Toward a Coordinated Global Observing System for Seagrasses and Marine Macroalgae

Frontiers in Marine Science, Journal Year: 2019, Volume and Issue: 6

Published: July 4, 2019

In coastal areas around the world, dominant primary producers are benthic macrophytes, including seagrasses and macroalgae, that provide habitat structure food for diverse abundant populations communities, drive ecosystem processes. Seagrass meadows macroalgal forests economically central to human particularly in developing contributing fisheries yield, storm protection, blue carbon storage, important cultural values. These services threatened worldwide by activities, with substantial of seagrass kelp lost over last half-century. Tracking status trends marine macrophyte cover quality is an emerging priority ocean management, but doing so has been challenged limited coordination across numerous efforts monitor which vary widely goals, methodologies, scales, capacity, governance approaches, data availability. Here, we present a consensus assessment recommendations on current state opportunities advancing global observations, integrating contributions from community researchers broad geographic disciplinary expertise. The time ripe harmonize observations building existing networks identifying core set common metrics approaches sampling design, field measurements, taxonomy, governance, capacity building, management. A observation would then be facilitated ensuring rigorous documentation, archiving open-access sharing protocols resources at all stages workflow, surveys provision data. Realizing these will produce more effective, efficient, responsive observing, accurate picture change systems, stronger international sustaining observations.

Language: Английский

---

Temperate fish detection and classification: a deep learning based approach

Applied Intelligence, Journal Year: 2021, Volume and Issue: 52(6), P. 6988 - 7001

Published: March 22, 2021

Abstract A wide range of applications in marine ecology extensively uses underwater cameras. Still, to efficiently process the vast amount data generated, we need develop tools that can automatically detect and recognize species captured on film. Classifying fish from videos images natural environments be challenging because noise variation illumination surrounding habitat. In this paper, propose a two-step deep learning approach for detection classification temperate fishes without pre-filtering. The first step is each single an image, independent sex. For purpose, employ You Only Look Once (YOLO) object technique. second step, adopt Convolutional Neural Network (CNN) with Squeeze-and-Excitation (SE) architecture classifying image We apply transfer overcome limited training samples improve accuracy classification. This done by model ImageNet classifier via public dataset (Fish4Knowledge), whereupon both are updated interest. weights obtained pre-training applied post-training as priori. Our solution achieves state-of-the-art 99.27% using model. accuracies also high; 83.68% 87.74% augmentation, respectively. strongly indicates viable more extensive dataset.

Language: Английский

---

Seascape ecology: identifying research priorities for an emerging ocean sustainability science

Marine Ecology Progress Series, Journal Year: 2021, Volume and Issue: 663, P. 1 - 29

Published: March 2, 2021

Seascape ecology, the marine-centric counterpart to landscape is rapidly emerging as an interdisciplinary and spatially explicit ecological science with relevance marine management, biodiversity conservation, restoration. While important progress in this field has been made past decade, there no coherent prioritisation of key research questions help set future agenda for seascape ecology. We used a 2-stage modified Delphi method solicit applied from academic experts ecology then asked respondents identify priority across 9 interrelated themes using 2 rounds selection. also invited senior management/conservation practitioners prioritise same questions. Analyses highlighted congruence discrepancies perceived priorities research. Themes related both concepts management practice, those identified include change, connectivity, spatial temporal scale, ecosystem-based technologies metrics. Highest-priority (upper tercile) received 50% agreement between respondent groups, lowest (lower 58% agreement. Across all 3 tiers, 36 55 were within ±10% band present most determined by proportion votes received. For each theme, we provide synthesis challenges potential role These serve roadmap advancing during, beyond, UN Decade Ocean Science Sustainable Development (2021-2030).

Language: Английский

---

Unlocking the potential of deep learning for marine ecology: overview, applications, and outlook

ICES Journal of Marine Science, Journal Year: 2021, Volume and Issue: 79(2), P. 319 - 336

Published: Dec. 9, 2021

Abstract The deep learning (DL) revolution is touching all scientific disciplines and corners of our lives as a means harnessing the power big data. Marine ecology no exception. New methods provide analysis data from sensors, cameras, acoustic recorders, even in real time, ways that are reproducible rapid. Off-the-shelf algorithms find, count, classify species digital images or video detect cryptic patterns noisy These endeavours require collaboration across ecological science disciplines, which can be challenging to initiate. To promote use DL towards ecosystem-based management sea, this paper aims bridge gap between marine ecologists computer scientists. We insight into popular approaches for analysis, focusing on supervised techniques with neural networks, illustrate challenges opportunities through established emerging applications ecology. present case studies plankton, fish, mammals, pollution, nutrient cycling involve object detection, classification, tracking, segmentation visualized conclude broad outlook field’s challenges, including potential technological advances issues managing complex sets.

Language: Английский

---

Class-Aware Fish Species Recognition Using Deep Learning for an Imbalanced Dataset

Sensors, Journal Year: 2022, Volume and Issue: 22(21), P. 8268 - 8268

Published: Oct. 28, 2022

Fish species recognition is crucial to identifying the abundance of fish in a specific area, controlling production management, and monitoring ecosystem, especially endangered species, which makes accurate essential. In this work, problem formulated as an object detection model handle multiple single image, challenging classify using simple classification network. The proposed consists MobileNetv3-large VGG16 backbone networks SSD head. Moreover, class-aware loss function solve class imbalance our dataset. takes number instances each into account gives more weight those with smaller instances. This can be applied any or task imbalanced experimental result on large-scale reef dataset, SEAMAPD21, shows that improves over original by up 79.7%. Pascal VOC dataset also outperforms model.

Language: Английский

---

Nearshore seascape connectivity enhances seagrass meadow nursery function

Ecological Applications, Journal Year: 2019, Volume and Issue: 29(5)

Published: May 24, 2019

Abstract Diverse habitats composing coastal seascapes occur in close proximity, connected by the flux of materials and fauna across habitat boundaries. Understanding how seascape connectivity alters important ecosystem functions for fish, however, is not well established. For a seagrass‐dominant seascape, we predicted that configuration composition adjacent would alter access trophic subsidies, enhancing nursery function juvenile fish. In an extensive Zostera marina seagrass meadow, established sites to (1) highly complex productive kelp forests ( Nereocystis luetkeana ), (2) unvegetated sand habitats, (3) meadow interior. Using SCUBA , conducted underwater observations young‐of the‐year YOY ) rockfish Sebastes spp.) recruitment sites. generalized linear mixed effects models, assessed role adjacency relative provisions (habitat complexity prey) on recruitment. collections were used trace sources allochthonous vs. autochthonous primary production food web, via δ 13 C 15 N isotopic mixing model, prey consumption using stomach contents. Overall, was strongly influenced associated subsidies. Allochthonous . greatest source energy assimilated within N. forests, s consumed higher quality prey, which corresponded with better body condition sand. Moreover, forest enhanced suggesting key feature influencing nearshore habitats. general, promote connectivity, conservation restoration should prioritize inclusion mosaics high structural productivity. We illustrate emphasize importance seascape‐level approach considers linkages among management functions.

Language: Английский

---

The influence of seafloor terrain on fish and fisheries: A global synthesis

Fish and Fisheries, Journal Year: 2021, Volume and Issue: 22(4), P. 707 - 734

Published: Feb. 27, 2021

Abstract The structure of seafloor terrain affects the distribution and diversity animals in all seascapes. Effects on fish assemblages have been reported from most ecosystems, but it is unclear whether bathymetric effects vary among seascapes or change response to modification by humans. We reviewed global literature linking species (96 studies) determined that relief (e.g. depth), complexity roughness), feature classes substrate types) morphology curvature), widespread assemblages. Research ecological consequences focused coral reefs, rocky continental shelves deep sea ( n ≥ 20 studies), are rarely tested estuaries = 7). Fish associate with a variety attributes, variation depth aspect features reef shelf seascapes, sea. different also respond distinct metrics, fluctuations slope (coral reefs), rugosity (rocky reefs) (continental shelves, sea) each linked changes assemblage composition. Terrain simplification coastal urbanization dredging) resource extraction trawling) can reduce abundance, recover inside effective marine reserves. these for fisheries are, however, measured key challenge now examine how conservation combine alter distributions productivity across diverse

Language: Английский

---

Remote underwater video reveals higher fish diversity and abundance in seagrass meadows, and habitat differences in trophic interactions

Scientific Reports, Journal Year: 2019, Volume and Issue: 9(1)

Published: April 29, 2019

Abstract Seagrass meadows play a key ecological role as nursery and feeding grounds for multiple fish species. Underwater Visual Census (UVC) has been historically used the non-extractive method to characterize seagrass communities, however, less intrusive methodologies such Remote Video (RUV) are gaining interest could be particularly useful habitats, where juvenile camouflage among vegetation easily hide or flee from divers. Here we compared performance of UVC RUV in assessing communities two with low high canopy density. We found that detected more species individuals than UVC, on habitat higher density, which sheltered juveniles, especially herbivorous, adult piscivorous commercial importance, evidencing significant differences energy flow macrophytes predators between also ecosystem services they can provide. Considering ongoing worldwide degradation ecosystems, our results strongly suggest surveys using ecologic fisheries programs would render accurate information adequate inform conservation planning around world.

Language: Английский

---

The role of macroalgae as nursery areas for fish species within coastal seascapes

Cambridge Prisms Coastal Futures, Journal Year: 2022, Volume and Issue: 1

Published: Oct. 10, 2022

Abstract One of the most critical ecosystem functions provided by shallow coastal habitats is as nurseries for juveniles fish. Many studies that have assessed nursery function structurally complex compared seagrass with unstructured sand and mud such, has emerged important habitat juvenile fishes. Although considerably less work focussed on provision macroalgae within seascapes, recent started to highlight canopy-forming may in fact be comparable seagrass. This review collates research published role both tropical temperate seascapes highlights importance smaller brown algae from Fucalean genera (particularly Sargassum spp.) core areas fishes, particularly emperors (Lethrinidae), rabbitfishes (Siganidae), wrasse parrotfishes (Labridae), goatfishes (Mullidae), groupers (Serranidae), surgeonfish (Acanthuridae) damselfish (Pomacentridae) back-reef systems. Similarly, fucoid ( Cystoseira macroalgae-dominated reefs were Chromis chromis ), numerous species sparids (Sparidae). overall density fish was not shown higher kelp relative other habitats, recruitment Notolabrus celiodotus (wrasse), Paralabrax clathrus Brachyistius frenatus (Embiotocidae), Heterostichus rostratus (Clinidae) Sebastes spp. (Scorpaenidae). interchangeable (fish communities often different), found similar seascapes.

Language: Английский

---

Ecosystem-level effects of large-scale disturbance in kelp forests

Marine Ecology Progress Series, Journal Year: 2020, Volume and Issue: 656, P. 163 - 180

Published: July 20, 2020

Understanding the effects of ecological disturbances in coastal habitats is crucial and timely as these are anticipated to increase intensity frequency future due increasing human pressure. In this study we used directed kelp trawling a scientific tool quantify impacts broad-scale disturbance on community structure function. We tested ecosystem-wide BACI design using two 15 km 2 areas. The had substantial impact forests study, removing 2986 tons causing 26% loss total canopy at trawled stations. This created 67% reduction epiphytes, an 89% invertebrates altered fish populations living within habitats. effect habitat was variable depended how different species structure. Our results show that large-scale experimental habitat-forming have consequences extend beyond decline single affect multiple trophic levels broader ecosystem. findings relevance for understanding anthropogenic disturbances, including increased storm caused by climate change, may alter ecosystem

Language: Английский

---