological innovations that improve the speed, scale, reproducibility, and accuracy of monitoring surveys will allow for a better understanding global decline in tropical reef health. The DiveRay, diver-operated hyperspectral imager, complementary machine learning pipeline to automate analysis imagery were developed this purpose. To evaluate use imager underwater, automated classification benthic taxa communities was tested. Eight reefs Guam surveyed two approaches employed: high taxonomic resolution categories broad categories. results from DiveRay validated against data concurrently conducted photoquadrat determine their utility as proxy surveys. classifications did not reliably predict when compared those obtained by standard analysis. At level categories, however, comparable This particularly true estimating scleractinian coral cover, which accurately predicted six out eight sites. annotation libraries generated study insufficient train model fully account biodiversity on Guam’s reefs. As such, prediction is expected with additional surveying image annotation. first directly compare underwater scanning traditional survey techniques across multiple sites levels identification different degrees certainty. Our findings show dependent well-annotated library, imaging can be used quickly, repeatedly, monitor map dynamic using

Language: Английский

---

Seeing the Forest for the Trees: Mapping Cover and Counting Trees from Aerial Images of a Mangrove Forest Using Artificial Intelligence

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(13), P. 3334 - 3334

Published: June 29, 2023

Mangrove forests provide valuable ecosystem services to coastal communities across tropical and subtropical regions. Current anthropogenic stressors threaten these ecosystems urge researchers create improved monitoring methods for better environmental management. Recent efforts that have focused on automatically quantifying the above-ground biomass using image analysis found some success high resolution imagery of mangrove sparse vegetation. In this study, we focus stands with dense vegetation consisting endemic Pelliciera rhizophorae more widespread Rhizophora mangle species located in remote Utría National Park Colombian Pacific coast. Our developed workflow used consumer-grade Unoccupied Aerial System (UAS) forests, from which large orthophoto mosaics digital surface models are built. We apply convolutional neural networks (CNNs) instance segmentation accurately delineate (33% average precision) individual tree canopies species. also CNNs semantic identify (97% precision 87% recall) area coverage as well surrounding mud water land-cover classes. a novel algorithm merging predicted tiles trees recover shapes sizes overlapping border regions tiles. Using segmented ground areas interpolate their height model generate elevation model, significantly reducing effort pixel selection. Finally, calculate canopy combine it inventory derive each tree. The resulting forest, P. information, crown shape size descriptions, enables use allometric equations important metrics, such carbon stocks.

Language: Английский

---

Advanced imaging for microalgal biotechnology

Algal Research, Journal Year: 2024, Volume and Issue: 82, P. 103649 - 103649

Published: Aug. 1, 2024

Language: Английский

---

Safe AI for coral reefs: Benchmarking out-of-distribution detection algorithms for coral reef image surveys

Ecological Informatics, Journal Year: 2025, Volume and Issue: 90, P. 103207 - 103207

Published: May 20, 2025

Language: Английский

---

RapidBenthos: Automated segmentation and multi‐view classification of coral reef communities from photogrammetric reconstruction

Methods in Ecology and Evolution, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 26, 2024

Abstract Underwater photogrammetry is routinely used to monitor large areas of complex and heterogeneous ecosystems, such as coral reefs. However, deriving data on benthic components (i.e. sand, rubble, algae) from products has remained challenging due the highly time‐consuming process manual extraction. We developed a machine learning approach quantify community composition in reefs orthomosaics, which requires no delineation for training or implementation. The current study presents RapidBenthos, an automated workflow that segments classifies large‐area images. Our pipeline (1) uses pre‐trained segmentation model, eliminating need manually generated fine‐scale segmented data, (2) resulting multiple views using underlying survey images, allowing classification fine taxonomic levels. Within test photomosaic built reef area 40 m −2 , model automatically detected 43 different classes. Validation resulted overall accuracy 0.96 0.87, when compared digitised replica. RapidBenthos was 195 times faster than classification. Additional validation 524 Acropora colonies 11 additional plots 0.92 0.88 coarser ‘ Acropora’ group. capability extract unprecedented level photomosaics other environments, sustainably scale photogrammetric monitoring technique both replicate extent, consequently can lead new research questions more informed ecosystem management.

Language: Английский

---

Comparison of DenseNet-121 and MobileNet for Coral Reef Classification

Matrik Jurnal Manajemen Teknik Informatika dan Rekayasa Komputer, Journal Year: 2024, Volume and Issue: 23(2), P. 333 - 342

Published: March 8, 2024

Coral reefs are a type of marine organism that has beauty and benefits for other sea creatures’ ecosystems. However, despite its usefulness, coral vulnerable to damage such as bleaching, which can impact reef This research aims classify digital images healthy, bleached, dead reefs. method is DenseNet-121 MobileNet based on Convolutional Neural Networks. uses dataset from 1582 image data with three main classes: 720 were 150 dead, 712 healthy. The testing process carried out using several forms split datasets, namely 60:10:30, 50:10:40, 70:10:20. test results obtained sharing percentage 60:10:30 show architecture achieved 88.00% accuracy, 91.57% accuracy. Using 84.51% DenseNet- 121 90.52% Meanwhile, separation 70:10:20, 85.48% 92.74%

Language: Английский

---

A Scalable, Cloud‐Based Workflow for Spectrally‐Attributed ICESat‐2 Bathymetry With Application to Benthic Habitat Mapping Using Deep Learning

Earth and Space Science, Journal Year: 2024, Volume and Issue: 11(11)

Published: Oct. 29, 2024

Abstract Since the 2018 launch of NASA's ICESat‐2 satellite, numerous studies have documented bathymetric measurement capabilities space‐based laser altimeter. However, a commonly identified limitation point clouds is that they lack accompanying spectral reflectance attributes, or even intensity values, which been found useful for benthic habitat mapping with airborne lidar. We present novel method extracting bathymetry from data and automatically adding values Sentinel‐2 imagery to each detected point. This method, leverages cloud computing systems Google Earth Engine SlideRule Earth, ideally suited “big data” projects products. To demonstrate scalability our workflow, we collected 3,500 segments containing approximately 1.4 million spectrally‐attributed points. then used this set facilitate training deep recurrent neural network classifying habitats at photon level. trained two identical models, one without investigate benefits fusing photons Sentinel‐2. The results show an improvement in model performance 18 percentage points, based on F1 score. procedures source code are publicly available will enhance value new product, ATL24, scheduled release Fall 2024. These may also be applicable upcoming CASALS mission.

Language: Английский

---