University of Tasmania
Browse
Final Thesis - GROS.pdf (9.27 MB)

Quantifying and predicting vulnerable marine ecosystems on the Antarctic continental shelf

Download (9.27 MB)
thesis
posted on 2024-07-01, 23:57 authored by Charley GrosCharley Gros

Human activities put our oceans under pressure, leading to significant repercussions on their biodiversity. As a result, there is a growing international call on protecting fragile benthic biodiversity hotspots in the deep sea, collectively known as vulnerable marine ecosystems (VMEs). Identification of VMEs plays a crucial role within the management framework for bottom fishing activities. VME protection was largely instigated in 2006 by the 61/105 Resolution of the United Nations General Assembly (UNGA), which aimed to prevent adverse impacts on VMEs caused by bottom fishing activities. While the Food and Agriculture Organization (FAO) has developed international management guidelines, challenges remain in implementing criteria for identifying and protecting VMEs when encountered by fishing gear. In the Southern Ocean, VMEs’ protection falls under the responsibility of the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR), the international organisation regulating regional fishing activities in the high seas around Antarctica. VMEs identification is of high relevance to CCAMLR as it aligns with their commitment to both the UNGA resolution and the 2004 resolution to establish a network of Marine Protected Areas (MPAs) across the Southern Ocean. The urgency to protect this region is further heightened by the compounding pressures of accelerating anthropogenic climate change and commercial fishing activities.
Conservation efforts to protect VMEs face substantial challenges due to the prevailing lack of information regarding their specific locations. To address this challenge, species distribution modelling (SDM) has increasingly been recommended to extend our knowledge beyond sparse and fragmented VME observations by predicting the location of potential VMEs. SDM can estimate the probability of presence of certain taxa, known as VME indicator taxa, that act as proxies for VMEs in areas that have not been sampled. Nevertheless, the adoption of predictive VME distribution models in spatial management planning and conservation remains limited. Chapter 2 critically reviews VME distribution modelling studies and suggests ways to enhance the relevance and impact of these models on policy and management decision-making. I provide specific guidelines for seven common applications of VME distribution modelling to better align the models with user requirements. Through this literature review, I encourage scientists to base their models on specific and quantitative definitions of VMEs, assess site conservation value in relation to spatial predictions for multiple taxa, and explicitly map their vulnerability.
Underwater imagery offers a reliable means of sampling the benthic sessile fauna in the deep sea, making it a highly promising alternative to relying on by-catch reports from fisheries to map the distribution of VMEs. The recently released AS-AID dataset consolidates seafloor imagery datasets from 19 research cruises conducted around Antarctica. I manually delineated VME indicator taxa in over 1,800 georeferenced downward facing images within the AS-AID dataset. This effort generated over 40,000 annotations, where the morphotype of each taxon was based on the CATAMI (Collaborative and Automated Tools for Analysis of Marine Imagery) classification scheme. These annotations encompass 53 VME indicator morpho-taxa. From these annotations, I derived both presence-absence data and proportion data, representing the coverage of seafloor fauna.
Using this imagery-based dataset to develop a proof-of-concept, Chapter 3 presents a multi-criteria approach to VME identification that acknowledges that VME indicator taxa are not all equally disturbed by fishing activities. This approach is based on the computation of a VME index. To quantify the VME index, I combined the VME indicator morpho-taxa richness and abundance measured from underwater imagery data to map the vulnerability of benthic assemblages to fishing. Specifically, the contribution of each morpho-taxa is weighted by its estimated vulnerability to fishing. The implementation of this quantitative method is intended to enhance VME identification and contextualise the bycatch events.
Using biological data from underwater imagery and relevant environmental predictors, Chapter 4 models and maps the spatial patterns of VME indicator taxa richness and abundance across the entire Antarctica continental shelf (depth range: -200 to -2,500 m). I used a recently published Bayesian and spatial approach that jointly models the distribution of all VME indicator taxa together, called Hierarchical Modelling of Species Communities (HMSC). Predictions and associated model uncertainties are generated on a 2km x 2km spatial grid. The spatial distribution of VME indicator taxa hotspots is discussed, compared to the existing and proposed MPAs, and overlapped with the fishing footprint.
Undoubtedly, the benthic ecosystems of the Southern Ocean face more than just the threat of fisheries. In Chapter 5, I expand the concept of vulnerability to encompass other threats linked to human-induced climate change, including ocean warming, ocean acidification, iceberg scouring, and significant fluctuations of the ice shelf. Through my multi-threats analysis of benthic vulnerability, I identify new core areas with high VME index scores and suggest some management implications concerning the Southern Ocean marine protection network.
This work is particularly timely given the increasing global attention towards the sustainable use of ocean resources. International negotiations within the Convention on Biological Diversity and Biodiversity Beyond National Jurisdiction process have reached a level of maturity where the development of a science- and data-based framework is important, with VME distribution playing a key role within it. Notably, this work represents the first circumpolar mapping of VMEs in the Southern Ocean, which holds significant importance for CCAMLR when evaluating MPA proposals, conducting benthic ecosystem assessments and evaluating risk. Although presented in the Southern Ocean context, the data-driven framework has broader applicability in other oceans to guide towards a more holistic conservation of our fragile benthic ecosystems.

History

Sub-type

  • PhD Thesis

Pagination

xii, 163 pages

Department/School

Institute for Marine and Antarctic Studies

Publisher

University of Tasmania

Event title

Graduation

Date of Event (Start Date)

2024-03-15

Rights statement

Copyright 2024 the author

Usage metrics

    Thesis collection

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC