Imagery has become one of the main data sources for investigating seascape spatial patterns. This is particularly true in deep-sea environments, which are only accessible with underwater vehicles. On the one hand, using collaborative web-based tools and machine learning algorithms, biological and geological features can now be massively annotated on 2D images with the support of experts. On the other hand, geomorphometrics such as slope or rugosity derived from 3D models built with structure from motion (sfm) methodology can then be used to answer spatial distribution questions. However, precise georeferencing of 2D annotations on 3D models has proven challenging for deep-sea images, due to a large mismatch between navigation obtained from underwater vehicles and the reprojected navigation computed in the process of building 3D models. In addition, although 3D models can be directly annotated, the process becomes challenging due to the low resolution of textures and the large size of the models. In this article, we propose a streamlined, open-access processing pipeline to reproject 2D image annotations onto 3D models using ray tracing. Using four underwater image datasets, we assessed the accuracy of annotation reprojection on 3D models and achieved successful georeferencing to centimetric accuracy. The combination of photogrammetric 3D models and accurate 2D annotations would allow the construction of a 3D representation of the landscape and could provide new insights into understanding species microdistribution and biotic interactions.

BACKGROUND: Coral-associated microbiomes vary greatly between colonies and localities with functional consequences on the host. However, the full extent of variability across the ranges of most coral species remains unknown, especially for corals living in deep waters which span greater ranges. Here, we characterized the microbiomes of four octocoral species from mesophotic and bathyal deep-sea habitats in the northern Gulf of Mexico, Muricea pendula, Swiftia exserta, Callogorgia delta, and Paramuricea biscaya, using 16S rRNA gene metabarcoding. We sampled extensively across their ranges to test for microbiome differentiation between and within species, examining the influence of environmental factors that vary with depth (53-2224 m) and geographic location (over 680 m) as well as the host coral\'s genotype using RAD-sequencing.
RESULTS: Coral microbiomes were often dominated by amplicon sequence variants whose abundances varied across their hosts\' ranges, including symbiotic taxa: corallicolids, Endozoicomonas, members of the Mollicutes, and the BD1-7 clade. Coral species, depth, and geographic location significantly affected diversity, microbial community composition, and the relative abundance of individual microbes. Depth was the strongest environmental factor determining microbiome structure within species, which influenced the abundance of most dominant symbiotic taxa. Differences in host genotype, bottom temperature, and surface primary productivity could explain a significant part of the microbiome variation associated with depth and geographic location.
CONCLUSIONS: Altogether, this work demonstrates that the microbiomes of corals in deep waters vary substantially across their ranges in accordance with depth and other environmental conditions. It reveals that the influence of depth on the ecology of mesophotic and deep-sea corals extends to its effects on their microbiomes which may have functional consequences. This work also identifies the distributions of microbes including potential parasites which can be used to inform restoration plans in response to the Deepwater Horizon oil spill.

The three-dimensional structure of habitats is a critical component of species\' niches driving coexistence in species-rich ecosystems. However, its influence on structuring and partitioning recruitment niches has not been widely addressed. We developed a new method to combine species distribution modelling and structure from motion, and characterized three-dimensional recruitment niches of two ecosystem engineers on Caribbean coral reefs, scleractinian corals and gorgonians. Fine-scale roughness was the most important predictor of suitable habitat for both taxa, and their niches largely overlapped, primarily due to scleractinians\' broader niche breadth. Crevices and holes at mm scales on calcareous rock with low coral cover were more suitable for octocorals than for scleractinian recruits, suggesting that the decline in scleractinian corals is facilitating the recruitment of octocorals on contemporary Caribbean reefs. However, the relative abundances of the taxa were independent of the amount of suitable habitat on the reef, emphasizing that niche processes alone do not predict recruitment rates.

Seascape connectivity increases carbon and nitrogen exchange across coastal ecosystems through flow of particulate organic matter (POM). However, there are still critical gaps in knowledge about the drivers that mediate these processes, especially at regional seascape scales. The aim of this study was to associate three seascape-level drivers which could influence carbon and nitrogen stocks in intertidal coastal seascape: connectivity between ecosystems, ecosystem surface area, and standing vegetation biomass of ecosystems. Firstly, we compared whether connected mangrove and seagrass ecosystems contain larger carbon and nitrogen storage than isolated mangrove and seagrass ecosystems. Secondly, we compared autochthonous and allochthonous POM in mangrove patches and seagrass beds, simultaneously estimating the area and biomass relative contribution to POM of the different coastal vegetated ecosystem. Connected vs isolated mangrove and seagrass ecosystems were studied at six locations in a temperate seascape, and their carbon and nitrogen content in the standing vegetation biomass and sediments were measured. POM contributions of these and surrounding ecosystems were determined using stable isotopic tracers. In connected mangrove-seagrass seascapes, mangroves occupied 3 % of total coastal ecosystem surface area, however, their standing biomass carbon content and nitrogen per unit area was 9-12 times higher than seagrasses and twice as high as macroalgal beds (both in connected and isolated seascapes). Additionally in connected mangrove-seagrass seascapes, the largest contributors to POM were mangroves (10-50 %) and macroalgal beds (20-50 %). In isolated seagrasses, seagrass (37-77 %) and macroalgal thalli (9-43 %) contributed the most, whilst in the isolated mangrove, salt marshes were the main contributor (17-47 %). Seagrass connectivity enhances mangrove carbon sequestration per unit area, whilst internal attributes enhance seagrass carbon sequestration. Mangroves and macroalgal beds are potential critical contributors of nitrogen and carbon to other ecosystems. Considering all ecosystems as a continuing system with seascape-level connectivity will support management and improve knowledge of critical ecosystem services.

Each winter, the North Atlantic Ocean is the stage for numerous cyclones, the most severe ones leading to seabird mass-mortality events called \"winter wrecks.\"1-3 During these, thousands of emaciated seabird carcasses are washed ashore along European and North American coasts. Winter cyclones can therefore shape seabird population dynamics4,5 by affecting survival rates as well as the body condition of surviving individuals and thus their future reproduction. However, most often the geographic origins of impacted seabirds and the causes of their deaths remain unclear.6 We performed the first ocean-basin scale assessment of cyclone exposure in a seabird community by coupling winter tracking data for ∼1,500 individuals of five key North Atlantic seabird species (Alle alle, Fratercula arctica, Uria aalge, Uria lomvia, and Rissa tridactyla) and cyclone locations. We then explored the energetic consequences of different cyclonic conditions using a mechanistic bioenergetics model7 and tested the hypothesis that cyclones dramatically increase seabird energy requirements. We demonstrated that cyclones of high intensity impacted birds from all studied species and breeding colonies during winter but especially those aggregating in the Labrador Sea, the Davis Strait, the surroundings of Iceland, and the Barents Sea. Our broad-scale analyses suggested that cyclonic conditions do not increase seabird energy requirements, implying that they die because of the unavailability of their prey and/or their inability to feed during cyclones. Our study provides essential information on seabird cyclone exposure in a context of marked cyclone regime changes due to global warming.8.

Human activities degrade and fragment coastal marine habitats, reducing their structural complexity and making habitat edges a prevalent seascape feature. Though habitat edges frequently are implicated in reduced faunal survival and biodiversity, results of experiments on edge effects have been inconsistent, calling for a mechanistic approach to the study of edges that explicitly includes indirect and interactive effects of habitat alteration at multiple scales across biogeographic gradients. We used an experimental network spanning 17 eelgrass (Zostera marina) sites across the Atlantic and Pacific oceans and the Mediterranean Sea to determine (1) if eelgrass edges consistently increase faunal predation risk, (2) whether edge effects on predation risk are altered by habitat degradation (shoot thinning), and (3) whether variation in the strength of edge effects among sites can be explained by biogeographical variability in covarying eelgrass habitat features. Contrary to expectations, at most sites, predation risk for tethered crustaceans (crabs or shrimps) was lower along patch edges than in patch interiors, regardless of the extent of habitat degradation. However, the extent to which edges reduced predation risk, compared to the patch interior, was correlated with the extent to which edges supported higher eelgrass structural complexity and prey biomass compared to patch interiors. This suggests an indirect component to edge effects in which the impact of edge proximity on predation risk is mediated by the effect of edges on other key biotic factors. Our results suggest that studies on edge effects should consider structural characteristics of patch edges, which may vary geographically, and multiple ways that humans degrade habitats.

Human activities are considered a critical impact factor for decision-making in coupled human-nature systems, such as conservation of coastal systems. Identifying key human activities that cause significant habitat degradation for coastal species remains challenging. We improved the spatial subsidy approach to identify and prioritize control strategies for human-caused distribution shifts of marine species. We applied this method to a threatened Indo-Pacific humpback dolphin (Sousa chinensis) in Xiamen Bay, China. Our results indicate that (1) a significant distribution shift for humpback dolphins from existing nature reserves to peripheral waters occurred from 2011 to 2014; (2) coastal tourism and industrial and urban construction had more significant negative impacts on humpback dolphins than maritime transportation and reclamation; and (3) proactive management should be implemented for maritime transportation and reclamation, while reactive management should be implemented for coastal tourism and industrial and urban construction. Human impact analysis, combined with spatially explicit modeling, contributes to determining the spatial alternatives for conservation planning. In response to possible ecological damage caused by human activities, the improved spatial subsidy results help provide knowledge and platforms for ecological compensation.

The eelgrass Zostera marina is an important foundation species of coastal areas in the Northern Hemisphere, but is continuing to decline, despite management actions. The development of new management tools is therefore urgent in order to prioritize limited resources for protecting meadows most vulnerable to local extinctions and identifying most valuable present and historic meadows to protect and restore, respectively. We assessed 377 eelgrass meadows along the complex coastlines of two fjord regions on the Swedish west coast-one is currently healthy and the other is substantially degraded. Shoot dispersal for all meadows was assessed with Lagrangian biophysical modeling (scale: 100-1,000 m) and used for barrier analysis and clustering; a subset (n = 22) was also assessed with population genetic methods (20 microsatellites) including diversity, structure, and network connectivity. Both approaches were in very good agreement, resulting in seven subpopulation groupings or management units (MUs). The MUs correspond to a spatial scale appropriate for coastal management of \"waterbodies\" used in the European Water Framework Directive. Adding demographic modeling based on the genetic and biophysical data as a third approach, we are able to assess past, present, and future metapopulation dynamics to identify especially vulnerable and valuable meadows. In a further application, we show how the biophysical approach, using eigenvalue perturbation theory (EPT) and distribution records from the 1980s, can be used to identify lost meadows where restoration would best benefit the present metapopulation. The combination of methods, presented here as a toolbox, allows the assessment of different temporal and spatial scales at the same time, as well as ranking of specific meadows according to key genetic, demographic and ecological metrics. It could be applied to any species or region, and we exemplify its versatility as a management guide for eelgrass along the Swedish west coast.

Macroalgal blooms are becoming an increasing problem in coastal regions worldwide and have been associated with a widespread decline of seagrass habitats. It is critical to measure macroalgal bloom (MB) impacts at broad spatial scales since seagrass seascape characteristics can influence feedback processes that regulate the resilience of seagrass ecosystems. We assessed the broad-scale spatial impacts of an MB formed by Anadyomene spp. on the seagrass seascapes in Biscayne Bay (Miami, US) using a multi-scale seascape approach. By integrating field and remote sensing data, our multi-scale approach showed significant reductions in seagrass foliage cover and a seascape structure transformation across the bloom extent. The landscape cover and patch extensiveness declined after the MB peak. Other spatial pattern metrics also showed that the seagrass seascape structure got fragmented. We demonstrated that a persistent MB could transform the structure of seagrass seascapes, hindering the resilience of seagrass habitats.

Many freshwater organisms have a life-history stage that can disperse through seawater. This has obvious benefits for colonization and connectivity of fragmented sub-populations, but requires a physiologically challenging migration across a salinity boundary. We consider the role of landscape boundaries between freshwater and seawater habitats, and evaluate their potential effects on traits and developmental histories of larvae and juveniles (i.e., dispersing life-history stages) of an amphidromous fish, Galaxias maculatus. We sampled juvenile fish on their return to 20 rivers in New Zealand: 10 rivers had abrupt transitions to the sea (i.e., emptying to an open coastline); these were paired with 10 nearby rivers that had gradual transitions to the sea (i.e., emptying into estuarine embayments). We reconstructed individual dispersal histories using otolith microstructure, otolith microchemistry, and stable isotope analysis. We found that fish recruiting to embayment rivers had distinct dispersal and foraging histories, were slower growing, smaller in size, and older than fish recruiting to nearby non-embayment rivers. Our results indicate that landscape edges can affect dispersal capabilities of aquatic organisms, potentially leading to divergent life-history strategies (i.e., limited- versus widespread-dispersal). Patterns also suggest that dispersal potential among landscape boundaries can create heterogeneity in the traits of individuals, with implications for metapopulation dynamics.