Search

Environmental heterogeneity, understood as spatial or temporal variability in environmental conditions, influences biodiversity and ecosystem processes over multiple scales, including at the deep seabed. However, as a result of the inaccessibility of the ocean beyond conventional SCUBA depth (> 30 m), key knowledge gaps remain on the biotic and abiotic patterns that influence the occurrence and distribution of seabed-associated taxa, habitats and resulting ecological processes and ecosystem services. Although seabed and habitat mapping documents seabed environmental and habitat heterogeneity, very little research explicitly quantifies it to draw conclusions on its ecological consequences. To study the character and effect of environmental heterogeneity on biodiversity and marine ecological processes in the deep sea, this thesis explored the potential of combining existing seabed and habitat mapping practices with the theoretical framework and analytical techniques from land- and seascape ecology. This thesis first reviews seascape ecology and its potential to study the ecological implications of environmental heterogeneity at the deep seabed, and identifies theoretical focal areas for the application of tools and concepts from seascape ecology deeper than 30 m (Chapter 1: Introduction). The objectives of this thesis, based on these focal areas, can be divided in three main themes: 1) characterising spatial heterogeneity using spatial pattern metrics; 2) assessing the ecological relevance of spatial heterogeneity quantified using spatial pattern metrics; and 3) applying this knowledge to inform environmental management. Objectives are addressed through a set of case studies that provide the opportunity to explore the multi-scale relationship between seabed structure and ecology in habitats at seamounts (km-m scale, Chapter 2 and 3) and reefs found between 30 m-250 m on atoll slopes (m-cm scale, Chapter 4 and 5) in theWestern Indian Ocean. Case studies test specific ecological hypotheses using spatial pattern metrics quantifying seascape composition, configuration and terrain structure, which function as predictors for the occurrence and distribution of benthic assemblages and demersal fish. Chapter 2 combines habitat mapping and spatial pattern metrics from seascape ecology to quantitatively test for and compare differences in seascape composition and configuration between five seamounts on the Southwest Indian Ridge (SWIR). Results quantitatively demonstrate that seamounts are highly variable in morphology, even when part of the same geological feature. As heterogeneity in the relative proportion and spatial relationships of habitats may influence ecological functioning, habitat mappers and marine managers focusing on representational protection of seamounts could benefit from such spatially-explicit approaches to quantify seabed heterogeneity. Chapter 3 examines the influence of multi-scale seabed spatial heterogeneity on 15 commercially important fish families at three SWIR seamounts, focusing on patch affinity, patch complexity, patch size and seascape aggregation. Although strongly driven by site and depth, demersal fish respond to unique combinations of seascape composition, configuration and terrain structure depending on their family. Further, seascape composition and configuration (i.e. habitat size, shape and structural connectivity) had higher predictive power than terrain derivatives commonly used in developing proxies for deepwater fish biodiversity. These outcomes indicate the importance of incorporating spatial pattern metrics when identifying environmental predictors of fish distributions and suitable habitat in deep-sea environments. Chapter 4 tests whether multi-scale geomorphology can act as a reliable spatial proxy for deeper reef assemblage (30 m-250 m) distribution. It found that assemblage occurrence and distribution is determined by a combination of environmental parameters, explained by the functional characteristics of each assemblage. Depth and structural complexity were main predictors, and broad scale predictors (25 m) proved more informative than finer scale predictors (2 m). Findings addressed geographical gaps in our knowledge of the distribution of deeper reef habitats and generated insights into ecological relationships. Complex geomorphological structures, including terraces and paleoshorelines, supported particularly high densities of mesophotic benthic assemblages and could be considered priority habitats for management. Chapter 5 investigates the effect of fine-scale (cm-m) environmental heterogeneity on fish associated with mesophotic reefs (30 m-120 m). Spatial pattern metrics quantifying benthic composition, configuration and terrain structure were extracted from transect terrain models and orthomosaics produced with Structure-from-Motion (SfM) photogrammetry. In addition to known drivers (depth and geographic location), results show a combination of fine-scale seascape metrics of terrain structure, patch composition and patch configuration best explains mesophotic fish assemblage structure. Overall, sites with steep slopes and high terrain complexity hosted highest fish abundance and biomass. Across case studies, spatial pattern metrics allowed quantification and comparison of seascape structure and functioned as reliable predictors for the occurrence and distribution of benthic assemblages and demersal fish at deeper reefs and seamounts in the Western Indian Ocean. Overall, spatial pattern metrics facilitated a better understanding of biodiversity-environment relationships in heterogeneous environments, in some cases functioning as the main explanatory variable. This thesis therefore recommends their further application in deep sea ecology, monitoring and ecosystembased management and conservation, whilst accounting for the biological phenomenon under consideration and scale- and context dependency in survey and analysis.

Monowai is an active submarine volcanic center in the Kermadec Arc, Southwest Pacific Ocean. We show, using cross-correlation and time-difference-of-arrival techniques, that low-frequency underwater sound waves from the volcano travel in the Sound Fixing and Ranging (SOFAR) channel and can be detected by bottom-moored hydrophone arrays of the International Monitoring System (IMS), a global sensor network maintained by the Comprehensive Nuclear-Test-Ban Treaty Organization. Hydroacoustic phases associated with the May 2011 eruption at Monowai are identified in the record of the IMS station at Ascension Island, Equatorial Atlantic Ocean. The source-receiver distance of ~15,800 km is the furthest documented range of any naturally occurring underwater signal ever observed. Our observations are consistent with results from transmission loss modeling, which suggest that acoustic propagation at southern latitudes is facilitated by the anomalous temperature regime of the Antarctic Circumpolar Current. Subsequently, we examine the 3.5-year record of the IMS hydrophone station near Juan Fernández Islands, Southeast Pacific Ocean, for volcanic activity at Monowai. Density-based clustering of arrivals during the time periods when data is available, i.e. from July 2003 to March 2004, and between April 2014 and January 2017, reveals 82 discrete episodes that are spaced days to weeks apart, typically ranging from a few hours to days in length. The resolution of the hydrophone data for seismic events at the volcano is estimated at 2.2 mb and exceeds regional broadband networks by one order of magnitude. Considering the results and techniques developed in the study of Monowai, we investigate the 2014 submarine eruption of Ahyi volcano in the Northern Mariana Islands. Acoustic phases of the 15-day episode are identified in the record of an IMS hydrophone array located at Wake Island in the northwestern Pacific Ocean. Explosive volcanic activity occurred in two bursts, accompanied by a decrease in low-frequency arrivals that is interpreted as a shift in signal source parameters. Acoustic energy released during the event is on the order of 9.7 1013 J.